Better Know Your Dependent Variable: A Multination Analysis of Government Support Measures in Economic Popularity Models

Three aggregate measures of government support are commonly used in the study of the impact of economic conditions on that support: survey measures of vote intention and approval of the government/president, and electoral outcomes. The appeal of monthly time-series data over long periods has attracted many analysts to the first two of these measures, collectively known as popularity measures.Whether approval or vote intention is used, it is usually the same vote-based theories that are being tested. Some academics argue that measures of vote intention and approval capture different phenomena.1 And contextual factors, suggested by Bingham Powell and Guy Whitten, and Torsten Persson and Guido Tabellini, that have been theorized to mitigate the effect of economic conditions on vote intuitively will not have the equivalent effect on approval.2 Moreover, leading scholarship in the United States points towards the fact that the economic considerations that matter for vote intention are different from those for approval.3 In order to understand the consequences of using one measure versus another in a model, this Note asks: what is the difference (if any) between government approval and vote intention, why is there a difference, and when will it matter?