6
government partisanship (Lange and Garrett 1985, 1987; Alvarez, Garrett, and Lange 1991;
Scruggs 2001). The crucial point here is the strategic interactive relationship between
government and union actor. When centralized unions meet a pro-labor government, they will
restrain wage increases with an expectation of favorable macroeconomic policies, which in turn
strengthens the virtuous circle of government spending, investment promotion, job creation, and
economic growth. When centralized unions meet a rightist government, however, they will not
expect such a political commitment. Thus, they will choose to maximize their short-term wages
by utilizing their organizational powers. A similar logic can be applied to the case of
decentralized unions. When atomized unions meet a leftist government, they are likely to push
for the wage maximization strategy with an expectation that the government would eventually
loosen fiscal and monetary polices to buffer the negative effects of high wage increase. But, this
would likely result in inflation without much growth because wage militancy dampens the
investment promoting role of government expansion. When they meet a rightist government,
however, they expect macroeconomic austerity and will not risk job losses by pushing up their
wages beyond the market averages.
Coordinated wage bargaining, political institutions, and economic growth
While empirical studies have provided statistical evidences for the hump-shaped
interaction effects of union centralization and government partisanship on economic growth, this
paper argues that the significance of the effects are mainly due to incomplete model specification.
More specifically I will show that the effects virtually disappear with very insignificant P-values
once economic and structural variables are taken into the mode. But, there does exist an
interaction effect at institution level, that is, between wage bargaining institution and political
Convention