So you may be wondering why it has taken me so long to post the results of the Justice and Mercy survey. The truth is, the results themselves were uninteresting.
Negative results all round, almost astonishingly so. There was no association between any of the demographic factors and mercy. There was no association between your perception of your past or current behaviour and mercy. Disturbingly for the validity of our our measure, there was no association between tendency to condemn and total mercy.
With a total of N=181 respondents to the final version of the survey, I would have thought we had plenty of power to detect effects worth detecting at critical threshold P=0.01, but alas no.
Our measure of mercy did, at least, predict preference for punishment harshness for all categories except wealthy offenders, including poor offenders, juvenile offenders, murderers, rapists, thieves, illegal drug suppliers and those guilty of assault.
The following hypotheses were not sustained:
“Mercy will be associated with both a lower tendency to condemn and a lower ideal tendency to condemn- validating it as a measure of mercy in a social context.”
“Heavy use of social media will be associated with a tendency to condemn.”
“The worse you view your own past behaviour, the more inclined you will be towards mercy.”
“The worse you view your own current character, the more inclined you will be to mercy”
“Belief that you might have committed a horrific crime had circumstances been different will be positively associated with mercy”
“When controlling for degree of mercy, being left-wing will make you more likely to endorse harsher punishments of rapists, and when controlling for degree of left-wingedness, being merciful will make you more likely to endorse”
“The overall (not controlling for mercifulness) relationship between leftwingedness and attitudes towards sexual assault perpetrators will vary with gender. Being leftwing for men will, overall, reduce the desire for harsh punishments for sexual offenders whereas being leftwing for women will, overall, increase the desire for harsh punishments of sexual offenders”
“Lower log income will be associated with greater mercifulness, control for education and age.”
“Education will be associated with mercy, controlling for income and age.”
Most of these hypotheses weren’t even close to significance.
Pretty much the only association that did show up was a positive association between mercy and self-rated left-wingedness. This is scarcely surprising, as many of the questions about mercy and punishment are directly related to left-wing policy positions- rehabilitation, restorative justice, leniency in juvenile justice and so on. Leftwingedness was associated with greater mercy at a rate of about R^2=17% P=<0.001. A reasonably sizeable relationship for individual level effects in the social sciences, though far from enormous.
I also verified that in our sample: “Even adjusting for age, heavy social media users will be more left-wing”, P=0.008. It is important to note however that this may be a quirk of our sample, since some participants were recruited via social media, and given my social media connections it is probable they were unusually leftwing.
So what remains to be said?
First of all I am very glad to have ran this survey, in which a total of 259 people participated and thought about politics, justice, mercy, redemption, religion, ethics. The ideals of mercy and forgiveness are conceptually tied to religion, and I worry that this means we’ve ceased to think enough about them as our society has embraced secularism. Secularism seems unlikely to end anytime soon so we need to find new secular ways to imagine mercy. Until very recently the United States had an incarceration rate of 1%. Any “better way” than this is going to require us to think through justice, mercy and punishment.
Second of all, even though I failed to find significant trends, some of these questions are at the heart of our understanding of interesting social questions so I urge further investigation.
Thirdly, given that most of my volunteers were either personal friends, blog readers and/or Slate Star Codex Subreddit users, we shouldn’t underestimate the possibility that these negative results reflect oddities of the sample.
If you enjoyed this article please consider joining our mailing list: https://forms.gle/TaQA3BN5w3rgpyqeA also, a collection of my best writing between 2018 and early 2020 is available as a free e-book “Something to read in quarantine: Essays 2018-2020”. You can grab it here.
DePonySum 