Download (PDF)

Building and Using Strategic Capacity: Labor Union Federations and Economic Policy

Mancur Olson made strong claims about the importance of ``encompassing'' groups for economic performance. He argues that broadly-based interest groups, especially those that transcend regional or industrial barriers, are more likely to push for policies that yield longer-term benefits for a larger proportion of the population. Groups with narrower constituencies tend to pursue parochial goals with the aim of securing a bigger share of the national income for themselves. Observers of labor market institutions have relied on this line of reasoning, especially pertaining to the centralization of wage bargaining, to explain the variance in macroeconomic performance of rich democracies since the oil shocks. Even though the ``encompassingness'' of unions drives these theories no one has attempted to explain how union federations develop the capacity to bargain at the national level or the wide variation they display in their abilities to influence outcomes they care about. This project begins to explain how economic actors aggregate their interests and how these aggregates develop the capacity to influence national-level policy, with an empirical focus on confederations of national labor unions.

The first part of the project asks why labor unions form confederal organizations and what explains the variation in the confederal authority over affiliates. I devise a model of confederal organization in which I conceive of the confederal problem as akin to a repeated Stag Hunt game. Unions have an incentive to cooperate in collective projects but they vary in how much they value the collective project. To enable ongoing cooperation, unions have the option of signing a contract that allocates decision rights about how best to invest collective effort. Weaker affiliates have an incentive to yield decision rights to the stronger. For their part, stronger unions will only permit confederal influence over wage bargaining when there are pronounced economies of scale in cooperation and/or they are compensated in other ways, e.g., through political concessions. The model also predicts a trade-off between the size of the confederation and the number of domains over which it has competency. Subsequent chapters explore the model empirically. Chapter~\ref{chap:AFL} is a paired historical study of the American Federation of Labor and the Knights of Labor. Chapter~\ref{chap:labcent} presents the first econometric study of union confederal centralization. I find consistent evidence for the size-scope trade-off as well as evidence that the structure of political authority, especially federalism and more frequent Left government, affect degree to which unions centralize membership and authority in confederal organizations. I also find large and significant effects for structural variables like trade exposure, size of the economy, and cultural/linguistic fractionalization.

The second half explores the emergence of formal policy agreements between governments and unions in 20 OECD countries, 1974-2000. In chapter~\ref{chap:pactmod} I model these agreements as self-enforcing contracts that allow political parties to make policy promises credible to voters. Unions agree to abide by the pact so long as economic conditions make policy a better tool than industrial action for achieving gains in living standards. The model implies that 1) the emergence of pacts should follow the electoral cycle and that the 2) organizational linkages between unions and parties are critical for providing both the incentives for partisan politicians and unions to sign pacts and for making agreements self-enforcing in equilibrium. Chapter~\ref{chap:pactdet} is an event history analysis of pact onset using an original dataset of social pacts covering 20 OECD countries between 1974 and 2000. I find that, as expected, pacts are strongly influenced by the electoral cycle. In chapter~\ref{chap:OzNz} I apply the model to an empirical puzzle in the cases of Australia and New Zealand. Both countries faced similar economic crises in the early 1980s yet a pact emerged in Australia but not in New Zealand. Relying on original interviews with key policy makers in both countries, I conclude that the organizational relationship between the union peak associations and the Labor parties were the critical differences that made intertemporal political transactions possible (and self-enforcing) in Australia but not in New Zealand.

The findings in the dissertation have implications across several areas. Part I endogenizes a key variable in current models of wage-price bargaining: union centralization. Part II links the literatures on social pacts and wage bargaining with the larger debates on political business cycles in policy and outcomes. I provide better micro-foundations for theories relating the organizational structure of unions to macroeconomic outcomes. Overall I contest Olson's pessimism about the emergence of strategically meaningful encompassing groups. There are also normative implications for inequality and democracy. Findings in the study show that the cost of implementing needed economic reforms has been both lower and more equitably shared in countries where unions and employers were capable of behaving as unified strategic actors in national-level politics and industrial relations.