How “the Grim Reaper effect” stops our government from saving lives

Last summer, the Congressional Budget Office released a report under the unassuming name “Budgetary Effects of Policies That Would Increase Hepatitis C Treatment.” I read it because I am the type of person who is interested in the budgetary effects of policies that would increase hepatitis C treatment.

Embedded in the report, though, was a point that will be important for just about anything the federal government tries to do to save the lives of Americans.

Hep C is a nasty viral infection whose effects are, for a virus, unusually long-lasting. Untreated, it causes serious liver damage over the course of decades, leading to much higher rates of cirrhosis and liver cancer, all of which is very expensive to treat.

But in the 2010s, a number of extremely effective antivirals, which randomized trials show cure upwards of 95 percent of chronic infections, came on the market. Like most new drugs, these antivirals are under patent and quite expensive; as of 2020, the cost of an eight-to-twelve week course of the drugs, usually enough to cure an infection, was between $11,500 and $17,000.

Yet CBO concludes that the drugs are so effective, and the costs of treating patients with hep C who haven’t been cured are so massive, that expanding treatment with these drugs reduces federal spending on hep C treatment and associated complications overall. Doubling the number of Medicaid patients getting the drugs would increase federal spending by $4 billion over 10 years. But over the same decade, the federal government would save $7 billion through reduced need for treatments like liver transplants and ongoing care for chronic cases.

Put like that, this starts to sound like one of the rarest discoveries in federal budgeting: a free lunch. That means a policy that is good on its own merits (saving lives and preventing debilitating chronic disease) but also saves the government money.

But the most interesting part of the report to me comes at the end. “An increase in hepatitis C treatment could also affect the federal budget in other ways—for example, by leading to improved longevity and lower rates of disability,” the authors note. The latter point is pretty straightforward: If hepatitis C leads to disabilities that make people eligible for disability insurance and subsidized health coverage, then reduced hep C means lower spending on those programs. But (and this is me speculating, so blame me and not the CBO if I’m wrong) that effect is probably swamped by that of “improved longevity.”

Simply put: curing hep C means people live longer, which means they spend more years collecting Social Security, Medicare, and other benefits. That could mean that whatever cost savings the actual hep C treatment produces might be wiped out by the fact that the people whose lives are being saved will be cashing retirement checks for longer.

I like to call it the Grim Reaper effect. The US runs a large budget deficit. It also provides far more generous benefits to seniors than to children or working-age adults. Per the Urban Institute’s regular report on government spending for children, the ratio of per capita spending on senior citizens to per capita spending on children is over 5 to 1. Put together, the deficit and the elder-biased composition of federal spending implies something that is equally important and macabre: helping people live longer lives will, all else being equal, be bad for the federal budget.

In an increasingly aging country, hep C is not the first place where the Grim Reaper effect has been felt, and it won’t be the last. I don’t have an easy fix for the situation, but it feels important to at least understand.

One of the first and clearest cases of this longevity dilemma in budgeting came with cigarettes.

The history of mass cigarette smoking in the US is surprisingly short. Per the CDC, American adults were only smoking 54 cigarettes annually per capita as of 1900. By 1963, that number had grown to 4,345. The development of automatic rolling machines, milder forms of tobacco, and mass marketing meant millions of working and middle-class Americans became pack-a-day smokers.

But while the per capita average floated around 4,000 from the late ‘40s to the early ‘70s, it then began a precipitous decline. In 2022, the most recent year for which the Federal Trade Commission released data, Americans bought 173.5 billion cigarettes, or 667 per adult, less than a sixth of the peak, while fewer than 12 percent of American adults now smoke.

Cigarettes are, of course, deadly, but they kill with a lag, usually after decades of regular smoking. That meant that in the late 1980s and 1990s, the US started to hit peak cigarette deaths, as adults who came of age during the smoking era started to get lung cancer and emphysema en masse, at numbers that less-addicted subsequent generations wouldn’t match. The male death rate from lung cancer peaked in 1990, and the female death rate peaked in 1998.

A flurry of economic research at the time tried to make sense of what this meant for the federal budget. Smoking harms your health. But it also shortens your lifespan. A useful 1998 Congressional Budget Office report noted that most research found that, over their lives, smokers spend more in health care costs (including more that goes on the federal tab) than non-smokers, even accounting for their shorter lifespans.

But that picture changed once you added in pensions and other non-health spending. Economists John Shoven, Jeffrey Sundberg, and John Bunker in 1989 estimated that the average male smoker saved Social Security $20,000 (about $60,000 today) in benefits not paid. The figure for women, who live longer than men on average but earn less in wages and thus in Social Security, was about half that.

“It seems likely that the Federal budget currently benefits from smoking,” two Congressional Research Service researchers concluded in 1994, when the “benefits” of early death to Social Security and Medicare were included. Malcolm Gladwell, in a thoughtful 1990 treatment of the problem in the Washington Post, was catchier: “Not Smoking Could be Hazardous to Pension System.”

Decades later, the CBO did a fuller analysis of the budgetary consequences of smoking in the aftermath of the large cigarette tax increase President Obama signed in early 2009 and proposals for further hikes. At first blush, the revenue raised from a cigarette tax should be easy to estimate: multiply annual cigarette sales by the amount of the tax. But obviously raising the price of the good will reduce the amount people buy; one major reason for cigarette taxes, after all, is to deter smoking. The CBO used a price elasticity of -0.3, meaning that a 10 percent increase in cigarette prices reduces the number sold by 3 percent.

But the 2012 report was meant to go a step or two further, according to then-director Doug Elmendorf, who explained the backstory in a recent conversation with me. “The effects of making people healthier are good for those people, obviously, but also perhaps good for the federal budget because the federal government pays for a lot of health care. If you’re healthier, you don’t need so much health care.” But at the same time, “It was clear that if people were healthier, they would live longer, and that could have budgetary costs. It wasn’t obvious offhand what the balance of those effects would be.”

The 2012 CBO report tried to put all these effects together: the effect of lower smoking on reducing health-care spending (including government-funded spending) due to a healthier population, the effect on Social Security and other benefit spending from resulting longer lifespans, the effect of lower smoking rates on wages, and tax revenue from those wages. (The latter is often not included in formal CBO scores, as it tips closer to “dynamic” scoring where the effect of legislation on the overall economy is included.)

Over the first 10 years after a hike in the cigarette tax, they found that having a healthier population was more of a blessing than a curse, budget-wise. The health effects of a cigarette tax hike reduced federal health spending by over $900 million over a decade, even after accounting for people living longer and claiming more years of Medicare. By contrast, retirement programs only spent $183 million more because people lived longer. Swamping all that was a $2.9 billion increase in tax revenue from a healthier population capable of working and earning more.

But that’s just the 10-year effect. As the decades pass, the effect of longevity would grow and grow. First, Medicare costs would start to rise, as the cost of a longer-lived population began to swamp the cost savings of that population being healthier overall. (Even people who’ve been healthy for a long time can run up major health spending at the end of their now longer lives.) Social Security costs would keep rising, too.

Fifty years in, these costs would overwhelm the benefits, and the cigarette tax’s health effects would start costing the budget, on average.

The point isn’t “cigarette taxes are good” or “cigarette taxes are bad.” The point is that even a policy that saves lives isn’t necessarily a slam dunk from the hard-eyed perspective of budget policy.

Recent years provided a possibly even darker example. In 2022, the Medicare Trustees pushed back the date they expected the program’s Hospital Insurance Trust Fund to be depleted by two years. They had several reasons, but a major one was that Covid-19 had killed hundreds of thousands of Medicare patients prematurely. Not only that, but “Medicare beneficiaries whose deaths were identified as related to COVID had costs that were much higher than the average Medicare beneficiary prior to the onset of the pandemic.” Put another way: Covid killed off Medicare’s sickest, and most expensive, enrollees. That meant the program was left with an overall healthier population, which by itself lowered medical costs by 2.9 percent in 2021.

Similarly, a paper by a team of health economists earlier this year estimated that the 1.4 million excess deaths in the US due to Covid had the net effect of boosting the Social Security trust fund to the tune of $156 billion. That represented $219 billion in benefits that no longer needed to be sent, minus $44 billion in lower payroll tax revenues and $25 billion in new benefits to surviving family members.

It all reminds one of Logan’s Run, in which people are killed off upon hitting age 30 lest they take up too many of society’s resources. That movie is a dystopia — but as a budget proposal, it’d score very well.

It’s good to save lives, actually

The economists and agencies doing this math are, of course, only doing their jobs. We need to know what government programs will cost over the near- and long-run. These effects on health and life and death matter to those calculations.

“Members of Congress regularly thought that we were ghoulish for talking about how, if people live longer, there’ll be higher benefits for Social Security,” Elmendorf recalls. “But it’s not ghoulish. Obviously, we want to live longer and members of Congress should try to help all Americans live longer. CBO’s job — an analyst’s job in general — is just to be honest about the likely effects.”

But the fact that increased human longevity on its own worsens the budget picture should lead to some reflection. For one thing, it suggests that sometimes we should embrace policies simply because they’re the right thing to do, even if they don’t pay for themselves. Recall the hepatitis C treatments that prevent expensive long-term expenses for Medicaid, but might add on new costs by extending the benefits’ lifespans. It’s possible that, upon taking the latter into account, expanding access to hep C drugs costs the government money on net. It’s a free lunch no longer.

That’s not a reason not to embrace the policy, though. Lots of things the government does cost money. The military doesn’t pay for itself. K–12 schools don’t pay for themselves. Smithsonian Museums don’t pay for themselves. That doesn’t mean those aren’t important functions that it makes sense to put some of our tax dollars toward. Hep C treatment, I think, fits in that list, even if it’s not literally free from a budget standpoint.

Congress should also allow agencies like the CBO to do more to symmetrically account for the positive budgetary effects of longevity, along with the negatives. People who live longer, after all, often earn wages in those new years of life, wages that generate income and payroll tax revenues for the federal government. Moreover, people at the end of their careers are earning more money and hence paying more taxes than young people, meaning life extension helping people in their 50s and 60s might be especially good for tax revenue.

The problem is that the CBO generally considers “how many workers paying taxes are there” to be an economic effect and only considers it in special “dynamic” scores of legislation, in which the economic consequences of them are taken into account. Dynamic scoring has been a topic of great controversy for decades, going back at least to the Bush II administration, but the rule Congress sets for CBO on when to use dynamic scoring results in CBO applying dynamic scoring very rarely in practice.

A middle ground option, though, would be something called “population change” scoring, in which CBO considers the direct effects of a change in the population (through longer lifespans, say, or immigration) on the level of employment and tax revenue, without doing a full, more complicated dynamic score. That would make its accounting of the effects of longer lives less biased: the budgetary benefits would be counted alongside the costs.

We should also consider the aspects of our budget situation that make the longevity effect a reality. One is the US’s long-standing, bipartisan choice to run massive budget deficits, even during relative boom times. One arithmetic consequence of that choice is that it makes the continued existence of every American a net loss for the country’s books. That’s not the main reason to avoid large deficits during booms, but it’s a somewhat toxic byproduct all the same.

The other aspect driving this effect is the choice to invest government resources very heavily in seniors relative to other age groups. This is due in large measure to the US choice to provide universal health care for seniors but not other age groups, and due to our lack of investment in very young children and working-age adults compared to other rich nations.

There is no law of nature saying the US has to weigh its priorities that way. As long as we do, the numbers will imply that it’s better for the budget for people to die before they get old.