Left behind?, part 2

 After I finished my last post, it occurred to me that if people in rural areas felt "left behind," that should lead to a decline of confidence in institutions.  The General Social Survey has a series of questions about confidence in "the people running" various institutions.  For most of them, there's been a long-term decline (see this post):  the question I'll consider here is whether that decline has been larger in rural areas.  By and large, it has:  taking the twelve that have been asked about since the 1970s, the median correlation of confidence with time is -.115 for people living in MSAs, -.15 for people living in "other urban" counties, and -.18 for people living in rural areas.  Two of them stood out as especially large gaps:  medicine and education.  The figures show means by place of residence over time:

There's a lot of sampling variation in the individual years, so I include smoothed estimates.  Their exact shape should not be taken too seriously, but it's clear that in the 1970s and 1980s confidence in education was generally higher in rural areas and "other urban" than in MSAs, and in recent years the difference has disappeared.  

In the 1970s, confidence in medicine was higher in towns and rural areas than in MSAs:  now it is lower.  

Why would those two institutions show a particularly large decline in confidence in rural areas?  One possibility is that it's part of a general decline in trust in science and experts, maybe because they are all seen as part of a distant elite.  However, there's also a question about confidence in "the scientific community," and the trends for that are about the same in all types of places.  That suggests that the explanation is something specific to medicine and education.

There has been a trend towards geographic concentration of medical services, so people in rural areas are less likely to have medical providers in the community or nearby.  This could help to explain the relative decline in confidence.  For education, there was also a process of consolidation as small school districts were merged into larger districts.  That is, people in rural areas didn't have as much connection to and control of the schools as they previously had.  Although most of the consolidation took place in the 1950s and 1960s, the regional schools started with the teachers and administrators who had been in the local schools, so it may have taken a while for perceptions to change.  

The GSS also had a question on how much satisfaction you get from "the city or place you live in."  Unfortunately, they dropped the question in the mid-1990s, but there's a striking trend in the years before then.  In the 1970s, people in rural areas were substantially more satisfied than people in MSAs; that advantage was gone by the 1990s.  Hopefully, there are questions that are similar enough to take the story up to the present:  if I find any, I'll talk about them in a future post.

Left Behind?

 A few weeks ago, the New York Times had an article on rising support for the National Rally in rural France, which it said was due to a feeling of being neglected by the national government:  "Residents in this sparsely populated region ... - describe what is happening to their community as 'desertification,' by which they mean an emptying out of services, and of their lives."  Rural areas in the United States are also moving to the right, and people often give a similar explanation--they are "left behind" by changes in the economy, think the government isn't interested in their problems, and in some accounts resent the places and groups that seem to be moving ahead of them.  Economic growth has clearly been lagging in many small towns and rural areas, but I haven't seen much information about changes in political views, so I'll look at that issue in this post.  

The General Social Survey classifies places into six groups:  central city of the twelve largest metropolitan statistical areas, suburbs of the 12 largest MSAs, central city of the 13-100 MSAs, suburbs of the 13-100 MSAs, counties with towns of 10,000 or more, and counties without towns of 10,000 or more.  I combined the first four, so that there are three groups.  Then I calculated the correlation between year and opinions for each question.*  I'll start with two general measures of political orientation:  self-rating as liberal, moderate, or conservative; and whether the government is trying to do too many things that should be left to individuals and private business.  Each is coded so that a positive correlation means that opinions have tended to move in a liberal direction since the question was first asked (in the 1970s).

                          MSA     Town      Rural

POLVIEWS       .00        -.05         -.07
HELPNOT         .01        -.01         -.07

So urban areas have stayed about the same, while rural areas have become more conservative on both (generally, differences of about .03 or more are statistically significant).  

Now a few "social issues," again with positive signs meaning liberal trends:  legal abortion, whether sex between two adults of the same sex is wrong, whether there should be prayer in public schools, and whether a police permit should be required in order to buy a gun:

                             MSA     Town      Rural

ABSINGLE           -.05       -.05        -.03
HOMOSEX            .37        .32          .32
PRAYER                .07        .02          .01
GUNLAW             -.03        -.03        -.06

Rural areas have become more conservative relative to MSAs on three of the four.

Now some opinions related to race and ethnicity:  whether racial inequality is mostly due to discrimination against blacks, whether it's because blacks don't have the motivation and will power to get ahead, and whether the number of immigrants should be increased or reduced.  Again, a positive sign means a liberal trend:

                              MSA           Town         Rural
RACDIF1                 .06              .04            -.06
RACDIF4                 .22              .15              .17
LETIN1A                  .18              .15             .21

Rural areas have become relatively more conservative on the two questions about race, but there's no clear difference on immigration (that question has been asked only since the 1990s, so the standard errors are larger).

Now some questions about government spending on various issues.   Positive numbers mean a trend towards support for more spending.  There are a lot of questions, so I won't list them all, just talk about the general pattern and some notable cases.

                             MSA    Town    Rural

Nataid                .15  .10         .07
Natpark                .01  .05         .03
Natarms                .01  .05         .12
Natroad                .09  .08         .07
Nateduc                .13  .15         .16
Natrace                .14  .10         .09
Natheal                .00  .01         .06
Natenvir                .02  .01         .04
Natcity               -.02  .02         .05
Natmass                .05  .04         .10
Natfare                .11  .08         .06
Natsoc                .03  .05         .09
Natfarey                .06  .01         .05
Natspac                .19  .18         .22

Almost all of the trends are positive (towards favoring more spending)--the 1970s were a time of backlash against government spending.  The average across the questions is about the same for all groups, but there are some differences for individual items.  The upward trend is stronger in rural areas for spending on defense and the military (NATARMS), and weaker for foreign aid (NATAID), spending to help blacks (NATRACE), and welfare (NATFARE).  So far, it's a relative conservative shift in rural areas.  But the upward trend is also stronger for social security (NATSOC), health care (NATHEAL), mass transity (NATMASS), and even "solving the problems of big cities" (NATCITY).  

Finally, there's a question on the amount of federal income taxes you pay.  A positive number means that it's too low or about right rather than too high.

                              MSA       Town      Rural

TAX                       .06            .04          -.01

People in MSAs have been getting more satisfied (or less dissatisfied) and people in rural areas have stayed about the same.  That is, although people in rural areas have become more favorable towards spending on a lot of things, they haven't become more willing to accept taxes.

Overall, the movement in rural areas hasn't been a straightforward conservative one.   The spending questions suggest that people in rural areas have moved towards wanting the government to do more for Americans in general, but not for blacks specifically, and not for foreign nations.  There doesn't seem to be growing resentment against cities--if anything, it's diminished (see NATCITY and NATMASS).  And although rural areas have become relatively more conservative on race (NATBLACK, RACDIF1, RACDIF4), they haven't generally turned against "outsiders" (see LETIN1A).  

*I limited the analysis to whites.

The elites vs. the public?

A negative view of "elites" has become a central part of American conservatism.  It was always present to some degree, but it used to be focused on specific groups, especially journalists and intellectuals.  Now it's become a more general vision of a "deep state," "blob," or "uniparty."  In some versions, the elite has left-wing views and despises ordinary Americans--I wrote about one example a few weeks ago.  But there's also another version, which doesn't see the difference purely in terms of left and right, and where the elite is just not very aware of the public.   According to Oren Cass, the underlying problem in America is "elites who remain fully committed to their own preferences, to pulling the levers of power for their own benefit and to offering candidates in both parties who would preserve the status quo."  By "candidates in both parties" he doesn't mean Biden and Trump, but Biden and Nikki Haley--Trump represents the people, although he does that "imperfectly."  Although Cass doesn't offer any data on elite opinion, he is right in suggesting that Trump appeals to some views that are popular in the public but not among elites.  

The figure is from a Chicago Council on Foreign Relations (Chicago Council on Global Affairs) survey that asked people and members of "foreign policy elites" to rank the importance of various possible foreign policy goals (1-3, higher means more important).  The ones below the diagonal line were ranked higher by the public than by elites; the ones above the line were ranked higher by elites.   The biggest gaps are for "helping to improve the standard of living of less developed nations" (ranked higher by elites), "controlling and reducing illegal immigration" (ranked higher by the public), and "protecting the jobs of American workers" (ranked higher by the public).  

Here is the corresponding figure for 2016.  The question on improving the standard of living wasn't included in this survey, but illegal immigration and jobs were, and again they were ranked more important by the public.  The 2016 survey added a question on "attaining US energy independence," which was ranked higher by the public (the 2004 survey had a question on "securing adequate supplies of energy, which was also ranked higher by the public).  The biggest gap in the other direction was "limiting climate change"--that question wasn't in the 2004 survey, but "improving the global environment" was ranked higher by elites in both surveys.  Compared to the general public, elites also placed a higher priority on "defending our allies' security."

So there are large and persisting gaps between elite and public opinion, and the goals which the public ranks more highly are ones that Trump has emphasized.  Cass says that  "leaders might seek to shape public opinion and alter preferences — indeed, that is part of leading — but they must yield to the outcome. Their obligation is to pursue the community’s priorities, not their own." The problem with this principle is that different potential goals are not independent of each other--people think protecting American jobs is important, but it's safe to say that they also think that having low prices is important.  On the other hand, sometimes goals are complementary--for example, a higher standard of living in less developed countries might reduce illegal immigration.   So if elites don't follow public opinion on specific policies, it's not necessarily because they are thinking about their own benefit:  it could be because they think they have a better understanding of how to achieve "the community's priorities," as they understand them.  Of course, those beliefs could be wrong, and Cass suggests they are:  his essay is called "This is what elite failure looks like."  Many other observers have offered similar accounts:  the idea is that decades of policy failures led to popular discontent and a revolt against the elites.  But have elites actually done so badly?  I'll look at that question in a future post.  

Vice-precedented

Since becoming vice-president, Kamala Harris hasn't been very popular with the public--but how much of that is the result of her own qualities, and how much is because she has served with a president who hasn't been very popular?  I looked for data on approval ratings of previous vice-presidents.  For this analysis, I just used a single observation for each one--if Joe Biden decides to step aside, I may include more.  I tried to use approval ratings at about this point in the term (fourth year of the first term), but questions on vice-presidential approval haven't been asked all that often, so I couldn't follow that closely.   The figure shows net approval ratings (approve minus disapprove) for the vice-president and president.*

There is an association--if the president has a good approval rating, the vice-president is likely to have one too.  A regression of vice-presidential approval on presidential approval gives:

VP=9.8+0.64P; t-ratios are 2.6 and 3.2.  Three vice-presidents are substantially below the predicted values--Harris, Dick Cheney, and Spiro Agnew.  That company is not good news for Harris.  The biggest surprise is Dan Quayle, who was above the predicted value, with 50% approve and only 33% disapprove, when the survey was taken (January 1992).  

Presidents have opportunities to stand for the whole nation--doing things like making 4th of July speeches--and vice-presidents are often given less appealing tasks.  So I expected a negative intercept--if people were evenly divided on the president, they'd be predominantly negative on the vice-president.  Of course, you can't be too confident about any conclusions from such a small sample, but the evidence points in the other direction.  

 

*There was no approval/disapproval question for Mondale, so I used a favorable/unfavorable one.  I thought it was important to include something for that case because of the parallels with the current situation (unpopular president running for re-election).  

[Data from the Roper Center for Public Opinion Research]