People Go to Church More on Easter
And a quirk about time use data collection means we don't have much information on Christmas
American Time Use Survey data confirms it: people are more likely to attend religious services on Easter than non-Easter Sundays.
From 2013 to 2023, excluding 2020, 21% of sample respondents on non-Easter Sundays reported attending religious services. On Easter Sundays, that jumped to 27% (p=0.005).
Dis-aggregated by sex, the change was only significant for men. 17% of men attended religious services on non-Easter Sundays versus 23% on Easter Sundays (p=0.05). For women, it was 25% versus 30% (p>0.05). On non-Easter Sundays, women were significantly more likely to attend religious services.
Easter Sunday religious attendance didn’t vary much by marital status, parental status, or 4-year college degree status. That separates it from non-Easter Sundays, where differences on marital status (24% for married, 19% for unmarried) and parental status (20% for parents with minor children, 22% for respondents without minor children) were both statistically significant. Some of that could be down to sample size — there are fewer Easter Sundays in the sample than non-Easter Sundays.
Initially, I had hoped to write this about both Christmas and Easter religious attendance, but it turns out the sample has very few responses for Christmas and none from Christmas Eve. Why is that? To create the ATUS, Census employees call respondents from the Current Population Survey and ask them about what they did the previous day. Christmas is a federal holiday, so Census employees don’t call people to ask what they did on Christmas Eve. And while there are a few responses from Christmas Day itself, the sample is much smaller than for Easter (127 over 10 years versus 495), probably because employees are more likely to take vacation on December 26 than Easter Monday.
My hopes are dashed on NHES-ECPP data
Awhile back, I wrote about how something was odd with child care cost data in the 2023 sample of the Early Childhood Program Participation survey, and that I hoped the microdata, once released, would provide more clarity.
Unfortunately, the Trump administration reportedly fired everyone in the National Center for Education Statistics, which is responsible for this survey.
Sin makes you stupid.