David Henderson has a recent post discussing the likelihood of the US government defaulting on its debt, at some point in the future:
That motivated me to go back to an article that San Jose State University economist Jeff Hummel and I had published in Independent Review back in 2014 that, for some reason, I don’t seem to have posted about here. It’s titled “The Inevitability of a U.S. Government Default.” In it, we argue that the feds are likely to default, that money creation as an alternative is not likely to get them out of their fiscal fix, and that default is actually better economically than massively high inflation.
There is much that I agree with in their paper, but in the end I am somewhat skeptical of the claim that the US government will default. In the paper cited above, David and Jeff make this comment:
Nevertheless, the spending increases in the three federal programs highlighted–
Medicare, Medicaid, and Social Security–cannot go on forever. As one of author Henderson’s previous bosses, Herb Stein, put it, “If something cannot go on forever, it will stop.”Because these spending increases won’t go on forever, they will stop. How will they stop? Of the answer to that, we are less sure. A reasonable guess is that eligibility for Medicaid will be tightened, and Medicare and Social Security will be means tested, all well before 2050.
But if these reforms are not made well before 2050, then a very likely outcome is a government default on the federal debt. The default could range from outright repudiation to partial repudiation.
This is precisely why I think default is unlikely. The fiscal trends in the US are unsustainable. They are unsustainable even if the US government defaults. Thus I see two options:
1. Stop running excessive budget deficits, and do not default.
2. Wait until immediately after defaulting to stop running excessive budget deficits.
Obviously the first option is much better, as default solves no problems, and causes much future distress. So why not avoid default?
A cynic might point to Greece as a cautionary tale. Perhaps the government will not cut back on deficit spending until it is forced to. A cynic might also point to the current trajectory of fiscal policy, which has become extremely irresponsible over the past 12 months—indeed worse than anything previously seen in America (when taking account of where we are in the business cycle.) So why am I less cynical?
The Greek government hid its fiscal problems through dodgy accounting tricks. The US fiscal problem is too big to hide. Thus the bond market would begin to show signs of worry long before a default was imminent. And this would put pressure on the federal government to slow the increase in spending.
Note that it does no good to say that both tax increases and spending cuts are politically impossible. Not only are they possible, they are certain to occur for precisely the reason provided in the Herb Stein quotation. Some combination of tax increases and entitlement reform will occur during the 21st century. The only question is when. Because doing this reform before a default is a far more sensible and far less painful option, I continue to see that outcome as the most likely, while acknowledging that default remains a possibility if the political system remains highly dysfunctional.
READER COMMENTS
Carl
Apr 20 2018 at 11:45am
“Obviously the first option is much better, as default solves no problems, and causes much future distress. So why not avoid default?â€
I think default is likelier because of the dynamics of blame.
If on responsible party (or a group of responsible politicians in both parties) reduce deficits before default, the other party or their primary opponents will blame them for the tax increases and spending cuts.
If the US defaults, politicians can begrudgingly reduce deficits and blame speculators/foreigners/free markets/whatever for forcing their hands.
Philo
Apr 20 2018 at 1:23pm
The bond market vigilantes will have brought enough pressure to bear on the politicians to make them tighten up long before we reach default.
Ray
Apr 20 2018 at 3:34pm
Agree with Carl. There are no individuals with true skin in the game, and austerity is an incredibly difficult political proposition.
The American people don’t truly care enough either (well, they care, but just don’t touch “my” goodies).
Charlie
Apr 21 2018 at 12:27am
A market monetarist would have to say that wouldn’t they! As the market shows no sign of default risk in treasuries. And, I agree.
Scott Sumner
Apr 21 2018 at 4:26am
Carl, Yes, I concede that’s possible, but I think it unlikely.
Ed Hanson
Apr 21 2018 at 9:14am
Scott
I was surprised to see government debt projections had Medicare growth as the biggest, by far, problem. Medicaid and Social Security, my choices,have growth per GDP was quite small.
Medicare is the hole and its time to quit digging. Face the fact that the government is a terribly inefficient provider of health care.
Declare a cut off age, those younger, no Medicare. Instead issue a bond for their contribution toward Medicare they have made, to be used to supplement the cost of long term medical care. Those above the cut age should have the option of a bond or Medicare. Costly now but should relieve any future default.
Ed
bill
Apr 21 2018 at 11:22am
I definitely agree that the US is extremely unlikely to default (see current CDS pricing as confirmation), at least due to expenses. If we do default, it will be due to political grounds (a choice). For instance, Social Security already “balances” under current law. Current projections have the Trust Fund running out in 12 or so years. At that time, under current law, then current benefits automatically drop to match then current revenues. To pay more benefits than the then current revenues (or to increase the taxes or some combination of the two) would require the Senate, the House and the President to all change the law. Medicare will run into similar issues. We can bring in death panels then. Or raise taxes. Or change the law and just start borrowing more. The important distinction is that Medicare will not go BK the way any other insurer would and Social Security can not implode like any other pension plan. They don’t have a contract with beneficiaries (the Supreme Court has already ruled that our contributions to these programs do not give us a legal claim to benefits). I’ll add that our demographics are better than a majority of the other OECD countries, so until I see one of them (one that controls its own currency in particular) default, I will not expect the US to default.
Todd Kreider
Apr 22 2018 at 7:21am
“They are certain to occur”
Scott makes the same mistake that most social scientists do on this topic — assume that technoogy of the 2030s and 2040s, including medical, will look about the same as today. Tyler Cowen is a good example of an economist holding this myopic view as he said in 2009 that there won’t be much innovation or productivity growth until the 2040s(!) Cowen also said in 2011 that there won’t be any medical breakthroughs before 2030. A few years later, CRISPR became a household name and stem cell therapies are only two or three years away from becoming widespread treatments with effective health pills just around the corner to help prevent cancer, heart disease, stroke and even eliminate Alzheimer’s disease.
Scott Sumner
Apr 22 2018 at 8:21am
Todd, You said:
“Tyler Cowen is a good example of an economist holding this myopic view as he said in 2009 that there won’t be much innovation or productivity growth until the 2040s(!)”
So far he’s been right, as productivity growth has been quite slow since 2004.
mike davis
Apr 22 2018 at 9:23am
Scott: You’re surely right about Tyler being right but only if you’re sure that productivity growth is properly measured. I’m sure that it is not. I’m less sure whether the mis-measurement of productivity means that growth since 2004 has been ok. And I’m especially confused about whether or how we should measure changes in leisure productivity. (Suppose, for example, the internet and video games make it possible for adolescent boys to have much richer leisure-time experiences with much less money. That would be very hard to measure and even if we could, many would argue that the pleasure the kids get from killing virtual mutants and watching porn just shouldn’t count.)
Todd Kreider
Apr 22 2018 at 9:51am
Scott,
Cowen’s predictions begins the year after he states them. For productivity, that is 2010, and he has 23 years left to go:
2010 to 2017 has averaged 1.0%. Here is the historical picture:
1950s: 2.8%
1960s: 2.9%
1970s:-0.2%
1980s: 1.8%
1990s: 2.0%
2000s: 2.6%
2010s: 1.0% (?)
1950s, 1960s, 1970s: 1.8%
1980s, 1990s, 2000s: 1.9%
Cowen’s view of the future since “the low hanging fruit has been picked.(It wasn’t low hanging for past inventors/innovators):
2010s, 2020s, 2030: 1.0% max
Three weeks ago, Noah Smith made a similar argument about Japan, but it extends to the US and EU as well:
“And in the U.S., the 1980s were already giving rise to computerization and the internet, which would ultimately power high productivity growth in the 1990s and early 2000s; Japan, in contrast, doesn’t obviously have any such breakthrough technology in the pipeline.”
(Note that productivity was no higher in the 1990s than in the 1980s, nor was GDP per capita growth.)
Japan doesn’t have any breakthrough technology on the horizon except for revolutions in energy, health care and medicine, robotics, virtual reality and rapidly improving A.I. systems.
If you, Smith, Cowen and add Krugman, etc. completely ignore rising virtual reality, robotics, A.I. that beat the top Go player, stem cell rejuvination, the emerging heath pills and CRISPR type technologies, then you have a point.
bill
Apr 22 2018 at 3:09pm
Scott,
How can we accurately measure productivity growth if we can’t measure inflation?
From your post of 12/7/2014 (and this makes total sense to me):
The government tries to use “hedonics” to adjust for quality changes, but in truth there is no solid theoretical support for their attempts. It’s all a big mess.
What does it mean to say that product X is 47% higher quality than product Y?
Patrick
Apr 23 2018 at 1:07am
“Thus the bond market would begin to show signs of worry long before a default was imminent.”
Reinhart, Rogoff, and Savastano’s “Debt Intolerance” (2003) shows this claim to not always be true.
What odds would you put on financial repression as an alternative to explicitly addressing taxes or entitlements?
Justin
Apr 23 2018 at 10:14am
In a fiat money economy, very high deficit and a very high debt burden (expressed as a % of GDP) are infinitely sustainable if rates are sufficiently low.
3% annual NGDP growth (1.5% real growth, 1.5% inflation) and 6% annual deficits produce a long term debt ratio of 206% of GDP. If bond rates average 2%, that’s 4.1% of GDP spent on interest expense, meaning the following tax/spending levels are sustainable:
27% of GDP spending
22.9% of GDP non-interest spending
21% of GDP taxation
Or take a very slow growth economy targeting 0% inflation. 1% NGDP, 5% deficit, 0% inflation, 0.5% bond yield. It produces a long term 505% debt to GDP ratio but only 2.5% of GDP is spent on interest expense.
25.4% of GDP spending
22.9% of GDP non-interest spending
20.4% of GDP taxation
To the degree the bond market gets jitters, all the Fed has to do is step in with temporary purchases (sort of a Greenspan put for the bond market) and bond investors will realize that selling into a rate spike is a good recipe to take a loss while not being able to enjoy the subsequent upswing as they are no longer invested.
I don’t see such situations as ideal (for philosophical reasons I prefer a balanced budget when unemployment is in the 4.5%-5.0% range and surpluses when higher), but I don’t see default as inevitable even if deficits are high and persistent.
Mike W
Apr 23 2018 at 10:39am
H/H’s claim in their paper appears to be out of date:
Overall federal government spending, including interest on the debt, could exceed 40 percent of GDP by 2050.
According to the CBO’s most recent Long-Term Budget Outlook:
All told, under CBO’s extended baseline, federal spending would increase from today’s 21 percent of GDP to 23 percent in 2027 and to 29 percent by 2047.
Comments are closed.