THE DETERMINATION OF WAGES IN THE UNION AND NON‐UNION SECTORS
| Date | 01 July 1977 |
| Author | George E. Johnson |
| DOI | http://doi.org/10.1111/j.1467-8543.1977.tb00085.x |
| Published date | 01 July 1977 |
British
Journal
of
Industrial Relations
Vol.
XV
No.
2
THE DETERMINATION
OF
WAGES IN THE
UNION AND NON-UNION SECTORS
GEORGE
E.
JOHNSON*
I.
INTRODUCTION
The purpose of this paper is
to
attempt
to
assess the relative importance of the
union and non-union sectors in the determination of changes in the aggregate wage
level in the United
States.
Opinions expressed in the literature range from the
view that unionsdominate the
wage-setting process
to the view that they are of little
importance. An example of the former view is the model of Eckstein and Wilson
in which non-union wages are assumed
to
follow what has been happening in
the union sector.’ On the other side, Lucas and Rapping argue that unions
merely attempt
to
maintain
a
certain wage position relative
to
the non-union
sector and that it is therefore appropriate
to
analyse aggregate wage determina-
tion solely in terms of competitive factors.’ Another approach is taken by
Phelps, who points out that since only about
a
quarter of all workers belong
to
unions it is reasonable
to
ignore the influence of unionism on the determination
of wages.3 Public policy during the past fifteen years has generally reflected the
view that unions
are
of
major importance in the
wage
determination process; for
example, the focus
of
the KennedyJohnson Wage-Price Guideposts was on
keeping large contract settlements ‘within reason’.
If
the view that the union
sector is dominant is correct, such
a
policy can shift the inflation-unemployment
trade-off
to
the left (or lower the natural unemployment
rate).
But if the union
sector merely follows the non-union sector,
a
jiolicy of wage restraint by unions
will not
be
very effective.
Despite the importance of this question, there has been little systematic
attempt
to
resolve it. Most studies have either focused on the union or the
non-union sector. In this paper I will show in Section
I1
that on
a
priori
grounds
one cannot rule out either view
of
the importance of unionism (or some
intermediate position that both sectors
are
important) and
set
up
a
test
for the
resolution of the question. This
test
requires data on union and non-union wage
rates over time, and the construction of this series
is
discussed in Section
111.
The basic empirical results are set out in Section
IV,
and their implications are
explored in Section V. The bottom line of this exercise is the following: although
both sectors are important in the wage determination process the non-union
sector has been the more important during
the
past twenty years. This has not
always been
so
and will not necessarily be
so
in the future, and this qualifica-
tion is given in Section VI.
11.
CONCEPTUAL
ISSUES
Suppose
first
that wages in the non-union sector are determined indepen-
dently of wages in the union sector. In this case business firms will attempt
to
maintain
a
certain
relative
wage position in the labour market. Firms will
therefore change their wage offers over time by an amount equal
to
the rate
at
which they
expect
the average wage offer of other firms in the market
to
*
Department
of
Economics,
The University
of
Michigan.
211
To continue reading
Request your trial