Wage Compression in Collectively Bargained Labour Contracts

Transcription

1 Wage Compression in Collectively Bargained Labour Contracts Theory and some evidence for Germany Michael C. Burda, Thorsten Vogel, and Nicole Walter Humboldt-Universität zu Berlin 2. Klausurtagung, Motzen, Juni 2006 Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

2 What do unions do? Do unions increase wages? (McDonald and Solow 1981, Oswald 1985, Nickell and Andrews 1983, Espinosa and Rhee 1989) Do unions compress the wage distribution? Pro (+): pay standardisation role of unions (Freeman and Medoff 1984, Lewis 1986, Horn and Svensson 1986, Agell and Lommerud 1992, Burda 1995) Contra (-): monopoly aspect of unionism (Friedman 1962, Minford 1983) Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

3 In Anglo-Saxon countries (US, Canada, UK) (Card, Lemieux and Riddell 2003, DiNardo, Fortin and Lemieux 1996, Machin 1997) union workers earn higher salaries than non-union workers wage dispersion (standard deviation of log wages) is lower in unionised sectors Problem to derive comparable empirical results for continental European countries because union coverage (Tarifbindung) is much higher than union membership (Mitgliedschaft) microdata have only recently become available to the scientific community (Germany: (1) linked employer-employee data, LIAB; (2) Verdienststrukturerhebung) Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

4 In Anglo-Saxon countries (US, Canada, UK) (Card et al. 2003, DiNardo et al. 1996, Machin 1997) union workers earn higher salaries than non-union workers wage dispersion (standard deviation of log wages) is lower in unionised sectors Problem to derive comparable empirical results for continental European countries because union coverage (Tarifbindung) is much higher than union membership (Mitgliedschaft) microdata have only recently become available to the scientific community (Germany: (1) linked employer-employee data, LIAB; (2) Verdienststrukturerhebung) Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

5 Coverage and membership (international perspective) Union density rate Bargaining coverage rate Country Period t 1 t 2 t 1 t 2 Austria (85-99) France (85-98) Sweden (85-98) Italy (85-98) Germany (West) (85) Germany (98) UK (85-97) EU (15) (85-97) Canada (85-98) United States (85-00) Source: Visser (2003) Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

6 Coverage and membership (Germany) Branchentarifvertrag Firmentarifvertrag Kein Tarifvertrag (davon Orientierung an Tarifvertrag) NUD Branche West Ost West Ost West Ost West Ost Landwirtschaft (Bergbau/Energie) (33.1) 67.5 (58.1) Bergbau/Energie (95.4) 6.9 (50.1) Grundstoffverarbeitung (63.7) 48.5 (61.0) Investitionsgüter (72.2) 55.0 (54.1) Verbrauchsgüter/Nahrungsmittel (50.0) 55.2 (54.8) Baugewerbe (62.4) 49.5 (69.2) Handel (46.4) 57.8 (48.0) Verkehr/Nachrichten (51.1) 41.7 (33.2) Dienste für Unternehmen (36.1) 57.5 (47.7) sonstige Dienste (51.4) 45.7 (54.4) Org. ohne Erwerbszweck (47.0) 52.7 (53.7) Dienste für Haushalte gesellsch. Dienste (z.b. Krankenhaus,Schulen) gesellsch. Dienste (z.b. Reinigung/Verbände) Gebietskörperschaften/Sozialversicherung (44.9) 1.7 (47.8) Insgesamt (49.3) 44.6 (53.8) Source: Kohaut/Schnabel IAB-Betriebspanel. 8. wave West and 5. wave East NUD: own calculations using SOEP data. Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

7 Wage distribution (West Germany) Median (p50) p80/p20 p50/p20 p80/p50 Branche Landwirtschaft (Bergbau/Energie) 2,390 2, Grundstoffverarbeitung 2,299 2, Investitionsgüter 2,355 2, Verbrauchsgüter/Nahrungsmittel 2,094 2, Baugewerbe 2,073 2, Handel 2,123 2, Verkehr/Nachrichten 2,186 2, Dienste für Unternehmen 2,709 3, Dienste für Haushalte 1,827 1, gesellsch. Dienste (z.b. Krankenhaus,Schulen) 2,397 2, gesellsch. Dienste (z.b. Reinigung/Verbände) 2,299 2, Gebietskörperschaften/Sozialversicherung 2,099 2, Insgesamt 2,237 2, Source: IAB-Beschäftigtenstichprobe. Monthly wages of full-time employed men in the old Länder (including West Berlin). Constant year 2000 prices. Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

8 Decomposing wage differences into age, skill, and industry components ln w it = age i + year t skill i + year t industry i + ε it Tobit Median (q50) Branche Grundstoffverarbeitung * -.023*.008 (.012) (.013) (.012) (.010) Investitionsgüter * -.023*.025** (.011) (.013) (.010) (.008) Verbrauchsgüter/Nahrungsmittel -.111** -.064* -.126** -.075** (.012) (.014) (.012) (.011) Baugewerbe -.118** -.121*.141** -.122** (.013) (.014) (.010) (.011) Handel -.120** -.099* -.142** -.106** (.013) (.014) (.011) (.011) Verkehr/Nachrichten -.086** ** -.093** (.013) (.015) (.012) (.012) Dienste für Unternehmen.032* -.076* (.014) (.015) (.012) (.008) Dienste für Haushalte -.284** -.315* -.246** -.253** (.023) (.027) (.028) (.029) gesellsch. Dienste (z.b. Krankenhaus,Schulen) -.162** -.113* -.127** -.041** (.015) (.016) (.014) (.011) gesellsch. Dienste (z.b. Reinigung/Verbände) -.134** ** -.093** (.017) (.020) (.020) (.020) Gebietskörperschaften/Sozialversicherung -.185* -.103* -.187** -.074** (.012) (.015) (.012) (.013) # obs 44,169 # persons 31,740 Source: IAB-Beschäftigtenstichprobe. Full-time employed men in West Germany in 1985 and 2001, spells weighted by length of employment spell. Constant year 2000 prices. Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

9 There are (economically) significant difference in wage rates between industries. union wage premia? Estimate ln w i = age i + skill i industry i + ε i separately for 1985 and Concentrate on full-time employed men aged between 20 and 35. (Leaves only 5 per cent of observations censored.) 1 Union wage premium: regress coefficients of skill industry dummies on net union density (NUD) (skill industry) j = c 1 + α 1 NUD j + ε 1 j 2 Union wage compression: compute q80/q20 ratio for each skill industry group. Regress this ratio on NUD ( ) q80 = c 2 + α 2 NUD j + ε 2 j q20 j Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

10 There are (economically) significant difference in wage rates between industries. union wage premia? Estimate ln w i = age i + skill i industry i + ε i separately for 1985 and Concentrate on full-time employed men aged between 20 and 35. (Leaves only 5 per cent of observations censored.) 1 Union wage premium: regress coefficients of skill industry dummies on net union density (NUD) (skill industry) j = c 1 + α 1 NUD j + ε 1 j 2 Union wage compression: compute q80/q20 ratio for each skill industry group. Regress this ratio on NUD ( ) q80 = c 2 + α 2 NUD j + ε 2 j q20 j Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

11 (1) Union wage premium estimated net union density / NUD (mean) individualeffect y Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

12 (1) Union wage premium estimated net union density / NUD (mean) individualeffect y Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

13 (1) Union wage compression 1985 // p80 p20 difference estimated net union density / NUD diff8020 y Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

14 (1) Union wage compression 2001 // p80 p20 difference estimated net union density / NUD diff8020 y Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

15 What do unions do? Why should we expect unions to affect the wage distribution? Insurance effect of unionisation (Horn and Svensson 1986, Agell and Lommerud 1992, Burda 1995) When bargaining over labour contracts workers are ignorant of their future locale, say, their future skill, occupational status, regional effects, and the like. Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

16 Model 1 Large number of locales (qualifications, regions, industries) 2 After labour contract has been agreed on, each locale is hit by shock 3 Firms and unions can commit to the contract 4 Labour contract covers wage rates w (θ), and employment rates, ρ(θ) 5 Unit mass of workers 6 Expected utility EU = [ρ(θ) U (w (θ),0) + (1 ρ(θ)) U (b, 1)] g (θ) dθ 7 Given profits of firms π Problem of the union [θf (ρ(θ)) ρ(θ) w (θ)] g (θ) dθ (1) max EU s.t. (1) {ρ(θ),w(θ)} Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

17 Counterfactual: Wage distribution on spot labour markets Moments: w m = θf (1) E [w m ] = E [θ] f (1) Var [w m ] = E [(θf (1) E [θ] f (1)) 2]. = Var [θ] [f (1)] 2 Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

18 Optimal labour contract For each θ the state contingent labour contract {w (θ),ρ (θ)} maximises the Lagrangian F (w,ρ;θ) = U (b, 1) + ρ [U (w, 0) U (b, 1)] λ [θf (ρ) ρw] where λ < 0 is a constant for all θ(hestenes 1980, theorem 5.1, p 215). First-order conditions U c (w, 0) + λ = 0 U (w, 0) U (b, 1) λ [θf (ρ ) w ] 0 0 < ρ 1 Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

19 Findings 1 Wages are constant across states of the world. 2 Suppose there is no disutility of working and benefits b are financed through taxes on labour. Then there is full employment across states of the world. 3 The greater the bargaining power of firms (i.e., the larger π), the smaller wages w and employment ρ. Strong unions get higher wages and more employment than weak unions. Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

20 Result 1 may be too strong. Introduce positive finite cost, x, for switching from the unionised to the non-unionised sector. Additional constraint to the union maximisation problem ( Günstigkeitsprinzip ) w (θ) w spot (θ) x Result 4: If utility is strictly concave in w (under some regularity conditions) there is a productivity θ θ such that for all θ θ the union wage rate is constant. This wage may be referred to as a minimum wage. For all θ > θ both wages and migration increase in θ. Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

21 Conclusion 1 Empirical finding for Germany ambiguous ( use better data that have just become available!!!) 2 Simple wage insurance model explains union wage premium (that is found in Anglo-Saxon countries) as other right-to-manage models do. 3 Strong prediction concerning wage compression 4 Adding Günstigkeitsprinzip results in minimum wage (agreed on by the union, not the government) Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

22 Appendix real Euros log real Euros median (West) median (East) log of median (West) log of median (East) IAB Beschäftigtenstichprobe, year 2000 prices, Land Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

23 Appendix real wages % 50 % 80 % IAB Beschäftigtenstichprobe; West: 1984=100, East: 1992=100, Land Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

24 Appendix 4500 west 4000 real Euros ohne berufsausbildun abitur ohne berufsau fachhochschulabschlu BMG IAB Beschäftigtenstichprobe, year 2000 prices, Land volks, haupt, real abitur mit berufsaus hochschulabschluß Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

25 Appendix % 1% 3% 3% 21% 1% 4%5% 7% 13% 71% 71% ohne berufsausbildung abitur ohne berufsausbildung fachhochschulabschluß Graphs by Jahr volks, haupt, realschule mit berufsausbildung abitur mit berufsausbildung hochschulabschluß Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

26 Appendix %3%2%6% 4% 3% 11% 8% 3% 5%3%5% 7% 6% 14% 22% 25% 10% 7% 8% 12% 10% 11% 11% Graphs by Jahr Landwirtschaft/Bergbau/Energie Grundstoffverarbeitung Investitionsgüter Verbrauchsgüter/Nahrungsmittel Baugewerbe Handel/Reparatur Verkehr/Nachrichten Dienste für Unternehmen Dienste für Haushalte gesellschaftsbez. Dienste (Krankenhaus/Schulen) gesellschaftsbez. Dienste (Reinigung/Verbände) Gebietskörperschaften/Sozialversicherung Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

27 Agell, J. and Lommerud, K. E.: 1992, Union egalitarianism as income insurance, Economica 59(235), Burda, M. C.: 1995, Unions and wage insurance, CEPR Discussion Paper Series No Card, D., Lemieux, T. and Riddell, W. C.: 2003, Unions and the wage structure, in J. T. Addison and C. Schnabel (eds), International Handbook of Trade Unions, Edward Elgar, Cheltenham, UK, chapter 8, pp DiNardo, J., Fortin, N. M. and Lemieux, T.: 1996, Labor market institutions and the distribution of wages, : A semiparametric approach, Econometrica 64(5), Espinosa, M. P. and Rhee, C.: 1989, Efficient wage bargaining as a repeated game, Quarterly Journal of Economics 104, Freeman, R. B. and Medoff, J. L.: 1984, What Do Unions Do?, Basic Books, New York. Friedman, M.: 1962, Capitalism and Freedom, University of Chicago Press, Chicago. Hestenes, M. R.: 1980, Calculus of Variations and Optimal Control Theory, Robert E. Krieger Publishing Company, Huntington, New York. Horn, H. and Svensson, L. E. O.: 1986, Trade unions and optimal labour contracts, Economic Journal 96(382), Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24

28 Lewis, H. G.: 1986, Union Relative Wage Effects, University of Chicago Press, Chicago. Machin, S.: 1997, The decline of labour market institutions and the rise in wage inequality in britain, European Economic Review 41, McDonald, I. and Solow, R. M.: 1981, Wage bargaining and employment, American Economic Review 71(5), Minford, P.: 1983, Unemployment: Causes and Cure, Martin Robertson, London. Nickell, S. J. and Andrews, M.: 1983, Unions, Real Wages and Employment in Britain , Oxford Economic Papers 35(Supplement), Oswald, A.: 1985, The economic theory of trade unions: An introductory survey, Scandinavian Journal of Economics 87, Burda, Vogel, Walter (HU-Berlin) Wage Compression Motzen / 24