The Weekly Sift

Conservative blogs often post a graph more-or-less like the one below, which I got from the blog of Keith Hennessy, who is currently at the Stanford Business School and used to be Director of the National Economic Council under George W. Bush. (So: not somebody I usually agree with, but probably not a dummy either.) He claims that the numbers were computed for him by Bush’s Office of Management and Budget in 2007. (So: probably not a fabrication.)

It looks bad. Taxes as a percentage of GDP have stayed in a relatively narrow band since World War II, only occasionally peaking over 20%. But starting in about 2016, spending as a percentage of GDP starts to take off, reaching the incredible level of 40% by 2080 with no end in sight.

The typical liberal response to this, which I have given myself, is not that graphs like this are wrong, but that they hide the real problem: Government spending goes out of control because healthcare costs go out of control. But just capping what the government spends on Medicare and Medicaid (i.e., the Ryan plan) doesn’t fix anything. If healthcare costs are unsustainable, then what does it matter whether we’re paying those costs through government, through private insurance, or out of our pockets? Personally, it’s all the same to me whether I go broke paying taxes, paying health insurance premiums, or paying my doctor.

So a liberal would rather imitate the countries who already get better healthcare for less money than we do and increase the government’s role, ideally with a single-payer system.

Summing up: Liberals and conservatives agree that we have a long-term problem, but they argue about what kind of problem: a government spending problem or a healthcare cost problem.

Recently I ran into a potentially game-changing question: What if there is no problem? In other words, instead of being trapped in the dismal liberal/conservative argument about which apocalypse we’re headed towards, what if we’re actually not headed towards an apocalypse at all?

“That’s crazy!” That was my first reaction too. I mean, look at that graph. But the guy making the claim (William Baumol in the recent book The Cost Disease: Why Computers Get Cheaper and Health Care Doesn’t) has a track record that earns him a hearing.

Baumol is an economist who is most famous for identifying Baumol’s Cost Disease in the 1960s. His observation is that although the economy as a whole becomes more productive with the advance of technology, not all sectors progress equally, and some don’t improve their productivity at all. For example, a 21st-century farmer feeds far more people than a 19th-century farmer. Likewise, a worker at a modern shoe factory makes more shoes than a 19th-century cobbler. But it still takes four talented musicians to perform a Beethoven string quartet, and they don’t do it any faster than they did in Beethoven’s day. String quartets have not seen a productivity increase.

The economic consensus of the 1960s said that wages were tied to productivity. If that were true, then classical musicians would have seen their incomes crash relative to farmers and shoemakers, who would by now be vastly wealthier than the lowly performers of the New York Philharmonic or the Boston Pops.

In fact that didn’t happen, because in the long run the labor market has a supply side as well as a demand side, the result being that every profession has to pay enough to induce talented people to make whatever sacrifices are necessary to enter that profession. But something has to give somewhere, so we see the productivity difference as inflation: The price of a New York Philharmonic ticket is going to rise much faster than the cost of a loaf of bread or a pair of shoes.

So Baumol’s observation is that industries with a large component of personal service are not going to increase their productivity as fast as the rest of the economy, and as a result those industries are going to see higher inflation than the economy as a whole. Year-by-year those higher inflation rates might just be a nuisance, but over time exponential growth works its dark magic: If two products each cost $1 today, but one is subject to a 2% inflation rate and the other 10%, in 100 years the low-inflation product costs $7.25 and the high-inflation product costs $13,781.

Health care. Health care has a high component of personal service. It does not have high productivity growth.

Now this part gets a little tricky, because we all know how much medical technology has improved over the decades. But the improvement is almost entirely on the outcome side rather than the productivity side. Adrian Peterson could tear up his knee and be better than ever the next season, where half a century before Gale Sayers was never the same. But the amount of attention patients need from doctors and nurses has not gone down. Health professionals are doing better for their patients, but they are not processing more of them faster.

And most of us wouldn’t want them to. If you heard that one local hospital had one nurse for every five patients and another “more productive” hospital had one nurse for every 50, which would you choose for your surgery? If one doctor sees 30 patients in an hour of clinic time and another doctor only six, which would you pick as your PCP?

So back in the 1960s, Baumol looked at this situation and concluded that medical inflation was here to stay. Not because doctors are greedy or health insurance companies are evil or socialized medicine is inefficient, but just from the nature of health care. While other factors undoubtedly matter, the exponential growth would happen anyway.

This is borne out by the inability of any country to tame medical inflation. France, for example, is often held up as a model healthcare system. But its costs are also rising exponentially.

Government spending. And it isn’t just health care. Government services in general tend to be in what Baumol calls “the stagnant sector” — not due to bureaucratic waste or the power of public-sector unions, but because the services themselves require one-on-one attention.

In education, we call productivity by another name: students per teacher. But nobody wants his third-grader in a 150-student lecture hall. Everybody’s happy when an hour of labor builds more cars or mines more copper. But it’s not necessarily a good thing if social workers, public defenders, parole officers, or cops on the beat handle more cases faster.

So Baumol predicts that over time government spending will rise as a percentage of the economy.

But we can afford it. So far this is just a different spin on Hennessy’s graph. But here’s the difference: In Baumol’s model, government spending isn’t crowding out everything else. As a society, we aren’t doing without manufactured goods because health care is soaking up all our money; we’re just using less of our labor to produce the manufactured goods we want.

Despite their ever increasing costs, stagnant-sector services will never become unaffordable to society. This is because the economy’s constantly growing productivity simultaneously increases the community’s overall purchasing power. … If governments cannot be led to understand the ideas presented here, then their citizens may be denied vital health, education, and other benefits because they appear to be unaffordable, when in fact they are not.

In other words, even though orchestra tickets cost more now than in the 1800s, it’s ridiculous to claim that past societies could afford orchestras and our far richer society can’t.

Think about food. America’s Farmers estimates that an American farmer today feeds 155 people. By contrast, in colonial times a farm family barely did more than feed itself. Imagine going back to colonial times and telling people that by 2013 the non-farm part of the economy would grow so much that it would force a single farmer to feed 155 people! They would undoubtedly picture some cancerous expansion in the non-farm economy that could only be checked by mass starvation.

But that’s not what happened. The non-farm economy came to dominate GDP, but we’re not starving. That 1 farmer is providing his 155 eaters with too many calories, not too few.

This conclusion — that our descendants will likely be able to afford more health care and education as well as more of all the other goods and services they consume — may seem strikingly implausible … if health-care costs continue to increase by the rate they have in the recent past, they will rise from 15 percent of the average person’s total income in 2005 to 62 percent by 2105. This is surely mind-boggling. It means that our great-grandchildren in the year 2105 will have only a little less than forty cents out of every dollar they earn or otherwise receive to spend on everything  besides health care — food, clothing, vacations, entertainment, and even education! Yet as this book will show, this prospect is not nearly as bad as it sounds.

There are many possible objections to Baumol’s argument. (I wonder how it’s affected by the way that wages in general have come unstuck from productivity.) But here’s the message that I take from his book: When someone presents a graph like Hennessy’s and acts like the conclusion is obvious — say, that government spending can’t reach 40% of GDP by 2080, and so some catastrophe will have to intervene before that point — don’t buy it without a more compelling explanation.

The economy of 2080 or 2105 will be different from today’s in many, many ways. Maybe current trends will persist until then or maybe they won’t. But you can’t conclude anything from the mere fact that some statistic from the far future looks implausible.

The far future is going to look implausible to us, if we manage to survive long enough to see it. That’s the one prediction I have complete confidence in.