S7 Ep34: Why farmers struggle to adopt new agricultural technology

How many maize seeds should farmers plant in each hole? Ask the farmers and they say two or three. Agronomists can show them more reliable seeds, where they only need one. 

But change the seed and everything else changes too; the fertiliser, the spacing, the whole system. This is why getting better technology into the hands of African farmers, and helping them to find ways to improve their profits, is so much harder than it looks.

Rachid Laajaj (Universidad de los Andes) and Karen Macours (Paris School of Economics) tell Tim Phillips about an experiment in Kenya in which farmers ran trial plots on their own land for three seasons, comparing input combinations side by side, and were followed for five seasons more.

Farmers adopted the new inputs and their profits fell. But they kept experimenting anyway, season after season, until the losses became gains. The most skilled farmers went first, made the most mistakes, and paid the highest price; their neighbours watched, copied, and adopted at a fraction of the cost.

Each failed attempt costs a farmer a season. Lower that cost, Laajaj and Macours argue, and you change what is possible for millions of farming households.

The research behind this episode:

Laajaj, Rachid, and Karen Macours. 2026. "The Complexity of Multidimensional Learning in Agriculture." Econometrica 94 (2): 465-503. 

To cite this episode:

Phillips, Tim, Rachid Laajaj, and Karen Macours. 2026. "Why farmers struggle to adopt new agricultural technology" VoxDev Talks (podcast). 

About the guests

Rachid Laajaj is Associate Professor of Economics at Universidad de los Andes in Bogota, with research spanning technology adoption in agriculture, corruption, and human capital, studied from a micro-development perspective with particular attention to the role of information. He received his PhD in Agricultural and Applied Economics from the University of Wisconsin-Madison.

Karen Macours is a professor at the Paris School of Economics, a senior researcher at INRAE, and a Research Fellow of the Centre for Economic Policy Research in the Development Economics programme. Her research spans agricultural productivity, rural poverty, social programmes, and early childhood development. She chairs the Standing Panel on Impact Assessment of the CGIAR and is co-editor of the Journal of Development Economics.

Research and concepts cited in this episode

The trials. In 48 randomly selected villages in Kenya, ten farmers per village were invited to run a three-season agronomic trial on a small parcel of their own land, guided by an agronomist. Each trial plot was divided into six subplots; five tested different combinations of modern inputs and one served as a control. The trials occupied a tiny share of each farm, so the direct economic effect was negligible; the point was the opportunity to learn. A further 48 villages served as controls, giving a sample of 960 farmers followed across six seasons of data collection.

Integrated Soil Fertility Management (ISFM). The input combinations tested in the trials followed ISFM principles; a sustainable intensification approach that combines improved seed with mineral fertiliser and organic inputs, adapted to local conditions. Its logic is precisely the multidimensionality the paper studies; the components work through their combination, not in isolation.

The International Institute of Tropical Agriculture (IITA) provided the agronomists who guided farmers through the trials. The paper builds on joint work with IITA.

The target-input model. The workhorse theory of learning in agriculture, in which farmers learn the optimal quantity of a single input through experience; it dates to work by Foster and Rosenzweig in 1995 and underlies much of the literature on social learning, including Conley and Udry's study of pineapple farmers in Ghana. Laajaj and Macours extend it to many interdependent inputs, where adopting one requires re-optimising the others; what Laajaj calls the curse of multidimensionality.

Measuring farmer skills. The origin of this project; Laajaj and Macours spent years developing measures of farmers' cognitive, non-cognitive, and technical skills, work that circulated as CEPR Discussion Paper 13271, "Measuring Skills in Developing Countries" (gated). Asking farmers agronomic questions revealed that many had no single right answer; the answer depended on everything else the farmer was doing.

Employment in agriculture. Macours notes in the episode that agriculture accounts for half of employment in sub-Saharan Africa, rising to as much as two-thirds when the wider agro-food sector is included.

Conservation agriculture. A prominent example of a knowledge-intensive sustainable practice; it combines minimal tillage, crop rotation, and the retention of crop residues, so a farmer must learn several new practices at once and find a combination that works.

More VoxDev Talks episodes

African agriculture's underappreciated supply side. Hope Michelson on the other half of the adoption puzzle; the markets and firms that supply seeds, fertilisers, and pesticides to smallholders.

Strengthening climate resilience in agriculture. Tavneet Suri on how small-scale farmers respond to extreme weather, and the seed varieties that protect crops against floods and droughts.

Combining improved seed varieties and index insurance to address drought losses. Paswel Marenya on a trial in Mozambique and Tanzania that paired drought-resistant seeds with insurance.

Related reading on VoxDev

Agricultural Technology in Africa, the VoxDevLit surveying the evidence on what drives, and blocks, technology adoption by African farmers.

Improving agricultural extension and information services in the developing world, drawing lessons from nearly fifty randomised evaluations of constraints on smallholder productivity.

Encouraging agricultural technology adoption through autonomy: evidence from Mexico, on how giving farmers ownership of the adoption process makes new practices stick.

Harnessing the benefits of digital agriculture for smallholder farmers in East Africa, on text-message-based extension programmes evaluated in Kenya and Rwanda.