Behavioural assumptions in labour economics: Analysing social security reforms and labour market transitions

Abstract

The aim of this dissertation is to test behavioural assumptions in labour economics models and thereby improve our understanding of labour market behaviour. The assumptions under scrutiny in this study are derived from an analysis of recent influential policy proposals: the introduction of savings schemes in the system of social security. A central question is how this reform will affect labour market incentives and behaviour. Part I (Chapter 2 and 3) evaluates savings schemes. Chapter 2 examines the behavioural effects of replacing the existing unemployment insurance system with unemployment accounts (UAs). Previous studies argued that such a reform will improve employment incentives considerably and thereby lead to a dramatic decrease in unemployment levels and durations. The chapter shows that this expected impact depends critically on the assumptions on time preferences. Using recent insights from behavioural economics, the study demonstrates that the theoretical impact of UAs is ambiguous rather than substantially positive. Next, Chapter 3 investigates the labour market effects of life course schemes. In contrast to UAs, life course schemes are savings schemes that can be used as general income-smoothing devices. The chapter surveys existing evidence on the impact of savings schemes on wealth accumulation and on the relation between wealth and labour market behaviour. Part II (Chapters 4-6) tests several behavioural assumptions empirically. All the empirical analyses are based on the Dutch DNB Household Survey (DHS), a representative longitudinal survey collected annually since 1993 by CentERdata. Chapter 4 analyses the effects of time preferences on job search behaviour of the unemployed. The aim of the study is to test the exponential versus the hyperbolic discounting model within a labour market context. The findings are in line with the hyperbolic discounting model. Next, Chapter 5 tests the exponential versus the hyperbolic discounting model within a framework of on-the-job behaviour. This study examines theoretically and empirically the effects of time preferences on two types of career investments: work effort and on-the-job search activities. The results presented in this chapter also provide support for the hyperbolic discounting model. In Chapter 6, the effects of wealth on labour supply are examined, focusing on adjustments atthe intensive margin. The empirical findings point out that wealth significantly affects labour supply behaviour. The dissertation points out that the introduction of unemployment accounts in the system of social security may backfire: although the alternative system improves long-term incentives, it weakens short-term incentives. Since the empirical results provide support for the hyperbolic discounting model, the opposite policy strategy is likely to be more effective. In addition, the findings suggest that life course schemes may have a limited impact on labour market behaviour. A general message of the study is that introducing more realistic behavioural assumptions matters crucially to economic analyses.