Photo: Women study the family framing insurance booklet near Isiolo, Kenya
Photo: Women study the family framing insurance booklet near Isiolo, Kenya
In a prior post, we reported on the substantial impacts that the BOMA Project’s Rural Entrepreneur Access Project had on the assets, income, and well-being of women and their families in Kenya’s arid pastoralist regions. Unfortunately, those hard-earned gains are under constant threat from the increasingly frequent droughts that sweep through the regions, destroying as much as 50% of total household wealth in the span of a few months. Thus, sustainably empowering women would require building and protecting women’s assets.
Women are well aware of their exposure to drought risk and its impact on them and their families. During an informal conversation with the research team, one woman reported,
“When there is drought, our children fall out of trees, break their legs, and require medical care.” Believing that something had gotten lost in translation, we asked for further clarification, which the woman patiently gave. In her area, the last vegetation animals can consume when all others have been destroyed are the upper leaves of a local tree species. Tree-climbing, therefore, becomes part of a family survival strategy, albeit one with collateral damage to children.
Previously, we reported on developing a strategy whereby an available index-based livestock insurance (or takaful) contract (IBLT) could be made relevant and available to women. IBLT issues payouts based on the forage conditions in the open rangeland where the insured party usually grazes their animals1.
Unfortunately, this kind of transhumant, open rangeland grazing is socially defined in northern Kenya as an exclusively male activity. This strict gender-typing, along with the fact that insurance has been sold exclusively in male-oriented cattle units, raises the question as to whether livestock insurance resonates with women and their immediate concerns and responsibilities. In our earlier work, we used a tablet-based game to remind women of their indirect exposure to rangeland risk. This alternative framing of insurance doubled women’s demand for livestock insurance in our experimental game.
But would the same strategy work in the real world, with women’s hard-earned money at stake?
Working with our insurance partner, Takaful Insurance Africa, we developed a variant of livestock insurance called IBLT for Families. This insurance was sold in family units of coverage, where one unit was designed to cover the basic needs of a single family member during a drought spell. We also developed educational materials to explain the novel IBLT for Families concept to potential users.
To evaluate the effectiveness of this novel product, we randomly divided the communities in the BOMA REAP program’s ongoing impact evaluation into two groups. One group received the conventional IBLT messaging (“protect your livestock”), while the other received the novel “protect your family” framing. This novel framing was first offered to communities during the August – September 2021 IBLT sales window and then again in the January – February 2022 sales window2.
Notably, the secondary sales window was the first in which IBLT was offered without any of the subsidies that had been used as part of the broader impact evaluation research design. These subsidies were randomly distributed to study households as an incentive to boost engagement and experimentation with the insurance. The withdrawal of all subsidies prior to the January – February 2022 sales allows us to study the impact of the family framing and its interaction with the prior provision of learning subsidies.
The impact evaluation study covers 1,710 women in Samburu country in Northern Kenya. Six hundred fifty women had fully graduated from the REAP program by the early 2022 sales window. The remaining women included women eligible for REAP but who were never offered the REAP program (717) and slightly better-off women who were not eligible for REAP based on the program’s means test (343). All three groups of women were equally eligible for insurance discount coupons in the years preceding 2022.
Using data on insurance purchases by these 1,710 women and their families3, we can evaluate the impact of the family framing and the insurance coupons. Figure 1 shows the results of this analysis.
Figure 1. Insurance purchase rates and amounts paid by sales window and framing treatment
Both graphs have a similar structure. The last two sets of bars (2021-2 and 2022-1) display the impacts following the introduction of the family framing to half of the sample population. Different bars report purchases on those offered subsidies in the seasons before the primary season of 2022 (2022-1 in the figure). Note that no one had a subsidy for purchases in the 2022-1 season. However, as we can see, the difference between those who had subsidies before the 2022-1 season and those who never had subsidies is striking.
The left graph shows that almost no one purchased insurance in that last season unless they had previously received a subsidy. In the earlier seasons when subsidies were active, the only buyers were those with subsidy coupons. The econometric results indicate that past receipt of a subsidy increases the probability of buying insurance from zero to a statistically significant 19%. While other work is investigating the impact of subsidy receipt on understanding IBLT, this finding suggests that the learning effects of the subsidy were substantial. Ultimately, those with coupons experimented with the insurance, learned its value, and purchased it even when they had to pay the full market cost.
Both graphs have a similar structure. The last two sets of bars (2021-2 and 2022-1) display the impacts following the introduction of the family framing to half of the sample population. Different bars report purchases on those offered subsidies in the seasons before the primary season of 2022 (2022-1 in the figure). Note that no one had a subsidy for purchases in the 2022-1 season. However, as we can see, the difference between those who had subsidies before the 2022-1 season and those who never had subsidies is striking.
The left graph shows that almost no one purchased insurance in that last season unless they had previously received a subsidy. In the earlier seasons when subsidies were active, the only buyers were those with subsidy coupons. The econometric results indicate that past receipt of a subsidy increases the probability of buying insurance from zero to a statistically significant 19%. While other work is investigating the impact of subsidy receipt on understanding IBLT, this finding suggests that the learning effects of the subsidy were substantial. Ultimately, those with coupons experimented with the insurance, learned its value, and purchased it even when they had to pay the full market cost.
The graph on the right of Figure 1 draws this out further. This graph indicates how much of their own money families spent purchasing IBLT. As can be seen, during the subsidy periods before the 2022-1 season, families spent very little money on insurance, with the bulk of the premium paid by subsidy coupons. After subsidies were eliminated, those who had paid for their insurance in the past using coupons spent an additional 1000 and 1500 Kenyan Shillings each ($10-$15 using the market exchange rate, or about $20-$30 using purchasing power parity adjusted exchange rates). This difference with the group that did not receive subsidy coupons is again statistically significant.
Finally, and most interestingly, we see that the family framing has a very strong interaction effect with the subsidy coupons. By itself (without past receipt of a subsidy coupon), family framing had no impact on insurance demand. However, when family framing was combined with a past receipt of a family coupon, the probability that the family would buy insurance increased by ten percentage points, up to almost 30%. Comparing purchase probabilities between 2022-1 and 2021-1, we see that the family framing kept insurance demand steady after the withdrawal of the subsidy. In contrast, demand fell by a third under the conventional livestock framing, dropping from 30% to just under 20%. The right graph in Figure 1 shows that family framing induced 50% higher insurance expenditures than livestock framing. This positive impact of family framing is statistically significant.
Given the importance of protecting assets for women’s economic health and empowerment, this field study offers an important lesson. For exotic risk-management products like insurance, some level of at least temporary subsidy appears to be necessary for any experimentation and learning about insurance to take place (Cai finds similar results in China, as do Boucher et al. in Tanzania and Mozambique). However, learning is highly complemented by efforts, like the family insurance framing, that makes insurance relevant to rural women’s needs. Putting the two together offers a promising way forward to empower women sustainably.
Boucher, S., Carter, M.R., Flatnes, J.E., Lybbert, T., Malacarne, J., Marenya, P., & Paul, L. (2022). Bundling Genetic and Financial Technologies for More Resilient and Productive Small-scale Agriculture. NBER Working Paper no. w29234 (revised). https://www.nber.org/papers/w29234
Cai, J., de Janvry, A., & Sadoulet, E. (2020). Subsidy Policies and Insurance Demand. American Economic Review, 110(8), 2422–2453. https://www.aeaweb.org/articles?id=10.1257/aer.20190661
Chantarat, S., Mude, A., Barrett, C. B., & Carter, M.R., (2013). Designing Index Based Livestock Insurance for Managing Asset Risk in Northern Kenya. Journal of Risk and Insurance, 80(1), 206-237.
By Michael Carter & Julian Arteaga (University of California, Davis), Andrew Hobbs (University of San Francisco), Nathan Jenson (International Livestock Research Institute), and Sam Owilly (The BOMA Project)
