The plague years

 I'd heard about a decline of self-reported happiness in recent years, but I hadn't looked at the numbers until I saw a post by Noah Smith a few days ago.  He gave a figure showing that the percent who said they were "not too happy" was almost twice as large in 2021-2024 as it had been in 2015-9.  My first thought was that the numbers should be broken down by year--it's understandable that people were unhappy during the Covid pandemic, but things probably bounced back in 2024.  However, it turns out that average happiness (3=very happy, 2=pretty happy, 1=not too happy) in 2024 was lower than in 2022 and not much higher than in 2021 (data are from the GSS, which is conducting their 2026 survey now but hasn't released the data).  

My next thought was that it was because of political polarization.  Traditionally, partisanship affected views about the state of the country--people were more favorable when their party had the presidency--but had little or no effect on individual happiness.  But as partisanship gets stronger, it might affect a wider range of things--that is, if you think the economy is in bad shape and we're at risk of war, it's reasonable to be unhappy.  However, once again my thought was wrong:

Republicans were consistently a bit happier, but the gap didn't change depending on the party in power, even in recent years.  

I looked at differences by various demographic variables--age, gender, race, education, and more.  Although changes weren't exactly parallel, they were similar in all groups--a big decline between 2019 and 2021 and only modest recovery (see this paper for a more detailed analysis).  Then for some reason I compared people born in the United States to people who were born outside.

The 2019-2021 drop was smaller and the 2021-2024 recovery was larger among people born outside the United States.  Their 2024 average was just about the same as the average over the whole period (2.13 vs. 2.16).  Why the difference?  My hypothesis is that it reflects a difference in expectations:  because of a combination of the nation's wealth and power, a government that functions pretty well, and good luck, Americans have become used to thinking that things will work out.  So when Covid turned out to be as bad or worse than people had expected and recovery was slower, it was a shock.  Most of the people who were born outside were from places with more history of misfortune and misgovernance, so their expectations were lower.  

Of course, I developed this hypothesis after looking at the data, but Gallup collects annual data on happiness (using a somewhat different question) in a large number of nations.  My hypothesis has implications for the size of post-Covid changes in happiness in different nations.  I may look at that in a future post.