Abstract:
Universal social benefits seem to contradict important notions in economics. They are poorly targeted and must be paid for by what seem to be high taxes. This paper describes the costs of universality and then proposes two competing explanations for why an electorate might wish to pay these costs. It may be harder to identify the poor through targeted social programs than to simply give everyone social benefits and withdraw part of these benefits through the tax system. Or, universality may be a form of political insurance that protects any one group of voters from being exploited by others. Each conjecture leads to different predictions about the manner in which government benefits will vary with the incomes of the recipients. I use a model of tax and spending incidence for Canada in 1990 to see which conjecture helps best to understand the data. I find mixed evidence in favor of the notion that universality is a form of political insurance.