Ross Douthat had a post on "The Missing Right-of-center Media" in which he noted that "outlets that mix political and cultural coverage" had liberal readerships; the other side, which he didn't say explicitly, was that outlets with conservative readerships were pretty much exclusively focused on politics. His proposed explanation is that "well-educated and well-informed conservatives are often businessmen (note the Wall Street Journal’s near-center position) whose reading interests are more practical and professional, who don’t cultivate a member-of-the-intelligentsia self-image, and who treat their media consumption mostly as a source of information rather than identity."
But political news and opinion is of no more practical value than cultural coverage (arguably less), so his explanation implies that there wouldn't be a significant audience for conservative political media, which there obviously is. It seems to me that there are two possible explanations.
One is historical: the "old media" model was to cover a range of topics, and most of the surviving "old media" have become politically liberal (the Wall St. Journal is an exception). So they have a mostly liberal but ideologically mixed readership (in the case of the WSJ, mostly conservative but mixed)--some who read them mostly for the political coverage, but others who don't agree with their politics but read them for other features. The "new media" model is to focus on one thing, and a number of conservative outlets have appeared to supply the demand. They have an overwhelmingly conservative audience.
The other is that it reflects demand: liberals like cultural coverage more than conservatives do. Douthat says something like this, although in pejorative terms--the "liberal clerisy" likes to "cultivate a member-of-the-intelligentsia self-image." Of course, if there is a difference of this kind, you could explain in a way that's favorable to liberals, or take the sensible course and say that you don't know why it exists.
There are a few surveys that have asked about interest in "cultural" things. One of them is a Pew survey from 2005, which asks people if they have visited the following in the last 12 months:
an art museum; a science or technology museum; a zoo or aquarium; a planetarium; a natural history museum; a public library. Unfortunately, the only questions on political views are ideological self-rating (liberal... conservative) and party identification. However, liberals are more likely to say that they have visited each one of those institutions. The differences for science and technology museums are not statistically significant, but all of the others are, most by a comfortable margin. The differences remain after controlling for education. There is little or no difference by party identification.
As I've said in several posts, a lot of people don't seem to understand the terms "liberal" and "conservative," or to understand them in non-political ways. But the relationships are so strong that it seems likely that liberals really are more interested in "culture."
[Data from the Roper Center for Public Opinion Research]
Sunday, October 26, 2014
Monday, October 20, 2014
Earthquakes and inflation
In November 2010, an open letter to Ben Bernanke warned that quantitative easing would "risk currency debasement and inflation." Inflation was running at an annual rate about 2% then, and is still about 2%. A recent Bloomberg news story found that none of the signers of the letter said they had changed their mind: 9 stood by the letter, 13 didn't respond, and one had died. The basic defense (among those who had a more or less coherent response) was that the letter just said that it was a risk, not a certainty. Cliff Asness, who didn't respond to the Bloomberg reporters but later posted a reply, puts it this way: "if you believe the risk of an earthquake is 10 times normal, but 10 times normal is still not a high probability, it's rational to warn of this risk, even if the chance such devastation occurs is still low and you'll look foolish to some when it, in all likelihood, doesn't happen."
I was struck by the earthquake analogy, which I've seen before in discussions of inflation. Earthquakes are sudden--you go from nothing to a disaster in a matter of minutes. Another image that frequently comes up is "playing with fire," which is the same idea--it can suddenly go from apparently safely under control to completely out of control. Is inflation really like that?
I took annual data on inflation in 40 countries since 1950 (from the OECD) and divided it into six categories: deflation, low (0-3%), medium (3-6%), fairly high (6-10%), high (10-20%), and very high (20+%). My question was how often countries had gone from low to high inflation in the space of a year. There were 574 nation-years with low inflation. Of those, 5 were followed by a year of fairly high inflation, 3 by a year of high inflation, and none by a year of very high inflation.
The cases of a "jump":
Year 0 Year 1
2.9% 13.3% India 1963-4
3.0% 10.6% Norway 1969-70
2.7% 10.5% Canada 1950-51
2.7 % 7.0% Sweden 1969-70
2.7% 6.5% Finland 1970-1
3.0% 6.3% Czech Rep. 2007-8
2.5% 6.3% Indian 1978-9
2.8% 6.1% New Zealand 1966-7
Six of those cases occurred in small nations and two in India, which I'm guessing did not have well developed institutions for economic management. None occurred in the major economic powers. France almost made the list, going from 3.1% in 1957 to 15.3% in 1958 (that was the year of the collapse of the 4th Republic).
So if there is a theory implying that quantitative easing created a substantial but not overwhelming probability (say 20% for each year the policy was followed) of a jump in inflation, it's not refuted by the experience of the last four years. But a theory like that wouldn't fit the behavior of inflation in the past, in which jumps from low to high inflation are rare.
I was struck by the earthquake analogy, which I've seen before in discussions of inflation. Earthquakes are sudden--you go from nothing to a disaster in a matter of minutes. Another image that frequently comes up is "playing with fire," which is the same idea--it can suddenly go from apparently safely under control to completely out of control. Is inflation really like that?
I took annual data on inflation in 40 countries since 1950 (from the OECD) and divided it into six categories: deflation, low (0-3%), medium (3-6%), fairly high (6-10%), high (10-20%), and very high (20+%). My question was how often countries had gone from low to high inflation in the space of a year. There were 574 nation-years with low inflation. Of those, 5 were followed by a year of fairly high inflation, 3 by a year of high inflation, and none by a year of very high inflation.
The cases of a "jump":
Year 0 Year 1
2.9% 13.3% India 1963-4
3.0% 10.6% Norway 1969-70
2.7% 10.5% Canada 1950-51
2.7 % 7.0% Sweden 1969-70
2.7% 6.5% Finland 1970-1
3.0% 6.3% Czech Rep. 2007-8
2.5% 6.3% Indian 1978-9
2.8% 6.1% New Zealand 1966-7
Six of those cases occurred in small nations and two in India, which I'm guessing did not have well developed institutions for economic management. None occurred in the major economic powers. France almost made the list, going from 3.1% in 1957 to 15.3% in 1958 (that was the year of the collapse of the 4th Republic).
So if there is a theory implying that quantitative easing created a substantial but not overwhelming probability (say 20% for each year the policy was followed) of a jump in inflation, it's not refuted by the experience of the last four years. But a theory like that wouldn't fit the behavior of inflation in the past, in which jumps from low to high inflation are rare.
Tuesday, October 14, 2014
College and class
There have been a lot of articles lately about how students from working-class backgrounds are under-represented in American colleges and universities. Most of them imply that this is to some extent new--that colleges used to be more inclusive. The Roper organization did a survey of college students in 1949 which sheds some light on this issue. It asked students about their father's occupation, using a fairly detailed classification. Other Roper surveys of the general public taken at about the same time used the same classification, so you can compare the fathers to the general public:
Students' Fathers
Public Veterans Non-vets
Professional 5.2% 12.4% 19.3%
Salaried-executive 10.7% 13.7% 18.5%
Proprietor-other 11.9% 15.3% 18.4%
Salaried-minor 10.4% 12.7% 15.3%
Proprietor-farm 5.1% 3.9% 5.4%
Wages-other 18.2% 11.7% 8.2%
Wages-factory 20.9% 8.3% 4.5%
Wages-farm 6.1% 0.2% 0.4%
Students from middle-class (especially professional) backgrounds were substantially over-represented, but there was a difference between students who were veterans and those who weren't. For example, among non-veterans, those whose fathers were professionals outnumbered those whose fathers were factory workers by more than 4:1; among students who were veterans, the ratio was less than 1.5:1. Apparently the GI Bill of Rights had a big impact.
[Data from the Roper Center for Public Opinion Research]
Students' Fathers
Public Veterans Non-vets
Professional 5.2% 12.4% 19.3%
Salaried-executive 10.7% 13.7% 18.5%
Proprietor-other 11.9% 15.3% 18.4%
Salaried-minor 10.4% 12.7% 15.3%
Proprietor-farm 5.1% 3.9% 5.4%
Wages-other 18.2% 11.7% 8.2%
Wages-factory 20.9% 8.3% 4.5%
Wages-farm 6.1% 0.2% 0.4%
Students from middle-class (especially professional) backgrounds were substantially over-represented, but there was a difference between students who were veterans and those who weren't. For example, among non-veterans, those whose fathers were professionals outnumbered those whose fathers were factory workers by more than 4:1; among students who were veterans, the ratio was less than 1.5:1. Apparently the GI Bill of Rights had a big impact.
[Data from the Roper Center for Public Opinion Research]
Saturday, October 4, 2014
The 14 percent
In sportscasters' language, opponents of the Washington Redskins name have been moving the ball at will recently. This isn't because of overwhelming public support. In 1992, an ABC News/Washington Post poll asked: "Some people say that the Washington Redskins should change its team name because it is offensive to native American Indians. Others say the name is not intended to be offensive, and should not be changed. What about you: Should the Redskins change their team name, or not?" The same question was also asked in 2013 and 2014 in an Associated Press/Gfk poll. The results:
Should Should Not
1992 7% 89%
2013 11% 79%
2014 14% 83%
Although support for a name change is growing, it's still a pretty small minority. What sort of people support a name change? Unfortunately, the individual-level data for the 2013 and 2014 surveys are not available, but a 2010 Vanity Fair/CBS News poll asked a similar question: "The University of North Dakota has just retired its mascot, the Fighting Sioux, on the grounds that it gives offense to Native Americans. What should happen to other Native American references that are common in sports, such as the Washington Redskins and the Atlanta Braves?...Do you think they should be retired because they are offensive to Native Americans, or should they be kept because this kind of political correctness has gone too far?" 12% said that the names should be retired and 78% said that they should not.
Party identification and self-rated political ideology made a substantial difference: about 20% of Democrats and liberals, and only 6 or 7 percent of Republicans and conservatives supported a name change. Age made a difference: people 18-29 were more likely to support a change. So did education: people with graduate degrees were more likely to support a change. So did ethnicity: blacks, Hispanics, and Asian-Americans were all more likely to support a change. Region didn't make much difference--southerners were a bit less likely to support a change, but there were no clear differences between the Northeast, Midwest, and West. Income didn't make much difference, either. However, people living in cities were more likely to support a name change.
Putting it together, Washington DC has the demographics that would maximize support for a change. I suspect that another factor is that the team has been generally disappointing for a long time.
[data from the Roper Center for Public Opinion Research]
Should Should Not
1992 7% 89%
2013 11% 79%
2014 14% 83%
Although support for a name change is growing, it's still a pretty small minority. What sort of people support a name change? Unfortunately, the individual-level data for the 2013 and 2014 surveys are not available, but a 2010 Vanity Fair/CBS News poll asked a similar question: "The University of North Dakota has just retired its mascot, the Fighting Sioux, on the grounds that it gives offense to Native Americans. What should happen to other Native American references that are common in sports, such as the Washington Redskins and the Atlanta Braves?...Do you think they should be retired because they are offensive to Native Americans, or should they be kept because this kind of political correctness has gone too far?" 12% said that the names should be retired and 78% said that they should not.
Party identification and self-rated political ideology made a substantial difference: about 20% of Democrats and liberals, and only 6 or 7 percent of Republicans and conservatives supported a name change. Age made a difference: people 18-29 were more likely to support a change. So did education: people with graduate degrees were more likely to support a change. So did ethnicity: blacks, Hispanics, and Asian-Americans were all more likely to support a change. Region didn't make much difference--southerners were a bit less likely to support a change, but there were no clear differences between the Northeast, Midwest, and West. Income didn't make much difference, either. However, people living in cities were more likely to support a name change.
Putting it together, Washington DC has the demographics that would maximize support for a change. I suspect that another factor is that the team has been generally disappointing for a long time.
[data from the Roper Center for Public Opinion Research]
Wednesday, October 1, 2014
beautiful hypothesis vs. ugly fact
I followed up my last post by looking at the correlation between perceived corruption in 2006 and opinions on whether politics played a part in the administration of relief in 1937. My idea was that states would have an enduring political culture, so that the ones that were corrupt (or perceived that way) in 1937 would still be in 2006. The correlation between the scores at the different times was .05, with a p-value of .75. The means are from samples, pretty small samples in many cases, but when you start from that near zero, correcting for sampling error won't make much difference.
Then it occurred to me that the alleged role of politics in the administration of relief was a Republican issue in the late 1930s, so you should control for the political inclinations of the states. I regressed 1937 perceptions on vote for Roosevelt in 1936 and perceived corruption in 2006. This produced an R-square of 0.3 and an adjusted R-square of zero. It's not looking good for my hypothesis.
Then it occurred to me that the alleged role of politics in the administration of relief was a Republican issue in the late 1930s, so you should control for the political inclinations of the states. I regressed 1937 perceptions on vote for Roosevelt in 1936 and perceived corruption in 2006. This produced an R-square of 0.3 and an adjusted R-square of zero. It's not looking good for my hypothesis.