Did the Fed, or Asian Saving, Cause the Housing Bubble?
Just about the only good thing to come out of the housing bubble is that many financial analysts are coming to see the virtue of the Austrian theory of the business cycle. Specifically, though Greenspan did his best to blame deregulation and foreigners who saved too much, many people now think that the Maestro's ultra-low interest rates in the wake of the dot-com crash may very well have sowed the seeds for our current crisis.
Ironically, at the very moment of the free-market economists' intellectual victory, some in our camp want to take away the champagne. Specifically, Tyler Cowen has repeatedly argued on his very popular blog that it was not the Fed but rather an increase in foreigners' savings and appetite for risk, that caused the boom. And in a recent Cato paper, David Henderson and Jeffrey Rogers Hummel defend Greenspan's record, going so far as to say that "Alan Greenspan stands out as the most competent—and arguably the only competent—helmsman of United States monetary policy since the creation of the Federal Reserve System."
Let me warn the reader that I am going to have nothing nice to say about this defense of Greenspan; I think his policies caused or at least made possible the housing boom. As I walk through the specifics of the Henderson and Hummel (H&H) defense, I conclude many of their statements are either misleading or outright falsehoods. Because my critique will be so harsh, I want to stress that I actually know H&H personally, and acknowledge that they are both intelligent and very courageous advocates of economic liberty. So all I can say regarding this particular Cato paper is that either they or I am suffering from a bout of temporary insanity—and the little green elf who hovers over my laptop assures me I'm not the one who's crazy.
Many Austrians Were On to Greenspan For Years
Since their objective is to defend Greenspan's overall record, it is appropriate that H&H remind everyone that "two years ago, on leaving office, Greenspan was widely heralded as a financial wizard whose wise, discretionary macromanagement had brought an unprecedented two decades of low inflation, high prosperity, and infrequent and mild recessions."
But then the fickle public turned on poor Alan—and all because of a pesky little financial crisis not seen since the Great Depression. Now all of a sudden the Maestro is the bad guy! H&H explain: "Recently converted critics are now charging Greenspan with having carried on an excessively expansionary monetary policy, particularly following the recession of 2001 and possibly during the dot-com boom that preceded it" (emphasis added).
To cater to those who cherish "falsifiable predictions" as the mark of a good economist, I want to point out that many subscribers to the Austrian theory of the business cycle were warning of a Fed-induced housing boom years before others realized there was a problem. Some of these prescient warnings came from Mark Thornton, Frank Shostak, Stefan Karlsson, Peter Schiff, and Robert Wenzel. (Sadly I myself was very late to the game, as I explain in this mea culpa.) With the benefit of hindsight, it is truly impressive to read the articles and watch the video linked above. These gentlemen hit the ball out of the park, and if the world were a meritocracy they would have governments and Ivy League universities beating down their doors.
The Fed Doesn't Really Control Interest Rates?
Let us return to H&H's defense of Greenspan:
Why do people now believe Greenspan was an "inflationist"? For one main reason: they note how low interest rates were from 2002 through 2004. But interest rates have never proved an adequate gauge of what the Fed is doing: not during the Great Depression, when rates were very low despite a collapsing money stock; not during the Great Inflation of the 1970s, when rates were high despite an expanding money stock; and not under Greenspan. A focus on interest rates not only obscures the well-known distinction between nominal and real rates (nominal rates equal real rates plus expected inflation), it also ignores the simple fact that interest rates can change as a result of real factors involving supply and demand.
The market ultimately determines interest rates. Although central banks are big enough players in the loan market (and the quintessential noise traders to boot) that they can push short-term rates up or down somewhat, that ability is increasingly diminished, even for a major central bank like the Fed, as globalization integrates world financial markets. In defending his actions, Greenspan is correct in attributing the unusually low interest rates early this decade mainly to a massive flow of savings from emerging Asian economies and elsewhere.
For the record, even after adjusting for price inflation, interest rates during the housing boom were at negative levels, and in fact were the lowest they had been since 1979, as I show in a chart in this response to an earlier H&H article. However, in subsequent email exchanges H&H have told me that they never meant to deny that interest rates really were low during the housing boom under Greenspan, even after adjusting for the lower price inflation of the 1990s and 2000s (as compared to the 1970s and early 1980s). Their main point was that the low interest rates were not caused by Greenspan's whims but rather by the global supply and demand for savings.
Before we examine this claim, let us pause to note that even if it were true, then H&H are defending either a lunatic or a liar. Alan Greenspan himself in his testimony over the years sure made it sound as if he had a lot to do with interest rates in the United States. He'd throw in a bunch of Greenspan jargon, of course, to keep the financial press swooning, but he sure seemed to buy into what most commentators believe, and what I and just about every other college professor taught our students in Intro to Macro, namely, the Fed can "stimulate" the economy by lowering interest rates, and this is accomplished through flooding the credit markets with more reserves (created during open-market purchases). But if price inflation starts getting uncomfortably high, the Fed can raise rates back up by selling assets and thus sucking reserves out of the banking system.
Now don't get me wrong. I am all for saying the emperor has no clothes, and that governments cannot trump economic laws. However, I still believe that central bankers really do exercise a large influence over the interest-rate structures in their respective jurisdictions.
It's true that the options at Greenspan's disposal in, say, 2000 were much different from the options facing Fed Chairman Volcker in 1980. But even though they are ultimately subject to supply and demand, the sad fact is that central bankers have the ability to create new money out of thin air. Yes, the federal funds rate is set in the marketplace, but when the Fed can create or remove billions in the supply of reserves, it can thereby exercise a large degree of control over the "market" rate of interest.
H&H are right to note that central banks don't create real savings, and so they can't alter the "natural" interest rate that would balance the supply of real savings with the demand for investable funds. But that's the crux of the problem: by pushing down the market rate of interest below this correct "natural" rate, central banks distort the market process and set in motion an unsustainable boom. By simply asserting that the market can set interest rates (largely) independently of central bankers' actions, H&H are really assuming away the very issue under discussion. So let us now examine whether this "savings-glut" hypothesis can really explain what happened.
Does An Increase In Savings Explain the Housing Boom?
In their Cato article, H&H don't actually provide empirical evidence for the claim that it was a foreign "savings glut" that caused the housing boom, though in a footnote they give several references; here is a good representative. Many of the proponents of the "savings-glut" hypothesis point to the "World Economic Outlook" put out by the IMF to make their case. (Specifically, look at Table A16 in this pdf.) The alleged smoking gun is that savings as a percentage of GDP rose sharply among emerging and developing economies just as the housing boom kicked into full gear. But even using the graphic they themselves construct, we see something puzzling:
Now before I showed you the above chart, weren't you expecting savings to shoot way up during the housing boom, and then fall back down when the bust kicked in? Yet as the chart above shows, savings rates among the economies in question rose throughout the boom and bust in housing. If the increased savings rates from 2001–2006 are supposed to explain the rapid appreciation in home prices, then one would think that the much slower increases from 2006–2008 would at best correspond to a stabilization of house prices, not their collapse.
Here's another way to show the puzzle. Check out the movement in 30-year conventional mortgage rates in the United States:
Conventional mortgage rates were pushed to their lowest levels during 2003, and then rose substantially up through early 2006. Isn't that odd, since the savings rate in the emerging economies continually rose from 2000–2008? We are being asked to believe that a V-shaped graph (mortgage rates) can be explained by an inclined plane (developing economies' savings rates).
Ah, but wait. I've got yet another curveball for you. Remember that H&H want to exonerate Greenspan by (correctly) pointing out that world capital markets have become far more integrated in the last decade than ever before. And also remember, they want to attribute low long-term interest rates in the United States to the global forces of supply and demand. Since that is their strategy, why should we be looking at the savings rate among emerging and developing economies? What happens if we look at the whole world? Why, I'd expect that the "global savings glut" would go hand in hand with an increase in the global savings rate, wouldn't you?
At first glance, it seems that you do find this: the global savings rate increases steadily, rising from 20.5% in 2002 up to 23.7% in 2007. But hang on a second: there's something strange here. The same Table A16 also shows that during the period 1986-1993, the global savings rate averaged 22.7%, and during the period 1994–2001, the global savings rate averaged 22.1%. I think it's best to reproduce the relevant line from the table, so the reader can see exactly what I am talking about:
Look at the table above, and consider what it means for the proponents of the global-savings-glut thesis. They are forced to argue that the biggest speculative bubble in world history was fueled by savings rates in the early 2000s that were lower than their average value over the previous 15 years. Then, the bubble started popping just around the time when world savings rates finally exceeded their 1986-93 average, in the year 2006. And as world savings rates continued to rise, the speculative bubble that they were fueling continued to collapse.
Does the Austrian Theory Fit the Facts Any Better?
Naturally, just because the global-savings-glut hypothesis has some trouble fitting the facts, that doesn't mean we should completely throw it out. After all, this is macroeconomics, not quantum physics. There's no such thing as a controlled experiment in the realm of the social sciences, especially when we're talking about global outcomes.
Even so, look at how much better the Austrian theory makes sense of the observations. The following chart shows year-over-year (i.e., twelve-month changes) in the famous Case-Schiller Home Price Index, plotted against the federal funds rate (the interest rate that the Fed targets):
The relationship between the two series is almost perfect in explaining the housing boom and bust. The annual appreciation in house prices was at its highest when interest rates were at their lowest, and then the price rises tapered off as the Fed's targeted rate came gradually up. Home prices started falling (i.e. the blue line crossed the 0% point on the left axis) once interest rates had stabilized at 5.25%. It's true, the housing boom started earlier than Greenspan's infamous rate cuts; from the graph it looks to have started in about 1996 or 1997. But that is also precisely when President Clinton exempted owner-occupied homes from capital-gains taxes (for the first $500,000), and so it's no wonder that these particular assets saw their prices rising at that point.
H&H claim that Greenspan, perhaps unwittingly, implemented the monetary rule that free-banking writers Selgin and White advocate. Well, this was news to Selgin himself. H&H also make two empirical claims, the first of which is very misleading (and self-serving for their hypothesis), and the second of which is simply false. First, the misleading claim. In defending their assertion that Greenspan "came close to freezing the domestic monetary base," H&H write:
Between December 1986, 8 months before Greenspan became Fed chairman, and December 2005, 19 years later, the monetary base rose by a hefty amount, from $248 billion to $802 billion (no figures are seasonally adjusted). True, that doesn't sound like a freeze. But virtually the whole increase was in currency in circulation…During that same time, total bank reserves grew from $65 billion to $73 billion, for an average annual growth rate of a mere 0.65 percent. (These figures are unadjusted for any changes in reserve requirements and—unlike the somewhat misleading reserve totals reported by the Fed's board of governors—include all vault cash, clearing balances, and float.) In some years aggregate reserves rose; in others they fell, with the major bump surrounding Y2K, when the accumulation of reserves by banks appears to have induced the Fed to accommodate a 40 percent jump followed by a 30-percent drop.
In a footnote, H&H explain how they calculated their preferred measure of "total reserves." I confess I did not bother reproducing their results. But if we instead look at the Board of Governors chart of "Total Reserves, Not Adjusted for Changes in Reserve Requirements" (and which are not seasonally adjusted), the picture shows that H&H's claim above is rather misleading in this context:
To repeat, the above chart is not what H&H are discussing in their text; they find the particular series that I graphed above to be "somewhat misleading" because of its treatment of currency etc. Even so, I find it very odd that H&H exonerate Greenspan from the housing bubble by claiming "he kept total reserves constant," when a look at the standard graph of "total reserves" shows them skyrocketing just as the housing boom really kicks in, and then falling as the boom tapers off. It suggests to me that their theory is flawed, and maybe solely looking at what happened to (adjusted) bank reserves isn't the way to judge whether Greenspan had anything to do with the housing bubble. (For one thing, didn't H&H just get through telling us that we are in a globally integrated capital market? So why would the condition of domestic base alone tell us whether the Fed were pushing down interest rates?)
Finally, we come to—what appears to be—an outright falsehood. Following on the block quotation above, H&H write: "Total reserves are also the one monetary measure that show a slight uptick into 2003, when interest rates were down."
I have now reread this passage at least eight times, and I have no idea how to square it with the truth. Every other monetary measure that economists use, rose sharply during the housing boom, including the year 2003. The only way I can make sense of H&H's statement is to assume they meant that all other monetary measures were growing at more than a slight uptick into 2003—and hence total reserves would be the only one showing merely "a slight uptick into 2003":
In the graph above, the reader should note that the axis for M1, the blue line, is on the right, while all other monetary aggregates are charted with reference to the left axis.
David Henderson and Jeffrey Rogers Hummel have tried to exonerate Greenspan from his alleged role in the housing bubble. However, to do so they must rely on novel transformations of the conventional monetary measures, in order to demonstrate "tight" Fed policy, when all of the conventional indicators show a very "loose" policy precisely when the boom was at its greatest. Furthermore, Henderson and Hummel offer an alternative thesis—a global savings glut—that hardly fits the most basic of facts. Global savings rates were lower for most of the housing boom than in the preceding decade, and global savings rates are higher now (amidst the bust) than they were during the boom that they allegedly fueled.
Putting aside the details, we can also step back and ask ourselves, does it really make sense that one of the worst financial crises to grip the world was caused because people started saving too much? Or does it make a lot more sense to blame it on huge central-bank injections of phony credit into the markets? How two economists who appreciate the strength of the free market could opt for the former hypothesis—despite the dubious evidence for it—is puzzling indeed.
Naturally, there is more to the story of the housing boom than simply saying, "The Fed chairman did it." A full explanation would involve the Chinese government, Fannie Mae and Freddie Mac, and the ratings agencies that gave high marks to dubious mortgage-backed securities. But the original Misesian insight has withstood the test: it still seems that the Fed was a necessary condition for the worst speculative bubble in world history.