Estimating the proportion of Puritan genes in America's white population

Ellsworth Huntington estimated about 5500 people bearing the surname Huntington were alive in America circa 1930.

Census Bureau estimates put the number of whites bearing the surname Huntington in 2000 at 6325.

So the absolute frequency of New England genes might have increased something like 6325/5500 = 1.15 times in a 65 year period in which the American "non-Hispanic white" population overall increased 188 128 296 / 110 286 740 = 1.7 times. While there is uncertainty in these numbers, there's little doubt the relative representation of New England genes amongst American "whites" has fallen over the past century.

As for what the actual proportion of Puritan genes might be, we can make a crude estimate as follows.

There were 992 421 whites in New England in 1790, perhaps 7/8 of whom were of Puritan stock.

There were 687 Huntingtons in the US in 1790, about 85% of them in New England.

6325 / 687 = 9.2

9.2 * 992 421 = 9 130 273

Adding say 20% to account for people of New England stock outside of New England in 1790 brings us to 10 956 328.

10 956 328 / 188 128 296 = 5.8% (of overall US white ancestry)

As of 1930, just over 1/4 of Huntingtons in Ellsworth Huntington's tabulations were located on the "East Coast", so the fraction in New England proper will have been well under 1/4 even at that date. But if, to find an upper limit, we assume 1/4 of New England genes remain in New England today, we're left with the equivalent of 10 956 328 / 4 = 2 739 082 individuals. That's against a total New England "non-Hispanic white" population of 11 686 617 in 2000.

2 739 082 / 11 686 617 = 23.4% (upper limit for New England) [the correct estimate is about half this number; see update below]

And, again, that's an overestimate of the proportion of New England genes in New England's white population today. [see update for more accurate estimate], but it's pretty clear "Yankees" are not the dominant force in voting in slates of Irish Catholic Democrats.

Update: Corrected math error and checked a couple more names.

Coolidge - 2663.892 / 225 = 11

Conant - 4671.9 / 551 = 8.48

Huntington + Coolidge + Conant - 13661 / 1463 = 9.34

More names can be added, but these seem similar enough.

Update 2: It appears something like 613 / 4,571 = 13.4% of Huntingtons were living in New England at the time of the 1940 Census.

So with better numbers the estimate for the average degree of New England ancestry of today's white New England residents becomes 12.6%.

14 comments:

JayMan said...

I have been reading your recent posts with interest. While fascinating, nothing you've posted so far is necessarily meaningful – yet, but I await additional posts.

However on this post, you've built a bridge way too far:

"Ellsworth Huntington estimated about 5500 people bearing the surname Huntington were alive in America circa 1930. Census Bureau estimates put the number of whites bearing the surname Huntington in 2000 at 6325. So the absolute frequency of New England genes might have increased something like 6325/5500 = 1.15 times in a 65 year period in which the American 'non-Hispanic white' population overall increased 188 128 296 / 110 286 740 = 1.7 times."

Hold on here. I don't think you can make that conclusion. This is somewhat informed guess, but only a guess.

"While there is uncertainty in these numbers, there's little doubt the relative representation of New England genes amongst American 'whites' has fallen over the past century."

I don't know about that either. You're using a surname analysis with the obvious limitations that come with it.

That said, I will grant that liberal fertility, primarily concentrated in the Old North, has lagged behind conservative fertility for much of the previous century.

"As for what the actual proportion of Puritan genes might be, we can make a crude estimate as follows. There were 992 421 whites in New England in 1790, perhaps 7/8 of whom were of Puritan stock. There were 687 Huntingtons in the US in 1790, about 85% of them in New England. 6325 / 687 = 9.2 9.2 * 992 421 = 9 130 273 Adding say 20% to account for people of New England stock outside of New England in 1790 brings us to 10 956 328. 10 956 328 / 188 128 296 = 5.8% (of overall US white ancestry)"

A very crude estimate. And likely quite wrong for a variety of reasons.

These include:

1. Surname only traces ancestry through the male line, which loses the vast majority of individuals in a lineage
2. There's no guarantee this Huntington family is representative the reproductive success of New Englanders as a whole

Let's assume you've (partially) addressed the second problem by looking at additional surnames, you're still stuck with the first.

Decent attempt, but I wouldn't put much stock in it, yet.

Vanishing American said...

How, exactly, could anyone verify this to the satisfaction of the skeptics? A lot of other sources (census maps, for example) indicate that Anglo-Americans are becoming more rare in New Englad, apart from some rural areas which have not been ethnically cleansed by later immigrant waves.

However the census maps are based, I think, on self-reporting of ethnic ancestry, which is not reliable. It is a widely accepted (though unproven) belief that Germans are the majority among White Americans, and that Anglo-Saxons are fewer, having melted into the melting pot. But neither of these beliefs are proven, and absent widespread DNA testing I don't think it's possible to be dogmatic about American ancestry. The sad fact is, most Americans who are not of recent immigrant stock do not know their ancestry for sure. They rely on family rumors like 'Cherokee ancestry' and other such fairytales.

I suspect that there are many more "puritan stock" descendants in Utah than anywhere else, but I can't prove it. I do know that many of my New England ancestors migrated Westward, to Utah or other Western and Midwestern states long ago.

I just don't get why there is this bizarre fixation on ''Puritans'' as the villains in our national story. I can think of other more likely explanations. But it's ''in'' to cast blame on "Puritans'' and "Yankees'' from New England.

TGGP said...

JayMan, I don't think it matters too much if tracing ancestry through the male line loses "the vast majority". A single surname will also leave out the vast majority of Puritans, the more important question is your second one: is it representative? And the way to check that would be to see if other Puritan surnames had a similar shift in representation.

chris said...

"I just don't get why there is this bizarre fixation on ''Puritans'' as the villains in our national story. I can think of other more likely explanations. But it's ''in'' to cast blame on "Puritans'' and "Yankees'' from New England."

Puritans have become a cultural trope for self-righteous,(and/or) hypocritical, scare-mongering/witch-burning/witch-hunting, e.g. Salem witch trials, the novel The Crucible.

JayMan said...

People in the blogosphere fixate on the Puritans because it's simpler and cooler to do so. A lot of the multiculturalist sentiment that exists today isn't Puritan, since they were xenophobes. Much of it is actually Quaker & German (collectively Midlander).

n/a said...

VA,

Glad to see you're back.

TGGP,

JayMan, I don't think it matters too much if tracing ancestry through the male line loses "the vast majority".

Exactly. It's a sample. It makes no difference that we're sampling the direct paternal line, unless there are systematic differences in fertility between female and male lines. Down to around 1850 or 1900 (by which point most of the population growth had occurred), this would not have been possible to any meaningful degree, since New Englanders married primarily among themselves. Subsequently, considering that education tends to suppress female fertility more than male fertility, New England surnames would if anything tend to overrepresent the fraction of New England genes remaining in the US population.

JayMan said...

@n/a:

Who's to say male fertility and female fertility didn't systematically differ? This is obviously not possible in a group that mates endogamously, but it is quite possible in a group that mixes with other groups.

By the way, female fertility was *not* suppressed by education. White American men and women have the same number of children. (This doesn't mean that there isn't within group skews).

n/a said...

JayMan,

Pretend you're wrong. I know it's hard, but just pretend you haven't really thought this through, and your objection is not an important one. Then look for any evidence that would disprove that, and post the evidence (rather than going directly from imagining a particular objection "could" be devastating to posting a comment).

Note: I was interested here in making a rough estimate. If I wanted to make a more precise estimate, I can think of about half a dozen other things I would work on before worrying about male-female differences in fertility.

JayMan said...

@n/a:

You are the one who needs to support your criticism. When I find a hole in it, the burden of proof returns to you.

Look my friend, your line of reasoning is similar to that used by the medical establishment in giving us the nonsense wisdom that they have; they assume their crude and often useless methods are capable of garnering definite truth, and disregard the "could be" objections that folks like me raise. It doesn't matter if you think my "could be" objections are likely are not; as long as they're possible, your "proof" isn't proof. The burden is then on you to show that they're not possible.

"Note: I was interested here in making a rough estimate."

Rough indeed.

"If I wanted to make a more precise estimate, I can think of about half a dozen other things I would work on before worrying about male-female differences in fertility."

Go for it. I look forward to it, because I am indeed following this topic with interest.

Anonymous said...

What happens in these "debates" is that some people simply cannot bear the thought that their pet theories are wrong and so they try to introduce all sorts of red herrings into the discussion without actually doing the basic research required to see that their opponents are basically correct on just about any relevant measure you care to use.

They continually move the goalposts and demand a burden of proof they would never conceive of applying to their favored narrative.

The goal isn't to think through the issues logically, but to desperately cling to their pet theory. The whole thing is done in bad faith from the start.

JayMan said...

@Anonymous:

Look, n/a has presented some fascinating enough information, but not enough to "settle" the debate, yet. That might change in the future. Be patient.

JayMan said...

@Anonymous:

"They continually move the goalposts and demand a burden of proof they would never conceive of applying to their favored narrative."

And, for the record, to counter this charge, in this debate, I've always maintained that the definitive piece of evidence would be genetic analysis. I haven't moved any goalposts. N/a hasn't provided such evidence, yet.

Now, that's not to say that it's not possible to demonstrate what he's claiming with other types of evidence, but, so much for saying I've been moving goalposts.

Steve Sailer said...

The Huntingtons are an amazingly accomplished bunch in America. Around 30 have their own Wikipedia pages. Southern California, for instance, has Huntington Beach, Huntington Park, and the lovely Huntington Library and Gardens.

Huntington ranks up there with Eliot.

Hail said...

12.6% (N/A's final estimate)

In Reuters' 2012 exit-polling asking about respondents' race and religion, 21% of those in New England, New York, and New Jersey claimed to be White-Protestants, I found.

Summary of Voter Ethnoreligious Self-ID for NY, NJ, New England, 2012
_____________
21% of voters, White-Protestant (see link for denom.)
32% of voters, White-Catholic
8% of voters, White-Jewish
6% of voters, White-"Other-Religion"
12% of voters, White-No-Religion
_____________
21% of voters, Nonwhite
______________

Caveats:
(1) 21% refers to entire voter population, according to the Reuters Exit Poll. Looking at only Whites, this makes Protestants 27% of white voting-population in NY, NJ, New-Eng.
(2) Some share I won't hazard to guess of those White-Protestants (i.e., 27% of White voters) are not "Puritans".
(3) Descendants of Puritans, I'd conjecture, are more likely to vote for a lot of reasons, and may be slightly overrepresented vs. overall population

If we trust the Reuters Exit Poll, and if we assume that less than half the White-Protestant voters of 2012 of those states are of "Puritan" stock, the 12.6% estimate for New England is very reasonable.