Selected Working Papers
with Cassidy, R., Groot Bruinderink, M., Janssens, W.J.
Online Appendix here
Abstract: Women may face systematically greater benefits than men from adopting certain technologies. Yet women commonly hold lower bargaining power, which may constrain adoption when decisions are joint. Introducing a version of the technology that is second-best in terms of cost or effectiveness, but more acceptable to men, may increase adoption and welfare. We conduct a field experiment introducing female condoms – which are less effective and more expensive than male condoms, but often preferred by men – in a setting with high HIV prevalence. We find strongest adoption among women with low bargaining power, who were previously having unprotected sex.
with Anderberg, D.
Abstract: Transfers motivated by altruism, guilt, and norms of giving play an important role in supporting individuals who suffer losses due to risk. We present empirical evidence from an artefactual field experiment in Ethiopia in which we introduce formal insurance in a setting where donors make redistributive transfers to loss-suffering recipients. We find that donors tend to reduce their transfers to recipients who do not take up insurance when offered, and that transfer reductions are larger for donors who firmly expect that the recipient would take up insurance. The results are consistent with a framework in which the introduction of insurance erodes norms of giving by revealing underlying population heterogeneity.
with Berg, E., Blake, M.
Abstract: Households, organisations and governments commonly engage in risk sharing. However, the residual risk is often considerable, especially in the case of aggregate shocks in low-income countries. In response, many policy makers consider the introduction of parametric or index insurance. This raises the question of how demand for these insurance products depends on the extent of pre-existing risk sharing. We contribute to the literature in two ways. First, we develop a simple and parsimonious theoretical framework which shows that whilst indemnity insurance is a substitute to risk sharing, index insurance acts as a complement. Second, we provide the first direct experimental evidence on how the extent of risk sharing affects demand for both indemnity and index insurance. In an artefactual field experiment with Ethiopian farmers, the predictions of the theoretical framework are borne out.
with Harrison, G.W., Martínez-Correa, J., Ng, J.M., Swarthout, J.T.
Abstract: Experimental evidence shows that people exhibit preferences for compound risks that are different from preferences exhibited for their actuarially-equivalent counterparts. We study the potential link between index insurance demand, probability weighting, and attitudes towards compound risks. We test the hypothesis that the basis risk of index insurance negatively affects both the demand for the product and the welfare of individuals making take-up decisions. We measure the expected welfare of index insurance to individuals while accounting for their risk preferences, and structurally decompose the sources of the welfare effects of index insurance. Our results show that the compound risk in index insurance decreases the welfare of index insurance choices made by individuals. The behavioural inability to process compound risks decreases welfare when there is a compound risk of loss, whereas loss probability, basis risk and premium only impact the welfare of insurance choices when the risk of loss is expressed in its reduced, non-compound form. We also see that take-up is not a reliable indicator of welfare.
with Harrison, G.W., Schneider, M.
Abstract: There is widespread concern in developing countries with the expansion of formal insurance products to help manage significant risks. These concerns arise primarily from a lack of understanding of insurance products, low levels of financial literacy, and the need to use complex products in order to keep costs down for low-income households. We rigorously measure the understanding of insurance products, as well as general financial literacy, cognitive reflection and intelligence and evaluate the extent to which they determine the demand for insurance. A feature of our approach is the identification of welfare-improving and welfare-worsening insurance purchase decisions, and benchmarking the effect of our measures of understanding on insurance demand, with the effect of standard actuarial characteristics. We find that both general financial literacy, as well as domain-specific literacy about insurance help to increase the welfare of insurance purchase decisions.
Work in Progress
Do No Harm: The Welfare Consequences of Behavioural Interventions
with Harrison, G.W., Schneider, M.
We evaluate the consumer welfare implications of a wide range of behavioural interventions that are typically used in the promotion of insurance products. Based on laboratory experiments where subjects make risky choices, we estimate subject's individual risk preferences, and then randomly assign subjects to our behavioural interventions before they make insurance purchase decisions. We estimate the expected consumer surplus gained or foregone from observed take-up decisions and compare these across the intervention arms. We also elicit measures of subjects' cognitive functioning, and bias and confidence in their understanding of general financial questions and specific insurance purchase tasks and consider to which extent these explain heterogeneity in gains and losses in consumer surplus. We show that while our treatments typically increase take-up on average, they reduce consumer welfare.
Male and Female Norm Change, Intra-household Bargaining and Domestic Violence.
with Cassidy, R., Janssens, W.
We ask whose preferences within the household attitudinal interventions should target in order to reduce violence and promote education for girls. We conduct a cluster-randomized control trial across 200 rural villages in Pakistan, in which an attitudinal intervention is targeted at either men, women or both genders. We use incentivized artefactual field experiments to elicit beliefs, attitudes, beliefs about others’ attitudes (“social norms”) and intra-household bargaining.
Intimate Partner Violence, Preferences, and Economic Decision-making.
with Anderberg, D., Cassidy, R.
We ask to what extent violence in the household affects preferences of the spouses, and thereby their cooperation and economic decision making. We conduct an artefactual field experiment with spouses in urban slums of Addis Ababa, Ethiopia to examine the effects of priming women’s experiences of intimate partner violence on their cooperation and economic decision-making.
Evaluating the Welfare Effects of Behavioural Insurance Interventions in Ethiopia
with Barrett, C.B., Harrison, G.W, Jensen, N.D., Schneider, M., Swarthout, T, Upton, J.
The rapid expansion of financial products such as insurance, along with low levels of financial literacy, and the increasing complexity of products, raises serious concerns about the extent to which the demand for these products increases welfare for consumers. We apply two insights from behavioural welfare economics in a cluster-randomized control trial with 2000 Ethiopian pastoralists in 240 communities in which we try to improve and evaluate Index-Based Livestock Insurance. The first insight is a recognition that many widely used observable measures, such as product take-up, are not in fact measuring welfare when we allow for the preferences and beliefs that are actually driving choices to purchase or not. The second insight is that heterogeneity of welfare impact might provide an under-utilised and general basis for normatively designing better, conditional interventions. Rather than just providing one-size-fits-all interventions, our behavioural interventions will be tailored to provide specific interventions designed for the behavioural characteristics of those making insurance decisions. These are ``behaviourally smart'' conditional interventions, not unconditional ``nudges.''
Addressing Intimate Partner Violence: Evaluating Interventions with Male Perpetrators vs Female Victims
with Bhalotra, S., Carvalho, J.R., Biroli, P., Vecci, J.
Intimate Partner Violence (IPV) is a marker of conflict, a breakdown of cooperation between partners. Often both perpetrator and victim experienced IPV in childhood and repeatedly selected into violent relationships, and both sides are caught in patterns of behaviours that perpetuate cycles of violence. Therefore, interventions targeting both partners have more potential for success, but engaging perpetrators has proven challenging. We collaborate with the Brazilian judiciary to conduct a randomised control trial which introduces interventions with both partners in parallel. The male intervention targets men’s violent attitudes and behaviours. The female intervention focuses on empowering women. We investigate their independent and joint impact on relationship status, cooperation, and IPV.