Considering the amount of research on the consequences of central bank independence, it is remarkable how few insights have been generated into causes of cross-national variance in independence. Instead, the new classical literature presents us with a puzzle: If there are benefits to having an independent bank, but no apparent costs, why do we see so much institutional variance?1 By contrast, if monetary policies have real effects that vary systematically across bargaining systems - as argued in this book - then there is good reason to expect monetary regimes to also vary systematically. In particular, while governments may seek to retain policy autonomy in decentralized systems — where the economic costs are either small or negative if flexibility reduces volatility - in intermediately centralized systems sociotropic voting would furnish governments of all stripes with a strong electoral incentive to adopt nonaccommodating regimes. By the same logic, governments would have a strong incentive to adopt flexibly accommodating regimes in highly centralized bargaining systems.

The institutional design argument also pertains to bargaining centralization. In neo-corporatist theory, centralization is explained as an attempt to control wage costs, but the present analysis highlights the possibility that conservative monetary regimes offer an alternative form of cost control without centralization. This is important because bargaining systems are not distributively neutral. Centralized bargaining typically involves intrusive restrictions on employers’ discretion over firm-internal wage structures, and it strengthens the bargaining power of low-wage workers relative to high-wage workers. Institutional design is therefore a political game that involves distributive struggles between partisan governments and sectoral interests of employers and workers.

The theory of contested institutions presented in this chapter hypothesizes that, over time, stable coalitions will be formed behind either centralized bargaining with an accommodating regime — Keynesian centralization — or intermediately centralized bargaining with a nonaccommodating regime - monetarist decentralization. In its weak form, this thesis simply states that efficiency gains will lead to a convergence around one or the other institutional equilibrium. This is a proposition that is easy to test with quantitative data, and the second section of this chapter presents such a test. That section also illustrates the notion of institutional equilibria with reference to four empirical cases.