Solidaristic Wage Bargaining
91
1983 by making a separate offer to the
metalworking union. In 1990, the Swedish
employers' confederation closed its bargai-
ning department to render a return to cen-
tralized, Solidaristic bargaining impossible
(Pontusson and Swenson 1995).
What happened? One explanation
points to changes in the organization of pro-
duction that purportedly make centralized
pay-determination more difficult if not
impossible. In particular, it is argued that the
replacement of large-scale production of
standardized products by "flexible specializa-
tion" or "diversified quality production" has
increased employers' need for more discreti-
on in wage-setting (Pontusson and Swenson
1995, Lindbeck and Snower 1995).
A less speculative explanation is that
the policy changed. Over time the unions'
egalitarian goals increased. As Hibbs and
Locking (1991) describe the change, the
policy grew from "equal pay for equal work"
to "equal pay for all work". In the 1950s and
1960s, Solidaristic wage agreements reduced
inter-industry and inter-firm wage differenti-
als as we described in the previous section,
but left intra-firm wage differentials unregu-
lated. From 1969 forward, however, central
agreements in Sweden included provisions
for the equalization of wages inside each
plant unless the union and the firms agreed
otherwise at the industry level (Pontusson
and Swenson 1995). In the 1970s, wage
compression began to be applied across
occupations in both Norway and Sweden,
above all in the public sector.
There were a number of reasons why
the unions sought to increase the scope of
Solidaristic bargaining at the end of the
1960s. The unionization of the public sector
that occurred in the 1960s changed the com-
position of the union movement in a manner
that increased the relative weight of low-paid
workers. The large increase in female partici-
pation in the labor force (most of whom
entered the public sector) and the subse-
quent demands for wage equality between
genders pushed in the same direction. The
pressure for wage compression within firms
came from some of the private sector unions
as well, most notably the metalworkers in
Sweden, who were not low-paid on average
(and thus got little out of wage equalization
among industries) but who had a significant
number of low-paid members (Pontusson
and Swenson 1995).
According to the estimates of Hibbs
and Locking (1995), the impact of wage
compression between occupations was to
reduce productivity growth (in contrast to
wage compression between firms and indus-
tries). Why the compression of inter-occupa-
tional wage differentials should affect pro-
ductivity and profitability negatively is not
clear. Edin and Topel (1995) suggest that
lower wages for high-skilled work reduced
the supply of skilled workers. It may be diffi-
cult to induce workers to work hard when
employers have only a limited ability to eit-
her reward good workers with higher pay or
to punish substandard workers with dismis-
sal. Perhaps the most compelling reason for
the negative effects of inter-occupational
wage compression is the resentment the poli-
cy generated among the higher paid.
Bargaining among the unions over relative
wages became as Conflictual as bargaining
between labor and management (Hernes
1991, Moene and Wallerstein 1993).
Interunion wage disputes became a common
cause of strikes in the 1980s.
The medium-run effect of the
expansion of the scope of Solidaristic bargai-
ning was to significantly reduce overall wage
inequality. The longer-run effect was to
undermine the policy's support. The ability