As an urban trauma center, our hospital’s ER sees its share of murder victims every year. Murder happens for all sorts of reasons but one thing all murders have in common is eventually, someone will politicize them. If a murder happens in a Republican stronghold, then Democrats will blame lax gun laws and far-right hate groups. If a murder happens in a Democrat stronghold, then Republicans will blame lax crime laws and undocumented immigrants or the homeless. The finger pointing is inevitable. But what is the truth? Does how a city or a state votes in political elections affect its homicide rate?
To answer this question, I looked at whether a state or major U.S. city voted Republican or Democrat in the 2024 presidential election. For city data, I used the result of how the county that the city lies in voted since county voting information is more readily available than city voting information. There are two Federal agencies that track U.S. homicides: the FBI and the CDC. The FBI’s database is derived from reports from local law enforcement agencies but this reporting is not mandatory so many homicides do not get reported to the FBI. The CDC’s database is derived from death certificates that are filled out by coroners. Because coroner reporting of homicides is mandatory, the CDC data is more complete and will show a higher homicide rate than the FBI for any given state or city.
States that voted Republican have a higher homicide rate than states that voted Democrat
Homicides are best reported by their rate per 100,000 population instead of their total number. New York City with its population of 8.8 million has more people than the entire states of Rhode Island, Montana, Delaware, South Dakota, North Dakota, Alaska, Vermont, and Wyoming combined. Therefore, it is to be expected that there would be far more homicides in New York City than in, for example, Montana, simply based on the sizes of the populations.
There is tremendous variation in homicide rates between states ranging from a low of 1.8 per 100,000 population in New Hampshire to a high of 20.7 per 100,000 in Mississippi. In other words, an average person is about 12 times more likely to be murdered in Mississippi than in New Hampshire. The map below shows 2022 homicide rates by state with the darker the shade, the higher the homicide rate.
This map shows that homicide rates are the highest in southern Mississippi Valley states plus New Mexico. Homicide rates are lowest in New England states, upper Midwest states, and Mountain West states. We. But are homicide rates higher in states that voted Republican in the 2024 presidential election or in states that voted Democrat? The graph below shows the homicide rate per 100,000 population in all 50 states. States that voted Republican are in red and states that voted Democrat are in blue.
The average homicide rate for states that voted Republican was 8.5 per 100,000 and the average homicide rate for states that voted Democrat was 5.6 per 100,000. This is a statistically significant difference with a P-value of 0.018 by unpaired T-test. In other words, a person is 50% more likely to to be murdered in a state that voted Republican than a state that voted Democrat.
Most U.S. cities voted Democrat
There is a striking difference between rural America and urban America when it comes to political preference. This was evident to anyone driving across the United States last October – inside city limits, yard signs were mostly for Democrat candidates and outside city limits, yard signs were overwhelmingly for Republican candidates. The Pew Research Center found that in 2023, 60% of urban counties vote Democrat and 60% of rural counties vote Republican; suburban counties are evenly split between Republican and Democrat. Researchers at the Carsey School of Public Policy found that this trend also held in the 2024 presidential election with 61% of large core metropolitan areas voting Democrat and only 26% of non-metropolitan rural areas voting Democrat. The graph below shows the 50 most populous cities in the U.S. in 2023 based on data from the U.S. Census Bureau.
Of these 50 cities, 38 voted Democrat (blue bars) and 12 voted Republican (red bars) in the 2024 presidential election. However, the total population of all 12 Republican cities combined (7,715,274) is lower than the population of the nation’s most populous city, New York City (8,804,199). Note that for all of these cities, their metropolitan area population is much larger than the city population.
There is not a relation between the size of cities and the homicide rate of cities. The graph below shows the homicide rate per 100,000 population for the same 50 largest U.S. cities by population in 2022 based on CDC data. New York City with the largest population has approximately the same homicide rate as Arlington, Texas which has the smallest population.
Notably, many of the nation’s largest cities have homicide rates below the nation average, including New York City, Los Angeles, San Diego, Austin, San Jose, and San Francisco.
Cities that voted Democrat have a higher homicide rate than cities that voted Republican (but it is not statistically significant)
There is no statistically significant association between a city’s political preference and its homicide rate, although there is a trend for cities voting Democrat to have higher homicide rates than cities voting Republican. The graph below shows the same information from the previous graph for the 50 largest cities in the U.S., but instead of being listed left to right by population, the cities are listed left to right by homicide rate. The average homicide rate for cities that voted Democrat in the 2024 presidential election was 13.3 per 100,000. The average homicide rate for cities voting Republican was lower at 9.2 per 100,000; however, the difference between Democrat (blue bars) and Republican (red bars) cities was not statistically significant (P = 0.17).
Another way of examining homicide in the nation’s cities is to look at the cities with the highest homicide rates, regardless of population. This includes many cities that are not in the top 50 cities by population. For example, the U.S. cities with the highest homicide rates in 2022 were New Orleans (58.4 per 100,000) and St. Louis (57.2 per 100,000). However, New Orleans is only the 52rd most populous city in the United States and St. Louis is only the 76th most populous city.
Of these 50 cities with the highest homicide rates, 44 voted Democrat (blue bars) and only 6 voted Republican (red bars) in the 2024 presidential election. Indeed, the top 18 cities with the highest homicide rates all voted Democrat.
Many large cities have lower homicide rates than entire states
New York City is often accused of having a lot of murders. And it is true that because of its population of more than 8 million, a lot of murders do occur in New York City. However, when expressed as the homicide rate per 100,000 population, New York City has a lower homicide rate than 33 of the states in the U.S. In other words, a person is more likely to be murdered in the state of Texas (7.6 homicides per 100,000) than in New York City (5.3 homicides per 100,000). Similarly, in Florida, a person is more likely to be murdered outside the city limits of Tampa (7.2 homicides per 100,000) than inside the city limits (6.8 homicides per 100,000).
Income matters
There is an association between income and homicide rates, both for states and for cities. However, the association is weak and there are plenty of exceptions. The graph below shows homicide rates (blue bars) versus median individual income in 2022 (red line) for the 50 states. The dotted line shows that there is a trend for states with lower incomes to have higher homicide rates but there are exceptions. For example, Maryland has a very high median income but also has a high homicide rate. On the other hand, Maine and Nebraska have relatively low median incomes but also have low homicide rates.
There is a similar trend for cities. The graph below shows city homicide rates in the blue bars and median individual income in the red line. Cities with the lowest median incomes tend to have the highest homicide rates. But once again, there are exceptions – for example, Washington DC has a very high median income but also has a high homicide rate. Given the proximity of Maryland and Washington DC, it is likely that the same factors accounting for the income/homicide rate discrepancy in Maryland also account for that some discrepancy in Washington DC. On the other hand, Grand Rapids, Michigan and Riverside, California both have relatively low median incomes but also have low homicide rates.
What can we as physicians do?
A few years ago, I was interviewing a medical student who was applying to our internal medicine residency program. He was a foreign medical graduate, having trained in the United Kingdom. He said he wanted to come to the U.S. because of our exceptional residency training programs but was afraid to because he was worried that he would be shot and killed. His impression was that everyone in the U.S. had a gun and most Americans carried a gun with them at all times. His fears are not entirely unfounded – the murder rate in the U.S. is 8-times higher than the murder rate in the U.K. And the homicide by firearm rate is 340-times higher in the U.K. than in the U.S. I tried to assure him that most Americans do not own guns and that our streets are actually reasonably safe but he ultimately decided to stay in London for his residency.
As physicians, we should advocate to reduce all preventable deaths. And homicide is a leading cause of preventable death in the U.S. For Americans between the ages of 15 -24, it is the 3rd most common cause of death and for those between the ages of 25-44, it is the 5th most common cause of death. There are striking racial/ethnic differences in homicide. Homicide is the 21st most common cause of death for Asian Americans of all ages, the 20th most common cause of death for White Americans, the 13th most common cause of death for Native Americans, the 12th most common cause of death for Hispanic Americans, and the 7th most common cause of death for Black Americans. For Black Americans between ages 15 – 24, homicide is the number one cause of death. Black Americans account for 14% of the U.S. population but account for 58% of all American homicide deaths.
For these reasons, all physicians should be concerned about our homicide rates. We need to advocate for protection of victims of domestic violence. Because most U.S. homicides are carried out using guns, we need to advocate for common-sense firearm legislation, such as mandatory background checks for all gun sales. We need to advocate for gun safety measures such as firearm training and use of gun safes. And we need to ask about guns in the homes for our patients with mental health disorders or family members with mental health disorders. Lastly, we need to ensure that our hospitals and medical clinics have policies and procedures in place to protect healthcare workers and patients from harm.
There are dozens of reasons why people kill other people. The Brookings Institute identified three correlates with high homicide rates in different communities: poverty, teenage boys not in school, and young men out of work. We need to hold our community leaders responsible for addressing these root causes of high homicide rates including measures to keep teens in school and keep young men employed.
Homicide is not a disease but it results in death, no different than cancer or myocardial infarctions. In this sense, reducing homicide rates should be a priority for all physicians. The nation’s high homicide rate is not a Democrat problem or a Republican problem. It is an American problem.
January 3, 2025
