Recent debates on the US farm bill have me going back to some data I collected in 2010.  Rather than discussing the details of the 1,000+ page farm bill, let’s ask a basic question.  Who supports farm subsidies?

Along with a team of other academics we fielded an internet survey through the Cooperative Congressional Election Survey on support for US agriculture policy.  We asked two questions about agriculture policy in the United States to a representative sample of 1,000 US residents.

The 30 second background on US agriculture is:

1. The US “Farm Bill” includes a lot of things that aren’t directly related to farming.  One of the largest components of this bill is the food stamp program.  The decision to include food stamps into the bill began in the 1970s and is well documented by John Ferejohn.  But the simple logic is that you can build a coalition of representatives in Congress from both urban and rural districts by linking these policies.  Ezra Klein has a pie chart the distribution across programs in the proposed bill.  [Although I have concerns about these projections since many are based on expected prices of crops, the number of people on food stamps, and ignores any supplemental farm legislation that adds more money into the pot down the road.]
2. While the US provides high levels of financial support to agriculture, less well known by most Americans (including academics) is that the US actually provides lower levels financial (and trade protection) support than most of the other developed countries.  A great book on the topic here.

In our survey we utilized the logic of a survey experiment, changing the “frames” or “primes” that respondents are exposed to.  The logic is that by exposing respondents to a quick story on US agriculture in one way or the other or changing a word here or there, we can examine how support for agriculture policy varies.  We specifically tested how the inclusion of food stamp programs affect support for agriculture policy, and how information on the relativelevel of US agriculture subsidies affects public support.

In this blog post I’m going to focus on the question of how the inclusion of food stamps affects public support for US agriculture policy.

In this question we asked a random sample of ½ of the 1,000 respondents question A (including the words “food stamps”):

Q1A:  The U.S. government provides a number of different types of support to farmers and the agriculture industry.  The last farm bill included financial payments to farmers, food stamps, and limits on foreign agriculture imports.

The other ½ received question B (excluding the words “food stamps).

Q1B:  The U.S. government provides a number of different types of support to farmers and the agriculture industry.  The last farm bill included financial payments to farmers and limits on foreign agriculture imports.

Then all respondents were asked:

In general, how supportive are you of these types of policies?

1. Very supportive
2. Supportive
3. Against
4. Strongly against
5. Don’t know

The difference between the two questions is only one word.  Respondents were randomized and the groups receiving the “food stamp” wording were roughly the same as respondents who received the control of no “food stamp” language.

Ignoring the differences between the food stamp and non-food stamps treatments, what do the responses look like in general?  Here is a quick table.

Support for US Agriculture Policy

Democrats are more supportive of ag policy than Republicans.  But for all groups, current US agriculture policy has majority support.

What happens when we include or exclude food stamps?  In aggregate, very little.  Support for US agriculture programs (summing Very Supportive and Supportive) is about 70% when food stamps are excluded, and about 71% when food stamps are included.

More interesting, and in some sense obvious, is how this affects the most partisan of voters.  When food stamps are included in the question, support by Democrats increases an additional 7%, from 77.67% supporting to 84.30% supporting.  Republicans are the mirror image.  Support drops a little under 7%, from 59.5% to 52.9%.  While independents are in the middle of the road in their support for agriculture programs (64%) they behave a lot more like democrats when we include food stamp programs.  Support increases 71%.

What does this exercise tell us?  Two things.

1. The inclusion of food stamp programs not only helps build a coalition in Congress, it also has some (partisan) impact on support for the farm bill.  But the tradeoff is that increased support by Democrats and independents are countered by a decrease in support by Republicans. The net impact is quite small.
2. No matter how we cut it, agriculture policy is quite popular.  Even amount Republicans, when we include the food stamp language, support remains above 50%.

In my next blog post I’ll write up the summer of a little experiment on framing this in competitive terms affects support.  Spoiler alert.  It matters.  A lot.