Liberals and Conservatives Differ in their Perceptions of Humor
New research shows how people judge jokes in line with their ideological positions. But is everything as it seems?
In recent years, a number of high-profile comics have stopped playing gigs on university campuses, citing concerns over political correctness and a lack of life experience as reasons why. At the core of these concerns is that comedians who tell jokes that are “close to the bone” risk reprisal, and a subsequent loss of air time and, ultimately, earnings.
Commentators on this issue have suggested a potential ideological reason as to why some jokes (and their comedians) might be subject to close scrutiny. Those who make jokes about perceived lower-status groups (such as women, LGBT+ groups, black and minority ethnic individuals) have faced the most criticism, with these target groups being those chosen for greater protection by those on the political left. This is consistent with a broader societal “cancel culture” that has taken hold since 2018, whereby those acting counter to the ideological status quo in academic and liberal arts contexts have been removed from broadcasts or prevented from presenting.
Until now, there has been a lack of research into ideological judgments of stereotypical jokes. Some research has found that women who identify more strongly as feminists are more likely to judge jokes about men as being funnier than are women who do not identify with the feminist label. Stereotypical jokes about ethnic minorities have also been found to prime discriminatory judgments about them, such as more exaggerated perceptions of criminality in a mock criminal case.
A study published earlier this year, led by Dean Baltiansky under the supervision of noted social psychologists John Jost and Maureen Craig, investigated whether people’s system justification orientation was associated with their judgments of the funniness of jokes about high- and low-status groups. It was recently published in the journal ‘HUMOR’.
System justification refers to the tendency to uphold social hierarchies and structures. On the standard measure of system justification, those scoring high will agree with statements such as “In general, I find society to be fair,” while those scoring low are more likely to endorse statements such as “American society needs to be radically restructured.” High levels of system justification are associated with a greater propensity to self-label as politically conservative, while low levels are linked to political liberalism.
In this study, the authors began with two simple hypotheses in line with system justification theory:
Those scoring high on this index were predicted to find jokes about low-status groups funnier than jokes about high-status groups. This is a motivated response designed to maintain social hierarchies, disparaging those at the bottom of society and defending those at the top
Those scoring low on system justification were predicted to find jokes about high-status groups funnier than those about low-status groups. This would also be a motivated response, but this time designed to bolster the position of low-status groups and bring down those at the top of the social ladder
A total of 179 Americans took part in the study, hearing jokes from high-profile comedians about a range of high- and low-status groups. They were asked to judge how funny each joke was, as well as completing the system justification scale mentioned earlier. Importantly, all of the jokes were presented in an audio-only format to blind the participants to the racial background of the comedian.
Baltiansky et al. (2021) reported two significant main effects and one interaction in their data. The first main effect related to participant sex, with men rating jokes (regardless of their target) as funnier than women. The second main effect related to the targets of jokes, with jokes about low-status groups being rated as less funny than jokes about high-status groups.
The interaction within their analysis came between ‘joke target’ and ‘system justification score’. This means that the extent to which somebody scored high or low in relation to system justification changed how funny they found jokes about high- and low-status groups to be. Interpreting this, Baltiansky et al. (2021) said:
The results of this study suggest that the tendency to appreciate stereotypical jokes perceived as targeting high vs. low status groups is associated with individual differences in system justification. Specifically, we obtained consistent evidence that high system justifiers found jokes disparaging low status groups such as women, poor people, African Americans, Muslims, and Mexicans to be funnier than did low system justifiers. To a much lesser extent , we also observed— in at least a few cases— that low system justifiers were more likely than high system justifiers to appreciate jokes targeting high status groups such as European Americans and rich people.
I’ve added some emphasis in the above quote to highlight exactly what the authors are arguing. I’ve also created a modified version of the interaction graph provided by Baltiansky et al. (2021) in their original paper:
The interpretations presented by the authors do not explicitly link to the data as they are presented. Correlational data (looking at the association between system justification and joke funniness) are interpreted in the context of group-based differences, with the conclusions being drawn seemingly driven by a desire to tell a particular story about the nature of stereotypical humor perception.
Let’s revisit what the authors said.
Conclusion 1: high system justifiers found jokes disparaging low status groups […] to be funnier than did low system justifiers
This appears to be true. The line for funniness ratings about jokes targeting low-status groups ends higher on the right hand side of the graph (among high system justifiers) than on the left hand side of the graph (among low system justifiers).
Conclusion 2: to a much lesser extent […] low system justifiers were more likely than high system justifiers to appreciate jokes targeting high status groups
Again, this seems correct. The line showing funniness ratings for jokes about high-status groups is slightly higher on the left hand side of the graph (for low system justifiers) than on the right hand side (for high system justifiers), but the slope of the line is barely noticeable.
So… what’s the problem? This is where my annotated graph comes into the story.
You’ll see that among high system justifiers (on the right hand side of the graph) within the circle labeled ‘Section B’, all of the data-points are roughly the same. That is, the gray bar around the lines (signifying the confidence interval of probable true values) are all around the same. This is even the case for jokes about high-status groups, where the confidence interval is close to those for the other joke types.
The part of the graph labeled ‘Section C’ shows how all of the data points on the right hand side of the graph are roughly equivalent to those on the left hand side in relation to judgments of jokes about high- and mixed-status groups. This then leads us to ‘Section A’ on the left hand side, which shows how the only errant data-point is in relation to jokes about low-status groups. Here, it appears that these jokes are viewed as markedly less funny than the other joke types.
I have worked with a colleague of mine (Dr. Harry Purser) to re-analyze the Baltiansky et al. (2021) data. Consistent with our suggestions, we find that high system justifiers did not make significantly different judgments about the funniness of jokes about low-, high-, or mixed-status groups. However, we did find that scoring low on the system justification scale led to substantially lower funniness judgments about jokes targeting low-status groups than jokes about high- or mixed-status groups.
Baltiansky et al. (2021) suggested that their data reflect a tendency among high system justifiers to find disparaging jokes about low-status groups to be more funny, comparative to low system justifiers. Although this appears to be an accurate description of the general relationship between system justification scores and judgments of the funniness of these jokes, the framing of this argument is misleading. This relationship is driven not by exaggerated funniness judgments among high system justifiers (typically conservatives), but by exaggerated judgments of their lack of funniness among low system justifiers (typically liberals).