Published online by Cambridge University Press: 13 June 2011
An influential line of argument holds that globalization causes economic uncertainty and spurs popular demands for compensatory welfare state spending. This article argues that the relationship between globalization and welfare state expansion is spurious and that the engine of welfare state expansion since the 1960s has been deindustrialization. Based on cross-sectional-time-series data for fifteen OECD countries, the authors show that there is no relationship between globalization and the level of labor-market risks (in terms of employment and wages), whereas the uncertainty and dislocations caused by deindustrialization have spurred electoral demands for welfare state compensation and risk sharing. Yet, while differential rates of deindustri-alization explain differences in the overall size of the welfare state, its particular character—in terms of the share of direct government provision and the equality of transfer payments—is shaped by government partisanship. The argument has implications for the study and the future of the welfare state that are very different from those suggested in the globalization literature.
4 There has been, however, significant cross-country variation in exposure to these risks. Those countries that were relatively early “deruralizers” were confronted with less problematic conditions, which helped them cope with structural change. Among these conditions were small entenng-age cohorts, relatively low female participation in labor markets, and relatively buoyant labor markets. All of these conditions eased the problems of structural change. For a discussion of the timing of “derural-lzation” and its consequences, see Esping-Andersen, Gosta, Social Foundations of Postindustnal Economies (Oxford:Oxford University Press, 1999), 24–27CrossRefGoogle Scholar.
5 Blossfeld, Hans-Peter, “Is the German Dual System a Model for a Modern Vocational Training System?” International Journal of Comparative Sociology 33, nos. 3—4 (1993Google Scholar).
6 Kohli, Martin, Rein, Martin, Guillemard, Anne-Marie, and van Gunsteren, Herman, eds., Time for Retirement: Comparative Studies of Early Exit from the Labor Force (Cambridge:Cambridge University Press, 1992Google Scholar).
7 Esping-Andersen, Gesta, The Three Worlds of Welfare Capitalism (Princeton:Princeton University Press, 1990Google Scholar).
9 Rodrik (fn. 2), 13.
10 Garrett (fn. 3,1998), 7.
11 Cameron (fn. 1); Ruggie, John G., “International Regimes, Transactions and Change: Embedded Liberalism in the Postwar Economic Order,” in Krasner, Stephen D., ed., International Regimes (Ithaca, N.Y.:Cornell Univerity Press, 1983Google Scholar); Katzenstein (fn. 1).
13 See Rodik, Dani, Has Globalization Gone Too Far? (Washington, D.C.:Institute for International Economics, 1997Google Scholar), chap. 3.
14 Rodrik (fn. 2).
15 The government sector is less volatile, but it does not make sense to include it in the comparison since this sector is supposed to be growing as a consequence of high volatility in the exposed sectors.
16 This constraint could come on both the spending and the revenue side. Thus, there is broad consensus that capital market integration has raised the price of running deficits, while others (including Rod-rik) have argued that it has made taxation of capital more difficult. The latter argument, however, is disputed by Duane Swank, who finds no evidence that business taxation has become less important for revenue generation. See Swank, , “Funding the Welfare State: Globalization and the Taxation of Business in Advanced Market Economies,” Comparative Political Studies 46 (September 1998Google Scholar).
17 Katzenstein (fn. 1).
18 Cameron (fn. 1).
19 Rodrik (fns. 2 and 13).
20 Indeed, trade openness may be more salient for less-developed countries because trade in many of these countries, unlike trade between OECD countries, has led to heavy dependence on a few primary commodities that are subject to high international price volatility.
21 Garrett (fn. 3).
22 Our thanks to Geoffrey Garrett for generously providing us with the data he used in his analyses, thereby allowing us to replicate his findings. See Table 4.4 in Garrett (fn. 3,1998), 90. Note that Garrett uses levels of spending on the left-hand side, but this formulation gives an estimate for R-squared that is uninformative since the lagged dependent variable will pick up most of the cross-national variance. Using changes in spending on the left-hand side avoids this problem while leaving the estimated coefficients the same. (In mathematical terms, we are simply subtracting the lagged dependent level variable on both sides of the equal sign, which obviously leaves the coefficients for all other variables unchanged.)
23 Garrett (fn. 3, 1998), 80; Roubini, Nouriel and Sachs, Jeffrey D., “Political and Economic Determinants of Budget Deficits in the Industrial Democracies,” European Economic Review 33, no. 5 (1998Google Scholar).
26 Cusack (fn. 24, 1997).
27 For recent examples see Cusack (fn. 24); Alexander Hicks, and Swank, Duane, “Politics, Institutions, and Welfare Spending in Industrialized Democracies, 1960–82,” American Political Science Review 86 (September 1992CrossRefGoogle Scholar); and Huber, Evelyn, Charles Ragin, and John Stephens, “Social Democracy, Christian Democracy, Constitutional Structure and the Welfare State,” American Journal of Sociology 99, no 3 (1993CrossRefGoogle Scholar).
28 Esping-Andersen, Gøsta, Markets against Politics: The Social Democratic Road to Power (Princeton:Princeton University Press, 1985Google Scholar); idem (fn. 7).
31 See the pathbreaking study by Isabela Mares on the role of employees in social policy formation. Mares, “Negotiated Risks: Employers' Role in Social Policy Development” (Ph.D. diss., Harvard University, 1998).
32 See, for example, ibid; Jo Martin, Cathie, “Nature or Nurture? Sources of Firm Preference for National Health Reform,” American Political Science Review 89 (December 1995CrossRefGoogle Scholar); and Swenson, Peter, “Arranged Alliance: Business Interests in the New Deal,” Politics and Society 25, no. 1 (1997CrossRefGoogle Scholar).
33 Wilensky, Harold L., The Welfare State and Equality (Berkeley:University of California Press, 1975Google Scholar).
34 Flora, Peter and Alber, Jens, “Modernization, Democratization, and the Development of Welfare States in Europe,” in Flora, and Heidenheimer, Arnold, eds., The Development of Welfare States in Europe and America (London:Transaction Books, 1981Google Scholar).
36 See Wilensky (fn. 33), 47.
37 There are passages in Harold L. Wilensky and Charles N. Lebaux's Industrial Society and Social Welfare that concern the problem of relocating and retraining workers made redundant in the process of industrialization in the U.S. This logic is clearly in line with our argument. We thank an anonymous reader for pointing this out. Wilensky, and Lebaux, . Industrial Society and Social Welfare: The Impact of Industrialization on the Supply and Organization of Social Welfare Services in the United States (New York:Russell Sage Foundation, 1958Google Scholar).
38 Garrett (fn. 3).
39 The attraction of the median voter depends on the electoral system. In two-party systems the mechanism is vote maximization, which leads to centrist appeals. See Downs, Anthony, An Economic Theory of Democracy (New York:Harper and Row, 1957Google Scholar), chaps. 7–8; and Cox, Gary W., “Centripetal and Centrifugal Incentives in Electoral Systems,” American Journal of Political Science 34, no. 4 (1990CrossRefGoogle Scholar). In multiparty systems the mechanism is office maximization through which governing coalitions come to recognize the importance of appealing to the median voter. See Laver, Michael and Schofield, Norman, Multiparty Government: The Politics of Coalition in Europe (Oxford:Oxford University Press, 1991Google Scholar).
40 Meltzer, Alan, and Richard, Scott, “A Rational Theory of the Size of the Government,” Journal of Political Economy 8, no. 9 (1981Google Scholar). See also Hicks and Swank (fn. 27).
41 Lijphart, Arend, “Unequal Participation: Democracy's Unresolved Dilemma—Presidential Address, “American Political Science Review 91, no. 1 (1997Google Scholar).
42 Esping-Andersen (fn. 7).
43 See Esping-Andersen, Gøsta, Changing Classes: Stratification and Mobility in Postindustrial Societies (London:Sage, 1993Google Scholar); idem, Social Foundations of Postindustrial Societies (Oxford:Oxford University Press, 1999Google Scholar); Iversen and Wren (fn. 8); and Glyn, Andrew, “Low Pay and the Volume of Work” (Manuscript, Corpus Christi College, Oxford University, 1997Google Scholar).
44 See Beck, Nathaniel and Katz, Jonathan N., “Nuisance vs. Substance: Specifying and Estimating Time-Series-Cross-Section Models,” in Freeman, John R., ed., Political Analysis, vol. 6 (Ann Arbor:University of Michigan 1996), 11Google Scholar.
45 The reason is that in the model using levels of spending on the left-hand side, much of the variance will be accounted for by the lagged dependent variable, showing simply that current spending depends on past spending.
46 See Beck, Nathaniel, “Comparing Dynamic Specifications: The Case of Presidential Approval,” in Stimson, James, ed., Political Analysis, vol. 2 (Ann Arbor:University of Michigan Press, 1992Google Scholar).
47 This does not have to be the case. One of the control variables, unexpected GDP growth, can in principal rise indefinitely.
48 The basic results we report below for the deindustrialization and globalization variables are robust to a large number of alternative model specifications. Thus, we have tried to group the data into time intervals using different periodizations; we have run the regression with and without the political variables, with and without the change terms; and we have tried to exclude one country at a time. In all cases the results for deindustrialization are strong and statistically significant, whereas for the globalization variables, they are weak and statistically insignificant, with the exceptions noted in the text.
49 An F-test indicates that the country dummies belong in the model.
51 Specifically, we used Roy E. Welch's “bounded-influence technique,” which weights individual cases if they exceed a size-adjusted cut-off point. Welch, “Regression Sensitivity Analysis and Bounded-Influence Estimation,” in Kmenta, Jan and Ramsey, James B., eds., Evaluation of Econometric Models (New York:Academic Press, 1980), 164Google Scholar–66. See also Fox, J., Regression Diagnostics (London:Sage Publications 1991), 31CrossRefGoogle Scholar.
52 Wood, Andrian, North-South Trade, Employment and Inequality: Changing Fortunes in a Skill-Driven World (Oxford:Oxford University Press, 1994Google Scholar); Learner, Edward E., Sources of International Comparative Advantage: Theory and Evidence (Cambridge:MIT Press, 1984Google Scholar); idem, “Trade Economists View of U.S. Wages and ‘Globalization’” (Paper presented at the Political Economy of European Integration Study Group Meeting, University of California, Berkeley, January 1996).
53 Rodrik (fn. 2), chap. 4.
54 We tested whether the effects of employment losses in agriculture differ from those in manufacturing by breaking the deindustrialization variable into two component variables: one variable defined as 100 minus manufacturing employment as a percent of the working-age population, and another defined as 100 minus agricultural employment as a percent of the working-age population. The results for both variables are similar to those for the deindustrialization variable, but when both variables are entered simultaneously, the manufacturing variable has a somewhat stronger effect (0.045 versus 0.035 for transfers, and 0.035 versus 0.022 for consumption). One reviewer suggested that a possible reason for this pattern can be found in social mobility studies, which show that a greater decline in agriculture occurs through intergenerational mobility than in manufacturing.
56 Esping-Andersen (fn. 7), 115.
58 A typical left government is defined here as one that is 1 standard deviation to the left of the mean on the partisan government variable. A typical right government is defined similarly.
59 See Appelbaum, Eileen and Schettkat, Ronald, “Employment and Productivity in Industrialized Countries,” International Labor Review 134, nos. 4–5 (1995Google Scholar); and Iversen and Wren (fn. 8).
60 Rowthorn, Robert, “Corporatism and Labour Market Performance,” in Pekkarinen, Jukka, Pohjola, Matti, and Rowthorn, , eds., Social Corporatism (Oxford:Clarendon Press, 1992Google Scholar); Waller-stein, Michael, “Wage-Setting Institutions and Pay Inequality in Advanced Industrial Societies,” American Journal of Political Science 43 (July 1999Google Scholar); Rueda, David and Pontusson, Jonas, “Wage Inequality and Varieties of Capitalism” (Working paper, Institute for European Studies, Cornell University, 1998Google Scholar).
61 Both figures are slightly exaggerated because there are small negative indirect effects of deindus-trialization (through growth and other variables) that reduce the net impact of deindustrialization. Yet these indirect effects do not affect the gap in the effect of deindustrialization between countries with centralized and decentralized bargaining structures, which is what is of interest here.
62 See, for example, Wood (fn. 52); Steven S. Saeger, “Globalization and Economic Structure in the OECD” (Ph.D. diss., Harvard University, 1996); and idem, “Globalization and Deindustrialization: Myth and Reality in the OECD,” Weltwirtschaftliches Archiv 133, no. 4 (1997Google Scholar).
63 Saeger (fn. 62,1997), 604.
64 Streeck, Wolfgang, “German Capitalism: Does It Exist? Can It Survive?” in Crouch, Colin and Streeck, , eds., Political Economy and Modern Capitalism: Mapping Convergence and Diversity (London:Sage, 1997Google Scholar).
65 Rowthorn, Robert and Ramaswamy, Ramana, “Deindustrialization: Causes and Implications” (International Monetary Fund Working Paper, no. 42, Washington, D.C., 1997Google Scholar); idem, “Growth, Trade, and Deindustrialization, International Monetary Fund Staff Papers 46 (March 1999Google Scholar); Krugman, Paul, Pop Internationalism (Cambridge:MIT Press, 1996Google Scholar).
66 Rowthorn and Ramaswamy (fn. 65,1999), 19.
67 We recognize that the debate over trade with LDC countries is broader than the question of whether it has led to widespread deindustrialization. It also involves issues such as the relative performance of particular manufacturing industries, relative prices, and above all, wage inequality. For a good introduction to the debate see the special issue of Journal of Economic Perspectives 9 (Special issue, 1995). For our purposes, however, the key issue is whether or not our main independent variable, deindustrialization, is in large measure the result of competition from low-wage countries.
68 Bacon, Robert and Eltis, Walter, Britain's Economic Problem: Too Few Producers (London:Macmillan, 1976Google Scholar).
69 See, for example, Meltzer and Richard (fn. 40); and Alesina, Alberto and Perotti, Roberto, “The Welfare State and Competitiveness,” American Economic Review 87 (December 1997Google Scholar).
70 As in the previous analysis, problems of heteroskedasticity led us to employ Beck and Katz's method for deriving panel corrected standard errors. Beck and Katz (fn. 50).
71 We could have used change in the log of the deindustrialization variable on the left-hand side in Table 4 without affecting the results (except for switching the signs of parameters, of course). We chose the current setup to stay as close as possible to the analysis by Rowthorn and Ramaswamy (fn. 65).
72 The countries include: Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Japan, Netherlands, Norway, Sweden, the United Kingdom, and the United States. Missing data problems precluded adding Switzerland. The time frame is the maximum possible given the availability of data.
74 Investment is measured as a percentage share of GDP. It is taken from the Summers, Robert and Heston, Alan, Penn World Tables, version 5.5, datafile (Cambridge:National Bureau of Economic Research, 1993Google Scholar).
75 The results of alternative specifications are available from the authors upon request.
76 Rowthorn and Ramaswamy (fn. 65, 1999).
78 OECD, Labour Force Statistics (Paris: OECD, various years); and transfer data.
79 Summers and Heston (fn. 74).
81 OECD (fn. 78).
83 International Institute for Democracy and Assistance, Electoral, Voter Turnoutfrom 1945 to 1997. A Global Report on Political Participation (Stockholm:IDEA Information Services, 1997Google Scholar).
84 OECD (fn. 78); and transfer data.
85 Thomas Cusack, “The Changing Contours of Government” (Wissenschaftszentrum Berlin fui Sozialforschung, WZB Discussion Paper, no. 304,1991); OECD (fn. 78); Stockholm International Peace Research Institute, The SIPRI Year Book (Stockholm:SIPRIGoogle Scholar, various years).
86 Cusack (fn. 85); and OECD, , National Accounts, Part II: Detailed Tables (Paris:OECDGoogle Scholar, various years)
87 Summers and Heston (fn. 74).
88 Cusack (fn. 24, 1997).
89 Castles, Francis and Mair, Peter, “Left-Right Political Scales: Some ‘Expert’ Judgments,” Europeai Journal of Political Research 12, no. 1 (1984Google Scholar).
91 Density data: Visser, Jelle, European Trade Union in Figures (Deventer, Netherlands:Kluwer Law and Taxation Publishers, 1989Google Scholar); and idem, “Unionization Trends Revisited” (Manuscript, University of Amsterdam, 1996). Centralization data: Iversen, Torben, “Wage Bargaining, Central Bank Independence and the Real Effects of Money,” International Organization 52 (Summer 1998CrossRefGoogle Scholar).
92 IMF, , Direction of Trade Statistics Yearbook (Washington, D.C.:IMFGoogle Scholar, various years).
93 OECD (fn. 86).
94 Roubini and Sachs (fn. 23); and OECD (fn. 86).