We use a survey of unemployed people to examine how a job loss impacts on household expenditures. The principal focus is on the effect of the level of income replacement provided by Unemployment Insurance. We restrict attention to a sub-sample of respondents who are still in their first spell of unemployment after six months. For this group we find large consumption falls, averaging about 16% of total expenditure. The actual fall depends on a variety of factors of which the most important is the pre-job loss ratio of the respondent's income to household income. The effects of varying the replacement ratio are relatively small. We only find effects for those who did not have assets at the job loss and even for them the elasticity of total expenditure with respect to benefit is small. We conclude that for most of our sample, small changes in the benefit level will have no effect on living standards within the household and hence on other facets of behaviour such as job search, unemployment duration and the quality of any new job taken