Please use this identifier to cite or link to this item:
Year of Publication:
45th Congress of the European Regional Science Association: "Land Use and Water Management in a Sustainable Network Society", 23-27 August 2005, Amsterdam, The Netherlands
Understanding the evolution of earnings inequality is a major research topic with obvious policy implications. While there is widespread belief that institutions are largely responsible for the limited rise in inequality in some European countries, it is also recognised that little or no growth in inequality could be the outcome of market forces alone. However, the role of these market forces in different institutional environments is still not entirely understood. Is the small growth in inequality at the country level reflecting small increases in inequality within and between groups or is it the result of large offsetting changes in different components? Using a large longitudinal matched employer-employee dataset we produce several measures of within and between groups inequality in Portugal for the 1986-1998 period. We focus our attention on changes in the returns to observable characteristics (gender, age and education) of workers and test the hypothesis that these changes reflect developments in the labour market. However, we depart from previous research by shifting focus from the supply side to the demand side of the labour market. Drawing on the results of the by-now large literature on plant turnover we investigate the link between plant entry and exit and changing returns to observable worker characteristics. We argue that Portugal is an interesting case study because, despite very tight regulation and a centralised wage setting system, the level and changes of earnings inequality in recent years make the Portuguese case akin to the US and UK cases rather than to other European cases that share with Portugal similar labour market institutions. Furthermore, high firing costs have previously been identified as the cause for a larger share of employment adjustment occurring in Portugal through plant openings and closings as an alternative to the expansion and contraction of continuing plants. Our analysis is done at the regional level - 28 regions in mainland Portugal (NUTS III) are considered. The advantages of working at the regional level are twofold. First, data for all regions come from the same source - the Personnel Records - which eliminates all issues of comparability that plague many studies dealing with international comparisons. Second, regional comparisons within the same country guarantees a common institutional background which is appropriate given our focus on the role market forces play in shaping patterns of changing earnings inequality. Personnel Records are an administrative survey administered by the Ministry of Employment which is mandatory to all plants with at least one wage earner. Data reported by respondents include characteristics of the plant (location, industry and size), the firm they belong to (location, industry, total employment, annual sales and legal status) and each worker in the plant. Reported worker characteristics include gender, age, education, skill, tenure, earnings and weekly hours of work). Because each plant is assigned a unique invariant identifier, plants can be followed across waves and entries and exits can be identified. On average, the data contains information on 200 thousand plants and 2 million workers per year.
Appears in Collections:
Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.