The Added Worker Effect and Intrahousehold Aspects of Unemployment

The added worker effect states that unemployment of a household member leads to an increase in labour supply of another household member. This paper investigates whether there is such an effect in a developing country. We use a rich data set for urban Ethiopia. We first give a brief description of who is unemployed within the household and find that they are mostly related to the household head. Men are not more likely to be unemployed than women once we control for being family in law. The eldest remaining sons in the household are more likely to be unemployed, but this may be due to a selection bias. The oldest remaining unemployed have no higher job aspirations than their younger brothers, suggesting that if older brothers have more entitlements, waiting in unemployment for a good job is not one of them. We carry out two separate analyses to investigate the added worker effect. First we analyse the effect using actual labour supply and find no evidence for an added worker effect once we take unobserved individual effects into account. We then investigate whether there is an added worker effect using desired labour market participation and find that there is none. The combined evidence indicates that there is no added worker effect. This suggests that households have other ways to cope with unemployment and is consistent with results from previous analysis which shows that the use of savings (by selling assets) and consumption smoothing are important mechanisms to cope with unemployment.