On-the-ground research must guide climate adaptation investments

Kyle Emerick
Saturday 5 Nov 2022

As research from across the globe shows increasing threats to staple food yields from climate change, investment for climate adaptation must be guided by on-the-ground research from the real-world.


Projections show substantial declines in future yields under unmitigated warming. 

The negative effect of climate change is not limited to staple foods. Production of fruits and vegetables is projected to decline sharply as climate changes. Importantly, these effects are uneven over space. Countries in the Global South --- where temperatures are already high --- suffer a disproportionate burden because extreme heat is much worse than moderate heat for crop yields.

Egypt, like many other countries, faces the challenge of how to adapt the agricultural sector to climate change. Along these lines, Egypt’s Ministry of International Cooperation has launched the Nexus of Water, Food, and Energy Program (NWFE) to implement multiple projects targeting the water, food, and energy sectors. These investments have the potential to support programs and policies that trigger climate adaptation.

But the policy approaches that we take should be based on rigorous on-the-ground evidence that studies both the efficacy of adaptation investments and the incentives that farmers face to take them up. The Abdul Latif Jameel Poverty Action Lab (J-PAL), where I am an affiliated professor, works with governments across the world to carry out randomized evaluations studying what works and what does not work for lowering poverty.  Since 2020, J-PAL’s Middle East and North Africa office (J-PAL MENA), based at the American University in Cairo, has been working with the government, NGOs, and other stakeholders to use randomized evaluations for informing policy in Egypt and the broader region. The current climate crisis presents a similar challenge. Namely, what policies and interventions protect agriculture the most from climate change? 

Our goal should be to direct our adaptation investments to policies with proven effectiveness. At the same time, on the ground randomized evaluations should be done to tell us which of our adaptation investments are achieving the desired objectives. If not, we run the risk of investing in interventions that are either ineffective or less effective than competing approaches. Examples from randomized evaluations in several countries show how new technologies and policy interventions can lead to climate adaptation, particularly when there is an enabling environment that creates incentives for climate adaptation. In addition, these studies can tell us when something does not provide the adaptation benefits that we expected.    

As one example, some of my work in Bangladesh studies Alternate Wetting and Drying (AWD), a water-saving technique in rice irrigation that is included as part of the country’s Nationally Determined Contributions. The technique serves as both adaptation and mitigation because in theory, it preserves water and reduces methane emissions from rice paddies (which are flooded fields used to cultivate rice). Significantly, the study finds that the AWD technique --- utilizing a simple plastic PVC pipe --- lowers groundwater use by about 20%! However, there are clear differences across farmers. Farmers who pay for water by the hour of pumping face incentives to adopt and save water. Farmers who do not pay for water, or who buy water with seasonal charges, do not adopt, and do not save water. In short, the benefits of climate adaptation were only realized in certain areas where there was the right enabling environment.

This example of efficient water use in agriculture has relevance in Egypt. Research suggests that by 2100, the annual flow of the Nile could become more variable by 50%. This greater variability creates a need to maximize the efficiency of irrigation water use. But interventions that on the surface would seem to achieve this do not always preserve water in farmer’s fields. Farmers seek to maximize the profits from cultivating their land. Forms of climate adaptation that immediately contribute to their bottom line will likely deliver the greatest impact. Policy decisions in this space should be guided by randomized evaluations that rigorously measure which investments meet adaptation goals of more efficient water use.

Extreme weather is predicted to become more common as the climate changes. Improved seed varieties that are more tolerant to droughts and floods provide a technological mechanism for farmers to adapt to climate change. Some of my work in India finds that new flood tolerant rice seeds from the International Rice Research Institute reduce yield losses during flooding. They also increase the investments farmers make on their plots. Researchers have found that drought tolerant maize seeds, when coupled with weather insurance, make farmers in Mozambique and Tanzania more resilient to bad weather events. 

But technologies for adaptation do not get adopted fast because it takes farmers time to learn. Some interventions help farmers learn and adapt faster. Others do not. It is not enough to assume that climate adaptation will happen automatically when lab results with new technologies are promising. These new technologies need to be tested with experiments with actual farmers. Evaluations often tell us more about how to improve learning and speed the uptake of adaptation technologies.  

As COP27 arrives in Egypt, climate adaptation will continue to be an important part of the discussion, as it has been in previous years. Like many countries, Egypt faces important challenges on how to direct investments targeted towards adaptation. The experience from other countries suggests that there is no one-size-fits all approach to climate adaptation. Rather, policies to trigger adaptation need to be evaluated in real-world settings with farmers.

*The writer is affiliated professor at The Abdul Latif Jameel Poverty Action Lab (J-PAL), and an associate professor of economics at Tufts University, USA​

Short link: