This paper analyzes the connection between economic context, voters' individual assessment of the economy, and party identification, taking the German national elections of 2009 as an example. Based on the results of multinomial and multilevel regression models, we show that partisanship moderates how the economy affects the vote. In case of increasing unemployment in the constituency, voters are
... [Show full abstract] less likely to vote for the social democrats instead of the CDU and the opposition. This effect is particularly strong amongst voters who report that they identify with the social democrats. Regarding the effect of voters' assessment of the national economy, the results are different. In this case, a better evaluation of the regional economy increases the probability to vote for right parties – notably amongst voters without partisanship. These findings suggest that we should not overestimate the moderating effect of partisanship on the economy's impact on voting behavior. What is more, this analysis implies that research on economic voting should always account for contextual factors and voters' personal 2 assessment of the economy together as these measures reveal different dimensions of economic voting.