My Enemy's Enemy

The headline at the top of the front page of this morning's Pittsburgh Post-Gazette said, “Romney booed before NAACP.” His speech was described as “a blunt pitch,” and as “offering tough policy prescriptions that are unpopular with [African-Americans].” You've probably seen this video of Romney being booed on the nightly newscasts. What's going on?

Although the June polls show substantial drops in support for Obama among almost all demographic groups, he still has a huge lead among African-Americans. It's unlikely that Romney hoped to win many converts among those attending yesterday's NAACP convention. The real target of Romney's speech was the general public who read and watched yesterday's news.

In 1946, Fritz Heider, one of the pioneers of social psychology, proposed a theory of cognitive balance. He pointed out that people prefer both their interpersonal relationships and their cognitive structures to be in a state of balance, or consistency, with one another. We expect to like and dislike the same people as our friends, and to share their positive and negative attitudes. To see how this works in interpersonal relations, consider a three-person social network consisting of Alice, Betty and Carol. If Alice likes Betty, Alice likes Carol, and Betty likes Carol, the system is in balance.

A + B, A + C, B + C

It is also balanced if Alice likes Betty and dislikes Carol, and Betty also dislikes Carol.

A + B, A – C, B – C.

On the other hand, the network is unbalanced if Alice likes Betty and Carol, but Betty dislikes Carol; or if all three of them dislike one another.

A + B, A + C, B – C; and A – B, A – C, B – C

A simple mathematical rule for determining balance is to multiply the signs of the relationships. If the product is positive, the system is balanced [(+) (–) (–) = +], and if it is negative, the system is unbalanced [(+)(+)(–) = –]. Systems that are unbalanced are unstable, and there is cognitive pressure to balance them by changing one of the three relationships.

The system also makes predictions about new and developing relationships. For example, suppose Alice and Betty have a mutual relationship, Alice has an attitude toward Carol, but Betty and Carol are unacquainted. If Alice and Betty are friends and Alice dislikes Carol, Betty can balance the system by disliking Carol. Among the three other possibilities is this one: If Alice and Betty dislike one another and Alice dislikes Carol, there is a tendency for Betty and Carol to be attracted to one another. This latter principle is illustrated by the saying “my enemy's enemy is my friend.” The U. S. and the Soviet Union were allies during World War II not because we liked Communism or Stalin, but because the Soviets were also fighting against Nazi Germany.

This may shed some light on Romney's motive for attending the NAACP convention and going out of his way to antagonize those in attendance—including referring to the Affordable Care Act as “Obamacare,” which many of Obama's supporters find insulting. I suggest he went there expecting and hoping to be booed, knowing that the corporate media would act as a megaphone for any rude treatment he received.

There are many Republicans and Independents who are unenthusiastic about Romney. Tea Party members, for example, probably would have preferred a different candidate. Many of these folks also have strong attitudes toward African-Americans, and they aren't very pretty. A University of Washington survey finds that Tea Party membership predicts a 25% increase in racial resentment, even after controlling for conservatism and Republican partisanship—which are both also predictive of racial resentment. If black people are disliked (my enemy), and if black people booed Romney (my enemy's enemy), some of these undecided folks may have found a new friend. Of course, yesterday's events might also have energized members of the Republican base who were already committed to Romney.

In a Fox News interview later in the day, Romney indicated that the boos he received were not a surprise. “We expected that, of course,” he said. The NAACP didn't let him down. Take a look at his facial expression as he was being booed.

The Media Select Their Candidate

Mitt Romney has charged that the news media are biased in favor of Barack Obama and against him. As usual, the facts are otherwise. (See my previous posts on media bias here and here.) It's not clear whether the Elephants actually believe the liberal bias claim, or are “working the refs.”

This week, the nonpartisan Pew Research Center's Project for Excellence in Journalism released is first survey of 2012 campaign coverage in over 11,000 news outlets. The survey analyzed two aspects of news coverage: (1) the frames used to cover campaign news between November 1, 2011 and April 15, 2012, and (2) the tone (positive, negative, or neutral) of the coverage received by each candidate, which was analyzed only from January 1 to April 15. Framing refers to the media's tendency to select certain aspects of the campaign and make them more salient, as if to suggest what is important about presidential elections. Frames were determined by human coders. Tone was analyzed with a computerized content analysis program which recognizes positive, negative and neutral words and phrases. (When the tone of a sample of stories is coded by researchers, the computer and the human coders agree 97% of the time.)

As in previous elections, the media are framing the campaign as a horserace. Pew calls this the strategy frame, and it includes reports on momentum, polling, advertising and fundraising. This frame accounted for 64% of campaign stories. Critics charge that the strategy frame does not give voters the information they need to determine the candidates' positions on issues. That information is included in the categories of domestic issues, foreign issues and the candidate's public record, which collectively account for 16% of the coverage. (Of course, in Romney's case, his past public record is all but meaningless for determining his current position on issues.)

The tone of the news stories about Obama and Romney is summarized below:

	Romney	Obama
Positive	39%	18%
Neutral	29%	34%
Negative	32%	34%

Contrary to Romney's claims, he has received much more favorable coverage than President Obama. In fact, I find it hard to imagine that a sentient being could read, listen to or watch the corporate media and actually believe they favor the President.

As shown above, the tone of stories about Romney has changed over time. They became much more favorable after the Michigan primary on February 28, when it became clear that Romney would be the Elephant nominee. The coverage of President Obama, as you can see, has been unrelentingly negative throughout all 15 weeks of the survey.

My interpretation of these data is fairly straightforward. The owners of the corporate media in this country are overwhelmingly wealthy and politically conservative, and their preferences exert a compelling influence over the behavior of reporters and editors. Previous research has shown that the media traditionally favor candidates from the Elephant Party. This accounts for the negative tone of stories about Obama. Prior to February 28, the media owners were probably split over which Elephant candidate they preferred. As long as Gingrich and Santorum were in the race, they ran quite a few negative stories about Romney. However, once Romney's nomination was inevitable, media ownership closed ranks behind the nominee of their party.

I find it particularly interesting that the media frequently report charges of liberal bias made by Elephant candidates, but almost never report studies such as this one which exonerate them of the charge.

There is no reason to think that either the predominant use of the strategy frame or the tone of the coverage of the candidates will change for the remainder of the campaign. It is also likely that the media will continue to report whatever the candidates say without evaluating whether their statements are true or false, and that they will not investigate and report whether the candidates' current statements are consistent with their past public record. All of these policies will translate into a significant advantage for Romney.

Romneycare: The Bottom Line

Sometimes we get so caught up in the details of health care reform that we forget the real purpose—to improve people's health. A new study by Charles Courtemanche and Daniela Zapata, published by the National Bureau of Economic Research, examines the health effects of Massachusetts' 2006 health care reform bill, commonly called “Romneycare.” This effects of this reform are important because it is nearly the same as the Affordable Care Act (ACA).

It may sound silly to ask whether expanding people's access to medical care improves their health, but the outcome of the study was not obvious. Critics of health care reform sometimes claim that people will not take advantage or make good use of medical care and their health will be unchanged. More importantly, the moral hazard hypothesis claims that when health care is available at low cost, people will take unnecessary health risks, such as smoking and overeating. The result could be a decline in their overall health, accompanied by an increase in medical spending.

Previous studies have generally found improvements in health following increased access to care. The most impressive is the Oregon Health Study, which found a 13% increase in the number of people reporting themselves to be in good or excellent health following enrollment in Medicaid.

The data for this study come from a survey conducted by state health departments and the Center for Disease Control between 2001 and 2010. It includes answers from 2.8 million respondents from all 50 states and the District of Columbia. The study is a time series design. Health care reform in Massachusetts went into effect in March 2006, but the implementation was gradual and was not completed until July 2007. The research question is whether there were any changes in the health of Massachusetts residents around those times that did not occur in other states. The analysis controls for irrelevant variables such as age, income and marital status.

The main dependent measure was a self-report health question asking respondents to classify their overall health as either poor, fair, good, very good, or excellent. The results showed an improvement in health while the reform was being implemented (from April 2006 to July 2007), and approximately twice as large an improvement after it was fully implemented. These changes did not occur in other states at the same time. To put this into perspective, it is estimated that 1.4% of the population went from being in either poor, fair or good health to either very good or excellent health. Considering the overall cost of the program, Massachusetts spent $9,782 per year for every individual whose health improved from poor, fair or good to very good or excellent. Of course, having more people in very good or excellent health might save money in the long run.

This analysis includes everyone. However, if you look specifically at those people who acquired health insurance as a result of the reform, their probabilities of being in poor, fair or good health went down by 6.2%, 9.8% and 8.5% respectively, while their probability of being in very good health and excellent health went up 8.5% and 16%. This is comparable to the results of the Oregon study.

The main concern about this measure is that self-report questions are subjective and can be influenced by various biases. For example, people might have expected their health to improve due to the reform. On the other hand, access to medical care might make people more aware of the health problems they have. Therefore, the survey included several other measures. People were asked the number of days out of the past 30 that they were not in good physical health, that they were not in good mental health, and that they experienced health-related functional limitations. They were also asked the number of minutes per week they spent in moderate and vigorous physical activity, whether they experienced joint pain, and whether they smoked. Finally, their body mass index (weight/height²), or BMI, was calculated. Since these questions are more specific, they should be less subject to bias.

The results showed significant improvements on all of these measures with the exception of vigorous exercise and smoking. The fact that smoking did not increase and that BMI was reduced casts doubt on the moral hazard hypothesis that people would take more health risks. In fact, the overall pattern suggests that people were heeding medical advice.

Finally, internal analyses showed that, while almost every subgroup showed improvements in health, women improved more than men, and people between 55 and 64 (the oldest group not eligible for Medicare) showed the greatest improvement. Those in the lowest income category, who were eligible for a state subsidy to help purchase their insurance, showed greater improvement, and blacks improved more than other races. The authors estimate that health care reform reduced black-white health inequality by 21.5%.  

Does this study predict a positive effect of the ACA on health if it is implemented? Maybe. But the ACA includes cost-cutting measures that were not part of the Massachusetts plan, such as reductions in Medicare spending, which might reduce the gains that would otherwise be expected. On the other hand, Massachusetts had one of the lowest percentages of uninsured citizens of any state, so implementing these same reforms nationwide might result in a greater improvement in the health of the nation.

I presume these successful results will be embarrassing to Governor Romney, who has repudiated his own health care reform in search of the approval of the bewildered herd of Elephants.  

Correction.  If you read "What the Data Say" yesterday, please note that I have corrected a mathematical error.  The effect of the error was to make the increase in 2012 campaign spending appear greater than it actually was.  I should have followed my own advice and thought more slowly.

Romney's Gaffe

Governor Mitt Romney's statement, "I like being able to fire people," has been widely misinterpreted.  He did not say that he enjoys firing his employees. The reality is much worse. The remark was made in the context of a discussion of health insurance. It shows how out of touch Romney is with the health insurance system in this country.

Please read this comment by Aaron Carroll, from The Incidental Economist blog.  He points out why most Americans can't fire their insurance company. I have nothing to add. I just wanted to call it to your attention.