+ All Categories
Home > Documents > teport No. EC-128 - World Bankdocuments.worldbank.org/curated/en/338101468138276720/pdf/mul… ·...

teport No. EC-128 - World Bankdocuments.worldbank.org/curated/en/338101468138276720/pdf/mul… ·...

Date post: 30-Apr-2020
Category:
Upload: others
View: 3 times
Download: 0 times
Share this document with a friend
39
teport No. EC-128 i report may not be published nor may it be quoted as representing the ve o the Bank and its affiliated organizations. They do not accept responsibility for its occuracy or completenegg 1NTERNATIONAL BANK FOR RECONSTRUCO110N AN- DVELOPMEN-1 INTERNATIONAL DVELOMEN ASSOCIA ATION THE EVALUATION OF AGRICULTURAL PROJECTS: A STUDY OF SOME ECONOMIC AND FINANCIAL ASPECTS May 7, 1964 Economic Department Prepared by: Herman C. van der Tak Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized
Transcript

teport No. EC-128

i report may not be published nor may it be quoted as representing the ve othe Bank and its affiliated organizations. They do not accept responsibility for itsoccuracy or completenegg

1NTERNATIONAL BANK FOR RECONSTRUCO110N AN- DVELOPMEN-1

INTERNATIONAL DVELOMEN ASSOCIA ATION

THE EVALUATION OF AGRICULTURAL PROJECTS:

A STUDY OF SOME ECONOMIC AND FINANCIAL ASPECTS

May 7, 1964

Economic Department

Prepared by: Herman C. van der Tak

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

PREFACE

1. The role of economic analysis in the evaluation cf an investmentpro.ject is to determine whether the project is worth doing in view othe possible alternative uses for the resources involved. To ao thisefiectively, it is necessary to consider econodc aspects at all stagesof project design and appraisal with due regara to the technical,administrative and financial considerations with which they are closelyassociated. Sensible conclusions depend on a careful evaluation of whatis known about all aspects of the project in relation to each other.What may appear to be a critical factor in one case may be of minorimportance in another. In some cases it may be the estimate of themarket, in others the productive effect of increased irrigation or theeffect on foreign exchange earnings, or the costs of construction.Each appraisal must be soundly basea on the realities of the particularcase. This is all very well in practice, but to present a systematictreatment of the role of economic analysis in the process it is necessaryto simnlifv realitir for the sake of clp-ity. Tt is necesnarv to abstractfrom the general in order to concentrate on the particular. This paperanproe-itIp t.he problemr i n this wayu It. deals oznlyi with ecnor -ndfinancial aspects as they pertain to the evaluation of agriculturalprojects.

2. Tn pra-tice the substan of ecn ic o appaisl' may differ-.* j.

4 C- 4J. I4'J1 - A fJU Jlt . . J4 . 'JI 04 JJF - 4.- f - -O A. .

considerably from case to case. In some cases, only a crude analysismay be possib1le lbeceuse oil lpcrk of information, tim opene l toaevote to the task. Neverthelecs, this may be satisfactory in the contextol a prirt-icular case where other factorJs mayCJ.Y dominat thU prIob LU-lm adLIU

alternatives are limited. In other cases, the importance of the issueand the possibility or securing the required ino aton may maeelaborate treatment both desirable and possible. Quantification may beextremely difficult in some cases - much less so in others. However,although the problems encountered in practice are varied, it is possibleto distinguish two elements common to all appraisal problems - those ciconcept ana method and those of' application. Among the conceptualproblems: What is to be measured? What are the appropriate units olmeaourement? What values should be used? How should data be organizedto indicate the relative merits of a project? Among problems ofapplication: How can the estimates be made? Is the data comparable'Mhat is the appropriate degree of refinement in a given case?

3. This paper deals primarily with the first or conceptual problem.It is concerned with a systematic presentation of methods for quanti.y-ing the appraisal o agricultural projects. The problems o applicationare recognized but no attempt has been made to deal with them in anysystematic way. It should be clear, of course, that the paper should notbe interpreted as implying that all the considerations set forth arebqually relevant or applicable to all cases.

- i.i -

4. The application of general econoic principles to specific usesin project cppraisal is a difficult process. General principlebi must betranulated into terms that are both relevant to the specil*ic problenm andcound in economic terms. rrogress has been mcde alcng these lines inrEcent :.ear - ::d current deve1onments re revflcted in t.his nar. How-ever, aithouga tne paper draws neavily on the existing state of the art,il, doen not nllo)w thnt. there is aeneral agreremnt. nff.,-cr econcmists onall the views expressed in it. The appraisal problem itsell tends tobehseen r different pn. view - d n in large part on the

level o- involvement of the viewer. This in turn is partly responsiblefor de f erent aroache widch -a be taken to such prb, -les s thc useo shadow prices, the handling oi rent and the role of project inducedsavings in apraisals. No doubt more could be said on these points.Indeed, a continuing discussion on problems ol method and, in particular,

prLoblem 01 JLLctdi, aog tiful, wLlorig LIn Islu LidLU W.1l U leuu

i economics is to play its full role in the practice ol project appraisal.I iLi hoped that this paper my make some contribution towards this end.

5. The author has benefitted greatly from comments oi Bank staff.Particularly valuable have been criticisms and suggest4ons o' members 0lthe stal o the Agricultural Division in the Department ci TechnicalOperations, and also, of the Ban]:'s Economic Advisors. However, theviews expressed in the study should not be taken to represent the viewoi the Bank.

6. This stafi study on agricultural projects is secohd in a seriesdcaling with methodological problems of appraisal as they are encounteredin dillerent economic sectors. The earlier one concerned power ("TheEconomic Comparisons o Hydroelectric ?rojects with Alternative Develop-ments of Thermal Electric Power" by Herman G. van der Tak, July 15, 1963).

K.A. BohrEconomic Department

TABLE OF CONTENTS

Page No.

PREF CE ........... ****.***** ***********. **.....**

I. INTRODUCTION ....................................

II. ECONOMIC EVALUATION ........................... 1

1. Investment Costs ............................. 2Lroject Design . .. .Items to be Included in Investment Costs ... 3Ihe 1Veal Value of, Investment Costs .........

2. nefits ....................................Increase in Output and Yields .............. 7rricng oi Agricultural Output .............

.urreU IL UU UO ................. .............. 11

14. The Rate o 1 eturn ........................... LCriteria and Neasurement 13Interpretations ol Return Calculations ..... 1;

. Conclusions ................................. 17Quantitative Significance of Suggested

AUjustments ............................. 18

III. OPERATING RESULTS AND FINANCIAL CONSEQUENCES 20

1. The Significance of Operating Results ........ 20

2. Benefits to the Farmer ..................... 21

3. Public Finance and Savings ................... 23

L. Pricing of Irrigation Water .................. 25

IV. SUM ARY ......................................... 28

APPENDIX

The Savings (Tax) Criterion for Project Selection:Some Furtlier Comments ............................... I

THE EVALUATION OF AGRICULTURAL PROJECTS: A STUDYOY SOME ECONi=l AIN FINMCLtSP ECT3

T TNTrODUTIOTTN

1. In recent years there has been a growing interest in methods andproblems of economic appraisal and considerable uork has been done on this-uhjet The nim of thi.q nnnti- is; to nresnnt In a systematic way soma ofthe current thinking as it relates to the appraisal of agricultural pro-jcts. It ineluie aeneral d-iscussin off ser enomic and financitlaspects of appraisal, as a basis for considering the economic logic ofappraisal methods. Although a paper of this kindin to serveas a guide to pro4ect appraisal, it can serve a useful purpose by focusingattention on some of the econOMi4c issueS inoled Iti4 moratt

the right questions are asked, and attempts at quantification, insofar aspossiblc, 4Ap-ose a aliscipline vhich may contribute totecotne dvlpmant of more effective appraisal practices.

2. Althou%h man- of the issues considered are comaon to a -.ide rangeol projects, the paper .*L directly concerned only with irrigai pjectsand 7jith schemes 1or land settlement and development. The first part dealswith the ecournmic evaluation of projects in terms of their real economiccosts and benefits to the country concerned, in other *Yords, it attempts toconsider such investment p-ojects in terms of the allocation of resourceswithin the country as a whole.

3. The second part of the paper discusses various financial aspectsincluding the operating results of the project and the results of theproject on the farm level, i.e., the adequacy of the benefits to the in-dividual faaer in terms of his incentives. 2he impact on public financeis also discussed, as !ell as the related problea of pricing irrigationwater.

Q. As iill be argued at length, the economic and financial analysisof a project deal uith different questions. Consequently, each type ofanalysis re-uires a different measurement of costs and benefits and eachshould be handled separately. Nevertheless, they are closely related: theincentive aspects of the benefits to the farmer are a koy factor in estime-ting increase in output that may be expected from the pfoject.

II. ECOFOI-LIC EVALUATIOT

5. Thn economic evaluation of projects serves to determine the desir-ability of the proposed investments frorl the point of vie'7 of elterinativ.opportunities for investment in the economy. A usual measure of desirabilitvis th- return on the investment. To compute returns from the point of vietwof the economy as a Yhole, it is necessary to evaluate both costs and benefitsin terms of their real volues, which may differ from the values relevant forvarious financisl aspects of a project. In part, this is a question of theitems to be included in costs and benefits- and in nart. of the valitinn of

- 2 -

these items. To the extent that market prices clearly do not reflect rela-tive scarcities for the country concerned, some adjustment is required. Theimportance of such adjustments and difficulty of determining what they shouldbe Till differ from case to case. However, in principle, so-calle6 shadowprices - i.e., prices giving a better measure of the value of inputs and out-puts from a national point of view - should be substituted for market pric-sfor the piirpoje of comparing the economic costs and benefits of a project.l/Part I discusses, from the economic point of view, the evaluation of invesE-nent costs, benefits and current costs, the measurement and interpretationof the rate of ieturn, and finally, some tentative conclusions that may bedrain from return calculations.

1. Investment Costs

Project Design

6. The estimation of investment costs is to a large extent the task ofengineers. The economist's role is to insis tha+. f.hAs technicpl

alternatives are considered in relation to their respective costs and benefits,and secondly, that the proper prices- are used for emStin+ng, theser- rcst (nde

benefits), and that all the right cost elements, and no other, are included.Thus, the analysis should, in principle, consider not only major alternatives,say, dam versus tubewell, but also possible variants after the major choicehas been made. Although technical necessities, tLme, and the resources avail-able for designing projects may severely limit the area of choice, there are

unlkl to beC no1- alternative poIDssib.LtiLes at a ll1. Lf L,2 V r tLeIi a1nlys t L

becomes involved in a project only at a late stage, consideration of suchAlt +ernatives- ceases to be relevant-. It* hUl be stese tha - conomic .a I~d U-L VU3 kltzcLouJ I UU U 1 LU -AIOU.LU U U k UIAA~ U UIU7

analysis must start at an early stage in the process of formulating and da-signing a project. If economic evaluation takes place only after the majordecisions have been made, its usefulness is much diminished.2/

1/ These costs and benefits to the country uill be referred to in this paperas real, or economic, costs and benefits, or occasionally as opportunitycosts, as distinct from market costs and benefits; and prices measuringthese real costs and benefits are referred to here as economic prices orshadov prices, as distinct from actually paid or roceived market prices.Although the concepts are different the values used may often coincide."Iarket prices" may often be "economic prices".

2/ The ihol2 process of project preparation from preliminary survey, formu-lation and design, to final appraisal is a lengthy process of su-cessiveapproximations. Institutional factors may limit the scope of analysisby outside lending agencies to the final stage of app:aisal. The commentsin the text concern the general need for economic anal7sis of projectsuithout regard to the particular stage at which the analyst is presentedtith the problem.

-3-

* ~ ~~~~ , nr. 1rs u )J ý4> .u. _J * -- -- ic A.. - r~- --7. The ronsider- "tíon of al ternative-s is pIarticu11lry important in th1

: o P-roj.ect3 relaýting to settlement schenes or, more gener,,lly, to a-i--nlioa! eelopnnt. uchi Schtemes do not constitlute Single rjeJ, u

ari t.d set nL' rojects. They envisae rur.cl reconstruction, uith dif-f. --nt ter n tcnure, irrigation, cattle breeding cent1r0, nEr .uulturl

hii tin-s, rod7, pcrcessing facilities, and sonetimcs even jchools nndhosTit-ls. 'ìuite rightly, the LIUhole complex of thse and sinil measuresis usually considered together, as many of them in isolation, Yithout thcAThfrfl, inui, be of little use. The appraisal 5hould erbain that aLJ..IrU2i-1 elements of the "pac'cage"I are provided for. Dryond this, hoi'ver,

e nppraja1 :ould consider .,hether the ingredients cf juch comeix cemescannot, be allered. Roads, for exajnple, may be essential for the produce tobe trnsported to the rrket, but ther# remains a considcrabie margin offreedom in dcciäing on the type and capacity of the road to be built. Simi-lnrPr, th, pnt.tern of the inter-related system of development ine,suras isnrt completel. rHgi-.: although man: of the elementc enunerated may be in-d:ne-nsnb ., 1/ _he empharis given to each of them can probably be variedn good deall. Thoe alternctive sub-projects and alte,riative patterns nedto be nonside,-'d if agricultural development pro,jects are not to be 3t'onllyrejeŽcte, nr tn, readily accepte1.

3. Therr are, of rourse, practical Iimitations to the numnber ()f altrn'-tiv nhnL c-Ii- h- c,nsidered. Preparation anc evaluntion of alternativeot, ,1.; tiir. T1he r-quired expertise is sclrce and oxpensive. It is al:aYsn,onf,ry to ba]!nce possible gains from considering alternatives or mckingfurther improvennts in the project 2rainjt the cot involve(d in doin i-,i,aLnd against, the lelays before the project can be startci. grceesi'e per-iectionism should be avoided. A ood project now may be preferablk to nbetter proect some years later.

Ttco tr Includd in Inves;tment Costs

5. Gion thn design of n proj,t, ther-, vre sor i.-i2rltYes encouIcrci.n decidin "hat lhoulj b2 included in investmtnt rosts,. nrt f:om nbvlouril.,ns such as lunr, equirnent, and materials, there ie ual an item 1colr"overhead-, and 3n;ineerin'", and onc for Icontingenci½s". The frmer. jorme-times consisting of a fiat surcharge on the total cost of _quipment, laborand maLerinls. and somotimes based on nrevious exnerierce. allovs for thecost of domestic and foreign consultants, and the time devoted to the projectby ivil snrrvAnt- in the rIpnnrtments conncrned Tho rontingn!is !_hich maybe up to 20 per cent of all other costs (or of some m.ajor Dtegories) areadded to nllnw for changos in pr-ces, 2/ miscalcultions nisions ni

I/ Snme, uc s-hools and hospitals, ay not be indsren1,l t hprojert, and should then be considererl on their ni,n nerits.

/ A grn-rnl chnnge in the price level, due to inflationary cir'ýumstnnces,ln'n not imply ,r- g-eater use of re'sources in the prnject and ir, of onuroc,irrelevant for the economic appraisal. Onlv the po;sibility that nnc2ificprIres-, turn thigLer thjan expectLed, OVer aJnU dcaUV an13 gnr 1ri n-crons, nhoffl Ve taken into account in the continguncy allowince.

unforeseen circumstances. Both items reflect, of course, normal Fngineveingpraticcs, and the uncertainty inherent in cost estimates.

10. Another issue concerns the treatment of purely financial items, suchas the possible loss of land tax caused by flooding of land in the reservoirarea of a proposed dam. or, conversely. the gain in revenue resulLina fromthe imposition of a betterment levey on land which is to benefit from newirrigation facilities. The subtraction or addition of the capitalized valueof such items in the determination of the investment cost of the project ismisleadinr, from the economic point of view. The loss of land revenue is nota real cost in the sense of a claim on resources; and the proceeds of ? betterment levy cannot be considered An nffst. to invest.mnt cost heaurn it rpnrp-sents simply a part of the economic benefit of the project which is taken bythe government through a snpifin tax measure.1/ Logically, they sholdr beleft out of account in the economic evaluption of the project, and be con-sidered only among thA financial consquens of the proj-t. On the otherhand, the loss of production in the flooded area (part of which is representedby the Ia xind E;4 ) ~~ i qA r-.Q n nn~A chrm IA )-%ez ,Ap ftaA 4 n +1- manalysis of the project. /

11. A somewhat different question arises concerning the inclusion of thevalue of the land to be used for a settlement scheme in the investn cts.The fact that land used for settlement may have to be bought from the presentorners for resale to the n se'ers - n. the atu-lI p.ic paid f it -is obviously important for the financial consequences of the project, and forthe share of the benefits accruing to th new settlers; but this may not berelevant for the evaluation of the return on the project for the economy as a

khole. The purchase price paid, if any, nay be a poor measure of the realvalue of the claim on real resources involved in the use of th3 land concernedand should not, therefore, be included among the investment costs for thepurpos-. of evaluating the economic return on the project. But this does notmean that the real cost of using land may be ignored in the economi- evalua-tion of settlement projects.3/

1/ In other words, betterment levies affect the distribution of the benefitsfrom the project betueen State and farmers (landoners).

2/ This may be treated as an addition to investment ccsts or, preferably,as an offset to the gross production benefits of the project. The lossof the land tax itself should be included in the financial analysis ofthe project as part of the impact on government finances.

3/ On this point, see the discussion of land rents below, paras. -)IO.Uhether the cost of land is treated as a current cost (rent) or q qn imv1e.t eno . (capitalized rent) is, of course, simply a question of con-venience and does not affect the basic issue.

The Rcal Value of Investment Costs

1?. Having decided what to include in investment costs, the various

items should be valued at appropriate prices. This raises much greatcr

difficulties. As is well-known, mnrket p-ices in developing coantries

frequently do not reflect the real cost of inputs used, and an evalua-

tion of investment costs in terms of market prices may therefore impart

a considerable bias to the economic return on a project.

13. Supplementary private investments necessary to complete the pro-

ject, such as diggint of field channels on farms so that. irrigation water

can reach individual plots are usually evaluated as costless in economic

terms, at least in less developed countries. Such investments are deemed

to cost nothing because they are to be executed by farmers themselves,

)ith their o-n enuipment if necessary, in time during ihich they vould

oi'herise have been "idle". In other .jords, it is assumed undeo these

circumstances that the real cost of farm labor is zero. However, this

line of argument is seldom extended to other cost items; the labor costs

of the bull, of the project are usually evaluated at thatever going uagerate has to be paid by the contractor. For institutional reasons, in de-

veloping cuntries, these -ages uill often exceed the valuo of unskilled

labor in alternative uses (the opportunity cost of labor), given iide-

qnread in(Hr) emp1oyment. in agriculture. The wageC paid for s.:illed

labor, on the other hand, im,ay sometimes understate its real costs. The

practical significance of this is not esv to determine on the basis of the

usual p:oject report, as it is not customary to sho, labor costs as a sepa-

rate item even less +n give a rnugh hr,1dontii het.ijen skilled and un-

skilled labor. Furthermore, even though the L'oregoirg line of argument is

accepted in prinCipl., there is room for on!:zirb disagreement as to

the proner levol of labor costs in particular countries.

1. 2ero marginal productivity of labor may be considered extr'Dne. The

extent of underciriploymient certainly varies a good deal from countryto o,,ln-

try. But even whcre thare is subst=ntial unemployment, it may not be Ju:.i-i.Ul0 to impueU ,er a. s Uae> U unsk%lled lCbor, as -heLe

may he real cost3 involved in putting people to work auay from home.

1. A mora g3neral argument is sometimes based on the assumption thatdionl,' con-ur,ton by JJ preiuSlV UI(d.)I t=AlUyUk1 rkers JJLAU to LJ

on the investment project is a "cost". Its relevance would appear to de-p3nd on particular institutional arrangements. If tue project is. i1ple-m-nted by voluitary labor on the local level, no additional consumptionensus; ani the same is true if soldiers are put to uork on the prDjest.Through devices such as a national labor service the incraase in consumptionmn-r aiso be k,pt. to a minimum. The imnact on .iditional consumptin tendto be greatest if uorkers have to be paid regular uages. But ev(-n if con-sumption is increased as a result of additional employment on the project,the Qxtent to ihich this should be counted a "cost" mayr he questioned.Consiimers' goods industries may be iorking at under-capacity, and free La-pnrt: of surplus food miy be available. Furthriiore fiscal ieasures may heu>ed to counterict. any undesirable increase in consurption resulting frorthe use of iuo,-ployed labor on investment projects. In general, tnerefore,it, wuld anpc)r Uat not much weight can be attached to this argumrent.

16. Similar arguments apply to the foreign exchange costs of the pro-ject. It is usual to.give the foreign exchange component of the invest-ment costs, 1/ but its counterpart in local currency is arrived at by con-version at the official (market) rate of exchange. As is well knoin, manycurrencies arc overvalued, and are only maintained at their official valueby means of stringent import restrictions, currency regulations, high tariffs,etc. Thus the foreign exchange costs of the project are often undervaluedin terms of local currency, and the total investment costs underestimated.

17. Labor and foreign exchange costs, together with interest costs (,hichwill br dis cn d Iter) r the major examples of distortions ihich may re-sult from the use of market prices in the evaluation of investment costs.The efPfc of diqt.rrtP c of mteial nf and niiinipent nrndued loc.-illyare probably less important, on the whole, and will often have to be ignoredin practice. Rough correction for the major distortions mentioned (one couldherdly expect an-y great precision) would provide some impression of the im-pFort-once of t1hCse distortions on 4-ota-l invesme_+nt co S. This reqtui-cs,

houever, the availabilitv of data on the relevant categories of investmentcosts. IT shkouldl be no+ed nevertheless th-a+ the correctio-s called for, int ' UJ. )V I ~ tJ u` J IJ. u l ~ u tl&t~ V V Va.C I Jfh.)1 L LL tt±1

most cases, are not all in the same direction - cost of unskilled labor -ouldgo down, and that of other i tems would go up - so that t1hey ,ould cancel out

to some extent. 1ore important, perhaps, than the impact of these correcticnson total cost,s Toul b11I,e t-he change in relative costs of Jaiusiputs,-4 hchUUL AJL L U _L~JUU UU V1 U14t .LCULUU A.~ u -L VL vCL1 L%U V .)LLZ) LlJJUL->,~ '.JLIL Ll~

might lead to a reconsideration of the type of installation to be built andthe method of its construction. Louer costs of unskilled labor and highercosts of imported machinery would put a premium on the use of more labor-intensive methods, rather than labor-Saving machinery. The practical possi-bilities of alternative solutions can, of course, only be judged in specificcases.2/

2. Benefits

18. what are the effects of the proposed investment, and what are they.. o"th? For many agricultural projects the obvious and straightfor,ard Pnsweris an increase in (the value of) agricultural output. Even if the directproduct is water, as with irrigation projects, the value of the water is corm-monly determined in terms of its contribution to agricultural output, becausethe price, if any, in the form of water rates, etc., depends on an essential-ly arbitrary adiii.nistrative decision, and does not refloct The real economicvalue of the iater made available. As stated, the latter value is obviouslyrelated to the impact of the better supply of water on agricultural output.'/

1/ This is sometimes an accounting charge only, insofar as it is based on thjdepreciation of previously imported machinery, and not on actual importsnecessitated by the project.

2/ See also "quantitative significance of suggested adjustments" paras. 5'-5r.

/ This is a simple example of the use of "composite sets", i.e., considera-tion of separate but related activities together, so as to avoid havingto estimate the value of intermediate products. Similar problems concern-ing the real value of the output of a project arise with certain tynes ofpover and transport projects, but %ithout the relatively simple solutionanplicable 4- irrigation projects.

11. The benefits of improved drainage are similarly assessed in termsof the higher agricultural output resulting from the project. But suchdrainago schemes also have other benefits. The same is true of othermeasures to prevent the deterioration of irrigation schemes, and of floodcontrol measures as well. They may not only raise output, but prevent afuture decline, or intermittent losses, in output. Jithout such measuresoutput from already irrigated land may gradually decline because of 1ater-logging or salinization, or damage may be caused to crops by floods. Suchbenefits m-.y be (and sometimes are) dealt i.ith in terms of preventing apotential decline in output, rather than in terms of an increase in output,but do not differ otherwise from "ordinary" benefits.

20. In general, the principal problem in evaluating benefits concernsthe estimation, first, of the increase in physical output expected from thaproject, and second, of the agricultural prices to be applied to this out-put. each of these will be considered in turn.

Increase in Output and Yields

21. nio major physical factors determining the potential increase inoutput resultine from an irritational project are the crea which becomesirrigable and the amount of water w_hich, on the average, can be made aval-able. These arA tiohnialv HAterminAd and an he rPn.qnahly iAl PqtAmnt.P.A different set of questions is rhether, in fact, all the available -:aterisill be used, whether it till be used over the full possible area, and if so,whether this till take place immediately after the water becomes available,or only after a certain las of time On past experience, it usually takeso) nrlir -n'a-n " ni f n )a good deal of time, 10 yezars or even more in many cases, before full use ismade of ne, facilities. This is partly a-us-e--- o d e n i -h neces-

sary supplementary works on the farm level, and partly - and usually more4mportant'll, - becaus conservats as to be overcome, and fanaers have tcget useed to new ways of farming. Similar difficulties and time lags ar; en-countered with many other types of agricultural projects.

22. As will be discussed later, this transitional period is of consider-able importance for the return on the project. Therefora, estimates are com-monly made, as best one can, of the gradual increase in benefits (yieldcurves) and used in the return calculations. H1orever, details of develop-ments consideved most likely during the transitional period - apart from itslength - are usually mitted from appraisal reports.

23. The gross benefits expected during and after the interim period arecalculated by estimating the increases in output of different crops, or otherproducts, and by valuing these at a certain price. The basis for the choiceor a certain cropping pattern (or, more generally, composition of outprt) isnot always indicated. It reflects the analyst's best judgnnt as to iwyhat isfeasible, in the light of demand, changes in cultivating practice, pric.-s,etc., that may be expected. It is sometimes selected so as to be fairly closeto the pre-project pattern. This may not give the best technically possibleresults, but it tends to facilitate the transition from non-irrigated farming.and may be the best feasible pattern, especially if the time elemont is tacuninto account. Undiscounted benefits may be smaller, but discount2d >enre'fitslarger, if a cropping pattern is chosen which provokes least resistance tochange by farmers.

- 8 -

2L. Further exploration of the issues involved in arriving at an eco-nomically optimum croppine pattern for the purDose of proiet annraisal.would appear useful. If technical, econoical, or social circumstanceslimit the range of alternativemin practice. this noint mirht we11 hestressed.l/ The choice of the best cropping pattern is likely to dependheavily on the estiinated nricAs of out.ut,. a;nd the. epected yield. P-icesactually received by the farmer are, of course, an important element amongthe factors deterining which crrpq in f-ct, will be produced./ On theother hand, as will be argued below, market prices may riot be the most rele-vant prices for issessing the onmic 1beit of J p-oject t hras a whole.

25. The yields expected after a certain interval from, say, irrigationprojects aeusuall- re'erred to as "cons ervative " and quite rightly so in_relation to the yields technically possible. Nevertheless they are, in manycases, much higher (sometimes triple or more) than yields prevailing beforethe project. These higher yields are expected to result from not just irri-gation, but also such additional input5 a a greater application of ferti-lizer, harder work and better practices resulting from a greater effort bythe extension services. In a given project area all these changes frequentlyform an inseparable !hole: more fertilizer, or better practices, may not bevery productive without irrigation, and vice versa.

20. On the other hand there may be other areas - perhaps already irri-

.ated - 'here similar efforts to introduce a more intensive use of fertilizers,improved seed, and pesticides, and better practices may by themselves yieldsubstantial benefits - without any large new investment in irrigation. It mayalso be possible to upgrade the cropping pattern in such areas to grow morevaluable crops than at present. These possibilities must be considered ifresources are to be used efficiently.

27. This is another example of the need to consider altarnatives. It isdifficult to handle this kind of problem while looking at individual p:.3jects.It requires essentially a judgment as to the existing possibilities of raisingagricultural output without large investments in irrigation or land reclama-tion, i.e., consideration of the agricultural program as a whole rather Lhanseparate projects. In analyzing irrigation projects account can be taken ofthis, Lrosso modo, by con3idering what is the economic value of the supple-mentary inputs. This is normally done at least to some extent. Estimates of

1V To some extent, the design of the proiect., with its bearing on the quan-tity and seasonal pattern of water made available, will in itself limitthe ranre of alternative cropping patterns.

2/ On the question of incentives, see "Benefits to the Farmer" and "Pricingof Irrigation I.ater", para. 64, et seq.

- 9 -

g-ross benefits are coromted by deting the increase in cash ePenditures

of the farmer for such items as fertilizer, and occasionally an allowance ismade for overhead costs of the government on ex-si -Aevces. But it"I- I~-. r.r..-, n,. -C ta n an5 .. nbm aam n rao ...n -,^ ~r oneQW4

seems likely that, in many cases, the real value of these inputs is heavilyunderstated: Uhe market price %or contro.l_led price)l o.f 1erti.LL4.Z.lizer1, =my notreflect its real cost as derived from its import price (at the economic rate01 fori.gnl excnange). Similarly, skilled extension workers tend to be veryscarce and the salaries they receive may grossly understate their value interms of potential increases in output. It has been widely argued that,perhaps in most cases, shortage of skills may be a greater obstacle bo higheragricultural output than lack of physical capital. If so, this should somehoube reflected in the prices used for evaluating additional inputs of suchscarce factors for purposes of project appraisal, or in the interpretation ofthe calculated return on the investments.

28. At any rate, it seems clear that the expected increase in yields isprobably the factor of greatest uncertainty in the appraisal of many agri-cultural projects. If things turn out well, i.e., if farmers do use theirrigation water, adapt themselves to the changes in circumstances, and adoptthe improvements in farm techniques envisaged, the increase in yields may bespectacular. But a less favorable response is far from excluded for a variet-of economic and social reasons. This large margin of uncertainty should some-hoi, be taken into account in comparing the returns on irrigation projects withother investments.

Pricing of Agricultural Output

29. The expected increase in output (or potential loss of output), finall,has to be valued. Local market orfann gate prices are normally used for thipurpose, except for export crops for which export prices are considered.Local market prices suffer from distortions due to a more or less elaborateset of restraints, subsidies. etc. They frequently differ widely hetweenvarious areas of the same country, not simply due to transportation :osts,but to the malfunctioning of the distrihufion sirtem. From a natio-al pointof view, investment decisions based on such prices tend to be misleading.

30. It would seem preferable, for the economic appraisal, to use alsofor the valuation of outputs (shadow) prices reflecting their real value iothe country concerned. In many cases, these can be estimated on the basisof (exnentd) export or imnnrt. prices, 1/ adJusted for transport costs andprocessing margins where appropriate, irrespective of whether or not the in-creased omtmut ictually is. intanei forM eprt or evyen hether it displacesimports or not. Agricultural world markets are also, of course, in mostcases, riddled qth qotas, subsidies, international agreements, regionl preferences, residual markets, etc. However, export or import prices, dis-torted as they may be may eesent the real value of the com-modities foe the country concerned. For internationally traded com,nodities

1/ Converted on the basis of the real rate of exchange. Import dutie7 are,of course, to be excluded and export taxes to be included for thispurpose.

- 10 -

imports or exports are always the alternative to domestic production or sole:on the local market. Prices of some other agricultural commodities, not enter-ing into international trade, may perhaps be estimated by conparing them v.ithclosely related international products. Similar link--ups may be necessa-y toaccount for quelity differences and consumer preferences. This raises, ofcourse, difficult but perhaps not insuperable problems.

31. In still other cases, hoievW, such internationally-based price: irillgive little guidance as the national market is more or le-s isolated on ac-count of interrational transport costs. At import prices, domestic d2mand,if any, is then only small, and potential export prices would be very lou.For such "national" conmodities, relevant shado, prices can only be estimated,as best one can, on the basis of national circumstances.l/

32. It was suggested above that prices should be adjusted for processingand transport costs. For commodities actually exported this is relativelysimple: the costs of processing and transport to the port of shipment arededucted from the export price.2/ It should be noted that, to the extentfeasible, real costs should be used here, not the costs in terms of marketprices: if. e.g., transport facilities are under-utilized anL the cost ofthe additional transport is consequently very small, only these small costsshould be deducted and not the perhaps much hiaher freight charzes actuallypaid.3/ For output destined for the domestic market, adjustments for pro:ess-ing costs if any, are the same. but a idment as to t.h relevant transportcosts is loss clear-cut. In principle, it iould seem most reasonable toarrive at ex-farm nrices by correcting international prices for the din real cost of transport betueen area of production and that of consumption.,nnd ht.teen porf. of entrnd ara nf consuption, rsetively. Vr otenof course, it will only be possible to mace rough estimates.

33. is hatever prices are used - internationally-based or local-mar'retpJIces - forecasting the future l -e 4 be necessary, and diffic lIt. Tosome extent it might be thought sufficient to use present prices (or thoseexpected in the near future) as cLangZ in relative prices miht be compen-sated by changes in the cropping pattern without affecting the total value of

I / 7 . L _ __A __ - - -- 1 - . 1

±/ 1Ljurt adiu export prices for sucn COMMOiies indicate only the up-)erand loier limit to the shadow price.

2/ Or, alternatively, the processing and transport costs should be includedamongst the current production costs.

z/ Temporary under-utilization of transport capacity can, of course, be ig-nored. On the other hand, account should be takcn of chronic unu-utilization arising, for example, from the fact th7t, oiing to the struc-ture of tcade, traffic to a port may greatly exceed or fall belot., tr-iffT:from a port, leading to low capacity utilization in one direction.

- 11 -

benefits. Quite apart from practical limitations on the possibilities foraltcring cropping patterns, this would certainly not be satisfactory ifchangeL were envisaged in the average price level of ,he relevant agriculturalcommodities (relative to other, especially industrial, products). Futurep rice develom,ents should therefore be evaluated to the best extent possible.(The best guess ma:, be, of course, that prices vil2 not change much, as se?msoften to be assumed.) This is perhaps somewhat easier for !:orld market prices,than for local prices, and estimates of local price trends are, in fact,usually geared to world market trends.

3. Current Costs

34. The current costs of agricultural projects may be broken do,ni intotwo categories. In the first place, there are the operating costs of the"overhead facilities", e.g.. of the irrigation scheme proper, or of processinaplant or of agricultural roads. These costs consist of such items as power,maintenance, iages and overheads, and do not require further comment. On theother hand, there are additional inputs on the farm level, to uhich referencehas already been made. It was suggested above that the analysis mayv attributebenefits to investment which, in fact, result from better fanning practices;or, in other iords. that ariditionl inniif.q oth4r than pnit.l (uch na ny-tension efforts, fertilizers) are undervalued, and net benefits of the ".nvest-mont. projec.t overestiiated. SomA furthnr c o point are now inorder.

35. The projects always enviic.o more work for the farmer (and his family).This additional input of laboni, i less developed c isiner levaluated at, zero costs, however, because of the absence of alternative em-ploym-ent nossibilities. This; ma or -a not be+ apro-atl ri-ht Por the

4' . WL 4,&L.Y I&VJU WlJ ajJJ.L .L LJIka UCl .LJ I.SAU VL W,L1

present, but to assume that 10 or 20 years hence this would still be true,implics a ralhor pessUsi vire of he. p r ospect fr . co i dev -l-pme

of at least some of the countries concerned. It -:ould then seem more reaisticto asuAetha btee uill bie a s10 ut Ulstefadcly inrein labor costs.

Reasonable assuptions as to the expected development of labor costs are, ofcourse, difficult to make. "Ievertheless, this would appear to be a morerealistic procedure than to assume that the labor costs of the farmer arezero A uring 'Vte tlhole long J..Le of the project.. Te latter clear`l over.-estimates the return on the investment.

3 . The valuation of labor of farmers is of special significance forsome settlement schemes, which may be highly selective: in some casesthe farmers accepted for settlement may be superior in skills and capabilityto the average farmer. W'ithdraval of such scarce skills from farming else-where in the country will not be without repercussions on agricultural out-put outside theo project area. Fonmally, this means, of course, that the useof such skilled farm labor should be treated as a high-price cost factor, thusconsiderably reducing the return on the investment proper. In practice, onedoes not knou, of course, what the real value of such skilled farmers is tooutput; one can only question whether their use elsewhere, on other projects,iould not have given an even higher return.

- 12 -

37. Sinilar considerations apply to the use of extension vorkers rdthe necezsary sunerstructure of agronomists and other specialirts. Some-times such inputs are not accounted for at all. At other times, onL nm-ysuspect that the use of their salaries as - measi,re of their re-l costs,in terms of outrut for:!one elseuhere, is probably groosly inadcauate.If, in fact, 'raired personnel is scarce, and this lacl is a severoobstacle to raising agricultural output, the appraisal of agriculturalprojects should, in principle, reflect this in the price put on the inputor such resournes. This is nartinularlv irnortant if one insists onp-,io:ity treatment for a project iith respect to obtain;ing the -equisite

i r ncd tecnica.pl nprsnnAl- thus safe-fuarding its roperexecution. Such P concentration of resources on particular projects mayor may not beagood policy for the country as a ol hut. At qnv rnte-the high real cost involved in the use of this personnel should be takcn

33. Occasionally, allowance is made for the uages ::hic have to bepaid for addition,l hired labor. In terms of real lrbor costs, this wouldappear incors-itent with the assump-on of zero costs for the labor of tnefarmer himself (and his family). It would seem better to use the samee-;luation fo: the labor of the (Wnskilled) larmer and fo hired1 labo.1 /In rare instances also tha share of the farmer in the proceeds of the in-creased output is deducted from the gross benefits. This is done in se

where the institutional set-up provides for a share of the farmers rcco:dJngto fixed rules, tne farmers theruure resue±in hire J lab rs. Te some-

what anomalous result of this treatment is thA.t the economic return on theinvestment appears to be much greEter if ifcners a.- consid .eU to be "UsLf-employed" than if they are considered to be laborers.

39. Land rents are prices paid for the use of land. here land is no,a f2ce good its 7eneral scarcity is reflected in the 2-ent of land. rn(Ulend is abundant, this rent element w:ill be zero. In addition to thisgeneral scarcity element, if any, there are in most caces di_fer-en.ial -entelements reflecting the greater fertility, better site, or love2 productioncosts of a plot. The greater yield (return) from such plots resvilc f,om theuse of a scerce resource, i.e., "superior" land. he use of lanI in a pre-ject is therefore nearly always a cost.

O. Increases in land rents might be treated, in principle, as currentcosts. The problem, in practice, is that their real level is impossib?eto estilarte independently of the process of 'inding the best of the fecsibleprojects. 7 ents actually paid depend on many institutionrl factors and maynnt he indicative of the -eal value of land. In piact4 ce, it is the17:iorCprobably better to treat rent increases as part of the return on the project.This rent element in the return should then be taken into account lhen can-sidering whether the return on the investment in the project is aducuze.2/

1/ This holds true also if seasonal demand fo: hired labor is sufficient tueliriinate terorarily an excess supply of labo-.

2/ Sec "Interpretations of [eturn Calculations", pprir. lj7-b3.

- 13 -

Apart from the.e queAtons of Ahat, in principl, should be incluidii rostL, and at :.hat price, there are the usual practical lifficulties ofhow to obtain a reasonable estimate. In come cases, and for some items,thcze can be ba.3ed on farm budget data, but in others one may be frrced, forl.ck of any beter information, to apply rules of thul4b for estimating cur-rent cots, such as the assumption that current costs make up 60 per centof gross output. It would appear, hooever, that this raises questions asto the consistency of input and output (yelds) estitantes. The share ofcurrent costs is likely to vary ith the level of expocted yields.1/

4. The Rate of leturn

Criteria and Eeasurement

t2. The expected benefits and costs are finally compared and expressedas a roturn on the investment to be made. In the past, a custonary measlreof this return '*as the gross (increase in) output/investment ratic that willbe achieved ihen the benefits of the project have been fully realized. A"favorable" ratio indicated then that the project ias econoically justifi6d.This method is open to question because it ignores both durrent costs andthe runnin--in period, hich normally is required, after construction iscompleted, beror7e maximum expected output is achieved. It presupposes that,as a matter oL e::pe:'ience, there is some sort of a fixed relationshir) betwecnthe gross output/investment rztio and a net output/investment ratio dJetermiiiedby taking into account current costs and discounting all future costs andbenafits in order to reduce them to present values. This raquires. in turn,that not bonofits are a more or less constant proportion 6f gross benefits,as bctyeen diffe-ont nroiectr 2/ and that the time path of benefits and corts.in particulc: the rate at uhich~benefits accrue during the running-in period,is also much the same. Thi.s linits the usefuness of t.ie gars oif.niit./inves._rent ratio even as a rough rule of thurib.

L3. Increases in the current costs necessary for F project are al rasdiscuss ed. How.ever, they have not alony formally been taken into accountin the measure of return on the project. The net benefit/investment ratiohas hen cirni t.ed -'nere thrvalb nfomatio on c urr ncst isdertoo unrelIable to Tarrant use of this ratio. It is true, of course, thatulle Mr,gin of -ror attached to the current cost stmt-es *s lar. Heewith consistent esti-mates of output and current cost, the error in both typesof data 1ill run in the same direction: both tend to be overestheted orunderesti:iated. If certain chanees in agricultural methods, certain apnlica-tions or Fe-tili-ers or ot her current inputs arc not na-le't-nd t4 inr4 s

/See preas. 2 3

./ Soie such relationship is, in fact, used sometimes to estimate the netbQnefits if better information is lacking. But the relation between netand gross benefits seems to vary a good deal among projects.

- 14 -

in current costs is correspondingly smaller), the e:pected increases inv-ields are probably not forthcoming, and vice versa. Thus i 7r)uld zer

preferable to make a judgment, as best one can, about tho most likely irnnA-output combination to be realized. It does not seem very con:i-tDn" tonake estimltes of future output, but not uf the correspon ing incoe in

current inputs.

t4. Both the gross output/investment ratio and thM not output./i% t_: :nratio usually relate to the benefits w1hich are expected to be rcape(d :i"L(,r

the project is in full operation. Normally, however, it takes some unebefore, for e::ample, maximum use is made of the available ir:-ig:ti-)n --aur,and before farmers have adjusted themselves to the differrnt ways of Larmin;

now necessary. During this time benefits become gradually available, reach-ing their maximum only after a conside2able time. As discussed above., es4 i-mates are commonly made of the developient of benelits and costs during therunning-in period (but often not shoin in reports). This risin- (not) benefistream is then to be compared with the investment cost.

hf. Current nractice uses for this purnose the disconted-earh-flomethod to obtain the internal rate of return of th3 project, i.e., therate of discount which equalizes the present discounted value of b7rnefttsand costs. Alternatively, benefits and costs may be dIscounted by the rnteof interest and the nresent iorth of the t!o st,rs comnred (-resent w.rth

method). For most practical applications, and as long as it may b- reasonallAspmd thnt. the rate f in i q cnt.nt over ethe actul eo

used seems of secondary importance. In conside'ing whether the return onthe project is adequate both methods led to the sane practical reiolt, p:a-vided that both use the same (constant) rate of interest.1/ This rate actsas a ym,rdst-ic-' by) ich l the intern' rate obtained by the fi:j -4ho stl

be measured, and as the proper rate of discount to be employed for the senr

1/ nd provided both methods are properly applied. In decilin. .t:' --mutually exclusive projects (project design) one zhould net :the project 'Jith the highest over-all internal return (el.n >c .the rate of interest). The project -ith lower :vr-l int .:.ny nevertheless be superior (for ,ra'ple i the lwe r:turnon a much lrger investment or over a much longcr :)er'Od). ihtest is ihother the internal caturn on the incrncet-1 iveon.the diCference betueen the two projlects), or:ce-Ci t, ,..tc,ost. Pissible misinterpretation 4n such comec is _z ilthe present. worth method.

- 15 -

-11- 3 4 . . ..i ..~ 4-~ - 1,4 1 A -- P1 -- + + V, 1one. This particular ret of& intres shul reflectL Ih rea cost A. O '

c?pital 1/ t7hich, as is well knoion, often deviates considerably fromm7rket interest "ates. Various attempts have been made to estAtJ1 e therate (s) of interest, none of which is very satisfactory. Neverthcle%s,5ome judgment as to the real rate (s) of interest has to be madc. Nomethod of project appraisal can by-pass this question.

4A. It should be noted here, in parentheses, that the constructionperiod is sometimes treated separately even when the internal rate ofreturn is used. The latter is then calculated oack to the capital in-vested at the end of the construction period. This estimate of capitalcosts allows for interest during construction, but often only in part,since the interest rate used - reflecting the rate charged on nationalor interest loans - generally understates the opportunity cost of capital.Interest during construction should be calculated at the real rate of in-terest (and, of course, over all construction costs). Or it should beavoided altogether b-, following the current practice of calculating theinternal rate of return from the beginning of the construction period,,rather than from the date of completion.2/

Interpretations of Return Calculations

h7. More important than the method of calculation is the interpret:a-tion of the return on the project, .Yhether expressed as an internal rateor a net present worth. The following comments will be based on the in-t-rnal rate of return, out apply equally, in essence, to other measuresof the return on the project. If there are good reasons for thinking thatthe real rate of interest is of the order of, say, 10 Der cent. and thein'ernal return on the project is found to be 20 per cent, does this meanthLt the project is excellant? It does, !ith some reservations. As hasbeen discussed at len,th, there is considerable uncertainty concerning thevaluatiDn of various cost itemF. It is obvious that any uderestimationof the real costs will inflate the return on the project. It was stressedcoevo that, apart from many "ordinary" dif"icilties of estination, thi. islikely to be of special significance in relation to scarce skills.

1/ In other words, it is the "shadou" rate of interest. Strickly speaking,the shadow rate of interest, like other prices, is likely to vary overtime, and the future income and cost streams should then be discountednot at a constant but at varying (compound) rates of interest. Andstrictly speaking, therefore, Lhe internal rate of return is not a usefulmeasure of the return on the project, as there is no feasible yardstickagainst which it can be measured; only a net worth calculation, P,t thevarying rates of interest. gives then a "good" investment criterion. If,a change in interest rates over time is ruled out - if only for olbviou3practical reasons - the internal rate of return has some advantage ofconvenience: it obviates an explicit choice of the recl rate of return.It faces this same issue only afterwards when decdina ihether the in-terral ra+,e of return on the project is adequate.

?/ This avoi2s the imlibit usp of t'ro difeer-n.t `ntbreLt eates - one durinthe ir .. t p.ri(i nne ~no -n~inn.pa~ no intrv-n r -n+q

- 1'S -

LJ~) * I,ven Lif all such1 co t assocUr~.L 1U VJLU111 U11A ~J _~L ~U.LA L,"; P, ~i. ven f al suh cots asocated with the project couldbeporlaccounted for, the project might still show a return far above the real -,ostof capital. If so, this would then signify that an additional rent woui boearned on the land benefiting from the project, reflecting it. choice loc L-tion, fertilit;, or simply the general scarcity of lan!.1/ In princi)le,this "surplus" iould be eliminated if increasBsin real land rento could bedetermined independently and included as a cost iteri.2/ Jut, as argued above,real rents Pre difficult if not impossible to estimate beforehand and onlyemerge from the project evaluations themselves. Consequently, in practico,the expectation of a "surplus" to be earned on a project over and above thereal cost of capital is no guarantee of its economic merit. Other projectsmay have even larger "surpluses". In other words, if the value of land usewere properly accounted for as a cost, the return on the project might evenbe below the real cost of capital. Selection of projects on the basis of asimple comparison with a generalized real cost of capital is, thorefore, notsufficient. Specific alternatives should be considered. Of course, to theextent that the land has no other "second-best" uses, the surplus. does indi-cate the economic superiority of the project.

49. This brings us to another, and related, question. It has been pointedout that the increase in gross output resulting from the project is pIrticu-larly uncertain, and hinges on the attitudes and reactions of the farmers.

It is this very large human element in the determination of the returns, whichmakes the outcome so uncertain. It may be uncxpectedl, favorable, or bad.The estimates underlying the benefits should be based on ubau is consideredto be the "most likely" outcome, and not be either overly optimistic or ex-cessively ressimistic. So much goes without saying. But there remains ar.element of uncertainty (risk) inherent in most agricultural projects to anextent that may be absent from some others.

1/ This may or nay not result in an increase in land rents actually ppi-.urh incrases are nhviouslv of iznnnrtanrP for t.h- disqt-hiution a oo

net benefits of the project, and will be discussed further in ccnnection

2/ In principle, if real prices were used for all cnputscapital, labor,land, extension services, management, etc. - there would be no surpls.Each factor ,ould then receive its appropriate "return" and, bdei-tion, for the best project total costs would equal total benefits. tcrinfcrior prOjects,h , total costs -. U,ould excA eed benefits or othrr-wise some factor or factors would earn less than iheir value in altc-na-tive uses. In practice, it is not possibe to determine such prices al,

the usual computation of return on an investrent proje-ct include nameelements that should properly be ascribed to other input,. Thus, Thoughthe retu-n may exceed the real cost of capital, it m4ght not c: Ifthvalue of other inputs were properly allowed for.

- 17 -

r0. This iniist. sonm hP Intlo for in the appraisal of the prn-eirt.

Tt, co,ld appear that such adjustments for risks must depend, in principle,on thr views of the decision-maker (ay ,h rovernren.t) n to the prnh-bility distribution of the possible high and lou returns on the projccta-nd on IhiS val11Uation of uncertainty. Roghy peng,nr th+ re h e -I

(epectation of possibly lou returns, and the greater the unwillinne.:- torun such risks, the ri be th "ris dic (nd ce r

Evaluating risl: in particular cases is largely a matter of judment; butiiijmch iremains to be learned on ho-.v 61t th ,L51%.e isk eleImtJ1 Ce-n b incUJL4)a" -into

return calculitions in a systematic way.

,. Conclusions

51. From the preceding discussion it should be clear that there is nosimple, easily applicpble, mechanical test of whether the benefits of aproject are sufficient to warrant its implementation. Uncertainties abound,and professional judgment and common sense cannot be replaced by a formula.Nevertheless, it is important that the right questions be asked so that noimportant considerations are overlooked. It is probably also important bhatthis process be formalized, as far as possible, in order to delineate thearea of judgmant. It is common practice, for example, to supplement thecnlculation of a return with statements to the effect that benefits in thuform of an improvement in the balance of payments, or in the employmentsituation, rhould also be taken into account. But costs, such as the re-quired agricultural extension effort, 7hich similarly may not have beenaccounted for in the formal analysis of the rate of return are seldom re-ferred to. The risk of such oversights is diminished, and quantificationof such additional costs and benefits, :herever possible, encouraged ifattempts -re made to incorporate all costs and benefits in the formalanalysis of the rate of return on the project. There are factors, of course,which are impossible to quantify. These includa, for example, the educa-tional value of i pilot project, the impact of a project on social changeor its political repercussions. But the number of these factors .ill Valimited, and the areas where such judoments are involved -ill be pinrx-ntedand brought into sharper focus, by stretching the more formal approach :sfar as it can Zo.

52. Another conclusion emerging at many noints in this nanper is titit is virtually impossible to appraise the economic merits of projects inisolation from the general frametiork of the nationl en-rmy or even theworld at large. This follows from the need to make projections of thedemand for rommoditios to he produced- and to estimate export nd importprices for agricultural commodities. It follows from the need to considei,alternati ves in snpeci fit- r a cPneoral sns which in turn involves +need to assess the real cost of, at least, labor, capital, and foreign ex-

z- costs whirch ini d no epenA +n alone. T hsOS- .- - --- . _ .- f V _ a . 'as

been stressed that mony of these estimates of real prices can only inilcateorders of magnitude. Clearly no great p o can or should be expeted.BuL it seems reasonably certain that in cases where market prices areobviously distorted even partial corrections in the direction of more eco-nomic prices will lead to a better allocation of resources than otherwisewould be achiovnd. This presupposes, of course, that projects in any ivencountry -re evnlurted in the same manner and on the basis of the same -sti-mater of the re-al costs of identical inputs. If not, comperisons of returnsbcttucen projectr have lttle real meaning.

- 18 -

. ThJ ned for considering alternative possibilitics has bee2n otcrat several places in this paper, and the need, therefore, of econonic -nol.si- at. ar early stage in the preparation of projects. This m,7y te rspeci l.rel1vant for large composite projects, such as schemes for genoral -gricul-tural developmcnt and settlement schemes. These consist, in fart n? , In 4on-,moer of interrelated sub-projects. In such cases it is not. suifieent oconsider only the over-all return on the project. For exaipie, it is ou:tetrue that iieaLures to increase agricultural output may not be very usefulunless roads are built so that the products can be marketed. Sir.:.ikrly Lin.ro-d cannot be justified iithout the traffic resulting from the agricultuialproject. It is clear that the total benefits of such a joint proj :et shouldbe appraised tog-ther. But this does not mean that the over-all rctiurn britself provides a sufficient criterion for the economic merit of the project.A favorable over-all return may hide badly conceived sub-projezts, coripenratecby others with a very high return. In terms of the abov: examr.plc, the nei:agricultural road may have been over-designed in reThLaton to the e::rectedtraf."ic: but the unsatisfactory return on this part of the total projDct maybe made good by a very favorable return on the agricultural sub-pr;ject.Although such interrelated sub-projects cannot be 3valuatud in isolction(because of the impossibility of deternin4ng the econoic valu-3 nd costs oftheir outputs and inputs), this does not preclude a forther separate assors-ment of them after the interrelationships have been taken into account in theover-all appraisal. Inueea, such a course appears both possible ard neccss.r,in the !bove example, it is then known hor much has to be transported., thelo-est-cost solution to this problem can be found.

Quantitative Significance of 3uggested Adjustments

Th. Finally, the question naturally arises what difference the vario::sadjustments in appraisal methods sur7ested here could make for the hoiceof projects. It is difficult to generalize on this score. The dine -n-

bet!cen market and economic orires 'or incivrnl -neK itPm-, nd thmr - l-tive neight in the total varies a great deal armong projects Pnd ci.untric..1/

1/ Istimates of shado;: prcsaea best approxination. Corcin .n

in pr7ctice, only be made for some itcms, and ray go cnly ofrt OF thv. This cle7irly affects the possibl- Jpact of the suggesLt.d lju7t-

ments on the choice of projects. The significanc2 oi: various "i u +for th ieturn on the prject nay be analyzed along lines sinar ttset forth in "The conomIc Comparison of Hydroelct:ic Pro,ects :rthAlten.tiv l)-ivelopm-nts of Thurmal 11ectric Poor, Econonio L f ppa-ed by Hie.,Pan G. v-n der Tak. July 15, 19,3 (EC-117). '1ch "sntL<v.analysis" focusses a'tention on itens for which diff: ront .rtim,tquantities or prices uould have a deeisive influrice on thme -ot:i-r. r rProject. This -voids wasting time and effort ma'ing rined Prat. ts:of items i.ith - netligible impact on the outcome of th; c.oculat c.

- 19 -

Nevertheless, one may hazard some general statements on systematic biasesinherent in project appraisal on the basis of market prices. In general,a.ong the components of investment costs, the use of market prices tendstn underestimate foreign exchange costs and to overestiaate unskilled laborcosts (except for on-farm labor). Nothing much can be said about inputs ofdcmestically produced materials; and the inclusion of "financial" items anongthe investrent costs is usually of secondary importance only. On balance itdoes not seem certain whether investment costs thus tend to be systematicallyover- or underestimated. This clearly depends on the importance of the re-quirements of foreign exchange and skilled labor in relation to those of un-skilled labor, and on the degree of divergence between their real and marketcosts.i/ On the other hand, it seems certain that evaluation at market pricess-stematically favors using mechanized construction methods (with high costsin terms of foreign exchange and skilled workers) and disfavors the massiveemployment of unskilled labor.2/ whatever the use of prices more representa-tive of the real cost of inputs may mean for total investment costs and thechoice of projects, it certainly would tend to affect their method ofconstruction.

55. Current costs probably tend to be significantlv underestimated byappraisal at market prices. The cost of such items as fertilizers andplant protection materials in terms of foreion eyhange are likelv to beconsiderably higher than their market prices. In some cases it can perhapsalso be ar.-ued that the real csts of power used for irri gation is under-stated. And for current costs there is no significant offset in the formof overvalued unskilled farm labor which, in the usual apprai"al methods,enters the calculations only in e::ceptional cases (e.g., for hired labor).On the cornt.rar-, the cost of farm labor tends occasIonally to be consider-ably underestimated, as uhen projects involve the use of specially selectedand trained farmers. Most important of all probably is the undervaluation(if not complote neglect) of the extension effort and personnel necessaryto achieve the be"nefts expected from +he projeCt-. Even ro-Uhqat tvestimates are difficult to make on this score, but it iould certainly appearthat this undervaluation may resul I consiuurauljy ila returns onagricultural projects.3/ In particular it tends to make the returns on

1/ If intcrest during construction is added-- at reaivl lo-aso-1 V _tu LA6 k %I tl. L ued a, rl~atively low rates of:interest (say 6,' or lower), this tends to impart, in most cases, a

sizeble o,,m-rd bias to total capital costs.-, ~ k .. LJL A L.e L L, ujU k, k.u~J U a± C S

2/ The litter may entail a longer period of construction and delay in theavailability of the benefits. This may add sufficiently to "interest

din constu ud/or dutract enough from che value of the benefits,to offset the basic advantage of large-scale employment of unskilledlabor indicated in the text. This may also run into organizationaldifficulties in some cases.

3/ Or alternatively, this may simply reflect the over-optimistic assumptionthat an effective extension effort will be forthcoming, so that yieldsire overestimated, rather than the extension effort undervalued.

investment in irrigation projects and settlement schemes look better than

1' 12! ly aLe s' comp ared ui returns utuht my LJt oLUainal Lm. UJi-

pr ven4 in agricultural practIces ,ithout lorge capital investments.

5'. The focagcing discussion relatee to the revaluation of project costs.7 vrther co::.lications arise if the benefits are considered. It is, perhaps,a fair -enralization that in most underdeveloped countries industrial pricesaro high relative to agricultural pr-ices, as compared w,jiti relative pricer onthe "'orld i.arket". The degree of protection of industry is usually greaterthen that of agriculture (both being measured at the official exchange r.te).As far as output prices are concerned, revaluation of domestic narket pricesso as to make them more "realistic" tends therefore to favor agriculturalprojects over industrial ones, thus correcting the previous bias. It is un-ccrtain, houever, whether this remains true if the inpact of the revaluationof the cost factors is also taken into account. lore is involved than apassible shift bet,7een industrial and agricultural projects. At any ratean appraisal of invastment projects (and programs) based on a revaluation ofinput p.ices as suggested here, might result in drai.ing into productiveactivitis a large numer of otherwise un(der) employed vokers, thus makingpossitle an expansion of the total investment program.

III. OPELlTING -ZiULTS :.D FIHAACIAL CONSZC!U.714C'7S

f7. Thus far the economic appraisal of agricultural projects has beendi,,cussed in terms of their real economic costs and benefits. This isfundaiental -oi the economic justification of the proJect. But other aspectsconcorninC the financial consequences of the project must also be taken intoaccount. These are the operating results of the project, th3 share of thebenefits accruin- to the farmer, the impact of the project on public financeand savings, and related ouestions concerning the nricing of irrigation .,ater.These espects 'ill be considered in the remainder of this paper.

1. The Significance of Operating Results

53. The econenic evaluation aims at detennining the economic return onthe project, i.e., the return on the project above on the basis of :%he realvalues of costs and benefits. As discussed above, such real values, interms of shado p ces, may differ from values measured br marke plices.Similar calculations on the basis of market prices - and including somefinancial items exclueabov-_ - do not necessar i a at the eco-nomic benefits of the project are for the economy as a ihole. They shou the

-alld op4Pin result of- th`roet1Te-infcance of such ralru-

lations for agicultural projects is discussed in the follouing paragraph.

59. In the first place, the operating results oC' a project show itsprofitability from the point of view of a "private? investor. If theproject is to be 'inanced .holly privately, this is relevant for the question

hAther t offers sufficient incentives to entrepreneurs to underta'-e theprject. morver, mnst -,ri cultural projects are l-rgely financed fronpublic funcis so that tho question of incentives does not arise in this fo,"m.The relnvant incentives are those affecting the farmers and depend on their,ire in the bene,its of the project, rather than on the profitability of

the project as a 7hole.

- 21 -

C. Secondly, the operating results are a test 1 L1te s -1luiati..rcha-acter of the project, i.e., whether the capital and current costs caibe lefrayed from the proceeds of the project. If a project is self-liqui-dating, this indicates that any loans made for it can be paid back overits lifetime from income generated by the project itself. However, thismay not be decisive for the creditworthiness of the borrover, since it maybe impaired by already existing indebtedness. For public projects in par-ticular - such as many agricultural projects - or for projects !ith a publicguarantee, tUe question whether financial obligations on the p:ojact can bhonored over its lifetime depends not only on the particular project, butalso, and mainly, on the over-all financial position of the government. Thebroader issues concerning the impact of a project on public finance will beconsidered separately below.

61. By definition, a self-liquidating project does not need to be zub-sidized. It is far from certain, however, that subsidies are necessarilyobjectionable in all cases. The need for a subsidy ma:; simply reflect thedivergence between the market and economic evaluation of costs and benefits(i.e., between market prices and shado, prices). In such a case, the pay-ment of a subsidy is required in order to ma!e a project with adequateeconomic benefits for the country as a ihole attractive to a private entre-preneur guided by market prices. On the other hand, consumers may be sub-sidized even Yhen this is not evident on the basis of market prices. It isa good general nrinciple that users pay the real econoiic costs of the goodsand services they want; if they do not, they are being subsidized in a realsense. It is very possible for a project to be self-liquidating, but never-theless to be subsidized in the sense that the market price of its outputdoes not cover its real costs.

62. It. w-ould nnpar thnt the nornt.ing result.s are of ov nly limiteisignificance for the appraisal of many agricultural projects: they usuallydo not anser the relevnt q tins cnnerning incentives, are only pawrtilIrelevant for the subsidy issue, and are not necessarily decisive in determin-ing whpther an enterprise or agency carrying out the project is "crediturthyto receive a loan in terms of ability to service this loan.

2. Benefits to the Farmer

63. As suggested above, crucial for the incentive aspects of agriculturalprnoet.-zI fin nalysis ofP the benefts f-om the proje acc-n to 4he

farmer. The (expected) increase in costs to the farmer resulting from theproJect, and the %epcd inceas in hiso benefUUits.Lk O, shoVULd beV copaediorder to make sure that the not benefits to the farmer are large ancugh toJustify the expectation that his indi spensable cop-to in--- thA.ojcis forthcoming. '!ithout this cooperation the return on the project 1!ill, atbest, be small. Financial incentives aro, of courso, not tha only, opossibly even the most important, factor in the motivation of the farners.Even though the benefits to the farmar seem reasonable, the farmers ma stillbe reluctant to cooperate and other obstacles, of a pbychological and socialnature, my have to be overcome. Material incentives, however, -ould at anyrate seem necessary.

)4. It is therefore necessary to assess the increase in costs to b-borne by,l the farmer, aU Lt inLra inC thI value V oU i U pr 1 duciLProblems of yields and associated costs are similar to those discuconabove for the econM.2C eCeValuation, and the uncertainty port.aining t othese estimates is the same. But there are some questions concerning theprices to be used and the items to be included.

6$. On the cost side, prices for fertilizer and other inputs shou-ld,of course, be those actually paid by the farmer, including taxes and s3ub-sidies, if ony. If the farmer has to borrou money to equip himself, fc,rexample, to buy machinery, seed, fertilizer, etc., the cost of thi: b-rrw-ing should be taken into account. Similarly, iiater chrges, land better-ment levies, or other charges for the improved facilities provided, shouldbe added to the costs of the farmer. And on the benefit side, the increasein output should be evaluated in terms of the prices actually received bythe farmer. So much is standard practice. The estimate of the farmer'sbenefits should Plso take account of the, possibly increasing, share goingto middlemen. The higher incomes accruing to farmers may result in an in-crease in land rents being paid, in one fona or another, to landlords (ormoneylenders). If so, this will obviously tend to neutralize any incentivefor the farmers to cooperate in the prcject. In fact, idespread fear,even if unwarranted. that any increase in their output will only benefitthe landlords and middlemen irill have the same effect. This question thcre-fore deserves thorough investigation in the appraisal of agriculturalprojects.

66. The assessment of the benefits to farme-s requires, of course, acomparison between their post-project incomes and the incomes they other-iise might have enjoyed. For most projects, this turns on a comparison offarmers' incomes in the project area tith and without the project, As afirst approxim,rtion, the latter may be assumed to be incomes (or net outr,tper he,iu). For settlement schemes, however, the relevant comp-ri3on is nit,rith incomes in the area, but with incomes and income expectations of theprospective settlers in their place of origin. Expected incomes in thesettlement area should be sufficiently favorable by comparison to compenoatenot only for the larger effort frequently required, including possibly someinput of cqnitq1_ but also for the nonn-pecuniar7 factors involved in themove. 1/

1/ -I so-e c Pes afr exmple, whr adsettlement clLims ;:,lev-J t- t,i.dl-

spread land hunger, the incentive for the farmer to move may be obvious.I-)utu ev en t h en, c om p a ris3on ofI e xpe ctIIedU s e t tle-r' i nom es :ih a nd U-::i thou -itthe project may be necessary if later setback: due to disanpointrnt orthie settlers a"e to be avoided; this would C.-e__. ' UFUcill tru fi?te

prospective settlers are not landless laborers but rolatively 'e11 cst2-b-lished fanc,ers-.

- 23 -

67. In considering the benefits of the project accruing to the fannersthe time element should not be forgotten. In many, if not in most cases,the net benefits vill not be a steady flow from the beginning of the project,hut will come in 7if.h a cnsiderable time lag.1/ This raises the familiarquestions concerning the relevant time discount rate in a somewhat differentform. To be adequate as incen+Av fl henenef'its sh1rd in the, opinion of

the ftnner, become available reasonably soon. Furthermore, even where thisi so, the io pi ir resn roble. In the i.nItial stages

of the project, the farmer may be required to make relatively large cormnit-mont A Ihich, in 1 or -in his Opinion, It s in-

portant, therefore, to study farm budgets, on a sample basis, and the furtherUevelopmtent of the farmers' financial position ulri.ng1 LW Je A i wAeA the

benefits gradually become available. It may be necessary to give the farmersome reLief in this period by ikeeping Such payments aand debtservice lo7.

68. Questions concerning the adequacy of the incentives to other partici-pants in the project - such as private licencees of irrigation facilities,investors in new processing facilities, or marketing agents, etc. - may besimilarly dealt with, if relevant and sufficiently important. These are re-latcd to the evaluation of the operating results of appropriate sub-items.This part of the financial appraisal should, of course, not consider the fi-nancial return on all capital invested in such projects, but the return onthat prrt of capital (and labor) contributed by the participant. For in-centive purposes this should compare favorably vith returns obtainable else-,here in the private sector.

3. Public Finance and Savings'

69. It is sometimes suggested that the economic appraisal of a projectshould be based not only on its economic return but should take account alsoof the government revenue or, more generally, the public and private savingsgenerated by the project. The underlying idea is that many countries ex-perience Zreat di.ficulties in raising their savings ratio from present loulevels, and that governments are faced iith serious problems vhen attemptingto raise revenue in line with increasing current expenditure, let alone suf-ficientl- to make a contribution to national savings. As higher savinCs (andinvestments) are crucial for higher rates of output in the future (and to the

1/ Even if the economic benefits have a time lag, this is not necessarilyso for the benefits to the farmer, depending on the financial arranga-ments made and on the policy as to, for example, water charge; duringthe running-in period.

2/ Strictly speaking, this sub-section deals with the question whether theimpact of a project on public finances should be part of the economiccriterion foi selection of projects, rather than with its purelyfinancial aspects.

extent that -overnmen.t c e exenditum i c nsd`rri more "1iprrant"th2n Trivate concurption), the argument iould seem to tmply that a pezmium

_ahes to roj cgeneratng more v= -n t ne o s g nr tav revenui. Ty the

extent that these projects are also those with highe: returns therc it, ofrs n rue iseen are onl y when i - +n an ai; suarei at i

,ith lo'or returns. The argumant implies that projects*thich are in'eriorn termsi~ of direc contriution to output may be super- ior on account of their

"feed back", i.e., if their indirect effects on future output, through highersavings and investment induced by the, are taken into account.

70. There is clearly something to this idea, bUt its pract3cal significanceiould seem rather limited - and it has seldom, if ever, been applied in theLori_al econoric 'ccQarison of projects. It assumes that nationail saving's ,ndgovernment revenues cannot be raised, or not sufficiently, bi- alternativemeans which do not interfere (or interfere less) with the efficient allocationof investment resources; that an induced increase in savings and G,overinmentrevenue through the choice of projects (the pattern of investment) is feasiblein circumstances precluding autonomous increases through fiscal policy; andthat the indirect gains through higher savings generated by the "inferior"project, outueigh the direct losses in output. It se?ems doubtful whetherthis combination of conditions is often met vith in practice.

71. Countries do have difficulties, of course, in raising their nationalsavings rates and public revenues. The major problem, houever, wouldfrequently appear to be the existence of political obstacles to increases intaxation, and a lack of determined efforts on this score, rather than anorganizational incapacit,l/ to raise tax rates (including prices of publicitilities) or to e::ploit new sources of public revenue. In such cascs, itiould seem open to question whether substantial increases in the averagetax rate through systematic shifts in the composition of tax objects t7ou3lpass unnoticed and without political opposition. A selection of projectssuch as to favor profit over wage income, in order to increase savings, isalso likely to have political repercussions.2/

72. Furthermore, a systematic shift in project selection in favor of asector uith higher savings, but lover direct vields. and avay from a sectorwith lover savings but higher direct yields, would tend to uiden the gap indirect vields bettY(en the tio sectnrs. Over-investment in higher saviags/lower yield sectors increases the direct losses in output relative to the

gains in savngs. Tiis- limit, the ran in which a shift in inrer;tment to-,:ards

1/ Or prouh , 1- iitivLe costs involved in raising adu"it-ional revenues; o,r Lick oftechnical data such as cadastral surveys as a prerequisite to land

2/ Th absolute volumes of savings and taxes should, oi' courso, he hig!--,not just the average savings and tax rates.

such sectors inay be adv-intageous. Finally, projects which are apparentlynor-revenue ( i-avng ) po ducing mny, in fact, have largebilt4-in svings(reinvestment) rates if they are sloi-yielding. This ,ould seem furtherto rcstrict the- scp _or possible 4plcto oth"edbc"ideq as a~-' ~ ~X J-t. LI;LC Zl%UP UJ. VV,1Q.U.LU CJJ.LUCaLU.LU1L UJL kl4~ lu Uc L U ~ -dcriterion in project analysis.

73. To give much weight to the savings and tax effects as a criterionin project analysis, it would appear necessary that direct measures to raisesavings and public revenues -are extremely difficult for organizationalreasons, and that, moreover, "automatic" increases in zavings and tax quotas,induced by the selection of projects, 1/ would not soon meet with politicalopposition. This particular combination of circumstances does not ceemlilkely to prevail in many countries. And whore it does, the quantitativesignificance of possible advantageous shifts towards lower-yi.eld-but--higher-s:nvings projects would seem rather limited. It would appear therefore thatthe tax-curi-savings criterion is not often relevant to the actual selectiono projects.2/ This does not imply, of course, that public revenue aspectsof a pro,Ject - once it has been selected on the basis of economic evaluationas discussec earlier in this paper - are not important. They play a largerole, for instance, in the issues concerning pricing of public utilities,i.e. irrigation water.

4. Pricing of Irrigation later

74. Although the impact of a project on public finance is unlikely toplay a substantial role in the selection of the pro4ect; it is neverthelessan inportant aspect when considering the appropriate price policy to follow.At several points in this paper we have already touched on the question ofwater rates: the price paid by the farmer for irrigation water made avail-able by the prcject.3/ The rates charged are obviously important for theoperating results of an irrigation proJect: they influence the benefitsreceived by the farmer from the project; and they enter into the revenuesfrom the project accruing to tha government. These are all financinl asnectswater rates influence the economic evaluation of the project only indirectlythrough their impact on the incentive to the farmer and on the use of water,and thus on expected output. The difficulty of ascertaining the economicprice for irrigation water is, in fact. the rpannn that thb onnmit I-vMi1n._tion has to be conducted in terms of the value of the increase in agricul-tural output.

1/ I.e., without any change in savings or tax rates: such increases inaverage savings and tax quotas result simply from the change in output(or input) mix.

2/ See the Annendix for some further wnments of the quantitative as.ectsof this issue.

3/ Similar problems may arise with other facilities provided by publicinvestment in agriculture, but irrigation water is the most important

example.

- 26 -

75. Tt "as argued above that the operating results are of limited -ini-ficance for public irrigation projets The water rates issue is t!hereforeconfined to t,,o major aspects: the impact on farm incomes and behaviour -on

the one hand, and fiscal considerations on ohe. The f-ll-winw hrief

comments refer to each in turn.

76. As noted before, the benefits accruing to the farmer ray be a de-

cisive factor in the success or failure of the project. The human factor

plays a large role in deterriing the increase in a-ricultural output Lhatmay be expected from the i4provement in irrigation .acilitieo. If th

farmers' cooperation is not forthcoming, the fact that large increases inproduction are physically possible will be of ro avail. To elicit his co-oneration material incentives are a necessary, but perhaps not a sufficientcondition. It may also be necessary to overUme psycoo and zocialrosistances to change in the accustoned ways of doing things.. But maturialincentives are no doubt an imoortant aid in inducing such chang--s. Thelevel of the water rates charged is one, but by no mieans the only factor in

determining the farmers' b3nefits from the project. Other taxes, possib:!increases in land rents, the terms at which necessary credit is made avail-

able, and, of course, the price he receives for his produc's, may well beas, or more, important.

77. 1'ith these p-ovisos, then, water rates may be manipulated to en-courage the use of irrigation water - and thus ensure the cucces: a theproject. To that end it would appear a good idea to charge little, oreven nothing at all, for the use of water in the initial stages of theproject. That is the time when farmers might be reluctant to use tne waterout of fear that the promised benefits may not actually occur, and thatthey would be saddled with the extra charges. Later, when the benefitshave proved themselves, water rates can be charged without any unto-ardeffect on the use of irrigation water. The exact modalities of such arrange-ments may change from case to case, but as a general principle, token pay-ments in the early years and a gradual increase thereafter until the fulilevel is reached, would appear logical. Some such suggestions are, in fact,frequently found in appraisal reports of irrigation projects. But it wouldseem that this idea inight be pushed even more vigorously, ztressing louerpayments in the beginning - and h4gher ones later on.l/

78- This leads to the question of the appropriate level of the waterrates after the initial running-in period, then their incentive aspectsbegin to lne in imnortance and the fiscal aspects gain pr2doninance. Itwas argued above that differences in the contribution to savings and fiscalrevenues are not, in general, significant fActors in the selection )f

1/Even after the running -in per"od it might be advi sable omk pyeb

24 11VA 1 V, I ,11 I UAUfl11LB .US V 11 '4 .-- - 1- 1 -- 1..~~J S ~ JJhL~. ~ .t&.

of water charges conditional on realization of the benefits, in order toI .~ I- i - _ _ - . _ _ C. rk. -L A. I~ _. - -*I-- V..avoid ha.LrusiI.Lps in1 case o1 curup .ailure. DU. tl s .oU.U aLs) fet otht'.,

fixed charges on the farmer, and other measures than annual adjustments iithe water zates might be more suitable to cope with such vagarics.

- 27 -

projects. This does not mean, of course, that fiscal possibilities openedup by ne: projects should not be exploited. 1ater charges may offer an ad-ministratively convenient instrument of agricultural taxation. As withother public utilities, there is a thin line between the profit element andthe turnover tax element in the total price charged but this is irrelevantfor revenue purposes. From an economic point of view, it is a good generolprinciple that charges on water users cover the real cost of providing water.Over and above these costs they mi2ht possiblv make a contribution to gener-alrevenue.

79. Several considerations hovever tend to limit the extent to which thisnrincinp c.n be npliid in practice. ThA o-verridqing imnnrtanr#e o)f enr agnir,1-ing the use of water during the transitional period was stressed already.And in the later stages the requirements of efficient use of water should notsimply be sac-ificed to revenue needs. The structure of water rates is as im-portant as their level, if not more so. Careful consideration should be giienin devising water rate systems to maintaining a proper balance between effi-Cieont waterY usen nndl Yrvnue comside.ation...

farm incomes are very low so that the farmers' capacity to pay is limited.Thi4s does not seer, consistent, however, wit th lag inrae in -r in-A 1Id) U.U' 1'~ .UJI LL, 41UVUVtL:L W-I UM L UI-Kt .L1L;1LLUC±it .A .1 r U 1 1-

comes expected from many projects. Even if water rates were to incorporatea definite tax element, they :ould only absorb a relatively small part ofthese increases in incomes. In view also of the fact that the agriculturalsector is large in less developed countries and that its contribution to taxrevenues is frequently relatively small, there would seem to be good reasonsfor levying higher tpxes through water rates. Administrative problems, fre-quently invoked against suggestions to raise taxation in developing countries,hardly apply in this case. If it is possible to levy low water rates, it isalso administratively possible to collect higher charges - and at lesser costper unit of revenue. Nevertheless, it should be admitted that, in some cases,it may be difficult for the fariner to make cash payments, even though hi- in-come rises substantially as a result of the project.

l. These problems should, of course, be seen in a broader framework. Itis hardly reasonable to make the beneficiaries of new projects pay high waterrates, if other farmers who derive their water from older schemes are assessedvery lightly. Equity requires that roughly equal treatment be meted out toall. In many countries water rates on existing schemes are relatively low,and would have to be raised substantially before rates on new irrigation pro-jects could be set on the basis suggested above. This may.vell involve poli-tical difficulties which can only be overcome gradually. People everywh"reare most reluctant to pay more for what they have received hitherto at moreor less subsidized rates. However, provided a general revision of water ratesis started early enough, and pressed vigorously enough, this would not inter-fere with the application to new projects of the "new" policy, which, i'. willbe remembered, envisages during an interim period very low rates to encouragethe rapid achievement of full utilization of all irrigation water.

- 28 -

IV. SUMMARY

1. In recent years th3re has been a gro-ing interest in problems of2conomic appraisal of projects and conniderable wo:k 1:s been done onthis subject. The aim of this paper is to present in a systematic waysome of the current thinkiny A it. relata to the apprnisnl of ngriculturalprojects. It includes a general discussion of some economic and financialaspects of appraisal, as a bass for considering the economic logic ofappraisal methods. Although a paper of this kind is not intended to serve

sau to project apprai4sal 4 + can See useful ppeb fcus',ingas a mil j -- -- )j.. -w art aaa -u - ~ ' t~ ~ jtJ -'J 1 _ t

attention on some of the economic issues involved. It is important thatthe right questions ar asked, and-A' attemptl,s at q -ua,,n i o ins of4 -+ J'- ' a. as

nossiblo, impose a discipline which may contribute to the continued de-

vepnent of more effective aPraialJ pracices

* 4 L s~ Jmr tan'~.LL LOILW.~d~±4 .L La.II LI UekaWt:UL UJLt:: t:UULU.U2.~~ ItiUiprtn to make a clear distinction btenteeooiapproisal of a project on the one hand and the appraisal of its financialconsequences on the other. rThetotpso1apaslar ietda

different questions and require a different measure of costs and benefits.The econoiruc apprisal is corcerned with the real economic costs aid bene-fits of the project and with the problem of allocating resources so as tomaxinize the benefits to the country as a whole. The financial appraisalconsiders the costs and benefits of a project in terms of actual receiptsand cxpenditures only, in order to assess the self-liquidating characterof the project, the ndequacy of its incentives, and its impact on govern-ment finance.

In the process of economic appraisal, the need to consider, to tneextent feasible, alternatives at all stages of the formulation of the pro-ject is emphasized; e.g., in designing its major features, in considcringthe merits of various sub-projects, in deciding the best cropping pattern,and in considering alternative uses of such scarce input3 as organizationalskills, farm talent, aZricultural experts end extension workers. This isnecessary because of the impossibility of putting appropriate values onsuch cost items as land rents, site values, entrepreneurship, extensionefforts, etc., which tend to be undervalued if not ignored in practice.Thus, a project which might look attractive in isolation could neverthelesscompare unfavorably with an alternative use of the same resources.

4. The question of which items to include in the evaluation of ecoroniccosts and benefits and the prices at which they should be valued are con-sidered. Purely financial items are irrelevant for the economic evaluat.onof 2 project but, of course, not for its financial results. lhera marketprices obviously do not reflect economic values, adjustments should be madeand the use of so-called shadow prices should he considered. The nrarticaldifliculties are substantial and estimates of shadow prices can only be ap-nri,:imate! nevertheless. In many Fass .h ene of shadon prices at 1pAetfor r--jor categories such as foreign exchange, cost of capital and (unskilled)

labor, may give a more realistic measure of the economic merits of a pro-!ert, thpn ,ould be obtained if market prices were used. Use of shado)w)ri:es for the economic valuation of the agricultuiLal outputs expectedr)m Lnno poct (and perhans for some materials used) mar also be appropri-

a.e in :om, c-ses. On the other hand, the shadoi price of some other inputs-zich ns f,r orample, the required extension effort. is irnpsihlA to assess

.ithout considering alternative projects or, in effect, alternative programs.

5. As a measure of return, the customary internal rate and net present-orth criteriI. are endorsed ith some qualfications. Mensures used in thep:.st - the raAo of (increase in) gross or net annual output to investment- are open to question asa h +eio +4th 4

ie niral from +hn ped nf +th

construction period until the full increase in output is achieved, or theinc--s in curetcot, or both.I S-U rati -- V-_...s ca onl b. inerr 1e by.JA 1 -MC>U 4.11 L(I - -A A &J. t' O '.- L _ ~ VAJ~ . LJUSI.L& I M . 0.L % 4L4~ V L L,) 4 4 JJ - J

rcference to current costs, and length of running-in period, rate of interest,

L. It is sugted U thUt, I Prini-±p±le .U, ta eVaLULUIn of Le IaL1CiCJLaLoperating results of a project may shou, first, profitability of the projectfrom tle point of view of a private investor (and the adacuacy of his in-centives) and, secondly, the self-liquidating character of the project or theneed for subsidies. It is argued, however, that for many agricultural pro-jects the operating results are of only secondary significance. The questionof incentives depends on the benefits to the farmer rather than on the fi-nancial results of the project; and the "credit.orthiness" of such publicprojects does not depend on the self-liquidating character .of the projects.Self-liquidation of agricultural projects, moreover, is seldcm insisted onin practice.

7. On the other hand, the return to the farmer for his labor and possiblyhis capital, is of great importance for the outcome of the project. LandrentF paid by thD farmer to the landlo-d may be raised in one uiy or ano+Fer,or, more gene.al1V, middlemen may appropriate part cf the increase in output'esulting from the project, thus loweiing the benefits accruing to the farmer.The resulting net income to the farmer, after taxes and Yater charges, shouldbe c-omplred with his income without the project, possibly not in the projectarea but else!here. To be adequate as incentives, benefits to the farmershould come in 7ithin a reasonable period of time. Close attention should bepaid, moreover, to the financial position of the farmer during the periodihen benefits gradually become available. The project may entail "large"initial commitments on his part. Even though the benefits to be reaped byhim after only a short while are large, the former will :lot be interested if,in fict or in his opinion, he cannot affoid this initial outlay.

8. In general, differences in their contribution to savings and fiscalrevenues a0e not significant factors in the selection of projects. But thisdoes not mean, of course, that fiscal poscibilitieo onened un bv net uro",ctsshould not be exploited: water charges may offer a conv.enient instrument ofagricuiturol taxation. From an economic point of view, it. is a good generalprinciple that charges on water users cover th3 real cost of providing water- and the,r might possibly make an additionna cont.ribution t. 7o general taxrevenue. oiever, during the running-in period incentive consi.derations arguefor lon "ator rate.- And the need f efficit use of water, the possible

- 30 -

difficulty. for the farmer to make cash payments, and political and soci.2considerations may also limit the extent to -.hich this principln c,n beapplied in practice. In particular, it may require a gencral ch:ne innater rate policies: it is hardly reasonable to milze the beneficiaries ofnur projects pa- high water rates if other farmnrs !;ho derive the;i -atecfrom older schemes are assessed very lightly.

- 1 -

The Savings (Tax) Criterion for Project Selection:Some Further Comments

1. 11hat revenues are to be taken into account if the selection of aproject is to be based not only on its direct return, but also on itsindirect return, i.e., through its contribution to naticnal savings(and further investments)? Aq far a taY vipld ar# eonnnrnedreference has already been made, when discussing the operating resultsof the nrniect. to revenues renresentinry nament fnr ervices renderPd_such as water rates and land betterment levies. These should, ofcourse, be included to the extent that they accrue to the GnVenment.Otier indirect taxes on inputs and outputs of the project, such as importt.ne, ePornt. axes lnndi tfxsc etc, areev-nt i"nsofn" fts theryield increases as a result of the project. Increised income tax pay-r.ents by the beneficiaries of the project must also be takn io aOn the other hand, some revenues may decline on account of the project.

This may. .sp be the resul. of the disoearance of tle Itax object, "isin the case of land taxes on land flooded by an irrigation reservoir.T" e effect may also le more indirecto i the project results in impiortsuhstitution it seems reasonable to allow for the loss of import taxesin evaluating the impact of the project on public finance. But wheredoes one draw the line? Are all changes in incomes and expenditures inother sectors brought about by the project (and the concomitant changesin receipts) to be taken into account? And subsequent multipliereffects? it seems clear that this leads to insurmountable practicaldifficulties; even theoretically, allocation of the indirect effects tospecific projects is only possible under strict ceteris paribus assump-tions.

2. Apart from these considerations, however, the impact of a projecton public finance consists not only of gains and losses of revenue butPLa,, also involve financial contributions to cover further cash require-ments in the initial stages, or permanent subsidies to cover operatinglosses. Such payments should be deducted from gross revenue receiptsin order to obtain estimates of the net increases in revenues from theproject accruing to the Government. As far as the contribution tonational savings is concerned, this net increase in revenue is onlyrelevant to the extent that it does not lead to higher current expendi-ture, nor to lower private savings.1/ Changes in private savingsresulting from the project can be similarly assessed. Differences inprivate savings between projects depend on the size and distribution ofincome after tax resulting from each project and differences in savingsrates of various groups of inee receivers.

1/ For a more r,eneral argument, see below, page 2.

-2-

3. Finally, these additional (public and private) savings streams haveto be evaluated and made comparable with the direct economic benefits ofthe project in order to determine its total benefits. The value ofthese additional national savings derives from the further investmentand increase in output made possible thereby, and should therefore bebased on the productivity of this additional capital. The increase ininvestment made possible by savings resulting from the project yieldsa whole series of additional income streams, starting in successive years.It should be noted that the time shape of the income stream of variousprojects is of major significance in this respect. Slow-yieldingprojects have lcrge built-in savinga/reinvestment rates. The sum ofthese additional income flows discounted back to the present, equals thetotal indirect "savings-benefits" of the project. Thus, if the directreturn of the project is smaller than that of another project, but its"savings-benefits" are m6re than enough to compensate this inefficiency,it is the better project in the conditions postulated above.

4. A simple example may be helpful. Let project A and B have thefollowing direct benefit streame:

Project Investment Net Direct Benefits in Year

A 100 10 10 10 10.....10........ 10B 100 9 9 9 9..... 9....... 9

The return on A is 10% and on B 9%. with a rate of interest of 107..A is "in" and B is "out". However, net savings from project A (out of bothprofits and labor income, etc.,) are, say, h per year and from B_ 5 ner year,so that the additional savinps from B amount to 1 per year. 'hat is tevalue of these additional savings? For small amounts of additional savingsand investment, the productivity of extra capital may be taken to equal therate of interest (in a final return on investment) of 10 per cent.-Assuming for simplicity's sake this rate to remain constant over time, theadditional saviivs of year 1 produce an addiltiAnnl inome qtAmm of 1 tjme10% - 1/10 per annum ad infinitum. Further additional income streams of1/10 per annum start in year 2-3-b- etc. Total benefits of Bare there-fore, in our example, as follows:

Year 1 2 3 4........ .20..100...

Direct benefits 9 9 9 9_ 9Indirect benefits 1 1/10 1/10 1/10 1/10....l/10....1/10o....1/0...from additional I /I /-I 2n , A investment in *year 31/10 11,.11.,11.,/0,

4 ~1/10....1/10 .... 1/10....1/10...

10 1/10....1/10. ..o.1/10o.20 1/10).... 1/10...

10i110.

-3-

5. In year 10 total annual benefits of B ("direct output" lUs "reinvest-ment benefits") catch up with the "output benefits" of A. Kii only inyear 20 will the accumulated (undiscountedJ) reinvestment benefits of B

__ _ - 4. L_ _ _ - _ tvA _4. _4 - .4. 1 -- - -- A -4+V A TPWipe! oUL UTV acULUULted dr JAJU.Lutp lsseJ s aS c r.AA A. I-

the "reinvestment benefits" are properly discounted (at 105) it takes muchlonger. Ultimately, however, inU.[ uUrIample, the total return on Bequals that on A.

6. This is clearly not necessarily so. Far "small" additional aimuntsof reinvestment, as assumed above, the outcome depends on -behther thedirect "output loss" is larger or smaller than the "additional" savingrenerated by the project. If the additional savings are "large" theaverage productivity of the additional reinvestment will exceed theirterest (discount) rate. In that case the "direct output losS" may bebigger than the "savings gain" but nevertheless smaller than the "reinvest-ment benefits". It is difficult to see how these considerations can begiven concrete operational significance.

7. The argument thus far has been developed in terms of different contribu-tions to national savings; higher tax revenues were considered relevantonly if they provide a net addition to national savings, not if they trans-fer private savings to public savings (or, a fortiori to public currentexpenditure). It mig"it be possible to refine somewhat further the valua-tion of the increase in savings and public revenue brought about by theproject, by allowing for possible differences in the relative social valueof increased public expenditure on the one hand, and private investment orconsumtion on the other. Thus, if iucreased revenue is likely to resultin higher public current expenditure and less private codsumption, onemight differentiate according to the relative importance attached to morepuLlic expenditure (on education, on defense, on civil service salaries?)and private consumption (of beer, or of education?). Or, if higher publicrevenue is associated with lower private savings and investments, thismight mean less private investment in luxury flats and more public currentexpenditure on education (traditionally classified as "consuption"), butit mirht also mean less private investment in a textile plant, and morepublic investment in a steel mill or an increase in diplomatic representa-tion allowances.1/ The valuation of the savings and revenue effects ofte project would thus depend on the use that is made of such funds byprivate investors and public authorities. It is doubtful whether thiscan be given any concrete content for purposes of project evaluation.

8. In view of this discussion it would appear that appraisal reportswisely do not tAke Accnunt of thene qavinngs and mihlic finance aspects intheir evaluation of the return on agricultural projects. Occasionally,ho.PvPr_ there are TAfPrences to the increase in general taxationattributable to the project. These are usually included amongst thegeneral benefits of the r"ciat ot thNn t.hoe%a aercte in the rate ofreturn, but without indication of the value (weight) given to these taxhnefits in ralAtin +n the return proper. Increases in savings appeL:never to be consicered at all in this context.

1/ In general, an increase in public revenue is likely to entail an increasein public investment or other development expenditure, ard a decline in bothprivate consumption and savings investment). Thus, it is likely to producea net increase in total savings and investments), or at least in develop-ment expenditure.


Recommended