Agricultural Assistance and Policy Feedback: Experimental Evidence from Farmers

• In Event: Experimental Methods in Public Policy

Abstract

Research on policy feedback shows that the design of public policies can bolster public support for those policies, increase political engagement, and improve citizens’ views of government. However, the lessons that can be drawn from this literature are currently limited in three important ways.

First, much of this literature relies on surveys to both 1) determine which citizens received which government benefits and 2) ask citizens for their views about government and the programs providing the assistance. This may be problematic. Individuals who remember having received government benefits may also be the ones who have more positive views about those programs—whether because the benefit influenced their views, or because preexisting favorability toward the program made them more likely to remember receiving the benefits. Further, other research designs (e.g., examining county-level variables) require making ecological inferences.

Second, research on the political effects of public policies has mainly focused on broad social welfare programs such as Social Security and the major healthcare programs. This is an important set of policies, but the U.S. government does more than provide Social Security, Medicare, and Medicaid. It also provides assistance to citizens who are probably very different in key respects than, for example, the typical recipients of Medicaid.

Third, existing research focuses heavily on how policies affect public support for those very same policies. Only some research investigates their effects on people’s views of government. And almost no research evaluates how policy benefits in one area affect citizens’ views on policies in another area—and there is good reason to think such spillover effects might exist. Citizens tend to respond differently to abstract questions about principles than they do to concrete questions about specific policies. And so while a citizen might report a negative, pessimistic view of government in general, she might also be positively inclined toward specific government programs. Thus, there is much to be learned about how receiving government assistance of one kind affects individuals’ views about other government programs.

In this paper, we evaluate how receiving agricultural assistance affects recipients’ views of agricultural assistance programs, their views of U.S. government generally, and their views on other issues, in particular climate change and trade. Not only will our study address all three of the aforementioned limitations of the existing policy feedback literature; it will also shed new light on how a wide range of government beneficiaries—Republican, Democrat, rural, and urban—view different policies as well as government itself.

To do so, we are conducting a large-scale mail survey of a unique and theoretically relevant population: American farmers who have received some kind of assistance from the U.S. Department of Agriculture between 2012 and 2019. As far as we know, no previous large-scale survey effort has ever been conducted on this population. Further, we are administering a randomized experiment embedded in the survey. This, by itself, is a major step forward from existing research, because we can track which individuals received which benefits rather than relying on respondents to remember and report the information accurately. It is also an advance in that the programs and population being studied differ from those typically covered in studies of policy feedback.

The survey instrument contains an experiment in which we remind some respondents that they received agricultural assistance, complete with the amount they received. Other respondents are assigned to either a control condition or a statement about the importance of agricultural assistance more generally. We then ask respondents a series of questions about their support for and favorability toward government activity—with the subject matter ranging from general (trust in government) to specific (views toward agricultural assistance, trade policy, and climate change). We also will examine the effects of assistance on vote intention, speaking to the literatures on retrospective voting and clientelism.

Our analysis will assess whether respondents’ views depend on whether they are reminded they received agricultural assistance, and also how their responses vary by the amount and type of assistance they received, their reported party affiliation and ideology, and characteristics of the farms, including how rural they are, whether they have been affected by natural disasters or trade shocks, and the partisanship of the county.

This paper will therefore answer a number of questions about policy feedback that build on what we know from existing research. Especially in today’s political environment, it is important that we advance knowledge about how the design of a wide range of public policies might affect citizens’ views of government and their support for public programs.

Authors

• Sarah F. Anzia, University of California-Berkeley

• Jake Alton Jares, Stanford University Graduate School of Business

• Neil Malhotra, Stanford University