SURVEY EXPECTATIONS AND STOCHASTIC TRENDS IN MODELLING THE EMPLOYMENT — OUTPUT EQUATION

AuthorJaakko Pehkonen
Published date01 November 1992
DOIhttp://doi.org/10.1111/j.1468-0084.1992.mp54004007.x
Date01 November 1992
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 54,4(1992)
0305-9049 $3.00
SURVEY EXPECTATIONS AND STOCHASTIC
TRENDS IN MODELLiNG THE EMPLOYMENT
- OUTPUT EQUATION
Jaakko Pehkonent
I. INTRODUCTION
Optimal behaviour in the face of adjustment costs implies a partial adjust-
ment model whereby a choice variable is gradually adjusted to changes in a
control variable. As Gould (1968) pointed out, any model involving optimal
distributed lags requires a theory of expectation formation. The conventional
employment-output equation, which relates employment to its own lags as
well as current and lagged output, is an example of a partial adjustment model
in which expectations are assumed to be adaptive.1 Although this conven-
tional specification has an appeal of its own in its simplicity, its empirical per-
formance in explaining movements in employment and, in particular, in
forecasting, has given rise to debate and further studies. For instance Wren-
Lewis (1984), using UK data, shows that the forecasting performance of the
conventional employment-output equation deteriorated substantially in the
late 1970s and 1980s and that it can be improved by exploiting data on firm's
short-term output expectations. Bond (1988) in turn finds that the inclusion
of data from macro-economic forecasts, which serve as a proxy for firms' out-
put expectations, enhance the performance of the employment-output
equation.
In addition to the fact that modelling expected output as some unspecified
function of current and lagged output is theoretically unsatisfactory, there are
other factors which cast doubts on the appeal of the conventional employ-
ment-output equation. In particular, the commonly-used procedure to proxy
the effects of technical progress on employment by including a time trend in
the employment-output equation can be questioned since it implicitly
assumes that productivity growth follows a steady trend. This issue is recently
illustrated in Harvey, Henry, Peters Wren-Lewis (1986) who show that a
fI am grateful to Kan Hämäläinen for research assistance and Trjö Jahnsson Foundation for
financial support.
'Early studies employing the specification include Ball and St Cyr (1966) and Brechung
(1965).
579

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT