VoxEU Column Politics and economics

Sex, money, red states and blue states

Cultural generalisations are dangerous. Sometimes, however, statistical relationships are so strong that it is worth pondering their significance. Are some groups more likely to take responsibility for their personal behaviour than others with respect to their sexual behaviour, physical fitness, financial dependency on the federal government? The fearless social scientist will not shrink from confronting these questions.

Mitt Romney, the Republican candidate for US president, said in now-infamous comments that 47% of the electorate is dependent on the federal government, that he will never be able to teach them to take personal responsibility for their lives, and that they are certain to vote for Barack Obama in November. He continues a tradition in his party that goes back at least three decades: building political campaigns around the proposition that folks in the heartland exhibit the American virtues of self-sufficiency and personal responsibility, with the implication that more urban regions display decadent social values and dependency on government.

It is a good general rule to judge people on their own merits and not on the supposed attributes of the racial, socioeconomic or geographic groups to which they belong. Cultural generalisations are dangerous. But since questions have been raised, the fearless social scientist will not shrink from confronting them. Are residents of 'red states', who tend to vote Republican, indeed more likely to take responsibility for their personal behaviour than those who live in 'blue states' and tend to vote Democratic?

The statistical reality is that the red-staters are, on average, less prone to pay income taxes, more prone to receive subsidies from the federal government, less physically fit, less responsible in their sexual behaviour, more prone to inflict harm on themselves and on others through smoking, drunk driving and misuse of firearms, and more prone to freeride on the healthcare system, compared to blue-staters.

Economists have long known that, in spite of the rhetoric about 'getting the government off our backs', the red states receive more federal spending, net of taxes, than the blue states. Alaska, Mississippi, Louisiana, West Virginia, and the Dakotas topped the list of moochers in 2005. Despite Romney’s comments, it is the states with high percentages of people who pay no income tax that tend to vote Republican. Mississippi is number one, followed by Georgia, Alabama, Arkansas and South Carolina. The Democratic-leaning states of New England (comprising Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont), New York, New Jersey, and California are the ones paying into the federal treasury and subsidising everyone else. Updated data show the same pattern in 2007. The mid-Atlantic states, New England, Minnesota, and Illinois are the biggest net givers. Alaskans are the most dependent on the federal government, receiving $7,448 in spending (net of taxes) per capita. (That is not even counting the handouts of oil revenues that they get from their own state government). Those who claim to be fiscally conservative are the ones who in truth tend to be the biggest sources of deficits vis-à-vis the federal government.

Figure 1 puts on the horizontal axis each state's tax receipt of federal government spending, net of taxes paid in, per capita. The vertical axis shows the ratio of Democratic to Republican votes by state. The red states (low in the graph) tend to be on the receiving end (right of zero). The blue states (high in the graph) constitute a majority of the ones that foot the bill (positive contributions to the nationwide kitty). The relationship is of great statistical significance.

Figure 1. Federal spending received, minus taxes paid, by blue and red states

Note: Average of votes in 2000, 2004 and 2008 presidential elections.

Most people don’t know that similar relationships hold regarding various other measures of personal responsibility. The numbers show that states where residents suffer more from obesity, in part because they get less physical exercise and have worse eating habits, tend to be the states that vote Republican. To illustrate, a mere one percentage point decrease in a state’s obesity rate is associated on average with an estimated increase in the ratio of Democratic to Republican voters from 1.00 to 1.07 – enough to make the difference in many an election. The statistical confidence interval is slim enough to exclude a zero effect. A physical fitness index, combining absence of obesity with physical exercise, shows a strong relationship with voting. (See statistical appendix at my website.)

Figure 2. Obesity (percentage of population) among blue vs. red states

Note: Average of votes in 2000, 2004 and 2008 presidential elections.

Figure 3. Smoking rates (adults) among blue and red states

Note: Average of votes in 2000, 2004 and 2008 presidential elections.

Other measures of personal responsibility show similar patterns. States with high rates of smoking vote Republican too, as Figure 3 illustrates. States with high rates of fatal accidents from drunk driving vote Republican (Figure 4). States with high rates of firearms assaults – not gun ownership, but armed assaults – vote Republican (Figure 5). Again, these relationships are highly significant statistically. An index of risky behaviour, combining the drunk driving and firearms assaults numbers, shows strong relationship (appendix).

Figure 4. Drunk driving fatalities among blue and red states

Note: Average of votes in 2000, 2004 and 2008 presidential elections.

Sex is interesting. Notwithstanding fervent proclamations about the importance of abstaining from sex before marriage, evidence suggests that young people in red states do not act the way they talk. They do have sex before marriage, and it is less likely to be safe sex than among those in blue states. States that vote Republican have higher pregnancy rates among girls aged 15 to 17, as Figure 5 shows (the difference is highly significant statistically). They also have higher rates of the sexually-transmitted disease chlamydia (this difference, unlike the others, is not statistically significant at the aggregate state level, but is when combined into an overall measure of unsafe sex).

Apparently the gap between what they say and what they do is particularly wide for teenagers who describe themselves as evangelical Christians. According to research by Mark Regnerus (a sociologist at the University of Texas, Austin), white evangelical adolescents usually (74%) state a belief in abstinence before marriage, but in fact are surprisingly more active sexually, compared to mainline Protestants and Jews who do not tend to state such a belief. When the believers in abstinence do engage in sex, they are less likely to use contraception than others. The gap between word and deed is high for the millions of teenagers who take a formal pledge to remain celibate until marriage, typically in a ring ceremony, according to “Red Sex, Blue Sex,” a New Yorker article by Margaret Talbott. The majority of them, though holding out for a while, “end up having sex before marriage, and not usually with their future spouse.” Two other sociologists, Peter Bearman (Columbia University) and Hannah Bruckner (Yale) find a positive correlation between the abstinence pledge and sexually transmitted diseases. Pledgers are less likely to use a condom if and when they first have sex and overall are slightly more likely to contract a sexually transmitted disease. (Under George W Bush, the federal government subsidised such abstinence pledge programs despite their questionable effectiveness.)

Figure 5. Teenage pregnancy rates (among girls aged 15 to 17) among blue and red states

Note: Average of votes in 2000, 2004 and 2008 presidential elections.

Utah is the most conspicuous outlier in most of these relationships. It has a high population of Mormons. Apparently they follow the strictures of their religion more closely than those of other religious denominations. (Could this be why evangelicals tend to resent Mormons so much, according to opinion polls?) But, on average, the relationship holds nonetheless.

It is particularly striking that the states where the most residents exhibit behaviour that endangers their health and that of others – with many of these unhealthy people later freeriding on their fellow citizens when they show up uninsured in the hospital emergency room – are also the states where congressmen tended to vote against the Affordable Care Act (Obamacare) in 2010. (See Figure 1 in my blogpost on Obamacare or the graph in an op-ed, or associated regressions in appendices.) This risky behaviour includes poor physical fitness (as measured by rates of obesity, lack of exercise, and poor diet), careless sexual behaviour (as measured by rates of pregnancy among girls aged 15 to 17 and rates of chlamydia), smoking, drunk driving (as reflected in fatalities) and irresponsible use of guns (as reflected in armed assaults). Each obese American incurs medical costs 42% higher than those of normal weight. Often others are stuck with the bill because they are uninsured. They may, for example, be unable to get health insurance because they are so overweight. These people are freeriders on the healthcare system even if they don’t want to be. Obamacare was designed to fix this problem.

Most US citizens don’t know what the Affordable Care Act does. Many think that it reduces personal responsibility for healthcare, but the truth is the opposite. Under the current system, hospitals are required to treat patients who show up at the emergency entrance with a heart attack – even if their condition is partly their fault, due to habits of overeating and under-exercising. The hospitals have to pass the costs on, and the rest of the US ends up footing the bill. The individual mandate is designed to fix that, by making everyone pay for the health care they get. One benefit is that it will encourage them to see a doctor, who will typically advise them to adopt a healthy lifestyle by exercising, eating better, stop smoking, and deal with alcoholism. Establishing personal responsibility, not socialised medicine, is the reason why conservative think tanks such as the Heritage Foundation proposed the idea of the universal mandate in the first place, and why Mitt Romney enacted it in Massachusetts when he was governor.

What makes all these patterns so remarkable of course is that the central animating principle in the campaigns of most red-state political candidates is the claim that they, representing the American heartland, believe in personal responsibility, family values, and “getting the government off our backs,” while the decadent blue states are accused of being free-riders who lack a sense of social morality.

It seems to me that, until now, 'liberals' who live in the blue states have, out of politeness, held back from suggesting that those who live in red states exhibit behaviour in their personal lives that falls short of the conservative rhetoric of personal responsibility in which their politicians revel. It would be unseemly and elitist to point fingers at any category of fellow Americans and imply that they are promiscuous, fat, gluttonous, lazy, uneducated, or that they are more prone to divorce and shootings. (Residents of blue states tend to be more educated and to have higher incomes than residents of red states, thus qualifying us as 'elites'.)

Charles Murray features such statistics in his book, Coming Apart. Murray argues that those who live in the 'super zip codes' – the areas with high levels of income and education – are maintaining the traditional American values of hard work and family values. Those who live elsewhere show 'crashing' rates of industriousness and marriage. (His statistics look at white people only, so as to control for race.) Earlier studies had documented that divorce rates have declined among blue-state households, while they have continued to rise in red states. Carbone and Cahn (2010): “The areas of the country most committed to traditional values have the highest divorce and teen pregnancy rates.” (The decline in divorce is associated with educated women who delay marriage in the first place. Carbone and Cahn 2012, and Goldin and Katz 2002.)

Many academic researchers and news media fear accusations of liberal bias. Murray may be immune from this fear. He is well-known as a conservative/libertarian whose book The Bell Curve dealt with black-white differences in test achievement.

While some reviewers have taken away from Murray’s book the message that those who live in the super zip codes should feel guilty for their virtuous qualities, professional achievements, and economic success, Murray is not worried that he will be accused of elitism. His criticism of those living in the super zips is quite different. Although they tend to live responsible and virtuous lives, they are failing to exhort others to do the same: “The new upper class still does a good job of practicing some of the virtues, but it no longer preaches them. It … preaches non-judgmentalism instead. Non-judgmentalism is one of the more baffling features of the new-upper-class culture. The members of the new upper class are industrious to the point of obsession, but there are no derogatory labels for adults who are not industrious” (p.289). The problem is “an unwillingness on the part of any significant portion of the new upper class to preach what they practice”.

This judgment is remarkable. Many residents of red states have been doing the opposite, i.e. preaching personal responsibility, family values, self-reliance and small government, while engaging in risky behaviour, failing to form stable families, and reaping federal subsidies. Surely a failure to practice what you preach is more objectionable than a failure to preach what you practice. But red-state politicians have long used this hypocrisy as a successful strategy. And they use it against responsibility-promoting policies to help their own citizens, such as sex education, Obamacare, and support for education.

Do these statistical relationships between personal responsibility and voting behaviour have analytical implications? How can one explain such counter-intuitive results? I have pondered this puzzling question. I am still not sure of the answer. But here is what I have to offer.

One major reason for countries having trouble maintaining fiscal discipline is that voters consider the specific government spending that benefits them personally to be well-deserved while considering most other spending to be wasteful. Florida retirees believe in social security benefits, Midwestern farmers believe in agricultural price supports, Californians believe in water subsidies, Texans believe in oil subsidies, coastal communities believe in fishing subsidies, Michigan auto workers believe in support for their industry, and so forth.

It stands to reason that some people are less self-aware than others, and less knowledgeable on how the budget adds up, for whatever other reason. We hear repeatedly of seniors who tell politicians to keep the federal government away from their medicare, of ranchers who support the Tea Party or right-wing militia while collecting farm subsidies.

The people who suffer the biggest gap between their perceived and actual share of the federal pie are likely to be getting a disproportionate share and yet to believe the opposite. If they believe that others are getting more than they themselves are, they are more likely to buy into the angry belief that other social groups are freeriding on society, and the ideology that government spending is wasteful and needs to be cut back, without realising that this includes the benefits they themselves receive. Perhaps these people are more likely to vote for Republican politicians, who tell them what they want to hear. It’s a theory, anyway.

This article draws in part on an op-ed published by Project Syndicate. Statistical details are available here.


CM Atlas, TW Gilligan, RJ Hendershott , 1995, “Slicing the federal government net spending pie: Who wins, who loses, and why”, American Economic Review, Vol. 85, No. 3, Jun..

Hannah Brückner, and Peter Bearman (2001). “Promising the Future: Abstinence Pledges and the Transition to First Intercourse”, American Journal of Sociology, 106: 859-912.

Hannah Brückner and Peter Bearman, “After the promise: the STD consequences of adolescent virginity pledges”, Journal of Adolescent Health 36 (2005) 271-78.

Naomi Cahn and June Carbone, Red Families v. Blue Families: Legal Polarization and the Creation of Culture, 2010 (Oxford University Press).

June Carbone and Naomi Cahn “Red vs. Blue Marriage,” Jan. 30, 2012 SSRN paper no. 1995822.

Jeffrey Frankel, 2003, “Republican and Democratic Presidents Have Switched Economic Policies”, Milken Institute Review. vol.5, no.1, 1st Quarter, pp.18-25.

Jeffrey Frankel, 2012, "Obamacare champions personal responsibility. The states that hate it don't", Christian Science Monitor, Sept. 6.

Claudia Goldin & Larry Katz. (2002). “The Power of the Pill: Oral Contraceptives and Women’s Career and Marriage Decisions”, Journal of Political Economy, 110, 730-770.

R Hingson, et al, 1996, “Reducing alcohol-impaired driving in Massachusetts: the Saving Lives Program”, American Journal of Public Health, June, 86, no. 6.

Charles Murray, 2012, Coming Apart: The State of White America, 1960-2010 (Crown Forum: NY).

Mark Regnerus, 2007, Forbidden Fruit: Sex and Religion in the Lives of America’s Teenagers, Oxford University Press.

Mark Regnerus and Jeremy Uecker, Premarital Sex in America: How Young Americans Meet, Mate, and Think about Marrying, Oxford University Press, 2011. esp. Ch.7 “Red Sex, Blue Sex: Relationship Norms in A Divided America”, pp.205-35. 

14,068 Reads