Wednesday, May 16, 2018

Fully, accurately, and fairly

In 1972, a survey conducted by the Gallup Poll asked "how much trust and confidence do you have in the mass media--such as newspapers, TV and radio--when it comes to reporting the news fully, accurately, and fairly--a great deal, a fair amount, not very much, or none at all?"  The question was asked again in 1974 and 1976.  Then there was a long gap until it was asked again in 1997, but since that time it has been asked pretty regularly.  The means, with a great deal=4 .... none at all=1:

There were a few surveys which asked about the "news media" rather than the "mass media":  they are shown in red.

In December 2016, I wrote about other Gallup questions about confidence in various institutions, include newspapers and TV news.  Both showed a downward trend, so it's not surprising that confidence in the "mass media" does too.  However, the  question that I just discovered helps to shed light on the nature of the trend.  The following graph shows them all together (the mean is adjusted so it'a on a comparable scale):

With newspapers, there is an unusually high figure in one year (1979).  If you exclude that, there is very little trend from the 1970s until the early 2000s.  Similarly, TV has one unusually high year, which happens to be the first year it was asked, and then no trend until to the early 2000s.  When you add the question on the media, it's pretty clear that there was a decline from the 1970s to the 1990s, but that confidence then held up for several years before starting to decline again.

Another interesting point is that confidence in the media rose from 2016 to 2017 (September in both years).  This also happened with confidence in newspapers and TV news.  An obvious possibility is that the gain was a result of reporting on Donald Trump.  That might have pleased some liberals who in 2016 thought that the media was too hard on Hillary Clinton and/or Bernie Sanders.  I think there may also be a general tendency for it to be lower in election years:  it rose between 2012 and 2013, 2008 and 2009, and 2004 and 2005 (it was the same in 2000 and 2001).  That could be because people get tired of "horse race" coverage.

[Data from the Roper Center for Public Opinion Research]

Saturday, May 12, 2018

What's the alternative?

I saw one more article on culture vs. economics in the 2016 election and thought I should say more about my own position.  One popular view, which is advocated in the article by Diana Mutz that I have mentioned in previous posts, it that support for Donald Trump was not driven by economic distress.  Mutz pointed out that economic conditions were considerably better in 2016 than they had been in 2012 or 2008.   Another popular view, which is advocated in the latest article (by Dave Leonhardt) is that the lack of economic progress for less educated people over the last 40 years produced a gradual buildup of anger and frustration--Trump appealed to that feeling, and turned it against immigrants and racial minorities.  So they agree about the immediate motivations of Trump voters--racial and ethnic fears, and they agree that those fears stemmed from distress about the way things were going in the country; where they differ is on their ultimate source.

I don't agree with either of these analyses.  First, there is no evidence that people were particularly discontented with the way things were going in the country (see this post).  Second, opinions are not becoming more hostile to immigrants or racial minorities (see this post, among others).  Of course, race and ethnicity played an important role in this election, but they always do.  What was different about 2016?  I think it was the very low level of confidence in government.  That made people more interested in outsiders.  Also, there are a number of issues on which public opinion consistently diverges from policy.  An important example is immigration--people always think that more should be done to prevent illegal immigration.  Another important example is trade--people always are suspicious of trade agreements, and suspect that other countries are taking advantage of us.  However, when confidence in government is high, people are willing to give it some slack, and accept assurances that this particular trade agreement is good, or that the government is doing all that it reasonably can to stop illegal immigration.  When confidence is low, they'll credit claims that government officials are selling us out. 

I have calculated a measure of confidence in government which indicates that it fell to low levels in the early 1990s and then rebounded before falling to even lower levels in 2016.  A paper by J. Eric Oliver and Wendy Rahn calculates a measure of confidence which uses different data sources, but shows the same pattern.  While 2016 had Donald Trump, 1992 had another outsider candidate, Ross Perot.  Although he didn't have much lasting impact, Perot's electoral performance was arguably more impressive than Trump's.  He got almost 19% of the vote, which was the most by any third-party candidate since Theodore Roosevelt in 1912.  Unlike the other third party candidates who cleared 10% (Roosevelt, Robert LaFollette, and George Wallace) Perot had no political experience and lacked a regional base.  He was not particularly charismatic, and was much less well-known than Trump when he started his race.  The most plausible explanation for his strong performance is that voters were looking for an outsider.

Although Perot appealed to the same nationalist sentiments that Trump did, he drew about evenly from all educational levels.  On that point, I think that the difference is style.  Perot was kind of eccentric, but basically conducted himself as a "respectable" candidate; Trump didn't.   

I think this account makes sense of a lot of things, but there is one aspect still puzzles me.  It's easy to understand why people lacked confidence in government in 2016, but not why they did in the early 1990s. 

Tuesday, May 8, 2018

Other cultures

Last week I had a post on the idea that Donald Trump's gains among "working-class" (less educated) white voters were because of their anxiety about maintaining social dominance.  I mentioned that I wasn't convinced by the paper by Diana Mutz  that has been cited in support of this claim, but I didn't go into detail.  Yesterday I saw a piece by Andrew Cherlin in the New York Times, which said that "these conclusions, faithful as they may be to the survey data that underlie them, exemplify a misguided debate about whether culture or economics was the driving force in Mr. Trump’s win."  I agree that the debate is misguided--I've had a number of posts arguing that public opinion about economics includes a large dose of moral considerations.  However, I don't agree that the conclusions about social dominance are faithful to the survey data.

Mutz had a panel survey--the same people were asked the same questions in 2012 and 2016.  She found that "switches" (Obama in 2012 and Trump in 2016 or Romney in 2012 and Clinton in 2016) could be explained by position on three issues:  free trade, deportation vs. path to citizenship, and view of China as a threat or economic opportunity.  For each one, people were asked what they thought the position of the Democratic and Republican candidate was, as well as about their own position. 

Between 2012 and 2016, voters moved away from support for free trade and towards support for a path to citizenship.  The first shift helped Trump, while the second helped Clinton.  The overall effects of those two shifts almost exactly offset each other.  On China, there was no change in average public opinion, but the perceived position of the Republican candidate moved in the direction of average public opinion.  That is, Trump took the popular position on China, which helped him. 

That's the data--now on to the interpretation.  Mutz says that anxiety about social dominance should make people turn against "outsiders"--that is, against trade, against illegal immigrants, and against China.  People did turn against trade agreements, but became more sympathetic to illegal immigrants and didn't change on China.  So in terms of the hypothesis, one change was in the expected direction, one was in the "wrong" direction, and one didn't change.  In other words, what actually happened didn't match what should have happened if people were defending social dominance.   

What's my interpretation?  Social scientists are always attracted to the idea of having an interpretation that ties different things together, but I don't think that's possible here.  For immigration, the move continues a long-term shift towards more "liberal" views (the opposite of what the social dominance hypothesis predicts).  For trade, I think it was a short-term change resulting from the combination of criticism from Trump and Bernie Sanders, and the lack of a strong defense from Clinton.  And on China, there's an enduring gap between public opinion, which is tends to be sympathetic towards "America First" positions, and elite opinion, which tends to be more internationalist.  Trump seized an opportunity that previous candidates (except Ross Perot) had ignored. 

Wednesday, May 2, 2018

The Moral Economy of the American Crowd

I've had a number of posts observing that people are not very enthusiastic about free trade.  Why not?  One view is that it reflects fear--this New York Times article, summarizing research by Diana Mutz, mentions isolationism, nationalism (a belief that "the United States is culturally superior to other nations"), and ethnocentrism.  A recent article, also by Diana Mutz, has gotten a good deal of attention in the media. It offers a refinement of the earlier work:  opposition to free trade is about white men's fears of threats to their dominance in American society.  Basically, my view is that she provides some valuable information, but it doesn't support her interpretation.  However, rather than expanding on my criticism, I decided to suggest an alternative.  The alternative is that people see international trade as a form of competition, in which being in first place is important.

A 1990 NBC News/Wall Street Journal poll had the following question:

"Here are two situations that might occur.  Which would you prefer?

Situation A, in which the U. S. economy grows at a high rate, but the Japanese economy grows even faster, and over a period of several years, Japan becomes the world's leading economic power.

Situation B, in which the U. S. economy grows at a slower rate than in 'Situation A,' but faster than the Japanese economy, and the U. S. continues to be the world's leading economic power."

9% favored A, 86% favored B, and 5% weren't sure.  That is, 86% favored a lower standard of living for themselves just so that the United States could stay ahead of Japan.  Moreover, preference for Situation B was overwhelming among all kinds of people (84% among non-whites, 82% among people with graduate degrees, 82% among people who reported voting for Michael Dukakis in 1988).  That is, it wasn't just whites, or less-educated whites, that felt that way, it was people in general.

The point of my title is that suspicion of free trade isn't just a reflection of prejudice, but is part of the way that ordinary people think about economics.    In fact, it's pretty much the way that even sophisticated people thought about economics until well into the 20th century.

[Data from the Roper Center for Public Opinion Research]