Invent your own research

 On March 30, Nicholas Kristof had wrote   "Survey data indicates that married couples on average report more happiness, build more wealth, live longer and raise more successful children than single parents or cohabiting couples, though there are plenty of exceptions."  The most popular reader comment, with over 2000 likes, said
"I notice that you didn't talk about the research that shows differences in gender in marriage happiness levels.
    Marriage is generally GREAT for men, who report being far happier in marriage than being single.  Much research indicates the reverse is true for women.  Single women report being happier, in general, than married women are." 
Most of the other leading comments were along the same lines.  

Since the 1970s, the General Social Survey has asked whether "taken all together, how would you say things are these days."  41% of married women say that they are very happy, and 8% say they are "not too happy" (the rest chose "pretty happy"); among single (never married) women, 23% say they are very happy and 17% not too happy.  This is just one survey, but it's one that puts a lot of effort into selecting a representative sample and gets a high response rate, so it can be considered to be pretty definitive.  Of course, the difference in happiness between married and never-married women is not necessarily caused by marriage, but it's there.  

But maybe the difference is smaller for women--married women are somewhat happier than single women, but married men are a lot happier?  The figure shows the averages for married and never-married men and women over the years (higher numbers mean happier).  There's a lot of sampling variation, so I also show the smoothed averages.  The pattern seems similar for men and women.

The next figure shows the difference between married and never-married people among men and women:  

Up until about 2000, the difference was a little bigger for men, but since then it's been about equal.  

I looked to see if the gap varied by other characteristics--for example, education, race, political views.  To make a short story even shorter, I found nothing worth mentioning.  

In the course of doing this analysis, I noticed that the GSS had a question on whether you thought that married people were generally happier than unmarried people.  45% agreed (or strongly agreed), 24% disagreed (or strongly disagreed), and 32% chose "neither agree nor disagree."  There were some group differences in average views on this question:  men were more likely to agree than women; whites more likely to agree than blacks or people of other races; conservatives more likely to agree than liberals; married and widowed people more likely to agree than never-married, with divorced people least likely to agree.  Despite what is sometimes said about "elites," there was no discernible difference by education, and people in higher status occupations were more likely to agree.  The question was asked only four times, most recently in 2012, but it seems like the gender difference was growing:  the means for men were 3.6 in 1988, 3.43 in 1994, 3.47 in 2002, and 3.32 in 2012; the means for women were 3.50, 3.26, 3.17, and 2.97 (higher means more agreement)--that is, the gender difference went from .10 to .35.  There were also signs that the gender difference varied among groups--for example, agreement was particularly low among black women (20% agreed and 48% disagreed).

The problem is you?, part 2

 I won't try to give a table of regression coefficients in this post, just summarize the differences between the analysis of the geographical origins of January 6 insurrectionists by Pape, Larson, and Ruby and my re-analysis of their data.  

1.   Control variables:  my main change was to use the logarithm of population rather than population as a predictor variable, for reasons discussed in my previous post.  I also created a variable for people living within driving distance, which I defined as 700 kilometers (which includes Boston, Cincinnati, and Detroit) and an interaction between distance and that variable.  My idea was that (a) if you were in driving distance you could make the trip without spending much money and (b) with driving, the cost in time and money is strongly related to the distance; if you have to fly the relation is weaker (a lot depends on distance to the nearest airport, whether you can get a nonstop flight, and the mysterious pricing decisions of airlines).

2.  Points in common:  the number of insurrectionists increased with the percent of the county that was non-Hispanic white; decline in manufacturing employment didn't make any clear difference; number of insurrectionists was higher in urban areas (although the estimated effect was much smaller in my analysis).

3.  Points of divergence:  a decline in the white population led to more insurrectionists in their analysis but had no effect in mine; the percent who voted for Trump led to fewer insurrectionists in their analysis but more in mine.  They also considered the difference between percent for Trump in 2020 and percent for Romney in 2012--that is, voters who were specifically attracted by Trump.  That also didn't have an effect in their analysis.  I ran a model including both Romney support in 2012 and the difference, and found that they both had similar positive estimates.  I think this is important--it suggests that the insurrectionists were drawn both from new Trump followers and traditional Republicans.  However, if you think of the population at risk of being insurrectionists as Trump voters in a county, it would be log(np)=log(n)+log(p), where n is the number of people and p is the percent for Trump (of course, there are people who weren't eligible to vote and people who were eligible but didn't vote, but suppose they are roughly constant across counties).  That suggests that both log(n) and log(p) should have coefficients of about 1.0.  Log(n) did, but log(p) was about 0.55.  The fact that it's not zero is interesting, but so is the fact that it's less than 1.  My thought is that a lot of people, especially those who aren't very interested in politics, just go along with the local climate.  That is, if you're a Trump voter in Manhattan, you're probably highly committed; if you're a Trump voter in Wyoming, you may just be following the crowd.  So the proportion of Trump voters increases, the fraction who are highly committed may decline.  

Overall, they conclude that participation in the insurrection was largely a response to perceived ethnic threat, and that the sources of "violent populism" are very different from those of "electoral populism."  My conclusion is that the sources are similar--after you control for population and distance, the places where Trump got votes were also the placed where he got supporters on January 6.

The problem is you?, part 1

 The Atlantic recently published a critical review of the new book by Tom Schaller and Paul Waldman, White Rural Rage: the Threat to American Democracy. The review, by Tyler Austin Harper, concluded by saying that they were not just wrong, but had it backwards--the threat is from the cities and suburbs:  

"Schaller and Waldman are right: There are real threats to American democracy, and we should be worried about political violence. But by erroneously pinning the blame on white rural Americans, they’ve distracted the public from the real danger. The threat we must contend with today is not white rural rage, but white urban and suburban rage.

Instead of reckoning with the ugly fact that a threat to our democracy is emerging from right-wing extremists in suburban and urban areas, the authors of White Rural Rage contorted studies and called unambiguously metro areas 'rural' so that they could tell an all-too-familiar story about scary hillbillies. Perhaps this was easier than confronting the truth: that the call is coming from inside the house. It is not primarily the rural poor, but often successful, white metropolitan men who imperil our republic."

The report that Harper links to says:  "the more rural a county, the lower its rate of sending insurrectionists, a finding which is significant with a p-value <.01%."  A  just-published paper by Robert A. Pape, Kyle D. Larson, Keven G. Ruby in PS: Political Science and Politics gives a more detailed analysis.  The results are from a negative binomial regression in which the  dependent variable is the number of people from a county who were charged with crimes related to the January 6 attack on the Capitol.  The number is estimated to be 2.88 times as large in urban than in rural counties, controlling for other factors.  

Of course, the population of the county is one of the other factors.  But a negative binomial regression predicts the logarithm of the dependent variable and their control is population (in 100,000s).   The estimated coefficient for population is .148, meaning that the natural log of the predicted number of insurrectionists goes up by .148 for every 100,000 increase in county population.  If the natural log of the predicted number goes up by .148, the predicted number goes up by about 15%.*  If you're starting from a population of 1,000, an increase of 100,000 means that population goes up by a factor of of about 100; if you're starting from a population of 1,000,000, it's 10%; if you're starting from a population of 5,000,000, it's only 2%.  So the model controlling for population builds in a relationship between county population and the chance that a person will be an insurrectionist:  declining and then increasing.  The figure shows the nature and size of the relationship using their estimate:

The number 1 on the y-axis represents the rate in a county of average size (about 100,000).  In a county with population of 10,000, the rate is about 8.5; in a county with 500,000, it's about .4, and in one of 5,000,000, it's about 80.  The biggest county in the United States (Los Angeles) has a population of about 10,000,000, but I don't extend the x-axis that far because it would make the figure too hard to read.   Of course, there is no reason to expect that there really is a relationship of this form.

A straightforward alternative would be to model the rate--number of insurrectionists (x) divided by county population (n).  But log(x/n)=log(x)-log(n), so you could express that by a regression with log(x) as the dependent variable and log(n) as one of the predictors.  Then a coefficient of 1.0 on log(n) would mean that the rate was the same across different county populations; a coefficient of less than one would mean it was higher in counties with smaller populations and a coefficient of greater than 1.0 would mean it was higher in counties with larger populations.  

What happens if you use log(population) rather than population as a control variable?

                                                                        Population                    Log

% white population decline                            .111***                    .035
                                                                        (.019)                        (.020)

manufacturing employment decline                .011                            -.006
                                                                        (.0054)                        (.006)

extra Trump %                                                    -.039***                .003
                                                                            (.0081)                    (.0082)

% non-Hispanic white                                           .009***                .014***
                                                                            (.0033)                    (.003)

Metro county                                                        1.095***             .326*
                                                                           (.1335)                    (.135)

Distance to DC                                                    -.304***            -.210***
                                                                            (.0623)                (.051)

(log) population                                                    .148***            .999***
                                                                             (.0210)                (.056)

  The fit of the model with the logarithm as control is better.  Several of the estimates for the other variables change substantially.  The estimate for metro counties is still statistically significant, but not overwhelmingly so (p=.019), and is much smaller than when using population.  So I don't think that the evidence justifies sweeping condemnation of urban and suburban men.

I have experimented with other specifications of the model, but this is enough for one post.  

*My figures are from my analyses using their replication data file, which are slightly different from the numbers implied by their tables.  
  

Did he do it?

 

After O. J. Simpson's death last week, I looked for surveys that asked for views about whether he was guilty of the murders of Nicole Brown Simpson and Ronald Goldman.  The figure is based on surveys which offered options of definitely guilty, probably guilty, probably not guilty, and definitely not guilty (counted as +2,+1, -1, and -2).*  All of the figures are positive, meaning that more people thought him guilty than not guilty.  Over the whole period, there was a move towards thinking that he was guilty--the latest surveys found that about 40% thought he was definitely guilty and another 40% thought he was probably guilty.  I thought that the civil case that found he was responsible for the murders might have helped to move opinion in that direction, but if anything it seems to have been followed by a short-term shift away from thinking he was guilty.  One survey taken at the time of the controversy over the release and withdrawal of his book, "If I did it," showed a move towards thinking him guilty.  But basically, it seems to have been a gradual movement over a long period.  

The next figure is the same data, but limited to the period up to the end of the criminal trial.  I don't recall most of the details of the trial, but the way the story is often told now, there were a number of dramatic moments in which the defense succeeded in raising doubts about his guilt.  You don't see much sign of this in the survey results--there seems to have been a small move towards thinking him guilty during the trial, followed by a drop after the verdict.  The absence of ups and downs during the trial surprised me--a lot of people followed it closely, and although many had strong opinions, it wasn't a partisan issue, so I would expect them to be more open to changing their minds.  

During and after the criminal trial, many people noted that there was a large division by race--black people were a lot less likely to think that Simpson was guilty.  This raises the question of whether that was there from the beginning or emerged during the trial, and whether it remained in the later surveys.  I'll look at this issue in a future post.  

*Some asked if the charges were definitely true, probably true, probably not true, or definitely not true.  Immediately after he was found not guilty, some prefaced the question with "whether or not you agree with the jury's decision."  There was no evidence that these variations made a difference.  

The happiness gap

 Conservatives typically report being happier than liberals do, but the size of the gap appears to change over time.    In 2015, I looked at data from the GSS  and found that the "happiness gap" became larger during the GW Bush presidency, but fell in the Obama presidency.  A few days ago, Ross Douthat had a column called "Can the Left be Happy," which said that "the left-right happiness gap is wider than before"--that is, the relative happiness of the left has declined in the last decade or so.   He made a plausible case but didn't offer any systematic data, so I'll take another look.

The GSS asks people to rate their political ideology on a seven point scale, from very liberal to very conservative, which I collapsed into three groups:  extremely liberal or liberal; slightly liberal, moderate, or slightly conservative; conservative or extremely conservative.  The liberal and conservative groups are both about 15-20% of the sample.  

The figure is hard to interpret, partly because of sampling error in individual years and partly because of the big drop among all groups in 2021 and 2022, so here's a figure showing the difference between the averages for liberals and conservatives (positive numbers mean conservatives report being happier than liberals do):

The higher reference line is the average difference.  The gap was larger than average in 2000, 2002, 2004, 2006, and 2008, and then fell in 2010 (liberals were actually happier than conservatives).  It's generally remained smaller than average since then.  

Here's the corresponding figure for the moderate/conservative gap:

A similar story:  the gap became a lot smaller in 2010, and has generally remained below average since then.  

So it's not liberals who have become relatively unhappier in recent years, but conservatives.  Going back to the original figure, there was little or no happiness gap in the 1970s.  Conservatives pulled ahead in the 1980s, and the size of the gap seemed to be gradually increasing until 2008.  Then conservatives became less happy in 2010, and the gap has been smaller since then.  

You could say that the shifts in the 21st century are just another example of increasing political polarization:  liberals are relatively happy when a Democrat is in office, and conservatives are relatively happy when a Republican is.  But I don't think that fits the pattern very well.  Although there were signs of growing polarization under Bush, they were pretty small--you didn't get a big increase until the Obama years.  And although conservatives became relatively happier in 2018, the change was not that big, and the gap remained smaller than it had been under Bush. 

But considering the whole period suggests that there may be a connection to the general political and social climate.  In the 1970s, the tide seemed to be running to the left, but with the tax revolt of the late 1970s and the election of Ronald Reagan in 1980, things seemed to stabilize and maybe even reverse.  Before the 2008 election, conservatives could feel pretty good--Republicans had won five of the last seven presidential elections, and no Democrat had received a majority of the popular vote since Jimmy Carter got 50.1% in 1976.  Republicans had also gained parity in Congress, and the courts had moved to the right.  Then Obama was elected with a lot of popular enthusiasm, a solid Congressional majority, and an economic crisis that provided a rationale for vigorous government action.  Prominent conservatives reacted to the threat by a strategy of scorched-earth opposition--e. g. they denounced Obamacare not just as ineffective and expensive, but as the end of the American way of life.  Since then, conservatives have felt like they are on the defensive, even when Trump was president--fighting against the deep state, the media, Big Tech, etc.  So my idea is that although ratings of happiness are primarily affected by individual factors, there is some spillover from feelings about the general direction of society.

When I was young

 In 1951, the Gallup Poll asked "Comparing your present family circumstances with those when you were a child, would you say you are better off, or worse off, than your parents were then?"  They asked it again in 1991, and other organizations asked it in 1994 (twice) and 2016.  The results:

                    better      worse    same (vol.)
Feb 1951       60%         24%      14%
Dec 1991      78%         12%        9%
Aug 1994      65%        22%      11%
Sep  1994      72%        21%        6%
Dec  2016      72%        20%        7%

The distributions are all about the same except for 1991, when people were more positive.  I can't think of a plausible reason that opinions would change that much between December 1991 and 1994 (economic conditions were similar, but somewhat better in 1994), so I think that difference is at least partly sampling error.  The important point is that opinions weren't more pessimistic in 2016 than in 1950 or 1994.  It's often suggested that Americans always felt like they were making economic progress from generation to generation until recently, and that the loss of that sense has led to a variety of social and political problems, like "deaths of despair," the decline in rates of marriage and childbearing, and the rise of Trump.  But I don't think that's the case--people are discontented with some aspects of society, but most still believe that there's an upward trend in their standard of living.

[Data from the Roper Center for Public Opinion Research] 

Going right?, part 2

 I want to follow up on my previous post, which proposed that the recent rightward move of Black, Hispanic, and less educated voters might have a common cause--concern about foreign affairs.  My idea is that these groups tend to have a more suspicious view of the world--to feel that you can't count on other people and have to look out for yourself.  When applied to foreign affairs, this means an affinity for an "America first" approach--don't get involved unless our interests are at stake, and when we do get involved, rely on military or economic threats rather than diplomacy and alliances.  That's exactly in line with Trump's view of foreign policy.  So Trump has an advantage with these groups in this sphere (I don't call it an issue, since it's more of a general orientation than a policy position).  Most voters aren't that concerned with foreign affairs most of the time, so if things were going well, it wouldn't have much influence.  But now, with the wars in Ukraine and Gaza and the rise in illegal immigration, people are more concerned, and Trump's approach makes sense to a large number of voters.

So that's my idea--what about evidence?  I alluded to some last time, but didn't give details.  On the "suspicious view of the world," there are three questions that have been regularly included in the General Social Survey, whether "most people can be trusted, or that you can't be too careful in life," "most of the time people try to be helpful, or that they are mostly just looking out for themselves," and "most people would try to take advantage of you if they had a chance, or would they try to be fair."  If you make an index of positive views of human nature by counting "can be trusted," "try to be helpful," and "try to be fair" answers, the means are:

Black            0.9
Hispanic       1.0
NH White     1.6

Non-grad       1.4
Coll Grad       2.0

The GSS doesn't have many questions on foreign affairs, but one was included regularly until 2006:  "Do you think it will be best for the future of this country if we take an active part in world affairs, or if we stay out of world affairs?"  More recently, a question asking if they agree or disagree that international organizations are taking too much power from the American government has been asked several times.  The table shows opinions by scores on the index:

       Active part           Taking power (yes minus no)
0     54%                          17
1     60%                          13
2     69%                          -3
3     82%                          -30

So general views of people are strongly related to these foreign policy opinions.  The racial/ethnic and educational differences in general views of people are longstanding and haven't changed much, which suggests that an "America first" policy has always had the potential to appeal to less educated people and racial minorities (and repel educated people).  However, from the 1940s until the rise of Trump, that kind of policy was almost universally rejected by elites of both parties, so it wasn't offered as a choice.  

Going right?

 There's been a lot of talk recently about Blacks and Latinos shifting to the Republican party (an example).   Most of the accounts I've seen say that it results from a weakening of party loyalty, so that Black and Latino conservatives start moving to their "natural" home.  However, you would expect that to be a gradual process, much of it the result of generational replacement--why would it be happening now more than it was 10, 20, or 30 years ago?  I looked at the Echelon Insights surveys, which ask people about their views of Joe Biden and Donald Trump, comparing the latest one available (Feb 2024) to October 2021, which I chose as the baseline so that we would be past Biden's "honeymoon" and the reaction to January 6 would have had time to fade.  The figure shows change in mean favorability (4=very favorable, 3=somewhat favorable, 2=somewhat unfavorable, 1=very unfavorable), for groups that were included in  reports (party, race/ethnicity, sex, income, education).

Biden's ratings have declined among most groups, and Trumps have risen, but three stand out:  blacks, Hispanics, and people without a college degree.  These groups have had the largest anti-Biden and pro-Trump shifts.  People without a college degree are different from Blacks and Hispanics because they already gave strong support to Trump--that is, the decline of party loyalty explanation can't account for their change.  Is there any common factor that might explain the recent shift of these three groups?  I think that it might be the increase in foreign conflicts and illegal immigration.  

As I've mentioned before, many people think that other countries take advantage of the United States--they we try to be fair but they don't.  The political importance of this feeling is likely to increase when times are tough--people will think that under the circumstances we can't afford to worry about other countries and their problems and need to focus on our own national interests.  This point is relevant to the group shifts because blacks, Hispanics, and less educated people tend to be more suspicious of others--less likely to think that other people can be trusted or will generally treat you fairly--and the differences are big by the standard of individual-level survey data.  That suggests that they will respond more strongly to the kind of appeals that Trump makes.  But it also suggests that they may shift strongly in the other direction if Biden seems to be getting things under control.  

The road to Mar-a-Lago

Donald Trump now has enough delegates to clinch the Republican nomination for the third time.  How did we get here?  One view is that it came out of the blue--Trump staged a "hostile takeover" of the Republican party by appealing to ordinary voters, and the Republican leadership has gone along because they can't stop it.  Another view, which I think is the correct one, is that it's the culmination of a long development--that by 2016, the party was ready for someone like Trump.  Of course, it's hard to identify a precise starting point for a long-term shift, but some popular choices are Nixon's "Southern strategy" and Reagan's election in 1980.  I want to suggest another possibility-- the 1990 budget agreement, in combination with George H. W. Bush's 1988 "no new taxes" pledge at the Republican convention and his loss in the 1992 election.  The obvious effect of this experience was to strengthen Republican opposition to raising income tax rates, but I think it had a larger effect.

The figure shows net public support (favorable minus unfavorable opinions) for the proposed budget agreement in October and November (the last Congressional vote was October 27, and Bush signed it on November 5).  On the whole, public opinion was negative, but not all that negative--averaging across the ten surveys that asked, 36% were in favor and 41% opposed.  The figures were more negative in the two final surveys, but they just asked about views of the budget agreement with no further detail, while all of the previous surveys said something about bipartisan agreement.  People like bipartisanship, so it's likely that the difference in responses was because of the difference in questions rather than because of a real change in attitudes towards the agreement.  Opinions were also getting more favorable over time until the last two questions, which suggests that people were happy that Congress seemed to be getting things done.*  That is, people didn't like the idea of a tax increase, but liked the idea of parties working together on a plan to reduce the budget deficit.   

 The result of this episode was that the Republicans became less willing to cooperate with Democrats in being "responsible"--doing things which most knowledgeable people say are necessary but which the public is inclined to oppose.  Bush and the Republican leadership in Congress set aside politics (Bush's "read my lips" pledge) to do the "responsible" thing, and it wound up hurting them.  That started a move towards a strategy of uniform opposition--make the Democrats do it and don't give them the cover of bipartisanship.  Of course, this isn't an absolute--there are some cases in which the "responsible" thing is in line with Republican ideology, but they have become less common as the party has increasingly turned against "elites" and experts.  

[Data from the Roper Center for Public Opinion Research]

*There were two questions that had more detailed descriptions of what was in the agreement, which are indicated in the figure. The descriptions were different, and the second one sounds more favorable to me than the first, which may explain some of the difference between them.  But the evidence for a time trend is still pretty strong if you set them aside.  

Two part twos

 1.  My last post observed that in 1970, people with higher incomes were less likely to say that people who are successful get ahead because of luck, but that in 2016 no relationship was visible.  How about education?  In 1970, there was no relationship between education and opinions after you controlled for income.  In 2016, there was:

                        Luck
Not HS grad     19%
HS only            16%
Some college    10%
College grad     14%
Grad educ         16%

That is, the middle levels of education were least likely to say that luck was what mattered.  This supports my general point about people at the top becoming less likely to assert their superiority (also see this post).

2.  In January, I tried to understand the persistence of Republican support for Donald Trump by comparing it to views of Watergate.  Views on whether Watergate was a "very serious matter" or "just politics" shifted towards "a very serious matter" during the earlier part of the Watergate investigation, but then stabilized.  They were pretty evenly divided in the summer of 1974 (not long before Nixon resigned) and in several later surveys that asked people to look back.  

In that post, I said "Of course, Nixon was ineligible to run for President,  but no one said he should remain a major voice in the party and no one sought his endorsement when running for office."  I later found some questions from 1979 that were prefaced "there has been some talk of President Nixon getting back into active political life" and then asked what they thought about several possible ways.  (I don't remember whether there actually was such talk or whether the people doing the survey just thought it was an interesting question).  The percent saying it would be a good idea for him to:

                                                              Republicans      Independents     Democrats

Run for office                                          18%                   14%                     7%
Be appointed to high post                       14%                   10%                    5%
Speak out on issues                                 55%                    33%                   27%
Take active role in party                         31%                     19%                   14%

About 8 or 9 percent said they weren't sure--that was pretty constant across all the party/question combinations.  So a substantial number of Republicans supported the idea of Nixon taking "an active role in the workings of the Republican party," and some even thought he should run for office.  

  After Trump's loss, Republican elites seemed to expect their voters to spontaneously turn against him, and then be surprised that he retained substantial popular support (or interpreted it as evidence that he had a unique personal appeal).  This example shows that they shouldn't have been surprised:  Nixon had significant support from Republican voters even after he had acknowledged wrongdoing, resigned, and kept a low profile for several years.  

Why was elite behavior different in the two cases?  The most apparent factor is that in the 1970s there was a stronger core of leadership that could speak for the majority of the party in Congress.  But I think that there's also another:  there's now a stronger sense of "team spirit" among both politicians and opinion leaders.  That is, Republicans were reluctant to make common cause with Democrats.  As a result, even those who aren't Trump supporters have promoted the idea that he's being unfairly treated, that Democrats have done similar things in the past or are trying to do them now--e. g. Ross Douthat's indignation about efforts to remove Trump from the ballot ("antidemocratic and incompetent at once, signifying ... a general elite fear of the voting public").  That makes it easier for ordinary Republicans to conclude that it's all "just politics."

[Data from the Roper Center for Public Opinion Research]

The secret of success

 In 1939, the Gallup poll asked "Do you think people who are successful get ahead largely because of their luck or largely because of their ability?"  They asked the question again in 1970, and CBS news asked it in March and September 2016.  

                Luck         Ability        DK/NA
1939         16%            80%           4%
1970           8%            86%           6%
3/2016       14%            81%          6%
9/2016       11%            80%           6%

At each time, overwhelming majorities said that ability was the major cause.  That's important in itself, but differences by social standing are also of interest.  The 1939 survey contained an interviewer rating of economic standing, and people who ranked higher were more likely to say that ability was what mattered.  The 1970 and 2016 surveys didn't have the interviewer rating, so that can't be used for comparison, but both the 1939 and 1970 surveys asked about occupation, so you can compare occupational differences:

The figure shows the log of the odds ratio of ability to luck answers by occupational group.  I classified the occupations as higher to lower:  business and professionals, then farmers and white collar workers, then skilled manual workers, with semi- and unskilled manual workers at the bottom.  At both times, people in "higher" occupations were more likely to say ability, but the relationship seems weaker in 1970--the estimated slope is about 1/2 to 2/3 as large, and there is more scatter around the line.  

The 2016 survey didn't ask about occupation, but it did ask about income, which can be used to compare it to 1970.  Of course, the income categories were different, but there happened to be 11 each time, so to keep things simple I'll just number them as 1 through 11.

There was a clear relationship in 1970--people with higher incomes are more likely to say that success depends on ability--but not in 2016.  

Over the whole period, the relationship between social standing and opinions about the cause of success has become weaker, and maybe even disappeared.  Rather than looking down on the working classes, as critiques of "meritocracy" claim, people in the upper and middle classes have become less likely to assert their superiority.  I've argued this before, but didn't have such direct evidence.  

[Data from the Roper Center for Public Opinion Research--with special thanks for obtaining the 2016 survey]

You never know

 A few weeks ago, I wrote about a claim that "only about four percent of all marriages today are between a Republican and a Democrat."  I gave some results from surveys that asked people how their spouses voted, but mentioned that in some cases people might be mistaken.  There aren't many surveys that ask members of couples separately about their votes, but I found one from 1993 by Nancy Burns, Kay Lehman Schlozman, and Sidney Verba (data in ICPSR).  Among couples in which both members voted in the 1992 presidential election, 76% voted for the same candidate and 24% split their votes.  Of course, that election was unusual because an independent candidate (Ross Perot) got a large number of votes.  Limiting it to Bush and Clinton, 67% voted for the same candidate, and 7% split their votes.  

The survey didn't ask people how they thought their spouse voted for President, but they did ask whether they disagreed on any race.  In 47% of the couples, both said no, and in 22% both said yes.  In the other 31%, one partner thought that they didn't disagree on any races and one thought they did.  Breaking votes for president down by perceived disagreement:

                                 Same            Different

Both No                   96%              4%                      [89%-0%]
Split                         68%              32%                    [60%-15%]
Both yes                   46%              54%                    [32%-12%]

The figures in brackets are limited to Bush and Clinton votes.  Looking at it another way, almost half of the split votes, and more than half of the Bush/Clinton split votes, came from couples in which one member thought that they agreed on all the races (there were no clear gender differences in accuracy of perception).  

Although the data are old, I think that the general point is still relevant--if there's ambiguity, people tend to assume that their friends and family vote the same way that they do.  It's possible that increased political polarization has made people pay more attention to evidence about what their friends and family actually think, but it's also possible that it's made people more likely to avoid political discussions and more likely to assume that reasonable people agree with them.  

Two nations?

 A couple of weeks ago, the New York Times had a piece on the upcoming election that said "Red and Blue Americas are moving farther and farther apart geographically, philosophically, financially, educationally and informationally."  It went on to say "In 1960, about 4 percent of Americans said they would be displeased if their child married someone from the other party. By 2020, that had grown to nearly four in 10. Indeed, only about 4 percent of all marriages today are between a Republican and a Democrat."  I already wrote about the last point.  I think it refers to perceived voting differences from one's spouse (there aren't many political surveys which interview both members of a couple):  they are rare today, but were also rare in the previous years for which I had data (1944, 1960, and 1984).  This post will consider the hypothetical question about a child marrying someone from the other party.  This is a summary of all the relevant questions I could find:

Year                       Positive               Negative             Survey

1960                     14%                       4%                         Almond/Verba
2008                                                     24%                     YouGov               
2010                                                     41%                     YouGov
2014                     28%                       17%                       Pew
2017                                                     14%                      PRRI
2018                     69%                       35%                       PRRI     
2020                                                     38%                      YouGov

The three figures in boldface involve the same question:  the others all had different questions and different response categories.  Some of the surveys just asked one question about marrying someone from the other party, but others asked everyone two questions--one about marrying a Republican and one about marrying a Democrat.  For the ones with two questions, I also show positive responses--people who say they would be pleased if the hypothetical child marries in the party (rather than saying it wouldn't matter to them).  There's clearly been an increase in both positive and negative reactions, although the differences among the questions makes it hard to say much about its exact timing (I wish people had stuck with the 1960 question).

Why do we have stability with (perceived) party differences within actual marriages but increasingly negative reactions to party differences in a hypothetical marriage?  People generally know something about their spouse's political views (although there's undoubtedly a tendency to exaggerate agreement).  But with the hypothetical question, they have to come up with an idea of what an unspecified Democrat or Republican would be like.  The most likely source for that would be prominent Democratic or Republican political leaders.  In a time when ideological differences between parties were small and political leaders tried to show respect for the other side, that wouldn't seem so bad.  But with larger differences and more conflict between the parties, it would.  That is, increased objections to a hypothetical marriage to someone from the other party don't necessarily reflect increased social distance between Democrats and Republicans in the public--they could just reflect increased differences at the elite level.

[Data from the Roper Center for Public Opinion Research]

All together now?

 A few days after the 2008 election, a USA Today/Gallup poll asked "In dealing with the problems facing the country, do you think Barack Obama will make a sincere effort to work with the Republicans in Congress to find solutions that are acceptable to both parties?" and parallel questions about whether the Republicans would make a sincere effort to work with the Democrats and Obama and whether the Democrats would make a sincere effort to work with the Republicans.  In March and September 2009 they asked about whether the various parties had made a sincere effort to work with each other, and in February 2010 they asked about working together on health care reform.  They also asked the forward-looking questions after the 2010, 2012, and 2016 elections.  The figure shows the percent who said that Obama, the Democrats, the Republicans, or Trump would (or had) worked with the other party:

Obama consistently ran ahead of both the Democrats and Republicans in Congress, and the Democrats were generally somewhat ahead of the Republicans, but they all rose and fell together.   In principle, you might expect that they would move in opposite directions at least some of the time:  that people would see one side as being obstructionist and give the other side credit for trying, but that didn't happen with any of these surveys.  Rather, the public seemed to blame both sides about equally when there was disagreement (if anything, Obama's ratings might have fallen more than the Congressional parties).  

I think this data helps to explain why Republicans turned against the Senate immigration deal.  If it had passed, Biden would have gotten some of the credit from the public, and most Republicans are unwilling to do anything that will make Biden more popular (several of them said as much).  A few years ago, I suggested that a strategy of uniform opposition had driven down Obama's popularity.   Republicans have continued with that under Biden.  Of course, there has been some important bipartisan legislation, like the American Rescue Plan Act, but they were mostly early in his term and my impression is that the Republicans have tried to avoid publicizing them.  It used to be that when popular legislation was passed on a bipartisan basis, both parties would talk about it and try to claim some of the credit.  But more recently, people seem to have realized that elections are more about the President than about Congress, so for an opposition party, denying credit to the President is more important than claiming credit for yourself.  And blocking potentially popular legislation might make your side less popular, but it will probably make the other side less popular as well.

A few other observations:

1.  The numbers expecting the parties to make a sincere effort to work with each other were higher than I expected. 
2.  Just after the 2016 election, 58% expected Trump to make an effort to work with the Democrats, which was somewhat ahead of the number who expected the Republicans in Congress to work with the Democrats.  That might be because in 2016, many people saw Trump as a "dealmaker" rather than a traditional conservative, or it may be that there is a tendency to have hopes for a new president.  
3.  It's not possible to be sure, but it seems that the questions that asked about the future produced more positive responses than those that asked about the past.
4.  Following from the previous points, it's unfortunate that this question hasn't been asked since 2016--I would like to see how expectations have changed.

[Data from the Roper Center for Public Opinion Research]

Republicans and reality

 If you go by the standard statistics, the economy is doing very well--unemployment remains low, inflation and interest rates are coming down, the stock market is rising.  Yet the public still isn't impressed, although ratings have been improving recently.  Paul Krugman suggests that the reason is politics--that Republicans have rated the economy as bad ever since Joe Biden took office, and have continued to do so even as economic conditions have improved.  He says that Republican "beliefs are nearly impervious to reality."  

The Michigan Consumer Surveys have a monthly "index of consumer sentiment" which goes back to the 1940s.  Since April 2017, they have regularly asked people about their party identification.  The figure shows consumer sentiment by party during the Trump administration:

There wasn't much variation among either Democrats or Republicans until Covid hit in early 2020 and ratings declined among both before rising a little in the fall.  I count January 2021 as the beginning of the Biden administration, so changes in the last two months are probably a response to the election results.  Through October 2020, the correlation between Democratic and Republican ratings is about 0.9.  

During the Biden administration:

Ratings diverged in the first few months and Democrats became more positive and Republicans became more negative--since the summer of 2021 they have tracked each other pretty closely.  Over the course of the Biden administration, the correlation has been about 0.75.  That is, Republicans and Democrats responded to reality in similar ways during both administrations (and to similar degrees--the standard deviations were about the same).  Of course, politics make a difference--Democrats were more favorable under Biden and Republicans were more favorable under Trump.  It's possible that the change of administration had more impact on Republicans--this seems to be the case for beliefs about crime rates and the prospects of the next generation.  But it's hard to say, because economic conditions also changed substantially between late 2020 and mid-2021:  unemployment fell but inflation rose.  

The general point is that negative perceptions aren't just Republican partisanship:  something was making Democrats feel worse about the economy between mid-2021 and mid-2022.  

And something has made Republicans feel better since mid-2022:

 That still leaves the question of why views were so negative in mid-2022, and are still pretty negative despite recent improvement.  In August 2022, unemployment was 3.7%, annual inflation was 8.2%, and the index was 53.2.  In November 1973, unemployment was worse (4.8%), inflation was essentially the same (8.3%), but the index was twenty points higher (76.5).  I'd say that the two most plausible explanations are (a) a general shift towards more negative assessments, maybe because of more negative media coverage and (b) a stronger reaction to the inflation of the 2020s because it came suddenly and people weren't used to it.  

News not fit to link

 A recent piece in the New York Times says "In 1960, about 4 percent of Americans said they would be displeased if their child married someone from the other party. By 2020, that had grown to nearly four in 10. Indeed, only about 4 percent of all marriages today are between a Republican and a Democrat."  That is, they included a link to the source of the second piece of information, but not the first and the third.  But my standards of what's fit to link are lower, so I'll try to provide them.

The first is from The Civic Culture, by Gabriel Almond and Sidney Verba, but the third was new to me.  Ideally, you would have a survey that had separate interviews of both members of married couples, but there aren't many like that, so I looked for surveys that asked people about spouse's politics.  I couldn't find any that asked about their spouse's party identification or registration, but there was one from 2016 (just after the election) that asked people about how they and their spouse had voted.  There were also some earlier surveys that had parallel questions.  The results:

            Same     Different      Ratio  
1944   72%         4%                   18
1960   67%         5%                   13.4
1984   64%        6%                    10.7
2016   63%        4%                    15.8
[2016  68%       10%]                   6.8

Although I can't be sure, I'd guess that the 2016 survey was the source of the statement in the Times.  In any case, it makes it possible to compare things to the past.    In 1944, 1960, and 1984, almost all votes were for the Democratic or Republican candidate--the columns don't add to 100 because some people said that their spouse hadn't voted and others didn't know how they'd voted.  But in 2016, about 6% of the vote went to other candidates, and the figures in brackets include those votes.  If you don't count the 2016 "others," there's no clear pattern--the samples are only one or two thousand, so there's a good deal of sampling error.  If you count the "others," there was more intra-marriage disagreement in 2016 than in earlier elections.  But maybe at least some of those should be counted as intermediate (e. g., one for Trump, one a write-in) rather than disagreement?  I won't get into that--I'll just observe that the surveys don't provide evidence that married couples are more likely to vote the same way now as they were in the middle of the 20th century.  But what about the question about how you would feel if a child married someone from the other party, where there is evidence of change?  I'll consider that in a future post.  

[Data from the Roper Center for Public Opinion Research]

.   

Indictments

 After the 2022 election, it looked like Donald Trump's support in the Republican Party was finally weakening. As Trump made a comeback in the middle of 2023, some people said that one reason for his resurgence was that Republicans were rallying around him because he had been indicted on various charges.  Now this seems to have become conventional wisdom:  a news story in the NY Times says "But far from diminishing the former president’s standing with Republicans, the charges actually rallied the party around him."

A few months ago, I looked for relevant data.  Lots of surveys asked if you had a favorable or unfavorable view of Trump, but I wanted ones that asked for degree of favorability--my idea was that the indictments wouldn't convert people from unfavorable to favorable, but they might make people who were already favorable more strongly committed.  Surveys that ask people for degree of favorability or unfavorability are less common, and I didn't find enough for an analysis.  After the New Hampshire primary, I tried again and found a source I hadn't known about before:  a company called Echelon Insights has monthly polls that include a question about views of Trump (very favorable, somewhat favorable, somewhat unfavorable, and very unfavorable) and breaks them down by party identification.  Very and somewhat favorable ratings of Trump among Republicans:

The first indictment came on March 30, after data collection for the March survey was complete.  There was a lasting increase in very favorable ratings and decline in somewhat favorable ratings starting in April.  Of course, in principle the pattern could be the result of something else that happened at around the same time, but I can't think of any other obvious candidate.  There was no lasting change after the second and third indictments, but it seems reasonable that the first one would have more impact.  

There also has been some polarization of ratings among independents, with both very favorable and very unfavorable ratings becoming more common, but this was a gradual change--there's no sign that the indictments had any special impact.  The latest figures among independents are 17% very favorable and 46% very unfavorable.   For Democrats, very unfavorable ratings have been steady at about 85 percent over the whole time period.  

You don't dislike me, you really don't dislike me

 A few days ago, David Brooks wrote that Donald Trump "has an advantage that Haley can't match.  He is reviled by the coastal professional classes. That’s a sacred bond with working-class and rural voters who feel similarly slighted and unseen."  Today, Damon Linker offered a similar analysis: "fundamental to Mr. Trump’s strength is populist anger at 'them' — the progressive-leaning elites who graduate from the country’s most selective universities, control the commanding heights of culture, run America’s leading public institutions and media outlets and sneer at him and his supporters, calling them racists, xenophobes, misogynists and fascists."  I could go on piling up citations--Bret Stephens wrote something similar a few days ago--but this is a blog post, not a research paper, so I'll get to the data.  In 2021, a Pew survey asked how people felt about various possible characteristics of political leaders, including "Has a degree from a prestigious university, such as Harvard or Stanford."  Overall, 7% said they liked that a lot, 14% that they liked it a little, 5.6% that they disliked it a little, and 5.8% that they disliked it a lot.  The majority (about 67%) said they neither liked nor disliked it.    What if we break it down by groups?  To make comparison easier, I show the percent who say they like it minus the percent who say they dislike it--e. g., among college graduates, 23% say like and 10% say disklike, for a net of +13.

College grad        +13
Some college         +8
No college             +8

Metropolitan       +12
Non-metro             -3

Educational differences are small, and there are more likes than dislikes in all groups; the metro/non-metro difference (coded by the survey organization, not a self-report) is bigger, and people outside of metropolitan areas are more likely to see an elite degree as a minus than as a plus. 

There are substantial partisan differences:

Democrats               +18
Republicans                 0
Neither                      +5

Turning to some other groups:

aged 18-29                    +15
         30-49                   +15
         50-64                   +4
         65+                      +3

White                            +7
Black                             +7
Hispanic                        +20
Asian                             +17

Men                                +8
Women                           +11

I was a bit surprised at the age differences:  my impression is that anti-elite rhetoric has been growing in recent years, so younger people would be more attuned to it.  (For example, Donald Trump seems to admire elite universities:  he frequently boasts that he's not just a graduate of the University of Pennsylvania, but of the Wharton School, and a secondary part of his "birther" campaign was a demand that Obama release his academic records to prove that he deserved to be admitted to Columbia and Harvard Law School).  The Pew survey contains a question that has some bearing on this issue:  who was the best president of the last 40 years.

Reagan                +4
GHW Bush        +13
Clinton               +14
GW Bush              -0
Obama                +20
Trump                    -5
Biden                    +8

As the differences by party identification suggest, people who name one of the Democratic presidents as the best have more favorable views of degrees from elite universities, but the people who name Trump are more negative than those who name Reagan or one of the Bushes.

Some of the partisan division is undoubtedly created by the parties--Republican leaders denounce elite universities, so Republican voters follow them and Democratic voters react against them--but the substantial divisions by age, ethnicity, and metro status suggest that opinions have some independent basis.  However, education itself doesn't make much difference, and only 13% of people who didn't attend college see having a degree from an elite university as a negative.    

[Data from the Roper Center for Public Opinion Research]