would have reason to expect on the basis of the theory. Inflation and unemployment did not hurt the government, nor did increases in income favour it. 'This finding firmly supports the predictions of the global judgment model and discredits the personal grievance model. The evidence indicates that individuals do not vote their own money problems but instead respond to how the government is handling the economy in general.'60
The comparative research on economic voting can be summed up in the words of a leading authority who, in a work on the state of knowledge about a large number of countries, notes that there is general agreement on the fact that sociotropic assessments are Significant for voting behaviour yet the pocket-book hypothesis is still debated. What keeps the discussion alive is not the discovery of any evidence to support it but rather the fact that it seems so plausible: 'a good hypothesis facing resistant data'.61
How would sceptical philosophers regard these results of electoral research? Would they agree that the picture painted by the publicchoice school of voters who are primarily motivated by self-interest is a misleading, not to say an inaccurate one?
Probably not all would do so. They would quite likely concede that empirical research had achieved a higher degree of sophistication and had thereby succeeded in avoiding many of the pitfalls that lie in wait for the political scientist who tries to determine whether it is self- interest or public interest that plays the predominant role in voting. They would find it satisfactory that these conclusions are not based solely on observations of the individual's 'behaviour', 'expressions of will', and 'motivations', all of which can be deceptive sources of knowledge about self-interest, and that the respondents have also been required to express 'judgements' about economic changes both in their private lives and in society at large, as well as their 'preferences' by indicating which party they choose. Nevertheless, even intimate information can be unreliable. Not unlike the fox who declared the grapes to be sour, the voter perhaps does not even admit to himself that his economic circumstances have deteriorated; however, this objection might be rebutted by arguing that subjective assessments are more important than objective realities in the theory of economic