Frame-of-reference bias in subjective welfare regressions

Past research has found that subjective questions about an individuals'economic status do not correspond closely to measures of economic welfare based on household income or consumption. Survey respondents undoubtedly hold diverse ideas about what it means to be"poor"or"rich."Further, this heterogeneity may be correlated with other characteristics, including welfare, leading to frame-of-reference bias. To test for this bias, vignettes were added to a nationally representative survey of Tajikistan, in which survey respondents rank the economic status of the theoretical vignette households, as well as their own. The vignette rankings are used to reveal the respondent's own scale. The findings indicate that respondents hold diverse scales in assessing their welfare, but that there is little bias in either the economic gradient of subjective welfare or most other coefficients on covariates of interest. These results provide a firmer foundation for standard survey methods and regression specifications for subjective welfare data.