Transfers and the Price Level Under Fixed, Free, and Indexed Exchange Rates--A Monetary Approach

International Finance Discussion Papers Number 111

October 1977

TRANSFERS AND THE PRICE LEVEL UNDER FIXED, FREE, AND INDEXED EXCHANGE RATES -- A MONETARY APPROACH

by.

James A. Hanson

NOTE: International Finance Disqussion Papers are preliminary materials circulated to stimulate discussion and critical comment. References in publications to International Finance Discussion Papers (other than an -acknowledgment by a writer that he has had access to unpublished material) should be cleared with the author or authors.

SN ll tthe aR ET Ghana oe en gehen ate a cans Semen MTA emma tenn se nbmenm eo Sa Oe, on ee Ae

Transfers and the Price Level Under Fixed, Free, and Indexed Exchange Rates -- A Monetary Approach

by

James A, Hanson*

‘1. Introduction

This paper presents a comparative statics, monetary analysis of the classical transfer problem under three different exchange rate regimes--~ fixed, floating and indexed. Szccifically it compares, under the three different exchange arrangements, (1) the degree to. which an unrequited transfer is completed in a small, open economy, (2) the domestic price level and resource - allocation effects of the transfer, and (3) the form taken by the under effected portion of the transfer. The first two regimes serve as standards with which

to compare the increasingly popular indexing scheme,

: ' The economy receiving the transfer is assumed to "small" in international commodity markets--the prices of its exports and imports are quoted “in foreign currency and unaffected by local activities. ‘This means that the transfer cannot be effected through the mechanism of changes in the international terms of trade.t However, the economy includes a nontraded goods sector and changes in its relative price do form part of the transfer mechani’s..

in what is now well-known fashion.-

The transfer is assumed to be a fixed (net) amount of foreign currency. Our interest centers on the portion which is spent on local goods; we ignore

that. portion immediately spent on imports as already transferred and any saving:

*This paper was completed while the author was a Visiting Scholar in the International Finance Division of the Board of Governors of the Federal Reserve System. The paper represents the views.of the author and should not be interpreted as reflecting the views of the Federal Reserve System ot its staff.

b See Keynes for a statement of this transfer mechanism.

gee’ Viner and Samuelson.

ene meme ne

out of the transfer. For example, we would be concermed with immigrant remittances or the portion of program aid spent on lecally produced goods and its effects on the local economy, as opposed to skipments of goods

by emigrants or the import component of project aid. Since the first type of transfer is of great importance, it secms wise to investigate how it "works" under the different exchange rate artamgements. which

currently exist.

In comparing the local effects of the transfer under the three "pure" exchange regimes we are cognizant of the fact that most economies use some intermediate ‘system for regulating internatiomal payments. Dirty floats have been well known since 1971 and countries such as Argentina, Brazil, Chile, Colombia, and Peru which claim to pursue indexing schemes, in fact, deviate substantially from what we define

as indexing in this paper--an exchange rate which moves. proportionately ‘to the difference in the inflation rates of foreign and local goods.> Nonetheless, there scems to be some value in exploring the Limiting cases, since one might represent intermediate cases by some average

of the three, or some modification of the "pure" indeximg, scheme

described here.

The basic conclusion of the paper is that under indexed rates transfers cause the most disturbance to home goods prices ‘and to the price level; under flexible rates, the least. Fixed rates gemorate an inter-. mediate rise in-home goods prices and the price level. To see these results

first note that under free vates prices of tradeables fall, while home goods prices rise. By comparison under fixed rates local prices of tradeables remain |

constant. The larger price rise under fixed rates generates a wealth effect

I A

3See Donges, Hanson, Langoni and Kogut and Suplicy.

. vere

.

ate ng eh RR GENER LETTS TC INU Tm aes

a ha nee te oe 2 og

and a correspondingly larger nominal: demand for money.

This demand is assumed to be satisficd by the monetary authorities’ increased is suance of local money in exchange | for foreign exchange reserves. Indeed the results of

our flexible exchange case are based on tie assumption that _ the transfer causes no additional, private, iocal demand for foreign exchange and requires no central bank. intervention. Thus in the flexible exchanye rate case

the transfer is fully effected by increased purchases

of imports (motivated by the pure substitution effect of an appreciated exchange rate on import prices cembined with a rise in home goods' prices) anda decline im exports ‘(also motivated by the appreciated exchange rate and the rising home goods'! price). In the fixed rate case the transfer is not fully effected since there is some

general price rise which requires some new money and

acquisition of foreign exchange by the central bank.

Next, an indexed rate cause is shown to.cause a larger rise in home goods prices and the price level’ than fixed “rates. The reason is that indexing eliminates the substitution effect (the indexing rule being assumed to keep the exchange rate proportional to local prices). “Therefore, the demand for home goods which arises from -the transfer can be. satisfied only as all prices rise, reducing private demand for home and foreign goods, and increasing nominal money demand. Households add to their “nominal money balances to a greater degree than in either of the other two cases, because prices of _tradeables, as well as

home goods rise.

In fact, we show tht theoretical increase in money and reserves exceeds the original transfer. Since there is no substitution effect, domestic goods prices increase -Sufficiently to make room for the transfer motivated demand ‘via an equivalent increase in household money holdings and internation: reserves. However, indexing also means traded goods prices rise proportionately, creating additional, nominal demand for money and therefore a total reserve accumulation which exceeds the original transfer. - This result

may explain Colombian accumulation of reserves inthe

TET ane egemntee i © NS aN a RORY TERY NOL se Fe <eameNNONR RRR sae pe Pane epee mere = we ree map ne meer emer eg eter oe eames iat: aiidadatl "

- 4 -

early 70's" and the recent Brazilian accumulation, despite

i

“See Barro and Hanson.

a "large" current account deficits, which, however, were less than capital inflows.

This accumulation of foreign exchange represents a net loss to society, as shown below. Thus indexed exchange rates are not good arrangements for countries which receive large and fairly constant transfers, though, as argued in the conclusion, they may be useful in situations where transfers

are highly variable, particularly in inflationary economies. ‘2. ©The Comparative Statics Model

Consider a simplified model of an economy in which three equations can be used to represent demand-supplv equilibrium in the markets for home goods, money, and foreign exchange. Since only two équations need be considered explicitly because budget constraints make all three interdependent, we will ‘focus on the first two for ease of analysis. ‘Yo simplify the analysis we do not include a nonmonetary asset market and neglect the labor market as discussed below.

Demand for home gooas is the sum of (the price deflated value of) the transfer and of househola demand, which is assumed to be a function of relative price of imports vis-a-vis home goods, real income, and real wealth, as shown in equation (1).° Imports demand is a function of the same arguments aN ‘This function may be derived from a demand function

‘ homogeneous of degree one’ in nominal income, nominal wealth, and prices which in turn may be derived from intertemporal maximization of utility, subject to a budget constraint.

, Aggregate demand is assumed to be independent of the functional distribution of income, which is suppressed in the model. See

Lucas and Grossman, Hanson, and Lucas.”

i

as household demand for home goods. For simplicity we neglect

any effects of the’ transfer on money demand or on imports.

cc TT AAR en tat stisuaenenvencaas tn hacer ee en rasan een riagmeeneeminnenae eteneccanenisona demas anheseh teeceaiticaneeecettaaeettaninaie tneatmenemneiatiatiaie: -otaemecaheammeatennns meneame tas. - san ideal aad

ae d S, = , _ASs oS Re) oo (1) (eT/P ) + Co (e == x” (eP_(/P_) +c", PLP on 7? 4) + $+ t so = |. : . -~C (eP/? A) = 0

—

where C™ = household demand function for home goods e = exchange rate, local currency per unit "world" currency “PL = price of exports, fixed in “world” currency PQ = price of home goods ; . of . . ~ o> PY = price of imports, fixed in “worid" currency fe) rar , . M” = initicl money holdings = transfer, fixed in "world" currency © s . = supply function of home goods Ss . . pos ws xX” ='export supply function

and signs beneath variables indicate (assumed) signs of the partial derivative of the function with respect to that “variable.

. The arguments of this paper hold exactly in

a tradeable-nontradeable goods model. However, exports and imports are distinguished ina simplified way to permit comparisons with the usual models in which the terms of trace are important. Only imports’ relative price enters household demand directly because of the simplifying assumption that exports are ‘not purchased locally. Note that the partial @erivative wit! respect to relative prices includes wealth as well as substitution effects (but see below). Note also that the price Of tradeables could be substituted for import »rices throughout the analysis of this paper, .since international prices are

held fixed and local prices are determined by multiplying international prices by the exchange rate. The same rationale would permit the use of tradeables' production instead of only exports |

in-the definition of real income; here we simplify and assune-

no imports are produced locally.

The choice between supplying home ang export goods is assumed to be a function of relative pricz, as described ‘by a production possibility curve. Except fox a brief discussion of transfer problems uncer £lexible rates jn note 9 we assume factor prices are sufficiently Flexible to . guarantee full employment of the fixed supplies of factors

of production.

-Local money is assumed to be the only fexzm of wealtn and to be held oniy by househclds. Its use may be justified

by reference to intertemporal utility max simizacion cr to

facilitate transactions. Househoids are assumed te onptain their desixed levels of money balances within ‘he period of analysis Yo simplify the analy 31S we assume che demand for money depends oniy on nomin al income® and weaitn

Ia ®Purchases could be included in addition to or instead of

income without mach change in the resulvs. For simplicity

we have assumeé relative price changes are comsidered vermanent by myopic consumers and do not. affect money demand. See Kemp,

Borts and Hanson, and Rod riguez.

naan TOS nee

en aaa Nennnnnennnnnnnnneneannnaaa

and ignore any changes in transactions velocity or switches between local and foreign money, though the cualitative nature of the conclusion described below wouid be

unchanged under the usual assumptions about the directions of these processes. By assumotion any changes in nominal balances can be made only through tramsactions

in foreign exchange with the central bank. THnLS assumption is made to fochs the analysis on etfects generated by the foreign-sectcr, with out the com-

plication of domestically based monetary ereation. Thus

“we may represent equilibrium in the money market as

%

tenner eee rene enecreraneeeenecen eestor etn LOCA OEE COC COCOA en Oe Are mewn ee i ete Ri tg Ba egy metre Ai Memmi cement — ITT

d er. s S fo) (2) M” (> x + C°, M’/P_) - aM = 0 c + _

where ua = addition to money balances desired over the pexiod aM .

U

addition to money balances = e multiplied by

additional central bank holdings of foreign

excnange (dR). , or Equilibrium in the market for foreign exchange is

Simply written as:

“

> .S _ = Go (3) PLox” - P_v~ + 7 - dR= 06 x Vv d - ; where v” = import demand function

but this is implied by (1), (2) and the household: budget constraint: ,

7 d os ¢€_ isis s- (4) ue + PC” + eP_v = aP_x”™ + PC”. co Vv x c . 3. The Comparative Statics of a Transfer Unéer S

Alternative Exchance Regime

“To determine the effects of a transfer under the various exchange regimes we first differentiate the system totally, which yields the following matrix system:

oO r Ms r — 1M oe . | -a = - ct a o| | ap (. ©. gp 1 Po 3 p2 1 c| Ps c . | | de j; = = S . ° = S : eP_x - e) P_xX ~Mo "#2 yt MO yd x Ly aM 0 2 Pp | 1 p?2 ; 2 Pp 1 c | L ( Cc c !

where subscripts indicate partial derivatives of a

function,

RE NRA oe, et ROL RTR CNY atte NNEE NRCC RNR Y We Ht Sw UA RNY AAA Ai RCN YRS YD RE Ne oe a A ARN ERAT GS ener SR, mie. Sa ee

‘where a= Cc

and we use the production condition,

s ax" /aP Po

“a 4 ac’yap, OPx ‘to eliminate derivatives of production with respect to a price. , The effects of a transfer under fixed, flexible,

and indexed rates may now be compared by using side condi-

tions to eliminate various rows of the matrix. Flexible Rates: It is assumed there is no change in the holding of foreign currency. Therefore, AM = ~eAR = 0. Thus we eliminate

the third ‘column and require prices to adjust to both clear the commodity market and leave households satisfied with existing money balances. In this case the system determines

two prices’ and is quite stable.

The general equilibrium effects of a transfer are then

Bia ap far = (0/P/MIC) MO/P2) ~ (Mba M°/R2) /U (MOP, 2°/PQ))) > 0

do. ; dp /dT < ’ 2 (/aT < (e/P)/1(C Mm /Pa) + (a, e/P .)). since eP x” 1,0 ~ ut ae mt E> 9,7 1 ey 2. p2

7?his can be seen by noting that the m> gunction, (2), is derived from a general money function which includes

nominal income, nominal wealth, home goods prices and

era erent atecatns eestor ANN GSaONNERARAee of e-store mb seetshmesnenreereee + annemaneprer we:

7Footnote continued

-, import prices. Presumably the partial derivative of this function with respect to home goods prices is positive

. which implies the above condition. Essentially we are saying that when home goods prices rise, household try to spread their increased nominal income over more than

one period and/or refrain from consuming’ "expensive"

home goods now) by holding more money.

5.b de/at = —————=-—, 5 —______"—.. « 0

5.c dM/dT = 0.

Fixed Rates. Thé exchange rate is held fixed by the central bank's activity as a buyer or seller of last

resort, exchanging local money for foreign exchange;

thus de = 0 and the second column now becomes irrelevant. The system determines Po and aM. ; e/P . 6.a ap /dT = > 0 re e a me Cc 3 p2

6.b . de/dT = 0

e/P (ane Px°/Po) + oe m°/p2))

-. 6.0 | aM/daT = a > 0 " a e aM apo tc 1 Bo - 3 p2 ._ ¢c

Ce ee eee eee

.7 10 -

Indexed Rates: _It is assumed that the exchange rate is adjusted so as to maintain equal rates of inflation betwéen home. goods prices and import-export prices. Assuming no inflation in the rest of the world:

de= > ap. C

Such price setting of foreign exchange, as in the case

of fixed rates, requires.a buyer or seller of last resort: the central bank. Thus, once again, the money supply

is affected by international conditions and the third. column becomes relevant. The second column, Gescribing

. the effects of exchange rate changes is addea to

the first column, describing the effects of changes in

* home goods' prices.

; _ d oe) 2 J.a dp (/at = (e/P)/C. M /P 6 a ap

7.b de/a? = = . iS Cc 7.c aM/dt = (-e/P.) (Me) / (C2)

which implies

dM = edT (1 + eP. vic P_) > eat? ; . : Vi3°°3 ~-C since Mt + eB vi + pct = 0 Lo 2 Vvo30

by the household budget constraint.

A comparison of 5.a, 6.a and 7.a shows that flexible exchange rates raise home goods prices the least of the

three regimes and actually lower local import. {and export)

oe eS Son, oon : - . . : - : eer ham ar ncn agement eaten mee glee ARENAS MR ns FR ARI OR ARNE FE OR nS REE ye UR cm nee iar! he emg sae ema ne EE <-

‘

prices. Fixed exchange rates have an intermediate effect

and indexed rates raise home goods prices most. ‘The

reason for this result is that under indexed rates the

‘excess demand for home goods, generated by the’ transfer,

is met solely through a reduction in household demand

for all goods, generatec by a pure wealth effect.

The proportionate movement.of the exchange rate and

home goods prices Means that the real value of transfer,

in terms of home goods, is unaffected by the rise in

their price, and that consumers do not switch

to imports and producars do net expand home good -

-

production. In contrast,

ixed rates permit some

substitution in production and consumption as well as a

decline in the "reai" vaiue of the transfer, though

overall prices still rise and there is some wealth effect.°

®8Notice this result depends crucially on the assumption

that households increase their money holdings when domestic

goods' prices rise.

fy

Flexible rates provide the mest substitution-—export

-and import prices actually fall. There P .

is

no wealth effect and all adjustment takes place through

substitution in production and consumption and through

Changes in the real value of the transfer.

Comparing the effects of the transfer on the pattern -

of production, we see that, correspondingly, £lexible

rates cause the greatest change of production patterns,

with fixed rates generating a lesser shift toward home

goods production and indexed rates none whats@ever, as

there is no change in. relative prices. This is the

usual result and, of course, provides-a commom argument

Le cL NE aS TUR ACL Tm Oe tho OS Ot Ok ROE. at emma Renn ee A ee NI OE veg ue

ill iaik tetelnanesenmnentinate icmmemetmethanelineaineteatemennenedeeatioee nan nari naed

against flexible rates--that they generate

in production, with corre

costs.?

- 12 -

substantial shifts

espondingly higher total adjustment

°It is interesting to note that another commom argument

-against free rates--that thev reduce labor demand because

of currency appreciation in the face of Capital inflows--

becomes much less potent

induced by the capita

and employment in their production rise

any loss of employment in t that re

affect saving propensities

Under our assumption

demand for mcnev and,

be constant, with the switch

home goods to imports

government demand for nome coods,

transfer. If,

once we ailow for the

inflows.

therefore,

as a first avproximetiocn,

spending Demand for hame goods | tenciing to offset duction.

- Gea

ble goods prod

.

©,

ative price changes do not nominal aqgregate income,

household spending would

in household demand from

exactly matched by the rise in

financed by the

we assume that

Marginal value products of labor in each industry are

equal, then the unemployment generated by the zeduction in

the nominal value of export production will be Just matched

by increased employment

in home goods production.

For nonmarginal changes or fixed factor prices one also

would have to consicer the elasticity of the marginal

productivity schedule in the two industries.

goods were Savy,

initially would rise.

lack of aggregate demand,

more labor intensive,

If production of home goods

as described in Grossman,

Ef home’

then employment

were constrained by

Hanson

and Lucas, then the value of the marginal product of

labor would exceed its wage and, thsarefore, the value of

the marginal product in export. production, where there is

enn er ne eee RR NR GR eS enn ALI PP mat ak fe eee OE let eel ine manana artes enh oP chad

-13-

no shortage of buyers. In this case the capital inflow ‘again leaves nominal inccme constant but the switch in the pattern of production would permit-less use of labor, because of the higher value of the marginal product in . the home goods industry. This argumenc implies that _ the: degree to which the transfer increases unemployment depends on the ratio of marginal products in the two industries and suggests the problem mav have been

overstated.

-Against the higher adjustment costs imposed by the necessarily greater shifting of resources under free and fixed exchange rates, we must ccnsider the loss in purchasing power and utility which gcvernment use of the transfer imposes on households. ‘This loss, valued at original prices, can be approximated by the degree to which the transfer remains uneffected in the classical sense,. for that measures the degree to which households refrain from nsrmai spending in order to rebuiid

inflationary losses on real money balances.

By assumption there is no change in reserve holdings under flexible rates so the transfer must be fully effected. Thus households are able to substitute imports for the reduction in their consunmetion of home goods. Comparing fixed and indexed rates (6.¢ and 7.c) we see that reserve holdings increase by a greater amount under ‘indexed rates...In fact, under indexed rates the transfer is not only incomplete, its local effects actually stimulate an additional increase ine reserve holdings, over and above the original transfer (see 7.c).

Under indexed rates the process through which the transfer provides the governmont with increased real command over goods is exactly the same as monetary @xpansion-~house—

holds reduce demands for all goods in order to reconstruct

£1 ene anette A ERTS ANE ES RN TR ERE: MON em Re RN ERECT a RN cane mR cae a cmp ee came me meme range ne

~ 14 - -

the real value of money balances. Inased ja tne Colombian "experience with indexed ratcs, as described by Larro

and by Hanson, the fiscal sector of the government us ed foreign transfers to create additional moner at the central bank, rather than simply borrewing dixcectly.

‘The nominai value of the loss to households created py the transfer is not sclely their reduction in

consumption of home goods, matched by increased government purchases, but also their reduction in conswmustion of

imports.?}°? This secona reduction is equal to the gain

lOphis reduction in desired import consumption heips to explain why real Colombian imports grew slowly in 1972 and 1973 and in 1375, despite attempts at import liberalization. See Hanson.

re Et tenner = re

in reserves, over and above the trahsfer and represents a net loss (at base prices) .to the economy under

indexed rates.?! This loss would not cecur if the increased

tlwe have not attempted to derive explicitly Bailey's inflation tax burden because we abstract from infiation, the resulting variations in the velocity of money and correspondingly the resources necessary to economize on the use of money, which is pres sunanly reflected in

the. downward sloping demand curve for money. ee

a o

government purchases of home goods were financed by domestic monetary creation and exchange rates were free.

Under those circumstances it can be shown that relative

prices of home goods rise, rather than remain constant}?

te amen s

ne ann

~—

12This can be seen by using the matrix equation system

for flexible rates, after substituting dM for Te/P

in equation (1) and solving for GP ./oM ana de/dM.

and therefore, househvoias do not have to regince their total consumption to finance the rise in the local value of export as well as home goods production, as is the

cas@> under indexed rawes.

A similar type of analysis can ke used to determine the social loss imposed by fixed regimes. Comvarison of 6.c and 7.c shows that a greater percentage of ‘the transfer is completed uncer a fixed rate regiine than under an indexed rate regime. lowever, 6.c also shows that the transfer is not fully effected, as

reserves increase. It uiso is possibie that reserves

may increase by more than the transfer

13This may be seen by using the total derivative of the

household budget constraint, (4), to observe that the

‘

.aifference between the numerator and denominator of vi.c

(ignoring T) is equal to. the unsigned expression — a = Ss .

eP_v eP_x a _ doo ——. : — : P P + evr: ? Pe [ev sy y EV ep / c (ev M 7d

where exry = elasticity of x w.r.t y. If demand for imports is very price elastic, the transfer wiil be s :

partially completed.

aren ni cermin eS ake eminence amen RE SEEPS ce ee nee

- 16 -

4. Conclusion and An Extension

This paper nas deionstrated the coms arative static ‘Proposition that flexible exchange rates permit a transfer to be fully effected, through an adjustment of relative prices of home and traded gooas which ieaves the price level unchanged, but which generates a substantial reallocation of factors of production coward he hrcma

re wees

tn

ier

1

goods industry. In contrast, fixed exchange r and fully indexed rates raise the price level, forcing

part of the transfer to pe absorbed by the foxeign

currency desk of the central bank in exchenge for

increases in domestic money, rather than to be used to increase imports and decrease exports. These increased nominal holdings of domestic money come at ‘the

expense of housshold consumption. The reduction -ir

home: goods consumption makes room for transfer motiveced goods demand and thus both fixed and indexeé races require a smaller realilocacion of factors of production than flexible rates to satisfy tnat new demand; indeed,

indexed rates require no redilocation at all for there

is no change in relative prices. However, the reduction

in consumption of imports represents a real loss to society, since it is wnnatched by @ corresponding

rise in government consumption, as is the case with reduced consumer spending on home goods. Such a lo OSs may occur. under a fixed rate recine if the substitution effect of relative price changes is small, and will certainly

occur under an indexed regime where there is mo change in relative prices. The loss shows up as an increase in

- reserves over and above the transfer. To prevent such a

loss flexible exchange rates are in order.

The above analysis is one of the comparative statics of the price level. It might, therefore, be asked whether

the’ conclusions are robust if the transfer were (2) a

- 17 - ae one shot affaix, or (b) continued.

If the transfer were a one, shot affair, then once it is stopped the analysis of Section 3 would simply run in reverse, and the above recommendation for the use of flexible rates is much less correct. Under flexibie rates, the end of the transfer would return relative prices to tneir former values, and factors of production to their former uses, with associated adjustment costs. In contrast, undex other regimes the fali in demand for home goods would lower their prices and the pyri.ce level and househoid, finding themselves with excess money balances, would proceed to buy goods, eventually xunning down foreign exchange reserves to their pretransfer level, thereby completing, the transfer. Here monev could: serve

as a buffer’* ' spreading purchasing power

‘*See Darby and Borts and Hanson.

intertemporally and reducing acitustment costs, anc it seems

”

more appropriate to use fixed or even indexed rates

In contrast, flexible rates seem most appropriate if the same transfer is to be continued for some time periods. In this case the adjustment costs can be spread over many perjodc and there need be no reduction of . household consumption; it is almost &s if income has risen. Fixed rates create a one time loss: to households while they accumulate money balances in accordance wit the necessities of the higher price level that must be achieved to adjust relative prices. Once that new

set of relative prices and higher nominal money balances

TOE ame AON te RGAE A La Le CETTE OT, AE RS Es NE OEE SE RR SOU fC, SOS ARR AO IETINREREROT NR rsa |. me or + oo negara,

is obtained, then the transfer would be conpbieced in all ‘subsequent periods.?®

. . . -=- ._.s -— d ?20bserve from (3) that the derivative or Px - Pov .

holding M/? constant, is strictly negative.

since they involve no relative price changes and thus the extra flow demand for home goods can oily be accommodated throuch @ continous rise in prices, which forces household saving, «edditions to reserves and a corresponding loss of welfare in each perio. Ne benefit is gained from the transfer under imfoxed rates, the same government comand over purchasing

could be obtained at a lower cost to society through domestic credit creation and free rates. .Thus

our conclusion thet free rates are the most desirable exchange regime for effecting transfers which occur over

a number o* time periods.

Indexed rates are most undeasirab!

1

{

Bailey, M., "The Welfare Cost of Inflacionary Finance," - Journal of Poli Ltieal Reonony, Vol. 66, No. 2 (April,

1956), pp. 93-116. °

Barro, R., “El Dinero y la Bass Monetauriu cm Colombia,"

srollo,. Vol. V, No. 2,

Revista | Ge Phancacién y Desa:

(abril- junio, 1973), pp. 68-87.

Borts, G. and J. Hanson, "The Monetary Approach to

the Balence of Payments with Empirical Froplicat.on to the Case of Panama," IbPLS, Ministerio Ge Planificacién

~

de Panama, NBER, Conference on Planning and Short-Term

Macro Policy in Latin America to apyear in Annals of

Foonoiwic. and Social Measurement.

om

Darby, Michael, "The Ailocation of Transitory Income a . .

al Teeter eet ge Wo Raye Rosa Ves mn 1 -2 GS ASSEUS, ANCHLESN RMconm@mice Review,

Among Consumer

Vol. 63, No. 5, (December, 1973), pp. 871i-8ct.

Donges, J., Brazil's trotting Peg: A New Amproach ' to Exchange Rate Flexibility in Less Developed Countries, Washington, AEI, 1971. Grossman,. H., J. Hanson, anc R. Lucas, "Bifects of Demand .Disturbancés Under Alternative Exchange Rate Reqimes," Brown University Working Paper No. 76-12, revised. ‘Hanson, J., "The Colo:bian Lxperience with Imdexed Exchange

Rates and Capital Inflows 1967-1975," Brown University |

‘Working Paper No. 77-17.

af. Anata ee

Kemp, M., The Pure 'hedry of International. Trade ,

| -Englewood Cliffs, New Jersey, Prentice-Hall, 1964.

Keynes, J.M., "he German Transfer Problem," in H. Ellis and L. Metzler, eds., Readings in the hoorv of International Trade, Philadelphia, Blakiston Co., 1949. oS ,

Langoni, C. and E. Kogut,"Development, Policies and Problems: The Brazilian Experience} Journal of Monetary Economics, Vol. V. Supplement 1977, pp.’ 79-116.

o -~- 20 -

Lucas, R., “Monetary Union: An Altexnatiye to Free ~ Exchange Rates," Unpublished Ph.D. Dissertation, Brown University, 1975.-

Rodriguez, C. "Tha Yorms of Trade and the Balance of Payments - ¢ la jn the Short Run," Amoricar Sooronic Review, LNVI,

No. 4 (Septenber, 1976), pw. 710-716. .

Samuclson, Paul, "The Transfer Problem and Transport Costs," in R. Caves and Il. Johnson, edas., Readings in jahernational. fconombhcs, Homewood, Tllinois, Irwin, 106Gc.

Suplicy, E., "The Hffects of Mini-devaluations on DBranilian Exports," Instituto de Pesquisas EPcononicas, U. de Sado Paulic ana NBER Seminax on Indeuation, Feb. 26-26, 1975.

Viner, 7., Studies ju Internationa; Trade, Chapter 6.

2