Letters

This art icle was downloaded by: [ Universit y of Glasgow] On: 04 June 2014, At : 07: 36 Publisher: Taylor & Francis I nform a Lt d Regist ered in England and Wales Regist ered Num ber: 1072954 Regist ered office: Mort im er House, 37- 41 Mort im er St reet , London W1T 3JH, UK Proj ect Appraisal Publicat ion det ails, including inst ruct ions f or aut hors and subscript ion inf ormat ion: ht t p: / / www. t andf online. com/ loi/ t iap18 Letters a John Robert s , Ian Livingst one b & Dr. Michael Tribe c a Overseas Development Administ rat ion , 94 Vict oria St reet , London , SW1E 5JL , UK b School of Development St udies, Universit y of East Anglia , Norwich , NR4 7TJ , UK c Development and Proj ect Planning Cent re , Universit y of Bradf ord , Bradf ord , ED7 1DP , UK Published online: 17 Feb 2012. To cite this article: John Robert s , Ian Livingst one & Dr. Michael Tribe (1996) Let t ers, Proj ect Appraisal, 11: 2, 133-139, DOI: 10. 1080/ 02688867. 1996. 9727028 To link to this article: ht t p: / / dx. doi. org/ 10. 1080/ 02688867. 1996. 9727028 PLEASE SCROLL DOWN FOR ARTI CLE Taylor & Francis m akes every effort t o ensure t he accuracy of all t he inform at ion ( t he “ Cont ent ” ) cont ained in t he publicat ions on our plat form . However, Taylor & Francis, our agent s, and our licensors m ake no represent at ions or warrant ies what soever as t o t he accuracy, com plet eness, or suit abilit y for any purpose of t he Cont ent . Any opinions and views expressed in t his publicat ion are t he opinions and views of t he aut hors, and are not t he views of or endorsed by Taylor & Francis. The accuracy of t he Cont ent should not be relied upon and should be independent ly verified wit h prim ary sources of inform at ion. Taylor and Francis shall not be liable for any losses, act ions, claim s, proceedings, dem ands, cost s, expenses, dam ages, and ot her liabilit ies what soever or howsoever caused arising direct ly or indirect ly in connect ion wit h, in relat ion t o or arising out of t he use of t he Cont ent . This art icle m ay be used for research, t eaching, and privat e st udy purposes. Any subst ant ial or syst em at ic reproduct ion, redist ribut ion, reselling, loan, sub- licensing, syst em at ic supply, or dist ribut ion in any form t o anyone is expressly forbidden. Term s & Condit ions of access and use can be found at ht t p: / / www.t andfonline.com / page/ t erm s- and- condit ions zyxwvutsrqponm zyxwvutsrqp zyxwv P r q e d Apprarsal, volume 1 1 , number 2, June 1996, pages 133-139, Beech Tree Publishing, 10 Watford Close, Guildford, Surrey GUI 2EP, England Letters Projects with long time horizons: a rejoinder zyxwvut John Roberts Downloaded by [University of Glasgow] at 07:36 04 June 2014 zyx N THEIR ARTICLE on “Projects with long time horizons: their economic appraisal and the discount rate” (ProjectAppraisal, 10(2), June 1995, pages 66-76) Professor Livingstone and Dr Tribe argue that it is wrong to apply a standard discount rate to projects in all sectors because this could discriminate against, and eliminate from investment budgets, projects with long pay-off periods. Livingstone and Tribe (LT) are concerned that using test discount rates, which are typically in the range 8-15%, is inconsistent with a responsible use of investible resources in the interests of future generations, and sometimes also with sustainable development in the lifespan ofthe presentgeneration. Their particular concern is with projects with potential global environmental benefits, such as those for promoting the conservation of forests and for abating greenhouse gas emissions. They also mention that population projects might be rejected if decisions on them were taken on cost-benefit grounds using conventional discount rates, and that the technology chosen in infrastructure projects would be likely to lead to subsequent maintenance problems. Their recommendation is that resources should be allocated between sectors on “general grounds”, and that sectoral allocations should then be optimised using sectorally specific discount rates. By implication, though this is not made explicit, the authors had in mind the use of discounted cash flow in the allocation of public budgetary resources. They are not concerned about the (sometimes more important) task of influencing private-sector decisions. Unfortunately, their conclusion is neither very helpfd from the point of view of public policy, nor docs it provide the basis for a consistent and responsible approach to project choice. I Jolm Roberts is a Senior Economic Adviser in the Overseas Development Administration, 94 Victoria Street,London SWlE 5JL,UK, at present on secondment to the European Commission (EC). The views expressed in this paper are personal, and do not necessarily represent those of the ODA or the EC. Inter-sectoral resource allocation It is in practice often the case, as LT recommend, that budgetary resources are allocated bctween ministries or public service functions on “general grounds” (that is, on the basis of political compromise behveen competing interests). The results can be very unsatisfactory. One of the essential purposes of cost-benefit analysis in government is to apply uniform standards in project selection, and thus to provide a franiework for assessing inter-sectoral trade-offs. To fulfil this purpose CBA must use common shadow prices and decision criteria - including the discount rate. There are no sound reasons for claiming that this is not possible: the opportunity costs of resources and intertemporal trade-offs behvecn costs incurred or benefits realised at different times are the same, irrespective of sector. As a tool CBA may be blunt, not least when there is no ready monetary valuation of outputs but it is at least capable of highlighting the more prominent cases of intersectoral resource misallocation, and so mitigating the effects of capture of the budgctary process by special interest groups and lobbies. It is a tool which should not be discarded lightly. Choice of discount rate LT helpfidly summarise the distinction between the social time preference-cum-consumption rate of interest (SRTPKRI) and the social opportunity cost (SOC) rationales for the choice of discount rate, and show the reconciliation between the two through the shadow price of investment.‘ In essence, the largcr the difference between SOC and SRTP the higher the shadow price of savingshnvestment; and the higher the share of investment financing taken from savings (as opposed to consumption) the closer the rate of discount should be to SOC. This presentation neatly subsumes in a single framework the two alternative approaclics to zyxwvutsrqp Project Appraisal Jitrie 1996 0268-8867/96/020133-7 US$08.00 0 Beech Tree Publishing 1996 133 Letters zyxwvutsrqpo discounting proposed over 20 years ago by Little and Mirrlees (1968) on the one hand and the authors of the UNIDO Guidelines (1972) on the other. It also helps to confirm that the economics profession has added little to its analytical baggage on discounting in the intervening years. The rationale for the practice of the majority of aid donors who, at least notionally, use a discount rate based on a public sector SOC. stands out clearly from this presentation. It rests on the propositions that: Downloaded by [University of Glasgow] at 07:36 04 June 2014 0 is invalid over long time horizons. In particular, environmental assets which are becoming scarcer, or which are becoming more highly prized as personal incomes rise, can be expected to rise in price relatively to other goods and services.2 Projects to preserve or plant forests or to limit GHG (greenhouse gas) emissions should thus exhibit rising benefit streams. h s i n g output unit values may offset the negative impact of discounting at the SOC which Livingstone and Tribe decry in their paper. If public investment decisions are optimised, at appropriate shadow prices, using a uniform SOCbased discount rate, real incomes will be higher than under alternative prescriptions. This will lead not only to (probably) higher valuations of environmental assets but also to higher savings from which to finance remedial action - if this is needed for subsequent environmental protection. zyxwv zy zyxwvutsr finance for public-sector development projects is from a development budget, and thus from domestic or foreign savings, not from consumption, capital market imperfections and macroeconomic policy considerations limit the scope for switching savings between the public and private sectors, so the development budget is de fucto subject to a capital ration, the job of public-sector resource planners is to maximise the present value of net welfare benefits from public investment, subject to the capital ration, this means discounting at the rate of return of the marginal project which just exhausts the capital ration. Putting these principles rigorously into practice is not easy because capital rations are not fixed for all time, the boundary between public and private saving is permeable (and become more so as economic reforms bite) and the universe of potential public-sector projects is only partially known. Nevertheless, it is clear that an opportunity cost of capital rationale for the choice of a discount rate, whether conceived solely in terms of alternative public-sector options or in terms of returns achievable in the private sector, is the only realistic one for public investment planners. Using a consumption-based rate of discount would not be appropriate, at least not unless accompanied by ersatz capital rationing devices such as the investment premium or shadow price of investment proposed by the UNIDO authors, because it would over-commit available resources. Using an artificially low discount rate in public investment decisions may also have the same effect as offering the private sector a below-market interest rate. Technology choices can be biased towards excessive capital-intensity, to the detriment of employment creation, as the ranking of project alternatives is upset. Problem of long time horizons Are these arguments fora uniform, SOC-based public discount rate invalid for projects with long time horizons? The answer is ‘no’ for two main reasons: The common assumption of relative price stability made in analysis of medium time horizon projects 134 0 These propositions do not trivialise environmental concerns. They are, in fact, perfectly consistent with taking as a maximand in project appraisal an environmentally corrected concept of national income. Indeed, present and fiture environmental assets and liabilities, and the effects of projects thereon, should always be assessed, valued and incorporated into the analysis. If there is sufficient consensus that precautions should now be taken to avoid some present or future hazard of potentially major proportions this too can be built in, though it is hard to give pecuniary expression to the precautionary principle. The correct way of handling investments in the environment for future generations - or for that matter investments for the health and safety of contemporaries - is to make accurate and explicit statements about the physical effects which are sought and about the values which present and future generations may put upon them. If the effects are unquantifiable but major, the precautionary principle should be invoked, though with sensitivity analysis to establish the acceptable extent of precautions to be taken. Discretionary tinkering with discount rates is quite the wrong way of facing the longer-term future. It obfiscates the issues, and, if we are not carefil, it debases cost-benefit analysis by making it into a tool for giving spuriously scientific justification for decisions already taken using other criteria. This is certainly the conclusion reached by Pearce, Markandya and Barbier (1989, chapter 6) who write: zyxwv “There are compelling arguments why this [adjusting discount rates for environmental projects] should not be done: (i)calculating the appropriate rate is extremely difficult;. . . (ii) a selective lowering of the rate for environmental projects is inefficient and administratively cumbersome and difficult; (iii) there are alternative ways of dealing with z Project AppraisalJune 1996 zy zyxw Letters many of the environmental concerns that are probably more effective.” and: be chosen which offers the most appealing combination of expected NPV (net present value) and probability of success. Fiddling with the discount rate would be doubly harmful tothe decision-making process. It would both lead to inconsistency in project rankings and obscure the real issue of risk which concerns decision-takers. zyxwvutsr zyxwv zyxw “It is better to define the rights of future generations, to use these to circumscribe the overall cost-benefit result, leaving the choice of discount rate to current-generation oriented considerations.” Notes 1. Downloaded by [University of Glasgow] at 07:36 04 June 2014 Infrastructure projects The argument for rejecting discount rate variation as a tool for ensuring that infrastructure projects are sustainable, and that their viability is not undermined by poor maintenance, is rather different. The crux of it is the proper use of risk analysis in project appraisal. The experience of aid donors amply confirms LT’s observation that road maintenance, for example, is often rendered inadequate by institutional factors, so that maintenance-intensive techniques of road construction which theoretically offer the least lifetime costs turn out in practice not to be least cost. The problem is in fact so common that it has become instinctive for donors to identify inattention to maintenance as a major risk factor. The problem should be tackled from two angles. First, by applying the precepts of risk management in the course of project design, technical and institutional options for mitigating the risk of poor maintenance should be identified. Second, the options should be compared using cost-benefit analysis with risk analysis, showing both the expected value of project NPV, and the risks of project failure, under alternative configurations.That configuration should 2. The public sector SOC may be higher than the SRTP/CRI either because savingsare suboptimal (for reasonsof myopia or capital market imperfection) and/or because of the deadweight cost of mobilising public-sector resources. Public savings investment may therefore stand at a premiumwith respect to consumption in public resource allocation decisions. If returns from public investment are fully consumed the premium (4can be expressed as q/i where q is the SOC and ithe consumption rate of interest-cum-time preference rate. If some of the return on investment is saved, the premium is higher (UNIDO 1972. chapter 14). LT &ow. following Cline (1992), using a consumption numeraire,that the discount rate (3 should be the weighted sum of q and i, where the weights are shares of savings and consumption respectively from which project financing is drawn, using the formula: r=a*qw(l-a)’i Environmentalasset prices (such as those of timber in forests) would rise in the market place if property rights are properly assigned. If there is market failure appraisal economists should attempt, for instance by contingent valuation, to estimate the values which a well-functioning market would set, and, in addition, assign values to non-pecuniaryexternalities. References zyxwv W R Cline (1992), The Economics of Global Wem’ng (Institute of InternationalEconomics, Washington). I M D Little and J A Mirrlees (1968), Manual of Industrial Project Analysis in Developing Countries (OECD Development Centre, Paris). D Pearce, A Markandya and E Barbier (1989), Blueprint fora Green Economy (Earthscan, London). UNIDO (1972) , Guidelinesfor Project Evaluation (UN, New York). Projects with long time horizons: response Ian Livingstone and Michael Tribe vv ’E RESPOND HERE to Roberts’ comment (Roberts, 1996) on our paper which appeared in the June 1995 issue of Project Appraisal (Livingstone and Tribe, 1995). The main objective of our paper was to draw attention to quite a number of different categories of investment choice in which existence of long time horizons raised speProfessor Ian Livingstone is at the School of Development StudUK and Dr ies, University of East Anglia, Norwich Mi4 TJ, Michael Tribe is Lecturer at the Development and Project Planning Centre, University of Bradford, Bradford ED7 I DP, UK. Project Appraisal June 1996 cial questions regarding the ready applicability of conventional ‘official’ project appraisal methods based on discounting.We did not offer strong conclusions regarding the best way of dealing with these different problems, though we identified various possibilities. Roberts does not deal closely with the specific cases that we identified, and essentially reproduces the orthodox case, already described in our paper, following the exposition of Birdsall and Steer (1993). There is now a sufficiently large body of economists who are uncomfortable with the results generated by 135 Letters zyxwvutsrqpo this orthodoxy that there is clearly a need to reconsider the relationship between the economic principles which it espouses and the empirical reality within which it seeks to give practical guidance for investment decisions. In addition to identifying the different categories or examples of problems, our paper also compared the two different bases for calculating the ‘shadow’ discount rate which are given in the literature, the social opportunity cost (SOC) and the social rate of time preference (SRTP), since these usually give significantly different discount rates, typically in the order of 8-15% for the former and 2-3% for the latter,’ in turn producing different economic preferences in the case of projects with long time horizons. Our comments were not particularly innovative, although we attempted to make some of the implications clearer than they are in much of the literature. Roberts’ basic argument is that the conventional ‘official’ orthodoxy, as explained by himself and by Birdsall and Steer, can adequately handle all conceivable economic characteristics of projects in such a way that rational and consistent investmenthesource allocation is possible on this basis alone. This ‘of€icial’ view is based on the use of an SOCderived shadow discount rate of 8-15% per m u m , the implication being that the ‘marginal project’, using funds which would otherwise have been employed on alternative investment, has an ex ante economic internal rate of return (E1RR)somewhere in the range 8- 15%. No verifiable evidence has been presented in support of this view? Roberts asserts that: recourse to the calculation of the EIRR for the programme as a whole, although individual projects within the programme may be subjected to systematic investment criteria (CBA or cost-effectiveness).Reference may be made, lastly, to the adoption of ‘process project’ methodology, explicitly recognising that regular economic criteria for resource allocation may not be appropriate in some areas of official aid disbursement.3 Downloaded by [University of Glasgow] at 07:36 04 June 2014 zyxwvuts zyxw zyxwvutsr Conventional theory With respect to Roberts’ central contention, that conventional theory can handle the particular problem cases which we enumerated within the same basic framework, the following may be a useful way forward. Concern is not generally expressed with the logic of the conventional theory applied over limited time periods. The problem cases are all associated with extended time horizons. If, therefore, some assumptions ofthe model do not hold over such periods, where precisely do these assumptions break down? One possibility, relaxing the assumption of perfect forward foresight,4 is that ‘myopia’ exists on the part of government in relation to public-sector investment. This myopia might spring from concern for the immediately tangible, compared with more distant, and therefore less certain and clearly identifiable, benefits, or from deliberately favouring current over future generations (who have, as yet, no vote). This would specifically operate against projects with long time horizons. The same considerations would affect investments which protect against possible future catastrophes, the occurrence of which at any future time is by nature uncertain and not seen as immediate. More general environmental effects will also share these characteristics. We cited investment in research as another area in which returns may accrue over long periods, such delays also being associated with uncertainty of the returns. A recent paper from a World Bank source (Anderson and Purcell, 1996, page 4) notes that “the high returns to agricultural research [in less developed countries] are symptomatic of persistent under-investment in this form of public spending”. We can gain a feel for possible undervaluation of future environmental effects under conventional in-vestmentcriteria if, instead of assuming that a catastrophe occurs in year 20, and calculating its present value, we supposes that it occurs in year 1, and that the cost is compounded to year 20 using the conventional 8-15% annual rate of return on initial value: compounding to an amount up to 16 times as great, so that, in the case of the higher rate, one ‘Chernobyl’ in year 1 is equivalent to 16 Chernobyls in year 20. A second assumption ofconventional theory is that interdependenciesand linkages are limited and definable: They can be catered for by making limited allowances for ‘externalities’. As we indicated previously, development strategies are, in contrast, zyxwvu “to fulfil this purpose [of efficient resource allocation] CBA [cost benefit analysis] must use common shadow prices and decision criteria including the discount rate. There are no sound reasons for claiming that this is not possible: the opportunity costs of resources and intertemporal trade-offs between costs incurred or benefits realised at different times are the same, irrespective of sector.” This is not entirely accurate. Investment decisions within official aid programmes are not based solely on economic criteria - ‘multi-criteria analysis’ (MCA) is effectively in operational use to the extent that judgemental factors supplement pure economic analysis. Also, CBA has never been the sole economic investment criterion used within aid programmes: costeffectiveness analysis is used to complement CBA in those sectors where ‘revenues’ are difficult or impossible to estimate. While in principle costeffectiveness analysis uses the same test (shadow) discount rate as CBA, the marginal EIRR cannot be calculated within a cost-effectiveness framework (that is, there is no ‘marginal efficiency of capital’ in these cases). Thirdly, programme aid is allocated without 136 zy Project Appraisal June 1996 Downloaded by [University of Glasgow] at 07:36 04 June 2014 Letters recommended on the basis of iust such interdependencies and linkages. Infrastru&al investment‘s will be based on similar judgements. Interdependencies can also lead to divergent processes. Another of our earlier observations was that undesirable income distribution effects can be compounded from one generation to another, with income differentials between sections of the community or parts of a country being divergent over time in the absence of corrective social or regional policies. Public investment criteria might be modified to take account of income distribution through weighting (as expounded, for example, in Squire and van der Tak, 1975, chapter 7) in the initial period or periods, but the effect ofweighting beyond that will be completely ‘dissipated’ by discounting. In their “Author’s preface to the sixth printing” Squire and van der Tak (1992, page v)themselves explain that the methods outlined in their book with respect to distribution weighting were hardly used at all in any investment analysis in the 13 years after publication. Inter-generational conflicts of interest obviously involve long time horizons. While the utility of the second generation is discounted relative to that of the first (Kula, 1992, pages 240-263), conventional theory would argue that the income and welfare of the second will be maximised by selecting investments with the highest rates of return now. If income per head is higher for the second generation this is itself a reason for discounting. While this may hold for investment in the economy as a whole, it will not apply to specific situations, particularly involving the environment, in which effects on future generations are underestimated or undervalued for reasons already given.s zy Remedial action may not be possible retroactively: many environmental impacts are irreversible, and anticipatory action is needed; if forests are to be in existence in 30 years they may need to be planted now market itself will perform the necessary adjustment and we need only make accurate forecasts of future prices. The problem in practice is again myopia: lack of full anticipation and of allowance for longdelayed future events which require anticipatory action. This myopia perhaps amounts to a special form of market failure associated with long project horizons. Roberts himself does not appear to give much weight to future hazards, saying only that “[ilf there is sufficient consensus that precautions should now be taken to avoid some present or future hazard of potentially major proportions this too can be built in, though it is hard to give pecuniary expression to the precautionary principle”. Roberts refers to the possibility of incorporating the valuation of environmental (and other) extemalities within the conventional framework using a range of techniques which have been developed especially in the last 10 to 15 years, of which contingent valuation is only one. Unfortunately these techniques are not at present used sufficiently universally or consistently that we can justify the level of confidence exhibited by Roberts, and use of them is not likely to develop substantially in the near future (the situation may be similar to that described by Squire and van der Tak for income distribution weights). zyxwvu zyxwvutsr zy zyxwvuts Remedial action A further point made in our paper was that remedial action may not be possible retroactively.For example, many environmental impacts are irreversible, and anticipatory action is needed; if forests are to be in existence in 30 years they may need to be planted now. Similarly, and more generally, if a country is to be at acertain point on an identified development path in 10 or 20 years, specific policies and programmes will need to be in put in place now. In the former case Roberts’ comment is that: “. . .environmental assets which are becoming scarcer, or which are becoming more highly prized as personal incomes rise, can be expected to rise in price relatively to other goods and services . . . ksing output unit values may offset the negative impact of discounting at the SOC which Livingstone and Tribe decry in their paper.” Local investment Roberts makes no reference to what we consider to be an important question raised in our paper: that of locally available investment opportunities.The standard theory assumes the local availability of indefinite investment opportunities at the conventional 8-1 5% level of the shadow discount rate. It has sometimes been argued that even ifthese are not available locally the opportunity cost of capital should be calculated assuming capital is internationally mobile, making the location factor irrelevant. In fact the objective of most LDCs must be to establish economic activity, employment opportunities and a cumulative development process domestically - and perhaps differentially domestically to incorporate backward regions. The question of which SOC value should be selected to achieve development objectives becomes an issue: national or international? national or local? We also emphasised that zyxwvutsrqpo This argument suggests that it will not even be necessary to assign values to externalitiesin such cases: the Project Appraisal Jiorc. 1996 137 Letters zyxwvutsrqpo discounting across long time horizons, together with an exclusive focus on the analysis of individual separate projects, marginalises development strategies, linkages and development sequences over time -the long term factors in economic development. Apart from the location issue with reference to the choice of SOC and thus of the discount rate, there is the assumption that the current SOC/rate of return on investment can be extrapolated over extended periods oftime. If we do not feel confident enough in an LDC situation to project indefinitely an 8-15% rate of return and feel it wise to err on the side of caution,this may be done by using a lower rate of discount for very long-run projects. zyxwvu zyxw zyxwvutsr zyxwvuts Tackling problem areas Downloaded by [University of Glasgow] at 07:36 04 June 2014 long-term impacts of different kinds and categories of projects in different country situations to provide a sounderbasis for the ex ante appraisal of projects with long time horizons. For the moment we remain uneasy about results which are generated by the routine application of a standard rate of discount in the range of 8-15% per annum for the economic analysis of projects, without regard to the long-term economic consequences. Given the various investment areas affected by long time horizons (infrastructure, research, distributional, strategical, environmental and so on) the question then arises as to the best way of making allowance for these considerations. In all the problem areas mentioned the values or weights which might be assigned are subject to a great deal of uncertainty. There is likely to be a judgemental element, therefore, in any method which takes them into account. A policy of accepting projects solely on the basis of current project worth calculated using the 'official orthodoxy', and assuming that this approach will put us in the best position developmentally and environmentally at some point well into the future, itself involves a judgement regarding the (un)importance of interdependencies and sequences over time, since it is most unlikely that in practice these can be suf€iciently quantified and incorporated. The danger here lies in the rigid application of formulae which give the appearance of greater precision and meaning than the underlying project circumstances can reliably support. Allowing a lower rate of discount on particular categories of project will certainly be arbitrary unless the reduction has a quite specific transparent basis for the judgement involved. This applies to the application of a lower rate to particular categories of projects, the research category for example, as well as to individual long-term projects. Any judgement will need to be made on the basis of the best research, evidence and experience available.6 Special allowance might be made to projects or programmes on grounds of anticipated linkages and long-run development effects benefiting particular areas or sections ofthe community, orto projects with favourable but distant environmental effects. This allowance could be made directly by requiring alower EIRR for a particular category of project (or accepting this for an individual project) or by a reduction in the discount rate, which is equivalent to requiring a lower EIRR but with an effect which is a direct function of the time horizon. More research is requircd on the short- and 138 Notes 1. The rangesof 8-1 5% for the SOC and 2-3% for the SRTP are identified in Livingstone and Tribe (1995). Roberts takes the range of 8-15% as indicative for the SOC value of the sociaVshadow discount rate, but MacArthur/ODA (1988, page 48)takes values of 8-12%. 2. We should remind ourselves that the implication of taking a value of SOC in the range of 8-1 5% per annum is that there is a comparatively abundant supply of local investment opportunities in specific LDCs which have ex ante economic internal rates of return in excess of these values. This does not appear to be the situation in a large proportion of LDCs. The SOC is the EIRR of the marginal project. Potts (1994, page 174) presents evidence for an historically high value of the longterm international real interest rate for LDCs at between 7 and 8% per annum for the period 1980-90. An SOC based on a marginal return to investment in the range of 8-15% (that is the minimum return to capital invested is around 8%) would imply much higher rates of economic growth in LDCs than have been observed in practice, in turn implying that such estimates of the SOC are unrealistically high (we are grateful to David Potts for this point). 3. This discussion should not obscure the fact that we are concerned with the economic appraisal of projects in all parts of the economy, including the private sector (where government decisions on taxes, subsidies and other elements of policy depend on a view of the impact of projects on the economy as a whole or on a particular region). The concerns of our paper therefore transcended the allocation of aid budgets in particular, or of public-sector resources in general. 4. The fact that perfect foresight does not exist in realQ is recognised in project analysis by the adoption of sensitivity analysis as a means of handling the phenomenon of uncertainty. This issue is conceptually, and practically, quite distinct from the concept of time preference which is embodied in the discount rate. 5. Both Roberts (1996) and Birdsall and Steer (1993, page 6) express concern that lower-than-orthodox discount rates would unduly bias investment decisions in favour of capital-intensive projects. This would not be the case if lower discount rates (reflecting time preference) were used in conjunction (correctly) with a savings premium (reflecting the scarcity value of investment funds). 6. Such an adjustment has been used consistently by the United Kingdom Government over a long time period for the forestry sector leading to discount rates of between 3 and 5% per annum (Kula, 1992, chapter 3, pages 68-69). Bibliography J R Anderson and D Purcell (1996), "Institutional issues in getting agricultural RBD back to work: lessons from World Bankoperations", paper presented to the ESRC Development Economics Study Group Annual Conference, University of Leicester, March. N Birdsall and A Steer (1993), "Act now on global warning - but don't cook the books, Finance and Development, 30(1), March, pages 6-8. E Kula (1992), Economics of Natural Resources and the Environment (Chapman and Hall, London). I Livingstone and M A Tribe (1995), "Projects with long time horizons: their economic appraisal and the discount rate", Project Appraisal, 1O(2). June, pages 66-76. zyxwvu zy Project Appraisal .June 1996 z zyxwvutsrqp Letters Project Appraisal, 11(2), June, pages Oo(M00. L Squire and H van der Tak (1975), Economic Analysis of Projects (Johns Hopkins Universlty Press for the World Bank, Baltimore). L Squire and H van der Tak (1992), Economic Analysisof Projects (Johns Hopkins University Press for the World Bank, Baltimore, 7th printing). zyxwv Downloaded by [University of Glasgow] at 07:36 04 June 2014 J D MacArthudODA (Overseas Development Administration) (1988),Appraisal of Projects in Developing Countries: A Guide for Economists (HMSO, London). D Potts (1994), "Real interest rates and the pattern of agricuttural investment", in N Maddock and F A Wlson (editors), Project Design for Agricultural Development (Avebury, Aldershot). J Roberts (1996), "Projects with long time horizons: a rejoinder, zyxwvutsrqpo Project Appraisal Jioie 1996 139