Seasonal liquidity, rural labor markets and agricultural production: Evidence from Zambia

Transcription

1 Seasonal liquidity, rural labor markets and agricultural production: Evidence from Zambia Günther Fink Harvard T.H. Chan School of Public Health B. Kelsey Jack Tufts University Felix Masiye University of Zambia preliminary and incomplete: do not circulate, do not cite Abstract Small-scale farming remains the primary source of income for a majority of the population in developing countries. In rainfed agricultural settings, income typically arrives only once a year. To meet consumption and investment needs during the rest of the year in the absence of formal financial markets, households adopt a range of strategies for financing consumption. Family labor sold in local labor markets offers an immediately available source of liquidity to households, but may lower harvest output if increased off-farm labor supply is associated with a decrease in the quantity or quality of labor inputs on the family farm. In a two-year trial, we provide small scale farmers in randomly selected villages in rural Zambia with access to loans during the lean season to identify the causal impact of seasonal liquidity constraints on rural labor markets and agricultural output. We find that small unrestricted loans reduced the time household members spent working off-farm by over 40 percent, increased total labor invested on the family farm by 4 percent, and increased agricultural output by 5 percent. We also find substantial increases in local wages in treatment villages in response to the observed reductions in off-farm labor supply. Our findings suggest an additional effect of local credit markets on agricultural output, through the allocation of household labor, with implications for the distribution of incomes within rural farming communities. We thank audience members at Boston University, Georgetown University, IFPRI, Northeastern University, University of Maryland, University of Massachusetts - Amherst, University of Zambia and IGC Growth Week for comments and suggestions. We are grateful to the Growth and Labor Markets in Low Income Countries (GLM-LIC), the International Growth Centre, the Agricultural Technology Adoption Initiative (JPAL/CEGA) and an anonymous donor for financial support, and to Innovations for Poverty Action for logistical support. Many thanks to Rachel Levenson for careful oversight of the field work and to Daniel Velez Lopez, Chantelle Boudreaux and Carlos Riumallo Herl for assistance with the data. 1

2 1 Introduction In Zambia, like in much of Sub-Saharan Africa, agriculture employs the vast majority of the rural population, in spite of low levels of productivity and farming income. 1 A lack of irrigation infrastructure combined with a long dry season means that harvest income arrives only once per year, and must cover household needs for the subsequent months. Scarcity in local markets is synchronized by rainfall patterns, leading to a pronounced hungry season in the months leading up to the harvest. With limited access to formal capital markets, households frequently turn to alternative strategies for smoothing income and consumption, including livestock and asset sales (Rosenzweig and Wolpin 1993; Janzen and Carter 2013,), migration (Halliday 2012; Bryan et al. 2014), or restricting food intake (Kazianga and Udry 2006; Kaminski et al. 2014). 2 One common strategy to address cash needs in agricultural settings is the sale of family labor in local labor markets (Kochar 1995, 1997; Jayachandran 2006; Ito and Kurosaki 2009). Given that households can choose among financing mechanisms, income-maximizing households will only sell family labor if the expected loss in future harvest income is smaller than the costs of alternative financing strategies. Higher formal or informal interest rates on borrowing increase reliance on family labor sales to finance consumption, and, all else equal, lower both capital and labor investment on the family farm. On-farm labor inputs may be further reduced if households are exposed to unanticipated negative income or expenditure shocks that cannot immediately be smoothed through other financing mechanisms and may result in households accepting wages below the (discounted) marginal product of labor on-farm (Kochar 1995, 1999; Rose 2001; Ito and Kurosaki 2009). 3 Anticipating these constraints, utility-maximizing households will alter both the total land under production and the crop mix relative to an unconstrained environment (Fafchamps 1993; Rosenzweig and Binswanger 1993). To investigate the extent to which local credit markets shape consumption and labor allocation during the hungry season, as well as the agricultural output implications of these decisions, we conducted a two year cluster-randomized experiment with 3,140 small-scale farmers from 175 villages in rural Zambia. 4 Zambia s agricultural cycle is centered around the rainy season from November to April. Harvest takes place in May and June, and comprises the primary income for the calendar year. As illustrated in Figure 1, household (food) reserves gradually decline between July and December, and are most scarce from January to April. Early crops start to become available for consumption in April. The January to March period is what is referred to as the hungry season by 1 A recent study in the region of proposed research shows average gross production value of USD 500 for a family of six (Fink and Masiye 2012). Once capital input, land and labor costs are considered, net profits for many of these families may be negative. 2 For summaries of the consumption smoothing literature, see Morduch (1995) and Besley (1995). 3 Typical income shocks in the study area include the loss of stored food reserves due to pests or theft; expenditure shocks include funerals and medical costs. 4 A pilot for the current study is described in Fink et al. (2014), which uses a slightly different intervention design and is not powered to detect effects on agricultural output. 2

3 farmers, and is the period where cash-on-hand liquidity constraints are most binding. We directly targeted this period with two related interventions. In treatment group 1 (59 villages), households could borrow 200 Kwacha (approximately USD 33) of cash in January. 5 In the second group (58 villages), farmers could borrow three bags of maize over the same period. The three bags of maize were roughly equivalent in terms of financial value to the cash loan, and theoretically provide a sufficient amount of calories to feed a family of five for at least two months. 6 In both loan groups, repayment was due after the harvest in late June to early July. In year 1, all households were given the option to repay in cash (260 Kwacha) or in kind (4 bags of maize). 7 The remaining third of the villages were assigned to a control group. 8 To assess the impact of access to lower interest borrowing on labor allocation, consumption and agricultural output, we develop a series of predictions through a simple multi-period agricultural production model, and test them empirically using our experimental data. The main focus of our stylized model is the interaction between local credit and agricultural labor markets. The main intuition of the model is relatively straightforward: farmers can finance consumption during the hungry season either through borrowing at local interest rates or through family labor sales off-farm, which borrow against future harvest income. An increase in the local interest rate mechanically leads to a shift in labor allocation from on-farm to off-farm labor (more labor sold). As long as some farmers remain unconstrained, reductions in local interest rates decrease aggregate labor supply and increase local wages. We begin our empirical analysis by showing that the interest rates offered in the experiment were below locally available alternatives: both the demand for and the willingness to repay the loans was high in both years. Around 98 percent of eligible households took up the loans and close to 90 percent repaid in both treatment arms. We then document the impact of the loans on labor allocation. Both the extensive and intensive margin of family labor sales falls, with a reduction of 43 percent in the hours of labor sold per week to other farms in treated villages. We see little impact on hiring, but an overall increase in on-farm labor inputs. The reallocation of labor also directly affects local labor market clearing wages, which increase by 17 percent in treated villages. Consumption also increases during the hungry season, and the combined effect of more hours of labor investment and potentially higher quality effort leads to a 5 percent increase in agricultural revenue at the end of the year. This increase in agricultural output corresponds to around 32 USD, an amount that is not sufficient to repay the full amount of the loan. The benefits of the loans extend beyond 5 We report values in Kwacha in the paper and provide USD equivalents based on an exchange rate of 6 Kwacha per USD, which reflects the average exchange rate over the study period. 6 One kilogram of maize provides approximately 3600 kcal. Three bags would thus provide 5 household members about 1800 calories per day over a 60 day period. 7 Official rates set one 50 kg bag of maize at 65 Kwacha. However, local seasonal fluctuations in maize prices affect the relative value of the two loan offers. In Section 4, we calibrate the interest rate and value of each loan. 8 In a small sub-sample of these control villages, households were given a gift of 60 Kwacha to directly measure potential positive income-effects the loan may generate. 3

4 consumption and agricultural output: we also document improved subjective well-being, reduced morbidity and lower exposure to high interest informal borrowing. Overall, the net welfare benefits of participating in the program appear sizable as evidenced by almost universal re-enrollment in the loan program in the second year of the study. Most of the positive treatment effect on agricultural output can be attributed to increased labor investment on-farm, since we rule out increases in other inputs, both empirically and through program design. The magnitudes appear plausible: in year 1, the estimated increases in total labor invested on farm and in agricultural revenue imply a marginal product of labor of approximately US2.95 which is very close to casual labor wage rate (USD 2.58) observed during the hungry season. The positive impacts that we document, together with high rates of take up and repayment, raise the question of why similar hungry season loans have not arisen outside of the study. We document and discuss two primary reasons: high transaction costs associated with rural lending that undermine profitability and the dependence of impacts on rainfall outcomes. The positive average loan impacts observed mask a substantial amount of heterogeneity, both across years (aggregate productivity shocks) and seasons, and across households (idiosyncratic income shocks). The first year of the study (the 2013/2014 agricultural cycle) enjoyed considerably better rainfall than the second year of the study (2014/2015), resulting in average agricultural output in the control group that was around 12 percent lower in year two. Our stylized model of labor allocation predicts that aggregate productivity shocks compress wages but have ambiguous effects on local labor allocation. The empirical patterns observed are largely consistent with these predictions. While off-farm labor supply falls in both years, we find that total labor inputs do not respond much to treatment in the second year, i.e. households do not reallocate family labor toward their own farm. Correspondingly, the average effect on agricultural output is driven entirely by impacts in year 1; the estimated impact on agricultural output in year 2 is not significantly different from zero. Second, we observe differences in impacts across the agricultural calendar. Consistent with our conceptual framework, the largest effects on labor allocation and consumption were observed during the hungry season, when resources are most constrained. Third, we observe differences in treatment effects based on household initial endowments and the realization of idiosyncratic income shocks. As predicted, the loan programs has larger impacts on less well endowed (in savings and labor) farms. Finally, we observe few differences in outcomes between the maize and cash loan treatment arms. 9 However, implementation costs varied substantially by lending modality. In Year 2 of the study, we introduced additional randomized sub-treatments to help isolate mechanisms. First, the loans were phased out inroughly half of the 57 villages eligible in year 1, and introduced in 20 of the former control villages. This treatment rotation was designed to identify persistent effects of the program as well as year-specific treatment effects, which allow us to identify 9 We find no evidence that the cash loan induced more wasteful consumption or lowered repayment rates. In general, observed consumption levels of non-essential goods such as alcohol and cigarettes were very low in the sample, with no evidence of additional liquidity changing patterns. 4

5 impact heterogeneity across years and to improve generalizability (Rosenzweig and Udry 2016). As mentioned above, treatment effects for agricultural output varied considerably across years, with relatively sizable impacts in Year 1 and mostly zero impacts in Year 2; differences on other margins of adjustment were much smaller across the two years. Second, to assess the importance of ex ante adjustments in the production plan, half of the treated villages in year 2 learned about the loan at the start of the planting season in September, while the other farms were informed about programs in January. While the effects of early notification are suggestive of additional gains from ex ante adjustment, overall treatment effects on agricultural output were small in the second year, limiting power to precisely estimate these differences. Recent evidence on the impacts of capital access on agricultural productivity is mixed. 10 Ghana, Karlan et al. (2014) find no evidence that liquidity constraints impede agricultural investments. On the other hand, Beaman et al. (2014) find that relaxing credit constraints through grants increases agricultural investment and yields among rice farmers in Mali, but that the same is not true for loans. Both studies focus on capital for farming inputs (seeds, fertilizer or pesticides) as the primary mechanism through which credit impacts yields. The results presented in this study suggest that a lack of access to financial resources also may affect labor inputs, as as well as local labor market equilibria (similar to Jayachandran 2006; Mobarak and Rosenzweig 2014). Our results are also linked to the literature that documents the role of family labor sales for consumption smoothing in response to unanticipated liquidity needs and poor credit access (Kochar 1995, 1999; Rose 2001; Ito and Kurosaki 2009). 11 Our findings suggest that these patterns emerge more broadly and independent of unanticipated shocks in settings were incomes are seasonal and households ability to transfer resources across seasons is limited. In addition, we generate novel evidence that selling family labor to smooth consumption affects agricultural output, an effect that may be partially driven by changes in labor quality due to changes in consumption. 12 A related, largely qualitative literature suggests that the local labor markets we study are often associated with deviations from household income maximization (Kerr 2005; Bryceson 2006; Orr et al. 2009; Michaelowa et al. 2010; Cole and Hoon 2013). The results in this study are also closely linked to a growing literature investigating the impact of seasonal transfer or loan programs. Burke (2014) offered farmers in Kenya a loan product that allowed them to exploit seasonal variation in maize prices and finds significant effects on total maize 10 An earlier literature uses observables to define whether households are credit constrained, and compares productivity and consumption across constrained and unconstrained households (e.g. Feder et al. 1990). 11 The general relationship between income seasonality and consumption smoothing is not well established in the literature. While some studies suggest that precautionary savings are sufficient to smooth consumption even if income is highly seasonal (Paxson 1993; Chaudhuri and Paxson 2002; Jacoby and Skoufias 1998), others have highlighted the pronounced consumption differences over the year (Dercon and Krishnan 2000; Khandker 2012). Kaminski et al. (2014) point to recent evidence from Africa that seasonal consumption patterns are closely linked to seasonality in staple crop prices, which suggests that households are not able to adequately smooth consumption. 12 The relationship between nutrition and productivity has been debated extensively in the literature (see Pitt and Rosenzweig 1986; Strauss 1986; Behrman et al. 1997; Schofield 2013). In 5

6 revenues and household expenditures. Bryan, Chowdhury and Mobarak (2014) show that providing credit and grants leads to large increases in seasonal labor migration in Bangladesh, arguing that credit market failures and highly uncertain returns likely keep long-distance labor supply below optimal levels. Most similar to our study, Basu and Wong (2015) evaluate a seasonal food credit and improved storage program in Indonesia; similar to the results presented here, they find that food loans increase non-staple food consumption during the hungry season and income from crop sales at harvest, but do not analyze impacts on labor allocation or yields. Our findings contribute to that literature by providing the first direct evidence that capital market interventions timed to coincide with the hungry season not only affect consumption, but also change local labor allocation and increase agricultural output. These types of programs may not, however, be viable unless coupled with interventions that lower transaction costs, such as mobile-based savings and borrowing or piggybacking on existing networks through contract farming companies or other institutions. Other strategies for decreasing the cost of consumption smoothing, such as more secure savings, may also decrease reliance on family labor for consumption smoothing, and improve agricultural production. The paper proceeds as follows. In the next section, we present a simple model that highlights the linkages between capital markets, labor allocation and agricultural output, and generates testable predictions for our empirical analysis. Section 3 describes the study context, and relevant background on local credit and labor markets. Section 4 explains the experimental design, data collection and identification strategy. Section 5 presents the main results, following the predictions of the conceptual model. Section 6 shows heterogeneity, other margins of adjustment and several robustness checks. Section 7 discusses cost effectiveness. Finally, Section 8 concludes. 2 Conceptual Framework Consider a simple model of agricultural production, where rational farming households maximize utility over the agricultural cycle. Households start off with an endowment consisting of previous assets and their most recent harvest at the beginning of each cycle, and maximize utility by optimally allocating labor and financial resources across periods. The agricultural cycle is divided into three periods: period 1, the post-harvest season (August to September in the Zambian context), where farm activities are limited due to lack of rainfall; period 2, when fields need to be prepared and crops need to be planted (October - December); and period 3, where fields need to be weeded and maintained (January - April). Harvest begins at the end of period 3. Forward-looking farmers maximize the following utility function: U(c) =u(c 1 )+ u(c 2 )+ 2 u(c 3 )+ 3 V (Y,B) (1) where u(.) is a generic concave utility function, <1 is the subjective discount factor, and V (.) is the indirect utility derived from the final harvest value Y net of borrowed resources B. We measure 6

7 all inputs and outputs in monetary units and normalize all prices to one. 13 Farms have an initial capital endowment A 0, which comprises previous assets and savings as well as net harvest outcomes from the previous cycle. Farms can earn a return r s on savings, and can borrow locally at a rate r b r s In period 1, farms can consume or save. In periods 2 and 3, farms can consume, save or invest in their fields. Investment and consumption in periods 2 and 3 (I 2,I 3 ) can be financed through savings or through borrowing, which needs to be fully repaid at the end of the season. In addition, farms can generate short-term income by selling labor on the local labor market at a period-specific wage rate w t. For simplicity, we normalize total labor available to each farm to 1, and assume that labor is fully employed during the main farming periods 2 and 3. Given this, the consumption constraints in periods t =1, 2, 3 are given by c t + I t + S t apple S t 1 r s + B t + w t (1 L t )+ t (2) where 0 apple L t apple 1 is the share of the farm s labor used on the farm s fields, 1 L t is the total amount of family labor sold at the local wage rate w t,and t is a period specific income shock. The total debt payments due at the end of the season are given by B = B 1 r b 3 +B 2 r b 2 +B 3 r b, and net harvest income is given by Y ( 2, 3,I 2,I 3, L 2,L 3 ) B, where t are period-specific aggregate productivity shock, which drive the marginal product of investment and labor, such 0 (I)/@ > 0,Y 0 (L)/@ > 0. In each period, farms decide on consumption and borrowing. In periods 2 and 3, farms also need to decide on labor and capital investments. Given that labor is traded freely on the labor market, farms will choose a labor allocation such that the marginal product of labor on the farm equals the period-specific wage times the marginal product of capital investment, i.e. E[Y 0 (L t )] = w t E[Y 0 (I t )], (3) where L t is the total labor input, and L t L t is the net labor hired. The local wage w t is determined endogenously such that local markets clear: NX L it(w t )=N, (4) i=1 where N is the number of farmers in each community. For any given wage rate, income (rather than utility) maximizing farms will choose labor inputs such that the discounted future increase in revenue equals current wages, i.e. 13 By normalizing prices to one, we suppress seasonal variation in prices. Empirically, prices tend to be highest in the hungry season (period 3), explaining at least some of the strong seasonal patterns in consumption that we observe. We fix prices but allow for endogenous wages because the latter are more likely to be determined through very local markets while the former are more likely to vary regionally. 14 Note that r corresponds to the interest rate on borrowing or savings plus one. 7

8 E[Y 0 (L t )] r 4 t = w t. (5) For any given interest rate r, utility maximizing farms will choose consumption and investment level such such that u 0 (c t ) E[V 0 (Y )] =( r)4 t (6) and choose a labor allocation such that the discounted marginal loss in future utility equals the wage times the marginal utility of current consumption: 4 t E[V 0 (Y 0 (L t )) u 0 (c t ) = w t. (7) If utility is linear in consumption and the subjective discount factor is equal to the market rates ( = 1 r ), optimality condition (7) is the same as condition (5). If either assumption does not hold, utility maximization will lead to differential labor allocations; the higher the marginal utility of current consumption and the higher the subjective discounting rate, the lower on-farm labor investment compared to the income maximizing optimum. Interest rates shape consumption patterns across the agricultural cycle: the lower the savings rate, the larger first period consumption compared to consumption later in the cycle. This pattern of consumption results in a distinct hungry season associated with period 3, when savings are most depleted and consumption levels are at their lowest. Higher borrowing interest rates reduce farms ability to invest, and increases their propensity to sell labor to finance consumption as shown in equation (5). This is most obvious in a setting where savings are zero, as is the case for most small-scale farmers in our study setting during the hungry season. If all farms have to borrow to finance hungry season consumption and investment, higher interest rates mechanically lower investment in both capital and labor inputs. As a result, demand for labor falls and off-farm labor supply increases, resulting in a decrease in local wage rates until the labor market equilibrium in equation (4) is restored. Higher interest rates lower disposable incomes both through lower immmediate wage income and lower subsequent harvest outcome, leading to a reduction in consumption in all periods. Individual vs. Aggregate Level Shocks To explicitly account for the uncertainty faced by small-scale farmers, our model allows for two types of shocks. Aggregate productivity shocks, such as temperature and rainfall patterns or local pest outbreaks, directly lower the marginal product of labor and capital investment. Depending on the wage elasticity of the aggregate labor supply relative to the decline of the marginal product of labor on each farm, the net effect of aggregate productivity on total labor allocation on each farm in ambiguous, and can be positive or negative. With lower returns to labor investment, wages will fall 8

9 in equilibrium. The priors for capital investment and consumption are less ambiguous: with a lower marginal returns in terms of future output, investment will decline. With lower current (wage) and future (harvest) incomes, consumption will fall in all periods. Second, unanticipated period- and farm-specific income shocks, t, include health related costs, field specific crop needs or other family-related (unanticipated) expenditures. These lower disposable income, and therefore consumption in all subsequent periods. For any given production plan, farms will increase borrowing and lower consumption to accommodate the reduction in overall resources available, but will not change labor allocation. Predictions Our intervention provided farming households with access to seasonal food or cash loans at an interest rate below the informal lending rate available locally. Given the loans were not announced until the start of period 3 (the hungry season) in our main treatment arms, decision making in periods 1 and 2 should not have been affected by the intervention. In practice, most farmers have completed planting and basal fertilizer application by January. The main task during the hungry season is weeding, with some additional fertilizer and chemical application if such inputs are available. Further details on the treatments and timing are provided in subsequent sections. The main objective of the project was to test the following hypotheses: 1. Access to lower interest borrowing lowers households propensity to sell labor and increases their likelihood of hiring labor, increasing local wages. 2. Access to lower interest borrowing increases overall on-farm labor. 3. Access to lower interest borrowing increases consumption. 4. Access to lower interest borrowing increases agricultural output Predictions 1 and 2 emerge from equation (5). Lower interest rates result in a marginal adjustment away from household labor sales, with a corresponding increase in on-farm labor inputs. Prediction 3 is directly related to these adjustments: as available borrowing rates fall and labor sales decrease, household consumption is adjusted upward in the hungry season. Prediction 4 follows mechanically from prediction 2, providing the marginal product of labor is positive. Negative aggregate level shocks change the predictions: while farms should still adjust their labor allocation towards their own fields (less selling, more hiring), we expect a relatively larger share of the additional resources to be used for consumption, and an accordingly smaller impact on total output. Farmers exposed to idiosyncratic shocks should experience similar labor adjustments to farmers without such shocks, but access to lower interest borrowing should lessen the impact of short-term shocks on consumption. 9

10 3 Background and Context The study was implemented between October 2013 and September 2015 (with data from three agricultural harvests) in Chipata District, Zambia. Chipata District is located at the southeastern border of Zambia, with an estimated population of 456,000 in 2010 (CSO 2010). Approximately 100,000 people live in Chipata town, the district and provincial capital; the remaining population lives in rural areas, with small-scale farming as primary source of income. According to the 2010 Living Conditions Monitoring Survey (CSO 2010), 47 percent of household were classified as very poor in Chipata District compared to 32 percent in the rest of Zambia. In rural Chipata, 63 percent of households classified as very poor compared to 42 percent in rural Zambia overall. Average monthly expenditure of rural households was estimated at US$ 122 in 2010 (US$ 0.8 per person and day), corresponding to about one third of the national average (US$ 389). Access to electricity and piped water is close to zero. 3.1 Local credit and labor markets As described in greater detail below, the study implementation targeted on small-scale farmers, i.e. households growing crops on 5 hectares (12 acres) or less. The label small-scale is somewhat misleading since it suggests that these farmers are unusually small; in fact, small-scale farmers represent the overwhelming majority of households in rural villages in Zambia. In our study villages, over 90 percent of listed households fall into this category. The intervention affects local capital and labor markets, which we describe in some detail to aid the interpretation of the results. Capital markets In terms of borrowing opportunities, the study setting is similar to many rural areas in developing countries, where credit markets are absent or costly to access. In the baseline survey, 2 percent of household respondents report accessing formal loans for something other than inputs. 15 Input loans are more common: around forty percent of baseline respondents accessed an in-kind input loan, typically seeds and pesticides provided by companies purchasing cash crops like cotton and tobacco from small scale farmers. For accessing cash, informal borrowing channels are slightly more common: around 7 percent report taking high interest loans, locally referred to as kaloba, with monthly interest rates of 50 percent on average. Informal loans from friends and family are reported by around 8.5 percent of baseline respondents. In terms of other microcredit institutions, rotating savings and credit associations (ROSCAs) and village savings and loan associations (VSLAs) are each reported by around 1 percent of baseline respondents. Only 5.6 percent of households report saving in a bank; slightly more (9.1 percent) report saving with friends, family or employers. By far the most common savings strategy, reported by 76.7 percent of households, is saving money at home. The median self reported cash savings (a measure 15 Formal lenders include banks, credit unions, government sources, NGOs, and agricultural companies 10

11 likely to be reported with substantial error) at baseline, at the start of the planting season, is 80 Kwacha or around 14 USD. Savings also occurs through grain storage, which typically occurs in a thatch (28 percent of respondents) or bamboo (62 percent of respondents) granary. Sixty percent of households report storage losses and the median grain in storage at baseline is only four bags, or enough to last a family of four until February or March at most (see Figure 1). Thus, both cash savings and grain storage are insufficient to last most households until the next harvest. Local labor markets Local wage earning opportunities for study households are defined largely by casual or piece-wise labor locally referred to as ganyu. In focus groups, a majority of small-scale farmers in Chipata described ganyu both as the most common strategy to cope with temporary liquidity shortages, as well as an activity most farmers would rather avoid if possible. 16 In the baseline survey, the most common response to why an individual in the household worked ganyu during the previous agricultural season was to obtain food. The second most common reason was to access cash for a personal purchase, and the third was to deal with an emergency. When asked what the household will do in the coming year if they run out of food, 56 percent report that they will do ganyu. The next most common answers include borrow from friends or family (28 percent), using savings (22 percent) and sell assets or livestock (17 percent), all of which may be difficult to rely on during the hungry season. Households appear reasonably accurate in their forecast of whether they will have to engage in ganyu in a given year. Among control group households that predicted at baseline that they would have to do ganyu in the coming year, around 76 percent did; among those that predicted not doing ganyu, around 41 percent ended up working off-farm. Households that sell ganyu one year are not necessarily sellers in all years. Among control group households that did not engage in ganyu the year before the study, 40 percent sold ganyu the following year. Further detail on how ganyu participation varies with age and gender of household members is discussed in Fink et al. (2014), which describes a pilot for the current study. The majority of casual labor transactions take place in or near the worker s own village, which may be explained by the relatively large distances and general absence of motorized transport. The overwhelming majority of farms hiring ganyu are small (i.e. fewer than 5 hectares of land), with some farms acting as both buyers and sellers during a single season (thought typically at different points in the season). Reciprocal labor arrangements are relatively rare. Wages are seasonal, with the highest wages reported during planting (October to early December) and at harvest (May to June). Wages are determined by local market clearing, which either means within a village or within a small group of villages. Ganyu wage rates are typically negotiated on a case-by-case basis, and anecdotally are highly responsive both to demand and supply shocks. 16 For instance, in our baseline survey, around 90 percent of households disagreed with the statements Doing ganyu increases people s respect for you in the community and Successful farmers do lots of ganyu. Conversely, around 60 percent of households agree with the statements Lazy people do lots of ganyu and People who can t budget do lots of ganyu. 11

12 In the hungry season (when resources are most constrained), wages are likely to be suppressed both by increased supply from cash-constrained farms and by reduced local demand (also potentially due to cash constraints). As a result, farms with sufficient resources to hire ganyu may be able to hire labor at rates below the marginal product of ganyu on their land; as such, ganyu contracts may constitute a within-village or within-community transfer from smaller or more constrained to larger or less constrained households. 3.2 Study sample The study sample was constructed to be representative of small-scale farmers living in rural areas Chipata District. The district is divided into 8 administrative blocks, each of which contains a number of camps. We randomly sampled 5 villages from 50 of the 53 camps in the district, omitting the camps that contain Chipata town. The village list was assembled from the Ministry of Agriculture s farm registry, which included 99,000 registered farms in the district in To facilitate sampling, villages with less than 20 or more than 100 farms listed in the registry were excluded from the initial village selection. IPA enumerators visited the sampled villages in order, recording the number of households, and screening for eligibility. Villages were ineligible if: (1) IPA had worked there before, (2) the village bordered a village that was in the study pilot, (3) the village bordered a village already listed, (4) the village had fewer than 17 households, or (5) it was impossible to get a 4x4 vehicle within a 5km radius of the village during rainy season. Enumerator screening visits stopped once 201 villages met all eligibility criteria. During the baseline survey, 25 additional villages were eliminated for a failure to meet one or more of the eligibility criteria that had been overlooked during the screening process. In addition, one village refused to participate in the baseline survey. This left us with a sample of 175 villages for the study. Within each eligible village, households were sampled from the village rosters collected during the initial screening visits. Only small farms less than 5 hectares according to the Zambian Ministry of Agriculture were eligible for the program. 17 Eligible households were randomly sorted and the first 22 selected for the baseline survey. A total of 3,701 households were sampled for the baseline and 3,141 were surveyed (84.9 percent). 18 the baseline sample, in Section 4.4 below. We describe attrition, conditional on being in 17 We restricted our sample to households with at least 2 acres of land to distinguish households with very small scale home gardens from households engaged in crop production, and also to increase the likelihood of sufficient harvest to repay the loan. 18 The most common reasons that listed households were not surveyed were that they were temporarily or permanently away from the village (N=219) or that they were ineligible when land size was verified with the household head (N=146). 12

13 4 Experimental Design Study implementation began with baseline data collection in October 2013 and lasted for two years, with some differences in the study design across years, as described in this section. 4.1 Loan treatments The main objective of the project was to estimate the impact of short run loans offered during the hungry season on household-level outcomes. The study took place over two years and was designed to coincide with the agricultural calendar (see Appendix figure A.1), which starts with field preparation in September, followed by planting activities around the time of the first rains in November. Planting is followed by weeding between January and April, which is also the time referred to as the hungry or hungry season. In April, early crops start to become available and harvest begins in earnest in May. Between August and October, few agricultural activities take place. We refer to study year 1 as covering the agricultural cycle (beginning in September) and study year 2 covering the agricultural cycle. The study design is summarized in Figure 2. The study included two main loan treatment arms: a cash loan treatment and a maize loan treatment, both at the start of the hungry season (January). Repayment was due at harvest, and loans could be repaid in either cash or maize (or both). Access to cash versus in-kind borrowing presents trade-offs that these two arms were designed to investigate. On the one hand, if hunger is the primary driver of ganyu labor, then access to maize loans offers a more direct solution with a relatively low risk that the resources would be diverted toward wasteful consumption. On the other hand, hunger may not be the only driver of ganyu; cash for health or other household or farming needs would more readily addressed by a loan in cash. As discussed in further detail below, the implementation cost of both programs varied widely, with important implications for scalability and sustainability. Both treatment arms were rolled out in January Of the 175 study villages, 58 (1033 farms) were assigned to a control group, which received no intervention, 58 (1092 farms) were assigned to a cash loan program, and 59 (1095 farms) were assigned to a maize loan program in the first year of the program. In the second year of the program, the treatment groups were rotated: 20 villages that were in the control group in year 1 were rotated to either the maize loan or cash loan treatment arms (10 each), and 29 cash loan villages and 28 maize loan villages were rotated to the control group. To test ex ante adjustments in the production plan, we also varied the timing of the loan announcement in the second year of the program. Half (40) of the treated villages received notification before the start of the planting season, in September, while the other half of treated villages was only informed about the loan program in January. In addition, to identify the extent to which repayment modalities affected uptake and repayment, half (40) of the treated villages in Year 2 13

14 were required to repay the loan in cash only (i.e. no maize repayment allowed). In addition to the main loan treatments, a small number of villages (6 villages, 91 farms in year 1 and 5 villages, 81 farms in year 2) were assigned to an income effect control group, which provided a cash gift of 60 Kwacha, which was the median value assigned to participation in the loan groups in our choice experiments. 19 Cash gift villages were randomly selected within geographic blocks from villages initially assigned to the control group. Details of the cash and maize loans In the maize loan treatment arm, households were offered three 50-kilogram mags of unpounded maize. This quantify was chosen so that the typical family of five would be able to cover their basic maize needs for at least two months during the peak hungry season. In the cash loan treatment arm, households were offered 200 Kwacha (~ USD 33), which corresponds to value of the three maize bags at official government prices. In both treatment arms, repayment was due in July when most harvest activities were completed. In the first year of the program, households could repay either 4 bags of maize or 260 Kwacha (or a mix at K65 = 1 bag). As described above, some villages were required to repay in cash only during the second year of the program. While both treatment arms were designed to reflect an interest rate of about 30 percent (over 5 months), actual interest rates are hard to compute in practice due to substantial regional and seasonal price fluctuations in grain prices, as well as limited information on transaction costs associated with buying and selling maize locally. As shown in Table 1, interest rates in the maize arm vary between -11 and 33 percent depending on which maize price is used in the calculation, and also depending on what repayment modality is chosen by farmers. To make the two loan programs as comparable as possible, we conducted a series of hypothetical choice experiments in non-target villages within the district in November In these choice experiments, respondents (N=72) were asked a series of dichotomous choice questions on whether they would prefer a loan for three bags of maize over a cash loan of x Kwacha, with x varied between 50 and 600 Kwacha percent of respondents preferred a maize loan over a cash loan of 175 Kwacha, while only 36 percent preferred the maize loan over a cash loan of 250 Kwacha. As part of these choice experiments, we also asked about timing and acceptable interest rates. Specifically, respondents were asked if they would take up a maize (cash) loan that paid 3 bags (200 Kwacha) in January with a repayment of 4 bags (265 Kwacha) due in subsequent months. While only 27.8 (maize) and 20.8 (cash) respondents were interested in a loan with repayments in May, acceptance rate jumped to to 81.9 and 83.3 with repayment in June for maize and cash loans, respectively. Responses to these questions on value and timing determined final design decisions 19 For further details on choice experiments, see Appendix C Hypothetical loan dates were consistent with program offered (pay out in January and repayment in June), but the hypothetical loans involved no interest. 14

15 for the treatments. Further detail on the implementation of the choice experiments is provided in Appendix C.1. In year 1, treatments were assigned at the village level using min-max T randomization (Bruhn and McKenzie 2009), checking balance on both household and village characteristics. The approach relies on repeated village-level assignment to treatment and selects the draw that results in the smallest maximum t-statistic for any pairwise comparison across treatment arms. Balance was tested for 14 household level variables, village size and geographic block dummies, with results described in Section 4.4. The smallest p-value for the pairwise comparisons observed in the final draw was p = In year 2, treatment assignment was balanced on the same variables plus yields from year 1, and stratified by year 1 treatment. In other words, year 2 treatment assignment was carried out within each year 1 treatment arm, with assignment to both the main treatment arms (control, cash loan and maize loan) and the sub-treatments (income effect control, early notification and cash repayment). 4.2 Loan program implementation The loans were administered under the project name Chipata Loan Project (CLP) to distinguish them from the survey visits, which were conducted by Innovations for Poverty Action (IPA) surveyors. This distinction between CLP and IPA was made to assure participants that survey responses would not affect loan eligibility. Eligibility in both years was determined by participation in the baseline survey (see Section 3.2). In year 2, eligibility also depended on year 1 repayment in the sixty villages treated in both years, i.e. any household that did not fully repay in year 1 was not eligible for a loan in year 2, though they were still included in data collection. The loan intervention was announced to households during a village meeting to which eligible households were invited. 21 At the meeting, project staff began by describing eligibility for the program to clarify why only some households were invited to the meeting. The terms of the loan were then described, followed by details on how the loan distribution would be organized. Loan enrollment and consent forms were provided to eligible households. If a household wished to join the program, they were required to present both forms with a signature of the household head when picking up the loan. Loans were distributed between 3 days and one week after the village meeting at a location convenient for transportation, selected in cooperation with the village headman. Project staff registered attendees, confirmed their identity using the national registration card, 22 and collected their signed enrollment and consent forms. Before finalizing the transaction, project staff confirmed 21 Ineligible households were not barred from listening in. Eligible households could send an adult representative if the household head was not available to attend. All village headmen were eligible for the loan, even if they were not sampled for the baseline survey (and are therefore not in our study sample). In addition, the baseline data for 3 households who were surveyed was lost. They are dropped from the sample. 22 In select cases, a household representative picked up the loan. In these cases, the representative needed to carry the loan-holder s NRC card with him or her. 15

16 that the participant understood the terms of the loan. The loans (cash or maize) were handed over and a receipt was provided to the household and kept for project records. Repayment was due in early July. Villages were notified in advance about the date of repayment as well as the central locations at which repayment would be collected. Two attempts at collecting repayment were made. Households were provided with a repayment receipt upon full repayment. Throughout the project, households were told that the program might or might not continue in future years. Further summary statistics on repayment patterns are described below. 4.3 Data We rely on both survey and administrative data in our analysis. Administrative records include loan take up and repayment, two key outcomes in our data. Comprehensive surveys of all study households were conducted at baseline (November 2013), harvest of year 1 (August 2014) and harvest of year 2 (August 2015). We refer to these as long recall surveys since they ask questions about the preceding agricultural cycle. Additional sub-sample surveys on labor activities, consumption and farming practices were collected on an ongoing rolling sample between comprehensive survey rounds. We refer to these as short recall surveys since they primarily ask about activities in the past two days to two weeks. A total of 15,044 observations from the sample of 3,141 households were collected over the course of the study. Appendix B.1 summarizes sample sizes and key content collected in each survey. Main outcome measures We focus on three main outcome types, based on the predictions in our conceptual framework: (1) measures of labor allocation and local wage rates, (2) consumption indicators, and (3) agricultural output. Labor allocation outcomes include (a) family labor sold to other farms (ganyu sold), (b) labor labor purchased (ganyu hired) and (c) family labor invested on-farm. For ganyu hiring and selling, wee construct binary measures from the long-recall surveys (baseline, harvest and endline) which asked farms to report retrospectively on the preceding agricultural season. We use the short-recall surveys (labor surveys) to construct measures on both the extensive and intensive margin, at the household level, for labor activities over the past week. Our intensive margin measure is in hours, summed across all individuals in the household (i.e. a total hours measure), to account for the fact that ganyu does not always last the full day and a partial day of ganyu sold might still allow for some time invested on-farm. We can use adjustments in family hours on-farm and hired hours to construct a measure of total labor hours on-farm over the past week. We also construct a measure of equilibrium wages at the village level during the hungry season (when most ganyu is reported). We again use the labor surveys, which also collect earnings from ganyu sold at the individual level, 16