Tax and Benefit Reforms in a Model of Labour Market Transitions

Abstract

Summary: We present a method for taking advantage of labour market transitions to identify effects of financial incentives on employment decisions. The framework we use is very flexible and by imposing few theoretical assumptions allows extending the modelled sample relative to structural models. We take advantage of this flexibility to include disabled people in the model and to analyse behaviour of disabled and non-disabled people jointly. A great deal of attention is paid to appropriate modelling of financial incentives on the labour market. This in the case of disabled people turns out to be an extremely complex process but one which in the end turns out well worth the effort. The model is used to compare reactions on the labour market to marginal changes in financial incentives and also to model one of the most important reforms of the UK Labour government, the introduction of the Working Families’ Tax Credit. The methodology relies on matching transitions and incomes data between cross-sectional and panel surveys, and could be used in other countries where detailed reliable income data are not collected in a panel format. Zusammenfassung: In diesem Beitrag stellen wir eine Methode vor, mit der die Wirkungen von finanziellen Anreizen des Steuer- und Transfersystems auf die Arbeitsmarktpartizipation anhand von Veranderungen im Beschaftigungsstatus geschatzt werden. Der Ansatz ist flexibel, wenige theoretische Annahmen erlauben es, die analysierte Population gegenuber Strukturmodellen auszuweiten. Dadurch konnen behinderte und nichtbehinderte Personen gemeinsam analysiert werden. Die finanziellen Anreize werden detailliert abgebildet. Mit dem Modell konnen Beschaftigungswirkungen von geringfugigen Anderungen der Grenzbelastungen als auch grosere Reformen der britischen Arbeitsmarktpolitik – wie etwa die Einfuhrung des Working Families’ Tax Credit durch die Labour-Regierung – analysiert werden. Die Methode basiert auf Matching-Verfahren, mit denen Querschnitts-Erhebungen und Panel-Daten zusammengefuhrt werden. Sie kann auch in anderen Landern eingesetzt werden, in denen keine detaillierten Einkommensdaten im Rahmen von Panel-Studien erhoben werden. * This paper is based on methodology and results developed under two research projects carried out for several UK Government Departments while both authors were employed at the Institute for Fiscal Studies. The methodology was initially developed in the project called “Fiscal policy and labour supply” conducted for the HM Treasury, the Inland Revenue (currently HMRC) and the Department for Work and Pensions, while its extensions and further results were developed under the project called “Including disabled people and their partners in a dynamic model of labour supply” carried out for the Department for Work and Pensions. We would like to acknowledge the financial support of all three Departments who contributed to this research. We would like to thank the civil servants involved in these projects for their advice and extremely useful comments at various stages. We also thank an anonymous referre for valuable comments which helped to improve the paper. ** Both authors have since moved from the IFS, Michal Myck is a senior economist at DIW Berlin, while Howard Reed is the Research Director at the Institute for Public Policy Research. Michal Myck would like to thank for financial support through the REVISER project, an RTN project financed by the European Commission (contract no. HPRN-CT-2002-00330) which made completing of this final paper possible. Data from the Family Resources Survey and the Labour Force Survey used in this paper were supplied by the UK Data Archive, who bear no responsibility for its analysis and interpretation. Micro-simulations for the UK were conducted using the IFS’s tax and benefit model TAXBEN – we are grateful for making it available to us. We are also grateful to our colleagues from the IFS for their suggestions and comments during the development of the projects. The usual disclaimer applies.