Houston Says Crime Is Down. But Are Houstonians Simply Calling the Police Less?

Houston officials have good reason to be pleased that reported crime is falling. Mayor John Whitmire has repeatedly pointed to lower crime numbers as evidence that the city is moving in the right direction. In March, his office cited Major Cities Chiefs Association statistics and declared that “crime is down in Houston.” (City of Houston)

But there is a problem with those three words.

The Houston Police Department itself describes its public crime database more precisely. These are offenses “for which HPD wrote police reports.” (City of Houston)

That is not necessarily the same thing as every crime that occurred.

And in Houston in 2026, there is a growing reason to ask whether the difference is becoming important.

The man who did exactly what police wanted

On August 9, Juan Francisco Monroy Turuy heard gunshots in a Houston Home Depot parking lot. He called 911, stayed at the scene, talked to HPD officers, and helped identify the man police subsequently arrested.

Then ICE arrested Monroy.

HPD says it did not summon ICE. Federal agents happened to be there and detained him while he was cooperating with the police investigation. Harris County District Attorney Sean Teare warned that incidents like this can make victims and witnesses afraid to call police. (ABC13 Houston)

Teare says he is already seeing something in Houston that deserves attention.

“We’ve seen calls for service for domestic violence, for aggravated robbery, for a lot of the really dangerous crimes out there fall off a cliff.” (KPRC)

That is not an academic theory. That is the Harris County district attorney describing what his office is seeing.

And Univision found something even more revealing when it interviewed another immigrant worker near the place where Monroy was detained.

“Aunque vea que maten yo mejor no digo nada” — even if I see someone killed, I would rather say nothing. The worker said he was now afraid of dealing with police. (Univision)

There is the problem in one sentence.

The person being beaten, robbed, or shot does not have to be undocumented.

The victim could be an American citizen.

If the undocumented man across the parking lot sees the assault and decides that calling 911 could bring unwanted attention to himself, his wife, his brother, or his coworkers, he may simply walk away.

The American citizen still got assaulted.

The difference is that nobody called the police.

HPD has already documented the connection

Between April and June, HPD recorded 103 contacts with ICE, with 19 people ultimately transferred to immigration custody. These numbers came from HPD’s required quarterly transparency report, not from a leaked document. (Telemundo Houston)

ABC13 found an even more relevant detail: 34 of those immigration contacts began with calls for police service. They included calls involving assaults, sexual assaults, thefts, suspicious persons, and welfare checks. Six ultimately resulted in someone being transferred to ICE, although HPD’s data do not establish whether the person transferred was the original caller, a victim, a suspect, or someone else encountered during the investigation. (ABC13 Houston)

Legally, those distinctions matter.

On the street, they may matter much less.

The story people remember is simpler:

Somebody called the police. ICE became involved.

So I went looking in HPD’s crime numbers

I downloaded Houston’s 2025 and 2026 NIBRS offense files. They identify the offense, occurrence date, ZIP code, and type of premise where HPD wrote the report.

The 2026 data currently cover January through June. That is important: we have only half a year. Nobody knows yet what July through December will produce.

I concentrated on ZIP code 77081 in the Gulfton area and 77011, 77012, and 77023 in the East End.

This was not an arbitrary choice. The City of Houston describes Gulfton as home to many recent immigrants from Mexico and Latin America. City demographic data show Spanish is spoken at home by nearly 30,000 of roughly 41,500 Gulfton residents age five and older. Magnolia Park, in the East End, is described by the city as one of Houston’s oldest Hispanic neighborhoods. (City of Houston)

Then I tightened the test.

I removed stores, supermarkets, restaurants, offices, hotels, gas stations, and other commercial premises. I also excluded shoplifting.

Businesses have insurance claims, loss-prevention procedures and financial reasons to report crime. A supermarket is unlikely to decide not to call HPD because somebody in the manager’s family might encounter ICE.

Instead, I looked only at offenses occurring at residences and on streets — places where reporting depends much more heavily on ordinary victims and witnesses.

Here is what HPD’s data show for the first six months of 2025 compared with the first six months of 2026:

That is not proof that people stopped reporting crimes.

But neither is it nothing.

Reported aggravated assaults in these heavily Hispanic ZIP codes fell almost twice as fast as Houston overall.

The difference becomes larger when the analysis is limited to aggravated assaults occurring inside residences.

There were 211 residential aggravated assaults reported in these four ZIP codes during the first six months of 2025.

There were 136 during the same six months of 2026.

That is a 35.5 percent decline.

Houston citywide residential aggravated assault fell about 17.7 percent during the same period.

Again, perhaps there really were dramatically fewer assaults in Gulfton and the East End.

That would be good news.

But HPD’s numbers by themselves cannot tell us that.

They tell us how many aggravated assaults resulted in HPD reports.

The East End comparison is particularly curious

There is another way to test whether we are simply seeing normal seasonal movement.

During January and February 2025, the three East End ZIP codes recorded 37 residential aggravated assaults. During April and May, the number rose to 43.

The same comparison in 2026 went in exactly the opposite direction.

January and February: 46.

April and May: 25.

A year earlier, residential aggravated assault increased as spring arrived.

This year, it fell 46 percent.

That does not prove ICE caused the decline.

But it makes “that’s just the normal seasonal pattern” a considerably less satisfying explanation.

And then there are the stolen cars

Motor-vehicle theft provides an interesting comparison.

It is not a scientifically perfect control. Some stolen vehicles certainly go unreported.

But a person whose car disappears has unusually strong reasons to call the police regardless of immigration fears. There may be an insurance claim. The vehicle may be financed. The owner wants it recovered. And the owner certainly does not want responsibility if someone later uses the stolen vehicle in a crime.

In the four ZIP codes we examined, reported motor-vehicle thefts at residences and streets did not fall.

They rose from 80 to 91 — up nearly 14 percent.

Meanwhile, residential and street aggravated assaults fell 31 percent.

That divergence is worth asking about.

Houston’s Hispanic audience is hearing this story repeatedly

There is another factor that citywide numbers cannot show.

Houston does not have a single common news audience.

English-language television stations are covering ICE. ABC13, KPRC, and Fox 26 have all reported the detention of Monroy and the potential effect on crime reporting. So this is not a story being hidden from English-speaking Houston. (ABC13 Houston)

But immigration enforcement is an especially prominent and continuing local story on Houston’s Spanish-language television stations.

Univision 45 currently maintains a standing “Inmigración en Houston 2026” section. Its recent coverage includes HPD’s 103 contacts with ICE, the detention of the man who called 911, the detention and release of witnesses in the Lorenzo Salgado case, immigration arrests following traffic stops, what to do if ICE comes to your door, whether people may record ICE operations, and what to say when an agent asks where you were born. (Univision)

Telemundo Houston has repeatedly reported the same local enforcement stories: Monroy’s detention after calling 911, HPD’s 103 ICE contacts, the detention of witnesses following the fatal Lorenzo Salgado operation, and the continuing investigation of that shooting. Mayor Whitmire himself gave Telemundo an interview in July in which he described aspects of ICE enforcement as “out of control.” (Telemundo Houston)

That is not criticism of Univision or Telemundo.

They are doing their job.

Immigration enforcement directly affects their audience, so they report it heavily and explain what viewers need to know.

But it means a Spanish-speaking family in Gulfton or the East End may encounter these stories repeatedly — not as an occasional Washington immigration story, but as local news about people being detained on Houston streets, at traffic stops, after speaking with police and during neighborhood ICE operations.

Fear does not require ICE to arrest everybody who calls 911.

Fear requires people to believe that contact with police might carry a risk.

And Houston’s Spanish-language audience has now seen enough actual local cases to know that the possibility is not imaginary.

Researchers have already found this effect elsewhere

There is also substantial evidence that immigration enforcement can change crime reporting.

An August 2026 study in the American Economic Review examined Secure Communities, the federal program that increased cooperation between local law enforcement and immigration authorities.

Researchers found that increased deportation fears reduced the likelihood that Hispanic victims reported crimes to police. They also found that reduced reporting encouraged more offending against Hispanics: their estimate was that a 10 percent reduction in reporting produced a 7.9 percent increase in offending. (AEA Web)

A separate study using Dallas Police Department records found the reverse effect. When federal immigration enforcement became more narrowly targeted, serious violent and property crimes reported by Hispanic complainants increased about 4 percent. The researcher found no corresponding evidence that Hispanic criminal activity or population growth explained the increase. (ScienceDirect)

In other words, this is not merely someone imagining how frightened immigrants might behave.

Researchers have measured it.

So is Houston crime really down?

Almost certainly, at least some of it is.

Homicides have declined. Houston is part of a broader national decline in killings. There is no justification for pretending that every reduction appearing in the police statistics is fictional.

Our own numbers don’t support that conclusion either. Some categories declined citywide and in Hispanic neighborhoods, while others behaved differently.

But Houston officials are claiming something the database cannot actually prove.

HPD’s database does not count every crime that occurred.

It counts crimes for which HPD wrote police reports.

And now Houston has a district attorney saying calls involving serious crimes have “fallen off a cliff,” a 911 caller taken into ICE custody, 34 calls for police service that resulted in HPD contacts with immigration authorities, and a worker telling Univision that even if he saw somebody killed, he would rather say nothing.

Then we looked at the numbers.

In Gulfton and the East End, reported residential aggravated assault has fallen 35.5 percent from the first half of 2025 to the first half of 2026 — roughly twice the citywide decline.

At the same time, reported motor-vehicle theft in those same neighborhoods increased.

Maybe the neighborhoods suddenly became much safer.

Maybe fewer people are calling.

Most likely, some combination of both is occurring.

We cannot tell from HPD’s crime reports alone.

And that is why City Hall should be more precise when celebrating Houston’s numbers.

For now, there is one statement the evidence unquestionably supports:

Reported crime is down.

Whether actual crime has declined by exactly the same amount is a very different question.

And Houston may be conducting, in real time, an experiment that tells us why.