Gluten aversion is not limited to the political left

Abstract

Despite a heightened political discourse surrounding food choices, few studies have identified connections between political beliefs and consumer perceptions. Using gluten as an example, this article identifies how political opinions relate to opinions of food products. If an avoidance of gluten is a biological condition and not a social construct, there should be no correlation between political opinions and gluten avoidance. Our study uncovers a complex relationship between the social construction of gluten avoidance and the potential role of political views. Perhaps most surprising, we find that supporters of Donald Trump are more likely to identify as avoiding gluten, relative to non-supporters. Findings suggest that future research might benefit from considering the political beliefs of consumers when estimating models of food demand.

Appendix

Although data on aversion to gluten and political affiliation is easily collected, interpreting any correlation between the two is difficult because there are multiple reasons why a correlation might exist. This section describes four stylized links between political affiliation and gluten-avoidance. Each link is an event or series of events that causes one to choose a specific political affiliation and begin to avoid gluten. The event could be a personal experience like a sickness, exposure to new information, or parental influence. The event might occur during the person’s life (a contemporary link), or before the person was born (historical link). The distinction between contemporary and historical events do not have to be during or before a person’s birth, but could instead be interpreted as demarcating a recent versus a past event. Also, the link might be mediated in that the event causes one to adopt a political view first, and that political view then induces them to form a specific opinion about gluten, or it might be unmediated in that the view on gluten is formed separately from the view on political issues, but both are ultimately caused by the same event. See Fig. 4 below for these four categorization of links as they relate to liberal a belief system. They are best understood with a series of examples. These specific examples are contrived to show how someone could become more liberal and begin avoiding gluten. Of course, the opposite examples might show how someone could become more conservative and begin consuming gluten.

Fig. 4

Four conceptual links between political ideology and gluten avoidance

Link A Suppose an individual develops a preference for a particular form of wheat bread, and soon after experiences poor health. Regardless of whether the health symptoms are caused by the gluten, the person may suspect that gluten is to blame for their poor health, and they thus develop an aversion to gluten. As they make repeated trips to the doctor and undergo expensive medical tests and scans, their growing medical bills makes them alter their views on health care, a highly politicized issue. Now desiring government-funded health care, they become more liberal. Both the aversion to gluten and embrace of liberal politics happened at roughly the same time. In this instance, politics did not cause gluten avoidance even though politics and perceptions of gluten are correlated. The link connecting gluten perceptions and political affiliation is poor health, and the link is contemporary to the individual’s life. It is also unmediated, in that the link helps form both political affiliation and gluten-avoidance directly. That is, the sickness would have led to more liberal political beliefs even if the person did not believe the sickness was caused by gluten.

Link B A person is raised in a conservative home but then attends college where they are exposed to and become attracted to liberal views. This happens before the person has heard of gluten-sensitivity. As they begin socializing in liberal circles, they meet an increasing number of people claiming to be gluten-sensitive. According to Identity Theory, the person will seek to verify their new political identity to both themselves and others (Burke and Stets 2009), and begin to prefer gluten-free food like their new friends. That practice is related to the economic theory of signaling, which suggests that people might engage in practices in order to gain a within-group advantage (Spence 1973). The link here is exposure to new ideas. It is a contemporary link, but it is mediated in the sense that the only reason they avoid gluten is because they adopted new political views. Without a new political outlook, their perceptions towards gluten would have remained the same.

Link C This individual is raised in a liberal household, and due to exposure to their parents adopts liberal views as well. If liberals are more likely to claim gluten-sensitivity, the individual may make this claim as well in the self-verification process. Like Path B the link is mediated, in that the only reason they avoid gluten is because they first chose liberal politics. The link is also historical, though, as their parents’ political identity (assuming they do not change political views after having children) was caused by the events in the parents’ lives, from before the person was born.

Link D Perhaps there are a variety of genes increasing the odds one is both a liberal and sensitive to gluten. The link, genetics, is direct in that it causes both, but is unmediated in that the individual does not need to be exposed to liberal views first to develop an aversion to gluten: it was inherent in their genes. This differs from Link A in that Link A is assumed to be a recently contracted illness whereas Link D is assumed to be a health problem the person experiences throughout their life, and caused by an event prior to their birth.

Any correlation between political affiliation and gluten sensitivity could be the result of any or all of these paths—or numerous paths not listed. Moreover, there could be many different varieties of any one link. For example, a person may be exposed to multiple events inducing them to adopt a liberal political view, such that there could be a dozen different types of Path B. The figure above described Link D as activated by genetics, but it could be activated by the happenstance of the region one was raised.

See Tables 4, 5 and 6.

Table 4 Regression results used to create the control function instrument

Table 5 Model estimates for political ideology expressed separately for economic and social issues where the dependent variable is the constructed gluten avoidance scale

Table 6 Model estimates for presidential favorability where the dependent variable is the constructed gluten avoidance scale (N = 1086)