Solidaristic Wage Bargaining
89
of construction. But changes in the bargai-
ning system can alter the market value of
existing plants, and thereby affect employers'
wealth. Whether or not employers would
prefer Solidaristic bargaining or local bargai-
ning depends, in general, on both the value
of alpha and the age of the plant. If α is suffi-
ciently low such that collective bargaining
only raises wages with decentralized bargai-
ning, the market value of newer plants is hig-
her with Solidaristic bargaining, and the
value of older plants is no lower. If, at the
other extreme, α is sufficiently high that
even owners of newly built plants prefer local
bargaining (OC>OC* in Figure 4), employers are
unanimous in preferring decentralized bar-
gaining. When OC is greater than αo but less
than CC* in Figure 3, there is some vintage
(which depends on ɑ) such that owners of
newer plants prefer Solidaristic bargaining
while owners of older plants prefer local bar-
gaining. The closer is CC to OC0, the larger the
proportion of plants whose value would be
increased with Solidaristic bargaining.
Average Wages
Wage earners are affected differently by soli-
daristic bargaining, depending on the age of
the plant in which they work. Average wages
under the two bargaining systems can be
compared, however. With Solidaristic bargai-
ning, the wage is a share of average producti-
vity in the industry, and average productivity
increases as workers' share increases since
only the most efficient plants can survive.
With decentralized bargaining, the average
wage also approaches a simple share of avera-
ge productivity as OC increases. In contrast
with Solidaristic bargaining, however, average
productivity remains constant as workers'
share increases with local bargaining since
the age at which plants are retired remains
constant. Thus, the average wage is higher
with local bargaining when CC is sufficiently
close to αo, but higher with Solidaristic bar-
gaining when OC is sufficiently close to one.
Summary of Efficiency
and Distributional Effects
Solidaristic bargaining prevents less efficient
firms from paying lower wages than the most
efficient firms. As a result, older plants are
shut down earlier than they would have been
with decentralized bargaining. At the same
time, Solidaristic bargaining prevents workers
in newer, more productive plants, from
obtaining higher wages. The result for firms
is higher profits during the first period of a
plant's life.
Whether or not Solidaristic bargai-
ning is more efficient than decentralized bar-
gaining, in the sense of raising productivity,
employment and output simultaneously,
depends on the value of OC. If OC is close to CC0
the union wage is close to the wage that cle-
ars the labor market. In Norway and
Sweden, the national wage agreement cove-
red the entire blue-collar labor market. In
both countries, unemployment rarely rose
above 3 per cent until the very end of the
1980s. It follows that the difference between
the union wage and the market-clearing
wage was not large. Therefore, the relevant
case for Norway and Sweden until very
recently is OC close to 0Co, the case in which
Solidaristic bargaining increases employ-
ment, output and productivity. Rehn and
Meidner's belief in the relative efficiency of
Solidaristic bargaining compared to local
wage-setting in the Scandinavian context is
supported by our model.
The same reasoning on the elimina-
tion of wage differences among plants within
an industry might be applied to the elimina-
tion of wage differentials between industries.
With industry-level bargaining, wages will