What happens to incomes after job-loss?

Two issues motivate research on income support programs available to those who lose their jobs: among economists it has been the possibility of disincentives to take on employment provided by benefits thought to be overly generous; among sociologists and political scientists it has been the effects on the incidence of poverty. Much of this research uses data on a small number of features of unemployment insurance programs to rank countries by benefit generosity and to infer what happens to incomes after job loss. The indicators often used are: replacement rates, benefit durations, and benefit qualifying periods. The data used mainly come from OECD sources. There are problems with this approach. i) The programs themselves are typically complex. ii) There are often benefit penalties associated with quits, dismissals, and refusals to take an available job (or enrol in a training program). The severity of these penalties and their enforcement varies considerably across countries. iii) The income loss caused by job-loss usually causes qualification for a number of transfer programs other than unemployment compensation. iv) What happens to income after job loss varies with the duration of unemployment. Unemployment durations vary greatly across countries.  A better way (than using summary OECD indicators) to analyse the effects of job-loss on income is to use the panel data sets that are increasingly available. I illustrate the different conclusions this leads to using data from the British Household Panel Survey.