Credit, subsidies, and cash transfers can improve yields and revenues for smallholder farmers in low- and middle-income countries, but translating these gains into higher profits proves far harder. The evidence suggests that credit works best not as a standalone intervention but when paired with new agricultural technologies that give farmers something genuinely worth investing in.
Editor’s note: You can listen to this podcast on Spotify, Apple Podcasts, or wherever else you get your podcasts.
Can relaxing credit constraints boost farmers' profits in low- and middle-income countries? It is a question that sits at the heart of agricultural development policy – and one to which the answer turns out to be frustratingly elusive. In this episode of VoxDevTalks, Craig McIntosh discusses a recent J-PAL policy insight that takes stock of the evidence from randomised controlled trials on credit, subsidies, and cash transfers for smallholder farmers, arriving at conclusions that challenge some of agriculture's most persistent development assumptions.
Why standard microcredit fails farmers
The obvious starting point is microcredit, long held up as the solution to financial exclusion in the developing world. Yet the evidence here is discouraging. Across studies in Morocco, Ethiopia, Bangladesh, and Malawi, take-up of microfinance loans among farmers is just 13–33%.
The explanation is structural. Standard microfinance products require repayment to begin within a week or two of disbursement – a cash-flow timeline that is simply incompatible with long-cycle agriculture. Joint-liability lending structures, designed to spread default risk among borrowers, also break down in farming communities, where weather shocks hit everyone at once.
"These products are really designed to keep farmers out in many ways."
McIntosh suggests this is less an oversight than a deliberate feature: microfinance lenders have long understood the dangers of agricultural lending and designed accordingly.
Tailored credit products can raise take-up
There are, however, better-designed alternatives. Studies from Tanzania, Ghana, and Kenya demonstrate that restructuring repayment around the harvest cycle – allowing farmers to defer the bulk of payments – roughly doubles take-up rates, from around 20–25% to 40–50%. Self-collateralised lending using durable agricultural assets, such as water tanks for dairy farmers, has produced similarly strong demand.
A particularly striking example comes from Kenya, where a study by Burke et al. addressed the 'sell low, buy high' trap: farmers are forced to sell at harvest when prices are lowest, then buy food back during the lean season when prices peak. By offering loans at harvest time to allow farmers to defer their sales by a few months, the study found take-up of 64% and returns of 29%.
"By allowing some people to come off the market when prices are low and come onto the market when prices are high, they actually flattened out those price trajectories in treatment villages."
Yields go up, but profits usually don't
Better-designed credit does appear to raise yields and revenues. In 11 out of 16 RCTs offering credit or grants, either yields or revenues increased. But the more important question for household welfare is whether profits rise – and here the evidence is sobering.
Only two out of ten RCTs that measured agricultural profits found a statistically significant increase. The explanation, McIntosh argues, lies in how farmers had been behaving before the intervention. If households had already invested up to the point where marginal revenue equalled marginal cost, then additional capital will generate additional output, but at a cost that mirrors the revenue it produces. The intervention is not wasted, but neither does it generate surplus.
"It is harder to move profit than it is to move yield, and that once you've seen that, it really does cause you to back up and ask yourself: why has there been so much focus on yield in the first place?"
The mixed record of input subsidy programmes
Government-backed input subsidy programmes (ISPs), particularly for fertiliser, have absorbed enormous resources across sub-Saharan Africa – accounting for around a fifth of the entire national agriculture budget in some countries. A meta-analysis by Jain et al., drawing on 80 non-experimental studies, finds the returns questionable in most contexts.
Targeting is a recurring problem: subsidies frequently flow to politically connected farmers or electorally valuable regions rather than to those who would benefit most. Where ISPs have worked well – notably in a Mozambique study by Carter et al., which found sustained revenue gains for years after the programme ended – the context was unusual: two-thirds of farmers had never used fertiliser before the intervention. The lesson is that ISPs are most likely to generate real returns where fertiliser use was negligible to begin with.
Cash transfers: Investment, but not always in farming
Unconditional cash transfers work in a broadly positive sense – resources are not wasted. But evidence from six studies tracking both agricultural and non-agricultural outcomes is evenly split: half found households investing more in farming, while the other half found them using transfers to diversify into non-farm enterprises.
A study in Zambia found household income doubling, driven primarily by investment in off-farm businesses. From a narrow agricultural lens, this looks like a failure. Viewed through the lens of household welfare and risk management, it is precisely what economists might hope to see. Farming in many low-income settings is a high-risk, low-return activity, and households that can use transfers to escape that trap are acting rationally.
Bundling credit with technology
The most promising results in the portfolio come not from relaxing credit constraints in isolation, but from using credit to facilitate the uptake of new agricultural technologies. A study in Tanzania found that offering soil testing and fertiliser vouchers separately had no measurable impact, but combining the two – new information paired with the liquidity to act on it – produced significant improvements.
"Credit is a very important vehicle to helping people adopt these expensive packages of productivity enhancement."
This points to a reframing of how credit should be understood in agricultural development. In wealthier farming economies, debt is a constant feature of farm life, but what it finances is movement along a technological frontier – new seeds, equipment, and market access. The same logic, McIntosh argues, should apply in developing-country agriculture: credit matters most when there is something genuinely worth investing in.