What Don't We Want?

The New York Times has had several articles recently noting that growing inequality over the last 40 years hasn't led to growing support for redistribution.  One of them, entitled "Why Americans Don't Want to Soak the Rich," says that since the 1970s "Americans’ views on whether the government should work to redistribute income — to tax the rich, for example, and funnel the proceeds to the poor and working class — have, depending on which survey answers you look at, either been little changed, or shifted toward greater skepticism about redistribution."  I agree with that--I'd favor the "little change" summary, but you could argue it either way.  Then it says "in other words, Americans’ desire to soak the rich has diminished even as the rich have more wealth available that could, theoretically, be soaked."

I don't agree with that--at least the "in other words" part, because taking from the rich and giving to the poor are two different things.  There are lots of survey questions about aid to the poor, but few about taking from the rich.  There are some that combine the two issues, for example by asking about "reducing differences between people with high incomes and people with low incomes," but then you don't know which aspect people are focusing on in their answers--my guess is that it's the low incomes side.  So we don't really know how opinions about "soaking the rich" have changed. The same is true for international comparisons:  there's a lot of evidence that Americans are less favorable to aid to the poor than people in Western Europe, but very little on opinions about  taking from the rich.

But the 2009 edition of the International Social Survey Programme, which I've written about before, contains some interesting information.  It had the following questions:


1.  Agree or disagree that "differences in income in [country] are too large."
2.  Agree or disagree that "it is the responsibility of the government to reduce differences in income between people with high income and those with low incomes." [the 1999 question]
3.  Agree or disagree that "the government should provide a decent standard of living for the unemployed."
4.  Agree or disagree that "the government should spend less on benefits to the poor."
5.  "Do you think people with high incomes should pay a larger share of their income in taxes than those with low incomes, the same share, or a smaller share?"
6.  "Generally, how would you describe taxes in [country] today for those  with high incomes?  ... much too high, too high, about right, too low, much too low"
7.  "Is it just or unjust--right or wrong--that people with high incomes can buy better health care than people with lower incomes?"
8.  "Is it just or unjust--right or wrong--that people with high incomes can buy better education for their children than people with lower incomes."

Overall, the United States is one of the least egalitarian nations, but there are differences among the questions.  I give the standardized scores, which are comparable across questions, with higher meaning less egalitarian:

2.83  government responsibility to reduce differences
2.20  government should provide for unemployed

1.27 differences in income are too large
*1.00 people with high incomes should pay a larger share in taxes

0.75  unjust buy better education
0.69 unjust buy better health
0.45 spend less on poor
*0.15 taxes for those with high incomes

The two items with asterisks involve the rich.  Americans are less favorable to "soaking the rich" than average, but the differences aren't that big--in fact, Americans are barely different from the average in views about whether taxes for people with high incomes are too high or too low.   The American mean is  3.29 and the mean for all nations is 3.34 on a scale of 1-5, with 3 meaning about right and higher scores meaning too low).

Where the United States really stands out is the two questions that mention the government's responsibility.

More about mobility

I was surprised by the results on educational mobility in my earlier post on the subject , so I did some more investigation.   People progress through educational categories in a definite order:  everyone starts at the lowest level and moves up one step at a time.  So you can regard educational mobility is as a series of transitions--how much does father's education predict your chance of going beyond grade school, going to high school, graduating from high school, etc.?  I estimated the association by generation for each transition.  This figure shows the first three (going beyond grade school, going to high school, graduating from high school).

There is an upward trend in the influence of father's education on all three transitions.  More exactly, there was an upward trend, although it's not clear that it's continued in the last couple of generations.

This figure shows the next two--attending college and graduating from college given that you attended.  Again, both show an upward trend.    

There's one more transition--getting graduate education--but the effect of father's education is near zero in all generations.  That is, if you've graduated from college, how much education your father has doesn't predict whether you will get graduate education.  

Although there may be some deviations, like the jump in the influence of father's education on going to college in the last generation--the overall picture is of a steady increase in the influence of father's education at all levels.  However, father's education has more influence on the lower-level transitions--the numbers on the y scale aren't meaningful in an absolute sense, but they can be compared to each other.  Given the rise in average educational levels, the lower level transitions became numerically less important--today almost everyone at least makes it into high school--and that masked the decline in mobility at each level.

That raises the question of why would there be a trend over a 100 year period, through depressions, prosperity, declining economic inequality, rising economic inequality,  urbanization, suburbanization, etc?

Religion and American Life

In a recent article, Thomas Edsall gave a graph showing that the percentage of people saying that religion was losing influence had increased substantially between 2012 and 2014.  I thought that the question went farther back than that, so I searched for it.  I got more than I bargained for, because it was asked no less than 78 times between 1957 and 2014, first by Gallup, which still asks it from time to time, and then by other organizations, notably Pew.  The exact question is:  "At the present time, do you think religion as a whole is increasing its influence on American life or losing its influence?"  Sometimes people volunteered that it was staying the same, or some other intermediate answer like "some of both," but those number varied widely (from 1% to 17%), so I think that organizations must have differed in the extent to which they pressed for an answer.  To take them out of the picture, I took the odds of "increasing" to "losing."  For example, the last time it was asked, 22% said increasing and 72% said losing, for odds of about 1:3.3.  Here are the results over time (shown on a log scale):

There are a couple of obvious changes:  a big shift towards "losing influence" in the 1960s and a large but short-lived shift towards "increasing influence" after 9/11.  But there are also several other shifts that took place over a substantial period of time:  towards "losing influence" from the mid-1980s to the early/mid 1990s, then back towards "increasing influence until about 2000, then towards "losing influence" since 2002.  I can't think of a plausible reason for them:  they don't seem to follow the political climate, or trends in things that people might think of as moral problems like the crime rate or divorce.  But given the amount of data here, they're not just chance variation.

[Data from iPOLL, Roper Center for Public Opinion Research]

Serious sociology

This is a follow-up to my post of a couple of weeks ago.  In the general media, I've encountered a lot of claims that educational mobility is declining--that it's become less common for children of parents with little education to move "up" or children of parents with lots of education to move "down".*  Although I'm reasonably familiar with the scholarly literature on occupational and income mobility, I don't know of anything on educational mobility, and a search of Google Scholar didn't come up with anything very promising.  

So I did a basic analysis using the General Social Survey, which asked people how much schooling they had, and also how much their father had.  I made a tables of father's education by own education for six groups:  born in 1909 or before, born 1910-19, 1920-29, 1930-54, 1955-69, and 1970-90.  Both were classified into seven categories:  6th grade or less, 7th or 8th grade, some high school, high school graduate, some college, college graduate, and graduate education. I restricted this to white men because it seems pretty likely that the trends will be different for women and non-white men, and to people age 25 or over because many of those below that age are still in school.

I fit a model that described the changes with a single number representing the difference from the last cohort.  The number represents the association between father's and son's education, so high numbers mean less mobility.  Here are the results as a figure:

The conventional wisdom seems to be right:  the association is substantially higher in the last generation (people born 1970-90), although there was some upward trend before then.  The scale of the vertical axis isn't meaningful, but in a rough way you could say the association is about twice as strong in the youngest cohort as in the oldest.  

  *I mean those in relative terms--how much education a person has relative to what other people in their generation has.  In an absolute sense, upward movement has been much more common, because average levels of education rose from generation to generation until the 1970s.  

Caring about college

Scott Walker, an all-but-declared candidate for the Republican presidential nomination, didn't graduate from college.  That's unusual--as far as I know, the last "major" candidate for president without a college degree was Paul Simon, who sought the Democratic nomination in 1988.  Ted Cruz, a declared candidate for the Republican presidential nomination, got an undergraduate degree from Princeton and a law degree from Harvard.  That's not unusual--George W Bush had degrees from Yale and Harvard.  But most Republicans with degrees from elite schools have downplayed them--Bush talked about Midland, Texas, a lot more than he did about Harvard or Yale-- while Cruz has made them a central part of his story.

Do voters care about this sort of thing?  A 2007 Pew survey asked respondents if they would be more or less likely to vote for a candidate with various characteristics.  One was "attended a prestigious university such as Harvard or Yale," and another was "did not attend college."  People were split on a candidate who hadn't attended college:  49% said it would make no difference and 47% said they'd be less likely to support the candidate (only 3% said they would be more likely to).  Most people (74%) said that attending a prestigious college wouldn't make any difference; 20% said more likely and only 5% less likely.

A common perception is that liberals value educational credentials, while conservatives are less interested, or maybe even hostile.  In fact, that was why I started thinking about the issue:  would Cruz's association with elite colleges hurt him with "the base"?  According to these questions, there might be some difference in the expected direction, but it's small,  Another potential influence is education:  you could make arguments for an effect in either direction.  It turns out that, less educated people may be a little more likely to say that not attending college wouldn't matter; education has no clear effect on views about having attended a prestigious university.

I had no reason to expect gender to make any difference, but since nothing very interesting was going on, why not try?  Gender didn't make much difference for reaction to a candidate who hadn't attended college, but made a substantial difference for reactions to one who attended a prestigious university:  8% of men, and only 2% of women, said they would be less likely to support one.  To put it another way, more than three-quarters (60 out of 79) of the people who said they would be less likely to support the candidate were men.

Even among men, attending a prestigious university was much more likely to be a plus than a minus, but the gender difference needs an explanation.  My thought is that a "man of action, not words" image has appeal to some men, and there's no real parallel to that image for women.

[data from the Roper Center for Public Opinion Research]

Forgotten Beauty

In March 1951, the Gallup Poll had a question in which they gave people a picture (at that time, the polls were done in person rather than over the phone) and said:  "Here are pictures of six different girls all dressed alike.  I'd like you to look at each one and tell me which ONE you think is most beautiful." #4 was the most popular choice with about 57%, followed by #3 and #5 with about 15% each.   The others were at 5% or less.

Preference varied by education:  #4 was uniformly popular in all groups, but  #3 was the choice of less than 10% of the people with a grade school education and 28% of the college graduates, while #5 was the choice of more than 20% of those with a grade school education and only 5% of the college graduates.  There were similar, although less dramatic, differences by occupation and economic level:  #3 got more support in the "higher" groups and #5 got more support in the lower.  There were also big differences by race:  among blacks, #5 led #3 by about 30%-8%; among whites #3 led by about 16%-13%,

There were no gender differences in preferences for #3, but compared to men, women were more likely to prefer #4 and less likely to prefer #5.  Older people were less likely to prefer #3, but only a little more likely to prefer #5--they were more likely to prefer #1 and #2 (age was the only factor that had a clear connection to preferences for #1 and #2).

At this point, I should show the picture, but I don't have a copy and have never seen it.  It's not included in the documentation at the Roper Center.  The Gallup Poll used to have a newspaper column in which they highlighted findings from their polls, but I did a search and couldn't find it there.  The question is prefaced with "we have been asked to find out", which suggests they did it for some client.  Maybe it was a magazine that published an article about the poll, but if so I haven't found it.  Or maybe it was a commercial client (they occasionally put "market research" questions in their regular polls), in which case it's probably lost for good.

In conclusion, there seems to be something interesting going on, but I have no idea what it is, and little hope of ever finding out.

[Data from the Roper Center for Public Opinion Research]

If you're so smart.....

In a review of Robert Putnam's new book, Jason DeParle (a reporter for the New York Times) says "where Putnam succeeds is in describing the diverging life chances of children in rich and poor families. ('Rich' parents finished college; 'poor' parents have high school degrees or less.) "   Of course, education is not really equivalent to riches, as many adjunct professors could tell you, but this led me to wonder exactly how much overlap there is the incomes of different educational groups.  

It's easy to find statistics on differences in average income by education, but harder to find information on the whole distribution.  Here is a table, which I made from Bureau of Labor Statistics data on "usual weekly earnings of full-time wage and salary workers."

                           Percentiles

                  10th   25th  50th    75th   90th

Not HS Grad       301     374 488     657    887

HS Grad           367     482 668     960   1355

Some College      398     532 761    1111   1560

Bachelor's        529     744 1101   1647   2368

Advanced Degree   588     953 1386   2009   2974

Education raises the "ceiling" more than the "floor"--the 10th percentile for people with an advanced degree is less than twice the 10th percentile of people who aren't high school graduates, while the 90th percentile is about 3.3 times as high.  You can get a sense of the overlap by comparing percentiles--for example, someone who's at the 90th percentile for those who aren't high school graduates would be well below the median for people with a bachelor's degree.  

For a more detailed comparison, I assumed that earnings followed a log-normal distribution (which is usually approximately true except in the highest levels) and estimated the mean and standard deviation from the data above.  The resulting distributions are shown in this figure (I scaled it up to earnings per year):

The figure makes it clear that people without a high school degree have little chance of being rich, or even affluent.  People with just a high school degree, however, have a reasonable chance to make it to the upper middle class.