Essays on tax evasion and savings: Evidence from three randomized experiments in Chile

This dissertation presents three randomized field studies in the area of public policies for developing countries.;The first essay studies tax evasion by firms, with a particular focus on the value added tax (VAT). Randomly selected firms receive a letter with a deterrence message aimed at increasing the perceived audit probability or a motivational message aimed at affecting the perceived social norm of compliance. I find a large and highly significant effect of the deterrence letter on VAT payments, but no significant impact of the social norm letter. In line with predictions on the self-enforcing aspect of the VAT, the effect is much stronger on transactions between firms and final consumers than on transactions between two firms, and this effect of the VAT chain is especially important among small firms.;The second essay (with Felipe Kast) evaluates the role of peer groups to promote formal savings by informal or self-employed workers. Previous studies found mechanisms such as defaults and direct deposits from wages to be effective. However, these mechanisms can only be applied to formal employees. We find that a commitment device based on accountability within peer groups strongly increase deposits into formal savings accounts and significantly increase the savings balance, particularly for those who believe that they are better than their peers at the task at hand. In contrast, a 5% higher interest rate does not lead to a substantial increase in savings and participants do not reallocate their portfolios towards the higher interest rate account.;The third essay (with Felipe Kast and Stephan Meier) studies whether peers as a commitment device can be generalized beyond physical meetings to more widely applicable peer-based text message reminders, and analyzes the channels through which they affect behavior. We find that peer-related text-messages strongly increase savings deposits, but there is no difference in impact between a message informing participants of their own achievement and the success rate of others, compared to a treatment in which both the participant and a savings buddy receive a message informing them about the participant's achievement. The regular follow-up and taking stock may be more important than actual peer pressure in making peer groups an effective tool to encourage savings.