JOB REALLOCATION, EMPLOYMENT CHANGE AND AVERAGE JOB TENURE: THEORY AND WORKPLACE EVIDENCE FROM AUSTRALIA
| Date | 01 August 2004 |
| Author | Peter N. Smith,Karen Mumford |
| DOI | http://doi.org/10.1111/j.0036-9292.2004.00312.x |
| Published date | 01 August 2004 |
JOB REALLOCATION, EMPLOYMENT
CHANGE AND AVERAGE JOB TENURE:
THEORY AND WORKPLACE EVIDENCE
FROM AUSTRALIA
Karen Mumford and Peter N. Smith
n
Abstract
We explore determinants of job reallocation, employment change and average job
tenure in this paper. A model which associates technological advances with the
process of economic growth is modified and analysed. The features of this model
allow for the possibility of asymmetric behaviour when looking at the net change in
employment. Workplace data from Australia (AWIRS95) are used to test the
predictions generated from the model for both employment change and average job
tenure. Results are presented that provide clear evidence as to the nature of
workplaces in which net employment growth is concentrated. We find that
employment growth is asymmetrically related to expected changes in demand for
the output of the workplace. We also find that employment is lower for workplaces
that are larger, older, have greater union density, offer higher relative earnings, or
are operating in a more competitive environment. The impacts on average tenure
are as expected from the model and from these changes in employment. Overall
there is a substantial degree of agreement between the predictions of the model and
our empirical results.
I Intro ductio n
The empirical reality that the labour market is characterised by continuous flows
of workers through jobs, across employment states and in and out of the labour
force is now well documented for many countries (OECD, 1994; Burda and
Wyplosz, 1994). In this paper
1
we concentrate on two important outcomes resul-
ting from the flow of workers in and out of jobs at the workplace level (Davis
and Haltiwanger, 1990, 1992 and 1998; Davis et al., 1996): employment growth
and average job tenure. These two labour market outcomes, when taken toge-
ther, can be used to evaluate the implications of job creation and destruction.
Descriptive empirical studies of employment growth and job flows are rare,
although there is a study of manufacturing in Australia using disaggregated
n
University of York
1
This paper is a revised and extended version of our paper ‘Job Reallocation: Theory and
Workplace Evidence,’ centre for Economic Performance Discussion Paper No. 360, July 1997.
Scottish Journal of Political Economy, Vol. 51, No. 3, August 2004
rScottish Economic Society 2004, Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK
and 350 Main Street, Malden, MA 02148, USA
402
industry sector data (Borland, 1996) and a limited number of studies for other
countries are available (Contini and Revelli, 1997; Konings and Pissarides, 1994;
CEPR, 1995). Some multivariate empirical analyses of job flows have also been
carried out resulting in a limited number of stylised facts, for example,
employment change is associated: with firm size and age; with the extent of
product market competition; with different industry groups (Blanchflower and
Burgess, 1996; Davis et al., 1996); and with positive and negative firm-specific
demand shocks, although the relative sizes and the cyclical properties of these
flows are debatable (Boeri, 1996; Foote, 1997; Broersma and Van-Ours, 1999).
Papers which attempt a theoretical explanation of job reallocation are even
fewer in number. Amongst the most important of these papers is the partial
equilibrium approach of Caballero and Hammour (1994) and the general
equilibrium explanations provided by Aghion and Howitt (1994), Mortensen
and Pissarides (1998) and Caballero and Hammour (1996). These papers all rely
on economic growth resulting in a process of creative destruction which
generates reallocation in the labour market.
A flexible theoretical model is needed to assess the relative importance of the
above stylised facts in order to give them structural interpretations. A key
feature of such a model is that it should allow the firm to control hiring of new
workers and firing of existing workers separately. This will allow for the
possibility of asymmetric behaviour when looking at the net change in
employment. We use the Caballero and Hammour (1994) model as a vehicle
to analyse both employment growth and average tenure. In so doing, we make
minor modifications in order to examine additional features beyond those that
the original model explicitly addressed. We also re-calibrate the model for the
Australian context and translate the model into an empirical model suitable for
modelling workplace employment change and average job tenure in Australia.
We test the predictions of the model with data generated by the Australian
Workplace Industrial Relations Survey 1995 (AWIRS95). This data source
provides important information on total job movements across workplaces,
allowing for the calculation of employment growth and average tenure.
AWIRS95 respondents also provided information on a range of topics, resulting
in a rich source of additional information used to test the predictions generated
from the theoretical model.
In Section II of this paper, we consider the creative destruction model
discussed above and develop a range of hypotheses we wish to explore
empirically. Following a discussion of the data set in Section III, the predictions
of the model using our measures of net employment growth and average job
tenure are investigated in Section IV. Conclusions and suggestions for further
work are presented in the fifth, and final, section of the paper.
II Modelling Job Flows
Following Caballero and Hammour (1994), firms combine labour and capital in
fixed proportions to create a new productive unit (a new job) which they endow
with the latest technology. The exogenous continuous technological progress is
JOB REALLOCATION AND EMPLOYMENT CHANGE: AUSTRALIA 403
rScottish Economic Society 2004
To continue reading
Request your trial