Teacher Careers Education Systems Coherent Learning

Teacher Careers in
Education Systems That

Are Coherent for Learning
Choose and Curate Toward Commitment
to Capable and Committed Teachers (5Cs)
A Primer
Yue-Yi Hwa and Lant Pritchett

December 2021
Teacher Careers in Education Systems
That Are Coherent for Learning
Choose and Curate Toward Commitment to Capable
and Committed Teachers (5Cs)
Yue-Yi Hwa
Blavatnik School of Government, University of Oxford
Lant Pritchett
Blavatnik School of Government, University of Oxford
Acknowledgements:
We would like to thank Barbara Bruns, Clare Leaver, Joan
DeJaeghere, and Bich-Hang Duong for their very useful
comments on an earlier draft, without implicating them in
RISE—the large-scale education systems research
its remaining flaws. Marla Spivack, Michelle Kaffenberger,
programme—is supported by funding from
Jason Silberstein, and Joe Bullough also gave constructive
the United Kingdom’s Foreign, Commonwealth
input. Lillie Kilburn and Katie Cooper did invaluable work
and Development Office (FCDO), the Australian
with copyediting and formatting this primer, respectively.
Government’s Department of Foreign Affairs
and Trade (DFAT), and the Bill and Melinda Gates
Foundation. The Programme is managed and
Please cite this paper as:
implemented through a partnership between
Hwa, Y. and Pritchett, L. 2021. Teacher Careers in Education Oxford Policy Management and the Blavatnik
Systems That Are Coherent for Learning: Choose and Curate School of Government at the University of Oxford.
Toward Commitment to Capable and Committed Teachers
(5Cs). Research on Improving Systems of Education (RISE).
https://doi.org/10.35489/BSG-RISE-Misc_2021/02
Use and dissemination of this working paper is encouraged;

however, reproduced copies may not be used for commercial
purposes. Further usage is permitted under the terms of the
Creative Commons License.
The findings, interpretations, and conclusions expressed

in this document are entirely those of the author(s) and do
not necessarily represent those of the RISE Programme, our
funders, or the authors’ respective organisations. Copyright
remains with the author(s).
For more information, visit: riseprogramme.org
Photo credits: BigStock (cover pages)

Contents
Part 1. Introduction: Back from the brink 6

1.1 The 5Cs: Choose and curate toward commitment to 6
capable and committed teachers
The 5Cs as a set of principles for teacher careers in education systems
coherent for learning
Five premises for practising the 5Cs
1.2 Five current realities for teacher careers 10
1.3. Readers’ guide 16
What this primer is and is not
Overview
Part 2. We need to start with purpose and priorities 19

(because teaching and teacher careers are
complex)
2.1 The teaching profession is complex: Exploring teacher 20
motivation, teacher career dynamics, and education
bureaucracies
Teacher motivation is driven by multiple, interacting factors
Teacher careers are hedonic and dynamic, and any

model of teacher careers must reflect this
Effective organisations have a shared purpose

and technical expertise at their core
2.2 Typical teacher career reforms are often incoherent 37
with the complexity of teacher careers
Piecemeal changes that are incoherent with existing
education systems and organisations
PREMISE FOR PRACTICE #1: on clear, consensus-based prioritisation of
learning
Changes that are incoherent with the complexity

of classroom teaching
Changes that are incoherent with the complexity

of teacher motivation
Part 3: Why the 5Cs matter (and how they fit together) 44
3.1 The 5Cs contrast markedly with the conventional civil 46
service approach
‘Curation’ is common in comparable professions—but
unusual in teaching in developing countries
3.2 Why should teachers be not only ‘capable’ but also 52
‘committed’? Because motivation and purpose matter
4 The 5Cs: Choose and Curate Toward Commitment to Capable and Committed Teachers
3.3 Why ‘choose and curate’? Because the early career 53

phases are pivotal for systemwide norms of capability
and commitment
PREMISE FOR PRACTICE #2: on cultivating norms during the novice teacher
phase
3.4 Why ‘curate’, when there’s already ‘choose’? Because 56
classroom teaching is complex
PREMISE FOR PRACTICE #3: on the importance of curation
3.5 Why ‘commitment’? Because continuous curation is 59
costly
3.6 The 5Cs encapsulate the hedonics and dynamics 61
of teacher careers—toward the goal of capable and
committed teachers
Part 4. The 5Cs in practice: Varied examples of common 63

principles
4.1 Contextual differences can fundamentally shape 64
education systems (and the possibilities for reforming
them): the example of spider and starfish systems
In the long run, starfish systems may fit more naturally with purpose-
driven teacher careers ...
… yet the 5Cs can be applied equally in starfish and spider systems
4.2 Aspects of purpose-drive teacher career structures 66
that are fundamental regardless of context
In purpose-driven teacher career structures, decision-makers invest in—
and act on—‘thick’ information about teachers’ capability and commitment
PREMISE FOR PRACTICE #4: on information and support for teaching quality
In purpose-driven teacher career structures, ‘fairness’ in compensation is

based on shared goals (rather than administrative defaults)
PREMISE FOR PRACTICE #5: on purpose-driven definitions of fairness

4.3 The journey may not look like the destination: 71
examples of feasible ways forward
Examples of the 5Cs at the ‘destination’ of empowered,
purpose-driven teaching professions
Examples of the 5Cs on the ‘journey’ toward empowered,

Part 5. Conclusion: Where do you want to go? 75
References 78
The 5Cs: Choose and Curate Toward Commitment to Capable and Committed Teachers 5
Tables
Table 1. The 5Cs: A set of principles for teacher careers in education systems coherent for learning
Table 2. The 5Cs differ from the status quo approach to teacher careers in several fundamental ways
Table 3. Premises for practice: what to do differently in building teacher careers under the 5Cs
Table 4. A hedonic and dynamic model of teacher career structures
Table 5. Most design elements of teacher career structures can influence either the pecuniary or
psychosocial sources of teacher motivation—but delegation can directly affect all four sources
Table 6. An application of the 5Cs would vary the design elements for each teacher career phase, to suit
phase-specific and overall systemwide goals for the teaching profession (illustrative)
Table 7. Unlike commitment-only and careerlong curation approaches, the 5Cs directly support all four
sources of teacher motivation
Figures
Figure 1. Unlike typical civil service approaches, teacher career structures that apply the 5Cs are designed
such that each phase of the teacher career cycle optimally serves systemwide goals
Figure 2. Many education systems are failing to cultivate learning for most children, whether they are
enrolled in school or out-of-school
Figure 3. Schooling leads to wildly varying learning outcomes across developing countries, from near-
universal literacy to widespread illiteracy
Figure 4. While virtually all countries have improved access to schooling, few have concurrently improved
student learning—and most have instead seen declines in school quality
Figure 5. In some middle-income countries, only a small (and, in some cases, declining) proportion of
teachers feel that they are valued by society
Figure 6. The factors affecting teachers’ motivation and job satisfaction (e.g.,, urban vs. rural school
locations) can vary across contexts
Figure 7. The technical core—comprising a shared purpose and purpose-driven technical practices—is
central to organisational effectiveness
Figure 8. A possible instantiation of the 5Cs principles
Figure 9. Teacher career structures in PISA-D countries appear to diverge substantially from the ‘choose
and curate toward commitment’ elements of the 5Cs
Figure 10. There are five premises for practice about what education authorities and organisations should
do differently when applying the 5Cs for teacher career reform
Figure 11. The 5Cs are a subset of the conditions for empowered teaching professionals within education
systems that are coherent for learning, and which are embedded wider contexts
Boxes
Box 1: How do the four sources of teacher motivation differ from other classifications of motivation in
teacher careers?
Box 2: Resources on teacher educators and mid-tier education bureaucrats
Box 3: Examples of strategies for building clear, consensus-based prioritisation of learning
Box 4: Research on how education authorities and organisations should choose teachers during the pre-
service and novice phases
Part 1. Introduction: Back from the brink
It is important that we know where we come from, because if you do not know where you come
from, then you don’t know where you are, and if you don’t know where you are, you don’t know
where you’re going. And if you don’t know where you’re going, you’re probably going wrong.
—Terry Pratchett, author’s note in I Shall Wear Midnight (2010)
Part 1. Introduction: Back from the brink 6

1.1 The 5Cs: Choose and curate toward commitment to 6
capable and committed teachers
The 5Cs as a set of principles for teacher careers in education systems 6
coherent for learning
Five premises for practising the 5Cs 10
1.2 Five current realities for teacher careers 10
1.3. Readers’ guide 16
What this primer is and is not 17
Overview 18
A guiding vision needs to be clear-eyed about both the socialisation, or the distribution of patronage—that do
present reality and the possible future. A clear-eyed view not contribute (and often inhibit) teaching practices that
focused solely on the present reality can lead to pessimism promote student learning.
and even cynicism and inaction. On the other hand, turning
a blind eye to current reality prevents effective action 1.1 The 5Cs: Choose and curate toward
because effective action responds to fact—not fiction— commitment to capable and committed
about actual conditions. A useful map must show the teachers
desired destination—but it must also show the steps from
where you are now to that destination. The 5Cs as a set of principles for teacher
careers in education systems coherent for
Our vision of the future is one of empowered, highly
learning
respected, strongly performance-normed, contextually
embedded teaching professions that cultivate student To guide education systems beyond these dispiriting and
learning. As we discuss below, the present reality in damaging present realities, we propose a set of principles
all too many education systems is that the structure for building empowered, respected, well-compensated
of teacher careers neither expects nor cultivates good teacher careers in education systems that are coherent
performance. These career structures often treat teachers for learning. We call these principles the 5Cs: choose and
as interchangeable civil servants rather than professionals curate toward commitment to capable and committed
with a specialised craft. Moreover, these structures often teachers. The 5Cs are summarised in Table 1.
facilitate other purposes—such as compliance with
bureaucratic processes, centralised control over children’s
The
7 5Cs: Choose and Curate Toward Commitment to Capable and Committed Teachers 7
Table 1. The 5Cs: A set of principles for teacher careers in education systems coherent for learning
The 5Cs What they imply
Education authorities and organisations should initially choose teachers

Choose … based on their potential capability and commitment to cultivating student
learning.
The novice phase of teaching should be viewed as a period of curation, during

which there is ongoing attention to teachers’ professional development—and
… and Curate …
also to identifying those teachers who are most willing and able to make
careerlong contributions to children’s learning.
After the initial curation, education authorities and organisations should

… toward Commitment … make a long-term employment commitment to those teachers who have
demonstrated capability and commitment in classroom practice.
Effective teachers need to be technically capable and equipped for cultivating

… to Capable …
student learning in their specific school, classroom, and curricular contexts.
Teachers also need to be motivationally committed to the systemwide goal of

… and Committed teachers
cultivating student learning.
The 5Cs suggest that teacher career structures should seek professional skills and do not rely on short-term pay-for-
to identify those teachers who are capable of enacting performance based on standardised ‘thin’ indicators
effective classroom teaching and learning practices and (either of inputs, outputs, or outcomes). High-performance
who are committed to the educational purposes of the organisations in both not-for-profit and in purely private,
organisation. The evidence suggests that capable and for-profit domains, from hospitals and medical practices
committed teachers can only be accurately identified to architects to religious organisations to military to
through the teacher’s actual practice (and certainly not universities, all use variants of this 5Cs approach.
exclusively through ‘thin’1 indicators like pre-qualification
The 5Cs are a set of principles—not a recipe or formula or
training or examinations). Selecting the best achievable
blueprint—because empowered and strongly performance-
pool of teachers entails not only some initial process of
normed, contextually embedded teaching professions that
choosing pre-service and novice teachers based on the
cultivate student learning can look very different from each
available non-classroom-practice-based information,
other. Singapore’s and Finland’s teaching career structures
but also an ongoing process of curation during the early
are both lionised as exemplifying ‘best’ practices—despite
phases of teacher careers with a moderate to high degree
large differences between these two sets of ‘best’ practices
of turnover (nearly all voluntary) as it becomes apparent
in teaching career structures. And, yet, crucially, these
which novice teachers are best suited to remaining in the
stark differences are entirely coherent and appropriate to
profession. After the initial choose-and-curate period, the
the very different contexts in which they are embedded
organisation makes an employment commitment to these
(Hwa, 2021; for other examples, see Voisin & Dumay, 2020).
capable and committed teachers.
This variability in context-coherent forms applies not just
These 5Cs principles are not a new or novel idea of ours.
to teaching careers, but also to education systems more
Rather, they are just our description of what is common
broadly, whether among the most celebrated systems or
in effective organisations that require specialised, highly
otherwise. Consider these education systems at good-but-
1 Our use of ‘thin’ draws on three sources: Clifford Geertz’s (1973) notion of ‘thick description’; James Scott’s (2008) distinction in Seeing
Like a State between the constrained vision of bureaucratic high modernism that attempts to reduce reality to narrow classifications
versus the richness of actual lives; and the idea in principal-agent models that some indicators are easily observable and verifiable (like
age or ‘seniority’) versus other indicators that involve the use of judgment.
not-stellar performance levels: the federalised education service approach. (We return to this example in more detail
system in Germany, the pillar-choice-based system in the in Section 3.1.)
Netherlands, and the centralised system in France yielded
Crucially, the 5Cs differ from the status quo of teacher
remarkably similar performance in PISA 2012 (Pritchett,
careers in ways that go much deeper than the structure of
2014; this similarity in performance levels holds true in the
compensation and teacher cohort sizes illustrated in Figure
more recent PISA 2018). Given the different goals, features,
1. We argue that one of the reasons why the discourse on
and historical path dependencies across education
teacher compensation has been unproductive is that it has
systems, these contextual differences are not only an
been assumed that the design of teacher compensation
inevitable feature of effective teacher career structures,
could be discussed as a matter of public sector ‘personnel
but also a desirable one.
economics’ more or less independently of that nature
This is why we call the 5Cs a set of principles: principles of the education system, its structure and objectives,
can (and often should) be applied differently in different and the nature of teaching as an activity. Some of these
contexts. fundamental differences are summarised in Table 2.
With that in mind, Figure 1 shows a hypothetical example These differences between the status quo and the 5Cs are
of just one of any number of possible instantiations of the explored throughout this report. For now, we preview our
5Cs principles, presented in contrast to the typical civil arguments for just three of them.
Figure 1. Unlike typical civil service approaches, teacher career structures that apply the 5Cs are designed such that
each phase of the teacher career cycle optimally serves systemwide goals
(a) Typical civil service approach (b) Choose and curate toward commitment to
(commitment from day 1), capable and committed teachers (5Cs),
contrasted with typical private sector compensation one illustrative example
Notes on panel (b): Hypothetical example illustrating one of many possible instantiations of the 5Cs; adapted from Pritchett &
Murgai (2006), Figure 9.
Table 2. The 5Cs differ from the status quo approach to teacher careers in several fundamental ways
Status quo The 5Cs
Piecemeal policies and reforms System-oriented policies and reforms
Financial compensation is artificially separated from The approach is fully hedonic and fully dynamic in considering
other elements of teacher motivation the full range of sources of teacher motivation
Teaching as straightforward task Teaching as complex
Education as indistinguishable from other aspects of Education as a distinct field with a distinct purpose and
public service delivery specialised technical practices
Teacher career and compensation structures as a single Teacher career and compensation structures as tailored to
salary scale that varies little across contexts systems and purposes in question
First, we say that the 5Cs are a fully hedonic, fully dynamic and hedonic preferences across the full pool of teachers,
approach. By fully hedonic2 we mean that the 5Cs take into as well as the fact that policies and perceptions of prior
account all of the features that affect teachers’ satisfaction cohorts can affect subsequent cohorts of teachers. In
with their work: not only monetary compensation, but also this primer, we make the case that such a hedonic and
the details of school assignments, roles, responsibilities dynamic perspective on the complex interactions within
and school-specific work conditions, their satisfaction teacher career systems substantially raises the likelihood
from effective accomplishment of an important purpose, that education authorities and organisations will attract,
and the respect and validation that others in their society retain, and motivate teachers who will carry out their goals.
accord to them. The 5Cs are also fully dynamic3 in that
Second, we take it as a given that classroom teaching is a
they look beyond isolated points in the career cycle
complex task (see, inter alia, S.A. Brown & McIntyre, 1993;
(e.g., starting salaries or annual bonus schemes) and
DeJaeghere, Duong, & Dao, 2021). It requires the teacher
aggregate measures (e.g., whether teachers are overpaid or
to somehow align curricular expectations and available
underpaid), and instead consider the interplay of hedonics
instructional materials with students’ varied levels of
across all phases of the teacher career cycle. For example,
prior knowledge, attention spans, and moods, all of which
some jobseekers in Indonesia are willing to accept
can change in real time within the open system of the
unstable, poorly paid positions as contract teachers in
classroom. In many low- and middle-income countries, the
the hope of eventually gaining secure, well-compensated
challenge of this complexity is compounded by curricula
tenure as civil service teachers (Huang et al., 2020; Alifia,
that are far more advanced than the average students’
Pramana, & Revina, forthcoming). Beyond the individual
mastery levels (e.g., Pritchett & Beatty, 2012; Muralidharan
teacher, a fully hedonic, fully dynamic approach must also
& Singh, forthcoming) and by learning goals that far exceed
take into account the heterogeneity of motivational levels
2 ‘Hedonic’ analysis is used by economists as a way of determining ‘prices’ for the gains/losses from a specific product (or job) by
breaking it into a detailed list of characteristics. For instance, there is a massive literature on hedonic pricing of houses that assigns a
value to characteristics of a house like square footage, number of bedrooms, number of bathrooms, age, etc. There is also a massive
literature in labour economics examining how risky job require greater compensation and hence there is a ‘hedonic price’ of risk (e.g.,
Smith, 1979; Thaler & Rosen, 1976; Viscusi, 1993). Hedonic wage analysis has long been used in the economics of education in the United
States, where school districts can set their own wages and conditions and compete for teachers, especially since it is not uncommon for
multiple school districts to exist within the same metropolitan area (Chambers, 1981).
3 That is, a fully dynamic approach assumes that prospective teachers will aim to maximise their utility over the full course of the
career cycle.
teachers’ own subject-matter competencies (e.g., Venkat & careers are shaped by numerous factors. In 2019 alone,
Spaull, 2015). major reports on teacher career policy were published
by the World Bank (Béteille & Evans, 2019), the Education
Third, even though we focus in this synthesis on
Commission (2019), UNESCO’s International Institute for
teachers within large-scale, public-sector education
Educational Planning (Tournier & Chimier, 2019), and the
bureaucracies (rather than mom-and-pop standalone
International Task Force on Teachers for Education 2030
schools that are both simpler in organisational structure
(UNESCO, 2019).
and directly responsive to families rather than enmeshed
in various administrative or jurisdictional levels), we We join the conversation and hopefully add value by
nonetheless believe that teachers should not be viewed as focusing on the fact that teacher careers are embedded in
interchangeable civil servants who could be frictionlessly education systems. Prior studies of teacher careers have
reallocated to or from a different government ministry. similarly observed that teacher-related policies interact
Rather, teachers and teaching should be regarded as a in complex systems (e.g., Vegas & Ganimian, 2013), and we
distinct professional field with a distinct purpose and further this line of thinking. We show that teacher career
specialised technical practices—precisely because structures involve complexity at multiple levels of analyses,
classroom teaching is complex. and that typical policy approaches are often incoherent
with this complexity. We then propose a set of principles—
These second and third characteristics of teacher careers—
the 5Cs—that fully acknowledges this complexity in
the fact that classroom teaching is complex and the
managing teacher careers.
importance of establishing a clear, consensus-based
purpose and a set of specialised technical practices—in
Five premises for practising the 5Cs
turn underscore how important it is for teacher careers
structures to incorporate a process of curation during the The 5Cs go hand in hand with five premises for practice.
early career phases. Specifically, the complexity of teaching Each premise for practice highlights an area in which
means that no prospective teacher (nor the education education authorities and organisations should change the
authorities and organisations that may wish to employ this typical status quo approach in order to apply the 5Cs and
teacher) can fully know whether they will thrive amid this realise the vision of empowered, highly respected, strongly
distinct professional purpose and specialised technical performance-normed, contextually embedded teaching
practices until they have spent a significant amount of professions that cultivate student learning.
time in engaging in classroom practice. Moreover, such
These premises for practice are summarised in Table 3.
engagement in classroom practice will undoubtedly reveal
More details on premise for practice can be found in boxes
that some novice teachers may instead be better suited to—
throughout this primer, and Figure 10 in the conclusion of
and happier in—other occupations.
this primer shows how each premise for practice relates
Hence education authorities and organisations have the to the 5Cs.The reason why the 5Cs and these premises for
choice of either (a) sticking to the typical civil service practice matter is that conventional solutions to improving
approach, which does not disrupt the status quo but does teaching and teacher careers—such as raising teacher pay
mean that the teaching profession will include a nontrivial or improving pre-service training—do not go anywhere
proportion of people who may not be particularly near far enough to truly reform and re-form the teaching
committed to the purpose of the profession nor especially profession in many underperforming education systems.
capable at the specialised technical practices that can In such systems, the premises in Table 3 represent a
further that purpose but who remain in the profession due radical departure from business as usual. In the rest of this
to inertia, sunk costs, or a lack of appealing routes out of primer, we justify why such a radically different approach
the teaching profession; or (b) designing teacher careers is needed to get back from the brink of such deeply rooted
that allow for a nontrivial amount of turnover during a and persistent dysfunctions.
probationary novice period, for the sake of having an
empowered teaching profession comprising teachers 1.2 Five current realities for teacher careers
who have made an open-eyed choice to contribute to the
At this point, you might think that the 5Cs sound good but
profession long-term. We further discuss both the value
that the principles outlined above—with the extensive
and the challenges of curation in Part 3.
turnover of novice teachers during ‘curation’ (premise for
We are far from the first to emphasise the need for new practice #3) and of substantial investments in identifying
approaches to teacher compensation or that teacher and supporting good teaching (premise for practice #4)—
Table 3. Premises for practice: what to do differently in building teacher careers under the 5Cs
Typical approach What to do differently
No consideration of whether the education

Premise for practice #1: Clear, consensus-based prioritisation of
system is actually committed to universal
learning, delegated from education authorities and organisations to
achievement of acceptable student learning
schools and teachers, is fundamental to teacher career reform.
outcomes.
Premise for practice #2: Contextually embedded, learning-oriented

No particular attention to the novice phase. teacher professional norms must be cultivated throughout the novice
teacher phase.
Premise for practice #3: The pre-service and novice phases should be
Teacher security in job tenure is de facto
a period of curation, such that, as with nearly all other professions, a
awarded from day one.
substantial proportion of initial entrants do not persist in the career.
EMIS indicators, years of service, and formal Premise for practice #4: Education authorities and organisations should
certifications are the main (or only) sources invest in building multi-component ‘thick’ information systems about
of information about teaching quality used in teaching quality and in supporting teachers to continually improve their
structuring teacher compensation. pedagogical competencies.
Fairness in teacher compensation is defined Premise for practice #5: Fairness in teacher compensation should be
almost exclusively in terms of seniority, defined based on what, in the specific context, will attract, retain, and
perhaps with some consideration of formal motivate capable and committed teachers who make the best possible
qualifications and official responsibilities. contributions to student learning.
will never be implemented. You might be tempted to collections in the world, then soaking the stamps off of your
dismiss the approach as ‘impractical’. letters is not a practical strategy for getting there. It may
be a realistic, easy-to-implement, conventional strategy—
The reality is the exact opposite. The current conventional
but it would not at all be an effective strategy for building
approach to public-sector teacher careers may be
pre-eminence in world of stamp collecting.4
‘practical’ only in the sense that it can be realistically
implemented, but it is far less practical for achieving Similarly, an approach that is practical for employing a
results. It isn’t working, has never worked, and will never large number of people called teachers may be completely
work. And it is ludicrous to tack on piecemeal reform impractical if your vision is building up empowered, highly
attempts to this fundamentally broken approach, on the respected, strongly performance-normed, contextually
baseless assumption that these piecemeal reforms will embedded, teaching professionals who cultivate student
somehow revolutionise classroom teaching to become the learning. The fact that existing practices in the status quo
empowered, respected, technically specialised profession can be implemented year in and year out doesn’t make
that it can and should be. them ‘practical’ if they do not deliver the outcomes they
are designed for. ‘Practical’ must imply not only ‘realistic’
Here’s an illustration of what we mean by ‘practical’. If you
but also ‘effective’—or else we regress into the cynical
want to have a stamp collection, then the usual approach
outlook that although the status quo doesn’t deliver, it is
of soaking the stamps off any letters that are sent to your
the only option. In this section, we show just how absurd
house is a practical way of going about it. But if your vision
the conventional approach has become, as demonstrated
is to have one of the biggest and most exciting stamp
in five shaping conditions for teacher careers in many low-
4 For reference, according to the Guinness World Records, the largest collection of stamps featuring birds includes 14,558 stamps
from 332 countries. The largest collections of stamps featuring paintings weighs in at 19,284, while another collector has amassed 2,740
stamps featuring eyeglasses.
and middle-income countries. are a shameful disservice to students and teachers alike.
Surely, we can do better—and surely we should.
If, instead, you think that the 5Cs sound obvious, then
there’s no reason for you to read the rest of this primer The first of the five current realities for teacher careers in
(although you may enjoy it). Also, if the teachers in your underperforming education systems is that the learning
education system are already good at cultivating student crisis is widespread and severe—but not inevitable. In many
learning—such that, (a) nearly all students are still enrolled low- and middle-income countries, education systems fail
in school at age 15, (b) average performance on measures to cultivate learning for most children, even those enrolled
of learning is near the OECD median, and (c) student in school. The World Bank estimates that at age 10, nearly
learning is steady or improving5—then this is not aimed at half of all children who are enrolled in school across all low-
you. We’re offering principles for guiding the journey from and middle-income countries are unable to meaningfully
a disempowered teaching corps to an empowered and engage with a simple text (44 percent of children, out of
effective teaching profession. The teaching profession in 91.3 percent who are enrolled in school; World Bank, 2019),
your education has already reached that destination. The as shown in Figure 2. By age 15, many of those children
journey ‘back from the brink’ of low-performing education will have either dropped out of school or fallen far behind
systems, which we focus on in this primer, may look very their expected grade level—and even among those who
different from the iterative exploration of possibilities by are formally progressing in their school careers, the
an effective teaching profession, which would be the case proportion who are meeting minimum proficiency levels
in your already functional and at least mostly effective will have shrunk drastically. According to PISA-D data in
education system. seven upper- and lower-middle-income countries, only
9.1 percent of 15-year-old are enrolled in Grade 7 or above
The current realities for students and teachers in many
and are meeting minimum proficiency levels in the SDG,
education systems around the world are truly dire and
whereas 33.5 percent children of the cohort are enrolled
in Grade 7 or above but are nevertheless not meeting
minimum proficiency standard of the SDG (Pritchett &
Viarengo, 2021).
5 If, for instance, your average PISA scores on math, reading and science are above 425 (roughly the level of Turkey) and are steady
or rising (or, at least, not falling fast) then this paper really is not meant for you. This of course rules out nearly all OECD countries. But
if your average scores are near those of the recent participants in PISA-D (which, as Pritchett & Viarengo, 2021, show are typical of the
developing world) and are around 350 or below, then this paper is for you. We mention PISA and other cross-country assessments here
and elsewhere in this paper not because we think PISA should be the be-all-and-end-all for educational quality, or because we think
it is worth paying attention to small differences in PISA scores (as is generally the case for OECD countries) or to relative ranks in the
PISA horse race among OECD countries. Rather, we mention these learning assessments because many countries that fail to cultivate
learning for all children also fail to maintain rigorous localised benchmarks for locally determined student learning goals. Hence our
reliance on these cross-country assessments as benchmarks that are incomplete and imperfect but far better than the alternative of
no benchmark whatsoever.
Figure 2. Many education systems are failing to cultivate learning for most children, whether they are enrolled in school
or out-of-school
Sources: Authors; based on data from World Bank EduAnalytics (2019), Tables and Figures for Learning Poverty technical paper,
accessed 13 April 2021; and Pritchett & Viarengo (2021).
While learning crisis conditions are common across ‘Farming is hard work’) in a language of their own choosing
the developing world, they are not inevitable. Learning ranged from 97 percent in Rwanda to 11 percent in Nigeria.
outcomes around the world range from literally the worst The tremendous variation shows that low learning levels
possible to mediocre to impressively high. As shown in are not an inescapable fate of poor countries. Vietnam,
Figure 3, across 51 countries in the Demographic and for instance, reaches near-OECD performance levels even
Health Surveys, among young women at the time of after adjusting for the lower enrolment rates and sample
the survey who had attended Grade 6 (but no higher), composition (Dang et al., 2020).
the proportion who could read a simple sentence (e.g.,
Figure 3. Schooling leads to wildly varying learning outcomes across developing countries, from near-universal literacy
to widespread illiteracy
Source: Pritchett & Sandefur (2020), Figure 1(a).

Second, while schooling completion has expanded schooling access while also improving the percentage
impressively and this has raised the overall education of of adults who completed Grade 5 and no higher who are
the global population, it appears that in most countries literate. Instead, the large majority of countries have seen
in the world the learning outcomes for those enrolled have declines in the literacy rates of those with five years of
gotten worse. In an analysis of DHS and MICS survey data schooling. Moreover, in many countries these declines
from 87 developing countries, Le Nestour, Moscoviz, & have been more than 20 percentage points. In other words,
Sandefur (2021) found that every country in the sample education systems in many low- and middle-income
had expanded completion of at least Grade 5 between the countries have been effectively aligned for expanding
mid- and late-20th century, with some countries charting access to schooling in the logistical sense of building
schooling expansions of over 50 percentage points (e.g., more schools, hiring more civil servants, and getting more
Bangladesh). Yet, as shown in Figure 4, only a small number children into classrooms
of these countries have managed to substantially improve
Figure 4. While virtually all countries have improved access to schooling, few have concurrently improved student
learning—and most have instead seen declines in school quality
Source: Authors’ calculations based on data from Le Nestour, Moscoviz, & Sandefur (2021), Table A.4.
Notes. Each datapoint represents a country from the Demographic and Health Survey (DHS) or from the Multiple Indicators
Cluster Survey (MICS) for which information on literacy and schooling were available. For each country, the first birth cohort
decade in the sample was either the 1950s or the 1960s, and the last birth cohort decade was either the 1980s (rarely) or the
1990s.
This situation of sustained expansion in grade attainment improving, or even maintaining, the quality of student
accompanied by stagnating or declining learning per year learning in the classroom (Pritchett, 2013).
of schooling suggests that most education systems have
achieved ‘alignment’ and ‘coherence’ for the logistics of Third, current teacher career structures fail to support
expanding access but are evidently not fit for purpose in capable and committed teachers. To give just one example
of the failure to cultivate capable teachers, nationally Bryson, Corsini, & Martelli, 2020, on Italy).7 Few would
representative data from SACMEQ 2007 indicates that argue that teachers who are contributing less to students’
Grade 6 teachers in South Africa had completed an average learning growth should be paid more than their more
of 3.3 years of teacher training, and 74 percent of them had effective colleagues.8 (Instead, as we propose in premise
completed senior secondary school or above (Makuwa, for practice #5, fairness in teacher compensation should be
2011)—but only 21 percent of teachers demonstrated defined based on what, in the specific context, will attract,
mastery of the Grade 6 maths content that they were retain, and motivate capable and committed teachers who
supposed to be teaching (Venkat & Spaull, 2015). make the best possible contributions to student learning.)
In Pakistan, Bano (forthcoming) has characterised the
As an example of the failure to cultivate committed teachers,
internal school environment created by the fact that many
across the eight African countries surveyed in the Service
teachers are hired by political patrons as an ‘anti-work’
Delivery Indicators, an average of 40 percent of randomly
culture.
selected teachers were not present in the classroom during
an unannounced survey visit (World Bank, 2017).6 Schipper Fourth, in many low-performing education systems,
& Rodriguez-Segura (2021) find similarly disappointing teacher professional norms—and societal respect for the
classroom absence rates in a separate study in Tanzania teaching profession—have been eroded. A 2007 study of
and show that more than half of classroom absence is due teacher motivation in 12 countries in Sub-Saharan Africa
to teachers that are physically on school grounds but not and South Asia observed that:
in the classroom, which is some ways is even more striking
The general perception of stakeholders and
than not having come to work at all. In India, one estimate
teachers in all countries is that the teaching
suggests that US$5.0 billion of government expenditure on
profession no longer commands the high status it
teacher salaries is lost due to teacher absences (Datta &
enjoyed 30 years ago and that teachers, especially
Kingdon, 2021).
primary school teachers, are now ‘undervalued by
Empirical data suggests that poorly designed teacher society’, The country studies confirm that teaching
career structures are contributing to—or, at least, failing is very much regarded as ‘employment of last resort’
to mitigate—these systemic failures to cultivate capable by most school leavers and university graduates
and committed teachers who, in turn, cultivate student (Bennell & Akeyeampong, 2007, p. 38).
learning. For instance, in Andhra Pradesh, India, there
is a negative (though insignificant) correlation between Also, as shown in Figure 5, among five middle-income
teacher value-added and pay, likely due to a subset of countries that participated in the Teaching and Learning
veteran teachers who are getting higher pay but exerting International Survey (TALIS) in both 2013 and 2018, less
less effort (Lemos, Muralidharan, & Scur, 2021; see also than half of all teachers agreed or strongly agreed with the
6 Based on the mean of country-level absence rates in Kenya, Madagascar, Mozambique, Nigeria, Senegal, Tanzania, Togo, Uganda.
Some of these teachers were absent from school, and some were present in school but absent from the classrooms in which they were
scheduled to teach.
7 Even more striking is that being a ‘civil service’ teacher per se (versus variants of renewable contracts) appears to lower teacher
performance on promoting student learning, a finding from both experimental (RCT) evidence in Kenya (Duflo, Dupas, & Kremer, 2011)
and in observational evidence in Uttar Pradesh, India (Atherton & Kingdon, 2010). The evidence from Uttar Pradesh suggests students
learned half as much with a teacher with a civil service appointment but in all other respects observationally equivalent to a teacher
on a renewable contract. This suggests, shockingly, that being engaged in the typical existing teacher career structures in and of itself
reduces performance. That said, this does not mean that teachers should never receive permanent contracts. As we discuss in Section
3.5, continuous short-term contracts are costly for individual teachers and for the education system as a whole. Rather, as we argue
throughout this paper. low-performing education systems need to radically reform of what it means to be a civil service teacher in their
context.
8 Some of the shortcomings of teacher careers and compensation structures become apparent when assessed relative to the wider job
markets in which they are embedded. Using administrative data on public school teachers in Florida, Martin West and his collaborators
have found that teachers who enter the profession during economic recessions tend to be more effective at raising test scores, which is
likely due to a greater supply of high-skilled individuals opting into teaching, implying that teaching is a relatively unappealing profession
under normal circumstances (Nagler, Piopuinik & West, 2020). They also found that teachers who are more effective at raising their
students’ test scores tend to receive a larger boost in pay after switching careers, relative to less effective teachers who also exited the
profession—such that the most effective teachers have the strongest incentives to leave the classroom (Chingos & West, 2012). This
implies an embedded perverse structure wherein those who emerge as the best teachers have larger pay raises from being a good
teacher outside of teaching rather than in the profession.
statement ‘I think that the teaching profession is valued benchmarks became less and less attainable, affecting
in society’, with less than 12 percent of Brazilian teachers the personal satisfaction that teachers could derive from
agreeing with that statement in 2018. Moreover, between their jobs (to the extent that teachers’ job satisfaction
these two survey rounds, Mexico and Chile saw sizeable was still defined by the ideals of a few students passing
declines in teachers’ beliefs that their profession was high stakes examinations). Often, this coincided with an
socially valued. In contrast, in high-performing Vietnam, expansion of top-down, ‘rational’ bureaucratic control over
92.8 percent of teachers in TALIS 2018 agreed or strongly school systems in line with postcolonial modernisation
agreed that their profession was socially valued. (Metzler, 2009; Harper, 2011)—whether driven by local
leaders or by the donors and lenders ingrained in the
Figure 5. In some middle-income countries, only a small international development orthodoxy—with the side
(and, in some cases, declining) proportion of teachers effect of re-orienting the norms in the teaching profession
feel that they are valued by society from fostering autonomous and creative educators toward
generating rule-following bureaucrats. (We discuss this
further in Section 2.1 on ‘donut’ organisations that are
hollow inside as they have lost their core purpose.)
Fifth, piecemeal approaches to improving teaching and
teachers are largely failing because they are neither system-
oriented in considering the complex interactions between
different parts of the education system nor fully hedonic in
considering all elements of teacher motivation. As noted
above, the problems facing the teaching profession in low-
performing education systems are large, interconnected,
and entrenched. Given the magnitude of these problems, it
is both tragic and unsurprising that numerous standalone
attempts to reform teaching and teachers have failed to
realign the profession as a whole for student learning. This
has been demonstrated time and again, and is discussed
further in Section 2.2. For example, the past few decades
Source: Authors, using TALIS data extracted from OECD.Stat. have seen multiple reform attempts in Indonesia, but
incremental improvements in the technical quality of in-
service teacher training have failed to meaningfully shift
In many education systems, this erosion of professional the norms of classroom practice (Revina et al., 2020). More
norms and societal respect for teachers is closely strikingly, a national policy that effectively doubled the
intertwined with the rapid expansion of schooling salaries of teachers did not lead to any gains in student
enrolment described above. Or, more specifically, this learning (de Ree et al., 2018). Similarly, despite being
shift in norms and professional status coincided with lauded by many economists, teacher ‘performance pay
systemic inadequacies in support for teaching and learning for learning outcomes’ schemes have a decidedly mixed
as education systems expanded toward mass enrolment track record, at least in developing countries, of leading
without the concomitant shift in the understanding of to sustained student learning gains (Breeding et al., 2021).
purpose. As teaching evolved from a profession that was The issue with these piecemeal solutions is that they
for the elite and by the elite toward a profession serving all don’t account for complex, longstanding interactions
children at scale, the profession lost a substantial amount throughout the wider education system—so they don’t
of prestige and social validation—which was not inevitable go deep enough to reset teacher professional norms or to
(the Vietnam case offers a counterexample), but occurred establish systemwide consensus about student learning as
in many contexts that lacked a deep consensus about the the overarching priority of teacher careers specifically and
importance of universal learning (see, e.g., Béteille et al., of the education system as a whole.
2020, Chapter 4, on the weakening of teacher professional
norms amid rapidly growing demand for education in 1.3 Readers’ guide
a number of South Asian countries). Concurrently, as a
In the rest of this primer on teacher careers, we make
more diverse body of teachers began serving a more
the case that teaching and teacher career reform are
diverse body of students, elite or colonial-era learning
complex, and that the 5Cs can serve as an approach for in the conclusion, the teacher career structures are a
guiding reform amid these interconnected and sometimes subset of the larger conditions in which an empowered,
unpredictable systemic interactions. respected, professional teaching force can individually
and collectively act for students’ sakes. Neither do we look
What this primer is and is not in detail at every single aspect of teacher careers. Instead,
To shape reader expectations, it is worth noting four points we prioritise the aspects of teacher career structures that
about this primer.9 are most distinctive to the 5Cs (as compared to elements
that are common across many teacher career approaches,
First, what you are reading is not a typical academic journal
e.g., high-quality teacher education and training). In some
article or even a typical academic or research ‘paper’. We
instances, we include boxes that suggest, as a starting
present the 5Cs principles, as along with key premises for
point, some resources on related aspects that we do not
translating those principles into practice, by synthesising
cover in detail.
insights and identifying patterns across dozens of studies
in a wide range of academic papers and country contexts. Third, when it comes to framing, the 5Cs and the five
We also draw on pertinent examples from occupations that premises for practice are designed as guidelines for the
are not teaching, but that share with it certain recognisable education authorities and organisations that have the
characteristics. This cross-disciplinary, cross-context authority to make systemic decisions that shape teacher
synthesis approach means that we do not offer estimates careers. This inevitably means that this primer does not
of effect sizes for student learning gains under the 5Cs give teacher agency the amount of airtime that it deserves.
(because there are far too many interacting factors across This is not because teacher agency doesn’t matter. As
far too many contexts; and teacher career structures that stated at the very beginning, our vision of the future is of
apply the 5Cs principles should look different in different an empowered teaching profession, and our exploration
contexts). Neither do we point to a single education system of complexity in the teaching profession in Part 2 below
that perfectly embodies every principle of the 5Cs and all starts with a discussion of motivation from the teacher’s
five of the premises for practice about what to do differently perspective of view rather than from a policy standpoint.
(because principles can and should be applied differently in Rather, the 5Cs are framed around education authorities
different contexts, and no such archetype exists). However, and organisations because (for better or worse) their
we do offer numerous empirical examples throughout the decisions and actions can shape the lived experiences—
primer to substantiate different aspects of the approach. and room for agency—of teachers and students throughout
This primer is not an academic paper that is too long and the education system.
tries to cover too much with too much structure and not
Finally, this primer is not intended to be the final word
enough precision, but an analytic synthesis.
on teacher careers. It is not a report that results from
Second, despite being very long, this primer is not a a multistakeholder consultative process, but rather a
comprehensive discussion of teachers and teaching. Rather, narrative written by two people trying to make sense of
we call it a ‘primer’ because we give an overview of many this subject by drawing on research across a wide range
(but not all) topics related to teacher career structures, of academic disciplines and contexts. We have written
including system-level policies, processes, and practices this primer to provoke discussion about teacher careers.
that affect the composition of the pool of teachers in an We hope that you engage with it and, perhaps, disagree
education system and their performance and progression with it—in addition to being a ‘primer’ in the sense of an
throughout the career cycle. We do not look in detail at introductory guidebook on a subject, we hope that this will
many other important aspects of teaching (e.g., curriculum be a ‘primer’ in the sense of an initial layer of material that
and classroom materials), teachers (e.g., teacher wellbeing prepares the ground for further development.
during crises and pandemics), and teacher-related policy
(e.g., teacher standards).10 As we illustrate in Figure 11
9 An intriguing analysis of which movies get an ‘F’ rating by opening-night audiences via CinemaScore (something only a handful of
movies have ever done) suggests that violating genre expectations is a way of eliciting strong disapproval, so we want to be clear that
we know what this primer is not. See: Lincoln, K. (2017, September 20). What the 19 Movies to Ever Receive an ‘F’ CinemaScore Have in
Common. Vulture. https://www.vulture.com/2017/09/here-are-the-only-19-movies-to-ever-receive-an-f-cinemascore.html.
10 For one example of teacher standards, see the Education International/UNESCO Global Framework of Professional Teaching
Standards.
Overview incoherences identified in the previous section, and what

this means for hedonics and dynamics.
In the rest of this primer, we make the case that teaching
and teacher career reform is complex, and that the 5Cs can In Part 4, we explore concrete examples of the 5Cs
serve as a set of principles for guiding reform amid these in practice. We discuss how elements of the 5Cs are
interconnected and sometimes unpredictable systemic applied differently in different contexts, and also look at
interactions. commonalities in teacher career reform across contexts.
This part of the primer includes examples both from
In Part 2, we first discuss some of the complexities
education systems that have high-performing teacher
that affect teacher careers, focusing on three levels of
careers—representing the desired ‘destination’ of the 5Cs
analysis: the sources of teacher motivation, the design
and of teacher career reform more generally—and from
of teacher careers, and the drivers of organisational
education systems that are on the journey back from the
effectiveness. Against the backdrop of this complexity,
brink.
we identify three types of incoherence that frequently
afflict attempts to reform teacher careers (and education Part 5 is the conclusion.
systems more broadly). We also argue that a fundamental Throughout this primer, we use standalone boxes to
way of strengthening the teaching profession amid highlight some premises for practice. These premises for
these complexities is by emphasising a consensus-based, practice propose key areas in which education authorities
systemwide purpose. and organisations can change the status quo in line with
In Part 3, we lay out the 5Cs, discussing why each of the the 5Cs. Other boxes suggest further reading on aspects of
‘C’s is a necessary part of the whole, how they collectively teacher careers that we do not address in detail.
accommodate the complexities and overcome the
Part 2. We need to start with purpose and priorities (because teaching and teacher
careers are complex)
Life now is so complicated … just buying a tomato at a grocery store means that you are
unwittingly supporting toxic pesticides, exploiting labour, contributing to global warming.
Humans think that they’re making one choice, but they’re actually making dozens of choices
they don’t even know they’re making. …
Look, there are plenty of awful people and thoughtless jerks … But there are also people who
are really trying to be good.
—Michael, The Good Place, Season 3 Episode 11 (2019)

Part 2. We need to start with purpose and priorities 19
(because teaching and teacher careers are
complex)
2.1 The teaching profession is complex: Exploring teacher 20
motivation, teacher career dynamics, and education
bureaucracies
Teacher motivation is driven by multiple, interacting factors 20
Teacher careers are hedonic and dynamic, and any 24

model of teacher careers must reflect this
Effective organisations have a shared purpose 32

2.2 Typical teacher career reforms are often incoherent 37
with the complexity of teacher careers
Piecemeal changes that are incoherent with existing 37
education systems and organisations
PREMISE FOR PRACTICE #1: on clear, consensus-based prioritisation of 39
learning
Changes that are incoherent with the complexity 40

of classroom teaching
Changes that are incoherent with the complexity of teacher motivation 41

In this section, we explore two key insights from systems is an open system, with complex interactions between
thinking and what these insights imply for teacher careers. students, teacher, and lesson content. In this section,
This lays groundwork for understanding why the 5Cs we look at three levels of complexity that affect teacher
matter in reforming teacher careers. careers.
The first insight is that teacher careers are a complex First, we zoom in on individual teachers, each of whom has
system. By complex, we mean not only that teacher their own goals and motivations. We group these under
careers comprise a wide range of elements, but also four sources of motivation: like any other humans, teachers
that these elements interact via feedback loops that can legitimately want to derive an adequate combination of
yield unexpected outcomes.11 Sometimes, these complex finances, material circumstances, personal satisfaction,
interactions mean that changing one element of a system and social validation from the effort that they put into their
will have no effect at all on other parts of the system work (discussed further below). Each teacher’s desired
(because those other parts can maintain a longstanding combination of motivational factors depends on individual
equilibrium despite some changes) or it will have perverse idiosyncrasies as well as contextual influences.
effects (because the change did not adequately account for
Next, we look at the design of teacher career structures. We
some of the elements or interactions within the system).
identify five design elements that education authorities
In Section 2.1, we discuss three (out of the many) levels of
and organisations13 can influence in teacher career policy,
complexity that affect teacher careers.
and four phases of the teacher career cycle that each have
The second insight is that coherence is central to effective distinct requirements.
systems. By coherence, we mean that different elements
Finally, we zoom out to consider education bureaucracies
and interactions within a system are (more or less) aligned
as organisations. Here we focus on the observation that
toward a common purpose rather than working at cross
organisational success is driven by a combination of a
(or even contradictory) purposes (Pritchett, 2015; see also
common purpose and technical practices that advance
Besley & Ghatak, 2005; Fullan & Quinn, 2015; Meadows,
that purpose, which together compose the technical core
2008; Roberts, 2004; UNESCO, 2019).12 Unfortunately, given
of the organisation.
how complex teacher careers (and education systems more
broadly) are, there is no shortage of potential incoherence Teacher motivation is driven by
that can constrain the teaching profession. We discuss multiple, interacting factors
three types of incoherence in Section 2.2.
In a well-designed teacher career structure, the overall,
This discussion of complexity and coherence in education subjectively assessed well-being of an individual teacher
systems leads to premise for practice #1, on the centrality is higher when they are fulfilling the objectives of those
of a clear, consensus-based systemwide sense of purpose authorities or organisations who designed the system (and,
and prioritisation of children’s learning. hopefully those authorities or organisations are reflecting a
broad social consensus about the objectives of education).
2.1 The teaching profession is complex: That is, a teacher career structure should aim to align
Exploring teacher motivation, teacher career (broadly) what teachers want with what students, parents,
dynamics, and education bureaucracies communities, authorities, and education organisations
Teachers are embedded in a range of interconnected, want. Designing such teacher career structures entails
nested systems. As mentioned above, every classroom creating a working model of teachers’ individual and
11 i.e., the phenomenon of emergence, such that a system as a whole has distinct properties that are not properties of any individual
element of the system, but which emerge in the interaction between these elements.
12 Our emphasis on ‘coherence’ neither ignores nor disputes the classic Weick (1976) paper that characterises education organisations
as ‘loosely coupled’. Neither should ‘coherence’ or ‘alignment’ be seen as an argument that education organisations should be ‘tightly
coupled.’ We argue that loosely coupled organisations in a complex system can nevertheless be ‘coherent’ for purpose and that this
can be a much more desirable situation that organisations that are designed to be tightly coupled but only around process compliance
and ‘thin’ inputs.
13 By education authorities and organisations, we mean education ministries, other pertinent ministries (e.g., finance, human
resources), related government agencies (e.g., the national examinations board), the head offices of large private school chains, and
other organisations in positions of significant decision-making influence over teacher careers and over frontline education provision
more generally.
collective goals. In this section, we propose one such Economists have a helpful structure for thinking about the
model, which treats the motivations embedded in a difference between ‘finances’ and ‘circumstances’. If one
teacher career structure as a system (with the attendant thinks about well-being as ‘utility’ (in completely general
complexities). rather than value-laden terms), then economists define the
expenditure function as the money income it would take to
Finances, circumstances, satisfaction, achieve a level of wellbeing U at the prices one faces, p, in a
and validation as complementary given set of circumstances c.
sources of teacher motivation
e = e(U , pc)
Teacher motivation is complex, in that it is driven by
multiple factors that interact in complex ways (ILO, 2012, For instance, if a person lived in a very high-rent city and
para. 70; Watt & Richardson, 2007; see also Evans & Yuan, another in a low-rent city, then the money income to
2018, for a look at the range of teacher working conditions achieve the same standard of housing would be lower in
in developing countries and how they may affect the low-rent city, and hence the money income needed to
motivation and performance). achieve the same well-being (all other prices equal) would
be lower, or, equivalently, the ‘real’ (price-adjusted) wage
For the purpose of designing teacher career structures, we
would be higher in the low-rent city as the same money
propose four sources of teacher motivation. These four
wage could produce higher wellbeing. We can interpret
sources are framed from the perspective of the teacher.
‘prices’ very broadly to include access to infrastructure,
That is, each motivational source represents a different set
remoteness, amenities, etc. For instance, if someone
of reasons why a teacher could care about performance
live in one locale where there was reliable access to
in their profession—which, as we illustrate in Box 1, is
electricity from the grid at a constant price pelectricity, and
distinct from a framework that classifies aspects of teacher
another person lived where there was no electricity grid
motivation based on the types of policy instruments
and hence the effective price of electricity was a higher
involved in influencing it. Of the four sources of teacher
pno grid access, then the money income one would need to
motivation, two are pecuniary, and two are psychosocial: 14
achieve the same well-being is higher in the area without
Finances. This category pertains only to money, the grid access would be higher; and one could calculate the
purely pecuniary compensation that teachers receive for ‘compensating differential’ of grid access as the amount
doing their jobs. Finances receive the bulk of the attention by which, for a given person, money wages (i.e., finances)
in many large-scale, system-level studies of teacher would need to be higher for a person to be just indifferent
motivation—which, in our judgement, is disproportionate. between the two circumstances.
(We discuss this further in Section 2.2 on policy reforms that
Personal satisfaction. This category, the first of the two
are incoherent with the complexity of teacher motivation.)
psychosocial sources of motivation, spans a range of
Material circumstances. This category pertains to the individual non-pecuniary sources of psychological well-
quasi-pecuniary, i.e., the things that money can buy, such being, including goal fulfilment and identification with
as the added quality of life that might come from being a larger cause.15 To give one example of the influence of
assigned to a school in a well-connected modern city (or, personal satisfaction on teachers’ career choices, the
alternatively, in the unpolluted, tranquil countryside close 2019 teacher selection process in Peru incorporated an
to their hometown). Such circumstances can be a powerful RCT of how different types of behavioural priming affect
motivator. A recent field experiment demonstrated the pre-service teachers’ interest in applying for vacancies
use of postings as an incentive and saw large gains in in disadvantaged schools. Candidates were primed with
tax revenue growth under a merit-based intervention reflection exercises and messages that emphasised
where Pakistani tax officials who demonstrated better either the hardship allowances and career acceleration
performance were more likely to be posted to their associated with employment in disadvantaged schools
preferred locations (Khan, Khwaja, & Olken, 2019). (finances and circumstances), or the altruistic nature of
14 We are deliberately avoiding (at least for now) the more familiar categorisation as ‘intrinsic’ and ‘extrinsic’.
15 In Ryan and Deci’s (2000) taxonomy of motivation, such personal satisfaction could include a combination of intrinsic motivation
(pleasure from an inherently interesting or enjoyable task), integration (pleasure from a task that is fully congruent with your values and
needs), identification (pleasure from a task that serves a personally valued goal), or some forms of introjection (pleasure from a task
that reduces guilt or enhances self-esteem).
such service (personal satisfaction). Relative to a control Teachers’ career decisions are based on a
group that received neutral messages, both treatments complex combination of motivational sources
led to small but significant increases in the proportion
of candidates applying for vacancies in disadvantaged Teachers make decisions about whether to enter and
schools—with the altruistic prime also leading to a remain in the profession, which goals to prioritise, and how
significantly higher likelihood of being matched with much effort to channel toward those goals depending on
a disadvantaged school in the next selection phase the combination of finances, circumstances, satisfaction,
(Ajzenmann et al., 2020).16 In other words, personal and validation that they expect to gain from their work. It
satisfaction was at least as strong a source of motivation is important to underscore that finance is not the only, nor
as finances and material circumstances in the preferences even necessarily the most important, source of teacher
of these pre-service teachers—despite the outsized policy motivation.
and research attention to the two pecuniary sources. We can extend the simple expenditure function framework
Social validation. The second category of psychosocial to include personal satisfaction, PS, and social validation,
things that teachers might want (and, consequently, SV. In comparing teaching to another occupation, the
that might affect teacher motivation) is the respect or compensating wage differential would be the money
appreciation they receive from those around them for income difference to make a person just indifferent in
fulfilling socially valued goals. This overlaps with what Fehr working in the two occupations:17
and Falk (2002) describe as the desire for social approval. Compensating wage differential =
For example, a teacher performance pay intervention in
e(U , p , PST , SV T) -e (U , p , PSOther , SVOther)
Tanzania successfully raised student outcomes while being
viewed positively by teachers (impact evaluation in Mbiti Compensating wage differentials can also be used to
et al., 2019; teacher perceptions survey in Mbiti & Schipper, compare different teacher career structures or working
2021). Notably, an interview study of the same intervention conditions. For example, discrete-choice experiments with
found that one factor underlying these positive teacher 2,200 teachers in England found that moving from a school
views were visits by the external implementor to the culture with unsupportive school leadership to one with
school and community, which felt like a recognition of the supportive school leaders was valued as equivalent to a 9
teachers’ status—which was especially influential given percent increase in annual pay, with a similar valuation of
that many of the teachers felt that the profession has lost a supportive peers (Burge, Lu, & Phillips, 2021). Additionally,
great deal of public respect over time (McAlpine et al, 2018). if a teacher were to move from a school with good student
Additionally, more than half of the interviewed teachers behaviour to one where poor behaviour disrupts most
said that they shared some of their performance bonuses lessons, it would take a 26 percent pay rise for the teacher
with other teachers in their schools who were not eligible to feel adequately compensated (ibid). Such collegial
for the performance pay intervention, again indicating support and congenial classroom conditions relate to
the social nature of motivation (ibid). More generally, personal satisfaction and social validation from teaching,
the role of social validation has been incorporated into rather than to the pecuniary sources of teacher motivation.
economic models in which the amount of effort that an
That said, a few complications confront any education
agent is willing to exert (i.e., how motivated the agent
authority or organisation seeking to optimally attract,
is) is influenced by signals from the principal about how
retain, and motivate teachers by anticipating teachers’
favourably they regard the agent (Ellingsen & Johannesson,
motivational patterns. For one thing, these motivational
2008) and by opportunities for the agent to broadcast
factors don’t necessarily combine in a straightforward
reputation-enhancing signals about their own altruism
linear sense. An interview study of teachers in Tanzania
(Bénabou & Tirole, 2006).
found that the limited finances and reduced circumstances
16 For a review of research on attracting teachers to disadvantaged schools, which includes a discussion of the Anjenmann, et al. (2020)
paper, see Evans, D. K., & Acosta, A. M. (2021). How to Recruit Teachers for Hard-to-Staff Schools: A Systematic Review of Evidence from
Low- And Middle-Income Countries (No. 595; CGD Working Paper). Center for Global Development. https://www.cgdev.org/publication/
how-recruit-teachers-hard-staff-schools-systematic-review-evidence-low-and-middle-income.
17 The idea of the compensating wage differential goes back all the way to Adam Smith.
of teachers in village environments also compromised the different structures of teacher performance pay in a private
satisfaction they derived from their work, because these school chain in Pakistan found that the performance pay
material inadequacies prevented them from embodying scheme had a positive and significant treatment effect
a socially desirable example of an educated person, on the test scores of those students whose teachers had
which undermined their professional identity (Barrett, stated, in a pre-treatment questionnaire, that they would
2005). Conversely, higher salaries for NGO leaders in prefer performance-based pay to a fixed salary, whereas
Pakistan reduced their social validation because would-be treatment effects on those students whose teachers
supporters were less likely to view them as altruistic, even preferred a fixed salary were insignificant (C. Brown &
though these salaries were earmarked funds from well- Andrabi, forthcoming).18
meaning external donors (Bano, 2012).
For a context-level example, Figure 6 shows urban and rural
Another complication is that desirable combinations of teachers’ level of agreement with the statement ‘I would
motivation vary across contexts and individuals. For an recommend my school as a good place to work’, based
individual-level example, a survey of teacher motivations on PISA-D teacher questionnaires. In Senegal and Zambia,
in England identified four broad types of teachers teachers in rural schools are much more likely to disagree
(practitioners, moderates, idealists, and rationalists), who or strongly disagree with that statement than their urban
were driven by different sets of motivations (Menzies et counterparts, thus indicating that material circumstances
al., 2015; see also Barrett, 2008, and Hagos et al., 2018, on can make a significant difference to the hedonics of teacher
teacher identity in Tanzania and Ethiopia, respectively). careers. However, these urban-rural differences are not
Such individual inclinations can strongly affect teacher consistent across countries, thus indicating (with the usual
practice and student outcomes. An RCT examining caveats about self-report survey data) that the hedonics
Figure 6. The factors affecting teachers’ motivation and job satisfaction (e.g.,, urban vs. rural school locations) can
vary across contexts
Source: Authors, using PISA-D data (OECD, 2018).

Notes. Sample comprises teachers in both publicly and privately funded schools that teach 15-year-old students. Data are
unweighted because PISA sampling weights are calculated for representativeness at the student level, not the teacher level.
Missing data are deleted listwise.
18 However, in a teacher performance pay experiment in Rwanda, Leaver et al. (2021) do not find any evidence that teachers who were
recruited to the teaching profession under an advertised performance pay contract exert any more effort in response to performance
pay than their peers who were recruitment under an advertised fixed wage contract (but who subsequently experienced performance
pay in-service). As noted above, teacher motivation is complex and can vary across contexts.
of teacher careers vary across contexts. More generally, a career structures that can encompass both hedonics and
key contextual feature in the relationship between teacher dynamics. This model consists of five design elements of
motivation and teacher career structures are the social teacher careers across four career phases.
and professional norms influencing teachers in a given
setting. These will be discussed in the next subsection. Design elements of teacher careers
As shown in Table 4, this model comprises five design
Teacher careers are hedonic and dynamic, and
elements—delegation, information, support, finance, and
any model of teacher careers must reflect this
norms, which are modified from RISE education systems
As noted in Part 1, teacher careers involve a range of framework (Pritchett, 2015; Spivack, 2021; both of which
‘hedonics’ (degrees of preference or aversion toward are an evolution of the World Bank’s World Development
different parts of the whole of compensation), and teacher Report 2004 ‘accountability triangle’)—and four phases of
careers also have an important element of dynamics. We teacher careers—pre-service, novice, experienced, and
now lay out a way of mapping the components of teacher veteran.
Box 1: How do the four sources of teacher motivation differ from other
classifications of motivation in teacher careers?
While some frameworks look at teacher motivation within teacher careers through the lens of the policy instruments that
influence teacher motivation, the four sources of teacher motivation in this primer focus on what a teacher might want
and how these desires shape their decisions about whether to enter the profession and what to prioritise once they are
within the profession. These perspectives are complementary, with the former focusing on education authorities and
organisations and the latter focusing on teachers. Neither is better than the other. Rather, they serve different purposes.
For example, Bruns and Luque (2015) developed a framework for classes of incentives that motivate teachers, building
on Vegas and Umansky (2005). These three classes of incentives, as shown in Table O.21 in Bruns & Luque (2015), are:
• professional rewards, such as intrinsic motivation, recognition and prestige, mastery and professional growth, and
well-equipped, congenial working conditions;
• accountability pressure, such as managerial feedback, client feedback, and job stability policies;
• financial incentives, such as bonus pay, pension and benefits, and salary differentials.
By classifying different policy levers for influencing teacher motivation, this framework can aid education authorities and
organisations in choosing the appropriate configuration of incentives for influencing teacher motivation in their context.
This can include evaluating whether the configuration of incentives is balanced across the three different classes and
ensuring that all of the deployed incentives are coherent and well-aligned rather than working at cross-purposes.
In turn, the four sources of teacher motivation in this primer—finances, material circumstances, personal satisfaction,
and social validation—focus on teachers’ goals and decisions. The goal of this classification is to contribute to the larger
understanding of hedonics and dynamics in teacher careers, as developed in this primer. Additionally, it can also aid
education authorities and organisations in thinking through how teachers might respond to changes to the career
structure.
Given the subjectivity and context-specificity of individual-level perspectives, the four sources of teacher motivation
and the three classes of incentives do not map neatly onto each other. For example, making job stability contingent
on a certain level of performance would constitute a form of accountability pressure from a policymaker’s perspective.
However, from a teacher’s perspective, how they respond to such performance-based job stability depends on how
they perceive it. Most teachers would value, albeit to different degrees, the long-term gains in finances and material
circumstances that job stability represents. Some teachers may be motivated to work toward permanent job tenure
not only for financial and material reasons, but also because of the social validation that comes with having attained a
desirable status and performance benchmark. Conversely, some teachers may regard performance-based job stability
policies as an illegitimate intrusion on what should be a peer-regulated (rather than externally dictated) professional
career, and this implied decline in social validation of the profession may instead demotivate them.
Table 4. A hedonic and dynamic model of teacher career structures
Phases of teacher careers

(approximate duration)
Pre-service Novice Experienced Veteran
<0 years Novice 0–5 years 5–25 years >25 years
Delegation Systemwide
Who does what purpose
where—and Individual
why? assignment
Information
How well are they doing it?
Design Support
elements What would help them to do it
of teacher better?
careers
Finance
Does compensation attract, retain,
and motivate good teaching?
Norms Professional
How should
Organisational
‘good teachers’
act? Social
Source: Authors, with the design elements modified from Pritchett (2015) and Spivack (2021).
The implicit theory in this model is that education Two clarifications are in order. First, when we say
authorities and organisations can best achieve their ‘teachers’ in this framework, we also mean educational
goals by building and sustaining a teaching corps who administrators in schools, district offices, and central
are motivated and equipped to engaged in practices that governments, especially when such roles are potential
cultivate student learning. This entails: stages in a teacher’s career. In systems where education
staff move between roles in teaching and administration,
• dentifying and communicating clear, consensus-based,
the full complement of roles needs to be included in the
achievable priorities, through appropriate delegation
teacher career approach. Even in education systems where
of systemwide purpose and individual assignments
such administrative roles are not part of the conventional
and, indirectly, though the priorities signalled in the
teacher career pathway, they still need to be taken into
information used to assess teachers’ performance;
account in the design of teacher career structures because
• providing teachers with resources for achieving of the crucial roles that they play in the teaching profession,
these priorities, through support for classroom not least in supporting teachers to improve classroom
instruction, finance as compensation for delivering practice and student learning (Childress et al., 2020;
that instruction, and information for feedback on how Education Commission, 2019).
they can improve their work;
Second, while each design element is conceptually distinct,
• ensuring that these four design elements are coherent they all act in concert (that is, the design elements do not
with organisational, social, and professional norms act sequentially), and coherence across various elements
that shape local understandings of what constitutes and career phases is key to the overall effectiveness of
appropriate and desirable teacher practice—and using teacher career structures, as discussed below in Section
policy design and communication to reshape those 2.2.
norms where they are incompatible with student
The two design elements that are arguably the most
learning.
pivotal—i.e., delegation and norms—are also the least
prominent in typical discussions about teacher careers. Teacher careers also fall within the larger complexity of
While research and policy debates swirl around questions education policy. Within a given education system, policy
of the structure of teacher salaries (finances), how teachers structures interact with each other, and the same policy
should be appraised (information), and, to a lesser but still can have different effects depending on whether it is
significant extent, pre-service teacher training and incoherent with other, pre-existing policies (Pritchett, 2015,
service professional development (support), delegation 2017; Bruns & Luque, 2014).
and norms typically receive short shrift.
Finally, teacher careers also constitute the intersection—
The linchpin among these five design elements is and, ideally, the convergence—of the goals of three
delegation. ‘Delegation’ is a term of art in the literature sets of actors, who themselves are heterogeneous. As
about accountability and reflects the systemwide purpose noted at the start, when education authorities and
of the whole teacher career structure, and indeed the organisations have clear goals, they want to attract, retain,
entire education system and its constituent elements and motivate a pool of teachers who will further their
and organisations. If clear delegation is absent, finance, goals. The teachers themselves want to get a reasonable
support, and information can become untethered. That is, combination of finances, circumstances, satisfaction, and
one cannot even assess whether someone is going a ‘good’ validation for their efforts. However, a third set of actors
or ‘bad’ job unless the job has a clear and shared purpose. is involved: communities, families, and students, whose
Without a clear purpose, everything can trend to a lowest goals may or may not overlap with state authorities’ and
common denominator of mere process compliance. In bureaucrats’ goals. To use the language of principal-agent
turn, the degree to which these other design elements accountability, teachers are the ‘agents’ in accountability
are coherent with purposes and priorities can influence relationships with two different sets of principals: not
whether the stated delegation is seen as a credible only education authorities and organisations, but also
statement of expectations to be upheld (and hence as a communities and families. Delegation is the means
set of expectations around which both formal information, through which these distinct relationships and goals may
financial rewards, and informal norms are constructed)—or, be brought into a workable degree of consensus-based
conversely, whether ‘everyone knows’ that the statements alignment.19 (For other some other actors who can strongly
about certain ‘goals’ are toothless, de jure, rhetorical influence consensus and coherence in teacher careers
statements that bear little resemblance to reality and but are not direct principals or agents in these two main
have few implications for anyone. This systemwide accountability relationships, see Box 2.) In short, this is not
determination of purpose is the why of delegation. a technocratic model of how the education ministry should
organise teacher compensation. It is an integrated social
At a more granular level, the systemwide purpose
model of how a society supports a corps of motivated,
manifests in the individual assignment aspect of
capacitated teachers to produce education.
delegation—the specification of who becomes a teacher,
what they are expected do in their individual allocations The social nature of this model comes to the fore in the
of subjects and tasks, and where they do it—also shapes fifth but far from least important design element: norms.
some underappreciated sources of teacher motivation (as Norms are dominant and often entrenched perceptions,
discussed below). beliefs, and understandings about who teachers are and
what they should (and shouldn’t) do. They may involve
A key reason why delegation and purpose are crucial is that
perceptions of teachers’ status in the occupational
teacher careers are complex, in a few different ways.
pecking order, beliefs about acceptable ways of punishing
To begin with, classroom teaching itself is a complex an intransigent student, and tacit understandings about
human endeavour, as noted in the introduction. At the whether a directive from the district office needs to be
level of policy design, as shown in Table 1, teacher career obeyed in practice or merely on paper. (For a review of
structures involve multiple design elements and multiple the economic literature on occupation-related norms,
career phases. Tweaking any element for any phase might including effort norms, professional norms, and identity
alter the configuration in other parts of the structure.
19 Relatedly, the International Labour Organization states that ‘social dialogue’, i.e., consultations and negotiations between
employers, workers, and government representatives, is ‘the glue for successful education reform’ (ILO 2012, p. 201).
norms, see Rebitzer & Taylor, 2011.)20 categories of norms that affect teachers and teaching:
professional norms that are specific to teachers and their
Here are two empirical illustrations of the power of norms.
practice; organisational norms that influence the education
First, a quantitative study based on surprise visits to a bureaucracy or the civil service more generally; and social
nationally representative sample of early-grade teachers norms about societally valued aspects of education, day-
in Tanzania found that only 40 percent of teachers were to-day interpersonal interactions, and other facets of local
present in the classroom—but also that a teacher was life that affect the schools and classrooms embedded in
much more likely to be in the classroom (i.e., there was them.22
a statistically and substantively significant relationship)
Regardless of their sources, these norms can shape
if a larger proportion of their colleagues at the school
how any changes to delegation, information, support,
were also present in the classroom (Schipper & Rodriguez-
and finance may be interpreted and implemented,
Segura, 2021). This at least suggests the possibility of norm
sometimes overturning the theory of change in apparently
convergence within schools.
well-designed policy reforms (for one example, see
Second, a qualitative case study of a public-school Muralidharan & Singh, 2020).23 Two cross-country patterns
improvement programme run by a philanthropic in teacher career policy that have negatively affected
foundation in Pakistan found that many low-performing teacher norms are the widespread single salary structure
schools had, at the outset, ‘anti-work’ norms rooted in with annual increments based solely on seniority, which
the fact that teachers and principals were appointed can demotivate those committed teachers who see their
primarily through a system of political patronage—such less committed colleagues gaining identical recompense
that teachers often sat with colleagues drinking tea during (Crehan, 2016); and rapid, insufficiently supported
school hours, or sent children to run personal errands for expansions of schooling and of the teaching profession
them (Bano, forthcoming). Under these circumstances, that led to a weakening of professional norms (Béteille et
teachers who had entered the profession out of a desire to al., 2020; Fuller, 1991).
cultivate student learning (rather than to reap the benefits
While imprudent policy decisions can compromise teacher
of partisan support) were often subject to colleagues’
professional norms, effective policy decisions about
active attempts to demotivate them, i.e., to get them to
teacher careers can also reshape and revitalise teacher
comply with prevailing norms, by ridiculing their efforts or
norms (Tournier & Chimier, 2019. Aiyar et al., 2021, is a
questioning why they are working so hard (ibid).21
richly detailed, three-year, embedded study of a reform
Teacher norms have numerous sources, whether in Delhi. They show that implementation was hampered
longstanding or recent, internal to the profession or shaped by pre-existing norms but that sustained effort managed
by wider societal views. In this model, we highlight three to shift some entrenched norms in crucial ways. Such
20 In some other conceptualisations, norms are either equivalent to, or a subset of, culture. For example, Roberts (2004), describing
the culture as one of four components of organisations (the others being people, architecture, and routines) writes that: ‘Culture is the
“softer” stuff, but it is no less important for that. It involves the fundamental shared values of the people in the firm, as well as their shared
beliefs about why the firm exists, about what they are collectively and individually doing, and to what end. It also encompasses the
special language used within the firm, which shapes thought and action. Culture also involves the fundamental mindsets of the firm’s
members and the mental models they have, which determine how they see themselves and the firm and how they interpret events. Most
significantly, it involves the norms of behavior that prevail in dealing with other members of the firm and with outsiders. Culture defines
the context in which the relations among people develop and operate and set the basis for the implicit contracts that guide and shape
decisions. It operates as a social motivation and control system’ (p. 18).
21 Encouragingly, this study found that the school improvement programme often succeeded in shifting schools toward a ‘pro-
work’ culture by through a combination of embedding highly motivated teachers in the school who were tasked with identifying and
encouraging existing ‘pro-work’ teachers and respectfully building relationships with less motivated colleagues, alongside establishing
and enforcing a set of basic rules for good teaching practice (e.g., fulfilling a certain number of teaching periods, discouraging rote
learning, not having breakfast during working hours, and not collecting funds for any purpose; Bano, forthcoming).
22 Beyond these three broad categories, other sources or forms of teacher norms may affect particular subgroups of the population
(see, e.g., Jacinto & Gershenson, 2021, on how American children whose mothers are teachers are more likely to enter the profession).
23 It is also worth noting that the prevailing norms in an education system can profoundly shape the socialisation that children
experience in school. Given that the classroom and the school are the social ecosystems where school-going children spend the bulk of
their public life during their formative years, the consequences of inhabiting social ecosystems shaped by demotivated adults and unruly
peers, neither of whom gain much satisfaction from the interaction, may be far-reaching. We explore this point a bit more in Section 3.2.
reshaping will necessarily be a fundamental part of overarching goal needs to be manifest in specific and
transforming a pool of teachers from a low-performing, differentiated priorities for teachers at different phases of
ill-equipped, and demotivated occupational group into a their careers.
strongly performance-normed, contextually embedded
Dynamics matter because teachers typically have
profession.
different challenges and capabilities at different stages of
Phases of the teacher career cycle their careers (e.g., Hargreaves & Fullan, 2012; Huberman,
1988; Pritchett & Murgai, 2006). A targeted approach
Besides the five design elements that represent the
allows administrators to strategically allocate limited
‘hedonics’ of attracting, retaining, and motivating teachers,
administrative and financial resources to support
the teacher career model also includes four phases of
teachers and optimise for the overarching goal of student
teacher careers: pre-service, 24 novice, experienced, and
learning across these differentiated profiles. The task is
veteran. As noted in the introduction, the overarching
not over when education authorities and organisations
systemwide goal of teacher workforce management for
have attracted appropriate candidates to the teaching
systems coherent for learning should be to attract, equip,
profession; they also need to ensure that career structures
retain, and motivate teachers who consistently engage in
retain effective teachers for long-term careers and
practices that effectively cultivate student learning. This
Box 2: Resources on teacher educators and mid-tier education bureaucrats

Besides teachers, education authorities and organisations, and communities, families, and students, some of the
other key actors who can affect the capability and commitment of teachers include:
Teacher educators and teacher education institutions, whether those providing pre-service or in-service training.
Open-access research resources on teacher educators and education institutions include:
• Akyeampong, K., Pryor, J., Westbrook, J., & Lussier, K. (2011). Teacher Preparation and Continuing Professional
Development in Africa: Learning To Teach Early Reading and Maths. University of Sussex Centre for International
Education.
• Popova, A., Evans, D. K., Breeding, M. E., & Arancibia, V. (2021). Teacher Professional Development around the
World: The Gap between Evidence and Practice. The World Bank Research Observer. (See also tools from the World
Bank’s Coach initiative for improving in-service teacher training.)
• Tatto, M. T., et al. (2012). Policy, practice, and readiness to teach primary and secondary mathematics in 17
countries: Findings from the IEA Teacher education and Development Study in Mathematics (TEDS-M). International
Association for the Evaluation of Educational Achievement.
Mid-tier bureaucrats (i.e., those in between the frontline and the central ministry), such as district-level administrators,
instructional coaches, and inspectors. Open-access research resources on the role of mid-tier actors and other
intermediaries include:
• Childress, D., Chimier, C., Jones, C., Page, E., & Tournier, B. (2020). Change agents: Emerging evidence on
instructional leadership at the middle tier. IIEP-UNESCO; Education Development Trust; Education Commission.
• Education Commission. (2019). Transforming the Education Workforce: Learning Teams for a Learning Generation.
New York: Education Commission.
• Mundy, K., & Asim, M. (forthcoming). Of agents and agency: The missing middle in educational reform in
developing countries. RISE Working Paper Series.
24 Although those in the pre-service phase are not actively contributing to student learning (except during the practicum or practice
teaching stints), this phase is critical for the selection of teachers into the profession, for the support and socialisation of teachers, and
for the career-long value proposition for prospective entrants to teacher training courses.
motivate teachers to continually cultivate student learning goals for veteran teachers include creating appealing
throughout the span of their careers (Pritchett & Murgai, pathways (a) to retain and motivate highly effective
2006). teachers to amplify their contributions to student learning
by offering mentorship and instructional leadership to their
The ranges for years of experience for each phase of a
colleagues, and (b) to encourage those teachers whose
teacher career are only approximate.
competencies or motivation have diminished to retire early
Pre-service. During the pre-service phase, potential rather than remaining in the classroom—or, alternatively,
entrants to the teacher workforce receive their initial to be reassigned to administrative and supervisory roles.
training. System-level goals for the pre-service phase
The different needs and priorities of different teacher
should include attracting high-potential candidates to
career phases come to the forefront in teacher career
initial teacher training; then equipping the pool of potential
progression reforms, which have been recommended
teachers toward foundational instructional competencies
in flagship teacher policy documents from a range of
and desirable professional norms via training and
organisations (e.g., Education Commission, 2019; Bruns
socialisation; and finally selecting novice teachers from this
& Luque, 2014, and Evans & Betaille, 2019, for the World
pool based on demonstrated pedagogical competencies.
Bank; Crehan, 2016, and Tournier & Chimier, 2019, for IIEP-
Novice teachers range from new entrants to those who UNESCO).
have approximately 5 years of teaching experience. The
novice phase is typically a period of rapid growth in Teacher motivation and the design
teaching competencies (Araujo et al., 2016; Bau & Das, elements of teacher careers
2017; Hobbis et al., 2020; Kraft et al., 2020; Papay & Kraft,
As shown in Table 5, each of the five design elements of
2015; Podolskly et al., 2019). System-level goals specifically
teacher careers can affect teacher motivation.
targeting novice teachers include continuing to equip
them with instructional competencies; motivating novice Delegation can and should affect all four motivational
teachers as they acclimate to classroom challenges; and sources. For example, the systemwide purpose aspect of
encouraging the retention of promising teachers, alongside delegation influences what a teacher would regard as a
(tacit) mechanisms for encouraging the (overwhelmingly fair compensation structure (i.e., finances and material
voluntary) turnover of teachers whose performance and circumstances) for the objectives they are expected to
motivation are incompatible with the education priorities meet, while the degree to which they identify with this
of the system. systemwide purpose can shape their sense of professional
fulfilment (i.e., personal satisfaction), and the degree
Experienced teachers have between 5 and 25 years of
to which the systemwide purpose aligns with dominant
teaching experience. System-level goals for experienced
societal values affects the respect that they enjoy (i.e.,
teachers are retaining experienced teachers who
social validation).
cultivate student learning, and continuing to equip and
motivate them by fostering their continuous professional To give an example related to the individual assignment
development rather than allowing them to settle into aspect of delegation, assigning a teacher to a school in a
performance plateaus (Hobbiss et al., 2020). Part of socioeconomically privileged neighbourhood can increase:
retaining experienced teachers will almost certainly be their likelihood of earning a performance-based salary
compensation that rises with seniority, but not merely increment (finances), the lifestyle amenities that they
because of seniority per se, but because the efficacy and can conveniently access (circumstances), the magnitude
contributions of teachers to educational goals rise with of the learning gains that their better-prepared students
seniority. demonstrate (satisfaction), and the recognition that they
receive from parents who can afford to pay more attention
Veteran teachers have been in the profession for over 25
to their children’s education (validation).
years. At this career phase, the range of possible teacher
contributions to student learning widens. On one hand, As for the other design elements, information, support,
many veteran teachers have cumulatively built a thorough, and norms affect the psychosocial sources—i.e., teachers’
contextually relevant body of knowledge about how to personal satisfaction and the social validation that
effectively cultivate student learning. On the other, some they receive—by conveying feedback about whether
veteran teachers may have depleted levels of commitment they fulfilled professional goals, by influencing their
to student learning, due to gradually accumulated erosions capacities for fulfilling those goals, and by influencing
to their motivation over time. Accordingly, system-level the identification and interpretation of the goals. Most
obviously, finance shapes teachers’ finances and finances by determining the allocation of positions across
circumstances, but it also affects the social validation that individuals and by benchmarking their performance in
they receive because these pecuniary benefits can also be those positions. Locally embedded norms can powerfully
powerful signals of status. shape how any teacher career design element is perceived
and, consequently, how its motivational effects are felt
All that said, the relationships of influence summarised
by individual teachers. In particular, it is important that
in Table 5 drastically simplify the complexity of these
overall compensation structures are perceived as fair and
relationships by focusing only on direct influences. In fact,
equitable, and such perceptions are tightly intertwined
indirect influences of design elements on motivational
with norms (see Section 4.2 for some examples).
sources abound. Delegation and information also affect
Table 5. Most design elements of teacher career structures can influence either the pecuniary or psychosocial sources of
teacher motivation—but delegation can directly affect all four sources
Can the design element influence the sources of teacher motivation?

Pecuniary sources Psychosocial sources
Material Personal
Finances Social validation
circumstances satisfaction
Delegation ✓ ✓ ✓ ✓
Information ✓ ✓
Design elements
Support ✓ ✓
of teacher careers
Finance ✓ ✓ (✓)
Norms ✓ ✓
A key takeaway from Table 5 is that delegation is interact to influence teacher motivation.
fundamental to teacher motivation. 25 Delegation is
Besides the many ways in which the design elements
not only essential to setting the purpose(s) for which
of teacher careers can influence the sources of teacher
teacher careers are oriented, as discussed above, but it
motivation, another significant interaction between
is also fundamental to teacher motivation because the
design elements and motivational sources is the parallel
determination of ‘who does what, and where, and why’ can
importance of, on one level, delegation and norms in
affect all four sources of teacher motivation. On the surface,
safeguarding the purpose of teacher career structures and,
education authorities and organisations’ expectations of
on another level, personal satisfaction and social validation
teachers may look straightforward: teachers should teach.
in safeguarding the purpose of teachers’ individual and
But classroom teaching is a complex task that needs to
collective work.26 Specifically, we posit that when teachers
be supported by many interacting but discrete activities.
in an education system have depleted levels of professional
Managing and maintaining healthy school communities
satisfaction and validation (or when the delegation and
entails another set of activities, including administrative
norms of teacher career structures are coherent for
work that falls partially on teachers. At the scale of a school
purposes unrelated to student learning), then the system’s
system, these numerous activities are cross-cut by the
immune system for withstanding dysfunctional pressures—
need to allocate different activities to specific people in
such as pressures to prioritise goals other than children’s
different places, and to indicate the relative importance
learning and development—will be compromised. This is
of each responsibility. In short, there are many possible
because purpose, expectations, and goals are inherent
combinations of teachers, tasks, and settings, all of which
25 For more on the role of purpose in sustaining intrinsic motivation, see Ryan and Deci (2000) and Jeevan (2021).
26 For both levels, one component in each pair pertains to the internal calibration of priorities (whether by education authorities
and organisations in the case of delegation or by individual teachers in the case of satisfaction); and the other component (norms and
validation, respectively) pertains to the influence of wider social patterns on how these priorities are interpreted.
to satisfaction and validation, whereas the utility derived incentives for teaching in disadvantaged schools.
from finances and circumstances are much less contingent
A similar dynamic operates in the allocation of teachers
on purpose.27 In a subsequent subsection, we discuss this
to classes. This can take different forms. In schools where
dynamic at the level of the organisation.
classes are streamed by ability within grades, being
When ‘delegation’ isn’t purpose-driven, the assigned to teach higher-ability classrooms is sometimes
used as a reward for favoured teachers, or as a way to boost
hedonics of teacher careers can heighten inequity
the number of top scorers in national exams—but this can
To illustrate the complexity of the relationship between widen within-school inequity, besides demotivating the
delegation and motivation, consider the many potential less-favoured teachers (Gamoran & Berends, 1987). In the
inequities in the allocation of teachers to schools and to many countries where the early-grade classrooms are
classrooms within schools, i.e., the ‘where’ of individual immensely overcrowded but higher grades typically have
assignment within delegation. As noted above, there are very small class sizes (due to a combination of repetition of
many hedonic advantages to teaching in a school with the early grades and dropout at higher grades), the natural
more socioeconomically privileged and/or academically hierarchy of seniority often means the most effective
prepared children. These advantages have tangible teachers assign themselves to teach higher-grade classes
effects on teachers’ career choices. For example, a quasi- that are easier and more satisfying to teach both because
experimental study using a change in upper secondary they have fewer children and because those children
school admission rules in Stockholm found that an who have persisted in school are the most academically
increase in the GPA of incoming students to a school inclined of their cohorts—when these teachers would
substantially and significantly reduced the likelihood of contribute far more to aggregate student learning if they
teachers choosing to leave the school, whether to seek taught the early grades (Schiefelbein et al., 1999; see also
employment in other schools or to leave the profession Crouch & Merseth, 2017).
(Karbownik, 2020). An analysis of nationally representative
All of these potential inequities arising from the hedonics
teacher surveys in the U.S. found that teachers are
of allocation to schools and classrooms are a particular
more likely to leave lower-performing schools if these
danger for the many education systems in which public-
schools are additionally subject to performance-based
sector teachers are all on a common salary scale that moves
sanctions, although the effects of both sanctions and of
in lockstep with years of service (or other such criteria that
lower performance can be counterbalanced by supportive
fit easily into bureaucratic databases but have little to do
school leaders and greater teacher autonomy (Ingersoll,
with student learning). In such systems, the allocation to
Merrill, & May, 2016). An unfortunate corollary is that those
schools and classrooms is one of the few areas of systemic
teachers who have the most influence (whether due to
variation in the hedonics of teacher careers—which
pedagogical performance, social standing, years of service,
heightens the relative incentives for teachers to use time,
or other traits) within the education bureaucracy are often
effort, social capital, and any available loopholes to attain
disproportionately concentrated in these more desirable
favourable assignments.28 In this case there is an ‘illusion
schools. This can occur both in decentralised systems
of equality’ (Pritchett & Viarengo, 2009) and of fairness as
that allow all teachers to apply directly to the schools or
all teachers with the same objective characteristics have
districts that they want to work in (e.g., Engel & Cannata,
the same pay but there can be massive differences in the
2015, on the U.S.), and in centralised allocation systems
total compensation across teachers with the same pay as
that are vulnerable to patronage (e.g., Béteille, 2009, on
some have assignments they strongly prefer (e.g., teaching
India) or that prioritise the school assignment preferences
at a ‘good’ school near their home) whereas others have
of higher-performing teachers without commensurate
assignments they resent. This can lead to a reality in which
27 Another dynamic that may be at play is that experimental research on pro-social norms suggests that individuals in an anonymous
setting modify their behaviour in response to others’ self-serving norm violations but not their pro-social norm compliance, such that
the pro-social norms are eroded over time. However, when individuals know that they have characteristics in common with those they
are observing, they respond to others’ signals of both norm compliance and norm violation, such that the pro-social norms remain stable
over time (Bicchieri et al., 2020). In other words, social validation may shift norms.
28 Another area of systemic variation in hedonics is the allocation of teaching and non-teaching responsibilities, such as promotion to
school leadership, i.e., the ‘what’ of delegation. Other areas of variation, such as the quality of in-service training or respect for teachers
in the local community, tend to be idiosyncratic to the individual, school, community, or teacher trainer in question, rather than being
areas of systemic, policy-driven variation.
the administrative management of the schools spends procurement (to obtain the materials needed by people
nearly all its time on adjudicating claims about transfers in the core), and IT (to keep track of data and information
and postings and nearly no time on how to actually make to make informed decisions), and finance (to attract and
the schools better learning environments. deploy resources).
Effective organisations have a shared purpose However, organisations can sometimes lose their core
and become what we call ‘donuts’, as illustrated in panel
(b) of Figure 7. As the core has two elements, purpose and
While teachers and their various sources of motivation practices, the loss of core can happen in two distinct ways.
are embedded within classrooms and teacher career
An organisation could lose its core by losing the sense of an
structures, the classrooms and career structures in the
adequately shared purpose across the key actors. This loss
public sector are themselves embedded within education
of purpose can happen through erosion from the inside, as
bureaucracies. These bureaucracies span a range of actors
core stakeholders who are inside the organisation lose
(e.g., students, teachers, administrators), administrative
faith in the common goal, or through increasing conflict
levels and branches (e.g., schools, district offices, central
about multiple possible goals. This erosion of purpose
ministries, departments within ministries), and functions
can also happen from ‘outside’ if the organisation loses
(e.g., teaching, curriculum development, human resources).
the support of key stakeholders that provide necessary
Given this wide range of people and organisations,
resources (financial or other support).
education bureaucracies are enmeshed in many different
forms of organisational complexity (including the nature Organisations can also lose their core while maintaining
of relationships between central administrators and those a purpose if there is a loss of a commitment to a set of
on the front line, which we discuss below in Section 4.1). In effective technical practices. This can happen when
this subsection, we focus on how alignment across these external circumstances change and the organisation’s
complex, interacting elements is key to organisational agreed-upon technical practices are no longer effective.
effectiveness, whether in education bureaucracies or other
For instance, private sector firms often go bankrupt or
organisations.
shrink when the external market circumstances change
We posit that effective organisations are effective from and the core organisational practices are a mismatch
the inside out—and that this ‘inside’ or organisational for the current market conditions. Christensen’s (1997)
core has two elements. One, a purpose that is sufficiently classic analysis of ‘disruptive innovation’ argues that
shared by all actors, in their settings, and functions, many firms went bankrupt with the disruptive innovations
across the organisation (see also Besley & Ghatak, 2005, in the computing sector in the 1990s not because they
on organisational mission; Honig, forthcoming, on were badly managed or were not following industry ‘best
mission-driven bureaucrats; and Jeevan, 2021, on the practice’. Rather, they failed because they were well-
role of purpose in intrinsic motivation both individually managed and they were following best practices that were
and collectively). The second element of a core is a set of well-adapted to one set of market conditions. Since many
technical practices that are sufficiently share and expected early computing firms arose in a period in which being
by all actors to advance the organisation’s purposes. the industry technology leader was key to success, they
developed organisational practices consistent with that
This is illustrated in panel (a) of Figure 7. Scholars of
purpose. But ultimately market circumstances changed
organisational management have called this the ‘technical
as the available computing power actually exceeded the
core’ (Thompson, 2003/1967) or the ‘operating core’
needs of most users, and hence the ‘best practices’ for
(Mintzberg, 1979; see also Andrews et al., 2017). In any
staying at the technological cutting edge were ill-adapted
organisation that effectively fulfils the purpose at its core,
to the market—and were also impossible to change.
all of its other functions operate as services that support
the core—such as HR (to bring in and manage people),
Figure 7. The technical core—comprising a shared purpose and purpose-driven technical practices—is central to
organisational effectiveness
(a) Effective organisations are effective from the inside (b) If an organisation loses its core, it might continue
out: from their core purpose and the technical practices functioning as a donut, but it will likely get co-opted for
advancing that purpose. other goals.
(c) A core-less system cannot be fixed simply by fixing (d) But if HR and other service functions are not
HR or other service functions. These functions cannot coherent with the core, the core purpose must compete
fill the core. with other priorities.
Source: Authors
Another possibility is that an organisation is founded on economic study.29 There are also dynamics among not-for-
the expectation that a set of practices will be effective and profit organisations like private religious or charitable or
it eventually becomes clear that these practices were never educational organisations that also experience episodes
effective to begin with. of growth, or stagnation, or decline. Social movements
can arise, thrive, but then also disappear (for reasons both
A key element of the systems (or ‘ecosystem’) within
good and bad).30
which organisations exist is whether, and for how long,
donut organisations persist. In market systems for small However, organisations within the public sector (e.g.,
private firms there is often very large turnover (especially armies, police forces, regulatory agencies, etc.) are often
in their early years), and most new firms fail. At the unique and necessarily monopolies and hence cannot
same time, many large private firms become ‘too big to disappear, even if they have become ‘donut’ organisations
fail’ and shrink only slowly over time. Differences in the that have lost effectiveness and are hollow at their core.
dynamics of firm creation and destruction is a key area of
29 This is the process of ‘creative destruction’ that can lead to increases in economy-wide productivity even if firms themselves do not
learn over time if more productive firms are more likely to survive. Recent papers by Hsieh and Klenow (2009, 2014) suggests that a large
part of the difference in aggregate productivity across countries may well be differences in firm dynamics.
30 In U.S. history, for instance, the temperance movement seeking to limit or ban alcoholic beverages rose, successfully passed a
Constitutional amendment prohibiting alcohol, but then tides turned, the Prohibition was repealed, and the ‘temperance’ movement
waned.
There are three important phenomena that are common bribery directly by applicants to employees; rather, all of
among donut organisations. the side payments were intermediated by the touts. This
suggests that the organisation had not ‘lost control’ of its
First, many organisations that lack a core can nevertheless
agents but rather than the entire organisation had been
survive as donuts through inertia and process compliance.
‘repurposed’ as a way of generating revenue from extra-
Even though the organisation can no longer generate a
legal fees for the driver’s license. Also, since that revenue
shared and agreed-upon answer to the question of ‘why
was shared among various actors inside and outside the
are we doing what we are doing?’ for either ends (purpose)
organisation (e.g., politicians), the pooling of the collected
or means (practices), they can continue to go through the
extra-legal revenue had to be tracked carefully so that the
motions and, especially insofar as it projects an image to
parties could trust they were getting their share. When
the outside can pretend to be a ‘regular’ organisation and
organisations have a weak core, then alternative purposes,
not a donut.
like being used for ‘patronage’ for the employment of the
Second, many organisations that lack a core of shared politically connected, are incredibly difficult to resist (see,
purposes and practices nevertheless project themselves e.g., Bano, forthcoming).
as effective organisations through a phenomenon known
For example, some countries’ education systems, and
as ‘isomorphism’ (DiMaggio & Powell, 1983) or ‘isomorphic
therefore the education bureaucracies in those systems,
mimicry’ (Andrews et al., 2017) in which the organisation
lost a shared commitment to an achievable set of learning
imitates the outward appearances of other successful
outcomes as part of their core during periods of mass
organisations in their field. This reproduces the appearance
expansion of enrolment. This can be especially likely when
of an organisation without meaningfully advancing the
the expansion happened without enough support for
purpose for which it was intended (Pritchett, 2014). In this
teachers, at a time when the teaching corps and student
sense, even ‘donut’ organisations can use isomorphism to
body were getting more diverse in their backgrounds and
seem to be dynamic and actively engaged in ‘reform’. But
needs.
this isomorphism can generate decades of ‘reform’ efforts
that leave the organisation no more capable or effective Fuller (1991), drawing on fieldwork in Malawi and literature
than it was. Andrews (2013) documents that decades on other postcolonial education systems, argued that
of ‘reform’ in public financial management never really resource-constrained teachers in many weak postcolonial
improved the efficacy with which resources were spent as states inadvertently perpetuate top-down bureaucratic
governments often tactically adopted those changes that control rather than cultivating children’s development:
were superficial (like multi-year budgets) and which did
… the state legitimates and enforces bureaucratic
not affect the actual practices that mattered.
forms and rituals: telling headmasters to keep tidy
Third, ‘donut’ organisations can often be hijacked or class schedules, requiring teachers to punch in
captured by other actors or purposes while maintaining and punch out, demanding teachers to write out
a shell or appearance of effectiveness, when in fact each lesson plan. Such classic routinization and
the actors inside the organisation understand that it is breaking down seemingly complex tasks into simple
devoted to an entirely different purpose. As with many behaviors signals that the school, though poor in
phenomena, there is an interesting example from nature, material resources, can take on the attributes of a
where a fungus infects a certain species of ant and causes modem bureaucracy. … perhaps most importantly,
the ant to do what is in the best interest of the fungus and the state sets in place the mass conditions present
its reproduction—even though that behaviour kills the ant. in most Third World classrooms. Faced with up to
A fascinating study of the process of obtaining a driver’s ninety children, the teacher becomes preoccupied
license in Delhi, India, showed that the quickest and most with maintaining control, engaging a dense batch
reliable way to obtain a driver’s license was to hire a tout of youngsters, and reinforcing his or her own
who facilitated the process with the public agency and, in authority. The fragile state may spew rhetoric about
the process, subverted the step of taking an actual driving encouraging individual, personalized development
examination—such that 46 percent of study participants of the child. But the teacher must deliver a uniform
who had received driving licenses demonstrated so little package of facts and knowledge to all children, and
knowledge of how to operate a motor vehicle in a follow-up evaluate each along universal and simple criteria (p.
oral test that study administrators refused to let them take 132).
the practical driving part of the follow-up test (Bertrand
In other words, rather than having a shared purpose of
et al., 2008). Interestingly, there was almost no direct
cultivating student capabilities and a shared commitment the literature, such as: ‘making a measure into a target’31
to practices that promote that purpose, the core of or ‘seeking the wrong goal’ as in Meadows’ (2008) primer
education bureaucracies in some countries has been on systems thinking. The fundamental error here is
hollowed out and displaced by some combination of: reducing a ‘thick’, complex, principled purpose (such
a. process compliance enabled by service functions as student learning) to a ‘thin’, narrowly standardised
(e.g., class schedules, time clocks, and recorded target (such as student enrolment)—or, in other words,
lesson plans at the school level; and maintaining flattening accountability into mechanistic accounting
disciplinary control at the classroom level) that when accountability for complex tasks would be more
were initially intended to serve a core purpose but appropriately framed as detailed, embedded, relational
which are now simply compliance for compliance’s accounts or justificatory narratives between principals and
sake; agents (see Mintzberg, 1983, Chapter 10, on professional
bureaucracy or Honig & Pritchett, 2019, for examples in
b. isomorphic mimicry in which systems go through
education).
the motions of ‘reform’ and improvement and core
service functions like pre-service training (Warwick This is a fundamental reason why a donut organisation or
& Reimers, 1995) or in-service teacher training system cannot be rehabilitated simply by fixing the human
(Revina et al., 2020); and/or resources department or other service functions (such as
stronger and better EMIS), as shown in panel (c) of Figure
c. outright capture for other purposes like patronage 7. At the scale of a large education organisation that lacks
employment or contracts for construction or inputs. a core, HR can only focus on thin indicators, whereas
Organisations that have drifted from their core purpose classroom teaching and learning that leads to effective
can erode the motivation of mission-driven individuals learning is irreducibly thick and complex (see Honig,
within them. For example, in a qualitative study that forthcoming, for a similar argument).
followed novice teachers in Indonesia for the first two years To give one example, a large-scale randomised-control trial
following their graduation from a teacher certification in Madhya Pradesh, India, evaluated the effectiveness of a
programme, Alifia, Pramana, and Revina (forthcoming) school quality assurance intervention called Shaala Siddhi.
found that novice teachers often experienced a degree Shaala Siddhi was based on international ‘best practices’
of shock during their first year of teaching because of that were modified to (some aspects of) the local context.
mismatches between their understandings of what it The programme was implemented, in that compliance
means to be a good teacher and the non-teaching-related with the required steps was achieved. However, it changed
tasks that dominated their day-to-day work. The authors nothing in classroom practice and hence did not improve
further note that one reason why these non-teaching either teacher performance or student learning. In the
tasks were so dominant was that school funding was words of a headteacher at a school that was recognised as
contingent on the completion of administrative tasks—to an effective implementor of the intervention:
the extent that they took precedence over novice teachers’
There is a lot of documentation work. We have to
professional development as educators.
make a work plan and then upload it, get it printed.
Hollowed-out, ‘donut’ education There is so much paper work that by the time some
bureaucracies cannot be fixed without teachers figured that out they had forgotten what
was Shaala Siddhi itself. I do all the documentation
restoring their shared purpose
work at home because I have no time in the school
Many ‘donut’ education bureaucracies successfully funnel (quoted in Muralidharan & Singh, 2020, p. 18).
the overwhelming majority of children in their jurisdiction
Put differently, for teachers on the frontline, it felt like the
into schools, without corresponding success in helping
purpose of the intervention was the paperwork.
these children to cultivate foundational literacy and
numeracy, as shown in Section 1.2. A different illustration of the depth of the problem of
‘donut’ organisations in education is the prevalence of
This common phenomenon goes by a number of names in
31 One version of this is known as ‘Goodhart’s law’ (1984), in which the very act of turning a particular measure into a target changes
the behaviour of the organisation—such that, even if there was a close observed association between the measure and the desired
success before the measure became a target, this close relationship can be distorted by target-oriented changes in organisational
behaviour, and achieving the measure when it is a target no longer produces success.
widespread cheating on tests, which has been documented across nearly all schools. This illustrates that if the core
in India (Singh, 2020) and Indonesia (Berkhout et al., 2020), purpose of student learning had been displaced, then
among other countries. Singh (2020) shows that in Madhya adopting and attempting to promote a thin target of raising
Pradesh, India, even the apparently low-stakes (for student exam scores may well generates pressures to achieve that
or school) examinations are inflated by a factor of 2 in goal by whatever means available, including cheating. But
mathematics (so that in a school where the student can then it is obvious that widespread cheating on examination
correctly answer only 40 percent of questions, the official scores must undermine personal satisfaction and social
data report 80 percent success) and that this is widespread validation of teachers and erode professional norms.
Box 3: Examples of strategies for building clear, consensus-based prioritisation

of learning
While it is clear that fixing service functions alone (e.g., human resources or IT) cannot restore the core of an
ineffective organisation, it is much less clear how to rebuild this shared sense of purpose. This lack of clarity is partly
because shared purpose is a relational, subjective matter which is highly contingent on local context. Nonetheless,
emerging RISE research suggests some strategies to consider for building clear, consensus-based prioritisation of
learning:
• At the system level, conduct and/or fund learning assessments, which can sometimes catalyse political attention
or citizen-led pressure to prioritise student learning (Kaffenberger, 2021).
• At the local government level, organise deliberative forums for a wide range of actors in the education system to
discuss and agree on priorities for education (Castradori, 2021, on Nigeria, and Gaduh et al., 2020, on Indonesia).
• At the school level, embed mission-driven teachers within ‘donut’ schools in order to build relationships with
teachers and headteachers and act as role models with the goal of shifting school-level norms toward learning
(Bano, forthcoming).
Two caveats: as noted above, any such strategies need to be thoroughly rooted in the local context (and their success
or failure may be highly context-specific); and this list is far from exhaustive.
Other resources on restoring organisational purpose and building consensus-based prioritisation of learning include:
• The article on stakeholder consensus in Edsyclopedia: The Education Systems Wiki: https://edsyclopedia.org/
Stakeholder_consensus
• Honig, D. (forthcoming). Managing for Motivation as Performance Improvement Strategy: Closing the
Implementation Gap in Education & Far Beyond. RISE Working Paper Series.
• Jeevan, S. (2021). Intrinsic: A manifesto to reignite our inner drive. Endeavour.
Although clear, consensus-based prioritisation of learning is an essential aspect of teacher career reform (as we
outline below in premise for practice #1), it is not a prerequisite for teacher career reform. Even if an education
system currently lacks a widespread consensus around a shared learning-related purpose, purpose-driven actors
within the education system can effect specific, targeted changes to incrementally reorient aspects of teacher
career structures toward the purpose of cultivating student learning. Such incremental, purposive changes can
help stakeholders within the system to recognise the value of prioritising learning and to see the possibility of
improvement—thus building a shared sense of hope.
Throughout this primer, we suggest possible action areas for education authorities and organisations in the premises
for practice. We will discuss starting points for other education stakeholders in subsequent publications.
In short, initiatives that specifically aim to build stakeholder consensus do not necessarily need to precede initiatives
to reorient teacher career structures. Rather, they can be mutually reinforcing steps along the journey toward
consensus about, and coherence for, student learning within an education system.
Unfortunately, piecemeal improvements to service coherence in a number of ways.

functions often receive more attention than the
deeper challenge of restoring the shared purpose and 2.2 Typical teacher career reforms are often
technical practices in the organisational core because incoherent with the complexity of teacher
such improvements to service functions, especially in careers
the case of standalone improvements focused on thin
Before looking at how to develop a strongly performance-
indicators, are much easier to address than the complex,
normed, locally embedded, professional teaching career
entrenched challenges at the core. This is true of research
structure, we first look at what not to do. As observed in
on public sector improvement in developing countries,
the Teacher Task Force’s Teacher Policy Development Guide:
as in mainstream economics analyses that mention the
importance of organisational mission in the public sector, As things now stand, most education sector plans
yet focus in far more detail on the design of contracts, create more fragmentation and less coherence; they
monitoring, and incentives rather than on the prior address teacher policy dimensions only partially
question of whether the public sector organisations in and do not systematically include all the major
question have a shared sense of mission to begin with determinants of learning success linked to teachers,
(e.g., Finan, Olken, & Pande, 2017; Besley, Burgess, Khan, & i.e., recruitment; initial education; balanced
Xu, 2021). It is also true of many education reform efforts, deployment; continual professional development;
as with the Madhya Pradesh school quality assurance decent salaries; career prospects; and working
programme that improved neither teacher practice nor (teaching/learning) conditions (UNESCO, 2019, p. 19).
student learning, yet is regarded as an administrative In this section, we discuss three common types of teacher
success and has been scaled up to over 600,000 schools career and education policy reforms that are currently
(Muralidharan & Singh, 2020). dominant in education reform in some low- and middle-
All that said, although service functions like HR or IT income countries—but are fundamentally incoherent
or procurement cannot fill the core of an ineffective with the nature of teaching and/or the nature of effective
organisation, service functions can be vital to sustaining education systems.
the success of an effective organisation when these
Piecemeal changes that are incoherent with
functions are aligned to comprehensively support the
purpose and technical practices at the core. Conversely, if existing education systems and organisations
the service functions are not coherent with the purpose at One currently dominant mode of educational change
the core, then the core has to compete with other priorities. is reform driven by the introduction of new projects/
For example, EMIS data entry can compete with lesson programmes or the tweaking of existing projects/
planning and teacher professional development within programmes with the goal of incrementally raising
the limited number of hours in a teacher’s day. This is education system performance from a low baseline.32 The
illustrated in panel (d) of Figure 7. The key point here is that problem here is that interventions or ‘best practices’ that
coherence with the core purpose is pivotal to effectiveness were effective in one context or within tightly monitored
in complex organisations, as noted at the start of Part 2. and well-resourced experiments often fail to trigger the
More specifically, purpose is central to teacher careers. same causal mechanisms of improvement when they
Besides sitting at the core of effectiveness at the are tacked on to the business-as-usual complexities of
organisational level, purpose is also fundamental to education systems, teacher career structures, and teachers’
delegation at the level of teacher career structures. At motivations (see also Pritchett, 2017). Just as a rubber band
the level of teacher motivation, purpose shapes personal hanging on a nail can be stretched out by the application
satisfaction and influences social validation. In short, of force, but springs back as soon as that force is no longer
alignment with an overarching purpose is a hallmark of applied—so teacher practice and the operations of an
coherence in teacher careers and in education systems education system can be temporarily altered by a resource-
more generally. Unfortunately, in many low- and middle- and monitoring-intensive intervention, without producing
income countries education systems depart from such any enduring change to the purpose of cultivating student
32 The ‘piecemeal’ approach is popular because it is neatly compatible with the bureaucratic and technocratic approaches of many
donors who seek discrete, identifiable, funding opportunities. A minor twist on the piecemeal approach that has extended its popularity
is to base the project/programme design on the findings from one or more randomised control trials, with the implied (but false) that
the project/programme is therefore based on ‘rigorous’ evidence.
learning. If the goal is permanent change, then something the new teacher was hired as a civil servant the learning
needs to be done to fundamentally change the system at impact was much smaller (not statistically different from
hand: an additional nail to hold the stretched rubber band zero). This recommendation for ‘contract teachers’ was
in place; a transformation of the purpose for which the taken up by the government and there was an effort to
design elements of a teacher career structure are coherent. extend the programme nationwide. The scaling up was
In other words, a sustainable reorientation of an education itself tracked with a randomised control trial. In the scale-
system can only be achieved by coherent systemic change, up, some of the programme implementation was done by
not incoherent piecemeal change. an NGO and some by the government. The study found that
the class size reductions by contract teachers in the NGO
To illustrate incoherent piecemeal change, consider
implemented areas had almost exactly the learning gains
the school quality assurance RCT in Madhya Pradesh,
from the original study, but in the areas where the same
discussed in the previous section, as analysed by
programme was implemented by the government the class
Muralidharan & Singh (2020). This intervention was
size reductions through additional contract teachers had
based on a set of ‘best practices’ in school management
zero impact on learning (Bold et al., 2018).
(i.e., school performance assessment based on tailored
scorecards, customised school improvement plans based It should come as no surprise that the impact of doing the
on the assessments, and quarterly visits from subdistrict- ‘same’ programme/project/policy is different depending on
level administrators to follow up on the plans) that were the organisation that implements it. In one organisation, a
adapted to, and piloted in, the local context, before programme could be a piecemeal, superficial intervention,
being implemented across thousands of schools. Despite whereas in another organisational context the ‘same’
consistent implementation, the intervention had no programme could be an element in coherent systemic
effect on student achievement, nor on the determinants change that is tailored to the organisation in question.
of student achievement that were measured in the study,
In fact, any single element of coherent systemic change
including student absence, teacher absence, teacher
may, in isolation, look counterproductive, yet it may be
practices, or the quality of monitoring and support from
an indispensable element in a system that is coherent for
bureaucrats and school committees. The authors conclude
a highly productive purpose. Consider the compensation
that the intervention failed because of a mismatch—i.e., an
system and organisational structure of the Lincoln Electric
incoherence—between the programme’s theory of change
Company, as described by Roberts (2004). Lincoln Electric’s
and how the programme was perceived by teachers,
decades of consistent profitability made it the subject of,
headteachers, and bureaucrats.33 Rather than regarding
among other things, multiple Harvard Business School
it as a form of support for improvement in their practice,
case studies. 34 One component of the organisational
teachers viewed the programme as a source of paperwork
system at Lincoln Electric is piece-rate compensation,
that was disconnected from their classroom practice. They
coupled with substantial individual bonuses based on
also saw it as yet another fleeting policy change that would
supervisor ratings, which aim to offset the perverse piece-
soon be replaced—so it was piecemeal both in relation to
rate incentive to focus on quantity at the expense of quality
teacher practice and to overall education policy (ibid; see
and of complementary tasks such as team cooperation.
also Romero et al., 2021).
While performance-based pay (whether piece rates
Another example, which speaks specifically to the or bonuses) is a darling of some economics textbooks,
question of whether projects demonstrated to be other organisational choices at Lincoln Electric are
effective ‘piecemeal’ can be extended systemically, is an hardly textbook prescriptions. For example, to make the
experience with contract teachers in Kenya. In a modest- piece-rate incentives meaningful, the company tolerates
scale randomised control trial implemented by an NGO, considerable amounts of work-in-progress inventory—
the researchers (Duflo, Dupas, & Kremer, 2011) found that which runs counter to the ‘lean production’ perspective,
reducing class size in early grades from their very high but in this context allows the piece-rate incentives to be
initial levels produced large learning gains when the new meaningful because individual workers are not constrained
teacher was hired on a renewable contract, but when by the productivity of those preceding or following them
33 Put differently, the formal delegation of the school quality assurance programme was incoherent with organisational norms wherein
teachers were used to completing reports for the sake of documentation rather than to facilitate improvement, and bureaucrats are
accustomed to being conduits for top-down directives and bottom-up reporting rather than being enablers of change.
34 See https://hbr.org/search?N=4294923992&Ntt=lincoln+electric, accessed on 22 March 2021.

PREMISE FOR PRACTICE #1: on clear, consensus-based prioritisation of learning

→ What to do differently
Rather than assuming that the education system is independently committed to learning …
Clear, consensus-based prioritisation of learning, delegated from education authorities and organisations to schools
and teachers, is fundamental to teacher career reform.
→ Why this matters
Education systems involve a wide range of people who have many different goals that they want the system to
fulfil. Many of these goals are valuable and legitimate, since child development—and societal development—are
multifaceted. Yet student learning, especially the mastery of foundational literacy and numeracy, should be a central
priority in any education system, and it should not have to compete with other goals.
However, these multiple and often competing goals are one of the key reasons why attempts to reform education
policy are often piecemeal, rather than oriented toward deep, systemic, purpose-driven change. Other reasons
for such piecemeal reforms include short time horizons, misapplications of research findings across incompatible
contexts, and a failure to consider other elements of the system.
Instead, children in education systems deserve system-oriented, purpose-driven reforms to teacher career structures
and other aspects of the education system. A key way to raise the likelihood of such reforms is by aligning different
people and processes within the system around a shared goal of cultivating student learning. Such alignment cannot
be taken for granted in large, complex systems with many moving parts and competing goals. Yet initiatives to build
stakeholder consensus around the priority of student learning, together with policy reforms to increasingly align
teacher careers as well as other educational structures and processes with this priority, can jointly and iteratively
shift systems toward educational improvement.
→ See also
• ‘Examples of strategies for building clear, consensus-based prioritisation of learning’ in Box 3
• ‘Effective organisations have a shared purpose and technical expertise at their core’ in Section 2.1
• ‘Typical teacher career reforms are often incoherent with the complexity of teacher careers’ in Section 2.2
• ‘Why should teachers be not only “capable” but also “committed”? Because motivation and purpose matter’ in
Section 3.2
• Premise for practice #5 on purpose-driven definitions of fairness in teacher compensation
in the production process. Even more unusually, Lincoln the context of Lincoln Electric’s well-established coherent
Electric maintains a policy of avoiding employee layoffs, system.
even amid recessions. While avoiding layoffs may seem—
Note that incremental change need not necessarily be
and, in many contexts, may be—foolishly inefficient, it is
incoherent and piecemeal. Some incremental, self-
a coherent and necessary element of Lincoln Electric’s
contained programmes can be highly effective at boosting
norms of organisational trust that, among other things,
the learning levels in a selection of content areas for the
buttress employees’ confidence that supervisors will not
selection of students that they target; as with Teaching
use piece-rate driven improvements in productivity to
at the Right Level (TaRL), an instructional model that has
subsequently lower the compensation rate that employees
expanded and adapted from its origins in India to ten
receive. Introducing layoffs and streamlining inventory
countries in Sub-Saharan Africa (Banerji, 2015; Banerjee et
may be sensible policies in some settings, but they would
al., 2017).35 But TaRL and other comparably successful self-
undoubtedly constitute incoherent piecemeal change in
contained interventions are usually (a) tightly coherent with
35 See also https://www.teachingattherightlevel.org/tarl-in-action/, accessed on 22 March 2021.

the goal of student learning; (b) highly coherent internally incoherence of thin indicators with thick, purpose-
across programme elements and externally with local driven, coherent systemic change applies not only to thin
contexts (as with TaRL’s well-aligned curricula, materials, indicators of inputs, but also of outputs.
and teacher/instructor training); and (c) designed to
Changes based on thin inputs typically fail to yield
address specific systemic shortcomings (as with TaRL’s
purpose-driven reorientation, instead yielding isomorphic
catch-up programmes for children in the middle primary
mimicry, i.e., changes that make the organisation ‘look
grades who haven’t mastered the foundational literacy and
right’ (Pritchett, 2013, 2014; drawing on DiMaggio & Powell,
numeracy that, in a well-functioning system, they would
1983). One example of ineffective and largely isomorphic
have attained in earlier grades). Such locally coherent
change based on thin inputs is Indonesia’s 2013 initiative
incremental change can be part of broader systemic
for teachers’ continuous professional development, which
pushes toward coherence for learning. For example,
was intended to overcome the shortcomings of prior in-
Pratham, the Indian NGO that developed TaRL, uses and
service teacher training programmes by targeting training
continually refines a diverse set of initiatives, such that
modules to individual teachers based on their assessed
the catch-up instruction of TaRL complements community
knowledge and by linking successful completion of the
mobilisation and national advocacy through the ASER
modules to credit points for civil service promotion (Revina
assessment tool, as well as a range of initiatives to engage
et al., 2020). Consequently, this reform was systemic in
and mobilise individuals and communities to prioritise
attempting to link support, in the form of in-service training,
student learning and to be equipped to support it (Bano &
with delegation and finance, in the form of promotions and
Oberoi, forthcoming).
the associated salary bump, within teacher career system.
Still, even a coherent and demonstrably effective Yet module completion turned out to be a thin indicator of
intervention like TaRL can falter when incorporated into a programme that isomorphically looked right but did not
an education system with entrenched norms that are improve teacher practice, not least because the training
coherent for bureaucratic compliance and exam pass rates materials for many modules were theoretical rather than
rather than student learning. Banerji (2015) documents applicable to classroom practice. The study authors also
an instance in which TaRL, in spite of demonstrated and found indications that participants valued the module
acknowledged success in improving learning outcomes completion certificates—because of the concrete benefits
in a number of districts of a state, was abandoned as an that resulting from these certificates—more than the new
approach in favour of a return to more top-down and knowledge gained from the modules.
‘process-compliant’ approaches. A study of the Delhi school
For an example focused on thin outputs, Michael
system’s TaRL-inspired Chunauti programme (Aiyar et al.,
Barber’s ‘deliverology’ approach to educational change
2021) similarly shows the difficulties of implementation
is patently a systemic approach, with its mapping of
that require a change in organisational ‘delegation’ and
diverse stakeholders and functions across education
norms from a ‘top-down process compliance with thin
systems (Barber et al., 2010; Barber et al., 2016). It also
inputs’ mode to an engaged and empowered learning
clearly focuses on outputs, which is an improvement
approach.
over assuming that inputs will necessarily lead to student
Changes that are incoherent with the learning. Yet the insistence that every education system
aspiration be summarised in quantifiable metrics that can
complexity of classroom teaching
be monitored with red-amber-green dashboard indicators
Another dominant form of education reform is systemic is an unabashedly ‘thin’ approach. This thinness is also
(rather than piecemeal) but nonetheless incoherent with reflected in the fact that the ‘delivery pyramid’ for ‘moving
comprehensive, learning-oriented reorientation of teacher from aspiration to implementation’, with its linear flow
careers and education systems. In this case, the incoherence from aspirations to metrics and down through performance
emerges from the fact that teaching, and education more patterns, drivers, system activities, interventions, to plans
generally, are complex and multidimensional, or ‘thick’36— (Barber et al., 2010, p. 47), implies a conceptualisation
such that reform efforts that focus on ‘thin’ indicators of of education as complicated, with multiple, nested
quality are likely to yield more negative side effects than elements—but not necessarily as complex, with numerous
positive learning-oriented changes (Honig & Pritchett, multidirectional interactions that can have unpredictable
2019; see also ILO, 2012; Murnane & Cohen, 1986). The results. Moreover, deliverology proposes that ‘a strong
36 For an explanation of our use of ‘thick’ and ‘thin’, see footnote 1.

aspiration … must consist of goals we can assign to specific more generally—derives from a range of sources that
people’, such that each metric has a ‘person who spends go far beyond finances to also incorporate material
most of his/her time on the priority and has sleepless circumstances, personal satisfaction, and social validation,
nights worrying about hitting the targets’ (Barber et al., as discussed in Section 2.1. In settings where most teachers
2016, p. 14). This is firmly rooted in what Honig and Pritchett are demotivated, the root causes of this demotivation tend
(2019) call accounting-based accountability, which is based to be more complex than salary levels.
on top-down managerial control of thin indicators, in sharp
For example, in describing the challenges faced by novice
contrast to account-based accountability, which is founded
teachers in Ethiopia, Hagos (2018) noted that inadequate
in thick, relationship-embedded, narrative justifications
pay was an issue, but also that:
for performance. Thick account-based accountability is
far more coherent with the complexity of teaching and The community has a negative attitude towards
learning than thin deliverology-style accounting.37 the profession, as a result of which teachers don’t
want to be called teachers. And that is worrisome,
Strikingly, a recent study suggests that performance
because these are teachers who are joining the
management based on ‘thick’ indicators can be just as
profession as a last resort. … The novice teachers
good at incentivising improvements in a ‘thin’ indicator as
were telling us that, ‘When we go to the school, they
performance management based on the ‘thin’ indicator
discourage us. “Why do you come here? Why do you
itself. In a randomised control trial in Pakistani private
want to be a teacher? You don’t have to be a teacher.”
schools, Andrabi and C. Brown (forthcoming) find that
teacher performance pay based on school principals’ Bennell and Akyeampong (2007) suggest that this norm of
subjective evaluations of each teacher (a thick indicator teaching being a last resort goes far beyond Ethiopia. To
of teacher performance) was equally effective at raising use the language of the hedonic and dynamic model of
test scores as teacher performance pay based on teacher careers shown above in Table 4, the issues here
students’ test score gains (a thinner indicator of teacher are much broader than finances. They are entangled with
performance). Moreover, under the thinner, test-based entrenched norms and perceptions about the teaching
performance pay treatment, there were negative effects profession not only from the local community but also
on student socioemotional development, such as their from the experienced teachers and school leaders who
love of learning, and on how much students said they are inducting and socialising these novice teachers into
liked attending school—perhaps because their teachers the profession.
spent more time on test preparation and were observed In such a setting, raising teacher pay may bump up
to conduct less student-centred lessons (ibid). This offers a teacher motivation from finances, but it will do little to
clear cautionary tale—that thin indicators may jeopardise address teachers’ depleted personal satisfaction and
unmeasured but valuable aspects of education—alongside social validation—and may be insufficient to meaningfully
a reason for optimism—that thick, contextually embedded support the cultivation of student learning. A large
performance management systems may help to boost experimental evaluation of a (de facto, even if not de jure)
children’s academic learning alongside other valued unconditional salary increase that permanently doubled
outcomes. teacher pay in Indonesia found that teachers who received
the salary increase during the evaluation period ‘were
Changes that are incoherent with the
significantly more likely to be satisfied with their income,
complexity of teacher motivation significantly less likely to report financial stress, and
Discussions about improving teacher quality and significantly less likely to hold a second job than teachers
strengthening the teaching profession often focus in control schools’ (de Ree et al., 2017, p. 1034)—i.e., they
disproportionately on teacher pay: raising it, changing its derived greater motivation from finances—yet there was
distribution, making it contingent on performance, and ‘no effect on teacher effort toward upgrading their own
so on. Yet teacher motivation—and human motivation skills, no consistent evidence of changes in self-reported
37 For one description of the hierarchical dysfunctions and perverse incentives of thin-output deliverology in education policy and
practice, see Naviwala (2016), on deliverology in Punjab, Pakistan (see also Gewirtz et al., 2019, on deliverology in England). For an
interesting contrast, see Barber’s (2013) The good news from Pakistan: How a revolutionary new approach to education reform in Punjab
shows the way forward for Pakistan and development aid everywhere. It is also worth noting that after the change in political leadership
in Punjab, there has been massive pushback against the deliverology approach in education there, as it relied on a very narrow basis of
political authorisation and acceptability.
teacher attendance, and no effect on the ultimate outcome same teaching quality into the occupation:
of student learning’ (ibid, p. 1034)—i.e., this boost in
motivation was insufficient to change classroom teaching CW(more paperwork)
and learning. Shifting from the level of teacher motivation =e(U,pc,PS(more paperwork),SV)
to the design of teacher careers, this change in finances did -e(U,pc,PS,SV)>0
not alter professional norms.38
Alternatively, suppose over time teaching has come to seen
To be clear, we are not arguing that finances don’t matter. as just warehousing kids and not really being committed
Paying teachers adequately and appropriately is not only to helping every child learn, and that therefore the social
an ethical imperative, but also an indispensable part of validation from being a teaching is lower.
a sensible strategy for cultivating the future wellbeing of
children individually and countries collectively. The point CW(less shared purpose)
here is that finances are not the only thing that matters— =e(U,pc,PS,SV(less shared purpose))
and, in many contexts, they are certainly not the most -e(U,pc,PS,SV)>0
important factor in teacher motivation and practice. 39
One way of understanding this situation is ‘We need to
For example, one analysis of a national survey of U.S.
pay teachers more in order to attract adequate quality
schools suggests that private schools in this context tend
teachers.’ But we hope that the reader sees that this is
to be more effective than public schools at recruiting and
completely, totally, the wrong conclusion to draw, for two
retaining good teachers despite paying salaries that are,
deep and important reasons.
average, 35 percent below teacher salaries in the public
sector (Ballou & Podgursky, 1998). The authors attribute First, if in fact ‘more paperwork’ action does not improve
the private school advantage in teacher recruitment and education outcomes, then it is inefficient and ineffective to
retention to personnel policies such as flexible pay (rather compensate teachers because their personal satisfaction
than fixed salary schedules) and greater support for novice is lower. The question that needs to be asked is: ‘Are the
teachers in student discipline, classroom instruction, and benefits of forcing teachers to take actions X large enough
acclimation to the school environment (ibid). to make it worthwhile to pay a wage premium to keep the
actual teacher overall well-being constant despite the
An advantage of taking a fully hedonic approach is that one
teacher well-being losses from taking actions X?’ Posing
can pose clearly important questions about how to attract,
this question makes it clear that many things that are
retain, and motivate good teaching.
allowed to happen in education systems are ‘lose-lose’,
For instance, suppose a government has taken actions in that they don’t make outcomes better and they reduce
that make teaching a less personally satisfying occupation teacher satisfaction—and hence require either (a) higher
for excellent teachers (say, demanding more paperwork money wages or (b) allow the attractiveness of teaching as
that teachers perceive as unnecessary). This action has an occupation to decline. We are hoping the reader by now
an associated compensating wage differential from the can see that the hedonic approach to total compensation
expenditure function,40 which is how much higher money is a way of pointing out that if an education system is
wages would need to be in order to attract a person of the
38 Overemphasising finances may not only fail to improve teacher norms, but in some cases may actively harm these norms. In the
context of the U.S., Jal Mehta (2013) argues that focusing disproportionately on pay and working conditions has inadvertently reinforced
the disempowerment of the teaching profession: ‘Beginning in the 1960s, teachers, responding to their low pay, status, and respect,
organized themselves to bargain collectively. This tack brought about increases in pay and considerable political clout for the teacher
unions, whose relentless focus on bread-and-butter issues yielded significant gains for their members. But at the same time, by moving
towards an industrial bargaining model (a union), as opposed to a professional association, teachers defined themselves as labor in a
labor-management dispute, as opposed to professionals who controlled their practice. The emphasis on wages and job protection has
weakened their public claim to speak on educational issues and undermined their power in the eyes of legislators and the public. Groups
that are far less traditionally powerful than they, in terms of members and money, have been able to trump them in political debate
because of their greater moral power’ (pp. 33–34).
39 For an empirical exploration of whether teachers are relatively overpaid or underpaid in Sub-Saharan Africa, see Evans, Yuan, &
Filmer (2020).
40 For more on the compensating wage differential, see Section 2.1 above on complementary sources of teacher motivation.
doing things that are ‘lose-lose’ in that they both reduced to maintain a status quo system and organisation that are
learning and de-motivate/reduce satisfaction from being not committed to good teaching (and hence reduce the
a teacher the response to this of ‘raise wages’ to offset the personal satisfaction and social validation benefits of
‘lose-lose’ actions is completely wrong-headed. being a teacher for good teachers) is not the path to good
education outcomes.
Second, the composition of the teaching force is a function
of how overall motivation is achieved. That is, suppose (in It is also worth noting that many finance-focused changes
an illustrative, ideal-typical sense) that teachers are of just that are incoherent with the complexity of teacher
two types: ‘committed to teaching’ and ‘process compliant.’ motivation are, concurrently, piecemeal changes that
The second type of teacher just does what they are told are incoherent with the complexity of education systems
and does not derive personal satisfaction from achieving and teacher career structures. Conversely, successful
better education outcomes for the children they teach. reforms related to finance and teacher careers typically
Then a ‘more paperwork’ choice reduces the well-being entail a wide range of changes that go far beyond
of the ‘committed to teaching’ type, but not the ‘process the purely financial. For example, successful teacher
compliant’ type, who is indifferent between teaching and performance pay programmes in low- and middle-income
filling in forms. Therefore, if money wages do not improve, countries tend to incorporate a combination of mutually
then the deciding to institute more paperwork reduces reinforcing and mutually coherent elements, similar to
over time the fraction of teachers who are ‘committed to the career and compensation system at Lincoln Electric,
teaching’ and increases over time the fraction of those described above. In a review of teacher performance
who are ‘process compliant’. This then has second-round pay programmes in developing countries, some of which
feedback effects on everything. For instance, it is going were implemented as RCTs, Breeding, Béteille, and Evans
to be harder to maintain organisational, workplace, and (2021) found that only a few of the programmes had a
professional norms of commitment to teaching if the statistically and substantively significant effect on student
compensation structure differentially rewards process learning. These successful programmes tended to have
compliance over personal satisfaction. Also, in the next incentives that suited the context, plans for long-term
round of policy choices there will be less resistance to sustainability, multiple measures of performance (rather
actions that reduce personal satisfaction from teaching of than depending solely on test scores), and buy-in from
committed teachers since there will be fewer of them. teachers. In the language of the hedonic and dynamic
model, effective teacher performance pay interventions
So, perhaps paradoxically, when one is confronted with
were by nature systemic, affecting not only finances, but
a situation in which it is difficult to attract into teaching
also delegation and information (in defining multiple
people who are capable of and committed to good teaching
measures of performance) and norms (in ensuring that
practices, it can be precisely wrong for the first step to
programme design was sufficiently compatible with
be increased wages. The first step should be to diagnose
existing norms to secure teacher buy-in). Additionally, one
how much of the reduced attractiveness is due to reduced
of the programmes included in the review found modestly
personal satisfaction and reduced social validation from
positive but most insignificant effects on student learning
being a teacher because the system and the organisation
from a teacher bonus scheme on its own, but substantial
(ministry) have become a ‘donut’ for learning and are not,
and significant learning gains in a treatment arm that
and is not perceived to be, committed to teachers who
combined the bonus scheme with a school-level grant
engage in good teaching. Paying higher wages because
aimed at alleviating resources constraints (Mbiti et al.,
people are reluctant to work for your organisation because
2019)—suggesting that support (in ensuring that teachers
they are not convinced the education organisation is
have adequate resources for good practice) is similarly
truly committed to its education goals (or are convinced
vital to effective reforms, even if the reform nominally
the organisation’s goals have been supplanted by other
focuses on teacher pay.
purposes) is madness.
Good teaching is worth paying for, but paying more in order
Part 3: Why the 5Cs matter (and how they fit together)
Being a mortal was strange.…Time was inexorable and unstoppable.

I had always known that, but it had taken me fifteen years as a
mortal to understand what it meant.
–the Greek god Apollo, in Jo Walton’s The Just City (2014)
Part 3: Why the 5Cs matter (and how they fit together) 44
3.1 The 5Cs contrast markedly with the conventional civil 46
service approach
‘Curation’ is common in comparable professions—but 49
unusual in teaching in developing countries
3.2 Why should teachers be not only ‘capable’ but also 52
‘committed’? Because motivation and purpose matter
3.3 Why ‘choose and curate’? Because the early career 53
phases are pivotal for systemwide norms of capability
and commitment
PREMISE FOR PRACTICE #2: on cultivating norms during the novice 55
teacher phase
3.4 Why ‘curate’, when there’s already ‘choose’? Because 56
classroom teaching is complex
PREMISE FOR PRACTICE #3: on the importance of curation 58
3.5 Why ‘commitment’? Because continuous curation is 59
costly
3.6 The 5Cs encapsulate the hedonics and dynamics 61
of teacher careers—toward the goal of capable and
committed teachers
Having outlined some of the levels of complexity and the 5Cs principles aim to establish a strongly performance-
common incoherences affecting teacher career and normed, contextually embedded, professional, teaching
compensation structures, and argued that a systemwide career that cultivates student learning.
sense of purpose and consensus-based prioritisation of
To begin with the end in mind, the goal of the 5Cs is to
learning is crucial for teacher career reform in complex
cultivate a corps of capable and committed teachers.
education systems (premise for practice #1), we now
As discussed in Section 2.1, the core of any effective
propose a set of principles that takes into account both
organisation is a pairing of a shared purpose and a set of
complexity and coherence. We call this the ‘choose and
technical practices that advance the purpose. In the case
curate toward commitment to capable and committed
of an effective education bureaucracy, teachers who are
teachers’ approach, or the 5Cs. As noted at the beginning,
capable and committed embody that combination of theme or narrative of a specific exhibition.
technical capability and purpose. Similarly, Mintzberg
Curation—from the Latin root ‘cura’, to care—also entails
(1983) emphasises the need for organisations to equip
caring for the artworks or artefacts under the stewardship
jobholders both via training, i.e., ‘the process by which
of the organisation.41 Correspondingly, in the context of the
job-related skills and knowledge are taught’ (p. 39), and
teaching profession and the 5Cs, curation also involves not
indoctrination (or socialisation), i.e., ‘the process by which
only identifying those novice teachers who can become
organisational norms are acquired’ (p. 39). In the language
capable and committed teaching professionals, but
of the hedonic and dynamic model of teacher careers,
also providing all novice teachers with the best possible
capability is the intersection of the design elements of
support for developing their capability and commitment.
delegation (who does what where—and why?) and support
Put differently, curation involves an immersion in everyday
(what would help them to do it better?). Commitment is the
classroom realities, but it should not mean throwing
intersection of delegation and norms (how should ‘good
unprepared novice teachers in the deep end. Ideally, both
teachers’ act?), as well as teachers’ motivational senses of
the early-career teachers and those training and manging
personal satisfaction and social validation.
them should give their best shot during the pre-service and
Building a corps of teachers who are committed to, novice phases.
and capable of, cultivating student learning is not a
Also, the bulk of the turnover from curation should be
straightforward project, especially at the scale of a
voluntary, i.e., novice teachers deciding for themselves,
national school system. Within the normal constraints of
after having experienced the classroom, that they would
government and the institutions of governance, it is far
be better suited to other employment and professions.42
easier for education authorities to build a corps of teachers
who manage of the logistics of the expansion of available Following the processes of choose-and-curate during the
classrooms and engage in process compliance—which is pre-service and novice phases of the teacher career cycle,
why so many more countries have been able to expand the education bureaucracy then makes a commitment
access than have managed to create and maintain learning to the capable and committed teachers in the profession.
outcomes. This commitment can take the form of civil service tenure
or other types of long-term or permanent employment
Developing a capable and committed teaching profession
contracting. However, preceding the commitment with a
requires the application of different principles at different
period of choosing and curating is starkly different from
phases of the teacher career cycle. During the pre-service
the typical civil service approach.
and novice phases, this first requires using the best
available information—such as interviews, mock lessons, In Section 3.1 , we explore this difference by comparing
and tests of teacher content knowledge—to choose an example of the 5Cs with the typical approach. Next,
prospective teachers for pre-service training and an initial in Section 3.2, 3.3, 3.4, and 3.5, we delve into why each
period of classroom teaching. This choosing, or screening, of the ‘C’s is indispensable to the 5Cs as a unified set of
takes place at one (or a few) defined points in time during principles. These sections also include premise for practice
the pre-service and novice phases. #2 and premise for practice #3 about what to do differently
under the 5Cs. Finally, Section 3.6 wraps up this part of
Alongside the choosing, the pre-service and novice phases
the discussion by making the case that the 5Cs collectively
under the 5Cs also involve ongoing curation to discern the
constitute a fully hedonic, fully dynamic approach to
best possible fit between the pool of early-career teachers
teacher careers.
and the specific needs of the children and classrooms in
the education system in question. The term ‘curate’ here It is worth reiterating, from Section 1.3, that the 5Cs are
draws on the arts and heritage sectors, where the process framed as guiding principles for education authorities
of curating an art exhibition entails selecting from a pool because they have authority over systemwide teacher
of valued artworks or artefacts (that have already been career structures, and not because teacher agency is
chosen to be part of the larger collection of the museum unimportant. Far from it. The ‘match quality’ between any
or gallery in question) to identify those that best suit the given teacher and the education system in which they are
41 Also, in some Catholic and Protestant church traditions, a curate is an entry-level member of the clergy, who has been trained and
is expected to care for members of the congregation but does not yet hold the full responsibilities of a parish priest.
42 We recognise that this ideal may be far from reality in contexts where labour market conditions mean that there is a lack of desirable
alternatives to teaching, or where the quality of pre-service training is insufficient to be an asset in seeking employment in other fields.
embedded is a two-sided prospect. In other words, it is not student learning;

only education authorities and school leaders who choose
b. the organisation commits to permanent
teachers at the point of entry into the teaching profession,
employment only after teachers have clearly
but also the teachers themselves who make decisions
demonstrated this capability and commitment in
about whether to join the profession (as discussed below the classroom;
in Section 3.3). Also, while education authorities should
design teacher career structures to facilitate a period c. if compensation is variable, it varies based on these
of curation to further refine the identification of those demonstrated contributions to student learning
teachers who demonstrate capability and commitment and the goals and purposes of the education
in classroom practice, the bulk of this turnover should system; and compensation is calibrated to attract,
retain, and motivate the best possible contributions
be voluntary, i.e., based on individual teachers’ choices,
to student learning, across the full pool of teachers
as noted above. Moreover, as discussed below in Section
at various stages of their careers, and early
3.2, the 5Cs aspire toward a profession of teachers who
compensation is designed to encourage voluntary
are not only capable but also committed, i.e., who are separation of those not capable or committed to
autonomously motivated to cultivate student learning teaching (as illustrated in Section 3.3).
and who exercise their agency to best do so within their
respective classroom contexts. As a set of principles, the 5Cs can and should look different
in different education systems, given that the teaching
3.1 The 5Cs contrast markedly with the career-relevant features of each context vary in complex
conventional civil service approach ways. That said, in this section we outline one possible
instantiation of the 5Cs, which helps to illustrate the
Far and away the dominant model of teacher careers is a contrast between this approach and the conventional
‘commitment early and only’ approach that borrows a civil civil service approach. This instantiation draws heavily
service structure in which: on Pritchett & Pande (2006; see also Pritchett & Murgai,
a. teachers are hired based on formal qualifications 2006), which was shown above in panel (b) of Figure 1
and pre-service assessments (which may have little and is reproduced in Figure 8 on the following page. In
or no correlation with actual capability to perform); this application of the 5Cs principles, the education
bureaucracy initially employs novice teachers on fixed-
b. the organisation makes a lifetime employment
term contracts with a flat and relatively low—but above-
commitment, perhaps after an entirely nominal
probationary period, to individuals independently subsistence—pay scale, with a step-jump to a substantially
of the person’s capability to perform or whether higher pay scale and a permanent contract for those who
they are committed to the goals of the system; demonstrate, at the end of the novice phase, that they
have built up a sufficient, pre-determined baseline of
c. nearly all aspects of compensation (including instructional competencies. Thus, in this example, the
the important hedonic elements of assignment
organisation chooses at two points in time: at the point of
to districts, schools, subjects and classes) are
entry into pre-service training, and at the end-of-probation
determined by a points system that almost
transition from novice teacher status to experienced
exclusively focuses on seniority, with some
points for formal qualifications and certifications teacher status. However, a much larger proportion of
(which again may have little or no relationship to the streamlining of the cohort size occurs voluntarily, in
performance); and an ongoing period of curation, as pre-service and novice
teachers decide to exit the profession on their own as they
d. compensation is both excessively ‘frontloaded’ gain an increasingly thorough immersion in the profession.
and excessively ‘backloaded’, in that the seniority Those who successfully complete this novice phase will
structure of wages leads to wages that are ‘too high’
move on to become experienced and veteran teachers
early in the career and ‘too low’ late in the career—
on permanent contracts, with mutual commitment from
but pensions and benefits backload compensation
teachers who have demonstrated their motivation and
to encourage retention at high levels of seniority.
competence to children’s learning and from education
In contrast, the 5Cs aspire toward a teacher career system organisations that commit to continually hiring, paying,
in which: and developing these teachers.
a. teachers are initially chosen based on their potential Given that there will necessarily be heterogeneity in
for capability and commitment to cultivating teacher motivation and competence—whether across
Figure 8. A possible instantiation of the 5Cs principles
Source: Adapted from Pritchett & Murgai (2006), Figure 9.
the pool of teachers or, for individual teachers, across the context-specific trade-offs.) Additionally, rather than
range of their varied responsibilities and over the span of requiring a uniform retirement age for pension eligibility,
their full career—this particular 5Cs example also includes this approach also incorporates an attractive early
some performance-based rewards for experienced and retirement option for veteran teachers who, for any reason,
veteran teachers, such as annual salary increments based no longer have the capacity or inclination to adequately
on teacher appraisals, and selective alternative pathways fulfil their responsibilities.
for top-performing veteran teachers. (However, given the
Panel (a) of Figure 1 shows instead the conventional
complexity and contingency of teacher motivation, as
‘commitment from day 1’ civil service approach to teacher
discussed in Section 2.1, such performance-based career
careers. Teachers (and other fresh graduates) in the public
elements will need to be designed with a careful eye to
sector are often hired on starting salaries that are relatively
high for new entrants to the labour market43 Consequently, abysmally low; and where performing well or badly will
the profession will attract a motley mix of prospective not affect a person’s career in either their compensation
teachers, some of whom are genuinely passionate about or assignments?’ Phrased this way, the question seems
student learning, and some of whom are making a primarily absurd, but this is exactly what exercises like that of the
financial calculation about which occupation to enter. Also, Education Commission do when they recommend (and
there is little practical distinction between novice teachers build into their estimated financing needs) ratios of
and those who are fully equipped and socialised into the teacher salary to GDP. This ‘level of teacher pay inside
profession. Throughout the span of the career cycle, the commitment-only career structures’ approach ignores the
salary scale is relatively flat, with little variation besides facts that (a) some very poorly performing public sector
increments from seniority. Although some experienced and systems (like India and Pakistan) have extraordinarily high
veteran teachers may eventually move into administrative compensation levels for public sector teachers (with pay 8
roles, and although there is a small amount of natural to 10 times as high as private sector teachers) and yet these
attrition along the way, the teaching corps stays relatively do not lead to higher learning performance in the public
stable from the novice phase onward, at least until sector, and (b) uniform increases in teacher compensation
teachers gradually hit retirement age. have been shown both through experience and experiment
(e.g., de Ree et al., 2017; discussed in Section 2.2 above) to
Discussions of teacher wages or teacher compensation that
have little or no impact on teacher performance or student
take for granted the ‘commitment early and only’ approach
learning.
are bound to get everything wrong. That is, they (often
unknowingly) ask questions like ‘What should the teacher In contrast to this focus on compensation to the exclusion
pay level be to attract people who would be capable and of all other levers for supporting teacher capability
committed classroom teachers into a system in which and commitment, the 5Cs encourage a comprehensive
education system goals for learning are not clearly defined approach. Table 6 gives an example of how the 5Cs example
and/or not collectively pursued; where existing teachers in panel (b) of Figure 1 would map onto the hedonic and
are demoralised, the social approbation of teachers dynamic model of teacher careers (with the pre-service
has fallen, the norms of performance have deteriorated, phase left out for conciseness).
standards of recruitment are low, and learning levels are
43 As with any other aspect of teacher careers, this is not universally the case. For example, comparisons of teacher salaries to other
comparable professions show decidedly mixed results across countries in Latin America (Bruns & Luque, 2015; Mizala & Ñopo, 2016) as
well as Sub-Saharan Africa (Evans, Filmer, & Yuan, 2020). However, relatively high salaries for public-sector teachers as compared to
their private-sector counterparts have been documented in a range of countries, including the U.S. (Ballou & Podgursky, 1998), Pakistan
(Andrabi, Das, & Khwaja, 2008), and India (Lemos, Muralidharan & Scur, 2021)—and the example laid out in this section is based on
Pritchett & Pande (2006), which was written in the context of the Indian education system.
Table 6. An application of the 5Cs would vary the design elements for each teacher career phase, to suit phase-specific
and overall systemwide goals for the teaching profession (illustrative)
Phases of teacher careers

Novice Experienced Veteran
Cultivate student learning;
Cultivate student learning, with a Cultivate student contribute to school/system
Delegation reduced teaching load; consolidate learning; contribute to improvement;
instructional competencies school improvement optional: master teacher
pathway
Lesson observations and review discussions for ongoing development
Information End-of-phase review based on
Portfolio of instructional materials
multiple indicators of instructional
for periodic formal appraisals
competencies
Design Well-aligned curricula, assessments, teachers’ guides, and classroom materials
elements Support Structured induction to Annual in-service training requirement, matched to
of teacher consolidate competencies and instructional needs; specialised training for those on
careers develop specialisations selective pathways
Permanent contract
Fixed-term contract with a flat with a higher pay
As with experienced teachers;
Finance and relatively low (but above- scale with increments
optional: early retirement
subsistence) pay scale subject to periodic
appraisals
Teaching as an empowered, strongly performance-normed, contextually embedded
profession that cultivates student learning; with norm reinforcement via systemwide
Norms
communication, alignment of design elements, and provision of resources (earmarked time,
funding, networks, etc) for teacher collaboration
Besides the differences between this approach and the In prioritising commitment and norms rather than solely
typical civil service approach, it is also worth noting emphasising compensation levels, this approach pays due
that the 5Cs are also distinctly different from the three regard to the complexity of teacher motivation.44
common incoherences in teacher career reform discussed
in Section 2.2 In establishing a shared goal (capable and
‘Curation’ is common in comparable professions—
committed teachers) and implementing an interacting but unusual in teaching in developing countries
set of policies to progressively achieve that goal (choose
and curate toward commitment, as instantiated in the At this point, it is important to note that even though a
combination of design elements in Table 6, this approach novice period with high rates of curation-oriented turnover
is clearly systemic rather than piecemeal. In emphasising is unusual in the teaching profession, it is common
the importance of both capability and commitment, and in other occupations. It is also important to note that
the fact that a period of in-class curation is needed for curation benefits not only the education authorities and
discerning the best fit between teachers and system needs, organisations that employ teachers, but also teachers
this approach respects the fact that teaching is a complex themselves.
task, rather than a straightforward standardisable task Curation can benefit individuals who leave the profession
that can be delivered by interchangeable civil servants. and switch to an occupation for which they experience
44 For a complementary perspective, Ingersoll (2017) identifies three dominant (and competing) explanations for shortcomings of
teacher quality in the U.S.: that entry to the profession is too restrictive; that the supply of teachers in insufficient to keep up with
demand; and that teachers are underqualified and should face more rigorous coursework requirements. But he concludes that all three
explanations miss the root cause of the problem, i.e., that teaching is viewed as a straightforward semi-profession rather than complex,
high-status profession. He argues that ‘solving the problem of teacher quality will require addressing the underlying systemic roots of
the problem. … Rather than solely focusing on trying to force the existing arrangement to work better, this alternative suggests we view
teacher quality issues as a design problem: that we need a different arrangement, better built for those who do the work of teaching.
From this perspective, to improve the quality of teachers and teaching, it will be necessary to improve the quality of, and respect
afforded to, the job and occupation of teaching’ (p. 93, emphasis original).
better ‘match quality’. Using survey data of U.K. university and similar occupations in the U.S. also imply a reasonably
graduates, Malamud (2010) found that those who switch robust initial ‘curation’ period. For the teaching profession
to occupations that are unrelated to their field of study specifically, a review of the literature on teacher
initially earn lower wages, but also experience a faster recruitment and retention in the U.S. identified a consistent
rate wage growth so that the difference between switchers pattern in the literature of high attrition from the teaching
and non-switchers is insignificant after six years. In other profession among novice teachers, followed by low exit
words, curation can be costly for the individual at the rates until retirement age is reached (Guarino, Santibañez,
point of transition, but not in the long term. More generally, & Daley, 2006). Importantly, despite the frictions that this
Woodcock (2015) estimates that match quality, as distinct novice teacher turnover may generate,46 this early-career
from individual productivity and firm productivity, turnover does appear to have beneficial ‘curation’ aspects.
accounts for 16 percent of variation in earnings in a large, For example, in an analysis of teacher turnover in an urban
representative longitudinal dataset from two U.S. states. U.S. school district, Murnane (1984) concluded:
This suggests that switching jobs to seek better match
The findings reported in this paper provide no
quality can benefit an individual’s cumulative financial
support for the hypothesis that patterns of teacher
earnings—and, likely, their job satisfaction and motivation
turnover have a detrimental effect on the quality
as well.45
of public school teaching staffs. In fact, the results
To illustrate the fact that such occupational switches are indicate that there is selective attrition from one
a common phenomenon, we look at data from the U.S., urban public school system of the less productive
simply because disaggregated labour market data are elementary school teachers. All of the selective
more extensively analysed and more easily available for attrition that does occur takes place during the first
the U.S. than for most other countries, and because the years on the job, a pattern consistent with the view
American education system is an example of a system that important new information about the quality of
that performs at good-but-not-stellar levels, with a well- the job match is generated during the initial years on
established, learning-oriented teaching profession that the job, and this information causes administrators
nonetheless faces unresolved challenges. and/or teachers to take actions that influence the
duration of the job match (p. 517, emphasis added).47
Starting with a broad-brush data, occupational mobility—
i.e., the likelihood that an individual is in a different Notably, early-career turnover among teachers in the U.S.
occupation compared to the previous year (or, if they is comparable to those in professions that require similar
were unemployed in the previous year, compared to their levels of education and certification, such as nurses, social
most recent previous employment)—was between 15 and workers, and accountants (Harris & Adams, 2007). (In
20 percent in the U.S. in the 1990s, depending on how fact, among the four professions that Harris and Adams
narrowly occupations are defined (Kambourov & Manovskii, analysed, early-career switches to other professions
2008). The same study also found that occupational were substantially higher among social workers than
mobility rates are significantly higher for younger workers, among teachers, nurses, or accountants. This is perhaps
which would align with the idea of curation, i.e., an early- reflective of the fact that social workers, like teachers, have
career phase of higher initial job turnover as employees a complex and motivationally demanding job involving
explore how well-matched they are to their occupations. personal, purpose-driven interactions, but tend to be even
more poorly compensated than teachers.)
Going beyond general labour force averages toward
occupation-specific analyses, turnover rates in teaching However, robust early-career occupational turnover
45 It is worth noting that match quality matters not only for a teacher’s match to the profession and to the education system, but
also to the specific school in which they are embedded. Poor match quality between a teacher and their school can be problematic for
multiple reasons, not least that it complicates the curation process, e.g., a novice teacher who does not work well with the headteacher
in their assigned school may have potential for capability and commitment in a different school.
46 For more on frictions generated by teacher turnover, see Section 3.5
47 Note that the findings in Murnane’s (1984) study, in examining how long teachers stay employed within a single school district,
combined the effects of the quality of job match (i.e., the match between an individual teacher and their district/school) and occupational
match (i.e., the match between an individual and the teaching profession). In this primer, ‘curation’ relates primarily to occupational
match because we are looking at entry and exit to the teaching profession as a whole. Nonetheless, job match between individuals and
schools can significantly affect teacher performance (see Jackson, 2013, for an analysis using longitudinal data from a U.S. state).
appears to be relatively rare among teachers in low- and Senegal already hold permanent contracts, implying
middle-income countries. This is due in part to the fixed that there is no meaningful probation period. In contrast,
salary schedules that dominate education bureaucracies most novice teachers in Guatemala hold non-permanent
worldwide (Bruns, Filmer, & Patrinos, 2011; Crehan, 2016). contracts—but this also applies to most experienced
Focusing specifically on developing countries, available teachers, suggesting a lack of job security that may affect
data suggests that teacher career and compensation teachers’ commitment to the profession as well as the
structures diverge considerably from the 5Cs. As shown in strength of professional norms.
Figure 9, among public school teachers in seven middle-
Notwithstanding the worrisome implications of Figure 9
income countries in PISA-D, teacher cohorts tend to be
that these seven countries lack a meaningful probation
much larger among experienced teachers (those falling
period for teachers, it is also true that the distributions
between the two reference lines, i.e., those with roughly
of teacher experience in Figure 9 are shaped by a wealth
5–25 years of experience) than among novice teachers
of complex factors, such as demographic change and
(those to the left of the first reference line, i.e., those
politicians’ changing policy platforms, that go far beyond
with less than 5 years of experience). This is the reverse
the realm of the 5Cs and other such principles into the
of what would be desirable under the 5Cs, where novice
nitty-gritty of implementation. In what remains of Part
teacher cohorts should be substantially larger to allow for
3 of this primer, we take a step back from this empirical
both curation and for some organisational selection (i.e.,
snapshot in Figure 9 and from the specific example of the
choosing) of the most effective teachers in the novice–
5Cs in Figure 1 and focus on the 5Cs as a set of principles
experienced transition. Moreover, the vast majority of
more broadly, to consider what the various ‘C’s individually
novice teachers in Cambodia, Paraguay, Zambia, and
and collectively contribute to teacher careers.
Figure 9. Teacher career structures in PISA-D countries appear to diverge substantially from the ‘choose and curate
toward commitment’ elements of the 5Cs
Source: Authors, using PISA-D data (OECD, 2018).

Notes. Sample comprises teachers in publicly funded schools that teach 15-year-old students. N(teachers)=513 for Guatemala; 414
for Honduras; 1,865 for Ecuador; 989 for Senegal; 789 for Zambia; 1,472 for Paraguay; and 3,453 for Cambodia. Years of experience and
contract type are based on self-reported teacher questionnaire data. Data are unweighted because PISA sampling weights are calculated for
representativeness at the student level, not the teacher level. Countries are sorted in decreasing order of the proportion of teachers on non-
permanent contracts. Reference lines are at 5 and 25 years of teaching experience.
3.2 Why should teachers be not only ‘capable’ which they derive the most personal satisfaction and social
but also ‘committed’? Because motivation and validation). In political theory, Mansbridge (2014) argues
purpose matter that when it is possible to select a set of agents whose
personal goals are aligned with those of the principal,
So far, we have noted that the complementary then ‘sanction-based accountability’ should be rejected
characteristics of teacher commitment and capability are in favour of ‘trust-based accountability’, which entails
a manifestation, at the frontline, of the shared purpose and substantial upfront effort to select such aligned agents but
technical expertise that constitute the core of a successful subsequently demands much less effort to monitor and
organisation. Thus, as a feature of empowered, highly sanction them—especially when the agents’ work might be
respected, strongly performance-normed, contextually difficult to monitor closely or when it is desirable for the
embedded teaching professions that cultivate student agent to act flexibly (both of which, again, very much apply
learning, ‘commitment’ is the element most closely to teachers’ work).
associated with the shared purpose around which different In addition to the fact that the goals that teachers should
elements and levels of an education should be coherent. fulfil are not only complex, difficult to monitor, and
This is similar to how this shared purpose runs through facilitated by flexibility, a second and closely related
personal satisfaction and social validation at the level of reason why teacher commitment matters is that the goal of
individual motivation, and delegation and norms in the cultivating children’s learning demands not only technical
design of career structures. skill, but also socioemotional competence and effort. In
In addition to the centrality of teacher commitment other words, teacher motivation and commitment matter
to systemwide alignment, there are at least two other not only for organisational efficiency in a managerial sense,
reasons why any education authorities and organisations but also in a more fundamental sense because these
that aspire to cultivate student learning should care about psychosocial traits and ethical commitments can shape
not only teachers’ technical capability but also their children’s cognitive and socioemotional growth. This
motivational commitment.48 is true not only in the obvious sense that teachers serve
as role models for students’ affective, social, and moral
First, if agents are expected to fulfil a complex set of tasks
development, 50 but also in the sense that meaningful
(which, as noted in Section 1.1 and discussed further in
socioemotional engagement is an ingredient in classroom
Section 2.2, very much applies to teachers), then caring
interactions that cultivate academic learning (see
about their commitment facilitates organisational
DeJaeghere, Duong, & Dao, 2021, for empirical examples
efficiency. This argument has been framed in a few
of socioemotional engagement in high-performing
different ways.49 In economics, Besley and Ghatak (2005)
classrooms in Vietnam; for a review of teacher motivation
propose that when an organisation produces collective
research, including the links between teacher motivation,
goods (that have wider social benefits rather than only
teaching effectiveness, and student motivation, see Han &
private individual benefits), then it tends to attract
Yin, 2016).
motivated agents who prioritise the intrinsic benefits of
their work rather than just what they are paid (see also Just as the core of effective organisations constitutes not
Honig, forthcoming). Accordingly, agent motivation and only technical practices but also the shared purpose that
organisational productivity are optimised when there is is the basis for those practices in the first place, so should
the closest possible match between the organisations’ education systems that are coherent for learning actively
mission and agents’ mission preferences (i.e., the varieties prioritise and cultivate teachers who are not only capable
of mission that they value the most or, in our terms, from but also committed.
48 See also World Development Report 2018, Chapter 6: ‘Teacher skills and motivation both matter (though many education systems
act like they don’t)’ (World Bank, 2018).
49 For a more detailed discussion of this literature, see Honig (forthcoming).
50 In the domain of children’s socioemotional development, there is a third reason for valuing committed teachers, at the intersection
of the other two reasons. That is, given that the cultivation of beliefs is impossible or, at best, highly difficult to verify externally (because
people can pretend to subscribe to beliefs that they do not genuinely hold), then education authorities and organisations have little
choice but to trust that teachers are trying to foster children’s development of the beliefs and dispositions that the organisation
prioritises (whether national identity, religious foundations, ethical principles, etc.)—which, in turn, implies that it is optimal to employ
teachers who are personally committed to these beliefs and dispositions (see Pritchett, 2013; Pritchett & Viarengo, 2015).
3.3 Why ‘choose and curate’? Because the cycle, public-sector teaching offers a decent starting salary
early career phases are pivotal for systemwide with small but reliable increments throughout the career.
norms of capability and commitment Additionally, in many settings, public-sector teaching also
offers substantial pensions or other retirement benefits.51
Under the 5Cs, the choose-and-curate period during the
pre-service and novice career phases is an indispensable This hypothetical choice set reflects empirical realities, at
opportunity to establish strong norms of quality teaching, least in some settings. A longitudinal study of the career
both for the individual teacher and for the teaching movements of 130,000 teachers in Florida found that
profession more broadly. For example, the 1966 ILO/ the those who left the teaching profession had a much
UNESCO Recommendation Concerning the Status of more dispersed distribution of earnings as compared
Teachers—which is one of the key normative instruments to the relatively compressed wage distribution within
informing the global teacher union federation Education teaching (Chingos & West, 2012). Thus, career choice will
International’s work on teacher status (see Education likely be influenced not only by job-specific factors such
International, 2017)—states that: as the strength of interest in education, but also by risk
preferences—not unlike choosing between low-risk but
A probationary period on entry to teaching should relatively low-yield bonds versus high-risk stocks that
be recognized both by teachers and by employers as offer the prospect of much larger gain. A large-scale panel
the opportunity for the encouragement and helpful survey in Germany found that teachers tend to be more
initiation of the entrant and for the establishment risk-averse than their peers who made other occupational
and maintenance of proper professional standards choices (Ayaita & Stürmer, 2020), with similar findings
as well as the teacher’s own development of his about master’s in education candidates compared to other
practical teaching proficiency. The normal duration graduate students in a lab-based risk preference task in at
of probation should be known in advance and the a U.S. university (Bowen et al., 2015).
conditions for its satisfactory completion should
be strictly related to professional competence (ILO/ Thus, for prospective entrants to the profession who
UNESCO, 1966, paragraph 39, emphasis added). are weighing the option of pre-service/novice teaching
against other choices, the hedonic disutility (i.e., the
Similarly, one of the five principles of the World Bank’s unappealing nature) of the initial lower-paid, fixed-term
Global Platform for Successful Teachers is: ‘Promote contract, coupled with the uncertainty in whether they will
meritocratic selection of teachers, followed by a eventually make the grade for a permanent contract and
probationary period, to improve the quality of the teaching whether they will subsequently attain desirable levels of
force’ (Béteille & Evans, 2019, p. 2), which corresponds to professional responsibility and compensation, can act as a
choosing followed by curation. screening mechanism to inhibit would-be entrants who are
A key benefit of ‘choose and curate’ is that it mitigates the less committed to teaching as a long-term career.
system-level uncertainty about the eventual in-school Such a screening device is important because initial
motivation and effectiveness of prospective teachers by teacher training can be a costly investment for education
making some of this uncertainty an explicit part of the authorities and organisations (to say nothing of the hours
decision facing prospective teachers at the individual level. of children’s time spent in lessons with novice teachers). It
To explain what we mean by this, recall that panel (a) of is inevitable and appropriate that some of this investment
Figure 1 compared the typical careerlong compensation will not yield long-term results because some novice
that a hypothetical graduate could expect to earn by teachers will eventually discover that another career would
entering the civil service as a teacher with the careerlong be a better match for their inclinations and strengths,
compensation that could be expected from opting into which is a desirable feature of curation. However, it is less
the private sector (treated generically across industries desirable for some of this initial investment to be wasted
for simplicity). While the generic private-sector option on prospective teachers who cynically view their initial
offers a lower starting salary and much less certain but teacher training and in-school placements solely as a
potentially much higher increments later in the career springboard to a different career outside of education, or
51 In the U.S., for example, pensions constitute a large portion of education spending—an extreme example being the state of Illinois,
where subsidies for teacher pensions in 2016 amounted to one-third of total state-level spending on K–12 education (Costrell, Hitt &
Shuls, 2020). This is despite a lack of evidence that current teacher compensation and retirement benefits in the U.S. are structured to
optimise either student learning, teacher motivation, or teacher preferences (Costrell & Podgursky, 2009; Johnston, 2021).
who intend to treat their jobs as a low-effort civil service performance appraisals (a) should be cumulative, rather
sinecure. This may be a particular issue in contexts where than focusing only on the most recent school year (and
teaching is regarded as a relatively appealing career all the vagaries that may have afflicted it); and (b) should
choice (whether for favourable reasons like its traditional involve a range of information sources, such as information
prestige or unfavourable ones like a paucity of comparably on student learning (both ‘thin’ sources such as test
appealing career choices). In other words, applying the scores and ‘thicker’ sources such as reviews of student
5Cs principles cannot overcome the unknowability of coursework), lesson observations, other types of teacher
how a prospective teacher will respond to the complex peer review, and parent feedback. The probationary
stresses of the classroom, but it can help to mitigate the novice phase will only contribute to ‘the establishment
difficulty that education authorities’ and organisations’ and maintenance of proper professional standards’—or,
face in discerning what prospective teachers already know, put differently, it will only help to make teaching a strongly
ex ante, about their private preferences and plans. Thus, performance-normed, contextually embedded profession
‘choose and curate’ aligns with long-established arguments that cultivates student learning—if the post-probation
in labour economics about self-selection and turnover (e.g., promotion to the permanent employment contract is
S. Salop & J. Salop, 1976; for an overview, see Prendergast, based on ‘conditions … strictly related to professional
1999, on deferred compensation)—in addition to being a competence’ (ILO/UNESCO, 1966, paragraph 39). That
pivotal point for the development of professional norms, is, the 5Cs aren’t just about delegation (who does what
as argued above. where—and why?) and finance (compensation), but also
about good information on how well novice teachers are
Three caveats:
developing their instructional capacities, and about the
First, even during the choose-and-curate stage, novice norms that are reinforced through these other design
teachers must be paid above subsistence level. Above- elements, as illustrated above in Table 6.
subsistence compensation is crucial not only for novice
Third, also as stated in the ILO/UNESCO recommendation,
teachers to be able to focus their efforts on teaching rather
another crucial function of the ‘curate’/novice/probation
than on second or third jobs, but also to reinforce norms of
period is to consolidate ‘practical teaching proficiency’.
teaching as a profession with commensurate status. That
That is, the 5Cs are also about support, i.e., equipping and
said, in many fields, such as medicine in the U.S., there
enabling teachers to do their jobs better. This interplay
is an expectation of a long period of apprenticeship in
between the five design elements of teacher careers is
which earnings are low before doctors are fully certified in
a reminder that, as noted in Section 2.1, teacher careers
their profession. To a very significant extent it is the long
are interactive systems. Looking specifically at the design
probationary apprenticeship period followed by merit-
element of support, such support for novice teachers is
based selection that creates and sustains the social status
pivotal both for establishing professional norms and for
of a profession.
ensuring successful transitions from pre-service training to
Second, the performance appraisals used in promoting the classroom. For example, a study of novice teachers in
novice teachers to permanent status, and in awarding Indonesia based on a two-year series of reflective journal
annual salary increments for experienced and veteran entries and interviews found that many teachers, especially
teachers, need to be coherent with the expectations of those in public schools, felt inadequately supported—
the profession (which, in turn, should centre on cultivating and many also felt incompetent and inadequate despite
student learning, as emphasised in premise for practice having completed both a four-year undergraduate
#1). In other words, information must be coherent with initial teacher education programme followed by an
delegation. Additionally, given that teaching is complex, additional year of professional certification training
(Alifia, Pramama,&Revina, forthcoming). 52 Nationally or to leave the profession if they received more extensive
representative surveys of teachers in the U.S. indicate that induction supports such as mentoring, reduced workloads,
novice teachers are less likely to transfer to other schools and extra resources (Ingersoll & Smith, 2004). Conversely,
PREMISE FOR PRACTICE #2: on cultivating norms during the novice teacher phase
Rather than paying no particular attention to the novice phase …
Contextually embedded, learning-oriented teacher professional norms must be cultivated throughout the
novice teacher phase.
Conventional teacher career structures fall short in two fundamental ways (among others). First, they artificially
separate teacher salaries from other sources of teacher motivation. Instead, if we view all of these sources as an
integrated whole, it becomes clear that teachers’ personal satisfaction with their work and the social validation they
derive from it—i.e., the psychosocial sources of motivation—are strongly shaped not by money, but by the dominant
norms and shared sense of purpose within their profession.
Another shortcoming of conventional teacher career structures is that they draw no distinction between different
phases of the teacher career cycle (besides progression on the uniform salary scale and perhaps in formal
administrative responsibilities). This represents a massive lost opportunity because the novice phase of the teacher
career cycle can be a pivotal period for establishing those shared norms that can shore up individual teachers’
satisfaction and validation and orient the profession as a whole toward the goal of cultivating student learning.
→ See also
• ‘Teacher motivation is driven by multiple, interacting factors’ in Section 2.1
• ‘Teacher careers are hedonic and dynamic, and any model of teacher careers has to reflect this' in Section 2.1
• ‘Changes that are incoherent with the complexity of teacher motivation' in Section 2.2
• 'Why should teachers be not only "capable" but also "committed"? Because motivation and purpose matter' in
Section 3.2
• 'Why "choose and curate"? Because the early career phases are pivotal for systemwide norms of capability and
commitment' in Section 3.3
• ‘The 5Cs encapsulate the hedonics and dynamics of teacher careers—toward the goal of capable and committed
teachers’ in Section 3.6
52 Similar observations were reported by Akyeampong et al. (2011) in a study of teacher training in Ghana, Kenya, Mali, Senegal,
Tanzania, and Uganda, in which they note that: ‘newly qualified teachers, arriving in schools with the confidence of having successfully
completed their training, often find that it has equipped them to deal neither with the difficult realities of large classes and lack of
resources, nor to make the most of the possibilities offered by carefully planned and developed curricula. This can lead to demotivation
and diminution of ambition and is subverting the effort that is going into curriculum reform’ (p. 65). In the same study, the authors also
illustrate the degree to which the pre-service and novice phases in low-performing education systems can depart from the capability-
and-commitment orientation of choose-and-curate: ‘The structure of the pre-service training programs pivots around the teaching of
a body of propositional content knowledge with methods for teaching primary school content taught separately and assessed through
examination rather than practice. … Trainees in this study replicate the way they have been trained by their tutors … Their tutors rarely
have primary school experience themselves and often do not have direct access to the current primary curriculum, textbooks or teaching
and learning resources to support their teaching. … Orientation to the primary curriculum was short at two to three weeks and based on
preparing schemes of work and lesson plans for one particular topic of the primary syllabus only. The practicum was not long enough
to develop trainees’ pedagogical content knowledge and practice in reading and maths and sometimes they did not get to observe or
even teach a reading lesson in lower primary. There was little opportunity to engage as apprentices in the practices of the experienced
teachers they did observe and little proper mentoring or guidance in schools’ (pp. 66–67).
much of the teacher attrition due to inadequate and a wide pool of novice teachers, supporting them as they
unevenly distributed support would likely constitute a adapt to its complex challenges, and facilitating their exit
wasteful shrinking of the talent pool rather than mutually from the profession if they find that they are less well-
desirable exit of those who would be more inclined to suited to the classroom than they expected, is what we
other occupations. More generally, inadequacy in any one call curation.
of these design elements would hobble the efficacy of the
To illustrate why curation can make a difference, Bau and
5Cs principles.
Das’ (2020) panel data study of teachers and students in
government and private schools in Pakistan found that
3.4 Why ‘curate’, when there’s already
observed indicators of teacher characteristics accounted
‘choose’? Because classroom teaching is
for less than 5 percent of the variation in teachers’ value-
complex
added contributions to student learning outcomes in
So far, we have argued that there are substantial benefits government schools. This is despite the fact that the
to demarcating a novice, probationary phase of the teacher characteristics included in the model included
teacher career cycle, during which teachers do not yet hold traits that might reasonably be expected to enhance
permanent appointments but do receive tailored support student learning—and which could reasonably be used as a
to develop professionally. In this section, we will argue that basis for choosing candidates for the teaching profession—
one of the most effective ways of ensuring that this early- such as whether the teacher was from the local area,
career phase contributes to a strongly performance-normed, whether the teacher held a bachelor’s degree, whether they
contextually embedded professional teaching career that had completed any teacher training, how much teaching
cultivates student learning is by combining a process of experience they had, and, notably, even how they scored on
choosing teachers from among a wider pool at certain key the same content knowledge tests that were administered
points in the career cycle with a more extended process of to their students. 53 However, in the private sector, even
curation to discern which teachers are best suited to the though there was no systematic assessment of student
specific classroom needs in question. and teacher performance, schools and headteachers could
identify and reward effective teachers, as evidenced by the
The key point here is that choosing—i.e., selection or
fact that a 1 standard deviation increase in teacher value-
screening based on characteristics that are observable
added was associated with a 49 percent higher salary
ex ante—is a necessary but insufficiently accurate way
(see also Lemos, Muralidharan, & Scur, 2021, for a similar
of identifying capable and committed teachers. This is
finding in India). Taken together, these findings suggest
because classroom teaching is complex. Complexity here
that teaching quality can be more effectively identified
implies that (a) teaching is not a straightforward task for
using granular, day-to-day observation rather than using
which we can reasonably assume that everyone above
standardised metrics.54
a certain threshold of competence will deliver pretty
much the same quality of education; and (b) given that Beyond the Pakistan case in Bau and Das (2020), there
unexpected phenomena can emerge from the interactions are many other examples suggesting that conventional
in complex systems, it is impossible to make completely indicators of teacher characteristics that can be observed
accurate predictions about how any given prospective outside of the classroom—and that could be used to choose
teacher will perform in the classroom until they have among candidates for entry to the teaching profession—are
actually spent time in the pertinent classroom context. inadequate proxies for the commitment and capability that
This process of giving frontline classroom experience to distinguish effective teachers. For example, an analysis of a
53 Of these factors, teachers’ content knowledge and years of teaching experience were associated with higher teacher value-added.
Despite this association, these factors collectively explained little of the total variation in teacher value-added (Bau & Das, 2020).
54 It’s worth noting that: (a) Bau and Das (2020) focused on academic learning, which can be proxied with reasonable accuracy using
test scores. The fuzziness of approximating teacher quality using standardised indicators is probably far worse for the less tangible
but equally important aspects of children’s development, such as socioemotional learning and ethical understanding. (b) Although
observations from day-to-day practice are likely to give better information about teacher effectiveness than standardised, ex ante
credentials and scores, headteachers’ appraisals of teacher quality vary in accuracy and usefulness. For example, a study of teacher
performance pay in Pakistan found that student outcomes significantly improved, on average, when teachers were awarded salary
raises based on managers’ subjective appraisals of their performance—yet there were no student learning gains under managers who
were ranked in the bottom quintile of teachers’ perceptions about whether their appraisal scores were likely to be accurate (Andrabi &
C. Brown, forthcoming).
public-sector teacher evaluation system in Ecuador, which who should remain in the profession long-term. Or, in the
included a written test and a demonstration lesson, among language of Section 2.2 on coherence in teacher career
other elements, found that higher evaluation scores were reforms, introducing a curation phase into teacher career
not associated with higher levels of student learning (Cruz- structures is coherent with the complexity of classroom
Aguayo et al., 2017). A major review of research on teacher teaching.57
quality in the U.S. found that there was no systematic
To clarify, the importance of ‘curate’ does not mean that
relationship between whether a teacher held a master’s
‘choose’ doesn’t matter. Choosing candidates based on the
degree and how their students performed, and that the
best available ex ante data serves (at least) two distinct
relationship between teachers’ content knowledge scores
functions: reducing the costs of curation, which will be
and student outcomes was positive but weak (Hanushek &
discussed below in Section 3.5; and strengthening the
Rivkin, 2006).55
norm that teaching is a quality-oriented profession. On
To put this in 5Cs terms, if it were the case that standardised the latter point, an analysis of a change in teacher hiring
metrics could proxy for everything that we value in rules in Mexico found that those teachers were hired based
teaching, then teacher career systems could depend wholly on their scores on a test were more effective in raising
on ‘choosing’, and they would not need ‘curation’. 56 But student learning outcomes than those who were hired
since both theory and empirical data show that teaching under the pre-existing discretionary process that was
is too complex to be reduced to standardised metrics, highly influenced by powerful stakeholders (Estrada, 2019).
curation matters tremendously for teachers’ capability and Strikingly, teachers’ test scores had no predictive power
commitment. One of the most valuable aspects of curation for how much each teacher subsequently contributed
is that it creates an extended opportunity for both early- students’ learning outcomes, as observed in other contexts
career teachers and their managers to elicit ‘thicker’, more mentioned above—rather, the merit-based selection
accurate information about their classroom-embedded procedure appeared to attract more capable candidates to
capability and commitment. This is fundamental to how apply to the profession (Estrada, 2015; see also Béteille &
the ‘choose and curate’ stage can reinforce professional Evans, 2019, pp. 27–28). In other words, choosing teachers
standards, as argued in Section 3.3: thick information based on the best available, purpose-driven data helped to
about novice teachers’ classroom practice is vital not shift perceived norms about the teaching profession.
only as formative feedback for teachers to improve their
professional practice, but also as input for teachers and
their managers to jointly make good decisions about
55 This is not to imply that teachers’ cognitive skills are irrelevant for student learning. To give one example to the contrary, Hanushek,
Piopiunik, and Wiederhold (2019) find that country-level average teacher scores on the OECD’s Programme for the International
Assessment of Adult Competencies (PIAAC) are significantly associated with student learning as measured in PISA. Rather, the point here
is that conventional standardised indicators used in selecting candidates for entry into the teaching profession do not do adequately
proxy for the knowledge, skills, and dispositions of prospective teachers who have the potential to be capable and committed in the
classroom long-term.
56 Standardised metrics have the potential to be reasonably adequate proxies for some aspects of teacher quality for teachers
who have been in service for a few years, in settings where there is reasonably good longitudinal data on student test scores. Chetty,
Friedman, and Rockoff (2014) argue that teacher value-added measures that control for students’ prior test scores can give unbiased
predictions of teachers’ impacts on subsequent student achievement, based on their analysis of data from a U.S. school district. Yet
this does not account for aspects of student development that are less easily quantified and that may fall by the wayside in systems that
overemphasise test scores (Andrabi & C. Brown, forthcoming; see also Blazar & Kraft, 2017); nor for poor quality test score data in some
low-performing education systems (Berkhout et al., 2020; Singh, 2020). Most crucially for the argument at hand, teacher value-added
measures based on previous years’ track records cannot aid the process of choosing the best candidates from a pool of prospective
teachers who have yet to chalk up any classroom track records.
57 See also Jovanovic (1979) for an economic model of this phenomenon, framed as a theory of job turnover and job matching when
an individual’s productivity in a specific job is not known ex ante but is gradually revealed as they spend more time in the job.
PREMISE FOR PRACTICE #3: on the importance of curation

Rather than awarding permanent job tenure from day 1 …
The pre-service and novice phases should be a period of curation, such that, as with nearly all other
professions, a substantial proportion of initial entrants do not persist in the career.
Teacher career structures often reflect administrative defaults within the civil service. But teaching is not just any
other logistical, post-office-style58 aspect of public service delivery. Rather, good teaching is a complex craft that
needs to respond to the individual student and to the classroom, school, and education system environments in
which it is embedded.
Because of this complexity and embeddedness, there is no way for education authorities and school leaders—nor
even for prospective teachers themselves—to judge with 100 percent certainty whether a prospective teacher
will be technically capable in, and motivationally committed to, classroom teaching until the novice teacher has
actually spent some time engaging in classroom practice. This means that any ex-ante observable characteristic
of pre-service teachers (e.g., formal qualifications, test scores, interview performance, or even demonstration
lessons) will not give an adequate picture of who will be capable and committed to cultivating student learning
for the long haul. Accordingly, the process of identifying the best match between prospective teachers and long-
term teacher careers must not only start with pre-service screening, but also continue for the first few years of
classroom service.
→ See also
• ‘Changes that are incoherent with the complexity of classroom teaching’ in Section 2.2
• ‘The 5Cs contrast markedly with the conventional civil service approach’ in Section 3.1
• ‘Why “curate”, when there’s already “choose”? Because classroom teaching is complex’ in Section 3.4
• ‘The 5Cs encapsulate the hedonics and dynamics of teacher careers—toward the goal of capable and
committed teachers’ in Section 3.6
• ‘In purpose-driven teacher career structures, decision-makers invest in—and act on—“thick” information
about teachers’ capability and commitment’ in Section 4.2
58 For the source of this image, see Aiyar & Bhattacharya’s (2016) ‘The Post Office Paradox’.
Box 4 : Research on how education authorities and organisations should choose

teachers during the pre-service and novice phases
Even though the typical observable indicators have little relationship to teachers’ classroom effectiveness, as
discussed in Section 3.4, this does not mean that it would be futile for education authorities and organisations to
invest in refining ex ante assessments for choosing pre-service and novice teachers. We would hypothesise that
such investments would be especially worthwhile when the assessments are closely tied to the forms of capability
and commitment that would advance a given education bureaucracy’s core goals.
Recent studies suggest some observations about how to improve the quality of screening for entry into pre-service
teaching training and for entry into the classroom:
• To gauge how capable an early-career teacher may be, it may be worth developing instruments measuring not
only their content knowledge on the same tests that their students will take, but also their pedagogical content
knowledge of how students are likely to interact with the content in question (Shulman, 1986)—as in a study
in Peru that tested teachers on their ability to identify and explain students’ misconceptions in mathematics
(Cueto et al., 2017).
• While it is even harder to predict ex ante how committed a prospective teacher may be to the profession in the
long run, some researchers are in the process of developing situational judgement tests and other instruments
may be more informative about candidates’ non-cognitive attributes (e.g., empathy, conscientiousness) than
conventional interviews and written tests (Klassen et al., 2020).
• More generally, using multiple, tailored assessments can improve the accuracy screening processes. An
analysis of Peru’s centralised assessment for pre-service teacher education graduates, comprising assessments
on reading comprehension, logical reasoning, and curricular and pedagogical knowledge, found (perhaps
unsurprisingly) that the combined assessment score was more strongly correlated with teachers’ subsequent
value-added scores in the classroom than were any of the three assessment components individually (Bertoni
et al., 2020). (See also Malinen et al., 2012, on Finland’s multistage screening process for entry into pre-service
teacher education, as mentioned in Section 4.2 below.)
Although we believe that some amount of curation is a necessary part of an effective teacher career system,
improving the quality of choosing will help to mitigate the costs of this curation, which we discuss next.
3.5 Why ‘commitment’? Because continuous continuous curation is costly—whether for ministries,
curation is costly schools, and students, or for teachers themselves. To
quote a study of teacher turnover in Rwanda:
Having made a case for the choose and curate elements
of the 5Cs, and having argued that students and school Teacher turnover affects the production of student
systems need teachers that are both capable and learning not only indirectly, through the stock of
committed because organisational effectiveness requires teachers’ skill and motivation, but also directly,
technical capability aligned to a core purpose, we now because it introduces frictions in the allocation
focus on the remaining ‘C’, commitment. Specifically, of teachers to schools. … head teachers faced
if choose-and-curate can be so useful for building a with staffing shortages may be forced to allocate
strongly performance-normed teaching profession, and teachers to subjects outside of their areas of
if ‘commitment early and only’ teaching professions specialty, resulting in a patchwork of teaching
tend to be so dysfunctional, why should we bother with assignments that adversely affects learning
commitment at all? Why shouldn’t education ministries outcomes (Zeitlin, 2021, p. 82).
simply hire all teachers on short-term contracts for which Thus, from a service delivery perspective, careerlong
renewal is contingent on performance?59 curation implies constant turnover across the entire pool
The short answer is that commitment matters because of teachers, and teacher turnover generates substantial
frictions for students, school leaders, and school systems
59 For an overview of the incidence of contract teaching, see UNESCO (2020) on Sub-Saharan Africa and Kingdon et al. (2013) on
developing countries more generally.
as a whole (see also Ronfeldt, Loeb, & Wykoff, 2013, using uncertainty about compensation—apply to teacher career
data from New York City schools). structures that apply the choose-and-curate-toward-
commitment principles of the 5Cs. Yet, for all of the reasons
In addition to general turnover frictions, novice teachers
described in the previous sections, we believe that these
are typically less capable than their experienced peers
costs are worthwhile.
simply because they have had less time to consolidate their
pedagogical competencies (Araujo et al., 2016; Bau & Das, It is also worth noting that the opposite extreme, i.e., a
2020; Hobbis et al., 2020; Kraft et al., 2020; Papay & Kraft, commitment-from-day-1 approach, can be extremely
2015; and Podolskly et al., 2015). Consequently, constant costly as well. As discussed in Sections 3.1 and 3.3,
curation would mean that there is a larger share of novices such commitment-only approaches sacrifice a prime
across the full pool of teachers, which would mean lower opportunity to establish professional norms during the
the average quality of lessons delivered—which would probationary choose-and-curate period. Consequently,
negatively impact student learning. education bureaucracies that use such commitment-
from-day-1 approaches often lack the systemwide
From a teacher motivation perspective, continuous
professional norms that can buttress teachers’ personal
curation would entail a degree of job insecurity that
satisfaction and social validation across the full pool of
could be highly demotivating. As noted in Section 3.3,
teachers—leaving employers to depend only on two of the
certainty about your expected compensation can be
four sources of teacher motivation, as indicated in Table 7.
hedonically valuable over and above the hedonic value of
This implies that a commitment-only system that seeks to
the compensation itself. Conversely, constant uncertainty
attract more capable and committed teachers or to shore
about whether or not you will continue to be employed in
up teacher motivation within the system will have to pay
the same role from year to year can affect all four sources
them more, whether directly through finances or indirectly
of motivation that teachers derive from their jobs, as
through material circumstances. Such a system could be
indicated in Table 7. Moreover, the persistent denial of job
tremendously costly from a financial standpoint, without
security may communicate the message that teachers are
addressing the root causes of teachers’ demotivation.
disposable or, at least, that they are not highly valued as
To put this in the language of Section 2.2 on incoherent
professionals, which can be damaging for professional
teacher career reforms, commitment-only approaches
norms (see also Bennell & Akyeampong, 2007, on how
to teacher careers often get locked into finance-focused
employing para-teachers can have negative effects on the
teacher career reforms that are incoherent with the
status of regular teachers). Continuous curation can also
complexity of teacher motivation and of education systems.
impede teachers’ pursuit of long-term professional goals
As for service delivery, a commitment-only approach may
and of rich relationships between teachers and the local
have a smaller share of novice teachers who are still honing
community, as they may legitimately and prudently choose
their pedagogical skills, given the lack of a curation phase;
to limit the effort that they invest in their work to hedge
but we would still expect a 5Cs-based approach to have
their bets against the threat of contract non-renewal.
a higher average level of pedagogical quality because the
Of course, all of these costs of curation—turnover frictions, orientation toward teacher capability and commitment is
a relatively large proportion of novice teachers who have baked into the system.
yet to consolidate their pedagogical competencies, and
Table 7. Unlike commitment-only and career-long curation approaches, the 5Cs directly support all four sources of
teacher motivation
Commitment only Careerlong curation
Sources of teacher motivation The 5Cs
(typical civil service model) (contract-teacher-only model)
Finances ✓ ✓ ?
Material circumstances ✓ ✓ ?
Personal satisfaction ✓ ? ?
Social validation ✓ ? ?
Given both the significant costs and substantial benefits stable but not especially capable or committed corps of
of curation, the 5Cs propose a balance, with curation civil service teachers, largely because this is the prevailing
during the pre-service and novice phases of the teacher norm across the civil service. In turn, some education
career cycle in order to gain the benefits of curation at a bureaucracies that attempt to remedy teacher quality
high-leverage phase when teachers are being inducted issues by hiring large numbers of contract teachers are
into the profession, followed by commitment during the instead paying a premium to be able to fire their teachers,
experienced and veteran phases when those who remain which may be less costly in terms of direct payroll budgets,
in the profession have demonstrated that they can act in but may incur substantial costs in terms of professional
the classroom as capable and committed teachers. norms as well as teachers’ personal satisfaction and
social validation, which in turn have indirect but large cost
Built into the 5Cs and the associated premises for
implications.
practice are three measures for containing the costs
of curation, both in terms of costs of service delivery Relatedly, the point of this primer is to ask: how can
and costs to individual teachers. First, pairing ‘choose’ education authorities and organisations cost-effectively
with ‘curate’ helps to contain the costs of curation from develop empowered, highly respected, strongly
a service delivery perspective by screening out some performance-normed, contextually embedded teaching
candidates who would be obviously better suited to professionals who cultivate student learning? Our answer
other occupations. Second, restricting curation to the is that they should apply the principles of ‘choose and
pre-service and novice phases of the teacher career cycle curate toward commitment to capable and committed
helps to contain the costs of curation both from a service teachers’, or the 5Cs.
delivery perspective (turnover frictions, lost investments
The 5Cs imply that education bureaucracies should pay a
in teacher professional development, etc.) and from a
premium to: (a) immerse a relatively wide pool of early-
motivational perspective for individual teachers.60 Finally,
career teachers in the classroom, while allocating resources
investing in high-quality support for teachers to improve
to support their induction into the profession and to
their pedagogical competencies, as in premise for practice
identify those novice teachers who demonstrate the forms
#4 below, especially for pre-service and novice teachers,
of capability and commitment that are most closely suited
helps to mitigate the impact on average lesson quality
to systemwide goals (constituting the ‘choose and curate’
from having a larger proportion of less experienced
stage); and (b) retain and motivate those teachers who
teachers in classrooms. Besides mitigating the costs of
have demonstrated that they are capable and committed in
curation on the quality of service delivery, good support
the classroom (constituting the ‘commitment’ stage). Thus,
for teachers’ pedagogical competencies and professional
the 5Cs are fully hedonic in that they account for the full
development will also mitigate the costs of curation for
range of sources of wellbeing and motivation that teachers
individual teachers who leave the profession given that
would value, on the supply side; and, on the demand side,
their training and classroom practice will have equipped
they offer a principled consideration of how education
them with skills—such as communication, coaching,
bureaucracies should be allocating their resources in order
project management, and sequencing information in
to build a teaching profession that serves the overarching
cognitively appropriate ways—that can be unbundled
purpose of cultivating student learning.
from core pedagogical competencies and applied to other
occupations in which these former teachers may enjoy The 5Cs are also fully dynamic. On the supply side, this
better match quality. approach assumes that prospective teachers decide to
enter the profession by considering the value they expect
3.6 The 5Cs encapsulate the hedonics and to gain from their work over the full span of a career cycle.
dynamics of teacher careers—toward the goal On the demand side, the 5Cs unbundle the phases of the
of capable and committed teachers teacher career cycle and propose that education systems
should pay more, in both salary scales and job security, to
In economics, the point of a hedonic approach is to ask:
experienced and veteran teachers who have demonstrated
what are you paying for, and why? Unfortunately, some
that they are capable and committed in ways that advance
education bureaucracies are paying a premium for a
systemwide goals. Unproven novice teachers, in contrast,
60 Another way of mitigating the costs of curation for teacher motivation would be awarding a lump-sum exit payment to those
teachers who complete the novice phase but are not chosen for permanent contracts (as proposed in the context of India’s public service
by Muralidharan, 2016)
may draw lower salaries on limited-term contracts, which and Akyeampong’s (2007) study of teacher motivation
is cost-effective for education budgets while also serving in twelve countries in Sub-Saharan Africa and South
as a screening device to favour self-selection into the Asia found that low wages seriously undermined teacher
profession by those prospective teachers who believe motivation—but so did the low and declining status and
that they have a good chance of proving themselves to be social validation of the profession. Moreover, as noted in
capable and committed teachers in the long run. Section 1.2, these inadequacies in both teacher pay and
teacher status frequently derive from another form of
Having said that, we recognise that some countries
dynamics: interconnected systemwide changes over time,
face chronic teacher shortages, such that the notion of
wherein the rapid expansion of schooling led to an erosion
rigorously choosing from among a pool of applicants and
of teacher professional norms and status. Given all of these
further whittling down this pool via an extended period
mutually reinforcing factors, fixing the design element of
of curation sounds fanciful at best and wantonly wasteful
finance (i.e., teacher compensation) on its own, without
at worst. In these contexts, it may seem like the most
altering the delegation and norms that shape expectations
sensible and straightforward response to shortcomings in
and goals, would be unlikely to shift the profession away
teacher supply, quality, and motivation is to pay teachers
from its low-performance, low-motivation equilibrium. In
more. Yet, as we have argued throughout, financial
Part 4, we explore some alternative ways forward.
issues rarely occur in isolation. For example, Bennell
Part 4. The 5Cs in practice: Varied examples of common principles
“These textiles are so completely absorbed into the patterns of daily life in many
parts of Africa that they are everywhere but invisible,” observes an art historian.
“This is a major African art form, which is also a major European art form and a
major Asian art form. It is, in short, complicated.” Textiles tend to be. The cultural
authenticity of cloth arises not from the purity of its origins but from the ways
in which individuals and groups turn textiles to their own purposes. … Trying to
impose an external standard, heedless of consumers’ beliefs and desires, is not
merely futile but disrespectful and absurd.
—Virginia Postrel, The Fabric of Civilization: How Textiles made the World (2020)
Part 4. The 5Cs in practice: Varied examples of common 63

principles
4.1 Contextual differences can fundamentally shape 64
education systems (and the possibilities for reforming
them): the example of spider and starfish systems
In the long run, starfish systems may fit more naturally with purpose-driven 65
teacher careers ...
… yet the 5Cs can be applied equally in starfish and spider systems 66
4.2 Aspects of purpose-drive teacher career structures 66
that are fundamental regardless of context
In purpose-driven teacher career structures, decision-makers invest in—and 66
act on—‘thick’ information about teachers’ capability and commitment
PREMISE FOR PRACTICE #4: on information and support for teaching quality 67
In purpose-driven teacher career structures, ‘fairness’ in compensation is 68

based on shared goals (rather than administrative defaults)
PREMISE FOR PRACTICE #5: on purpose-driven definitions of fairness 70

4.3 The journey may not look like the destination: 71
examples of feasible ways forward
Examples of the 5Cs at the ‘destination’ of empowered, 72
Examples of the 5Cs on the ‘journey’ toward empowered, 72

Thus far, in making a case for the 5Cs for teacher careers, implementation context can affect teacher career reform,
we have focused on three levels of complexity: in the we focus in this discussion on the difference between
sources of teacher motivation, in the interaction between ‘spider’ and ‘starfish’ school systems. There are many other
design elements and phases of teacher careers, and in the types of variation that also matter for the implementation
technical core of education bureaucracies. of teacher career reform, but we focus on this one because
all school systems will fall somewhere along the spider–
Besides the influence of complexity, interactions, and
starfish continuum and, consequently, this is a useful
feedback loops, another key lesson from systems thinking
illustration of the importance of context.
is the importance of context. This is certainly true for
teacher career structures. Just as teacher motivation is So, for the purposes of this discussion, we categorise
influenced by the career structures in which teachers are school systems into either centralised spider systems that
embedded, so are teacher career structures influenced tend toward top-down management and decentralised
by the education systems in which they are embedded. starfish systems where management and decision-
As noted at the outset, the 5Cs are a set of principles that making functions are distributed across different
can and should be instantiated differently in different levels and branches of the system. Of course, this is an
contexts. This aligns with broad recognition in education oversimplification—and, given the complexity of goals,
and international development that context influences the roles, and relationships within education systems, many
efficacy of policy reform (e.g., GEAAP, 2020; UNESCO, 2017). systems are hybrids, resembling one ideal type for
certain policy domains and other ideal types for other
Also as noted above, the goal of the 5Cs is to create
ones.61 Simplification notwithstanding, this typology
and sustain empowered, highly respected, strongly
reflects dominant patterns in the organisation of school
performance-normed, locally embedded professional
systems and teacher career structures, which have far-
teaching careers. In this part of the primer, we give
reaching implications for teacher career reform.62
examples of what this can look like in practice. First, in
Section 4.1, we discuss just how extensively contextual The modal way in which school systems have expanded
differences can shape teacher career structures and their enrolments—and, indeed, the modal education
education systems more generally, giving the example of system in low- and middle-income countries today—is by
centralised ‘spider’ systems in comparison to decentralised organisational expansion in the form of a spider system.
‘starfish’ systems. Next, in Section 4.2, we focus on two Spider systems are highly centralised, comprising webs
aspects of teacher careers that matter tremendously that are controlled and coordinated by a single spider
regardless of context—i.e., ‘thick’ information systems and brain at the centre (Brafman & Beckstrom, 2006; Pritchett,
purpose-driven definitions of fairness in compensation— 2013). They can manifest in different forms. In the U.S.,
and show how these aspects can look very different in standalone schoolhouses were consolidated into spider
different contexts. This section also sets out premise for systems to serve urban-industrial development during the
practice #4 and premise for practice #5 about what to do Progressive Era from the late 19th to the early 20th centuries
differently under the 5Cs. Finally, in Section 4.3 we discuss (Tyack, 1974). In many low- and middle-income countries,
how the journey toward empowered, purpose-driven the expansion of spider school systems was designed to
teacher careers may look different from the destination, create a ‘modern’, postcolonial citizenry during the 20th
again giving examples from different education systems. century (e.g., Metzler, 2009, on Sub-Saharan Africa; Harper,
2011, on Southeast Asia; see also Fuller, 1991).
4.1 Contextual differences can fundamentally
Besides spiders, a different mode of organisational design
shape education systems (and the possibilities
and school system expansion is the starfish. Unlike
for reforming them): the example of spider and
centralised spiders, starfish systems comprise loosely
starfish systems
connected parts that have a great deal of autonomy
To show the extent to which differences in the (Brafman & Beckstrom, 2006; Pritchett, 2013). School
61 For example, Indonesia has a hybrid spider-starfish system that is aspiring-to-be-but-not-yet coherent for systemwide learning.
62 Some highly localised schools may operate as standalone entities that are not enmeshed in a wider education system, such that the
spider/starfish question is irrelevant because all management decisions occur at the school level. These were common in the premodern
periods before the advent of state education bureaucracies, as with one-room schoolhouses in the rural U.S. (Tyack, 1974) or the pondok,
pesantren, and madrasah schools of pre-colonial Southeast Asia (Harper, 2011), and continues to be true of private schools in some
contexts. However, our focus in this paper is on large-scale school systems that affect students and teachers at scale.
systems that expand starfish-style are often driven by in Section 2.1 that norms, i.e., dominant perceptions and
the goal of expanding a particular instructional model, beliefs about the acceptable and/or ideal teacher practice,
rather than expanding enrolment for enrolment’s sake. can strongly shape the effects of teacher-related policy.
Sometimes this takes the form of a ‘bring more children
On a more operational note, one clear difference between
into the selective elite system’ model of expansion, as with
spider and starfish systems is in who calls the shots. In
the growing proportion of children enrolled in Germany’s
a spider system, decisions about the number of teacher
pre-university gymnasium system (Phillips, 2000).
vacancies, allocation to schools and districts, metrics
Alternatively, another type of starfish-style expansion is
for performance, and eligibility for promotion would be
evident in Colombia’s Escuela Nueva, which was designed
determined centrally. But all of these decisions could
to serve underprivileged children, but also expanded
just as easily be distributed to different administrative
starfish-style at varying scales in 14 countries, with a
levels in a starfish system. As one of us proposed in
consistent emphasis peer learning among teachers and
Pritchett and Pande (2006; see also Pritchett, 2013, Table
responsiveness to local communities (Colbert & Arboleda,
6-6), approval of a teacher as eligible for hiring could be
2016; Le, 2018).
determined at a higher administrative level (e.g., the state
We draw attention to this bifurcation between spiders and or district), whereas allocation to a specific school could
starfish because complex systems have feedback loops be determined by the school itself. Such a separation
and path dependencies, such that any attempted change between approval and allocation has dual benefits.
is more feasible and sustainable if it is compatible with the First, it gives discretion to the administrative level best
general pattern of pre-existing relationships, structures, positioned to make a good decision (where the higher
and processes within the system. This distinction between administrative level would have more technical capability
spiders and starfish is far from novel. It has been proposed for evaluating teacher eligibility, and the school would
time and again across multiple academic disciplines have thicker knowledge of the teachers that would best
drawing on a wide range of empirical sources (see Table serve its students). Second, in contexts where patronage
1 in Honig & Pritchett, 2019, for a partial summary). To is common, it makes it more difficult for administrators to
give one example, Rowan’s (1990) review of research extract bribes from jobseekers, since granting approval for
on the organisational design of U.S. schools found that hiring does not guarantee allocation to an actual teaching
school improvement interventions typically followed post.
either a spider-style control strategy that ‘involves the
This points to an important trade-off: to the extent that
development of an elaborate system of input, behaviour,
we believe that teaching should be treated as nonroutine
and output controls designed to regulate classroom
because educational goals and classroom interactions
teaching and standardise student opportunities for
should be complex (and, as we have shown thus far,
learning, and the expected result is an increase in student
teachers’ motivations, teacher career structures, and the
achievement’ (p. 354) or a starfish-style commitment
education bureaucracies in which they are embedded
strategy that ‘rejects bureaucratic controls as a mode
are complex), then the most desirable pathway forward
of school improvement and instead seeks to develop
is commitment-based and starfish-oriented. But to the
innovative working arrangements that support teachers’
extent that most low-performing education systems are
decision-making and increase teachers’ engagement in the
spider systems that are coherent for the expansion of
tasks of teaching’ (p. 354). The review finds that both types
enrolment, then the most feasible pathway forward is
of strategies can support school improvement, but each
control-based and spider-oriented.
strategy tends to be most effective when it is implemented
intensively, rather than piecemeal. This trade-off between desirability and feasibility in
strategies for reforming teacher career systems is
In the long run, starfish systems may fit more heightened by the path dependence mentioned above.
naturally with purpose-driven teacher careers … That is, taking steps down one pathway might make it
harder to switch to the other pathway down the road,
Rowan (1990) further observes that control (spider-
because the people and structures within the system will
oriented) strategies tend to work in settings where
be increasingly oriented toward the original pathway, and
teachers regard their work as a set of routine practices, and
there may be inertia to fundamental changes in direction.
commitment (starfish-oriented) strategies tend to work in
For example, a study of the influence of social norms
settings where teachers regard their work as complex and
on district-level educational innovation in Indonesia
nonroutine. This observation strengthens the argument
described an intervention that deployed security guards
to every school in Gowa, a district with historically low are at least two aspects of teacher career reforms that are
levels of trust between citizens and the local government. common to across education systems that are coherent
This intervention succeeded in its aim of reducing teacher for systemwide learning: ‘thick’ information systems
absenteeism by tasking the security guards to monitor that are coherent with delegation; and compensation
teacher attendance and equipping them with motorcycles structures that are based on what will best attract, retain,
to pick up tardy or unjustifiably absent teachers (Nihayah, and motivate capable and committed teachers in the given
Revina, & Usman, 2020). However, this intervention seems context.
likely to reinforce the low levels of trust and the perception
We illustrate how each of these may manifest differently
that ‘it is best to show submission to government
in spiders and starfish using examples from the highly
programmes, as they do not want to get themselves in
successful teacher career structures in Singapore’s
trouble’ (ibid, p. 18)—both of which may further entrench
spider-y education system and Finland’s starfish-y one.
top-down spider-style governance. In the words of
These examples draw on prior research by one of us on
political theorist Jane Mansbridge (2014), ‘sanction-based
teacher accountability in Finland and Singapore (Hwa,
accountability not only stems from distrust; it creates
2019, 2021). For more on teacher career structures in
distrust’ (p. 55).
these countries, see Dimmock & Tan (2013) and Lee et al.
It is often true that the perfect is the enemy of the good, (2008) on Singapore, and Simola et al. (2009) and Sahlberg
in that pursuing overly demanding starfish-style reforms (2021) on Finland. Crucially, despite their deep contextual
in an under-resourced spider system may consume a great differences, both Finland’s and Singapore’s education
deal of effort with few returns; yet in the case of teacher bureaucracies are strongly oriented toward, and coherent
career reforms, the good may also turn out to be the for, the goal of cultivating student learning.
enemy of the perfect, in that pursuing relatively attainable
spider-style reforms in an already spider-oriented system
In purpose-driven teacher career
may make it increasingly hard to switch subsequently to a structures, decision-makers invest in—
starfish mode of governance. and act on—‘thick’ information about
… yet the 5Cs can be applied equally teachers’ capability and commitment
in starfish and spider systems High-quality information systems that are coherent with
student learning and with other systemwide goals are
Notwithstanding the respective merits and challenges
crucial for establishing and safeguarding professional
of spider and starfish systems, it is crucial to note that
standards (as noted in Section 3.3), and for aligning the
the 5Cs can be applied equally in a spider system or in a
numerous and cross-cutting actors and relationships in
starfish one. For example, Rowan (1990) observes that
an education system for shared purposes of cultivating
teacher career ladders in control-oriented (spider) systems
student learning—and of supporting capable and
are often justified on the basis that they facilitate close
committed teachers in their cultivation of student learning.
evaluation of teacher performance and close links between
Information also helps to orient the different sources
performance and pay; whereas the rationale for similar
of individual teacher motivation toward shared goals,
career ladders in commitment-oriented (starfish) systems
because performance-related information received by
is creating channels for teachers to mentor and support
headteachers and administrators may affect the pecuniary
one another outside of the typical school management
returns (finances and material circumstances) to teachers’
structures.
effort, and performance-related information received
In the next section, we illustrate this point further by by teachers and members of the wider community may
giving examples of how the same aspects of teacher affect teachers’ personal satisfaction with their work and
career structures have been instantiated differently—and the social validation they receive for doing it. Additionally,
effectively—in Singapore’s (spider) and Finland’s (starfish) relevant, regular, reliable, and rich information on teacher
approaches to teacher careers. practice also helps teachers to develop their craft. Put
differently, relevant information flows are indispensable for
4.2 Aspects of purpose-drive teacher career maintaining systemwide coherence with delegation.
structures that are fundamental regardless of
Effective and efficient ways of developing such information
context
flows will vary depending on the context of the education
Important contextual differences notwithstanding, there
PREMISE FOR PRACTICE #4: on information and support for teaching quality
Rather than relying on EMIS indicators, years of service, and formal certifications as the main (or only) sources of
information about teaching quality …
Education authorities and organisations should invest in building multi-component ‘thick’ information
systems about teaching quality and in supporting teachers to continually improve their pedagogical
competencies.
Education bureaucracies are distinct, complex organisations with a distinct purpose (i.e., cultivating student
learning) and specific technical practices in service of that purpose (e.g., pedagogical expertise appropriate to the
context in question). Rather than treating teachers as interchangeable civil servants who are generically involved
in public service delivery, education bureaucracies need to attract, retain, equip, and motivate teachers who are
capable of, and committed to, cultivating student learning.
To develop such teachers, it is crucial to identify quality teaching that facilitates learning, and to comprehensively
support teachers to continuously refine their pedagogical capabilities. In addition to high-quality support in the
form of pre- and in-service training, mentoring, well-developed teaching and learning materials, education systems
also need multi-component, ‘thick’ information about those aspects of teacher practice that are most important
for realising shared systemwide learning goals. Such information systems have the twofold purpose of allowing
education authorities and organisations to identify teachers (for entry into the profession, for permanent positions,
for additional targeted support to improve their classroom practice, etc.)—and of enabling teachers themselves to
iteratively improve their practice in response to feedback.
→ See also
• ‘Design elements of teacher careers’ in Section 2.1
• ‘Resources on teacher educators and mid-tier education bureaucrats' in Box 2
• ‘Changes that are incoherent with the complexity of classroom teaching' in Section 2.2
• ‘Research on how education authorities and organisations should choose teachers during the pre-service and
novice phases’ in Box 4
• ‘In purpose-driven teacher career structures, decision-makers invest in—and act on—“thick” information about
teachers’ capability and commitment’ in Section 4.2
system.63 In Singapore’s purpose-driven spider system, individual teachers at within-school appraisal panels, with
teacher appraisal is a standardised, annual, compulsory some additional benchmarking across schools provided
process for every teacher (and other education officers by occasional participation in appraisal panels by ministry
in the civil service) under the Enhanced Performance officials. Annual performance grades, which are collated
Management System (EPMS). EPMS appraisals for in a centralised ministry database, affect teachers’ speed
individual teachers are carried out by reporting officers of promotion up the career ladder, eligibility for paid study
(i.e., other teachers holding management-level positions), leave, and annual bonuses, which can exceed 3 months’
who then collectively triangulate their appraisals and rank salary for top performers.
63 Variation notwithstanding, there are probably advantages to having some amount of student assessment benchmarking conducted
by the central state. This is because having some genuinely systemwide standards can facilitate coherence, and also because designing
reliable, valid, and curriculum-aligned assessments is a technically demanding process requiring concentrations of expertise that are
easiest to accumulate at the central ministry level (see Gustafsson, 2019).
Note that this centralisation and standardisation in communicated by students to their parents, who would in
Singapore does not mean thin information. Reporting turn speak to school leaders.
officers are required to observe at least one classroom
In short, ‘thick’, purpose-aligned information is crucial—
lesson per appraised teacher annually, while also
whatever forms it may take in the education system in
checking a selection of students’ coursework, and holding
question.
three review discussions with each appraised teacher
at different points in the school year. In fact, in 2014, the In purpose-driven teacher career structures,
EPMS appraisal instruments moved away from a ‘thinner’ ‘fairness’ in compensation is based on shared goals
format, in which multiple competencies were each
(rather than administrative defaults)
assessed on a 4-point scale based on specific behavioural
indicators, toward a thicker appraisal instrument with In most education bureaucracies, whether in developed or
fewer categories that each required qualitative judgements developing countries, teacher compensation is determined
that were summarised in a performance grade (Kan, 2014; primarily by factors that have little to do with student
Low & Tan, 2017). Collecting and communicating such learning—and everything to do with conventional, civil-
thick information on teacher practice is highly resource- service-wide administrative processes. As Bruns, Filmer,
intensive—but it successfully accommodates the and Patrinos (2011) observe:
complexity of teaching and education within a centralised Almost universally, teacher recruitment and
spider system, thus sidestepping the pitfalls of thin-output, promotion are based on the number of years of
deliverology-style spider systems. preservice training, formal certificates, and years
Finland’s purpose-driven starfish system also has clear in service. Yet an extensive body of research has
flows of information about student learning and teacher documented the lack of correlation between
performance. However, rather than being centralised, the these ‘observable’ factors and teachers’ actual
type of information collected and the purposes for which effectiveness in the classroom … The clear
it is used are distributed—not only across different actors implication of available research is that most school
in the education system, but also across the phases of systems are recruiting and rewarding teachers for
teacher careers. For the individual teacher, those in the the wrong things, failing to encourage the capacities
pre-service phase are appraised the most intensively. To and behaviors that contribute most directly to
choose candidates for entry into the profession, admission student learning results, and unable to sanction
screenings for pre-service training involve a first-stage ineffective performance (pp. 143–143).
national exam based on a pre-determined and annually In other words, fairness in teacher compensation
varying selection of academic articles related to education, systems is conventionally defined based on seniority,
followed by second-stage interviews and other assessment whether seniority in years of service or seniority in formal
tasks designed and administered by individual universities qualifications. Such approaches provide an illusion of
(Malinen et al., 2012). Collectively, these admissions equality, in that teachers with the same standardised,
screening processes filter prospective pre-service teachers observable characteristics receive the same rewards.
for early indications of both capability and of commitment. This is the default in education bureaucracies that either
The pre-service phase also includes an extended teaching lack any clear, consensus-based purpose—or that have a
practicum, during which pre-service teachers are regularly profusion of overambitious and conflicting goals. In the
observed and coached by experienced teachers. absence of consensus about a core purpose, an education
After this heavily front-loaded appraisal (and training and bureaucracy would also lack a shared orientation toward
socialisation) of pre-service teachers, Finland’s education performance, since there would be no shared basis
system collects little formal information about teacher for judging what sort of performance matters. Hence
performance on the individual level. Instead, teacher the default toward generic, seniority or points-based
and school performance at a collective level is routinely characteristics as the definition of ‘fairness’ in teacher
benchmarked through sample-based student assessments compensation.
of national education quality (and, indirectly, though Yet, as argued in Section 3.4, such observable
the matriculation exam for university entry). Individual characteristics are tenuously—if at all—linked to teacher
teacher practice is informally monitored by headteachers capability and commitment. Consequently, such seniority-
and colleagues; with a baseline expectation that any based compensation structures can be tremendously
serious aberrations from professional standards would be demotivating for purpose-driven teachers who exert
considerable effort into cultivating student learning, productivity in Ghana, the least individualistic country of
yet they are not offered meaningful opportunities for the three (Bandiera et al., 2020).
professional advancement—and, instead, they see their
In some school systems, teacher performance pay may
less-motivated colleagues gaining identical, or even
be the way to go. This applies if such a compensation
superior, rewards (Crehan, 2016). Defining fairness based
structure would be regarded as fair and desirable by
on seniority can also damage teacher professional norms,
the majority of capable and committed teachers in the
as noted in Section 2.1.
system—such that it would bolster their motivation to
Instead, education bureaucracies that genuinely prioritise cultivate student learning. Returning to the Singapore
the cultivation of student learning and, by extension, that example, one Singaporean teacher, despite describing
prioritise capable and committed teachers should define himself as a ‘happy ordinary teacher’ who prioritises
fairness in teacher compensation based on what will best classroom teaching rather than pursuing the broader range
serve this core priority. A teacher compensation structure of accomplishments that are rewarded within the EPMS
should be deemed fair if (and only if) it advances the goal appraisal and career advancement system, nevertheless
of attracting, retaining, and motivating teachers who make remarked that: ‘We do recognise those who are deserving
the best possible contributions to student learning, in the of credit because [...] something about them enables them
given context. to go above and beyond for the students, and we don’t
begrudge them if they are rewarded accordingly’ (Hwa,
Note that this does not necessarily mean instituting
2021, p. 230). In other words, he regarded the EPMS as
a performance pay system based on standardised
fair, despite his lack of personal interest in climbing its
measurements of either inputs or learning outputs.64 This
incentive ladder. Similarly, another teacher in the same
is because implementation contexts are fundamental
study said of the EPMS that:
to fairness in compensation structures—just as they are
fundamental to the design of information systems, and When it’s used correctly, it motivates me a lot,
to the application of the 5Cs and education reform more regardless of whether I just get a normal C grade or
generally.65 not. But when you don’t feel it’s fair, [...] then you
feel trapped. You feel that whatever effort you put
To illustrate this contextual diversity of beliefs about
in is not worth it (Hwa, 2019, p. 174).
fairness in compensation, a survey of 20,079 teachers
in nine developing countries found wide variation in the Elsewhere, this teacher affirmed that she thinks the EPMS
proportion of teachers who agreed that they should is usually applied fairly, within acceptable limits of human
receive financial bonuses if their students did well in bias from ‘thick’ and, inadvertently, subjective judgements
official exams—ranging from less than 20 percent of from managers—a view echoed by most Singaporean
teachers in Argentina and Indonesia to more than 90 teachers in the study. All of this suggests that (a) fairness
percent of teachers in Pakistan and Tanzania (Sabarwal et in teacher compensation is paramount to motivation, (b) a
al., 2021). Such differences in perceptions likely have real- sufficient threshold of capable and committed Singaporean
world implications for teachers’ motivation and wellbeing, teachers regard the performance-based EPMS appraisal
as well as the quality of their work. To draw on an example and compensation structures as sufficiently fair to attract,
from a different domain, a randomised-control trial of flat retain, and motivate them.
wage rates versus piece rates for data entry in Ghana, the
In contrast, the same study found that teachers in Finland
Philippines, and India found that individual piece rates
were, almost without exception, opposed to the idea of
raised worker productivity by 20 percent relative to the
performance-based pay. For example, when asked how
flat wage control group in India, the most individualistic
teachers would respond to a performance-based system
country in the sample on Hofstede’s individualism–
like the EPMS, one Finnish teacher responded:
collectivism scale, but it had no significant effect on
64 Also, it emphatically does not mean instituting a performance pay system focused narrowly on ‘thin’ measures of outputs—see
Section 2.2 on changes that are incoherent with the complexity of classroom teaching, Section 3.4 on why judgements of teacher quality
should be practice-based and school-contextualised; and the previous subsection on the importance of ‘thick’ information under the
5Cs.
65 For a selection of recent experimental studies exploring variation in teachers’ preferences for performance pay in low- and middle-
income countries, see Leaver et al. (2021) on Rwanda, Brown & Andrabi (forthcoming) on Pakistan, Mbiti & Schipper (2021) on Tanzania,
and Muralidharan & Sundararaman (2011b) on India.
We are so independent here. And we like that said that teachers ‘should get paid equally’ because they
independence in our classrooms so much, that all had attained the same high calibre of professional
even that bonus would not make this system a good training, so ‘the baseline is that every teacher is as
thing. [...] And we are so equal, among teachers. competent as everybody else’—because she regarded this
[...] We do not want to give others the possibility shared standard of pre-service training as the wellspring of
of rushing higher. [...] We have done some studies teacher capability and commitment. In short, these Finnish
about rewarding people with money, and it gives teachers would consider performance pay to be neither
satisfaction for shorter period of time than when fair nor motivation-enhancing given the dominant norms
you are valued by the society you work in (Hwa, of their profession—which contrast distinctly with the
2021, p. 234). dominant social, professional, and organisational norms
shaping Singaporean teachers’ conceptions of fairness in
Besides concerns about autonomy and egalitarianism,
compensation. (For more on norms, see Section 2.1 on
another teacher posited that performance pay would
design elements of teacher careers and Section 3.3 on how
erode collegial collaboration. Yet another Finnish teacher
PREMISE FOR PRACTICE #5: on purpose-driven definitions of fairness

Rather than defining fairness in teacher compensation almost exclusively in terms of seniority, perhaps with some
consideration of formal qualifications and official responsibilities …
Fairness in teacher compensation should be defined based on what, in the specific context, will attract, retain,
and motivate capable and committed teachers who make the best possible contributions to student learning.
In line with widespread recognition that ‘context matters’, teacher career and compensation structures need to be
tailored to the education systems in which they operate. This may entail a radical departure from the conventional
single salary structures that are common across public-sector teacher careers and across public-sector jobs more
generally.
A fundamental element of any effective education system is the core purpose that animates and aligns the actions of
all stakeholders contributing to its performance. This core purpose—together with the professional, organisational,
and social norms and shared beliefs about the teaching profession—should be the primary basis for defining what
constitutes ‘fairness’ in teacher compensation.
In contrast, defining fairness in teacher compensation primarily on years of service or other measures of seniority
would suggest that teaching is being treated like a generic civil service occupation; rather than as an empowered,
highly respected, strongly performance-normed, contextually embedded profession with the overarching, shared
purpose of cultivating student learning.
→ See also
• ‘Design elements of teacher careers’ in Section 2.1
• ‘Resources on teacher educators and mid-tier education bureaucrats’ in Box 2
• ‘Changes that are incoherent with the complexity of classroom teaching’ in Section 2.2
• ‘Research on how education authorities and organisations should choose teachers during the pre-service and
novice phases’ in Box 4
• ‘In purpose-driven teacher career structures, decision-makers invest in—and act on—“thick” information about
teachers’ capability and commitment’ in Section 4.2
the novice teacher phase can shape careerlong norms). look very different depending on the particularities of each
implementation context.
To summarise, (a) fairness in teacher compensation
can be crucial to attracting, retaining, and motivating But another important source of variation is that sensible,
teachers; (b) fairness in teacher compensation should feasible, forward-looking steps along the journey of
be defined in terms of the shared goals (i.e., the specific teacher career reform may look very different from the
articulations and aspects of student learning and, in policies in education systems that have already reached
turn, of teacher capability and commitment) for which the ‘destination’ of empowered, purpose-driven teaching
the education bureaucracy wants to attract, retain, and professions.
motivate teachers; and (c) these conceptions of fairness—
For example, when Zimbabwe rapidly expanded the
and, consequently, the design of teacher compensation
provision of education in the years immediately following
structures—will vary across contexts because of variation
its independence in 1980—with primary school enrolment
in the social, professional, and organisational norms that
increasing by a factor of 2.76 in the seven years following
influence teachers and their career structures.
independence (Dorsey, 1989)—it needed a similarly rapid
A final note: one sensible way of finding out which career expansion in the number of capable teachers. To do so, it
and compensation structures would be regard as fair and successively introduced two new approaches for training
motivating in a given context would be to ask the teachers primary school teachers.66 First, in 1981, the government
themselves. Surveys of teachers in the U.S., the U.K., and launched a predominantly field-based teacher qualification
Australia tend to find that most teachers oppose the programme in which new teachers spent one term in
idea of performance-based pay (Leigh, 2013). However, a an intensive residential course, followed by 3.5 years of
majority of teachers in separate surveys in India, Pakistan, teaching in a primary school while participating in teacher
Rwanda, and Tanzania said that they would be in favour education via distance learning, followed by another
of bonus payments based on objective measures of term in an intensive residential course (Chivore, 1986).
student learning—and performance-based pay schemes Subsequently, the ministry introduced a different teacher
implemented for each of these samples of teachers did, training approach that allowed for more extensive initial
in fact, lead to student learning gains (Muralidharan teacher education, where new teachers spent their first
& Sundararaman, 2011a, 2011b, on India; C. Brown & year in residential teacher training, followed by a second
Andrabi, forthcoming, on Pakistan; Leaver et al., 2021, on year in schools as classroom teachers while participating
Rwanda; and Mbiti et al., 2019, and Mbiti & Schipper, 2021, in distance learning, followed by a third year of full-time
on Tanzania; see also Sabarwal et al., 2021, as described residential training, and a fourth year of teaching with
above). in-service distance learning (Chivore, 1986; Dorsey, 1989;
Nziramasanga, 1991). While these two approaches may not
be ideal teacher education models that we would expect
4.3 The journey may not look like the to see in education systems that have already reached the
destination: examples of feasible ways destination of an empowered, purpose-driven teaching
forward profession, they were valuable steps along the journey
As discussed in the introduction, we are proposing the of improving educational quality in Zimbabwe, at a time
5Cs—choose and curate toward commitment to capable when there were not only massive shortages of trained
and committed teachers—as a set of principles for teachers but also inadequate teacher training facilities to
navigating away from the present reality of disempowered, concurrently accommodate the necessary number of pre-
low-performing teaching professions in many low- and service teachers in residential programmes (Chivore, 1986).
middle-income countries, and toward the destination In this section, we give examples of how 5Cs-aligned
of empowered, highly respected, strongly performance- approaches to teacher careers may look very different
normed, contextually embedded teaching careers that depending on where an education system is in the
cultivate student learning. So far, we have given examples journey toward empowered, highly respected, strongly
of how 5Cs-aligned approaches to teacher careers may performance-normed, contextually embedded teaching
66 For a firsthand account from Dzingai Mutumbuka, Zimbabwe’s inaugural minister of education, a podcast interview hosted by Marla
Spivack is available at https://riseprogramme.org/podcast/dzingai-mutumbuka.
careers that cultivate student learning. as noted in Section 4.2, there is a well-established system
for choosing candidates for entry into pre-service teacher
Examples of the 5Cs at the ‘destination’ of training, based on their potential for both capability and
empowered, purpose-driven teaching professions commitment in the profession.
In both Singapore and Finland, teacher career policy Second, because of the high quality and state funding
reflects some aspects of the 5Cs while concurrently of higher education, the costs of curation at any point
departing from some of other aspects in ways that suit in the teacher career cycle are relatively low, whether
systemic needs and orientations. for teachers, who can typically pursue state-subsidised
For example, in Singapore’s spider-style system, novice retraining in another field, and for the municipalities and
teachers are civil servants upon their entry into the career schools who hire them, because exiting teachers can
progression. This may appear to be a departure from typically be replaced with entrants of comparable calibre.
premise for practice #3, which emphasises the importance So even without compensation and hiring policies that
of curation rather than de facto awarding permanent jobs generate a concentrated period of curation during the pre-
from day 1. Nonetheless, the teacher career system does service and novice phases, and even without structured
incorporate an abbreviated period of curation, in that careerlong monitoring of individual teacher practice,
a condition for entry into the postgraduate diploma in the wider education system encourages voluntary exit
education at the National Institute of Education (i.e., the of teachers who, at any point in their careers, lose their
main entry route in the country’s sole pre-service training commitment to the teaching profession.68
institution) is a ‘compulsory untrained teaching stint’ in Third, Finnish teachers who are willing and able can take
a regular public school, colloquially known as ‘contract on administrative responsibilities within their schools,
teaching’. This stint lasts several months and takes place which for which they receive separate salary supplements,
in between applying for admission and enrolling in pre- some of which are part of the union-negotiated
service training. The education ministry’s website states national salary structure and some of which are at each
the contract teaching stint ‘allows you to affirm your municipality’s discretion. Some teachers also choose to
interest in teaching, and lets us assess your suitability for reduce their teaching hours to take on such administrative
the profession’67—thus serving the same screening function responsibilities. Teachers can also opt into school
for both commitment and capability as the high-turnover principalship, which requires teaching experience and
novice phase in the hypothetical instantiation of the 5Cs additional certification. So even without the structured
described in Section 3.1. Once they enter the profession bundling of career pathways in the example of the 5Cs
as novices, teachers are subject to the centralised EPMS discussed in Section 3.1 or in Singapore-style teacher
appraisals described above, which specify performance career ladders, there is some room for self-selected
criteria depending on teachers’ career tracks (teaching, variation in teachers’ responsibilities based on preferences
leadership, or senior specialist) and their position on the and experience. This room for variation facilitates better
career ladder—thus representing a form of curation into matching between teachers’ specific capabilities and
different roles within the school system, as a complement commitments and systemwide needs.69
to the pre-service curation into/out of the teaching
profession. Examples of the 5Cs on the ‘journey’ toward
empowered, purpose-driven teaching professions
In contrast, teaching positions in Finland’s starfish system
comprise a mix of permanent tenure and fixed-term While Singapore’s and Finland’s present-day teacher career
contracts, determined variously by its more than 300 structures may represent context-specific instantiations of
municipalities. More importantly, three features of teaching the 5Cs in learning-coherent spider and starfish systems,
in Finland are worth noting in comparison to the 5Cs. First, these well-established and richly resourced systems
67 https://www.moe.gov.sg/careers/become-teachers/pri-sec-jc-ci/postgraduate-diploma, accessed on 24 March 2021.
68 To give an anectdotal example, one Finnish interview participant, who had retrained as a lower secondary school teacher following
a previous career in a large corporation, told one of us that: ‘I promised myself that when I’m tired of doing what I’m doing, then I will
quit. I will never become that teacher who’s just doing their job and not being interested in it .’ (Hwa, 2019, p.205)
69 As noted in Section 3.2, Besley and Ghatak (2005) argue that ensuring the closest possible match between the organisational
mission and the mission preferences of individual agents is optimal for the productivity of organisations that serve publicly beneficial,
collective goals.
may not be the most helpful lodestars for the typical goodwill gestures from the ministry such as bonuses for in-
developing-country education system. Instead, we now service teachers who opted into the career ladder (which
offer examples from education systems that have made was compulsory for new teachers), gradual increases in
progress along the journey from lower performance the salary scale, and support for teacher professional
toward empowered, purpose-driven teaching professions. development (Salazar-Morales, 2018).
Specifically, we highlight aspects of teacher-related policy
reform in Peru and Brazil. For each example, we also note a A point of caution to note from Peru’s experience of
point of caution for navigating the reform journey. quality-oriented reforms in teacher careers is the politics
of reform can be perilous. Although reformers successfully
Streamlining and refining teacher selection instituted the teacher career reforms described above,
and assignment processes in Peru recent years have seen substantial political resistance,
with a months-long strike in 2017 by teachers who, among
Peru’s teacher career system has been undergoing a
other goals, sought the repeal of the appraisal process that
series of reforms since 2006, coincident with broader
can lead to dismissal for underperformance (Chanduví
education reforms that have included a recentralisation
Jaña, 2017). In June 2021, the leader of these 2017 strikes
of the education bureaucracy and a new national student
was elected president of the country, suggesting that there
assessment (Saavedra & Gutierrez, 2020; Salazar-Morales,
is substantial popular support for this resistance to the
2018), indicative of a spider system that is introducing
teacher career reforms (Muñoz, 2021).
performance pressure. Throughout this period, Peru’s
average PISA scores have been on an upward trajectory,70 Alignment of teacher career structures
suggesting increasing coherence for systemwide student
toward local priorities in Sobral, Brazil
learning.
Besides the centrally driven, spider-style reforms that
Prior to these reforms, teacher policy in Peru was
Peru has used to gradually increase the coherence of
fragmented, with little trust between teachers and the
teacher career structures (and of the education system
ministry (Salazar-Morales, 2018). Major reforms to teacher
more generally) for systemwide student learning, another
career structures have included a centralised accreditation
pathway for making teacher career structures and
programme for teacher education programmes
education systems increasingly coherent for systemwide
(improvements in the design element of support) as well
student learning is the starfish option. Disaggregating
as a teacher career ladder offering higher compensation
decision-making discretion within an education system
(delegation and finance). Additionally, the government has
can involve at least three aspects: (a) distributing
instituted a systematic appraisal process (information),
discretion across administrative levels to improve the
which includes a national-level assessment of teacher
quality of information and reduce conflicts of interest, as
content knowledge (Bruns & Luque, 2015; Tournier &
in the example in Section 4.1 of separating the authority
Chimier, 2019)—both of which could help align the system
to approve teachers for hire and the authority to assign
toward achieving quality in teaching practice and student
them to jobs in schools (Pritchett & Pande, 2006), which
learning, by improving the process of choosing teachers
focuses primarily on the design element of delegation;
and the development of teachers’ capability.
(b) empowering teachers and schools with the resources
The system for assigning teachers to schools is managed and the authority to continually refine their practice, as
centrally by the education ministry, in typical spider style, with teacher professional learning communities and
but the allocation process includes in-person evaluations other professional development networks, which focuses
in schools or in local education administrative units primarily on the design element of support; and (c)
as well as some consideration of candidates’ rankings decentralisation of teacher policy areas to lower levels of
of preferred schools (Ajzenman et al., 2020)—which administration to allow more positive deviance, or locally
facilitates better curation in matching the pool of teachers coherent improvements, to emerge (Andrews, Pritchett, &
to system needs. Available data suggests that there have Woolcock, 2017). In this example that follows, we focus on
been improvements in teaching quality (Bruns & Luque, the third aspect.
2015), as well as growing trust and mutual commitment
Like Peru, Brazil has also seen some improvements in its
between teachers and the ministry, especially with
70 See https://www.oecd.org/pisa/publications/PISA2018_CN_PER.pdf, accessed 24 March 2021.

PISA scores over time, especially in mathematics, and influence norms) for good performance (Crouch, 2020). In
especially in the 2000s.71 Like Peru, it also introduced a other words, the changes to teacher policy helped to orient
common national assessment of student learning during local teachers to be committed to municipal educational
this period (Bruns, Evans, & Luque, 2012). Unlike Peru, priorities. This tightly coherent package of changes would
other reforms in Brazil did not tend toward a centralisation not have been possible in a spider-style system that
of control, but rather toward empowerment of its retained central control over these elements of teaching
municipalities, which hold considerable authority over and learning. Instead, Brazil’s starfish system empowered
early childhood and primary education (IBE-UNESCO, 2012). Sobral’s education authorities to design a set of policies
These starfish-oriented reforms included a guaranteed per that advanced progress toward systemwide student
student capitation grant, i.e., a ‘money follows the student’ learning, such that Sobral’s basic education system
policy (Bruns, Evans, & Luque, 2012), which reinforced the outperformed all of Brazil’s other municipalities (of which
performance pressure by incentivising municipalities there are more than 5,000) in the 2017 national assessment.
to expand education provision and quality; as well as In 2005, prior to the reforms, Sobral ranked 1,336th (ibid;
increased flexibility in how the funding was used (Pritchett, drawing on Becskehazy, 2018).
2013). This combination of flexibility and resourcing has
Here’s the cautionary note: Sobral is a positive outlier
allowed for the emergence of state- and municipality-
relative to other Brazilian municipalities not only in
specific teacher career reforms (see Bruns & Luque, 2015,
learning outcomes, but also in successfully tailoring
for some examples).
teacher policy to municipal purposes.72 While the inherent
Notably, in the 2000s, the municipality of Sobral flexibility of starfish systems can be an advantage in
introduced an ‘alphabetisation at the right age’ reform to accommodating subnational variation in educational
boost foundational literacy levels that included several priorities (and also in accommodating the complexities
teacher policy changes that were tightly coherent with the of the teaching profession, as noted in Section 4.1), this
literacy reform. These teacher policy changes included same flexibility is a vulnerability in that it may be harder to
well-aligned instructional materials and professional establish a systemwide baseline of quality or to orient the
development based on these materials (i.e., improvements whole system toward a shared purpose.
in the design element of support) as well as significant
bonuses (finance) and public recognition events (which can
71 https://www.oecd.org/pisa/publications/PISA2018_CN_BRA.pdf, accessed on 24 March 2021.
72 Email correspondence with Barbara Bruns, 24 April 2021.

Part 5. Conclusion: Where do you want to go?
Everything should be made as simple as possible, but not simpler.

Everything should be made as simple as possible, but not
—attributed to Albert Einstein73
simpler.
—attributed to Albert Einstein73
The driving question in this primer, as noted in Section 3.6, in their particular contexts by people who have detailed,
has been: how can education authorities and organisations thick knowledge of the day-to-day educational realities of
cost-effectively develop empowered, highly respected, that context.
strongly performance-normed, contextually embedded The 5Cs approach is admittedly and deliberately complex.
teaching professionals who cultivate student learning? While Einstein promoted simplicity and parsimony as
Our answer, in summary, is that they should apply the goals of scientific endeavour, it is worth noting that his
principles of choose and curate toward commitment to theory of general relativity as expressed in the Einstein
capable and committed teachers, or the 5Cs, adapting field equations is (Wikipedia): ‘a system of ten coupled,
these principles to the needs of their education system nonlinear, hyperbolic-elliptic partial differential equations’
and the shared goals that are articulated and affirmed by that cannot be solved exactly even in the simplest cases.74
people within the system. ‘Simple’ approaches to improving teacher performance are
While we have made a case for the 5Cs, we have also left likely to be wrong.
many other hard questions unanswered. Such as: how can Rather than answering these programmatic questions,
the school-level goal of local control over teacher selection the function of the 5Cs (and the goal of this primer) is to
be reconciled with the system-level goal of making a encapsulate some key principles for realising a vision of
permanent commitment to experienced and veteran empowered, highly respected, strongly performance-
teachers? How should you sequence a policy rollout, normed, contextually embedded teaching professions
if a teacher career pathway reform has very different that cultivate student learning. Alongside the principles
implications for incoming cohorts compared to teachers of choose and curate toward commitment to capable
who are already in the system? What should you do if and committed teachers, we have offered five premises
politicians won’t agree to reroute resources away from for practice as a starting point for identifying areas in
school computer labs and toward high-quality pre-service which education authorities and organisations should act
training, or if teacher professional norms are so eroded differently to reform teacher careers in line with the 5Cs. As
that no one wants to enter the profession at all? Yet a key shown in Figure 10 on the following page, each premise for
lesson of the systems thinking approach of this primer is practice is derived from different principles within the 5Cs.
that context matters. These questions, while important
and immensely difficult to address, need to be answered
73 According to the Oxford Dictionary of Quotations (Knowles, 2014), this is the commonly quoted version of a somewhat less snappy
statement documented in a 1933 lecture by Einstein: ‘It can scarcely be denied that the supreme goal of all theory is to make the
irreducible basic elements as simple and as few as possible without having to surrender the adequate representation of a single datum
of experience.’ Arguably, using the snappier version of the quote honours the spirit of the original one.
74 From the Wikipedia article on ‘Solutions of the Einstein field equations,’ accessed 19 Oct 2021: ‘Once equations of state are chosen
and the gauge is fixed, the complete set of equations can be solved for. Unfortunately, even in the simplest case of gravitational field
in the vacuum (vanishing stress–energy tensor), the problem turns out too complex to be exactly solvable. To get physical results,
we can either turn to numerical methods; try to find exact solutions by imposing symmetries; or try middle-ground approaches such
as perturbation methods or linear approximations of the Einstein tensor.’
Figure 10. There are five premises for practice about what education authorities and organisations should do differently
when applying the 5Cs for teacher career reform
Choose and Curate toward Commitment to Capable and Committed teachers

TYPICAL APPROACH 5Cs APPROACH PREMISES FOR PRACTICE
Assume that the PREMISE FOR PRACTICE #1

Articulate and prioritise a core purpose that
education system Clear, consensus-based prioritisation of learning,
is oriented toward student learning so that
is independently delegated from education authorities
teachers can be capable of, and committed
committed to and organisations to schools and teachers,
to, fulfilling that core purpose collectively.
learning. is fundamental to teacher career reform.
Recognize that professional norms developed

during the novice phase—including the norms PREMISE FOR PRACTICE #2
Don't pay any that are implicit in processes for choosing, Contextually embedded, learning-oriented
particular attention curating, and committing to teachers—can teacher professional norms
to the novice phase. shape teachers' careerlong beliefs about what must be cultivated throughout
it means to be a capable and committed the novice teacher phase.
teacher.
Prior to a permanent employment PREMISE FOR PRACTICE #3

commitment, include curation in the pre-
Award permanent service and novice phases so that both The pre-service and novice phases should be
job tenure from employers and teachers have the opportunity a period of curation, such that,
day 1. to discern, through engagement in classroom as with nearly all other professions,
practice, whether they will be capable and a substantial proportion of initial entrants
committed to classroom teaching long-term. do not persist in the career.
Rely on EMIS Earmark resources for 'thick' information

indicators, years systems and extensive professional support PREMISE FOR PRACTICE #4
of service, and so that teachers have the best chances of Education authorities and organisations
formal certifications being capable and committed in the complex should invest in building multi-component
as the main (or work of classroom teaching, and education ‘thick’ information systems about teaching quality
only) sources of authorities and organisations can make the and in supporting teachers to continually
information about best possible choices about whom to employ improve their pedagogical competencies.
teaching quality. long-term.
Define fairness PREMISE FOR PRACTICE #5

in teacher Make decisions about compensation, as
compensation in with all other aspects of teacher careers, Fairness in teacher compensation should be defined
terms of seniority based on the goal of cultivating capable and based on what, in the specific context,
(and perhaps formal committed teachers who, in turn, cultivate will attract, retain, and motivate
qualifications and student learning. capable and committed teachers who make
responsibilities). the best possible contributions to student learning.
The 5Cs are not a silver bullet for the low-performance, other levels of the system. As illustrated in Figure 11, we
low-satisfaction, low-pay equilibria that afflict the teaching can think of the 5Cs as being embedded within three other
profession in many low- and middle-income countries. In Cs:
fact, an implication of the 5Cs is that silver bullets are
• the 5Cs are principles that operate at the level of
usually ineffectual against complex, embedded, internally
teacher career structures;
coherent equilibria. Rather, the way forward is through
purposeful, systemic, iterative, context-specific change • these principles are a subset of the conditions
that gives due attention to the complexities of teaching, created by education authorities and organisations
motivation, and organisational change. at the level of the teaching profession, in order to
facilitate autonomous, supported classroom practice
One way to bear in mind all of this complexity and
byempowered teachers;
embeddedness is to think of the 5Cs as being nested within
Figure 11. The 5Cs are a subset of the conditions for empowered teaching professionals within education systems that
are coherent for learning, and which are embedded in wider contexts
• these teachers are, in turn, one of several sets of education bureaucracy where decision-makers have
actors who play a vital role in building coherence for uninhibited control over the number of teachers in the
the purpose of student learning at the level of the system at any given point, then a contract teacher model
education system; andeducation systems themselves should get you there. But if your desired destination is
are embedded within, and needto be coherent with, an empowered, purpose-driven teaching profession
wider contexts.75 within an education system that is coherent for student
learning and consistently equips students with the array
We conclude with a question: where do you want to go?
of competencies they will need in their lives, then consider
If your desired destination is a teaching profession that is
adopting the 5Cs—choose and curate toward commitment
coherent with the existing administrative procedures of
to capable and committed teachers—as an approach for
the rest of the civil service, then the typical, commitment-
navigating toward this destination.
from-day-1 approach to teacher careers and compensation
should get you there. If your desired destination is an
75 We are grateful to Joan DeJaeghere for proposing this framing of the 5Cs embedded within teachers’ working conditions and
broader contexts
References
Aiyar, Y., & Bhattacharya, S. (2016). The Post Office Paradox. Economic and Political Weekly, 51(11), 61–69.
Aiyar, Y., Davis, V., Govindan, G. and Kapoor, T. 2021. Rewriting the Grammar of the Education System: Delhi’s Education
Reform (A Tale of Creative Resistance and Creative Disruption). Research on Improving Systems of Education (RISE). https://
doi.org/10.35489/BSG-RISE-Misc_2021/01
Ajzenman, N., Bertoni, E., Elacqua, G., Marotta, L., & Méndez Vargas, C. (2020). Altruism or Money? Reducing Teacher Sorting
Using Behavioral Strategies in Peru (No. 1143; IDB Working Paper Series). Inter-American Development Bank. https://doi.
org/10.18235/0002625
Akyeampong, K., Pryor, J., Westbrook, J., & Lussier, K. (2011). Teacher Preparation and Continuing Professional Development
in Africa: Learning To Teach Early Reading and Maths. University of Sussex Centre for International Education. https://www.
sussex.ac.uk/webteam/gateway/file.php?name=tpa-synthesis-report-july2011.pdf&site=320
Alifia, U., Pramana, R. P., & Revina, S. (forthcoming). Becoming a Teacher: How Teacher Policy Shapes Teacher Identity
Construction in Indonesia. RISE Working Paper Series.
Andrabi, T., and Brown, C. (forthcoming). Subjective versus Objective Incentives and Employee Productivity. RISE Working
Paper Series.
Andrabi, T., Das, J., & Khwaja, A. I. (2008). A Dime a Day: The Possibilities and Limits of Private Schooling in Pakistan.
Comparative Education Review, 52(3), 329–355. https://doi.org/10.1086/588796
Andrews, M. (2013). The Limits of Institutional Reform in Development: Changing Rules for Realistic Solutions. Cambridge
University Press.
Andrews, M., Pritchett, L., & Woolcock, M. (2017). Building state capability: Evidence, analysis, action. Oxford University Press.
http://www.oapen.org/search?identifier=624551
Araujo, M. C., Carneiro, P., Cruz-Aguayo, Y., & Schady, N. (2016). Teacher Quality and Learning Outcomes in Kindergarten. The
Quarterly Journal of Economics, 131(3), 1415–1453. https://doi.org/10.1093/qje/qjw016
Atherton, P., & Kingdon, G. (2010). The relative effectiveness and costs of contract and regular teachers in India. CSAE
Working Paper Series. https://www.csae.ox.ac.uk/publication/1144655/manual
Ayaita, A., & Stürmer, K. (2020). Risk aversion and the teaching profession: An analysis including different forms of risk
aversion, different control groups, selection and socialization effects. Education Economics, 28(1), 4–25. https://doi.org/10.
1080/09645292.2019.1675592
Ballou, D., & Podgursky, M. (1998). Teacher recruitment and retention in public and private schools. Journal of Policy Analysis
and Management, 17(3), 393–417. https://doi.org/10.1002/(SICI)1520-6688(199822)17:3<393::AID-PAM2>3.0.CO;2-B
Bandiera, O., Dahlstrand-Rubin, A., & Fischer, G. (2020). Incentives and Culture: Evidence from a Multi-Country Field
Experiment. Manuscript in preparation.
Banerjee, A., Banerji, R., Berry, J., Duflo, E., Kannan, H., Mukerji, S., Shotland, M., & Walton, M. (2017). From Proof of Concept
to Scalable Policies: Challenges and Solutions, with an Application. Journal of Economic Perspectives, 31(4), 73–102. https://
doi.org/10.1257/jep.31.4.73
Banerji, R. (2015). How Do Systems Respond to Disruptive Pedagogic Innovations? RISE Working Paper Series. 15/002.
https://doi.org/10.35489/BSG-RISE-WP_2015/002
Bano, M. (2012). Breakdown in Pakistan: How Aid Is Eroding Institutions for Collective Action. Stanford University Press. http://
ebookcentral.proquest.com/lib/oxford/detail.action?docID=867724
Bano, M. (forthcoming). Beating the ‘Anti-Work’ Culture: Lessons from a Successful Attempt to Improve Performance in State
Schools in Pakistan. RISE Working Paper Series.
Bano, M., & Oberoi, Z. (forthcoming). In Need of Fresh Thinking: What Pratham’s Experience of Mobilising Communities Says
about Current Development Thinking about Community Participation in Education. RISE Working Paper Series.
Barber, M. (2013). The good news from Pakistan: How a revolutionary new approach to education reform in Punjab shows
the way forward for Pakistan and development aid everywhere. Reform.
Barber, M., Moffit, A., & Kihn, P. (2010). Deliverology 101: A Field Guide For Educational Leader: A Field Guide For Educational
Leaders. SAGE Publications.
Barber, M., Rodriguez, N., & Artis, E. (2016). Deliverology in practice: How education leaders are improving student outcomes.
Corwin.
Barrett, A. M. (2005). Teacher accountability in context: Tanzanian primary school teachers’ perceptions of local community
and education administration. Compare: A Journal of Comparative and International Education, 35(1), 43–61. https://doi.
org/10.1080/03057920500033530
Barrett, A. M. (2008). Capturing the différance: Primary school teacher identity in Tanzania. International Journal of
Educational Development, 28(5), 496–507. https://doi.org/10.1016/j.ijedudev.2007.09.005
Bau, N., & Das, J. (2020). Teacher Value Added in a Low-Income Country. American Economic Journal: Economic Policy, 12(1),
62–96. https://doi.org/10.1257/pol.20170243 (open-access version: https://riseprogramme.org/publications/misallocation-
pay-and-productivity-public-sector-evidence-labor-market-teachers)
Becskehazy, I. (2018). Institucionalização do direito à educação de qualidade: O caso de Sobral, CE [PhD thesis, Universidade
de São Paulo]. https://doi.org/10.11606/T.48.2018.tde-04122018-175052
Bénabou, R., & Tirole, J. (2006). Incentives and Prosocial Behavior. American Economic Review, 96(5), 1652–1678. https://doi.
org/10.1257/aer.96.5.1652
Bennell, P., & Akyeampong, K. (2007). Teacher Motivation in Sub-Sahara Africa and South Asia. DFID. http://sro.sussex.ac.uk/
id/eprint/26224/
Bertoni, E., Elacqua, G., Méndez, C., & Santos, H. (2020). Teacher hiring instruments and teacher value added: Evidence from
Peru (Working Paper IDB-WP-01199). IDB Working Paper Series. https://doi.org/10.18235/0003123
Besley, T., Burgess, R., Khan, A., & Xu, G. (2021). Bureaucracy and Development. Manuscript in preparation.
Besley, T., & Ghatak, M. (2005). Competition and Incentives with Motivated Agents. American Economic Review, 95(3), 616–636.
https://doi.org/10.1257/0002828054201413
Béteille, T. (2009). Absenteeism, transfers and patronage: The political economy of teacher labor markets in India [PhD
thesis, Stanford University].
Béteille, T., & Evans, D. K. (2019). Successful Teachers, Successful Students: Recruiting and Supporting Society’s Most Crucial
Profession. World Bank. http://pubdocs.worldbank.org/en/235831548858735497/Successful-Teachers-January-2019.pdf
Béteille, T., Tognatta, N., Riboud, M., Nomura, S., & Ghorpade, Y. (2020). Ready to Learn: Before School, In School, and Beyond
School in South Asia. South Asia Development Forum, World Bank. https://doi.org/10.1596/978-1-4648-1327-6
Berkhout, E., Pradhan, M., Rahmawati, Suryadarma, D., & Swarnata, A. (2020). From Cheating to Learning: An Evaluation of
Fraud Prevention on National Exams in Indonesia. RISE Working Paper Series. 20/046. https://doi.org/10.35489/BSG-RISE-
WP_2020/046
Bertrand, M., Djankov, S., Hanna, R., & Mullainathan, S. (2008). Corruption in Driving Licensing Process in Delhi. Economic
and Political Weekly, 43(5), 71–76.
Bicchieri, C., Dimant, E., Gächter, S., & Nosenzo, D. (2020). Social Proximity and the Erosion of Norm Compliance. In IZA
Discussion Papers (No. 13864; IZA Discussion Papers). Institute of Labor Economics (IZA). https://ideas.repec.org/p/iza/
izadps/dp13864.html
Blazar, D., & Kraft, M. A. (2017). Teacher and Teaching Effects on Students’ Attitudes and Behaviors. Educational Evaluation
and Policy Analysis, 39(1), 146–170. https://doi.org/10.3102/0162373716670260
Bowen, D. H., Buck, S., Deck, C., Mills, J. N., & Shuls, J. V. (2015). Risky business: An analysis of teacher risk preferences.
Education Economics, 23(4), 470–480. https://doi.org/10.1080/09645292.2014.966062
Bold, T., Kimenyi, M., Mwabu, G., Ng’ang’a, A., & Sandefur, J. (2018). Experimental evidence on scaling up education reforms
in Kenya. Journal of Public Economics, 168, 1–20. https://doi.org/10.1016/j.jpubeco.2018.08.007
Brafman, O., & Beckstrom, R. A. (2006). The starfish and the spider: The unstoppable power of leaderless organizations.
Portfolio.
Breeding, M. E., Béteille, T., & Evans, D. K. (2021). Teacher Pay-for-Performance: What Works? Where? and How? World Bank.
https://documents.worldbank.org/en/publication/documents-reports/documentdetail/183331619587678000/Teacher-Pay-
for-Performance-What-Works-Where-and-How
Brown, C., & Andrabi, T. (forthcoming). Inducing Positive Sorting through Performance Pay: Experimental Evidence from
Pakistani Schools. RISE Working Paper Series.
Brown, S. A., & McIntyre, D. (1993). Making sense of teaching. Open University Press.
Bruns, B., Evans, D., & Luque, J. (2012). Achieving World-Class Education in Brazil: The Next Agenda. World Bank. https://doi.
org/10.1596/978-0-8213-8854-9
Bruns, B., Filmer, D., & Patrinos, H. A. (2011). Making schools work: New evidence on accountability reforms (No. 60050;
pp. 1–272). World Bank. http://documents.worldbank.org/curated/en/483221468147581570/Making-schools-work-new-
evidence-on-accountability-reforms
Bruns, B., & Luque, J. (2014). Great Teachers: How to raise student learning in Latin America and the Caribbean. World Bank.
http://documents.worldbank.org/curated/en/833781468238477204/Great-Teachers-how-to-raise-student-learning-in-
Latin-America-and-the-Caribbean
Bryson, A., Corsini, L., & Martelli, I. (2020). Teacher Allocation and School Performance in Italy. IZA Discussion Papers, No.
13669. Institute of Labor Economics (IZA). https://www.iza.org/publications/dp/13669/teacher-allocation-and-school-
performance-in-italy
Burge, P., Lu, H., & Phillips, W. (2021). Understanding Teaching Retention: Using a discrete choice experiment to measure
teacher retention in England. https://www.rand.org/pubs/research_reports/RRA181-1.html
Castradori, M. (2021, June 25). Situating RISE in Nigerian Political Tradition: Historical Roots of Deliberative Politics Among
the Igbo [blog post]. RISE Programme. https://riseprogramme.org/blog/Nigerian-political-tradition-Igbo-education
Chambers, J. G. (1981). The Hedonic Wage Technique As A Tool For Estimating The Costs Of School Personnel: A Theoretical
Exposition With Implications For Empirical Analysis. Journal of Education Finance, 6(3), 330–354.
Chanduví Jaña, E. (2017). Peru Faces Crises in Politics and Education. NotiSur, Latin American Data Base (LADB). https://
digitalrepository.unm.edu/notisur/14553
Chetty, R., Friedman, J. N., & Rockoff, J. E. (2014). Measuring the Impacts of Teachers I: Evaluating Bias in Teacher Value-
Added Estimates. American Economic Review, 104(9), 2593–2632. https://doi.org/10.1257/aer.104.9.2593
Childress, D., Chimier, C., Jones, C., Page, E., & Tournier, B. (2020). Change agents: Emerging evidence on instructional
leadership at the middle tier. IIEP-UNESCO; Education Development Trust; Education Commission. https://unesdoc.unesco.
org/ark:/48223/pf0000374918
Chingos, M. M., & West, M. R. (2012). Do More Effective Teachers Earn More Outside the Classroom? Education Finance and
Policy, 7(1), 8–43.
Chivore, B. R. S. (1986). Teacher Education in Post-independent Zimbabwe: Problems and possible solutions. Journal of
Education for Teaching, 12(3), 205–231. https://doi.org/10.1080/0260747860120301
Christensen, C. M. (1997). The innovator’s dilemma: When new technologies cause great firms to fail. Harvard Business
School.
Colbert, V., & Arboleda, J. (2016). Bringing a student-centered participatory pedagogy to scale in Colombia. Journal of
Educational Change, 17(4), 385–410. https://doi.org/10.1007/s10833-016-9283-7
Costrell, R. M., Hitt, C., & Shuls, J. V. (2020). A $19-Billion Blind Spot: State Pension Spending. Educational Researcher. https://
doi.org/10.3102/0013189X20912754
Costrell, R. M., & Podgursky, M. (2009). Peaks, Cliffs, and Valleys: The Peculiar Incentives in Teacher Retirement Systems and
Their Consequences for School Staffing. Education Finance and Policy, 4(2), 175–211. https://doi.org/10.1162/edfp.2009.4.2.175
Crehan, L. (2016). Exploring the impact of career models on teacher motivation. IIEP-UNESCO. https://unesdoc.unesco.org/
ark:/48223/pf0000246252
Crouch, L. (2020). Systems Implications for Core Instructional Support Lessons from Sobral (Brazil), Puebla (Mexico), and
Kenya. RISE Insight Series. 2020/020. https://doi.org/10.35489/BSG-RISE-RI_2020/020
Crouch, L., & Merseth, K. A. (2017). Stumbling at the first step: Efficiency implications of poor performance in the foundational
first five years. PROSPECTS, 47(3), 175–196. https://doi.org/10.1007/s11125-017-9401-1
Cruz-Aguayo, Y., Ibarrarán, P., & Schady, N. (2017). Do tests applied to teachers predict their effectiveness? Economics Letters,
159, 108–111. https://doi.org/10.1016/j.econlet.2017.06.035
Cueto, S., León, J., Sorto, M. A., & Miranda, A. (2017). Teachers’ pedagogical content knowledge and mathematics
achievement of students in Peru. Educational Studies in Mathematics, 94(3), 329–345. https://doi.org/10.1007/s10649-016-
9735-2
Dang, H., Glewwe, P., Lee, J., & Vu, K. (2020). What Explains Vietnam’s Exceptional Performance in Education Relative to
Other Countries? Analysis of the 2012 and 2015 PISA Data. RISE Working Paper Series. 20/036. https://doi.org/10.35489/
BSG-RISE-WP_2020/036
Datta, S. & Kingdon, G. (2021). The Myth of Teacher Shortage in India. RISE Working Paper Series. 21/072. https://doi.
org/10.35489/BSG-RISE-WP_2021/072
de Ree, J., Muralidharan, K., Pradhan, M., & Rogers, H. (2018). Double for Nothing? Experimental Evidence on an Unconditional
Teacher Salary Increase in Indonesia. The Quarterly Journal of Economics, 133(2), 993–1039. https://doi.org/10.1093/qje/
qjx040
DeJaeghere, J., Duong, B. & Dao, V. (2021). Teaching Practices That Support Thinking and Promote Learning: Qualitative
Evidence from High and Low Performing Classes in Vietnam. RISE Insight Note, 2021/024. https://doi.org/10.35489/BSG-
RISE-RI_2021/024
DiMaggio, P. J., & Powell, W. W. (1983). The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in
Organizational Fields. American Sociological Review, 48(2), 147–160. https://doi.org/10.2307/2095101
Dimmock, C., & Tan, C. Y. (2013). Educational leadership in Singapore: Tight coupling, sustainability, scalability, and
succession. Journal of Educational Administration, 51(3), 320–340. https://doi.org/10.1108/09578231311311492
Dorsey, B. J. (1989). Educational Development and Reform in Zimbabwe. Comparative Education Review, 33(1), 40–58.
Duflo, E., Dupas, P., & Kremer, M. (2011). Peer Effects, Teacher Incentives, and the Impact of Tracking: Evidence from a
Randomized Evaluation in Kenya. American Economic Review, 101(5), 1739–1774. https://doi.org/10.1257/aer.101.5.1739
Education Commission. (2019). Transforming the Education Workforce: Learning Teams for a Learning Generation. New York:
Education Commission. https://educationcommission.org/transformingtheeducationworkforce/
Education International. (2017, March 9). Status of Teachers. Education International. https://www.ei-ie.org/en/detail_
page/4648/status-of-teachers
Ellingsen, T., & Johannesson, M. (2008). Pride and Prejudice: The Human Side of Incentive Theory. American Economic Review,
98(3), 990–1008. https://doi.org/10.1257/aer.98.3.990
Engel, M., & Cannata, M. (2015). Localism and Teacher Labor Markets: How Geography and Decision Making May Contribute
to Inequality. Peabody Journal of Education, 90(1), 84–92. https://doi.org/10.1080/0161956X.2015.988533
Estrada, R. (2015). Rules Rather than Discretion: Teacher Hiring and Rent Extraction. In Economics Working Papers
(MWP2015/14; Economics Working Papers). European University Institute. https://ideas.repec.org/p/eui/euiwps/mwp2015-
14.html
Estrada, R. (2019). Rules versus Discretion in Public Service: Teacher Hiring in Mexico. Journal of Labor Economics, 37(2),
545–579. https://doi.org/10.1086/700192
Evans, D. K., Yuan, F., & Filmer, D. (2020). Are Teachers in Africa Poorly Paid? Evidence from 15 Countries (CGD Working Paper).
Center for Global Development. https://www.cgdev.org/publication/are-teachers-africa-poorly-paid-evidence-15-countries
Evans, D. K., & Yuan, F. (2018, June 22). Working Conditions of Teachers in Low- and Middle-Income Countries. RISE Annual
Conference, Blavatnik School of Government, Oxford, UK. https://riseprogramme.org/sites/default/files/inline-files/Yuan.
pdf
Fehr, E., & Falk, A. (2002). Psychological foundations of incentives. European Economic Review, 46(4), 687–724. https://doi.
org/10.1016/S0014-2921(01)00208-2
Finan, F., Olken, B. A., & Pande, R. (2017). The Personnel Economics of the Developing State. In Handbook of Economic Field
Experiments (Vol. 2, pp. 467–514). Elsevier. https://doi.org/10.1016/bs.hefe.2016.08.001
Fullan, M., & Quinn, J. (2015). Coherence: The Right Drivers in Action for Schools, Districts, and Systems. Corwin.
Fuller, B. (1991). Growing-Up Modern: The Western State Builds Third-World Schools : The Western State Builds Third-World
Schools. Routledge.
Gaduh, A., Pradhan, M., Priebe, J., & Susanti, D. (2020). Scores, Camera, Action? Incentivizing Teachers in Remote Areas.
Research on Improving Systems of Education (RISE). https://doi.org/10.35489/BSG-RISE-WP_2020/035
Gamoran, A., & Berends, M. (1987). The Effects of Stratification in Secondary Schools: Synthesis of Survey and Ethnographic
Research. Review of Educational Research, 57(4), 415–435. https://doi.org/10.3102/00346543057004415
GEAAP (Global Education Evidence Advisory Panel). (2020). Cost-Effective Approaches to Improve Global Learning: What
Does Recent Evidence Tell Us Are “Smart Buys” for Improving Learning in Low and Middle Income Countries? UK Foreign,
Commonwealth & Development Office; World Bank. http://documents1.worldbank.org/curated/en/719211603835247448/
pdf/Cost-Effective-Approaches-to-Improve-Global-Learning-What-Does-Recent-Evidence-Tell-Us-Are-Smart-Buys-for-
Improving-Learning-in-Low-and-Middle-Income-Countries.pdf
Geertz, C. (1973). The interpretation of cultures: Selected essays. Basic Books.
Gewirtz, S., Maguire, M., Neumann, E., & Towers, E. (2019). What’s wrong with ‘deliverology’? Performance measurement,
accountability and quality improvement in English secondary education. Journal of Education Policy. https://doi.org/10.10
80/02680939.2019.1706103
Goodhart, C. A. E. (1984). Monetary Theory and Practice. Macmillan Education UK. https://doi.org/10.1007/978-1-349-17295-5
Guarino, C. M., Santibañez, L., & Daley, G. A. (2006). Teacher Recruitment and Retention: A Review of the Recent Empirical
Literature. Review of Educational Research, 76(2), 173–208. https://doi.org/10.3102/00346543076002173
Gustafsson, M. (2019). The case for statecraft in education: The NDP, a recent book on governance, and the New Public
Management inheritance (WP16/2019; Stellenbosch Economic Working Papers). University of Stellenbosch. www.ekon.sun.
ac.za/wpapers/2019/wp162019
Hagos Hailu, B., Abera, B., Desta, D., Calchisa, D., Lemma, G., & Yima, W. (2018, June 22). The Nexus between the Practice
of Induction and the Formation of Novice Teachers’ Professional Identity in Ethiopia. RISE Annual Conference. https://www.
riseprogramme.org/sites/www.riseprogramme.org/files/inline-files/IHagos.pdf
Han, J., & Yin, H. (2016). Teacher motivation: Definition, research development and implications for teachers. Cogent
Education, 3(1), 1217819. https://doi.org/10.1080/2331186X.2016.1217819
Hanushek, E. A., Piopiunik, M., & Wiederhold, S. (2019). The Value of Smarter Teachers International Evidence on
Teacher Cognitive Skills and Student Performance. Journal of Human Resources, 54(4), 857–899. https://doi.org/10.3368/
jhr.54.4.0317.8619R1
Hanushek, E. A., & Rivkin, S. G. (2006). Teacher Quality. In E. Hanushek & F. Welch (Eds.), Handbook of the Economics of
Education (Vol. 2, pp. 1051–1078). Elsevier. https://doi.org/10.1016/S1574-0692(06)02018-6
Hargreaves, A., & Fullan, M. (2012). Professional capital: Transforming teaching in every school. Teachers College Press.
Harper, T. (2011). The tools of transition: Education and development in modern Southeast Asian history. In C. A. Bayly, V.
Rao, S. Szreter, & M. Woolcock (Eds.), History, historians and development policy: A necessary dialogue. Manchester University
Press. https://www.manchesteropenhive.com/view/9781526151612/9781526151612.00021.xml
Harris, D. N., & Adams, S. J. (2007). Understanding the level and causes of teacher turnover: A comparison with other
professions. Economics of Education Review, 26(3), 325–337. https://doi.org/10.1016/j.econedurev.2005.09.007
Hobbiss, M., Sims, S., & Allen, R. (2020). Habit formation limits growth in teacher effectiveness: A review of converging
evidence from neuroscience and social science. Review of Education, early access. https://doi.org/10.1002/rev3.3226
Honig, D. (forthcoming). Managing for Motivation as Performance Improvement Strategy: Closing the Implementation Gap
in Education & Far Beyond. RISE Working Paper Series.
Honig, D., & Pritchett, L. (2019). The Limits of Accounting-Based Accountability in Education (and Far Beyond): Why More
Accounting Will Rarely Solve Accountability Problems. Research on Improving Systems of Education (RISE) Working Paper
Series, No. 19/030. https://doi.org/10.35489/BSG-RISE-WP_2019/030
Hsieh, C.-T., & Klenow, P. J. (2009). Misallocation and Manufacturing TFP in China and India. The Quarterly Journal of
Economics, 124(4), 1403–1448. https://doi.org/10.1162/qjec.2009.124.4.1403
Hsieh, C.-T., & Klenow, P. J. (2014). The Life Cycle of Plants in India and Mexico. The Quarterly Journal of Economics, 129(3),
1035–1084. https://doi.org/10.1093/qje/qju014
Huang, A.R., Revina, S., Fillaili, R., & Akhmadi. (2020). The Struggle to Recruit Good Teachers in Indonesia: Institutional and
Social Dysfunctions. RISE Working Paper Series. 20/041. https://doi.org/10.35489/BSG-RISE-WP_2020/041
Huberman, M. (1988). Teacher careers and school improvement. Journal of Curriculum Studies, 20(2), 119–132. https://doi.or
g/10.1080/00220272.1988.11070783
Hwa, Y. (2019). Teacher accountability policy and sociocultural context: A cross-country study focusing on Finland and Singapore
[PhD thesis, University of Cambridge]. https://doi.org/10.17863/CAM.55349
Hwa, Y. (2021). Contrasting Approaches, Comparable Efficacy? How macro-level trust influences teacher accountability
in Finland and Singapore. In M. Ehren & J. Baxter (Eds.), Trust, Accountability and Capacity in Education System Reform:
Global Perspectives in Comparative Education (pp. 222–251). Routledge. https://www.taylorfrancis.com/chapters/oa-
edit/10.4324/9780429344855-11/contrasting-approaches-comparable-efficacy-yue-yi-hwa
IBE-UNESCO. (2012). World Data on Education: Seventh edition 2010-11. http://www.ibe.unesco.org/en/document/world-
data-education-seventh-edition-2010-11
ILO. (2012). Handbook of good human resource practices in the teaching profession. International Labour Office.
ILO/UNESCO. (1966, 5 October). Recommendation concerning the Status of Teachers, 1966. http://www.ilo.org/sector/
Resources/sectoral-standards/WCMS_162034/lang--en/index.htm
Ingersoll, R. M. (2017). Misdiagnosing America’s Teacher Quality Problem. In M. Akiba & G. K. LeTendre (Eds.), International
Handbook of Teacher Quality and Policy (pp. 79–96). Routledge. https://doi.org/10.4324/9781315710068-6
Ingersoll, R. M., & Smith, T. M. (2004). Do Teacher Induction and Mentoring Matter? NASSP Bulletin. https://doi.
org/10.1177/019263650408863803
Ingersoll, R., Merrill, L., & May, H. (2016). Do Accountability Policies Push Teachers Out? Educational Leadership, 73(8), 44–49.
Jacinto, A., & Gershenson, S. (2021). The Intergenerational Transmission of Teaching. American Educational Research Journal,
58(3), 635–672. https://doi.org/10.3102/0002831220963874
Jackson, C. K. (2013). Match quality, worker productivity, and worker mobility: Direct evidence from teachers. The Review of
Economics and Statistics, 95(4), 1096–1116.
Jeevan, S. (2021). Intrinsic: A manifesto to reignite our inner drive. Endeavour.
Johnston, A. C. (2021). Preferences, Selection, and the Structure of Teacher Compensation. EdWorkingPapers, No. 20-202.
Annenberg Institute at Brown University. https://www.edworkingpapers.com/ai20-202
Jovanovic, B. (1979). Job Matching and the Theory of Turnover. Journal of Political Economy, 87(5, Part 1), 972–990. https://
doi.org/10.1086/260808
Kaffenberger, M. (2021, March 25). Aligning Education Systems for Learning: How Systems Shift [blog post]. RISE Programme.
https://riseprogramme.org/blog/aligning-education-systems-for-learning-how-systems-shift
Kambourov, G., & Manovskii, I. (2008). Rising Occupational and Industry Mobility in the United States: 1968–97. International
Economic Review, 49(1), 41–79. https://doi.org/10.1111/j.1468-2354.2008.00473.x
Kan, Y. (2014, April). Your appraisal and career planning process streamlined. Contact: The Teachers’ Digest, (14), 14–15.
Singapore Ministry of Education, Communications and Engagement Group.
Karbownik, K. (2020). The effects of student composition on teacher turnover: Evidence from an admission reform.
Economics of Education Review, 75, 101960. https://doi.org/10.1016/j.econedurev.2020.101960
Khan, A. Q., Khwaja, A. I., & Olken, B. A. (2019). Making Moves Matter: Experimental Evidence on Incentivizing Bureaucrats
through Performance-Based Postings. American Economic Review, 109(1), 237–270. https://doi.org/10.1257/aer.20180277
Kingdon, G., Aslam, M., Rawal, S., & Das, S. (2013). Are contract teachers and para-teachers a cost-effective intervention
to address teacher shortage and improve learning outcomes? EPPI-Centre, Social Science Research Unit, Institute
of Education, University of London. https://assets.publishing.service.gov.uk/media/57a08a1fe5274a27b200042d/
Parateachers2013Kingdon.pdf
Klassen, R. M., Kim, L. E., Rushby, J. V., & Bardach, L. (2020). Can we improve how we screen applicants for initial teacher
education? Teaching and Teacher Education, 87, 102949. https://doi.org/10.1016/j.tate.2019.102949
Knowles, E. (Ed.). (2014). The Oxford Dictionary of Quotations. Oxford University Press.
Kraft, M. A., Papay, J. P., & Chi, O. L. (2020). Teacher Skill Development: Evidence from Performance Ratings by Principals.
Journal of Policy Analysis and Management, 39(2), 315–347. https://doi.org/10.1002/pam.22193
Le, H. M. (2018). The reproduction of ‘best practice’: Following Escuela Nueva to the Philippines and Vietnam. International
Journal of Educational Development, 62, 9–16. https://doi.org/10.1016/j.ijedudev.2018.02.005
Le Nestour, A., Moscoviz, L., & Sandefur, J. (2021, March). Trends in school quality in 87 developing countries, from 1950 to
2000. Paper presented at CSAE Conference 2021: Economic Development in Africa, University of Oxford (online conference).
Leaver, C., Ozier, O., Serneels, P., & Zeitlin, A. (2021). Recruitment, Effort, and Retention Effects of Performance Contracts
for Civil Servants: Experimental Evidence from Rwandan Primary Schools. American Economic Review, 111(7), 2213–2246.
https://doi.org/10.1257/aer.20191972
Lee, S. K., Goh, C. B., Fredriksen, B., & Tan, J. P. (Eds.). (2008). Toward a better future: education and training for economic
development in Singapore since 1965. World Bank; National Institute of Education. http://documents.worldbank.org/curated/
en/121951468302724234/Toward-a-better-future-education-and-training-for-economic-development-in-Singapore-
since-1965
Leigh, A. (2013). The Economics and Politics of Teacher Merit Pay. CESifo Economic Studies, 59(1), 1–33. https://doi.org/10.1093/
cesifo/ifs007
Lemos, R., Muralidharan, K., Scur, D. (2021). Personnel Management and School Productivity: Evidence from India. RISE
Working Paper Series. 21/063. https://doi.org/10.35489/BSG-RISE-WP_2021/063
Low, E.-L., & Tan, O.-S. (2017). Teacher Education Policy: Recruitment, Preparation and Progression. In O.-S. Tan, W.-C. Liu, &
E.-L. Low (Eds.), Teacher Education in the 21st Century (pp. 11–32). https://doi.org/10.1007/978-981-10-3386-5_2
Makuwa, D. K. (2011). Characteristics of Grade 6 Teachers (No. 2; SAQMEQ Working Paper). Southern and Eastern Africa
Consortium for Monitoring Educational Quality. http://www.sacmeq.org/sites/default/files/sacmeq/publications/02_
teachers_final_05dec2011.pdf
Malamud, O. (2010). Breadth versus Depth: The Timing of Specialization in Higher Education. LABOUR, 24(4), 359–390. https://
doi.org/10.1111/j.1467-9914.2010.00489.x
Malinen, O.-P., Väisänen, P., & Savolainen, H. (2012). Teacher education in Finland: A review of a national effort for preparing
teachers for the future. Curriculum Journal, 23(4), 567–584. https://doi.org/10.1080/09585176.2012.73101
Mansbridge, J. (2014). A Contingency Theory of Accountability. In M. Bovens, R. E. Goodin, & T. Schillemans (Eds.), The Oxford
Handbook of Public Accountability. Oxford University Press. https://doi.org/10.1093/oxfordhb/9780199641253.013.0019
Mbiti, I., Muralidharan, K., Romero, M., Schipper, Y., Manda, C., & Rajani, R. (2019). Inputs, Incentives, and Complementarities
in Education: Experimental Evidence from Tanzania. The Quarterly Journal of Economics, 134(3), 1627–1673. https://doi.
org/10.1093/qje/qjz010
Mbiti, I., & Schipper, Y. (2021). Teacher and Parental Perceptions of Performance Pay in Education: Evidence from Tanzania.
Journal of African Economies, 30(1), 55–80. https://doi.org/10.1093/jae/ejaa012
McAlpine, K., Kagucia Omesa, N., Semwene, J., & Mbise, A. (2018). A Twist on Performance Theory: How rewarding individual
teachers may promote whole school improvement. Examining the effects of KiuFunza in Mbozi District, Tanzania. Twaweza.
https://www.twaweza.org/uploads/files/Kiufunza%20Qualitative_Final%20report%20March%202018.pdf
Meadows, D. H. (2008). Thinking in Systems: A Primer. Chelsea Green Publishing.
Mehta, J. (2013). The allure of order: High hopes, dashed expectations, and the troubled quest to remake American schooling.
Oxford University Press.
Menzies, L., Parameshwaran, M., Trethewey, A., Shaw, B., Baars, S., & Chiong, C. (2015). Why teach? Retrieved from LKMco
website: https://whyteach.lkmco.org/wp-content/uploads/2015/10/Embargoed-until-Friday-23-October-2015-Why-Teach.
pdf
Metzler, J. (2009). The Developing States and Education: Africa. In R. Cowen & A. M. Kazamias (Eds.), International Handbook
of Comparative Education (pp. 277–294). Springer Netherlands. https://doi.org/10.1007/978-1-4020-6403-6_18
Muñoz, P. (2021). Latin America Erupts: Peru Goes Populist. Journal of Democracy, 32(3), 48–62. https://doi.org/10.1353/
jod.2021.0033
Mintzberg, H. (1979). The structuring of organizations: A synthesis of the research. Prentice-Hall.
Mintzberg, H. (1983). Structure in fives: Designing effective organizations. Prentice-Hall.
Mizala, A., & Ñopo, H. (2016). Measuring the relative pay of school teachers in Latin America 1997–2007. International Journal
of Educational Development, 47, 20–32. https://doi.org/10.1016/j.ijedudev.2015.11.014
Muralidharan, K. (2016). A New Approach to Public Sector Hiring in India for Improved Service Delivery. India Policy Forum,
12(1), 187–236. https://www.ncaer.org/publication_details.php?pID=267&pID=267
Muralidharan, K., & Singh, A. (2020). Improving Public Sector Management at Scale? Experimental Evidence on School
Governance in India. RISE Working Paper Series. 20/056. https://doi.org/10.35489/BSG-RISE-WP_2020/056
Muralidharan, K., & Singh, A. (forthcoming). Improving Schooling Productivity through Computer-Aided Personalization:
Experimental Evidence from Rajasthan. RISE Working Paper Series.
Muralidharan, K., & Sundararaman, V. (2011a). Teacher Performance Pay: Experimental Evidence from India. Journal of
Political Economy, 119(1), 39–77. https://doi.org/10.1086/659655
Muralidharan, K., & Sundararaman, V. (2011b). Teacher opinions on performance pay: Evidence from India. Economics of
Education Review, 30(3), 394–403. https://doi.org/10.1016/j.econedurev.2011.02.001
Murnane, R. (1984). Selection and Survival in the Teacher Labor Market. The Review of Economics and Statistics, 66(3), 513–
518. https://doi.org/10.2307/1925013
Murnane, R., & Cohen, D. (1986). Merit Pay and the Evaluation Problem: Why Most Merit Pay Plans Fail and a Few Survive.
Harvard Educational Review, 56(1), 1–18. https://doi.org/10.17763/haer.56.1.l8q2334243271116
Nagler, M., Piopiunik, M., & West, M. R. (2020). Weak Markets, Strong Teachers: Recession at Career Start and Teacher
Effectiveness. Journal of Labor Economics, 38(2), 453–500. https://doi.org/10.1086/705883
Naviwala, N. (2016). Pakistan’s Education Crisis: The Real Story. In Woodrow Wilson International Center for Scholars. Woodrow
Wilson International Center for Scholars. https://eric.ed.gov/?id=ED570671
Nihayah, R. W., Revina, S., & Usman, S. (2020). Sociocultural Drivers of Local Educational Innovations: Findings from
Indonesia. RISE Working Paper Series. 20/043. https://doi.org/10.35489/BSG-RISE-WP_2020/043
Nziramasanga, C. T. (1991). Teacher Education Innovation in Zimbabwe. Action in Teacher Education, 13(3), 16–20. https://doi.
org/10.1080/01626620.1991.10462780
OECD. (2018). PISA for Development Database. https://www.oecd.org/pisa/pisa-for-development/database/
Papay, J. P., & Kraft, M. A. (2015). Productivity returns to experience in the teacher labor market: Methodological challenges
and new evidence on long-term career improvement. Journal of Public Economics, 130, 105–119. https://doi.org/10.1016/j.
jpubeco.2015.02.008
Phillips, D. (2000). The Enduring Nature of the Tripartite System of Secondary Schooling in Germany: Some Explanations.
British Journal of Educational Studies, 48(4), 391–412. https://doi.org/10.1111/1467-8527.00154
Podolsky, A., Kini, T., & Darling-Hammond, L. (2019). Does teaching experience increase teacher effectiveness? A review of
US research. Journal of Professional Capital and Community, 4(4), 286–308. https://doi.org/10.1108/JPCC-12-2018-0032
Prendergast, C. (1999). The Provision of Incentives in Firms. Journal of Economic Literature, 37(1), 7–63. https://doi.
org/10.1257/jel.37.1.7
Pritchett, L. (2013). The rebirth of education: Schooling ain’t learning. Center for Global Development.
Pritchett, L. (2014). The risks to education systems from design mismatch and global isomorphism: Concepts, with examples
from India (WIDER Working Paper). UNU-WIDER. https://doi.org/10.35188/UNU-WIDER/2014/760-8
Pritchett, L. (2015). Creating Education Systems Coherent for Learning Outcomes. Research on Improving Systems of
Education (RISE) Working Paper, No. 15/005. https://doi.org/10.35489/BSG-RISE-WP_2015/005
Pritchett, L. (2017). “The Evidence” About “What Works” in Education: Graphs to Illustrate External Validity and Construct
Validity. RISE Insight. https://doi.org/10.35489/BSG-RISE-RI_2017/002
Pritchett, L., & Beatty, A. (2012). The Negative Consequences of Overambitious Curricula in Developing Countries. Center for
Global Development Working Paper No. 293. http://www.cgdev.org/content/publications/detail/1426129
Pritchett, L., & Murgai, R. (2006). Teacher Compensation: Can Decentralization to Local Bodies Take India from the Perfect
Storm Through Troubled Waters to Clear Sailing? India Policy Forum, 2006–07. https://www.brookings.edu/wp-content/
uploads/2016/07/2006_pritchett.pdf
Pritchett, L., & Pande, V. (2006). Making primary education work for India’s rural poor: A proposal for effective decentralization
(No. 36965; pp. 1–115). The World Bank. http://documents.worldbank.org/curated/en/748351468042304557/Making-
primary-education-work-for-Indias-rural-poor-a-proposal-for-effective-decentralization
Pritchett, L., & Sandefur, J. (2020). Girls’ schooling and women’s literacy: Schooling targets alone won’t reach learning goals.
International Journal of Educational Development, 78. https://doi.org/10.1016/j.ijedudev.2020.102242
Pritchett, L., & Viarengo, M. (2009). The Illusion of Equality: The Educational Consequences of Blinding Weak States, for
Example. Center for Global Development Working Paper No. 178. http://www.cgdev.org/content/publications/detail/1422654
Pritchett, L., & Viarengo, M. (2015). The State, Socialisation, and Private Schooling: When Will Governments Support
Alternative Producers? The Journal of Development Studies, 51(7), 784–807. https://doi.org/10.1080/00220388.2015.1034109
Pritchett, L. & Viarengo, M. (2021). Learning Outcomes in Developing Countries: Four Hard Lessons from PISA-D. RISE Working
Paper Series. 21/069. https://doi.org/10.35489/BSG-RISE-WP_2021/069
Rebitzer, J. B., & Taylor, L. J. (2011). Extrinsic Rewards and Intrinsic Motives: Standard and Behavioral Approaches to Agency
and Labor Markets. Handbook of Labor Economics, 4, 701–772. https://doi.org/10.1016/S0169-7218(11)04114-1
Revina, S., Pramana, R. P., Fillaili, R., & Suryadarma, D. (2020). Systemic Constraints Facing Teacher Professional Development
in a Middle- Income Country: Indonesia’s Experience Over Four Decades. RISE Working Paper Series. 20/054. https://doi.
org/10.35489/BSG-RISE-WP_2020/054
Roberts, J. (2004). The modern firm: Organizational design for performance and growth. Oxford University Press.
Romero, M., Bedoya, J., Yanez Pagans, M., Silveyra, M., & De Hoyos, R. (2021). School Management, Grants, and Test Scores:
Experimental Evidence from Mexico. In Policy Research Working Paper Series (No. 9535; Policy Research Working Paper
Series). The World Bank. https://ideas.repec.org/p/wbk/wbrwps/9535.html
Ronfeldt, M., Loeb, S., & Wyckoff, J. (2013). How Teacher Turnover Harms Student Achievement. American Educational
Research Journal, 50(1), 4–36. https://doi.org/10.3102/0002831212463813
Rowan, B. (1990). Commitment and Control: Alternative Strategies for the Organizational Design of Schools. Review of
Research in Education, 16(1), 353–389. https://doi.org/10.3102/0091732X016001353
Ryan, R. M., & Deci, E. L. (2000). Intrinsic and Extrinsic Motivations: Classic Definitions and New Directions. Contemporary
Educational Psychology, 25(1), 54–67. https://doi.org/10.1006/ceps.1999.1020
Saavedra, J., & Gutierrez, M. (2020). Peru: A Wholesale Reform Fueled by an Obsession with Learning and Equity. In F. M.
Reimers (Ed.), Audacious Education Purposes: How Governments Transform the Goals of Education Systems (pp. 153–180).
Springer International Publishing. https://doi.org/10.1007/978-3-030-41882-3_6
Sabarwal, S., Abu-Jawdeh, M., & Kapoor, R. (2021). Teacher Beliefs: Why They Matter and What They Are. The World Bank
Research Observer, lkab008. https://doi.org/10.1093/wbro/lkab008
Sahlberg, P. (2021). Finnish lessons 3.0: what can the world learn from educational change in Finland? (Third edition). Teachers
College Press, Columbia University.
Salazar-Morales, D. A. (2018). Sermons, carrots or sticks? Explaining successful policy implementation in a low performance
institution. Journal of Education Policy, 33(4), 457–487. https://doi.org/10.1080/02680939.2017.1378823
Salop, J., & Salop, S. (1976). Self-Selection and Turnover in the Labor Market. The Quarterly Journal of Economics, 90(4), 619.
https://doi.org/10.2307/1885325
Schiefelbein, E., Wolff, L., & Schiefelbein, P. (1999). Cost-effectiveness of education policies in Latin America: A survey of
expert opinion. Bulletin of the UNESCO Regional Office for Education in Latin America and the Caribbean (Chile), 49. https://
unesdoc.unesco.org/ark:/48223/pf0000117890
Schipper, Y., & Rodriguez-Segura, D. (2021, September 20–21). Teacher incentives and attendance: evidence from Tanzania
[Conference presentation]. RISE Country Research Teams Virtual Meeting 2021.
Scott, J. C. (2008). Seeing like a state: How certain schemes to improve the human condition have failed. Yale University
Press.
Shulman, L. (1986). Those Who Understand: Knowledge Growth in Teaching. Educational Researcher, 15(2), 4–14. https://doi.
org/10.3102/0013189X015002004
Simola, H., Rinne, R., Varjo, J., Pitkänen, H., & Kauko, J. (2009). Quality assurance and evaluation (QAE) in Finnish compulsory
schooling: a national model or just unintended effects of radical decentralisation? Journal of Education Policy, 24(2), 163–178.
https://doi.org/10.1080/02680930902733139
Singh, A. (2020). Myths of Official Measurement: Auditing and Improving Administrative Data in Developing Countries. RISE
Working Paper Series. 20/042. https://doi.org/10.35489/BSG-RISE-WP_2020/042
Smith, R. S. (1979). Compensating Wage Differentials and Public Policy: A Review. Industrial and Labor Relations Review, 32(3),
339–352. https://doi.org/10.2307/2522263
Spivack, M. (2021). Applying Systems Thinking to Education: The RISE Systems Framework. 2021/028. https://doi.
org/10.35489/BSG-RISE-RI_2021/028
Thaler, R., & Rosen, S. (1976). The Value of Saving a Life: Evidence from the Labor Market (No. c3964). National Bureau of
Economic Research. https://www.nber.org/books-and-chapters/household-production-and-consumption/value-saving-
life-evidence-labor-market
Thompson, J. D. (2003). Organizations in action: Social science bases of administrative theory. Transaction Publishers.
Originally published in 1967.
Tournier, B., & Chimier, C. (2019). Teacher career reforms: Learning from experience. International Institute for Educational
Planning, UNESCO, Paris. http://www.iiep.unesco.org/en/publication/teacher-career-reforms-learning-experience
Tyack, D. B. (1974). The One Best System: A History of American Urban Education. Harvard University Press.
UNESCO. (2017). Accountability in education: Meeting our commitments (Global Education Monitoring Report 2017/18). http://
unesdoc.unesco.org/images/0025/002593/259338e.pdf
UNESCO. (2019). Teacher Policy Development Guide. UNESCO, Paris. https://teachertaskforce.org/knowledge-hub/teacher-
policy-development-guide
UNESCO. (2020). A review of the use of contract teachers in sub-Saharan Africa. Teachers Task Force for Education 2030.
https://teachertaskforce.org/knowledge-hub/review-use-contract-teachers-sub-saharan-africa
Vegas, E., & Ganimian, A. J. (2013). Theory and Evidence on Teacher Policies in Developed and Developing Countries. Inter-
American Development Bank. https://publications.iadb.org/en/publication/16577/theory-and-evidence-teacher-policies-
developed-and-developing-countries
Vegas, E., & Umansky, I. (2005). Improving teaching and learning through effective incentives: Lessons from education
reforms in Latin America. In Incentives to Improve Teaching (pp. 1–20). World Bank. https://doi.org/10.1596/978-0-8213-
6215-0
Venkat, H., & Spaull, N. (2015). What do we know about primary teachers’ mathematical content knowledge in South
Africa? An analysis of SACMEQ 2007. International Journal of Educational Development, 41, 121–130. https://doi.org/10.1016/j.
ijedudev.2015.02.002
Viscusi, W. K. (1993). The Value of Risks to Life and Health. Journal of Economic Literature, 31(4), 1912–1946.
Voisin, A., & Dumay, X. (2020). How do educational systems regulate the teaching profession and teachers’ work? A
typological approach to institutional foundations and models of regulation. Teaching and Teacher Education, 103144. https://
doi.org/10.1016/j.tate.2020.103144
Warwick, D. P., & Reimers, F. (1995). Hope or despair? Learning in Pakistan’s primary schools. Praeger.
Watt, H. M. G., & Richardson, P. W. (2007). Motivational Factors Influencing Teaching as a Career Choice: Development
and Validation of the FIT-Choice Scale. The Journal of Experimental Education, 75(3), 167–202. https://doi.org/10.3200/
JEXE.75.3.167-202
Weick, K. E. (1976). Educational Organizations as Loosely Coupled Systems. Administrative Science Quarterly, 21(1), 1–19.
JSTOR. https://doi.org/10.2307/2391875
Woodcock, S. D. (2015). Match effects. Research in Economics, 69(1), 100–121. https://doi.org/10.1016/j.rie.2014.12.001
World Bank. (2004). World Development Report 2004: Making Services Work for Poor People. World Bank. https://doi.
org/10.1596/0-8213-5468-X
World Bank. (2017). Service Delivery Indicators. Accessed at https://databank.worldbank.org/reports.aspx?source=service-
delivery-indicators
World Bank. (2018). World Development Report 2018: Learning to Realize Education’s Promise. World Bank. https://doi.
org/10.1596/978-1-4648-1096-1
World Bank. (2019). Ending Learning Poverty: What Will It Take?. World Bank https://openknowledge.worldbank.org/
handle/10986/32553
Zeitlin, A. (2021). Teacher Turnover in Rwanda. Journal of African Economies, 30(1), 81–102. https://doi.org/10.1093/jae/
ejaa013
December 2021

Teacher Careers Education Systems Coherent Learning

Uploaded by

Copyright:

Available Formats

Teacher Careers Education Systems Coherent Learning

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Teacher Careers Education Systems Coherent Learning

Uploaded by

Copyright:

Available Formats

Teacher Careers in

Education Systems That

Yue-Yi Hwa and Lant Pritchett

Use and dissemination of this working paper is encouraged;

The findings, interpretations, and conclusions expressed

For more information, visit: riseprogramme.org

Photo credits: BigStock (cover pages)

Part 1. Introduction: Back from the brink 6

Part 2. We need to start with purpose and priorities 19

Teacher careers are hedonic and dynamic, and any

Effective organisations have a shared purpose

Changes that are incoherent with the complexity

Changes that are incoherent with the complexity

3.3 Why ‘choose and curate’? Because the early career 53

Part 4. The 5Cs in practice: Varied examples of common 63

In purpose-driven teacher career structures, ‘fairness’ in compensation is

PREMISE FOR PRACTICE #5: on purpose-driven definitions of fairness

Examples of the 5Cs on the ‘journey’ toward empowered,

Part 5. Conclusion: Where do you want to go? 75

Part 1. Introduction: Back from the brink

Part 1. Introduction: Back from the brink 6

The 5Cs What they imply

Education authorities and organisations should initially choose teachers

The novice phase of teaching should be viewed as a period of curation, during

After the initial curation, education authorities and organisations should

Effective teachers need to be technically capable and equipped for cultivating

Teachers also need to be motivationally committed to the systemwide goal of

Status quo The 5Cs

Piecemeal policies and reforms System-oriented policies and reforms

Teaching as straightforward task Teaching as complex

Typical approach What to do differently

No consideration of whether the education

Premise for practice #2: Contextually embedded, learning-oriented

Source: Pritchett & Sandefur (2020), Figure 1(a).

Overview incoherences identified in the previous section, and what

Teacher careers are hedonic and dynamic, and any 24

Effective organisations have a shared purpose 32

Changes that are incoherent with the complexity 40

Changes that are incoherent with the complexity of teacher motivation 41

Source: Authors, using PISA-D data (OECD, 2018).

Table 4. A hedonic and dynamic model of teacher career structures

Phases of teacher careers

Box 2: Resources on teacher educators and mid-tier education bureaucrats

Can the design element influence the sources of teacher motivation?

Box 3: Examples of strategies for building clear, consensus-based prioritisation

Unfortunately, piecemeal improvements to service coherence in a number of ways.

34 See https://hbr.org/search?N=4294923992&Ntt=lincoln+electric, accessed on 22 March 2021.

PREMISE FOR PRACTICE #1: on clear, consensus-based prioritisation of learning

35 See also https://www.teachingattherightlevel.org/tarl-in-action/, accessed on 22 March 2021.

36 For an explanation of our use of ‘thick’ and ‘thin’, see footnote 1.

Being a mortal was strange.…Time was inexorable and unstoppable.

–the Greek god Apollo, in Jo Walton’s The Just City (2014)

embedded is a two-sided prospect. In other words, it is not student learning;

Figure 8. A possible instantiation of the 5Cs principles

Source: Adapted from Pritchett & Murgai (2006), Figure 9.

Phases of teacher careers

46 For more on frictions generated by teacher turnover, see Section 3.5

Source: Authors, using PISA-D data (OECD, 2018).

49 For a more detailed discussion of this literature, see Honig (forthcoming).