Comparability of web and telephone surveys for the measurement of subjective well-being

We test whether web survey mode alters people's evaluations of their well-being compared to telephone survey mode. We use unique, nationally representative data from Luxembourg in which five questions about subjective well-being were asked using web and telephone surveys. Oaxaca decomposition and multinomial logit with Coarsened Exact Matching indicate that the survey mode alters peoples' well-being scores. Web respondents are more likely to report low well-being and less likely to report the neutral category. However, the consequences for statistical inference are negligible. Our results support the view that web-surveys are convenient and reliable tools for collecting subjective data, such as people's well-being.