The other day I showed some data from the World Happiness Index, and guessed that, as in 2016, the 2018 data would show a significant negative correlation between the religiosity of a country and its happiness index: that the more religious the country, on average the less happy its inhabitants.
Now two readers have plotted the 2018 data and indeed saw such a correlation, which is expected given that the data wouldn’t change much in two years. First, here’s Greg Mayer’s analysis (see the link for the countries involved); his comments are indented:
Here’s the 2018 analysis: the correlation has become a bit stronger.
r=.56, p=.00002 (n=51) [JAC: the proportion of variance among countries in happiness explained by religiosity is simply r², which is 0.31 or 31%. That is, 31% of the variation in happiness among nations is explained by taking into account their religiosity.
The regression line: Happiness_2018 = 6.55 – .0174(Religiosity)
Happiness 2016 and 2018 are highly correlated: r=.91. Much of the discrepancy is due to Venezuela having become much unhappier.
I’ve attached a picture, not as fancy as Gluon’s—just the regression line and the line’s 95% confidence interval.
And reader Michael Coon got pretty much the same thing. His notes:
I noticed a request on WEIT to see if the UN Global Happiness scores you commented on today correlated with religiosity. In case someone hasn’t already provided you with the analysis, I had a few moments and plotted the scores against the “importance of Religion” scores reported in a Gallup poll cited in Wikipedia. I used data from table 2.2 in the UN report. The Gallup data is from 2009 so caveats there, though I doubt there is much difference today. Indeed—it doesn’t look much different to the one Gluon Springs posted.
Note that not all of the countries represented in one poll are represented in the other.
As far as “happiness” goes, there is a strong trend with religiosity but it is poorly fitted with an r² of only 0.337. [JAC: pretty much the same as Greg’s] This suggests (to me anyway) that although religiosity does correlate with happiness, other factors are in play. Not surprising, to me anyway.
I wouldn’t say that 0.34 is a “poor” contributor; it is a substantial proportion of the variance. And yes, of course other factors are in play, including the historical presence of Communism, the variance in social well being (the underlying cause of the variation in religiosity and happiness, in my view) and so on.
What remains is that these graphs giving data across countries shows that more religious societies aren’t happier. They are unhappier.
And that holds within countries, too, at least if you look at the correlation between the social development index of American states and their religiosity. In that case, which a kind reader analyzed in 2012, the correlation between religiosity and the “human development index” (HDI; a measure of social well being) among states was -0.669, with r² = 0447—even higher than the correlation for happiness across countries. Again we see that there’s a negative correlation between religiosity and well being, and although there’s no index of happiness for U.S. states, I’m pretty sure it would go along with the HDI.
If religion makes a society happier, there’s no data to show that. All we see is that on average countries that are well off are less religious and are happier; and that goes for U.S. states as well. Checkmate, religionists.
Nice analysis. Of course, there remains the question of causation. Is increased religiosity a cause or an effect of unhappiness/social dysfunction?
And oddly there have been reports that show religious individuals are happier than non-religious individuals (though not by all that much) so clearly this happiness does not scale up. It’s either a statistical screw up or a real issue when scaling up to the societal level that would be interesting to understand.
Simpson’s paradox, perhaps?
Yes, perhaps. Some confounding variables are most likely in there somewhere causing the issues.
Venezuela became much unhappier because it became much poorer. I doubt that it became much less religious. Perhaps an analysis of changes over time rather than correlations of levels at a point in time can unravel the causation issue.
Well, we have time changes for the GINI index versus religiosity in the U.S., and what those show is when the GINI index shows an INCREASE in inequality, religiosity INCREASES a year later. Same goes for decreases. This implies that if there is a causality, it’s the well-being that affects religiosity (I’ve written on this before). So one would expect that countries would become more religious as well as unhappier after the social support system frays.
It would be interesting to put those into a multivariate analysis
That makes sense. Educated people enjoying a good standard of living are less likely to fall for the false promises of religion.
Gini need not be in all caps, it’s named after its developer, not an acronym. The Gini index is equivalent to the Simpson diversity index for a full community (as opposed to a sample from the community).
It could be that extreme poverty leads to both (1) less happiness (2) amenability to religious belief. And it seems to be the case that missionaries target poor areas to proselytize, reinforcing this effect.
I’m not convinced that being religious makes you unhappy (if that’s even the suggestion here). The happiest people I know ignore politics, spend a lot of time with family, and go to church. :/
See my comment about individuals. At the individual level statistically, the religious are more happy but at the societal level this doesn’t seem to hold.
There is no logical reason the correlation at the two different levels of analysis should be the same, or even in the same direction. Possibly another example of the reversal “paradox” (it is not really a paradox) known as Simpson’s paradox.
That’s not Simpson’s paradox; it’s the ecologic fallacy.
Same thing: The ecological fallacy is a special case of Simpson’s paradox involving a cross-level inference from group level data to individuals (the reverse is know as the individualistic fallacy).
Extreme poverty and increased religious belief seem to go hand in hand and both will put you in the unhappiness category. How well a country takes care of it’s population seems to have great affect in happiness with Scandinavia at the top. Odd that another country that really takes care of it’s people is not at the top. Such is the case for Saudi Arabia.
Of course. The Saudi ultra-conservative monarchs have been funding the Wahabi version of Islam massively with petrodollars in Saudi Arabia, the Middle East and Europe. In Europe they have paid hundreds of Imans for decades, and many view the Saudi Wahabi faction as the main responsible for the failing of the Arab Spring and the enormous increase of religiosity among Arab and North African immigrants in Europe, and not only that, it encouraged the emergence of radical groups like Al queda, ISIS, and others.
It would be interesting to see side by side comparison of Saud-Wahabi and certain Oligarch – dark money transactions in the west. Anybody capable of producing accurate facts and figures on either would deserve a place in history.
… and is likely to get a 9mm problem solver in the back of the head.
I was a little surprised how high the happiness rating was for Saudi Arabia – considering half of their citizens are women. Then I wondered if they were allowed to participate in the survey.
Sure, but they would have to answer exactly as the husband or handler said.
Reminds me of the Iranian film, Secret Ballot. It follow a woman vote-collector as she goes around a tiny, remote island to collect ballots. A good film.
Am I not reading Michael’s graph right, or is Cyprus out of place on the happiness axis? It’s 61 on the UN chart, but is right up there with Finland and Norway on the graph.
That’s the only one that leapt out at me, but perhaps there are others?
I’d like to see if there’s a correlation between prosperity and religiosity, and also between prosperity and happiness. I wonder if prosperity is the underlying key to both.
very interesting study. I’d suggest trying to incorporate average national IQ measures and maybe breaking out the happiness and religious importance variables by religious affiliation within each country. From the tests I ran IQ is a strong predictor for happiness(.69***) and importance of religion(-.73***). However, this pattern does not hold across all religions. weak evidence suggests that Christianity is negatively correlated with the importance of religion(-.18*) and positively with happiness(.16*) and not correlation with IQ(-0.01). For Muslims this pattern is reversed and stronger(.40***), (-.25***) and (-.26***).
In line with your hypothesis % of “unaffiliated” is strong predictors of importance, happiness and IQ (-.74***), (.37***) and (.52***)
Perhaps you have cause and effect reversed. To quote Bierce:
INDIGESTION, n. A disease which the patient and his friends frequently mistake for deep religious conviction and concern for the salvation of mankind. As the simple Red Man of the western wild put it, with, it must be confessed, a certain force: “Plenty well, no pray; big bellyache, heap God.”