The Return of Lord Keynes

The Economist magazine is at it again. They asked, last week, if we are in a return to the 70s. Based on what ? Loose monetary policy in the U.S. and Japan, oil prices rising, reckless credit expansion in China, and . . . not much else.

From the back pages, I read: %change in CPI Japan, -0.4, EU 2.5, U.S. 3.1. Are you scared? Well, these are year-on-year rates, and current monthly rates are higher. On the other hand, oil prices are down 8% in the last month - from very high levels. All in all, this may be a bit like the early 70s, but it is not the stuff of a Central Banker’s nightmares.

This log is not about rebutting the Economist. Everybody has to sell their product, and the Economist deals in prognostication for effect as often as it does for accuracy. (Remember their prediction that oil prices might fall to, was it, $5 per barrel? Just five years ago, when oil was at $10?) But it is interesting to compare their complaints about excess liquidity to the money supply numbers. Broad money in the Euro area has been growing at only 5.6%, in the U.S. at 5.5% and in Japan at 2.0%. Narrow money looks a little worse: 10.9%, 5.1% and 4.2%. But the highest growth is in the region with the highest unemployment and the most slack in general.

Rather I want to focus on oil prices, and commodity prices in general, to see if they can tell us anything useful. It is a vexed subject. The Wall Street Journal, in the days of radical monetarism of the early 80s, proposed that central banks anchor the price of gold as a preventative against inflation (and a quick fix for expectations of high inflation that are otherwise painful to wring out.) Later in the decade, when they were beginning to absorb the logic of economists concerning the problems of being so inflexible, they shifted to settling for fixing prices of a basket of commodities. Then commodity prices began falling in the 90s, inflationary expectations finally passed away with the Clinton budget deal, and the call fell silent. (Too bad the Argentine quick fix, the Currency Board, wasn’t scrapped before it straitjacketed them into a depression. But then, they were more desperate to fight inflation than the U.S. was.)

All this to say, one interpretation of the current surge in commodity prices is that it simply represents a return to normalcy. Consider the Economist’s commodity price index. By comparison with 1995, the entire index (presumably with a proper weight on oil) is down 6%. It would be down more if you adjusted for inflation. This is no longer an artifact of the high dollar - the Euro index is only up 2% (in 9 years) and the yen index up 10%. Commodities in general, especially foodstuffs, are experiencing deflation as much as inflation, and on average are just catching up after their incredible fall in the 90s. We could clamp on the monetary brakes like the Fed did in response to rising prices in 1937. Or we could take the view of Franklin Roosevelt, when assuming office in 1933, that what the economy needed was to get prices back up again.

This idea of a proper minimum level of prices is a tricky one. It could easily lead to abuse. In its essence it captures the notion that demand deflation is a bad thing, and can leave a residual drag on the economy for a long, long time.

But it might be more manageable for the moment to translate into concrete microeconomic terms. Roughly speaking, it corresponds to the idea of pricing to cover long-run costs, i.e. costs taking into account the per unit cost of capacity investments. Costs have been falling during the 90s for most minerals, as exploration and exploitation got cheaper. In fact it is one of the basic regularities at the heart of development economics that primary commodities tend to fall in price over time. It is believed that this happens because better technology makes it easier to find and extract resources, but also because bulky goods take a declining share of income as people get more affluent. This second, unfamiliar idea is that higher income leads people to spend more mainly on the vast array of knowledge-intensive goods and services that become available, from medical care and ski gear to electronic entertainment and impressive looking packaging, rather than on proportional increases in food, housing, clothing, etc.

However these two trends should always be set against the more familiar basic fact that non-renewable natural resources get scarcer as they are used up. In fact, the price would always be expected to reflect a “scarcity rent” on the basic resources, so that the price is not just the explicit cost (for labor, machinery, energy, etc.) of extracting the resource, but also reflects two parts of the scarcity of the resource. The first part is the extra payment that is required to bring in higher cost resources to production, if they are needed for satisfying the next unit of demand. So for example, if demand is not met from low cost sources of oil, one must pay enough to cover costs of North Sea oil. These are “quality rents” or “rents to easy extraction,” and the highest cost supplies won’t receive any.

The second part is a rent for future demand. Because copper ore that is left in the ground can be used next year (or next decade), part of its price should reflect that foregone opportunity. Roughly speaking, we expect the price to rise by the rate of interest: if the price is higher than next year’s expected price discounted by the price of money, then it would pay to mine more now and the price this year would fall while the price next year would rise. Of course if next year the copper will be worth less, because of a foreseeable decline in the future cost of production, this can drive the future demand rent down. It can never drive the current price below current production cost, of course. But when there is no scarcity rent for future value, we might want to ask if demand is growing as fast as it should.

There is a case to be made that the commodity standard is best thought of as an indicator of when demand is too slow (rather than, as the Wall Street Journal thought, too fast.) By contrast with finished goods, which are mostly produced oligopolistically and so have prices that are sticky downward, competitively produced resources will lower prices when the economy is sagging. They sometimes drop well below the full cost, in fact. Presumably they can still be worth producing below full cost because the cost of extracting the next ton is often low once the ore is located and the infrastructure in place.

So the question of whether the price declines of the 90s represent mainly technological improvements or mainly slow demand is a vital one for assessing policy, and for assessing current trends. We know that prices fell below full historical costs for at least some producers. Oil prices, for example, fell below the price of extracting new oil from the North Sea, let alone from Greenland. But there is plenty of evidence that technological improvements were responsible (both improvements in location techniques and improvements in Russia’s functioning technology). (Someone should calculate how much the overstatement of reserves contributed to low oil prices). Likewise for minerals - liberalizing countries allowed in new investment, that achieved dramatically lower costs (though sometimes at the expense of the environment). Field crops took a dive when Republicans took over Congress, because they eliminated setasides and so increased supply. They are barely recovering now, presumably after enough marginal producers have given up.

But remember that prices of raw materials are not supposed to match the lowest cost available. There should be at least some “quality rents” to low cost sources. In general, economists are not comfortable with saying what is the “right” price. It smacks too much of just prices and attempts to repeal the laws of supply and demand. But I would like to suggest that raw materials prices are a kind of indicator. One way to think of the question is whether there is slack in the (world) economy. If some resources that are already under development begin to be priced out of the market by new sources, we know for sure that in that market at least, demand is not keeping up with supply. That would be an indicator of slack.

We like to point out the issue on the other side, when there is too little slack and economic stimulus seems to be creating pent-up pressure on prices. Public discussion of the issue has traditionally focused on unemployment rates, but those are really just one indicator. Central bankers in practice must also look at capacity utilization rates, queues for machinery making (and computer programming, if we had the data), and excess reserves in the banking system, to name a few of the most prominent indicators. But we can’t just measure these items and plug them into a formula. We need to assess them in light of the correct questions.

First, the question needs to be put in terms of growth rates. Since supply is always growing, the question is primarily whether demand is growing too fast. There will be a natural self-regulatory process in which demand that is growing too fast will slow itself down by raising interest rates. The essence of monetarism is that government can frustrate this process by generating too much artificial credit, based only on money creation, and not at all on setting aside current resources through the saving process. The result, after a period of low nominal and real interest rates, will be inflation.

The essence of Keynesianism is that there is no self-regulatory process to speed up demand if economic growth is too slow. Not lower interest rates? Well, yes, a mild slowdown will stimulate additional demand using lower interest rates. But it is very easy for the growth rate to fall below the rate that can be sustained by pushing down interest rates. Certainly when nominal rates hit zero that is a sure sign that demand for investment is too low to sustain growth near its potential. Many of us think it happens well above that point, (or the point of zero real interest rates). In other words, you can have inflation even though there is too little demand for the supply. The “sacrifice ratio” needed to squeeze inflationary expectations out of the economy is an example of this.

To understand why demand does not easily restore itself to proper levels, think about bottlenecks. If demand is going too fast, bottlenecks will develop. That is, there will be sectors of the economy that are slowing down overall growth because of their limited capacity. Suppose the economy needs lots more steel, but steel capacity is still too small, because previously, it wasn’t needed as much. Steel will be a bottleneck for the economy. And its prices will almost surely rise, rationing demand for steel to the highly valuable uses. So we have a story about the emergence of inflation.

But if you study development economics, you run into a somewhat different story about bottlenecks, in the context of growth theory. Hirschman suggested long ago that flexible prices provide a signal of where development would be most valuable. If steel prices are high, then investment in steel would be of high value. In this view, bottlenecks are there to be cleared. If you don’t have bottlenecks, then you don’t get signals.

It is easy to exaggerate this notion. In truth, industries may expand even though they already have excess capacity, and will not be raising prices. These are quantity signals to invest. The question is whether quantity signals alone are enough when the overall economy is too slow. Will declining interest rates tell you to invest despite some excess capacity, even though demand projections are fading at the same time as a “pause” sets in? Will quantity signals be enough to sustain investment at a level that generates equilibrium growth?

If you define equilibrium as zero inflation, the answer is probably yes. Recall that some industries, usually electronic these days, advance rapidly enough that they need to replace old capital before it begins to wear out. Thus demand would have to collapse to get this investment demand to disappear. And these same industries are generally lowering prices. This leaves room for some bottlenecks in other parts of the economy, consistent with zero inflation.

But the appropriate standard is whether demand is sufficient to utilize capacity without accelerating inflation. So, in other words, without excess credit creation. I am suggesting that investment can easily fall below levels at which this will happen. Note that it also implies reduced capacity in the future, so that a zero-inflation strategy will be forced to live within the constraints of the underbuilt capital stock. The question needs to be phrased broadly enough to encompass the issue of whether the economy is growing as fast as it can. My claim is that if no sectors are raising prices due to limited capacity, growth is almost certainly too slow. Even stronger, if more bottlenecks are being cleared than being generated, growth is probably still too slow.

Why bottlenecks? Aren’t they really proving that slack is gone? Isn’t it dangerous to push the economy so hard that it faces supply constraints, practically guaranteeing some inflation? But zero inflation is a goal only in the unrealistic world of economic theory, in which forecasts are infallible and all inflation represents an excess of monolithic, single-commodity “demand” over equally simple “supply.” In the messy world of actual investment decisions, we need a more subtle view of where inflation comes from and what policy (or at least what policy “experiments” in the Tobin-esque sense) will bring this inflation about.

Ask yourself whether, if demand suddenly doubled (all the Chinese, Japanese and European consumers suddenly began spending, instead of saving) whether there would be enough capacity to meet their demand. In the short-run, surely not. But in the long run, as we teach in microeconomics, the capacity would increase and it is possible that the doubled demand could be met. At least, ability to build factories would not be the ultimate constraint. Maybe we couldn’t train enough doctors for the resulting demand. Or enough carpenters. But maybe, we could train plenty of both, partly by adding productivity to all the service workers we have, and by convincing people to do without the ones that are easily replaced like service station attendants and bank tellers have been replaced. At any given point in time, there is probably plenty of excess potential capacity that could be met given a bit of time to build.

Inflation (not one-time price increases needed to signal for more resources) comes when the demand increase is persistently ahead of the rate at which capacity can be added across many sectors. And surely it is obvious in today’s economy that productive capacity, in the sense of physical capital, is not the constraint on the pace of growth. I propose that the pace of innovation is probably the limit today.

Isn’t it enough to have the innovation that is naturally generated by electronics and other fast-improving sectors? Isn’t other investment sufficient, if it occurs when firms are just keeping capacity ahead of demand? Maybe, but I think that vastly underestimates the innovative capacity of modern production. If there are bottlenecks, innovation will naturally be targeted at the spots in the economy that are limiting growth. An acceleration in demand that generates bottlenecks will be followed some years down the road by innovations that relieve them. The supply will naturally accommodate itself to the evolving pace of demand, but only with some pressure.

We can have a slow-growing equilibrium, with supply of funds equalling demand for them, where the demand comes only from the innovations that are most obviously profitable in a slowly evolving economy. In a different, fast-growing equilibrium, capital is allocated toward innovation that will best permit growth. In the slow-growing equilibrium, capital is allocated (at a lower interest rate) toward capacity that is under-innovative and accommodates the priorities of under-competitive oligopolies. And the cumulative under-innovation goes unnoticed, especially if all the economists have been trained to believe in the mystical power of equilibrium forces to reveal all.

Thus I am proposing that policy be deliberately aimed not at zero inflation, but at (reasonably low) inflation that is sufficiently far from price stability to avoid the situation of too few bottlenecks. It must also be sufficiently sustainable that many or most bottlenecks are being reduced by innovative activity faster than they are being increased by further stress on the overall economy. Whether this situation is possible is not yet known, but would (as the late James Tobin would have said) make an interesting experiment.

To interpret this for the situation of natural resources, it would suggest that there be plenty of quality rents, but that the pace of innovation and new capacity addition be sustained by the action of prices that are at least not rapidly falling, and probably are rising at a pace similar to the real interest rate.

Suppose we did a quick inventory of the types of resources, to see if the world economy has much slack or not.

China’s economy, bigger now than Britain or Italy, is sucking in more steel and raw materials than any other economy. That is the proximate cause of the increases in metals prices, and perhaps partly for oil price increases as well. Does that mean there is no excess capacity left in mining? Hardly. BHP and Rio Tinto are investing billions to increase mine capacity in Australia, and other locations are also expanding. The potential for expansion is significant. Remember that the issue is not current capacity so much as realizable potential capacity (and long-term sustainability of the pace of expansion).

At the same time, China is putting its formal economy into the modern world. Automobile purchases doubled last year. Obviously they cannot continue doubling forever. Equally obviously there are lots of potential automobile consumers. But how many can afford one? It depends on at what rate the country shifts into the higher gear of modern life. College graduates also doubled in the last few years. Chinese companies are competing world wide in technically sophisticated industries, not just lending cheap labor to foreign investors.

Yet the potential for expansion in China (and India) makes clear that there is no effective labor supply constraint on world production. What about skilled labor? Maybe. But education in developing countries is made cheap by the low cost of its primary input, which is the time of the learner. If you ask the question whether higher relative prices would be necessary to induce more production of educated or trained workers, it does not seem likely.

More likely the constraint is absorptive capacity. That used to be another word for effective aggregate demand. Now it has more to do with what the ability to adopt modern methods profitably. The main problem is not that labor is so cheap that labor-saving technology doesn’t pay. Outside of subsistence agriculture, nothing could be further from the actual situation. Rather, the situation is that today’s technology-intensive methods can move into any fast growing market and use a quality edge to cream off all the high end demand. This is not labor-saving technology in the sense of automating or mechanizing so much as saving on the variations and errors that creep in with straightforward mass production. As a result of this pre-emption effect of global capital, manufacturing employment is likely to grow slower than income, rather than providing positive feedback to growth. Automation of production was the engine of growth in today’s industrialized countries. Outside Mexico and China, there may be no place left in the world where this can happen again.

I conclude that the absorption of advanced economy methods is limited by the pace of opportunities for low-salaried local producers to put together income for buying modern stuff. And thus the creation of skilled labor is limited more by demand than by supply.

The other main constraint in the world today is environmental. What China is doing to its environment is criminal, even by the laws of China. We don’t really need carbon dioxide to be pouring into the atmosphere any faster. But as in the 60s, the problem is not so much overproduction as under-regulation. The environmental industry in industrialized countries is calling for pricing methods on effluents, to give manufacturers an incentive to adopt the pollution control technology that already exists. The shortage of hybrid vehicles is another example of how proper pricing could move demand quickly toward taking advantage of the technology we have. The damage to the environment should be paid for by polluters, period.

The kinship between development economics and Keynesianism is closer than even most economists realize. Much mischief has been done in the name of both, but behind this abuse is the genuine analysis of market failure. Today the world is reaching a stage where the linkage has come full circle, from closed economy macroeconomics, to development planning policy and back to world economy macroeconomics. The Keynesian insights that made their home in inward-looking development policy, mistakenly, now make sense in a world that has been globalized by the rejection of inward-oriented growth policies. By rejecting these policies, nations have forged a world labor market and a world capital market capable of the kind of rapid expansion that probably needs a government push to bring it to fruition.

Oil price shock fools monetarists at the Economist

Well, once again the Fed has it right and the Economist has it wrong. Gives you at least a bit of confidence in the process of choosing economic policymakers.

In response to the increases in oil prices, and the resulting inflation, the Economist’s ”Economics focus” column says interest rates should rise. Wrong. They do propose taking into account the speed of economic growth and the amount of slack, which means they are not totally out among the planets.

The basic case is explained in my second blog entry ("When supply takes a hit"). A contraction in Aggregate Supply calls, in general, for a decrease in interest rates, not an increase. To conclude otherwise is to mistake “cost-push” inflation for the effect of persistent expansionary monetary policy. I feel a touch of pride at having gotten to the question before they did, but then, that is what academics are for. That, and to get the ideas right. However, the column discusses some interesting ideas and, bravely, raises case history. So, this blog is my reply.

To discuss ideas first, they note accurately that the result for inflation depends on monetary policy. In other words, an oil price increase will only cause inflation if it is “monetized.” If not enough money is available to accommodate the higher cost of living, other sectors will suffer and most likely other prices will fall. If the Central Bank is vicious enough, it can enforce “price stability” or zero increase in overall prices, but the effect on output will be more than a little disturbing.

The Economist staff should have thought further about this principal when looking at another issue they raise, namely whether the price increase will be temporary or a pervasive, longer lasting effect.

To start with the extreme case, for an oil price increase to result in a permanent increase in the inflation rate, the conditions in the market would have to continue tightening on a perpetual basis. Of course it would also require monetizing, but more to the point it is almost exactly what one is not observing when prices jump. Rather they are signaling a “permanent” increase in oil scarcity (it may turn out to be temporary) requiring an ongoing rationing of demand to remain within the new, more limited, supply conditions. There is no basis for thinking that it has a long-term inflationary component.

Therefore, returning to the question of monetization, only if there is a steady increase in money creation, leading to prices rising for all goods, is there reason to think oil prices would continue to rise, and that would be a clearly demand-based change. Therefore the price jump that accompanies tighter supplies is almost by definition a one-time price increase, which is what Greenspan labeled it, rather than a source of inflation. Yet the Economist does not recognize this implication despite declaring “if there is no sign of a rise in the core rate of inflation, then there is no need to raise interest rates.”

In a second interesting line of thought, they turn to analyzing the issue in terms of how much slack there is in the economy, on the principle that slack will restrain inflation. While this is a sensible principle, the Economist makes such a hash of it that the stranglehold of monetarist myopia is made plain for all (trained eyes) to see.

Most seriously, they completely miss the obvious fact that an oil shock will, rather than removing slack as simple AS-AD analysis suggests, actually generate slack in the economy. This is because it has lots of spillover effects on demand (see Blog #2), in an illuminating example of the “short-side rationing” that we had the sense to discuss in the 70s. (It is illuminating because the rationing is not a crisp comparison of demand to capacity, but a messy cluster of sectoral drops in demand created directly by supply constraints but also indirectly by the demand shifts among complements and substitutes, the restriction in future prospects and the siphoning of spending abroad.)

It might be worth giving some thought to the implications of a complete “equilibrium” approach in this situation. The Long-Run Equilibrium story takes the view that the fall in investment following on an oil shock is entirely appropriate because profitable opportunities have in fact declined. While based on a valid point, if left to the market this would imply that any overshooting of investment is strictly a smooth adjustment to the new equilibrium capital level, which just happens to imply a low replacement level until the economy catches up to capacity.

The truth is that individual industries do not know what the long-run equilibrium level will be, and they are in serious danger of mistaking the temporary lull for the long-term prospects and thus confirming their collective pessimism. These are demand spillovers, they are a market failure, and they are at the core of Keynesian analysis. Put another way, slack has a tendency to be self-perpetuating, and the Central Bank has an obligation to lean against it, (though not to go to extremes or pretend that it can always be eliminated).

Less drastically, the Economist gets it wrong by waffling on whether Europe is more subject to broad-based inflation caused by cost increases than America is. On the one hand, Europe has more slack, they say, so that it is in less danger of the oil price increases becoming systemic. On the other hand, Europe has more labor market rigidities which will cause the higher cost of living to be turned into higher wages. Ignore, for the moment, that they express it in terms of this funny idea that a change in one key market can turn into ongoing inflation.

What else is missing here is a sense that wage (or price) rigidities create a ratchet that policymakers should be taking into account. In a time of falling real exchange rates, adjustment for the nation requires either falling prices of non-tradables (such as land and labor) or falling nominal exchange rates. But rigidities in wages will necessarily cause the fall in demand for local labor to translate mainly into unemployment rather than falling wages. There are two simple fixes - higher prices and lower nominal exchange rates. Lower interest rates promote both. These will not solve the problems on the real side of the economy, but they will prevent those real problems from being compounded by unnecessary extra contraction.

It is very useful to think of an oil shock as a decline in the real exchange rate. (The Economist recognizes this dimension in micro terms, and goes on to note the fall in demand from transferring income abroad. Pity they do not acknowledge any except the “real equilibrium” demand effects.) It is different in kind from a fall in the demand for a country’s exports, but the implications for macroeconomic policy are very similar. World goods become, on balance, more expensive. A given amount of consumption is therefore more expensive too. While the initial effect is not a fall in demand, as would be the case with exports declining, what we might call “real demand” (i.e. adjusted for prices) will have fallen. As I have written before, if you are not fixated on zero inflation, it is obvious that credit should loosen, not tighten.

Think of the situation of a small country under a world gold standard (or a state in the U.S.) whose products sell less. Real estate prices tend to fall, wages tend to fall. Capital flows out, as investors seek areas with higher returns. But since wages tend to resist downward pressure, employers compound the problem by laying off workers during the period of adjustment of wages. If instead of a fixed currency, a general loosening of credit specific to the area could occur, causing its currency to decline and its prices to rise, then no unnecessary closings would result. In the U.S., a region in decline sees considerable outmigration, but it also sees big falls in real estate prices and some attraction of financial capital to occupy the newly available physical capital. Why not give first call on the physical capital to local entrepreneurs, who will have more understanding of the area’s comparative advantages and will have more positive demand spillovers? A loosening of credit in the area would do that.

Interestingly, the Economist also takes up the issue of world slack. Their conclusion is that since oil prices are rising due to rising demand worldwide, especially in China but they should have mentioned Asia generally, then there is no scope for a stimulative policy. As I hope the preceding paragraphs indicated, this confuses the issue of appropriate world monetary conditions with that of appropriate U.S. monetary conditions. Although the Fed may be the closest thing the world has to a world Central Bank (the IMF has specifically been denied this role), it still has control and responsibility mainly within the US economy. It is true that demand in the outside world is raising the cost of natural resources, that some internal uses will be outbid and that the real exchange rate will fall. But again it is dangerous to try to force stable prices on the situation, and least disruptive to adjust with a declining currency and rising prices.

The Economist puts the matter in a different light by noting the rising exports to China that are resulting from world expansion, and the reduced slack that will result within the industrialized West. To that extent they have a valid point, and the core issue is exactly that - whether overall demand in the U.S. is rising so fast that interest rates need to rise to restrain it. I will return to this point at the end.

The example they cite from recent policy history is the oil shock of 1974. The story they tell is that the Fed first accommodated the rise with lower interest rates (sending short term rates lower than inflation for a negative real interest rate) and then found that interest rates and unemployment had to rise even more than they would have otherwise, as a result of the extra inflation that had been produced. In fact the picture is even worse than they paint it, but very little of the damage is to be attributed to accommodation of the oil shock in 1973-74, except as a part of the larger pattern of policy over the 12 years from 1966 to 1978. To see why, one should compare with Japan’s response to the same crisis. Japan fully monetized the oil price increase, saw inflation spike at a much higher level than in the U.S., but then saw it return to its normal low levels.

Inflation in the U.S. was systemic at the time, based on the Fed accommodating Keynesian fiscal policy. Because the inflationary result had been systematically underestimated, there was a catch-up process underway at the same time oil prices jumped. (It would not be too farfetched to label the oil price jump as part of the same catch-up process: as economic agents saw others turning to price increases, they began to view their options differently and raise prices themselves.) But economic policymakers had not expected such a big jump to materialize at once, and failed to anticipate the effect on the mortgage market.

In 1974 the Savings and Loan industry essentially shut down new lending, based on the first taste of the trouble that would later create the S & L crisis of the 80s. The problem was that their portfolio of loans was at low mortgage rates, often 6 to 7 percent, but they could not get funds except at higher rates. So they were buying money dear and selling it cheap, and had no way of escaping their previous loans. (This was before Fannie Mae began bundling mortgages into marketable "mortgage-backed securities", but even if that had been possible the losses might have wiped out the capital at most S&Ls). At first they treated the jump in (nominal) interest rates as temporary, and just waited it out by shutting off new loans. But as the higher inflation rates persisted, they eventually resumed lending - at high nominal rates.

By that time the U.S. economy had undergone the shockingly high unemployment rate of 11% in 1975. But the unemployment was created in housing and the heavy industries challenged by Japan’s economic emergence (made more dramatic by the demand shift to smaller cars). It was not the Fed “slamming on the brakes” so much as a convergence of serious shocks (add in demographic shifts in the labor market, environmental limits, industry shifting South, productivity slowing and the end of the Vietnam conflict) that caused the pain of 1975. By the time of the 1974 oil shock it would have been too late for the Fed to avoid a serious recession by raising interest rates (what a sentence! What a strange proposition to be rebutting!).

It is probably true that if the Fed had shifted earlier to fighting the built up demand pressures for inflation it could have avoided some of the trouble. The S & L adjustment would have been more gradual, and investment would have downshifted sooner, so that wage demands would not be out of line with the sudden new reality of competition from abroad. But this amounts to saying that hindsight is better than foresight. The problem was not the monetization of the oil shock per se.

That we are still going over the story of the 70s is not so surprising. The 30s live on in economic discussions as well. But it is sad to note that economists have generally bought into the most extreme version reinterpreting the experience, the Milton Friedman version that sees the short run as meaningless and markets as perfectly able to anticipate opportunities. To my mind this is a result of the disconnect between the world of policy economics, where only simple answers will have any effect, and the world of academic economics, where only mathematical modeling will be accepted. The most tractable models take over among academics (Gresham’s Law), so that more nuanced views are not allowed to be heard, and of course the nuanced views would sound too complicated for legislators anyway. It is ironic that Walter Heller, the Keynesian maven who sold Kennedy on tax cuts as a solution to economic stagnation, also argued for tax increases when government deficits rose and the economic slack was gone. There was a time when nuance was understood, at least by Ph.D.’s, and economists had more than one hand.

No doubt policy circles needed to be told “control your deficits” in clear terms. Maybe Keynesian doctrine is just too tempting for politicians, amounting to a soft budget constraint. But when we begin interpreting the experiences that do not fit neatly into the monetarist paradigm, and surely an Aggregate Supply contraction is chief among these, it is important not to try to force the round peg into our square theory.

So - is the U.S. economy growing so fast that interest rates need to rise? Of course. The yield curve is all out of whack. Short term rates are below inflation, and the recovery has taken hold not only in hiring but also, wonder of wonders, in Japan. It is time, as Greenspan said, to take a neutral stance rather than one that promotes growth.

I think it is also time to give Greenspan the first “Two-Handed Economist” award. We should give one every four years to the economist in policy circles who shows the most ability to think clearly about the real world. I am ready to nominate the next, once the obvious Mr. G. receives his.

IMF irony - entirely unintentional

Asia medicine still wrong

In the face of criticism from within the ”mainstream” of economics, most notably by Joseph Stiglitz in “Globalization and Its Discontents”, the IMF has admitted it may not have administered the right medicine in the 1997 Asia crisis. I come to the same conclusion, by way of an issue at once less deep and more focused than the one Stiglitz works from.

The IMF prescribed monetary and fiscal tightness. This standard remedy has logic behind it. It includes very high interest rates - like 30%. It also advises fiscal tightness, but the IMF position on that part changed reasonably quickly as the nature of the crisis became clearer. Fiscal deficits were accepted and remain “correct” in its retrospective views of the situation. However, it still defends the high interest rates (tight monetary policy) that was the heart of its advice. In fact, the program remains defensible -- but wrong.

What the point was

The situation they were addressing was a plummeting currency, caused by hot money trying to get out the door first because the devaluation was eroding the value of their investments. Although the reasoning behind the prescription of high interest rates is more complicated, think of it for the moment as a program of offering high enough interest rates to attract foreign investors to stay in the market. You might think of it as trying to stop a stampede with the smell of gourmet cattle food.

Of course, the panicked exit did not start without reason. The devaluation of the Thai baht that set off the entire crisis caused a self-perpetuating drain on the baht’s value. An excessively large number of Thai borrowers had taken out dollar loans (generally the only kind available on international capital markets), and were struggling to make their loan payments as these were getting more and more expensive in baht terms. Since their income was generally in baht – from rental property, department stores or hotels catering mainly to Thais, or even just taxicabs – they were broke but reluctant to admit it. They were hoping to get by until conditions improved.

The international investors in Thai stocks and bonds knew better, and wanted out. Somehow the downward pressure on the baht needed to be arrested, and the IMF demanded the only steps it feels it can justify. Unfortunately for Thailand, the IMF interest hike was seriously undermining the fundamental health of the economy, but the IMF felt confident that it was more important to send a signal – that the country was willing to bear pain.

More pain, more gain?

This has the effect of making the problem worse. Investors can easily see that the country’s economy is being further damaged, reducing the “fundamental” value of the currency further. Not only is the damage scaring them with the prospect of further devaluation, it is also undermining the confidence that any new loans will actually be repaid, so adding to the “risk premium” of additional interest needed.

The risk of adding to the risk premium is known. In fact, a long history of financial analysis has recognized that one simply does not lend at usurious interest rates, in large part because high rates damage the borrower’s means as much as they compensate for high risk. Unless you can credibly threaten to break the thumbs of the borrower, high interest rates are simply pointless.

Except that the IMF had tried them before and found that they worked, when the borrower is a government and the time frame is limited. The secret was the signal – high interest rates could show government resolve to follow responsible policy. To understand why this was the basic tool in its kit, we need to think about both the underlying IMF mission and the two most recent financial crises before 1997.

The nature of the IMF

The IMF was constructed after the Second World War as a fund to make stable exchange rates possible. The competitive devaluations of the 30s were seen as part of the crisis that brought Hitler to power, and stable exchange rates were one key to the orderly world of economic progress foreseen for the West. The point of a currency fund was to allow a country to stave off a speculative attack on its currency, by borrowing foreign exchange from the fund. In the meantime it might decide to devalue anyway - at a time and pace dictated by prudent judgement rather than market anticipation.

Over the next decades, the 60s and the 70s, the IMF also took on another role. By negotiating conditions for deep loans, it could in effect impose good policy on the countries. This was needed because a country in a payments crisis was almost always itself responsible for the pressure on their currency. The fundamental underlying pressure was nearly always inflation, generated by loose credit as a method of stimulating the country’s domestic economy. The last time an OECD country needed an IMF loan was Britain shortly before the Thatcher years began. With loose money but no willingness to allow the currency to devalue, the Labour government created an irresistible attack on the pound. IMF conditions included devaluation.

The mindset of the IMF has this problem deeply ingrained in it. A payments crisis is interpreted through a filter that almost assumes the problems to be excess liquidity. “Good policy” is therefore seen primarily as conservative monetary and fiscal policy. A disciplined monetary policy avoids inflation, and a (relatively) balanced budget avoids the need to print money. Bad policy, of which the fund has seen more than its critics usually admit, consists of spending money the government doesn’t have and “debauching the currency” to pay the bills. Under pressure to give immediate results, or to postpone pain in hopes that something will turn up, a large majority of countries have gone down this road at least once.

Approving policy

When the debt crisis of the 80s hit, the IMF’s role as policy judge came to the fore. Creditors insisted on an IMF loan before they were willing to accept write-downs, and, even more crucially, to resume short-term lending such as the trade credit that gets goods through the no-man’s land between jurisdictions. If a country got an IMF loan, it was demonstrating that it was sincerely trying to follow sound policy. (The thing that makes most criticism of the IMF from the left so ludicrous is that it makes out the IMF as some kind of international bully, nearly ruling over poor countries, while in fact the only enforcement mechanism it has is the threat of withholding new credit. If a country wants to go its own way, even to renouncing its debts, Cuba style, no one will come after it with gunboats. But it will certainly notice the lack of trade credit, among other carrots withheld by the folks with the money. Essentially, they don’t like to play with someone who makes up their own rules as they go along.)

The payments crises in the 80s were caused mainly by external factors: falling commodity prices, slowing exports, a rising dollar and rising interest rates. But profligate, over-optimistic investment and unwillingness to absorb the pain of higher oil prices had certainly contributed. Many countries needed debt forgiveness and new credit both, and they had not given creditors reason for confidence. The IMF used its conditionality power to push these debtors toward monetary and fiscal policy that would reduce inflationary pressure. It also asked for “structural adjustments” to reduce the role of government in allocating scarce resources, rationed foreign exchange being high on its list. By the time the mid-90s rolled around, the IMF knew a thing or two about what would inspire confidence among creditors.

As the debt crisis dragged on through the late 80s and early 90s, the IMF repeatedly dealt with floods of flight capital in Latin America. The stage was set for the 1995 peso crisis, and the standard medicine had appeared to work very well. If a country had truly repented its sins of the past, and took steps to convince creditors that it would not stumble into the bar again for another round of deficit spending and cheap money, its own economy had generally gotten better (after a painful adjustment period) and lenders had re-opened the credit lines.

(To cut short the period of pain necessary to convince investors, various methods of guaranteeing their good behavior have been proposed, notably the gold standard and Estonia’s “currency board”, which contributed mightily to the wrecking of the Argentine economy when its inflexibility made it impossible to deal with hard knocks of the late 90s. The IMF flirted with these guarantees, acknowledging the benefits of quick credibility. To its credit, it never fully bought in.)

It seemed to work

What was used instead was the rigor of high interest rates, sustained over a substantial period such as six months to a year. If a government had the political moxie to stick that out, the bond buyers were ready to trust them. In the peso crisis, a crisis of confidence par excellence, the pain got so bad that there was talk of its banks going under, and the government had to make real sacrifices to shore them up. But the medicine worked and the $40 billion bailout loans from the US were paid back early. Real interest rates were down to reasonable levels in about six months, and the economy was basically healthy within a year and a half, (due in part to NAFTA).

Unfortunately, the lesson learned by the IMF was that flight capital can be turned around by short-term pain, consistent with its previous experience but lifted completely out of its inflationary context. If it works in two different contexts, it must be a law of nature, right?

Ya gotta be tough. . .

So two years later, when the baht devaluation kicked off a similar spiral of capital flows out of the country, the same medicine was prescribed. Again giving due credit, the IMF did relent on fiscal deficits, perhaps in part because they were dealing with a country that did not have regular oil income to tap. But it maintained the fiction that high interest rates would staunch the capital exit, in the face of continuing evidence to the contrary. Congratulating itself that the currency stabilized with only market incentives, by contrast with the capital controls imposed by maverick Malaysia, the IMF treated the pain created by 30% interest rates as inevitable medicine that must be endured.

To be fair, the crisis in Southeast Asia would have been a tough time for ordinary people anyway. Malaysia’s experience was not much more pleasant, and the return of foreign investment took noticeably longer there, as the IMF had predicted. But the truth is that Malaysia had only imposed capital controls because the IMF medicine was such a bad prescription. Had the Fund simply recognized that the problem was not inflationary policy in the first place, it could easily have put its stamp of approval on a much more moderate program and had approximately the same effect on world confidence. To do this would have required giving up the “trial by fire” approach that had evolved to deal with governments accustomed to saying one thing and doing another.

Perhaps the Fund had itself been burned too many times – getting an IMF deal signed was usually the critical step for easing international credit, and many countries had then failed to meet the agreed “targets.” In addition, foreign investors looking into the soundness of Southeast Asian economies suddenly claimed to discover “crony capitalism” that had somehow escaped their notice before, and the violent change of regime in Indonesia made everyone face the turbulence following strongman government as a sobering prospect they had overlooked.

But that ain't enough. . .

Nevertheless, the IMF had in its power the possibility of asserting the expertise that supposedly gave it the authority to certify policy as sound. If it had done so, the countries could have avoided punitive levels of interest and capital controls. It is entirely possible that the help to the economy would have given more confidence than the attempt to demonstrate resolve.

It is not too much exaggeration to say the IMF behaved as if its own credibility and resolve were the issue. Its defensiveness in the face of Stiglitz’ broad attack is still on display in the “views and opinions” on its website. At least three of the 2002 responses are still there, including two rather wounded and rather personal in tone (in fairness, Stiglitz provoked it with the nature of his attacks, which were unfocused to the point of getting in each others’ way, and themselves included some personal innuendo.) The point at which its defensiveness is most obvious is when they allege that Stiglitz’ vocal dissent at the time of the negotiations undermined the confidence they were trying to establish. As if pointing out the devastating effects of sky-high interest rates would take Wall Street by surprise. As if proof of toughness was the only issue that mattered.

The causes of this single-mindedness should be given careful thought by pundits such as those at Fred Bergsten’s Institute for International Economics. They need to ask: how did the IMF get into a situation where it could not recognize an adequate macroeconomic plan and certify it, but instead resorted to the voodoo of outguessing foreign investor reaction? Or was it even worse – did the Fund know what was really needed and insist on more pain anyway, to shore up its own credibility? One of the things that ought to be told, when Stanley Fischer’s memoirs recount the events of the time, is how much effort went into actually checking with Wall Street.

The IMF quotes the neo-Keynesian

Another telling piece of the puzzle can be found in the reply to Stiglitz posted by Thomas Dawson, director of external relations at the Fund. He quotes Larry Summers (a favorite voice for replies to Stiglitz because of his own sound credentials as a Democrat and neo-Keynesian) saying that there is always a dilemma in a payments crisis, between long term credibility priorities and short-term “financial repair”. “It's a classic problem of a single instrument and multiple targets. Confidence is widely recognized as essential in combating financial crises," said Summers.

Summers had slogged through months of negotiations to line up the lending that tided Mexico over its 1995 liquidity crisis, and credibility was a very familiar problem for him. But in the classic fashion of generals re-fighting the last war, he and the other Washington officials who hammered together the policy for Southeast Asia seemed unable to recognize the differences in the situation they faced in 1997. (And it is sad that current IMF thinking is still willing to interpret his nuanced phrases as an endorsement of putting all the weight on credibility issues.)

I will name just a few crucial differences.
1. Mexico got into trouble by, in effect, lying to its creditors: it hid dollar loans it owed, using an ingenious approach in which the government indexed peso bonds to the dollar, making them “dollar” debts without the formal requirement of reporting them internationally as agreed in negotiations to resolve the 80s debt crisis;
2. It did so at the end of president Salinas’s term, to help ensure the re-election of the PRI, thus putting the credibility of the new Zedillo government squarely in the center of the issue;
3. Thailand’s problem was solvency, not liquidity, as it had taken on dollar loans to earn baht income at an unrealistic exchange rate (though one that had looked realistic before the dollar began its long rise in the mid to late 90s); and
4. Thailand had no large stream of oil income (or prospect of NAFTA) to bank on for long-term solvency, but rather needed the viability of the local economy to pay its bills internationally and to enable the continued progress up the value chain that investors had anticipated.

What could they be thinking?

Why did the IMF have such blinkers on? Stiglitz sees excessive trust in markets at the core of the problem. It is true that the economics profession overreacted to the macroeconomic lessons of the 70s, and that an exaggerated faith in markets (or at least distrust of governments) had taken root in international institutions during the 80s and 90s. The lessons of Enron should help provide some corrective, and economists (who tend to be American and therefore insulated from the real world) are gradually taking on board the spectacular failures of monetarist theory in Japan and, in most direct practice, Argentina. Stiglitz gets it right, I think, when he argues for proper attention to short-term dynamics and demand stimulus.

(Unfortunately he and the other neo-Keynesians have not given the world a solid enough version of how their “market failures” work, or a set of econometric tests that will demonstrate their point. They are too often left with one vague, hand-waving version of the world to counter the vague, hand-waving version of Milton Friedman and his fellow travelers, the neoclassical, competitive equilibrium, market-clearing, Rational Expectationists who pretend that math is an ideal substitute for understanding. “Globalization and its Discontents” reads like a case study in the Big Think Vagueness approach that I would consider even more fundamentally at fault than monetarism per se.)

Yes, I personally think the “market worship” version is too glib. Much of the rhetoric of the right is true, and much is motivated by genuine humanitarian concern, rather than by trying to feather a nest with right-wing think tanks or Wall Street firms. I am more inclined to blame simple human frailty in the form of a tendency to look for facts that confirm your current beliefs. Abetted, I must say, by a deep divide between policy and academic economics.

The academe ought to provide a reservoir of wisdom and a calibration of its advancement, as it would in biology or chemistry. Instead academic economists solve theoretical problems whose main interest is the career enhancement available for impressive displays of mathematical firepower. When it comes time to bridge the gap between theoretical and real issues, trained economists are almost as much at sea as lay persons, and far more likely to wander down the wrong path out of devotion to their theoretical compass. Indeed, Robert Rubin’s performance in government looks better in retrospect than either Stiglitz’ or Summers,’ despite the sincere efforts and incredible intelligence of the latter pair.

I would recommend Alan Greenspan as the rare example of someone who can bridge the worlds, and he has every reason to be proud of the eclectic approach of the Fed on his watch. Because they did not make the mistake of slavishly re-fighting the previous policy issues, we learned that unemployment can indeed go below 6 percent without inflation taking off. That Wall Street has the temerity to blame Greenspan for the bubble of the 90s, and the high tech crash that followed, is a good index of how much their views are to be trusted.

Uncle Milton is betrayed

As a final note on this rather rambling analysis, one of the richest ironies in the 1997 experience is apparent to a neo-Keynesian, but not to those who still have on the international version of monetarist blinkers. The policy dictates out of Washington were derived fairly directly from the ideas of Milton Friedman, most especially the supposed long-run macroeconomic impotence of monetary policy that justifies focusing only on inflation control. This view has never come to grips with the question of what to do when Aggregate Supply is contracting (see most recent blog). The closest I have seen was in a pre-Clinton textbook by neo-Keynesian (more or less) Alan Blinder, who plods through the story as if it is a mere classroom exercise and fails to demonstrate insight into the nature of the assertions being made and assumed.

Friedman, with confidence at least the equal of Stiglitz’, did face up to the question, in the context of explaining the Great Depression. His conclusion (a sort of monetarist mental backflip) is very instructive: that the Great Depression would have been avoided if only the Federal Reserve had actively increased money supplies in the face of the monetary contraction that was setting in as the margin buying unraveled. The Fed, it should be noted, was concerned with its credibility and a need for sound money, at the time. At least Friedman brought in the use of a proxy for Aggregate Demand, in the form of a falling money supply, and allowed as how it might matter.

Yet today’s international monetarists at the IMF, intellectual descendants though not disciples, came out advising tightening of credit in a contraction. Like the Fed in Hoover’s day, they could see the credibility issue all too clearly, and found that other business to be just too murky to be concerned with.

Of course if you believe that the “real” economy is impervious to financial variations, as monetarists supposedly do, then Summers’ trade-off is an illusory one, mattering only in the short term. But if you believe that, then the Great Depression was an unfortunate series of negative shocks, and Friedman’s view should be ignored (as, incredibly, it was.) I suppose we should take some comfort from the fact that the IMF is quoting neo-Keynesian Larry Summers to argue that their side was at least one horn of a dilemma. Perhaps next they will take on board the lesson that the other horn is just as real.