POSITIONAL CONCERNS - How and why, and with what effect, are we positional when comparing with those around us?

Abstract

We took an experimental approach to study what may contribute to such behavior, and what effect these choices have on individual wellbeing. We used three independent surveys to analyze this, all three measuring positional preferences in a broad spectrum of domains. If left uncorrected, there might be large unintended external effects from positional preferences, which is why this research is necessary. On a greater scale, we can use this understanding as a tool to design a policy instrument correcting for the suboptimal behavior when people are concerned with relative performance. Our first study focused on the effect of gender, closeness, and relevance of the reference group (with whom we compare) on positional preferences and we explored how social identification in the relevant domains contributes to such behavior. The focus of our second study was to test if the instruments commonly used in this research are robust to a variation in levels (endowment) or subject (deciding for the self or deciding for a grandchild). Finally, our third study focused on the effects of positional preferences on life satisfaction. To test our hypotheses we ran a set of logistic regressions, controlling for a variety of social-demographic indicators. Our findings suggest that comparing with an average in society elicits a higher share of positional preferences than comparing with someone close and relevant, and we found that social identification with a domain correlates with positional preferences in the same domain. We also found that gender of the individual, and the reference group matter for positional preferences in some domains, but not all domains are gendered. When we tested the robustness of these instruments, we found that they are relatively insensitive to a variation in levels and in the targeted subject, which is reassuring as it validates earlier findings and makes them replicable. Finally, our third study suggests that positionality in all the included domains has negative and significant effects on how we feel about ourselves.