Looking at https://en.wikipedia.org/wiki/List_of_countries_by_intention..., I suspect it has more to do with average wealth / standard of living combined with wealth inequality. "Social cohesion", whatever that means, is probably correlated, but I wouldn't be sure it's the underlying cause.

How do you measure social cohesion? Crime rate might be one way, but then you've got yourself a tautology.

Economic inequality perhaps?

If you sort https://en.wikipedia.org/wiki/List_of_countries_by_income_eq... by "CIA Gini %" and skip the countries where there is no data in that column, then sort https://en.wikipedia.org/wiki/List_of_countries_by_intention... by the "2010" column, then squint a bit at the top ten or twenty countries in each list, they look kinda correlated. (Or co-correlated with a whole bunch of other factors.)

According to Stephen Pinker, the correlation between inequality and violence is weak. It really surprised me, because I always assumed that they’d be linked.

That seems controversial considering all the people who found the opposite.

https://www.mitpressjournals.org/doi/abs/10.1162/00346530055...

Quoting from _Better angels of our nature_ p119:

”Among economic measures, inequality is generally a better predictor of violence than unemployment. But The Gini coefficient, the standard index of inequality, actually rose in the United States from 1990 to 2000, while crime was fallinh, and it had hit its low point in 1968, when crime was soaring. The problem with invoking inequality to explain changes in violence is that while it correlates with violence across states and countries, it does not correlate with violence over time within a state or country, possibly because the real cause of the differences is not inequality per se but stable features of a state’s governance or culture that affect both inequality and violence. (For example, in unequal societies, poor neighbourhoods are left without police protection and can become zones of violent anarchy.)

Thanks, very interesting.

If anything, as hinted above, there are lots of other confounding factors. Also, Wikipedia's wealth inequality list looks a lot less similar to murder rates than the income inequality list I posted above. Would you know off the top of your head if Pinker wrote about wealth or about income inequality? (Or yet another measure?)

I looked it up, since I got downvoted (see sibling comment). Since he’s talking about the Gini coeffecient, I’d presume he’s referring to income inequality, although it’s not explicit.

Right, thanks for the quote. The Gini coefficent is applicable both to income and to wealth inequality, so that doesn't answer the question. (Not your fault, of course.)

According to the data it is very strong. There are only few countries where there is both high economic inequality and low incidence of crime, and those countries are full blown dictatorships like North Korea.

Economic diversity, racial diversity, and religious diversity all spring to mind.

EDIT: Downvotes? For this?

I'm not claiming that lacking the diversity is a good thing, merely that having extra options decreases cohesion. The more diverse the economic situation is (say, due to vast wealth inequality) the less cohesion there is. The more diverse the religions are, the less common-ground people have based on religion--as in most of European history. The more racially diverse a place is, the more tensions due to race (LA versus Helsinki).

Is it worth the hit to cohesion to have better diversity? I'd argue that, in the absence of external threats, it absolutely is. But living in a fantasyland where diversity doesn't come at the expense of cohesion doesn't help.

> How do you measure social cohesion?

It's the inverse of diversity. Economic inequality, religious, philosophical & cultural diversity — make a measure which combines all of those things, then invert it. It'd at least make a start.

Willingness to trust strangers. https://ourworldindata.org/trust

Yes. Social trust was probably actually the term I should have used, e.g. https://www.bloomberg.com/view/articles/2018-02-23/you-can-t...

It seems to be somewhat inversely related to diversity but not entirely. As Megan has also written, Greece and Italy are also fairly homogeneous but don't have most of the social characteristics of the Nordics.

And Canada is very diverse but has relatively high trust.

Mostly irrelevant factoid - Lincoln, NE is reasonably close to Iceland in population, murder rate, gun ownership.