I'm guessing he's aggregating the numbers from two different data sets. The New York smokers data set with incomes <= $30,000, (which spent almost a quarter of their income on cigarettes) and the $13,000 a year and under data set who on average spent 9% of their money on the lottery.
That's not going to yield an accurate or meaningful result, but I'm guessing that's what happened.





Reply With Quote