ABSTRACT
Bundling micro-insurance with loans should help not only decrease costs and therefore increase take-up but may also decrease financial inclusion if the insurance and/or loan are refused. We implement a randomised control trial in which a voluntary crop micro-insurance product is offered jointly with a loan application or separately (at a later date). The delayed offer of insurance did not influence overall insurance take-up or coverage amount but had heterogeneous impacts by crop, supporting the idea that bundling microfinance products is an opportunity more than a constraint. Yet, low product understanding also highlights the need for well-designed and effective consumer protection policies.
Acknowledgements
We thank Kelsey Jack, Corinne Novell, Sugato Chakravarty, Leah Bevis and participants at the CGAP Financial Inclusion Research Fund Event, the Midwest International Economic Development Conference 2016 and the 91st Annual Conference of the Western Economic Association International for helpful comments and discussions. We thank the Crezcamos team, including Mauricio Osorio Sanchez, the insurance team and all the branch staff involved in the project, as well as Barbara Magnoni and Emily Zimmerman at EA Consultants and Michael McCord at the MicroInsurance Center. This work was supported by CGAP’s client at the Center Financial Inclusion Research Fund, which was managed by the MicroInsurance Center. Any remaining errors are our own.
Disclosure statement
No potential conflict of interest was reported by the authors.
Supplemental data
Supplemental data for this article can be accessed here.
Notes
1. See Eling, Pradhan and Schmit (Citation2014) and Cole (Citation2015) for excellent reviews of determinants of micro-insurance demand, including the role of trust in the distributor.
2. Bundling financial products also takes place in the developed world, although it is relatively new and has largely escaped researchers’ scrutiny. See Ricci (Citation2012) for a review of the literature specifically on ‘bancassurance’, or the combination of banking and insurance industries and the bundling of their products.
3. Moral hazard, or the tendency of insurance buyers to not take as good care of their crop after having bought insurance as they would have in the absence of insurance, could also be present. Moral hazard does not influence the decision to purchase insurance; however, it influences insured farmers’ claim rates postpurchase, an outcome that is beyond the scope of this paper.
4. The rate for most permanent crops is 3 per cent (it is higher for plantain, banana and palm, which were not insured by any farmer during the study). The rate for annual crops varies between 4 per cent and 9 per cent. This rate determines the total premium, which is then subsidised.
5. The rate was higher for the next most common plant covered, citrus trees, at $1.35 per $100 of coverage; it was also higher for common annual crops such as cassava ($2.70), corn ($1.88) and potato ($4.02). The highest rate in our data was that for potato.
6. The video is available online: https://www.youtube.com/watch?v=yANQ6YKIfkM.
7. This was a moot point for individuals assigned to the separate offer, but a key one for those assigned to the bundled offer group. In the interest of simplicity and compliance, we only designed one sales protocol for both groups.
8. These application uses were dropped for the following reasons. First, 16 clients were mistakenly processed through the app twice. In all analyses, we consider the earlier assignment, and the second app use was removed. Second, we removed 16 app uses that did not match with a loan application in the lender’s database. These individuals approached loan officers about a loan, and were processed in the app, but decided not to continue their loan application. Lastly, we removed 24 app uses for individuals who did not cultivate an insurable crop and were not eligible for the insurance.
9. Robustness tests of our base model excluding loan applicants who were denied a loan show results similar to our main results (Table F in the online supplemental material).
10. One may be concerned that this indicates heterogeneity in the approach to selling insurance policies by loan officers. However, with a take-up rate of 23 per cent, it is not unusual that many loan officers would not have sold any policies.
11. We do not present regression coefficients for annual crop growers due the small sample size (38 applicants only grew annual crops).
12. We also checked whether the delayed offer of insurance had heterogeneous effects along loan applicant characteristics by interacting the delayed offer indicator with the gender and age of the loan applicant. None of the coefficients were statistically significant at the 5 per cent level or lower (Table A in online supplemental material).
13. Another piece of evidence that loan officers did not pressure clients to purchase insurance is that several loan officers did not sell any policy, and no loan officer had a take-up rate of 100%.
Additional information
Funding
Notes on contributors
Jonathan Bauchet
Jonathan Bauchet is an assistant professor in the Department of Consumer Science at Purdue University. His research interests include international development, particularly focusing on the financial life of poor households in developing countries. He is also interested in impact evaluation and its methodologies. His recent work includes examining the demand for microinsurance, the impact of an ultra-poor programme in Andhra Pradesh, India, informal financial institutions in Taiwan, and cash (and in-kind) transfer programmes in Bolivia.
Amy Damon
Amy Damon is an associate professor in the Department of Economics at Macalester College. Her research broadly focuses on household poverty reduction strategies and food security issues. Research projects include investigating the impact of international migration on families living in Central America, studying the economics of bushmeat in East Africa, evaluating the effectiveness of education programs in Latin America, and examining food consumption patterns in the United States. Professor Damon has worked at a number of international organizations, including the World Bank, the International Food Policy Research Institute, and the Inter-American Institute for Cooperation on Agriculture.
Vance Larsen
Vance Larsen is a PhD student at the Evans School of Public Policy and Governance at the University of Washington. His research interests include financial literacy and education, and household financial decision-making.