DRAFT

The dramatic events in Asia and Russia, and the contagion they have spread, have generated a torrent of commentary about exchange rates, hot money and exchange controls. As someone who was deeply involved in Indonesia (Review Editorial,1998) and who predicted that the ruble would collapse by midyear (Hanke, 1998), I offer my views as to why most of the commentary has been either half-baked or simply wrong. Indeed, in most instances the international chattering classes have misdiagnosed the patient, and in consequence, prescribed the wrong medicine. A correct diagnosis requires an understanding of alternative exchange-rate regimes.

Exchange Rate Regimes

There are three types of exchange-rate regimes: floating, fixed and pegged rates. Each type has different characteristics and generates different results (see Table 1). Although floating and fixed rates appear to be dissimilar, they are members of the same family. Both are free-market mechanisms for international payments. With a floating rate, a monetary authority sets a monetary policy, but has no exchange-rate policy--the exchange rate is on autopilot. In consequence, the monetary base is determined domestically by a monetary authority. Whereas, with a fixed rate, a monetary authority sets the exchange rate, but has no monetary policy- -monetary policy is on autopilot. In consequence, under a fixed-rate regime, the monetary base is determined by the balance of payments. In other words, when a country's official net foreign reserves increase, its monetary base increases and vice versa. With both of these free-market exchange rate mechanisms, there cannot be conflicts between exchange-rate and monetary policies, and consequently balance of payments crises cannot occur. Indeed, under floating and fixed-rate regimes, market forces act to automatically rebalance financial flows and avert balance of payments crises.

While both floating and fixed-rate regimes are equally desirable in principle, it must be stressed that floating rates, unlike fixed rates, do not perform well in developing countries because these countries usually have weak monetary authorities and histories of monetary instability. For a recent dramatic example, we have to look no further than Indonesia. It floated the rupiah on August 14, 1997. Unfortunately, but not surprisingly, the rupiah did not float on a sea of tranquillity. Indeed, the rupiah fluctuated wildly and lost over 80% of its value against the U.S. dollar in six months.

Another example of the perils of floating is illustrated by the experience of Hong Kong (Culp and Hanke, 1993). It floated the HK dollar in November 1974, when the HK dollar was trading at about 5HK$/US$. Prior to the float, Hong Kong operated with a currency board and a fixed exchange rate, with sterling as the anchor currency.

After the float, the HK dollar was volatile and steadily lost value relative to the U.S. dollar. The volatility reached epic proportion in late September 1983, after the end of the fourth round of Sino-British talks on the future of Hong Kong. Indeed, financial markets and the HK dollar went into tailspins.

At the end of July 1983, the HK dollar was trading 7.3 HK$/US$. By "Black Saturday," September 24th, the HK dollar had fallen to 9.55HK$/US$, with dealer spreads being reported as large as 10,000 basis points. Hong Kong was in a state of panic, with people hoarding toilet paper, rice and cooking oil.

The chaos ended abruptly on October 15th, when Hong Kong reinstated a currency board system. The new anchor currency was the U.S. dollar, and the exchange rate has remained fixed at 7.8HK$/US$ ever since.

Fixed and pegged rates appear to be the same. However, they are fundamentally different. Pegged rates are not free-market mechanisms for international payments. Pegged rates require a monetary authority to manage both the exchange rate and monetary policy. With a pegged rate, the monetary base contains both domestic and foreign components. Unlike floating and fixed rates, pegged rates invariably result in conflicts between exchange rate and monetary policies. For example, when capital inflows become "excessive" under a pegged system, a monetary authority often attempts to sterilize the ensuing increase in the foreign component of the monetary base by reducing the domestic component of the monetary base. And when outflows become "excessive," an authority attempts to offset the decrease in the foreign component of the base with an increase in the domestic component of the monetary base. Balance of payments crises erupt as a monetary authority begins to offset more and more of the reduction in the foreign component of the monetary base with domestically created base money. When this occurs, it's only a matter of time before currency speculators spot the contradictions between exchange rate and monetary policies and force a devaluation. This is what happened in Turkey and Mexico in 1994. The same story repeated itself during the summer of 1997 and most recently in Russia. It's time for the international chattering classes to drop their dogma and learn a lesson from history, least history repeat itself.

The Proof of the Pudding Is in the Eating

Base money is the purview of monetary authorities, either central banks or currency boards. Central banks have broad latitude to create base money. Currency boards do not have that flexibility. A currency board system requires domestic notes and coins, as well as deposits held at the board, to be fully covered (usually at 100% to 110%) by foreign reserves denominated in the board's designated foreign anchor currency. Furthermore, the domestic currency must trade, without restrictions, at an absolutely fixed exchange rate with the anchor currency. An orthodox currency board cannot create credit. Therefore, it cannot extend credit to the fiscal authorities or act as a lender of last resort for the banking system. Currency boards run on automatic pilot, with changes in the monetary base determined solely by changes in the demand for domestic base money -- the balance of payments. Table 2 presents the characteristics of currency boards and contrasts them with those of central banks.

Currency boards have a rich history. The first currency board was established in 1849, and beginning about 1913, that system spread rapidly throughout most parts of the world. In the 1950s and 1960s, many currency boards were abandoned and replaced by central banks. This was a remarkable development. The performance of currency boards had been excellent. All had maintained full convertibility into their anchor currencies. Furthermore, countries with boards had realized price stability, respectable economic growth and balanced government budgets.

The demise of currency boards resulted from a confluence of three factors. A choir of influential economists was singing the praises of central banking's flexibility and fine-tuning capacities. In addition to changing intellectual fashions, newly independent states were trying to shake off their ties with their former imperial overlords. And the IMF, anxious to obtain new clients and jobs for the boys, lent its weight and money to the establishment of new central banks. In the end, the Bank of England provided the only institutional voice which favored currency boards. That was obviously not enough. Politics, not the economic record, prevailed.

The picture began to change in 1983. After the currency crises of May-September 1983, Hong Kong stopped floating and reestablished its currency board system. In the 1990s currency board systems have gained broader acceptance, with boards being established in Argentina (1991), Estonia (1992), Lithuania (1994), Bulgaria (1997), and Bosnia (1998).

These developments trouble some analysts who fret about the inflexibility of currency boards. The Economist summarized these sentiments in a piece, "The Great Escape," which appeared in the May 3, 1997 issue. That article asserted: that currency boards cannot cope with external shocks; that they are vulnerable to surges in inflation triggered by capital inflows; and that with limited lender of last resort capacities, they cannot deal effectively with financial emergencies.

The evidence does not support these oft-repeated assertions about the alleged drawbacks of currency boards. Let's look at the data from 98 developing countries during the period 1950-1993 (see Table 3). The data are separated into one of two categories: countries that have pegged exchange rates and those that have fixed rates. The latter category includes countries with currency boards, monetary institutes, and those that rely solely on foreign currency. Countries with currency boards or board-like systems (fixed exchange rates) have had average growth of rates that were 54% times higher than those with pegged exchange rates. Furthermore, the variability of those growth rates (as measured by their standard deviations) was virtually identical, indicating that the lack of discretionary monetary policy with fixed exchange rates did not result in any greater incidence or vulnerability to external economic shocks. As for inflation, fixed rates have proved far superior to pegged rates, with average inflation rates being 4.9 times higher in countries with pegged rates and 4.2 times more variable. In terms of budget deficits, those countries utilizing pegged rates have had deficits which on average were 65% larger. Finally, countries with fixed rates have experienced fewer financial emergencies.

Not only has the relative performance of fixed rates (currency boards) been favorable, but so has their performance over time. When currency boards have been introduced, credibility has been established, the demand for local currency has increased along with foreign exchange reserves. Consequently, through interest rate arbitrage, interest rates have fallen dramatically.

The superior performance of fixed rates (currency boards) has also been reported in (Schuler, 1996) and (Ghose, Gulde and Wolf, 1998). That said, it is important to acknowledge that currency board systems are not trouble free. Argentina provides but one case study.

Some Thoughts on Argentina's 1995 Crisis

The Convertibility Law (currency board law) was tested in November 1992 and April 1994, when unfounded fears of a devaluation briefly pushed short-term peso interest rates into high double digits. In November 1992, for example, financial markets interpreted a package of new tariffs and export subsidies as a signal that the next step might be a devaluation. The monetary base shrank 300 million pesos, or 3 percent, and interest rates on short-term peso deposits peaked at 85 percent. However, those tests were brief compared to the test Argentina underwent after the Mexican peso was devalued on 20 December 1994. (See Table 4 for a summary of the major crisis events in Argentina). Mexico's currency troubles caused many foreign investors to reappraise investments throughout Latin America in a rather indiscriminate manner. Financial markets slumped not only in Mexico but in Argentina, Brazil, and Chile.

The shock imposed on Argentina by Mexico's financial crisis had four distinct phases (Banco Central, Republic of Argentina, 1995). The first phase lasted from 20 December 1994 through February 1995. External drains form the currency board-like system occurred, with the central bank's liquid reserves falling from $15.8 billion before the crisis to $13.3 billion at the end of February. (Note that Argentina's system is not orthodox because only 80% of the system's reserves must be held in dollar denominated assets issued by foreign governments. These are called "liquid reserves." The remaining reserves must be dollar denominated, but can be issued by the Republic of Argentina. Both types of reserves must be valued at current prices.) There were also internal drains of peso deposits from commercial banks. Wholesale banks and small retail banks were most strongly affected. Two small wholesale banks, with a high proportion of their assets in Argentine government bonds, were suspended. (See Table 5 for a summary of key financial indicators.)

During the first phase, the broad money supply, M3 (pesos outside banks plus peso and dollar deposits), decreased by 3.6 billion pesos, or 6.5%, by the end of February. And bond prices fell sharply, with yields on peso denominated bonds moving from 22.6% before the crisis to 38.9% at the end of February. The prime rate on peso denominated loans also increased during the period, from 12.4% to 22.7%. Dollar denominated bond yields and the dollar prime rate also increased, but not by as much as those on comparable peso denominated instruments. The peso-dollar bond yield spreads, for example, increased from 480 basis points to 930 basis points (Pre 1 vs. Pre 2 bonds), reflecting an increase in the perceived exchange rate risk.

The central bank took steps to tighten the link between the peso and the dollar. On January 12th, it eliminated the spread between buying and selling rates for dollars, making the rate exactly 1 peso = US$1. It also required banks to hold their current accounts at the central bank in dollars instead of pesos (these accounts are used for clearing and reserve requirements). That reduced the central bank's potential gain from a devaluation.

To increase the liquidity of banks, the central bank temporarily reduced reserve requirements on deposits on the 28th of December and again the 12th of January. On January 12th, it also established a safety net (lender of last resort facility) financed with 2 percentage points of existing required reserves. These funds could be loaned to solvent banks with liquidity problems. Previously, there had been a private, voluntary safety net to buy loans from banks with temporary liquidity problems.

Subsequently, the government took further action to assure bank liquidity. Decree 286 of February 27th created the Fiduciary Fund for Provincial Development to help privatize banks owned by provincial governments, many of which were notoriously weak. Decree 290 of February 27th amended the Organic Law of the central bank to broaden its power to lend to illiquid banks, and amended the Law of Financial Institutions to allow the central bank to play a more active part in reorganizing troubled institutions.

The liquidity squeeze that the financial system had endured became a true crisis in the second, post-tequila phase, which started in late February 1995 and lasted through March. On February 27th, the international banks with branches in Argentina cut off credit lines to their branch operations, citing "country risk" as the rationale for such drastic action. This shocked branch bank managers and sent them scurrying unprepared into the domestic interbank market. The consequences were predictable. The interbank interest rates rose dramatically, from about 20% on peso deposits to over 50% within hours after the international credit lines were cut. Also, the massive entry of the international branch banks into the interbank market was interpreted as a vote of no confidence for Argentina's currency board-like system. Consequently, both external drains from the currency board-like system and internal drains from the commercial banks became great floods.

It is important to stress that the withdrawal of the credit lines from the international banks was the event that pushed Argentina from a liquidity squeeze into a liquidity crisis. Viewed from a historical perspective, this event is unusual, if not unique. With absolutely fixed exchange rate regimes, either currency boards or the classical gold standard (1880-1914), foreign banks have provided liquidity during times of liquidity squeezes (Hanke, Jonung, and Schuler, 1993). Indeed, private foreign banks have traditionally acted as lenders of last resort in absolutely fixed exchange rate systems. This was not the case in Argentina, because its currency board-like system was relatively immature and untested. Argentina had to pay a price for the sins of its monetary past.

Fuel was added to the crisis, when in the last days of February, rumors spread that the government might freeze deposits, as it had done with the "Bonex plan" of early 1990. Many people were also apprehensive that the decrees of February 27th were a prelude to broader powers that would reduce the rule-bound nature of the currency board-like system. There was also fear that a possible devaluation in Brazil would have effects in Argentina.

In March, the broad money supply fell by 4.4 billion pesos, or 8.5%; it reached its low point for the crisis at the end of the month. Peso deposits, dollar deposits, and currency outside banks all fell, reflecting a desire by many Argentines to hold their money outside the local financial system, either in dollar deposits abroad or dollar caches locally. Interest rates jumped; the peso prime rate increased from 22.7% at the end of February to a peak of 45% in March. Dollar interest rates also increased, with the prime rate peaking at 30%. The exchange rate risk exploded, with the peso-dollar interest rate spread moving from 930 basis points at the end of February to 1647 basis points at its peak. Since dollar interest rates were many times those prevailing in the United States, market participants evidently perceived that there was a high risk that borrowers of dollars would default or that the Argentine government would interfere with loans and deposits in dollars.

With the crisis worsening, the central bank took additional measures to further increase liquidity for banks. On March 10th, it temporarily allowed all banks to use up to 50% of their peso and dollar vault cash to fulfill reserve requirements, and allowed banks that had bought assets from institutions with liquidity problems to use the remaining 50% in the same way. The central bank also used its excess reserves to make short-term loans (discounts and repos) exceeding 900 million pesos to solvent banks with liquidity problems. As a result of its loans and the reduction in pesos outside banks, the central bank's liquid reserves fell to a low of 10.2 billion pesos at the end of March, compared with 15.8 billion pesos in late December. However, at no time during the crisis did the ratio of liquid reserves to the monetary base fall below the statutory minimum of 80%.

On March 14th, the government announced a package of measures, totaling 11.4 billion pesos, to contain the financial crisis. It accelerated privatizations. It reduced its planned government spending by 2 billion pesos, including cutting the wages of senior government employees by 5% to 15%. It increased several taxes, most importantly the value-added tax (a temporary increase of the rate, from 18% to 21%). And it announced plans to borrow up to US$7 billion--a three-year US$1 billion domestic "patriotic" bond issue from Argentines, a three-year US$1 billion bond issue from private foreign lenders and US$5 billion from the IMF, World Bank, and Inter-American Development Bank.

Argentina's success in putting together the (oversubscribed) $1 billion Patriotic Loan turned the corner. The loan was a convincing domestic affirmation of the creditworthiness of the Menem-Cavallo administration, and the IMF, et al, tagged along. This soon opened up the system with a fall in interest rates as dramatic as the February-March rise. The crisis was over.

From the beginning of April to the presidential election of May 14th, phase three, the crisis leveled off in some respects and eased in others. The broad money supply remained approximately constant: dollar deposits decreased, while peso deposits and pesos outside banks actually increased. Interest rates declined slightly; the peso prime rate eased from about 28.1% at the beginning of April to 24.5% in mid-May. Perceived exchange rate risk also declined, with the peso-dollar spread narrowing from 1320 basis points at the beginning of April to 710 basis points in mid-May. And on April 12th, the government announced the creation of a privately financed, voluntary Deposit Guarantee Fund for peso and dollar deposits.

Despite the crisis, the Convertibility Law continued to enjoy support across the political spectrum, including the three leading candidates and parties in the presidential election. Nonetheless, people were apprehensive about what might happen if President Menem were defeated or forced into a runoff election. However, President Menem's re-election in the first round, and the strong showing of his Justicialist Party in Congress, calmed the fears that a change in the government would undermine the Convertibility Law.

After the presidential election, all monetary indicators improved, marking the last phase the crisis, which ended by the beginning of August. Interest rates fell; the peso prime rate fell from 24.5% in mid-May to 14.4% by late July, while the dollar prime rate fell from about 19.4% to 12.2% in the same period. The broad money supply increased, though by the end of July it was still about 9% below the pre-crisis level. Pesos outside banks, peso deposits and dollar deposits all registered increases. Bank credit to households and businesses, which had fallen about 4% from the beginning of the crisis to the low point at the end of April, began to recover. Further, the central bank reduced its repos by about half between the end of March and the end of July, while its liquid reserves held against the monetary base rose above 90%. The perceived exchange rate risk also eased, with the peso-dollar spread narrowing from 710 basis points at the beginning of phase three to 500 basis points at the end of July.

At the end of July 1996, the Argentine monetary indicators were superior to their the "pre-tequila" magnitudes. The peso-dollar exchange rate remained absolutely fixed at 1 to 1. And inflation for 1995 was 1.6%, one of the lowest rates in the world and the lowest annual inflation in Argentina since 1944. The central bank's liquid reserves increased to reach $16.1 billion, slightly greater than the pre-tequila level of $15.8 billion. Peso denominated time deposits in July 1996 were $2.5 billion higher than in December 1994, and M3 exceeded its December 1994 level by $7.4 billion. And the peso and dollar prime interest rates were 10.5% and 9.1%, respectively. Both rates were lower than their pre-tequila levels. The perceived exchange rate risk was actually lower than the pre-crisis level, with the peso-dollar spread at 304 basis points, 176 basis points lower than the pre-crisis spread. Like tempered steel, Argentina's currency board-like system was toughened by the crisis, as the peso-dollar spread indicates. Although the financial indicators were superior to their pre-tequila levels (see Table 5), the real economy remained restrained. Real GDP declined by 4.4 percent in 1995, and didn t become positive again until the second quarter of 1996 (3 percent). Unemployment increased from 12.2 percent in October 1994 to a peak of 18.6 percent in May 1995. Although the unemployment rate fell to about 17 percent in July 1996, it remains stubbornly high.

Lessons from the Argentine experience

1) Currency boards, much like the classical gold standard, provide a constraint on monetary authorities. This is not to say that the monetary authorities do not attempt to maneuver around the constraints. Indeed, writing about the gold standard in 1932, Professor F. A. Hayek concluded that, "Every effort has been made to obviate [the gold standard's] functioning at any point at which there was dissatisfaction with the tendencies which were being revealed by it" (Hayek, 1984). Argentina's recent experience shows that the authorities also squirmed as much as they could within the confines of Argentina's currency board-like system. However, as much as the monetary authorities tried to maneuver and introduce discretion into Argentina's rule-bound, absolutely fixed-rate system, they failed to shake its irksome constraints.

2) Consequently, currency boards create price stability, even in times of severe crisis.

3) The hard budget constraints imposed by currency boards motivate deeper liberal economic reforms, as well as desired automatic adjustments in the economy. For example, before the Mexican crisis, Argentina's banking system was notoriously weak. Given that Argentina's Central Bank could not print pesos at will or act liberally as a lender of last resort, many analysts thought that the weak banking system would be Argentina's Achilles' heel. It was not and is not. Argentina's banking system is rapidly being strengthened and consolidated, with a few weak banks going to the wall and many more being bought by strong banks. By the end of 1996, 80% of all Argentinean deposits were held in the 25 largest banks. In addition, virtually all the weak banks owned by the provinces were on the block in a mass privatization (Fern ndez, 1996). The discipline imposed by the currency board has put in place a virtuous cycle. Indeed, Argentina's banking system is now rated as the strongest in Latin America.

4) And perhaps most importantly, the results produced by currency board systems can give politicians a platform from which they can win elections.

Argentina's Devaluation Risk Revisited

From the beginning, many observers have been skeptical that the exchange rate of one peso per dollar would be durable. Peso instruments have carried higher interest rates than dollar instruments with the same country risk; for example, Argentine government peso bonds have carried higher interest rates than the government's dollar bonds of similar maturity.

As late as November 1995, well after the Tequila effect crisis, Bank of America's "Global Financial Markets Forecast" was forecasting that in February 1996 the exchange rate would be 1.50 pesos per dollar. And in 1993, Deutche Bank conducted a risk assessment and concluded that the possibility of the peso being devalued within the next few years was 20-40 percent. In early 1995, financial markets agreed: spreads between peso and dollar financial instruments of similar risk and maturity widened from low single digits to as much as 16 percent at the height of the crisis. The country risk spread between Argentina and the United States was widening at the same time by about the same amount, reflecting fears that even dollar assets in Argentina were not safe because the government might repeat the Bonex plan and freeze all deposits, whether in dollars or pesos. The combined effect of country risk and exchange risk led to spreads of 30-40 percent over interest rates in the United States at the height of the crisis.

Why were so many people so wrong about the risk of devaluation and of default? A few market participants perceived no devaluation risk and were aggressive investors in Argentine bonds, but they were a small minority (Hanke 1994, 1995). For the rest, two factors seem to have been at work. One was a misunderstanding of how Argentina's currency board-like system functions. People thought that the peso had a pegged rate with the dollar rather than a truly fixed rate, so they feared that the Argentine peso would be as susceptible to devaluation as, say, the Mexican peso was. They did not appreciate that, thanks to the Convertibility Law, the Argentine Central Bank held sufficient dollar reserves to withstand a speculative attack. Moreover, with convertibility, the Argentine Central Bank was forced to allow the monetary base to automatically shrink in response to a drain of dollar reserves. This prevented the monetary base from getting far out of line with the demand for pesos. Under a pegged exchange rate, in contrast, central banks frequently keep expanding the monetary base until it is well out of line with demand, and a speculative attack and devaluation serve to bring the real value of the monetary base closer to demand.

It is interesting to note that overnight interest rates remained in mid-double digits even at the height of the crisis. Argentina did not experience the triple-digit interest rates that occur when a devaluation becomes almost a sure thing in the view of market participants. For example, overnight interest rates for the Swedish krona reached 500 percent (annualized) shortly before the krona was devalued against the German mark in 1992. Rates of hundreds of percent have also been common for other pegged exchange rate systems when they drew near to devaluation. The comparatively low rates in Argentina indicate that, although fear of devaluation was high, expectations of devaluation never became concentrated on a period of a single day or weekend.

Another misunderstanding was former Minister Cavallo's views about speculative attacks and devaluations. He understood that the ultimate weapon against an attack was the full dollarization of the Argentine economy, and he was prepared to use it. By eliminating the peso-dollar exchange rate through dollarization, the possibility of a devaluation would be eliminated. In consequence, even in the worse case scenario, a devaluation in Argentina was not a real possibility.

The other factor that led people to fear devaluation and default was Argentina's long history of botched monetary reforms. Under central banking, Argentina has had frequent devaluations and few periods of monetary stability, especially since the Second World War. The Menem government had to demonstrate that it was different from previous governments, which were fervently for sound money in their rhetoric but undermined it with their actions. Argentina's currency board-like system does not yet have the credibility of Hong Kong's currency board system, but its good performance in 1995 earned it new respect, as shown by the narrowing of country risk and devaluation risk. It has become a more mature system.

For a number of reasons, the probability of a devaluation of the peso is close to zero for the next several years. The Convertibility Law worked well through the financial crisis, and has support across the political spectrum; there is virtually no sentiment to change it. President Menem has said that the peso will never be devalued while he is in office (his term expires in 1999); given his record, his statement is highly credible. The financial system is now stronger than before because the weakest private banks have been merged into stronger ones and the provincial banks are being privatized. At the same time as the financial system is better able to withstand stress, it is less likely to encounter it. Foreign banks have learned that cutting off credit to their Argentine branches during a financial squeeze costs them profits. Other foreign lenders have also learned that a financial squeeze presents opportunities for interest-rate arbitrage with no risk of devaluation. People have a better understanding of how the system works and are better able to assess accurately the true devaluation risk. In consequence, interest-rate spikes in Argentina will become less frequent, less severe, and of shorter duration.

These conclusions about the peso's devaluation risk and the continuing maturity of the convertibility system have been confirmed by the markets. Specific confirmation comes from the Central Bank's successful establishment of a liquidity fund. Initially, the Central Bank set the fund size at $3 billion. However, it received bids for $6.3 billion from 13 international banks. The oversubscription represents a major vote of confidence and will insure that international credit lines are available to meet the type of liquidity crisis that was set off in February 1995, when international banks cut their credit lines to their Argentine branches.

Argentina's Unemployment Problem

Unemployment has been the great cloud over the economic progress Argentina has achieved since 1991. Unemployment averaged 5.5 percent from 1985-9, but has increased since then; it was 12.2 percent in October 1994 and peaked at 18.6 percent in May 1995, before falling to 16.4 percent in October 1995. At present, unemployment remains stubbornly high.

High unemployment in Argentina is a problem caused by laws that make labor markets inflexible and structural changes that have occurred in Argentina's economy. Convertibility and monetary policy have not been the causes of high unemployment. To appreciate this, consider that Argentina's economy grew 35 percent from 1991-4. In that period, almost 1 million new jobs were created in the formal sector of the economy, and employment in the informal economy also boomed. However, unemployment increased from 7.4 percent in 1990 to 11.5 percent in 1994. Convertibility, therefore, generated record levels of aggregate demand in Argentina and many new jobs. However, the demand side boom also increased the labor participation rate, from about 40 percent prior to convertibility to approximately 42.5 percent in May 1995. Even though new jobs were being created during the boom, the number of new jobs was not enough to absorb all the new labor market participants. In consequence, the unemployment rate continued to climb during the boom period. The economic slump of 1995 has simply added a business cycle component to the worsening long-term unemployment trend.

Until the current government came to power in 1989, economic policy had for many decades been based on "import substitution"--building domestic industry by protecting it from foreign competition with trade barriers. Argentina became a closed economy with many inefficient industries. Workers in protected industries benefited from greater job security and higher wages than they otherwise would have had. To gain the support of labor unions, a very important constituency, successive governments allowed the unions wide-ranging powers. One such power was the principle of "ultra-activity," meaning that if a contract between a union and employers lapsed before a new contract had been agreed to, the provisions of the old contract applied. In practice that meant that the gains unions made were cumulative, and that employers had no leverage to introduce more flexible policies. Labor laws made it difficult for employers to fire employees, required sector wide wage negotiations (convenios colectivos), and required high indemnification for employees. And social security taxes drove a wedge of almost 50 percent between before-tax and after-tax wages.

The economic reforms of the Menem government, including the extensive reduction in trade barriers and the Convertibility Law, introduced deep structural changes. The changes increased the ability of Argentine firms to import foreign capital equipment and to gain access to foreign financing. Labor laws remained largely unreformed because the government did not think the political timing was right, so labor markets remained rigid. From 1991-5, the cost of capital decreased 25 percent relative to the cost of labor. The federal and provincial governments and privatized former government enterprises, which had been overstaffed, released workers onto a labor market that was not flexible enough to create jobs for all. Additionally, the work force grew larger as more women and immigrants participated. The informal economy is fairly extensive in Argentina, and some officially unemployed people work in the underground economy.

With convertibility, the Central Bank imposed a hard budget constraint on the fiscal authorities. In consequence, the government could not resort to make-work, public works projects to soak up the excess labor released by restructuring the economy and increased labor force participation rates. This, in combination with inflexible labor markets, resulted in increased unemployment rates.

The government moved quite slowly on reforming labor laws until July 1994. In the aftermath of the Tequila crisis, the government seems more committed to labor reforms. Several acts have been passed that make the labor market more flexible. It is now easier for employers to hire workers for short periods and on a trial basis. Small firms are exempt from some of the rigid labor laws that apply to larger firms. Social security taxes have been reduced for inexperienced workers up to 25 years old and part-time workers. Employers are exempt from paying unemployment compensation to workers dismissed after less than three months. Disability compensation has been reformed to reduce costly litigation (Laws 24.013, 24.465, 24.467, 24.557). Even though the government is committed to the strategy outlined by the "Jobs Study" of the Organization for Economic Co-operation and Development (OECD, 1995), major labor reforms that would reduce the structural ("natural") rate of unemployment in Argentina remain bogged down, however.

Prices, Inflation and Argentina's Competitiveness:

Many analysts claim that the peso is overvalued. Although GDP per person in Argentina is roughly one-third the level of the United States, real estate and the prices of some services are as expensive in Buenos Aires as in New York. Travelers to Argentina notice that the taxi to their hotel is expensive, that the hotel is expensive, and that restaurants in downtown Buenos Aires are expensive, too. From their casual investigation, some foreigners conclude that the peso is overvalued and must eventually be devalued, so that taxi rides, hotels and restaurants will no longer be so expensive. Incidentally, foreigners often have their views confirmed by Argentine exporters who frequently assert that the peso is overvalued.

Under conditions of free trade and low transport costs, goods arbitrage insures the same price level and rate of inflation for traded goods in the currency board country and its anchor currency country. In consequence, wholesale price inflation should tend to be the same in both countries because wholesale price indexes are largely composed of traded goods, such as foodstuffs, minerals, and manufactures.

Unlike wholesale prices, consumer prices can diverge between the two countries on a sustained basis because consumer price indexes contain both traded and non-traded goods. Non-traded goods have local markets that are less closely linked to international markets because transport costs are higher compared to the cost of the goods. This makes arbitrage between markets for non-traded goods more difficult, so their prices can diverge more or less permanently.

This divergence between consumer prices in two countries is motivated when the rate of productivity growth in tradeables in one country is higher than non-tradeables in that country, and when it is also higher than the rate of productivity growth for tradeables in the other country. With a currency board and an absolutely fixed exchange rate between the two countries, goods arbitrage insures that the wholesale prices between the two countries will be the same. And the only way to keep relative prices in line with the relative productivity paths is to have higher inflation in the currency board country's non-tradeables and a higher rate of growth in the currency board country's consumer price index than in the anchor currency country's consumer price index.

Since Hong Kong reestablished its currency board in 1983, these consumer and wholesale price relationships have existed: Hong Kong's consumer prices have risen much more rapidly than wholesale prices and Hong Kong's consumer prices have risen more rapidly than those in the U.S., precisely because productivity in Hong Kong's tradeables has exceeded that for tradeables in the United States, Hong Kong's anchor currency country.

Has the Hong Kong dollar become overvalued, and have Hong Kong's exporters become uncompetitive? The answer to both these questions is a resounding "No" (Hawkins and Yiu, 1995). To understand this, one has to realize that to determine whether a country's currency is overvalued or undervalued a purchasing power parity calculation (or a closely related real effective exchange rate calculation) must be made. These calculations require the use of price indexes in the countries that are being compared. When focusing on competitiveness, it is the wholesale price indexes (tradeables), not the consumer prices indexes (tradeables and non-tradeables) that should be used (G.T. Management, 1995).

If the proper calculations are made with Argentina's and the United States' wholesale price indexes, the Argentine peso is not overvalued. Indeed, the real effective exchange rate has remained remarkably stable since the currency board-like system was installed in 1991. Argentina's strong export growth (exports increased 32.4 percent in 1995, a year of negative GDP growth, compared with 1994), confirms this conclusion.

Many analysts have come to the incorrect conclusion that the Argentine peso is overvalued because they use consumer price indexes when calculating either the peso's purchasing power parity or real effective exchange rate. Indeed, when the consumer prices indexes are used, the peso appears to be approximately 45 percent overvalued, and Argentine exports appear to be uncompetitive.

Some Thoughts on Indonesia

Both theory and evidence soundly support the use of currency boards in developing countries. So why is there such an intense debate about currency boards and fixed exchange rates? Politics, both domestic and international, has provided most of the fuel for the debate and controversy. Economics often has little to do with the choice of an exchange-rate regime. There is no better example of this than Indonesia.

By the first week of February, President Suharto knew that he would be finished if he failed to stabilize the rupiah at a reasonable level.

As Suharto saw it, the IMF, by its own admission, had botched the closing of sixteen banks in November. This aggravated Indonesia's economic troubles by setting off a financial panic and capital flight. In an attempt to stabilize the rupiah, Indonesia signed a second IMF agreement on January 15th. That agreement failed to address the rupiah's problems. In consequence, the markets promptly jumped all over the rupiah and put it into a free-fall. Loaded with external debt, Indonesia's private sector was bankrupt. Workers were losing their jobs. And if that wasn t bad enough, prices were rising as a result of the rupiah's devaluation and the Bank of Indonesia's (BI) November-January explosion of credit.

What antidote could counteract this deadly cocktail? As Suharto s Special Counselor, I proposed a comprehensive rupiah stabilization program. It's linchpin was a currency board system (CBS), an idea that was endorsed by Nobelists Milton Friedman, Gary Becker, Merton Miller and Margaret Thatcher's economic guru, Sir Alan Walters, among others. My program also included proposals for external debt restructuring, bank restructuring and recapitalization, privatization, a bankruptcy code overhaul and the break-up of crony capitalism.

The CBS proposal, however, created a firestorm of controversy. Most of the objections to my Indonesian stabilization program amounted to little more than ad hominem attacks on me (for a critique of these attacks, see Review Editorial, 1998) and do not merit comment. The so-called substantive critiques of my comprehensive proposal can be dismissed, too. Indeed, none of the critics had ever read the confidential documents which contained my proposals. This is clear from the fact that the critics made a great deal of general noise about currency boards, but never made substantive comments about my detailed CBS proposal and failed to mention one word about other essential parts of my stabilization program.

Some Thoughts on Russia

On his return from the Boar War, Winston Churchill remarked that Nothing in life is so exhilarating as to be shot at without success. That's how I felt in September of 1998, when Russia, with the blessings of the IMF, gave serious consideration to the CBS idea. This time around, the debate was more civilized than it had been in the case of Indonesia. But the level of substantive discourse was just as low. For example, The Economist of September 5th, 1998 contained an evaluation of a Russian CBS. This, and other commentary on that topic were meaningless without a CBS law to debate, however. In order to promote more substantive debate, a Russian CBS law, which I proposed in September is presented below.

A Currency Board System

To put the ruble on a sound competitive footing, the Russian government should produce a CBS law immediately and announce that it will be implemented as soon as possible. And to work in Russia, a CBS must be ultra orthodox. Indeed, if it is modeled after the currency board-like systems in Hong Kong, Argentina, Estonia, Lithuania, Bulgaria and Bosnia, it will probably fail.

To implement a CBS in Russia's truly unique situation, ultra orthodoxy must be embraced, so that the CBS commands the respect and confidence of the justifiably skeptical Russian people. The following orthodox CBS monetary constitution (law) is complete and suited for Russia's truly unique conditions:

The Russian CBS is hereby created. The purpose of the CBS is to issue notes and coins in CBS rubles, and to maintain them fully convertible at a fixed exchange rate into a reserve currency as specified in paragraph 6.

3. The CBS shall have its legal seat in Switzerland.

a) The CBS shall be governed by a board of five directors. Three directors shall be non-Russian citizens appointed by the Bank for International Settlements (BIS) in Basel. They shall not be employees of the International Monetary Fund or its member governments. Two directors shall be appointed by the government of Russia. least one of the directors chosen by the government of Russia.

b) A quorum shall consist of three member of the boards of directors, including at least one of the directors chosen by the government of Russia. Decisions shall be by majority vote, except as specified in paragraph 15.

c) The first two directors appointed by the Russian government shall serve terms of one and four years. The first three directors appointed by the BIS shall serve terms of two, three, and five years. Subsequent directors shall serve terms of five years. Directors may be reappointed once. Should a director resign or die, the BIS shall choose a successor to complete the remainder of the term if the former director was a foreigner, or the government should choose the successor if the former director was a Russian.

5. The CBS shall issue notes and coins denominated in CBS rubles. The notes and coins shall be fully convertible into the reserve currency. The notes shall be printed outside Russia. The CBS may accept deposits of the reserve currency.

6.

a) The reserve currency is the foreign currency or the commodity to which the CBS ruble has a fixed exchange rate. Initially, the reserve currency shall be the US dollar and the fixed exchange rate shall be one CBS ruble equal one dollar.

b) Failure to maintain the fixed exchange rate with the reserve currency shall make the CBS subject to legal action for breach of contract according to the laws of Switzerland. This provision does not apply to embezzled, mutilated, or counterfeited notes, coins, and deposits, or to changes of the reserve currency in accord with paragraph 13.

8. The CBS shall begin business with foreign reserves equal to at least 100 per cent of its notes and coins in circulation and deposits with it. It shall hold its foreign reserves in securities or other forms payable only in the reserve currency. These reserves shall be held on deposit at the BIS. The CBS shall not hold securities issued by the national or local governments of Russia, or by enterprises owned by those governments.

9. The CBS shall pay all net seignorage (profits) into a reserve fund until its unborrowed reserves equal 110 per cent of its notes and coins in circulation and deposits. It shall remit to the government of Russia all net seignorage beyond that necessary to maintain 110 per cent reserves. The distribution of net seignorage shall occur annually.

10. The head office of the CBS shall be in Moscow. The CBS may establish branches or appoint agents in other cities of Russia. The CBS shall also maintain a branch in Switzerland.

11. The CBS shall publish a financial statement, attested by the directors, monthly or more often. The statement shall appraise the CBS s holdings of securities at their market value.

12. The CBS may issue notes and coins in such denominations as it judges to be appropriate.

13. Should the annual change in the consumer price index in the reserve country fall outside the range 5 per cent to 20 per cent for more than two years, or 10 per cent to 40 per cent for more than six months, the CBS must, within sixty days, either:

a) devalue (if the change in the index is negative) or revalue (if the change in the index is positive) the CBS ruble in terms of the reserve currency by no more than the change in the index during the period just specified, or

b) choose a new reserve currency and fix the exchange rate of the CBS ruble to the new currency at rate then prevailing between the new reserve currency and the former reserve currency.

15. The CBS may not be dissolved nor may its assets be transferred to a successor organization except by unanimous vote of the board of directors.

16. Beyond an initial loan of reserves from the International Monetary Fund, the CBS may not accept loans or grants of reserves from international agencies or foreign governments.

17. Exchanges of currency by the CBS shall be exempt from taxation by the Russian governments.

18. CBS rubles and the reserve currency shall be legal tender for paying taxes and settling debts in Russia. However, they shall not be forced tender for contracts between private parties.

In addition to legalizing the use of foreign currencies and adopting a currency board law, the Russians must amend Russian contract law. With a currency board, the ruble-dollar exchange rate would be absolutely fixed. To facilitate relative price adjustments, all other prices in Russia should be allowed to fluctuate freely in relation to the fixed price of the ruble. The indexing of contracts in Russia should, therefore, be strictly prohibited.

The devil always resides in the details, particularly in Russia. Anything less than an ultra-orthodox CBS will not command the confidence of the Russian people and doom a Russian CBS.

Argentina. Laws (published in Boletin Oficial de la Rep£blica Argentina).

BCRA. Banco Central de la Rep£blica Argentina. Boletin Estadistico, various issues.

BCRA. Banco Central de la Rep£blica Argentina. Boletin Monetario y Financiero (English version: Bulletin of Monetary and Financial Affairs), various issues.

BCRA. Banco Central de la Republica Argentina. Indicadores Economicos, various issues.

BCRA. Banco Central de la Republica Argentina. Memoria Anual, various issues.

BCRA. Banco Central de la Republica Argentina. Communications [to financial institutions].

Culp, C.L., and Hanke S.H. (1993). The Hong Kong Linked Exchange Rate Mechanism: Monetary Lessons for Economic Development. Department of Economics, The Johns Hopkins University.

The Economist (1997). The Great Escape. 3 May.

The Economist (1998). Should Russia Try a Currency Board? 5 September.

Fern ndez, Roque B. (1996). "Briefing to the Finance Committee of the Chamber of Deputies", Bulletinof Monetary and Financial Affairs, Banco Central de la Republica Argentina, May-March

G.T. Management (Asia) Ltd. (1995). Emerging Market Trends, March.

Ghosh, A.R., Gulde, A.M., and Wolf, H. (1998). Currency Boards: The Ultimate Fix? Working Paper 98/8, Washington D.C.: International Monetary Fund.

Hanke, Steve H. (1994). "Arbitrage in Argentina." Forbes, 19 December.

Hanke, Steve H. (1995). "Why Argentina Is Solid." Forbes, 8 May.

Hanke, Steve H. (1998a). How to Establish Monetary Stability in Asia The Cato Journal, Vol 17.No. 3.

Hanke, Steve H. (1998). Is the Ruble Next? Forbes, 9 March.

Hanke, Steve H., Lars Jonung, and Kurt Schuler. (1993). Russian Currency and Finance: A Currency Board Approach to Reform. London: Routledge.

Hawkins, John and Matthew Yiu, (1995). "Real and Effective Exchange Rates," Money and Banking Hong Kong, Hong Kong: Hong Kong Monetary Authority.

Hayek, F. A. (1984). Money, Capital and Fluctuations: Early Essays. (Edited by Roy MacCloughry.) Chicago: The University of Chicago Press.

OECD. (1995). The OECD Jobs Study: Implementing the Strategy. Paris: Organization for Economic Co-operation and Development.

Review Editorial (1998). Monetary Mischief. Far Eastern Economic Review, 2 July.

Schuler, K. (1996). Should Developing Countries Have Central Banks? Research Monograph No. 52 London: Institute of Economic Affairs.