Working Paper

Adjusting for Scale-Use Heterogeneity in Self-Reported Well-Being, with Daniel Benjamin, Kristen Cooper, Ori Heffetz, and Miles Kimball. [Paper] [NBER] [SSRN] [Appendices A-J] [Appendix K]

Analyses of self-reported-well-being (SWB) survey data may be confounded if people use response scales differently. We use calibration questions, designed to have the same objective answer across respondents, to measure scale-use heterogeneity, both dimensional (i.e., specific to an SWB dimension) and general (common across questions). In a sample of ~3,350 MTurkers, we find substantial such heterogeneity that is correlated with demographics. We develop a theoretical framework and econometric approaches to quantify and adjust for this heterogeneity. We apply our new estimators in several standard SWB applications. Our framework sheds light on when and how adjusting for general-scale-use heterogeneity changes results.

Survey Evidence on Habit Formation: Existence, Specification, and Implication [Paper] [SSRN]

Habit formation is a staple of macroeconomics and finance, but insufficient micro evidence has led to controversies over its existence, specification, and implication. This paper documents new and extensive micro evidence for habit formation, through survey experiments eliciting ten preference parameters informative about habit formation. The evidence suggests that habit forms both internally and externally, depreciates by around two-thirds annually, and has an about equisized welfare impact as peer effect. I also propose and implement the first set of formal tests of additive and multiplicative habits and find that these ubiquitous preferences are rejected. Evidence-based simulations show that combining habit formation with peer effect could explain the Easterlin paradox.

Happiness, Reference Dependence, and Motivated Beliefs in U.S. Presidential Elections, with Miles Kimball, Collin Raymond, Junya Zhou, Fumio Ohtake, and Yoshiro Tsutsui. [Paper] [NBER]

Collecting and analyzing panel data over the last four U.S. presidential elections, we study the drivers of self-reported happiness. We relate our empirical findings to existing models of elation, reference dependence, and belief formation. In addition to corroborating previous findings in the literature (hedonic asymmetry/hedonic loss aversion, hedonic adaptation and motivated beliefs), we provide novel results that extend the literature in four dimensions. First, happiness responds to changes relative to both the political status quo (i.e., the incumbent presidential party) and the expected electoral outcome, providing support for two major hypotheses regarding reference point formation. Individuals exhibit hedonic loss aversion to deviations from expectations, but hedonic loss neutrality to changes from the status quo. Second, the speed of hedonic adaptation to deviations from the status quo is significantly slower than the speed of hedonic adaptation to surprises. Third, expectations affect happiness in a nonlinear way, consistent with Gul’s model of disappointment aversion, but contrary to other influential reference-dependent models. Fourth, both “objective” and motivated subjective beliefs matter for the happiness reactions, although subjective beliefs matter more.


A Social Interaction Model with Ordered Choices, with Xiaodong Liu, Economics Letters, 2017. [Paper] [Appendix]

We introduce a social interaction model with ordered choices. We provide a micro foundation for the econometric model based on an incomplete information network game and characterize the sufficient condition for the existence of a unique equilibrium of the game. We discuss the identification of the model and propose to estimate the model by the NFXP and NPL algorithms. We conduct Monte Carlo simulations to investigate the finite sample performance of these two estimation methods.

Work in Progress

Cost of Business Cycles When Unemployment Is a Worker Discipline Device

Measuring Altruism

Habit Formation Preferences Consistent with Survey Evidence: Axiomatics and Implications