Chart allegedly proved Democrat-led cities have more violent crime. We crunched the numbers

In the wake of the assassination of the conservative political commentator Charlie Kirk on Sept. 10, 2025, posts claiming that political party membership was a predictor of political violence spread on social media sites like Facebook and X. For example, on Sept. 18, conservative podcaster Tim Pool posted (archived) a chart that he claimed in an X post compared "2023 combined violence rates by party." The text of the post read, in part, "When you get rid of fringe wackos and count crime by party control it looks a lot different."

This is what violence rates look like depending on Political party

How f***ing weird am I rite?

When you get rid of fringe wackos and count crime by party control it looks a lot different pic.twitter.com/HbMUpkavbB

— Tim Pool (@Timcast) September 18, 2025

The chart's y-axis label said it showed the total 2023 violence rate as the "sum of cities" while the x-axis featured two bars labeled "Democrat" and "Republican." Within hours of Pool sharing the chart, Tesla and X CEO Elon Musk reposted (archived) it to his nearly 230 million followers with the caption "Wow."

Pool's YouTube channel and his podcast's website (archived) posted a video (archived) on the same day, Sept. 18, 2025. Around the 24:52 mark of the video, he seems to explain how he created the chart:

Yeah I made a graph earlier where I took the — it's like 13 major cities and then said 'break down all the violent crime incidents in 2023 — which is, usually you have the more accurate data — by political party control and preference' and, guess what: it's 95% Democrat.

Pool did not name a source for the data or specify which set of 13 major cities he included, or why he chose them. We reached out to Pool for clarification about his methodology and the accuracy of his chart and will update the story if we receive a response. During a discussion of the chart (archived) in a Nov. 18 episode of his podcast, Pool said he "asked ChatGPT" to generate it for him.

Our own investigation of the available data found that Pool's chart was misleading and significantly overstated a real, but less extreme, difference in crime rates between cities with Democratic mayors and those with Republican mayors. Pool appeared to suggest his chart was evidence that the higher violent crime rates in Democratic-led cities were a direct result of the political parties of their mayors — a causal connection that social scientists have investigated and found no evidence for. Rather, experts believe the difference in violent crime rates can be attributed to other factors like economic inequalities and city-level policies.

Breaking down the data

A dataset is a collection of data taken from one place or intended for one project. We were unable to find any datasets that directly compared cities' violent crime incidents in 2023 by political party control. It was also unclear exactly what Pool meant by "political party control," as people in political positions from the local to national level each play a part in "controlling" a city. Because the chart focused on cities, we took the phrase to mean that Pool was referring to the political affiliation of the cities' mayors.

In an attempt to reverse-engineer the methods behind Pool's chart, we collected population and violent crime data from the FBI's 2023 Crime in the U.S. (CIUS) reports, which we cross-referenced with the reported political party of the cities' mayors at the time. The FBI's violent crime category tracks "murder and nonnegligent manslaughter, forcible rape, robbery, and aggravated assault."

Why did we pick this dataset? The FBI is one of the only organizations in the U.S. that officially documents crime statistics on a large scale, and its database is considered the "most complete estimate of national statistics available, incorporating data from more than 16,000 local police agencies covering roughly 94 percent of the population," according to the Brennan Center for Justice at NYU Law, a nonpartisan law and policy institute.

For this reason, social scientists who study urban crime rates frequently rely on CIUS reports for their data. For example, the database underpins many of the academic studies on crime rates we'll discuss later in this story.

However, the FBI's data is not completely infallible. The FBI has retroactively revised data before, and a 2025 article from the Council on Criminal Justice explains how the bureau's crime data sometimes conflicts with the other major U.S. crime database, the National Crime Victimization Survey. Uniform Crime Reports (UCR), the FBI's crime database, is only tracked based on crime reported to the FBI by law enforcement. This introduces two issues: The dataset does not include unreported crimes, and some law enforcement agencies might report crimes and/or types of crimes differently than others.

It's worth noting that the CIUS report covered metropolitan statistical areas (MSAs), wider areas of geographical study consisting of cities and their surrounding counties based on economic interaction. The dataset provided wider MSA numbers and more specific city-based numbers when possible. The three original charts we discuss in the following section use mostly city numbers, though in some cases only the broader MSA numbers were available, following the FBI's organization of the data. We also used the FBI's Crime Data Explorer to find the 2023 violent crime instances in Los Angeles and Chicago. This data was absent from the 2023 UCR document.

Visualizing the data

We used the FBI violent crime data to produce three different charts. The first, an attempt to follow Pool's description of his methodology as closely as possible, compares the sum of violent crime rates by mayoral party of the 13 U.S. cities with the highest reported instances of violent crime. (We picked the 13 cities with the highest violent crime instances because none of the top 13 cities by violent crime rate had a Republican mayor in 2023.)

To produce this chart, we added up the violent crime rates for the 13 U.S. cities with the highest violent crime instances in 2023 by mayoral political affiliation. (Violent crime rates are calculated by dividing the number of violent crime instances by the city's population, and then multiplying the result by 100,000, resulting in a rate of violent crime per 100,000 residents.) Unlike Pool's chart, which included only "Democrat" and "Republican" categories, we also included an "Independent" category for cities with mayors unaffiliated with either major party.

(We also created a version of this chart broken down by city.)

Although the numbers themselves do not exactly match those in Pool's chart, the general proportions are similar. The bar representing cities with Democratic mayors is significantly higher than the one for cities with Republican mayors, as in Pool's chart.

However, both the above chart and Pool's are misleading.

Using the sum of cities' violent crime rates means we added the crime rates of each city individually without averaging them. In other words, the above chart compares the total sum of crime rates from a larger set of Democrat-run cities to the crime rates of a tiny set of Republican-led cities. (Our data compares 11 Democratic cities to a single Republican city, though we can't say for sure if Pool's chart does the same.)

The sum of violent crime rates from the large set is disproportionately larger than the sum of rates from the tiny set, and this comparison cannot be used to draw any meaningful conclusions. Because there are simply more cities with Democratic mayors, adding up their violent crime rates without averaging them and graphing the result naturally produces a higher bar for Democratic-run cities.

In a series of interviews, Thomas D. Stucky, a professor of criminal justice at Indiana University Indianapolis who has conducted research on a mayor's impact on violent crime, said handling the data this way is poor statistics. He likened using the "sum of cities" to adding up the average temperatures of 12 cities and comparing the result to the average temperature of a single city. In a chart made this way, the collective average temperatures of the 12 cities would naturally be "hundreds of degrees compared to the average of the one city," which says nothing about the actual temperature differences. According to Stucky, compiling data in this manner "gets nonsensical in a hurry."

So what could be a more accurate way to represent the data? Calculating the average violent crime rates across cities, instead of adding the rates together is a good start. Our second chart, embedded below, compares the average violent crime rate by party of the mayor running the same 13 cities as in the first chart.

To get the numbers for this chart, we added together the violent crime rates in cities with Democratic mayors, then divided the result by the number of those cities. We did this for each party, then compared the results. Although using averages makes more mathematical sense, this visualization still presents a flawed reflection of the data. Our dataset of 13 cities — 11 with Democratic mayors, one with a Republican mayor, and one with an independent mayor — is much too limited to say anything about urban violent crime trends in general.

The limits of what the data can say about the cause of higher violent crime rates are clearest when we look at the proportions in the chart: the heights of the bars appear to show that cities run by independent mayors have more violent crime on average than cities with Democratic and Republican mayors combined. But the bar for independent mayors represents the notoriously high violent crime rate of just a single city, Detroit. Even though the chart seems to tell us that cities with independent mayors have the most crime, the data from a single outlier city is not significant enough to draw any conclusions.

So, we made a third chart using a larger dataset of the 50 U.S. cities with the highest crime instances, again examining the average violent crime rate by party of the mayor in office in 2023.

(We also created a table breaking down the crime rates of these 50 cities.)

This chart is more representative of the actual differences between U.S. cities by the party of their mayors. But comparing 38 Democrat-run cities to eight Republican-led cities and four independent-led cities still does not yield information concrete enough to draw conclusions about what causes the difference.

The danger of creating charts like these is that they appear to say much more than they actually do. Stucky explained that the "comparison of raw averages doesn't say anything about all of the factors that could also affect crime besides political party," noting that a "serious" attempt to measure whether a mayor's political party affected violent crime rates would need to account for other possible factors, such as  "unemployment, social safety net, [and] activities of police."

Without knowing the sources Pool used, there is no way to confirm that he used the same dataset we used here — or that his numbers were accurate at all. Another issue is the fact that Pool did not explain why he chose the cities he used to make the charts, or which cities those were, beyond describing them as "major." 

The reason for the size of his dataset, 13 cities, is also unexplained. The addition or subtraction of even a single city would have a large influence on the shape of the chart. The U.S. has over 300 cities with more than 100,000 residents. Pool chose a set of less than 5% of U.S. cities to represent all of them. 

As Stucky told us: Pool's chart could be "mathematically true and at the same time highly misleading."

History behind the debate

The claim that cities with Democratic mayors have more crime spread long before Kirk's assassination. For example, U.S. President Donald Trump has used it as a talking point for his presidential campaigns since 2016, and cited it as justification for his decision to deploy the National Guard to multiple U.S. cities in 2025. Other politicians have used this claim in similar ways.

A milestone in the years-long public debate over whether the party affiliation of elected officials is causally linked with violent crime was the 2022 publication of two competing reports from think tanks. Pointing out that state-level data shows red states have more homicides (just one type of violent crime), Third Way, a non-profit and think tank that describes itself on its website as "moderate," published a report on the issue in 2022, titled, "The Red State Murder Problem," that claimed in part:

The rate of murders in the US has gone up at an alarming rate. But, despite a media narrative to the contrary, this is a problem that afflicts Republican-run cities and states as much or more than the Democratic bastions.

In response, the Heritage Foundation, a conservative research and educational institution, released a report in the same year, titled, "The Blue City Murder Problem," allegedly debunking the concept of the Red State Murder Problem and arguing in part that Democrat-led cities have more violent crime as a direct result of their elected officials' political party.

Kessler and Murdock [authors of "The Red State Murder Problem"] did their level best as political operatives to blame their political opponents for the very thing—rising crime—that leftist policies at the city and county levels have caused.

Both reports used crime data to establish a causal link between crime and the political party of different types of leaders, including mayors. We cannot concretely prove or disprove these reports as a whole — these issues are far too nuanced for a simple answer. However, many experts believe a mayor's party in particular is not a significant factor behind the differences in crime rates.

Do cities with Democratic mayors have more violent crime? What experts say

Short answer: yes, according to the available data. But, as our charts above showed, approaching the data in different ways produces different pictures of how much of a discrepancy exists. Even more importantly, the data does not prove that the political affiliation of mayors is the reason for the difference in numbers.

For one, correlation does not imply causation. 

In other words, a city's mayor's party and the city's violent crime rate may have some relationship, but that doesn't mean mayors' political party membership directly causes the difference in numbers.

Multiple reputable studies examining violent crime rates in U.S. cities have found no evidence that the political party of a mayor has any significant causal effect on a city's violent crime rate.

For example, in 2025, the peer-reviewed journal Science published a study, "The partisanship of mayors has no detectable effect on police spending, police employment, crime, or arrests," that found no significant link between the political party of a city's mayor and crime, concluding, "We find that the estimated impacts of partisanship on overall numbers of violent, property, drug, and other crime arrests are close to zero and are all statistically insignificant." 

The same study did not rule out the possibility that mayoral partisanship could have an effect on policy, which could "[lead] to changes in the way that police forces act in the conduct of their jobs," though their findings here were also not very statistically significant.

In an accompanying The Conversation article, "Crime is nonpartisan and the blame game on crime in cities is wrong – on both sides," two authors of the Science study discussed how members of both political parties misused crime statistics to further their rhetoric. 

A 2024 article from the Manhattan Institute, a conservative public policy think tank, also broke down different ways people have interpreted crime statistics in trying to assign blame to one party, concluding that "data can say whatever a researcher wants them to say." One of the article's authors, Robert VerBruggen, told us via email that "the raw numbers tended to make one side or the other look bad." Instead, he explained that differences in homicide rates at a state and county level "were largely explained by demographic and economic differences, not by party per se."

Stucky argued that some characteristics of a city's mayor might actually have an impact on violent crime in his 2003 paper "Local Politics and Violent Crime in U.S. Cities," which appeared in the peer-reviewed journal Criminology. For example, the paper found that a mayor's race could play a part in violent crime rates — the data suggested that cities with Black mayors tended to have lower violent crime rates.

Stucky's paper did not account for mayors' political party, but he explained to us that "the little research that has been done on the subject does not provide any consistent evidence across multiple studies that [a mayor's political affiliation] affects crime, once one controls for other factors associated with crime." Even if a causal link was found, it would likely be small in scale compared to other factors, he said. VerBruggen also echoed this claim.

There is likely more violent crime in cities run by Democratic mayors, in part, simply, because there are many more major cities run by Democratic mayors. Between 2016 and 2025, the top 100 cities of the U.S. by population were on average 64% Democrat-led and 29% Republican-led. According to a 2014 Pew Research study, liberals are nearly twice as likely as conservatives to live in urban areas, resulting in more votes for Democratic mayoral candidates.

VerBruggen explained that most crimes would occur in Democrat-run cities, even if the "risk of any given resident being victimized" was the same, as long as Democrat-run cities made up the majority of the dataset. For example, say we're comparing 10 Democrat-led cities with a combined total of 1 million residents, and one Republican-led city with 100,000 residents. Even if the crime rates were the same between all the cities, the Democrat-led cities would have 10 times more crime than the Republican-led city.

The actual cause of violent crime in U.S. cities remains debated. Some studies have found it's due to economic inequality linked to a loss of traditional family structure and lack of education, for example.

VerBruggen believes that it is not a mayor's party that makes a difference, but a city's policies. He cited policing levels and even street lighting as examples of things that do affect crime rates, supported by multiple recent studies. "Rather than polarizing the issue, I think it's most helpful to study which approaches reduce crime and encourage policymakers of all parties to adopt them," he said.