This paper tests the ‘valence hypothesis’ which claims a graduate’s pay tends to increase in steps, a few years after he or she graduates from university. Data from over a hundred EuroBarometer surveys are combined, to produce a data source with a very large sample. There are drawbacks to using EuroBarometer data to assess this hypothesis, and it appears that a more convincing source of data is needed for us to tell if the valence hypothesis is correct. But the evidence in this paper, while not perfect to assess this topic, does appear to support the valence hypothesis. This suggests there would be advantages to individuals, and to society, if more people are encouraged to attend universities.