Household unemployment and the labour supply of married women

This research, supported by the Leverhulme Trust, investigates the relationship between the employment status of husbands and the labour market behaviour of their wives. In the UK the unemployment insurance system encourages the wives of unemployed men who are in receipt of unemployment benefit (UB now, Job Seekers Allowance) to work part-time since low levels of earnings by the wife do not affect the husband.s unemployment benefit. But if the wife.s earnings are large then the husband can loose part of his benefit payment, so it only makes sense for women to work full-time if their wages are quite high. In contrast, when the husband has been unemployed for a long period and has exhausted his entitlement to unemployment benefit he may be entitled to Income Support (IS). However IS treats ANY earnings of the wife as unearned income and reduces the husband.s IS payments by 100% of those earnings. This reduces the incentive for the wives of such men to engage in paid market work, especially part-time or low wage work. Thus, the duration of unemployment for husbands has quite distinct effects on the incentive structure faced by married women - with the wives of UB recipients given an incentive to work part-time or not at all, and the wives of IS recipients having little incentive to work at all (except in a full-time and relatively high wage job). With the introduction of the Job Seeker.s Allowance (JSA) this distinction is now faced after 6 months rather than 12 months unemployment duration. The purpose of the JSA is to reduce the duration of unemployment and yet it has a perverse effect on the spouses of recipients relative to the old UB system. We estimate a model of the labour supply of married women and our estimates suggest that the move from UB (or JSA) to IS increases the probability of the wife participating by 3.5% points, mainly at the expense of part-time work. When the husband finds work the incentive for the wife to work improves and we predict that the part-time participation would rise by 2.3% and the full-time participation rate would rise by 9.0% - and about half of these changes are due to the welfare system. The net effect of the JSA (net of the adverse effect on spouses) depends on the extent to which it promotes low durations of unemployment and the available empirical estimates suggest that this is likely to be quite modest. Given our estimates of the effects on spouses it seems unlikely that the net effect would be a beneficial one.