Research Policy 46 (2017) 1505–1518

Contents lists available at ScienceDirect

Research Policy
journal homepage: www.elsevier.com/locate/respol

Innovation, quality management and learning: Short-term and longer-term MARK

effects
⁎
Jane Bourkea, , Stephen Roperb
a
Department of Economics, Cork University Business School, University College Cork, Ireland
b
Enterprise Research Centre and Warwick Business School, Coventry, CV4 7AL, United Kingdom

A R T I C L E I N F O A B S T R A C T

JEL Classifications: Quality-orientated management change and innovation are central strategies for firms. Implementing both
O30 quality improvement and innovation poses significant managerial, organisational and technical challenges, and
032 may also involve significant lags before benefits are realised. Here, using data on a large group of Irish manu-
L15 facturing plants and econometric analysis, we establish the short- and longer-term influence of plants’ adoption
Keywords: of quality improvement methods (QIMs) on product innovation performance. Our study highlights the short-
Innovation term disruptive and longer-term beneficial effects of QIM adoption on product innovation performance. In ad-
Quality improvement dition, we find evidence of complementarities and learning-by-using effects from QIM adoption. Our results
Management change
suggest that maximising the returns to innovation and quality improvement requires consideration of the soft
and/or hard nature of individual QIMs and the timing and sequencing of their adoption.

1. Introduction quality improvement methods (QIMs). Quality improvement and in-

novation are clearly inter-related although there is little agreement on
With increased market competition, the successful management of whether this is of a complementary or opposing nature. Nowak (1997),
change is crucial to firm survival and success (Todnem By, 2005). for example, envisages a complementary relationship, commenting
Quality improvement and innovation have therefore become estab- that: ‘quality and innovation processes are inter-linked and should not
lished strategies as firms seek to create and defend their competitive be treated separately. Technical change not enhancing quality is illusive
position (Pekovic and Galia, 2009). Some authors have argued that because it does not contribute to a sustained and improved strategic
quality improvement and innovation are the central concepts of new competitive advantage, nor does it increase the value creation potential
forms of economic theory of the firm and models of business behaviour of available resources through quality creation’. Other writers have
(Anderson et al., 1994; Black and Porter, 1996; Rungtusanatham et al., seen quality improvement processes − which may involve mechanistic
1998), viz. ‘Quality is a vital component of the business strategy, and routinisation and standardised business processes − as restricting
quality improvement is a strategic variable employed in the highly creativity and innovation (Glynn, 1996; Kanter, 1983; Perdomo-Ortiz
competitive international business world’ (Adam et al., 2001, p. 43). et al., 2009a,b; Prajogo and Sohal, 2004). Where the relationship be-
And, on innovation Baumol (2002 p. ix) comments: ‘firms cannot afford tween quality improvement methods (QIMs) and product innovation
to leave innovation to chance. Rather, managements are forced by has been explored empirically relationships are generally positive
market pressures to support innovation activity systematically … The (Martínez-Costa and Martínez-Lorente, 2008; Hung et al., 2011, Prajogo
result is a ferocious arms race among firms in the most rapidly evolving and Sohal, 2004; Hoang et al., 2006; Zeng at al., 2015). Other studies,
sectors of the economy, with innovation as the prime weapon’. however, have found either neutral or negative relationships between
Within the management change literature, two paradigms of ‘hard’ QIMs and product innovation (Terziovski and Guerrero, 2014). With a
and ‘soft’ management change emerge. Hard managerial changes typi- growing recognition of the complementary nature of hard and soft
cally emphasise rules, standardisation, conformity, discipline, stability, managerial processes, recent studies highlight the benefit of in-
formality, whereas knowledge sharing/diffusion, reflection engage- coporating a combination of hard and soft quality management prac-
ment, empowerment and intelligence gathering and are reflective of tices for product innovation (Hoang et al., 2006; Zeng et al., 2015) and
soft managerial changes. We use these contrasts to explore in more firm performance (Gadenne and Sharma, 2009; Calvo-Mora et al.,
depth the relationship between product innovation performance and 2013).

⁎
Corresponding author.
E-mail addresses: jane.bourke@ucc.ie (J. Bourke), Stephen.Roper@wbs.ac.uk (S. Roper).

http://dx.doi.org/10.1016/j.respol.2017.07.005
Received 17 January 2016; Received in revised form 13 July 2017; Accepted 13 July 2017
Available online 29 July 2017
0048-7333/ © 2017 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

One other commonality between quality improvement and in- schools of thought emphasising the inter-subjective creation of knowl-
novation processes noted in the literature is that both are often difficult edge – people are continually developing and refining their views
to implement leading to significant lags in the realisation of any related which informs their actions (Crawford and Pollack, 2004). These bot-
benefits. Pekovic and Galia (2009) comment, for example, that ‘im- tom–up models of organisational change recognise a non-linear, poli-
plementation of the ISO 9000 standard … concerns the whole organi- tical and irrational process. Such models may be characterised as dy-
sation and involves changes in the fundamental behaviour and applied namic and fragmented, albeit interconnected, composed of competing
routine of employees’ (Pekovic and Galia, 2009, p. 831). Likewise, in- perspectives and interested and supported by informal systems
novation may result in short-term disruption before any longer-term (Crawford et al., 2003).
performance benefits are accrued by the firm (Roper et al., 2008). While organisational practices can differ considerably, parallels
Understanding the performance benefits of innovation and quality im- exist in the demarcation of soft and hard practices across the spectrum.
provement, and their interactions, is therefore likely to require long- For instance, in project management, the hard paradigm assumes that
itudinal data covering a period of years in which causal mechanisms are goals and methods are already well defined, and the objective is to find
clearly identifiable. the best solution to a particular problem, however ‘best’ is defined and
Here, using data on a large group of Irish manufacturing plants we measured. Contrastingly, the soft paradigm suggests that the aspects of
focus on the relationship between product innovation performance and a situation that cause it to be problematic are not easily defined or
the adoption of quality orientated hard and soft management change. isolated. Therefore, it is necessary to engage with people at a qualitative
We ask whether, and over what period, the adoption of QIMs (specifi- level in the understanding that it is unlikely that there will be a unique
cally ISO9000, TQM and Quality Circles) impacts on plants’ innovation ‘best’ solution (Midgley, 2000). Within the HR literature, similar dif-
success (specifically sales generated from product innovation). Most, if ferentiations apply. In general, soft management practices encourage
not all, of the prior studies of the relationship between QIMs and in- knowledge sharing, engagement, empowerment and encourages in-
novation have been based on cross-sectional analysis making causality telligence gathering and reflection whereas hard management practices
difficult to identify, and providing little information on the nature of often are rule-based and require conformity, standardisation, discipline
the learning effects and lags involved in QIM adoption and its potential and stability (Jenkins and Delbridge, 2013). Furthermore, there is a
benefits for innovation. Our study makes several important contribu- growing realisation across the organisational change literature that
tions. First, our data allows us to identify the temporal profile of the hard and soft practices are more beneficial when introduced together.
performance benefits of individual QIMs, highlighting short-term dis- Within the project management literature, Crawford et al. (2003) report
ruption (negative) effects but longer-term (positive) benefits. Second, it the need for both hard and soft perspectives when managing complex
seeks to explain the short-term and long-term aspects of the quality- organisational change projects, particularly when changing aspects of
innovation relationship within the context of the contrasting paradigms organisations, such as working practices and culture. In addition,
of hard and soft managerial change. Third, it highlights com- Arvanitis et al. (2013) report that the combination of hard and soft ICT
plementarities and learning by using effects for product innovation capital has a positive effect on both process and product/services in-
performance arising from the adoption of quality-orientated hard and novation
soft managerial processes.
2.2. Quality orientated management change
2. Concepts and hypotheses
Many firms have responded to the challenges they face by in-
2.1. Hard and soft management change corporating quality-based strategies into their change management
approach (Foley et al., 1997). A commitment to quality can drive firms
With increased market competition and developments in tech- to make significant improvements in profitability, productivity and
nology, the characteristics of business have changed drastically competitiveness (Deming, 1986; Morgan and Vorhies, 2001). Hard
(Pekovic and Galia, 2009). The successful management of change is quality management is mechanistic in nature and emphasises stability,
crucial to survive and succeed in the highly competitive and con- conformity and discipline, and comprises processes such as work design
tinuously evolving business environment (Todnem By, 2005). Organi- and statistical process control. These hard components relate to the
sational change management has been defined as ‘the process of con- control of processes and products to maintain uniformity, comply with
tinually renewing an organisation’s direction, structure, and quality standards and satisfy manufacturing specifications (López-
capabilities to serve the ever-changing needs of external and internal Mielgo et al., 2009). Soft quality management stresses employee en-
customers’ (Moran and Brightman, 2001). It means entering new ter- gagement, partnerships, and comparison with the market leaders. These
ritory and “playing the game by new rules” and moving the organisa- soft aspects of quality management are more organic in nature and
tion from its current state to a more desirable improved state (Ragsdell, focus on leadership, empowerment and training, and encourage em-
2000). Two paradigms of organisational change emerge from the lit- ployees to scan the environment for new trends, approaches and tech-
erature. In general, objectivist, scientific approaches are hard, while nologies (Moura E Sá and Abrunhosa, 2007; McAdam, 2000). Soft
subjectivist, social approaches are soft. The terms hard and soft are quality management promotes the more human and developmental
commonly used across a broad range of organisational change prac- aspects of the quality system allowing the firm to adapt to its changing
tices, such as HRM practices (Storey, 1989), quality improvement environment and promoting continuous improvement (López-Mielgo
practices (Zeng et al., 2015), Information and Communications Tech- et al., 2009).
nology (ICT) (Arvanitis et al., 2013) and project management Three of the most widely recognised QIMs, which span the soft-hard
(Crawford et al., 2003). range of management change practices, are Total Quality Management
The hard, positivist, paradigm promotes an understanding of the (TQM), Quality Certification (such as ISO9000) and Quality Circles.
world as an objective reality – systems are mechanistic processes, with TQM has been described as a management philosophy that fosters an
stable or predictably varying, relationships between the relevant vari- organisational culture committed to customer satisfaction through
ables (Crawford and Pollack, 2004). In practice, the hard paradigm continuous improvement (Kanji, 2002). The TQM philosophy essen-
often takes a top-down approach, following a rational hierarchical tially comprises three key elements: customer focus, people involve-
model that emphasises control and is expressed through formal struc- ment and continuous improvement (Moura E Sá and Abrunhosa, 2007).
tures and systems. Its language acts to superimpose a logic, order, and Quality Certification initiatives, e.g. ISO 9000, require detailed review
structure on an otherwise irrational social process (Crawford et al., and documentation of a firm’s production processes, in accordance with
2003). The soft paradigm stems from interpretivist and constructivist the quality system requirements specified by ISO.1 The ISO 9000

1506
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

standard is based on eight principles that address the core values and management leadership, HRM and a flexible culture oriented towards
concepts of quality management: customer focus, leadership, involve- continuous improvement (soft factors) are key elements in the success
ment of people, process approach, system approach to management, of TQM initiatives. This commitment and involvement of the manage-
continual improvement and factual approach to decision making ment and the people with the results, quality and continuous im-
(Kartha, 2004). Quality Circles (QCs) are small groups of workers who provement must be embodied in the formulating and effective im-
meet regularly on a voluntary basis to discuss problems (not necessarily plementation of a set of strategies policies and actions related to the
quality related) and determine possible solutions. Members of Quality resources, partnerships and processes (hard factors).
Circles are generally given training in quality control and evaluation
techniques (Trott, 2008). QCs improve problem-solving capabilities 2.3. Innovation and quality management
through employees’ participation and team work (Bodas Freitas, 2008).
QCs are soft in nature where groups of workers meet on a voluntary Innovation has been identified as a critical driver of business pro-
basis to try to solve quality related problems. QCs aim to encourage a ductivity and economic growth (Schumpeter, 1934; Romer 1990).
participative culture and would not be considered particularly onerous Schumpeter (1934) argued that innovation involves the transformation
to initiate from a managerial perspective. TQM’s three key elements of of knowledge into new products, services or business processes. The
customer focus, people involvement and continuous improvement relationship between innovation output and innovation inputs has been
combine both soft and hard components with implications for a number explored extensively (Crepon et al., 1998; McCann and Simonen, 2005;
of management practices such as leadership, training, employee-man- Griffith et al., 2008; Roper et al., 2008). Numerous scholars have at-
agement, information and analysis, supplier management, process tempted to explain why some firms are more likely to innovate, with
management, customer focus, and continuous improvement (Moura E firm characteristics, such as size, sector, ownership, and location being
Sá and Abrunhosa, 2007). While many of the principles of Quality identified as influential drivers of innovation output (Audretsch and
Certification are similar to those of TQM, in practice, the programme’s Feldman, 1996; Boschma, 2005; Gordon and McCann, 2005; Jordan
focus is on ensuring that organisations create consistent, stable pro- and O'Leary, 2008; McCann and Simonen, 2005; Tether, 1998; Romer,
cesses through process documentation and adherence, which assures 1990; Roper et al., 2008). The importance of R & D to innovation ac-
the delivery of quality products or services (Pekovic and Galia, 2009). tivity within firms has been established by many authors (Roper et al.,
Quality Certification is the most rule-based, mechanistic or hard of the 2008; Freel, 2003). Firms engaging in R & D increase their existing stock
QIMs lacking the soft elements of either TQM or quality circles. of knowledge resulting in commercial gains from the introduction of
Although, there is no clear consensus as to the impact of QIM in new products, processes and/or organisational innovations (Roper
firms, many scholars conclude that TQM positively affects business et al., 2004). Likewise, managerial capabilities have been highlighted as
performance (Sousa and Voss, 2002; Kaynak, 2003). For instance, an important factor in firm level innovation. Successful innovation re-
Sadikoglu and Zehir (2010) in a comprehensive review of the literature, quires that firms and managers provide clear and consistent signals to
report positive relationships between TQM and business performance, employees about the goals and objectives of the firm (Barnes et al.,
including metrics such as market and financial performance, employee 2006). There is also considerable evidence of the importance of external
performance and customer satisfaction (see Table 1, p. 16). As with sources to innovation outputs (Mansury and Love, 2008). These ex-
TQM, there is considerable evidence that ISO certification can deliver ternal sources of knowledge may include linkages with customers,
advantages for the firm, such as quality improvement (Douglas et al., suppliers, competitors and/or research institutes (Roper et al., 2008)
2003), sales growth (Terlaak and King, 2006), business performance It has long been recognised that innovation in processes is necessary
(Terziovski et al., 2003), financial performance (Corbett et al., 2005), when a company wants to increase productivity (Martínez-Costa and
and firm productivity (Diaye et al., 2009). However, critics of ISO 9000 Martínez-Lorente, 2008), implying a potential link between innovation
have claimed that the adoption of ISO9000 is costly and time-con- and quality management. We review existing literature on the quality-
suming, and is particularly difficult for small firms (Pekovic and Galia, innovation relationship in the context of quality management’s hard
2009). While there is limited evidence of the influence of QCs on firm and soft dimensions, although it should be noted that many of the
performance, there is evidence that human resource management studies empirically examine a specific QIM and its impact on innovation
practices, such as QCs, which empower and involve employees posi- outcomes rather than considering the soft and/or hard aspects of the
tively, influence employee motivation and behaviour with positive approach. A notable recent exception is Zeng et al. (2015). See Table 1
consequences for firm performance (Subramony, 2009). for a summary of this review.
In an investigation of the key hard and soft quality management Focusing first on the hard QIMs and innovation, a small number of
practices adopted by Australian small and medium enterprises (SMEs), studies examine the links between Quality Certification (ISO9000) and
Gadenne and Sharma (2009) report improved overall performance innovation. Benner and Tushman (2002) find that the extent of process
appears to be favourably influenced by a combination of hard TQM management activities in a firm are associated with an increase in ex-
factors such as benchmarking and quality measurement, continuous ploitative innovations and exploitative innovation’s share of total in-
improvement, and efficiency improvement; and the soft TQM factors novations in the paint and photography industry. Using two French
consisting of top management philosophy and supplier support em- microeconomic surveys, Pekovic and Galia (2009) also find that ISO
ployee training and increased interaction with employees and custo- 9000 certification is significantly and positively linked to seven out of
mers. They conclude that it is necessary to focus on a combination of nine innovation indicators. A more recent study reports that ISO900
soft or behavioural aspects and the hard ‘system oriented’ aspects of certification stifles product innovation performance, but facilitates
quality management to achieve an improvement in overall perfor- process innovation performance (Terziovski and Guerrero, 2014). In
mance, and that to maintain customer satisfaction and return on assets addition, Zeng et al. (2015) reports the importance of hard quality
it is just as important to focus on employee involvement and training, as management for innovation.
it is to have a customer focus. In a more recent study, Calvo-Mora et al. In relation to the relationship between soft quality management and
(2013) examine the relationship between soft and hard TQM factors innovation, we are not aware of any quantitative studies which speci-
and key business results for 116 Spanish firms. They report that fically relate Quality Circles to innovation. However, small group pro-
blem solving, employee suggestion schemes and employee training,
have been shown to encourage joint problem solving, empower em-
1
ISO 9000 certification is undertaken by various certification bodies called registrars
ployees and encourage them to make suggestions for improvements to
such as government laboratories, private testing organisations, early adopters of ISO, processes and update employees’ skills and knowledge. Within the HRM
industry trade groups and accounting firms. and quality management literature, there is evidence of such soft

1507
 J. Bourke, S. Roper

Table 1
Summary of Empirical QIM-Innovation Studies.

Study QIM Hard/Soft Dimensions Innovation Measures Principal empirical results: QIM-Innovation
relationship

1 Benner and Tushman ISO9000 Hard QIM dimensions Exploitative innovations; Exploitative Innovation’s share of total Positive for both innovation outcome measures
(2002) only innovations
2 Pekovic and Galia ISO9000 & other certifications Hard QIM dimensions New or improved products for the firm; Turnover due to new or Positive for seven of nine innovation indicators
(2009) only improved products; New or improved products on the market; Share
of new or improved products to the market; Effect on processes; New
or improved processes for the firm; Technologically new process;
New process (non-technological); Effect on innovation activities;
Total innovation expenditure
3 Terziovski and Guerrero ISO9000 Hard QIM dimensions Product innovation performance; process innovation performance Negative for product innovation performance;
(2014) only positive for process innovation performance
4 Moura E Sá and TQM Combined Hard & Soft Adoption of technological innovation A positive (weak) relationship for technological
Abrunhosa (2007) QIM dimensions innovation
5 Martínez-Costa and TQM Combined Hard & Soft Product and process innovation A positive relationship between TQM and product
Martínez-Lorente QIM dimensions and process innovation
(2008)
6 Prajogo and Hong TQM Combined Hard & Soft R&D TQM positively influences R & D
(2008) QIM dimensions

1508
7 Hung et al. (2011) TQM Combined Hard & Soft Product, process and organisational innovation A positive relationship between TQM and
QIM dimensions product, process and organisational innovation
8 Abrunhosa and Moura E TQM Distinct Hard & Soft Adoption of technological innovation; Timing of adoption of Soft elements of TQM positively impact
Sá (2008) QIM dimensions innovations innovation
9 Prajogo and Sohal TQM Distinct Hard & Soft Product and process innovation Soft elements of TQM positively impact
(2004) QIM dimensions innovation
10 Hoang et al. (2006) TQM Distinct Hard & Soft Innovation Indicator consisting of (1) new product or new service (2) Both hard and soft TQM practices positively
QIM dimensions Use of new materials or intermediate products. (3) New functional influence firm-level innovation
solution for an existing product or additional service based on an
existing service. (4) New method of production (5) Entering a new
market (6) New source of supply (7) New ways of organizing (re-
arranging the company’s human resource)
11 Perdomo-Ortiz et al. TQM Distinct Hard & Soft Assessment of innovation performance relative to main competitors Some soft TQM practices, such as HRM, positively
(2009a,b) QIM dimensions influence innovation
12 Perdomo-Ortiz et al. TQM Distinct Hard & Soft Business Innovation Capacity A significant, negative interaction effect with
(2006) QIM dimensions respect to the hard elements of TQM with
different dimensions of BIC.
13 Zeng at al. (2015) Process management and Quality information; Small Distinct Hard & Soft Product innovation; measured by speed of new product introduction Hard QIM affects Innovation; Soft QIM has
group problem solving, Employee suggestion, and Task- QIM dimensions and product innovativeness indirect effect on innovation performance
related training for employees through its impact on hard QIM.

Note: We are not aware of any quantitative studies which specifically relate Quality Circles to Innovation.
Research Policy 46 (2017) 1505–1518
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

practices benefitting innovation activity within firms (Nakata and Im, influence of QIMs with respect to product innovation (Martínez-Costa
2010; Laursen and Foss, 2014; Leiponen, 2005; Zeng et al., 2015). and Martínez-Lorente, 2008; Hung et al., 2011, Prajogo and Sohal,
To date, the majority of studies have focused on the relationship 2004; Hoang et al., 2006; Zeng at al., 2015); albeit with incomplete
between TQM and innovation, which is typically found to be positive. evidence as to if it is the hard routinized QIM processes or the softer,
Moura E Sá and Abrunhosa (2007), in an investigation of the Portu- more organic elements of QIM which benefit product innovation.
guese footwear industry report a positive relationship between TQM
and innovation, although the relationship proves relatively weak. 2.4. Hypotheses
Martínez-Costa and Martínez-Lorente (2008), in a study of 451 Spanish
companies, also report a significant and positive relationship between Although previous empirical studies have generally found a positive
TQM and product and process innovation, while Prajogo and Hong link between QIMs (whether embodying hard and/or soft managerial
(2008) find that TQM positively influences R & D in South Korean firms. processes) and innovation, the use of cross-sectional data and analysis
More recently, Hung et al. (2011) examined the impact of TQM and (e.g. structural equation models or correlation analysis) limits their
organisational learning on innovation performance in the high-tech ability to provide causal insights.2 In particular, it has been suggested
industry in Taiwan. They report that TQM has significant and positive that innovation cannot be realised without first adopting effective
effects on organisational learning, and TQM and organisational learning quality practices. Perdomo-Ortiz et al. (2009a,b) state that ‘in general
both have significant and positive effects on innovation performance. business practice first incorporates the concept of quality management
While many of these studies do not explicitly differentiate between the and then gradually integrates innovation’. This argument draws on the
blend of soft and hard elements within TQM with respect to innovation resource-based and dynamic capabilities (RDBC) theory of the firm,
outcomes, there is evidence that particular aspects of TQM influence which suggests that management priorities are path-dependent and that
innovation activity. In relation to the softer or more organic aspects of improving innovation performance requires greater organisational
TQM, Abrunhosa and Moura E Sá (2008) report that communication, complexity than quality management (Perdomo-Ortiz et al., 2009a,b).
supportive people management practices and teamwork positively im- Our review of the literature suggests that there are conflicting argu-
pact on innovation performance, whereas autonomy and consultation ments concerning the relationship between (hard and/or soft) quality
do not. Similarly, Prajogo and Sohal (2004), in an examination of the management and product innovation performance. Arguments for a
impact of TQM on product innovation within Australian firms, con- positive relationship between quality management and innovation
cluded that two elements of TQM – leadership and people management suggest that companies embracing quality in their system and culture
– positively influenced innovation. will provide a fertile environment for innovation (growth) as quality
Although, Hoang et al. (2006) find that both hard and soft TQM embodies principles that are congruent with innovation (Pekovic and
practices positively influence firm-level innovation, and illustrate how Galia, 2009). The motivation of both processes, quality and innovation,
three specific dimensions of TQM, leadership and people management, is meeting customer requirements with the need to make continuous
process and strategic management, and open organisation have a po- improvements. Pekovic and Galia (2009) explain that the philosophy of
sitive impact on the innovation performance of firms in Vietnam. In- quality management is that employees will be more satisfied and pro-
terestingly, Perdomo-Ortiz et al. (2009a,b) find that only the soft ductive if they can contribute their thoughts and ideas to the achieve-
human resource management element of TQM is linked positively to ment of the company goals, suggesting that both processes are inter-
innovation in their study of 105 Spanish industrial firms. They conclude related. Consequently, quality can be seen as creating an environment
that TQM contains a set of best practices related to human resource that encourages innovation. On the other hand, it can be argued that
management that promote better innovation performance. They also quality improvement processes – which may involve mechanistic rou-
consider five other aspects of TQM – management support, information tinisation and standardised business processes – restrict creativity and
for quality, process management, product design, and relations with innovation (Kanter, 1983; Glynn, 1996; Prajogo and Sohal, 2004).
agents- finding no positive link to innovation. In an earlier study, The implication here is that the payoffs from the adoption of any
Perdomo-Ortiz et al. (2006) examine the relationship between TQM and QIM, whether a hard and/or soft QIM, may only occur in the longer
innovation while considering business innovation capacity (BIC) as term. As managerial attention is initially focused on successful adoption
both a moderating and mediating factor. They report limited evidence of the QIM, managers are likely to have less capacity to devote to in-
of a moderating effect. However, they find a significant, negative in- novation with the potential for short-term disruption and negative
teraction effect with respect to the (hard) process management di- product innovation outcomes. An initial negative effect may be as a
mension of TQM with different dimensions of BIC. This suggests that result of attention being drawn away from the development and/or
the emphasis on the control and improvement of processes, in parallel commercialisation of new product innovations, with energies and focus
with management practices of innovation, such as project planning, being consumed in the implementation of a new QIM. In the longer
formulation and assessment, developing new knowledge and skills and term the principles and practices of the quality system are likely to
external cooperation, may have a negative effect on technological in- become embedded in the mindset and behaviour of both management
novation. and shop-floor workers reducing the level of managerial resources ne-
In general, quality improvement and innovation are seen as corre- cessary to maintain quality freeing resources to concentrate on activ-
lated functions within firms, and while quality management typically ities such as product innovation (Perdomo-Ortiz et al., 2009a,b; Prajogo
appears to benefit innovation; that relationship is not always positive. and Sohal, 2003; Pekovic and Galia, 2009). In addition, previous re-
In addition, there may be very different behavioural or organisational search has demonstrated that market response to improved product
mechanisms in play with respect to the hard and/or soft elements of the quality is not instantaneous but occurs over time (Tellis and Johnson,
quality management practices when adopted. There is a growing re- 2007). While product quality is considered a crucial element to com-
cognition across the organisational change literature of the com- petitive advantage (Tellis et al., 2009; Choi and Pucik, 2005; Calantone
plementary relationship between hard and soft practices, whether et al., 1996), Molina-Castillo et al. (2011) highlight the delayed long-
project management, ICT, or HRM, for business performance. It is likely term benefit of improved product quality to performance.3 In view of
that this complementary relationship also exists in terms of hard and
soft quality management and innovation, although we are only aware
2
In fact, many of the studies reviewed highlight the static nature of their analysis as a
of one recent study which takes such an approach. Zeng et al. (2015)
limitation, with calls for dynamic analysis of the quality-innovation relationship
considered how soft quality management can play a supporting or (Perdomo-Ortiz et al., 2009a; Martinez-Costa et al., 2008; Pekovic and Galia, 2009).
mediating role in the hard quality management and innovation re- 3
Discussion of product quality tend to differentiate between extrinsic (external
lationship. In summary, the studies reviewed here report the positive quality) or intrinsic (internal quality) cues as per the Zeithaml (1988) framework.

1509
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

more organic QIMs tend to be more innovative (Santos-Vijande and

Álvarez-González, 2007; Moura E Sá and Abrunhosa, 2007; Abrunhosa
and Moura E Sá, 2008; Hoang et al., 2006; Perdomo-Ortiz et al.,
2009a,b) as the softer elements of QIM favour incremental innovations
(Abrunhosa and Moura E Sá, 2008). In addition, some studies report the
positive influence of the softer dimensions of QIM specifically with
respect to product innovation (Perdomo-Ortiz et al., 2009a,b;
Abrunhosa and Moura E Sá, 2008; Prajogo and Sohal, 2004). On the
other hand, there is evidence that the mechanistic dimensions of hard
QIMs may hinder innovation, in particular, radical innovation (Prajogo
and Sohal, 2004; Benner and Tushman, 2002; Perdomo-Ortiz et al.,
2009a,b). Terziovski and Guerrero (2014), in a cross-sectional study,
report a negative relationship between hard QIMs (ISO9000) and pro-
duct innovation. This is not particularly surprising as the rationality,
efficiency and strict control of tasks required by mechanistic quality
procedures inhibit creativity and improvisation (López-Mielgo et al.,
Fig. 1. Short and longer-term innovation effects of QIM adoption. 2009).
This suggests that the potential disruption effects of introducing
QIM will be strongest with hard QIMs, such as Quality Certification.
this relationship, it would be unwise to expect the adoption of QIMs to
This contrast is illustrated in Fig. 2. Conversely, the soft nature of QCs
instantaneously impact product innovation performance. For example,
would suggest a less severe disruption effect with the benefits of
it may be that the quality systems in place assist firms in making better
adoption of QCs occurring sooner than the more mechanistic QIMs
decisions in relation to the innovations brought to market. In the
(Fig. 2). We expect a smaller disruptive effect from soft QCs due to their
longer-term, when QIMs are ideally well assimilated within firms, such
purely organic nature, whereas hard QIMs such as Quality Certification
systems may help the firm identify the ‘best’ innovations for commer-
may involve significant short-term disruption. In summary:
cialisation. So an increase in product innovation sales may be due to the
commercialisation of ‘better’ innovations; or it may also be that an H2. Soft QIMs will have weaker short-term negative disruption effects on
increased number of innovations are marketed in the long-term due to product innovation performance than hard QIMs
efficiencies from the now established quality systems. We therefore
Strategically, firms do not always adopt an individual QIM in iso-
anticipate that the adoption of a QIM, whether comprising hard or soft
lation. Firms may adopt QIMs sequentially or simultaneously. Indeed, a
dimensions or both, by a plant may result in negative disruption to
crucial element in firms’ strategic decision-making is the identification
product innovation in the short-term but yield longer-term positive
and effective harnessing of complementarities between different man-
benefits. For example, suppose as in Fig. 1 a firm has stable innovation
agerial activities, optimising resource use (Milgrom and Roberts, 1990,
performance of I1, which we might think of as sales of new, innovative
1995).6 Previously, Zeng et al. (2015) report that soft quality man-
products.4 At time A the firm introduces a new QIM which requires the
agement has an indirect effect on innovation performance through its
focus of management and causes disruption to existing systems of
mediating effect on hard quality management. Thus, we might antici-
production or service provision. This disruption reduces the resources
pate the complementary benefits of QIM adoption being strongest when
available for innovation, causing a short term fall in innovative outputs
the quality mechanisms have contrasting attributes, e.g. QCs and
to I2 (Fig. 1). As the new QIM becomes more embedded in the firm’s
Quality Certification (Fig. 2). This suggests:
operating routines the extent of any disruption declines, and by time B
innovative outputs rise again either to or beyond their previous level at H3. Positive complementarities, benefitting product innovation performance,
I3.5 Hypothesis 1 relates to the potential for this short-term disruption will exist where QIMs adopted have contrasting soft and hard dimensions.
and longer-term beneficial effect of QIMs on innovation:
In the innovation literature, discussion of complementarities has
H1a. Adoption of QIMs will lead to a short-term negative, disruptive effect often been related to the benefits of experiential learning. Rosenberg
on product innovation performance. (1972), for example, describes how a firm increases its’ stock of
knowledge based on its previous experience with technologies as
H1b. Adoption of QIMs will lead to longer-term positive beneficial effects on
‘learning-by-using’. Previous studies have also highlighted the benefit
product innovation performance.
to firms of learning-by-using new technology for subsequent adoption
While we might anticipate this temporal profile of benefits from decision-making (Stoneman and Kwon, 1994; Colombo and Mosconi,
each type of QIM, variations may be evident between the soft and hard 1995; McWilliams and Zilbermanfr, 1996; Stoneman and Toivanen,
QIMs. For instance, some studies report that firms that adopt softer or 1997; Arvantis and Hollenstein, 2001). In the same way, the cumulative
learning from earlier QIM adoption should ease the disruptive effects of
subsequent QIM adoption. Fig. 3 depicts a situation in which a firm
(footnote continued)
(represented by the solid line) successively adopts two QIMs, experi-
External quality is based on customers’ perception regarding extrinsic cues like brand,
price, country of origin, or warranty. Whereas internal quality cannot be changed without
encing disruption and beneficial effects in each adoption episode (A1 to
altering the nature of the product itself and can be subjective (e.g. is based on customers’ B1 and A2 to B2). Here, we envisage there will be learning-by-doing
perception of features such as product image or product design) or objective (e.g. eval- resulting in less disruption in the second adoption episode and stronger
uates whether the product performs as it is supposed to, incorporates features customers beneficial effects. Alternatively consider a firm (represented by the
do not expect, or has a low probability of failing) (Molina-Castillo et al., 2011). It is
dotted line) which has no prior QIM adoption experience and is in-
important to note that data limitations mean that we are unable to consider this differ-
entiation in our research. volved in a single adoption episode (A2 to B2). Here, in the absence of
4
We are grateful to an anonymous reviewer for suggesting this type of diagrammatic
illustration.
5 6
It is important to advance our understanding of the effects of time in and for orga- In terms of human resource management (HRM), for example, Laursen and Foss
nisations (Mitchell and James, 2001). While there is a clear conceptual rationale for the (2003) consider complementarity between different HRM practices in terms of their
mechanisms which might drive a negative short-term and positive long-term QIM-product impact on innovation outputs (see also Michie and Sheehan, 2003), while in a more
innovation relationship, neither the empirical or conceptual literatures provide any general context Lhuillery (2000) examines the impact of a range of organisational prac-
consistent insight on the likely duration of these ‘short-term’ and ‘long-term’ effects. tices on the innovation capability of French companies.

1510
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

Fig. 2. Short and longer-term innovation effects of

hard and soft QIM adoption.

Table 2
Sample Descriptives.
Source: Irish Innovation Panel 1994–2008 (waves 2–6).

Variable Name & Description Observations Mean St. Dev.

Innovative Sales from New Products (%) 1358 18.493 23.110

Innovative Sales from New and Imp. Products 1356 31.827 30.971
(%)
Quality Certification (Q Cert) Use 1358 0.683 0.466
Quality Circles (QC)Use 1192 0.150 0.357
Total Quality Management (TQM) Use 1238 0.365 0.482
Current Q Cert Adoption 1358 0.356 0.479
Fig. 3. The benefits of successive QIM adoption. Previous Q Cert Adoption 1358 0.197 0.197
Early Q Cert Adoption 1358 0.130 0.130
Current QC Adoption 1232 0.062 0.364
learning-by-using effects, we anticipate that the disruption effect will be
Previous QC Adoption 1232 0.042 0.201
greater and the beneficial effect less significant. This suggests: Early QC Adoption 1232 0.053 0.224
Current TQM Adoption 1295 0.154 0.361
H4. Learning-by-using effects from prior QIM adoption will reduce Previous TQM Adoption 1295 0.107 0.309
disruption and increase the product innovation benefits of subsequent QIM Early TQM Adoption 1295 0.117 0.322
adoption. In-plant R & D 1358 0.606 0.489
Linkages with Clients 1358 0.345 0.475
Linkages with Suppliers 1358 0.370 0.483
Horizontal Linkages 1358 0.965 1.496
3. Data and methods Employment (Log) 1358 4.110 1.132
Plant Vintage 1358 28.261 28.196
Externally Owned 1358 0.304 0.460
Our empirical analysis is based on the Irish Innovation Panel (IIP) Workforce with Degree (%) 1358 10.354 11.190
which provides data on the innovation activity and QIM adoption of Government Support for Innovation 1358 0.323 0.468
around 1300 manufacturing plants in Ireland and Northern Ireland over Export Sales (%) 1358 26.964 34.665
the period 1994–2008. More specifically, this element of the IIP com-
prises five surveys or waves conducted using similar survey meth-
odologies and common questions. Each of the five surveys covers the innovation activity it also provides information on the timing of
innovation activities of plants with 10 or more employees over a three- adoption of QIMs. Data was collected on the three QIMs identified
year reference period.7 The resulting panel is highly unbalanced re- earlier: Quality Circles, TQM and ISO 9000. Respondents were asked:
flecting non-response in individual surveys but also the opening and ‘Please indicate if you use any of the following production techniques.
closure of plants over the period covered.8 Plants’ innovation activity in Also, please indicate the date when they were first introduced?’ In
the IIP is represented by the standard indicator used in the European terms of the timing of adoption, respondents were asked whether they
Community Innovation Survey: the proportion of plants’ total sales (at had first introduced each QIM in the three year period covered by the
the end of each three-year reference period) derived from products survey, the previous three years, or prior to this.9 For each respondent
newly introduced during the previous three years. This variable has this provides an indication of whether they are using each QIM and an
been widely used as an indicator of plants’ innovation output (Laursen indication of the length of time since it was adopted by the plant. This
and Salter 2006; Roper et al., 2008; Love et al., 2009; Laursen and allows us to capture the short- and longer-term effects of QIM adoption
Salter, 2006), and reflects not only plants’ ability to introduce new from each cross-section (wave) of survey respondents. For example,
products to the market but also their short-term commercial success. around 68% of IIP respondents reported adopting ISO 9000 with 35.6%
Across those elements of the IIP used in the current analysis, 17.4% of adopting it in the previous three years, 19.7% adopting 3–6 years be-
plants’ sales were derived from newly introduced products (Table 2). fore the survey, and 13% earlier than that (Table 1). Quality Circles had
One rather unusual feature of the IIP is that alongside plants’ been adopted by 15% of plants with TQM adopted by just over a third
of plants.
The IIP also provides information on a number of other plant
7
Individual survey response rates were: 1994–96, 32.9%; 1997–99, 32.8%; 2000–02, characteristics which previous studies have linked to innovation
34.1%; 2003–05, 28.7%; 2006–08, 38.0% (Roper et al., 1996; Roper and Hewitt-Dundas,
1998; Roper and Anderson, 2000; Hewitt-Dundas and Roper, 2008).
8
The structure of the estimation sample is as follows: 800 plants are present in one
wave; 201 are present in two waves; 41 are present in three waves; 7 are present in four 9
In what follows we use the term ‘adoption’ to refer to the first use of a QIM by each
waves; and 1 plant is present in all five waves of the panel. plant.

1511
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

outputs (Annex 1). For example, plants’ in-house R & D activities are current (three-year) period and in two previous periods, i.e.
routinely linked to innovation performance in econometric studies with
suggestions that the innovation-R & D relationship reflects both Ii = β0 + β10 QIMt + β11 QIMit − 1 + β12 QIMit − 2 + β2 RDi + β3 FSi + β4 BSi
knowledge creation (Harris and Trainor, 1995) and absorptive capacity + β5 HSi + β6 CONTi + δi (2)
effects (Griffith et al., 2003). 54% of plants were conducting in-house
R & D at the time of the IIP surveys (Table 1). Reflecting recent writing Our first hypothesis suggests that in the short-term the adoption of
on open innovation (Chesbrough 2007; Chesborough 2006) external QIMs might create disruption to plants’ innovation activity (H1a) with
innovation relationships have also been shown to play an important longer term benefits (H1b). Support for H1a requires β10 < 0, with
role in shaping innovation outputs (Oerlemans et al., 1998; Ritala et al., H1b requiring β11 > 0 and/or β12 > 0. H2 relates to the relative size
2013), complementing plants’ internal capabilities (He and Wong, and impact of the alternative QIMs in equation (2).
2012; Cassiman and Veugelers, 2006; Arora and Gambardella, 1990; Our third and fourth hypotheses relates to potential complementa-
Belderbos et al., 2006). Here, we include three separate variables re- rities and learning-by-using effects between QIMs, denoted here QIMA
presenting plants’ external innovation co-operation with customers, and QIMB. If QIMtB− 2 = 1 where a plant is an early adopter of QIMB and
suppliers and other organisations outside the supply chain. Around zero otherwise we estimate:
30.0% of plants reported having innovation cooperation with custo- Ii = β0 + β101 QIMtA *QIMtB− 2 + β111 QIMtA− 1 *QIMtB− 2 + β121 QIMtA− 2 *QIMtB− 2
mers, while 32.7% had backwards innovation cooperation with sup-
+ β102 QIMtA *(1 − QIMtB− 2) + β112 QIMtA− 1 *(1 − QIMtB− 2)
pliers (Table 1). Links outside the supply chain could be with a variety
of different types of organisation (e.g. universities, consultants) and + β122 QIMtA− 2 *(1 − QIMtB− 2)
here we construct a count variable representing the number of types of + β2 RDi + β3 FSi + β4 BSi + β5 HSi + β6 CONTi + δi
partner with which a plant was cooperating. On average, plants were
(3)
cooperating with around 0.8 organisations outside the supply chain
(Table 1). For Hypothesis 3, which reflects the complementary benefits of si-
We also include in the analysis a group of variables which give an multaneous adoption, we anticipate that early adoption of QIMA in
indication of the quality of plants’ in-house knowledge base – e.g. skills, period t-2 will have greater benefits where a plant also adopts QIMB in
plant size, multi-nationality, plant vintage, and whether or not a plant period t-2. Here, we test β121 > β122. For Hypothesis 4 which reflects
is exporting. Skill levels are reflected in the proportion of each plant’s the potential learning-by-using effects from early adoption of QIMB we
workforce which have a degree level qualification to reflect potential test whether β101 > β102 and/or β111 > β112.
labour quality impacts on innovation (Freel 2005; Leiponen, 2005; Our choice of estimation method is dictated largely by the fact that
Freel, 2005) or absorptive capacity. Multi-nationality is included here we are using plant-level data and that our dependent variables are
to reflect the potential for intra-firm knowledge transfer between na- percentages. We therefore make use of tobit estimators on observations
tional markets and plants, while plant vintage is intended to reflect the pooled from the different waves of the IIP. We include in each model a
potential for cumulative accumulation of knowledge capital by older set of sector controls at the 2- digit level, and a series of time dummies
establishments (Klette and Johansen, 1998), or plant life-cycle effects to pick up any secular differences between the waves of the IIP. Our
(Atkeson and Kehoe, 2005). Finally, studies of the impact of publicly estimation sample is restricted both by the structure of the IIP and
funded R & D have, since Griliches (1995), repeatedly suggested that missing values for some survey questions. Specifically, QIM adoption
government support for R & D and innovation can have positive effects data is collected in waves 2–6 of the IIP for plants which reported un-
on innovation activity both by boosting levels of investment (Hewitt- dertaking some process innovation during the previous three years. Of
Dundas and Roper, 2009) and through its positive effect on organisa- the 3918 observations in waves 2–6 of the survey, 1403 plants under-
tional capabilities (Buiseret et al., 1995). Here, we therefore include a took no process innovation, and so no QIM adoption data is available
dummy variable where plants received public support for innovation.10 for these plants. In addition, 295 process innovators provided in-
Our empirical approach focuses on the innovation or knowledge complete data on the timing of QIM adoption and we therefore exclude
production function which represents the process through which plants’ these plants from our estimation leaving a potential estimation sample
knowledge capital is transformed into innovation outputs (Griliches of 2220 firms. This is further reduced by missing values for our de-
1995; Love and Roper, 2001; Laursen and Salter, 2006; Griliches, pendent variable and control variables resulting in an estimation
1995). If Ii is an innovation output indicator for plant i the innovation sample of around 1350 firms. Our estimation sample therefore re-
production function might be summarised in cross-sectional terms as: presents a group of firms which are more strongly innovation-oriented
that the general population of businesses. This is clear if we compare
Ii = β0 + β1 QIMi + β2 RDi + β3 FSi + β4 BSi + β5 HSi + β6 CONTi + δi innovative sales from new products (and improved products) which
(1) averaged 18.5 (31.8)% among our estimation sample compared to 12.8
(19.9)% among non-process innovators which are excluded from the
Where: QIMi denotes plants’ adoption of quality improvement methods, estimation sample (see Table 1).
RDi are plants’ in-house R & D investments, FSi, BSi and HSi are for-
wards, backwards and horizontal knowledge search respectively, and 4. Results
CONTi is a vector of other plant level controls (Annex 1). Typical of
previous cross-sectional studies of the relationship between QIM and Replicating previous cross-sectional studies of the quality-innova-
innovation, a positive association between QIM and innovation would tion relationship, we first undertake a cross-sectional (static) analysis to
here require β1 > 0. Implicit in this formulation – and previous cross- determine whether QIM adoption benefits product innovation perfor-
sectional studies – is the restriction that the date of adoption of each mance (Eq. (1)). As presented in Table 3, TQM adoption positively
QIM has no impact on its effect on innovation. For each observation in correlates with product innovation performance, although neither
our data we have information on when plants adopted each QIM. To Quality Certification nor QC has a positive association with product
test our hypotheses – and inter alia this restriction − we therefor es- innovation performance. In terms of TQM our results reflect those of
timate a version of Eq. (1) explicitly identifying QIM adoption in the previous studies which have also reported the benefits of adopting
TQM, a QIM comprising hard and soft dimensions, for firm innovation
10
Elsewhere we profile the range of public support initiatives for innovation in Ireland
(Moura E Sá and Abrunhosa, 2007; Martínez-Costa and Martínez-
and Northern Ireland over the period covered by the IIP (Meehan, 2000; O'Malley et al., Lorente, 2008; Abrunhosa and Moura E Sá, 2008); and Zeng et al.’s
2008). (2015) study which reported innovation benefits when firms introduce

1512
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

Table 3 Table 4
Static models: Tobit Models of Innovative Sales of New Products. Short-term and longer-term effects: Tobit Models of Innovative Sales of New Products.

Model 1 Model 2 Model 3 Model 1 Model 2 Model 3

Q Cert Use −1.896 Current Q Cert Adoption −4.597***

(−1.401) (−1.643)
QC Use 1.527 Previous Q Cert Adoption −0.936
(−2.069) (−1.820)
TQM Use 4.601*** Early Q Cert Adoption 1.585
(−1.534) (−1.979)
In-plant R & D 5.194*** 6.559*** 5.829*** Current QC Adoption 0.503
(−1.364) (−1.441) (−1.43) (−3.039)
Linkages with Clients 1.621 1.391 1.692 Previous QC Adoption −3.700
(−1.779) (−1.884) (−1.883) (−3.630)
Linkages with Suppliers 4.270** 4.036** 4.343** Early QC Adoption 5.504*
(−1.709) (−1.784) (−1.800) (−3.006)
Horizontal Linkages 0.053 0.311 −0.201 Current TQM Adoption 2.111
(−0.538) (−0.567) (−0.563) (−2.096)
Employment (Log) −0.523 −0.445 −0.899 Previous TQM Adoption 3.932*
(−0.657) (−0.690) (−0.685) (−2.310)
Plant Vintage −0.091*** −0.100*** −0.092*** Early TQM Adoption 7.372***
(−0.022) (−0.024) (−0.024) (−2.084)
Externally Owned 3.987** 3.725** 3.519** In-plant R & D 5.354*** 6.579*** 5.750***
(−1.679) (−1.784) (−1.787) (−1.360) (−1.439) (−1.428)
Workforce with Degree 0.045 0.128** 0.086 Linkages with Clients 1.763 1.552 1.719
(−0.058) (−0.065) (−0.063) (−1.777) (−1.883) (−1.881)
Government Support 4.600*** 3.399** 3.919*** Linkages with Suppliers 4.042** 4.035** 4.260**
(−1.446) (−1.54) (−1.518) (−1.705) (−1.781) (−1.799)
Export Sales 0.028 0.014 0.019 Horizontal Linkages 0.127 0.342 −0.116
(−0.023) (−0.025) (−0.024) (−0.536) (−0.566) (−0.565)
Constant 11.584*** 8.102** 9.402*** Employment (Log) −0.315 −0.447 −0.797
(−3.186) (−3.374) (−3.341) (−0.658) (−0.689) (−0.687)
N 1358 1232 1295 Plant Vintage −0.083*** −0.101*** −0.090***
Chi-squared 160.286 165.573 160.887 (−0.022) (−0.024) (−0.024)
Pseudo − R2 0.014 0.015 0.014 Externally Owned 4.299** 3.743** 3.636**
(−1.676) (−1.782) (−1.786)
Notes: Standard errors in parentheses; ***p < 0.01, **p < 0.05, *p < 0.1. All the fig- Workforce with Degree 0.041 0.123* 0.082
ures in the table are marginal effects generated from Tobit models. All models include (−0.058) (−0.064) (−0.063)
industry and wave dummies. Government Support 4.474*** 3.340** 3.818**
(−1.441) (−1.538) (−1.516)
Export Sales 0.026 0.012 0.021
both hard and soft quality management processes (Zeng et al., 2015). (−0.023) (−0.025) (−0.024)
Our cross-sectional, static results contrast with the limited number of Constant 10.893*** 8.218** 9.364***
previous studies that have reported a positive association between the (−3.185) (−3.370) (−3.335)
hard Quality Certification and innovation (Benner and Tushman, 2002; N 1358 1232 1295
Chi-squared 170.69 169.872 165.252
Pekovic and Galia, 2009).
Pseudo − R2 0.014 0.016 0.014
A limitation of this static approach to the quality-innovation re-
lationship is that empirically it reveals the correlations between quality Notes: Standard errors in parentheses; ***p < 0.01, **p < 0.05, *p < 0.1. All the fig-
systems and innovation, but it does not investigate a causal connection ures in the table are marginal effects generated from Tobit models. All models include
between QIMs and innovation. The QIM coefficients in the static ana- industry and wave dummies.
lysis capture the effects of both current and lagged adoption. Our dis-
aggregated analysis (Eq. (2)) removes this implicit restriction and al- this relationship is not present for plants who adopted QC in the current
lows us to test H1 which envisages a short-term disruption (H1a) and a or previous time periods. QCs are primarily organic in nature and
longer term beneficial effect (H1b) from QIM adoption on product in- therefore their adoption should not cause particular disruption to the
novation performance. Temporal analysis of the impact of early (t-2), plant compared to the more mechanistic QIMs. Practices, such as QCs,
previous (t-1) and current QIM adoption on innovation performance is which empower and involve employees have been shown to positively
presented in Table 4 based on pooled cross-section estimates and uti- influence employee motivation and behaviour (Subramony, 2009), and
lising firms’ responses on the date of their first QIM adoption. In rela- therefore the lagged beneficial effect is as anticipated.
tion to Quality Certification, we see significant disruption effects, and For TQM, there is no significant disruption effect, but positive and
no evidence of longer-term beneficial effects. These significant disrup- significant beneficial effects. Plants that adopt TQM in the previous
tion effects, which reduce innovative sales by 4.6%, may be due to the period realise innovative returns in the current period, and early
formalised, mechanistic and hard nature of Quality Certification. The adopters of TQM realise significantly larger returns, increasing in-
absence of a positive relationship between Quality Certification and novative sales by 7.4% after six years (Table 4). The beneficial effects
product innovation performance in our temporal analysis contrasts with from TQM adoption are not particularly surprising given the strong
results from previous (static) studies (Benner and Tushman, 2002; positive relationship between TQM adoption and product innovation
Pekovic and Galia, 2009). However, this finding is not altogether sur- reported in our static analysis and previous (static) studies (Martínez-
prising as the adoption of Quality Certification concerns the whole Costa and Martínez-Lorente, 2008). Furthermore, the lack of an initial
organisation and involves considerable disruption to fundamental be- disruption effect may be due to the informal, participative or organic
haviour and routinised tasks (Pekovic and Galia, 2009). components of TQM. For instance, previous studies have highlighted
In relation to soft QC, we find no significant disruption effect, but the returns to innovation from softer TQM elements, such as resource
significant longer-term beneficial effects, increasing innovative sales by management, leadership, people management and open organisation
5.5%. Early adoption of QC positively impacts on innovation, although (Perdomo-Ortiz et al., 2009a,b; Hoang et al., 2006; Prajogo and Sohal,
2004). The advantageous temporal profile of TQM may result from its

1513
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

multi-dimensionality nature. We hypothesised that QIM adoption Table 5

would influence plant innovation in terms of a short term disruptive Complementarities and Learning by Using Effects: Tobit Models of Innovative Sales from
New Products.
effect (H1a) and a longer term beneficial effect (H1b). We find support
for H1a as Quality Certification adoption has a significant short-term Innovation Innovation
disruption effect on plant innovation; and we find strong support for Sales Sales
H1b with TQM and QC adoption resulting in longer-term beneficial
Simultaneous QIM Adoption: Complementarities
effects for plant innovation.
Early Q Cert Adoption: Early Q Cert Adoption:
In line with previous studies, we find plant characteristics, such as with/without early QC Adoption with/without early TQM Adoption
R & D, linkages with suppliers, plant vintage, eternally-owned plants
Early Q Cert with early QC 6.560 Early Q Cert with 11.70***
and an educated workforce, strongly impact on plant innovation (Roper early TQM
et al., 2004; Roper et al., 2008; Freel, 2003). In addition, we find that (5.574) (3.204)
government support for innovation strongly and significantly impacts Early Q Cert without early 1.148 Early Q Cert without −0.322
on innovation performance (Love et al., 2011). QC early TQM
(2.090) (2.276)

4.1. Hard versus soft QIM effects Sequential QIM Adoption: Learning by Using Effects
Current & Previous Q Cert Adoption: Current & Previous Q Cert Adoption:
with/without early QC adoption with/without early TQM adoption
Next, we consider our results in the context of H2 which suggests
that soft QIMs will have weaker short-term disruption effects than hard Current Q Cert with early QC 6.245 Current Q Cert with 0.489
early TQM
QIMs. Quality Certification, a hard QIM, results in significant short- (6.019) (4.252)
term disruption for plants’ innovative performance, whereas QCs and Current Q Cert without early −4.017** Current Q Cert −3.438**
TQM, QIMs comprising full or partial soft components, impose none of QC without early TQM
these short-term disruption effects on plant innovation. While there is (1.626) (1.587)
Previous Q Cert with early −2.889 Previous Q Cert with 5.023
no evidence of long-term beneficial effects with Quality Certification,
QC early TQM
we do find long-term beneficial effects for plant innovation in the case (7.985) (4.510)
of QCs and TQM. In addition, the beneficial effects from TQM adoption Previous Q Cert without 0.142 Previous Q Cert −0.367
arise more quickly than in the case of QCs, and the returns from TQM early QC without early TQM
adoption are greater. Our results therefore suggest that the soft or or- (1.899) (1.887)

ganic components of each QIM impacts the temporal profile of this In-plant R & D 5.343*** 5.461***
disruptive –beneficial relationship. We find strong support for H2 as the (1.362) (1.356)
Clients Linkages 1.764 1.631
softer QIMs, QCs and TQM, have no short-term disruption effects on
(1.775) (1.771)
innovation in contrast to the hard Quality Certification. Supplier Linkages 4.363** 4.304**
(1.707) (1.698)
4.2. Complementarity and learning-by-using effects Horz. Linkages −0.00295 0.0329
(0.538) (0.535)
Employment (Log) −0.421 −0.425
In our investigation of complementarities and learning by using (0.657) (0.654)
effects, we attempt to determine if simultaneous and sequential adop- Plant Vintage −0.086*** −0.083***
tion of QIMs benefit product innovation performance. We hypothesise (0.022) (0.022)
that simultaneous QIM adoption may generate positive complementa- Externally Owned 4.177** 4.055**
(1.677) (1.677)
rities increasing the benefits to innovation (H3), and that early adoption
Workforce w Degree 0.0371 0.0358
of one QIM will generate learning-by-using effects increasing the in- (0.0584) (0.0580)
novation benefits of subsequent QIM adoption (H4). Given the finding Govt. Support 4.588*** 4.153***
of a strong disruptive effect of Quality Certification on innovation (as (1.447) (1.441)
reported in Table 4), we focus on this QIM for our complementarity and Export Sales 0.0265 0.0271
(0.0232) (0.0231)
learning by using effect analysis. Constant 11.09*** 10.96***
Complementarities exist if the sum of the benefits of adopting QIMs (3.190) (3.174)
separately is less than the benefit of adopting them simultaneously. Observations 1358 1358
Empirically, we are examining the influence of simultaneous early DV Q Cert Q Cert
conditional on QC TQM
adoption of Quality Certification and softer QIMs (TQM and QC) on
Complementarities 0.88 10.7***
product innovation sales (see Table 5). For instance, in the first model LBU Previous 0.14 1.35
in Table 5 we examine if early Quality Certification adoption and early LBU Current 2.89* 0.83
QC adoption are complementarities for innovation. Specifically, we
include two variables, one which captures the plants that are early
Quality Certification and early QC adopters and another which captures subsequent QIM adoption. The motivation for investigating whether
those that are early Quality Certification adopters but not early QC learning-by-using effects impact on plant innovation is that early
adopters. The insignificant coefficients indicate that these two QIMs are adoption of one QIM creates the potential for learning and hence sub-
not complementarities. The next regression model examines if early sequent adoption of an additional QIM is likely to be less onerous.
Quality Certification and early TQM adoption are complementary. Our Essentially we are examining if early adoption of a soft (QC) or partially
analysis reveals that TQM and Quality Certification are complementary soft (TQM) QIM and subsequent adoption of a hard (quality certifica-
initiatives and the benefit of their simultaneous adoption is greater than tion) QIM benefit innovation. Empirically, we test for learning-by-using
if adopted individually. We find that benefits of TQM adoption are effects by including variables which capture sequential adoption pat-
conditional on the simultaneous adoption of Quality Certification, and terns. For instance, in the first model in Table 5, we examine if early
vice versa. Of particular interest is how simultaneous early adoption of adoption of QC and subsequent quality certification adoption, in both
Quality Certification and TQM offsets the short-term disruptive effects the current (Current QCert* early QC & Current QCert*no early QC) and
of Quality Certification. previous (Previous QCert*early QC & Previous QCert*no early QC) time
Next, we investigate whether early adoption of one QIM generates periods, influence innovative sales. In the next model, we are ex-
learning-by-using effects increasing the innovation benefits of amining if early TQM adoption and subsequent quality certification

1514
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

adoption benefits innovation. Interestingly, early adoption of QC and, subsequent Quality Certification adoption. The results from these ro-
to a lesser extent TQM, offset the disruptive effects of Quality Certifi- bustness tests are therefore similar to those reported for the narrower
cation. Therefore, early adoption of QC generates learning-by-using dependent variable of innovative sales from new products, although we
effects for current Quality Certification adoption; the benefits (although do not find support for H4 here.
still insignificant) of Quality Certification are conditional on prior In a second robustness test we sought to allow for the potential
adoption of QC. endogeneity of the adoption of each of the QIMs, i.e. the possibility that
We previously hypothesised that positive complementarities will be the determinants of adoption may also be the determinants of innova-
strongest where the QIMs adopted have contrasting hard and soft tion outcomes. We estimated two-stage models estimating first a model
components (H3) and suggested that learning-by-using effects from for the probability of adoption and then including the implied Inverse
prior QIM adoption will reduce disruption and increase the product Mills Ratio (IMRs) in Eqs. (1)–(3) (Heckman 1979). For both our main
innovation benefits of subsequent QIM adoption (H4). Our analysis and alternative dependent variables the IMRs proved insignificant with
reveals a complementary relationship between the hard Quality the coefficients of interest also remaining unchanged in sign and sig-
Certification and the soft-hard TQM for plant innovation. We find no nificance.
evidence of a complementary relationship between the hard Quality Finally, we undertake a robustness test to determine if the QIM-
Certification and soft QCs with respect to innovation performance. innovation relationship differs by time period of adoption. The ratio-
Therefore, we find little support for H3 that complementarities are nale for this test is that each of the QIMs may be subject to a diffusion
strongest when QIMs comprise of contrasting hard and soft compo- pattern which sees variation in uptake, peak and decline in adoption
nents. Our primary finding in relation to learning-by-using effects is through time and across the waves of the survey. We partition the three
that early adoption of QCs generates learning-by-using effects for sub- QIM measures using the five waves of the survey and re-run our static
sequent Quality Certification. Therefore, we find support for H4 that estimations (Table 3). Testing the equality of coefficients between
prior QIM adoption will reduce disruption and increase the product waves reveals F-tests which are all insignificant. This indicates that the
innovation benefits of subsequent QIM adoption. Interestingly, learning QIM effects on product innovation performance are not significantly
by using effects exist when a soft QIM is adopted prior to a hard QIM different between survey waves, i.e. through time.
benefitting product innovation performance.
4.4. Discussion
4.3. Robustness tests
Previous cross-sectional studies have suggested a positive correla-
We conducted three robustness tests to evaluate our results further. tion between QIMs and innovation, with a focus on the TQM-innovation
These included running the analysis with an alternative measure of relationship (Moura E Sá and Abrunhosa, 2007; Martínez-Costa and
innovative output; using an alternative estimation approach allowing Martínez-Lorente, 2008; Abrunhosa and Moura E Sá, 2008). In cross-
for the potential endogeneity of the ‘treatment’ represented by plants’ sectional terms our data also suggests a positive relationship between
QIM adoption (Maddala, 1983); and determining if our results are TQM and product innovation performance although we find no re-
sensitive to the time period of adoption. First, in our main analysis we lationship, however, between either QC or Quality Certification and
use a dependent variable which reflects plants’ sales derived from new product innovation performance (Table 3). As our analysis of the short
products. This reflects an emphasis on more radical innovation rather and long-term effects suggests, however, these cross-sectional re-
than either imitation or more incremental product change (Schnaars, lationships hide some rather complex temporal effects on product in-
1994). To consider whether our results also hold for more imitative novation performance, effects which differ markedly between QIMs
strategies we repeated the analysis using an alternative and more (Table 4). In particular, we find evidence that QIMs can cause short-
broadly defined dependent variable – innovative sales from new and term disruption to product innovation activity before the development
improved products. Results were broadly similar to those reported in of longer-term benefits.
relation to our main dependent variable. In relation to the static ana- In considering these results, it is important to acknowledge that due
lysis, and reflecting the weak results of the static analysis in Table 3, to survey limitations our analysis is based on a sub-sample of plants
none of the QIMs have a significant effect on the broader measure of which undertook some process innovation, and which are more
innovative sales. Likewise, temporal analysis with our alternative in- strongly innovation-oriented than the general population of plants.
novation output measure generates results which are broadly similar to Evidenced by higher than average levels of innovative sales, these firms
those reported earlier (Table 4). In relation to Quality Certification, the are also likely to have strong internal resources and routines associated
coefficient signs indicate the same pattern of disruption and long term with innovation such as R & D capacity and engagement with colla-
beneficial effects as for innovative sales from new products but the borative networks. Even for this group of innovation-oriented firms,
short term disruption effect is not statistically significant. For QC, we however, we still find that the adoption of Quality Certification causes a
still see insignificant disruption effects but evidence of longer term significant short-term disruption effect. For other, less innovation-or-
beneficial effects is insignificant. There is clear evidence of long term iented firms in the population, or firms with less experience of im-
beneficial effects from TQM adoption, although these are generally plementing process innovations, we might anticipate even stronger
weaker than for innovative sales. In summary, effects from QIM adop- disruption effects, and potentially weaker longer-term innovation
tion are stronger for innovative sales from new products than the benefits.
broader dependent variable which captures innovative sales from new We can also consider these results in the context of the soft-hard
and improved products. dimensions of the QIMs. For instance, these disruption effects are most
We also repeated the learning-by-using and complementarity ana- significant for hard Quality Certification, reducing innovative sales by
lysis using innovative sales from new and improved products as the 4.6%, but prove weaker for soft QIM, such as Quality Circles. TQM –
dependent variable. Results from this robustness test are broadly si- which combines soft and hard components also has no significant dis-
milar to those reported in relation to the narrower dependent variable ruption effect. Those QIMs with soft components – TQM and QC –have
of innovative sales from new products (Table 5). The strong and sig- the most significant long-term benefits increasing innovative sales by
nificant complementary benefit of early Quality Certification and early 5.5–7.4% (Table 4). Quality Certification has no significant longer-term
TQM is also evident in relation to the broader dependent variables of effect on innovation.
innovative sales from new and improved products. However, there is no Our findings are broadly in line with previous studies which have
evidence of a learning-by-using effect from early QC adoption influ- examined how the soft and hard dimensions of QIMs influence in-
encing the benefits to this broader definition of innovation from novation. Some studies report that firms that adopt soft QIMs tend to be

1515
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

more innovative (Santos-Vijande and Álvarez-González, 2007; Moura E performance.

Sá and Abrunhosa, 2007; Abrunhosa and Moura E Sá, 2008; Hoang Two main managerial implications follow from our analysis.12 First,
et al., 2006; Perdomo-Ortiz et al., 2009a,b) as the organic elements of it is clear that the adoption of QIMs has significant implications for
QIM favour incremental innovations (Abrunhosa and Moura E Sá, plants’ product innovation outputs, albeit with some time lags as in-
2008). The weaker disruption effects reported for soft QCs and the more ternal routines are optimised. Quality improvement strategies and im-
significant disruption effect for hard Quality Certification is therefore plementation plans need therefore to consider their innovation im-
not particularly surprising. The longer-term beneficial effects are plications and any consequent impact on firm performance. Secondly,
strongest for TQM which comprises both soft and hard components, and the synergies we identify between QIMs suggest the value in strategies
it is likely that the beneficial effects owing to its organic components which maximise complementarities and gains from learning-by-using.
may come into effect sooner and off-set the short term disruptive effect In particular, we find that the early adoption of Quality Circles – a
caused by its more mechanistic components. relatively straightforward and low cost QIM – significantly enhances
Two implications follow from our analysis of short-term and long- the value of Quality Certification where these are adopted subse-
term effects. First, our analysis suggests the potential trade-off between quently. It may be, for example, that the adoption of a soft QIM is
the short-term disruption and longer term product innovation benefits stimulating an initial focus or interest in quality improvement in the
which result from the adoption of QIMs. Benefitting from the adoption firm which is then formalised in the adoption of a hard QIM. QC or soft
of QIMs takes some considerable time as firms revise and optimise or- QIM adoption may also be helping firms to overcome attitudinal bar-
ganisational routines. Second, the relationship between QIMs and riers related to change and the implementation of more formal quality
product innovation performance differs markedly between those QIMs management systems. Our results, which highlight the temporal nature
which have a strong organic component (i.e. TQM, QCs) and more of the relationship between QIMs and product innovation performance
mechanistic initiatives such as Quality Certification: hard QIMs have and between QIMs themselves, also emphasise the limitations of ana-
negative product innovation effects, while QIMs with a soft dimension lyses based on cross-sectional data. In particular, cross-sectoral analyses
have strong long-term product innovation benefits.11 inevitably see QIM adoption − as a uniform treatment, obscuring op-
These contrasts between the implications of alternative QIMs are posing short-term and long-term effects and/or interactions between
also reflected in our results on the benefits of combinations of QIMs. For QIMs.
example, complementarities between QIMs adopted at the same time Our analysis suffers from some limitations. First, our analysis fo-
prove strongest between Quality Certification and TQM which share cuses on Irish manufacturing plants only and may therefore be influ-
hard dimensions, although TQM also has soft components. Conversely, enced by specific national circumstances. The 1994–2008 period con-
we find no evidence of complementarities between contrasting QIMs sidered here, however, was a period of rapidly changing institutions in
such as QC and Quality Certification. Quality Circles do, however, Ireland as well as marked changes in the nation’s economic fortunes –
generate significant learning-by-using effects, enhancing the innovation the Irish recovery of the late 1990s, the 2000–02 high-tech crash, and a
benefits of both Quality Certification. The implication is that adoption period of rapid subsequent growth. Second, due to the structure of our
of a soft QIM such as Quality Circles may have a dual advantage for survey data our analysis is restricted to plants which introduced process
innovation: a direct longer-term benefit and also an indirect longer- innovations. This group are likely to more innovation-orientated than
term benefit through its impact on enhancing the effects of hard QIMs. other non-innovating firms, and this needs to be borne in mind in
considering our estimation results. Third, unlike some other – albeit
static – studies we are unable to identify separately those elements of
5. Conclusion each QIM linked to changes in human resource management, quality
management etc. This limits our ability to investigate the relationships
Our paper contributes to the research literature on quality and in- between different dimensions of quality management and innovation
novation highlighting the short-term disruptive and longer-term bene- and between different elements of QIMs. Both are areas in which future
ficial effects of individual QIMs on product innovation performance. It research would be valuable. Finally, our analysis is based on survey
also highlights the role of complementarities and learning-by-using data and a number of our key variables of interest and controls are
effects in shaping the quality–innovation relationship. therefore categorical in nature. Exploring alternative data sources
Our empirical analysis of the quality-innovation relationship reveals which may provide a greater range of continuous variables would allow
complex temporal effects not evident from previous studies. The IIP a richer contextualisation of the QIM-innovation relationship. Our
data allows us to establish the temporal profile of the – short-term analysis also suggests a number of potential avenues for future research.
disruptive and longer-term beneficial – effects of QIMs on product in- First, the focus of the current paper is the relationship between QIM
novation performance (H1a & H1b). Of particular interest is our finding adoption and product innovation success. Future studies could also
that soft QIMs have weaker short-term disruption effects than hard usefully consider the impact of QIM adoption on process efficiency
QIMs on product innovation performance (H2). reflected in broader measures of productivity, unit cost or resource
Our examination of the benefits of combinations of QIMs supports costs. In addition, the spatial and temporal contrasts in the adoption of
our hypotheses that simultaneous and sequential QIM adoption gen- QIMs is itself an interesting avenue for future research. Finally, we have
erates positive complementarities and learning-by-using effects re- not considered here the potential for firms’ other characteristics (e.g.
spectively enhancing innovation benefits. However, we find little sup- skills, levels of prior innovation) to moderate the QIM-innovation re-
port for H3 that complementarities are strongest when QIMs comprise lationship. This too might be interesting to consider.
of contrasting hard (quality certification) and soft components (QCs);
although we do find a complementary relationship between the hard
Quality Certification and the soft-hard TQM for plant innovation. In Acknowledgements
addition, we also found significant learning-by-using effects for sub-
sequent Quality Certification were generated from early adoption of Useful comments on an earlier draft were received from Will White,
QCs indicating that learning by using effects exist when a soft QIM is Innovate UK and participants in DRUID 2015, and the Irish Economics
adopted prior to a hard QIM benefitting product innovation
12
The pseudo-R2 for the econometric models are quite low (see Table 3). Therefore we
11
This is not to say, of course, that adopting hard QIMs such as Quality Certification would advise caution in being overly explicit with respect to managerial implications,
has no positive effects on wider business performance. Simply, that it has negative effects particularly as this is the first empirical investigation of the short- and longer-term effect
on product innovation performance. of quality management on innovation.

1516
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

Association Conference in May 2015. We are also grateful for significantly improved the paper.
thoughtful comments from the editor and three referees which

Annex 1. Variable Definitions

Innovation

Innovative sales (new) (% An indicator representing the percentage of plants’ sales at the time of the survey accounted for by products which
sales) had been newly introduced over the previous three years.
QIM variables
Current user Static binary measure taking value 1 if the plant had adopted the QIM at the time of the survey and zero otherwise.
Current adopter Time specific binary variable taking value 1 if the plant had first introduced the QIM in the previous three years and
zero otherwise.
Early adopter Time specific binary variable taking value 1 if the plant had first introduced the QIM 4–6 years prior to the survey
date and zero otherwise.
Previous adopter Time specific binary variable taking value 1 if the plant had introduced the QIM more than 6 years prior to the survey
date and zero otherwise.
Control variables
In plant R & D A binary indictor taking value one if the plant has an in-house R & D capacity
Clients Linkages A binary indicator taking value one if the plant is co-operating with customers as part of its innovation activity.
Supplier Linkages A binary indicator taking value one if the plant is co-operating with suppliers as part of its innovation activity.
Horiz. Linkages A count indicator of the breadth of plants’ other innovation partnering activity. Takes values 0–7 depending on how
many different types of partner the plant is working with: consultant, competitor, joint venture, government
laboratory, university, private laboratory, industry research centre.
Employment Employment at the time of the survey.
Plant vintage The time in years between the establishment date of the plant and the date of the survey.
Externally owned A binary indicator taking value 1 if the plant was externally-owned. For plants in Ireland this means owned outside
the country. For plants in Northern Ireland this means plants owned outside the region.
Workforce with degree The percentage of the workforce with a degree or equivalent qualification
Government support A binary indicator taking value one if the plant had received government support for product innovation over the
previous three years.
Export sales A binary indicator taking value one if the plant was selling outside the UK and Ireland at the time of the survey and
zero otherwise

References success: a cross-national comparison. Market. Sci. 15, 341–358.

Calvo-Mora, A., Picón, A., Ruiz, C., Cauzo, L., 2013. The relationships between soft-hard
TQM factors and key business results. Int. J. Oper. Prod. Manag. 34, 115–143.
Abrunhosa, A., Moura E Sá, P., 2008. Are TQM principles supporting innovation in the Cassiman, B., Veugelers, R., 2006. In search of complementarity in innovation strategy:
Portuguese footwear industry? Technovation 28, 208–221. internal R & D and external knowledge acquisition. Manag. Sci. 52, 68–82.
Adam, E.E., Flores, B.E., Macias, A., 2001. Quality improvement practices and the effect Chesborough, H.W., 2006. Open Innovation: a New Paradign for Understanding Industrial
on manufacturing firm performance: evidence from Mexico and the USA. Int. J. Prod. Innovation. Oxford University Press, Oxford.
Res. 39, 43–63. Chesbrough, H.W., 2007. Why companies should have open business models. MIT Sloan
Anderson, J.C., Rungtusanatham, M., Schroeder, R.G., 1994. A theory of quality man- Manag. Rev. 48, 22.
agment underlying the deming management method. Acad. Manag. Rev. 19, Choi, H., Pucik, V., 2005. Relationship between innovativeness quality, growth, profit-
472–509. ability, and market value. Strateg. Manag. J. 26, 555–575.
Arora, A., Gambardella, A., 1990. Complementarity and external linkages: the strategies Colombo, M.G., Mosconi, R., 1995. Complementarity and cumulative learning effects in
of the large firms in biotechnology. J. Ind. Econ. 38, 361–379. the early diffusion of multiple technologies. J. Ind. Econ. 43, 13–48.
Arvanitis, S., Loukis, E., Diamantopoulou, V., 2013. The effect of soft ICT capital on in- Corbett, C.J., Montes-Sancho, M.J., Kirsch, D.A., 2005. The financial Impact of ISO 9000
novation performance of Greek firms. J. Enterp. Inf. Manag. 26, 679–701. Certification in the US: an empirical analysis. Manag. Sci. 51, 1046–1059.
Atkeson, A., Kehoe, P., 2005. Modelling and measuring organization capital. J. Polit. Crawford, L., Pollack, J., 2004. Hard and soft projects: a framework for analysis. Int. J.
Econ. 113, 1026–1053. Project Manag. 22, 645–653.
Audretsch, D.B., Feldman, M.P., 1996. R & D spillovers and the geography of innovation Crawford, L., Costello, K., Pollack, J., Bentley, L., 2003. Managing soft change projects in
and production. Am. Econ. Rev. 86, 630–640. the public sector. Int. J. Project Manag. 21, 443–448.
Barnes, J.W., Jackson, D.W.J., Hutt, M.D., Kumar, A., 2006. The role of culture strength in Crepon, B., Duguet, E., Mairessec, J., 1998. Research, innovation and Productivity: an
shaping sales force outcomes. J. Pers. Sell. Sales Manag. 26, 255–270. econometric analysis at the firm level. Econ. Innov. New Technol. 7, 115–158.
Baumol, W.J., 2002. The Free Market Innovation Machine. Princeton University Press, Deming, W.E., 1986. Out of the Crisis. MIT Press, Cambridge.
Princeton. Diaye, M.-A., Greenan, N., Pekovic, S., 2009. ISO 9000 Norm as a Club Good: Network
Belderbos, R., Carree, M., Lokshin, B., 2006. Complementarity in R & D cooperation Effect Evidence from French Employer Survey. mimeo.
strategies. Rev. Ind. Org. 28, 401–426. Douglas, A., Coleman, S., Oddy, R., 2003. The case for ISO 9000. TQM Mag. 15, 316–324.
Benner, M., Tushman, M., 2002. Process management and technological innovation: a Foley, K., Barton, R., Busteed, K., Hulbert, J., Sprouster, J., 1997. Quality, Productivity
longitudinal study of the photography and paint industries. Adm. Sci. Q. 47, and Competitiveness: The Role of Quality in Australia’s Social and Economic
676–706. Development. Standards Australia, Sydney.
Black, S.A., Porter, L.J., 1996. Identification of the critical factors of TQM. Decis. Sci. 27, Freel, M.S., 2003. Sectoral patterns of small firm innovation: networking and proximity.
1–21. Res. Policy 32, 751–770.
Bodas Freitas, I.M., 2008. Sources of differences in the pattern of adoption of organiza- Freel, M.S., 2005. Patterns of Innovation and skills in small firms. Technovation 25,
tional and managerial innovations from early to late 1990, in the UK. Res. Policy 37, 123–134.
131–148. Gadenne, D., Sharma, B., 2009. An investigation of the hard and soft quality management
Boschma, R., 2005. Proximity and innovation: a critical assessment. Reg. Stud. 39, 61–74. factors of Australian SMEs and their association with firm performance. Int. J. Qual.
Buiseret, T., Cameron, H.M., Georgiou, L., 1995. What differences does it make? Reliab. Manag. 26, 865–880.
Additionality in the public support of R & D in large firms. Int. J. Technol. Manag. 10, Glynn, M.A., 1996. Innovative genius: a framework for relating individual and organi-
587–600. zational intelligences to innovation. Acad. Manag. Rev. 21, 1081–1111.
Calantone, R.J., Schmidt, J.B., Song, X.M., 1996. Controllable factors of new product Gordon, N.R., McCann, P., 2005. Innovation, agglomeration, and regional development.

1517
J. Bourke, S. Roper Research Policy 46 (2017) 1505–1518

J. Econ. Geogr. 5 (5), 523–543. Nakata, C., Im, S., 2010. Spurring cross-functional integration for higher new product
Griffith, R., Redding, S., Van Reenan, J., 2003. R & D and absorptive capacity: theory and performance: a group effectiveness perspective. J. Prod. Innov. Manag. 27, 554–571.
empirical evidence. Scand. J. Econ. 105, 99–118. Nowak, A., 1997. Strategic relationship between quality management and product in-
Griffith, R., Huergo, E., Mairesse, J., Peters, B., 2008. Innovation and productivity across novation. Mid-Atlantic J. Bus. 33, 119–135.
four European countries. Oxf. Rev. Econ. Policy 22, 483–498. O'Malley, E., Roper, J., Hewitt-Dundas, N., 2008. High growth and innovation with low
Griliches, Z., 1995. R & D and Productivity: Econometric Results and Measurement Issues. R & D: The case of Ireland. In: Edquist, C., Hommen, L. (Eds.), Small Economy
Blackwell, Oxford. Innovation System; Comparing Globalization, Change and Policy in Asia and Europe.
Harris, R.I.D., Trainor, M., 1995. Innovation and R & D in Northern Ireland manu- Elgar.
facturing: a Schumpeterian approach. Reg. Stud. 29, 593–604. Oerlemans, L., Meeus, M., Boekema, F., 1998. Do networks matter for innovation? The
He, Z.L., Wong, P.K., 2012. Reaching out and reaching within: a study of the relationship usefulness of the economic network approach in analysing innovation. Tijdschrift
between innovation collaboration and innovation performance. Ind. Innov. 19, voor Economische en Sociale Gcografie 89, 298–309.
539–561. Pekovic, S., Galia, F., 2009. From quality to innovation: evidence from two french em-
Heckman, J., 1979. Sample Selection Bias as a specification error. Econometrica 47, ployer surveys. Technovation 29, 829–842.
153–161. Perdomo-Ortiz, J., González-Benito, J., Galende, J., 2006. Total quality management as a
Hewitt-Dundas, N., Roper, S., 2008. Ireland’s innovation performance: 1991–2005. In: forerunner of business innovation capability. Technovation 26, 1170–1185.
Quarterly Economic Commentary. ESRI, Dublin. pp. 46–68. Perdomo-Ortiz, J., Gonzalez-Benito, J., Galende, J., 2009a. The intervening effect of
Hewitt-Dundas, N., Roper, S., 2009. Output additionality of public support for innovation: business innovation capability on the relationship between Total Quality
evidence for Irish manufacturing plants. Eur. Plan. Stud. 18, 107–122. Management and technological innovation. Int. J. Prod. Econ. 47, 5087–5107.
Hoang, D.T., Igel, B., Laosirihongthong, T., 2006. The impact of total quality management Perdomo-Ortiz, J., González-Benito, J., Galende, J., 2009b. An analysis of the relationship
on innovation. Int. J. Qual. Reliab. Manag. 23, 1092–1117. between total quality management-based human resource management practices and
Hung, R.Y.Y., Lien, B.Y.-H., Yang, B., Wu, C.-M., Kuo, Y.-M., 2011. Impact of TQM and innovation. Int. J. Hum. Resour. Manag. 20, 1191–1218.
organizational learning on innovation performance in the high-tech industry. Int. Prajogo, D.I., Hong, S.W., 2008. The effect of TQM on performance in R & D environ-
Bus. Rev. 20, 213–225. ments: a perspective from South Korean firms. Technovation 28, 855–863.
Jenkins, S., Delbridge, R., 2013. Context matters: examining ‘soft’ and ‘hard’ approaches Prajogo, D.I., Sohal, A.S., 2003. The relationship between TQM practices, quality per-
to employee engagement in two workplaces. Int. J. Hum. Resour. Manag. 24, formance, and innovation performance. Int. J. Qual. Reliab. Manag. 20, 901–918.
2670–2691. Prajogo, D.I., Sohal, A.S., 2004. The multidimensionality of TQM practices in determining
Jordan, D., O'Leary, E., 2008. Is irish innovation policy working? Evidence from irish quality and innovation performance—an empirical examination. Technovation 24,
high-technology businesses. J. Stat. Soc. Inquiry Soc. Ireland XXXVII, 1–45. 443–453.
Kanji, G.K., 2002. Measuring Business Excellence. Routeledge, London, UK. Ragsdell, G., 2000. Engineering a paradigm shift? An holistic approach to organisational
Kanter, R.M., 1983. The Change Master: Innovation and Entrepreneuship in the American change management. J. Organ. Change Manag. 13, 104–120.
Corporation. Simon and Schuster, New York. Ritala, P., Henttonen, K., Salojarvi, H., Sainio, L.M., Saarenketo, S., 2013. Gone fishing for
Kartha, C.P., 2004. A comparison of ISO 9000:2000 quality systems standards, QS 9000, knowledge? The effect of strategic orientations on the scope of open knowledge
ISO/TS 16949 and baldridge criteria. TQM Mag. 16, 331–340. search. Baltic J. Manag. 8, 328–348.
Kaynak, H., 2003. The relationship between total quality management practices and their Romer, P.M., 1990. Endogenous technological change. J. Polit. Econ. 98, S71–S102.
effects on firm performance. J. Oper. Manag. 21, 405–435. Roper, S., Anderson, J., 2000. Innovation and E-Commerce – A Cross-Border Comparison
Klette, T.J., Johansen, F., 1998. Accumulation of R & D capital and dynamic firm per- of Irish Manufacturing Plants. Belfast.
formance: a not-so-fixed effect model. Annales de Economie et de Statistique 49-50, Roper, S., Hewitt-Dundas, N., 1998. Innovation, Networks and the Diffusion of
389–419. Manufacturing Best Practice: A Comparison of Northern Ireland and the Republic of
López-Mielgo, N., Montes-Peón, J.M., Vázquez-Ordás, C.J., 2009. Are quality and in- Ireland. NIERC, Belfast.
novation management conflicting activities? Technovation 29, 537–545. Roper, S., Ashcroft, B., Love, J.H., Dunlop, S., Hofmann, H., Vogler-Ludwig, K., 1996.
Laursen, K., Foss, N.J., 2003. New human resource management practices, com- Product Innovation and Development in UK, German and Irish Manufacturing.
plementarities and the impact on innovation performance. Camb. J. Econ. 27, Queen's University of Belfast/University of Strathclyde/ifo Institut.
243–263. Roper, S., Hewitt-Dundas, N., Love, J.H., 2004. An ex Ante evaluation framework for the
Laursen, K., Foss, N., 2014. Human resource management practices and innovation. In: regional benefits of publicly supported R & D projects. Res. Policy 33, 487–509.
Dodgson, M., Gann, D.M., Phillips, N. (Eds.), The Oxford Handbook of Innovation Roper, S., Du, J., Love, J.H., 2008. Modelling the innovation value chain. Res. Policy 37,
Management. Oxford University Press, Oxford. 961–977.
Laursen, K., Salter, A., 2006. Open for Innovation: the role of openness in explaining Rungtusanatham, M., Forza, C., Filippini, R., Anderson, J.C., 1998. A replication study of
innovation performance among UK manufacturing firms. Strateg. Manag. J. 27, a theory of quality management underlying the Deming management method: in-
131–150. sights from an Italian context. J. Oper. Manag. 17, 77–95.
Leiponen, A., 2005. Skills and innovation. Int. J. Ind Organ. 23, 303–323. Sadikoglu, E., Zehir, C., 2010. Investigating the effects of innovation and employee
Love, J.H., Roper, S., 2001. Networking and innovation success: a comparison of UK, performance on the relationship between total quality management practices and
german and irish companies. Res. Policy 30, 643–661. firm performance: an empirical study of Turkish firms. Int. J. Prod. Econ. 127, 13–26.
Love, J.H., Roper, S., Du, J., 2009. Innovation, ownership and profitability. Int. J. Ind. Santos-Vijande, M.L., Álvarez-González, L.I., 2007. Innovativeness and organizational
Org. innovation in total quality oriented firms: the moderating role of market turbulence.
Love, J.H., Roper, S., Bryson, J.R., 2011. Openness, knowledge, innovation and growth in Technovation 27, 514–532.
UK business services. Res. Policy 40, 1438–1452. Schnaars, S.P., 1994. Managing Imitation Strategies. Free Press, New York.
Maddala, G., 1983. Limited Dependent and Qualitative Variables in Econometrics. Schumpeter, J., 1934. The Theory of Economic Development. Harvard University Press,
Cambridge University Press, New York. Cambridge, Mas.
Mansury, M.A., Love, J.H., 2008. Innovation, productivity and growth in US business Sousa, R., Voss, C.A., 2002. Quality management re-visited: a reflective review and
services: a firm-level analysis. Technovation 28, 52–62. agenda for future research. J. Oper. Manag. 20, 91–109.
Martínez-Costa, M., Martínez-Lorente, A.R., 2008. Does quality management foster or Storey, J., 1989. New Perspectives on Human Resource Management. Routledge, London.
hinder innovation? An empirical study of Spanish companies. Total Qual. Manag. Subramony, M., 2009. A meta-analytic investigation of the relationship between HRM
Bus. Excell. 19, 209–221. bundles and firm performance. Hum. Resour. Manag. 48, 745–768.
McAdam, R., 2000. Quality models in an SME context: a critical perspective using a Tellis, G.J., Johnson, J., 2007. The value of quality. Market. Sci. 26, 758–773.
grounded approach. Int. J. Qual. Reliab. Manag. 17, 305–323. Tellis, G.J., Yin, Y., Niraj, R., 2009. Does quality win? Network effects versus quality in
McCann, P., Simonen, J., 2005. Innovation, knowledge spillovers and local labour mar- high-tech markets. J. Market. Res. 46, 135–149.
kets. Papers Reg. Sci. 84, 465–485. Terlaak, A., King, A., 2006. The effect of certification with ISO 9000 Quality Management
Meehan, E., 2000. Britain's irish question: britain's european question? British-Irish re- Standard: a signalling approach. J. Econ. Behav. Organ. 60, 579–602.
lations in the context of the European union and the belfast agreement. Rev. Int. Stud. Terziovski, M., Guerrero, J.-L., 2014. ISO 9000 quality system certification and its impact
26, 83–97. on product and process innovation performance. Int. J. Prod. Econ. 158, 197–207.
Midgley, G., 2000. Systemic Intervention: Philosophy, Methodology and Practice. Terziovski, M., Power, D., Sohal, A., 2003. The longitudinal effects of the ISO 9000
Plenum, New York. certification process on business performance. Eur. J. Oper. Res. 146, 580–595.
Mitchell, T.R., James, L.R., 2001. Building better theory: time and the specification of Tether, B.S., 1998. Small and large firms: sources of unequal innovations? Res. Policy 27,
when things happen. Acad. Manag. Rev. 26 (4), 530–547. 725–745.
Molina-Castillo, F.-J., Munuera-Alemán, J.-L., Calantone, R.J., 2011. Product quality and Todnem By, R., 2005. Organisational change management: a critical review. J. Change
new product performance: the role of network externalities and switching costs. J. Manag. 5, 369–380.
Prod. Innov. Manag. 28, 915–929. Trott, P., 2008. Innovation Management and New Product Development. Pearson
Moran, J.W., Brightman, B.K., 2001. Leading organizational change. Career Dev. Int. 6, Eduction, England.
111–118. Zeithaml, V.A., 1988. Consumer perceptions of price, quality and value: a means-end
Morgan, N.A., Vorhies, D.W., 2001. Product quality alignment and business unit perfor- model and synthesis of evidence. J. Market. Res. 52, 2–22.
mance. J. Prod. Innov. Manag. 18, 396–407. Zeng, J., Anh Phan, C., Matsui, Y., 2015. The impact of hard and soft quality management
Moura, E., Sá, P., Abrunhosa, A., 2007. The role of TQM practices in technological in- on quality and innovation performance: an empirical study. Int. J. Prod. Econ. 162,
novation: the Portuguese footwear industry case. Total Qual. Manag. Bus. Excell. 18, 216–226.
57–66.

1518