Unpacking the birth order effects

Abstract

A considerable number of studies have found negative effects of birth order on a range of individual outcomes, from earnings and employment in adulthood to cognitive and non-cognitive skills throughout childhood and adolescence. Nevertheless, studies in developing countries exhibit positive effects, suggesting that birth order estimates may be highly context-specific. Moreover, recent evidence on the strong causal effect of sibling spillovers implies that well-established birth order estimates are likely biased. This paper is the first to estimate birth order parameters on educational attainment after accounting for sibling spillovers. We exploit rich data sources in two different settings–Germany and Egypt–that allow us to link individuals to their parents and siblings. Using a variant of the standard linear-in-means model of peer effects, we find strong negative effects of increasing birth order on years of schooling. However, we show that the marginal effects of birth order are substantially attenuated once sibling interactions are accounted for. We find a negative spillover effect which points towards a preference for non-conformity (i.e. sibling rivalry) among siblings.