Last year, Christos Makridis and I used homicide data from the Centers for Disease Control to break down the 2020 homicide spike by geography and demographics. With another year’s worth of numbers now finalized—plus “provisional” numbers stretching into 2022—it’s time for a brief update. The CDC’s data, compiled from death certificates, are especially crucial in a year when the FBI completely failed to collect murder data from many of the nation’s police departments.

The good news is that, after spiking in 2020 and rising a little further in 2021, homicides seem to be falling again. The bad news is that this has been an extremely slow process, with recent numbers still well above pre-2020 levels, even if violence remains far from the sky-high levels of the early 1990s.

The CDC puts the national homicide rate at 7.8 per 100,000 for 2021, versus 7.5 for 2020 and 5.8 for 2019. Here are the month-by-month totals since 2018, including provisional data for the first half of 2022:

Monthly homicide deaths


In late 2021, the totals started coming in lower than those from the previous year. Data from Jeff Asher, who compiles more up-to-date murder numbers from big-city police departments, suggest that 2022 saw about 5 percent fewer murders in those cities than 2021.

Meantime, the sociological contours of the problem have changed little. Makridis and I noted in our brief that the spike worsened existing disparities.

To take one of the more dramatic examples, between the 2018–2019 and 2020–2021 periods, the black homicide rate went up by about 40 percent and the white one by 15 percent—already a glaring disparity. But since the black homicide rate started out so much higher than the white one, this translated to an increase of just 0.4 per 100,000 for whites and 9.7 per 100,000 for blacks—nearly 25 times as large. The increase in the black homicide rate was greater than the total homicide rate for the nation as a whole.

Homicide rates by race and ethnicity


Another aspect of our brief that gained attention involved politics: Are homicide rates, as well as the recent increase in them, primarily a Republican or a Democratic problem? Some have tried to push the narrative of a “red-state” murder wave, noting that many of the highest state-level murder rates are found in the GOP South. But as my colleague Rafael Mangual recently explained, states are a poor focus for this kind of analysis. Policing and prosecution mostly occur at the local level, and crime problems in red states are often concentrated in their blue cities.

The extra year’s worth of data allows me to reiterate a few findings from our 2022 brief, this time combining the two-year periods of 2018–2019 and 2020–2021 and looking at counties with populations above 250,000 as of the latter period. The CDC doesn’t release homicide counts below ten, so smaller counties are frequently missing, and combining multiple years shakes loose some extra data by pushing more counties above the ten-homicide mark.

For one thing, overall homicide rates in both periods are positively correlated with the Democratic share of the two-party presidential vote in 2020. The later period is pictured here; the points’ sizes reflect county populations.

Source: Author’s calculations based on data from CDC WONDER and MIT Election Data and Science Lab

How strongly the increase between the two periods is correlated with politics depends on how you measure. Since Democratic counties started with higher homicide rates and the surge generally deepened existing disparities, it’s perhaps unsurprising that Democratic counties saw bigger increases in their raw homicide rates (that is, simply subtracting the 2018–2019 rates from the 2020–2021 ones):

Source: Author’s calculations based on data from CDC WONDER and MIT Election Data and Science Lab

When the changes are presented in percentage increases to homicide rates, the partisan skew is a little less pronounced, if still evident:

Source: Author’s calculations based on data from CDC WONDER and MIT Election Data and Science Lab

Of course, another obvious detail from the above charts is that homicide went up pretty much everywhere, including in GOP strongholds. Data aren’t available for the smallest counties individually, but the homicide rate rose about 25 percent even in places with population clusters of 50,000 or below (what the CDC terms “micropolitan” and “non-core” areas). Further, basic correlations like these don’t take us much beyond the partisan finger-pointing stage.

It’s far trickier to sort out the extent to which actual policies drive these patterns, instead of just constituent demographics. Both matter: Republicans, in addition to being the party of less-restrictive gun laws, are the party of Southern white honor culture; Democrats, in addition to being the party of the police-defunders, are the party of big cities and African-Americans. For those interested, our paper last year had some modeling probing such connections, but it’s an area deserving of more study.

At minimum, though, the new CDC data show that heaping blame on “red states” still seems uncalled for. And the most pressing goal is to bring homicides rates back down—everywhere.

Top Photo by Raul Roa/Getty Images


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next