
How can better monsoon forecasts help farmers in India to adapt to climate change?
A study in India shows that accurate long-range forecasts can help farmers make informed farming decisions, and improve their productivity and wellbeing. Providing better access to these forecasts can enhance adaptation to climate change among farming communities around the world.
Climate change is making weather more variable, with rainfall patterns becoming less predictable and extreme temperatures occurring more frequently. Agriculture is particularly sensitive to these changing conditions, jeopardising the livelihoods of 67% of the world’s poor populations who depend on agriculture for survival. Effectively coping with this variability will be essential as the world continues to warm. Highly variable weather makes it challenging for farmers to tailor their planting decisions to the coming season, leading them to underinvest in things that could increase their profits, like high-yielding crops, fertiliser, labour, and more. By giving farmers a preview of the conditions of the coming growing season, accurate long-range forecasts have the potential to help farmers make the best possible investments. This study presents the first experimental evidence of the impact of such an accurate long-range monsoon forecast—a new climate adaptation technology—on farmers’ behaviour and well-being.
Studying the impact of monsoon forecast information on farming and investment decisions
We study a forecast of one key piece of information for Indian farmers: the timing of the onset of the Indian Summer Monsoon. Even though nearly two-thirds of the global population live in monsoon-affected regions, predicting when monsoons will arrive has remained elusive. To test whether information about the onset of the monsoon would change farmers’ behaviours, we used a new monsoon forecast developed by the Potsdam Institute for Climate Impact Research, one of the only forecasts able to predict monsoon arrival more than a month before onset accurately.
Specifically, we study 250 villages in Telangana, India, where more than half the labour force are farmers. To test the impact of forecast information on farming, we randomly assigned the 250 villages to three groups:
- a group that receives forecast information at least one month in advance of the monsoon season
- a group that receives insurance (used as a benchmark to compare willingness to pay for forecast information)
- a comparison group that does not receive the forecast
They tracked how the forecast information impacts the farmers’ beliefs about the monsoon onset pre-harvest, their up-front investment decisions, their wellbeing at the end of the growing season, and how their behaviour compares to those who received the insurance.
Findings
Farmers disagree widely about when the monsoon will start, highlighting the need for forecasts.
The timing of the monsoon is extremely important for agriculture, but most farmers do not have access to accurate information about its onset. The majority of farmers in the study were getting their information from other farmers, with only a small minority relying on the government or other services for information on when the monsoon would begin.
We first measured farmers’ own predictions about when the monsoon would begin before providing the forecast information. During this initial visit, the farmers’ predictions varied widely, even within villages. The more optimistic farmers believed the monsoon would come about 2.5 weeks earlier than the more pessimistic farmers. The stakes are high - an earlier monsoon typically means a longer growing season, suited to cash crops like cotton, while later monsoons are generally worse, forcing farmers to grow lower-value subsistence crops.
Farmers are persuaded by forecasts and find them valuable.
During the year of our study, the forecast predicted an average monsoon. As a result, around a third of the farmers’ initial beliefs turned out to be overly pessimistic. The forecast brought these farmers the “good news” that the monsoon/growing season would be longer than they had originally thought. Conversely, a third of the farmers’ initial beliefs were overly optimistic and they received “bad news” from the forecast - the growing season would be shorter than they expected. The final third of farmers had initial beliefs that turned out to be correct.
After we shared this forecast information with the farmers, we investigated whether their beliefs had changed. We found strong evidence that the farmers changed their beliefs to be more in line with the forecast. Moreover, the farmers showed that they highly valued this information. They were willing to pay as much for the forecast as for an insurance product that pays out approximately 20% of average crop revenue under a late monsoon.
Farmers change their farming behaviours according to the forecasts.
The forecast went beyond changing beliefs and had a large impact on the farmers’ behaviours. Overly optimistic farmers, for whom the forecast was “bad news” of a shorter-than-expected growing season, took steps to cut down on investments by reducing the amount of land they cultivated by 22% and were 32% less likely to add a new crop. They also reduced their expenditures by 10% —driven largely by cutting the amount of fertiliser they bought by 30%.
On the other hand, overly pessimistic farmers, for whom the forecast was “good news” that the growing season would be longer than expected, increased investments and expenditures. They increased the land they cultivated by 15%, were 14% more likely to add a new crop, and were 16% more likely to plant cash crops. They spent over a third more on up-front investments in total.
Figure 1: Impact of forecasts on farmers' farming and investment decisions.

Changes in agricultural investments lead to changes in agricultural outcomes and wellbeing.
Overly optimistic farmers who received “bad news” and reduced their farming investments saw their agricultural output and sales decline by 27% and 20%, respectively. Their agricultural profits also declined. However, these farmers tended to find other ways to make money (four out of seven of these farmers were more likely to own a non-agricultural business) and cut their debt in half, leading to net savings of more than USD 560 per farmer. They also increased their business profits by 47%. Together, these results imply that the forecast likely made these farmers better off - instead of doing unprofitable farming, they were able to diversify their activities.
In contrast, the forecast caused overly pessimistic farmers to do more farming. This led to a 22% increase in agricultural production but did not qualitatively impact profits on average. The forecasts also appear to have benefited the poor farmers the most. Farmers who were less well off at the beginning of the study had the largest increase in profits as a result of receiving the forecast.
Figure 2: Net savings for farmers after being guided by forecast information.

Insurance encourages optimistic farmers to invest more but does not guide smarter choices.
While accurate forecasts inform farmers about what is coming, allowing them to tailor investments to the growing season conditions, agricultural insurance—a widespread policy around the world—lowers farmers’ risk exposure but does not improve their information.
Overall, farmers who received insurance increased the land they cultivated and the amount they spent on up-front investments like seeds and fertiliser by 12% compared to those who did not receive insurance or forecast information. This is because overly optimistic farmers (incorrectly) believed it would be a good year. Given the safety net the insurance provided, they responded with a large increase in investments—even though the forecast would have caused them to reduce investments instead.
On the other hand, overly pessimistic farmers for whom the forecast would have been “good news” did not meaningfully change their investments (though it would likely have been beneficial to do so) despite having the insurance. These findings underscore the fundamental differences between insurance and forecasts and suggest that they could be used as complimentary climate adaptation strategies. Forecasts let farmers make the right investments for the coming year, while insurance protects farmers against downside risk.
Need for improving forecasts that can enable farmers to make informed farming decisions
Long-range forecasts enable farmers to make the best possible decisions about whether to plant at all, how much to plant, what to plant, and how to make adjustments across crops by providing critical information about the coming growing season. As such, providing farmers with better access to forecasts can help them adapt to the less predictable conditions that come with climate change. However, forecasts with proven accuracy are rare and often inaccessible. Improving forecasts and making them available to farmers could increase the wellbeing of farmers and boost the economies of farming communities around the world.
A version of this article was first published by EPIC.