To what extent do the political preferences of voters and non-voters diverge? This is a pressing question in democratic systems insofar as inequalities of turnout among different groups might translate themselves into asymmetries of representation. If voters and non-voters differ substantially in their political views, then elected bodies will be biased towards the views of the former at the expense of the latter. Various studies have measured the gap between the preferences of voters and non-voters, with conflicting findings depending on the topic and the national cases under consideration. Most of these studies examine the respondents degree of agreement towards abstract statements such as “governments should reduce income inequalities”. Using original survey data, this paper examines the preferences of a sample of the Portuguese population (n = 2405) with regard to economic redistribution by asking the respondents their views on current levels of taxation of two resources: income and housing property value. We employed an experimental survey design in which each respondent was confronted with the existing tax rate of one of three levels of income (low, medium or high) and one of three levels of house valuation for fiscal purposes (low, medium and high), and then asked about whether (s)he considered such rate too low/adequate/too high. Additionally, for the first dependent variable in order to determine the extent to which responses were likely to vary depending on the background context available to respondents, half of the respondents were provided information about the average income in Portugal, whereas the other half were not. The analysis shows only small differences between the preferences of respondents who tend to vote and those who tend not to vote. Indeed, non-voters tend to be marginally more likely to consider that real-existing taxation rates of property and, especially, income are too high, thus espousing on the aggregate a slightly less redistributionist attitude than voters. These results are relevant insofar as they show that the preferences of voters and non-voters with regard to a key issue might be relatively similar even in a context of moderate economic and turnout inequality.

--