Monday, July 4, 2016

Public debts, public wealth, public services

This is the first of what I hope will be a few posts on Thomas Piketty’s Capital in the 21st Century.

My guess is that there are a lot more people who are curious about the book than there are who have read it, so in my writing I won’t assume that you’re one of the ones who’s read it.

I’m also guessing that my modest readership is not majority economists, so I’ll aim to be comprehensible to people not versed in the linguistic conventions of our field.


The state plays a key role in Piketty’s Capital in the 21st Century.

The central fact of the book is the decline in inequality during the first half of the 20th century, the continued relatively low level after World War II, and the renewed increase after about 1980.

Income shares of the top 10%, top 1%, top 0.1%, and top 0.01% in the U.S.
Data from World Wealth and Income Database
Focusing in on just the top 0.1% and top 0.01%.
Data from World Wealth and Income Database
In grad school I was taught the Kuznets curve: as a society started to get rich, inequality increased, but then as it kept getting richer, inequality decreased again.

This was treated as sort of a natural regularity. In a primitive situation, nobody is very wealthy at all, and so inequality is low. As wealth increases, it accrues primarily to a few people, so inequality increases. But then past a certain point, a middle class develops and inequality comes back down.

In the charts above, the part from the beginning of the chart to the 1950s, when Simon Kuznets was developing the idea, show the period of declining inequality, with the earlier increase hidden outside the data.

Piketty changes Kuznet’s story in two ways. First, on the basis of the data he and his colleagues assembled, he argues that the normal situation is more like the high level of inequality found in the 19th century than the lower level of the mid-20th century.

Second, what brought inequality down and kept it down was the shocks of World War I, the Great Depression, World War II, and governments’ responses to those events. The wars and the depression destroyed some wealth, disproportionately affecting the wealthier members of society. The high taxes levied on large incomes to fight the wars and the social programs instituted in response to the Depression sustained the reduction in inequality. In the 1980s, rich countries started significantly reducing taxes on high incomes and trimming social programs, and inequality increased as a result.

So the role of the state is important, but I want to start in a slightly different place, with the trio of concepts mentioned in the title.

Public debt gets plenty of attention—concerns about it are a significant force in our politics—but I want to set it in the context of the broader economy, an exercise familiar to economists, but not usually part of the public discussion of the issue.

In contrast, people are somewhat aware of public services, but generally don’t have a clear idea of the role they play in the larger economy.

And public wealth gets almost no attention—it’s a concept I hadn’t really thought about before reading Piketty, so I’m guessing it’s even further from the attention of people whose job is not following the economy.

Let’s start with the public debt. The following graph shows the liabilities of all levels of government in the U.S. from 1913 to 2013—the government debt. (The data are taken from the World Wealth and Income Database, an incredibly valuable resource compiled by Piketty along with many colleagues, and made freely available to the public. For a full description, see

US public liabilities, billions of current dollars
It looks absurd: basically nothing for decades, then in the late 1960s it picks up speed, and after 2001 it really takes off.

One of the first things to do with data like this is to represent it in logarithmic form, so that you see rates of change rather than levels. When you do that, you get a very different story. We picked up debt with World War I, then a bit more during the Great Depression, and a lot more in World War II.

US public liabilities, logarithmic scale
If there’s something troubling about this second chart, it’s that after World War II we never went back to our earlier sober ways, but just kept piling on debt, just not as quickly as during the war.

Which leads to the third way of looking at the debt, which is to scale it by the size of the economy. Piketty’s database uses national income, and when you look at the debt as a share of national income, you get this:

US public liabilities, as a percentage of national income
You can see the increases with World War I, the Depression, and World War II, but for 35 years, from 1946 to 1981, our public debt made up an ever smaller part of the economy. For the 12 years of the Reagan and George H.W. Bush administrations it grew, for the 8 years of the Clinton administration it shrank, then for most of the George W. Bush administration and all of the Obama presidency it grew again.

What about public wealth?

“Wealth” in the sense Piketty uses it here is essentially “net worth”: the value of what you own, minus the value of what you owe. If you have a house worth $200,000, and a retirement account worth $50,000, then you own $250,000. You set against that the $80,000 you still owe on your mortgage, the $35,000 left on your student debts, and the $5,000 you owe on your credit card, and your net worth, or wealth, is $130,000.

You can do the same thing with government, but while we obsess about the debt, what government owes, we generally pay very little attention to what government owns.

It actually matters a lot that we overlook government’s assets, because they include basic items that are fundamental to the working of the economy. Things like roads and sewage systems.

Then there are the things that aren’t about the day-to-day basic functioning of the economy, but about minimal long-term well-being, such as schools so that the next generation can be literate, and public universities and community colleges, contributing toward the next generation being more than literate.

And there are police cars, fire stations, etc., that are used for the provision of public safety.

And aircraft carriers, military bases, etc., that contribute to national security.

And there are public amenities like parks, from the playground down the street to Yellowstone, offering chances for relaxation and enjoyment.

On top of all these tangible, physical assets, government can own financial assets as well.

Putting those all together for the U.S., we can see that the value of the government’s non-financial assets for most of the 20th century has been about the same as the size of the national income, and its financial assets have added another 20 to 30 percent of national income.

US public assets as percentages of national income
The data for the net worth of government in the U.S. go back to 1870, and since 1880 the government’s net worth has almost always been positive.

US public wealth, as a percentage of national income
The two exceptions were during World War II (when heavy borrowing to pay for fighting the war more than offset the acquisition of wealth in that same effort), and the aftermath of the 2008 crash (when heavy borrowing in response to recession coincided with a reduction in government assets).

Looking at assets and liabilities together, you can see that the current net-negative position is a significant change from the late 1970s, before liabilities started going back up, it is nonetheless small compared to the overall levels of the lines.

US public liabilities and assets, percentages of national income
Our government’s debt has been a political talking point for decades, but our public wealth has been comfortably positive since 1880, with the only two exceptions being the effort to fight World War II and the response to the worst economic downturn since the Great Depression.

Which brings up an interesting point. Our current crisis is bad, but arguably not as bad the experience of 1929-1939. GDP only fell for 18 months, not for four years; unemployment only reached 10%, not 25%. Yet the current crisis has seen continual decline in U.S. net public wealth since 2007, with the level running into negative territory in 2011. In contrast, net public wealth stayed positive throughout the Great Depression.

If you look at liabilities and assets in constant-dollar terms, you can see what was different.

US public liabilities and assets, constant dollars, logarithmic scale
During the Great Depression, government in the U.S. increased its liabilities—it borrowed money—which was a big change from the flat level of government liabilities throughout the 1920’s. But it also built things: its non-financial assets kept increasing right through the Depression and on through the war.

In response to the current mess, our liabilities had already been increasing since 2001, and we merely accelerated our rate of borrowing after 2008. But our non-financial assets peaked in 2007.

When the Great Depression hit, we borrowed and built. When the Great Recession hit, we borrowed and … bailed out banks (probably necessary in some form, but we could have done it by bailing out underwater home-buyers) and extended unemployment insurance (a good idea under the circumstances, but it left nothing tangible in its wake). What we didn’t do is embark on a serious effort at rebuilding public infrastructure.

These charts, useful as they are (at least for someone who likes visual representations of data), leave plenty of unanswered questions. An important one, given the political heat around government deficits, is the issue of what it is that government is doing with all that spending.

The short answer is: public services.

There’s a classic way of dividing up economic activities into three types:
  • agriculture (which is sometimes expanded to include other harvesting activities, such as logging and fishing);
  • manufacturing (physically making things); and
  • services (using the products of agriculture and manufacturing to do things for people).
A regularity of this classification is that agriculture’s role in the economy (measured by the value of output or by the share of people employed in it) has been shrinking for a couple of centuries, manufacturing’s role grew during the 19th century but shrank during the 20th, and services have become dominant. In rich countries today, services account for 70-80 percent of employment.

Piketty observes:
Note that an important part of these services, especially in health and education, is generally financed by taxes and provided free of charge. The details of financing vary from country to country, as does the exact share financed by taxes, which is higher in Europe, for example, than in the United States or Japan. Still, it is quite high in all developed countries: broadly speaking, at least half of the total cost of health and education services is paid for by taxes, and in a number of European countries it is more than three quarters. This raises potential new difficulties and uncertainties when it comes to measuring and comparing increases in the standard of living in different countries over the long run. This is not a minor point: not only do these two sectors account for more than 20 percent of GDP and employment in the most advanced countries—a percentage that will no doubt increase in the future—but health and education probably account for the most tangible and impressive improvement in standards of living over the past two centuries. Instead of living in societies where the life expectancy was barely forty years and nearly everyone was illiterate, we now live in societies where it is common to reach the age of eighty and everyone has at least minimal access to culture. [p. 91-92]
This doesn’t speak to the issue of public assets, but in a very real sense it is wealth, and as Piketty explains in the paragraph above, it’s wealth created by public expenditure, to a significant extent.

An avid proponent of the market would object that this position overstates the usefulness of government, but there’s a counter-argument to that. The first piece is to understand how government services are accounted for:
In national accounts, the value of public services available to the public for free is always estimated on the basis of the production costs assumed by the government, that is, ultimately, by taxpayers. These costs include the wages paid to health workers and teachers employed by hospitals, schools, and public universities. This method of valuing services has its flaws, but it is logically consistent [I might say “methodologically consistent,” or “operationally viable” - KS] and clearly more satisfactory than simply excluding free public services from GDP calculations and concentrating solely on commodity production. It would be economically absurd to leave public services out entirely, because doing so would lead in a totally artificial way to an underestimate of the GDP and national income of a country that chose a public system of health and education rather than a private system, even if the available services were strictly identical. [p. 92]
Here’s what that looks like in the data for the U.S. I’ve focused on four categories of government outlays (all levels of government combined): health, education, retirement (mostly Social Security), and other income security (that is, other than retirement: unemployment insurance, disability, welfare, and assorted other programs).

You can see that all four of these categories have increased since the beginning of the data, though the change is most pronounced and continued in health, which jumps in the mid-1960s when Medicare and Medicaid were created, and has continued rising since then. The other categories are higher now than in 1959, but on balance they haven’t changed a lot since the mid-1970s.

Health expenditures keep growing; other parts are large but stable

Looking at the next chart you can see that these “public service” and “social insurance” functions have risen fairly continuously, from 35% of government in 1959 to over 60% today.

Social insurance and education have become the dominant activities of government
For the federal government specifically, which doesn’t spend so much on education but carries most of the health and retirement expenditure, the story is even more dramatic. In 1959, health, retirement, and other income security accounted for 25% of the federal government’s outlays; in 2014 it was 61%.

Most of what the federal government does is social insurance, with a significant sideline in defense.

This is the reality behind the quip (apparently by Peter Fisher, undersecretary of the Treasury in 2002) that the US government “is an insurance company with an army.” (And over the last 55 years it’s gotten to be more of an insurance company, and less of an army, at least in terms of shares of the budget.)

Returning to our avid market proponent, he/she might counter that services provided by the government are inferior. Piketty is having none of this:
The method used to compute national accounts [measuring the cost of government-provided services by the cost of providing them] has the virtue of correcting this bias [that would come from omitting such services from the GDP]. Still it is not perfect. In particular, there is no objective measure of the quality of services rendered (although various correctives for this are under consideration). For example, if a private health insurance system costs more than a public system but does not yield truly superior quality (as a comparison of the United States with Europe suggests), then GDP will be artificially overvalued in countries that rely mainly on private insurance. … It may be that this method of accounting by costs underestimates the fundamental “value” of education and health and therefore the growth achieved during periods of rapid expansion of services in these areas. [p. 92]

When you look at how the government’s provision of health, retirement, and education is a growing share of the economy, it’s easy to spin a story about government out of control, expanding without limit.

The government spends 20% of our GDP providing health insurance, retirement benefits, other forms of income security, and education.
But if we think about the causes, we get a very different story.

First, we have a growing share of older people in the society, which drives up the amount of money we need to spend on retirement benefits and the amount we spend on healthcare, since we tend to use more medical services as we age.

Second, it’s not crazy to have government carry a significant amount of that burden. If we reduce government’s role in providing those services, that doesn’t make the need for them go away; either the private sector picks up the slack, or the services don’t get provided. And government can be a relatively efficient provider of both retirement benefits and health insurance, so if we reduce government’s role, we probably can’t save money through greater efficiency; we can only save money by having less.

Third, cost containment in health care isn’t necessarily a sensible goal in itself. It’s true that there’s plenty of room for providing health care in ways that are more cost-efficient (for example, see Atul Gawande’s famous piece on health costs in two Texas counties); part of the Affordable Care Act is designed to try to reduce those sorts of inefficiencies. But our society is getting richer, even if those gains aren’t trickling down very far. And as we get richer, it’s not crazy to put more effort into keeping ourselves healthy and increasing the length of our healthy lifetime. In other words, there’s nothing inherently wrong with spending a greater portion of our incomes on health care as we get richer.

(And there's the phenomenon of people having very strong opinions about what the government shouldn't be doing, while not having a clue what the government actually does:)
Sources clockwise from right: here, here, and here

As always in a democracy, we should be paying attention to what the government is doing. But the growth of government-provided services is not in itself a cause for a freak-out.

And when we’re talking about the public debt, we won’t get anywhere constructive if we don’t also consider the ways the government supports the economy in the things that it builds (and then owns) and in the things that it does.

We also need to think about the different ways government can obtain the funds it needs in order to play its role, but I hope to touch on that in a future post on the role of the rentier.

No comments:

Post a Comment