Divergent trends and different causal logics: The importance of bargaining centralization when explaining earnings inequality across advanced democratic societies

Politics and Society

Volumn 33, Issue 3, 2005, Pages 359-385

  Oskarsson, Sven ^a  

^a UPPSALA UNIVERSITY   (Sweden)

Author keywords

Bargaining centralization; Comparative political economy; Institutions; Varieties of capitalism; Wage inequality

Indexed keywords

EID: 24944542892     PISSN: 00323292     EISSN: None     Source Type: Journal    
DOI: 10.1177/0032329205278460     Document Type: Review

References (69)

1. Frank Levy and Richard J. Murnane, "U.S. Earnings Levels and Earnings Inequality: A Review of Recent Trends and Proposed Explanations," Journal of Economic Literature 30, (1992): 1333-81;
2. Gary Burtless, "International Trade and the Rise in Earnings Inequality," Journal of Economic Literature 33 (1995): 800-16;
3. OECD, "Earnings Inequality, Low-Paid Employment and Earnings Mobility," in OECD Employment Outlook (Paris: OECD, 1996), 59-108;
4. Peter Gottschalk, "Inequality, Income Growth, and Mobility: The Basic Facts," Journal of Economic Perspectives 11 (1997): 21-40;
5. Peter Gottschalk and Timothy Schmeeding, "Cross-National Comparisons of Earnings and Income Inequality," Journal of Economic Literature 35 (1997): 633-87.
6. Gottschalk and Schmeeding, "Cross-National Comparisons," 633-87.
7. The reason for choosing the 50-10 ratio is that some of the causal hypotheses to be elaborated below first and foremost concern the lower part of the earnings distribution. As will be discussed in the empirical part of the paper, the choice of using the 50-10 ratio as a measure of wage inequality does not unduly influence the results.
8. Adrian Wood, North-South Trade, Employment and Inequality (Oxford: Clarendon Press, 1994);
9. Burtless, "International Trade and the Rise," 800-16;
10. Richard B. Freeman, "Are Your Wages Set in Beijing?" Journal of Economic Perspectives 9 (1995): 15-32;
11. Davis J. Richardson, "Income Inequality and Trade: How to Think, What to Conclude," Journal of Economic Perspectives 9 (1995): 33-55;
12. Adrian Wood, "How Trade Hurts Unskilled Workers," Journal of Economic Perspectives 9 (1995): 57-80;
13. Matthew J. Slaughter and Phillip Swagel, "The Effect of Globalization on Wages in the Advanced Economies," IMF working paper no. 97:43 (Washington, D.C.: International Monetary Fund, 1997).
14. Freeman, "Are Your Wages Set in Beijing?" 15-32;
15. Slaughter and Swagel, "The Effect of Globalization."
16. John Bound and George Johnson, "Changes in the Structure of Wages in the 1980s: An Evaluation of Alternative Explanations," American Economic Review 82 (1992): 371-92;
17. Levy and Murnane, "U.S. Earnings Levels and Earnings Inequality," 1333-81;
18. George E. Johnson, "Changes in Earnings Inequality: The Role of Demand Shifts," Journal of Economic Perspectives 11 (1997): 41-54;
19. Robert H. Topel, "Factor Proportions and Relative Wages: The Supply-Side Determinants of Wage Inequality," Journal of Economic Perspectives 11 (1997): 55-74.
20. Richard B. Freeman, "Unionism and the Dispersion of Wages," Industrial and Labor Relations Review 34 (1980): 3-23;
21. Richard B. Freeman, "Union Wage Practices and Wage Dispersion within Establishments," Industrial and Labor Relations Review 36 (1982): 3-21;
22. John DiNardo, Nicole M. Fortin, and Thomas Lemieux, "Labor Market Institutions and the Distribution of Wages: A Semiparametric Approach," Econometrica 64 (1996): 1001-44;
23. Nicole M. Fortin and Thomas Lemieux, "Institutional Changes and Rising Wage Inequality: Is There a Linkage?" Journal of Economic Perspectives 11 (1997): 75-96.
24. Freeman, "Unionism and the Dispersion of Wages," 3-23;
25. Freeman, "Union Wage Practices," 3-21;
26. Jonas Pontusson, David Rueda, and Christopher R. Way, "Comparative Political Economy of Wage Distribution: The Role of Partisanship and Labour Market Institutions," British Journal of Political Science 32 (2002): 282-308.
27. Rune Åberg, "Market-Independent Income Distribution: Efficiency and Legitimacy," in Order and Conflict in Contemporary Capitalism, ed. J. H. Goldthorpe (Oxford: Clarendon Press, 1984), 209-30;
28. Francine D. Blau and Lawrence M. Kahn, "International Differences in Male Wage Inequality: Institutions versus Market Forces," Journal of Political Economy 104 (1996): 791-837;
29. Douglas A. Hibbs Jr. and Håkan Locking, "Wage Compression, Wage Drift and Wage Inflation in Sweden," Labour Economics 3 (1996): 109-41;
30. Torben Iversen, "Power, Flexibility, and the Breakdown of Centralized Wage Bargaining: Denmark and Sweden in Comparative Perspective," Comparative Politics 28 (1996): 399-436;
31. Karl Ove Moene and Michael Wallerstein, "Social Democratic Labor Market Institutions: A Retrospective Analysis," in Continuity and Change in Contemporary Capitalism, ed. H. Kitschelt, P. Lange, G. Marks, and J. D. Stephens (Cambridge: Cambridge University Press, 1999), 231-60;
32. Michael Wallerstein, "Wage-Setting Institutions and Pay Inequality in Advanced Industrial Societies," American Journal of Political Science 43 (1999): 649-80;
33. Douglas A. Hibbs Jr. and Håkan Locking, "Wage Dispersion and Productive Efficiency: Evidence for Sweden," Journal of Labor Economics 18 (2000): 755-82;
34. Lawrence M. Kahn, "Wage Inequality, Collective Bargaining, and Relative Employment from 1985 to 1994: Evidence from Fifteen OECD Countries," Review of Economics and Statistics 82 (2000): 564-79;
35. Miriam Golden and John Londregan, "Centralization of Wage Bargaining and Pay Dispersion: A Critique of Wallerstein" (University of California at Los Angeles, 2002).
36. Torben Iversen, Contested Economic Institutions (Cambridge: Cambridge University Press, 1999).
37. See also Wallerstein, "Wage-Setting Institutions and Pay Inequality," 649-80.
38. Jonas Pontusson and David Rueda, "Wage Inequality and Varieties of Capitalism," World Politics 52 (2000): 350-83;
39. Geoffrey Garrett, Partisan Politics in the Global Economy (Cambridge: Cambridge University Press, 1998);
40. Dani Rodrik, Has Globalization Gone Too Far? (Washington, D.C.: Institute for International Economics, 1997).
41. Pontusson and Rueda, "Wage Inequality and Varieties of Capitalism," 350-83;
42. Geoffrey Garrett and Christopher Way, "Public-Sector Unions, Corporatism, and Wage Determination," in Unions, Employers, and Central Banks, ed. T. Iversen, J. Pontusson. and D. Soskice (Cambridge: Cambridge University Press, 2000), 267-91.
43. Stephen Nickell and Brian Bell, "Changes in the Distribution of Wages and Unemployment in OECD Countries," American Economic Review 86 (1996): 302-8;
44. Gottschalk and Schmeeding, "Cross-National Comparisons," 633-87.
45. Pontusson and Rueda, "Wage Inequality and Varieties of Capitalism," 350-83, is an exception to this rule, and I will return to their work when discussing the results of the empirical analysis.
46. Wood, North-South Trade;
47. Nickell and Bell, "Changes in the Distribution of Wages," 302-8.
48. It should be emphasized that this argument does not contradict the argument made earlier about the indeterminate effect of public employment on pay inequality, given decentralized bargaining. That argument was driven by the claim that the overall impact of the within-sector compression of wages depends on the distribution of public sector employees across the wage hierarchy. However, to the degree that wage developments are linked through centralized bargaining, wage compression in the public sector will spill over into the private sector and result in decreased pay inequality across the labor market.
49. The results of the regression equations reported below also include the rate of unemployment as a control variable. The argument for including unemployment in the analyses is that employers are more likely to lay off unskilled than skilled workers during a recession (Katharine L. Bradbury, "Rising Tide in the Labor Market: To What Degree Do Expansions Benefit the Disadvantaged?" New England Economic Review 32 (2000): 3-33).
50. To enable estimation using panel techniques, Switzerland or Denmark had to be omitted from the data on earnings inequality presented in Table 1. The choice fell on Switzerland in order to maximize the number of observations.
51. Bernard Ebbinghaus and Jelle Visser, The Societies of Europe: Trade Unions in Western Europe since 1945 (Oxford: Macmillan Reference, 2000).
52. Giancarlo Barbiroli, "Measuring Technological Dynamics and Structural Change, Their Interrelationships and Their Effects," Structural Change and Economic Dynamics 6 (1995): 377-96.
53. Iversen, Contested Economic Institutions.
54. The centralization measure and a description of how it is calculated can be found at http://www.people.fas.harvard.edu/~iversen/centralization.htm. To make sure the results obtained are not merely driven by the choice of centralization indicator, I reran all models using an alternative measure developed by Golden, Lange, and Wallerstein (the variable CONINVfrom
55. Miriam Golden, Michael Wallerstein, and Peter Lange, "Union Centralization among Advanced Industrial Societies: An Empirical Study," http://www.shelley. polisci.ucla.edu/data/, data set). These results, not reported here but available from me upon request, are, apart from the estimated effects of research and development, very similar to the ones presented in this article and do not alter the conclusions drawn. The reason for presenting the models using Iversen's indicator is to ensure the highest possible validity. Unlike the Golden-Lange-Wallerstein measure, Iversen's indicator focuses on the degree of centralization itself rather than on the extent of centralization-related activities by confederations. However, a drawback with Iversen's measure is that it is restricted to the time span 1970-1993. Since comparative data on research and development are available only from 1977, the time span covered in the empirical analyses is 1977-1993. For a review of measures of wage-setting institutions used in earlier studies,
56. see Lane Kenworthy, "Wage-Setting Measures: A Survey and Assessment," World Politics 54 (2000): 57-98.
57. To account for the possibility of country-specific intercepts, the models include a full set of n - 1 country dummies. To obtain unbiased standard errors in the face of contemporaneous correlation and heteroskedasticity, I use Beck and Katz's panel corrected standard errors (Nathaniel Beck and Jonathan N. Katz, "What to Do [and Not to Do] with Time-Series Cross-Section Data," American Political Science Review 89 [1995]: 634-47). In order to separate short-term and long-term effects and, from a more technical point of view, to correct for time-wise autocorrelation, the models include a lagged dependent variable on the right-hand side (partial adjustment model). The estimates in Table 4 are short-term effects. "To obtain the long-term effects, divide the value of the coefficient for the variable of interest by one minus the coefficient for the lagged dependent variable." Interpolated observations account for 14 out of 171 used in models 1 and 2, below.
58. As with any interaction model, one has to be careful when interpreting the conditional effects of a variable at the extreme ends of the intervening variable. To be sure, the example above is an out-of-sample prediction since no one country in the sample scores 0 or 1 on the centralization variable. The empirically relevant conditional effects are instead presented in Table 5.
59. Robert J. Friedrich, "In Defense of Multiplicative Terms in Multiple Regression Equations," American Journal of Political Science 26 (1982): 797-833.
60. In this way, the cross-validation bounds can be seen as a less formalized way of conducting a sensitivity analysis. Furthermore, computing and reporting cross-validation bounds seem especially suitable for models using time-series cross-section data. It is, of course, problematic if the results of a model are driven by the inclusion or not of a single country. By repeatedly estimating the model and excluding one country each time, we can directly test to which degree the magnitudes and significances of the coefficients are dependent on the inclusion of a specific country. One must also decide upon some criterion for when the cross-validation bounds signal problems with results driven by single countries. In this article, I have opted for an informal rule of thumb. Special attention should be given to cases where the exclusion of a single country changes the sign of a significant coefficient - that is, when the cross-validation bounds around a significant effect include zero.
61. The vif-statistic for the centralization variable is alarmingly high-172-indicating that less than 1 percent of the independent variation is left after taking account of covariation with the other explanatory factors. It should also be noted that the positive effects of centralization in model 2 do not contradict the well-established findings of earlier research. When evaluating the conditional effect of centralization at the sample means of the five interacting variables, the estimated impact is negative: -0.08.
62. David Soskice, "Divergent Production Regimes: Coordinated and Uncoordinated Market Economies in the 1980s and 1990s." in Continuity and Change in Contemporary Capitalism, ed. H. Kitschelt, P. Lange, G. Marks, and J. D. Stephens (Cambridge: Cambridge University Press, 1999), 101-34;
63. Peter A. Hall and David Soskice, "An Introduction to Varieties of Capitalism," in Varieties of Capitalism, ed. P. A. Hall, and D. Soskice (Oxford: Oxford University Press, 2001) 1-68.
64. Pontusson and Rueda, "Wage Inequality and Varieties of Capitalism," 350-83.
65. See Pontusson and Rueda, "Wage Inequality and Varieties of Capitalism," 350-83. These are Austria, Belgium, Denmark, Finland, France, Germany, Italy, Netherlands, Norway, and Sweden. Pontusson and Rueda categorize France and Italy as mixed cases. However, judging by their own criteria, the two countries conform very well to the SME cluster. Thus, they are included in model 3.
66. These are Australia, Canada, United Kingdom, Japan, and United States.
67. The simulation procedure assumes that the centralization variable is constant over time, and I therefore use the mean centralization scores for Norway (1980-1993) and the United States (1977-1993) presented in table 2.
68. β is the coefficient of the lagged dependent variable. To calculate the simulated effects, I have to assume that the changes in the explanatory variables between 1977/1980 and 1993 are lasting. However, since it takes less than three years for the explanatory factors to reach 90 percent of their long-term effects, this assumption is not particularly constraining. Looking at the trends in the independent variables over the last three to four years before 1993 in Norway and the United States, the changes are only marginal compared to the overall change between 1977/1980 and 1993.
69. 10) as a measure of earnings inequality that can take on any value between negative and positive infinity. This specification did not alter any of the conclusions drawn above. Finally, the results of models 1 and 2 did not change when including Norway and the United States in the sample-compare the results of model 5 (including Norway and the United States in the sample) with the corresponding results of model 1.