Found while looking for something else

In the current budget debate, it seems like Republicans are convinced that the public wants budget cuts and doesn't mind tax cuts for high earners.  Democrats don't seem to have the same conviction about what the public wants.  Of course, this gives Republicans an advantage in bargaining.  This observation led me to wonder if there were survey questions about what people thought that most people thought.  There are a few, which I will talk about in a later post, but while searching, I was diverted by a question in a Gallup poll from 1949:

"How do you think that most people vote on a candidate?  Do they vote for a man they think will do the most good for the country as a whole--or do they vote for a man they think will do the most good for them and other people like them?" 

28% said voters thought about the country as a whole, 66% said themselves, 5% didn't know, and 1% gave a qualified answer. 

The biggest group differences involved race and education. 

                  Country    Self    DK      TOTAL

Whites              27%       67%     5%      100%

Blacks              40%       47%    12%      100%

Not HS Graduate     33%       59%     7%      100%

HS Graduate         27%       69%     2%      100%

College             17%       79%     2%      100%

Missing Data and Missing Research

John Taylor, a distinguished economist at Stanford, observes that there is a strong relationship between investment and unemployment and suggests that "the anti-business sentiment coming out of Washington" has been keeping investment down.  Paul Krugman replies that investment is low simply because the market is bad--unemployed people don't buy much.  The stuff about business being spooked by what the government is doing or saying is "just politically motivated mythology."

This is an old issue, one that's been debated since at least the 1930s, so you'd think that people would have collected data.  I did a search and to my surprise found very little.  The only data I found for the United States that covered a significant period is the NFIB Optimism Index, which has been collected since the 1970s.  Changes between 1979 and 2009 can be found in this report.  It would be pretty straightforward to look at the relations among investment, the optimism index, and government policy, but I couldn't find any research that tries to do that. 

Another look

After posting the figure on height and satisfaction with life, I started wondering if the model (a cubic polynomial) might be hiding some other differences between men and women.  The figure above gives average satisfaction by height (e. g. averages for people who said that were 65, 66, 67... inches tall), smoothed in order to reduce the effect of sampling variation.  It confirms that satisfaction drops for the very tallest people, for both men and women.  The people who are happiest, on the average, are tall but not exceptionally tall.

 But it also shows something that's hard to see in the polynomial estimates:  for women, there is a substantial range, from about 65 to 75 inches, in which height makes very little difference.  That is, short women are less satisfied with their lives, but there's little difference between women who are just a little taller than average and those who are a lot taller than average.  For men, there's a clearer peak at about 74 inches--that is, men who are tall are more satisfied than men who are pretty tall, who are more satisfied than men of average height.

The biggest losers

The Census Bureau recently reported that Detroit has lost about a quarter of its population in the last decade.  I started wondering about which cities had lost population over a long period of time.  Of the 100 largest cities in 1900, 15 are smaller today (the population of the United States is about four times as large as it was in 1900).  Another five have grown by less than 5%.  

1.  St. Louis, MO      -38%

2.  Hoboken, NJ        -31%

3.  Scranton, PA       -29%

4.  St. Joseph, MO     -26%

5.  Wheeling, WV       -26%

6.  Buffalo, NY        -23%

7.  Troy, NY           -22%

8.  Wilkes-Barre, PA   -21%

9.  Fall River, MA     -13%

10. Holyoke, MA        -12%

11. Harrisburg, PA      -5%

12. Wilmington, DE      -4%

13. Pittsburgh, PA      -3%

14. Albany, NY          -0%

15. Providence, RI      -0%

16. Covington, KY       +0%

17. Cincinnati, OH      +2%

18. Reading, PA         +3%

19. Utica, NY           +3%

20. Camden, NJ          +4%

Pittsburgh would be in second place, with about a 35% loss, if you include the population of Allegheny, a city which it absorbed a few years later.  Allegheny was the 27th biggest city in its own right in 1900.

The big surprise to me was St. Joseph, which was the 34th largest city in the country in 1900.  That didn't last long--it lost about a quarter of its population, and has stayed about the same ever since.  Another surprise was that Providence was the 20th biggest city in the country in 1900.  Its estimated population in 2009 was within 100 of its population in 1900. 

Height and happiness

A recent article in the New York Times looked at the relationship between height and happiness.  It said that taller people were happier, but the relationship was stronger for men.  The results didn't actually involve happiness in the usual sense, but scores on the Gallup Index of Well-Being, which combines a numbers of things, ranging from how you felt yesterday to how often you'd visited the dentist in the last year.  I don't like indexes like this--feeling happy and visiting the dentist are both good things, but they're not the same thing.  So I looked at the relationship using answers to the question "In general, how satisfied are you with your life?" from the 2008 Behavioral Risk Factors Surveillance System.  Here's a graph showing the estimated relationship (using a cubic polynomial regression) for men and women:

The red line represents women and black represents men.  The figure shows that taller people report being more satisfied with life, up until they are about six inches taller than average for their sex.  Beyond that point, satisfaction declines (I wouldn't take the exact shape of the predictions too seriously for very tall people, since there aren't that many in the sample, but there's clearly some decline).  As the Times article said, the relationship seems to be stronger for men. 

Free Speech

Americans are very positive about the Constitution in general, but not about some of the things that are actually in it.  For example, here's a question from a Gallup Poll taken in January 1941:  "Do you think newspapers should be allowed to take sides in their editorials during election campaigns?"  59% said yes, 33% said no, and 7% were undecided. 

But more educated people are more likely to favor civil liberties, or at least say that do, and that's the case here--about 80% of college graduates said it should be allowed.  So given the increase in educational levels, we might reasonably expect support to be much higher today.  There are no recent results, but the question was asked in 1985.  At that time, only 45% said that newspapers should be allowed to take sides, 48% said they shouldn't, and 7% didn't know. 

Ignorance, madam, pure ignorance

The idea that Barack Obama wasn't born in the United States has shown a lot of staying power, but few surveys have asked about it. One of those is a July 2010 poll sponsored by 60 Minutes and Vanity Fair magazine. In this survey, 25% said that they thought Obama was born in another country and 14% said they didn't know. The challenge in interpreting these results is to separate “pure ignorance” from “politicized ignorance.” 

By “pure ignorance,” I mean people who don't know that the constitution limits the presidency to “natural-born citizens.” Such people might be neutral or even favorable about Obama being born outside the United States—they might even think it shows that America is a land of opportunity. The survey doesn't ask who you voted for in the last election, but we can look at groups that are heavily favorable to Obama: blacks, liberals, and people who approved of the administration's handling of the BP oil spill (only about 35% did at the time):

                            Not born in US    Don't Know

Liberals                       11%               6%

Blacks                         12%              11%

Approve handling oil spill     11%               8%

The combined total of people who think Obama was born outside the United States and those who don't know is about 20% in all of these groups. So it seems that about half of the original total represents what I call “pure ignorance.” On the other hand, that suggests that about 20% of Americans are “birthers” to some degree.