The Power and the Money

The Purple Rose of Texas is here

All eyes are on Texas.

I love Texas! And I have for a very long time. Maybe this dates back to a trip to Dallas to visit my stepfather back in 1986? Or maybe there’s just something about the state that I like? The food? West Texas? (Yes!) The Germans? (Hell, yes!) The Tejanos? (Double hell yes!) The state’s utterly insane history? The weird way Texans know about their state’s utterly insane history? I can’t say. I know that I’d be happy to live there, and that’s weird for a die-hard New Yorker like myself who only likes Washington because it has a subway system.

So it gives me more satisfaction to imagine Texas turning blue than it would for, say, Georgia. (Although MARTA, hey. And Atlanta does have the best BBQ in America.) It makes me want to believe that things like the below really are astonishing:

Something astonishing is happening in Texas. It's hard to know what it will actually mean for the result, but it's highly likely that the number of early votes will eclipse the *total* number of votes cast in 2016. The tally is already up to a whopping 91% of the total 2016 vote.

— Brian Klaas (@brianklaas) October 29, 2020

The turnout numbers do not yet mean that we have a revolution in the Lone Star State. The polls may mean that we’re seeing a Lone Star revolution, but not one big enough to put Biden over the edge without some luck. But not the turnout data, not yet anyway.

Start with the fact that registrations in Texas are only up from 78.2% of the voting age population in 2016 to 78.5% in 2020. If turnout rises back to 2008 levels (59.0%, versus 56.7% in 2016) then we will see the number of total votes cast rise from 9.0 million to 10.0 million. As of yesterday, the total number of early voters was 9,009,850 . That would be about 90% of our “expected” turnout. (Today is the last day for early voting.) That is way up from the 4,497,431 (or 52%) recorded in 2016. 

But ... we do not know is how many of those voters are new and how many have simply decided to vote early due to Covid fears or because early voting has expanded. It is quite plausible that all those early voters simply means that fewer people are going to show up on Tuesday. If we wind up with a turnout of 10 million voters, then Biden-Harris will need to snag basically all of the 1.0 million “new” voters. 

Is there any good news in the data, then?

Well, one solace is that the Democrats do not really need to snag all of the new voters. Lots of people have either passed away or moved out of the Lone Star State. The actual number of new voters will be larger than 1.0 million. A reasonable back-of-envelop calculation (which assumes that people who die or move out-of-state vote the same as all other Texans) would cut the share of new voters that the Democrats need to win to 70%.  

In other words, it is plausible (but not likely) for Democrats to win the state without a massive turnout surge. Consider the number of votes that major party candidates have attracted over the past 44 years. The number of Republican votes has been basically flat since 2004:

Let me note two weird things in the Texas data. The first is the jump in turnout from 1972 to 1976. First, why was turnout so incredibly low in 1972 and what caused it to spike in ’76? Federal action under the Voting Rights Act? The 26^th Amendment? It’s weird.

The second is the gyrations in the voting-age population (including non-citizens) as a share of the total population. It goes from 73% in 2008 to 69% in 2016 before shooting back up to 74% in 2020. For a state with more than 20 million people that seems like a lot of variation in birthrates or migration. 

TLDR: the turnout surge in Texas means less than you think it does. But even if Texas does not go blue, as is likely, purple Texas is already here. 

What would it take to get one billion Americans?

What would it take to get one billion Americans by 2100? Matt Yglesias does a good job of explaining why that would be a worthy cause in his book titled … well … One Billion Americans, but he doesn’t really say what it would take to get there. At best you can say that he proposes a bunch of policies that would be good in-and-of-themselves and will raise population growth on the margin. That’s pretty weak tea.

So what would it take? Well, here I’m going to focus on 2060 and 2100, since the Census Bureau projects out that far and I’m marginally likely to be alive, while my kids will be in their forties and (they better!) have kids of their own. According to the Census Bureau, the United States will have about 404 million people by 2060. The census projection then creeps upwards to 426 million by 2099 before starting to shrink.

The problem with that projection, as Lyman Stone will tell you, is that the Census is relying on outdated statistics. Both birthrates and net immigration have plummeted since they put together their estimates. I constructed a simple population model and plugged in actual Trump-era birthrates — i.e., 1.74 children per women instead of 1.89 and net migration of 0.6 million per year instead of 1.1 million. (I assumed the same improvement in mortality.) The result is a United States with only 362 million people in 2060. And that 2060 population will be falling, shrinking to 331 million by 2100 … a million fewer people than today.

So what could we do? Well, we could adopt Canada’s immigration policy. Using the net migration numbers for the “status quo” scenario put out by Canada’s Conference Board, American net immigration would rise to 2.7 million people by 2030. That’s about 4½ times the current number. Now, the current number has been pushed down by the Trump Administration, but immigration would still run about 2½ times the amount we averaged in 2000-16. (My calculations assume a ten-year phase-in period.)

But it would mean that the United States would accept a number close to all net migration into high income countries. According to the U.N., high income nations took in about 3.2 million people per year in 2015-20, down from a peak of 4.7 million per year in 2000-05. Now, I have no doubt that there is a lot of pent-up demand to emigrate held back by U.S. restrictions. But the United States is so big compared to Canada or Australia that it’s a real question whether we could attract this level of immigration over the next 80 years even if we wanted to. It is doubly unsure whether we could attract those numbers using the Canadian-style points system; we would almost certainly need to give bigger weight to other factors than Canada does or lower the cut-offs. 

And even that wouldn’t get us to 1 billion. Rather, the U.S. would hit 465 million in 2060 and 594 million in 2100. That’s a healthier population than you get under current policy, but it isn’t exactly the kind of dynamic growth that Yglesias (or I) would want to see.

To get to one billion by 2100 (or 570 million by 2060) you would also have to raise the total fertility rate (TFR) to 2.4 children per women. That is not a historically high number, but the only developed nation to maintain numbers that high is Israel. And even there, the number for secular Jews is only 2.2. At a TFR of 2.2, a Canadian immigration policy would get the USA to 534 million by 2060 and 851 million by 2100. A more realistic (but still hard!) TFR of 2.0 would get us 503 million by 2060 and 730 million by 2100.

But 730 million is not one billion.

I understand that hitting that target wasn’t Yglesias’s point. Rather, his point was twofold. (1) Here is a frame for a bunch of good policies! (2) Why the hell are most Americans afraid of aiming big anymore? For me, (2) is the salient point. We’re a technophilic nation, so it’s not that we’re afraid of change in general. What we’re afraid of is a certain kind of change, one involving big collective endeavors. We cower from building new cities, new roads, skyscrapers in our back yards, nuclear power plants, the Green New Deal. That’s a bad thing. We should stop. We will have a better future if we stop.

But sadly, absent some incredible rise in fertility or an unlikely drop in mortality, that better future won’t involve one billion Americans.