As noted, India has three key agricultural programs. In the first program, the government sells highly subsidized fertilizers to farmers. Under the second program, which also directly supports farmers, the government buys a substantial share of key crops like wheat and rice at attractive prices known as Minimum Support Prices (MSP). Finally, in the third program, the government sells these grains to households at heavily subsidized prices through a public distribution system (PDS). Together, the three programs cost about 1.2% of India’s GDP and impact nearly 800 million people.
"Once I started thinking about applications of IO tools for answering development-related questions, it was hard to think about anything else"
Utilizing a structural model of supply and demand and large datasets on these three Indian programs, Saxena and his coauthor were able to analyze how the programs interacted and assess their overall impacts on farmers and households. They found that the programs, when taken together, were beneficial to low-income consumers. In particular, they determined that the MSP procurement program reduced household food spending by 15-20%. Likewise, they observed that the programs benefit many farmers – but that these benefits are primarily enjoyed by large-scale farmers, who are more able to access the programs. They found that the MSP procurement program, for example, benefits the largest farmers while slightly harming small-scale producers who are less able to find and secure government buyers.
Many of these results reflected what economists call “equilibrium effects” – or economic effects that only become visible when implemented widely, such as through policy programs. By raising crop output through its fertilizer subsidies, for example, the government would have more crops to purchase from farmers through its MSP procurement program and more food products to sell to low-income households through its PDS program. This would decrease low-income households’ reliance on private agricultural markets, however, lowering prices and thus farmer revenues in those markets - and since, as noted, small-scale farmers have less access to MSP procurement, they are more exposed to this potential adverse effect.
To account for these equilibrium effects, Saxena and his coauthor’s model had to consider the decision-making processes of individual farmers and how those decisions affect the industry as a whole. For any given farmer, their model factors in their choice and share of which crops to plant, the level of fertilizer subsidy, the prices they expect to earn through MSP procurement, and the price they expect to earn in private markets. By aggregating these choices across the entire industry, the researchers could assess how the three programs interact with each other – and whether any inequalities or inefficiencies hindered their implementation.
For example, there are large differences in the geographic distribution of these programs. Some regions – like Saxena’s home state of Uttar Pradesh – see less than 5% of rice and wheat being sold to government buyers through the MSP program. Farmers in states like Chhattisgarh or Punjab, by contrast, sell 30-50% of their output through MSP procurement. While the reasons behind these regional dynamics need to be explored in greater depth, Saxena notes that they may stem from regional differences in the availability of government buyers and information about the program. This highlights a central challenge with MSP procurement, which has been a source of political tension in recent decades: large-scale farmers who know the program, have access to government buyers, and can transport their crops to government warehouses often reap the lion’s share of the program’s benefits.
The theoretical nature of their empirical model also allowed Saxena and his coauthor to play around with different program configurations. They were able to analyze, for example, how the overall effects would change if one or more programs were increased, decreased, or eliminated altogether. These findings have broad policy implications – which, as Saxena notes, ultimately depend on what the goals of the programs are.
“When these programs started in the 1960s, the objective was food security,” he said. “In recent years, the focus and rationale has been farmer welfare.”