What climate-smart agriculture means for smallholder farmers

| Article

Smallholder farmers1 generate an estimated 32 percent of global greenhouse-gas (GHG) emissions from agriculture.2 They are also one of the populations most at risk from climate change. Our analysis shows that in three countries—India, Ethiopia, and Mexico—nearly 80 percent of all smallholder farmers could be affected by at least one climate hazard by 2050 (Exhibit 1). Moreover, climate change will affect land suitability for crop production. For example, by 2050, India could lose 450,000 square kilometers of land currently suitable for rainfed rice cultivation (Exhibit 2).

1
2
By 2050, India could lose 450,000 square kilometers of land currently suitable for rainfed rice cultivation, while Southeast Asia will gain suitable land.

Stakeholders have focused on climate-smart agriculture for the past two decades. Nonetheless, there is no clear road map for the types of mitigation and adaptation measures smallholder farmers can adopt and how to prioritize investments and efforts to support those measures. In this article, we try to fill this gap. Our work is informed by a geospatial analysis of climate risk in key smallholder markets; an extensive review of current technologies and tools that smallholder farmers can deploy to adapt to and mitigate climate change; and a prioritization of those measures based on agroecological and farming systems in different countries.

Identifying adaptation and mitigation measures that the world’s 510 million smallholder farmers can adopt is critical to the protection and support of their livelihoods in the face of climate-related hazards.3 They are also key to global food security. Smallholder farms produce a third of the world’s food, and global food demand is expected to increase by 60 percent by 2050.4 Meanwhile, climate change has already led to a 21 percent loss of agricultural productivity globally since 1961.5 In a world where temperatures could rise another 2°C by 2050, there could be large reductions in crop yields if no countermeasures are taken. For example, in Africa, pest-driven losses are expected to increase by 50 percent (compared to the baseline) for staple crops such as maize, rice, and wheat.6

These measures are also important for countries with large smallholder populations and those that are making low-carbon pathway commitments, given that they will likely need to help farmers transition to less carbon-intensive agriculture. For example, Kenya has announced a nationally determined contribution (NDC) of reducing 32 percent of emissions by 2030, relative to the baseline. Kenya’s agricultural sector is the country’s largest source (58.6 percent) of total emissions.7 So reaching its carbon reduction goals will require the participation of its 4.5 million smallholder farmers (about 80 percent of all farmers) and 600,000 pastoralists.

We identified more than 30 measures smallholder farmers can adopt to help adapt to and mitigate climate change. We also noted several approaches that governments, development partners, and the private sector could pursue to help scale those measures. We found that implementing a prioritized set of three measures at scale in each country could mitigate 45 percent of smallholder farmer–driven carbon emissions. For adaptation, almost every smallholder farmer can adopt at least one on-farm adaptation measure. But about 75 percent can adopt at least three—and the more measures they adopt, the more likely that greater climate resilience could be achieved.

We identified more than 30 measures smallholder farmers can adopt to help adapt to and mitigate climate change.

How smallholder farmers can adapt to and mitigate climate change

Adoption of adaptation and mitigation measures among smallholder farmers is complex. Smallholder farms are fragmented and often have limited access to inputs, new agricultural technologies, and financing. However, as previously mentioned, we identified more than 30 measures smallholder farmers can pursue for adaptation and mitigation. We divide these measures into five categories: animal production practices, rice-based measures, other crop-based measures, land-use change and intensification, and postharvest and processing loss (Table 1). We separate rice measures from other crops because emissions from rice production are anywhere from five to 30 times higher per hectare than those from other crops.8

Table 1. Thirty-three mitigation and adaptation measures relevant to smallholder farmers

Of these measures, some are limited to climate adaptation, such as eco-engineering reefs to protect coastlines from flooding and the introduction of pest-tolerant crop varieties. Others are exclusively mitigation measures, such as GHG-focused livestock breed selection and scaling solar-powered irrigation. But many measures have both adaptation and mitigation benefits. For example, introducing irrigation to increase productivity has an indirect mitigation effect, allowing farmers to grow more crops on less land and therefore reducing the amount of land needed for agriculture. But irrigation also has an adaptation benefit by allowing farmers to continue to grow certain crops despite climate change–related increases in water stress and drought. (See sidebar, “About the research,” for an explanation of adaptation and mitigation and the methodology under which these measures were derived and prioritized; see the technical appendix for the case studies on which each measure is based.) While we focus on carbon emissions and climate adaptation in this article, it is important to note that these measures also have important nature-related benefits. These include reducing land-use change, decreasing nutrient runoff into waterways by moderating fertilizer application, and adopting integrated pest management practices.

Each measure is based on a proven trial in a smallholder farming environment. For example, one measure related to animal production practices focuses on improving livestock breeding systems for increased productivity and reduced GHG emissions. The approach was used in Malawi and Uganda for a community-based goat-breeding program with 269 farmers. In the program, goats reached a higher average weight (from 16 kilograms to 19 kilograms with the improved breeding) and survival rates (from 72 percent to 91 percent); emissions were also reduced.9

As another example, in rice production, straw management can be used to maintain and enhance soil fertility and carbon storage. One case from Vietnam showed that the direct incorporation of rice residues into soils after harvest led to increased soil organic carbon by about three metric tons per hectare. The approach also significantly reduced the amount of chemical fertilizer required to achieve the same yields by returning nutrients to the soil.10 For other crops, greater use of a combination of six tillage, residue management, and intercropping practices in legume–rice–wheat cropping systems in India resulted in the lowest emissions. The practices led to 823 to 3,301 kilograms of CO2-equivalent sequestered per hectare per year compared with 4,113 to 7,917 emitted per hectare per year in typical farming practices, as well as a 29 percent decrease in water usage.11

Prioritizing investments in adoption of on-farm adaptation and mitigation measures

The choice of measures to invest in depends on multiple factors, including a country’s farming system; farmers’ access to markets, which is an indicator of their ability to access other actors such as sales agents for seed companies that sell new drought-resistant seed varieties; cost of adoption; and capabilities required. Taking these factors into account, we prioritized measures using geospatial analysis in three countries: India, Ethiopia, and Mexico. Together, they are home to more than 40 percent of the global smallholder farmer population and generate about one metric gigaton of GHG emissions from agriculture.12 The results highlight not only areas of commonality across smallholder systems but also the importance of a differentiated approach by country and subnational region.

How adaptation measures vary by country

Priorities for adaptation differ by country. These differences are mainly driven by varied exposure to climate change hazards and by the farming systems used in the exposed areas (Table 2). For example, using drought-tolerant seed varieties is much more applicable in drought-prone India or Mexico than in Ethiopia. This is because pastoral livestock systems dominate the drought-prone regions of Ethiopia rather than crop production. Moreover, as shown in Exhibit 1, more farmers in India (95 percent of the total) are exposed to at least one risk, and most are exposed to multiple risks, which requires the adoption of multiple adaptation measures. In Ethiopia, on the other hand, 37 percent of farmers are exposed to at least one risk, and few face multiple risks. Nonetheless, resilience is likely to increase for any farmer who adopts multiple measures. Our analysis shows that about 75 percent of smallholder farmers in these three countries could adopt the three highest-priority measures.

Table 2. Priority adaptation measures

How mitigation measures vary by country

From a technical point of view, the breadth of application of all the measures combined would allow the vast majority (90 percent) of farmers in the three countries to adopt at least one mitigation measure. However, the applicability of measures varies across and within countries, driven by different farming systems and practices that lead to different emission mixes (Exhibit 3). For example, fertilizer application rates are more than five times higher in India than in Ethiopia, which means that soil- and fertilizer-related mitigation measures are much more applicable in India.13

3
Smallholder farmer–driven emissions across India, Ethiopia, and Mexico differ greatly.

When layering on feasibility criteria, the top ten mitigation measures highlight important differences by country (Exhibit 4 and Table 3). In India, given its large crop production (and rice production, in particular), rice- and crop-based measures account for nine of the ten priority measures. About 50 percent of smallholder farmer–driven agriculture emissions in India could be mitigated by scaling agroforestry and transitioning to more sustainable rice production practices on smallholder farms. Agroforestry alone represents the largest opportunity, with a mitigation opportunity seven times that of the next most impactful measure of incorporating rice straw into soils.14 This is consistent with the launch of India’s National Agroforestry Policy in 2014. India is the first nation to introduce such a plan to mitigate climate change and increase the resilience of smallholder farmers.

4
India’s mitigation opportunities are highest for crop measures (especially rice), whereas Ethiopia’s and Mexico’s are with livestock.
Table 3. Priority measures by country

In Ethiopia and Mexico, where cattle production systems are more common, livestock-related measures dominate and could collectively mitigate up to 25 percent and 35 percent of emissions, respectively. In Mexico, 60 percent of land is considered arid or semiarid, with a substantial area dedicated to the range farming of livestock. As a result, one of the largest mitigation opportunities15 lies in regenerative rangeland management. In Ethiopia, the livestock sector is responsible for 60 percent of agricultural emissions.16 Thus, Ethiopia’s emission-reduction potential is mostly associated with livestock-based measures (eight of the top ten), with rangeland management, improved timing of livestock sales, increased adoption of veterinary services, and feed-based measures making up a significant proportion of the opportunity.

In aggregate, these countries could achieve about 455 metric megatons of CO2-equivalent emissions savings—collectively about 45 percent of the total smallholder farmer–driven agriculture emissions from India, Ethiopia, and Mexico—by implementing only the top three prioritized levers across smallholder farms in a comprehensive and widespread manner.

These countries could achieve about 455 metric megatons of CO2-equivalent emissions savings … by implementing only three levers.

Implications for actors seeking to support adaptation and mitigation for smallholder farmers

Governments, financiers, development organizations, and private-sector players have a key role to play in supporting the global smallholder-farming community’s shift to more sustainable practices. Our analysis highlights two important considerations for this support.

First, as described above, it is important to prioritize which measures to focus on at a subnational level given the heterogeneity of smallholder farmer production systems, the range of impact, and the feasibility of adoption. This prioritization exercise could enable the identification of clusters of smallholder farms in which multiple measures are feasible for adoption and piloting could begin.

On this point, concerns such as market access are critical. For example, in India, almost all farmer types are within four hours of a market by road. By comparison, in Ethiopia, as few as 10 percent of farmers have market access in some areas because of greater population dispersion and less-developed infrastructure. This low market access suggests a potentially higher cost per farmer to implement measures at scale.

Second, driving adoption of these measures will require solutions at the farm and agriculture-system levels. Not only will farmers have to consider changing on-farm practices, but national agriculture research systems will also have to reflect on how to develop and commercialize new technologies, such as drought-tolerant seeds. Additionally, stakeholders will have to consider investments such as improved infrastructure to build resilience in the face of climate volatility. Government and private-sector actors will also have to consider building market linkages for crops in different areas because farmers might switch crops due to changing land suitability, as described earlier.

Driving adoption of these measures will require solutions at the farm and agriculture-system levels.

We identified several cross-cutting approaches that could help scale priority measures (Table 4). These solutions start with building a climate risk–adjusted agriculture and land management plan that geospatially prioritizes adaptation and mitigation measures at a subnational level and that ties investments to that prioritization. The solutions also include developing financing and incentive mechanisms to encourage on-farm practice shifts (for example, redesigning subsidy schemes, offering tax incentives, and linking farmers to carbon markets); putting system enablers in place (investing more in R&D and scaling traceability systems); and mitigating climate-induced volatility (scaling up crop insurance and integrating climate modeling into food security planning).

Table 4. Cross-cutting solutions to scale adoption of the prioritized measures

Few of these cross-cutting approaches have been applied in practice, given that the discussion of adaptation and mitigation for smallholder farmers is relatively new. However, there are some pilots under way. In China, for example, the government changed its subsidy policies to discourage use of chemical fertilizer for specific crops and encourage the adoption of organic fertilizer substitutes, with a particular focus on reducing nitrogen overapplication. This policy has helped reduce the application of chemical fertilizers by 111.5 kilograms per hectare in pilot counties and increase the use of organic fertilizers by 346.36 kilograms per hectare for sampled farmers. One estimate found that such policy reforms could reduce fertilizer use by 30 percent compared with current rates.17

Additionally, the International Food Policy Research Institute’s modeling of historic data in Punjab, India, has been used to project the effect of removing the subsidy for groundwater extraction, eliminating minimum support price policies for water-intensive crops, and reallocating subsidies to climate-smart technologies such as crop diversification, low tillage, and on-farm rainwater-harvesting ponds with solarized pumps and microirrigation. The modeling suggests there is potential to reduce water consumption by 15 billion cubic meters per year and reduce GHG emissions by 23 million metric tons by 2050. The modeling also finds that there will ultimately be no change to Punjab’s budget if subsidies for groundwater extraction are reallocated as incentives for the adoption of climate-smart agriculture practices.18

Development partners and private actors are also implementing pilots to support adoption of climate-smart measures. One Acre Fund, a social enterprise that works with more than one million smallholder farmers in Africa, is expanding an agroforestry program, exploring the link to carbon credit markets to offer incentives for on-farm tree planting.19

Others are using financial innovation to support climate-smart agriculture and resilience. F3 Life and Financial Access piloted a Climate-Smart Lending Platform in 2017 with 10,000 farmers in Kenya and Rwanda. They worked with lenders to develop loan products that featured terms and conditions encouraging farmers’ uptake of climate-smart agricultural and land management practices and use of mobile technology to monitor the adoption of climate-smart farming in compliance with loan agreement requirements.20 Pula, an insurance tech start-up in Africa, has provided agriculture insurance to 6.8 million farmers,21 including products to address weather-related yield impacts.

Resilience-related infrastructure is being put in place in even the most remote and low-tech contexts. For example, “contour bunds” (low walls) combined with Zai pits22 have been established in 200,000 to 300,000 hectares of land across the Sahel. The approach almost doubled the yield of cereals, despite frequent droughts.23

This is not a comprehensive list of macroscale solutions. But they illustrate some powerful initiatives stakeholders can pursue to support the scaling of adoption of adaptation and mitigation measures among smallholder farmers.

Additionally, actors can support further research to inform decision making. For example, the cost to adopt and scale these measures is largely unexplored in the currently available literature. While we use a qualitative assessment on cost, understanding the true costs is critical in making trade-offs on what measures to choose. Another research question could explore the effectiveness of various measures, particularly regarding adaptation where there is no common metric or set of metrics. Finally, extensive piloting would be helpful to test which macrosolutions are most effective in encouraging farmers to adopt priority measures and to develop and derisk sustainable business models to support adoption.


Smallholder farmers can adopt a range of measures to mitigate and adapt to the risks of climate change. Governments and other stakeholders could consider supporting them in their efforts to adopt sustainable farming practices. In doing so, stakeholders could reflect on the national context in which they are working and collaborate to identify adaptation and mitigation priorities. The prioritization would ultimately act as a North Star and would feed into an agriculture land management plan to inform a more efficient allocation of investment and effort, from innovative financing mechanisms to targeted research and development and technology innovations. Climate change is already creating huge losses for smallholder farmers globally. To achieve a 1.5° pathway, responsible actors have no time to lose in supporting the sustainable smallholder farmer community.

Download the full PDF, which includes a Technical Appendix.

Explore a career with us