Monday, September 24, 2018

On the Houston Chronicle's 'misplaced pride' and flawed reporting on DWI deaths

In an unsigned staff editorial, the Houston Chronicle complained that, "misplaced pride needlessly costs the lives of hundreds of people every year" because of DWI. But their own pride caused the Opinion section of the paper to double down on flawed reporting from the news side on DWI deaths.

The editorial board joined the news section's call for a regressive tuff-on-crime response to DWI, flying in the face of recent evidence about factors impacting DWI fatalities.

And they continued to manipulate data on the topic, focusing on aggregated numbers over a sixteen year period without parsing annual trends. This obfuscation leads them to disingenuously pretend that DWI-related fatalities might go down "if local law enforcement agencies assigned more officers to the task" of arresting people.

But we know that's not true. As demonstrated in this Grits post, the paper's assertion of a link between police deployments, DWI arrests, and alcohol-related fatalities simply isn't backed up by the evidence.

Texas has just witnessed a natural experiment regarding the relationship between DWI enforcement and alcohol-related-traffic fatalities, and something unexpected happened: DWI enforcement declined significantly in the past few years, Texas' overall population increased, and per capita traffic fatalities declined.

That's the opposite of what would have occurred if the relationship between police staffing and fatalities posited by the Chronicle were true.

Chronicle writers are able to keep up this pretension by relying on 16 years of data, aggregated altogether, but ignoring points where recent trends belie law-enforcement assertions in the story. The approach masks clear flaws in their recommendations, which for the most part are recycled wish-list items from the local police chief (hire more officers, authorize traffic checkpoints, etc.).

If more staffing assigned to DWI enforcement would reduce deaths, then traffic deaths in 2010 would have been lower than they are today. After all, Texas cops were making tens of thousands more annual DWI arrests at the time.

But clearly, from the annualized data trends, police enforcement levels weren't a decisive factor. What might be? ¿Quien sabe? My guess is that public-service advertising, changing cultural norms, and the rise of Lyft and Uber had something to do with edging per-capita death rates down. But one certainly couldn't tell from the Chronicle's reporting.

Other commentary was just weird. For example, the staff editorial opined that, "Texas has some of the most lenient license suspension laws in the country." But in reality, as The Washington Post reported earlier this year, Texas has suspended more drivers' licenses than any other state: 1.7 million, all told. It's like the editorial board didn't even Google the topic before writing this stuff.

It was disappointing that, instead of parsing facts to discover what works and what doesn't, the Chronicle recycled a bunch of tired, tuff-on-crime memes that fly in the face of the data. It all feels depressingly familiar, like throwback coverage from the 1990s.

RELATED: Check out this Reasonably Suspicious podcast segment from May hypothesizing causes for the decline in DWI and drunkenness arrests across Texas.

4 comments:

Anonymous said...

Has the world turned completely upside down? It used to be that the Houston Comical was as reliable a purveyor of "hug-a-thug" tripe as, let's say..... Grits for Breakfast! LOL! WTF?

Gritsforbreakfast said...

And yet, here you are, a regular reader, 5:02. Funny, that.

Steven Seys said...

The yellow journalism of the Houston Chronicle is no surprise to me. In the early 1990s the Chronicle sent a photographer into the Coffield Unit to take "day-in-the-life" type pictures for a special piece. The shutter jockey asked the security supervisor to stop every prisoner with any hair on his head so they could take a shot of the baldies. Two weeks later that photo was printed with the caption, "Neo-Nazi skin heads take over the prison." Two out of three prisoners in the photo were Hispanic, although it was hard to tell from the back.

Anonymous said...

Old Saying in Mathematics:

Statistics are like an Old W***e, you can make them show anything.