Not all homework assignments are boring! This is a short paper I wrote for my health economics class on the relationship between alcohol and wages – a fascinating subject that has actually had health economists debating for decades!
The relationship between alcohol consumption and wages is not a simple one. The first challenge in understanding it is that working decisions and drinking decisions are made jointly, and that both affect each other. Staying out late and binge drinking can leave workers tired, hung-over and unproductive the next day, and we would expect workers who do this regularly to have lower wages. Conversely, a productive worker who puts in more hours may be stressed, and therefore more likely to drink, but still also able to maintain higher wages. This starts to illustrate how different people drink for different reasons; some may drink because they don’t care about work, others because they care too much, while for many drinking does not have a direct impact on their work.
Another factor complicating the study of this issue is that empirical research seems to deliver conflicting reports of the impact of drinking on wages. While some studies show that alcohol consumption has a negative relationship with wages, others show a positive relationship. Perhaps moderate drinking allows workers to develop relationships with other workers, de-stress, and generally lead healthier lives than non-drinkers. Of course, it is unlikely that alcohol abuse (chronic binge drinking) carries these positive effects.
This paper will introduce some applicable theory to address the question of how drinking alcohol affects wages, summarize some of the empirical research on the subject, and offer a potential policy change to lessen the negative impacts while amplifying the positive impacts of drinking for workers.
Becker and Murphy’s rational addiction model is the standard economic framework for understanding addiction. In economics, a good is considered addictive if current consumption of the good has a positive influence on an individual’s propensity to consume the good again in the future. Because alcohol consumption is habit forming, it makes sense to evaluate the explanatory power of the rational addiction model with respect to drinking behavior.
The rational addiction model is based on the discounting of utility across time, where present utility receives more “weight” than future utility. In finance, the “time value of money” principle allows for the precise calculation of present and future cash flow values based on a discount rate. Similarly, the rational addiction model asserts that individuals discount health utility, and that a given amount of utility today is valued more than the same amount of utility some day in the future.
From this perspective, an individual could make a rational decision to become addicted to a destructive good (e.g. binge drinking) if the utility they get today outweighs the (discounted) negative utility expected in the future. An individual with a higher discount rate would be more likely to rationally use an addictive good, because the future is more heavily discounted for them (i.e. future events are less significant to an individual with a lower discount rate).
With respect to drinking, the model would assert that individuals have a complete and accurate picture of the impacts (positive and negative) of their drinking behavior before they start to consume alcohol, allowing them to make a rational decision whether or not to start. So, while the question challenges economists on aggregate, in the model, individuals have a much clearer picture of how their drinking will affect their wages, along with their health, social lives, etc.
Empirical studies of the relationship between drinking and wages have produced conflicting results. In 1988, Berger and Leigh showed that drinkers (including both moderate and abusive) earned higher wages than nondrinkers across demographic and human capital groups. Then, in 1993, Mullahy and Sindeler found a negative relationship between alcohol use and both wages and labor participation in men aged 30 to 59, but not for older and younger men.
Shortly thereafter, some studies started to piece together the conflicting results. French and Zarkin found in 1995 isolated moderate drinkers, and found that they have higher wages than others. Hein’s 1996 paper “Do drinkers earn less?” showed a concave relationship where moderate drinkers earn more than non-drinkers or heavy drinkers. This has been reinforced by other studies as well.
More recently, Shao-Hsun and Wallace Huffman’s paper “Binge Drinking and Labor Market Success: A Longitudinal Study on Young People” (originally written in 2002, and most recently re-published in January, 2010) found that an individual’s decision to binge drink is alcohol-price responsive and that it is a rational addiction. In line with previous research about heavy drinking, the study also found that binge drinking had a statistically significant negative impact on earnings. Their model used a pair of least squared regressions, one modeling drinking behavior as a function of income and other factors, and the other modeling income as a function of drinking and other factors.
If you visit Southwest Airlines’ headquarters overlooking Dallas Love Field Airport on a Friday afternoon, you’ll surely hear music blaring above you, and the sound filtering down of some seriously happy workers chatting, laughing, and celebrating the end of the workweek. Every Friday during the summer months, Southwest has a “deck party” where workers can socialize, relax, and usually drink on a giant deck, overlooking the runway where colorful Southwest jets land and take off all day long.
Frequent work-sponsored events like this one (i.e. with alcohol) give workers an opportunity to relax, have fun, socialize, and break down divisional silos. And, because it is a work event, almost no one drinks so heavily that they embarrass themselves or negatively affect their work. In short, you get all of the good effects of alcohol consumption with none of the bad. Events like these build the culture and cohesiveness of an organization, and alcohol plays a very real role in cementing those crucial and productive bonds between human capital – thus improving productivity, retention, efficiency and quality.
As an intern at Southwest, I was shocked when I attended my first deck party to find workers “limbo-ing”, playing beer pong, dancing – actually having a good time at a work event. And so much of Southwest’s model culture is rooted in the company letting workers relax, let loose, and have fun. I can’t imagine a Southwest without alcohol; it is as much a part of their culture (chain smoking founder Herb Kelleher loves his Wild Turkey!) as the Boeing 737.
So my policy recommendation is for employers to create some regular social events for workers, and supply the booze. It’ll take some work for some firms get to the level of camaraderie and cross-divisional cohesion that Southwest has achieved, but letting your workers be drinking buddies is a great first step in that process.
One last fun fact about Southwest: just a few years ago that the company placed a 5:00 rule for workers – no drinking at the office until after 5pm. Prior to that rule it wasn’t uncommon to see coolers of beer wheeled in to meetings!
Having grown up in a household where alcohol (especially recreational drinking) was frowned upon, I would have been shocked to hear about the positive effects of alcohol use for workers documented in some of these studies. While the negative impact of problem drinking and alcoholism on wages would be expected, the fact that moderate drinkers earn more than nondrinkers is surprising.
I find that the bulk of research and theory reinforces my personal decision to enjoy alcohol in moderation and to take advantage of its positive social effects.