hollis b. chenery and nicholas g. carter foreign...

WORLD BANK REPRINT SERIES: Number Nine

REP9May 1973

Hollis B. Chenery and Nicholas G. Carter

Foreign Assistance andDevelopment Performance,1960-1970

Reprinted from The American Economic Review,Volume LXIII, Number 2, May 1973

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Volume LXIIINumber 2

May 1973

FOREIGN ASSISTANCE AND DEVELOPMENT

PERFORMANCE, 1960-1970

By HOLLIS B. CHENERY AND NICHOLAS G. CARTER

Reprinted from

THE AMERICAN ECONOMIC REVIEW

( The American Economic Association

Foreign Assistance and DevelopmentPerformance, 1960-1970

By HOLLIS B. CHENERY AND NiCHOLAS G. CARTER*

The analytical and philosophical basis would have been taken. (See K. B. Griffinfor the aid and development programs of and J. L. Enos and T. E. Weisskopf.)the past decade was formulated in the The policies of the aid donors have beenearly 1960's. In outline form it asserts mixed. While aid as a share of GNP hasthat: declined continuously in the United States

since 1963, this decline has been somewhat(a)nextern eloresourc ries cansbe u ased offset by increases in the European contri-underdeveloped countiea a basis butions. Official Development Assistance

fort a. gnoct (ODA) from the DAC countries as a group, mnteand growth now approximates 0.35 percent of their

rates requires substantial changes in GNP, and total public and private flowsthe structure of production and trade amount to $18 billion or 0.8 percent of

,c external capital can perform a c l ' their GNP-considerably short of the(c) e externa cpt a .ca p ormia ti cal United Nations targets of 0.7 percent androle In both resource mobilization and 10pret epcieystructural transformation; and 1.0 percent, respectively.str ctu t ranfor m ,ion; and Despite this shortfall in aid, the overall(d)nthe nhese forucnraessonay aid dele performance of the developing countriesonce ths strucr changes arewel has generally lived up to the expectations

indlow may be productive). of the early 1960's. For the decade as awhole, the net amounts of external capital

This rather optimistic diagnosis of the pos- supplied were not far short of the moresibilitie , , .... ., ..self-sustaining conservative estimates of the amountssibilities for achieving neeedfo aceertedgrwt, ndmake

growth and of the potential value of well- needed for accelerated growth, and markettimed capital transfers has had widespread access for less developed country (LDC)effects on the plans of both aid donors and exports has improved. Growth rates havedeveloping countries. accelerated in most countries. Over the

More recently there has been a variety decade a number of aid recipients haveof criticisms of both the performance of gone through the anticipated sequence ofcountries receiving aid and of the basic increased investment rates, structuralideas on which aid programs have been transformation, and declining aid require-conceived. It is asserted that aid is often ments, On the other hand, several coun-

off set by increased consumption, that aid tries have confirmed the suspicions of thedonorsetb inerferse withunational pror ait aid skeptics and show little benefit fromdonors interfere with national priorities, teassac eevdand that aid permits countries to defer the assistance received.diffcul poic chage tha othewis In this paper we present a summary

evaluation of the interrelations between* Development Policy Staff, International Bank for internal and external policies and the role

Reconstruction and Development. We would like to of foreign assistance in the developmentacknowledge the help of Lance Taylor, M. Agarwal, and experience of the decade. The bHazel Elkington. Statistical assistance was provided by past pro-F. F. Jen and R. Bhakta. lem is complicated by the lack of any

459

460 AMERICAN ECONOMIC ASSOCIATION MAY 1973

tested empirical methodology for deter- adjust the trade and savings limits hasmining the sources of growth in developing produced disequilibrium conditions; andcountries. The factor productivity ap- (4) the extent to which the insufficientproach focuses entirely on internal factors external capital has been provided to sus-and does not readily accommodate trade tain minimum rates of growth; andbottlenecks and other disequilibrium con- (5) the effects of the allocation mecha-ditions common to developing countries. nism on the distribution of benefits.At the other extreme, approaches that For this purpose, we utilize two sets ofcenter on a comparison of trade perfor- estimates: the projections of feasiblemance tend to exaggerate the importance growth and aid requirements compiled byof aid and exports. Chenery and Strout for a sample of fifty

In order to include both internal and countries for the period 1962-1975, andexternal aspects of performance, we will estimates of the actual values of thebase our evaluation on the projections of parameters in this model for the periodgrowth and aid made by Chenery and 1960-1970, covering thirty-seven of theA. M. Strout for the period 1962-1970. We principal countries in this sample.' Thehave discussed elsewhere the improve- omission of thirteen countries (mainlyments that have been made in this model small) does not materially affect our find-(Chenery and Carter). In the present pa- ings.per we use it as a convenient frameworkfor measuring the past performance of the A. Overall Comparisonsdeveloping world against projections made The Projectionsat the beginning of the decade as a basisfo esalihn adrqient. While The Chenery-Strout projections weretodestayb tiesh give greatirermwei. Wht made to determine the needs of developingtoday countries give greater weight to cutisfretra aia ne aiuother objectives such as employment and countries for externalcaptal under variousincome distribution, we will examine the assumptions as to external trade and aidperformance of the 1960's in the terms of policies and internal resource mobilization.the overall growth objectives that were They were based on four analytical ele-commonly accepted at the beginning of ments:that decade. (a) past performance of each country,

particularly in the preceding five-yearI. Development Performance in the 1960, 's period (1957-1962);

While we cannot estimate the structural (b) development programs of all countriesrelations in the aid-development model for which they were available;with any accuracy because of the periodic (c) intercountry econometric studies ofexistence of disequilibrium in most coun- the principal parameters of the systemtries, there are several aspects of the (capital-output ratios, import ratios,Chenery-Strout analysis that can be savings parameters); andevaluated: (d) independent estimates of export pros-

(1) the extent to which growth has been pects for the principal commodities,accelerated, and the relative importance of which were used to modify individualinternal and external factors in this result; country forecasts.

(2) the extent to which each economyhas been able to absorb external resources Since the primary purpose was to deter-for productive uses; I A description of the sample and a full set of statisti-

(3) the extent to which the inability to cal comparisons is given in Chenery and Carter.

VOL. 63 NO. 2 EXPERIENCE OF DEVELOPMENT 461

mine aid requirements as a function of analysis of export growth that was thengrowth objectives and domestic perfor- taken as exogenously given for each coun-mance, alternative assumptions were made try. These are summarized in Table 2. Thefor each set of policy variables, reflecting a export forecast, which was made on asubjective judgment as to the likelihood of commodity basis for developing countriestheir achievement. We will use the central as a whole, accurately predicted the slowset of "plan" targets and "plan" perfor- increase in primary nonfuel exports ofmance as a basis for the present evalua- about 3 percent. The main difference is intion, since they were then considered to be the growth of manufactured exports andthe most probable outcome. 2 services, which have grown at 15 percent

compared to the anticipated rate of 6 per-GNP Growth Rates cent. Total exports for the sample group

Almost all countries that had not al- have therefore grown at 5.9 percent com-ready achieved growth rates of more than pared to the projected rate of 5.1 percent.5 percent in the 1950's planned for ac- The more rapid growth of exports hascelerated growth in the 1960's. Even after largely offset the slower growth of externalthe downward revisions by Chenery and capital. Annual requirements for externalStrout to make the plan targets more capital for the thirty-seven countries wererealistic, a significant acceleration in the predicted to double with a total net inflowrate of growth was projected for forty of between 1962 and 1970 of $66 billion. Asthe fifty countries; in thirty-five of these shown in Table 3, the actual inflow wascases some acceleration was achieved. The about 40 percent less. Although the totalunweighted average for the thirty-seven flow of imports was approximately whatcountries in our sample was raised from was estimated to be necessary to supporthistorical rates of 4.4 percent in 1957-62 realistic plans of these developing coun-to 5.25 percent in 1960-70. tries, the aid component was financed on

A comparison of actual to projected considerably harder terms than was antici-growth is given in Table 1. In 25 of the pated, thus biasing the distribution ofcases, actual growth was within + 1.2 per- growth in favor of countries that were able

cent of the plan rate. The other twelve to expand primary exports or borrow oncountries are fairly evenly divided, with hard terms.five growing significantly faster than pro- Since the overall supply of foreign ex-jected and seven significantly slower. The change, which constitutes the principalaverage planned growth for the whole exogenous element in these projections,sample was about the same as that which has been roughly as predicted, our analysistook place. can concentrate on the factors affecting its

distribution among countries. The higher

lixternal Aspects growth of mineral and nonprimary exportshas been of considerable benefit to six of

The projections of GNP growth and re- the countries in our sample, while short-quired capital inflow started from an falls in primary exports have hampered

'These estimates were adjusted from the original another six.plans of the countries according to the authors' judg- The distribution of external capital is ament to make them more "realistic"-i.e., with a prob- complex phenomenon which is examinedability of achievement of .5. Other projections based on in detail below. In general, successful de-historical performance and an "upper limit" estimate-defined by a probability of .2-were also made, giving a velopment has led to increased supplies oftotal of 18 projections for each country. external capital-usually on harder


TABLE 1-PARAMETERS OF DEVELOPMENT 1960-1970

Growth of GDPMarginal

Historical Plan Actual Growth of Growth of SavingsRate Rate, Rate Exports Imports Rate

Country (1957-62) (1962-70) (1960-70) (1960-70) (1960-70) (1960-70)

I. High Growth(Actual greater than 6%)

A. PlannedIsrael 9.0 9.0 7.9 15.4 12.5 .00Greece 6.0 6.5 7.3 8.6 6.7 .22Mexico 5.0 6.0 7.2 4.7 5.5 .22CostaRica 5.5 6.0 6.5 9.7 11.4 .15Jordan 5.6 5.6 6.4 8.3 7.5 .03Turkey 5.3 6.0 6.4 7.7 6.3 .23Malaysia 4.0 5.0 6.2 5.4 5.7 .22Tanzania 4.2 5.0 6.1 5.6 6.4 .22

B. Accelerated(Actual-Plan > 1.5%)

Taiwan 6.0 7.0 10.0 20.7 15.5 .35Korea 4.3 5.0 9.4 27.1 17.8 .29Iran 4.4 5.5 8.3 10.7 13.3 .30Thailand 5.0 6.0 8.0 7.9 9.9 .24Kenya 1.7 3.5 6.7 2.4 6.7 .15

II. Normal Growth(Actual 4.9 to 5.9%)

A. PlannedPhilippines 5.0 5.5 5.9 9.2 9.9 .20Venezuela 4.5 6.0 5.8 4.3 8.6 .31El Salvador 5.0 6.0 5.4 7.5 7.9 .06Brazil 5.5 5.5 5.3 6.2 3.8 .13Ecuador 4.2 5.0 5.1 2.0 4.8 .07Guatemala 4.0 5.0 5.1 10.4 9.2 .25Jamaica 4.0 4.5 5.1 5.4 6.7 .24Pakistan 4.5 5.3 5.1 2.7 5.2 .14Uganda 1.7 4.0 5.1 4.5 6.0 .22Ethiopia 4.5 4.5 5.0 5.6 6.5 .13Colombia 5.0 6.1 4.9 3.8 3.9 .21

III. Low Growth(Less than 4.8%)

A. PlannedPeru 5.5 5.5 4.5 2.2 1.6 .(0Argentina 3.1 4.3 4.0 2.6 0.3 .27Ceylon 4.2 5.0 3.9 -0.7 0.7 .22Morocco 2.8 4.0 3.9 2.7 2.8 .21Chile 3.5 5.0 3.9 6.2 2.5 .30Indonesia 1.0 3.0 3.0 5.0 10.1 .02

B. Retarded(Plan-Actual2 1.2%)

Egypt 4.5 5.5 4.2 0.2 1.9 .01Sudan 5.1 5.5 3.9 2.6 3.4 .00India 4.3 5.3 3.5 2.0 -0.4 .10Tunisia 4.1 5.0 3.5 5.5 3.4 .38Nigeria 4.0 4.5 3.0 7.6 3.6 .38Burma 3.2 4.0 2.7 -11.7 -7.6 .11Ghana 4.5 5.5 2.2 -1.3 -4.2 .09

TOTAL SAMPIX 5.4 5.4 5.1 5.8 0.18

a Median Projections of Chenery-Strout


TABLiE 2-PLANNED AND AcTUArL PARAMETERS OF GRowTH

Group I Group IIIHigh Growth Group II Low Growth

NormalA B Growth A B

Growth of GDP (%): Planned 6.2 5.4 4.2 4.2 5.2Actual 6.7 8.0 4.9 3.6 3.2

Growth of Exports (%): Planned 6.71 6.59 3.63 3.58 4.16Actual 7.34 12.63 5.38 3.36 1.88

Growth of PrimaryExports (%): 5.89 7.33 4.04 1.33 -0.51

Growth of ManufacturingExports (%): 16.24 25.89 18.39 10.60 6.21

Growth of Imports (%): Planned 6.49 7.23 4.74 4.61 5.46Actual 7.36 13.13 6.52 3.86 0.48

Capital Inflow 1962-70: Planned 12.4 4.5 16.6 12.9 19.8(S bilions) Actual 13.4 5.5 5.8 4.1 11.0

Marginal Savings Rate: Planned .150 .138 .184 .227 .108Actual .185 .272 .200 .177 .131

TABiy 3-DISTRmBUTION OF AID AND GROWTH'

Total GNP ($ billions) GNP Growth CIatl

Population 1970 (In Percent) ($ billions)(nillions) 1960 Proj. Act. Proj. Act. Proj. Act.

A. 16 Countries over$190 per capita 321 74 135 129 6.1 5.7 30 13

B. 20 Countries under$190 per capita 606 43 69 72 4.8 5.3 24 20

C. India 538 35 59 50 5.3 3.5 12 6

Totals 1,465 153 263 251 5.6 5.1 66 40

Totals may not add because of rounding.

terms-whereas unsuccessful development in both cases, but the effect of acceleratedhas usually led to a.reduction in the aid growth in lowering this value in the fast-supplied. Therefore, although the total growing countries-analysed in Chenerysupply of public funds for external assis- and P. Eckstein-was underestimated.tance can be taken as given, its distribu- Interpretation of the savings results istion depends both on donor policy and on complicated by the existence of disequi-the performance of the recipients. librium in the ex ante trade and savings

gaps. The Chenery-Strout projections as-Internal A spects sumed a median value of the potential mar-

The savings and investment perfor- ginal propensity to save (MPS) of .24, butmance of the sample group was somewhat because of the trade limits the projectionsmore favorable than the parameters pro- resulted in a median realized value of .15.jected. The mean value of the incremental The median actual MPS was .21, which iscapital-output ratio.(ICOR) was about 3.3 consistent with the somewhat less restric-


tive export earnings noted-above. which assumes that capital and labor areConsidering total savings and invest- fully utilized and gives no role to external

ment for the thirty-seven country sample, factors.we find both appreciably above the pre- In the cases where a "pure" savings ordicted totals. In the fast growing countries trade constraint can be identified, thethe higher than predicted growth rates analysis can be based entirely on the cor-have led to substantially larger amounts responding disequilibrium submodel and isof savings and investment, even though relatively straightforward. Our main diffi-marginal savings rates were not generally culty lies in interpreting the experience ofhigher than predicted. In countries of countries that are partially trade con-retarded growth, on the other hand, there strained. In this survey, we will try to indi-has been less of a shortfall in savings and cate the relative importance of internal andinvestment rates than in growth of GNP. external factors in the countries havingAs discussed below, poor savings per- large deviations from the original projec-formance does not seem to have been a tions: five cases of accelerated growth-major factor in the failure to meet plan Taiwan, Korea, Iran, Thailand, Kenya-objectives. and six cases of retarded growth-India,

We have also computed an approximate Colombia, Ghana, Tunisia, Ceylon, andrate of growth in the capital stock of each Chile.country. There are a number of countries(Iran, Korea; Tanzania, Malaysia, Paki- Cases of Accelerated Growthstan, Kenya) in which the rate of GNP When there are multiple constraints ongrowth is substantially higher than the growth and limited opportunities forrate of capital growth. These cases suggest medium-term substitution, an accuratethat fuller use is being made of the existing assessment of the sources of improved per-stock of capital to secure an acceleration of formenc can inpcieb determgrowth over a mited period. In contrast, yformance can si prlnctple be determonedwhr grwt ha enrpdfralne only from a solution to a plannmng modelwhere growth has been rapid for a longer with alternative sets of assumptions. How-period-as in Israel, Taiwan, Mexico, ever, when the deviations from the planGreece, and Thailand-the capital stock assumptions are concentrated in two orhas had to grow at about the same rate as three parameters, we can give an approxi-the GNP. mate evaluation of their importance by

B. The Constraints to Growth less formal methods. We would assess theirrelative importance in the five cases of ac-

The Chenery-Strout projections were celerated growth as shown in Table 4.derived from a simplified two-gap modelwhich exaggerates the likelihood of dis-equilibrium between internal and external TABLE 4-FivE CASES OF ACCELERATED GROWTH

constraints to growth, since normally one (In Percent)

or. the other constitutes the dominant External Factors Internal Factorslimit. Although we and others have tested Capital Excess

various methods of determining the rela- Countries Exports Inflow Savings Capacitytive importance of these constraints in Taiwan 50 - S0

actual cases, we have found none that is Thailand 50 So - -

entirely satisfactory. Despite these diffi- Korea 40 20 20 20

culties, we cannot fall back on the method- Iran 20 - 40 40

ology of general equilibrium analysis, Kenya - 40 - 60


The most significant difference is be- Chile. In Tunisia, both exports anid aidtween Taiwan and Thailand on the one were above plan levels. Tunisia has de-hand-where rapid growth was established liberately allocated a large share of invest-in the 1950's-and the other three, where ment to less immediately productive usesthere was substantial acceleration in the over much of the decade, which caused a1960's. Taiwan had a very large increase in slowdown in growth even though the planboth export growth and savings, permit- level of investment was maintained. Chileting both an acceleration of growth and a also showed little evidence of a trade limit,reduction in capital inflow, whereas Thai- due to favorable copper prices duringland required large additions of external most of the period. The capital-outputcapital. In Korea the substantial increase ratio rose as a result of excessive importin external capital made possible a fuller substitution, while the savings rate fellmobilization of the economy's resources. below the plan level as a rcsult of failure toThe existence of excess capacity is indi- control inflation.cated by a substantial fall in the capital- In Ceylon and Ghana, the retardationoutput ratio from its previous levels. of growth can be largely attributed to mar-

In Korea and Thailand it is impossible ket conditions for their major primary ex-to separate the effects of the added growth ports-tea and cocoa, respectively-whichof exports from the additional external re- account for over 50 percent of total exportssources, since both were substantial. In in each country. In Ghana, the problemKenya, the problem is simplified since was compounded by misallocation of in-neither savings nor exports were higher vestment and the consequent reduction ofthan projected; capital inflow and better the inflow of external capital. Failure tointernal management were the principal anticipate and adjust to the slow growthsources of improvements over the plan.' of their principal export commodity must

be considered the primary cause of re-Cases of Retarded Growth tardation in both countries.

Analysis of the causes of the serious In India and, to a lesser extent, Colom-shortfalls from planned performance is bia, the reduction in external assistancecomplicated by the extensive interaction played a major role in retarding growth.between external and internal factors. The In both cases the resulting shortage of im-two-gap model predicts that, when there ports was more serious than the shortageis a shortage of foreign exchange relative of finance for investment. The foreignto minimum import requirements, there trade bottleneck was made worse in bothwill be a fall in the savings rate and-un- countries by trade policies that discrimi-less increased external capital is forthcom- nate against exports of manufactureding-a reduction in growth and under- goods, which their degree of industrial de-utilization of capacity. These symptoms velopment would otherwise have sup-were present to a greater or lesser degree in ported.Ceylon, Ghana, India, and Colombia. In retrospect, the plan growth rates for

Internal factors provide the primary ex- these six countries (which are a close re-planation of slow growth in Tunisia and flection of their own plans) seem entirely

reasonable. Among the several elements' In countries where the economy had previously been causing the shortfalls, a reduction in the

growing slowly, the ICOR based on this experience over- expected capital inflow was a major ele-states the capital requirement. A fall in the ICOR re-flects both use of excess capacity and other aspects of ment in India and to a lesser extent ininternal management. Colombia, Chile, and Ghana. Internally,


failures of resource mobilization played a Taiwan, Turkey and Malaysia have re-smaller role than failures of allocation, duced their external capital requirementswith too much reliance put on import sub- from the plan levels through improved ex-stitution and little attention to export pro- port performance.motion or diversification. The less successful countries have a

more limited choice. In several cases theyII. The Allocation of Aid have tried to increase the inflow of capital

The allocation of external resources is to offset slow export growth with onlythe result of the set of policies followed by limited success. Of the "poor" performers,the recipients as well as the allocation cri- only Indonesia, Tunisia, Sudan, andteria of the donors. In the aid planning of Burma received amounts of external capi-the 1960's, the donors were primarily con- tal as great as-or even close to-the pro-cerned with (a) the efficiency of use of jected decade totals. While the reductioncapital, (b) the risk of loss, and (c) inter- was usually justified by poor internal per-country equity. In recent years a fourth formance of the recipient, this was not al-criterion, the effect on income distribution ways the case.within countries, has been increasinglystressed, although it has not yet had much B. The Distribution of Beneftseffect on the results. The dozen or so agencies that provide

The Chenery-Strout projections provide the bulk of official assistance operatc onthe only known "plan" for aid allocation different sets of criteria and with differingamong countries based on a consistent set geographical and political preferences. Al-of criteria. Comparing the actual growth though there are some common elements-projections and corresponding aid alloca- such as favoring countries that make effec-tions to the projections therefore reveals tive use of aid-it is impossible to con-something about the working of the mech- struct a general rationale for donorpolicyanism through which external resources over the decade. We can, however, deriveare allocated. some conclusions as to the distribution of

benefits by comparing the overall resultsA. Recipient Policy to those projected.

The choices facing recipients of external On the benefit side, the decade growthcapital vary considerably according to of GNP for the whole group was fromtheir past success in development. Coun- $153 billion to $251 billion, compared to atries that are seen to be effective users of projected total of $263 billion. Most of theexternal resources are favored by both shortfall is in India. On the external costprivate suppliers and aid donors. The most side, the shortfall in capital inflow (fromsuccessful countries have therefore had the $66 billion projected to $40 billion actual)choice of (a) accelerating growth further can be allocated in part (perhaps 30 per-on the basis of additional external capital; cent) to the substitution of manufactured(b) reducing their capital inflow, as en- and service exports for aid in countriesvisioned in the prototype of the two-gap such as Brazil, Taiwan, and Turkey. Mostsequence. of the remainder is the result of a shift

The high growth countries have re- away from the less successful countries assponded differently to these alternatives: measured by their growth performance-Greece, Korea, Iran, Mexico, Costa Rica, Argentina, Chile, India, Colombia, Egypt,and Thailand have chosen to increase both Nigeria, Ghana.GNP growth and capital inflow, while The relations of growth and aid to the


initial level of per capita income is shown version of the Chenery-Strout model. Thein Table 3, in which countries are grouped addition of the lost exports adds onlyby income level. The principal distribu- about 0.5 percent per annum to the 3.5tional effect is brought out more clearly by percent growth achieved, but the projectedtreating India separately. amount of aid would have brought the

The shift in distribution is striking. The growth rate up to 6.8 percent per annum.higher income group as a whole fell short In -other words, India might well haveof its target growth by only 7 percent even achieved the target rate of 5.3 percent ifthough the capital inflow was only 43 per- aid had been allocated on a basis compar-cent of that planned. Within this group, able to other countries.success varied largely with the ability toexpand exports in substitution for external III. Efficiency of the Aid Processcapital. The negative association between capi-

The second group of poor countries re- tal inflow and savings has led some authorsceived 80 percent of the inflows expected to question the efficiency of the aid process.and achieved growth rates 10 percent The proper test of the effectiveness of aid,above those planned. India, in contrast, however, is its effect on growth or otherreceived only 55 percent of her estimated social objectives rather than on savings asneed, and this shortfall was probably the conventionally measured.4 The two-gapmain single factor in her inability to grow model demonstrates that with a trade con-more rapidly. In this case there is a clear straint it is optimal policy to increase thefailure of the system of international dis- capital inflow even though the effect willtribution of assistance, which is heavily be to raise consumption as well as invest-biased toward smaller countries. ment, and in this case the productivity of

Since India constitutes almost half the external capital is very high.'population of the poorest group of coun- We have conducted several tests to trytries, its performance deserves some fur- to determine whether the observed nega-ther comment. The shortfall of aid to India tive effect of capital inflow on savings canin 1962-70 was on the order of $6 billion. be explained as largely a two-gap phenom-In addition, exports fell slightly short enon, as suggested by L. Landau. There(about $500 million) of the amount pro- are only a few countries-Thailand, Vene-jected. We can use the Chenery-Strout zuela, Jordan-in which a trade constraintframework to predict what might have can be ruled out; here the negative coeffi-happened had India received the aid and cients in a regression of savings on capitalthe exports that were anticipated, using inflow can be taken as indications of in-observed performance parameters. The efficiency in transforming the capital in-actual import ratio was reduced following flow into increased investment. In the vastthe 1967 devaluation and the imposition majority of cases, where there is a con-of strict import controls. While savings straint other than savings (or where thewere not as high as predicted, this reflects constraints are mixed over the period), thethe ex post fall that accompanies a domi- negative association is to be expected as anant trade gap. However, it is likely that result of ex post savings falling below exhad all the projected aid been forthcoming,India might have run into a capacity con- 4 G. F. Papanek has demonstrated that much of the

straint. apparent association is explainable on purely statistical

To evaluate these several eff ects, we ran grounds.6 This result is demonstrated in the optimizing versionsimulation experiments using an extended of the model used in Chenery and A. MacEwan.


ante as the system is constrained elsewhere. and A. MacEwan, "Optimal PatternsThe association between aid and savings in of Growth and Aid: The Case of Pakistan,"these cases is therefore not direct. If we in Adelman and Thorbecke, eds., The The-were to reduce F, the savings rate would ory and Design of Development, Baltimorerise, but oustput, investment, and consump- 1966.rise butd oftpu, Invesent, and countsm_ and A. M. Strout, "Foreign Assistancetion would fall. In general, the countries and Economic Development," Amer. Econ.that have raised their savings rates as a Rev., Sept. 1966, 56, 679-733.result of the aid-supported growth process K. B. Griffin and J. L. Enos, "Foreign Assis-greatly outweigh the cases in which an un- tance: Objectives and Consequences," Eco-necessary diversion to consumption can be nomic Development and Cultural Clzange,demonstrated. (See supporting statistical Apr. 1970, 18, 313-327.analysis in Chenery and Carter.) L. Landau, "Saving Functions in Latin Amer-

ica," in Chenery, ed., Studies in Develop-

REFERENCES ment Planning, Cambridge, Mass. 1971.G. F. Papanek, "The Effect of Aid and Other

H. B. Chenery and N. G. Carter, "Internal and Resource Transfers on Savings and GrowthExternal Aspects of Development Perfor- in Less Developed Countries," Econ. J.,mance," World Bank Disc. Paper No. 141, Sept. 1972, 82, 934-950.Dec. 1972. T. E. Weisskopf, "The Impact of Foreign

and P. Eckstein, "Development Alter- Capital Inflow on Domestic Savings innatives for Latin America," J. of Polit. Underdeveloped Countries," J. of Interna-Econ., Aug. 1970, 78, 966-1006. tional Econ., Feb. 1972, 2, 25-38.

INTERNA TIONAL BANK FORRECONSTR UCTION AND DEVELOPMENT

1818 H Street, N. W., Washington, D.C. 20433, U.S.A.Telephone number: EXecutive 3-6360

Cable address: INTBAFRAD

hollis b. chenery and nicholas g. carter foreign...

Documents