It’s tax-day today. It is also the forever day for tobacco companies when their settlement requires a payment of $9 billion to the states for the cost of tobacco-related illnesses, cardiovascular disease, lung disease, and cancers.  A new paper suggests that merely reducing the prevalence of smoking by 1% among those that smoke the most, Medicaid recipients, would result in $2.6 billion in savings to those states.
About 25% of Medicaid recipients smoke as compared to our national average of roughly 14%, so focusing on their use is a reasonable approach. Smoking is not homogenous, few things are. There are more smokers in the Midwest and South, with far fewer in the Northeast and West Coast; Medicaid recipients smoke more in all these regions, mirroring the overall pattern, except in the Northeast where they smoke as much as those in the Midwest.
To convert decreased smoking into healthcare dollars, the authors made use of another paper which “quantified short-run changes in health care costs…” So, to really understand their calculations you need to read the paper creating the conversion. That conversion factor is a different model and begins by determining the healthcare spending per capita in a state using relatively firm data, CDC reported healthcare spending (both private and public) and state populations. But it veers towards fuzzy math when it attributes a portion of those costs to smokers based merely on their prevalence. It was further a guesstimate when the only adjustments were made for income, age, and race.
To be fair, the authors of the conversion factor acknowledge those problems; they are not attributing causality, just piling on to the discussion. In that study, the prevalence of smoking was 21.2% nationally, 372 packs per year with an associated healthcare cost of $6,426. A 1% relative reduction, to 21% was associated with a savings of $7.58. It is only when you apply the conversion to large populations that we see those $2.6 billion savings.
With a conversion factor in hand, it is relatively straightforward to calculate savings, get a big number and publish.
Smoking is bad for your health. Any additional motivation is worthwhile in ending this hazard to our health and budget. My concern with the paper is that the conversion of smoking reduction to healthcare costs occupied one sentence; only a skeptic would look to understand that assumption. If you are modeling disease and its impacts, simply identifying your assumptions is insufficient, you need to discuss using more than a reference the basis for that assumption.
 This is part of the Master Tobacco Settlement
Source: Estimation of 1-year changes in Medicaid Expenditures Associated with Reducing Cigarette Smoking Prevalence by 1%. JAMA Network Open DOI: 10.1001/jamanetworkopen.2019.2307