For better or worse

 A few years ago, I said that beliefs about changes in crime rates were affected both by actual conditions and partisanship, but that the effect of partisanship seemed to be bigger for Republicans.  This post will look at the issue in more depth.  

The Gallup Poll has two questions about change in crime rates; "is there more crime in the U.S. than there was a year ago, or less?" and "is there more crime in your area than there was a year ago, or less?"  Both of them have been asked almost annually since 1989, usually in October.  The percent who think that there is more crime in the nation, by party ID:

The effect of partisanship is clear:  for example, between 2024 and 2025, Republicans became less likely to think that crime was increasing while Democrats became more likely.  Judging the effect of actual crime rates is more difficult, because "crime" can include a lot of things, most of which aren't well measured, but here are homicide rates:

Here's the percent who think there's more crime in their area, by partisanship:

If you regress beliefs on party of the president and homicide rates:

                             Constant          Rate       D
Reps, nation:           32                 4.7          16
Dems, nation:          38                 5.0         -9.2

D is one if a Democrat is president, 0 if a Republican is.  So if a Democrat is president, Republicans are more likely to say that the crime rate is increasing and Democrats are less likely.  The estimated size of the partisan effect is larger for Republicans, although the difference is not statistically significant.

                             Constant       Rate          D
Reps,  area:             25                 1.9          18
Dems, area:             28                 2.4        -3.5

It's the same general pattern, but the difference in partisan effects is larger (and statistically significant).   It might seem odd that the effect of partisanship on beliefs about local conditions would be as strong as the effect on beliefs about national conditions, but even getting a clear view of "your area" goes beyond personal experience.  

The effect of partisanship seems to be growing.  The average gaps (Republican minus Democratic percnet) by administration:

                         Area       Nation
GHWB              -7.3         -1
Clinton                3.5          3.8
GWB                  -9.5        -14.8
Obama                10.3         15.4
Trump-1              -3           -2.5
Biden                  34.8         39
Trump-2                9            8

There's also a more subtle point:  in the two Bush administrations, Republicans were less likely to think that crime was increasing than Democrats were, but if you average the two Trump administrations, there's essentially no difference.  That doesn't mean the effect of partisanship has disappeared--look at the Biden administration--but that Republicans have become more negative in general.  Although the 2025 homicide rate was the lowest of the whole period, Republicans were more likely to see crime as increasing than they were in 1998-2004.  This fits with something Paul Krugman mentioned the other day:  that Trump seems reluctant to talk about the decline in crime:  although he talks about the decline in Washington DC, he suggests that things are terrible in the "blue cities."  But it's noteworthy that the increasingly negative mood among Republicans seems to apply to perceptions of your area as well.  I wonder if it's a general rule that it's easier to influence your followers in a negative than in a positive direction.  That seems plausible, although I don't know of any research on the subject.  

Note:  the regression estimates are  OLS, but regressions allowing for serial correlation in the residuals produced similar results.  

[Data from Gallup and the Roper Center for Public Opinion Research]

I don't need no doctor

 Between March 2020 and December 2022, several polls asked people how much trust they had in Dr Anthony Fauci:  a great deal, a fair amount, not much, or not at all.  Most of the questions mentioned Covid--"trust him to provide reliable information on coronavirus." The means by party identification, counting "great deal" as 4 .... "not at all" as 1.

In the first survey (3/21-34/2020) the average level of trust was almost identical among Democrats, Republicans, and independents.  Democratic trust seemed to increase a little in 2020 and then declined, but the changes were small.  Republican trust dropped sharply in the second survey (late August/early September 2020), and then continued to declined at a slower rate.  Independents were just about halfway in between.  The gap between average Democratic and Republican trust:

There's very little change between June 2021 and December 2022.  That is, the partisan polarization of trust took place between March 2020 and June 2021--nothing that happened in the next 18 months made much difference.  It's also a high degree of polarization:  in June 2021, 61% of Democrats said they had a great deal of trust in Fauci and only 5% said they had none; among Republicans, 11% said they had a great deal and 50% said not at all.  A few days ago, Bret Stephens wrote "I was prepared to cut Fauci and others in government a lot of slack in the first months of the pandemic, when the take-no-chances approach seemed like the safest course. But at some point many of the measures that he insisted were necessary .... felt both onerous and performative, and guaranteed to create a backlash."  Whether or not that was true for him, it doesn't seem to have been true for most Republicans--they shifted against him right away rather than waiting.  

There were also several questions about trust in national public health officials or the CDC.  They followed the same basic pattern:  no partisan differences at the beginning and then a split, but the eventual partisan gap was smaller than with Fauci.  Finally, there were some questions about trust in Donald Trump to provide reliable information about Covid.  Of course, there was a partisan split on Trump from the beginning.  But did it also get bigger, or had it already reached its ceiling?  I'll look at that in a future post.

[Data from the Roper Center for Public Opinion Research]