Assessing the stability of fiscal attitudes: Evidence from a survey experiment

The literature on attitudes toward government budgets has been dominated by two distinct approaches, jointly studying both sides of the ledger (holistic approaches) and studying attitudes over spending and revenue separately (singular approaches). Despite both approaches being widely adopted, scholars have given limited attention to testing empirically how methodological differences in the approaches may affect measures of fiscal attitudes and the inferences we draw from those measures. In this paper, we ask, "Do the different approaches to studying the budget alter mass attitudes toward spending and taxes, and if so, how?" Using data from an Amazon MTurk survey experiment, we find that spending choices differ significantly (attitude instability) across the two approaches. On the revenue side, our results show that choices over taxation tend to remain consistent and stable, regardless of whether the choices include only taxes or the combination of taxes and spending.