Risk-sharing to mitigate local environmental risks

We carry out two interventions.
1. Mutual Commitment Contract: This treatment is designed to reduce the commitment frictions frequently encountered in risk-sharing. The mutual commitment contract, designed as a 'water sharing coupon', is pre-printed with a statement of commitment to share water in the future. Each tubewell owner is provided a set of coupons to exchange with other well owners. Households are then asked to get into a mutual commitment contract by exchanging coupons with other households. These contracts are formed before wells are tested for arsenic.

By exchanging coupons, symmetric risk-sharing relationships are established between two households.
Ideally, households should be allowed to exchange as many coupons as they like. We restrict the number of coupons to 10 for each well for operational tractability in most cases, based on our pilot work. Well-owners receive ten coupons per well for them to share with other well-owners.

We gave households 3 days to exchange coupons with other well-owner households in their villages. On the first day of the visit, the field agents invited qualified households to exchange coupons with other qualified household. Three days after the coupon distribution, field agents return to the village and record the tubewell and household IDs from the coupons each well-owning household held. Households keep the coupons.

2. Peer Monitoring: The second treatment aims to increase the expectation of peer monitoring of commitment contracts in the village. We sent text messages about contract exchanges to the participated villagers.
We characterize the Nth-order-neighbor as the neighbor that is N steps away from a given household in the contract-exchange network. For example, the first-order neighbors are the well-owners that directly exchanged contracts with the household, and the second-order neighbors are the well-owners that directly exchanged contracts with the household’s first-order neighbors, but did not directly exchange with the household, and so on. We randomly drew one first-order neighbor and one second-order neighbor to send the text messages if the household exchanged contracts with more than one first-order neighbor and had more than one second-order neighbor. If the household only has one first-order neighbor and zero second-order neighbors, the message will be sent to the only first-order neighbor.

The text messages are sent to the participated households in treated villages after the tubewell testing. A local phone carrier sent the text messages to the selected neighbors, telling them whom the household exchanged contracts with. The text message will include two risk-sharers’ names and the total number of exchanges the household made. The treatment also informs each household whose information is shared.

We obtained consent from households regarding this peer monitoring in advance (before the commitment contracts are formed). However, to reduce any bias arising from this consent, the actual peer monitoring is implemented in only half the villages where such consent was obtained.

2021-03-01

2021-12-31

1. The number of water-sharing agreements made by households (i.e., number of coupons exchanged), 2. Whether the household discussed water sharing with other households, 3. Access to low-arsenic water at the household level (continuous and binary), 4. Whether households with unsafe wells switch to a low-arsenic well ("switching"), 5. Changes in norms and preferences about sharing low-arsenic water from private wells.

The number of water-sharing agreements: the number of water-sharing coupons exchanged by each household.
Whether the household discussed water sharing with other households: We asked households in the survey whether they had discussed with other households regarding water sharing before the testing.
Access to low-arsenic water: a binary variable denoting if a household's primary well used for drinking usage has low-arsenic.
Switch to a low-arsenic well: a binary outcome denoting whether an unsafe well-owner eventually switches to safe water sources.
Norms and preferences regarding sharing low-arsenic water from wells with others are constructed using a set of survey questions.

Whether two households make water-sharing agreement (dyad level)

Using the network of coupons exchanged, we will create dyad-level binary variable indicating whether two households have exchanged coupons.

Our research features a clustered-randomization design. We randomly assigned 135 villages to 4 treatment arms.

Control (C): 36 villages are randomly selected into the control group. In the control group, we did not implement any interventions. We will record the simultaneous well switching and sharing to measure the magnitude of business-as-usual risk-sharing in these villages.

Commitment Contract (T1): 33 villages are randomly selected into the mutual commitment contract group. In this group, we distribute coupon-style commitment contracts to households and ask them to exchange them with other households to form commitment.

Commitment Contract + Peer Monitoring Notification only (T2): 33 villages are randomly selected into the commitment contract group. In this group, we distribute coupon-style commitment contracts to households and ask them to exchange. Additionally, we notify households about the possibility that there will be a peer monitoring program implemented after the well testing and obtain their consent.. We do not implement the peer monitoring program eventually.

Commitment Contract + Peer Monitoring Notification only + Peer Monitoring (T3): 33 villages are randomly selected for the commitment contract group. In this group, we distribute coupon-style commitment contracts to households and ask them to exchange. Additionally, we notify households about the possibility that there will be a peer monitoring program implemented after the well testing and obtain their consent. We implement the peer monitoring program eventually.

We collect endline data in two rounds of surveys -- the first endline survey immediately after well testing, and the second endline survey about 1.5 years later.

Not available

Randomization done in office by a computer

Village based cluster randomization.

Yes

135 villages. In case of large villages, a predefined part of the village (called para) is considered as one village unit.

16,054 households from the villages surveyed at the baseline. Among these households, 11,975 households are targeted as the experiment subjects as they fully or partly owned at least one well.

36 control villages; 33 villages in T1, 33 villages in T2 and 33 villages in T3 group.

We primarily evaluate the impact of our intervention on well switching (i.e., a high-arsenic well household switching to a nearby low-arsenic well). According to the well listing conducted at baseline census in January 2020, on average each village has 89 private well owners. We set the cluster size to be 80, considering potential attrition. There are 36 control villages with no explicit risk-sharing agreement and 99 villages with an explicit risk-sharing agreement (T1+T2+T3). The baseline switching rate is set to be 0.28 based on Barnwal et al. (2017). Alpha and Beta are set conventionally to be 0.05 and 0.8. Then: MDE = 0.182 when ICC = 0.1; MDE = 0.250 when ICC = 0.2; MDE = 0.303 when ICC = 0.3; MDE = 0.348 when ICC = 0.4; MDE = 0.388 when ICC = 0.5. For comparison, Tarozzi et al. (2021) use the ICC for switching decision of 0.09 in their power calculation. The 0.09 ICC was calculated from the data collected by Bennear et al. (2013). Note that the switching rate depends not only on the number of coupons villagers shared but also on the arsenic level of their own wells. To understand the magnitude of these detectable effects, we use a simulation exercise to calculate the combinations of the average number of agreements and the likelihood of having high arsenic well that achieves these effects. For example, if each villager makes 5 commitment contracts with others in a village of only 30% of wells are safe, then each villager still has over 80% chance to get at least one safe well among the households they exchanged coupons with and thus achieving a switching rate of 80%. The simulation provides that to achieve an MDE of 0.388 (ICC 0.5), villagers need to make 5.4 agreements with others in a village with 20% of safe well, and about 1.7 agreements in a village with 50% of the safe well. Assuming that well owners make 5-6 agreements on average and the targeted villages are predicted to have over 30% of safe wells, the calculated MDE is achievable. These numbers are calculated by assuming that villagers are perfectly committed to the agreement. The required number of coupon-exchanges will increase marginally to achieve this MDE when the commitment to the agreement is near perfect. If we consider ICC of 0.09, as in Tarozzi et al. (2021), our experiment is sufficiently powered to detect a much smaller treatment effect. References: Barnwal, P., van Geen, A., von der Goltz, J., & Singh, C. K. (2017). Demand for environmental quality information and household response: Evidence from well-water arsenic testing. Journal of Environmental Economics and Management, 86, 160-192. Bennear, L., Tarozzi, A., Pfaff, A., Balasubramanya, S., Ahmed, K. M., & Van Geen, A. (2013). Impact of a randomized controlled trial in arsenic risk communication on household water-source choices in Bangladesh. Journal of environmental economics and management, 65(2), 225-240. Tarozzi, A., Maertens, R., Ahmed, K. M., & Van Geen, A. (2021). Demand for Information on Environmental Health Risk, Mode of Delivery, and Behavioral Change: Evidence from Sonargaon, Bangladesh. The World Bank Economic Review, 35(3), 764-792.