"It would be straightforward to examine the geography of intermixing and to investigate results for cohabiting couples too."
In conjunction with what you have given us here, I think the cohabiting couples (and geography) would definitely be interesting. In addition, since the topic is, broadly, that of Americans mixing, are there any good ways to look at the parentage of children or what percentage of Americans have had a relationship outside of their race?
Thanks for all your work, it is always interesting and well presented, even if I have to smash my head on the desk to remember the stats and R programming I learned in the distant past.
I can certainly look at the data for children, but the percentages of Americans who have had exogamous relationships are unavailable to me. All I can look at is whether someone is interracially married or cohabiting.
@sebjenseb on Twitter has looked at selection into reported interracial dating in the NLSY '97 and Add Health datasets, so I might be able to look into this in those, but the power and timeframe would be much more limited. Here's a link: https://mankindquarterly.org/archive/issue/64-2/8
For example, the spreadsheet shows that the never-married percentage among black women increased from 28% (2,248/8,108) in 1970 to 47% (8,709/18,339) in 2020, whereas your graph shows an increase from about 50% to 60% over the same time period.
One possibility is that your 2 "percentage of never-married" graphs contain people of all ages, including children, whereas the links I posted are limited to individuals aged 15 years or older (this would also explain why the never-married percentage decreased for some groups in your graphs whereas it increased for all groups in my links; over this time period, the adult-to-child ratio would have increased, which decreases the never-married percentage). You mention the results are "confounded by age", but it's not clear if that means it's unrepresentative or that it includes children. If the data includes children, I'm not sure how informative these 2 graphs are.
That result was posted as-is. It's just the default output of MARST == 6 when no adjustments are used. As such, it does include children. Adjusting for age by subsetting to 25-45, 18+, or using a spline dramatically alters the result and places it in line with what you've posted. As noted: "If there’s enough demand for a desired age correction or constraint, I can post it." I might get around to it if I can find the time, but the number of requests for additional analyses has piled up pretty high since I posted this.
What's interesting is the relative stagnation of White-Asian newlyweds yet the White-Latinx rates are up. Asian immigration has been pretty high in the post-2010 period.
Anecdotally, I noticed a significant increase in "Asian nationalism", often implicitly anti-White, in the early 2010s. Typically from Asian men but it coincided with the wokism trend. Another reason may just be that because so many Asian immigrants came, it was easier for Asians to pick their own ethnic kin for new marriages. But this argument also works for Latinx, but you don't see a stagnation trend for them but a steady increase... even an acceleration in the second half of the 2010s.
Can you do it for different detailed ethnic groups like that done by Twitter user @tcjfs and also can you do it for us born Asians and foreign born Asians.I would like to see that.
"It would be straightforward to examine the geography of intermixing and to investigate results for cohabiting couples too."
In conjunction with what you have given us here, I think the cohabiting couples (and geography) would definitely be interesting. In addition, since the topic is, broadly, that of Americans mixing, are there any good ways to look at the parentage of children or what percentage of Americans have had a relationship outside of their race?
Thanks for all your work, it is always interesting and well presented, even if I have to smash my head on the desk to remember the stats and R programming I learned in the distant past.
I can certainly look at the data for children, but the percentages of Americans who have had exogamous relationships are unavailable to me. All I can look at is whether someone is interracially married or cohabiting.
@sebjenseb on Twitter has looked at selection into reported interracial dating in the NLSY '97 and Add Health datasets, so I might be able to look into this in those, but the power and timeframe would be much more limited. Here's a link: https://mankindquarterly.org/archive/issue/64-2/8
Your percentage of women never-married by race differs sharply from that reported in these links by the Census. What do you think that is?
For example, the following chart shows that the percent of never-married women has increased from about 20% to 30% from 1970 to 2020: https://www.census.gov/content/dam/Census/library/visualizations/time-series/demo/families-and-households/ms-1b.pdf
You can see data by race based on Table MS-1 here: https://www.census.gov/data/tables/time-series/demo/families/marital.html.
For example, the spreadsheet shows that the never-married percentage among black women increased from 28% (2,248/8,108) in 1970 to 47% (8,709/18,339) in 2020, whereas your graph shows an increase from about 50% to 60% over the same time period.
One possibility is that your 2 "percentage of never-married" graphs contain people of all ages, including children, whereas the links I posted are limited to individuals aged 15 years or older (this would also explain why the never-married percentage decreased for some groups in your graphs whereas it increased for all groups in my links; over this time period, the adult-to-child ratio would have increased, which decreases the never-married percentage). You mention the results are "confounded by age", but it's not clear if that means it's unrepresentative or that it includes children. If the data includes children, I'm not sure how informative these 2 graphs are.
That result was posted as-is. It's just the default output of MARST == 6 when no adjustments are used. As such, it does include children. Adjusting for age by subsetting to 25-45, 18+, or using a spline dramatically alters the result and places it in line with what you've posted. As noted: "If there’s enough demand for a desired age correction or constraint, I can post it." I might get around to it if I can find the time, but the number of requests for additional analyses has piled up pretty high since I posted this.
What's interesting is the relative stagnation of White-Asian newlyweds yet the White-Latinx rates are up. Asian immigration has been pretty high in the post-2010 period.
Anecdotally, I noticed a significant increase in "Asian nationalism", often implicitly anti-White, in the early 2010s. Typically from Asian men but it coincided with the wokism trend. Another reason may just be that because so many Asian immigrants came, it was easier for Asians to pick their own ethnic kin for new marriages. But this argument also works for Latinx, but you don't see a stagnation trend for them but a steady increase... even an acceleration in the second half of the 2010s.
I wish your charts colored races as 23andMe does.
Can you do it for different detailed ethnic groups like that done by Twitter user @tcjfs and also can you do it for us born Asians and foreign born Asians.I would like to see that.
Any information on what the mixed race population will be over time? I’ve heard mixed race couples have lower birthrates.
Are Spaniards included in white or hispanic category in the study?
Wonderful trends