The Role of Knowledge for Policy Preferences: Evidence from Argentina

Abstract

In this dissertation, I study the role of knowledge in policy preferences. In democracies, voters are often criticized for not being sufficiently informed when making their political decisions. Political scientists disagree about how consequential this ignorance is for the quality of democratic government. One camp emphasizes that people are capable of learning new information and that knowledge makes them change their views. Following this line of reasoning, I suggest using fundamental knowledge relevant to a specific policy domain rather than measuring knowledge that is directly related to the political world.

I focus on one policy domain – economic policies. I use the Global Financial Literacy test developed by Standard & Poor’s as a proxy for economic knowledge. I ask respondents to complete this test at the end of the three original surveys that I conducted in Argentina in 2017, 2018, and 2019. In total, 10,457 individuals participated in the surveys. I choose this country to study the effect of knowledge on policy preferences because of a drastic change in economic policies in 2015.

I find that those who score higher on the test are more likely to support pro-market economic policies, such as elimination of trade barriers, elimination of subsidies, and integration into the world financial markets. My identification strategy includes using a measure of knowledge that is uncorrelated with the most important driver for policy preferences – partisan attachments- and conducting analysis in two different contexts – economic boom and recession. The latter helps address possible confoundedness between financial knowledge and other factors, such as social status, that are important both for knowledge and policy preferences.

In addition to observational evidence, I provide randomly selected respondents in the 2018 and 2019 surveys with survey treatments – passages in which consequences of economic policies in Argentina and Venezuela are discussed. The results largely support the view that information helps shape policy positions. I receive stronger results in 2018 than in 2019. In 2018, I find that respondents who received the treatment are more likely to support the open economy and the debt repayment. In 2019, the coefficients for the treatment dummies do not reach a conventional level of significance, although their signs are consistent with the hypothesis.

Overall, my findings suggest that fundamental non-partisan knowledge matters when it comes to policy preferences.