The Skill Loss of Older East Germans after Unification

Gathmann, Christina Article The Skill Loss of Older East Germans after Unification Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies Provided in Cooperation with: Duncker & Humblot, Berlin Suggested Citation: Gathmann, Christina (2005) : The Skill Loss of Older East Germans after Unification, Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies, ISSN 1865-5742, Duncker & Humblot, Berlin, Vol. 125, Iss. 1, pp. 7-16, https://doi.org/10.3790/schm.125.1.7 This Version is available at: https://hdl.handle.net/10419/292104 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Schmollers Jahrbuch 125 (2005), 7-16 Duncker & Humblot, Berlin The Skill Loss of Older East Germans after Unification By Christina Gathmann* Abstract This paper documents and analyzes changes in relative wages of older East Germans after unification. In 1990, employees above 45 started out with very low wages relative to younger East Germans or West Germans in the same cohort. Results suggest that older employees lost up to 30 percent of their earnings because socialist work experi ence fully depreciated after the regime change. Older employees lost further in relative terms because their retums to education and occupational skills were much lower than younger East Germans early in the transition. lEL Classifications: 131, 138, 161, P51 1. Introduction Older cohorts are often considered especially vulnerable to the sweeping changes of the transition as their labor market skills are obsolete or less adap table to the new economic environment. Prior evidence from Russia suggests that older employees suffered substantial wage losses relative to the average employees and especially recent labor market entrants during the l 990s (Brainerd 1998). Since East Germany experienced the most rapid transition to a market economy, older cohorts might be hit especially hard.1 * I would like to thank James Heckman, Steve Levitt, Michael Greenstone, and Gary Becker for their advice and suggestions. Holger Bonin, Dan Hamermesh, Dean Lillard, Derek Neal, Peter Orazem, Susan Vroman and participants at the ZEW Seminar in Mannheim, the SOEP User Conference in Berlin, the SOLE Meeting in San Antonio and the IZA / SOLE Transatlantic Meeting provided additional useful comments. I also thank the German Institute for Economic Research (DIW) in Berlin for its hospitality and the Henry Morgenthau Jr. Memorial Fund at the University of Chicago for financial support. All remaining errors are mine. 1 Gross Domestic Product declined by 15.6 percent in 1990 and another 22.7 percent in 1991 -more than in most other transition countries. Mirroring the initial collapse of production, aggregate employment plummeted by 25 percent in the first two years and declined a further 10 percent in 1992. Schmollers Jahrbuch 125 (2005) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.125.1.7 | Generated on 2023-04-04 12:34:44 8 Christina Gathmann This paper docmnents and analyzes the detenninants of changes in relative wages of older East Gennans shortly after unification. The data reveal two stylized facts. First, older employees start out with just 25 percent of the wage of West Germans in the same cohort in 1990. In contrast, young employees bom after 1965 eam more than 40 percent of their cohort counterparts in the West. Second, wages of older East Gennans rise at a similar rate as for young er employees after 1991. New results are presented that quantify the initial wage lass from limited transferability of skills across economic regimes. Calculations show that co horts with 35 or more years of experience in 1990 lose between 23 and 30 percent of their earnings power compared to a situation where returns had remained at 1989 levels. In addition, older employees also eam lower retums to educational and occupational skills than younger East Gennans. In contrast, differences in educational attainment across cohorts or their distribution across industries and occupations explain only a small fraction of the relative wage lass early in the transition.2 The next section describes the changes in the relative wages after unifica tion. Section 3 quantifies the wage lasses of older workers due to skill depre ciation, while Section 4 concludes. 2. Changes in Relative Wages after Unification The empirical analysis is based on the German Socio-Economic Panel from 1990 until 2001.3 East Gennany experienced remarkable aggregate wage growth over the 1990s with average growth rates of 14 log points. As a conse quence, there was substantial wage convergence between East and West Ger many over the 1990s. Focusing on aggregate wages as done by the literature (Burda and Schmidt 1997, Franz and Steiner 2000) however masks substantial heterogeneity in convergence across different cohorts. Figure 1 shows wages of East Germans relative to their cohort counterpart in West Germany. Two facts emerge: first, older cohorts in East Gennany start out with lower relative wages shortly after unification. The two oldest cohorts, between age 35 -44 and 45 -54 in 1990 respectively, earn just about one fourth of West Germans in the same cohort. In contrast, the youngest cohort under 25 in 1990 starts out with close to 40 percent of the wages eamed by their cohort counterparts in West Germany. Second, all cohorts experience considerable wage convergence over the course of the transition. Ten years after unification, the cohort bom after 1965, now 35 years and younger, eams 2 For a complete analysis of initial wage lasses and later recovery of older East Ger mans, see Gathmann (2004a). 3 For details on construction of the sample, see Gathmann (2004a). Schmollers Jahrbuch 125 (2005) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.125.1.7 | Generated on 2023-04-04 12:34:44 Tue Skill Loss of Older East Germans after Unification 0,85 gJ, 0,75 +------- ---- --- �- "§ 0,65 i------���,,,._--7"',lr_--::::���====!':==:::�'__j 0 I 0 055 t-------,--,.�--:.,,,""-----------------------1 0 1ii 0,45 -i-------,--,------,�-------------------------j � 0,35 ,--�-----,�---------------------------j 0,25 -i-----�----�---�----�---�---------j 1990/91 1992/93 1994/95 1996/97 1998/99 2000 Year 1-• •After1965 ......,._1956-65 -1946-55 -+-1936-45 I 9 Source: Author's calculations based on the German Socio-Economic Panel, Waves 1990-2002. Figure 1: East-West Wage Ratio by Birth Cohort ahnost 80 percent of West Gennan wages while the cohort bom between 1936-45 approaching retirement eam around 65 percent.4 Another way to illustrate the initial wage loss for older employees is to compare relative wages in East Germany before and after the regime change. Median wages for employees aged 45 or above fell by 10 percent relative to the 25 -34 age group between 1989 and 1991. 5 Similarly, median wages of 45 -54 year-olds drop from the 65th percentile in 1989 to the 57th percentile in 1991 in the wage distribution of under 35 year-olds. The evidence above is based on the sample of workers and potentially af fected by changes in the composition of the workforce. While employment has declined for all age groups early in the transition process, the decline has been most pronounced for older East Germans. If those dropping out of the labor market come from the bottom of the wage distribution, relative wages of older cohorts from a sample of workers are upward biased. This would under state their relative wage losses early in the transition and overstate their 'keeping up' later in the transition. Suppose however that all nonworkers in 4 A sirnilar picture ernerges for relative wages across age groups within East Ger rnany. For exarnple, rnen above 45 earn only about 10 percent rnore than younger ern ployees in 1990, whereas West German rnen above 45 earn 17 to 25 percent rnore. s Only relative wages are cornpared. In the absence of a price deflator that accounts for subsidies and quantity constraints in the socialist econorny, wage levels between 1989 and 1991 are not directly cornparable. Schmollers Jahrbuch 125 (2005) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.125.1.7 | Generated on 2023-04-04 12:34:44 10 Christina Gathmann East and West Germany are frmn the lower half of the wage distribution and the fraction of nonworkers in the sample does not exceed fifty percent. Under this assumption, the median of the füll wage distribution can be recovered from the observed wages by adjusting the median of workers for the fraction of censored observations from nonworkers (Neal and Johnson 1996).6 Figure 2 shows median relative wages between East and West including nonworkers under this assumption. The corrected plot still exhibits substantial wage growth for all cohorts during the 1990s, but younger cohorts now experi ence more rapid wage growth than older cohorts. As a result, the wage pro files of different birth cohorts fan out over the 1990s and the convergence gap between the oldest (born 1936-45) and youngest cohort (born after 1965) more than doubles between 1990 and 2000. 0,85 --- -■- - -■ 0,25 +----�----�---�---�----�---------, 1990/91 1992/93 1994/95 1996/97 1998/99 2000 Year 1-• •After1965 ......,__1956-65 -1946-55 -1936-45 I Source: Author's calculations based on the German Socio-Economic Panel, Waves 1990-2002. Figure 2: East-West Wage Ratio Including Nonworkers 6 The assumption that all nonworkers earn wages below the median for workers is strong. If some nonworkers are in fact high-wage workers, the corrected plot under states the wage gains of East relative to West Germans in the labor market. More im portantly, relative wages in East Germany are misleading if young and old labor market dropouts come from different parts of the wage distribution. For example, if young nonworkers are high-wage earners and older nonworkers low-wage earners, the fanning out would still understate the true relative wage gains of younger workers. Schmollers Jahrbuch 125 (2005) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.125.1.7 | Generated on 2023-04-04 12:34:44 Tue Skill Loss of Older East Germans after Unification 3. The Transferability of Skills across Economic Regimes 3.1 Depreciation of Socialist Work Experience 11 One explanation for the low relative wages of older employees documented above is that some skills are only partially transferable across economic re gimes. In particular, older cohorts worked a greater fraction of their career in the socialist economy, which often used outdated technologies or production methods. If those skills are not valuable in the new market economy, older employees suffer relative and potentially absolute wage losses. To estimate the retums to labor market experience carried over from the socialist regime, the following eamings equation is specified7 where OExp denotes old socialist work experience, NExp work experience since unification and X other control variables like education and demographic characteristics. 8 In the above specification, older East Germans might suffer wage losses for two reasons: first, low retums to old experience decrease the wages relative to younger cohorts. Second, older workers might also eam low er retums to new experience, if for example, their productivity to acquire new skills is lower than for younger workers. To capture the latter, new experience and new experience squared are also interacted with age dummies. The results for East German men and women are reported in Table 1. The estimates show that socialist labor market experience has no economic value in the post-unification labor market. Returns to 'socialist' work experience are not statistically significant from zero for men and small or zero for women across all specifications. In contrast, retums to work experience accumulated after unification are very large, especially for men. The much lower retums to new work experience for women suggest that transition affected men and wo men differently.9 Column (2) and (5) add occupation and industry dummies. Conditional on occupation and sector, retums to new experience are some what lower. Column (3) and (6) interacts the new work experience variable with age dummies with the reference category being those under 25. For both men and women, retums to new work experience are not statistically different 7 See Mincer and Ofek (1982) for an application to female labor force participation. 8 Since labor force participation was high and unemployment rates below 2 percent in the socialist economy, the empirical measure of old experience is essentially a dum my variable for people bom in the same year and the same years of schooling. The new work experience variable is derived from calendar files that report the actual employ ment status for each month. Variation in new experience across individuals arises from unemployment and temporary nonemployment spells after 1990. 9 For a detailed analysis of how the transition process affected East German men and women, see Gathmann (2004b). Schmollers Jahrbuch 125 (2005) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.125.1.7 | Generated on 2023-04-04 12:34:44 12 Christina Gathmann across age groups (not reported). The F-Test for joint significance reported in the last row of Table 1 is rejected at the 5 percent level. 10 Table 1 Depreciation of Socialist Work Experience after Unification Men Women (1) (2) (3) (4) (5) (6) Old Experience 0,002 0,0016 -0,0009 0,0124 0,0071 0,0029 (0,0017) (0,0017) (0,0027) (0.0018)** (0.0017)** (0,0030) Old Experience Squared 0 -0,0001 -0,0001 -0,0003 -0,0002 -0,0001 0,0000 0,0000 (0,0001) (0.0000)** (0.0000)** (0,0001) New Experience 0,5847 0,5384 0,5486 0,1942 0,1431 0,0864 (0.1874)** (0.1996)** (0.2087)** (0.0624)** (0.0616)* (0,0826) New Experience Squared -0,0351 -0,0336 -0,0386 0,0032 -0,0043 0,0009 (0.0123)** (0.0129)** (0.0158)* (0,0046) (0,0045) (0,0095) Vocational Training 0,1403 0,0885 0,075 0,305 0,154 0,1553 (0.0339)** (0.0338)** (0.0347)* (0.0278)** (0.0264)** (0.0265)** University Degree 0,4658 0,2416 0,221 0,6585 0,3133 0,3084 (0.0356)** (0.0375)** (0.0383)** (0.0310)** (0.0302)** (0.0307)** Migrant or Commuter to West 0,1703 0,2032 0,2034 0,1004 0,151 0,1496 (0.0173)** (0.0174)** (0.0175)** (0.0244)** (0.0231)** (0.0231)** Year Dummies Yes Yes Yes Yes Yes Yes State Dumrnies Yes Yes Yes Yes Yes Yes Age Dumrnies No No Yes No No Yes Other Controls Yes Yes Yes Yes Yes Yes Industry Dumrnies No Yes Yes No Yes Yes Occupation Dumrnies No Yes Yes No Yes Yes R-Squared 0,39 0,47 0,47 0,39 0,55 0,55 Observations 7001 6733 6733 6642 6398 6398 F-Test Age-Specific Returns 2,35 2,27 Prob> F (0,0159) (0,0204) Source: author's calculations based on the German Socio-Economic Panel, Waves 1990-2002. Notes: The results are based on a pooled regression of log hourly wages for East German men and women between 20 and 60 years of age in 1990. People in füll-time education or vocational training ("berufliche Lehre") are excluded from the analysis. Robust standard errors are reported in parentheses. Coefficients with * are significant the 5 percent, those with ** at the 1 percent level. New experience refers to actual work experience in years accumulated after unification, while old experience is potential experience before 1990. The reference educational group is no vocational degree and the reference age group are those under 25. Other controls are whether the person is married and in column (2)-(3) and (5)-(6) also firm tenure. The occupation and industry dummies in column (2) and (5) control for 7 occupational and 12 industry categories. 10 Very similar results were found when the sample was restricted to those working in East Germany. Interaction terms between old work experience and education were not significant. Thus, the depreciation of socialist skills appears to have affected all education levels in a similar fashion. Schmollers Jahrbuch 125 (2005) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.125.1.7 | Generated on 2023-04-04 12:34:44 Tue Skill Loss of Older East Germans after Unification 13 The high returns to new work experience could be driven by selection ef fects as variation in the new experience variable relies on unemployment and nonemployment spells. If selection into work is positive and covariances be tween new work experience and other control variables are ignored, this leads to an upward bias in (>-i, >.2). Estimation of a fixed effect model that controls for time-invariant unobserved heterogeneity in levels shows that returns to new experience are still substantial but around 30 percent lower than the least squares estimates in Table 1. 11 Another potential explanation for the high returns to new experience is that the data only covers the first twelve years of the transition process. If wage profiles with respect to new experience are steep at the beginning of the post1990 working career, the returns mainly reflect the steep portion of the wage profile similar to new labor market entrants. While data constraints prohibit a fully nonparametric approach, a spline function was used with the knot placed at four years of experience. The result confirm that returns decline with accu mulated experience: while the return to the first four years of experience after unification is 0.43, it falls to 0.07 for the years 5 to 11. To quantify the wage lass that can be attributed to the depreciation of socia list work experience, the following thought experiment was used. Suppose that the wall bad fallen but returns to experience bad remained at their 1989 level. What would wages of high-experience workers be? In a first step, wage regressions for 1989 and 1990 were estimated separately for men and warnen and the wage distribution in 1990 predicted conditional on experience and experience squared. The counterfactual log hourly wage for 1990 in the ab sence of skill depreciation was then calculated by adding labor market experi ence in 1990 evaluated at 1989 returns. The calculation shows that high-ex perience workers suffered substantial wage lasses from skill depreciation. For warnen with 35 or more years of potential work experience, the wage lass amounts to 23 percent of 1990 log wages. For men in the same experience category, the wage lass is almost 30 percent of the actual log wage in 1990. 12 11 With respect to age-specific returns to new experience (column (3) and (6)), selec tion effects work however against the expected finding of higher retums to new experi ence for younger workers. This is true as long as the larger employment decline among older workers translates into a more severe upward bias in the experience coefficients. 12 An alternative interpretation of the relative decline of wages after unification is that labor market experience was overvalued in the socialist economy. The fact that retums to work experience in socialist East Germany in 1989 were actually smaller than in West Germany (Bird, Schwarze and Wagner, 1994) speaks however against this argument. Schmollers Jahrbuch 125 (2005) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.125.1.7 | Generated on 2023-04-04 12:34:44 14 Christina Gathmann 3.2 Lower Returns to Education and Occupational Skills Experience is only one labor market skill whose value adjusted after unifi cation. If the value of formal education increased shortly after unification and older employees have on average less education, this compresses relative wages across cohorts. Older East Germans rnight have also been employed more than proportionally in sectors and occupations whose relative wages declined after unification. In both cases, older employees lose because the skills they possess earn lower returns after the regime change. Altematively, their relative wages would also decline with the regime change if older em ployees earn lower returns for the same observable skills than younger ones. This could for example occur if age acts as a proxy for adaptability to the new economic regime and employers value more adaptable workers conditional on other observable skills. 13 To distinguish between the two explanations, the log wage differential be tween older and younger workers in 1990 is decomposed into differences in characteristics and differences in retums: The results of the decomposition are reported in Table 2. For men, differ ences in educational attainment matter little in explaining the wage differen tial.14 The distribution across occupations leads to higher wages for older workers, mostly because older workers are more likely to work in high-wage occupations like administrators or professionals. The contribution of characteristics is however small compared to the differ ences in returns to those characteristics. Most importantly, returns to educa tion are much lower for older workers in 1990. Also, retums to occupations are lower for the age group 45 -54, while for men 55 and older, retums to the sector of employment are a more important deterrninant of the wage differen tial. For women, a somewhat different picture emerges. First, younger women actually earn higher wages than older women. This is mostly driven by the 13 After unification, East Germans, who were employed by the former secret service ("Stasi"), were banned from entering certain occupations, for example in the public sector. If those occupations pay high wages and older East Germans worked more often for the "Stasi" than younger ones, this would provide an alternative explanation for low retums to observable skills unrelated to employer preferences. I thank an anonymous referee for pointing this out. 14 Older East Germans are on average less educated. There is however surprisingly little variation in the distribution across econmnic sectors. The distribution among oc cupations is more dispersed: older men are much more likely to be employed as admin istrators or professional. At the same time however, the share of unskilled labor among the oldest cohort is more than double that of the young (16 percent compared to only 9 percent for men under 25). Schmollers Jahrbuch 125 (2005) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.125.1.7 | Generated on 2023-04-04 12:34:44