Perry

& Wilson:
The
Accord
and Strikes
Australian
Journal of
Labour
Economics,
Vol. 8, No. 2, June 2005, pp 181 - 201

181

The Insider-Outsider Theory:

Some Evidence from Australia
Michael Dobbie *
Economics Department, Macquarie University

Abstract
This paper uses Australian micro-data to test the insider-outsider model. As part
of this, the paper also examines whether the distinction between insiders and
outsiders has more relevance for males or females. Provisional support for the theory
is found. The paper finds that males have more insider power than females. It is
argued that this represents an indirect test in support of Lindbeck and Snowerss
(1988) turnover cost version of the theory. The paper pays particular attention to
specification and estimation problems associated with the research.

1. Introduction
Insider-outsider models have been advanced in recent years to explain a
range of phenomena, principally the persistence of unemployment. This
paper provides an empirical test of this theory using Australian micro-data.
The work reported here can be seen as a contribution to the growing body
of international evidence on the topic. Moreover, to the best of my
knowledge, it represents the only attempt to test this theory with Australian
micro-data.
The paper is organised as follows. Section 2 provides a brief overview of
the insider-outsider theory. Section 3 outlines the data and methodology
employed in the analysis. Section 4 describes the key variables used, as
well as some limitations of those variables. Section 5 discusses an important
problem to do with collinearity in this study. Section 6 discusses a number
of econometric issues, including the estimation procedures employed.
Section 7 describes the sample used in the study, as well as the main sources
of missing observations. The results of the empirical analysis are presented
in section 8. Section 9 contains some concluding comments.

2. Theory
All insider-outsider models share in common the idea that insiders are
highly insulated from competition by outsiders in wage setting. Insiders
are usually employed workers; outsiders are usually the unemployed. The
main implication of this is that wage outcomes, particularly in the aftermath
of negative employment shocks, may prevent a rapid return to the preshock employment level. Three broad approaches can be identified in the
literature.
Address for correspondence: Dr Michael Dobbie, Economics Department, Macquarie
University, NSW, 2109, Australia. Email: mdobbie@efs.mq.edu.au
* I wish to thank Bill Junor, Roger Tonkin and Bruce Chapman for comments on,
and assistance with, previous research efforts that have lead to this paper. I thank
two anonymous referees for helpful comments on the paper. Naturally any remaining
errors and faults are my responsibility.
The Centre for Labour Market Research, 2005

182

Australian Journal of Labour Economics, June 2005

The first approach is most commonly associated with the names of

Blanchard and Summers (1986, 1987). In this approach the ability of insiders
to ignore outsiders in wage setting is simply assumed. This approach is
best seen as an extension of the microeconomic trade union literature. In
that literature the indifference curves of the union are negatively sloped in
real wage employment space. In the union literature, employed and
unemployed union members have equal status. The main innovation in
the Blanchard and Summers model is to assume that unemployed union
members lose membership of the union (the insider group), and become
outsiders.
This emphasis on membership rules creates the possibility of persistence
in unemployment. This is so since unanticipated shocks, which cause a
change in actual employment, will then alter the size of the insider group
in whose interest wages will be set in the next bargaining round.
Unanticipated contractions in employment tend to generate upward wage
movements, since any future level of labour demand has to be divided
between a smaller, and as such more secure, insider group. Of course this
wage behaviour tends to make the original contraction in employment
persistent. The basic Blanchard and Summers model incorporates what
could be called an extreme membership rule. Under this rule insider status
is lost as soon as an insider becomes unemployed. This extreme membership
rule is one of several strong assumptions required to generate hysteresis,
the most severe form of persistence.1
The second approach is associated with the work of Lindbeck and Snower
(1988, 2001).
This approach is also capable, under the right assumptions, of generating
the predictions of the Blanchard and Summers model. The major
contribution of Lindbeck and Snower is to explain the source of insider
power, rather than simply assuming it. They argue that insider power comes
from a range of turnover costs. These costs mean that the firms incumbent
workforce cannot be costlessly exchanged for unemployed outsiders. These
costs include hiring, training and firing costs. Firing costs include direct
costs such as severance pay, but may also include more amorphous
considerations, such as the negative morale impact of turnover on remaining
employees. These turnover costs create a rent to be bargained over, and
therefore, the possibility of wage outcomes that make it unprofitable for
employers to employ outsiders. Models combining turnover costs and
membership rules can explain unemployment persistence.
Lindbeck and Snower also explore asymmetric membership rules in which
insider status is acquired and lost at different rates. The short-term
unemployed for instance, may retain insider status. The newly employed
can be assumed to require several periods of continuous employment before
they are considered insiders. Varying these membership rules can produce
interesting modifications to the predictions of the basic model.
1

See Dobbie (2003, chapter 2) for a detailed discussion of the meaning and implication
of hysteresis and persistence as used in this literature.

Dobbie: The Insider-Outsider Theory: Some Evidence from Australia

183

The final approach is associated with Layard et al. (1991), and identifies the
long-term unemployed as outsiders. There are several strands to this
outsider ineffectiveness hypothesis. Long-term unemployment could
result in skill atrophy, and or, demoralisation. Either or both of these will
diminish the ability of the long-term unemployed to compete in the labour
market. In addition employers may use unemployment duration as a
screening mechanism. Under this scenario employers interpret long-term
unemployment as a signal that a potential employee has already been found
wanting. All these strands of thought lead to the conclusion that the longterm unemployed have little, or no, impact on wage outcomes.
Under these conditions past employment shocks could have a persistent
impact on unemployment. This is so since the actual history of shocks to
employment determines, in part, the current duration composition of
unemployment, and hence the number of long-term unemployed in any
given pool of unemployment. The larger the proportion of total
unemployment that is long-term, all other things being equal, the more
favourable are wage setting conditions for insiders. This may in turn
generate wage outcomes that are inimical to future employment growth,
thereby making the effect of past shocks persistent. See Layard et al. (1991)
for more on this approach.

3. Data and Methodology

Data

This study uses data from two sources. The first data source is the 1995
Australian Workplace Industrial Relations Survey (AWIRS). The second
data source is unpublished Labour Force Survey data provided by the
Australian Bureau of Statistics. The AWIRS data is described and
summarised in Morehead et al. (1997). The primary task of AWIRS was
to provide a comprehensive and statistically reliable database on
workplace relations in Australia (Morehead et al., 1997, p.1).
Workplaces from the Defence industry, as well as those from the Agriculture,
Forestry and Fishing industry are not included in the survey. In addition,
workplaces with less than five employees were excluded from the survey.
AWIRS 1995 also contained a small workplace survey that collected
information on workplace characteristics for workplaces with between 519 employees. However, no information on employees at these small
workplaces was collected. As such these workplaces have been ignored in
the analysis conducted in this paper.
AWIRS 1995 is rich in information about the characteristics of the sampled
workplaces and their employees. This information enables detailed modeling
of each employees human capital and demographic characteristics.
Survey data is subject to some well-known limitations. Dobbie (2003, pp.123125) provides a detailed discussion of these, and how they impact on this
study. It is argued there that careful choice of the variables employed in this
study has reduced to a minimum any negatives associated with survey data.

184

Australian Journal of Labour Economics, June 2005

As discussed below, various unemployment measures are used in this study

to capture insider-outsider influences. These variables are constructed from
unpublished Labour Force Survey data, provided by the Australian Bureau
of Statistics. These data are used to construct a total unemployment rate, a
short-term unemployment rate, and a long-term unemployment proportion
for each state and territory cross-referenced on a metropolitan/nonmetropolitan basis. 2 These unemployment variables are presented in
appendix 1.
Methodology

The empirical analysis is conducted by estimating a cross-section log wage

equation that includes insider-outsider proxy variables among the
regressors. The equation estimated has the following form:
lnWij = + Xij 1 + Zj 2 + ij
i = 1,......,Nj j = 1,......,J
Where
Wij = the wage of worker i at workplace j.

(1)

Xij = a set of employee specific human capital and job characteristics. These
include the following; potential experience, tenure at current workplace,
whether the employee is from a non-English speaking home, whether the
employee is disabled, occupation, education, whether the employee is
casual, whether the employee is on a fixed term contract. These variables
are fully defined in appendix 3 under the heading Individual Variables.
Zj = a set of variables describing the workplace at which each individual is
employed. The variables include the following; workplace size, union
density, occupational composition of the workplace, industry, the degree
of product market competition faced by the workplace, whether the
workplace operates primarily in export markets, whether the workplace
faces competition from imported goods, percentage of the workplace
workforce that is female, a measure of labour intensity, whether the
workplace is foreign or Australian owned. All these variables are fully
defined in appendix 3 under the heading Workplace Variables. Also
included in this vector are the insider-outsider proxy variables. These
variables are discussed in the next section, and are also fully described in
appendix 3 under the heading Insider-Outsider Variables.

ij = a disturbance term.
Nj is the number of observations for workplace j, and J is the number of
workplaces.
The wage equation is estimated separately for males and females. This is a
common practice in the estimation of union wage mark-ups. In addition,
estimating separate equations for males and females allows for an indirect
test of the turnover cost version of the theory. On average females have
2

There is no metropolitan/non-metropolitan split for the Australian Capital Territory

and the Northern Territory.

Dobbie: The Insider-Outsider Theory: Some Evidence from Australia

185

lower labour force attachment than males, and as such, less on the job
training and experience. 3 It is therefore expected that turnover costs
associated with female employees will be lower than for males. Lower
turnover costs should be associated with less insider power, all other things
being equal.

4. Description and Discussion

of Key Variables
The dependent variable is the log hourly wage. The first outsider variable
used in the analysis is the total unemployment rate in the region in which
each worker is employed. If the unemployed are outsiders, then variation
in the total unemployment rate will not have a statistically significant impact
on log hourly wages. The first specification employed to test the insideroutsider model in this analysis, involves adding the total unemployment
rate in the region where the employees workplace is located, to the vector
of workplace characteristics, Zj.
Rather than all unemployed workers being outsiders, it is likely that there
exists a continuum, representing degrees of outsiderness. The degree of
outsiderness would be based on unemployment duration (Lindbeck and
Snower, 1988, pp. 253-254). Flatau et al. (1990), Flatau et al. (1991) and Kenyon
(1990) use this idea to test the insider-outsider distinction using Australian
aggregate data. This study also uses this idea by including the proportion
of total unemployment that is long-term as an outsider variable.4
If the long-term unemployed are outsiders, it is expected that a significant
and positive relationship between insider wage outcomes and the
proportion of total unemployment that is long-term should exist. It is also
expected that the inclusion of a variable to control for the long-term
unemployed proportion might result in the coefficient on the total
unemployment rate taking a negative sign and becoming statistically
significant. The second specification employed to test the insider-outsider
model therefore involves adding the total unemployment rate in the region
where each employees workplace is located, and the long-term
unemployment proportion in the region where each employees workplace
is located, to the vector of workplace characteristics, Zj.
A third specification is employed in the analysis to test an alternative
formulation of the duration of unemployment, and or, membership rules
approach. In some versions of the insider-outsider model, insider status is
not lost the instant a worker becomes unemployed. Rather it is assumed
that an insider needs to be unemployed for a period of time before he or
she becomes an outsider. Under this scenario it might be reasonable to
assume that the short-term unemployed may remain attached to the insider
group, while only those with longer unemployment durations are ignored
in wage setting.
3
This is evident from an examination of the summary statistics in appendix 3. In the
data used in this study males have an average of 20 years experience, while females
have 16 years.
4
The long-term unemployed have experienced 52 weeks or more of continuous
unemployment.

186

Australian Journal of Labour Economics, June 2005

Alternatively, the short-term unemployed may not belong to the insider

group as such. However, they may be actual, or perceived, substitutes for
employed insiders. This could be the case if the short-term unemployed
retain labour market contacts and skills, which longer durations of
unemployment tend to erode. On either account, an increase (decrease) in
the short-term unemployment rate may induce employed insiders to
decrease (increase) wage demands. The third specification employed in the
analysis undertaken here involves the inclusion of a variable in Zj to
represent the short-term unemployment rate in the region where each
employees workplace is located.5
Three potentially important limitations of these unemployment variables
should be noted.
First, it is a premise of this study that insider-outsider influences can
manifest themselves at the regional level. In other words, it is assumed
that regional variation in turnover costs, and or membership rules, and or
the factors underpinning outsider ineffectiveness, can explain, at least in
part, the variation observed in the regional unemployment data used in
this study. This is a strong assumption given the nature of Australian
industrial relations and labour market practices. Is it reasonable, for instance,
to assume that turnover costs or membership rules vary systematically
across regions? Given that this study finds evidence in support of the insideroutsider theory, caution dictates that this should be viewed as provisional
support.6
Second, under the membership rules version of the insider-outsider model,
it is implied that the short-term unemployed may be insiders because they
have recently been part of the insider group, i.e., recently employed. In
reality most of the short-term unemployed have not been recently employed,
but have come from outside the labour force. There is little that can be done
about this except to acknowledge that it introduces the possibility of
measurement error into the estimations.
On the other hand, the duration of unemployment based insider-outsider
model is not subject to this limitation. In this version the short-term
unemployed are insiders, not because they have recently been employed,
but because they have not been unemployed for long. How they managed
to arrive at this state of short-term unemployment (i.e., from employment
or from outside the unemployment pool) is not relevant.
Third, as already noted, the AWIRS data employed in this study only enables
the location of workplaces to be identified by state and territory crossreferenced on metropolitan non-metropolitan basis. This means that the
size and nature of these regional dimensions is potentially problematic since
the impact of unemployment on wages will be constrained to be the same
5

The short-term unemployed have been unemployed for 13 weeks or less.

I am grateful to an anonymous referee for pointing this out. The author is currently
exploring the use of industry and occupational unemployment rates as insideroutsider proxies. Doing so allows additional dimensions of the insider-outsider
distinction to be explored. It may also be more acceptable to assume variation in
factors such as turnover costs and membership rules, at the industry and occupational
level.
6

Dobbie: The Insider-Outsider Theory: Some Evidence from Australia

187

for workers in geographically distinct regions such as Cape York and South
West Queensland. The result of this is the potential introduction of
measurement error, which could bias upwards the standard errors. Apart
from acknowledging the potential problem, there is little that can be done
about this. It is a problem imposed by the nature of the data set.7
Each specification outlined above also includes an insider proxy variable
that is drawn from the AWIRS 1995 data. This variable, called employment
change, is a dummy variable equal to one if employment change at the
workplace in the proceeding twelve months was positive, it equals zero if
employment change was zero or negative. This variable implies a strict
membership rule whereby the insider group at a workplace consists solely
of incumbent employees. It implies that, unlike the case in which the shortterm unemployed retain insider status, insider status is lost as soon as an
insider becomes unemployed. If this strict membership rule holds the
coefficient on employment change should be significant and negative. As
already noted such a strict membership rule is employed by Blanchard and
Summers (1986, 1987), and is one assumption required for hysteresis.
As noted previously, each regression also contains a large number of control
variables. These variables control for a range of individual employee and
workplace characteristics. They are, by and large, standard inclusions in
wage equations of the kind estimated in this study. Dobbie (2003, pp.132140) provides a detailed discussion of these variables, including the reasons
for their inclusion as regressors. Appendix 3 includes a description of these
variables. For efficiency of exposition these control variables are not reported
in this paper. The estimates in relation to these variables are generally as
expected. Full results are available from the author upon request.

5. Regional Dummies, Regional

Unemployment Rates and Collinearity
The earnings regressions include three different unemployment measures.
These are proxies for the insider-outsider influences that are central to this
study. As just discussed these unemployment rates are regional rates. In
fact, there are 14 regions identifiable in AWIRS, one for the metropolitan
area of each state, one for the non-metropolitan area of each state, and one
for each Territory.
Regional influences on wages can be characterised as working in three ways.
First, region specific effects may include such things as housing costs. There
are significant differences in average housing costs across Australia. These
differences may in turn produce systematic wage differences across the
regions in the form of compensating differentials. Lifestyle considerations
between regions could also generate compensating differentials.
Second, the evidence suggests that there are steady state differences in
equilibrium labour market outcomes such as participation and
unemployment rates. These differences may be associated with wage
differences. Debelle and Vickery (1998), for instance, estimate that Tasmania
7
I am grateful to Bruce Chapman for pointing out the second and third of these
limitations in relation to the unemployment variables used in this study.

188

Australian Journal of Labour Economics, June 2005

and South Australia have had higher equilibrium unemployment rates than
the other states and territories, over the past two decades.
Third, cross-section results could reflect the impact of labour market
adjustment. Debelle and Vickery (1998, p.10) show that Western Australia
and Queensland enjoyed strong employment growth over a protracted
period during the 1980s and 1990s. Indeed these two states increased their
share of total employment from 9 to 10, and 14 to 18 per cent respectively
from 1981 to 1997. The other states experienced a steady decline in
employment share over this time. These employment share trends could
generate a positive wage premium associated with the need to attract labour
from states and territories with declining, to those with expanding,
employment opportunities.
These considerations generate a number of econometric issues. First regional
influences on wages which are not directly related to unemployment, but
which may be correlated with unemployment, need to be controlled for.
Failure to do this may result in omitted variables bias. The estimated
coefficients on unemployment will register the impact on earnings of
omitted but correlated regional influences, in addition to the impact of
regional unemployment.
Second, regional dummies need to be included in order to isolate the impact
of permanent and transitory aspects of unemployment on wages. This is
important since the insider-outsider mechanism clearly relates to the impact
of variations in the transitory component of unemployment on wage
formation. Card (1995) has argued that excluding regional dummies from
these kinds of wage equations involves the implicit assumption that wages
respond to these two components of local unemployment with the same
elasticity. Card shows that this assumption is invalid in the USA, and results
in the fact that USA wage curves which omit region dummies, invariably
produce low or even positive elasticities (Card, 1995, p.789).
The rationale for including regional dummies to control for regional fixed
effects in addition to unemployment is clear enough. Including 14 regional
dummies, corresponding to the 14 regions for which unemployment rates
have been defined, along with 14 regional unemployment rates would
however result in perfect collinearity. On the other hand omitting the regional
dummies may result in omitted variables bias. These issues have been dealt
with as follows in this study. Along with 14 regional unemployment rates, 8
regional dummy variables have been included, one for each state and
territory. This eliminates perfect collinearity, enabling the coefficients to be
estimated. It does not of course eliminate collinearity as such.
To gauge the impact of collinearity and omitted variables bias on the results
the following strategy has been adopted. For each model, an a priori
preferred specification is estimated. This specification includes the eight
regional dummies as defined above, along with the relevant regional
unemployment measure(s) used to proxy insider-outsider influences. This
is referred to as the a priori preferred specification since it is the specification
that is most consistent with the relevant underlying theoretical and empirical
knowledge.

Dobbie: The Insider-Outsider Theory: Some Evidence from Australia

189

Each time a model with regional dummy variables is estimated, it is reestimated without those regional dummy variables. This is done as an
attempt to gauge whether the results from the a priori preferred specification
are quantitatively or qualitatively affected by collinearity. If for instance,
both sets of results tell the same qualitative story, this is taken as evidence
that collinearity is not a serious problem. It is possible that the two
specifications could tell different, even conflicting stories. It may not be
possible to conclude whether the difference in the results is due to
collinearity in the a priori preferred specification, or omitted variables bias
in the specification that omits regional dummies.
In an attempt to obtain additional information a third specification is
estimated. Kennedy and Borland (2000, p.789) argue that property values
are a major source of interstate cost of living differences in Australia. Thus,
following Borland and Kennedy the third specification includes the 14
regional unemployment rates and a variable measuring real median house
prices in each state and territory. The house price data used is described in
appendix 2. This specification omits full controls for regional specific fixed
effects, but does control for one potentially major source of regional variation
in earnings.
Two arguments are offered to justify proceeding in this way. First, there is
abundant empirical evidence to support the view that individual wage
outcomes are the result of both the characteristics of individual employees,
and of the workplaces and industries in which they work. It can be argued
that earnings equations that do not control for both employee and workplace
characteristics are compromised by the omission of relevant variables.
AWIRS 1995 provides a unique opportunity with respect to Australian data,
to match the characteristics of workers with the characteristics of their
workplaces. The potential problems that collinearity might cause should
not deter this work.
Second, there exists an extensive literature on the wage curve (for example
see, Blanchflower and Oswald, 1990, 1994 and 1995). An examination of
this literature shows that sensible results can be drawn from single crosssections, notwithstanding the problems that may arise from collinearity
and omitted variables bias. The work of Winter-Ebmer (1996) provides one
example. Dobbie (2003, pp.144-146) contains a detailed review of this aspect
of the wage curve literature.

6. Estimation and other Econometric Issues

In order to estimate equation (1) using ordinary least squares (OLS) it must
be assumed that the random disturbances are independently distributed.
As discussed in Wooden (2001) and Wooden and Bora (1999), this may be
appropriate in a sample in which individuals are selected at random from
the population of employees. But in the case of AWIRS, the employee sample
was not drawn at random from the population of employees. Rather, a
random sample of workplaces was taken, and then the employee sample
was drawn from these workplaces.
A group of individuals from the same workplace are likely to have
characteristics that are more similar, than a group of individuals sampled

190

Australian Journal of Labour Economics, June 2005

from the population at large. This could result from any of a number of
factors. Workers from a given workplace, for instance, could share a slightly
unique workplace culture which impacts on their productivity. Many of
these common characteristics are unlikely to be measurable, and hence
controlled for. As a result they will be registered in the error term with the
result that there will be a correlation in the errors. Greene (1991) shows
that, under the circumstances just described, the error structure is given as
follows:

ij = ij + j
i = 1,.,Nj

(2)
j = 1,..,J

This error term has two components. The first component ij varies
independently across individuals both within and across workplaces. The
second component j varies across workplaces but is constant for workers
within the same workplace. This error structure describes the random effects
model and the efficient estimator is feasible Generalised Least Squares
(Wooden, 2001; Wooden and Bora, 1999; and Greene, 1991).
Given that much of the data in AWIRS 1995 are grouped, it was expected
that heteroscedasticity could be a problem in the regressions. An
examination of residual plots, and the Breusch-Pagan-Godfrey test (neither
reported in this paper) indicate that this is indeed the case. The presence of
heteroscedasticity means that the standard errors from the random effects
regressions are not efficient. An estimator using a procedure such as Whites
(1980) is not available for the random effects model. In an attempt to
compensate for this, OLS regressions with the t-ratios corrected using
Whites (1980) procedure are also estimated and reported. If the results from
the OLS and random effects regressions are qualitatively similar, the
conclusion is drawn that the econometric difficulties just discussed are not
of practical importance. The results suggest that this is the case. The OLS
and random effects estimates, reported in tables 1 and 2 are quantitatively
and qualitatively very similar. It is acknowledged that this discussion of
the random effects model, the problem of heteroscedasticity and the decision
to estimate OLS regressions using Whites (1980) procedure follows Wooden
(2001) and Wooden and Bora (1999).
Measurement error also results from the fact that the employee earnings
variable reported in AWIRS 1995 is grouped into 23 categories. The usual
practice of allocating midpoints to each earnings category has been followed
in this research. The top and bottom pay categories are open-ended. They
have been closed somewhat arbitrarily. Sensitivity tests show that the
findings are not sensitive to the end points chosen.
The work of Moulton (1986) highlights another potential problem with the
estimation of equation (1). This problem relates to the fact that the
regressions include regional unemployment measures, and workplace level
variables as explanatory variables. These explanatory variables are defined
at a higher level of aggregation than the dependent variable. Moulton has
shown that in this situation the t statistics may be biased upwards. In the

Dobbie: The Insider-Outsider Theory: Some Evidence from Australia

191

earnings equations estimated in this paper, this will occur where the
earnings of workers in the same region share some common component of
variation that cannot be completely explained either in terms of measurable
characteristics or the rate of unemployment. In this case there will be a
positive correlation in the error term for workers in the same region
(Kennedy and Borland, 2000, p.784).
One way in which this problem can be overcome is to estimate a cell means
regression. In this estimation the dependent variable and all of the
independent variables are defined as the average values for some common
level of aggregation, e.g., in the current application the average for each of
the 14 regions. On the assumption that unobserved determinants of
individual earnings are uncorrelated across the regions this approach should
generate unbiased estimates of the standard errors. Unfortunately this
approach cannot be used in this research since the number of explanatory
variables is larger than the number of regional cells (Kennedy and Borland,
2000, p. 85).

7. Refining the Sample and Missing

Observations
The original sample consisted of 19155 employees and 2001 workplaces.
Dobbie (2003, pp.125-127) provides a detailed discussion of how the final
sample of workplaces and employees used in this study was arrived at. In
brief, all non-commercial and Public Sector workplaces and employees were
eliminated from the sample. This was done in the belief that the insideroutsider model is not meant to apply to these workers and workplaces.
This decision cost one third of the sample. In the case of some of the variables
used in the study, missing or no response rates where as high as 50 per
cent. Once all cases with missing observations are eliminated, the sample
used in the analysis consists of 4001 employees (1507 females and 2494
males) spread across 444 workplaces. This large loss of observations could
raise questions as to the representativeness of the final sample. Appendix 3
presents the means and standard deviations of variables used in this study.
These means and standard deviations do not indicate any serious cause for
concern.

8. Results
Tables 1 and 2 present the results for the male and female samples. In each
table there are three panels, A, B and C. These present the results for the a
priori preferred specification (panel A), the specification that omits regional
dummies (panel B), and the specification that omits regional dummies, but
which includes regional median house prices (panel C). Each table has six
columns of results. The first two columns (1A and 1B) present the OLS and
random effects estimates for the model that includes the total
unemployment rate as an outsider proxy. Columns 2A and 2B present the
OLS and random effects results for the version that includes the total
unemployment rate and the long-term unemployment proportion. Columns
3A and 3B present the results for the version of the model in which the
short-term unemployment rate appears. For ease of exposition in what
follows, only the OLS estimates will be discussed.

192

Australian Journal of Labour Economics, June 2005

The discussion turns first to the male sample estimates reported in table 1.
The results in column 1A of panel A indicate that the coefficient of the total
unemployment rate is not significantly different from zero. The conclusion
drawn from this is that the unemployed are outsiders. None of the
coefficients on the regional dummies are significantly different from zero
in this specification. The results presented in column 1A of panel B indicate
that this conclusion that the unemployed are outsiders, is robust to the
exclusion of regional dummies. This makes it unlikely that the lack of
statistical significance of the total unemployment rate variable in the results
reported in column 1A of panel A is due to collinearity. The results reported
in column 1A of panel C attempt to partially control for the possibility of
omitted variables bias by including median house prices. Neither this
variable, nor the total unemployment rate, is significant.
The numerical magnitude of the estimated coefficients on the total
unemployment rate reported in panels A, B and C are very small at 0.012,
0.057 and 0.042 respectively. It is reasonable to conclude that in terms of
both size and significance, the results reported in column 1A of panels A, B
and C support the hypothesis that the unemployed are outsiders.
The results reported in column 2A of panel A, are from the specification
that includes the long-term unemployment proportion, the total
unemployment rate and the regional dummies. The coefficients on the longterm unemployment proportion and the total unemployment rate are signed
in line with the predictions of the insider-outsider model. Moreover, they
are statistically significant at the one and five per cent significance levels
respectively. The size of the coefficient on total unemployment increases
from 0.012 to 0.203 (see panel A, columns 1A and 2A) when the long-term
unemployment proportion is added to the model. The elasticity of wages
with respect to the long-term unemployment proportion is 0.39. An elasticity
of this magnitude implies that the outsider effect is economically, as well
as statistically significant.
The addition of the long-term unemployment proportion leads four of the
seven regional dummies to statistical significance. Overall the evidence is
that the long-term unemployed are outsiders. It is also reasonable to
conclude that the insignificance of the total unemployment rate in column
1A is not the result of collinearity between it and the regional dummies.
Rather it reflects the fact that some of these unemployed, the long-term
unemployed, are outsiders.
This conclusion receives additional support from the results reported in
column 3A of panel A. These results show that the short-term unemployed
have a significant and negative impact on hourly wages. Once again this is
consistent with the view that the short-term unemployed remain attached
to the insider group, or that they are potential substitutes for employed
insiders. The magnitude of the estimated coefficient on the short-term
unemployment rate is 0.12. This is ten times the size of the estimated
coefficient on the total unemployment rate reported in column 1A of panel
A. In addition, two of the regional dummies are significant in this regression.

193

Dobbie: The Insider-Outsider Theory: Some Evidence from Australia

Table 1 Males (OLS and Random Effects Estimates. Dependent

Variable is Log Hourly Wage. Number of observations = 2494)
OLS 1 A RE 1B

OLS 2 A RE 2 B

OLS 3A RE 3B

Panel A
Insider-Outsider Variables
Total unemployment rate
Long-term unemployed proportion

-0.012
-0.021 -0.203* -0.307#
(-0.163) (-0.212) (-2.367) (-1.870)
0.390** 0.445*
(3.459) (2.225)

Short-term unemployed rate

Employment change
Region (VIC omitted)
NSW
ACT
TAS
NT
QLD
SA
WA
R-squared
Adjusted R-squared
RESET
Model F
Insider-Outsider Variables
Total unemployment rate
Long-term unemployed proportion

0.014
(1.071)

0.021
(1.195)

0.012
(0.890)

-0.120* -0.236*
(-2.518) (-2.189)
0.015
0.017
0.017
(0.722) (1.275) (0.689)

0.002
(0.146)
0.084
(0.893)
-0.046
(-1.391)
-0.072
(-1.524)
0.020
(0.911)
-0.033
(-1.366)
-0.011
(-0.424)

-0.005
0.003
0.002
0.012
0.028
(-0.234) (0.199) (0.104) (0.677) (0.891)
0.078
0.339** 0.386*
0.107
0.190
(0.896) (2.823) (2.283) (1.156) (1.624)
-0.033
-0.065# -0.104 -0.047
-0.120
(-0.881) (-1.896) (-1.584) (-1.518) (-1.483)
-0.0006 0.109
0.137 -0.039
-0.008
(-1.089) (1.489) (0.397) (-0.830) (-0.021)
-0.0008 0.185** 0.202*
0.054* 0.084#
(-0.034) (3.468) (2.227) (2.100) (1.734)
-0.021
-0.015 -0.007
-0.041# -0.052
(-0.818) (-0.637) (-0.187) (-1.692) (-1.153)
0.014
0.174** 0.169
-0.0004 -0.012
(0.521) (2.839) (1.638) (-0.016) (-0.233)
0.426
0.429
0.428
0.410
0.412
0.411
0.126
0.070
0.113
134.29**
132.34**
133.51**
Panel B

-0.057
-0.087
-0.061
-0.161
(-0.908) (-1.058) (-0.844) (-1.389)
0.003
0.032
(0.101) (0.588)

Short-term unemployed rate

Employment change
R-squared
Adjusted R-squared
RESET
Model F
Insider-Outsider Variables
Total unemployment rate
Long-term unemployed proportion

0.014
(1.051)

0.015
(0.897)
0.425
0.410
0.033
43.93**

0.014
0.016
(1.054) (0.820)
0.425
0.410
0.031
43.42**
Panel C

-0.042
-0.062
-0.021 -0.087
(-0.649) (-0.728) (-0.268) (-0.635)
-0.015 -0.001
(-0.418) (-0.028)

Short-term unemployed rate

Employment change
Regional control
House Prices
R-squared
Adjusted R-squared
RESET
Model F

-0.026
-0.051
(-0.673) (-0.918)
0.012
0.013
(0.941) (0.789)
0.425
0.410
0.025
44.09**

0.015
(1.155)
0.028
(0.959)

0.017
(1.010)

0.042
(1.077)
0.425
0.410
0.035
43.33**

0.018
(0.886)

-0.029
-0.054
(-0.747) (-0.964)
0.015
0.017
(1.120) (0.981)

0.034
0.054
(1.064) (1.012)
0.425
0.410
0.050
42.71**

0.034
0.050
(1.214) (1.355)
0.425
0.410
0.025
43.45**

0.015
(1.161)

Notes: **, *, # indicate significance at the one, five and ten per cent levels respectively. The t-ratios
(in brackets) in the OLS regressions have been corrected for heteroscedastic error structures using
Whites (1980) procedure. Each regression reported in this table contains approximately 50 other
control variables, the coefficient estimates for which have not been reported here.

194

Australian Journal of Labour Economics, June 2005

Table 2 Females (OLS and Random Effects Estimates. Dependent

Variable is Log Hourly Wage. Number of observations = 1507)
OLS 1 A RE 1B

OLS 2 A RE 2 B

OLS 3A RE 3B

Panel A
Insider-Outsider Variables
Total unemployment rate
Long-term unemployed proportion

-0.047
-0.044
-0.131 -0.126
(-0.454) (-0.502) (-1.109) (-1.063)
0.141
0.041
(0.997) (0.386)

Short-term unemployed rate

Employment change
Region (VIC omitted)
NSW
ACT
TAS
NT
QLD
SA
WA
R-squared
Adjusted R-squared
RESET
Model F
Insider-Outsider Variables
Total unemployment rate
Long-term unemployed proportion
Short-term unemployed rate
Employment change
R-squared
Adjusted R-squared
RESET
Model F
Insider-Outsider Variables
Total unemployment rate

-0.019
-0.104
-0.018
0.004
(-1.143) (-0.694) (-1.087) (0.316)

-0.025
-0.013
(-0.370) (-0.197)
-0.019
-0.010
(-1.143) (-0.696)

0.017
0.026
(0.788) (1.431)
-0.116
-0.105
(-1.490) (-1.340)
0.015
0.007
(0.323) (0.140)
0.113
0.14
(0.860) (0.877)
-0.062* -0.044*
(-2.546) (-2.065)
-0.051# -0.029
(-1.771) (-1.100)
-0.021
-0.031
(-0.482) (-1.038)
0.354
0.324
0.001
15.71**

0.022
0.029#
(1.064) (1.680)
-0.101
-0.093
(-1.330) (-1.204)
0.008
0.001
(0.162) (0.029)
0.125
0.156
(0.943) (0.922)
-0.053# -0.039
(-1.860) (-1.465)
-0.054# -0.033
(-1.876) (-1.228)
-0.011
-0.023
(-0.304) (-0.824)
0.354
0.324
0.0007
15.79**

0.016
0.036
(0.724) (1.623)
-0.023 -0.027
(-0.192) (-0.447)
0.005
0.023
(0.115) (0.419)
0.179
0.156
(1.200) (1.048)
-0.004 -0.024
(-0.074) (-0.550)
-0.042 -0.031
(-1.481) (-1.137)
0.041
-0.021
(0.528) (-0.459)
0.355
0.324
0.0007
15.64**
Panel B

-0.051
-0.019
-0.214* -0.224**
(-0.629) (-0.256) (-2.269) (-2.672)
0.146** 0.049
(3.451) (1.248)
-0.091# -0.087
(-1.823) (-1.422)
-0.019
0.0007 -0.018
0.010
-0.017
0.007
(-1.145) (0.048) (-1.124)
(0.685) (-1.059) (0.473)
0.348
0.352
0.349
0.320
0.324
0.321
0.0046
0.001
0.064
16.49**
16.95**
16.81**
Panel C
0.015
(0.179)

Long-term unemployed proportion

-0.017
-0.144 -0.255**
(-0.228) (-1.312) (-2.720)
0.114* 0.145**
(2.419) (3.092)

Short-term unemployed rate

Employment change
Regional control
House Prices
R-squared
Adjusted R-squared
RESET
Model F

-0.020
(-1.204)

0.004
-0.019
(0.314) (-1.155)

0.107** 0.034*
(2.772) (2.073)
0.351
0.323
0.097
16.71**

-0.093# -0.091
(-1.854) (-1.519)
0.002
-0.017
0.005
(0.197) (-1.058) (0.337)

0.056
0.056#
(1.296) (1.866)
0.353
0.325
0.020
16.80**

0.106** 0.078**
(2.903) (2.680)
0.352
0.324
0.026
17.03**

Notes: **, *, # indicate significance at the one, five and ten per cent levels respectively. The t-ratios
(in brackets) in the OLS regressions have been corrected for heteroscedastic error structures using
Whites (1980) procedure. Each regression reported in this table contains approximately 50 other
control variables, the coefficient estimates for which have not been reported here.

Dobbie: The Insider-Outsider Theory: Some Evidence from Australia

195

None of the estimated coefficients on the unemployment variables, reported

in panels B and C, are significantly different from zero. Given that many of
the regional dummies in the regressions reported in panel A are statistically
significant, it is reasonable to conclude that the results reported in panels B
and C suffer from the omission of these regional dummies. The insider
variable employment change is never significant. This suggests the absence
of the extreme membership rule required for hysteresis.
Considering that cross section data is used in this analysis, the values of
the coefficient of determination reported in table 1 are large. Moreover, the
F values support the view that the a priori preferred specification, reported
in panel A, is also statistically preferred. Ramseys RESET statistic suggests
the absence of specification error in any of the regressions reported in Table
1. The results in table 1 give clear support to the idea that the insider-outsider
distinction is relevant for males.
The results for the female sample are reported in table 2. The estimated
coefficients of the unemployment variables reported in the three columns
of panel A are signed in line with the predictions of the insider-outsider
model. However they are all statistically insignificant. As usual, the
possibility that this lack of significance is due to collinearity is examined
by omitting the regional dummy variables and re-estimating the models.
The results reported in column 1A of panels B and C indicate that the total
unemployment rate remains insignificant when regional dummies are
omitted from the model. However, column 2A of panel B shows that the
total unemployment rate, and the long-term unemployment proportion,
are both statistically significant, and the estimated coefficients associated
with them are signed in line with the predictions of the insider-outsider
model. Column 2A of panel C shows that, once median house prices are
included in the model, the total unemployment rate is driven to
insignificance, while the coefficient of the long-term unemployment
proportion remains positively signed, and significant at the five per cent
significance level. In panels B and C the magnitude of the coefficient on the
long-term unemployment proportion is 0.14 and 0.11 respectively. This is
smaller than the corresponding estimates from the male sample. The insider
variable employment change is never significant. This suggests the absence
of the extreme membership rule required for hysteresis.
Considering that cross section data is used in this analysis, the values of
the coefficient of determination presented in table 2 are large. The F statistics
indicate that the models are overall significant. Unlike table 1, the F statistics
reported in table 2 are numerically similar in the results reported in each of
the three panels. This suggests that, from a statistical viewpoint, there is
nothing between the specifications in panels A, B or C. This is consistent
with the absence of a compelling case in favour of the a priori preferred
model in the case of the female sample. Ramsays RESET test suggests the
absence of specification error. In sum, the results in table 2 do provide some
evidence in favour of the insider-outsider model. However, that evidence
does not come principally from the models that employ the a priori preferred

196

Australian Journal of Labour Economics, June 2005

specification. On the contrary the evidence comes from the models that have
omitted the regional dummies. As a result this evidence could be subject to
omitted variables bias. There is no way to decide further on this issue.
In view of this, two alternative approaches are possible. First, the results
for the female sample could be regarded as econometrically unreliable, and
judgement regarding this aspect of the research not attempted. Second, some
tentative conclusions could be drawn on the assumption that the results
from panels B and C are accurate. If the second approach is taken, the
following conclusions from the work reported in this paper seem reasonable.
First, insider power and outsider ineffectiveness appears to be a feature of
male wage outcomes. This conclusion is drawn on the basis of results that
have the a priori preferred specification. This specification is clearly superior
to the alternatives reported in panels B and C of table 1.
Second, there is some evidence that insider-outsider considerations are a
feature of wage outcomes in the female sample. The total unemployment
rate, when entered on its own, is insignificant in all three panels. Moreover,
provided the results in panels B and C are accepted as legitimate, the longterm unemployment proportion and short-term unemployment rate act as
the insider-outsider theory predicts.
Third, the results suggest that males have more insider power than females,
all other things being equal. For males, a one per cent increase in the longterm unemployment proportion is translated into a 0.39 per cent increase
in hourly wages (see table 1, column 2A of panel A). The corresponding
figure for females is either 0.11 or 0.14, depending on which model is used
(see table 2, column 2A of panels B and C). These estimates imply that
employed males can turn an increase in the density of outsiderness into a
larger wage increase than can their employed female counterparts. This is
what would be expected if insider power is explained, either fully or
partially, by turnover costs.

9. Concluding Comments
This paper has tested the insider-outsider theory with Australian microdata. The evidence is consistent with this theory, and especially with the
idea that the long-term unemployed are outsiders in regional labour
markets. The evidence in favour of the theory was strongest in relation to
male wage outcomes. The paper uses regional unemployment rates as
proxies for insider-outsider influences. This involves the assumption that
the Australian institutional setup can support systematic regional variation
in turnover costs, and or membership rules, and or skill atrophy,
demoralisation and screening. This is a strong assumption, and it is,
therefore, prudent to conclude that the results in this paper should be seen
as providing provisional support for the insider-outsider theory.

Dobbie: The Insider-Outsider Theory: Some Evidence from Australia

197

Appendix 1
Regional Unemployment Rates Used in this Study

Region

Total
Unemployment
Rate
(per cent)

Short-Term
Unemployment
Rate
(per cent)

Long-Term
Unemployment
Proportion
(per cent)

9.07
10.76
10.49
9.87
8.58
10.73
11.29
8.07
9.34
7.59
13.25
11.09
8.89
8.83

3.88
4.23
3.68
4.00
4.75
4.86
3.93
2.00
4.74
2.66
5.45
2.78
4.72
4.59

34.63
40.05
38.61
34.89
21.60
26.41
35.38
43.24
22.60
20.46
41.45
44.09
21.78
17.33

NSW Met
NSW Non Met
Victoria Met
Victoria Non Met
Queensland Met
Queensland Non Met
South Australia Met
South Australia Non Met
Western Australia Met
Western Australia Non Met
Tasmania Met
Tasmania Non Met
Northern Territory
Australian Capital Territory

Notes: Short-term unemployment is defined as 13 weeks or less. Long-term

unemployment is 52 weeks or more. The data for these estimates was provided by
the Australian Bureau of Statistics and are based on unpublished Labour Force Survey
(6203.0) data.

Appendix 2
Real Median House Prices, Capital Cities, September 1995
City

$1000s

Sydney
Melbourne
Brisbane
Adelaide
Perth
Hobart
Canberra
Darwin

170.85
118.24
111.67
90.92
107.96
85.28
128.84
135.98

Source: Unpublished data supplied by the Real Estate Institute of Australia Ltd.
Deflated using Australian Bureau of Statistics Consumer Price Index, Catalogue no.
6401.0.

198

Australian Journal of Labour Economics, June 2005

Appendix 3
Variable definitions; Means and Standard Deviations
(Note: The first number in each cell is the mean, the number beneath is the standard
deviation)
Variable
Dependent variable
Log hourly pay

Definition

Male

Log of Gross pay per week divided by hours

2.82
worked each week.
0.40
Insider-outsider variables
Short-term
Log short-term unemployment rate. Regional
1.39
unemployed
unemployed for less than or equal to13 weeks, 0.15
as a percentage of regional labour force.
Total unemployed
Log regional unemployment rate. Regional
2.29
unemployed as a percentage of regional
0.10
labour force.
Long-term
Proportion of regional unemployment that is
3.49
unemployed
long-term. Log of regional long-term
0.22
unemployment, expressed as a percentage of
total unemployment in the region.
Employment change Dummy variable: Equals 1 if employment
0.51
change at workplace was positive in
0.50
proceeding 12 months.
Workplace Variables
Active union
Dummy variable: Equals 1 if the senior delegate 0.56
from the union with most members spends
0.50
one hour or more each week on union activities,
and either a general meeting of members is held
at least once every six months, a union committee
exists and meets regularly with management,
or delegates meet with management at least
once a month.
Unionisation
Proportion of employees at workplace who are 0.63
union members.
0.28
Workplace size
20-49
20-49 employees at workplace.
0.14
(omitted category)
0.34
50-99
50-99 employees at workplace.
0.21
0.41
100-199
100-199 employees at workplace.
0.26
0.44
200-499
200-499 employees at workplace.
0.24
0.43
500-1000
500-1000 employees at workplace.
0.12
0.32
1000+
1000 or more employees at workplace.
0.04
0.04
Labour intensity
Labour costs as a proportion of total costs.
2.35
The variable is measured as a scale going
1.08
from 1 to 6 representing less to more labour
intensive.
Foreign
Dummy variable: Equals 1 if workplace is
0.27
majority foreign owned.
0.44
Import competing
Dummy variable: Equals 1 if workplace
0.45
faces import competition for its major
0.49
product or service.
Export
Dummy variable: Equals 1if more than 50 per
0.15
cent of main product or service is exported.
0.36

Female
2.56
0.36
1.39
0.15
2.29
0.10
3.49
0.22

0.57
0.49

0.36
0.48

0.58
0.31
0.12
0.32
0.20
0.40
0.29
0.45
0.24
0.43
0.11
0.32
0.03
0.17
2.45
1.16

0.16
0.36
0.33
0.47
0.05
0.23

199

Dobbie: The Insider-Outsider Theory: Some Evidence from Australia

Appendix 3 (continued)
Variable definitions; Means and Standard Deviations (continued)
Variable

Definition

Occupational composition of workplace

Managers (per cent) Managers as a percentage of total employment
at workplace.
Professionals
Professionals as a percentage of total
(per cent)
employment at workplace.
Tradespeople
Tradespeople as a percentage of total
(per cent)
employment at workplace.
Clerical (per cent)
Clerical as a percentage of total employment
at workplace.
Salespeople (per cent) Salespeople as a percentage of total
employment at workplace.
Plant and mach
Plant and machine operators as a percentage
operators (per cent)
of total employment at workplace.
Labourers (per cent) Labourers as a percentage of total
employment at workplace.
Para-prof (per cent) Para-professionals as a percentage of total
(omitted category)
employment at workplace.
Competition in
Dummy variable: equals 1 if firm has many
product market
competitors in product market, equals zero if
firm has no competitors.
Per cent Female
Percentage of workplace workforce which
is female.
Individual variables
Experience
Age (years at school + 5)
Tenure

Years employed at workplace.

Non-English
Speaking home
Disabled

Dummy variable: Equals 1 if employee comes

from a non-English speaking home.
Dummy variable: Equals 1 if a health
condition or disability exists which is likely
to last beyond six months.

Occupation
Plant and machine
operators
Sales

Clerks
Tradesperson

Para-prof
Professional
Manager
Other occupation
Labourers (omitted
category)

Dummy variable: Equals 1 if employed in

occupational group, Plant and Machine
Operators and Drivers.
Dummy variable: Equals 1 if employed in
occupational group, Sales and Personal
Services.
Dummy variable: Equals 1 if employed in
occupational group, Clerks.
Dummy variable: Equals 1 if employed in
occupational group, Tradesperson and
Apprentices.
Dummy variable: Equals 1 if employed in
occupational group, Para-professionals.
Dummy variable: Equals 1 if employed
in occupational group, Professionals.
Dummy variable: Equals 1 if employed
in occupational group, Managers.
Dummy variable: Equals 1 if not able to be
classified in the other occupational categories.
Dummy variable: Equals 1 if employed in
occupational group, Labourers and Related
Workers.

Male

Female

5.41
4.62
4.60
7.00
18.91
21.63
11.04
13.71
9.72
21.77
23.80
25.92
19.08
23.21
7.40
12.96
0.53
0.49

5.89
4.84
4.06
9.18
7.11
11.88
13.24
19.22
33.97
37.96
9.93
19.99
17.42
24.06
8.35
17.19
0.65
0.47

22.80
21.17

55.06
24.75

20.63
11.83
7.82
7.76
0.06
0.24
0.08
0.28

16.99
11.83
5.02
5.34
0.07
0.25
0.06
0.25

0.23
0.42

0.03
0.17

0.08
0.27

0.33
0.47

0.05
0.23
0.22
0.41

0.27
0.44
0.02
0.13

0.08
0.28
0.08
0.26
0.08
0.28
0.01
0.08
0.15
0.36

0.07
0.25
0.06
0.24
0.04
0.19
0.01
0.07
0.19
0.39

200

Australian Journal of Labour Economics, June 2005

Appendix 3 (continued)
Variable definitions; Means and Standard Deviations (continued)
Variable
Educational level
Primary education

Definition

Dummy variable: Equals 1 if attended

primary school but not secondary school.
Some secondary
Dummy variable: Equals 1 if attended,
but did not complete secondary school.
Skilled vocational
Dummy variable: Equals 1 if highest
educational level is skilled vocational
qualification.
Associate diploma
Dummy variable: Equals 1 if highest
educational attainment is Associate diploma
/advanced certificate.
Degree
Dummy variable: Equals 1 if highest
educational attainment is undergraduate
degree or a diploma.
Postgraduate
Dummy variable: Equals 1 if highest
educational attainment is a Postgraduate
degree or diploma.
Basic vocational
Dummy variable: Equals 1 if highest
educational level is basic vocational
qualification.
Completed secondary Dummy variable: Equals 1 if completed not
school (omitted
higher than secondary school.
category)
Fixed term contract Dummy variable: Equals 1 if employment
contract ends on a fixed date.
Casual
Dummy variable: Equals 1 if not entitled
to both paid holiday or sick leave.
School*exp
Years of formal schooling times potential
experience
Region dummies and house prices
NSW
Dummy variable: Equals 1 if workplace is
in New South Wales.
ACT
Dummy variable: Equals 1 if workplace is
in the Australian Capital Territory.
TAS
Dummy variable: Equals 1 if workplace is
in Tasmania.
NT
Dummy variable: Equals 1 if workplace is
in the Northern Territory.
QLD
Dummy variable: Equals 1 if workplace is
in Queensland.
SA
Dummy variable: Equals 1 if workplace is
in South Australia.
WA
Dummy variable: Equals 1 if workplace is
in Western Australia.
VIC (omitted
Dummy variable: Equals 1 if workplace is
category)
in Victoria.
House Prices
Real Median House Price in each capital
city (log)

Male

Female

0.03
0.18
0.31
0.46
0.23
0.42

0.02
0.15
0.40
0.49
0.05
0.21

0.08
0.28

0.09
0.29

0.09
0.29

0.10
0.30

0.04
0.20

0.03
0.18

0.02
0.16

0.05
0.23

0.18
0.15

0.25
0.43

0.06
0.25
0.08
0.28
246.5
139.1

0.06
0.24
0.20
0.40
203.1
137.2

0.33
0.47
0.01
0.07
0.03
0.18
0.0004
0.02
0.17
0.38
0.08
0.27
0.08
0.28
0.28
0.44
4.86
0.21

0.32
0.46
0.01
0.09
0.02
0.15
0.01
0.04
0.19
0.39
0.08
0.28
0.06
0.24
0.30
0.46
4.86
0.21

Note: Unweighted means and standard deviations for workplaces with 20 or more
employees.

Dobbie: The Insider-Outsider Theory: Some Evidence from Australia

201

References
Blanchard, O. and Summers, L. (1986), Hysteresis and the European
Unemployment Problem, NBER Macroeconomics Annual, 15-78.
Blanchard, O. and Summers, L. (1987), Hysteresis in Unemployment,
European Economic Review, 31, 288-295.
Blanchflower, D. and Oswald, A. (1990), The Wage Curve, Scandinavian
Journal of Economics, 92, 215-235.
Blanchflower, D. and Oswald, A. (1994), Estimating a Wage Curve for
Britain 1973-90, Economic Journal, 104, 1025-1043.
Blanchflower, D. and Oswald, A. (1995), An Introduction to the Wage
Curve, Journal of Economic Perspectives, 9, 153-167.
Card, D. (1995), The Wage Curve: A Review, Journal of Economic Literature,
XXXIII, 785-799.
Debelle, G. and Vickery, J. (1998), Labour Market Adjustment: Evidence
on Interstate Labour Mobility, Research Discussion Paper 9801,
Reserve Bank of Australia.
Dobbie, M. (2003), Insiders and Outsiders: Wage Setting Institutions and
the Persistence of Unemployment in Australia, PhD Dissertation,
Macquarie University.
Flatau, P. Lewis, E. and Rushton, A. (1990), The Duration Composition of
Unemployment in Australia, Hysteresis and Insider-Outsider
Theory, Murdoch University Working Paper, No 43.
Flatau, P. Kenyon, P. Lewis, P. and Rushton, A. (1991), The Accord,
Corporatism and Outsiders in The Labour Market, in Johnson, M.
Kriesler, P. and Owen, A. (eds.), Contemporary Issues in Australian
Economics, Macmillan, Melbourne.
Greene, W. (1991), Econometric Analysis, Macmillan, New York.
Kennedy, S. and Borland, J. (2000), A Wage Curve for Australia,
Oxford Economic Papers, 52, 774-803.
Kenyon, P. (1990), Insiders, Outsiders and Corporatism: An Interpretation
of the Empirical Evidence on the Effects of the Accord, Murdoch
University Working Paper, No 45.
Layard, R. Nickell, S. and Jackman, R. (1991), Unemployment, Oxford
University Press, New York.
Lindbeck, A. and Snower, D. (1988), The Insider-Outsider Theory of Employme
and Unemployment, Cambridge, MA:MIT Press.
Lindbeck, A. and Snower, D. (2001), Insiders versus Outsiders, Journal of
Economic Perspectives, 15, 165-188.
Morehead, A., Steele, M., Alexander, M., Stephen, K. and Duffin, L. (1997),
Changes at Work: The 1995 Australian Workplace Industrial Relations
Survey, Longman, Melbourne.
Moulton, B. (1986), Random Group Effects and the Precision of Regression
Estimates, Journal of Econometrics, 32, 385-397.
White, H. (1980), A Heteroskedasticity-Consistent Covariance Matrix
Estimator and a Direct Test for Heteroskedasticity, Econometrica, 48,
817-835.
Winter-Ebmer, R. (1996), Wage Curve, Unemployment Duration and
Compensating Differentials, Labour Economics, 3, 425-434.
Wooden, M. (2001), Union Wage Effects in the Presence of Enterprise
Bargaining, Economic Record, 77, 1-18.
Wooden, M. and Bora, B. (1999), Workplace Characteristics and Their Effects
on Wages: Australian Evidence, Australian Economic Papers, 38, 276-289.