Lots of interesting data visualizations tell a compelling story. One of them is below, in my humble opinion. It shows that in 2021, 20 percent of U.S. newlyweds married someone outside their major race/ethnic group. The figure shows the upward trend for 5 groups since 2008, and the pattern of lower rates for Whites, then Blacks, Hispanics, and Asian/Pacific Islanders. You can clearly see the three components I like to highlight in a memorable trend graph: Level, trend, and difference.*
But before you can make sense of a cool data visualization, it really helps to have some background knowledge. People absorb a lot of basic demographics just wondering around, but a lot of the time those facts are wrong, misleading, or misunderstood. If you’re teaching undergraduates, I highly recommend working with them a little to take on some basic facts as building blocks for the rest of their knowledge. To put this graph in context, for example, it’s good to know that the U.S. population is 59% non-Hispanic White.
There may be a sucker born every minute, but that’s only 0.4% of all babies born in the world.
The numbers are rounded to reasonable units for easy memorization. All refer to the US unless otherwise noted. Links to the sources below.
| Number | Source | |
| World Population | 8 billion | 1 |
| U.S. Population | 336 million | 1 |
| Children under 18 as share of population | 22% | 2 |
| Adults 65+ as share of pop. | 17% | 2 |
| Official unemployment rate (July 2022) | 3.7% | 3 |
| Unemployment rate range, 1970-2018 | 3.5% – 15% | 3 |
| Labor force participation rate, age 16+ | 63% | 9 |
| Percentage of U.S. adults who smoke cigarettes | 11% | 6 |
| Non-Hispanic Whites as share of population | 59% | 2 |
| Blacks as share of population | 14% | 2 |
| Hispanics as share of population | 19% | 2 |
| Asians / Pacific Islanders as share of population | 7% | 2 |
| American Indians as share of population | 1% | 2 |
| Immigrants as share of population | 14% | 2 |
| Adults age 25+ with BA or higher | 34% | 2 |
| Median household income | $75,000 | 2 |
| Total poverty rate | 12% | 8 |
| Child poverty rate | 16% | 8 |
| Poverty rate age 65+ | 11% | 8 |
| Most populous country, China | 1.4 billion | 5 |
| 2nd most populous country, India | 1.4 billion | 5 |
| 3rd most populous country, USA (CIA estimate) | 340 million | 5 |
| 4th most populous country, Indonesia | 279 million | 5 |
| 5th most populous country, Pakistan | 248 million | 5 |
| U.S. male life expectancy at birth | 74 | 6 |
| U.S. female life expectancy at birth | 79 | 6 |
| Life expectancy range across countries | 54 – 85 | 7 |
| World total fertility rate | 2.3 | 10 |
| U.S. total fertility rate | 1.7 | 10 |
| Total fertility rate range across countries | 0.8 – 6.8 | 10 |
1. U.S. Census Bureau Population Clock
2. U.S. Census Bureau quick facts
6. National Center for Health Statistics
8. U.S. Census Bureau poverty tables
* At Andrew Gelman’s request in the comments, I remade the figure as a regular line graph. Hard to argue with this as more clear, I guess, but I do like the way the graded column figures look. Thanks, Andrew.
Family Inequality 
I recommend plotting this instead as a line plot, with time on the x-axis (ideally going back before 2008, if the data are easily accessible), and a separate line for each group, with labels directly on the lines.
Try it!
Or send me the data and I’ll try it, and we can compare the 2 graphs.
LikeLike
Hard to argue with this suggestion, now that you mention it. I can’t remember why I did it that way. Added to the post! Thanks. (Can’t go back further in a comparable way because they only added marital events to the ACS in 2008.)
LikeLike
Always a fan of popularizing these demographic facts. I bet it would boost disciplinary solidarity if sociologists collectively put together a list like this and everyone expected each other to memorize the values (within say 10%). We are so often working at the edge of our collective knowledge it seems healthy for us to remember areas of high consensus.
As for the two intermarriage figures, I think the line plot mutes the trend because the x axis is too long. The year:percentage aspect ratio in the bar plot is ~1 while for the trend line its is around ~0.2. Without studying the line plot, my first impression is that all the groups go up about 5 percentage points. Squaring (or almost squaring) the aspect ratio would make it easier to compare trends between groups.
LikeLike
Nick:
Sociologist Lane Kenworthy already put together such a list! He called it, 100 Things to Know; see here: https://statmodeling.stat.columbia.edu/2020/03/12/100-things-to-know-from-lane-kenworthy/
So maybe your profession just needs to publicized Kenworthy’s list.
LikeLike
Well, I’m not a sociologist, but I’ll say a couple things.
1. I don’t really think the comparison of intermarriage rates between different ethnic groups adds anything. Smaller ethnic groups are more likely to intermarry unless they live in single-ethnicity communities, and there are a lot of other confounding factors (African-Americans, who are rarely from immigrant families, already have a lot of mixed blood, the people who identify as mixed obvious confuse things, etc.)
2. On the other hand, the takeaway for me — that in last 14 years, intermarriage rates have been skyrocketing — is something I really think everyone should be aware of. (You would agree, wouldn’t you, that in demographic terms, a 6% increase in 14 years is skyrocketing?)
3. I already knew all little numbers. But I guess that’s because I’ve been reading this for some years 😉
LikeLike
“Mixed blood,” huh?
LikeLike
What are you suggesting, Andrew? Plainly there’s something I’m not understanding.
LikeLike