This thesis seeks to answer two questions. First, is government policy in contemporary democracies congruent with public opinion? Second, what are the factors that determine opinion-policy congruence? I conceptualize opinion-policy incongruence as the distance between actual government policy and the policy preferred by the median voter. I apply this measure of opinion-policy incongruence to international survey data that assessed citizen` preferences regarding government spending. For each of the 33 countries involved, I obtain a measurement of the distance between the median voter and actual government policy for eight different policy areas. The results suggest that opinion-policy congruence is more often absent than present in contemporary democracies. Nevertheless, there is significant variation both between countries and between policy areas, Among policy areas, opinion-policy congruence tends to be highest for defense and unemployment and lowest for health, education, and retirement. The variation in opinion-policy congruence among countries is explored using fuzzy-set Quantitative Comparative Analysis, a data analysis technique based on a set theory. I identify two causal paths that lead to opinion-policy congruence; to achieve opinion-policy congruence, a country must either be rich and have a relatively equal distribution of income or it must be rich, decentralized, and use a non-proportional electoral system. I end by discussing the implications of these results. I pay special care to the question of weather opinion-policy congruence, in the countries where it is present, is robust to increase in income inequality or to different institutional changes.