This will be the last time I post an excerpt from my manuscript, although I will add some graphs next week. But first a few comments on items in the blogosphere:
1. Sweden’s Success: Yglesias and Mulligan are both right (and Krugman’s wrong.)
I came across this interesting item in Matt Yglesias’ blog.
My rule of thumb for thinking about the global recession is that whenever you hear claims that some country has weathered it unusually well because of Favored Policy Initiative A, you ought to first ask yourself if it’s not really just an exchange rate issue. That seems to be the story of Israel’s relatively mild recession and I don’t think any effort to explain a country’s successs—especially a small country—that doesn’t take this into account isn’t very credible.
For example, Casey Mulligan seems to think that tax cuts and reductions in the size of the social safety net explain why Sweden’s unemployment rate hasn’t increased as much as Denmark’s. A different theory would note that though Denmark and Sweden are both small, open economies with generally high taxes and generous social expenditures, Sweden’s currency floats whereas the Danish Kroner is pegged to the euro. Consequently, the outbreak of the recession was associated with a substantial reduction in the price of the Swedish Kroner relative to the DKK/euro, the dollar, or the yen:
Despite being a non-economist, Yglesias has very good instincts regarding the importance of AD policies. Then I checked the Mulligan post, and found this quotation, which is actually from Stefan Karlsson’s blog:
“In Sweden, the centre-right government elected in 2006 has implemented significant income tax reductions (and payroll tax reductions), particularly for low earners while at the same time, unemployment and sick leave benefits have been cut and it has become much more difficult to be able to receive such benefits.
In Denmark, by contrast, even though it too has a formally “centre-right government” it has largely refrained from reducing taxes or government hand-outs, and it has more specifically lacked the focus of their Swedish counterparts in strengthening incentives to work by reducing unemployment benefits and taxation of low income earners.”
This is the sort of policy mix that I keep advocating; efficient supply-side policies, lower payroll taxes to offset nominal wage stickiness, and expansionary monetary policy (the Swedish kroner depreciation that Yglesias mentioned.) In contrast, Krugman has argued that lower unemployment benefits and reduced labor costs may reduce AD. And those effects can’t be offset with monetary policy because we are stuck in a liquidity trap. Somehow the Swedes found a way to adopt a more expansionary monetary policy than Denmark, despite near-zero rates. So how did things work out for Sweden as compared to Denmark, which didn’t follow my policy mix? Here’s Mulligan again:
Denmark’s employment has fallen 5% in the past year while Sweden’s is unchanged.
My favorite welfare state let me down last year.
2. Population growth and liquidity traps.
I hate it when I discover new blogs full of interesting stuff that I really need to read. Here’s Karlsson on Australia:
I have frequently discussed how growth in the Australian economy is driven by rising commodity prices. There is however another factor driving growth, namely high population growth.
Australia’s population grew by 2.1% in the year ending September 2009, a lot higher than in most other advanced economies (typically population growth is less than 1%, and even negative in for example Germany and Japan). This has a particularly positive effect on the housing sector, which continues its long boom, despite high prices and interest rates that are higher than in most other countries.
As this article points out, the rapid population growth has contributed to a housing shortage, something which implies that both construction activity and house prices will continue to increase.
As you may know, Australia is the only developed economy to avoid a recession last year. They also avoided a liquidity trap. And they avoided a big housing crash, despite having the sort of price run-up that we saw in this country. A population growth rate of 2.1% will support a fairly vibrant housing market. It also boosts NGDP growth. I’ve argued that the key to liquidity traps is not inflation/deflation, it is the rate of NGDP growth. This is why China never came close to a liquidity trap, despite deflation in the late 1990s. Australia’s population grows more than 1% faster than the US, and more than 2% faster than Japan. It’s interesting that Japan has had a long-lasting liquidity trap, we are in what will be a shorter liquidity trap, and Australia avoided it entirely.
Asset prices are forward-looking, thus Australian housing prices are presumably supported by not just current population growth, but also by expected future population growth. Population growth in the US has recently slowed down. Part of the slowdown appears to have been caused by a crackdown on illegal immigration after Bush failed to get immigration reform through Congress. And when did this crackdown occur? Hmmm . . . it was in 2007. In retrospect it might not have been the optimal moment to send out a signal to property markets in the Southwest that Hispanic immigration was about to slow sharply.
After seeing what happened to poor Mankiw, I want to emphasize that there are much better ways of avoiding liquidity traps than encouraging illegal immigration. Nor do I think immigration was the biggest factor in the housing crash—tight money after August 2008 was the biggest factor. As Mankiw said, I’m just trying to draw attention to some interesting facts.
BTW, at the end Karlsson mentions Ireland and Spain, which also had high rates of immigration before their housing markets tanked. The difference, of course, is that they lack an independent monetary policy. If all of Europe experienced population growth rates as high as Spain and Ireland, the Eurozone might well have avoided falling into a liquidity trap.
3. China saved the world economy.
Here is Tyler Cowen quoting Jim O’Neill:
As far as China’s involvement with the rest of the world goes, the real story since the worst of the crisis is not China’s recovering exports but China’s strong imports. The forthcoming trade release – interestingly due a few days before the Treasury report – is likely to demonstrate enormous import growth again, absolutely and relative to exports. This is seen not just in Chinese data, but in those from many other important trading nations. Indeed, quite remarkably, Germany’s trade with China is showing such strong growth that by spring next year, on current trends, it might exceed that with France. China last year reported a current account surplus of 5.8 per cent of GDP, significantly lower than apparently assumed as the current level by many people in Washington. In 2010, it could be closer to 3 per cent – incidentally below the 4 per cent level deemed as “equilibrium” by the Peterson Institute for International Economics.
The video is also worth watching; O’Neill and Roubini both make some good points. If only Greece produced the sort of capital goods that China is buying in large quantities from Germany.
Here’s the last section of chapter 6. I show how in the Great Recession we repeated the policy errors of the Great Contraction.
Given the unprecedented severity of the Depression, it seems implausible that any monocausal explanation is adequate. Even if the Great Contraction of 1929-32 was essentially a monetary phenomenon, the preceding account suggests that it was monetary in the broad sense of a breakdown in the international monetary system, (as emphasized by Temin and Eichengreen), rather than simply a result of inept Fed policy (the focus of Friedman and Schwartz.)
But the gold market approach goes beyond previous accounts that emphasized the role of the gold standard, by focusing on how shifts in the demand for gold impacted expectations of the future path of monetary policy, and not merely the fact that adherence to the gold standard constrained monetary policymakers. The large increases in the world gold reserve ratio during 1929-30, and 1931-32 did not reflect monetary policy being “constrained” by the gold standard. On the contrary, adherence to the “rules of the game” would have required far more expansionary monetary policies during these episodes. And although private gold hoarding cannot, by itself, explain the massive deflation of the early 1930s, the timing of these hoarding episodes played an important role in the latter stages of the Great Contraction, and (as we will see in chapter 12) even more so in the depression of 1937-38.
Although investors were focused on international monetary issues during 1931 and 1932, financial market responses need to be interpreted in the context of the contemporary institutional setting. Market responses to policy shocks merely show that the markets believed that the Fed was likely to act as if it felt constrainedby potential gold outflows. The issue wasn’t so much whether the amount of “free gold” was actually a technical constraint on policy, but rather whether the markets thought the Fed viewed it as a constraint. As Eichengreen (p. 316) observed; “Correctly anticipating that the Fed would draw back, producers and investors kept to the sidelines” (during the 1932 OMPs.)
This does not mean that the Fed could not have safely adopted a somewhat more expansionary policy (as Friedman and Schwartz argued), and it does not necessarily even imply that the markets agreed with the Fed’s cautious stance. But it does suggest that in the policy environment of 1932 gold hoarding reduced expectations of future monetary growth. And recent research suggests that what matters is not so much the current stance of policy, but rather the expected future path of monetary policy.
Perhaps the strongest support for the gold market approach comes from the period immediately following the British devaluation. The markets crashed on the onset of the dollar crisis in late September and the first few days of October as gold hoarding increased dramatically, but remained strangely passive in the face of sharp increases in the discount rate during mid-October. Friedman and Schwartz might argue that the markets already knew that the gold outflow would require discount rate increases, but that sort of deterministic interpretation would be hard to reconcile with their often dramatic account of Fed policy, which emphasized the vast consequences resulting from the death of a single individual.
During 1932 the monetary situation in the U.S. is especially difficult to decipher. We know that the markets welcomed Glass-Steagall, but this is consistent with the interpretations of both Temin, and Friedman and Schwartz. It also seems likely that markets were concerned that some of the more radical Congressional proposals could lead to an outflow of gold, but there is a frustrating paucity of evidence on the impact of the much-debated spring 1932 open market purchase program. Although there were a few instances where the OMPs appeared to boost stock prices, the overall performance of the markets during this period suggests that the program was perceived as being largely ineffective.
The analysis becomes even more complex when we consider hypothetical alternatives. Given the fact that several switches in Fed policy were at least partly motivated by worries that Congress would adopt more radical monetary proposals, it is unlikely that we will ever be able to know the precise extent to which a dollar crisis would have occurred with the Congressional budget fight, and without the OMPs, or vice versa. But there is also a sense in which it doesn’t quite matter whether it was fiscal or monetary policy that caused this crisis. In either case the net effect of the crisis was to weaken the impact of countercyclical policy.
It should also be noted that the Temin/Eichengreen critique of Friedman and Schwartz doesn’t really provide any comfort for (traditional) Keynesians. Keynes once warned that an expansionary fiscal policy aimed at boosting the economy might need to be curtailed if it led to a loss of confidence in the pound. Ironically, Keynes later became famous for advocating fiscal policy precisely because monetary policy might be ineffective in a depressed economy. But the example of monetary ineffectiveness provided in his General Theory was a situation where, if monetary policy was ineffective, it was not for liquidity trap reasons but rather because it had led to a loss of confidence in the dollar. This sort of “confidence trap” is equally applicable to fiscal policy—indeed perhaps more so, to judge by the market response to radical fiscal proposals during 1932. In chapter 13 we will see how confusion over the constraints of the gold standard indirectly shaped Keynes’ views of monetary policy ineffectiveness, and subtly influenced his entire approach to macroeconomics.
The gold market approach emphasizes the international dimension of monetary policy under a gold standard. On one level, this study is certainly supportive of those (such as Johnson) who criticized the French policy of hoarding gold. Eichengreen suggested that the Bank of France’s hands were tied by the French Stabilization Law of 1928, but there are two weaknesses with this argument. First, at this late date surely what matters is not whether the Bank of France was to blame for hoarding gold, but instead whether France was at fault. And second, Eichengreen’s own account undercuts any attempt to exonerate the Bank of France:
“January 1930 was the one occasion on which Emile Moreau, Governor of the Bank of France, proposed that the central bank buy or sell Bons de Caisse. Not only were the 2.5 billion francs of open market sales he proposed a mere 1.7 percent of the money supply, but they worked in the wrong direction. (1992, p. 197.)
Moure (2002, p. 262) accepts Johnson’s arguments that French gold accumulation contributed to the deflationary environment of the early 1930s, but then argues that “It is not French experience, but the broader systemic problems that French experience highlights, which are more worthy of note.” Moure (p. 183) also cites a recent study by Bernanke and Mihov that suggests France didn’t so much violate the rules of the game, as play a fundamentally flawed version of that game (which involved a gradual move from a gold exchange standard to a traditional gold standard.)
It is possible to agree with Moure and still view France as having grossly departed from one common definition of the rules—a stable gold reserve ratio. This suggests that the whole “blame” issue is mostly a question of semantics. Are we accusing countries of violating well-agreed upon ground rules (like a player cheating in a sports competition) or are we accusing countries of adopting ill-advised policies that hurt others, and themselves? France is innocent of the first charge and guilty of the second. And to a greater or lesser extent the same judgment could be made with respect to all of the other major central banks.
Of course the gold market approach to the Great Contraction is equally supportive of the French criticism of the U.S. banking system, or the view that Britain made a serious mistake by not setting the pound at a more realistic level after WWI. The basic problem with assigning blame among countries is that there was no consensus as to what the “rules of the game” actually were. In a sense, the interwar gold standard was both too flexible and not flexible enough. A rigid adherence it a fixed gold reserve ratio would have greatly reduced central bank hoarding during the early 1930s. Alternatively, a much more flexible regime which completely disregarded the gold reserve ratio would have allowed central banks to more easily cooperate to economize on the demand for monetary gold during a deflationary crisis. Instead there was enough flexibility to do a lot of damage, but not enough to easily repair that damage.
Friedman and Schwartz are probably correct that much more could have been done, even within the constraints of the gold standard regime. But policymakers had no simple roadmap for stabilizing the economy in a world where financial markets were very sensitive to any suggestion of policies that might put the dollar under a cloud of suspicion. And it’s also important to recognize that some of the worst crises of the Depression occurred precisely because markets feared policies that would be expansionary enough to provoke a run on the dollar, but not expansionary enough to assure recovery.
Of course, both Keynes and Friedman opposed rigidly fixed exchange rates. When viewed from their perspective, it is difficult to avoid being contemptuous of the budget balancers and tight money advocates of the early 1930s. But the policymakers of that era (particularly in America) were living in a world where devaluation seemed almost inconceivable. The entrenchment of the gold standard regime in America helps explain why the financial markets placed such high hopes on the possibility of international monetary cooperation, despite the ultimate ineffectiveness of those initiatives. They saw it as the only game in town.
The research for this book was completed well before the current crisis, which may end up being the worst since the Depression. Given the similarities I am about to discuss, it is ironic that the Fed is currently headed by one of the leading scholars of the Great Depression; Ben Bernanke. Although my views on the Depression are quite close to Bernanke’s in many respects, I believe that the Bernanke Fed repeated many of the same mistakes made in the 1930s.
One of the most important lessons of Friedman and Schwartz’s Monetary History is that neither interest rates nor the monetary base are reliable indicators of the stance of monetary policy. They did advocate use of the broader monetary aggregates, but after the early 1980s most of the profession also rejected those indicators as well. Even our best-selling money textbooks emphasize the unreliability of interest rates as monetary policy indicators:
“It is dangerous always to associate the easing or the tightening of monetary policy with a fall or a rise in short-term nominal interest rates.” (Mishkin, 2007, p. 606)
Unfortunately, despite the well-known problems associated with using nominal interest rates as indicators of the stance of monetary policy, most economists seem to have assumed that Fed policy during the current crisis was “easy,” mostly on the basis of the low nominal rates. Even worse, unlike during the 1930s, we do know what happened to ex ante real rates on government bonds. Real rates on 5-year indexed bonds soared from about 0.5% to 4.2% between July and November 2008, just when the recession intensified dramatically. Not so long ago economists liked to make fun of statements such as Joan Robinson’s claim that easy money couldn’t have caused the German hyperinflation as (she argued) nominal interest rates weren’t low. Yet by late 2008 most of the economics profession seems to have made exactly the opposite error, assuming tight money couldn’t have caused the sharp downturn in late 2008 because nominal rates were low.
Even worse, in 2008 the Fed did not have the gold standard excuse to fall back on, they had virtually unlimited ability to ease monetary policy under a fiat money regime. Again, quoting from the best-selling money text:
“Policy can be highly effective in reviving a weak economy even if short-term interest rates are already near zero.” (Mishkin, 2007, p. 607)
New Keynesian economists, including Ben Bernanke himself, had devised numerous “foolproof” strategies for implementing expansionary monetary policy once rates hit zero. And yet the Fed did not adopt a single one of these strategies, not even replacing inflation targeting with price level targeting, a policy Bernanke had recommended the Japanese adopt in the late 1990s. (Some people wrongly assumed that the Fed engaged in “quantitative easing.” In fact, the Fed sterilized the injections of new base money by paying banks interest on excess reserves. Later we will discuss parallels between this interest on reserves policy and the infamous decision to double reserve requirements in 1936-37.)
At the time, few people understood that the Great Contraction was caused by a severe monetary shock. Instead, it was much easy to blame the Depression on the much more visible symptoms of sharply falling nominal spending, including the stock market crash, the banking crises and the exchange rate crises. This time around economists fell into the same trap, although in this case their confusion was a bit more understandable. This time there really was an exogenous financial shock (in 2007), a shock that was not caused by economic weakness. The subsequent fall in nominal spending did further weaken the banking system, but it wasn’t the only problem.
Unfortunately, most economists seemed to miss the monetary nature of the second and more severe phase of the recent crisis, which began around August 2008. Nominal GDP feel nearly 3% over the following 12 months, 8% below trend. As in the 1930s, this fall in nominal GDP dramatically worsened bank balance sheets. With a more expansionary monetary policy that kept expected nominal GDP growth up around its 5% long term average, banks would have done much better, and the recession would not have spread beyond construction, to sectors such as manufacturing and services.
In retrospect, we now understand that falling nominal GDP was the root cause of the financial distress of the 1930s. We await a future Friedman and Schwartz to deconstruct the current crisis, and discover just how much was caused by the original sub-prime mortgage crisis, and how much was attributable to the Fed allowing nominal GDP growth expectations to plummet in late 2008. This policy error dramatically worsened the financial crisis, which in late 2008 spread to higher quality mortgages and commercial real estate loans.
Perhaps the most important lesson of the Great Contraction is that it is possible for most pundits and policymakers to be almost completely in the dark about the causes of a crisis. To those closest to the problem a severe fall in aggregate demand, or nominal GDP, almost never looks like it was caused by monetary policy. It is likely to be accompanied by very low nominal rates, and a bloated monetary base (as people and banks hoard currency.) That looks like “easy money” to the untrained eye.
Instead, people are powerfully drawn to explanations that put the symptoms of falling nominal GDP, such as banking turmoil, front and center as the cause of the recession. This didn’t just happen in the early 1930s, but in the late 1990s in Japan, and again in late 2008 in the US. While this confusion is certainly understandable, it was nonetheless distressing to see history repeating itself as I concluded this study.
In the late 1990s Milton Friedman also complained that modern economists had not really absorbed the lessons of the Great Depression:
“Low interest rates are generally a sign that money has been tight, as in Japan; high interest rates, that money has been easy. . . . After the U.S. experience during the Great Depression, and after inflation and rising interest rates in the 1970s and disinflation and falling interest rates in the 1980s, I thought the fallacy of identifying tight money with high interest rates and easy money with low interest rates was dead. Apparently, old fallacies never die.”
We need to always look beneath the surface, and keep in mind that no matter how severe a financial crisis, under a fiat money regime the central bank has virtually unlimited ability to prop up nominal spending, even at zero rates. If they do so, the financial crisis will be much milder. In the next chapter, we will see how between March and July 1933 Roosevelt used monetary policy to boost aggregate demand at perhaps the fastest rate in American history, despite near zero interest rates, and despite a banking crisis much more severe than faced by the Fed in 2008. It can be done, but it requires the courage to use unconventional monetary policy tools. In late 2008 the Fed wasn’t willing to take those sorts of risks. But in early 1933 the economy was in far worse shape than today, and FDR was willing to try almost anything.
In chapter 5 we saw that changes in the world monetary gold stock were positively correlated with changes in industrial production. But if production responds with a lag to monetary policy, then the contemporaneous correlations may prove misleading. Stock prices should respond immediately to policy shocks. If stocks responded positively to expectations for future economic growth (and during the 1930s it seems clear they did) and if private gold hoarding did reduce aggregate demand, then there should be a positive correlation between U.S. stock prices and the world monetary gold stock during periods when the U.S. was operating under a gold standard regime. The Cowles Index provides a nearly comprehensive monthly U.S. stock market index from 1871. In Table 6.2, the first difference of the log of real stock prices (DLRCI) is regressed on the first difference of the log of the world monetary gold stock (DLG) and the log of deposits of failed U.S. banks.
Table 6.2. The Relationship Between Variations in the Cowles Index of Stock Prices (DLRCI), the World Monetary Gold Stock (DLG), and the Deposits of Failed U.S. Banks (LDFB), January 1927 – April 1933, Selected Periods, Monthly.
Dependent Variable – DLRCI
(1) (2) (3) (4) (5)
Variables 1/27 – 4/33 1/27 – 6/31 6/31 – 4/33 1/27 – 6/31 6/31 – 4/33
DLG 5.921 1.338 7.772 2.005 7.711
(3.25) (0.55) (2.51) (0.82) (2.57)
LDFB -.0165 -.0284
Adj R2 .115 .000 .201 .015 .250
Durban- 1.56 1.26 1.81 1.35 1.68
Notes: T-statistics are in parentheses. The regressions included a constant term (not shown.)
One could argue that these correlations do not prove that causality ran from gold hoarding to stock prices. Perhaps stock market declines led investors to hoard gold. This hypothesis can be tested by splitting the sample in two. The period from 1927 to June 1931 was devoid of devaluation crises, and the relatively minor fluctuations in the growth rate of gold stocks may have partially reflected endogenous responses to changes in the world price level. Although the coefficient on monthly changes in the gold stock was positive, there was no significant correlation between real stock prices and the world monetary gold stock. Major stock market crashes in late 1929 and mid-1930 did not trigger substantial gold hoarding.
I have argued that between June 1931 and April 1933 changes in the world monetary gold stock were dominated by fluctuations in private hoarding associated with devaluation fears. After June 1931 the world monetary gold stock became much more volatile, falling during the first, second, and third runs on the dollar. And during this period changes in real stock prices became strongly correlated with changes in the world monetary gold stock. This pattern supports the analysis in Appendix 5a, i.e. that a positive correlation between gold stocks and aggregate demand would be expected only when changes in the world monetary gold stock are exogenous. It appears that after June 1931 devaluation fears caused hoarding, and hoarding reduced stock prices (and presumably aggregate demand as well.)
See Sargent (1983) and Eggertsson and Woodford (2003.)
 New York Federal Reserve Bank President Benjamin Strong died in 1928. Less well known is the fact that Irving Fisher and R.G. Hawtrey shared Friedman and Schwartz’s views regarding the significance of Strong’s death.
 See Keynes (1982 ), vol. 9, pp. 353-54.
See Bernanke and Mihov (2000, pp. 148-150.)
Hetzel (2009) Richmond Fed, was a notable exception.
 Real stock prices were generated by deflating the Cowles Index by the National Industrial Conference Board’s Cost of Living Index.