Then find a current study not one that uses data from two to three + decades ago, that only divides income into above and below the medium income levels (doesn't segregate poor as was my original claim) and even at that doesn't have income data from almost 15% of the respondents and admits doesn't control for many variables that effect cigarette consumption. Like I said I said demonstrate not say. That particular study does not demonstrate even though it says it does. BTW -- back in the day, in the years the study encompasses being below the median in income did not mean you did not pay Federal income taxes.