The average is a really misleading statistic here, because alcohol and drug use seem to follow a power law distribution. The top X% of drinkers drink orders of

The average is a really misleading statistic here, because alcohol and drug use seem to follow a power law distribution. The top X% of drinkers drink orders of magnitude more than the average. They skew the average much, much higher than what people think of as how much the "average person" drinks. According to this news article [1], the "average person" drinks (73.85 + 15.28 + 6.25 + 2.17 + 0.63 + 0.14 + 0.02) / 10 = 9.834 drinks per week! But the fifth decile (the middle of the distribution) drinks only 0.14 drinks per week --- less than one per month! I appreciate your analysis but I think the average is too lossy of a statistic to make a good argument here. 1. https://www.washingtonpost.com/news/wonk/wp/2014/09/25/think-you-drink-a-lot-this-chart-will-tell-you/ https://www.washingtonpost.com/news/wonk/wp/2014/09/25/think...