Smoking v. Obesity: The Economics Of Prevention And Its Dependence On Treatment

In 2012, the Congressional Budget Office (CBO) released an in-depth study of the health and budgetary effects of raising the excise tax on cigarettes. We commented on this study in our blog about the complex economics of disease prevention and longevity. CBO has since turned its attention to obesity and recently released a list of issues needing resolution in order for CBO to estimate the effects of federal policies impacting obesity.

In this Health Affairs Blog post, we summarize research we have done, under a grant from the Robert Wood Johnson Foundation (RWJF), on the value of reductions in cigarette smoking and obesity. Among our findings, as explained in more detail below, was how dramatically the returns to prevention could be impacted by changes in the cost and effectiveness of treatment for the conditions being prevented.

Specifically, we found that preventing tobacco use and obesity at age 20 yielded similar levels of federal budget savings over time horizons out to 60 years, although the savings came from different sources. Beyond 60 years, savings from the prevention of tobacco use begin to decline because of resulting increases in longevity, and thus government expenditures for older Americans. Savings from obesity prevention, on the other hand, continue to increase for longer time horizons because there is little impact on longevity due to new treatments that have largely eliminated differences in mortality rates between obese and non-obese people at older ages.

However, expensive new treatments for smoking-related diseases could similarly reduce life expectancy differences between smokers and nonsmokers. If this occurs, tobacco prevention could become more like obesity prevention, saving treatment costs in the short run while not increasing longevity and associated costs in the longer term. (Of course, the budgetary perspective is only one lens through which to view prevention efforts.)

This raises a general point that policymakers and researchers should keep in mind: Advances in the treatment of many chronic conditions could dramatically impact the value of preventing them.

Background

In our work for RWJF, we posed the following questions:

• What is the value of reducing the number of smokers at age 20 by 100,000?
• What is the value of reducing the number of obese persons at age 20 by 100,000?

We did not try to specify how these reductions would be accomplished (nor their cost). Instead, we focused on the benefits that would occur if they could be accomplished. Those considering specific programs for reducing the prevalence of smoking and obesity can use this information, with appropriate rescaling, to help determine if the implementation costs are justified by the expected benefits.

We developed a model implemented as a spreadsheet application to compute benefits in four basic dimensions: better health (linked to quality adjusted life year [QALY] calculations), reduced mortality, lower health care costs, and greater productivity and earnings. It also computes the economic value of these benefits from the perspectives of society, the federal budget, and state government budgets.

For the smoking application, we relied heavily upon data from the aforementioned CBO study of an excise tax on cigarettes. For obesity, we relied upon a variety of studies in the literature. Our technical report contains a more in-depth description of our model and data sources used in its application to these smoking and obesity scenarios. A recent description of the modeling framework is here.

The bar chart below shows aspects of the differential impact of smoking and obesity on mortality, health care costs, and earnings based on recent research. Smoking results in higher mortality at older ages, while obesity has little effect. Put another way, converting a would-be smoker to a non-smoker increases longevity.

In general, this should be an unequivocal positive but, from certain stakeholder perspectives such as the federal budget, this can be viewed negatively, since increased longevity into old age results in greater federal spending on Medicare and Social Security. Average annual per-capita health care costs are pushed up more by obesity than by smoking. However, smoking causes a greater reduction in average annual earnings (Note 1).

Since federal budgets are strongly affected by health care costs, earnings, and longevity, we focus here on the federal budget perspective to highlight differences in these two scenarios. We ask how federal revenues and outlays would be impacted under a smoking reduction and an obesity reduction scenario.

Smoking And Obesity Scenarios

For the smoking scenario, we focus on what it would be worth if we could reduce the number of smokers at age 20 by 100,000 this year and in each subsequent year. More specifically, suppose there are 100,000 individuals in each age-20 cohort who would have been lifetime smokers and who will instead become lifetime non-smokers. We simulate how this would impact mortality, health care costs, and earnings each year into the future.

In the first year of the simulation, we are only concerned with the current cohort of 20-year-olds. In each subsequent year, we add another 20-year-old cohort, and thus the size of the simulated population grows over time. It is worth noting that this general scenario fits quite well with what might happen if current proposals to raise the legal age for purchasing cigarettes were implemented.

The obesity scenario is essentially the same as that described above for smoking. We focus on what it would be worth if we could reduce the number of obese individuals at age 20 by 100,000, for this and all subsequent years. Specifically, we imagine that there are 100,000 20-year-olds who would have been obese throughout the remainder of their lives but who are converted to being normal weight. As with the smoking scenario, cohorts accumulate over time as the early cohorts age and new age-20 cohorts enter. This scenario fits well with policies focused on reducing childhood obesity, since such policies should reduce obesity at age 20 in the coming years but would not have much impact on obesity among current adults.

The Impact On The Federal Budget

The alternative life paths in the model were used to estimate, for each year of the simulation, the net effect of reductions in smoking and obesity on annual deaths, health care costs, and earnings. Using other data sources, we estimated how these changes would impact the federal budget.

Changes in health care costs impact the federal government for those individuals who are covered by Medicaid and/or Medicare. Changes in earnings impact federal tax collections and can cause shifting in and out of Medicaid and federally funded social programs. Lower mortality has some positive effects on the federal budget in the early years of the simulation (increased tax revenues) but, in the long run, can result in greater federal outlays for programs like Medicare and Social Security.

Results For Alternative Time Horizons

Our model produces annual amounts of net gains (or losses) to the federal budget over as many years as we choose to run the simulation. One could create a total net impact by summing this flow of gains and losses, but this would be misleading, since a dollar today is worth much more than a dollar in 50 years. Therefore, we take the present value of the projected annual amounts, using a discount rate of 3 percent per year. In the charts below, we plot these present values over a 100-year period, encompassing a full range of time horizons of interest to various stakeholders. For example, the Congressional Budget Office (CBO) usually examines the 10-year time horizon.

Smoking Obesity

The present values are depicted by the solid black lines. They show that the gains to the federal budget are remarkably similar between the two scenarios for time horizons up to 60 years. Over this time period, from a federal budget perspective, the gain from reducing the number of new smokers is about the same as the gain from reducing the same number of new obese persons.

The source of the gains differs, however, with smoking reduction contributing more to increases in federal tax revenues via higher taxable earnings, and obesity reduction contributing to lower federal health care spending. The gains are large, reaching $2 billion for the 10-year horizon, $7 billion for the 20-year horizon, and over $40 billion for the 60-year horizon. One would expect that the hypothesized reductions in numbers of new smokers or obese persons could be accomplished at a cost much less than these projected benefits, particularly with time horizons approaching 60 years.

For horizons beyond 60 years, the gains from reduced smoking begin to fall, while those for reduced obesity continue to rise. This can be traced to the finding that reduced smoking increases longevity at older ages while reduced obesity does not. The longevity effect, shown in red, reflects increases in federal spending on Medicare and Social Security as individuals live longer. The blue bars show the present values without the dollar impact of longevity.

The Strong Dependence On The Costs And Effectiveness Of Treatment

One of the key findings from our research was the way in which innovations in the treatment of obesity-related medical conditions have impacted the benefits of obesity prevention. To illustrate, the next chart shows the results of modeling an identical obesity scenario, but using life path estimates based upon published findings pertaining to the 1980s rather than the early 2000s. It shows a pattern much more like our current smoking scenario, with a large longevity effect in the longer time horizons that causes the net gains to decline for horizons greater than 60 years or so.

Obesity (1980s)

There are two reasons for the dramatic difference in the scenario using 2000s inputs versus the scenario using 1980s inputs:

1. The impact of obesity on health care costs was smaller back then; and,
2. The impact of obesity on mortality was higher.

We believe that these two findings are closely related. Since the 1980s, expensive new treatments for obesity and its associated medical conditions (such as new drugs to treat diabetes, hypertension, and hyperlipidemia, and new surgical procedures such as bariatric surgery) have driven up obesity-related health care costs. These expensive new treatments have also reduced mortality associated with obesity. As a result, reductions in obesity save more in health care costs today than in the 1980s, but they do not increase longevity costs because obese individuals now have roughly the same longevity as others.

Some Implications For The Impact Of Reduced Smoking On The Federal Budget

This raises the interesting question referred to at the beginning of this post regarding current estimates of the value of reductions in the number of new smokers, and more generally the value of reducing many chronic conditions. Recent advances in screening, surgical procedures, pharmaceuticals, and chemoradiotherapy show promise for improving the prognosis of lung cancer patients. Is there a chance that expensive new treatments for diseases associated with smoking will drive up health care costs and extend life? If so, both would increase the costs to the federal budget of smoking and, by implication, increase the gains to the budget from smoking cessation.

More generally, by ignoring emerging advances in the treatment of many chronic conditions, we risk under-estimating the value of preventing them from a federal budget perspective.

Note 1

These estimated effects are generally derived from multivariate analyses in the literature that control for other characteristics of smokers and obese persons that may also impact mortality, health care costs, and earnings.

from Health Affairs Blog http://ift.tt/1oDwZv6