TY - JOUR
AU - Daniel, Stein,
AB - Abstract This paper analyzes the dynamic nature of rainfall insurance purchasing decisions. Customers of the Indian microfinance institution BASIX who receive an insurance payout are 9 to 22 percentage points more likely to purchase insurance the following year. This effect cannot be satisfactorily explained by trust, learning, or direct effects of weather, leading to the conclusion that it is driven by behavioral effects. Overall, low repurchasing rates even after payouts suggest that the studied rainfall index insurance products are likely to continue struggling to achieve significant sales at market prices. Roughly 60 percent of India’s population is employed in agriculture, and over 50 percent of agricultural land is dependent on rainfall to nurture the crops.1 But the Indian monsoon is notoriously unpredictable, prone to droughts and floods that can have devastating effects on the livelihood of rural Indians. Although Townsend (1994) argues that Indian villages do an effective job of providing informal consumption insurance against idiosyncratic shocks, a poor monsoon will hit whole villages and districts at once, likely rendering intravillage transfers ineffective. Beginning in the early 2000s, rainfall index insurance was introduced in India as a potentially important tool to help poor farmers deal with rainfall risk (Hess 2004), but it has struggled to reach a critical mass of customers, especially when unsubsidized (Binswanger-Mkhize 2012). Unlike physical goods (or even credit), it is difficult for customers to evaluate the benefits of insurance because its main benefits only occur when a payment is received. If customers are unfamiliar with how insurance works, they may be influenced by their recent experiences with insurance and also by the experiences of their friends and neighbors. Evidence in the developed world shows that purchases of flood and earthquake insurance in the United States are greatly influenced by recent experiences with disasters and insurance payouts, that peoples’ insurance decisions are influenced by their friends and neighbors’ experiences with insurance (Kunreuther, Sanderson, and Vetschera 1985; Gallagher 2014). Reacting to low rates of rainfall insurance uptake in Andhra Pradesh, India, Giné, Townsend, and Vickery (2008) suggest that “over time, lessons learned by insurance ‘early adopters’ will filter through to other households, generating higher penetration rates among poor households.” This paper seeks to understand how previous insurance payouts can affect future insurance purchasing decisions, and what mechanisms can explain this behavior. Using data on three years of insurance purchasers from the Indian microfinance institution BASIX, I find that customers who received an insurance payout are 9 to 22 percent more likely to repurchase in the following year than customers who did not receive any insurance payments. I show that this effect cannot be accounted for by neo-classical explanations, such as wealth effects, changing expectations about weather, or trust in the insurance company. Instead, I argue that it is caused by the behavioral effects of receiving an insurance payout. I test two main hypotheses as to why receiving payouts could increase insurance demand the following year. First, several studies have shown that weather shocks themselves could have an effect on insurance demand (Kunreuther, Sanderson, and Vetschera 1985; Eling, Pradhan, and Schmit 2014; Turner, Said, and Afzal 2014). This could happen because weather shocks change customers’ beliefs about future shocks, change their wealth, or simply increase the salience of shocks. I look for direct effects of weather by testing how rainfall in the year before insurance was introduced affected insurance purchases and find evidence that previous rainfall shocks tend to decrease insurance purchasing. This provides evidence against the argument that it is weather shocks as opposed to insurance payouts that are driving insurance purchases. Next, I test whether receiving insurance payouts could induce trust in the insurance company or learning about the insurance product. Bryan (2010) suggests that index insurance take-up is low due to ambiguity aversion, which should decrease as customers learn more about the product. I assume that if trust and learning are driving purchases, one should be able to witness spillover effects on other people in the village. This is because people in a village who witnessed payouts but did not receive them should also have been able to learn about insurance and gain trust in the insurance company. I do not find convincing evidence that these spillover effects are present, and argue that this is evidence that insurance repurchasing is not being driven by trust or learning. I instead argue that insurance repurchases are being driven by the psychological effects of receiving an insurance payout. Repurchase behavior is consistent with a number of behavioral explanations, such as recency bias, shifting reference points, and viewing insurance as a system of balanced reciprocity. This paper is related to a few separate lines of research. First, it contributes to a growing list of empirical studies that attempt to determine demand for weather index insurance (Giné, Townsend, and Vickery 2008; Cole et al. 2013; Mobarak and Rosenzweig 2013). One overarching conclusion from previous studies on index insurance is that demand for index products is low when provided at market rates and only increases when prices are slashed significantly. However, most of these studies look at insurance as a static purchasing decision, seeing what factors lead people to become first-time customers. There are a few of studies that do look at dynamic rainfall insurance decisions. Cole, Stein, and Tobacman (2014) study dynamic decision for rainfall insurance in Gujarat, India, over a period of seven years. They find that insurance payouts lead to greater purchasing for everyone in the village in the short term, but that these effects only persist for people who actually received payouts. Their point estimates for the effect an insurance payout are similar to those which we find here.2Hill and Robles (2010) provide rainfall insurance for free as part of an experimental game in Ethiopia and then return the next year to sell the same insurance. Despite the fact that two-thirds of the people who were granted insurance during the experiment received payouts, this group had a low take-up rate of 11 percent the next season, which was a lower rate than those who had not participated in the experiment. Karlan et al. (2014) study a multiyear rainfall insurance program in Ghana and find that people who received insurance payouts in the first year of the program are more likely to purchase in the second year. Also, they find that these effects spill over into customers’ social networks. Cai and Song (2012) show that experiences in hypothetical insurance games affect future real-world insurance purchase decisions. This study differs from these other works in a number of ways. Most important, the cited papers are all on small-scale experiments with closely controlled marketing and pricing strategies. This paper instead studies a large-scale commercial insurance operation, providing a much larger data set and real-world conditions. Although the previous studies have very strong internal validity as a result of their closely controlled experimental setting, this study complements them by studying a setting that is more likely to mirror commercial-level insurance operations; therefore, it provides a higher level of external validity. The paper will proceed as follows: Section I explains the insurance policies and data that will be studied in the empirical section. Section II provides the main empirical evidence, and shows that recipients of insurance payouts are more likely to purchase insurance the following year. Section III searches for evidence of a number of mechanisms by which this could take place. Section IV discusses a number of behavioral explanations for the results. Section V concludes and offers policy recommendations. I. Index Insurance and Customer Data In this analysis I study monsoon rainfall index insurance policies underwritten by the insurance company ICICI-LOMBARD and sold by BASIX, a microfinance institution based in Hyderabad. The policies insure against excess or deficit rainfall; they are calculated based on rainfall measured at a stated weather station. By basing payoffs on just rainfall, the policies have low monitoring and verification costs and also should be free of adverse selection and moral hazard (Collier, Skees, and Barnett 2009). These attributes make policies inexpensive to create and administer, which allows them to be sold in small quantities and priced at levels affordable for poor farmers. BASIX’s policies are designed to pay out in situations where adverse rainfall would cause a farmer to experience crop loss; therefore, they are calibrated to the water needs of local crops. BASIX insurance policies are sold in April and May, which are the months that precede the monsoon in India. Insurance policies cover only one season, so customers must purchase insurance again if they want coverage for the following year. The ratio of the premium to the expected payout averages around 2.5. This and other summary statistics for the insurance policies are shown in table 1. Table 1. Insurance Policy Summary Statistics Year 2005 2006 2007 Number of weather stations 34 42 28 Average premium for three phases (Rs) 290 295 287 Expected payout (Rs, using rainfall 1961–2004) 119 73 80 Ratio of premium to expected payout 2.67 2.47 2.12 Mean percentage of years policy would have paid out 1961–2004 22.9 15.6 13.9 Year 2005 2006 2007 Number of weather stations 34 42 28 Average premium for three phases (Rs) 290 295 287 Expected payout (Rs, using rainfall 1961–2004) 119 73 80 Ratio of premium to expected payout 2.67 2.47 2.12 Mean percentage of years policy would have paid out 1961–2004 22.9 15.6 13.9 Notes: This table lists basic features of the insurance policies studied in this paper. Each policy offers three phases of coverage with different premiums, but customers are not required to purchase all three. However, as purchasing all three phases was the most common behavior, the average premium for purchasing all phases of coverage is listed. The expected payout is estimated using the APHRODITE dataset, for the period 1961–2004. As pricing strategies for ICICI-LOMBARD are proprietary, these estimates may not correspond to their internal estimates of expected payouts. However, the APHRODITE data set likely draws from the same historical data used to price the insurance policies. The ration of premium to expected payout is the average premium (averaged across all weather stations) divided by the average expected payout. Source: Authors’ analysis based on data from BASIX and APHRODITE. Table 1. Insurance Policy Summary Statistics Year 2005 2006 2007 Number of weather stations 34 42 28 Average premium for three phases (Rs) 290 295 287 Expected payout (Rs, using rainfall 1961–2004) 119 73 80 Ratio of premium to expected payout 2.67 2.47 2.12 Mean percentage of years policy would have paid out 1961–2004 22.9 15.6 13.9 Year 2005 2006 2007 Number of weather stations 34 42 28 Average premium for three phases (Rs) 290 295 287 Expected payout (Rs, using rainfall 1961–2004) 119 73 80 Ratio of premium to expected payout 2.67 2.47 2.12 Mean percentage of years policy would have paid out 1961–2004 22.9 15.6 13.9 Notes: This table lists basic features of the insurance policies studied in this paper. Each policy offers three phases of coverage with different premiums, but customers are not required to purchase all three. However, as purchasing all three phases was the most common behavior, the average premium for purchasing all phases of coverage is listed. The expected payout is estimated using the APHRODITE dataset, for the period 1961–2004. As pricing strategies for ICICI-LOMBARD are proprietary, these estimates may not correspond to their internal estimates of expected payouts. However, the APHRODITE data set likely draws from the same historical data used to price the insurance policies. The ration of premium to expected payout is the average premium (averaged across all weather stations) divided by the average expected payout. Source: Authors’ analysis based on data from BASIX and APHRODITE. BASIX conducts marketing visits through village meetings and door-to-door visits. The first step is to hold a group meeting in the village, where potential customers are shown a marketing video that includes details about rainfall insurance (and other BASIX products). It then speaks with visitors and answers questions. The BASIX team then makes a follow-up visit where it goes door to door, trying to sell BASIX products, including rainfall insurance. Data The data set consists of the entire set of BASIX’s purchasers of rainfall index insurance from 2005 through 2007, which covers six states.3 Though it ran small pilots in 2003 and 2004, BASIX began to mass-market rainfall insurance starting in 2005. The data contain limited personal information about each customer, including their location, how many policies they purchased, and what payouts they received during that season. The BASIX data covers 42 weather stations and includes a total of 19,882 customers during the period 2005–2007.4 After numerous rainfall shocks in 2006, BASIX realized that many customers who had purchased only a small amount of insurance were disappointed that they received small payouts. In response, BASIX instituted a rule in 2007 that required all customers to purchase insurance coverage with a maximum payout of at least Rs 3,000. This was meant to encourage people to buy a level of coverage that would actually provide meaningful payouts in the event of a shock, but it resulted in a sharp decrease in the number of customers in 2007. A summary of characteristics of BASIX customers is given in table 2. Table 2. Customer Summary Statistics Year 2005 2006 2007 Total Number of villages 949 1,440 437 2,826 Number of villages (for whom insurance is available in the following year) 851 683 NA 1,534 Total number of buyers 6,425 10,074 3,375 19,874 Total number of buyers (for whom insurance is available in following year) 5,579 5,418 NA 10,997 Total number of buyers (in villages where there was at least one buyer the following year) 2,462 1,739 NA 4,201 Number of buyers who repurchase in following year 453 364 NA 817 Buyers receiving payouts 351 1,346 529 2,226 Average Average sum insured (Rs) 3,055 1,612 3,547 2,738 Average payout (Rs) 11 60 88 53 Average payout (Rs, if payout>0) 195 360 553 370 Year 2005 2006 2007 Total Number of villages 949 1,440 437 2,826 Number of villages (for whom insurance is available in the following year) 851 683 NA 1,534 Total number of buyers 6,425 10,074 3,375 19,874 Total number of buyers (for whom insurance is available in following year) 5,579 5,418 NA 10,997 Total number of buyers (in villages where there was at least one buyer the following year) 2,462 1,739 NA 4,201 Number of buyers who repurchase in following year 453 364 NA 817 Buyers receiving payouts 351 1,346 529 2,226 Average Average sum insured (Rs) 3,055 1,612 3,547 2,738 Average payout (Rs) 11 60 88 53 Average payout (Rs, if payout>0) 195 360 553 370 Notes: This table lists summary statistics for the individual sample. The sample only includes people who purchased rainfall insurance from BASIX in a given year. Source: Authors’ analysis based on data from BASIX. Table 2. Customer Summary Statistics Year 2005 2006 2007 Total Number of villages 949 1,440 437 2,826 Number of villages (for whom insurance is available in the following year) 851 683 NA 1,534 Total number of buyers 6,425 10,074 3,375 19,874 Total number of buyers (for whom insurance is available in following year) 5,579 5,418 NA 10,997 Total number of buyers (in villages where there was at least one buyer the following year) 2,462 1,739 NA 4,201 Number of buyers who repurchase in following year 453 364 NA 817 Buyers receiving payouts 351 1,346 529 2,226 Average Average sum insured (Rs) 3,055 1,612 3,547 2,738 Average payout (Rs) 11 60 88 53 Average payout (Rs, if payout>0) 195 360 553 370 Year 2005 2006 2007 Total Number of villages 949 1,440 437 2,826 Number of villages (for whom insurance is available in the following year) 851 683 NA 1,534 Total number of buyers 6,425 10,074 3,375 19,874 Total number of buyers (for whom insurance is available in following year) 5,579 5,418 NA 10,997 Total number of buyers (in villages where there was at least one buyer the following year) 2,462 1,739 NA 4,201 Number of buyers who repurchase in following year 453 364 NA 817 Buyers receiving payouts 351 1,346 529 2,226 Average Average sum insured (Rs) 3,055 1,612 3,547 2,738 Average payout (Rs) 11 60 88 53 Average payout (Rs, if payout>0) 195 360 553 370 Notes: This table lists summary statistics for the individual sample. The sample only includes people who purchased rainfall insurance from BASIX in a given year. Source: Authors’ analysis based on data from BASIX. For rainfall data, I use a historical daily grid of rainfall, which is interpolated based on readings from thousands of rainfall stations throughout India. The data are provided by the Asian Precipitation Highly Resolved Observational Data Integration Towards Evaluation of Water Resources.5 This data set has daily readings of rainfall for the period 1961–2004, at a precision of .25.6 The data are extrapolated based on historical rainfall readings from weather stations primarily provided by the Indian Meterological Department. For each block, the data contain the amount of rainfall in millimeters and the number of stations within the grid that contributed to the data. These data are used to evaluate how the insurance policies would have paid out historically, which can be used as a proxy for past rainfall shocks.7 The three individual years of BASIX customer data were converted into a panel by manually matching individual customers using available identifying data. Errors in matching customers from year to year create the possibility of introducing nonclassical measurement error into the analysis. In section I discuss the possible consequences of such errors and how they affect interpretation of the estimates. II. Results: The Effect of Payouts on Take-Up In this section I address the central question: is receiving an insurance payout correlated with repurchasing insurance the following year? To do this, I examine BASIX’s customers in 2005 and 2006, and regress repurchasing on payout reception and a year dummy. The basic econometric specification is as follows: \begin{eqnarray*} {y}_{i}{,}_{t+1}=\alpha +{\beta }_{1}{P}_{i,t}+{B}_{2}{D}_{2006}+{{\epsilon}}_{t,i} \end{eqnarray*} Here, $${y}_{i}{,}_{t+1}$$ represents whether subject $$i$$ purchases insurance at time $$t+1$$, and $${P}_{i,t}$$ is a dummy variable that takes a value of 1 if person $$i$$ receives an insurance payout at time $$t$$.8 The sample is all buyers of insurance from 2005 and 2006, and I include a dummy ($${D}_{2006}$$) that takes a value of 1 for purchasers in the year 2006 to control for time effects. Also, I only include purchasers who have weather insurance contracts available in their area in the following year.9 These results are presented in table 3, and column 1 reports the baseline OLS results. It shows that receiving a payout is associated with a 9 percent increased chance of repurchasing insurance the following year, which means that those who receive an insurance payout are more than twice as likely to purchase insurance the following year than those who did not receive a payout.10 However, this coefficient is not statistically significant at standard levels (p = .18). The effect on repurchasing may depend on the size of the payout as well. In column 2 I add two new continuous variables to the regression: the ratio of the payout received to the premium paid (which I will call the “payout ratio”) and the payout ratio squared. In this specification, the dummy on receiving a payout flips to negative and significant. However, the payout ratio is positive and significant while the payout ratio squared is negative and significant. Together, this suggests that people who receive very small payouts are less likely to purchase insurance, while those who have large payouts are more likely to purchase. The overall effect of receiving a payout switches from negative to positive when the payout ratio reaches around .77. Table 3. Effect of Payouts on Repurchasing Dependent variable is customer repurchasing insurance (1) (2) (3) (4) Received payout dummy 0.090 −0.088*** 0.222*** −0.195** (0.058) (0.028) (0.077) (0.074) Ratio of payout to Premium 0.123*** 0.246*** (0.037) (0.023) Payout ratio squared −0.012 *** −0.024*** (0.004) (0.002) Year 2006 dummy −0.025 −0.040 −0.027 −0.034*** (0.036) (0.034) (0.064) (0.063) Constant 0.070** 0.077*** 0.165*** 0.168*** (0.027) (0.028) (0.033) (0.033) Effect at average payout 0.207 0.426 Marketing restricted sample NO NO YES YES Observations 10,997 10,997 4,201 4,201 R-squared 0.01 0.034 0`.02 0.058 Dependent variable is customer repurchasing insurance (1) (2) (3) (4) Received payout dummy 0.090 −0.088*** 0.222*** −0.195** (0.058) (0.028) (0.077) (0.074) Ratio of payout to Premium 0.123*** 0.246*** (0.037) (0.023) Payout ratio squared −0.012 *** −0.024*** (0.004) (0.002) Year 2006 dummy −0.025 −0.040 −0.027 −0.034*** (0.036) (0.034) (0.064) (0.063) Constant 0.070** 0.077*** 0.165*** 0.168*** (0.027) (0.028) (0.033) (0.033) Effect at average payout 0.207 0.426 Marketing restricted sample NO NO YES YES Observations 10,997 10,997 4,201 4,201 R-squared 0.01 0.034 0`.02 0.058 Notes: The dependent variable is a dummy variable that takes the value of one if the person purchased insurance the following year. Columns 1 and 2 contain the sample of all buyers of insurance in 2005 and 2006 for whom insurance coverage was offered in their area the following year. Columns 3 and 4 contain the marketing restricted sample, which is restricted only to villages where at least one person purchased insurance in the following year. The average ratio of payout to premium (for those who received payouts) for the full sample is 3.6; for those in the marketing restricted sample it is 4.5. All regressions contain state fixed effects. Errors are clustered at the weather station level. *** p < 0.01, **p < 0.05, *p < 0.1. Source: Authors’ analysis based on data from BASIX. View Large Table 3. Effect of Payouts on Repurchasing Dependent variable is customer repurchasing insurance (1) (2) (3) (4) Received payout dummy 0.090 −0.088*** 0.222*** −0.195** (0.058) (0.028) (0.077) (0.074) Ratio of payout to Premium 0.123*** 0.246*** (0.037) (0.023) Payout ratio squared −0.012 *** −0.024*** (0.004) (0.002) Year 2006 dummy −0.025 −0.040 −0.027 −0.034*** (0.036) (0.034) (0.064) (0.063) Constant 0.070** 0.077*** 0.165*** 0.168*** (0.027) (0.028) (0.033) (0.033) Effect at average payout 0.207 0.426 Marketing restricted sample NO NO YES YES Observations 10,997 10,997 4,201 4,201 R-squared 0.01 0.034 0`.02 0.058 Dependent variable is customer repurchasing insurance (1) (2) (3) (4) Received payout dummy 0.090 −0.088*** 0.222*** −0.195** (0.058) (0.028) (0.077) (0.074) Ratio of payout to Premium 0.123*** 0.246*** (0.037) (0.023) Payout ratio squared −0.012 *** −0.024*** (0.004) (0.002) Year 2006 dummy −0.025 −0.040 −0.027 −0.034*** (0.036) (0.034) (0.064) (0.063) Constant 0.070** 0.077*** 0.165*** 0.168*** (0.027) (0.028) (0.033) (0.033) Effect at average payout 0.207 0.426 Marketing restricted sample NO NO YES YES Observations 10,997 10,997 4,201 4,201 R-squared 0.01 0.034 0`.02 0.058 Notes: The dependent variable is a dummy variable that takes the value of one if the person purchased insurance the following year. Columns 1 and 2 contain the sample of all buyers of insurance in 2005 and 2006 for whom insurance coverage was offered in their area the following year. Columns 3 and 4 contain the marketing restricted sample, which is restricted only to villages where at least one person purchased insurance in the following year. The average ratio of payout to premium (for those who received payouts) for the full sample is 3.6; for those in the marketing restricted sample it is 4.5. All regressions contain state fixed effects. Errors are clustered at the weather station level. *** p < 0.01, **p < 0.05, *p < 0.1. Source: Authors’ analysis based on data from BASIX. View Large One point of concern with these results is that there are many cases where there are multiple purchasers of insurance in a certain village in one year and then zero in the next year. Although this could be the result of people simply being unsatisfied with insurance, the large amount of villages that suddenly drop to zero purchasers is suspicious, especially since the BASIX data does not contain information about whether marketing activities took place in a given village in a given year. For all the villages that had purchasers in one year and then none in the next year, it is quite likely that no BASIX representative visited the village; therefore the customer did not really have a chance to purchase the insurance. If this was the case, it would make sense to exclude these villages from the analysis, as the previous year’s payout would have no way to possibly influence a customer’s purchase decision. In columns 3 and 4 I exclude villages that had no purchasers the following year from the analysis, creating what I call the “Marketing Restricted Sample.” Restricting the sample this way results in a drop of the number of observations from 10,977 to 4,201, and it causes the coefficient on receiving a payout in the simple specification to more than double to .22, now significant at the 1 percent level. Column 4 follows the same pattern as the unrestricted sample, but with larger coefficients, and predicts the overall effect of a payout becoming positive when the payout ration is 87 percent. These specifications provide evidence that the omitted information about whether a village received marketing was downward biasing the results of the original specification. The coefficients generated in this restricted sample may be incorrect because the decision to market to certain villages and not others is most likely not exogenous. If the marketing teams decided whether or not to market to certain villages based on the previous year’s rainfall or experience with insurance then the results could be biased. For instance, assume that there were a number of villages that experienced a rainfall shock but received very low payouts, making them unhappy with insurance. If the marketing teams knew this, they may have decided to not market to as many of these villages, therefore censoring villages that received a payout but were likely to have few repeat buyers. Regressions that use previous years’ payout characteristics to try to predict whether insurance is sold in a village the following year do not reveal any patterns that would suggest selection bias (shown in table S.5 in the supplemental appendix, available at https://academic.oup.com/wber), but they may miss more subtle selection patterns. It is possible that the coefficient for the marketing restricted sample is upward biased and it therefore would be reasonable to regard the coefficients in columns 1 and 3 as lower and upper bounds, respectively. The payout ratio has a positive and strongly significant effect while the squared term is smaller and negative. This suggests that higher insurance payouts result in greater propensity to purchase the following year, but that the marginal effects flatten out for larger payouts. Also, the simple dummy of receiving a payout flips to negative, suggesting that small payouts have a negative effect on purchasing. In fact, payouts have a positive effect only when the payout ratio nears 1. This is consistent with an interpretation that the effects of an insurance payout are being driven by customers experiencing a net gain on their insurance transaction the previous year. The supplemental appendix (tables S.3 and S.4) contains a number of robustness checks of this main specification, and they all support the conclusions of table 3. This includes running the regression on the full sample (including areas where no insurance was available the following year), the balanced sample (where insurance was offered all years), and the balanced marketing sample (where insurance was purchased in all years). One may be concerned that the linear probability model may give biased estimates, especially since such a small percentage of the sample were repeat buyers. Therefore, the results from a probit model are also presented in the supplemental appendix and give similar results to OLS. As mentioned earlier, the dependent variable in this regression was generated by manually matching customers from one year to another, and therefore it is likely measured with some error. Although there is no reason to believe that this measurement error is correlated with any independent variables in the regression, since the dependent variable in the regression is a dummy variable, this can lead to downward bias on the estimated coefficients. In order to get a feel for the potential magnitude of this error, I run simulations where I assume that the BASIX data has been matched completely correctly and then induce “measurement error” by randomly changing the dependent variable of whether people purchased the following year or not. With the introduction of 10 percent matching errors (with an equal probability of a mismatch for buyer or nonbuyers), the coefficient on receiving a payout in the full sample (column 1) falls from .090 to an average of .072 over 1,000 simulations. For the marketing restricted sample in column 2, it drops from .222 to .178. In other words, if one assumes 10 percent matching errors, then the estimated coefficients are likely to be underestimated by around 20 percent. It also may be possible that most of the error came from being unable to identify positive matches, possibly due to different members of a household signing the insurance contract from year to year. Repeating the above simulation but assuming that only people who were found not to have bought the next year could have been errors, the coefficients become underestimated by around 10 percent. Although the exact form and structure of the matching errors cannot be known, it is possible that the reported coefficients are somewhat lower (in absolute value) than the true coefficients. Overall, the results indicate that receiving an insurance payout correlates with a roughly 9 to 22 percent higher chance of repurchasing the next year compared with someone who purchased insurance but did not receive a payout. They also suggest that higher payouts lead to a greater chance of repurchasing, and that very low payouts may actually have a negative effect. The next section will explore some possible mechanisms for these results. III. Explanations for Main Result The results of the previous section show a simple correlation; people who receive insurance payouts are more likely to purchase insurance the following year. In this section I explore a number of reasons consistent with a neoclassical framework as to why this may be the case. Generally, payouts are given in the context of a rainfall shock. Therefore, the measured effects of payouts shown above could be caused by the rainfall shock itself as opposed to the insurance payout. This may be the case because of people learning about the effects of shocks, or experiencing changes in wealth or liquidity. To do this, I look at villages in the first year they were offered insurance and see if a rainfall shock the previous year correlates with greater insurance take-up. To the contrary, I find that villages that had a rainfall shock the previous year were actually less likely to purchase insurance the following year, which provides strong evidence against the argument that weather as opposed to payouts are driving the main result. I next consider the widely considered hypothesis that receiving insurance payouts would cause people to gain trust in the insurance company and learn about insurance and thus make them more likely to purchase insurance in the future. To do this, I assume that in order to gain trust in the insurance company or learn how insurance works, one would not have to receive a payout themselves; witnessing a neighbor receive a payout should have similar effects. I therefore look for evidence of spillovers within a village and do not find evidence that witnessing a payout without receiving one too has a significant effect on the propensity to purchase the following year. I also consider the possibility that payouts cause increased take-up due to direct wealth and/or liquidity effects as opposed to psychological effects. Although I do not have data to empirically separate these possible mechanisms, I argue that because of the timing and circumstances of rainfall insurance payouts, wealth and liquidity are unlikely to play an important role. Finally, I address the concern of unobserved marketing variation. Although the effect of this omitted variable is admittedly difficult to measure, I argue it is unlikely to be driving the central results. Given that I fail to find support for all of these explanations, I argue in the next section that repurchases are being driven by behavioral responses to receiving insurance payments. Direct Effects of Rainfall Since most rainfall insurance payouts come at the same time as a rainfall shock, it is possible that the rainfall shocks themselves as opposed to the insurance payouts are what is driving increased take-up the following year. There is some evidence for this happening in developed markets; for example, Kunreuther, Sanderson, and Vetschera (1985) note that purchases of flood and earthquake insurance in the United States spike after a recent event, even if people were not insurance customers before. There are a number of theories that could explain this behavior. First, recent experiences with rainfall could change subjects’ beliefs about the probability of a rainfall shock the following year (this is proposed as “recency bias” in Karlan et al. [2014]). If there is actual autocorrelation of rainfall events or if the subject has limited knowledge about the effects of rainfall shocks, people may update their beliefs about shocks and therefore have more desire for insurance the following year. Alternatively, recently experiencing a rainfall shock could make shocks more salient, increasing the chance they will buy insurance the following year. Also, rainfall shocks may affect the wealth of the farmers. If farmers become poorer because of bad rainfall, CRRA utility would suggest that they would be even more risk averse the next year as a second shock would cause greater disutility. I start by examining whether there is actual autocorrelation in the rainfall data. To test for autocorrelation, I create a panel of various rainfall indicators for the period 1961–2004 for each weather station. For each indicator, I run a regression of six lags of the variable on the current value, including weather station fixed effects. This specification is given in equation: \begin{eqnarray*} {y}_{w,t}=\alpha +\sum _{j=1}^{5}{\beta }_{j}{y}_{w,t-j}+{v}_{w}+{{\epsilon}}_{w,t} \end{eqnarray*} $${y}_{w,t}$$ is the rainfall variable at weather station $$w$$ at time $$t$$. $${v}_{w}$$ is a weather station fixed error term, while $${{\epsilon}}_{w,t}$$ is an idiosynchratic error term. These results are presented in column 1 of table 4, with just the coefficient on the first lag shown. Although a fixed-effects regression with a lagged dependent variable is not generally consistent, it will converge to the true value as $$T\to \infty $$. As T is relatively large (38), these estimates are likely to suffer from little bias. I also run a regression of the first lag using previous lags as instruments, using the methodology proposed by Arellano and Bond (1991), with results presented in column 2. The results from both specifications are similar, and show a negative first-order autocorrelation in rainfall that appears to be driven by rains early in the season. The bottom two rows test for autocorrelation of rainfall shocks using the parameters of the 2005 insurance policy to determine shocks. “Would Have Been Payout” is a dummy variable that takes a value of 1 if the insurance policy of 2005 would have given a payout; “Total Insurance Payout” is the size of this payout. By these measures, shocks do not appear to exhibit significant positive first-order autocorrelation. Table 4. Rainfall Autocorrelation Fixed effects Arellano-Bond (1) (2) Total rainfall −0.106*** −.086*** (.030) (.021) Phase 1 rainfall −.090*** −.075*** (.030) (.029) Phase 2 rainfall −.018 −.026 (.030) (.028) Phase 3 rainfall −.029 .007 (.030) (.028) Would have been payout .023 .017 (.030) (.022) Total insurance payout −.0353 .004 (.030) (.028) Fixed effects Arellano-Bond (1) (2) Total rainfall −0.106*** −.086*** (.030) (.021) Phase 1 rainfall −.090*** −.075*** (.030) (.029) Phase 2 rainfall −.018 −.026 (.030) (.028) Phase 3 rainfall −.029 .007 (.030) (.028) Would have been payout .023 .017 (.030) (.022) Total insurance payout −.0353 .004 (.030) (.028) Notes: Coefficients reported are from separate univariate regressions. The Fixed Effects specification is OLS with six lags of the dependent variable. The Arellano-Bond regression contains one lag of the dependent variable, with this lag instrumented by five previous lags. Observation are years 1967–2004 for the fixed-effects regression, and years 1962-2004 for the Arellano-Bond regression. All regressions contain weather station fixed effects. Robust standard errors are in parentheses. *** p < 0.01, **p < 0.05, *p < 0.1. Source: Authors’ analysis based on data from BASIX and APHRODITE. View Large Table 4. Rainfall Autocorrelation Fixed effects Arellano-Bond (1) (2) Total rainfall −0.106*** −.086*** (.030) (.021) Phase 1 rainfall −.090*** −.075*** (.030) (.029) Phase 2 rainfall −.018 −.026 (.030) (.028) Phase 3 rainfall −.029 .007 (.030) (.028) Would have been payout .023 .017 (.030) (.022) Total insurance payout −.0353 .004 (.030) (.028) Fixed effects Arellano-Bond (1) (2) Total rainfall −0.106*** −.086*** (.030) (.021) Phase 1 rainfall −.090*** −.075*** (.030) (.029) Phase 2 rainfall −.018 −.026 (.030) (.028) Phase 3 rainfall −.029 .007 (.030) (.028) Would have been payout .023 .017 (.030) (.022) Total insurance payout −.0353 .004 (.030) (.028) Notes: Coefficients reported are from separate univariate regressions. The Fixed Effects specification is OLS with six lags of the dependent variable. The Arellano-Bond regression contains one lag of the dependent variable, with this lag instrumented by five previous lags. Observation are years 1967–2004 for the fixed-effects regression, and years 1962-2004 for the Arellano-Bond regression. All regressions contain weather station fixed effects. Robust standard errors are in parentheses. *** p < 0.01, **p < 0.05, *p < 0.1. Source: Authors’ analysis based on data from BASIX and APHRODITE. View Large This evidence casts doubt on the hypothesis that positive autocorrelation of weather events is driving increased insurance purchasing. It appears that total rainfall is actually negatively autocorrelated, whereas shocks (which are proxied by the insurance contract giving a payout) do not appear to be correlated at all. Even if there is no positive autocorrelation of rainfall, there may be other aspects about experiencing a shock that result in people having a higher propensity to purchase insurance. In order to look at the results of weather separately from the effects of insurance, I analyze how previous weather events affected insurance purchase decisions in the first year that insurance was offered to BASIX customers, which was 2005. To accomplish this, I first aggregate the purchasing data to the village level and then test to see whether villages that experienced a rainfall shock in 2004 had more insurance purchasers in 2005 than villages that did not experience a rainfall shock. A shock is defined using each location’s insurance policies in 2005: If insurance would have paid out in 2004 based on the structure of the 2005 weather policy, this is deemed a rainfall shock. As the quality of the rainfall data is related to the amount of nearby weather stations, I weight the observations based on the number of nearby rainfall stations.11 Also, I create a hypothetical payout ratio, similar to the “Ratio of Payout to Premium” variable presented in table 3. This is the ratio of the amount that the 2005 policy would have paid out in 2004 divided by the premium of the policy. The results of this regression are presented in table 5.12 Table 5. Direct Effects of Rainfall on Purchasing Dependent variable is number of buyers in 2005 (1) (2) (3) (4) Would have been payout in 2004 −3.843*** −4.592*** −5.045** −3.788* (0.987) (1.039) (2.173) (1.898) Ratio of hypothetical 2004 payout to 2005 Premium 4.365 −0.755 (4.610) (5.543) Payout ratio squared −1.991 −0.279 (1.814) (2.064) Constant 8.001*** 0.651 7.985*** 1.015 (0.714) (6.341) (0.713) (6.494) Weather station constants NO YES NO YES Observations 733 733 733 733 R-squared 0.073 0.094 0.075 0.097 Dependent variable is number of buyers in 2005 (1) (2) (3) (4) Would have been payout in 2004 −3.843*** −4.592*** −5.045** −3.788* (0.987) (1.039) (2.173) (1.898) Ratio of hypothetical 2004 payout to 2005 Premium 4.365 −0.755 (4.610) (5.543) Payout ratio squared −1.991 −0.279 (1.814) (2.064) Constant 8.001*** 0.651 7.985*** 1.015 (0.714) (6.341) (0.713) (6.494) Weather station constants NO YES NO YES Observations 733 733 733 733 R-squared 0.073 0.094 0.075 0.097 Notes: The dependent variable is the number of buyers in 2005, which was the first year rainfall insurance was offered to our sample. The unit of observation is the village. Would have been Payout in 2004 is a dummy that takes a value of 1 if there would have been a payout in 2004 had the 2005 policy been offered in that year. This is calculated based on daily rainfall data from APHRODITE and each policy's payout structure. The weather station constants are the premium in 2005, the average historical payout, the total number of historical payouts, and the standard deviation of rainfall. Historical variables are calculated for the period 1962-2004. Observations are weighted by the quality of rainfall data. If there are no rainfall stations contributing to the APHRODITE data in 2004 within a .75°x.75° grid around the desired BASIX weather station, the observation is given a weight of 1. If there is a least one weather station in this .75°x.75°, the observation is given a weight of 1.5. If there is a rainfall station within the .25°x.25° grid, the observation is given a weight of 2. The weighted results to not differ significantly from the unweighted results. The number of observations is 733 out of a total of 949 villages in the sample in 2005, as APHRODITE data was available for only a subset of locations. All specifications include state fixed effects. Errors clustered at the weather station level. *** p < 0.01, **p < 0.05, *p < 0.1. Source: Authors’ analysis based on data from BASIX and APHRODITE. View Large Table 5. Direct Effects of Rainfall on Purchasing Dependent variable is number of buyers in 2005 (1) (2) (3) (4) Would have been payout in 2004 −3.843*** −4.592*** −5.045** −3.788* (0.987) (1.039) (2.173) (1.898) Ratio of hypothetical 2004 payout to 2005 Premium 4.365 −0.755 (4.610) (5.543) Payout ratio squared −1.991 −0.279 (1.814) (2.064) Constant 8.001*** 0.651 7.985*** 1.015 (0.714) (6.341) (0.713) (6.494) Weather station constants NO YES NO YES Observations 733 733 733 733 R-squared 0.073 0.094 0.075 0.097 Dependent variable is number of buyers in 2005 (1) (2) (3) (4) Would have been payout in 2004 −3.843*** −4.592*** −5.045** −3.788* (0.987) (1.039) (2.173) (1.898) Ratio of hypothetical 2004 payout to 2005 Premium 4.365 −0.755 (4.610) (5.543) Payout ratio squared −1.991 −0.279 (1.814) (2.064) Constant 8.001*** 0.651 7.985*** 1.015 (0.714) (6.341) (0.713) (6.494) Weather station constants NO YES NO YES Observations 733 733 733 733 R-squared 0.073 0.094 0.075 0.097 Notes: The dependent variable is the number of buyers in 2005, which was the first year rainfall insurance was offered to our sample. The unit of observation is the village. Would have been Payout in 2004 is a dummy that takes a value of 1 if there would have been a payout in 2004 had the 2005 policy been offered in that year. This is calculated based on daily rainfall data from APHRODITE and each policy's payout structure. The weather station constants are the premium in 2005, the average historical payout, the total number of historical payouts, and the standard deviation of rainfall. Historical variables are calculated for the period 1962-2004. Observations are weighted by the quality of rainfall data. If there are no rainfall stations contributing to the APHRODITE data in 2004 within a .75°x.75° grid around the desired BASIX weather station, the observation is given a weight of 1. If there is a least one weather station in this .75°x.75°, the observation is given a weight of 1.5. If there is a rainfall station within the .25°x.25° grid, the observation is given a weight of 2. The weighted results to not differ significantly from the unweighted results. The number of observations is 733 out of a total of 949 villages in the sample in 2005, as APHRODITE data was available for only a subset of locations. All specifications include state fixed effects. Errors clustered at the weather station level. *** p < 0.01, **p < 0.05, *p < 0.1. Source: Authors’ analysis based on data from BASIX and APHRODITE. View Large Column 1 presents the baseline regression, which shows that villages that experienced a rainfall shock in 2004 actually had an average of 3.8 $$\textit{fewer}$$ purchasers in 2005. One worry with this regression may be that because the insurance policies and rainfall patterns of each location are different, the definition of a shock may vary from one place to another. Therefore, the estimates may be improved with the inclusion of location and policy-specific covariates, which I title “Weather Station Constants.” In column 2 I add controls for the historical average rainfall, historical rainfall standard deviation, the policy premium in 2005, historical average payout of the policy, and the percentage of historical years there would have been a payout. Note that all the “historical” data is calculated fof the period 1961–2000. With the addition of these controls, the coefficients on having a rainfall shock in 2004 remains negative, and even decreases slightly. Following previous results that suggest that the size of the insurance payout is important, in columns 3 and 4 I include variables for the severity of the shock in 2004 using the ratio of the hypothetical payout to the premium (the payout ratio) and the payout ratio squared. In both specifications these variables are insignificant, suggesting that most of the variation in purchasing in 2005 is explained by the binary shock variable. The main conclusion to be drawn from these regressions is that the data does not support the hypothesis that bad weather induces people to purchase insurance in the following season. If anything, bad weather seems to decrease insurance purchases. I can only speculate on the reasons for this; it may be caused by the fact that people recognize the actual negative autocorrelation of rainfall, or it may be that the rainfall shocks decrease the available liquidity to purchase insurance the following year. Regardless, these data provide relatively convincing evidence that the direct effect of weather is not causing people who receive insurance payments to purchase again the following year. Trust, Learning, and Spillover Effects It is also possible that the propensity to purchase insurance after receiving a payout results from learning about insurance and trusting the insurance company, as opposed to being a direct result of the payout. In order to separate the effects of trust and learning from that of receiving the payout, I make the assumption that if trust and learning are playing an important role in causing people to purchase insurance after they have received a payout, then one should be able to see a positive spillover effect of payouts within the village.13 This is because one shouldn’t need to actually receive a payout to gain the effects of trust and learning; someone who witnesses a payout gains all the same information as someone who receives a payout. To explore spillover effects, I use two sources of variation: within-village and across village. First, buyers of insurance could choose to only purchase certain phases of the insurance rather than obtaining coverage for the whole monsoon season. Therefore, within a village where there were payouts, not all insurance purchasers received them. I can therefore look at the effects of insurance payouts on the buyers in the same village that did not actually receive them. Next, I use across-village variation to determine whether villages where payouts occurred had a greater number of new buyers in the following year. If the purchasing decisions of new buyers are similar to that of repeat buyers, this is evidence that spillovers are occurring, and therefore trust and/or learning is a plausible explanation for repurchasing behavior. The intravillage analysis is presented in table 6. The specification is the same as table 3, but with an additional dummy variable that takes a value of 1 for all observations where there was any payout in the village in the given year. The coefficient on this variable represents the effect of payouts on insurance purchasers who purchased insurance in villages where there was a payout but purchased only phases of the policy where there was not a payout. There were 427 cases where this happened. The coefficient on this variable is negative and significant in all specifications, suggesting that spillover effects are not playing a large role (and are overwhelmed by any direct effect of rainfall shocks). Table 6. Spillovers to Other Insurance Purchasers (1) (2) (3) (4) Received insurance payout 0.124** −0.05** 0.313*** −0.11 (0.056) (0.026) (0.072) (0.065) Insurance payout made in village −0.04* −0.04* −0.1** −0.09** (0.021) (0.020) (0.035) (0.036) Ratio of payout received to premium 0.123*** 0.246*** (0.037) (0.023) Payout ratio squared −0.01*** −0.02*** (0.004) (0.002) Year 2006 dummy −0.02 −0.04 −0.03 −0.03 (0.037) (0.034) (0.064) (0.063) Constant 0.071** 0.078*** 0.167*** 0.169*** (0.027) (0.028) (0.033) (0.033) Marketing restricted sample NO NO YES YES $$\hat{\rm R}$$2 0.014 0.034 0.036 0.058 N 10997 10997 4201 4201 (1) (2) (3) (4) Received insurance payout 0.124** −0.05** 0.313*** −0.11 (0.056) (0.026) (0.072) (0.065) Insurance payout made in village −0.04* −0.04* −0.1** −0.09** (0.021) (0.020) (0.035) (0.036) Ratio of payout received to premium 0.123*** 0.246*** (0.037) (0.023) Payout ratio squared −0.01*** −0.02*** (0.004) (0.002) Year 2006 dummy −0.02 −0.04 −0.03 −0.03 (0.037) (0.034) (0.064) (0.063) Constant 0.071** 0.078*** 0.167*** 0.169*** (0.027) (0.028) (0.033) (0.033) Marketing restricted sample NO NO YES YES $$\hat{\rm R}$$2 0.014 0.034 0.036 0.058 N 10997 10997 4201 4201 Notes: The dependent variable is a dummy variable that takes the value of 1 if the person purchased insurance the following year. Columns 1 and 2 contain the sample of all buyers of insurance in 2005 and 2006 for whom insurance coverage was offered in their area the following year. Columns 3 and 4 contain the marketing restricted sample, which is restricted only to villages where at least one person purchased insurance in the following year. All regressions contain state fixed effects. Errors are clustered at the weather-station level. *** p < 0.01, **p < 0.05, *p < 0.1. Source: Authors’ analysis based on data from BASIX. View Large Table 6. Spillovers to Other Insurance Purchasers (1) (2) (3) (4) Received insurance payout 0.124** −0.05** 0.313*** −0.11 (0.056) (0.026) (0.072) (0.065) Insurance payout made in village −0.04* −0.04* −0.1** −0.09** (0.021) (0.020) (0.035) (0.036) Ratio of payout received to premium 0.123*** 0.246*** (0.037) (0.023) Payout ratio squared −0.01*** −0.02*** (0.004) (0.002) Year 2006 dummy −0.02 −0.04 −0.03 −0.03 (0.037) (0.034) (0.064) (0.063) Constant 0.071** 0.078*** 0.167*** 0.169*** (0.027) (0.028) (0.033) (0.033) Marketing restricted sample NO NO YES YES $$\hat{\rm R}$$2 0.014 0.034 0.036 0.058 N 10997 10997 4201 4201 (1) (2) (3) (4) Received insurance payout 0.124** −0.05** 0.313*** −0.11 (0.056) (0.026) (0.072) (0.065) Insurance payout made in village −0.04* −0.04* −0.1** −0.09** (0.021) (0.020) (0.035) (0.036) Ratio of payout received to premium 0.123*** 0.246*** (0.037) (0.023) Payout ratio squared −0.01*** −0.02*** (0.004) (0.002) Year 2006 dummy −0.02 −0.04 −0.03 −0.03 (0.037) (0.034) (0.064) (0.063) Constant 0.071** 0.078*** 0.167*** 0.169*** (0.027) (0.028) (0.033) (0.033) Marketing restricted sample NO NO YES YES $$\hat{\rm R}$$2 0.014 0.034 0.036 0.058 N 10997 10997 4201 4201 Notes: The dependent variable is a dummy variable that takes the value of 1 if the person purchased insurance the following year. Columns 1 and 2 contain the sample of all buyers of insurance in 2005 and 2006 for whom insurance coverage was offered in their area the following year. Columns 3 and 4 contain the marketing restricted sample, which is restricted only to villages where at least one person purchased insurance in the following year. All regressions contain state fixed effects. Errors are clustered at the weather-station level. *** p < 0.01, **p < 0.05, *p < 0.1. Source: Authors’ analysis based on data from BASIX. View Large Next, I present the intravillage analysis. Here all buyers are aggregated to the village level and divided into two types: repeat buyers and new buyers, where repeat buyers are people who purchased insurance the year before. I then regress the number of each type of buyer on payout statistics and the total number of buyers in the previous year. When there was an insurance payout in the previous year, most of the repeat buyers received money from the insurance company while new buyers did not receive anything.14 If there are similar effects of payouts on people who had not purchased insurance the year before, this would be evidence that insurance payouts are generating trust and/or learning about insurance.15 These results are presented in table 7. Table 7. Village-Level Spillover Effects Indpendent variables for previous year Dependent variable is the number of buyers in a village the following year Panel A: All villages Panel B: Villages with at least 1 repeat buyer Total buyers Repeat buyers New buyers Total buyers Repeat buyers New buyers (1) (2) (3) (4) (5) (6) Was payout in village −0.334 −0.709** 0.376 3.353 −2.446*** 5.799* (2.252) (0.338) (2.121) (3.275) (0.433) (3.413) Mean ratio of payout to premium 1.319 0.838* 0.481 0.522 2.653*** −2.131* (1.330) (0.463) (1.050) (1.094) (0.168) (1.095) Mean payout ratio squared −0.135 −0.0729 −0.0617 −0.0832 −0.249*** 0.166 (0.141) (0.0539) (0.106) (0.120) (0.0167) (0.114) Number of buyers in village 0.131*** 0.0512*** 0.0795* 0.197*** 0.101** 0.0959 (0.0480) (0.0178) (0.0434) (0.0530) (0.0427) (0.0596) Year 2006 dummy −2.994* −0.256 −2.738* −5.231 0.0105 −5.241* (1.661) (0.236) (1.448) (3.284) (0.514) (2.929) Constant 3.445*** 0.144 3.301*** 10.48*** 0.483* 10.00*** (1.140) (0.136) (1.065) (1.599) (0.249) (1.614) Effect of payout at average payout 3.46 1.68 1.75 4.17 3.55 0.63 Observations 1534 1534 1534 459 459 459 R-squared 0.061 0.118 0.047 0.084 0.285 0.069 Indpendent variables for previous year Dependent variable is the number of buyers in a village the following year Panel A: All villages Panel B: Villages with at least 1 repeat buyer Total buyers Repeat buyers New buyers Total buyers Repeat buyers New buyers (1) (2) (3) (4) (5) (6) Was payout in village −0.334 −0.709** 0.376 3.353 −2.446*** 5.799* (2.252) (0.338) (2.121) (3.275) (0.433) (3.413) Mean ratio of payout to premium 1.319 0.838* 0.481 0.522 2.653*** −2.131* (1.330) (0.463) (1.050) (1.094) (0.168) (1.095) Mean payout ratio squared −0.135 −0.0729 −0.0617 −0.0832 −0.249*** 0.166 (0.141) (0.0539) (0.106) (0.120) (0.0167) (0.114) Number of buyers in village 0.131*** 0.0512*** 0.0795* 0.197*** 0.101** 0.0959 (0.0480) (0.0178) (0.0434) (0.0530) (0.0427) (0.0596) Year 2006 dummy −2.994* −0.256 −2.738* −5.231 0.0105 −5.241* (1.661) (0.236) (1.448) (3.284) (0.514) (2.929) Constant 3.445*** 0.144 3.301*** 10.48*** 0.483* 10.00*** (1.140) (0.136) (1.065) (1.599) (0.249) (1.614) Effect of payout at average payout 3.46 1.68 1.75 4.17 3.55 0.63 Observations 1534 1534 1534 459 459 459 R-squared 0.061 0.118 0.047 0.084 0.285 0.069 Notes: Observations are aggregated at the village level. The dependent variable is the number of insurance purchasers in a village. Panel A includes the sample of villages in 2005 and 2006 in which insurance coverage was offered in the village the following year. Panel B is restricted to villages in which at least on person purchased insurance the following year. The effect at average payout is the overall increase in purchasers estimated at the average payout level. In Panel A, the average ratio of payout to premium is 2.68; in Panel B it is 3.25. All regressions contain state fixed effects. Errors are clustered at the weather station level. *** p < 0.01, **p < 0.05, *p < 0.1. Source: Authors’ analysis based on data from BASIX. View Large Table 7. Village-Level Spillover Effects Indpendent variables for previous year Dependent variable is the number of buyers in a village the following year Panel A: All villages Panel B: Villages with at least 1 repeat buyer Total buyers Repeat buyers New buyers Total buyers Repeat buyers New buyers (1) (2) (3) (4) (5) (6) Was payout in village −0.334 −0.709** 0.376 3.353 −2.446*** 5.799* (2.252) (0.338) (2.121) (3.275) (0.433) (3.413) Mean ratio of payout to premium 1.319 0.838* 0.481 0.522 2.653*** −2.131* (1.330) (0.463) (1.050) (1.094) (0.168) (1.095) Mean payout ratio squared −0.135 −0.0729 −0.0617 −0.0832 −0.249*** 0.166 (0.141) (0.0539) (0.106) (0.120) (0.0167) (0.114) Number of buyers in village 0.131*** 0.0512*** 0.0795* 0.197*** 0.101** 0.0959 (0.0480) (0.0178) (0.0434) (0.0530) (0.0427) (0.0596) Year 2006 dummy −2.994* −0.256 −2.738* −5.231 0.0105 −5.241* (1.661) (0.236) (1.448) (3.284) (0.514) (2.929) Constant 3.445*** 0.144 3.301*** 10.48*** 0.483* 10.00*** (1.140) (0.136) (1.065) (1.599) (0.249) (1.614) Effect of payout at average payout 3.46 1.68 1.75 4.17 3.55 0.63 Observations 1534 1534 1534 459 459 459 R-squared 0.061 0.118 0.047 0.084 0.285 0.069 Indpendent variables for previous year Dependent variable is the number of buyers in a village the following year Panel A: All villages Panel B: Villages with at least 1 repeat buyer Total buyers Repeat buyers New buyers Total buyers Repeat buyers New buyers (1) (2) (3) (4) (5) (6) Was payout in village −0.334 −0.709** 0.376 3.353 −2.446*** 5.799* (2.252) (0.338) (2.121) (3.275) (0.433) (3.413) Mean ratio of payout to premium 1.319 0.838* 0.481 0.522 2.653*** −2.131* (1.330) (0.463) (1.050) (1.094) (0.168) (1.095) Mean payout ratio squared −0.135 −0.0729 −0.0617 −0.0832 −0.249*** 0.166 (0.141) (0.0539) (0.106) (0.120) (0.0167) (0.114) Number of buyers in village 0.131*** 0.0512*** 0.0795* 0.197*** 0.101** 0.0959 (0.0480) (0.0178) (0.0434) (0.0530) (0.0427) (0.0596) Year 2006 dummy −2.994* −0.256 −2.738* −5.231 0.0105 −5.241* (1.661) (0.236) (1.448) (3.284) (0.514) (2.929) Constant 3.445*** 0.144 3.301*** 10.48*** 0.483* 10.00*** (1.140) (0.136) (1.065) (1.599) (0.249) (1.614) Effect of payout at average payout 3.46 1.68 1.75 4.17 3.55 0.63 Observations 1534 1534 1534 459 459 459 R-squared 0.061 0.118 0.047 0.084 0.285 0.069 Notes: Observations are aggregated at the village level. The dependent variable is the number of insurance purchasers in a village. Panel A includes the sample of villages in 2005 and 2006 in which insurance coverage was offered in the village the following year. Panel B is restricted to villages in which at least on person purchased insurance the following year. The effect at average payout is the overall increase in purchasers estimated at the average payout level. In Panel A, the average ratio of payout to premium is 2.68; in Panel B it is 3.25. All regressions contain state fixed effects. Errors are clustered at the weather station level. *** p < 0.01, **p < 0.05, *p < 0.1. Source: Authors’ analysis based on data from BASIX. View Large In order to compare results with the main specification in table 3, I again provide a dummy for whether there was a payout in the village along with a quadratic effect of the ratio of payouts to the premium. When aggregating the village data, I use the mean of the payout ratios in the village to create a payout ratio for the village.16 The overall results of the table tell a consistent story: significantly sized payouts drive repeat buyers but not new purchasers, showing few spillover effects. Columns 1 and 2 shows how payouts affect the number of total and repeat buyers, respectively, the next year, and the results are very consistent with the baseline results from table 3. A dummy for whether there was any payout is negative, but the payout size has a positive effect. This suggests that low payouts have a marginally negative effect on the number of repeat purchasers, but this effect flips to positive as the size of the payout ratio increases above approximately 1. Column 3 shows the effect of payouts on new buyers in a village. Here, all the payout coefficients show a different pattern than for new buyers, but due to large standard errors I cannot reject that they are the same as the effects on repeat buyers. In panel B, I restrict the analysis to villages that had at least one buyer the year after insurance outcomes, creating a sample analogous to the “Marketing Restricted Sample” in table 3. The logic behind this is, if a village had zero buyers, it is likely that insurance was not marketed in the village that year, and therefore customers did not have an opportunity to purchase insurance. Restricting the data set in this way gives a much clearer pattern. Column 5 now shows much stronger effects of payouts on repeat buying, though the pattern is the same as in column 2. Small payouts have a negative effect, whereas increasing the payout ratio increases repeat buying. The squared term on the payout ratio is now negative and significant, indicating that high payout ratios have diminishing effects. The coefficients for new buyers in column 6 are now all significantly different from the coefficients for repeat buyers. In fact, the coefficients in column 5 flip signs, suggesting that payouts have the opposite effect on people who did not receive payouts. These results suggest that low payouts actually induce more new buyers, but that these effects decrease and then turn negative as the payout in the village increases. The average mean village payout in this sample (for villages that received any payout) is 3.25, and the coefficients suggest that this level of payout will have roughly no effect on new purchasers compared to a village that did not experience a payout. The opposite effects of spillovers versus direct effects are somewhat perplexing, and they are not consistent with the hypothesis that receiving payout increases trust and learning in the village where it occurs.17 One important clarification of these results is that most potential buyers living in a village that had experienced payouts would have also experienced uninsured rainfall shocks during the same season. Therefore, it may be possible that there are effects of trust and learning, but they are outweighed by opposite effects of the weather. As shown in the previous section, rainfall shocks tend to have a negative effect on insurance demand, so the (lack of) evidence of spillovers may be a result of a more complex interaction between trust/learning and direct effects of weather. Overall, these results do not support the hypothesis that trust, learning, or any other effects of simply witnessing insurance payouts are driving increased purchasing. Although it is possible that the measurements of spillovers are too crude and miss more subtle effects, the data simply does not provide evidence that there are strong spillover effects. The lack of visible spillover effects casts doubt on the theory that repurchases are being driven by increased trust in the insurance company or learning about insurance payouts. Notably, this result stands in contrast to recent results that do document spillover effects of insurance payouts (Cole, Stein, and Tobacman 2014; Karlan et al. 2014). Direct Effects of Payouts on Wealth and Liquidity The previous two sections discount the possibility that trust, learning, or weather effects are driving the result that receiving an insurance payout is correlated with purchasing insurance the following year. This points to the actual reception of money from the insurance company as being the driving force behind greater take-up. The most natural explanation for this phenomenon would be that receiving an insurance payout could directly affect choices the next year due to its effects on wealth and liquidity. For instance, if insurance is a normal good, then increased wealth would result in greater insurance demand.18 Although the BASIX data set does not offer the opportunity to test the direct effects of a cash payment separately from an insurance payout, there are a number of reasons why it is unlikely that wealth or liquidity effects are driving the results. Most important, insurance payouts are given in the context of a rainfall shock, which would most likely result in a loss of income. It may help to recall that the empirical results are being driven by variation in rainfall across locations, not by levels of insurance within a village. Therefore, for wealth effects to be driving the results, one would need to think that experiencing an insurance payout in the context of a rainfall shock resulted in people becoming wealthier than those people who didn’t experience a shock at all. Given the fact that most buyers bought a relatively low amount of insurance coverage relative to their incomes, experiencing a rainfall shock, even when insured, would likely decrease future wealth. Therefore, wealth effects seem like a poor explanation as to why receiving payouts spurs future insurance sales. If people who received insurance payouts had a decrease in wealth, it is also unlikely that receiving the insurance payout would increase their liquidity the next season. Insurance payments were generally made in January, but people had the opportunity to purchase insurance for the next season only in May. It is doubtful that these payments would have a lasting enough liquidity effect to influence insurance buying decisions five months later. Although I cannot provide direct empirical evidence against the hypothesis that insurance payments drive increased take-up due to wealth or liquidity effects, given the structure and timing of insurance payments this explanation seems extremely unlikely. Omitted Marketing Intensity With the available data, it is not possible to observe the level of marketing that each person received, making “marketing intensity” an important omitted variable. As mentioned earlier, when BASIX markets rainfall insurance, it first holds a village-level meeting and then follows up with door-to-door visits. Unfortunately, there is no data on the specific marketing practices in each village. If the intensity of marketing was correlated with both previous years’ insurance payouts and current years’ sales, this omitted variable could be biasing the results. For instance, assume that the marketing staff at BASIX think that people who have just received a payout are more likely to repurchase insurance. In this case, as the marketing team has limited resources, it may make sense for them to direct these resources toward the area of highest return, which would be people who have already received payouts. If this was the case, the increased take-up rates from people who received payouts could simply result from increased marketing attention from the BASIX team. Although the results could be picking up some of this effect, there are a couple of reasons I believe it is unlikely to be a significant factor. First, regressions of observable marketing factors (such as a dummy of whether there were any purchasers in the village) do not show any significant correlations with payouts.19 Next, the BASIX marketing staff claim to not give any special marketing treatment to previous payout recipients.20 As they are trying to build long-term business, BASIX claims that they do not change their marketing practices for villages that have recently received a payout. Finally, if BASIX targeted payout recipients and they did not really have a higher tendency to purchase, one would think that the marketing team would quickly learn that this strategy was not effective and would stop it. While I only observed two marketing cycles and erroneous beliefs could survive throughout this short time span, it is telling that the effect of payouts on take-up is greater in 2006 than 2005, suggesting that the effect is increasing over time.21 If it was caused by erroneous expectations of the marketing team, I would expect the effect to decrease over time. Overall, while I must accept the possibility that increased marketing is driving the results, I regard it as unlikely. IV. Discussion The previous section rejects a number of theories about why recent insurance payouts could be driving insurance purchases. The data does not show support for a number of explanations (learning, trust, wealth, liquidity) that would be consistent with a neoclassical framework. However, there are a number of behavioral explanations that are consistent with the results. Although is has been established that salience or recency bias related to the weather is unlikely to be driving the results, it is possible that recency bias unrelated to the weather is playing a role. People may simply believe that because the insurance paid out in the previous year, it is more likely to pay out in the following year. This may be especially important if people consider insurance as an investment, as in Karni and Safra (1987). One might think that this type of bias would have spillover effects (which I do not observe), but may also be present at the individual level if people believe recent insurance payouts are a sign of individual luck. Another explanation could be that insurance purchasers behave as if they are “gambling with house money,” as in Thaler and Johnson (1990). If customers exhibit loss aversion, they may not view insurance premiums paid after receiving payouts as true losses, since they are still “in the red” in their relationship with the insurance company. If the assumption is that reference points adjust after receiving insurance payouts, the observed behavior is consistent with this theory. This explanation is bolstered by the fact that the data show negative effects of small payouts, with the effect of payouts becoming positive only when the payout is larger than the premium. If payouts counteract loss aversion, then one would only expect an effect when payouts brought customers into the gain domain, which only happens when the payouts are greater than the premium. A related explanation comes from an observational study on mutual insurance among fishermen in the Côte d’Ivoire by Platteau (1997). Platteau observes malfunctioning mutual insurance cooperatives and theorizes that they are failing because members view insurance as a system of balanced reciprocity, meaning that they expect to break even over the lifetime of the scheme. When members have not received the services (in this case sea rescue) of the mutual in a long time, they start to view the insurance as a bad deal and ask for their contributions back. The results in this paper are consistent with people viewing insurance as a system of balanced reciprocity, as they could see insurance purchases after receiving a payout as giving back to a system that has helped them previously. The data used in this study do not allow me to distinguish between these competing explanations. They remain as areas for future research. V. Conclusion After receiving an insurance payout, customers of rainfall insurance in India are 9 to 22 percent more likely to purchase insurance again the next year. This behavior seems to be driven by actually receiving the money from the insurance company, as the data do not support alternative theories of repurchase being driven by direct weather effects, trust, learning, wealth, or liquidity. This study brings to light a number of questions that would are ripe for future research. First of all, do insurance payouts have long-term effects on future purchases? And do payouts continue to have similar effects for people who have years of experience with insurance. To answer these questions, one would need a data set with a longer time frame. Also, a longer data set could shed further light on the question of whether customers learn about insurance over time. It is possible that people need a few years of experience with insurance to really learn about the product and gain trust in it, which would explain why this paper fails to see any spillover effects. These results point to a number of policy recommendations for the Indian rainfall insurance market and possibly for insurance markets in general. One of the main arguments made for the slow adoption of insurance in India is that people do not understand insurance and do not trust the insurance companies. If trust and learning were the crucial determinant of insurance adoption, then incentives could be given to encourage early adoption, and over time as people witnessed and experienced payouts, one would expect insurance adoption to grow. This paper fails to find any evidence of increased trust and learning driving insurance decisions, which suggests that incentivizing early adopting is unlikely to quickly spur insurance take-up. Historical evidence (as in Kunreuther, Sanderson, and Vetschera 1985) has suggested that an effective policy to spur insurance markets would be to target areas that have recently experienced a large shock. This paper does not support this notion in the case of rainfall index insurance in India, as places that recently experienced a shock were less likely to purchase insurance. With relation to the future of rainfall index insurance in India, one stark result is that the raw numbers of continuing customers of insurance are very low, calling into question the sustainability of the product. Even among people who received payouts in excess of twice their premium in 2006, only 18 percent bought again in 2007. With the proportion of repeat buyers so low, one would have to assume that many people are not satisfied with their experience of insurance, which suggests that the product or marketplace will need to evolve in order to survive. One factor to note is that this study looks at the early years of the first major scale-up of rainfall insurance in the world. Rainfall insurance is still a young product, and it is still evolving to meet the needs of customers. One particular point of attention is the massive loading on most policies offered. As shown in table 1, BASIX insurance policies had premiums of over twice the actuarially fair rate. With premiums this high, it is unsurprising that people are not signing up. Also, one may argue that the correlation between insurance payouts and crop outcomes were less than ideal in these early products. Around the world, index insurance policies are constantly evolving to better correlate with crop outcomes and avoid basis risk. While this study predicts that rainfall insurance in the form of BASIX’s policies for the period 2005–2007 are likely to fail, it is quite possible that innovations in products and pricing can create an insurance product that better meets the needs of small-scale farmers. Daniel Stein is a senior economist at IDinsight; his email address is Daniel.stein@idinsight.org. This work would not have been possible without the assistance of the insurance staff at BASIX, especially Sridhar Reddy, for help in obtaining and understanding their customer database. Thanks to Tim Besley, Greg Fischer, Gani Aldashev, and Xavier Giné for giving helpful advice and commenting on earlier drafts. Thanks to participants at the Munich Re 6th Annual Microinsurance Conference, the Midwestern International Economic Development Conference, seminars at the London School of Economics, University of Namur, and KU Leuven for helpful comments. All remaining errors are my own. A supplemental appendix to this article is available at https://academic.oup.com/wber. Footnotes 1 CIA World Factbook: India; Indiastat.com. 2 Cole, Stein, and Tobacman (2014) find among insurance purchasers, a payout of Rs 1000 per policy corresponds to an increase in the chance of purchasing insurance the following year by .864. The average payout per policy is 182, so the effect at the average payout (which is comparable to our setup) is .157. 3 The states are, in descending order of number of buyers, Andhra Pradesh, Maharashtra, Jharkand, Karnataka, Madhya Pradesh, and Orissa. 4 Note that BASIX also sold many policies in the district of Deogarh in Jarkhand, and those buyers are omitted from this analysis. The reason for this is that the policy for Deogarh is heavily subsidized, resulting in a policy that is completely different from all the others. For instance, the Deogarh policy for 2005 has an expected payout of Rs 1,140 compared to an average of Rs 149, although the policy does not cost more than average. Because of its incredibly generous terms, the Deogarh policy has huge payouts for all years of the study; therefore, it does not seem to be “normal” enough to warrant inclusion in the main data set. All the analysis presented excludes all buyers in Deogarh, though most results do not change substantially when it is included. 5 APHRODITE’s water resources project; http://www.chikyu.ac.jp/precip. 6 25° latitude equals about 27.5 kilometers. .25° longitude varies by latitude; over the range of latitudes in this survey it equals roughly 26 kilometers. 7 It is not clear if ICICI-LOMBARD used APHRODITE or another source of historical rainfall data in order to price their policies because this information is proprietary. The APHRODITE data set is based on historical data provided primarily by the Indian Meteorological Department (IMD), but similar gridded data sets provided by IMD contain vast amounts of missing data over both time and space. It is very likely that the ICICI-LOMBARD products were priced using IMD data, but it is likely the ICICI-LOMBARD used different techniques to account for missing data. 8 It makes sense to assume that the error $${{\epsilon}}_{t,i}$$ is correlated for the same person across time as well as across people in a given year. Ideally, one would like to include individual fixed effects to account for individual heterogeneity. However, in order to exploit this variation one would need to look at customers who purchased insurance in both 2005 and 2006, and received payouts in only one of those years. Unfortunately, because of the very low repurchase rate, this results in very little variation and is therefore an unsuitable method of analysis. 9 BASIX’s insurance coverage area varied somewhat from year to year. Results do not change significantly if all areas are included in the regression (see table S.3 in the supplemental appendix). 10 Most people that received an insurance payout also experienced a rainfall shock. Therefore, this coefficient should be interpreted as the effect of receiving a payout in the context of a rainfall shock. 11 The APHRODITE weather data provides information about how many local weather stations contributed to a certain rainfall reading. Since some of the rainfall observations are likely to be more accurate than others, I weight them according to accuracy. If there are no rainfall stations contributing to the APHRODITE data in 2004 within a .75°×.75° grid around the desired BASIX weather station, the observation is given a weight of 1. If there is a least one weather station in this .75°×.75° grid, the observation is given a weight of 1.5. If there is a rainfall station within the .25°×.25°grid, the observation is given a weight of 2. The weighted results do not differ significantly from the unweighted results. 12 Note that while it is reasonable to think that village-specific characteristics (such as village size) may have an effect on village-level insurance take-up, village-level covariates are not included in the regression. When the regressions are run with the village characteristics from the 2005 Indian census, the coefficients of interest do not change significantly. Also, most village-level characteristics had insignificant coefficients, with the exception that a more literate population was correlated with higher take-up. Since village-level coefficients were only available for around 50 percent of the villages, these variables are not included in the main specifications. 13 If people can only gain trust and learning by actually receiving a payout themselves, then the data gives us no way to separate trust and learning from other possible mechanisms of receiving a payout. 14 Some buyers may not have received money if they bought one phase of the insurance policy but one of the other phases paid out. This is the variation exploited in table 6, and it happened in 427 cases. Removing these individuals from table 7 does not change the results. 15 Certainly, purchasers and nonpurchasers represent different populations, so one would expect that payouts would affect these populations differently because of selection. However, it is reasonable to assume that the general pattern of how payouts affect future purchasing would be consistent if trust and/or learning are playing an important role. 16 The results are not sensitive to using the mean, and they are very similar using the median, maximum, and mode. 17 It is certainly true that insurance purchasers and nonpurchasers are selected samples that may not be directly comparable. For instance, nonpurchasers may be less sensitive to learning offered by insurance payouts because they may be relatively uninterested in insurance. Although this would certainly attenuate any learning effect, I would still expect large insurance payouts to increase new purchasers if trust and/or learning were playing a large role. 18 This is consistent with the empirical findings of Cole et al. (2013). 19 This analysis is available in the supplemental appendix, table S.5. There is no correlation between payouts and a dummy for our “Marketing Restricted Sample.” However, BASIX is more likely to offer insurance in areas where there were insurance payouts in the previous year. Since all of this paper’s analysis is restricted to this sample where insurance was available the following year, the estimates are internally valid. However, the fact that some purchasers are omitted from the analysis likely introduces some bias in extrapolating the results to the full population. 20 Conversation with Sridhar Reddy, assistant manager for insurance at BASIX, January 2009. 21 Results not shown. References Arellano M. , Bond S. . 1991 . “ Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations .” The Review of Economic Studies 58 ( 2 ): 277 – 97 . Google Scholar Crossref Search ADS Binswanger-Mkhize H. P. 2012 . “ Is There Too Much Hype About Index-Based Agricultural Insurance? ” The Journal of Development Studies 48 ( 2 ): 187 – 200 . Google Scholar Crossref Search ADS Bryan G. 2010 . “Ambiguity and Insurance.” Cai J. , Song C . 2012 . “Insurance Take-up in Rural China: Learning from Hypothetical Experience.” Available at SSRN 2161649. Cole S. , Giné X. , Tobacman J. , Topalova P. , Townsend R. , Vickery J. . 2013 . “ Barriers to Household Risk Management: Evidence from India .” American Economic Journal: Applied Economics 5 ( 1 ): 104 – 35 . Google Scholar Crossref Search ADS PubMed Cole S. , Stein D. , Tobacman J. . 2014 . “ Dynamics of Demand for Index Insurance: Evidence from a Long-Run Field Experiment .” The American Economic Review 104 ( 5 ): 284 – 90 . Google Scholar Crossref Search ADS Collier B. , Skees J. , Barnett B. . 2009 . “ Weather Index Insurance and Climate Change: Opportunities and Challenges in Lower Income Countries .” The Geneva Papers on Risk and Insurance-Issues and Practice 34 ( 3 ): 401 – 24 . Google Scholar Crossref Search ADS Eling M. , Pradhan S. , Schmit J. T. . 2014 . “ The Determinants of Microinsurance Demand .” The Geneva Papers on Risk and Insurance-Issues and Practice 39 ( 2 ): 224 – 63 . Google Scholar Crossref Search ADS Gallagher J. 2014 . “ Learning About an Infrequent Event: Evidence from Flood Insurance Take-up in the United States .” American Economic Journal: Applied Economics 6 ( 3 ): 206 – 33 . Google Scholar Crossref Search ADS Giné X. , Townsend R. , Vickery J. . 2008 . “ Patterns of Rainfall Insurance Participation in Rural India .” World Bank Economic Review 22 ( 3 ): 539 – 66 . Google Scholar Crossref Search ADS Hess Ulrich. 2004 . “Innovative Financial Services for India, Monsoon Indexed Lending and Insurance for Smallholders.” ARD Working Paper 09. World Bank. Hill R. V. , Robles M. . 2010 . “Flexible Insurance for Heterogeneous Farmers: Results from a Small Scale Pilot in Ethiopia.” International Food Policy Research Institute Working Paper. Karlan D. , Osei R. , Osei-Akoto I. , Udry C. . 2014 . “ Agricultural Decisions after Relaxing Credit and Risk Constraints .” The Quarterly Journal of Economics 129 ( 2 ): 597 – 652 . Google Scholar Crossref Search ADS Karni E. , Safra Z. . 1987 . “‘ Preference Reversal’ and the Observability of Preferences by Experimental Methods .” Econometrica 55 ( 3 ): 675 – 85 . Google Scholar Crossref Search ADS Kunreuther H. , Sanderson W. , Vetschera R. . 1985 . “ A Behavioral Model of the Adoption of Protective Activities .” Journal of Economic Behavior and Organization 6 ( 1 ): 1 – 15 . Google Scholar Crossref Search ADS Mobarak A. M. , Rosenzweig M. R. . 2013 . “ Informal Risk Sharing, Index Insurance, and Risk Taking in Developing Countries .” American Economic Review 103 ( 3 ): 375 – 80 . Google Scholar Crossref Search ADS Platteau J.-P. 1997 . “ Mutual Insurance as an Elusive Concept in Traditional Rural Communities .” Journal of Development Studies 33 ( 6 ): 764 – 96 . Google Scholar Crossref Search ADS Thaler R. H. , Johnson E. J. . 1990 . “ Gambling with the House Money and Trying to Break Even: The Effects of Prior Outcomes on Risky Choice .” Management Science 36 ( 6 ): 643 – 60 . Google Scholar Crossref Search ADS Townsend R. M. 1994 . “ Risk and Insurance in Village India .” Econometrica 62 ( 3 ): 539 – 91 . Google Scholar Crossref Search ADS Turner G. , Said F. , Afzal U. . 2014 . “ Microinsurance Demand After a Rare Flood Event: Evidence from a Field Experiment in Pakistan .” The Geneva Papers on Risk and Insurance-Issues and Practice 39 ( 2 ): 201 – 2 . Google Scholar Crossref Search ADS © The Author(s) 2016. Published by Oxford University Press on behalf of the International Bank for Reconstruction and Development/The World Bank. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Dynamics of Demand for Rainfall Index Insurance: Evidence from a Commercial Product in India
JF - The World Bank Economic Review
DO - 10.1093/wber/lhw045
DA - 2018-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/dynamics-of-demand-for-rainfall-index-insurance-evidence-from-a-WNXFrPHFBo
SP - 692
VL - 32
IS - 3
DP - DeepDyve
ER -