The heir and the spare: Impact of birth order on risk attitudes, discount rates, and behaviors

Abstract

Chapter one gives an introduction to the literature on whether birth order has a significant impact on both cognitive and non-cognitive outcomes. Conceptual models of why birth order might matter are presented from the fields of economics, sociology, and psychology.

The second chapter uses data from the Children of the National Longitudinal Survey of Youth to determine the impact of birth order on initiation into the activities of smoking, marijuana usage, sexual intercourse, and drinking. Significant birth order effects are found for the last born of two and three child families across all four activities. A subset of the data is used to see whether an older sibling’s participation in the activity has a significant effect on the hazard of the younger sibling’s initiation.

Chapter three uses experimental data to determine if there are any differences between subjects of different birth orders in terms of: laboratory elicited measures of risk aversion, participation rates in the four activities described above, and age at first initiation into the activities. Significant differences are found between oldest and non-oldest subjects in terms of risk preferences. Birth order had an insignificant impact on most of the field behaviors studied in this sample and no strong correlation is found between laboratory elicited measures of risk aversion and risky field behaviors.

Chapter four examines to what extent risk perceptions and risk preferences differ between subjects of different birth orders across contexts using the Domain Specific Risk Taking Scale (2006). No significant differences were found between non-oldest and oldest children in terms of risk perception or willingness to take risk across domains. No consistent relationship is found between the Holt and Laury measure of risk aversion and the willingness to take risks measured in the DOSPERT.

The fifth chapter estimates differences in temporal preferences between subjects of different birth orders. Older children are found to be more patient compared to non-oldest children when using a new approach to estimate discount rates under conditions of risk neutrality. Significant differences are found between discount rates elicited in traditional ways compared to those elicited with the new method.