¹ OECD, Taxing Wages (Paris: OECD, 1999)Google Scholar.

² OECD, Statistical Compendium: Revenue Statistics (Paris: OECD, 1995)Google Scholar.

³ Scharpf, , Governing in Europe: How Effective? How Democratic? (Oxford: Oxford University Press, 1998), 142Google Scholar.

⁴ See, for example, Conseil Supérieur de l'Emploi, L'allègement des charges sociales sur les bas salaries (Paris: La Documentation française, 1996).

⁵ For an overview of these policy developments, see Bourguignon, François and Bureau, Dominique, eds., L'architecture des prélèvements en France: Etat des lieux et votes de réforme (Paris: La Documentation française, 1999)Google Scholar.

⁶ See, for example, Wanger, Gert, “Soziale Sicherung im Spannungsfeld von Demokratie und Arbeitsmarkt,” in Nubler, I. and Trabold, H., eds., Herausforderungen an die Wirtschaftspolitik an der Schwelle zum 21.Jahrhundert (Berlin: Sigma, 1999)Google Scholar.

⁷ For a formulation of these arguments, see Cameron, David, “Social Democracy, Corporatism and the Representation of Economic Interests in Advanced Capitalist Societies,” in Goldthorpe, John, ed., Order and Conflict in Contemporary Capitalism (New York: Oxford University Press, 1984)Google Scholar; Esping-Andersen, Gosta, Three Worlds of Welfare Capitalism (Princeton: Princeton University Press, 1990), 105–43Google Scholar.

⁸ Calmfors, and Driffill, , “Bargaining Structure, Corporatism, and Economic Performance,” Economic Policy 6, no. 1 (1988)Google Scholar.

⁹ As will be shown below, in the Calmfors-Driffill setup, the presence of the hump is conditional on some technical details related to the elasticity of substitution of goods produced in different sectors. The hump is more pronounced in less homogenous economies (that is, producing goods that are lesser substitutes of each other). In the limit case of an economy in which all sectors produce goods that are very close substitutes for each other, the relationship between the centralization of the wage-bargaining system and the level of unemployment becomes monotonic.

¹⁰ For representative examples of research that attempt to model the interaction between the cor-poratist literature on wage bargaining and the literature on macroeconomic policy, see Scharpf, Fritz, Crisis and Choice in European Social Democracy (Ithaca, N.Y.: Cornell University Press, 1991)Google Scholar; Hall, Peter and Franzese, Robert, “Mixed Signals: Central Bank Independence, Coordinated Wage Bargaining and European Monetary Union,” International Organization 52, no. 3 (1998)CrossRefGoogle Scholar; Iversen, Torben, “Wage Bargaining, Central Bank Independence and the Real Effects of Money,” International Organization 52, no. 3 (1998)CrossRefGoogle Scholar; idem, Contested Economic Institutions: The Politics of Macroeconomics and Wage Bargaining in Advanced Democracies (Cambridge: Cambridge University Press, 1999)Google Scholar; Soskice, David and Iversen, Torben, “The Non-Neutrality of Money with Large Price or Wage Setters,” Quarterly Journal of Economics 115, no. 1 (2000)CrossRefGoogle Scholar.

¹¹ Iversen (fn. 10, 1998), 48–49.

¹² See Cameron (fn. 7); and Esping-Andersen (fn. 7). This specification of the utility of trade unions distinguishes this model from the Daveri and Tabellini model; see Daveri, Françesco and Tabellini, Guido, “Unemployment and Taxes: Do Taxes Affect the Rate of Unemployment?” Economic Policy 15, no. 30 (2000), 98Google Scholar.

¹³ This assumption of full unionization can easily be relaxed. More specifically, one can assume that for workers who are nonunionized, wages are determined in a decentralized setting (in other words, both I and J are very large). The equilibrium wage level of this economy is a combination of the wages set in the unionized part of the economy and of the wages that are determined in this decentralized setting.

¹⁴ I follow both the assumption of Iversen (fn. 10, 1999) and that of Iversen and Soskice; see Torben Iversen and David Soskice, “Monetary Integration, Partisanship and Macroeconomic Policy” (Paper presented at the annual meeting of the American Political Science Association, Atlanta, Ga., 1999), 5. To make the comparison of the results easier, I use the same notation.

¹⁵ Iversen and Soskice (fn. 14) reformulate this assumption as follows. Assume M = P ^1-β where M is the nominal money supply, P is the aggregate price level, and β is the parameter measuring the degree to which the central bank follows a nonaccommodating policy rule. “If the central bank it completely accommodating, the central bank fixes the real money supply by setting M equal to the price level, whereas if the central bank is completely non-accommodating, it fixes the nominal money supply and sets M equal to unity” (p. 5). In other words, in the case of nonaccommodating monetary policy (β = 1), the money supply is independent of the price level, while in the case of accommodating monetary policy (β = 0), the money supply is set equal to the price level.

¹⁶ Mas-Colell, Andreu, Microeconomic Theory (New York: Oxford University Press, 1995), chap. 3Google Scholar.

¹⁷ Iversen and Soskice (fn. 14) assume that the union in sector i is , where is the average wage for the union members in this sector and e_i. is the employment rate in sector i (p. 7); Calmfors and Driffill (fn. 8).

¹⁸ For a formulation of these arguments by “corporatist scholars,” see Cameron (fn. 7); and Esping-Andersen (fn. 7).

¹⁹ For these views, see Bruno, Michael and Sachs, Jeffrey, Economics of Worldwide Stagflation (Cambridge: Harvard University Press, 1985)CrossRefGoogle Scholar; Cameron (fn. 7); and Soskice, David, “Wage Determination: The Changing Role of Institutions in Advanced Industrialized Societies,” Oxford Review of Economic Policy 6, no. 4 (1990), 36–61CrossRefGoogle Scholar.

²⁰ See Calmfors and Driffill (fn. 8).

²¹ More specifically, the models make different assumptions about the relative wage elasticities of demand, in other words, about the effects of changes in the level of wages in one sector on the level of employment in another sector. The models predicting a monotonic relationship between the centralization of the wage-bargaining system and the level of unemployment assume that changes in the level of wages in one union will have the same effect on the level of demand (and thus employment) in the remaining N-1 unions in the economy. By contrast, the Calmfors-Driffill model assumes that changes in the level of wages of one union will have different effects on the level of employment across different unions. Mares, Isabela, Taxation, Wage Bargaining and Unemployment (New York: Cambridge University Press, forthcoming), chap. 2CrossRefGoogle Scholar.

²² Given that the predictions of the theoretical model developed in this article extend (and challenge) the work of Calmfors and Driffill and Iversen and Soskice, the article pursues the same empirical strategy as the earlier research. I too test the model using unemployment as the dependent variable.

²³ The countries included in the analysis are Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Japan, Netherlands, Norway, Sweden, Switzerland, the United Kingdom, and the United States. As is standard in the literature, I use four-year averages for each country; Iversen (fh. 10, 1998, 1999).

²⁴ For an overview of these recent measures, see Kenworthy, Lane, “Wage-Setting Measures: A Survey and Assessment,” World Politics 54 (October 2001)CrossRefGoogle Scholar.

²⁵ Golden, Miriam, “The Dynamics of Trade Unionism and National Economic Performance,” American Political Science Review 87, no. 2 (1993), 444CrossRefGoogle Scholar.

²⁶ Schmitter, Philippe. “Interest Intermediation and Regime Governability in Contemporary Western Europe and North America,” in Berger, Suzanne, ed., Organizing Interests in Western Europe: Pluralism, Corporatism and the Transformation ofPolitics (Cambridge: Cambridge University Press, 1981)Google Scholar. Schmitter's “corporatism score” measures both the organizational centralization and the associational monopoly of the labor movement. While the first term is a measure of the level of the wage-bargaining authority, the second term captures the number and importance of competing unions at each level of bargaining or what Golden refers to as “union monopoly”; Golden (fn. 25), 444.

²⁷ Cameron (fn. 7). Cameron aggregates three institutional characteristics of the labor movement: the power of the labor confederation in collective bargaining (a proxy for the locus of decision-making authority), the “organizational unity of labor” (a measure of union monopoly), and an average measure of union density.

²⁸ Calmfors and Driffill (fn. 8), 52–53. Calmfors and Driffill measure the level of bargaining and the degree of coordination within organizations on both the union and the employer side.

²⁹ Iversen (fn. 10, 1999), 83. Iversen's measure of centralization is computed as , where w, is the weight accorded to each bargaining level j and P_ij is the share of workers covered by union (or federation) i at level j.

³⁰ Hall and Franzese (fn. 10), 530.

³¹ Traxler, Franz and Kittel, Bernhard, “The Bargaining System and Performance: A Comparison of Eighteen OECD Countries,” Comparative Political Studies 33, no. 9 (2000)CrossRefGoogle Scholar. I used the “bargaining centralization” index from the Traxler/Kittel data set.

³² Miriam Golden, Peter Lange, and Michael Wallerstein, “Union Centralization among Advanced Industrial Societies: An Empirical Study,” Data set from http://www.shelley.polisci.ucla.edu/data (downloaded November 1998). I have used the “overall wage-setting centralization” measure from the Golden, Lange, Wallerstein data set. To create a time-invariant measure, I have recoded the Golden, Lange, Wallerstein measure as follows. First, I created a time-invariant score for each country. Next, I rank ordered the economies (with economies having the most centralized labor-market institutions taking the highest values on this centralization score).

³³ The OECD measure of centralization of wage bargaining is computed as the average of two separate indices: a bargaining centralization index and a “coordination index”; see OECD, Employment Outlook (Paris: OECD, 1997), 71Google Scholar.

³⁴ Golden (fn. 25), 444.

³⁵ I average the time-varying indices, such as Iversen (fn. 10) and Golden-Lange-Wallerstein (fn. 32).

³⁶ Denmark and Finland are, of course, borderline cases. For a discussion of the “organizational fragmentation and the conflictual labour relations” in Finland, see Ebbinghaus, Bernhard and Visser, Jelle, Trade Unions in Western Europe since 1945 (London: Macmillan, 2000), 201CrossRefGoogle Scholar.

³⁷ Beck, Nathaniel and Katz, Jonathan, “What to Do (and Not to Do) with Time-Series Cross-Section Data,” American Political Science Review 89, no. 3 (1995)CrossRefGoogle Scholar.

³⁸ For the development of this measure, see Iversen (fn. 10, 1999), 57–60.

³⁹ Franzese, Robert, Macroeconomic Policies of Developed Democracies (New York: Cambridge University Press, 2002)CrossRefGoogle Scholar.

⁴⁰ I am grateful to an anonymous reviewer for recommending this strategy.

⁴¹ These results are robust to the inclusion of a number of additional control variables. For addi- tional results, see Mares (fn. 21).

⁴² See Lucio Baccaro, “The Organizational Consequences of Democracy: Labor Unions and Economic Reforms in Contemporary Italy” (Ph.D. diss., Sloan School of Management, MIT, 1999); Julia Lynch and Karen Anderson, “Internal Institutions and the Policy Preferences of Organized Labor: The Effects of Workforce Aging on Unions' Support for Pension Reform” (Paper presented at the Conference of Europeanists, Chicago, April 2004).

⁴³ Volkerink, Bjorn and Haan, Jacob de, Tax Ratios: A Critical Survey (Paris: OECD, 2000)Google Scholar; Bjorn Volkerink, Jan-Egbert Sturm, and Jacob de Haan, “Tax Ratios in Macroeconomics: Do Taxes Really Matter?” Working Paper 7/2001 (Madrid: European Economy Group, 2001); Eurostat (Statistical Office of the European Communities), Structures of the Taxation Systems in the European Union, 1970—1995 (Luxembourg: Eurostat, 1997)Google Scholar; European Commission, “Effective Taxation and Tax Convergence in the EU and the OECD,” Memo 5/1997 (Brussels: Directorate General II of the European Commission, 1997); OECD, Tax Burdens:Alternative Measures (Paris: OECD, 2000)Google Scholar.

⁴⁴ Mendoza, Enrique, Razin, Assaf, and Tesar, Linda, “Effective Tax Rates in Macroeconomics: Cross-Country Estimates of Tax Rates on Factor Incomes and Consumption,” Journal of Monetary Economic 34, no. 2 (1994)Google Scholar. The time series used in this model updates Mendoza's average labor tax rates series and has been developed and generously provided by Tom Cusack. Corporate taxes play only a minor role in financing the major welfare state programs; see Martin Schludi and Steffen Ganghoff, database (Cologne: Max-Planck Institute for the Study of Societies, 1998), Table 2.2.2.2. Thus, labor tax rates are an appropriate measure of the tax effort necessary to finance social policy commitments.

⁴⁵ A number of recent papers have examined the impact of the growth in the tax burden on the employment performance of OECD economies. These papers differ, however, in the specification of the theoretical model and in the sample size. See Daveri, and Tabellini, , “Unemployment and Taxes,” Economic Policy 15 (2000)Google Scholar; Nickell, Stephen and Layard, Richard, “Labour Market Institutions and Economic Performance,” in Ashenfelter, O., and Card, D., eds., Handbook of Labour Economics (Amsterdam: North Holland, 1999)Google Scholar. In the case of Nickell and Layard, the sample consists of twenty OECD countries over two five-year periods, 1983—88 and 1989—94. Thus, the sample size is much smaller than the sample size of this article. Daveri and Tabellini examine fourteen countries over six five-year periods. The labor-market variables used in their analysis are a variable measuring employment protection and a variable measuring the duration of unemployment benefits. As such, they do not test hypotheses about the hump-shaped relationship between labor-market institutions and unemployment.

⁴⁶ Evelyne Huber, Charles Ragin, and John Stephens, Comparative Welfare States Dataset, 1997, available at http://www.lisproject.org/publications/welfaredata/welfaredata/welfareaccess.htm, updated April 2004; Browne, Eric and Dreijamis, John, eds., Government Coalitions in Western Democracies (New York: Logman, 1982)Google Scholar.

⁴⁷ Hibbs, Douglas, “Political Parties and Macroeconomic Policy,” American Political Science Review 71, no. 4 (1977)CrossRefGoogle Scholar; idem, “Partisan Theory after Fifteen Years,” European Journal of Political Economy 8 no. 2 (1992)Google Scholar; Alesina, Alberto, “Politics and Business Cycles in Industrial Democracies,” Economic Policy 8 (1989)Google Scholar.

⁴⁸ Castles, Francis and Mair, Peter, “Left-Right Political Scales: Some ‘Expert’ Judgments,” European Journal of Political Research 12, no. 1 (1984)CrossRefGoogle Scholar.

⁴⁹ The correlation between these two variables is .42.1 also ran separate models (not reported here) that examined only the impact of higher taxes on unemployment. The analysis produced results that are very similar to the results reported here.

⁵⁰ Mas-Colell (fn. 16).