Federal revenue in 2013
Jason Smith has a post that points to a very interesting fact about 2013—federal revenue from asset sales soared, primarily due to dividend payments by Fannie Mae and Freddie Mac. By my calculations the revenue from asset sales increased by about $111 billion, from $53.6 billion in 2012 to $164.7 billion in 2013. This is certainly a non-trivial portion of the $500 billion in deficit reduction that occurred during 2013, and in my view offers a far better argument than pointing to things like state and local government (which is just as endogenous to the Federal government as private investment.)
I don’t claim to understand the Keynesian model, so I won’t even pretend to comment on the importance of these payments. Until recently I thought Keynesians viewed deficit reduction done via reduced transfers and higher taxes as being “contractionary.” You know, the sort of austerity that you see in Greece. Now I’m not so sure what they think. FWIW, at the time Matt Yglesias thought these GSE dividends were a contractionary “disaster.”
The only problem is that this gusher of federal revenue is actually an economic disaster.
In normal times, government coffers filled with dividends would be good because they could be put to some use. The government could spend that money on building Hyperloops or repairing schools or vaccinating children. Alternatively, the government could do the exact same things it was doing before, but reduce taxes and put more money in working peoples’ hands. But that would require a functioning political system. Today’s gridlocked Congress isn’t doing anything with the money.
Still, under ordinary circumstances the reduced government borrowing that results from a dividend windfall could be useful. A smaller deficit often allows the Federal Reserve to run lower interest rates without sparking inflation. That makes it easier for people to buy houses or for firms to invest in new production. Today, though, the Fed’s preferred measure of inflation is running at its second-lowest level on record, even though short-term interest rates have been at zero for years now.
So the Treasury is earning tons of Fannie/Freddie money. But the profits aren’t letting us spend more, they aren’t letting us tax less, and they aren’t freeing up private investment capital either. They’re doing nothing. It’s as if the money were sitting around as cash in a storage locker somewhere.
PS. There are all sorts of “Ricardian equivalence” arguments as to why Matt might be wrong—that those payments to the Federal government don’t matter. But do Keynesians believe in Ricardian equivalence?
PPS. Yglesias also has an excellent new post on the Chinese stock market:
On the other hand, I should say that when I went to China in 2008 I heard from a lot of smart foreign observers that the country was in the midst of an unsustainable stimulus-driven boom that would surely crash someday soon. Now it’s seven years later, and all the smart foreign observers say China is in the midst of an unsustainable stimulus-driven boom that’s in the midst of collapsing. And since no country goes forever without an economic contraction, surely China really will see its long boom come to an end and the economy fall into recession one of these days. Maybe even tomorrow!
But it’s dangerous to be too confident you know what’s going on. Nobody really predicted the boom that’s unfolded over the past six months, so nobody really knows what the future holds.
Tags:
27. June 2015 at 10:54
Here’s the deficit chart adjusted for changes in unemployment and asset income:
http://research.stlouisfed.org/fred2/graph/?g=1lzn
For comparison, adjusted only for unemployment:
http://research.stlouisfed.org/fred2/graph/?g=14Kr
27. June 2015 at 12:36
What is the difference between Freddie and Fannie making transfers to the government via dividend payments and transfers in the form of tax payments? What difference should it make which label one attaches to those payments? If you think that those dividend payments shouldn’t factor into the austerity formula, then taxes shouldn’t either. Isn’t the source of the payment and the recipient exactly the same? Or, is it the fact that these payments are “one-off” that makes all the difference?
27. June 2015 at 13:03
So is it contractionary for the treasury to send coupon payments to the fed and have them recycled to the treasury as fed dividends ?
Would it be stimulative for the treasury to simply declare a higher interest payment to bond holders? Is offering short dated debt instead of long dated debt contractionary ?
BTW, on the latter point the term structure of publicly held debt has changed dramatically.
This is basically the opposite view of the one that drove operation twist…
27. June 2015 at 14:39
I have the same question also Vivian. Could the answer be that the Fed basically bailed out the GSEs with purchases using created reserves? Therefore payments from the GSEs to the treasury after that really acted in a similar way to the Fed monetizing some of the deficit?
27. June 2015 at 17:15
The Fed shoveled $100 billion to the Treasury also.
27. June 2015 at 18:30
Sumner: “I don’t claim to understand the Keynesian model…” – yet he is willing to do an entire post on it?
As for Ricardian equivalence, it’s subject to all kinds of conditions, and is limited as well by bounded rationality. As for asset sales, an eminent economist and I corresponded by private email once, and the conclusion was that while it’s true an asset sale could solve a lot of deficit problems ‘painlessly’ (as the US Republicans found during the Reagan years), it’s politically difficult to do. I would imagine the Keynesian model treats an asset sale in a neutral manner, since no ‘money illusion’. But I don’t know (though you would think Sumner should know).
Seems economists play such games with their models that even fellow economists don’t know what they’re talking about. (Job) security by obscurity.
27. June 2015 at 19:04
On the other hand, I should say that when I went to Greece in 2008 I heard from a lot of smart foreign observers that the country was in the midst of an unsustainable stimulus-driven boom that would surely crash someday soon.
28. June 2015 at 06:24
E. Harding, Not sure what that is showing.
Vivian, I’m also confused. In the simple Keynesian model it makes no difference. In a more complex version there might be some sort of important distinction.
Jon. The Fed is a de facto part of the Federal government, so I don’t think it makes much difference who holds Fed earnings.
Jerry, I don’t know enough about the details of the bailout to comment.
Ben, See my answer to Jon. The real drain was when the Fed pulled $100 billion out of the private sector. But what makes the GDP dividends different is that they increased by over $100 billion in one year. The Fed’s earnings did not increase sharply in 2013.
Ray, I certainly know what the textbook model implies, but the real world Keynesians don’t seem to follow that model, unless it’s convenient to do so.
TallDave, Which shows that you just never know.
28. June 2015 at 07:15
OT- why Sumner refuses to cite an econometrics article that shows the Fed controls the economy. First and foremost, it doesn’t exist. Second, monetarists are like Austrians in that they eschew evidence for ‘deductive reasoning’, see below. To his credit however Sumner still lets me post here, most bloggers would have censored me long ago, as DeLong and Krugman (his editors) do. – RL
ECONOMIC METHODOLOGY by TIMOTHY A. WUNDER in 21st Century Economics: A reference handbook (2010) University of Texas at Arlington… “The monetarist ideas were different from the ideas of the neoclassical synthesis; however, the general methods still remained relatively the same. Starting with the foundations of the neoclassical school, there have been consistently four major components to economic method. The economy is seen as a mechanistic system with general equilibrium being the outcome. The basic unit of economic analysis is the individual, not groups. Deductive reasoning based on only limited observations is the best way to create economic theories, and internal consistency is to be valued over external evidence.”
28. June 2015 at 08:34
Scott, exactly they are integrated, so if the Fed receives them the public doesn’t. How are coupon payments by the treasury different from other transfer payments for the purposes of the fiscal multiplier.
28. June 2015 at 11:05
Some comments in the thread, including one from Jason, were deleted.
28. June 2015 at 17:42
This is the primary 2013 did not disprove Keynesian. Most of the deficit reduction was a lot things going right versus actual policy decisions. (And yes the Sequester and ending extended Unemployment were necessary.) Frannie, Freddie and the Fed handed the US a lot more money. (In reality I disagreed Matt on his analysis because the US lost a bunch on the original bailout of Fannie & Freddie.)
29. June 2015 at 11:33
@ Federal expenditures divided by Federal revenue adjusted for unemployment. It’s easier than deficit as a percentage of potential GDP because nobody knows what potential GDP is. The first one excludes Federal asset income, the second includes it.
29. June 2015 at 12:30
Ray, If you aren’t going to be goofy, why would we bother to read your comments?
Jon, I don’t know.
Collin, And the evidence in favor of Keynesianism is?
30. June 2015 at 11:38
[…] Scott Sumner is running victory laps, over those broken records who called it a “bubble” but didn’t give the precise […]
30. June 2015 at 16:25
While a better argument, I’m not sure this changes the story all that much. Is the dividend payment from the GSEs to the federal government economically any different from tax payments from a corporation? Should we take the multiplier to be zero for this payment from a Keynesian perspective? I’m open minded on this particular topic as I haven’t thought much about it.
Still, if you take $111 billion of deficit reduction out as having a zero multiplier, there’s still $389 billion needed to account for (yes, some deficit reduction typically occurs in an economic expansion, but if you’re Keynesian you also need to take into account the multiplier effect of exogenous fiscal contraction).
We also don’t have to speculate as to whether or not much of the decline in the deficit was endogenous or exogenous. There were a number of well known actions to limit the size of the deficit. Military spending was being cut. Spending on automatic stabilizers was going down. FICA taxes went up 2%. The Bush tax cuts for high earners expired.
The CBO ran the numbers in August 2012 using a Keynesian model and projected GDP to fall 0.5% during 2013 Q4/Q4 and unemployment to rise to 9.1% by 2013Q4. That compares with GDP growth of 3.1% and unemployment falling to 6.7%. Granted, the CBO was running the numbers inclusive of allowing all the Bush tax cuts to expire and not just for high income households, but if you ask me, it’s a dramatic miss given that we got ended up with most of the baseline deficit reduction yet with growth/unemployment performing better than the alternative fiscal scenario which eliminated most of the deficit reduction ($810 billion deficit projected for FY2015 vs. $412.0 billion over the last 12 months).
Is it enough to prove the Keynesian model false? No.
But the basic storyline which remains is that the Keynesian model requires that 2013 was on track to be a (quite uniquely, for this expansion) a red hot year of economic growth with at something like 4%+ real GDP growth ex-fiscal action. This compares with GDP growth of 2.6% from 2009Q3-2010Q4 (when ARRA stimulus was in full effect), 1.7% in 2011 Q4/Q4, 1.6% in 2012 and 2.4% in 2014, and with trend real GDP growth clearly in the mid 1% range so far this decade.
I’m skeptical.
1. July 2015 at 14:35
Justin, Every time they add another epicycle it makes the model seem more and more ad hoc. I don’t even understand the rationale for the multiplier. Is it still the Keynesian cross?