The Danger of Biased Drinking Studies

I got a very controversial email about a week ago, stating that day drinking – prior to 3pm – causes worse hangovers, more road accidents and generally a greater incidence of rowdy public behavior. As someone who has always enjoyed a glass of wine with a good lunch, I found it a bit suspicious.

The study is called Data on Day Drinking and was published by Alcohol.org which is a subsidiary of the Tennessee-based American Addiction Centers' (ACC) network. It is a self-defined "organization that specializes in addiction and recovery", shared notes spokesperson Tyler Burchett on behalf of Alcohol.org. He added that while the group supports "sober lifestyles… we also recognize that not everyone abstains from drinking".

The AAC is a provider of both residential and outpatient addiction treatment services, he adds. What is more the "study was funded with the hopes of starting a conversation around the dangers of day drinking versus night drinking… when people think of the negative aspects of drinking, they are most often associated with nighttime drinking [however] … The goal of the study is to shine a light on an aspect of drinking that is often overlooked."

The industry reaction

Most analysts in the business were quick to note that the data provided – based on a pool of one thousand respondents – seems to stem from an anti-alcoholic beverage perspective. A wine industry analyst, who preferred not to be identified, notes that the study's data looks like it was provided by an "anti-alcohol organization, so I would question how questions were drafted and put to the participants. Also even the sourcing of the participants could sway the outcomes".

He went on to continue that, as an "industry we were very proactive 25 years ago to address anti-alcohol interests. I think our focus has slipped in this regard and the industry needs to be more aware and responsive to this re-emerging issue."

Christian Miller, proprietor of the Californian number-crunching firm Full Glass Research, said that he was "instantly suspicious" of this study. He adds: "The average number of drinks [cited as] consumed per occasion – it dwarfs what you see in other surveys, which typically range in the two-to-three area."

He adds that, "In fact, I suspect the entire day drinking volume data is skewed by people referencing individual events. Like reporting 'I had six beers at Coachella [the music festival] and I have a daytime alcoholic drink twice a week,' therefore twice a week I have six drinks."

Other analysts in the drinks business also expressed concerns about the study's findings. "The report is also being funded and run by an anti-alcohol organization … which leads to questions about intent and bias of results," shared Rob McMillan, the St. Helena-based executive vice president and founder of Silicon Valley Bank Wine Division. "I don't see a breakout of the locations of respondents either. … The only information we are given is the respondents are in the US. The answers to those questions would add additional survey bias."

What is more, he adds that the Maryland-based National Institute on Alcohol Abuse and Alcoholism "data suggests roughly 56 percent of the adult population of 200,000,000 people in the US – or about 112 million people – consume alcohol regularly. So their sample size is 0.001 percent of the consuming population and perhaps a little thin to expect a high degree of statistical significance in the findings."

What is more, Miller notes: "Most of the footnotes and citations refer to popular news or 'news' articles, not academic or government or industry research." What is more, both Miller and I noticed that when visiting the page that "there was an instant popup that offered help with drinking problems. And that they named a link to an article on brunch 'boozy' brunches, which is not the title or sub-title of the article. And the sponsor is American Addiction Centers."

Investigating further

Some of the top investment bankers who follow the wine business also have questions about these types of studies. While, Bourcard Nesin, a beverage analyst at Rabobank's New York City-based Food and Agribusiness Research, adds that he has, "no particular qualm with alcohol.org, but these kind of surveys are not scientific and considering how much it costs to do high-quality polling [let alone scientific research] I would take these statistics with a grain of salt and receive them in the good-humored way they are presented… oh, and don't drink and drive," shares Nesin.

He adds that assumptions about when and what types of alcohol one drinks during the day – and how much that consumption might harm the individual consumer – are questionable. "Any suggestion that alcohol type or time of consumption affects hangovers is apocryphal. In my mind the only thing that really impacts a hangover is the amount of alcohol you consume. If people drink more because they started earlier, then it would make sense that day drinking could give you a worse hangover… not because of when you drank, but how much."

However, he goes on to add that daytime drinking and driving has continued to cause more damage and accidents. He notes that, "According to a National Highway Traffic Safety Administration (NHTSA) roadside survey from 2014, about 1.5 percent of weekend nighttime drivers test over the legal 0.8 BAC limit. [while] 0.4 percent of daytime weekend test over the legal limit of 0.8…. According to 2017 data, alcohol is a factor in 36 percent of fatal weekend nighttime crashes and it is a factor in 13 percent of daytime weekend fatal crashes."

"NHTSA data suggests that daytime alcohol impairment is more likely to result in a fatal crash, particularly single vehicle accidents. While nighttime drunk driving has been dropping steadily for decades, daytime drunk driving has not declined over the last decade or so and may even be increasing."

So perhaps the anti-alcohol folks might have some pertinent data to share – ideally through a more thorough study conducted with a larger sample base.