By Matthew J. Durán
In 2023, more than 700,000 people were arrested in the United States for driving under the influence. Even though this seems high, we know it only represents the people who were caught — the “tip of the iceberg.”
A question people often ask is how much of the “iceberg” are we seeing. More specifically, how many people, in total, drive under the influence in the United States?
We have an answer to that question.
According to a national survey that same year, about 15.6 million Americans admitted they drove under the influence at least once in the past 12 months. If one were to use this number as a benchmark for the population of DWI offenders, that means only about 5% were arrested by law enforcement.
Turning now to the city of El Paso, we could ask those same questions: During the last year, how many El Pasoans drove drunk? Is the total number of arrests by the police also only 5%?
The answer is, no one knows. There has never been a citywide effort to estimate how many people drink and drive in El Paso.
That means the total number of DWI arrests made in a year is without a benchmark – it does not tell us how much of the iceberg we are seeing. It does not tell us if the DWI problem is getting better or worse. So, what do DWI arrests tell us?
DWI arrests tell us more about the response to the problem than the problem itself. The changes in DWI arrests reflect changes in enforcement activity, not the behavior we seek to change.
It’s easy to confuse the two. To better explain, take the following example: imagine a police department doubled its capacity to make DWI arrests, increasing from 30 in one week to 60 arrests the following week.
Does this increase mean twice as many people were out drinking and driving?
Not necessarily.
The same logic would apply if DWI arrests decreased. For instance, if the total DWI arrests in one week went from 60 to 45, again, it would not necessarily mean less people were drunken driving. The decrease might be explained by changes in enforcement activity.
In fact, total DWI arrests is not the right information we need to answer this question, because we have no idea how many people were actually out committing the behavior.
So, if arrests are not a good measure of the problem (it’s a better measure of the response to the problem), then what is?
Researchers who evaluate DWI prevention strategies often use alcohol-related vehicle crashes as a measure of the DWI problem.
The figure below shows four-and-a-half years of alcohol-related crash trends in major Texas cities, from January 2021 to July 2025. Each line represents the monthly rate of alcohol-related crashes per 100,000 people. Although crashes, just like arrests, show us only a portion of the people who engaged in the behavior (because not everyone who drinks and drives crashes), one can argue this is a better measure of the problem we aim to solve.
Except, comparing El Paso with other cities seems like a mistake.
Crash rates in one city are not meant to be pitted against others. It’s not a competition, after all. And just like arrests, perhaps there are a host of other potential causes for why a city might experience increases or decreases in vehicle crashes. In fact, comparing El Paso with other cities seems beside the point. The goal is to figure out how to get the rate of crashes down to zero, regardless of how high or low El Paso is relative to other cities.
If a “vision zero” outcome is indeed our goal (but starting with even just a 10% decrease is something to be proud of), what follows is probably the most important question we should be asking:
How do we get there? More generally, what is the solution to the DWI problem in El Paso?
Several groups in our city are fighting to be the solution to the DWI problem (many of them are probably on to something). However, most of their efforts have never been empirically evaluated. That is, we cannot say with confidence what does or does not prevent DWI in El Paso.
I believe the path forward is to evaluate those local efforts, so we can expand our understanding of the problem we intend to solve. We should engage in formal research to help us learn about the problem in a systematic, objective way.
Another idea is to conduct a citywide survey, where residents are invited to provide confidential responses to questions about their attitudes and behaviors relating to drunken driving.
Taken together, evaluations and community surveys are how we can find out whether our current strategies are reaching the people who need them, missing the mark, or even making the problem worse. It is how we identify the moments of highest risk, the communities carrying the greatest burden, and the messages that actually change behavior.
We need local research to generate evidence, where we listen to residents, collect data, and let evidence – not guesswork – drive the response to a problem.
Let’s solve DWI with evidence.
Matthew J. Durán is a criminal justice researcher providing technical assistance to government and non-government organizations. He is currently a Ph.D. student studying criminology and criminal justice at Arizona State University.
The post Opinion: To solve El Paso’s drunken driving problem, start with evidence appeared first on El Paso Matters.
Read: Read More