ON THE ROLE OF STRIKES VARIABLES IN U.K. WAGE EQUATIONS
| Author | George Zis |
| Date | 01 February 1977 |
| Published date | 01 February 1977 |
| DOI | http://doi.org/10.1111/j.1467-9485.1977.tb00406.x |
Scottish
Journal
of
Political
Economy,
Vol.
24,
No.
1,
February
1977
ON
THE
ROLE
OF
STRIKES VARIABLES
IN
U.K.
WAGE EQUATIONS
GEORGE
ZIS*
INTRODUCTION
Economists investigating the determinants of wage inflation have sought to
isolate the independent influence
of
trade unions on the rate of change of
money wage rates by using strike activity as a proxy for trade union pushful-
ness, e.g. Godfrey (1971), Taylor (1972), Johnston and Timbrell (1973) and
Ward and Zis (1974). The choice
of
these particular four studies for con-
sideration reflects the limited objectives of this paper. We are not concerned
with the question
of
whether
or
not strike activity is
a
satisfactory index of
trade union militancy, nor do we consider the relative merits of the alternative
ways by which strike activity can be measured. The studies already cited all
used the number of strikes, in preference to days lost through industrial
disputes
or
the number of workers involved in strikes, as the relevant measure
of strike activity. Similarly, we refrain from discussing the propriety of
choosing this particular index
of
trade union pushfulness rather than, say,
the change in the proportion of the labour force that is unionised which has
been advanced as an adequate measure of trade union militancy (Hines,
1964). Finally, we are not interested in presenting some wage equation which
we would wish
to
defend as being, in some sense or other, the “truth”
regarding the process of wage determination.
Our
objective is confined to
providing
a
reconciliation between the conflicting empirical results that have
been presented on the statistical performance of strike variables in wage
equations. Thus, while Godfrey and Taylor argue that the number of strikes
does make a statistically significant contribution in the explanation of wage
inflation in the
U.K.
the empirical findings of Johnston and Timbrell and
Ward and Zis contradict this conclusion, their estimated coefficient on the
strikes variable not being significantly different from zero.
In section
I
the description of the differences among the four studies
provides the basis for the selection of a single wage equation as the most
appropriate for further investigation. In section
I1
we przsent our empirical
results while section I11 contains
our
conclusions.
1
The task of arriving at some reasonably confident conclusions regarding the
statistical performance of the strikes variable in wage equations is not facili-
tated by the multitude of differences among the four studies that provide our
*
I
am deeply grateful
to
Jon Stewart for his guidance and advice without which this
paper would never have been finished.
I
am, also, indebted to
M.
Summer and
G.
Smith
for numerous helpful comments on an earlier version and
to
R.
Ward for excellent research
assistance. However, responsibility for any errors is entirely my own.
43
To continue reading
Request your trial