Tuesday, June 30, 2020


Princeton University has just removed Woodrow Wilson's name from its School of Public and International Affairs.  It seemed like a tough call to me--Wilson had significant accomplishments. both as president of Princeton and president of the United States, but also significant negatives.  However, contemporary opinion seems to be overwhelmingly in favor of removing his name and almost no one has anything good to say about him--even Ross Douthat, who favors keeping his name for historical reasons, calls himself a "Wilson-despiser."  That led me to wonder about what the general public thought about Wilson and other presidents.  Between 1956 and 1999, the Gallup Poll occasionally asked "which three United States presidents do you regard as the greatest?"   Of course, most people don't remember all of the presidents, so there's a tendency to pick the current president, or a recent president.  Also, as time goes on, the number of presidents to choose from increases.  Putting those together, you would generally expect a decline for any given president. 

With that in mind, here are some results.  First, the classics:


 Next, a second level of 19th and early 20th century presidents:

Support for Theodore Roosevelt and Thomas Jefferson was pretty steady, but Wilson had a sharp decline,. from 13% in 1956 to 5% in 1975, and about 1% in later years. 

Then, some mid-century presidents:

FDR and Eisenhower both decline steadily, but Truman shows an unusual pattern:  he rose from 13% in 1956 to 37% in 1975.  He then declined, but still had almost as much support in 1999, almost 50 years after he left office, as he did in 1956.

Finally, some more recent presidents: 

Kennedy declined, although from a high level--in 1999 he was still second, behind only Lincoln.  The contrast between Regan and Bush is interesting.  Both got 20-30% when they were in office, but Reagan rose after he left office, to 34% in 1999.  Carter started from a low level, but held steady or increased. 

Returning to Wilson, he does have a distinctive pattern--in 1956, he was tied for fifth, behind Lincoln, Washington, FDR, and the current president (Eisenhower), but his support declined rapidly after that.  Truman, and maybe Reagan, stand out as presidents whose reputation grew after they left office.  

It's been more than twenty years since Gallup asked the question, so they ought to try it again.  My guess is that Obama would get a lot of support (35% or 40%) and Kennedy would still be high but would decline because of competition from Obama, who has the same general type of appeal.  Trump would be on the low side for a sitting president (maybe 15%), and Reagan would hold steady or increase. 

[Data from the Roper Center for Public Opinion Research]

Thursday, June 25, 2020

Police brutality?

There are a couple of survey questions about views of the police which go back to the 1960s.  One is "In some places in the nation there have been charges of police brutality. Do you think there is any police brutality in this area, or not?"  It was asked in 1965, 1967, a few times in the 1990s, and in 2005.  I have written about it before, and noted that the belief that there was police brutality in "this area" was higher in the 1990s and after than in the 1960s, but I didn't break it down by race, so I'll do that here.*  The individual level data for 2005 are not available, so I use surveys from 1965, 1967, 1991, and 1999.
              Is                   Isn't
        White  Black         White    Black
1965        7%      34%          82%      46%
1967        5%      15%          83%      59%
1991       36%      60%          58%      39%
1999       35%      64%          60%      30%

I computed log-linear models (treating don't know as a middle answer) and got the following estimates of the racial gap in opinions:

 The difference is largest in 1965.  There's enough uncertainty in the estimates so it's hard to say if it's a downward trend or a difference between 1965 and all later years.  But the major thing is that belief in the existence of police brutality in "your area" increased among both blacks and whites.  Why?  One possibility would be that it actually did.  However, I think it's more likely that it's because media coverage increased and people talked about it more.  In support of this interpretation, there were a lot of "don't know's" in the 1960s (20%-25% among blacks), but few in the 1990s.  I think that there's been a general tendency for don't know's to decline, although I haven't studies it systematically or seen any research, but this change is unusually large.

Although the number of blacks in the samples is small, the decline in belief in police brutality between 1965 and 1967 is too large to be plausibly explained by sampling error.  1967 was the "long hot summer" of urban unrest, and the survey was taken in early August, after the Newark and Detroit riots and many smaller ones.  Although the response to those included many examples of police brutality, the riots may have made people more sympathetic to police.  Although many accounts focus exclusively on white backlash, support for a "get tough" approach is a natural reaction to disorder and crime.  

*My figure in that post has a mistake--it shows that belief that there was police brutality was a little higher in 1967 than in 1965, but it was actually a little lower.

[Data from the Roper Center for Public Opinion Research]

Tuesday, June 16, 2020

Nothing happened?

A few days ago, Michelle Alexander wrote that "after decades of reform, countless commissions and task forces and millions of dollars poured into 'smart on crime' approaches, the police behave with about as much brutality today as they did in 1966 when a group of young black men . . . created an organization called the Black Panther Party for Self-Defense."  She didn't give any source for this claim, but it reminded me of a New York Times story I had mentioned in a recent post, I had mentioned a New York Times story which said that in New York City, the number of people shot and killed by the police had fallen from 66 in 1972 to eight in 2014.  After a little searching, I located the source of those figures:  the NYPD's annual "Use of Force Report" reports, which started in the 1970s as the "Firearms Discharge Report."  It has annual statistics for the numbers killed and injured by police shootings.  Here they are (on a logarithmic scale):

The decline is pretty steady.  It probably has some connection to the crime rate--the number killed and injured rose during the crack epidemic of the late 1980s--but started in a time when crime was steady or increasing.  It also doesn't seem to differ much by mayoral administration. 

The report also gives the number of police killed and injured by shootings.


Again, there is a downward trend, and there's a strong correlation between the numbers of police killed or injured and civilians killed or injured by police.  This may provide some insight into why it was possible to change things.  Most killings by police involve situations in which someone is armed and posing some kind of threat, so police are at risk too.  Once the department adopted a policy of investigating and reporting on all incidents, it was possible to learn how to keep those situations from developing or getting out of control.

Of course, this is just shootings--it doesn't show whether there were changes in other kinds of police brutality.  Still, it seems that reform is not as futile as Alexander suggests.   I'm not that familiar with crime and justice data, but there are some surveys about perceptions of police treatment which I'll talk about in a future post.