Tag: crime rates

El Paso Homicides Spiked After Border Fence Was Completed; The Fence Didn’t Cause It

At a recent rally in El Paso, Texas, President Trump again claimed that that city had a high crime rate before a fence was built between it and Mexico in 2008 and 2009.  Many people have pointed out that El Paso has long been a more peaceful city than others, before and after the border fence was built.  This post adds just a few more visuals to hammer home the point made by others and an odd anomaly in homicides.

I constructed the figures using local police department crime data from the Uniform Crime Reports (UCR) database, focusing on departments that policed populations of between 500,000 to 1 million.  That population size was appropriate as El Paso’s population was 683,577 in 2017.  The local police departments variable identifies city police departments in the UCR, but it also includes several large urbanized counties in Maryland, Georgia, Virginia, North Carolina, and elsewhere.  In total, 41 different cities and counties met the criteria, including El Paso.  I further relied on the FBI population estimates to calculate the crime rates.  Lastly, I set the base year of 2000 at 100, and compared the crime rates in El Paso over time with the average crime rates across the other 40 other jurisdictions.

These charts convinced me that I did not need to run any regressions nor was the story significantly than that which was already reported, with one exception.  First, the consistent findings.  Figure 1 shows the overall crime rate in El Paso versus the other 40 jurisdictions.  The gray shaded area is when the El Paso border fence was under construction.  Looks like crime continued to decline in El Paso and in the other cities after the wall was built at about the same rate as it declined prior the government’s construction of a border fence there. 

 

Figure 1: All Crime

 

Figure 2 shows the violent crime rates in El Paso relative to the other jurisdictions, with more of a pre-fence dip in violent crime in El Paso relative to other cities, followed with a bit of a rise before construction began.  The construction of the border fence there looks uneventful.  Figure 3 shows property crime rates and it looks even less impressive than the first two figures.

 

Figure 2: Violent Crime

 

Figure 3: Property Crime

 

Figure 4 shows the homicide rate in El Paso versus the 40 comparison cities – and it spiked more than a year after the government constructed the fence between El Paso and Mexico.  The homicide rate in El Paso was 2.8 per 100,000 in 2008 when fence construction began, fell to 1.9 in 2009 when fence construction, fell again to 0.8 in 2010, spiked to 2.4 in 2011, and climbed again to 3.4 in 2012 before coming back down.  The big decline happened right after the fence was built, but the huge spike also occurred when the fence was fully constructed.  Without a lot more econometrics, I’m unable to even provide hypotheses to explain the crash and spike in homicides in El Paso.  Illegal immigration is probably not a factor as the number of apprehensions in El Paso crashed, partly because of the border fence and partly because of the end of mass illegal Mexican immigration.  Homicide rates in San Diego and Tucson, two other border cities, do not show a similar pattern. 

 

Figure 4: Homicide

 

Regardless of the potential explanations for the spike in homicides shortly after the government completed the border fence in El Paso, President Trump’s story is even less true than has been reported by others – at least according to the empirical standards of this public debate.  Please don’t take this post or what I wrote here as arguing that the border fence in El Paso caused the spike in homicides, as I see no evidence to support that claim and I have not carried out nearly enough statistical work on this issue to confidently state that.  

President Trump’s Curious Obsession with Crime

In his Election Day tweet attacking Rep. Mark Sanford, President Trump declared that Sanford’s opponent, Katie Arrington, “is tough on crime and will continue our fight to lower taxes.” Well, maybe. She doesn’t mention either issue on her campaign website. (In fact, she has nothing but bland buzzwords about any issue.)

This tweet is typical. It seems like every time Trump tweets an endorsement or a criticism of a candidate, he calls the candidate “strong (or weak) on crime.” I count 60 Trump tweets since his inauguration that use the word “crime.” Some complain that he is being investigated for a “made up, phony crime” or charge Hillary Clinton with “many crimes.” But most seem to relate to a candidate: Dan Donovan is “strong on Borders & Crime.” Kevin Cramer of North Dakota is “strong on Crime & Borders.” Doug Jones is “WEAK on Crime.” Adam Laxalt is “tough on crime!” “Chuck and Nancy…are weak on Crime.” Ralph Northam is “weak on crime.” Also “VERY weak on crime!” “Keep our country out of the hands of High Tax, High Crime Nancy Pelosi.” And so on.

It’s not obvious that this makes political sense. Candidates aren’t talking much about crime, perhaps because they recognize the substantial decline in crime rates. In numerous Gallup polls over the past year, only 2 to 4 percent of Americans have identified crime as the country’s most important problem. Though about 50 percent of people say they worry a great deal about crime when asked that question directly.

But here’s the thing. Crime in the United States is in fact way down

Here’s a long-term look at the most visible crime, homicide:

U.S. Homicide Rates, 1960-2011 

Here’s a picture of broader crime rates:

U.S. Violent Crime Rate, 1973-2011

And yet, as the same source illustrated, at the very time when crime rates had fallen steadily and substantially for 20 years, 68 percent of Americans said the national crime rate was getting worse. (Crime rates continued to fall after 2011, though there was an uptick in murders in 2015 and 2016. The rate appears to have fallen in 2017.)

Of course, the president is better informed than average Americans. Surely White House staff have explained the crime statistics to President Trump. So why does he talk about “this American carnage” and pound away at the “crime” issue when endorsing candidates who never talk about it? Perhaps it’s part of his continuing use of racially charged language. Perhaps “crime and borders” is just shorthand for the kinds of social change he thinks his voters fear. Or maybe it reflects the fact that he grew up in New York City during a time of sharply rising crime. We all get ideas in our youth (“American cars aren’t well made”) that may stick with us even us as the facts change.

Whatever the reason, it seems curious that he so often cites “strong and crime” as the reason to support political candidates who haven’t talked about crime.

 

287(g) Does Not Fight Crime, But It Does Increase Assaults against Police Officers

Fear of immigrant criminality is driving many changes to domestic immigration enforcement programs during the Trump administration.  One of the earliest such changes was the reactivation of the 287(g) program that allows state or local law enforcement agencies to enforce federal immigration law after entering into a partnership with Immigration and Customs Enforcement (ICE).  The Obama administration substantially scaled back 287(g) after numerous government reports found serious flaws in the program.  Gaston County, North Carolina sheriff Alan Cloninger said his sheriff’s office enrolled in 287(g), “for the protection of the citizens of Gaston County.”  Sheriff Cloninger’s desire to increase public safety is the primary reason, if not the only reason, why 76 local and state level law enforcement agencies across the country have enrolled in 287(g).

Surprisingly, there is little research on whether 287(g) had any effect on crime.  To test whether 287(g) had its intended effect, Cato research associate Andrew Forrester and I investigated whether 287(g) adoption actually lowered crime rates in North Carolina counties where it was established.  From 2003 through 2013, we find no statistically significant relationship between crime rates in counties that adopted 287(g) agreements relative to those that did not in North Carolina.  Importantly, we look at the number of deportations due to 287(g) enforcement by county, which allows us to examine 287(g)’s specific effects.  This means that 287(g) failed to reduce crime in counties where it was activated prior to 2013 when the Obama administration canceled many 287(g) agreements across the country.

In North Carolina, the crime-prevention justification for 287(g) does not hold but neither does the primary critique that it would raise crime rates by reducing citizen cooperation with the police.  It is possible that immigrants in 287(g) counties reported fewer crimes due to fear of immigration enforcement and, thus, an increase in crime would not be recorded in official statistics.  However, some crimes, like murder, are difficult to hide and tend to be reported regardless of local immigration enforcement policies.  To account for this, we further break down the crime rates by the offense and find no relationship between 287(g) and murder or any other individual crime.  Since crime rates did not increase after 287(g) adoption in North Carolina counties, it did not impact trust between local police and the population enough to affect crime rates.   

The only statistically significant relationship that we did find was an increase in the average number of assaults against police officers in 287(g) counties.  We do not know why 287(g) is causally related to the increase in assaults against police officers and we do not know the identities or characteristics of those who committed them.  Besides otherwise law-abiding illegal immigrants who are deported as a result of 287(g) and their American friends, families, consumers, employers, and landlords, police officers in North Carolina also appear to be victims of this program that fails to reduce crime.

Almost 62 percent of the 287(g) agreements currently in effect, 47 out of 76, were signed after President Trump took office.  In the coming months and years, many more state and local law enforcement agencies could also enroll in 287(g) out of the desire to reduce crime.  Charlotte-Mecklenburg Police Chief Kerr Putney recently said:  

The intent [of 287(g)] was to make sure we’re taking felons and gang members, who are violent, out of play … If you apply [the program] specific to those reasons, I think you’d have a totally different outcome.  If you’re asking everybody about their national origin, I think it’s a different application.  And so if it were as it were designed, I think it’s a good tool. I don’t know that it’s being applied that way.

Our research addresses Chief Putney’s concern that 287(g) is not an effective anti-crime tool.  The experience of North Carolina’s counties where 287(g) failed to reduce crime while it increased the number of assaults against police officers should at least be a warning to other counties and police agencies that are considering joining this program: It will not reduce crime.

New Research Finds that Immigrant Crime Is Still Low

Last year we published a blog summarizing the research on how immigrants affect the crime rate in the United States. There are two major types of studies that examine this question.

The first uses Census data of the institutionalized population to investigate immigrant versus native incarceration rates. Although the Census evidence isn’t perfect because of potential issues with reporting immigration status and different types of incarceration, these studies show that immigrants are less likely to be incarcerated than similarly-aged natives.  The second type is a macro-level or area study that looks at the crime rates in places that have experienced large waves of immigration.  These generally find that immigration either lowers or has little effect on crime rates.  The research on unauthorized immigrant crime rates is poor.

A few recent papers recently extended these findings.  The first by David Green seeks to determine whether immigrants affect violent and drug-related crime in the United States on the state-level.  It looks at state-level rates of violent crime and drug arrests pooled for the 2012-2014 years against pooled statistics on foreign-born and Mexican nationals by immigration status, specifically legal versus unauthorized immigrants.  Green finds no association between immigrant population size and increased violent crime.  However, he does find a small but statistically significant association between unauthorized immigrant population size and arrests for drug offenses.

Latest Immigration Reform Bulletin Examines Immigrant Crime Myth

The June issue of Cato’s monthly newsletter on immigration reform, just released, tackles the timely topic of “Immigrants and Crime: Perceptions vs. Reality.” The bulletin finds that, contrary to public perception, immigration has not caused higher crime rates, in Arizona or in the nation as a whole. In fact, one new study even suggests that a rising level of immigration in a city actually leads to lower crime rates.

According to bulletin editor and author Stuart Anderson, a Cato adjunct scholar, “National studies have reached the conclusion that foreign-born (both legal and illegal immigrants) are less likely to commit crimes than the native-born.” It’s an important fact to consider as other states look to copy Arizona’s tough new law against illegal immigration, which was in large part motivated by fears of crime.

The latest bulletin is the third in a series Cato plans to publish through 2010 and into 2011. The May issue analyzed the pluses and minuses of a Senate Democratic proposal to reform U.S. immigration law, and the April issue critiqued efforts to impose a national ID card and the E-Verify system.

You can sign up here to receive the bulletin each month by email.

Topics: