What nudges will matter? An empirical study of female joblessness

The purpose of this empirical study is to use an interdisciplinary approach across labour economics, behavioural economics and social economics to explain female labour market statuses, in particular joblessness choices and conditions. We propose a new theoretical framework, based on Sen’s capability approach, new derived variables for the British Household Panel Survey and a new empirical methodology to estimate the probabilities of different labour market categories. Our results show that, in addition to the standard human capital variables, labour market statuses are markedly influenced by the interplay of environmental, social, psychological factors and personal views and values, such as disposition and conformity towards local social rules and to social relations’ (network) norms, degree of confidence and optimism, risk attitude, job and life satisfaction, preferences for the status quo and plans for the future. We find that the predicted probability of joblessness improves substantially when we move from the traditional into the multi- dimensional approach. Adding new variables to allow for more constraints and opportunities provides a richer and more refined view of factors affecting non employment and helps to recognize and explain better status of people within “similar” groups as well as to identify more accurately those people whose status in not in line with the “expected” (such as the employed against the odds and the voluntary” non- employed). We found that there are some clear idiosyncrasies across the different types of joblessness in particular when it comes to the category of carers, a unique feature of female joblessness. Our model predicts that being more embedded in a local community, having stronger values for family life and stronger ties with close “inactive” friends, facing potential income losses in changing labour market status (as measured by counterfactual labour income versus actual non labour income), are all factors that increase the odds of being a carer relative to being unemployed and that prompt carers to choose or preserve their status quo. Social environment matters in reinforcing personal attitude producing a sort of confirmation bias effect that suggests that social influence is more effective across similarly minded people. The results suggest that policies focused on areas where high levels of inactivity is present, working with large friendship groups within these areas could be used to motivate groups of individuals into work. Prospective financial losses (of receiving a labour income lower than non labour income if a carer became employed) increase the odds of being a carer while potential labour income gains would always decrease the odds of being carers versus any other category, particularly when the other categories are students and unemployed (i.e., those more potentially “attached” to the labour market). The results suggest a role for in work benefits such as the tax credits system, the national minimum wage and benefit reforms as policy options to help make work pay and provide incentives to work. Finally, designing active labour market policies that are more tailored to the characteristics of the individual job seeker could prove a fruitful avenue for policies such as the Work Programme.