Comment on Carmen M. Reinhart and Takeshi Tashiro, "Crowding Out Redefined: The Role of Reserve Accumulation"
Attention Conservation Notice: 2600 words reiterating the points made by Reinhart and Tashiro (and also by others like Raghu Rajan) that the East Asian financial crisis of 1997-98 appears to have triggered a permanent downward trend break in economic growth on the Asian Pacific Rim.
November 3-5, 2013: Federal Reserve Bank of San Francisco Asia Economic Policy Conference 2013
Brad DeLong: Comment on Carmen M. Reinhart and Takeshi Tashiro, "Crowding Out Redefined: The Role of Reserve Accumulation"
Let me second what Alan Taylor said. This, by Carmen Reinhart and Takeshi Tashiro, is another high-quality paper. And it shares the four standard high-quality characteristics of Carmen Reinhart papers: 1. It takes data that we have not looked at before or that we have not looked at in this way before. 2. It presents the data in a very interesting and thoughtful manner that makes us think very hard about important questions. 3. It does not focus on either the trend or the cycle exclusively, but looks hard at the interrelationships between them--interrelationships between the cycle and the trend that are traditionally ruled out, or at least not at the forefront of, our back-of-our-envelope first-cut. 4. It does not bow to current theoretical perceptions, but attempts to focus our attention on what the important and interesting features of the economy are.
This paper, in brief, is about the long and dark shadow cast by the cycle and the trend. In this case, the cycle is the 1997-1998 Asian Pacific Rim financial crisis. The trend is what that crisis has meant for the development of Asia’s Pacific Rim since. The at-least-partial motivation for this paper is that, right now, the European periphery is going through something somewhat similar to the sudden stop experienced by Asia's Pacific Rim in 1997-1998. The point is to look at the European periphery today in the the Asian Pacific Rim 1997-1998 mirror, and to see what we see. What do we see?
The origin of the sudden stop is that something bad happens to foreigners’ willingness to invest in a particular region. Something bad happens to foreigners’ willingness to hold assets located in a particular country. The financial system has to dal with this disruption to the pervious pattern of financial intermediation. And the economy then has to deal both with the changed financial situation and with the real side concomitants of that shift in foreigners' preferences.
When we back up and assume the 30,000-feet-up-in-the-sky economist’s perspective, of course, there is no reason why a shift in foreign investors' preferences and even a disruption of financial intermediation should materially disturb the real economy of production, employment, and investment. The market--or the central bank, or the government--should be able to build, should have already built, firewalls financial disruptions and real production, employment and investment. The marginal rate of time preference hasn’t changed. The marginal profit of capital hasn’t changed. The marginal product of labor hasn’t changed. All that has happened is that foreigners have suddenly lost their taste for buying assets denominated in a particular currency, located in a particular place. And we have a price--the value of the currency--that is supposed to smoothly match demand and supply in the market for assets located in a particular place, and markets are well-designed to deal with both the partial and the general-equilibrium adjustment to preference shifts.
Markets deal with such a shift in preferences by lowering the price. In the case of demand for Asian Pacific Rim or European periphery assets, the market's natural adjustment path is to bounce the value of the currency down. Bounce the value of the currency far enough that speculators think its next move is at least as likely to be slightly up as it is down. There is then no reason for safe interest rates in the region to rise. The situation stabilizes. And you have, at worst, a short, sharp, V-shaped downturn followed by an export boom. And yet that is not what happened. On the Asian Pacific Rim in 1997-8, the fact that so much of the region's debt was denominated in dollars meant that bouncing the value of the currency and thus of domestic production down far enough raised universal and valid fears of bankruptcy, and sharply raised risk premia: the Asian Pacific Rim thus had to, to a certain extent at least, defend its currency. And in Europe's periphery nations are tied by treaty, by the deep and close technical integration of the financial system, and by hopes for a united and peaceful European future into the euro zone. Thus when the crisis comes both regions must generate rapid adjustment of the current account: a sudden stop.
The problem is general. There are lots of reasons why the natural market's bounce-the-value-of-the-currency-down adjustment mechanism will not work. Overwhelming reasons to maintain a fixed parity. High levels of harder-currency debt. A tight coupling of import prices to domestic inflation and a belief that the costs of accepting domestic inflation are unacceptable--cough cough, why we all today feel sorry for Raghu Rajan. In any of these cases, when the crisis comes you must generate a rapid adjustment in your current account, and the easiest and the most straightforward way to do this are via domestic investment collapse. This is the first failure of the veil of the financial system to be merely a veil--the first coupling of financial distress to destructive real economic consequences.
But this should be temporary. This should produce a "V". But what Carmen and Takeshi impressively document is that investment does not come all the way back. We have a long-run 6%-point of GDP delta in the investment share in the post-financial crisis period relative to the pre-crisis normal on the Asian Pacific Rim. We have a long-run 2.5%-point per year delta in real GDP growth. We do not have a V. We have an L. And there does not appear to be anything going in in terms of exogenous breaks in long-run trends that would lead us to say that the trend break was going to come anyway, and that the financial crisis disruption and according depression was not the cause but the consequence of the trend break.
Now when I was young I was taught that the trend was the trend and the cycle was the cycle, and that sometimes breaks in the trend caused cycles, but not vice-versa--or at least not vice-versa enough that we needed to worry about it. When I was young I was taught that Say's Law held in the long-run even if not on the short-run, because the interruptions of Say's Law that caused demand cycles were driven by sudden excess or deficient demands for the stock of liquidity, and those had a naturally end because flows accumulated to make up stocks.
Furthermore, when I was young I was taught that central banks were large and powerful enough to make Say's Law roughly true in practice even though it wasn't true in theory even in the relatively short run--that the short run of aggregate demand shortfalls, and the duration of "V"s, was limited to two or at most three years--they could be surprised, and long and variable lags kept them from offsetting demand shocks immediately and instantly, but that the duration of the surprise was limited to the period of predetermined prices.
And I was taught that capital should flow downhill; that at the level of international economics, governments really, really were agents of the citizens so we could view the country as a whole; and that we could expect to find conditional convergence of living standards and productivity levels across nations--convergence conditional on getting good institutions, that is.
Yet these do not appear to be so. The fallout from the 1997-98 financial crisis has been very large, very persistent. This is even more puzzling because, as Alan Taylor said, the post-crisis policy reaction of building up of reserves should not be a drain on national savings at all.
The way such situations were supposed to be handled was that when there was, in the words of the Articles of Agreement of the IMF, a "fundamental disequilibrium", the IMF was suppose to step in. The IMF was there to take the blame for getting the country to do the things that had to be done in the long run to balance resources against commitments--to take the political fallout, and thus induce politicians to do what was necessary rather than kicking the can down the road while they hoped for a miracle. The IMF was there to get the country back onto its proper long-run trend growth path. And the IMF was there to provide bridge financing in order to make the process of adjustment as painless as possible.
In this context, the Asian Pacific Rim reserve accumulation after 1997-98 is best viewed as the product of a recognition by the Asian Pacific Rim that they did not trust the IMF to do its proper job.
The view in Asia in the aftermath of 1997-98 was that the IMF had been late to the party (although it did bring lots of refreshments when it did arrive). In 1997 and 1998, from the Asian Pacific perspective at least, first-world international speculators suffered a great failure of nerve, an irrational panic, and fled the region. This panic was irrational: those who held on did fine, and those who bought into the crisis did enormously well. But in the crisis the IMF did not do its job--did not provide enough funding fast enough without inappropriate conditionality. And the sovereigns of the Asian Pacific Rim reacted to this by deciding to build up their own reserves so they would never be forced to rely on the IMF again.
That decision should create added confidence. That you can now invest in Malaysia or Korea or Indonesia without worrying about any kind of financial disruption because its own central banks stands ready to smooth adjustment and does not need the IMF cavalry should lead to a higher investment share, not a lower one. Thus, relative to what is my and what I suspect is Alan's counterfactual, the investment gap is even larger than Carmen makes it sound.
Could the investment gap possibly be crowding-out by China’s production--the idea that there is a niche in the world economy for export-oriented rapid-development by some emerging market countries with low-valued currencies, but that in 1997-98 it became clear that China was going to occupy that niche, and bigfoot everyone else out of it. In response to that recognition, the argument would go, the returns to investment in the Asian Pacific Rim would drop and so investment would drop? I do not believe so, largely because I have sat at the feet of Chang-Tai Hsieh and learned from him how very tightly coupled the Chinese export sector is to the other economies of Pacific Asia. More exports from China mean more work and more profits elsewhere on the Asian Pacific Rim. China actually getting its act together and taking up a position as the supplier of the low-wage component of the value chain seemed and seems to me more likely to raise the marginal product of the capital in the rest of the Asian rim rather than to lower it. Could it be that there was a negative shock to expected pace of globalization in the aftermath of 1998? Again, I find this hard to credit. If anything, globalization has proceeded faster since 1998 than anyone had previously imagined it could before. Before 1998, we were gradually realizing that the moving of industrial production out of the North Atlantic core to the periphery was a thing that was now possible. We were realizing that containerization had produced another great downward leap in costs of transport. But we had little idea then on how much difference the coming of modern telecommunications would make in tying the world together.
Thus I find myself puzzled when I try to think of how 1997-98 could have seen a shift in trend that then caused the cycle. I cannot see any exogenous change in the trend that would validate either the pos-1998 slowdown in growth and investment or the financial crisis itself as a reaction to bad news about future growth in Asia.
Thus I am left hunting for other explanations. One of them is definitely Carmen and Takeshi’s: that this particular 1997-98 financial crisis is casting an extraordinarily long shadow on the economy. This may in some way be tied up, in a manner I do not understand, with the astonishing role that the dollar has played over the past 15 years. There is an economist in the front row who, back in 1979 and 1980, made me read at extended length about Robert Triffin and the Triffin Dilemma. And at the time I thought that this was a waste of my time--how could we ever again get into a situation in which it would once again be useful to characterize the world economy as suffering from a dollar shortage--a shortage of dollar-denomionated and U.S property- and taxing capacity-backed liquid assets? Yet, now, we have seen not just Bretton Woods II but the return of Robert Triffin at a scale that even back in 1998 I would never have believed possible in a thousand years The bottom line, I think, is that we have a difficult task before us. Our old belief that you have a trend, and you calculate the trend; that you have a cycle, and you argue over whether the cycle is falls below the trend or fluctuations around it; but that you can carry this discussion along two separate and largely-disconnected tracks--that belief looks to be simply wrong. Carmen and Takeshi’s paper here is another nail in the lid of its coffin, another demonstration that by trying to think in such ways over the past generation we have not done ourselves or the world a very good service.