Essays in financial and behavioural economics

Abstract

This thesis comprises three essays. The first two essays (Chapters 2 and 3) describe randomised controlled trials which experimentally test interventions to arm consumers with informational tools to navigate the promise and peril of the financial landscape more optimally. The third essay (Chapter 4) uses pan-European survey data to shed new light on an old puzzle in consumer finance: widespread non-participation in stock markets.

Chapter 2 tackles the well-documented failure to refinance puzzle, which can cost mortgage holders tens of thousands of euro over the lifetime of a loan, leave them with unnecessarily elevated debt-service burdens, inhibit the effective transmission of monetary policy, and reduce the incentive for competitive market entry. The chapter reports results from a large-scale field experiment with 12,000 borrowers with a large retail bank in Ireland. It tests the impact of smart enhancements to mandatory consumer disclosure in prompting take-up of refinancing opportunities. The best performing alternative increases the amount of refinancing by 76\% compared against the pre-existing standard. With average savings of ?1,209 achieved by refinancers in the first year, this represents no small achievement. To interpret results, the chapter extends and estimates a mixture model of inattentive financial decision-making to allow for disclosure treatment effects. The analysis shows that a simple reminder decreases the likelihood mortgage holders are inattentive by 15 percentage points. The results suggest that reminders could have larger effects on household refinancing than a 200 basis point rate cut and that reminders could strengthen the refinancing channel and stimulate local consumption even when policy rates are at the zero-lower bound or set in a monetary union.

Chapter 3 addresses the pernicious problem of fraud faced by small business owners using digital financial services (DFS) in low and middle-income countries, which can lead to severe financial cost, psychological damage, and frustrate the development of the strong and trusted financial ties critical for economic development. In a lab-in-the-field experiment in Northern Nigeria with 780 participants in the network of a partnering digital market platform, this chapter tests the impact of a series of learning interventions, and a technical solution for the authentication of inbound communications, in improving the ability of small business owners to accurately discern fraudulent and genuine communications, as well as in building trust in DFS. We show the difficulty of improving discriminant ability between genuine and fraudulent communications, and the severity of the challenge faced by users in navigating the navigating the noisy landscape of competing communications. The results cast doubt on the utility of light-touch quick-fix remedial learning interventions to combat fraud susceptibility in DFS.

Chapter 4 examines a new dimension to another established puzzle in household finance: the role of private emergency financial safety nets in helping to explain widespread non-participation in stock markets. This topic has continued to exercise minds due to the considerable opportunity cost implied by non-participation, and its consequent implication for long-term patterns of wealth-inequality, as well as the implication of engendering divergent attitudes towards tax policy as well as risk-sharing and redistribution among those with and without financial assets. Consistent with a theory of insurance-induced consumption, the analysis shows that those households that enjoy the option of emergency financial support are more likely to participate in the stock market, and to report high financial risk appetite. In illustrating a mechanism of advantage compounding advantage, the results point towards the importance of effective and sustainable policies for financial inclusion, and highlight a potential use case for downside protection policies in replicating the insurance role provided by private safety nets in inducing stock-market participation.