EUROPEAN i?iK&Yc ELSEVIER

European Economic

Review 41 (1997) 647-657

The decline of labour market institutions and the rise in wage inequality in Britain Stephen Machin a*by * a Department of Economics, b Centre for Economic Pelformance,

University College London, London WClE 6BT, UK London School of Economics, Houghton Street, London WC2A 2AE, UK

Abstract At the same time as the role of labour market institutions declined very dramatically in Britain there was a very sharp rise in wage inequality that, since 1980, has increased at a faster pace than in most advanced countries. Britain in the 1980s and early 1990s therefore provides a very good testing ground for evaluating the importance of labour market institutions in explaining the evolution of the wage structure. I look at this in terms of two institutions that have traditionally propped up wage levels at the bottom end of the wage distribution, unions and minimum wages. I find that their weakening has played an important part in the rise in wage inequality in Britain. 0 1997 Elsevier Science B.V. JEL classification: J5 Keywords:

Unions; Minimum

wages; Wage inequality

1. Introduction

Since the late 19’70s the pace of decline of labour market institutions in Britain has been unmatched amongst advanced countries. In 1980 54 percent of the labour force were trade union members. By 1995 this had fallen to 32 percent. In 1980 61

* Fax: (+44)

171 955 7595; e-mail: [email protected].

0014-2921/97/$17.00 Copyright PII SOO14-2921(97)00027-5

0 1997 Elsevier Science B.V. All rights reserved.

648

S. Machin/European

Economic Review 41(1997)

647-657

percent of manual employees worked in workplaces where a trade union was recognised for collective bargaining purposes. By 1990 this had fallen to 48 percent for manuals (for non-manuals it fell from 50 to 43 percent). The importance of minimum wage legislation, via the industry-based Wages Council system, was severely downgraded following 1979 with the ratio of the Wages Council minimum pay rates to economy-wide average wages falling from 0.48 in 1979 to 0.40 by 1992 (Dolado et al., 1996). The minimum wage system was then abolished in August 1993. Related to these declines, various commentators have pointed to the labour market becoming more flexible (Beatson, 1995), increased job insecurity (Gregg and Wadsworth, 1996) and an increased demand for skilled employees commensurate with the adoption of new technologies that require specific skills (Machin et al., 1996). At the same time the gap between the highest paid and lowest paid workers in Britain is now much larger than it was in the 1970s (and, indeed, since records began in the latter part of the 19th century). Wage prospects on offer in new jobs are severely worse, suggesting that the rapidly becoming widespread view of deteriorating labour market prospects for recent labour market entrants has some foundation. This pattern of rising wage inequality and a rapidly declining role for labour market institutions provides a very good testing ground for evaluating the importance of labour market institutions in explaining the evolution of the wage structure. This is what I do in this paper, focusing in some detail on the implications for the British labour market of institutional decline. To examine this issue I draw on the ideas of some earlier work (Gosling and Machin, 19951, on some recent interesting extensions of this work (Bell and Pitt, 1996) and on the literatures that have addressed this question using US data (e.g. Card, 1996; Freeman, 1983) and drawn Canada-US comparisons (Lemieux, 1993). In Section 2 I describe in a little more detail the nature of the declining role of institutions in Britain. In Section 3 I look at the importance of the declining role of trade unions and minimum wage protection for explaining the rise in wage inequality in Britain. The impact of decline is seen to be important, though clearly by no means the whole story in explaining the observed changes in the structure of wages. Finally, Section 4 concludes.

2. The declining importance

of institutions

in the British labour market

In the 1980s and early 1990s the role of labour market institutions has been considerably diluted. The two I focus on here (as I can obtain concrete measures of their importance) are unions and minimum wage legislation and I now discuss some aspects of their weakening.

S. Machin / European Economic Review 41 Cl997) 647-657

649

2.1. The decline of unions

Following the 1979 election of the Conservative government that has since remained in power a number of pieces of anti-union legislation were passed in Britain. Machin and Stewart (1996) discuss these and a brief summary of the legislative measures introduced is as follows: (i) 1980 Employment Act: placed additional restrictions on picketing, made unions liable to be sued for damages, enlarged the grounds on which to claim unfair dismissal for refusal to join a closed shop, and repealed the statutory recognition procedure. (ii) 1980 Social Security Act: reduced benefits paid to strikers’ families and the codes of practice on picketing and closed shops introduced in 1980 imposed additional restrictions on the numbers and conduct of pickets and on the operation of closed shop agreements. (iii) 1982 Employment Act: reduced the range of disputes considered lawful, increased trade unions’ liability for damages, and introduced a requirement of 80 percent support in a ballot to legalise a closed shop (although this latter condition did not come into effect until 1984). (iv> 1984 Trade Union Act: required secret ballots before industrial action, executive elections every five years by secret ballot and political fund ballots every ten years. (v> I986 Public Order Act: introduced a criminal offence in relation to picketing. (vi> 1988 Employment Act: made action to preserve a post-entry closed shop unlawful, placed further restrictions on industrial action and election ballots, required unions to compensate members disciplined for not complying with majority decisions, opened union finances to inspection, and prevented unions from paying the fines of officials and members. (vii) 1990 Employment Act: made it unlawful to refuse to employ non-union members (i.e. effectively undermined the pre-entry closed shop), made all secondary action unlawful, and made selective dismissal of strikers taking unofficial action possible. (viii) 1993 Trade Union Reform and Employment Rights Act: allows individuals to seek injunctions against unlawful action, requires all strike ballots to be postal, and requires written consent for check off to be re-obtained every three years. These legislative measures, though difficult to quantify, represent a severe onslaught on trade unions by the Conservative government and they seem to have had an impact on trends in unionization. It is well known that the time period in which they were introduced saw a sharp decline in union activity in Britain, following the quite substantial rise of the 1970s. As described in the introduction

650

S. Machin / European Economic Review 41 (1997) 647-657

Fig. 1. The toughness of Wages Council minimum wages 1975-92.

unionization fell sharply, irrespective of the measure one considers (union recognition, membership, strike activity etc.). Recent examinations (Disney et al., 1995; Millward, 1994) have stressed that much of the decline appears to be due to a failure to organise new establishments set up in the 198Os, many of which now operate in an environment free of unions. Against this backdrop it is of considerable interest to assess changes in the influence of trade unions on wage structure.

2.2. The reduced importance

of minimum wages

The Wages Council system of minimum wages that used to operate in Britain set minimum wages for workers in a number of low-wage industries. The Councils were initially set up in 1909 and set minimum wage rates up to the abolition of the system in August 1993 (in the 1993 Trade Union Reform and Employment Rights Act mentioned above). At the time of abolition there were 26 Wages Councils and they covered about 2.5 million workers. The largest Councils were in retail, hotels and catering, clothing manufacture and hairdressing. Fig. 1 shows the mean ratio of minimum wages to average wages in the largest Wages Councils between 1975 and 1992. The figure (taken from Dickens et al., 1993) shows a fall in the ‘toughness’ of minimum wages from 1979 onwards. It is clear from this that the declining importance of Wages Council minimum pay rates occurred at the same time as the rise in wage inequality. I consider the

S. Machin/ European Economic Review 41(1997)

647-657

651

empirical importance of this, and the fall in unionization, in the next section of the paper.

3. Effects of the declining importance of institutions on wage inequality 3.1. The decline of unions and the rise in wage inequality Even though the effect can theoretically go either way, ’ trade unions are traditionally thought to have a compression effect on wage inequality as most evidence points to a union effect that raises wages by more at the bottom end of the earnings distribution. * I examine the link between unions and wage structure in Britain at two points in time, 1983 and 1991, using the only two British cross-sectional data sources that permit such an investigation to take place. These are the 1983 General Household Survey (GHS) and the first wave of the British Household Panel Survey (BHPS) which took place in 1991. Whilst I am interested in the cross-sectional impact of unions my main focus is on how this cross-sectional impact has changed through time and how this relates to the coincident rise in wage inequality. Table 1 reports (in the lower panel) the variance of log(male real hourly earnings) in 1983 and 1991, showing a substantial increase from 0.229 to 0.272. This pattern is robust to other definitions of inequality (e.g. the 90/10 log hourly wage differential rises from 1.13 to 1.23) and has been confirmed elsewhere (see, inter alia, Bell and Pitt, 1996; Gosling et al., 1995; and Machin, 1996). Nor is it due to using two different data sources: in the 1991 Family Expenditure Survey the variance of log(male hourly earnings) is 0.299 and the 90/10 differential is 1.30. At the same time the proportion of workers in workplaces with recognized unions fell sharply. For the sample of male workers I consider here, the proportion fell from 0.71 to 0.56. Whilst I prefer to use recognition as my unionization variable a similar fall in union membership rates occurred over this time period. For the same sample of men, the proportion who were union members fell from 0.57 to 0.45. A natural question is, therefore, how much of the rise in wage inequality can be attributed to the fall in unionization?

’ See the discussion in Hayek (1980) or Minford (1983) who stress the monopoly aspect of unionism that widens wage differentials or Freeman and Medoff (1984) or Lewis (1986) who argue that the pay standardisation role of unions dominates such that they have an equalising impact on wage dispersion. * See, inter alia, Freeman (1980), Freeman (1982) for the US or Gosling and Machin (1995) for Britain.

1.675 0.181 1.024 3129 1983

0.229 - 0.058

2. Union efict on wage inequality

Variance Union effect 0.272 -0.041

1991

0.148 0.011 0.051

Change

union

- 0.057

1991, but replace

1.479 0.318 1.278 1282

1983

0.157 0.036 0.083

1991 with 1983 unionization

1.636 0.354 1.361 1020

union Change

without recognised

1991

Work in workplace

a Based on General Household Survey in 1983 and the British Household Panel Survey in 1991. Hourly real wages for all men (in 1991 prices). The method of calculating the union effect on wage inequality is given in the Appendix.

1.823 0.192 1.075 1303

1991

1983

1. Descriptive statistics

Average Log(real hourly wage) VtiLog(real hourly wage)) 90-10 log real hourly wage differential Sample size

’ with recognised

by union recognition Work in workplace

Table 1 Changes in the structure of male wages 1983-91,

& u

5

: c. i 2 2 % Y

a.

@ !+

E

: F -%

B

s

w

S. Machin/ European Economic Review 41 (1997) 647-657

653

The figures in the upper panel of Table 1 also make it clear that wage dispersion grew much faster in the non-union sector between 1983 and 1991. As such the fact that the nonunion sector increased its relative size vis-a-vis the union sector may well have contributed to the overall rise in wage inequality. The lower panel of Table 1 therefore reports a set of estimates of the impact of unions on wage inequality in 1983 and 1991. This is based on a computation derived in Lemieux (1993) (more details are given in Appendix A) where unions have an impact on wages that can differ for different skill levels and where they can affect the distribution of wages within and between skill levels. Skill levels are defined for the full sample on the basis of imputed wages from wage equations estimated on samples of non-union workers, with the distribution of imputed wages being split into four quartiles to reflect skill groupings (see Card, 1996, and Lemieux, 1993, for similar approaches). In both years, 1983 and 1991, unions have a stronger mean increasing and variance reducing effect in the lower skill groups (see Appendix A) and are thus associated with reduced wage inequality. In 1983 the union effect on the variance of log(male hourly earnings) is -0.058, or 20 percent of the variance one would expect to prevail in the absence of unions (i.e. 0.058/[0.229 + 0.058]). In contrast, by 1991 the union effect had lessened to - 0.041, or 13 percent of the variance where unions were absent (= 0.0411/[0.272 + 0.0411). Hence, the union compression effect seems to have been moderated over time. A natural question to ask is: what would the union compression effect be if the 1983 union recognition levels were still in place in 1991? The final row of the Table shows the union effect would have been essentially the same as the 1983 effect at -0.057. This is interesting as it shows that, where they still exist, the union effect is much the same (this is certainly backed up by the estimated union effects at different points in the wage distribution for 1983 and 1991 given in Appendix A). This operates through a stronger wage compression effect at the bottom end of the distribution. Unions have less of an impact in 1991 because the fraction of workers who are unionized has fallen, not because their impact where they still exist has been quelled, at least up to 1991 (see also Stewart, 1995, who reports that the mean union wage mark-up stayed relatively constant through the 1980s). To express these results another way they suggest a rise in wage inequality of 0.016 less, or about 2/5 less, than the 1983-91 increase had the level of union recognition stayed at its 1983 level. 3.2. The decline of minimum wages and the rise in wage inequality Table 2 uses New Earnings Survey data to consider the importance of declining Wages Council minimum wages between 1979 and 1992 (the year prior to Wages Council abolition). In line with the reduced toughness of the Wages Council minima, the wage gap between Wages Council and non-Wages Council workers

654

S. Machin/

European Economic Reuiew 41 (1997) 647-657

Table 2 Changes in the structure of male wages 1979-93,

Average Log(real hourly wage) Var(Log(real hourly wage)) 90- 10 log real hourly wage differential Residual variance Residual 90- 10 Sample size

by Wages Council coverage

a

Covered by Wages Councils

Not covered by Wages Councils

1979

1993

Change

1979

1993

Change

1.232 0.149 0.871

1.498 0.189 0.948

0.264 0.040 0.077

1.607 0.174 1.030

1.929 0.263 1.281

0.322 0.089 0.251

0.136 0.826 14622

0.153 0.974 15115

0.017 0.048

0.178 0.962 126865

0.239 1.206 116766

0.061 0.244

a Based on New Earnings Survey microdata. I thank Richard Dickens for research assistance in producing these numbers. The residual variance and 90-10 differential are based on residuals from a regression of log(hourly wages) on age and age squared.

fell sharply, from -0.38 log points in 1979 to -0.43 log points in 1992. However, like the union/non-union comparison above wage dispersion rose by less in the sector that offers wage protection to low-wage workers. So, by 1992 Wages Council minima were still having a compression effect on the wage distribution but, because the average wage gap fell, their importance had declined. Unfortunately the lack of skill controls in the New Earnings Survey (the employer-reported NES has no information whatsoever on education levels of workers) prevents me from carrying out a more formal analysis. But it is clear that the declining importance of Wages Council minimum pay rates reduced the wage compression effect and thereby contributed to the rise in wage inequality. 3

4. Conclusions

Britain in the 1980s and early 1990s is a natural testing ground for evaluating the view that labour institutions have an important impact on wage structure. At the same time as the role of labour market institutions declined very dramatically in Britain there was a very sharp rise in wage inequality that, since 1980, has increased at a faster pace than in most advanced countries. I look at this question for two institutions that have traditionally propped up wage levels at the bottom end of the wage distribution, unions and minimum wages. I find that their weakening has played an important part in the rise in wage inequality in Britain.

3 See also Dickens et al. (1993) and Machin and Manning (1994) for regressions showing that, within the Wages Council industries, reduced toughness of the minimum wage system went hand-inhand with rising wage inequality.

S. Machin/ European Economic Review 41 (1997) 647-657

655

Acknowledgements

I thank Richard Dickens for excellent research assistance on this paper. This work was funded by the Leverhulme Trust and was produced as part of the Industrial Relations Programme of the Economic and Social Research Council Centre for Economic Performance.

Appendix A. Calculation 1983 and 1991

of the estimated

union effect on wage inequality,

To estimate the union effect on wage inequality I split the (predicted non-union) wage distribution into quartiles to form skill rankings. The union effect on wage inequality is then composed of three parts (this is based on the derivation in Lemieux, 1993) the first two of which are union effects within skill groups, the third of which reflects the effect of unions on between-skill group wage differentials: (i) The effect of unions on the (residual) variance of wages within skill groups. This is given by U, A”V(s), where U, is the mean unionization rate within skill group s and A”V(s) is the estimated impact of unions on the variance of wages of skill group s workers. Averaging this over skill groups gives the overall union effect. This can be positive or negative but would be expected to be negative if unions standardise wages for unionized workers within skill groups. (ii) The effect of unions on the variance of wages within a particular skill group s. This is given by the following term: U,(l - Us>[(Au(raw)ws)2 - (A”(ruw)w, A” wJ2] (where A”(raw)w, is the raw union wage gap and A” w, is the estimated union wage gap) which can be averaged over skill groups. This will always be positive. (iii) The effect of unions on the dispersion of wages between skill groups. This will be given as V( U, A” w,) + 2 Cov( nw,, U, A” w,) where nw, is the non-union wage in skill group s (which equals the average wage in the skill group less the product of the fraction unionized and the union wage mark-up) and Cov denotes a covariance. This can be positive or negative depending on whether unions raise wages by more for higher or lower skill groups. The total union effect on wage dispersion is given by the sum of these three terms which are given in rows 6, 7 and 8 of Table A.1 .Table A. 1

656

S. Machin / European Economic Reuiew 41 (1997) 647-657 Bottom quartile

1983 General household survey (data on 4411 men) 1. Mean union recognition 2. Raw union/non-union hourly wage gap 3. Estimated hourly wage gap 4. Estimated variance gap 5. Union effect within-quartiles on within-sector variance (row 1 X row 4) 6. Union effect within quartiles on between-sector variance 7. Union effect on between-quartile variance 8. Total effect (row 6 + row 7 + row 8 for all)

2nd quartile

3rd quartile

Top quartile

0.613 0.735 0.744 0.745 0.305 0.105 0.096 0.002 0.155 b 0.084 b 0.080 b - 0.005 0.044 -0.011 -0.001 - 0.002 0.027 -0.008 -0.001 -0.001 0.017

0.002

0.002

0.000

1991 British household panel survey (data on 2323 men) 1. Mean union recognition 0.464 0.609 0.623 0.547 2. Raw union/non-union hourly wage gap 0.281 0.170 0.132 0.015 3. Estimated hourly wage gap 0.172 b 0.150 b 0.091 c 0.017 4. Estimated variance gap - 0.020 - 0.014 - 0.032 - 0.010 5. Union effect within-quartiles on within-sector - 0.009 - 0.009 - 0.020 - 0.005 variance (row 1 X row 4) 6. Union effect within quartiles on between-sector 0.017 0.007 0.004 0.000 variance 7. Union effect on between-quartile variance 8. Total effect (row 6 + row 7 + row 8 for all)

All

0.709 0.196

- 0.009 0.005 - 0.054 - 0.058

0.561 0.187

-0.011 0.007 - 0.037 -0.041

a The hourly wage gap is obtained from the coefficient on a union recognition dummy variable included in a logfreal hourly earnings) equation that includes age, age’, a set of five education qualification dummies, nine region and nine one-digit industry dummies. The estimated variance gap is taken from the gap in the variance of the residuals from log(real hourly earnings) equations (with the same controls) estimated separately for the union and non-union sectors. b Denotes the union effect is significant at the one percent level or better. ’ Denotes the union effect is significant at the five percent level.

References Beatson, M., 1995, Labour market flexibility, Report (Employment Department, London) April. Bell, B. and M. Pitt, 1996, Trade union decline and the distribution of wages in the UK: Evidence from kemal density estimation, Mimeo. (Nuffeld College, Oxford) Card, D., 1996, The effect of unions on the structure of wages: A longitudinal analysis, Economettica 64.957-979. Dickens, R., S. Machin and A. Manning, 1993, The employment effects of minimum wages: Theory and evidence from the UK, Discussion paper 93-13 (University College, London). Disney, R., A. Gosling and S. Machin, 1995, British unions in decline: An examination of the 1980s fall in trade union recognition, Industrial and Labor Relations Review 48, 403-419. Dolado, J., F. Kramarz, S. Machin, A. Manning, D. Margolis and C. Teulings, 1996, Minimum wages in Europe, Economic Policy, forthcoming. Freeman, R., 1980, Unionism and the dispersion of wages, Industrial and Labor Relations Review 34, 3-24.

S. Machin/ European Economic Reuiew 41 (1997) 647-657

657

Freeman, R., 1982, Union wage practices and wage dispersion within establishments, Industrial and Labor Relations Review 36, 3-21. Freeman, R., 1983, How much has deunionization contributed to the rise in male earnings inequality?, In: S. Danziger and P. Gottschalk, eds., Uneven tides: Rising inequality in America (Russell Sage Foundation, New York). Freeman, R. and J. Medoff, 1984, What do unions do? (Basic Books, New York). Gosling, A. and S. Machin, 1995, Trade unions and the dispersion of earnings in British establishments, 1980-90, Oxford Bulletin of Economics and Statistics 57, 167-184. Gosling, A., S. Machin and C. Meghir, 1995, The changing distribution of male wages in the UK, Discussion paper 270 (Centre for Economic Performance, London). Gregg, P. and J. Wadsworth, 1996, Mind the gap, please? The changing nature of entry jobs in Britain, Discussion paper 303 (Centre for Economic Performance, London). Hayek, F., 1980, 1980s unemployment and the unions, Hobart paper 87 (Institute for Economic Affairs, London). Lcmieux, T., 1993, Unions and wage inequality in Canada and the United States, In: D. Card and R. Freeman, eds., Small differences that matter: Labor markets and maintenance in Canada and the United States (University of Chicago Press, Chicago, IL). Lewis, H.G., 1986, Union relative wage effects (University of Chicago Press, Chicago, IL). Machin, S., 1996, Wage inequality in the UK, Oxford Review of Economic Policy 12, 47-64. Machin, S. and A. Manning, 1994, Minimum wages, wage dispersion and employment: Evidence from the UK wages councils, Industrial and Labor Relations Review 47, 319-329. Machin, S. and M. Stewart, 1996, Trade unions and financial performance, Oxford Economic Papers 48, 213-241. Machin, S., A. Ryan and J. Van Reenen, 1996,Technology and changes in skill structure: Evidence from an international panel of industries, Discussion paper 297 (Centre for Economic Performance, London). Millward, N., 1994, The new industrial relations (PSI Publishing, London). Minford, P., 1983, Unemployment: Cause and cure (Martin Robertson, London). Stewart, M., 1995, Union wage differentials in an era of declining unionisation, Oxford Bulletin of Economics and Statistics 57, 143-166.