- Charles Dahan, Ph.D. Candidate, UF Department of Political Science
In June 2014, the Pew Research Center released the most comprehensive study to date of partisanship in America. Their findings confirmed widely held beliefs about the expanding partisan divide, with pessimistic predictions, including, “increasing ideological uniformity and partisan antipathy affects politics, compromise, and everyday life.” Moreover, the most partisan Americans were the ones who participated most in politics and to whom politicians were most responsive, while these individuals also narrowed the definition of ‘compromise’ to mean their side getting, “more of what it wants.”
The federally funded Supplemental Nutrition Assistance Program (SNAP) offers nutrition assistance to millions of eligible, low-income individuals and families. The program is commonly known as “food stamps,” although in Florida the benefits are now delivered via a plastic Electronic Benefits Transfer (EBT) access card that looks like a standard credit card. Food stamp expenditures have been regularly shown to offer a positive return on investment, with over $1.70 in positive economic impact for each $1 spent.
Both conservatives and liberals have made arguments in favor of these expenditures on economic grounds, most famously during the 1990s with landmark legislation spearheaded by GOP presidential nominee Bob Dole. Yet contemporary rhetoric regarding food assistance programs has led the issue serving as a partisan heuristic; the invocation of “food stamps” may serve as a cue triggering partisan identity, leading to information on the issue having opposite effects for Democrats and Republicans. Moreover, ideological and social identity priming—in the form of one’s individualism and the importance one’s religiosity plays in their daily life—may affect social welfare preferences, as well as potentially activate partisan identification, altering such preferences.
BEBR Survey: Findings Similar, Different Conclusions
Recent surveys from YouGov find, “40% of Americans [want] the food stamps budget decreased, [and] were outnumbered by the combined 48% who said spending on food stamps should either be increased (24%) or kept the same (24%). This conclusion frames the study to be interpreted that Americans tend to be in favor of expanding food stamp expenditures, when the opposite is in fact true.
Respondents during BEBR surveys conducted from July 1st to August 30th were asked, “Do you think the amount of money the government spends on food stamps should be significantly increased, increased, kept the same, decreased, or significantly decreased?” This study finds, at least amongst Floridians, to misidentify the appropriate story told from the data: the intensity of Americans to decrease food stamp funding far outstrips any desire to increase funding. In fact while it is true a significant number of respondents (27.2% in the BEBR studies, 28% in national survey) wish to maintain the current level of food stamp spending, far more Americans wish to significantly decrease expenditures (21.2%) than significantly increase expenditures (5.1%). Moreover, 44% of Americans wish to decrease or significantly decrease funding, while only 29% wish to increase or significantly increase funding. Thus, by a large margin (71% vs. 56%) the appropriate interpretation of both surveys is that opposition or ambivalence to food stamp spending is greater than any desire to increase spending.
The story, however, is more complicated when breaking down respondents by party identification. As expected, partisanship is the primary explanatory variable.
Most sophisticated voting models, popularized by sites such as Nate Silver’s 538 blog and Washington Post’s Monkey Cage, include a retrospective evaluation variable, such as retrospective economic evaluation. While they should affect evaluations of an incumbent’s job performance, retrospective evaluations—such as are you better off economically than you were a year ago—should not logically play a role in future expenditure preferences.
Moreover, not only do Republicans and Democrats hold opposing views of food stamp expenditures, but perceptions of the trajectory of the economy—that is, economic optimism—cause Republicans and Democrats to reach different conclusions about their preferred levels of food stamp expenditures. While Democrats who view the economy as improving believe expenditures should increase, Republicans believe expenditures should be reduced as their consumer confidence rises.
The greater importance religion plays in an individual’s life, the more likely are they are to support cuts to food assistance expenditures. Even controlling for partisanship, when respondents state their religious faith plays a greater role in their daily lives, their support for food stamp expenditures decreases.
Interestingly, however, when individuals think about their religiosity, their support for food stamp expenditures increases.
In addition to all respondents being asked about their opinion regarding preferred level of food stamp expenditures, the respondents were broken up into three groups: a control group that was only queried about their preferred food stamp expenditures; a treatment group that was primed with an individualism-communitarian scale question; and a treatment group that was primed with a question regarding their religiosity.
Priming is a common technique in social science experimentation, defined as, “activating particular representations or associations in memory just before carrying out an action or task.” Individuals do not have set preferences; rather, an individual’s preferences are tied to the parts of their identity with which they associate the question. For example, if a respondent attended a parent-teacher conference with an admired instructor prior to being asked about their preferred level of education funding, they are likely to answer differently than if they are asked immediately after filing their taxes.
Many priming effects—particularly those related to feelings—have been thoroughly investigated in both social psychological and political science literatures. However, what has not been explored is the manner in which political partisans react to questions regarding policy preferences when primed.
Simply being religiously primed appears to decrease support for food assistance expenditures, again controlling for partisanship. This effect holds for all respondents (p = 0.09) as well as exclusively Republicans (p = 0.07), despite very small sample sizes (271 and 86, respectively). Religious priming does not appear to affect Democrats’ support for food stamp expenditures.