The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« Good News Takes A Break | Main | More On Labor Market Slack »

August 17, 2005

Did I (And My Kind) Cause Argentina's Tailspin?

Another Econ Journal Watch article comes my way, this one implicating U.S. economists in the truly painful travails of the Argentine people around the millennial turn.  According to the Kurt Schuler, the article's author:

Economists whose work in other areas I admire failed to do the research necessary for understanding Argentina’s situation accurately. As a result, their analysis was faulty. When Argentina followed the main recommendations of the consensus, the economy’s rate of decline accelerated...

A review of what U.S. economists said about Argentina shows that many failed to define key terms in their arguments; most ignored readily available data that contradicted the consensus view about Argentina’s economy; and nearly all neglected to examine the legal and statistical material, available for free online, necessary for understanding how Argentina’s monetary system worked. The episode is important because it raises the question of whether the public can trust economists who claim expertise on controversial issues of economic policy.

Yikes!  If legitimate, this is a stinging indictment indeed.  But I have some doubts, at least about part of the Schuler argument.  I, and my co-author Owen Humpage, are fingered for an Economic Commentary article we published in 1999 titled "Dollarization and Monetary Sovereignty: The Case of Argentina."  Here, according to Schuler is where we went awry:

... among the 100 most active commentators on Argentina, 91 of 94 who mentioned the topic called the convertibility system a currency board. Yet examination reveals important differences between the convertibility system and an orthodox currency board. The system was a central bank that mimicked some currency board features; it is perhaps best termed a currency board-like system, or even a pseudo currency board.

I don't know about the other 91 on the list -- a list that includes Brad DeLong, Nouriel Roubini, and Brad Setser, for example --  but I have to enter a plea of "not guilty."  Here is what Owen and I wrote:

Although the new monetary institution created by the Convertibility Law is not a pure currency board, such an unadulterated arrangement is a useful benchmark from which to begin thinking about Argentina’s monetary structure.

To be fair, Schuler provides a footnote to a table the admits we "Occasionally mentioned that the convertibility system was not an orthodox currency board, but on balance seemed to consider the system a currency board."  But I still argue that this is not quite true to the plain message of our article. We followed up our description of this "useful benchmark" with a section describing the actual institutional monetary arrangement in Argentina, which we titled "Argentina's Almost Pure Currency Board".  Maybe the "almost pure" qualifier is not sufficient for Mr. Schuler's tastes, but the main point of our article was that it was precisely the special non-currency-board-like provisions of the Convertibility Law that provided a response to critics who claimed that a currency board arrangement, or even dollarization, was too rigid for Argentina's own good.

Schuler refers to several other "mistakes" made by others --  opinions on whether the currency was overvalued, whether exports from Argentina were uncompetitive, was dollarization technically feasible.  He treats the answers to these questions, the last two especially, as definitive, but it seems to me that there is more room for honest disagreement than the author's views allow.

The advice given by economists may have been good or bad, and the record on that score is well worth exploring.  Furthermore, it is hard to argue with Schuler's plea that those who proffer such advice make the effort to truly understand the institutional arrangements (and socio-political realities) with which the targets of their attentions must deal.  But my suspicion is that a close and objective reading of the record will reveal a better score on along this dimension than he claims.

UPDATE: Brad Setser pleads not guilty as well, and if I was the jury he'd walk.  Be sure to read his thoughts in the comment section below.

August 17, 2005 in Latin America/South America , This, That, and the Other | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Did I (And My Kind) Cause Argentina's Tailspin? :


I'll second you final comment. I don't think i actually wrote anything about argentina until after convertibility collapsed. i worked on Argentina for the government at the time. And since Argentina called its arrangement a currency board, it was rather awkward to call it something else. Saying "argentina's not quite pure currency board" or using "convertibility" and putting a footnote saying it was not a pure currency board is rather awkward.

But we certainly were aware that it was not a pure currency board. Hell, Cavallo spent most of 2001 trying to make it less of a currency board, and to introduce greater flexibility into the regime in various ways. The making it a basket peg to the euro/ $ for example. the variable tariffs. some changes in domestic reserve requirements.

Moreover, it was totally clear that in 2001, a pure currency board would have implied a much more rapid fall in the money base and more deflation to generate the real depreciation needed to bring argentina's current account deficit (note -- current account deficit, not trade deficit) down to a level that was consistent with available external financing. and if argentines wanted to shift funds abroad, they needed to run a current account surplus.

Argentina wasn't willing to do this. It meant shrinking nominal government spending to match the fall in nominal tax revenues. That did not happen -- rather the government, particularly local governments, started issuing local quaisi-currencies.

Posted by: brad | August 17, 2005 at 05:20 PM

Three other comments:

1: Hanke and Schuler argue that the failure of convertibility in no way casts doubt on the desirability of currency boards, 'cause convertibility was never a true currency board. But remember, that in the fact of an external financial shock, a true currency board would imply bigger falls in base money and faster deflation. i.e. the internal adjustment would have been more brutal. I also don't remember Hanke loudly criticizing Argentina's not quite a currency board back when Argentina was doing well ...

2: To me, the trade argument is a bit of a red herring. Argentina had large external debts relative to its exports, and paid a high and rising interest rate on its external debt. So the net interest payments part of the current account was large and negative. Those interest payments could have been financed if Argentina was running a trade surplus. But it wasn't. Even after the 99-00 recession, Argentina's trade was roughly in balance. and in 2001, Argentina ran a current account deficit despite experiencing a $20 billion (private) capital outflow -- a deficit that was financed by running down reserves (gross reserves fell $10 b), borrowing from the imf and running down borrowed reserves(net reserves fell by $20b) and by running down various external assets (banks foreign assets fell by $10 b).

3: Moody's reports nominal exports grew by 1% in 98, fell 10.5% in 99, grew 11% in 00, and then were flat in 01. I think the 11% growth in 00 had something to do with the global rebound in commodity prices. 2000 was a good year for the world economy, and particularly good year for commodity exporters. Argentina exports oil and gas, along with wheat and soybeans. Over a four year period though, exports were basically flat. Schuler I think used a much longer time series to argue that export growth was not hurt by the currency board. For me the real question was whether exports could be expected to grow fast enough to get rid of the current account deficit after the series of shocks in 99 (devaluation in real, $ appreciation). I don't the data provided much support for a sustained export boom in the face of the stronger real peso after 99.

Posted by: brad | August 17, 2005 at 05:35 PM

"But remember, that in the fact of an external financial shock, a true currency board would imply bigger falls in base money and faster deflation. i.e. the internal adjustment would have been more brutal."

Yep, I'd say that this was the point. Why, hell, Italy has now got itself a spanking bells and whistles version 3.0 currency board (or should this be deutschmark-isation), and what's the betting this one is going to work.

Isn't the substantive point that plenty of people read Argentina completely wrong with disastrous results for the Argentinians, or what's the difference between Argentina and Turkey? Meanwhile I have the horrible feeling too many people are making the same mistake with the euro. Mind you, no-one will be able to accuse Marty Feldstein of getting it wrong :).

Posted by: Edward Hugh | August 18, 2005 at 08:30 AM

"Did I (And My Kind) Cause Argentina's Tailspin?"

This is a bit like the question "did the UK participation in the Iraq war cause the London bombings"?

Clearly it didn't, the terrorists caused it. Iraq involvement may well have been an aggravating factor.

Likewise Argentina's tailspin would be down to the institutional system and the political class in Argentina, but some of the advice they received most probably aggravated things. Especially the decision at the begining of 2001 to extend the credit and try to defend the peg, that's the one I really don't understand.

Posted by: Edward Hugh | August 18, 2005 at 08:53 AM

Oh I am tiresome, here I am again. The thing is it worries me that the lessons of Argentina have not been assimilated, and I keep shoving the boat in on the euro as I think if people were clearer about what went wrong in Argentina they might be able to understand what happens next in Europe a bit better. Reading your article Dave, you yourself were aware of the parrallel at the time:

"The costs of lost policy discretion are substantial. If this were not so, the United States and other large industrialized countries might forgo fiat money and return to a system based on gold or some other commodity. But the costs of low confidence and uncertainty about the exchange value of a currency can also be large. If this were not so, each of the European Monetary Union states would not have agreed to forgo issuing its own currency under its own independent monetary policy."

As you rightly say the costs of low confidence and currency uncertainty can be large. The issue is whether abandoning monetary policy autonomy resolves these problems by allowing them to be attacked at root, or simple shelves them for a time.

The latest relevant research on structural reorms in the eurozone by Romain Duval and Jørgen Elmeskov suggests that EMU may be an impediment rather than a stimulus to the needed structural reforms. That's the theory, and the reality seems to back it up if you look at Italy, Greece and Portugal: balance of payments issues, lots of debts, slow reforms, low (or in Italy's case negative) productivity improvements, lethargic growth.

Maybe you think the jury is still out, as far as I'm concerned the guilty verdict is in and its just a question of waiting for the sentence to be passed. Meantime I'm afraid until we make a collective adjustment on this one the ghosts of times past (in this case Argentinian ones) will still come back to haunt us.

Posted by: Edward Hugh | August 18, 2005 at 09:17 AM

Sorry, I'll stop trying to do some serious work and get back to blogging soon, honest I will. That way I won't be clogging you up with inane comments. But continuing my one man campaign against the *dismal* science, and with all those other Melanie Safka fans out there in mind:

I've Got a Brand New Pair of Roller Skates (You've Got a Brand New Problem)

For roller skates read yatch, rolex, mobile phone, washing machine, TV and anything else Berlusconi may have added to the list while I've had my back turned.

Posted by: Edward Hugh | August 18, 2005 at 09:34 AM

Schuler.... I think we know what to make of somebody who argues that he is right and Milton Friedman and Anna J. Schwartz are wrong on issues of monetary economics.

Posted by: Brad DeLong | August 18, 2005 at 12:49 PM

David, here are my responses to the thoughtful points you and Brad Setser raised. They express my personal views, which are not necessarily those of the U.S. Treasury Department.

1. I did not claim that the economists whose writings I surveyed caused the severe problems Argentina experienced in 2002. Rather, they provided intellectual cover for decisions by the Argentine government. Those decisions made 2002 a much worse year than it need have been.

2. A “pure” currency board is presumably one that measured along certain dimensions is 100%. Your article did not define “almost pure” numerically. I consider a generous allowance to be within 20 percentage points of “pure.” Table 3 of my article shows several ratios calculated from IMF statistics of balance sheet numbers for Argentina’s central bank. *No* median of the monthly statistics fell within the “almost pure” range. The two statistics that I consider most important are (a) the correlation between the change in net foreign assets and the net monetary base and (b) reserve pass-though, that is, how much of every dollar of change in foreign reserves passes through to a change in the monetary base. For Argentina during the convertibility system, (a) was 47% and (b) was 31%.

3. My criticism is not only that economists got important points wrong about Argentina. Even where I think they were right, they often failed to provide evidence. For example, most economists who expressed an opinion thought, as I did, that dollarization was feasible towards the end of the convertibility system. Few showed signs that they had examined the central bank’s balance sheet, which had a direct bearing on the subject.

4. Turning now to Brad Setser’s points, I reviewed postmortem commentary on Argentina’s depression, such as his, along with commentary written at the time of the crisis.

5. Although Argentine government officials sometimes referred to the convertibility system as a currency board, conscientious economists should compare words with deeds. Perhaps even government officials and agencies should do so in their public statements; from what I could find, the U.S. Treasury at the time did not. After the fact, Domingo Cavallo, who was minister of economy when the system began, wrote, “El texto de la Ley de Convertibilidad es muy claro al establecer la multiplicidad de monedas y nunca creó un “Currency Board” convencional, como lamentablemente muchos economistas interpretaron.” (The text of the Convertibility Law is quite clear in establishing [the right to use] multiple currencies; it never created a conventional currency board, as many economists unfortunately interpret it to have done.)

6. Yes, other things equal, a currency board would have produced a larger contraction of the monetary base in 2001 than the convertibility system did. But would other things have been equal under a currency board or the still more “cast-iron” system I proposed starting in early 1999 — official dollarization? (I proposed official dollarization after it became clear that Argentina would not change the convertibility system into a currency board.) No direct evidence is available, but the experience of Ecuador provides some indirect evidence. Ecuador dollarized in January 2000 in the midst of an economic meltdown. Bank deposits soon began rising and the monetary base apparently grew. (“Apparently” because no accurate breakdown of dollars in circulation in Ecuador is available.) The Argentine and U.S. governments estimated that “mattress money” held in dollars exceeded the monetary base in pesos. Under a monetary arrangement that generated more confidence than the convertibility system did, mattress money could have flowed into the banking system, as it seems to have done in Ecuador.

7. Steve Hanke wrote as early as October 1991 that the convertibility system should be changed into a currency board, and he and I were pretty consistent in stressing throughout the life of the system that we considered it an improvement over what it had replaced but thought its deviations from currency board orthodoxy invited problems. Footnote 8 of the article documents this point and Appendix 2 supports it with half a dozen or more sample quotations.

8. In my opinion, the problem by mid 2001 was not that the peso was overvalued, but that the government was too heavily indebted. Hence I favored default, if necessary, in preference to devaluation. Devaluation unnecessarily punished the private sector for the mistakes of the government.

9. Table 4 of my article shows that after 1999, a very bad year because of Brazil’s devaluation, Argentina resumed a pattern of growing export volume. Argentina’s share of the dollar value of world exports also grew, indicating that Argentina did well relative to the world average. The value of Argentina’s exports would have been higher in 2001 had bank deposits not been frozen for the month of December, causing economic activity to grind to a halt. For the first 11 months of the year they were 2.5% ahead of 2000, but in December 2001 they were almost 16% below December 2000. Export growth for the full year 2001 was therefore 0.8%. Even that was respectable given that world trade actually shrank in 2001. Since the convertibility system ended, Argentina’s share of world exports has fallen every year despite big increases in the prices of its major commodity exports.

10. Anybody who has read this far other than David, Brad and me needs to get a life.

Posted by: Kurt Schuler | August 18, 2005 at 12:50 PM


1) Ecuador dollarized after a massive depreciation of the sucre. That makes a difference. The comparable policy proposal for Argentina was "Devalue and dollarize." Ecuador also benefited from a major rebound in oil in 2000.

2) I do not quite see how a default would be consistent with dollars under the mattress (or dollars in Miami) flowing into the banking system. Remember, one of the main assets of the banking system was GOA dollar bonds. For more details, see the 2004 balance sheet paper from the IMF (rosenberg, keller, nystedt, setser and a cast of thousands). Moreover, if the economy would have continued to deflate, and with a shrinking nominal GDP and constant dollar debts, i suspect lots of private debtors would have defaulted as well. The quality of the assets in the argentine banking system was deteriorating rapidly. (I'll leave the banks loans to utilities that could not only price in dollars but index their price increases to us inflation in a deflating argentine economy aside, but suffice to say that i don't think continuation of that bargain was politically feasible)

3) Argentina had sufficient dollar reserves to replace pesos in circulation with dollars. But presumably, all peso denominated bank deposits would also have been dollarized -- and the banking system's dollar liquidity would have been constrained. Realistically, after the default, I suspect that the dollar deposits in at least some banks would have started to run, and barring a massive infusion of dollar liqidity from the IMF, the dollar deposits would have had to be frozen and restructured.

4) Argentina was not willing to live with the constraints of convertibility. It started issuing quaisi-currencies rather than cut salaries to match falling peso revenues. Barring a miraculous recovery/ the end of capital outflows -- even with a default -- the same deflationary forces would have remained. to fund capital outflows, Argentine needed to squeeze its economy to generate a current account surplus. A restructuring that cut external interest payments would reduce net interest payments in th current acocunt, but would (in my judgement) increase capital outflows. So the squeeze on the domestic economy would have remained. I am not convinced that a dollarized Argentina would not have started to issue quaisi-currencies rather than accept the constraints of dollarization.

Posted by: brad | August 18, 2005 at 03:27 PM

David, BRAVO!!!!!!!!!! This is FANTASTIC, thanks. It's reminded me how tough self-examination & change are for any profession. I don't have a clue how one would transpose the argument to U.S. relevance, but what you've done is what's called for. GREAT JOB, thanks again. (I WAS right.)

Posted by: bailey | August 19, 2005 at 06:36 AM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

Google Search

Recent Posts



Powered by TypePad