Essays in welfare reform : an empirical evaluation of some UK policies

• Main Author: Fisher, Paul
• Published: University of Essex 2013
• Subjects: 361.68
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.605174

The UK welfare system has undergone important changes over the last decade, including the introduction of three major conditional cash transfers: The Working and Child Tax Credits in 2003 and the Educational Maintenance Allowance in 2004. The first reform. involved substantial structural change to the UK in-work benefit system. A particularly interesting aspect, from the point of view of the household decision making process, was that a substantial proportion of a couples benefit income was redirected towards mothers. The second reform, for the first time, introduced cash support for the poorest students staying on in post-compulsory education. This thesis consists of three distinct papers. Each paper begins by carefully documenting the key aspects of the policies. It then goes on to present original empirical analysis of specific behavioural responses to the reforms, with each paper stressing a number of important findings that have implications for the design of welfare policies. In addition, alongside the empirical evaluation of the third paper, a methodological contribution is made. Here, a potential outcomes framework discusses the conditions. under which, log-linearisation will lead to misleading causal inference, in the popular difference-in-differences setting. To give a summary of the papers, the first presents evidence on family spending from the 2003 tax credit reform for a sample of low income couples with children. An interesting consequence of the reform is that tax credit payments were split between partners in couples, causing a rare wallet to purse transfer. It is often argued that households will give greater emphasis to child-related expenditures and investments when mothers control a greater share of a given household resources, although recent bargaining theory has questioned this assumption. The important evidence presented here is, however. supportive of the gender targeting of benefits for child development purposes. The second paper, explores the 2003 reform but in a different direction. The introduction of the Working and Child Tax Credits changed incentives associated with a range of socioeconomic behaviours. Some of the responses potentially run counter to the child poverty objective motivating the reform. To give one example, the expansion of the Working Tax Credit to people without children, increased [he incentives for families to live apart. This paper estimates behavioural responses to the reform for: fertility and living arrangements, childcare usage and benefit receipt, in addition to labour market outcomes. One intriguing finding is that households with more and younger children made greater use of formal childcare services following the introduction of the new tax credits. Again, as with the results from the first paper, it is speculated that payment of the childcare element of tax credits directly to women, played an important role in this fact. Finally, the third paper begins with a methodological discussion of the common practice of taking logarithmic transformations of continuous outcomes, in the difference-in-differences setting. The model the researcher has in mind is usually one with multiplicative effects, that are linearised taking logs. It is shown that, as long as we are interested in the treatment effect on the average, the assumptions needed for causal interpretation of a difference-in-differences exponential specification, are not equivalent to the log-linearised case. Evidence from the Educational Maintenance Allowance introduction in 2004 -is then presented. Estimates from the log-linearised model show that the reform had no impact on households expenditure, while results from estimating the exponential model directly show that low income households allocated the extra resources to transport spending