Can Marginal Rates of Substitution be Inferred from Happiness Data?

Résumé du livre

To what extent do marginal rates of substitution estimated from subjective well-being (SWB) data reflect the tradeoffs that individuals would deliberately choose to make? Surveying 561 students from U.S. medical schools shortly after they submit their choice rankings over residencies to the National Resident Matching Program, we elicit both choice rankings and anticipated-SWB rankings over residencies (using three common SWB measures). We find substantial differences between the two rankings in the implied tradeoffs between different features of the residencies. For example, while residency prestige-and-status weighs more in choice, expecting life to seem worthwhile during the residency weighs more in all SWB measures. At the same time, tradeoffs estimated from anticipated SWB are relatively highly correlated with those estimated from choice, and we find no sign reversals between a feature's relationship with anticipated SWB and its relationship with choice. We also find that evaluative measures (life satisfaction and Cantril's ladder) imply tradeoffs closer to choice than does affective happiness. We further investigate a multi-period happiness index and a multi-measure SWB index and do not find that they generate tradeoff estimates closer to those generated by choice. Finally, despite the differences in implied tradeoffs, we find that SWB questions predict pairwise choice reasonably well in our data, and often substantially better than alternative questions. We discuss implications of our findings for the use of SWB data in applied work.