Milton Friedman: Float or Fix?

This article appeared in Wainwright Economics on September 2, 2008.

With the passing of Milton Friedmanon November 16, 2006,we lost one of the great champions offree markets. Obituaries and commentarieson his life's work and enormousinfluence have invariably mentionedhis advocacy of floating exchangerates, leaving the impression that healways favored floating rates. This wasnot the case.

Free-market versus managed regimes.For Friedman, there are threedistinct types of exchange-rate regimes:floating, fixed, and pegged—each withdifferent characteristics and differentresults (Table 1). Indeed, in his responseto the opening question posedin an eight-part debate on exchangerates with Robert Mundell, Friedmaninsisted that the dichotomy (floating or fixed) be replaced by a trichotomy(floating, fixed, or pegged).1 WhatFriedman meant by these terms differsfrom the meanings they are oftengiven, and to understand Friedman'sthinking, one must understand thedifferences.

In Friedman's sense, strictly fixedand floating rates are regimes in whichthe monetary authority is aiming foronly one target at a time. Althoughfloating and fixed rates appear dissimilar,they are members of the same freemarketfamily. Both operate withoutexchange controls and are free-marketmechanisms for balance-of-paymentadjustments.

With a floating rate, a centralbank sets a monetary policy but hasno exchange-rate policy—the exchangerate is on autopilot. In consequence,the monetary base is determined domesticallyby a central bank. With afixed rate, or what Friedman often referredto as a unified currency, thereare two possibilities: either a currencyboard sets the exchange rate, but hasno monetary policy—the money supplyis on autopilot—or a country is"dollarized" and uses a foreign currencyas its own. In consequence, undera fixed-rate regime, a country's monetarybase is determined by the balanceof payments, moving in a one-to-onecorrespondence with changes in itsforeign reserves.

With both of these free-market exchange-rate mechanisms, there cannotbe conflicts between monetary andexchange-rate policies and balance-ofpaymentscrises cannot rear their uglyheads. Floating- and fixed-rate regimes are inherently equilibrium systems inwhich market forces act to automaticallyrebalance financial flows andavert balance-of-payments crises.

Table 1
Friedman's Foreign-Exchange Trichotomy
Friedman's Foreign-Exchange Trichotomy

Click on chart for larger view

Pegged rates. Most economists use"fixed" and "pegged" as interchangeableor nearly interchangeable termsfor exchange rates. Friedman, however,saw them as "superficially similar butbasically very different exchange-ratearrangements."2 For him, pegged-ratesystems are those where the monetaryauthorities are aiming for more thanone target at a time. They often employexchange controls and are not free-marketmechanisms for international balance-of-payments adjustments. Peggedexchange rates are inherently disequilibriumsystems, lacking an automaticresponse mechanism to produce balance-of-payments adjustments. Peggedrates require a central bank to manageboth the exchange rate and monetarypolicy. With a pegged rate, the monetarybase contains both domestic andforeign components.

Unlike floating and fixed rates,pegged rates invariably result in conflictsbetween monetary and exchange ratepolicies. For example, when capitalinflows become "excessive" under apegged system, a central bank often attemptsto sterilize the ensuing increasein the foreign component of the monetarybase by selling bonds, reducingthe domestic component of the base.And when outflows become "excessive,"a central bank attempts to offsetthe decrease in the foreign componentof the base by buying bonds, increasingthe domestic component of the base.

Balance-of-payments crises eruptas a central bank begins to offsetmore and more of the reduction inthe foreign component of the monetarybase with domestically createdbase money. When this occurs, it isonly a matter of time before currencyspeculators spot the contradictionsbetween exchange-rate and monetarypolicies (as they did in the Asian financialcrisis of 1997-98) and forcea devaluation, the imposition of exchangecontrols, or both.

When Friedman first distinguishedamong fixed, pegged, andfloating rates, fluctuating exchangerates were rare, and in fact the InternationalMonetary Fund discouragedthem. By the 1990s, many countrieswere practicing what is often termedmanaged floating, in which the monetaryauthority does not promise tomaintain any particular level of the exchangerate, but intervenes from timeto time to influence the rate. Despitehaving a fluctuating rate, managedfloating falls under what Friedmantermed pegged exchange rates, becausethe monetary authority is aimingat more than one target at a time.Perhaps today it would be better to usethe term "intermediate" to describethe gamut of arrangements betweena Friedman-style fix and a Friedmanstylefloat. What Friedman meant bya floating rate is what is now usuallycalled a clean float, to distinguish itfrom a managed float.3

An advocate of both fixed andfloating rates. Contrary to what mostpeople think, Friedman was not simplyan advocate of floating exchangerates. His exchange rate trichotomymakes this clear. As a matter of principle,Friedman favored both floatingand fixed rates, and rejected peggedrates as "worse than either extreme."4

Friedman, however, laid greatstress on the fact that a fixed exchangerate administered by a central bankis dangerous. There is always the potentialfor a central bank to engagein discretionary monetary policy andto break the one-to-one link betweenchanges in foreign reserves and changesin the money supply.

For example, after Argentinapassed its Convertibility Law in April1991, Friedman insisted that Argentina'scentral bank was the Achilles' heelof convertibility. He didn't trust thecentral bank. Even though the systemworked perfectly well for years, Friedmanthought that the central bankwould eventually adopt a discretionarymonetary policy and convertibilitywould get into trouble. He was laterproved right: Argentina's central banksimultaneously attempted to maintaina rigid exchange rate and engaged inan active monetary policy. This culminatedin a balance-of-payments crisis,exchange restrictions, the end of convertibility,and a peso devaluation inJanuary 2002.5

Friedman's first and most famousforay into the exchange-rate debate wasas much an attack on exchange controlsand a case for free trade as anythingelse. He originally wrote "TheCase for Flexible Exchange Rates"6as a memorandum in 1950, whenhe served as a consultant to the administering the MarshallPlan. At the time European countrieswere imposing a plethora of controlson cross-border flows of trade andcapital. Friedman opposed these restrictions.He concluded that adoptingfloating exchange rates across Europewould remove the need for exchangecontrols and other distortionary policiesthat impeded economic freedom.

It is important to stress that economicfreedom was also a primary motivator for Friedman's advocacy ofunified currency regimes for developingcountries. He concluded:

While the use of a unified currencyis today out of fashion, it has manyadvantages for development, as itssuccessful use in the past, and evenat present, indicates. Indeed, I suspectthat the great bulk, although not all, ofthe success stories of development haveoccurred with such a monetary policy,or rather an absence of monetarypolicy. Perhaps the greatest advantageof a unified currency is that it is themost effective way to maximize thefreedom of individuals to engage inwhatever transactions they wish.7

Even though the title of Friedman'srenowned 1953 article has contributedto the misperception that hewas a dogmatic proponent of floatingrates, a close reading makes it clearthat he was not arguing so much infavor of floating exchange rates as infavor of full convertibility. He simplysaw floating exchange rates as thebest way to achieve full convertibilityquickly in Western Europe. The overriding'free-trade" motivation is madeclear when Friedman discusses thesterling area:

"In principle there is no objectionto a mixed system of fixed exchangerates within the sterling area andfreely flexible rates between sterlingand other countries, provided thatthe fixed rates within the sterlingarea can be maintained withouttrade restrictions."8

Another factor that led people topigeonhole Friedman as a dogmaticadvocate of floating rates was the factthat Harry Johnson and other economistsassociated with the Universityof Chicago were strong, and accordingto most observers, one-sided intheir advocacy of floating rates. Manyincorrectly concluded that Friedmanespoused the same views as some ofhis colleagues.

An advocate of fixed exchangerates for developing countries. Inthe 1960s, Friedman turned his attentiontoward monetary problems indeveloping countries, where inflationand exchange controls were pervasive.For many of these countries, Friedmanwas skeptical about floating exchangerates because he mistrusted their centralbanks and doubted their abilityto adopt a rule-based internal anchor(such as a money-supply growth rule).To rid developing countries of exchangecontrols, his free-market elixirwas the fixed exchange rate (an externalanchor).

The surest way to avoid using inflationas a deliberate method of taxation isto unify the country's currency [via afixed exchange rate] with the currencyof some other country or countries.In this case, the country would nothave any monetary policy of its own.It would, as it were, tie its monetarypolicy to the kite of the monetarypolicy of another country—preferably amore developed, larger, and relativelystable country.9

In many cases, he advocated fixedexchange rates rather than floating.For example, in response to a questionduring his Horowitz lecture of 1972 inIsrael, Friedman concluded:

The great advantage of a unifiedcurrency [fixed exchange rate] is that itlimits the possibility of governmentalintervention. The reason why I regarda floating rate as second best forsuch a country is because it leaves amuch larger scope for governmentalintervention ... I would say youshould have a unified currency as thebest solution, with a floating rate as asecond-best solution and a pegged rateas very much worse than either.

Friedman was clear and unwaveringin his prescription for developingcountries:

For most such countries, I believe thebest policy would be to eschew therevenue from money creation, to unifyits currency with the currency of a large,relatively stable developed country withwhich it has close economic relations,and to impose no barriers to themovement of money or prices, wages,or interest rates. Such a policy requiresnot having a central bank."10

Friedman clearly favored both floatingand fixed exchange-rate regimes inprinciple. However, as a matter of practice,for most developing countries hefavored fixed over floating rates.11 Yetmost economists and financial journalistsbelieve that he espoused floatingrates as the sole solution. Friedman'sreal position was that an exchange ratedriven by a free market was best, andthat both fixed and floating exchangerates had equal claims to be consideredmarket-determined.

The original versionof this report appeared in the Cato Journal, Volume 28(2), Spring/Summer 2008, pp. 275-285.

  1. "Nobel Money Duel," The National Post, December 11, 12, 13, 14, 15, 16 and 21, 2000.
  2. Milton Friedman, "As Good as Gold," National Review, June 11, 1990, pp 28-35.
  3. Milton Friedman, Dollars and Deficits. Englewood Cliffs, N.J.: Prentice-Hall, 1968, pp. 267-72.
  4. Milton Friedman, "Canada and Flexible Exchange Rates," in Revisiting the Case for Flexible Exchange Rates, Bank of Canada, 2000, p28.
  5. H. Hanke, "On Dollarization and Currency Boards: Error and Deception," The Capitalist Perspective, H. C. Wainwright & Co. Economics Inc.,December 22, 2005.
  6. Milton Friedman, "The Case for Flexible Exchange Rates," in Essays in Positive Economics, Chicago: University of Chicago Press, 1953, pp. 157-203.
  7. Milton Friedman, Money and Economic Development, New York: Praeger, 1973, p.47.
  8. "The Case for Flexible Exchange Rates," p. 193.
  9. Milton Friedman, "Monetary Policy in Developing Countries," in P. A. David and M. W. Reder (eds.) Nations and Households in Economic Growth, NewYork: Academic Press, 1974, pp. 265-78, at p.270.
  10. Money and Economic Development, p. 59.
  11. It should be noted that, for Friedman, "most" means most, not all. For example, in an interview published posthumously, Friedman responded withan unambiguous "yes" to the interviewer's first question: "Should China float the yuan?" "Milton Friedman @ Rest: Email from a Nobel Laureate,"Wall Street Journal, January 22, 2007, Tunku Varadarajan.

Steve H. Hanke

Steve H. Hanke is a Professor of Applied Economics at The Johns Hopkins University in Baltimore and a Senior Fellow at the Cato Institute.