Workers or Employers: Who is Shaping Wage Inequality?
| Author | Ana Rute Cardoso |
| DOI | http://doi.org/10.1111/1468-0084.00081 |
| Date | 01 November 1997 |
| Published date | 01 November 1997 |
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 59, 4 (1997)
0305-9049
WORKERS OR EMPLOYERS: WHO IS
SHAPING WAGE INEQUALITY?
Ana Rute Cardoso*
I. INTRODUCTION
Evidence on growing wage inequality within industrialized economies
began to be reported in the empirical literature in the 1980’s, contra-
dicting previous evidence and challenging economic theory. There is now
wide consensus over the idea that this trend was widespread, having hit
economies with contrasting wage setting institutions. As in most other
OECD countries, rising inequality characterized the evolution of labour
returns in Portugal (see OECD, 1993). Moreover, inequality in the Portu-
guese labour market increased very sharply, from the already high values
of the early 1980’s. Indeed, the Gini index of hourly wages increased by
16 percent, from 0.32 to 0.38, between 1983 and 1992, while the coeffi-
cient of variation reported a more pronounced change of 26 percent, as
the wage distribution reinforced one of its characteristics of the beginning
of the decade — a very stretched upper half (Cardoso, 1996).
This study looks at both worker and employer attributes as sources of
wage dispersion and of its rising trend in Portugal during the 1980’s and
early 1990’s, a period initially marked by an economic crisis, and after
1985 characterized by high economic growth, low unemployment and
rising activity rate. Three main reasons contribute to the relevance of this
topic, and each will be dealt with separately.
Studies of earnings inequality have concentrated mainly on worker
characteristics, encouraged by an economic theory dominated by the
human capital approach and by the development of household surveys
providing detailed data on household attributes and rewards. The growing
awareness of the fact that labour economists have disregarded the
demand side of the market (see namely Hamermesh (1993) and Freeman
*I am grateful to John Micklewright for very incisive comments. Thanks are also due to
Paulo Madruga, Elena Bardasi, Sandrine Labory, the participants in the 1995 annual confer-
ences of the European Economic Association, European Association of Labour Economists
and European Society for Population Economics, as well as the participants in seminars held
at the London School of Economics (STICERD) and at the University of Essex (ESRC
Research Centre on Microsocial Change). This paper was revised while visiting STICERD,
whose hospitality I gratefully acknowledge. Financial support was provided by the Minist´erio
dos Neg´ocios Estrangeiros, Portugal. 523
© Blackwell Publishers Ltd, 1997. Published by Blackwell Publishers, 108 Cowley Road, Oxford
OX4 1JF, UK & 350 Main Street, Malden, MA 02148, USA.
(1989) led to the study of the role of the firm in wage inequality, namely
by Groshen (1986) and Davis and Haltiwanger (1991), who have detected
that inequality among firms accounts for a major share of the wage
dispersion. Bringing together worker and employer attributes in a study
of the determinants of earnings inequality thus seems a fruitful line of
research.
However, most of the work quantifying the impact of the firm on
earnings inequality has dealt with the USA, a labour market character-
ized by institutional arrangements quite different from those prevailing in
Europe, where firms are thought to be granted less autonomy when
bargaining over wages. In fact, the decentralized bargaining mechanisms,
the low safety net and the traditionally lower unionization rates character-
izing the deregulated and flexible American labour market contrast with
the more centralized bargaining system and the relatively higher
minimum wage levels enforced in Europe and with its traditionally higher
unionization rates. ‘I found it inconceivable that European style national
collective bargaining or extension of labor contracts from some employers
to their competitors would work in the US, outside of a mass mobilization
war environment’ (Freeman, 1995). In particular, this practice of exten-
sion of contracts is thought to have a major direct impact on the role of
the firm in wage inequality — ‘[introduction of] [e]extension of contracts
[in the USA] would reduce wage inequality among firms’ (Freeman,
1995). To the extent tht Portugal shares with its European counterparts
their framework of industrial labour relations, for example the contract
extension mechanism, analysis of the country can provide evidence on the
relevance of the firm in shaping wage inequality under an institutional
setting that diverges from that of the USA.
A particularly appropriate data set for the study of this issue is avail-
able for Portugal, which matches data on the firm, the establishment, and
each of the workers. This extensive data set is gathered annually by the
Portuguese Ministry of Employment and Social Security (MESS), based
on a questionnaire that every establishment with wage earners has, since
1982, been legally obliged to fill in. Appendix A provides more detailed
information on the data set and the sampling procedure used.
Section II discusses the degree of centralization in the Portuguese
collective bargaining system and the constraints imposed on the firm wage
setting behaviour by the minimum wage legislation. Section III presents
empirical evidence on the impact of the minimum wage on the dispersion
of earnings. Sections IV and V explain the methodology followed and the
results obtained regarding the quantification of the impact of employer
and worker attributes on earnings dispersion. In particular, they justify
the choice of the Theil index and its decomposition as tools of analysis.
Presentation of the results distinguishes between the determinants of
inequality and of its trend. Concluding remarks are presented in Section
VI.
© Blackwell Publishers 1997
524 BULLETIN
To continue reading
Request your trial