This paper empirically investigates the reasonability of assuming subjective wellbeing (SWB) data are cardinal. The inability or reluctance to assume cardinality implies limitations to use of data and methodology, which has been demonstrated to yield potentially biased results. This analysis uses the concept of transitivity to investigate the likely functional form of the SWB reporting function via a second alternative wellbeing measure. Here, data on mental health are used for this purpose. Results indicate that the SWB reporting function cannot deviate strongly from linearity, implying that the cardinality assumption is reasonable in most research contexts. An auxiliary analysis examines the bias that may result from possible nonlinearities in the SWB reporting function, which gives an indication of the potential cost of wrongfully imposing cardinality upon these data.
|Name||Economics Discussion Papers|