90
Karl Ove Moene and Michael Wallerstein
differ by industry in accordance with indus-
try-level differences in productivity or profi-
tability. Solidaristic bargaining, applied over
the national economy, limits the ability of
the most efficient industries to pay a wage
premium, and prevents the least efficient
industries from staying in business by lowe-
ring wages. In fact, the elimination of wage
differentials between industries can be
understood as a subsidy for new industries
and a tax on older ones (Agell and
Lommerud 1993). The result of nation-wide
Solidaristic bargaining is to force older indus-
tries to shut down while encouraging the
growth of new industries. The consequence
is a national economy composed of more
modern industries than would be the case
with industry-level bargaining.
Solidaristic bargaining, whether at
the industry level or the national level, has
two important distributional effects. The
first effect is to increase the wages of low-
paid workers relative to highly paid workers.
The second effect is to alter the distribution
of income between wages and profits. The
conditions under which Solidaristic bargai-
ning unambiguously improves efficiency are
also the conditions under which Solidaristic
bargaining increases employers' wealth and
lowers the average wage. Under the conditi-
ons that prevailed in Norway and Sweden,
our results indicate that the beneficiaries of
the egalitarian wage policy were low-paid
wage earners -whose relative wages increased
rapidly -and employers, particularly employ-
ers with modern plants. The principal losers
were the relatively highly paid wage and sala-
ry-earners whose incomes were held back in
the name of wage equality.
These conclusions fit well with
recent historical work on the origins of soli-
daristic bargaining in Sweden done by Peter
Swenson (1989, 1991a). The political coali-
tion that prevailed in the 1950s and establis-
hed the pattern of centralized and Solidaristic
bargaining that was to last for 25 years was
comprised of the low-wage unions inside the
LO and Swedish employers organized in the
Swedish Employers' Confederation, the
Svenska Arbetsgivareforeningen or SAE
High paid unions were prevented from lea-
ving the centralized negotiations by the thre-
at of lockouts. In 1955, a strike by workers
in the Paper Workers Union, a union with
rapidly growing productivity and relatively
high wages, was met by a threat to lockout a
majority of workers in all Swedish industry
(Swenson 1989: 54—55). Again in 1957, the
SAF had to threaten the unions with a loc-
kout in order to force them to bargain in a
centralized manner. It is unlikely that the
low-wage unions and the LO leadership
would have been able to force the high-wage
unions to accept an egalitarian wage policy
without the backing of Swedish employers
and the threat of lockouts against recalcitrant
unions.
Our results also fit with the econo-
metric study of wage inequality and produc-
tivity growth in Sweden by Hibbs and
Locking (1995). Hibbs and Locking find
that the reduction of wage differentials bet-
ween plants and between industries was asso-
ciated with above average productivity
growth, unlike the effect of the reduction of
wage differentials between occupations that
came later.
The Decline of Solidaristic Bargaining
In the 1950s and 1960s, we have argued that
employers were quiet but critical supporters
of Solidaristic bargaining. In the 1970s and
1980s, employers were increasingly vocal
opponents. In Sweden, employers in the
metalworking sector took the initiative to
end the system of centralized bargaining in