Download BibTex file | 

Our paper evaluates the labour supply effects of 'in work' benefit reforms in the UK, using the means-tested benefit reforms of July 1995 as an example. The reforms increased the generosity of all UK 'in work' benefits for those working more than 30 hours a week . This was a response to concerns that Family Credit, the main UK 'in work' benefit, was generating labour market inflexibility at part time hours due to high marginal withdrawal rates associated with the reduction of benefit as hours and gross income rose. Our paper evaluates this reform in terms of its distributional impact and, crucially for a reform intended to change behaviour, its effect on female participation and labour supply.

We begin by showing how Family Credit has grown since its introduction in 1988 to represent 0.5% of Government expenditure, and that nearly half the recipients are lone parents. More relevantly we show how the hours threshold of 16 hours work a week has created a small but perceptible peak in the hours distribution of Family Credit recipients. This helps explain the motivation for the reform; to provide a further incentive to work more than 30 hours.

We follow this description of the broad incentive objectives of the reform with a theoretical section setting out our estimations of a continuous and a discrete choice model of female labour supply. We chose two types of model in order to compare simulation results of the different sorts of models. The main reason for this was that the continuous model, while providing exact predictions for hours worked of those working has limited power in simulating the participation decision. In comparison, the discrete model explicitly addresses the participation decision but fails to give an exact prediction of hours worked.

Using these two models we simulate the effects of the Family Credit reform and compare the results of each simulation technique. This exercise showed the importance of a simulation model to understand the full effect of the changed incentives offered by a reform to the tax and benefit system. But because the reform modelled had little impact on the participation decision, the limitations of the continuous model in simulating the participation decision were not particularly evident for this reform. The advantages of exact results from using the continuous this model were also evident because it enabled the labour supply results to be entered into our tax and benefit model to calculate the second round costing of the reform.

Ultimately, however, we found that labour supply effect of the reform was small relative to the initial distributional impact. The reform targets poor single earner couples and working lone parents for extra financial help, which would not disappoint the Government. It might, however, be disappointed that the reform had relatively small labour supply effects, with one of the major effects being to discourage labour force participation of some married women.