Go to Brad De Long's Home Page
Comment on Robert Pollin, "Can Domestic Expansionary
Policy Succeed in a Globally Integrated Environment?"
Keynesianism in One Country?
University of California at Berkeley
January 3, 1997; revised
Let me focus on Mexico, because Mexico is our most recent example, and a brutal example, of the limits placed on domestic expansionary policy by the integration of the world's financial market.
In the late 1980s, the authoritarian PRI party ruling Mexico shifted its economic policies in a "neoliberal" direction. Instead of extremely tight restrictions on imports (even imports of the capital goods that serve as one of the major channels of technology transfer to the third world), they adopted policies of freer trade. Instead of the high-inflation high-deficit high-spending mix that had characterized Mexico's fiscal and monetary policies in the past, they sought monetary stability and balanced budgets. Instead of expansion of the publicly-owned sector (both to channel income and wealth to politically favored groups, and to try to make sure that windfall profits were not skimmed off by foreign elites), they sought "privatization" both to attract capital from abroad and to try to give organizations incentives to improve their productive (as opposed to their political) efficiency.
The hope was that such a shift to "neoliberal" policies would pay benefits in a number of directions.
First, the privatization program might--it was hoped--lead to sharp increases in productivity in the formerly publicly-owned sector. Firms whose principal goals had been to transfer money to favored clients of the PRI oligarchy would find themselves, instead, with the principal goal of earning profits--and satisfying consumers as a step toward earning profits. Second, the liberalization of domestic financial markets might--it was hoped--shift the allocation of investment. Instead of access to capital being limited to those who knew the right people, it might become open to those who could see how to profitably and productively satisfy a strong consumer demand. Third, the reduction in government deficits might--it was hoped--boost the social rate of investment, and thus boost the rate of productivity and real income growth.
Fourth, freedom to buy imports would raise middle-class standards of living, and might make Mexico's upper middle class happier with the PRI. Fifth, a shift to "neoliberal" policies might attract large enough capital inflows to allow Mexican investment to outpace domestic savings, further boosting the rate of productivity and real income growth. Sixth, a shift to "neoliberal" policies would make high Mexican officials heroes in the world financial governance community--boost their reputations abroad at least.
Thus Mexico's steps toward "neoliberal" policies were taken for a bunch of reasons: some good, some bad, some naive, some subtle.
I happen to think that the policy shift toward neoliberalism was a good--perhaps the best of a number of bad options, certainly not the worst--bet for a Mexican government to have undertaken in the late 1980s. It appeared likely to have led to somewhat faster aggregate economic growth than many other alternatives, albeit at a social cost (as Mexico's subsistence agriculture sector would wither under the force of competition from Iowa). But the social cost could be kept small with appropriate redistributional programs, and the PRI would have every incentive to keep the social cost small.
Were there preferred alternatives? The East Asian "developmental state" model is very attractive: it promises a relatively egalitarian distribution of income, rapid productivity and production growth, rapid transformation from a comparative advantage based on cheap labor to one based on mastery of manufacturing technologies, increasing democratization, and a government able to make the redistributions and carry out the infrastructure investments to diminish the social costs of industrialization.
But as Lant Pritchett has observed, there are few things worse than state-led development carried out by an antidevelopmental--a parasitic--state. It seems clear to me that the PRI government in Mexico does lack that relative autonomy and bureaucratic capability found in East Asia; it seems clear to me that the U.S. government lacks the relative autonomy and bureaucratic capability to successfully carry out the tasks of the "developmental state" model. And it seems likely that a neoliberal development strategy is not the worst way to play out the relatively bad hand held by those whose governments lack the prerequisites to follow in the footsteps of the East Asian miracle.
But we can argue about what would happen if a "parasitic" state tried to turn itself into a "developmental" one. Maybe the PRI made a mistake in opting for a "neoliberal" rather than a "developmental" policy mix. That is not my subject: my subject is the consequences of the fact that in 1994 Mexico took a policy step away from neoliberal orthodoxy.
Mexico took a step away from neoliberal orthodoxy for a number of reasons. First, there was the forthcoming "election"--which I hope will turn out to have been the last Mexican "election" that one will have to enclose in quotation marks. Second, aggregate economic growth had been disappointing over the previous several years: as Rudi Dornbusch and Alejandro Werner put it, Mexico had wound up with "reform, stabilization, and no growth". In part for these and in part for other reasons, the Mexican government shifted its monetary policies in a strongly expansionary direction in early 1994.
Some ( for example, Calvo, Leiderman, and Rinehart; Dornbusch and Werner) saw this step away as dangerous. In the view of Calvo et al., the tightening of monetary policy in the industrial core that took place in 1994 required not a loosening but a sharp tightening of monetary policy in Mexico in order to avoid a substantial devaluation, and the consequent loss of confidence and spiral of renewed inflation. In Dornbusch and Werner's view, the exchange rates that the Mexican government had set in the early 1980s left the peso overvalued: in spite of propaganda to the contrary, Mexico was not attractive enough as an export platform for the U.S. market. In their view, the sooner the devaluation came, the better for Mexico.
Yet the consensus--of which I was a very small and unimportant part--was that Mexico's step away from neoliberal orthodoxy in economic policy in early 1994 was unlikely to be an event with major consequences.
Capital flows into Mexico had put pressure pushing the peso upward, not downward, in 1993. There was reason to hope that the shift to neoliberal policies and the implementation of NAFTA would excite strong capital inflows over the following decade. Such strong capital inflows would help finance Mexico's industrial revolution: if even a small portion of Mexico's export manufacturing sector could use foreign-financed capital to invest and attain the productivity found in, say, Hermosillo then Mexico's economic growth future looked very bright. And rapid manufacturing production growth would provide the best circumstances for Mexican industrial workers to push for substantial real wage increases.
The neoliberal consensus view was (i) that Mexican economic policies had made it attractive to global capital; (ii) at this attraction was based on a sober and likely prediction of Mexico as a productive and profitable platform for exports to the U.S. market (iii); that these fundamentals were working in Mexico's favor--boosting investment, strengthening the peso--and (iv) that given these strong fundamentals a minor deviation from neoliberal orthodoxy like the loosening of monetary policy in an election year (a loosening that appeared likely to be temporary should the PRI retain or the PAN attain control of the presidency, and that was not large in absolute magnitude) should have little if any effect on the Mexican economy. Continued growth depended on foreign investors' continued confidence in Mexico as an export platform. And while confidence in Mexico might be undermined by large fiscal deficits funded by money-printing, it would not be undermined by the more-or-less standard election-year operation of the political business cycle.
This neoliberal consensus view was very wrong.
Monetary expansion in 1994--coupled with political assassinations and guerrilla rebellions--put downward pressure first on Mexico's stock of reserves, and then on the exchange rate. The Mexican government paid out some $20 billion or so in eight months trying to defend the peso.
December 1994 saw a devaluation--a small devaluation. It seemed to those who modeled the Mexican economy that this small devaluation of 20-30 percent or so, the size recommended by Dornbusch and Werner, would be sufficient to contain the effects of the monetary expansion and the political instability of 1994. If Mexico had been a profitable export platform at the higher exchange rate, it was an even more profitable export platform at the lower. If foreign investors had been eager to pour capital into Mexico when keen-eyed observers like Dornbusch, Calvo, and their co-authors saw devaluation as likely, they would be even more eager to pour capital into Mexico now that no one was arguing that the peso was overvalued.
A classic liquidity panic followed. Never mind that to an investor interested in Mexico over the long term, Mexico post-1994 (with its lower real exchange rate) would appear a better bet than Mexico pre-1994. Never mind that everything that had been said in the early 1990s remained true. The Mexican economy had guaranteed access to the largest consumer market in the world. Politics had shifted in a pro-market pro-capitalism direction. Mexico had become one hell of an export platform for any kind of manufacturing where technology is sufficiently routinized to be transferrable out of the industrial core.
But in spite of "fundamentals" stronger in 1995 than in 1993, the dominant forces governing financier decision-making in early 1995 were (a) let's get out of Mexico so that I no longer have to explain what it's doing in my portfolio, and (b) let's get out of Mexico before everyone else's running for the exit brings on a Great Depression in Mexico.
You could point out that those investors with long enough horizons to avoid the panic selling would avoid much of the losses from the crisis (albeit at the cost of running enormous risk). You could point out that those willing to be "contrarian" and put money into Mexico would (if Mexican politics did not degenerate into near-anarchy) realize very high returns. You could point out that an investor who thought that in terms of long-run fundamentals Mexico was a good deal in 1993 could not think that Mexico was a bad deal in 1995 without violating the canons of consistency and rationality in financial decision making.
So much the worse for consistency and rationality in financial decision making.
A lot of money was left on the table by investors in New York, London, Frankfurt, and Tokyo who folded and withdrew from the Mexican part of the global financial poker-game-cum-casino. Large expected profits (accompanied by enormous risks) were earned and are being earned by those who stayed in the game: in the fall of 1996 U.S. Vice President Al Gore defended U.S. support for the peso in terms of the high ex-post returns earned by the U.S. Treasury.
Either the flush of investment in Mexico in 1993 or the stampede out in 1995--or both--reflect substantial failures of those who make decisions in world financial markets to perform their social function of being a means-ends rational social capital allocation mechanism. All these things are true.
But that didn't help Mexico, which saw real output fall by between five and ten percent between 1994 and 1995.
The most likely explanation is that investors, fund managers, and financiers lack the sophistication that economic theorists believe they ought to possess in assessing the financial effects of macroeconomic policies. Economists focus on "structural" and "primary" deficits as indicators of whether countries' public finances are or are not in long-run balance, and whether or not countries will ultimately have to resort either to massive inflation or to massive tax increases to fund their spending. But financial markets appear to pay much more attention to reported deficits--and to fail to comprehend or pay attention to the more sophisticated measures.
Economists focus on the determinants of the money supply process--not only whether the central bank is allowing the money stock to grow, but why the central bank is allowing the money stock to grow and what the money stock will do in the future. Thus they make a sharp distinction between temporary increases in money growth (as in, for example, an election year) and permanent increases (as when a central bank trying to keep interest rates low monetizes large persistent government budget deficits): the second should be inflationary; the first should not. Yet financiers and fund managers appear to examine the news with a much less sophisticated, much more knee-jerk model--in which all deviations from financial orthodoxy are equally sinful, no matter whether economists' models classify them as mortal or as only venial sins.
In some ways this is an old story. The U.S. Treasury Department has in its art collection a magnificent portrait of 19th century financier Jay Cooke, without whose fundraising and bond selling skills Sherman's armies would never have gotten to Atlanta and Grant's armies would never have gotten to Richmond. He went spectacularly bankrupt in 1873 having played double-or-nothing with his, his bank's, and his clients' money one time too many while trying to build the Northern Pacific Railroad.
After the Jay Cooke bankruptcy British investors fled the United States. And perhaps one in ten nonfarm jobs in the American economy vanished as a result of the collapse in railroad construction alone. To rely on foreign capital inflows to finance one's industrialization was an extremely risky development strategy for the U.S. in the 1870s. It is an extremely risky development strategy in the 1990s as well.
So why run the risks? Why not simply close off all capital flows--except for a small selection? Because the temptation is enormous: when foreign investors are willing to finance domestic industrialization, extraordinary gains in productivity--private and public--are possible at a very low cost. Capital is cheap in the first world. Capital is expensive in the third world. There is every temptation for those searching for means to fund industrialization in the third world to try to tap the cheap capital of the first.
So what is to be done? Three strategies:
I don't know which of these strategies is best. My hunch is that the second is--because the benefits of foreign capital as a source of financing for development are so large (when it flows in) as to make the first seem unwise, and because I doubt that governments today can run regulatory regimes effective enough to accomplish the third. But I really do not know--and would not be surprised if evidence pushed me to any of the three positions over the next decade or so.
But what I do know--after Mexico 1994--is that even what I saw as the little bit of running room left to macroeconomic policy in developing countries is probably not there if they want to draw on global capital to finance industrialization. It is quite possible--it is highly likely--that the expansion of Mexican monetary policy in 1994 was a good thing for foreign investors in Mexico: increasing the likelihood that the PRI will hold on for an extra election cycle, and that the PRI's successor will be on its right rather than on its left. A government willing to boost the economy to try to convince voters that pro-market pro-foreign investment policies were working seems to me preferable from the viewpoint of a foreign investor to one that holds to monetarist orthodoxy in all cases.
But apparently not. Great Britain (in 1992) could get away with announcing that it was abandoning some of the contractionary restraints it had placed on its own policies, devalue the pound against the mark, and have lower interest rates six months later.
But Mexico in 1995 could not. The world is not fair.
Go to Brad De Long's Home Page