That study as be criticized for using poor methodology and excluding data that does not agree with their conclusions. There are also several counter examples. dspace2.creighton.edu/xmlui/bitstream/handle/10504/64409/… Secular results are completely relevent as religions are quite capable of instituting secular methodolgy. Also many would argue religion is the baseline as secular societies are relatively recent inventions.

That shift in culture is why it is becoming legalized in so many places, but 50 years ago, only 12% of people in the USA thought it should be legal. In fact, the point where approval peaked 50% is almost exactly when states started legalizing it. While specific laws don't keep up with rapid shifts in morality "in real time", when treated as a whole, legal systems do tend to reflect the ethics of their societies.

@John Your citation shows different results because it is studying belief in God and frequence of church visits, not the specific effects of belief in Hell. It also draws national borders segmenting secular developed nations from the less developed religious nations which confounds its results with social equity issues; whereas my citation aggregates behavior relative to people of shared beliefs. Yours also only measures Homicide. More heinous crimes like homicide and human trafficking are less affected by religion as my citation shows.

@Nosajimiki actually it doesn't since they did not control for many factors (your study does not control for social inequality at all, separating developed and less developed countries is what you should do otherwise you ar including a known confounding variable.) and used the wrong analysis (linear regression is a poor choice when you have large outliers), a more limited study is more accurate especially when done with a proper analysis as opposed to just dropping a large pile of confounding variable in a linear regression.

@Nosajimiki Using "crime" is also problematic since the definition of each crime varies wildly across nations, homicide is used because it is one of the few crimes with nearly identical definitions, You can even check the criticism of the study provided by the site. if you don't believe me. journals.plos.org/plosone/article/comments?id=10.1371/…

@John The criticized methodology of the study specifically measures the difference of people who believe in Heaven and Hell vs Heaven and no Hell. But, as the researchers explain, this choice was to eliminate factors that faithful people have in common and isolate just the factors around a belief in Hell as part of a belief in an afterlife. A more accurate conclusion of the study is that a belief in an afterlife correlates to higher crime, but adding hell to that faith reduces it.

@jamesqf While I do believe there is a correlation between law and a society's ethics, having reviewed my citation further, I found that it accounts for social variance by only evaluating for laws that are illegal in all of the tested countries such as murder, robbery, etc. I've revised my answer accordingly.

@Nosajimiki thats not the big confounding variable, not accounting for differences in standard of living, a known factor in many forms of crime as well as using general crime statistics in general are some of the biggest flaws. Also the dramatic differences in the time period used for each category, (1981belief 2008 crime?), Also no using the disparity between heaven and hell belief as the calculable factor results in only comparing belief in one and not the other, people who believe in both get excluded from the calculation. The flaws are so numerous the study is essentially worthless.

If you don't understnad the problem with using both general crime statistics and large disparate sample times this paper should help. It compares only two countries the US and England over the same time variation your study uses. spire.sciencespo.fr/hdl:/2441/f6h8764enu2lskk9p48i5a4sj/…