How birthrates color the electoral map

The Baby Gap: Explaining Red and Blue

by Steve Sailer

The American Conservative, December 20, 2004

American Conservative

iSteve.com

email me

Cover of American Conservative magazine, Dec. 20, 2004: "The Baby Gap: Beyond Red and Blue "  by Steve Sailer

"The Baby Gap: Explaining Red and Blue," my cover story in the Dec. 20th issue of The American Conservative is now available here

 

Here's David Brooks' New York Times op-ed that was based on my article: "The New Red-Diaper Babies."

 

And here's my sequel to "Baby Gap" from VDARE.com where I unveil a demographic factor even more important in deciding the vote than the baby gap: the marriage gap.

 

Here, you can find graphs and data tables, along with reader responses and additional information.

 

 An excerpt from my article:

 

Clearly, the "issues" that so excite political journalists had but a meager impact on most voters... If a demographic or regional group supported Bush's "humble" foreign policy in 2000, they supported his Alexandrine ambitions in 2004, and vice-versa.

Still, this doesn't mean voters are choosing red or blue frivolously. Indeed, voters are picking their parties based on differing approaches to the most fundamentally important human activity: having babies. The white people in Republican-voting regions consistently have more children than the white people in Dem
ocratic-voting regions. The more kids whites have, the more pro-Bush they get...

The single most useful and understandable birthrate measure is the "total fertility rate." This estimates, based on recent births, how many children the average woman currently in her
childbearing years will end up with. The federal National Center for Health Statistics reported that in 2002 the average white woman was giving birth at a pace consistent with having 1.83 babies during her lifetime, or 13 percent below the replacement rate of 2.1 children per woman. This below-replacement level has not changed dramatically in three decades.

States, however, differ significantly in white fertility. The most fecund whites are in heavily Mormon Utah, which, not coincidentally, was the only state where Bush received over 70 percent. White women average 2.45 babies in Utah compared to merely 1.11 babies in Washington D.C., where Bush earned but 9 percent. The three New England states where Bush won less than 40 percent -- Massachusetts, Vermont, and Rhode Island -- comprise three of the four states with the lowest white birth rates, with little Rhode Island dipping below 1.5 babies per woman.

 

Here's a scatter plot showing how closely Bush's share of the vote in a state correlates with the number of babies per white woman. The blue dot way down in the lower left corner represents Washington D.C. and the red dot way up in the right corner is Utah.

Bush carried the 19 states with the highest white fertility (just as he did in 2000), and 25 out of the top 26, with highly unionized Michigan being the one blue exception to the rule...

In sharp contrast, Kerry won the 16 states at the bottom of the list, with the Democrats' anchor states of California (1.65) and New York (1.72) having quite infertile whites.


Among the fifty states plus Washington D.C., white total fertility correlates at a remarkably strong 0.86 level with Bush's percentage of the 2004 vote. (In 2000, the correlation was 0.85). In the social sciences, a correlation of 0.2 is considered "low," 0.4 "medium," and 0.6 "high."

You could predict 74% of the variation in Bush's shares just from knowing each state's white fertility rate. When the average fertility goes up by a tenth of a child, Bush's share normally goes up by 4.5 points.

 

The rest of the article explains how people come to sort themselves out geographically by family size. It also looks at how this perspective can shed light on gun control, vouchers, and environmentalism. The Dec. 20th edition of The American Conservative should be on newsstands shortly. You can subscribe here. You can obtain the magazine electronically by subscribing to the Electronic Edition.

 

Here's the data:

 

White Total Fertility Rate Bush % 
USA      1.83 50.9%
Utah      2.45 71.1%
Alaska      2.28 61.8%
Idaho      2.20 68.5%
Kansas      2.06 62.2%
South Dakota      2.02 59.9%
Nebraska      2.02 66.0%
Oklahoma      2.01 65.6%
Wyoming      1.99 69.0%
Indiana      1.94 60.0%
Arkansas      1.94 54.3%
Texas      1.93 61.2%
Arizona      1.92 54.9%
Mississippi      1.92 59.6%
New Mexico      1.90 49.8%
Georgia      1.90 58.1%
Iowa      1.89 50.1%
Missouri      1.89 53.4%
Ohio      1.89 51.0%
Louisiana      1.88 56.7%
Michigan      1.88 47.8%
Montana      1.87 59.1%
Colorado      1.86 52.0%
Nevada      1.85 50.5%
Kentucky      1.85 59.5%
North Carolina      1.84 56.1%
Alabama      1.84 62.5%
Minnesota      1.83 47.6%
New Jersey      1.83 46.4%
Tennessee      1.83 56.8%
Virginia      1.82 54.0%
Maryland      1.81 43.3%
West Virginia      1.80 56.0%
Illinois      1.80 44.6%
South Carolina      1.80 58.0%
Florida      1.78 52.1%
North Dakota      1.78 62.9%
Wisconsin      1.78 49.4%
Oregon      1.76 47.6%
Connecticut      1.75 44.0%
Washington      1.72 45.6%
Pennsylvania      1.72 48.6%
New York      1.72 40.5%
Delaware      1.71 45.8%
New Hampshire      1.69 49.0%
California      1.65 44.5%
Maine      1.65 44.6%
Vermont      1.63 38.9%
Massachusetts      1.60 37.0%
Hawaii      1.59 45.3%
Rhode Island      1.50 38.9%
District of Columbia      1.11 9.3%

[Ethan Herdrick has kindly plotted the Total Fertility Rate - Whites and Bush's Share of the Vote on a nifty map of the U.S. -- Just hover your cursor over the two white boxes on the top and the map will flip from one variable to another. We're still fooling around a little bit with the color schemes, but you'll see that California, for example, doesn't change color because its white fertility and Bush share fall right on the best fit line.]

http://www.iSteve.com/04NovB.htm#babygaptac

More reader responses to my article "The Baby Gap:"

 

Adam Carstens of North Star Leadership Group looked up some useful information in the General Social Survey database. He found that for incomes below $50k (in 1998), white Republicans only have a very small advantage in number of children over white Democrats. But at higher incomes, Republicans have significantly more children. For example, white "Strong Republicans" with incomes of $50k or more average 2.16 children versus 1.62 children for white Democrats of either "Strong" or "Not Strong" fervency of the same income range. That's 1/3 more children. 

 

At $90k and above, "Strong Republicans"  average 2.47 children versus 2.04 kids for "Not Strong Republicans," and 1.56 for Democrats as a whole. The sample sizes are little small for slicing and dicing too narrowly, but the pattern seems apparent.

*

 

Readers write:

 

The only county-level natality data (SPSS or SAS) I know of is Local Area Summary Data Files, but you have to be a member of an ICPSR-affiliated institution to download it, or you have to be affiliated with institutions that have the data like UCLA.

 

Anybody know how I can get my hands on this data?

*

 

I wonder if it is possible to merge your lifetime fertility theory with some data about domestic migration. I have long believed that the profound domestic migration southward and westward is disproportionately made up of people prone to vote Republican (simplistically, it seems quite likely that the union workers, government workers and blacks stay behind). My guess would be that those moving consist of large corporation employees (dozens of Fortune 500 companies have moved to Texas and Georgia) and the small business entrepreneurs that spring up to service them. And guess what--the women have subordinated their careers to that the husbands can have the necessary career mobility.

*

 

You mention in passing how the cost of private schools factor into attitudes in the South, and you're correct that most whites, especially those dedicated to education, find public schools unacceptable. Part of it is that public schools have become vehicles for social engineering rather than education, and there's also less geographic distance between social classes or races than elsewhere. Tracking might offer a solution, but the educational profession won't permit it and the stakes of ending in the wrong track are just too high. While wealthy, professionals with bright, well-adjusted children might take the risk, most people won't; as you say, people want to keep their daughter's off the pole and their son's out of trouble. So I've noticed lot of working class families work an extra job to pay for tuition at a private academy, and the local academy where I live is a lot more diverse in terms of income and parental education level than the private schools my children attended in the Northeast.

Health insurance is another big issue that's as important as private schools. Few jobs provide family coverage, and the employee contribution for a family eats into monthly salary. While I haven't done the numbers, it would be interesting to see (1) whether the cost is constant across the country and (2) how it relates to average salaries in different regions. Self-employed people and small businesses have a very hard time with insurance. Living without insurance risks catastrophic costs in case of major surgery. While young singles may run the risk when they're healthy and getting by with low salaries as they enter the professional work force, its just not the sort of thing responsible parents do. And the whole point of your article is that middle class parents have middle class standards and aspirations for the children.

The cost structure of housing, health care, and schooling (including college, because there's a vast gap between institutions that have competitive and non-competitive admissions) provides a massive disincentive for families to have children. Indeed, it makes having a family a potential step toward downward social mobility unless you have two upper-end professional salaries. How can this be fixed? Or does everyone have to live in Idaho and homeschool?

*

 

Maureen Dowd's siblings and the Baby Gap: Nicely illustrating my new article "The Baby Gap: Explaining Red and Blue," snippy NYT columnist Maureen Dowd lets her ultra-Republican brother write her column for her. Maureen, of course, is an unmarried 52-year-old liberal woman who lives in Washington D.C. (average number of babies per white woman: 1.1; not coincidentally, Bush's share of the vote: 9%). The underlying theme running through her writing is her desperate effort to silence the little voice in her head that tells her she has wasted her life by not getting married and having babies.

 

Maureen comes from what I presume is a big Irish Catholic family (she's a 1973 graduate of Catholic U.) and her brothers and sisters are staunchly Republican. Her brother Kevin, a salesman, writes:

 

My wife and I picked our sons' schools based on three criteria: 1) moral values 2) discipline 3) religious maintenance - in that order. We have spent an obscene amount of money doing this and never regretted a penny. Last week on the news, I heard that the Montgomery County school board voted to include a class with a 10th-grade girl demonstrating how to put a condom on a cucumber and a study of the homosexual lifestyle. The vote was 6-0. I feel better about the money all the time.

 

Now, if only Kevin lived in suburban Virginia (a red state) instead of suburban Maryland (a blue state), the Dowd clan would fit my thesis perfectly. 

*

 

One important point that I might not have made hugely clear in my article is that this red-blue fertility breakdown probably works even better at the county level than at the state level that I used. All  states are a mix of urban, suburban, exurban, small town, and purely rural counties, so all we can do at the state-level is look at a continuum from blue New Jersey and Rhode Island at one end of the density scale to red Alaska and Wyoming at the other end. The one exception is purely urban Washington D.C., which is off by itself with an ultra-low white fertility and ultra-low Bush percentage. It's probably fairly representative of big cities, although I suspect it's a bit of a caricature. 

 

If you know where I could find fertility or family size data by county, please let me know.

*

 

Reponses to my article "The Baby Gap" (see below for an excerpt from my article):

 

One thing that I find very frustrating is the way journalists use statistics to create an argument.

TWO IMPORTANT DEFINITIONS:

- Fertility rate: the number of children born to women of child bearing years (typically described as 13-48 years old)

- Birth rate (which you do not mention, but should!): the number of births in a given year per 100,000 women

Your argument selects one statistic (FERTILITY RATE) that has no meaning unless it is correlated to the BIRTH RATE. Here's why:

Nebraska (a red state) has a relatively high FERTILITY RATE. However, the BIRTH RATE is quite low. Why? Because many YOUNG (white) WOMEN leave Nebraska and there is a larger percentage of older/non-fertile women there. Nebraska has one of the oldest populations in the country. It has fewer white children per 100,000 white people than Massachusetts has.

Utah (a red state) has the highest FERTILITY RATE, and has a VERY HIGH BIRTH RATE because Utah has tons of young people as a percentage of the population.

If you correlate BIRTH RATE into the picture, you'll be surprised to learn that the percentage of white children in states like Massachusetts and Washington are not that different from Nebraska and South Carolina and Virginia. (Utah is and will probably always be an exception because of Mormonism). You will find that white children make up a larger PERCENTAGE of the population in many blue states than they do in some red states.

I think that the failure to include BIRTH RATE into your calculations is key. You may want to run the numbers again. THE STATISTIC YOUR REALLY WANT TO USE TO SEE IF YOUR PREMISE HOLDS ANY WATER IS: The number of white children per 100,000 white residents.

 

Thanks. Very helpful and informative.

However, I think the lifetime total fertility of white women remains the statistic of relevance because it captures precisely the kind of behavior I'm interested in, which is: what causes people to move from one type of state to another or what causes them to stay home. My model suggests that the key figure in terms of affecting social and political attitudes is not the % of children in a population but the average number of children per family.

As you say, Nebraska is losing many of its more career ambitious young women to blue states, leaving behind the ones more oriented toward having bigger families (and Nebraska also picks up a few refugees from high density blue states looking for a cheap place to raise a large family). This churning of the population will have a big impact on the type of adults living in Nebraska and on their social and political attitudes, which they will likely continue to espouse even after their children eventually leave home.

In contrast, Massachusetts attracts a lot of young people for college or technology jobs, but has a hard time holding on to the ones who want more than one or two children and are willing to give up the more sophisticated culture of an adult-oriented state to move to a family-oriented state.

That's why total fertility works so well as a predictor of Bush's vote in the last two elections, while share of the population that is children does not.

Indeed, I think the statistics actually underestimate the extent of the family size gap between the states. 

First, white women in red states tend to have their first child at an earlier age. I believe blue Massachusetts recently became the first state ever where the age of first time mothers was over 30. So, people in red states tend to spend less of their lives as single and/or childless. 

Second, young urbanites often have one or even two children in the city, and then flee to cheaper and safer surroundings when they decide the toddlers need a yard or are ready to start school. So, the births get credited to the blue region but the bulk of the child-rearing is done in the red region.

*

 

http://www.iSteve.com/04NovB.htm#bgup

The Importance of the Baby Gap Is Growing: In my American Conservative article, I pointed out the extraordinarily high correlation between Bush's share of the vote by state in 2004 and the states' total fertility rate (estimated average number of babies lifetime) for white women in 2002: 0.86. You square that number to get the percentage of the variation in Bush's share predictable from the fertility rate: 74%. Has this always been the key to explaining the outcome of Presidential elections by state?

 

To find out, I've gone back into the National Center for Health Statistics document (warning: it's a big PDF), and pulled out the corresponding white TFRs for 1995 and 1990. (I couldn't find the figures for other years). If we use the 1990 white fertility rates by state and compare them to George H.W. Bush's share of the vote by state in his 1988 victory over another Democrat from Massachusetts, Michael Dukakis, then the correlation, while still strong, is significantly lower than in recent years: r=0.71, r-squared=51% versus the 74% in 2004. So, fertility by state was only about 70% as strong a factor 16 years ago compared to this year's election. The Baby Gap has always been a big deal, but it's turning into a bigger deal.

 

Bush the Elder's correlation with the 1990 TFRs dropped sharply in 1992 down to an r-squared of only 28%, but that was mostly because of Ross Perot's strong run. If we sum Bush's and Perot's shares, the R-squared goes back up to 59%, up from 51%. 

 

Dole was up to 61% in 1996, correlated with the 1995 TFRs, and Dole + Perot was at 68%. 

 

Bush the Younger hit 73% in 2000 versus the 2002 TFRs, and 74% in 2004. 

 

Election Share of Vote Non-Hispanic White Total Fertility Rate  r  r-Squared
2004 Bush 2002 NHW TFR        0.86 74%
2000 Bush 2002 NHW TFR        0.85 73%
1996 Dole + Perot 1995 NHW TFR        0.82 68%
1996 Dole 1995 NHW TFR        0.78 61%
1992 Bush + Perot 1990 NHW TFR        0.77 59%
1992 Perot 1990 NHW TFR        0.53 28%
1992 Bush 1990 NHW TFR        0.53 28%
1988 Bush 1990 NHW TFR        0.71 51%

 

Another question is whether changes in fertility per state are driving changes in voting behavior by state? The answer appears to be: a little, but not a huge amount. The correlation between change in white TFR from 1990 to 2000 and change in Republican share of the vote from 1988 to 2004 is only 0.31 or 9%. 

 

However, if this was weighted by population size of the states, it might be more impressive because of the huge change in California. Nationally, white fertility is down 1% from 1990 to 2002, but in California, it plummeted 14%. From 1988 to 2004, the GOP candidate's share of the vote dropped 2 percentage points nationally, but 7 points in California.

 

Generally speaking, white fertility from 1990 to 2002 has dropped the most in the Far West and upper New England. It has grown the most in Washington D.C., New Jersey, and Connecticut. My guess is that the big drop in crime with the end of the crack epidemic made the cost of insulating children a little less in those densely populated areas. 

***

 

One interesting question is how much differences in the abortion rate account for differences in the birth rate by state. A reader sent me this table, which is for all races. Another complication is that abortion statistics are often recorded by the state of the clinic not the state of the client. So, a woman from red West Virginia might drive into blue Pennsylvania for an abortion. Still, by looking at mostly white blue states like Vermont and Oregon versus mostly white red states, it does appear that pro-life red states do indeed practice what they preach -- the denizens of Nevada being an obvious exception to that rule. (I continue to be amazed that large numbers of parents of little girls are moving their families to booming Las Vegas. As Chris Rock says, "Fathers, your prime duty is to keep your daughters off The Pole.")

 

I was a little surprised by this. I sort of expected Massachusetts women to be well enough organized that they wouldn't have many unwanted pregnancies, but that turned out not to be true.

 

Per Capita - No. per 1,000 women aged 15-44
State Pregnancies Births Abortions
US Total 103 65 23
Dist. Of Columbia 166 62 83
Nevada 135 76 41
New York 122 65 40
New Jersey 118 65 36
California 176 75 33
Maryland 108 60 33
Florida 111 65 31
Massachusetts 100 57 29
Hawaii 117 73 27
Illinois 111 69 25
Rhode Island 94 57 24
Connecticut 99 62 23
Michigan 97 61 22
Washington 98 62 21
Virginia 93 58 21
Arizona 115 77 21
Oregon 97 63 20
Georgia 99 65 20
Texas 112 75 20
Colorado 97 64 19
North Carolina 96 63 19
Delaware 91 60 18
New Hampshire 83 53 17
Alaska 104 72 17
New Mexico 105 72 17
Pennsylvania 86 57 16
Ohio 90 60 16
Wyoming 89 60 16
Vermont 77 50 15
South Carolina 87 60 15
Alabama 91 62 15
Indiana 91 63 14
Mississippi 95 66 14
Missouri 89 62 14
Tennessee 88 61 14
Louisiana 93 65 13
Minnesota 87 61 13
Montana 85 59 13
Wisconsin 84 58 13
Arkansas 94 67 12
Kansas 92 65 12
Maine 73 50 12
Oklahoma 91 65 12
Nebraska 89 64 11
Iowa 84 61 10
Idaho 96 71 9
Kentucky 81 59 9
West Virginia 72 53 9
North Dakota 81 61 8
South Dakota 88 67 8
Utah 116 89 8

 

It's not clear whether legalized abortion actually reduces the total fertility rate among whites by very much. A recent Rand Corp. study estimated that outlawing abortion would raise the average number of babies per white woman only from 1.83 to 1.89. Without legal abortion to fall back on in case of unplanned pregnancies, white people would plan better and thus avoid unplanned pregnancies, according to the Rand researcher.

 

Ethan Herdrick has done another US state map visualizing the information in this table.

*

The American Conservative's Film Critic, Steve Sailer also writes for VDARE.com and iSteve.com.

 

iSteve.com

email me

 

 

 

 

American Conservative

iSteve.com

email me

 

 

 

American Conservative

iSteve.com

email me

 

 

 

American Conservative

iSteve.com

email me

 

 

 

American Conservative

iSteve.com

email me

 

 

 

American Conservative

iSteve.com

email me

 

 

 

American Conservative

iSteve.com

email me

 

 

 

American Conservative

iSteve.com

email me

 

 

 

American Conservative

iSteve.com

email me

 

 

 

American Conservative

iSteve.com

email me

 

 

 

American Conservative

iSteve.com

email me