Increasing the tenant’s share in output encourages profitable risk-taking, in addition to large effects on input levels
Arthur Young travelled to France in 1787, shortly before the onset of the French Revolution. The prominent English agronomist and writer described what he encountered in his Travels in France (1792): the miserable situation of French farmers, the low levels of agricultural production, and the emerging unrest. He also described the wide-spread ‘métayage’, a form of contracting that was new to him: in the France of the time, large landowners provided land for free to the farmers, and the fruits were equally shared in return. Young felt this contractual form, now known as ‘sharecropping’, was surely detrimental to agricultural productivity. He famously concluded that “the magic of property turns sand into gold”.
A century later, Alfred Marshall presented a precise version of Young’s ideas. In his Principles of Economics (1890), he wrote: “For, when the cultivator has to give to his landlord half of the returns to each dose of capital and labour that he applies to the land, it will not be to his interest to apply any doses the total return to which is less than twice enough to reward him”. In a remarkable footnote, he provided a graphical version of this argument. Marshall therefore not only provided a qualitative, but also a quantitative theory of the incentive effects of output-sharing contracts; an idea that lives on in much of modern contract theory and public economics.
When Young and Marshall described the incentive effects of sharecropping contracts, access to land had been at the heart of political disputes for centuries, going back to at least classical antiquity. Their arguments provided an important new impetus to the power struggles over land: redistributing land from large landowners to tenant farmers might not just be equity-enhancing, but also efficiency-enhancing. The promise of that idea, and the prevalence of sharecropping arrangements, spurred land-reform policies around the world, from pre-revolutionary Russia to Japan and South Korea in the aftermath of World War II, from Italy to India and Iran, from de-colonised Africa to post-independence South America. It continues to fuel such efforts, with recent examples ranging from Zimbabwe to Scotland.
Evaluating the overall equity and efficiency effects of those reforms has been an elusive goal of development economics, and it remains so today. A more feasible research goal is to seek to understand individual elements of those reforms, and in particular, testing and quantifying the mechanism at the heart of Marshall’s theory: is it true that making a farmer more of a residual claimant on output changes the farmer’s behaviour, and leads to an increase in agricultural output?
Existing evidence on marshalling incentive effects
Over the past decades, a number of studies have addressed this question in various ways: Rao (1971) observed that in Indian farms, output is higher in owner-operated farms relative to sharecropped farms; Bell (1977) and Shaban (1987) used plot-level data to compare output and input levels across plots with different tenancy statuses within the same household; Banerjee et al. (2002) showed that a tenancy reform which simultaneously changed the legal output share of registered tenants, and reduced their likelihood to be evicted by the landlord, increased agricultural output in West Bengal, India relative to Bangladesh. All of these research projects find potentially large output effects from increasing the farmer’s share in output. They also applied and advanced the best-practice methods of their time on this question. However, to a sceptic, the results of those projects were also plagued with doubts as to whether the differences in output that were uncovered might be reflective of other, confounding differences across the farmers, plots and regions studied, and were maybe entirely unrelated to the contractual structure (Arcand et al. 2007; Braido 2008; Jacoby and Mansuri 2009).
New experimental evidence: Large increases in input usage, risk-taking and output
We revisited this classic question of development economics in our study (Burchardi et al. 2019), conducted in collaboration with BRAC-Uganda. In early 2013, BRAC decided to rent out land across rural Uganda and offer it to young female farmers under a standard sharecropping contract, thus acting as a landlord and requiring the farmers to provide BRAC with 50% of output at harvest. Shortly after signing this initial contract, a randomly selected subgroup of farmers were offered a better contract which they unsurprisingly accepted: they were told they could keep 75% of the output at harvest. Our research team visited them and their plots during and after the harvest to collect data on input usage, farming techniques, and output levels.
Our data show that farmers who were allowed to keep 75% of output yielded a staggering 60% higher output from their plots compared to farmers who were only allowed to keep 50% of output from theirs. While farmers in the control group produced crops worth US$95 (PPP) on the rented plots, farmers in the treatment group produced crops worth $151 (PPP). This increase corresponds to roughly 30% of monthly household income, a large effect keeping in mind that household income also comes from other sources than the plots rented from BRAC.
This raises the question of what the former group of farmers did differently from the latter group. We found that they used more capital inputs, in particular a large increase of fertiliser usage over very low baseline levels, and they purchased more tools; we also see that the farmers received more (unpaid) help from family and friends on their plots. Using standard estimates of the returns to those inputs, we would expect an increase in output of 30% relative to the control group, i.e. the increased usage of inputs can explain about half of the total increase in output. But what explains the rest? Our data show that farmers in the treatment group took on substantially more risk; for example, they planted crops which are more heavily dependent on rainfall. Output in the treatment group also responded more strongly to rainfall, and in general has a higher variance, over and above what would be expected from increased input usage. Reasonable estimates of the returns to risk-taking in African agriculture suggest these choices by the farmers can explain the remaining 30 percentage points of the output increase.
A consistent picture emerges
Naturally, these findings were obtained in a specific setting, and results might be quite different in other contexts, with other farmers, and at another time. But they are consistent with the existing research. The input and output effects are qualitatively consistent with Marshall’s theory of the incentive effects of sharecropping contracts; the magnitude of the output effects we find is surprisingly similar to the results in Banerjee et al. (2002), the closest analogue to our work. And the finding that farmers without a high share in output shy away from taking risks lines up with recent work by Karlan et al. (2014), who find that farmers in Ghana avoid riskier agricultural choices in the absence of weather insurance. In the absence of perfectly functioning insurance markets, our results suggest that increasing the tenant’s share in output may also encourage profitable risk-taking, in addition to the incentive effects on input levels.
Yet, our findings should not be interpreted as evaluating the overall effect of land reforms. Land reforms not only affect the farmers’ share in output, but typically also decrease the average size of farms, redistribute land to a different set of farmers, and might increase land fragmentation. All of these aspects of land reforms are likely to affect agricultural productivity (Adamopoulos and Restuccia 2014, Adamopoulos and Restuccia 2020, Bryan et al. 2019), and the effect of any given land reform will depend on the specifics of the policy. What ours and existing research suggests is that any policy which increases farmers’ share in output – either through a change in contracts, formal or informal taxes, or the protection of output from theft – will likely result in substantial increases in agricultural yields.
References
Adamopoulos, T and D Restuccia (2014), "The size distribution of farms and international productivity differences", American Economic Review, 104(6): 1667-97.
Adamopoulos, T and D Restuccia (2020), "Land reform and productivity: A quantitative analysis with micro data", American Economic Journal: Macroeconomics, 12(3): 1-39.
Arcand, J-L, C Ai and F Ethier (2007), “Moral hazard and Marshallian inefficiency: Evidence from Tunisia”, Journal of Development Economics, 83: 411–445.
Banerjee, A, P Gertler and M Ghatak (2002), “Empowerment and efficiency: Tenancy reform in West Bengal”, Journal of Political Economy 110: 239-280.
Bell, C (1977), “Alternative theories of sharecropping: Some tests using evidence from Northeast India”, Journal of Development Studies 13: 317-346.
Braido, L H B (2008), “Evidence on the incentive properties of share contracts”, Journal of Law and Economics 51: 327-349.
Bryan, G, J de Quidt, T Wilkening and N Yadav (2019), “Can market design help the World’s poor? Evidence from a lab experiment on land trade”, Working Paper.
Burchardi, K, S Gulesci, B Lerva and M Sulaiman (2019), “Moral hazard: Experimental evidence from tenancy contracts”, The Quarterly Journal of Economics, 134(1): 281–347.
Jacoby, H G and G Mansuri (2009), “Incentives, supervision and sharecropper productivity”, Journal of Development Economics, 88: 232–241.
Karlan, D, R Osei, I Osei-Akoto and C Udry (2014), “Agricultural decisions after relaxing credit and risk constraints”, The Quarterly Journal of Economics, 129(2): 597-652.
Marshall, A (1980), Principles of Economics, London: Macmillan.
Rao, C H Hanumantha (1971), “Uncertainty, entrepreneurship and sharecropping in India”, Journal of Political Economy 79: 578-595.
Shaban, R A (1987), “Testing between competing models of sharecropping”, Journal of Political Economy 95: 893-920.
Young, A (1792), Travels in France. During the Years 1787, 1788, 1789. Republished in London: George Bell and Sons, 1909.