Friday, February 26, 2021

Gradually, then suddenly

 In 1974 and 1976, the Gallup Poll asked "More generally, how much trust and confidence do you have in the American people as a whole when it comes to making judgments under our democratic system about the issues facing our country--a great deal, a fair amount, not very much, or none at all?"  This question didn't appear again until 2001, but since then it's been asked fairly often, mostly by Gallup but also by the Monmouth University Polling Institute.  The averages, with higher values indicating more confidence:

It was lower in 2001 than in the 1970s, then rebounded for a few years, maybe as part of the general feeling of solidarity after 9/11 (the 2001 survey was taken just a few days before), then declined.  So there was definitely a downward trend.  But there was a sharper decline between 2010 and 2016, and then it stayed low before rebounding in the latest survey (Jan 21-24, 2021).  

A question about "trust and confidence in the wisdom of the American people"  showed a similar pattern of gradual decline and then sharper decline during the Obama administration.  This is different from the  decline in confidence in institutions, which has been pretty steady.

[Data from the Roper Center for Public Opinion Research]

Saturday, February 20, 2021

You could look it up, part 3

 My last couple of posts have argued that popular support for activist government didn't decline with the rise of the civil rights movement--if anything, it's somewhat higher today than it was in the 1950s.  So what is my explanation for what Heather McGhee called "the end of America’s high-tax, high-investment growth strategy"?  I don't have one, because it didn't happen.  Government spending and taxes as a percentage of GDP have grown since the 1960s (this is a convenient chart for spending and this is one for tax revenue). McGhee says that "tax revenue peaked as a percentage of the economy in 1969 compared with the average O.E.C.D. country," which may be true (although the NY Times story didn't include a link to the source for that either).  However, that's not because taxes have gone down in the US, but because they've gone up faster in other OECD nations.  

This piece illustrates a view that's popular on the American left--that the only reason that Americans don't support an extensive welfare state is racial animosity.  Although that's almost certainly a factor, it's not the whole story.  In 2012, I had a post comparing opinions about redistribution in thirty-eight nations.  New Zealand was least favorable, the US was third least, Britain was fourth, and Australia was sixth.  That is, it's not the United States that is distinctive, but Britain and nations that were settled from Britain.  (Canada wasn't included in the survey).  That is, there seems to be some lasting cultural difference in nations with a British heritage. 


Friday, February 19, 2021

You could look it up, part 2

 My last post was about a claim that white support for welfare state programs dropped sharply between 1960 and 1964, when the civil rights movement called for directing more of their benefits to black people.  I pointed out that there was a potentially important change in the question (from the ANES) between 1960 and 1964:

(Same introduction as in VCF0805 [CARD WITH RESPONSES SHOWN]).
'The government in Washington ought to see to it that everybody who
wants to work can find a job.'
In general, some people feel that the government in Washington should
see to it that every person has a job and a good standard of living.
Others think the government should just let each person get ahead on his
own.'  Have you been interested enough in this to favor one side over
the other.  (IF YES:)  Do you think that the government --
To complicate things further, the ANES has run another question starting in 1972:
Some people feel that the government in Washington should see to it
that every person has a job and a good standard of living.  (1972-
1978,1996-LATER: Suppose these people are at one end of a scale, at
point 1).  Others think the government should just let each person get
ahead on his/their own. (1972-1978,1996: Suppose these people are at
the other end, at point 7.  And, of course, some other people have
opinions somewhere in between, at pints 2,3,4,5 or 6.)
Where would you place yourself on this scale, or haven't you thought
much about this?  (7-POINT SCALE SHOWN TO R)

A few other surveys have asked each of these questions over the years, so I tried to put them all together to get a sense of change over the whole time since 1956.  

I use the log of the odds ratios of support to opposition, since the percent of don't know/middle responses varied substantially.  For the seven point question, I counted 1-3 as support and 5-7 as opposition.  

It's clear that support is higher for the first form, which doesn't mention standard of living and also says "everyone who wants to work" rather than "every person."  With that form, support was higher in the 1970s than in 1956-70 or in 2012.  Since it's been asked only once since the 1970s, we can't be too confident about changes in that period, but it clearly didn't decline after the 1950s (the 2012 estimate is slightly higher than the 1956, 1958, and 1960 estimates).  

The second and third forms show about the same level of support (oddly, the ANES codebook treats 1 and 2 as variants of the same question, and 3 as a different question).  In order to see changes in the those variables more clearly, here's the figure omitting type 1:

Two of the results are from a NORC survey from early 1974.  It seems like that survey had an unusually liberal sample--I've looked at some other questions from it, and they all seem out of line with results from most contemporaneous surveys.  If you exclude those two, there might be an upward trend--towards more support for government responsibility.  If you include them, there's no evidence of a trend.   

So overall, there are two interpretations that are consistent with the data (1) support for government action increased from the 1950s to the 1970s, and then declined, so that there's no trend over the whole period, or (2) support for government action increased from the 1950s to the 1970s and then stayed about the same or increased slightly, so that it's now higher than in the 1950s.  The one possibility we can rule out is that popular support for government action has declined since the 1950s.  

That leaves the question of how I would explain "the draining of public resources and investments" since the 1950s.  I'll take that up in my next post.

[Data from the Roper Center for Public Opinion Research]

Sunday, February 14, 2021

You could look it up

 There was an opinion piece by Heather McGhee in the NY Times this morning which contained the following passage:  "In the late 1950s, over two-thirds of white Americans agreed with the now-radical idea that the government ought to guarantee a job for anyone who wants one and ensure a minimum standard of living for everyone in the country. White support for those ideas nose-dived from around 70 to 35 percent from 1960 to 1964, and has remained low ever since."  This statement seems to be based on a question in the American National Election Studies: 

(Same introduction as in VCF0805 [CARD WITH RESPONSES SHOWN]).
'The government in Washington ought to see to it that everybody who
wants to work can find a job.'
In general, some people feel that the government in Washington should
see to it that every person has a job and a good standard of living.
Others think the government should just let each person get ahead on his
own.'  Have you been interested enough in this to favor one side over
the other.  (IF YES:)  Do you think that the government --

 White support for the first option was about 55% in 1956, 1958, and 1960, and a little over 25% in 1964 and 1968. (Both are lower than the numbers McGhee gives, but she may have been omitting the "some of both" or "don't know" answers). But  is that the same question in 1960 and 1964? The 1964 question omits "wants to work" and adds "good standard of living" as something for the government to do, and also gives an alternative--"just let each person get ahead on his own." To me, those changes sound like they could make a difference.

 Fortunately, there's some evidence on this point--in 1974, an NORC survey asked the earlier form, and found that 78% agreed, 10% disagreed, and 12% had no opinion.  It also asked the 1964-8 form to a different part of the sample, prefaced by a question asking "whether you have been interested enough in this to favor one side over the other?"  Combining those two questions, 43% chose "job and standard of living," 26% "get ahead on his own," and 27% didn't have an opinion. So the two different questions get substantially different responses. The job and good standard of living question has been asked a number of times since then, usually finding a majority on the "get ahead on his own" side. The "see to it that everybody who wants to work can find a job" was asked several more times in the 1970s, with about 75% agreeing. So it seems pretty clear that the story of a lasting decline in white support for government action starting in the early 1960s is not correct. If anything, support for government action was somewhat higher in the 1970s than in the 1950s. 


PS:   My title is sort of misleading.  I could look it up because I'm familiar with public opinion data, so I knew where to look.  But the story didn't have a link or any kind of information about the source, so the average person couldn't.  

Saturday, February 13, 2021

Winners and losers

 The Pew Research Center recently released a poll which asked whether certain kinds of people would gain or lose influence now that Joe Biden was president.  In addition to the overall figures, they showed breakdowns by party.  For example, 32% expected older people to gain influence and 35% expected them to lose influence (the rest expected no change), but there was a big difference between supporters of different parties:  among Republicans, 16% expected them to gain and 57% expected them to lose; among Democrats, 45% expected them to gain and only 15% expected them to lose.  

You might regard the gap between partisan perceptions of a group as a measure of esteem for that group--the extent to which supporters of a party want to claim it as one that they are concerned with.  So I computed the difference (Democrats gains-Democrats lose)-(Republican gains-Rep lose):  for older people, that's (45-15)-(16-57)=71.  The ranking, from most to least:

People like yourself        89

Military                    78      

Poor people                 76

Older people                71

Children                    63

Asian-American              54

Hispanics                   51

Whites                      46

Women                       42

Blacks                      39

Evangelicals                32

Young people                31

Men                         28

Unions                      22

Gay & lesbian               10

Corporations                -6

Wealthy people              -51

"People like yourself" isn't really comparable to the others, since it doesn't have a common meaning for Democrats and Republicans, but I show it anyway.  The general ranking of the groups is pretty plausible, although there are some surprises, particularly the high ranking of "poor people"--Democrats overwhelmingly thought that poor people would gain influence, but 42% of Republicans thought they would lose and only 29% thought they would gain.  Wealthy people stand out as a group no one wants to be associated with--Democrats think they'll lose influence under Biden and Republican think they'll gain.  It's interesting that they get a more negative reaction than "corporations"--often people are more favorable to people than to institutions.  

What about the relatively low ranking of men?  You could also rank the groups on another dimension:  the percent of people who thought they would either gain or lose.  This could be regarded as the extent to which the influence of the group was politicized--that it would rise or fall depending on which party was in power.  A scatterplot of the two dimensions:

Men rank low on what I'm calling politicization--most Democrats and almost half of Republicans don't expect their influence to change either way.  In contrast, about 80% of Republicans and 80% of Republicans expect a change in the influence of black people.  The rankings on this dimension seem to fit well with  my interpretation:  men and whites are well represented in positions of power regardless of which party is in office.  Asian-Americans don't have a high political profile, and children aren't seen as actors in politics (perceptions of which party is better for children might be more politicized).  

Pew has asked the same general question at the beginning of administrations going back to 1992 (although the list of groups has changed).  I will look at changes in partisan perceptions in a future post.  


Sunday, February 7, 2021

I don't make the rules, part 2

 My last post involved a question asked in the GSS:  agreement or disagreement with the statement:  "people like me don't have any say about what the government does."  The same question has appeared regularly in the American National Election Studies starting in 1952, but with only two or three options (agree or disagree, plus "neither agree nor disagree" starting in 1988) rather than the five offered by the GSS (strongly agree....strongly disagree). This post is about the ANES question.

I used the logarithm of the ratio of disagree to agree responses (in the early years, over 90% of people with a college degree disagreed, so ceiling effects are an issue).  Here are the trends among college graduates and everyone else:

A pretty steady decline in "efficacy" (rejection of the statement that you have no say) among both groups, although if you look more closely there are deviations in the 1990s and 2000s).  But the decline seems faster among college graduates.  Here is the difference in the logs of the odds ratios:

The first two (1952 and 1956) are much higher, then a long period without much trend, and then apparently a decline in the gap in 2012 and 2016.  The GSS and ANES series have only four dates in common, but they match up well in those years.  Both agree in showing a decline in relative feelings of efficacy among college graduates in recent years.  Many accounts of contemporary society hold that there's a growing gap between classes, in which the educated middle classes remain pretty satisfied while the working classes sink into despair.  I've had some posts questioning that claim, and this is another point against it.