Pollcy, Research, and External Affairs WORKING PAPERS Macroeconomic Adjustment and Growth Country Economics Department The World Bank March 1991 WPS 606 Adjustment Policies and Investment Performance in Developing Countries Theory, Country Experiences, and Policy Implications Luis Serven and Andr6s Solimano Adjustment without growth has been, for many developing countries, the outcome of the debt crises of the 1980s. Macro- economic stability, policy credibility, and adequate external financing are among the key ingredients for achieving a strong ;i-vestment response to adjustment measures. The Policy, Resurch. and Extemal Affairs Complex distibutes PRE Working Papes to diuaminate the findings of wodk in progress and to encourage the exchange of ideas among Bank staff and aU others interested in development issues. These papers carry the names of the authors, reflect only their views, and should be used and cited accordingly. The findings, interpretationa, and conclusios are the auLhors' own, They should not be attributed to the World Bank, its Board of Dirctors, its managanent, or any of its member cotintries. Policy, Re"arch, and External Affalrs Macroeconomic Adjustment and Growth WPS 606 This paper - a product of the Macrocconomic Adjustment and Growth Division, Country Economics Department - is part of a larger effort in PRE to investigate the response of private investment to macroeconomic adjustment measures. Copies are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. P;ease contact Emily Khine, room NI 1-061, extension 39361 (63 p3ges). Serven and Solimano analyze the response of tent or suspected to be only temporary - in private and public investment to extemal shocks, which case investors prefer to wait and see macroeconomic adjustment, and structural before committing resources to irreversible fixed reform in three sets of countries: (1) countries investments. that pursued structural reform and liberalization in Latin America in the 1970s (Chile) or the * The sequence of adjustment measures is 1980s (Mexico and Bolivia); (2) countrit that thus important. Trade liberalization measures experienced severe macroeconomic instability undertaken while macroeconomic instability and did not pursue macroeconomic reform persists are likely to be viewed as purely transi- (Argentina and Brazil); and (3) East Asian tory, and thus might actually distort the invest- countries with high-growth, outward-oriented, ment pattem. state-active economies that adjusted to the shocks of the 1980s and maintained high growth, * Even well-designed, consistent adjustment low inflation, and remarkable macroeconomic programs might have to overcome a lack of stability (Korea, Singapore, and Thailand). credibility, especially in their early stages. If enough external resources are available, the Drawing on the literature and their econo- private sector may be more confident about the metric analysis of the determinants of private viability of adjustment efforts, which could investment in developing countries using cross- facilitate investment recovery. country data for 1972-87, Serven and Solimano conclude (among other things) that: * Even if policy changes are perceived as permanent, inadequate infrastructure may pose a * Macroeconomic stability and policy cred- significant obstacle to the recovery of private ibility are essential for achieving a strong investment. The implementation of well- investment response. Investment is likely to be targeted public investments in infrastructure limited under great macroeconomic uncertainty projects can stimulate the private sector's or if policy measures are perceived as inconsis- response to adjustment measures. i Thc PRE Working Paper Series disseminates thc findings of work under way in the Bank's Policy, Rescarch, and Extemal Affairs Complcx. An objective of the serics is to ge these findings out quickly, cven ifpresentations are less than fully polished. The findings, interpretations, and conclusions in thcsc papers do not necessarily represcnt official Bank policy. Produced by the PRE Dissemination Centcr Adjustment Policies and Investment Performance in LDCs: Theory, Country Experiences, and Policy Implications by Luis serven and Andres Solimano* Table of Contents 1. Introduction 1 2. Investment in Developing Countries, 1970-88 3 2.1 The overall picture 3 2.2 Private investment and macroeconomic adjustment: 7 some country stories 3. Macroeconomic Policies and Private Investment: Theory 29 and Emp;rical Evidence 3.1 Demand management policies and investment 29 3.2 Exchange rate policy and private investment 33 3.3 Trade liberalization and investment 38 4. The Incentive Structure and Investment Response: 40 Credibility, Uncertainty, and Irreversibility 4.1 Irreversibility, uncertainty, and investment 41 4.2 The role of credibility 44 5. Econometric Analysis 48 6. Conclusions and Policy Implications 54 References 59 * We thank Jaime de Melo for kindly providing his data base on investment. We are indebted to Max Corden for helpful conversations and to Dani Rodrik for detailed comments and suggestions. Bela Balassa and Felipe Larrain also contributed with comments to an ea.iier version. Walter Novaes provided able research assistance. 1. Introduction Adjustment wlthout growth has ben, for many developing countries, the outcome of the debt crisis of the eighties. The adaptation to the reduced availability of external financing has not led to a signifLcant increase in domeotic savings, but to a reduction in private and public investment rates. WLthout a sufflelnt recovery of investment, a sustained resumption of growth is unlikely. In such conditLons, the attempts at structural reform of many countrles may be endangered, an in the absence of an Lnvestment response their intended efficiency gains cannot materialies, and thus the only visible result of the reforms is their adverse short-run social and distributivo cost. The decline in external financing is not the only factor behind the invostment slowdown. In many cases, the fiscal adjustment required to reduce the external imbalance or to bring down inflation has taken the form of a reduction in public investment. Also, the increased macroeconomic instability associated with the external shocks of the eighties has made the economic environment more uncortain, and hence more adverse, for investment decisions. One Important source of uncertainty has been the external debt overhang, especially in highly indebted countries, which may also have contributed to discourage investment through its implicit tax' effect, as part of the future returns on lnvestment must be collected by the creditors in the form of debt repayment. In goneral, the macroeconomic adjustment and reform efforts of most countries have not been rewarded with an adequato response of private lnvestment. Yven when substantLal progress has boen made in the correction of macroeconomic imbalances and in the restoration of profitability -- often through drastic ck e. in real wages -- the impact on private investment has bcen very weak and slow In this paper we investigate the contrlbution of these factors to explaininq the recert Investment performance In developing countries. The paper is organized as follows. First, in Section 2 we present the empirical record of investment in LDCa in the 1970m and '980a. The response of private and public Investment to external shocks, macroeconomic adjustment and structural reform is analyzed by comparing three sets if countries. The first group includes countries that have pursued structural reform and liberalization in Latin Amerlca either in the 1970. (Chile) or in the 1980a (Mexico, Bolivia). The second group is composed by Argentina and Brazil, that in the 1980s have experlenced severe macroeconomic instabllitv. Moreover, these countries have not attemptod the kind of structural reforms pursued by the first group. The third group consists of three 'success stories, in East Asia: Korea, Singapore and Thailand. These high-growth, outward oriented, state-active economies were able to adjust to the adverse external shocks of the 1980. while keeping a record of high growth, low inflation and, in general, a remarkable degree of macroeconomic stabillty. In Section 3 we discuss the llterature on macroeconomic policies and prlvate investment, examining the effect of monetary, fiscal and exchange rate policy on private investment, and emphasizing some economic and institutional features specific to LDC (e.g., the degree of intervention in financial markets, the posslble complementarities between public and private Lnvestment, the high reliance on imported capital goods) that may affect the transmission mechanisms through which standard macropolicy measures lnfluence private investment. In the -3 - fourth section we examine in more depth the recent literature on credibility, uncertainty and Irreversibility in Investment decislons. We discuss how such factors can contribute to determine the Investment response to a glven set of economlc Incentives, which is the key lemone in the transltion from stabilization and reform to sustalnable growth. In Section 5 we present an econometric analysis of the dotermlnants of prlvate investment in developing countrlie usLng croa-country data for the period 1972-1987 for a selected group of LDCs. Finally, Section 6 presents some concluding remarks. 2. Investment in Develovina Countries, 1970-1988 2.1 - The overall picture Between 1970 and 1988, investment rates in developing countrles exhibit two dlstlnct patterns, with 1982 the point of demarcation (figure 1). lor seventy-eight developing countries, the average share of investment In GDP (in constant prices) increased from about 22 percent In 1970 to 25 percent In 1981, and for most of this period investment rates were historically high. With the rise in International real interest rates In 1981 and the onset of the debt crisis in 1982, the rate of Investment fell sharply. Investment started to fall earlier for the hlghly Indebted countries than for other developing countries, and the declLne was also larger (see table 1). For all groups of developing countries, the decline In Investment was accompanled by a slowdown in growth (tables 1 and 2). The fall in investment has been so severe that some countries may not even be fully replacing depreciating capital. For example, in Africa the minimum investment neoded to replace depreciated capital is estlmated at 13 - 4 - Figure 1 Share of investment in GOP for developing counies (unweighted averages) 0 3 021 0.26 " -0RX _>_L~~~~~~, 0.22 4/ % 0.18 '%% _-._N" * \ ',P% -p 0.16 v.14- 1970 1971 1972 1973 1974 167' 1976 1977 1M71979 10? 0 1911 1663 16 1664 196 196 1967 19U . Middle I (ao@-HIC) - AU Cmoi" -- - -ICa .. ...... Low lIcome Table I Investmient, saving, and powth in developing countries, 1970-8 IndicAw' C17p I 0 19IJ142 19834J4 1985s Gras domestic invutm tP AUl 2. 24.0 20.2 19.6 (psreentap of GDP HiiLy Lndebted 22.8 23.0 15O 18.4 at cuzelt price) Mddle Ulcoume 25.5 28.6 24.4 21.9 Low ucotoe 19.7 20.3 17.0 17.4 Croa doametic svtig All 16.1 13.7 13.9 14.9 (perentagp of CDP Highly indebted 20.3 20l 19. 20.2 at cuwt prim) Middlo inewtns 163 17.3 17.7 17.J Lowtinaoaw 12.S 7.6 8.0 9.9 Rmoue balanIce deicit AU 6.4 tU.3 6.2 4.6 (peNst of COP Hklhy indebted 2 2.9 .1.7 .1.S at crnwt pricm) Middlo irtcoue 72 11.1 6.7 4.0 Low inmote 72 12.7 .9 7.3 Crow dotatmc Lnvustmnt AL 23.4 2t 20.6 19.6 (p.renutag of CDP Hihly 0debted 21.1 U3 17.1 16.8 Alt cotitAntpficls) Middle incomi 25.7 28.6 24.9 2U1 Low income 21.5 207 17.6 16.0 Rate of owth of rl GDP AlU 4.7 2.7 1. 3.3 (peurcatage per yer) Highly indebted 5.0 .0.3 .0.4 2.7 Middle incoms 6.1 43 3.9 3.2 Low ineoe 33 2L 0. 3 3 Table 2 Growth and investment Ral OP VrovJ Invstment rato 1965-88 1980-88 1965-88 1980-88 Sub-Saharan Africa 3.3 05 17.6 15.9 Asia 6.3 7.4 27.7 31.1 Europe/Middle East/ N. Africa 4.6 2.8 28.4 27.3 Latn America and Caribbean 4.5 1.6 19.7 17.9 Source: Short-gm Owitlook, Tternational Monetary Fund, 1989, Table 15. percent of GDP, and sevdn countries in sub-Saharan Africa had investment rates below that level In 1987. Similarly, the minlmum investment rate to replace capital in Latin America is estimated at 14 percent, and three countries were below that level in 1987.1 Investmont declined both because of the reduced availability of financing and lower domand for investment. There were important changes in the resource balance deficit (dofined as the dLfforence betwoon domestic investment and domestic savings) following the debt crisis in 1982 (table 1). The decline in the resourco balance deficit (becauso of lower extornal financing) was not matchod by a sufficient increase in domestic savLngs, and so the deficit was almost entirely reflected in reduced investment. Investment demand declined for several reasons. Puklic investment contracted because of the deterioration in fiscal conditions as a result of the cut Jn foreign lending and the lack of adjustment In other fiscal expenditures, the rise in international and domestic interest rates, and accelerating inflationi and also because in some cases it was unsustainably high and of dubious productivity. Private investment was discouraged by the slower or negative growth and by the increase in macroeconomic instability associated with the adverse external shocks, the uncertainty about the now cortiguration of relative prices and incentives, and the inability of governments to stabilize the economy. In additlon, the debt overhang may have discouraged investment both through the uncertainty it created and through its Implied 4tax" on 1W. Easterly, "Fiscal Adjustment and Deficit Financing during the Debt Crisis,' in Dealing with the Debt Critss,* edited by I. Husain and I. Divan. (Washington, D. C.s the World Bank, 1989). - 7 - future output ard the accompanying credit rationing in International capltal markets. Analysis of a set of twenty-nine countries shows that the share of private investment in GDP (in current prices) was ralatively stable until 1980 and then declined, followed by a modest recovery after 1985 (figures 2 and 3).2 The decline was larger in the highly lndebted centries than Ln the other countrles. Public lnvestment as a share of GDP and of total lnvestment rose until 1980 and thon foll after 1982, two years later than prlvate investment (table 3). Unlike private lnvestment, public investment rates declined steadlly untll 1988. 2.2 Private Investment and MacrooconomLc Adiuetment: $ome Country Storles in this s*ctlon we organlzo the dlicussion of the behavLor of private investment during the course of adjustment azound three groups of countries in Latin Amorican and Ln Sast Asla. Tho first group le composed by Chlle, Mexico and Bolivia. These three share the adoption of docisLve stabilization polLcies orlented to eliminate basLc macroeconomlc lmbalances together wlth policies of structural reform orlented to liberalLze foreLgn trade, and to derogulate crodit and labor markets along free-market lines. The second qroup we consider li constitutod by ArgentLna aV. Brazil, two countries that in tho 1980e have been unable to stabilize the economy and correct, ln a sustainable 2The breakdown if investment into private and public components draws on G. Pfeffermann and A. Madarassy, 'Trends ln PrLvate Investment in Thlrty Developlng Countrie3," IPC Working Paper No. 6, 1989. They calculated private investment by subtracting from the national accounts data the investment of the consolidated public sector. The latter was obtained from World Bank reports and government sources. - 8 - Fipwe I public wd private investment for 29 Couztries (uAweighttd avenge, pecent of CDP) 31.0~~~~~~dd 2.0 11.0*. 10.0 **., 0 '^q^ __, 1573 101 left I0 ¶4 1 1 " l $ 100 0 17 1 136 1311 11130 1s36 1148110 t" PIpu a 3 NbUc Sa prIvate IatsWt ft U3 laCa Cuaweia N Deo*.A pmmat of X0?) 1s.a.~~~~ b 4B - a- 10.0 . 1073ar101 101 10 137 101 '0 107 § so1 101 17 1333 1 1 _6 36 I 13636136 3 _ C 0s_ t . *., ., P,4wa 0iEei lawmi ............U... . ... . . Igqveue-* -9- Table *3 Public and private Investment for a group of 29 developing countries, 1970-88 (percentage of GDP at current prices) CGoup 1970-80 1981-82 19834 198548 29 countries Tota 20.3 22.2 18.8 17.6 Private 12.2 11.7 9.7 9.6 Public 8.2 10.5 9.0 8.0 13 highly indebted countues Total 20.1 20.2 15.1 15.2 Private 12.3 10.9 8.1 8.7 Pblic 7.8 9.2 7.0 6.5 ampl.: AzgendNza, Bargladesh, Bouvia', Bazil, Chile', Colombia', Cost Rica', Ecutador, Cuatemals, Huzswy, India, Ind onesa, Kenya, Korea, Mulaysla, Ma.dco", Nigera', Patan, Paul, P.dIppnes', Portugal. Si uLanka, Thailand, Tursa, Turkey, YUNgUaY, Vez~ua1', Za . Zbabwe. * Hghly Indebted Cowets .a Source: "Adjustment Lending Policies for Sustainable Growth," Policy and Research Series #14, The World Bank. - 10 - way, baslc macroeconomic imbalances, and that have not attempted comprehenalve structural reforms and liberalization of the type adopted by the countries in the flrst group. The thire ;roup consists of three success stories in East Asla and includes Korea, Singapore and Thailand, namely economies that have managed to sort out the external shocks and the debt crisis of the elghties without sacrlficing high growth and domestic macrooconomic stabillty. Adjusting cum Liberallzing Countries in Latin Americas Chile, Mexico and ollvia This group of three Latin American countries share several coimon features regarding macroeconomic policies and structural reforms. At the level of macroeconomic policy, they implemented either in the seventies (Chile) or In the middle and late eighties (Bolivia and Mexico) restrictivo fiscal and monetary policies oriented to reduce high inflation rates and unsustainable current account deficitse The three of them used fiscal adjustment (with better results in terms of permanent deficit correction in Chile and Mexico) as a centerpiece of the stabilization effort. The comprehensive use of incomes policies for stabilization purposes was present just in the Mexican "Pacto de Solidaridad Zcon6mica" of late 1987 though in the cases of Chile and Bolivia some form of exchange rate stabilization and/or 3Bolivia's inflation between 1984-85 was a case of hyperinflation rather than high-chronic inflation. - 11 - wage controls were used to help dislnflatlon at difforent time. during the course of stabillzation.4 On the front of structural reforms the three countrles implomented (to a dlfferent extent) trade liberalization, financial deregulation, privatizatlon and labor market flexibilization. The degree, timing and results of these policy reforms variod in each country, though there was a general free-market orientation in the throe cases. Some of tho policies such as trade liberalization and labor market flexibilization (coupled with wage controls) were used as anti-Inflationary devices in additlon to their intended nature of long run transformations required to improve economic ffliciency and speed-up economic growth.5 A common feature of the reforms -- particularly in Chile and Mexico--is that they were implemented by strong governments whose roputatlon in avoiding pollcy reversals was at stake.6 Following the swings of the world economy ln the 1980's those three countrles suffered the cycle of over-borrowlng, the sharp cutoff of foreign lending and the onset of the debt crisis. In Chile the bulk of the external debt was originally contractod by the private sector, while in Mexico and Bolivia the public sector was the actor that borrowed abroad most heavlly. 4For a reference on the Chilean experience with stabillzation ln the last two decades see Corbo and Solimano (1990). The Bolivia story with stabilization and reform is told in Morales (1990). For a comparative analysis of stabilization experiences in Latin America and in other regions, see Solimano, (1990). 5Rodrik (1990) calls attention, however, to the weak links between trade liberalization and growth both at an analytLcal and empLrical level. 6 The crisis of 1982-83 put under heavy stress some of these policies in Chile. Some reversals took place, such as increases in tariffs and direct intervention of the financial system. However, as the crisis receded, tari"s were lowered again and the financial system gradually deregulated. - 12 - Given this background, what have been the salient featureo in tho behavior of private and public investment ln theme three economies? As table 4 shows, total Lnvestment ln the period 1985-88 declined by 4 to S points of GDP with respect to the pro-crials perlod 1976-81. Public investment declined by almost 5 points of GDP both in Mexico and 8olivla over the saoe period. In contrast, in Chilo the lovel of public investment in 1985-88 was hLgher than before the debt crisis. Private investment rates are still below their pro- crisis level both in Bolivia and Chile, though the data shows a recovery ln private lnvestment ln Chile and Mexico towards the late 1980s. In Bolivia, however, no upsurge of prlvate investment has taken place in the aftermath of stabilization. what accounts for this performance of investment? What is the role played by the foreign debt crilss that hit these throe countries ln the behavior of investment? Was a decline in investment the toll pald for correcting the macro dilequillbria in those economies? What difference do the structural reforms and a more stable macro environment make for a quicker response of private Lnvestment? These are certainly difficult questions, though some hypotheses may be advanced. rirst, it is clear that the pattern of investment followed the "debt cyclo." Public Lnvestment in Mexico and prLvate investment in Chile increased sharply during the borrowing-led boom of the late seventies and early eightles. in 1982, when the access to extornal lending was abruptly cut off and the countries were forced to a rapid reduction in the current account deficit through tight demand policies along wlth real devaluation, investment fell sharply. Thus the adjustment was carried out basically through cutting - 13 - CHILE FICGU 4 MEXICO 0.3 v., gas ,@r. ~ ~ ~ lo lv Iostn1* " s1 " sn Ion It Iw *ni * N" in IN* YEARS YE _ ~~~~~~~~~~~~~~~~~~~~~. . . . . . ....... ... _ ................. ... BOLVIA 0._ 0U 0 9I I I I I I I It I S I II ! YEARS KS*P KGO . 4n Kr tOALO AWvP - 14 - TABLE 4 NAUC r NIWLIIa, IA4lsrUf Sw %&amms lelur Awk*" Avsqa4AI( Cot"i SU~~~~44r4 '.*e r'rrU,. e. , ............................................................. vo,.~~ n4sv~~sf IO 33 I 57 i .11 5 ,. . . . . .. . . . . .. . .. . . . . . .. . . .. .. . . . .. I . .. . .. . . .. . .. . . .. . . .. . . .. .sauC .vIsr of I o s n to P4|XX.alg _" 'mgl I' Zi lU g 5s 35 28 toh 29 UstUwm . 1.43 No t W 7tl P'l--8 w 'IM ,12 7r u1' 1g511 0AV3 CLaillar iCC WANCIIa 04.45 6,5* U IS 3D so$ O"SIcS loot sea' U.ac III,w It 30 295 I s. *ws. isa ws. S ) 01,1q 43c5 "Fe lra. r4 IGet I mr* *r| ~ ~ ~ ew AWII daisr (-lhs.I 97581 1900 9 .................................................................... 10as. ,E4dIVUI * -. 335w U A 9 3. larur reaI we *- rSrl. IWA"Nof, we rmsoomm loi" 4rl ..................................................................... 141AAs.LC 141UUf ISIefi i.u0 ...............................................I...................... Wall ItirUl 53.9 19 "I of so I fla t s o 3555 8.0 t4.54 e.s (Si CC" 2411.011) CAt I CC 6AAs.M '.0. 3. 11 M 3D 395 DaR1tGI eggs 19. 0 Is. 75 6s.7 ,l . ap go INIAL, sea. gIs 603.51 116.5 t4?.d 1USl l.t ag *4O55g .. WA 0~ e, A. (1U,1 Glow *w,mmm r o "wig *.Il q PY I he metAt IFt.NO an M O M C Xc MNOW AdUISIS41 ~ IWIA : A I Ar f t$ ig -l I4 % t r r- wdt114 i5975m 1.01 It." te.1 51_ . ra} | rU t rtrb jrF ri.t rd "[Vat s di r_rf 6.'.s . 4.3r 94R116 u9 OWN OI1 .3.7 Si) INOW1A111 54.0? 161.5?Pa CS of so) Z r4 br I*IS9 aCt Sol4 77 . 5z 5' 25 11!A,I 'e 5I o 1411 "A P10. , p.p. of As .I......... . . *.. .. I .. . . 0111111cl Prtw.-p I_r * we "f- o A 0 40 II ,O1 4 W06- 31m. - 15 - Investment demand ratner than by increasing domestic savlngs, a trend already detected In the pre.Lous sectlon. The response of private lnvestment ln the aftermath of the crisis of 1982-83 diffored ln the three countrles under scrutlny. Private lnvestment recovered in Chile and Mexico in the second half of the 1980e, a rather puzzllng phenomenon Ln the case of Mexlco, slnce there lt took place ln splto of very hlgh real lnterest rates. In addltlon, the recovery of private investment occurs when both countrLes carry-out an important resource transfer abroad.7 In contrast, in Bolivla the response of private investment ln the aftermath of stabilization has boon much weaker than in Chlle and MexLco. A fragLle macroeconomic equLlLbr!um (Lnternalized by the private sector) and high real Lnterest rates seem to be the chlef factors behind the slow recovery of private lnvestment Lr BolivLa. Morales (1990) explains the hlgh real lnterest rate ln Bolivia in the aftermath of stabilization by two factores the policy of tight money, and microeconomic problems In the banking and financial sector. Risk factors and credibility problems on the permanence and consolidation of the reforms also may have played a role in the observed high real interest rates. What can we concludo on the effects of the reforms on the performance of private investment in these economies? The experience of these countries ln tho 1980e shows clearly that the reforms may enhance private investment if they are accompanied by a stable macro *nvironment. High real interest rates (reflecting, in part, the existence of underlying macro imbalances) and other 7An econometric analysis of the behavior of private investment in Chile in the 1980s appear in Solimano (1990b). For a discussion of the recovery of private investment in Mexico in the mid-late 1980s see Ortiz (1990). - 16 - fundamental imbalances, tend to harm private investment. Chile in the second half of the 1980. is a good example of how fiscal balance, moderate real lntereet rates and competitive real exchange rates provide a good framework for private investment to respond to the incentives generated by the structural reforms.8 On the contrary, in the case of Bolivla, where disinflation le consolldated but the fiscal accounts and the financial system are regarded as in rather fragile condition (Morales, 1990), expectations of policy reversal have a depressing effect on private investment9 A second factor that is important for the structural reforms to be associated with a positive response of prlvate investment is the adequate availabLlity of external financing. In the three cases there is a debt overhang and the countries carry out a sizeable resource transfer abroad. From simple savings-investment idontities we can conclude that without a corresponding increase in domestic savings a high level of investment can hardly be achieved. In addition, the foreign debt service acts like an Implicit tax on investment. A third factor, generally down-played in the academic literature but to which investors in the real world seem to pay a lot of attention, refers to 8The development plans of the late sixties in forestry and agro-industrial activities, and the new land-property structure following the agrarian reform are also elements behind the strong export response of agricultural goods in Chile in the mid to late 1980s. 9Rogarding a supportive macroeconomic environment for private investment. the MHxican case is in between Chile and Bolivia. The fiscal reform has been by far more comprehensive in Mexico than in Bolivia. However, real interes rates in Mexico have remained much higher than in Chile. -17 tho favorable "business cllmate" generated with the liberalization process In fact, privatization measures as well as other liberalizing policies adopted in these countries refloct a renewed faith in free markets and private initiative. The distinctive featuro is that governments now perceive these principles as the "new engine to growthO. Two non-adiustina casOs in Latin Americas Argentina and Brazil Brazil and Argentina stand in the Latin Amrican landscape of the 1980z as two countries that have not been able to stabiliLz thelr economies, in particular to abate a stubborn process of hlgh inflation that in some episodes (e.g., Argontina in 1989) slild into outright hyperinflation. Brazil managed to grow at an Lpressive 7% per year between 1940 to 1980, and her developymnt strategy was that of a diriglste state supported, in the sixties and seventLes, by forelgn dlrect lnvestment and abundant external credits. Brazils external borrowing in the seventies largely went to finance her ambitious development plans that required high investment rates to speed-up rapid growth. In contrast, since the early seventies Argentlna started to experience a noticeable economic declino, reflected in a slowdown of growth and in mounting economic and political instabilLty. Toward the end of the 1970e and in the context of an Lll-conceived exchange rate-based stabilization 10fsynes (1936), ch.12, referred to lt as 'the state of confidence, is a matter to which practical men always pay the closest and most anxious attention. But economists have not analyzed it carefully... - 18 - plan, foreign borrowlng was (basLcally) used to flnance the acquisltlon of forelgn assets by nationals e.g., capltal flilght.11 The adverse external shocks of the early eightloe and the onset of the debt crlels severely hlt Argentlna and Brazil. The correctlon of the external and flscal lzbalances took the form of an acceleratlon of inflation and a slowdown ln growth. In contrast wlth Mexlco and Bolivla (Chle- undertook lts structural reforms in the mLd-1970.), domestlc authorltles ln Argentlna and Brazil dld not seize the opportunlty of the crlele to attempt comprehonslve structural reforms ln the publlc sector, the trade regime or other areas. The bulk of the energles of the domestlc governments ln these countrles were devoted to flght lnflatlon and to the management of thelr large external debt.12 Argentlna was the ploneer wlth hetorodox stabillzatlon wlth the launchlng of the Austral Plan ln mid 1985, followed by Brazil wlth the Cruzado in early 1986. After lnltlal success those plans were followed by a resumptlon of lnflatlon and the repeated use of prlce controls and emergency fiscal measures to curb (transLtorLly) escalatlng lnflatlon. Examples were the Bressor Pereira and Summer Plans ln Brazll ln 1987-88 and the Primavera Plan and other partial attempts ln Argentlna. The eltuatlon worsened ln 1989 for both economles as the lnflatlon rate approached hyperinflatlonary levels ln a context of domestlc recesslon and polltical dlearray. It is certainly not surprlilng to flnd a poor investment record ln the 1980. ln countrles like Argentlna and BrazlI affected by large economic llSee C.A Rodriguez (1989) for an analysis of the foreign debt problem in Argentina. 12See Heyman (1990), Kiguel and Liviatan (1990), and Cardoso (1990) on these tvo experiences vith stabilization. - 19 - Instabllity. However, there are some differencee between these two experiences. Ae table 5 Illustrates, the drop In investment rates is far larger in Argentina than in Brazil. In fact, In Argentina totai investment in the period 1985-88 Is nearly 9 points ot GDP lower than in the period 1978-81; this drop In total investment Is decomposed in a reduction of private Investment by S percentage points of GDP and a cut .'n public investment by 4 percentage points of GDP. Moreover, this decline in Investment has continued (on average) In the second half of the 1980s, in contrast to other Latin American countrLes. In Brazil the drop in total lnveetment Is less sorlous than In Argentina (its share in GDP is 3 points lower in the period 1985-88 than In 1978-8) and private investment started to recover after 1984 though public investment is stlll below Its pre-crLsis level. Argentina provides an almost text-book (though dramatic) case where protracted economic instablllty is a powerful deterrent to private Investment. An figure S shows, the downward trend in private investment -- as well as In public investment -- sta: ed in Argentina already In the mld-seventioe. Clearly, the preference for investlng resources abroad rather than at home war at work before the debt crisls, and to a large extent is responsible for the absence of recovery afterwards. On top of that lack of private invoestment, the data shows a public investment decline in the 1980e, a phenomenon tied to the fiscal crisis that Argentina suffer. 13 131t is already a well known story that the quality of public services has deteriorated sharply in Argentina in recent years. No doubt that this is related to the inability of the state to improve the collection of fiscal revenues from the tax system. - 20 - FIGURE 5 ARGENTINA 0 3 0 25 O OA,,, ~0 00U o 1170 1972 1¶74 Mys lore low0 113 198' I9" 393 117 137 1379 177 ¶373 ¶1 Io i" 3 YEARS "JNVAO PMIN TOrALMOO BRAZIL 02Si 0.3 0.0I as1s O ' ''' I,, A *97 ion 1974 1W7 173 low * lol1 ¶73 ¶ses is" '3p3 Iasi I 1O 113n YEARS ".USVIOO POtW NV GOO 1roTL INV GOP - 21 - TABLE 5 PuILIC AMC PRIVATI IMVIUIV INT AMG AC0IMCSiIC IWICATMS ANNIUAL AVIRUAGs(l) AIGMyTlWA ................................................ ......................................................... .......... vARIABLIE 197341 , 962-S 1 -U ,..................................................................... TOTAL ItVItSTMiT 22.03 16.30 13.08 (2 OF GOP) PLELIC IMVESTYNET t0.1l 6.40 6.SO (2 OF owP) PRIVATE IMVISTMEMT 11."5 7.90 4.56 (2 of GOP) REAL GOP CROTI4 -0.29 -C.04 0.20 (2) :iFLATION 126.81 396.55 319.60 (GOP OEFLATOR) CURRENT ACC EALANCE *3.32 *3.70 -3.01 (% OF MOP) fCOREIGN Oit 45.41 70.01 49.? CT. OF GOP) REAL EXC)4. RATE 124.13 209.91 239.78 ('980 a 100) ..............................................................I....... SOURCES: Inv*stfnMt: Pf.tfernam, O.P. NW N rasay, A. (1969) Other variOt l fr= Worltd BWr 0etsse PUSLIC AMO PRIVATE i1IVISTMENT ANO PAC*ICONMIC INOICATORS ANNUAL AVINAUGAI(S) BW IL ..................................................................... VARIABLE 1978-61 1912-14 9-IMI TOTAL INVESTMENT 23.10 ¶3.40 20.43 (2 OF GOP) PUBLIC 1IiVESTMENT 6.90 4.93 6.J (2 OF GOP) PRIVATE INVESTNEMT 14.20 11.17 13.73 (2 OF GOP) REAL GOP GROWTH 3.4 0.79 4.11 (2) [OIFLATION 72.97 154.05 317.19 (GOP OFfLATOR) CURRENT ACC BAIANCE -4.49 -3.0 -0.63 (2 Of GOP) FOREIGN OUST 23.47 43.51 9.72 (2 OF GOP) REAL EXCH. RATE 67.19 6.05 101.91 (1950 . 100) ..................................................................... SOURCUS: Ircnv An: t: PfgffefWV , 0.. P V disday, A. (,0199) Othier vBiebL" from W.rtd Bw* 0stalmo - 22 - In the came of Brazil the same downward trend in public invostmont, starting in the early eighties, is observed. Such reduction in public investment has been (part of) the fiscal response to the reduced external financlng as well am to the enlarged burden posed by the internal public debt. Macroeconomic Stabillty in East Asias Korea, Singapore, Thalland Let us leave Latin America and take a look at some of the "success stories" in East Asia. Let us consider the cases of Korea, Singapore and Thailand. These are high growth *conomios, outward oriented, with active state intervention in economic affairs and, in contrast with several Latin American experiencae, with a remarkable degree of macroeconomic stability. Korea since the mid-sixties has been a high growth country, strongly oriented toward the expansion of manufacturing exports. Income distribution is relatively egalitarianl4, though at the political level the country has boen governed slnce the sixties until 1987 by authorltarian military regimes. Hlgh investment rates were guided by a series of five-year economic plans where the government intervened actively controlling (among other things) the allocation of credit to firms with an overwhelming focus on exports. The close link between government and business, ln turn, crevced large conglomerateo and a high degree of industrial concentration.15 The trade regim has been far from liberal in Korea, with both tariffs and quantitative restrictions in place, although in the eighties a relaxation of 14The agrarian reform is credited as an important explanation behind the relatively even income distribution in Korea, see Collins and Park (1989). 15 See Collins and Park (1989) for a good description of the Korean case. - 23 - theso barriora has taken placo. Exchange rate policy has boen oriented to malnt,An tho oxtornal competitivoness of Koroan exports, though some episodoe of real appreciation have takan place (e.g., In tho late 1970s). Korea was hit by the debt crisis In the period 1979-82 but rocovered quickly afterwards. In contrast with most highly indebted countrier6, Korea has boon able to reduce her current account deficits after 1982 while restoring high growth, malntaining low inflation and avoldlng fiscal imbalances. The caso of Singaporo in rather particular. It is a city-state, with a high growth economy, completely open to foreign trade and with (almost) unrestricted capital mobility operating under a fixed exchange rate regime. Per capita income is comparable to that of low income OECD countries and the distribution of Income is considered to be relatively even. Singapore did not suffer a dobt crimes in the eighties and has boon running current account surpluses since the mid 1980a in the context of high growth and very low Inilation. Thailand borrowed in the late seventies and adjusted gradually afterwards taking advantage of a good record of creditworthiness . In the eighties, the reduction in the current account deficit took place in a macro envlronment of sustalned growth, while maintaining inflation low and the fiscal budget in check. This is certainly a case of sorting out adverse foreign shocks without going through a macroeconomic criaLs and domestic instability. 16 Chile is perhaps an exception in this respect. 17 See Corden (1990). - 24 - FIGt'RE 6 KOREA SINGAPORE io,0 ,3.,, .,,,.. 0; 043 0 a z a as 57 ! t S , 'O 'w ,g ,3 ¶gg Ig It It Je I et U, I w Iw I II o1 r e1 ,9t4 Ig* 1gg i* lgi o 19w s?3 nit" 'ste lae it3 t'I 'we we 193 fs it" 17 it"I low low Igo? 14 Ion19 t m Is_ " IN4 *"o Io YEARS YEARS -J6U lO WV lWV0P ?O?AL MViP ,POP O * lwOP P. ftV ?T*s NV OrP THAILAND 13 01-r 931 90e U," 0 o 0 I . . I I 19 19 , 1U 19 I 19 iete s*tev ste 'gi 'w 'S '34 we 'U YEARS NV LOP P~yr TOTA L$@aa * lOOP - 25 - TABLE 6 .t a "1S& I t N w MOt 4f a c Ig I6ICAtU AIu. a4SlaUsid) co" . .... .. . . . .. .... .......I.......... ... . ...I.... ....... ................ ,*tsI'A& 919 WI WI M.o .... .......... ......... ................................... *.6;.C !.41(48 9 13 s o a a *8 .a" o*C 1'4I iS 1 ii U II 9 .1 es:. as*AP v .1 es .1 :C IAIKI ' . I U S -06%49 5t3) 'I 21 2. afaa UoC. t Pe t I.W ... ... . .................................................. 3AtM. qt9in. *s'om - 4 _ _n.wa A. (1t a,w,akI t$79 S~ ft WI_ W W MIUIIC _ 1 40 43 1ACPUR_ _ _ C I-ICAtT ... ...... ..... ...... ij s s ....................................... ,&R[Mkl 911|t 1911w 1 luU , . .......................................................... CIa' 3 INVI S10 7^ It aie WI) 0' IC IA stu I.' i.U0. tol. r IW SM4 16 t flat1 act G T" 9. on,l NB?IN0. 4.30 cIl cc W I i s eta' Sit 5*4 U 9' 'I (tICS * 0ee1 . 1'S " ..................................................................... VINA : l0s , Of"tf, cI. , *aiw v A. 4t1s lAfI . _u ,mAI IVUI WI NO STCI ICICAT ......................... ..................................... ,MiCwk Ml-II 1686 U '. . . . . . . . . .3 . . i. U . . ... Na'. e WU' lo 1,6 *71 0m, ezwunv B. 7 .unlcS 9|@ nt 3 3sl t Is OF M I or WI GP' am(5 SItS '@ 15 i'6.31 iA , A"S t1CfW4 *. ........ C . I) T@iCs W - U WIq~ CW * t,n} (tSr 0._ SP)_e$, - 26 - What about Investment in these economies? Two main features are worth noting. Firt, particularly Korea and Singapore are high-investmont-high- growth economies. In the period 1978-88, Korea sustained n average a rate of investment near 30% of GDP and grow at an annual average rate of 6.5 percent. Slngaporo invested, on average,around 40 per cent of GDP over the same period and grew at an averago annual rate of 7.5 percent. Invostment was not immune to the cycles of economic activity experienced in these economies in the elghties, and some volatility in investment is shown in the data. To make a judgement of tho relatlve efficioncy of capital in these countries would require some difficult international comparisons, though the Implicit ICORIs do not look particularly low. Second, the data shows that in these countries private investment is by far more Important than public invostment as a share of total investment. In Korea around three-fourths of total capital accumulation is privatel in Singapore and Thailand the share of private investment in total capital formation Is around two-thirds. Those provide intereating cases of strong prlvate sectors backed by active, growth-oriented governments. An Overall Assessment From the diversity of experiences examined before some conclusions follow. There are som clear differences in the levol and composition of investment between the Latin American and Bast Asian countries examined. During the 19890 (and also earlier) investment rates of the order of 30% of GDP and more (40 percent on average in - 27 - Singapore) were not unusual in the East Asian countries. The growth record was also remarkable for the 1980s, with annual average rates of growth of the order of 6.5 - 7.5 percent. In terms of compoition, private investment li overwhelmingly dominant, representing between 2/3 and 3/4 of total capital accumulation. In Latin America, historically, investment rates have been of the order 18 of 20-25% to support rates of growth of GDP of 5.5-6.0% per year In the 1980e averago annual GDP growth decoleratod sharply to around 1.5% and investment rates centered in the range of 15-18% of GDP. In goneral, the share of public investment in capital accumulation is higher in Latin America. * The analysis suggests that a high degree of macroeconomic stability --low and predictable inflation, external and internal balance -- is of paramount importance to ensure a strong response of private investment to economic incentives. The last Asian cases examined provide a good example of this. In contrast, in some Latin American countries we find evidence that macroeconomic instability may be largely responsible for the poor performance of private investment. * The evidence on the effects of structural reforms -- e.g., liberalization, -- on private investment is, so far, still sketchy. Chile experienced a rapid recovery of private investment in the lato 1980o as real interest rates receded to "normal levels, the roil 18The average annual rate of GDP growth for the period 1950-80, was 5.8 Z vith output measured in adjusted purchasing power terms. GDP per capita in the same period grew at an annual rate of 3 Z. These calculations correspond to an average of 19 Latin American countries, see Cardoso and Fishlow, 1989. - 26 - exchange rate was kept at highly competitive levels, the economy was free of major micro distortlons and aggregate demand was hlgh followLng a boom ln copper prlces. MexLco -- whlch adopted far reachLng reforms ln the areas of trade liberalLsatLon, fLical reform and privatization ln the eLghtLes -- also saw a revLval of prlvate investment ln *plte of still hLgh domestLc real Lnterest rates. BolLvia, however, that also llberalLi-d trade, deregulated credit and labor markets and eliminated an hyperinflation ln the mLd-1980s, has not wltnessed an upsurge of prLvato Lnvestment. * A delilne ln public lnverstment has been observed ln several (adjusting and non-adjustLng) Latin Amorlcan economioe durLng in the 1980s. Chll- li one exceptlon ln this regard, though publie investmnt also declined sharply in the seventies when the structural reforms were adopted. This suggests that public investment may be squeezed ln the process of balancLng the fiscal and external accounts. SimLlarly, hlgh domestic real lnterest ratoe along wlth a hlgh level of public debt eventually impose fLscal tLghtenLng, whlch also tends to crowd-out publlc lnvestment both in adjustLng and non-adjusting countrles. - 29 - 3. Macroocon&Lc PolLeLs and Private Investment: Theory and emiirical *vLdemnce" In this sectlon revLew the literature on the offeets of macroeconomic polLcLeo on private Lnvestment, that can be useful to understand some of the experiences discussed before. In particular we are concerned wLth the impact on Lnvestment of different tools of monetary, fLscal and exchange rate polLcy aimd at correcting unsustaLnable macroeconomic lmbalances. 3.1. Demand Mana_ment PolLcles and Snvestment (L) Monetary Policy and Private Investment The restrictive monetary ar' credit polLcLes usually Lncluded ln stabilization packages affect investment through two *price channels. One is the riLs ln the real cost of bank credit, a major sourceo of Lnvestment financing ln LDCs. The second is the Lncrease Ln the opportunity cost of retaLned earnLngs -- also an important source of Lntvetment financing ln most developLng countrles -- due to higher real interest rates. Both mechanisms raise the user cost of capital and lead to a deciLne of investment. The empirical rolevance of this effect has been confirm_d ln a number of studLes (e.g., do Melo and Tybout (1990), Greene and Villanueva (1990), Solliano (1990)), but others do not find a significant effect of Lnterest rates on Lnvestment demand. The reason li that ln the repreod financial markets that characterls many LDCo, credit policy affects invesotnt diroectly, through the stock of credit avaLlable to firma with access to preforentLal Lnterest rates, rather than through the Lndirect Lntereot rate channel -- although the latter 19ThLi section draws, partly, from Serven and Soliano (1990). - 30 - will also operate for the firms that borrow in the unofficial money market (see Van Wljnbergen (1983a and 1983b)). This direct role of credit availability is found in many *mplrical studies (e.g., van Wijnb.rgen (1982), Blejer and Kahn (1984), Llm (1987), Dallami (1990)). Hence, the institutional set-up of the financial markets in developing countries is certainly an important feature determining the impact and t. numission mechanisms of monotary and credit policy on investment. (ii) fiscal Policy, Public Investment and Private Investment Hlgh fiscal deflcits push up interest rates and/or reduce the availability of credit to the private sector, and thus tend to crowd out private investment. Hence, the reduction of the public deficit usually achieved in adjustment packages should allow an expansion of private investment. However, as the experiences of several Latin American countries in the SOs show, fiscal adjustment often takes the form of reduced public lnvestment, some of whose components (especially infrastructure investments such as roads or coaiuunicatLons) may be complementary with private investment. As a result, private investment would fall as well. From the policy viewpoint, this would underscore the need to protect public infrastructure expenditures during the adjustment process, in order to facllitate the recovery of investment and growth. Several empirical studies have attempted to shed light on this issue. The results obtained by Blejer and Kahn (1984) from cross-country data indlcate that public investment In infrastructure is complementary with private investment (and other types of public investment are not). More - 31 - recently, Greene and Villanueva (1990) have arrlved at similar conclusions using a panel of 23 developing countries. Musalem (1989) flnds evidence of complementarity between private and public investment in a tLme-series study of Investment in Mexico. However, lalamma (1988) reports cross-section estimates showing that public and prlvate investment are negatively related, with an Increase in public lnvestment leadlng to a decllne ln private lnvestment. Furthermore he flnds a negative correlation between the share of publLc lnvestment in total lnvestment and the size of incremental output- capital ratios, whlch lndlcates a lower efflciency of publlc lnvestment as opposed to private lnvestment. Khan and Reinhart (1990) reexamlne the issue of the dlfferentials ln productlvlty between private and publlc lnvestment for a sample of 24 developlng countries, findlng that the marglnal productlvlty of publlc sector capltal ls negative, although not slgnlficantly no, whlle that of private lnvestment is slgnlflcantly positlve. The maln drawback of most of these emplrlcal studles Li thelr fallure to conslder lnfrastructure lnvestment separately from other types of publlc investment. Whlle ln most cases thli may be due to the unavallablilty of lnformation, such dlsaggregatlon would help identify more preclsely the relationship between public and private investment. (iii) Outout changes and Investment Emplrlcal studles of lnvestment behavlor show a strong response of lnvestment to changes ln output. Investment ln LDCr is no exception to this rule, and most econometrlc studles conclude that output fluctuations are the most important determlnant of prlvate lnvestment (see e.g., Blejer and Kahn - 32 - (1984), faini and de melo (1990), Greens and Villanueva (1990)). To a certain extent, this in a puzzling flnding, since a non-negllgLble part of output fluctuatlons ap.iSr to be transitory (therefore they should not affect Lnvestment), and lt is costly to Lnutall capltal (so adjusting to tranmitory shocks is also costly). Thus, this excessive, output-related varLability of Lnvestment in the cycle remalns largely unexplalned (see Blanchard's discussion of Shaplro, (1986)). Whatever the cause for this excessLve output sonaLtivity of lnvestment, the clear implication li that the usual demand-reduclng monetary and fiscal policies lntroduced as part of an adjustment package are likely to have an adverse short-run impact on Lnvestment through their negatlve effect on output growth. ThLs is apparent ln the context of the Q approach to lnvestment: as the econometric evldence shows (see Solimano 1989, for the case of Chile) aggregate profltablilty li hLghly procyclical -- TobLn's Q lncreases ln upturns and falls ln downturns -- so we should expect the market value of capLtal, and hence invostment, to fall ln the short run ln response to a slowdown ln economlc actlvlty followlng restrLctlve domand pollcles. This lnltlal downturn ln economlc activlty often assocLated with macroeconomlc adjustment may also affect lnvestment through its effect on expectations. In fact, a current recession could form the basis for "persimlitico expectatlons, that lead Lnvestors to postpone Lnvestment until the recovery arrLves; this, ln turn, may prevent the take off of lnvestment (particularly of projects wlth short gestatlon lags) and delay the recovery ltself, and the economy may get stuck ln a low Lnvestment equilbrlum because %of LnsuffLcient lnvestment arlsing from self-fulfilling preslmlim. How to - 33 - avoid such an outcome is an Important consideration in the design of restrictlve demand polLcles that mLnimLse the potentially adverse impact on Lnveutment and growth. 3.2. Exchange rate policy and PrLvate investment To reduce the external imbalance, adjustment progrms use a combination of expendLture-reducing and expendlture switching policLes. The latter typLcally include a real devaluatLon; thus ln the 80s many LDCs undertook sharp real depreclatlons as part of the adjustment to the debt crLiLs. A real deprecLatlon affects investment through several channels: 1i) The profitabliLty of investment - Investment goods can be viewed as a composite commodlty produced by combining domtic (i.e., construction or Lnfrastructure) and forelgn components (i.e., machinery and equlpment). In this setting, a real depreclation of the exchange rate raises the real cost of the Imported component and acts llke an adverse supply shock ln the "production" of investment goods -- wlth the magnitude of the shock being glven by the import content of investment. As argued by Buffie (1986) and Branson (1986), the effect of a real devaluation then li to ralse the real cost of new capital goods Ln terms of domestic goods; ceterls paribus, this effect tends to depress Lnvestment in the nontradable actlvltles. However, in the traded goods sector the opposlto happenst the real cost of new capltal goods falls, and Lnvestment rlsoe. The result for aggregate investment is therefore uncertaLn. Despite thli theoretlcal amblqulty, most emplrlcal studies conclude that ln the short run a real depreciation has an adverse Lmpact on Lnvestment - 34 - through this cost-of-capital-goods effect (although lts long-run effect may be positlve). For example, Musalem (1989) finds an adverse Investment effect of devaluation ln the case of Mexico. Falnl and de Molo (1990) arrlve at iLmllar reoults usLng data for 24 developlng countries. Branson (1986) explicitly calculates the lmpact of a devaluation on Tobnins Q in the home goods sector, concludlng that proflts fall (and along wlth them the market value of capital) whlle the real cost of new capltal goods rlers followlng a real deprecLatlon. Uslng an emplrlcal simultaneous equation modol for Chleo based on an extended Tobin's Q approach, SolLmano (1989) also concludes that a real deproclatlon reducoe lnvestment ln the short run. His results show that economy-wlde Q falls when the real exchange rate dcpreclates, an the adverse replacement cost effect domlnates tho market value effect. In general, a hlgh dependence on lmported capltal and intermedlate goods and a relatively low share of the traded goods sector Ln total investment would make the contractlonary result hold. ThL is made expliclt by Lizondo and MontLel (1988), who distinguish betwoon Lnvestment Ln the traded and non- traded goods sectors Ln a model in whlch capital is s*ctor-speclflc. They decompose the effect of devaluatLon on profitabillty lnto three elements: a) the Lmpact on the cost of capital; b) the effect on the product wage ln both sectors (also examlned by Van Wljnbergen (1986) and Rliager (1988)); and c) the Lmpact on the cost of Lmported Lntermedlate inputs. They show that the not effect of a real depreclatLon Ls generally ambiguous, slnce Lt tends to Lncrease investment ln the traded goods sector and reduce lt Ln the hom goods sector. - 35 - Another channel through which devaluatlon may affect the profltab$lity of inveutment is the real interest rate. Consider first the came of an unantlcipated devaluation (we discuss below the antilcipated devaluation case), and assume that interest rates are determined in domestic assets markets (I.e., in the money market). Devaluation raises the prlce level through its impact on the cost of imported intermediate Inputs and wagea under indexation; if monetary pollcy does not fully accommodate the increase in the price level, j._l money balances fall, pushing up the real Interest rate for a given rats of (exp3cted) inflation. In this way, devaluation depresses the market value of existing capital and exerts an adverse effect on investment. On the other hand, if devaluation was anticipated and if it succeeds in ellminating devaluation expectations, then it ma. result in an investment expansion, since the required return on capltal would tend to fall reflecting the reduction in the anticipated rate of depreciation. Whether this will be so depends or the degree of capital mobility and also on the Import content of investment (see below). (ii) Devaluation, activity levels and investment: Devaluation may also affect investment through its impact on aggregate demand. ThLi may be especially lmportant when firms face sales constraints, so that the degree of capacity utilization or other variable representing demand considerations has a strong systematic effect on investment (as noted above, such effect is often found empirically). If devaluation reduces aggregate demand ex-ante, then ex-post investment is likely to fall. Moreover, if Lnvestment has a sLgnificant import content, then output expansion is likely to be a necessary (but not - 36 - sufficient) condition for lnvestment not to fall ex-post (S-rven (1990)). The literature on contractionary devaluation (Krugman and Taylor (1978), Van Wijnbergen (1986), Zdwards (1987), Solimano (1986), Lizondo and Montiel (1989)) emphasizes the slow working of sabstitution effects arising from devaluation; hence in the short run the impact of a real devaluation on aggregate demand is dominated by its advers income effects. The latter operate through two main channelas one arisoe from the likely Initial trade Imbalance, which results in a real incomo transfer to the rest of the world (even at given terms of trade)t tho other, from the negative impact on consumption of real income redistribution from wages to proflts. On tho supply side, three transmission mechanisms may contribute to output contraction: the Increased real cost (Ln torms of domestic goods) of imported inputs, the rise of working capital costs (due to increased interest rates), and real wage resistance. If the net effect of a currency devaluation is contractionary, i.e., GDP falls, then the slump in economic activity is likely to form the basis f-- investors to cut investment spending -- unless they clearly perceive the slump to be transitory. However, with sufficiently strong substitution effects (e.g., a large impact of devaluation on net exports) an expansionary outcome wlll result, and so devaluation may raise real incom and stimulate investment spendLng as the dogree of capacity utilization increases. This outcome becomes more likely as time passes and substitution effects gradually come into play (see solimano, 1986, for an evaluation of such J-curve type effects of devaluation on output in Chile). - 37 - (iv) Anticipated depreciation and the timing of Investment - The discussion until now has focussed on the effects of devaluation without making any explicit distinction between anticLpated and unantLeipated devaluation. An anticipated devaluation can also have a substantial impact on the tiAing of investment. This results from the combination of two opposing effects ot devaluation expectations: the effect on interest rates, and the effect on the future cost of capital goods imports (for a detailed exposition, se* Sorven (1990)). The effect of an anticipated devaluation on interest rates depends on the degreo of capital mobility -- that is, on the costs of portfolio adjustmen$ In the general caso of lmperfect capltal mobility, the domestic real Interest rate is an increasing function of the forelgn real interest rate plus the expected rate of depreciation of the real exchange rate (it may also depend on the relative or absolute stocks of financlal assets). The perception by the public that a real depreciation is imminent will be reflected in hlgher real interest rates -- and more so the largor the degree of capltal mobility. In this way, devaluation expectations represent a transitory investment disincentive; pending the depreciation, the real interest rate is high and investment is low. once devaluation has taken place, the transitory investment disincentive is eliminated and investment rises. The import content of capital goods operates in the opposite direction. When a real depreciation is antLeipated, the real cost of imported capital goods is expected to risel pending tho depreclation, capital goods imports are transitorily cheap and hence investment must be transltorily high (the mechanism is entirely mimilar to an anticipated increase in tarlffs on - 38 - investment goods). As pointed out by Dornbusch (1986), this represents a transitory Investment incentive, that disappears once the dopreciation is actually implemented. Obviously, the not effect on investment depends on the dogroe of capital mobility relativo to the import content of investment. Wlth high capital mobility, the interest rate effect dominates, and devaluation expectations lead to an investment slump that will persist until the depreciation is actually undertaken. Wlth low capital mobility and high import content of investment, an anticipated depreciation may roesult in a transitory investment boom, and the actual depreciation may give way to a drop in investment. As described in Serven (1990), these concluslons are consistent wlth the empirical evidence for Chile and Uruguay. 3.3 - Trade liberalization and investment Trade liberalization is one of the structural reforms that eften accompany macroeconomic adjustment measures. In principle, a permanent trade liberalization should reduce investment In the previously protected import- competing sector and encourage Investment in the export sector. Hence, its impact on aggregate investment is uncertain, as it depend. on the relative capital intenolties of the different economic sectozs.20 In practice, in many LDCr the protected sector is relatively capital-intensive, and thus trade liberallzation could well result in roduced aggregate investment -- which of 20This is empirically confirmed by Lopez (1990), who doos not find any significant effect of import and export restrictions on capital accumulation in a sample of 35 developing countries. - 39 - course may be consistent with enhanced growth due to the increased efficiency with which investment would be used. However, when liberalization is perceived as temporary, its results can be very different. In such case, the removal of trade barriers can introduce important distortions in both the intertemporal and the sectoral allocation of investment. The timing effect is similar to the one examined in the previous subsection: if irvestment goods have a high import content, a temporary liberalization amounts to a transitorily low cost of investment goods, and hence to a temporary investment incentive. This may lead to a transitory investment boom, vyhich, in addition, is likely to be allocated to the 'wrong' sectors: if trade restrictions are expected to be reintroduced shortly, the increased investment will be directed to the protected sector and not to the export industry -- exactly the opposite effect to what the liberalization intended to achieve. Thus a trade liberalization suspected to be only temporary can have very adverse consequences for investment. As several authors have emphasized (see van Wijnbergen (1985), Rodrik (1989)), this is especially so when investment is irreversible: then there is an incentive to halt investment in all sectors, to avoid the irreversible mistake of investing in what can turn out to be the 'wrong' activity. We explore the issue of irreversibility in more depth in the next section. - 40 - 4. The incentive structure and investment response: credibility, uncertainty and irreversibilLty A key ingredient of most macroeconomic adjustment packages is a change in economic incentives that switches spending towards domestic goods (offsetting the deflationary bias of the usual monetary and fiscal restraint) and raises profitability in the tradable sector. This change in incentives is expected to lead to an outburst of investment in the tradable goods sector, Lncreasing productive capacity and enhancing economic growth -- and thus ensurLng the sustainability of the adjustment effort. In practice, however, the investment response often is unexpectedly weak, and involves long delays (a clear example is the case of Bolivia in the late 80s). This poses major diffLculties for the adjustment effort, since in the absence of an investment expansion the short-run deflationary consequences of the expenditure-restraining measures may be magnified, leading to a persistent reduction in growth. In this way, the lack of an adequate investment response in the tradable sector to the change in economic incentives increases the cost of the adjustment ln terms of employment and growth; ultimately, it may render the stabilization effort socially unacceptable and thus unsustainable. Conventional Lnvestment theories cannot provide a satisfactory explanation for this slow reaction of investment. To justify the latter, one would have to assum that firms face rapidly increasing adjustment costs to investment -- whlch does not sees to hold true empLrically -- or that Lnvestors' expectations adapt very slowly to changes ln the economic envlronment -- but there is no clear rationale for such suboptiml behavior by investors. A more satisfactory explanation can be offered by emphasizing the importance of uncertainty factors in investment decisions. - 41 - 4.1 Irreversibility, Uncertainty, and Investment As an emerging literature has emphasized (see Pindyck (1989) for references), the key role of uncertainty in investment decisions follows directly from the irreversible nature of most investment expenditures. These can be viewed as sunk costs, because capital, once installed, is firm- or industry-specific and cannot be put to productive use in a different activity (at least without incurring a substantial cost). The decision to undertake an irreversible investment in an uncertain environment can be viewed as involving the exercising of an option -- the option to wait for new information that might affect the desirability or timing of the investment. Thus, the lost value of this option must be considered as part of the opportunity cost of investment -- an issue which is overlooked in the conventional net present value calculations (which would therefore underestimate the opportunity cost and overpredict investment). As recent studies have shown, this opportunity cost can be substantial, and is also very sensitive to the prevailing degree of uncertainty about the economic conditions that determine the future returns to the investment. As a consequence, changes in uncertainty can have a very strong impact on aggregate investment; from a policy perspective, the stability and predictability of the incentive structure and the macroeconomic policy environment may be as important as the level of the taz incentives or the interest rate. In other words, if uncertainty over the economic environment is high, tax and related incentives my have to be very (or even prohibitively) large to have any significant impact on investment. It is important to note that this effect of uncertainty is completely independent of investors' risk preferences or of the extent to which their - 42 - risks may be diversifiable. Investors may be risk-neutral (as assumed by most of the irreversibility literature) and their risks completely diversifiable; yet investment would continue to depend negatively on the perceived degree of uncertainty. The latter becomes important here simply because the fixed investment decision cannot be 'undone' (at least at zero cost) if future events turn out to be unfavorable. In general, there will be a value to waiting (i.e., an opportunity cost to investing today rather than waiting for information to arrive) whenever the investment is irreversible and its returns evolve stochastically over time. The relevance of these results for macroeconomic policy, especially in developing countries, cannot be overemphasized. Consider, for example, the problem of relative price volatility. Many developing countries suffer from high and unpredictable inflation, which is usually matched by high relative price variability. The irreversibility approach suggests that this would reduce the effectiveness of relative price changes in stimulating investment. Specifically, a history of frequent relative price swings would make investors extremely cautious in reacting to a policy-induced change in sectoral incentives; substantial time may elapse before investors become convinced that the change is permanent -- and before they are willing to give up their option to postpone investment. Notice also that the implementation of an adjustment program may well increase uncertainty in the short run, as private agents start receiving mlxed Lncentive signals -- some associated with the previous policy rules, aome with the stabilization package, and som with the structural reforms aimed at restoring modium term growth. An example along these lines is provided by van Wijnbergen (1985), who shows that a trade - 43 . reform which is suspected to be only temporary can in fact lead to a fall in investment -- as economic agents postpone investment in both the home and traded goods sectors in order to receive additional information. The debt overhang faced by many high-indebted countries creates a similar problem, which has been emphasized by Sachs (1988). It &rises from the need to carry out an external transfer to the country's creditors, and represents another source of LnstabliLty of the macrooconomlc envLronments in a context of uncertainty, the level of the real exchange rate and/or the demand management policLes consistent with the required transfer also become uncertain: the size of the transfer itself is not known with certainty, as it depends on uncontrollable factors such as the future level of world interest rates and the terms of trade. Carrying out the transfer may require future real exchange rate changes, fiscal contraction, or both. Thus investors must face the risk of large swings in relative prices, taxation, or aggregate demand; as we argued above, each of them would lead to reduced investment. In practice, this effect may be hard to identify, since foreign debt may affect investment adversely through two additional channels (emphasized by Borenzstein (1989)). First, the debt overhang, which acts as an anticipated foreign tax on current and future income: since part of the future return on any investaent will accrue to the creditors as bigger debt service payments, it discourages capital accumulation and promotes capital flight. Second, the credit rationing effect: a highly indebted country is likely to face credit constraints in international capital markets, which is equivalent to facing higher real interest rates, and this will also discourage Lnvestment. Empirical studies have confirmed the adverse impact on investment of the . 44 - foreign debt burden (e.g., Faini and de H lo (1990), Greene and Villanueva (1990)), though still more research is needed to identify the specific mechanisms at work. 4.2 The role of credibility From a policy perspective, a very important source of uncertainty is the imperfect credibility of polLcy reforms. The latter is related to the public's perceptions about both the internal consistency of the adjustment program and the government's villingness to carry out the program despite its implied social costs. Unless investors viev the adjustment program as fully credible in both senses, the possibility of a future policy reversal vwll become a key determinant of the investment response. As argued by Dornbusch (1988), Rodrik (1990), the policy measures of an adjustment program can easily be reversed -- while investors cannot undo their fixed capital decisions. In such conditions, the value of waiting arises from the losses (the 'irreversible mistake', in Bernanke's (1983) terminology) that investors would incur if policy were in fact reversed in the future. Clearly, the larger the perceived probability of a future policy reversal, the less willing investors will be to undertake fized investment projects -- or the larger the current return they vill require in order to compensate for the possibility of an irreversible mistake. Moreover, such increase in the required return on investment can be substantial *ven when the perceived probability of reversal is moderately low, as Dornbusch (1989) and Rodrik (1989) have shown. Thus, when investment is irreversible policy uncertainty can have disastrous - 45 - consequences for private investment (Rodrik, 1990).21 This also implies that any gLiven set of policy easures can have widely different effects on investment depending on the prevailing degree of 'confidence' of the public. In particular, stabilization may entail large social and economic costs if credibility ie low -- sLnce the investment response will be insufficLent to offset the deflationary bias of the usual fiscal and monetary restraint masures: thus, a persistent receseLon may dtvelop before investors become confident enough that the adjustment measures will be maintained. This may be particularly relevant in economies with a past history of frequent policy swings or failed stabillzation attempts -- two features shared by many highly lndebted countries -- in which the ptlvate sector has learned to view adjustment programs with considerable skepticism. Hence setting the right economic incentives ls a required precondition for investment and growth, but it does not guarantee that they will in fact take place. Bolivia and Mexico provide examples of a rather slow lnvestment response, while Korea and Singapore are cases of strong private sector response to economic incentives. Obviously, high credibility would help speed up the investment response and reduce the costs of the adjustment. However, the question of how can credibility be affected by government actions remains largely unresolved. Specifically, an important issue here is the choice between gradual and abrupt stabilization. The former would set initially modest objectives, which can be achieved with near certainty, in order to build up the government's reputation. The latter would start with an 21Thia adverse impact of uncertainty on private investment in LDCs has been empirically verified ln several recent studies (see Sollmano (1989), Faini and de Melo (1990), Lopez (1990). - 46 - overadjust.ant (e.g., an over-depreciation of the exchange rate) to frontload the incentlves to resource reallocation (but also the costs of the adjustment). As argued by Edwards (1988), the choice may largely depend on the specifics of each countryl the social distribution of adjustment costs implicLt in the program, together with past policy experience, are likely to be important lssues here. It is important to emphasize that policy reversal ic an endogenous out- come in this framework, since current private sector decisions affect the opportunity set of future policy actions and ultimately determine the sustainability of the adjustment policy. As an example, consider again the case of a large real depreciation that due to low confidence fails to attract investment to the tradable sector. Its only visible effects will be a deflationary real income cut and an income redistribution from labor to capital, especially in the traded goods sector; however, because the depreciation is not sufficient to compensate for the lack of credibility, the increased profits will be reflected in increased capital flight. Social pressure and balance of payments problems may eventually force policy reversal, thus confirming the initial skepticism of investors. The alternative situation starts with high confidence, which allows an investment boom and validates the adjustment program. Thus, there are two possible outcomes, and the final result of the adjustment measures is - 47 ; indeterminate22. This is due to the existence of an externality that creates a wedge between the social and private returns to investment: higher aggregate investment helps sustain the adjustment effort and therefore results in higher returns to investment, a mechanism that will be ignored by the individual investor. If left to its own resources, the economy may get stuck in the 'low investment-adjustment failure' equilibrium23. Since the 'high investment- adjustment success' equilibrium is clearly better in a meaningful sense, it is crucial to investigate what specific policy measures can lead the economy to this superior outcome. There is no simple answer to this question. While transitory investment incentives would appear as the most appropriate tool to address the investment externality, in practice they run the risk of destabilizing public finances, which often are a key element in adjustment programs. On the other hand, sufficient external support to the stabilization effort may play an important role by raising investors' confidence in the sustainability of the adjustment, thus giving way to the Livestment takeoff. In fact, the lack of external resources has been a negative element that probably contributed to weaken the private sector's confidence in some stabilization attempts in highly indebted countries. 220beerve that in both cases expectations are self-fulfilling, which reflects the existence of multiple rational expectations equilibria. Such result is familiar from the literature on investment under monopolistic competition (Kiyotaki (1988), Shleifer and Vishny (1989)). An eazmple of indeterminacy similar to that in the text, but focussed on the consequences of trade liberalization, is provided by Rodrik (1989). 23Hovever, when multiple equilibria are present there is no clear rule to determine which of the possible outcomes will in fact prevail. An attempt to shed some light on this issue is made by Krugman (1990). 48 - 5. Econometric analysis In the preceding discussion we have examined from the theoretical viewpoint the effect on private investment of a number of factors. The inmmediate question is to what extent can these factors contribute to explain the observed performance of investment in LDCs in recent years. To investigate thl issue, in this section we estimate a simple investment equation using pooled cross section-time series data for a group of developing countries. We postulate that real private investment is a function of real output growth, the real exchange rate, real public investment, the foreign debt burden, and the degree of macroeconomic uncertainty/instability: IP/Y - F(AY, e, IG/Y, Di/Y, o, (IP/Y)_1) where IP is real private investment, Y is real output, e is the real exchange rate, IG is real public investment, D- /Y is the foreign debt/GDP ratio, and a represents an appropriate measure of instability. For empirical purposes, we also introduce lagged private investment among the explanatory variables, in order to allow for some dynamics arising from adjustment and/or installation costs. According to our previous discussion, we would expect real output growth to exert a positive effect on the private investment rate; in contrast, an increase in the degree of economic instability or in the burden of foreign debt should reduce investment. On the ^ther hand, the effect of the real exchange rate is uncertain, as discussed before; the same applies to the public investment rate, which can have an expansionary or contractionary - 49 - effect on private investment depending on whether public investment is ,rimarily complementary with or substitutive for private investment.24 To estimate this investment equation, we use data for the years 1972-1987 for twelve developing countrles so our sample is constituted by 192 observations. The choice of sample period was dictated by data availability; the countries considered are essentially those whose performance was reviewed in section 2.2 above, to which we add Colombia, Kenya, Turkey and Uruguay. Thus tho total sample comprises these four countries plus Argentina, Bolivia, Brazil, Chile, Korea, Mexico, Singapore and Thailand, and, as we noted before, represents a six of positive and negative adjustment exporiences. To masure mncertainty o, we followed other authors ln using the sample variability of some key macroeconomic variables. In particular, we experimented with the variability of the real exchange rate and of real output growth; the corresponding results are reported below. Because for each country the uncertainty variable a is time-invariant, the investment equation was estimated using a two-step procedure. First, we compute the estimates of the coefficients on the time-varying variables (i.e., all explanatory variables except a) using an instrumental variable procedure. In the second stage, we recover the coefficient estimate for the uncertainty variable e (for details see e.g., Anderson and Hsiao (1982)). 24We should note that our empirical equation does not include the real interest rate among the explanatory variables. Our experiments with alternative measures of the ex-ante real interest rate proved unsuccessful. The usual dlfficulties in measuring such variable are in our case likely to be compounded by the vLde differences in flnancial market arrangements across the countries in the sample, and also across time periods. Thus, we opted for excluding interest rates from our final specification. - 5o - We experimented with different dynamic specifications, allowing for lags in the effects of the explanatory variables. However, the results were in all cases very similar. Table 7 presents the first-stage estimation results for the preferred specification, which was selected on the basis of the overall significance of the estimates. - 51 - Table 7 Determinants of Private Investment (;972-87) (dependent variable: log(IP/Y) ) Variable Coefficient T-statistic Real output growtha 3.532 3.853*** Real exchange rateb (lagged) .002 .031 Public investmentc (lagged) .058 1.170 Foreign debt/GDP ratiod -.104 -2.633*** Lagged dependent variable .584 9.845*** R2 .584 SEE .203 N. obs 180 Notes: a - First difference of the log of real GDP (sources World Bank). b - Log of the real exchange rate index (source: World Bank and IFS). Increase means depreciation. c - Log of the real public investment/real GDP ratio (source: World Bank). d - Log of the ratio of foreign debt to GDP in U.S. dollars (source: World Bank). - Coefficient significant at the 5 percent level. - 52 - Table 8 The Effect of Uncertainty on Private Investment (1972-87) (dependent variable: country-specific effect from Table 5.1 )a Variable Equation 1 Equation 2 Equation 3 Constant -7.413*** -7.592*** -7.373*** (-50.566) (-63.710) (48.495) Output growth variabilityb -.130 ----- -.134*** (-1.)24) (-1.986) Real exchange rate ---- -.011 -.013 variabilityb (-.755) (-.993) Notest a - T-statistics in brackets b - Heasired by the coefficient of variation. - S3guificant at the S percent level - 53 - As in most empirical studies, we find that real output growth has a strong positive impact on private investment. In contrast, the effect of the real exchange rate is very small and insignificant, even after allowing for a one-year lag; this is in accordance with our theoretical discussion in which we identified several channels through which the real exchange rate affects investment in opposite directions. Public investment has a positLve effect on private investment after a one-year lag, suggesting that complementarity relationships between both investment categories dominate in our sample. However, the effect is only moderately (i.e., at the 25 percent level) significant. As expected, the foreign debt burden has a strong negative effect on the private investment ratio. As we discussed above, this result may reflect a combination of the increased macroeconomic uncertainty arising from the need to carry out an increased resoirce transfer, or also from credit rationing effects in world capital markets. Finally, we also find substantial inertia in private investment, as indicated by the large and highly significant coefficient of the lagged dependent variable. Using the resuits in Table 7, we can estimate the impact of uncertainty and instability on the private investment ratio. Our two proposed measures of instability are the variability of the real exchange rate, and the variability of real output growth: in both cases, variability is measured by the coefficient of variation of the corresponding variable. The empirical results for three alternative specifications (using real exchange rate variability, output variability, or both, as the relevant uncertaLnty measure) appear in - 54 - Table 8. Because the sample for the sesond-stage regression is very small (only 12 observations), the results should be viewed only as suggestive. Nevertheless, as Table 8 shows, we do find that in all cases the uncertainty measures have a negative effect on private investment; thus, countries with higher real exchange rate instability and/or higher growth variability tend to have lower private investment ratios -- although only the output variability effect is statistically significant at conventional levels. 6. Conclusions and Policy Implications One of the most troublesome features of the experience with macroeconomic adjustment in LDCs in the eighties has been the adverse impact on investment. In most cases, the adjustment measures have not been rewarded by a vigorous response of private investment, and this creates the risk of a persistent growth slump and an eventual failure of the adjustment effort. The cross-country comparison carried out between several Latin America and East Asia countries suggest the following results regarding the performance of investment: * There are some clear differences in the level and composition of investment between the Latin American and East Asian countries examined. During the 1980s (and also earlier) investment rates of the order of 30Z of GDP and more (40 percent on average in Singapore) were not unusual in the Bast Asian countries to support growth rates of GDP of the order of 6.5 - 7.5 percent. In terms -5 - of composition, private investment is overwhelmingly dominant, representing between 2/3 and 3/4 of total capital accumulation. In Latin America, in the 1980s average annual GDP growth dec lerated sharply to around 1.5Z and investment rates centered in the range of 15-182 of GDP (historically, in Latin America, investment rates have been of the order of 20-25S to sustain rates of growth of GDP of 5.5-6.0S per anum). In addition, on average, the share of public investment in capital accumulation is higher in Latin America. The analysis suggests that a high degree of macroeconomic stability --low and predictable inflation, external and internal balance -- are of paramount importance to ensure a strong response of private investment to economic incentives. The East Asian cases examined provide a good example of this assertion. In contrast, in several Latin American countries macroeconomic instability may be largely responsible for the poor performance of private investment. * The evience on the effects of structural reforms -- e.g., liberallzation, -- on private investment is, so far, still sketchy. Chile experienced a rapid recovery of private lnvestment in the late 1980s as real interest rates receded to nor.al' levels, the real exchange rate was kept at highly competitive levels, the economy was free of major micro distortions and - 56 - aggregate demand was high following a boom in copper prices. Mexico -- which adopted far reaching reforms in the areas of trade liberalization, fiscal reform and privatization in the eighties -- also saw a revival of private investment in spite of still high domestic real interest rates. Bolivia, however, that also liberalized trade, deregulated credit and labor markets and eliminated an hyperinflation in the mid-1980s, has not witnessed an upsurge of private investment. * A deelLne in public investment has been observed ln several (adjustlng and non-adjusting) Latin American economies during in the 19609. Chile is one exception in this regard, though public investment also deelined sharply in the seventies when the structural reforms were adopted. This suggests that public invostment may be squeezed in the process of balancing the fiscal and external accounts. Similarly, high domestic real interest rates along with a high level of public debt eventually lmpose fiscal tightening, which also tends to crowd-out public investment both in adjustLng and non-adjusting countries. On the other hand, we can summarize our econometric results as follows: * Real output growth has a strong positive impact on private lnvestsent. In contrast, the effect of the real exchange rate is mll and statistically LisnipLficant in our sample, even after allowLng for a one-year lag. - 57 - * Public investmert has a positive effect on private investment after a one-year lag, suggesting that complementarity relationships between both investment categories dominate in our sample. However, the affect is only moderately (i.e., at the 25 percent level) sLgnificant. * The foreign debt burden has a strong negative effect on the prlvate investment ratio. This result may reflect a combination of the increased macroeconomic uncertainty arLsing from the need to carry out an increased resource transfer, or also from credit rationing effects in world capital markets. * Our tvo proposed measures of instability (the variability of the real exchange rate, and the variability of real output growth, in both cases, measured by the coefficient of variation) have a negative effect on private investment; thus, countries wLth higher real exchange rate instability and/or higher growth variability tend to have lower private investment ratios -- although only the output varLabllity effect is statistically significant at comventLonal levels. What can we conclude for the desLgn of growth-enhancing adjustment programs? First, macroeconomic stability and policy credibllity are key ingredients for the achievement of a strong investment response. In a context - 58 - of high macroeconomic uncertainty, the reaction of investment to incentive changes is likely to be very limited. The same will happen if the policy measures are perceived as inconsistent or suspected to be only temporary. In such circumstances, investors will prefer to wait and see before committing resources to irreversible fixed investment. Second, this has important implications for the sequencing of adjustment measures. In particular, macroeconomic stability is a prerequisite for the success of many types of reforms. For example, trade liberalization measures undertaken in a context of large macroeconomic imbalances are likely to be viewed as purely transitory, and thus can have very adverse consequences on the intertemporal and intersectoral allocation of investment. Third, even vell-designed, consistent adjustment programs may have to overcome, at least in the early stages, the consequences of lack of credibility. The availability of sufficient exttrnal resources can play an important role here, by raising the private sector's confidence in the viability of the adjustment effort, thus contributing to facilitate the recovery of private investment. Fourth, even if the policy changes are perceived as permanent, the lack of adequate infrastructure may pose a significant obstacle to the recovery of private investmnt. The implementation of well-targeted public investments in infrastructure projects that complement private investment can play an important role to stimulate the private sector's response to the adjustment measures. - 59 - REFERENCES Anderson, T. and C. Hsiao (1982): 'Formulation and Estimation of Dynamic Models using Panel Data,' Journal of Econometrics 18. Arrow, K. (1968): 'Optimal Capital Policy with Irreversible Investment,' in Value, Capital, and Growth: Essays in Honor of Sir John Hicks, J. Wolfe, ed. Edinburgh University Press. Balassa, B. (1988): 'Public Finance and Economic Development,' PPR Working Paper 1 31. The World Bank. Bernanke, S. (1983): 'Irreversibility, Uncertainty, and Cyclical Investment,' Quarterly Journal of Economics. Bertola, G. (1989): 'Irreversible Investment,' mimeo, Princeton University. ------ and R. Caballero (1990): 'Irreversibility and Aggregate Investment,' mimeo, Columbia University. Bizer, D. and D. Sichel (1988): 'Irreversible Investments An Empirical Investigation,' mimeo. Blejer, M. and M. Khan (1984): 'Government Policy and Private Investment in Developing countries,' I.M.F. Staff Papers 31. Borensztein, E. (1989)s 'Debt Overhang, Credit Rationing, and Investment,' mimseo I.M.F. Branson, W. (1986): 'Stabilization, Stagflation, and Investment Incentives: The Case of Kenya, 1979-1980,' in S. Edwards and L.Ahamed eds. Economic Adjustment and Exchange Rates in Developing Countries, N B.E.R. The University of Chicago Press. Buffie, E. (1986): 'Devaluation, Investment and Growth in LDC's,' Journal of Development Economics, Vol. 16. North-Holland. Caballero, R. and V. Corbo (1988): 'Real Exchange Rate Uncertainty and Exports: Multi-country Empirical Evidence,' Working Paper 414, Department of Economics, Columbia University. Cardoso, 3. (1990)t 'From Inertial Inflation to Megainflation: Brazil in the 1980e. in Bruno, M. et. al.(eds) Lessons of Economic Stabilization and its Aftermath M.I.T Press (forthcoming). ------- and A. Fishlow (1989): 'Latin American Economic Developments 1950- 1980,' NBER Working Paper # 3161, Cambridge MA, Collins, S. and W.A. Park (1989): 'External Debt and Macroeconomic Elerformance in South Korea,' in J.Sachs (ed.) Developing Country Debt and the World Economy The University of Chicago Press. - 60 - Corbo, V. and A. Solimano (1990): 'Stabilization PoliCies in Chile Revisited,' in Bruno *t. al. Cordon, H. (1990): 'Macrooconomic Policy and Growth: SOme Lessons of Experience.' Paper presented at the World Bank Annual Conference on Development Economics, 1990. Chenery, H. and H. Bruno (1962): 'Development Alternatives in an Open Economy: The Case of Israel,' Economic Journal 72. Chhibber, A. and N. Shafik (1990): 'Does Devaluation Hurt Private In estment? The Indonesian Case,' World Bank PPR Working Paper 0 418. Dailami, H. (1987): 'Expectations, Stock Market Volatility, and Private Investment Behavior: Theory and Empirical Evidence for Brazil,' mimeo, The World Bank. ------- (1990a)s 'Financial Policy and Corporate Investment in Imperfect Capital Markets: the Case of Korea,' World lank PPR Working Paper 0 409. ------- (1990b)t 'Policy Changes that Encoura. Private Investment in Colombia,' World Bank PPR Working Paper # 226. ------- and M. Walton (1989): 'Private Investment, Government Policy and Foreign Capital in Zimbabwe,' PPR Working Paper 0 248, The World Bank. de Melo, J. and J. Tybout (1990): 'The Effects of Financial Liberalization on Savings and Investment in Uruguay,' Economic Development and Cultural Change. Dixit, A. (1987): 'Entry and Exit Decisions of Firms under Fluctuating Real Exchange Rates,' mimeo, Princeton University. Dornbusch, R. (1986): 'Overborrowing: Three Case Studies,' in International Debt and the Developing Countries, G. Smith and J. Cuddington eds. Washington, D.C.: World Bank. ------- (1987): 'Open Economy Macroeconomics: Now Directions,' N.B.E.R. Working Paper 2372. ------- (198) 'Notes on Credibility and Stabilization,' N.B.E.R. Working Paper 2790. ------- (1989): 'From Stabilization to Growth,' mimeo, MI. Edwards, S. (1988): 'Stabilization, Macroeconomic Policy and Trade Liberalization,' *mimo. Faini, R. and J. do Melo (1990): Adjustment, Investment, and the Real Exchange Rate ln DevelopLng Countries,' Economic Policy (forthcoming). . 61 - Greet., J. and D. Villanueva (1990)t 'Prlvate Investment in Devoloplng Countrles: An Emplrlcal Analysis,' siamo, IZN. Heymann. D. (1990)t 'From Sharp Disinflatlon and Hyper and lacks the Argentine Experience, 1985-69,' ln Bruno, M. et. al. Kahn, M. and C. Reinhart (1989): 'Prlvate Investment and Zconoaic Growth ln Developing Countries, IDM Staff Papers. Keynes, J. M. (1936)s The Coneral Theo"n of Eagloyment Interest and Money. A Harvest BHJ Book, Now York. Kiguel, M. and Livlatan, M. (1990): 'The Inflation StablILuatLon Cycles in ArgentLna and Irail,' in Bruno et. al. KiyotakL, (1988)t I MultLple Ezpoctational EquliLbrLa under MonopolLstLc Competition,' Quarterly Journal of Economics 103, p.696-713. Krugman, P. (1988): Exchanse Rate InstabliLty. M.I.T. Press. ------- (1989)s 'HLstory vs. ExpectatLons,' NbER WorkLng Paper * 2971. Lelte, P. and R. Saoe-Zadeth (1986): 'CredLt Allocation and Investment DecLsions: the Case of the Manufacturlng Sector in Koroa,a World Development 14. LLz, J. (1987): 'The Now StructuralLst CrltLque of the Monetarist Theory of Inflation,' Journal of D*veloment Economices. Lizondo, J.S. and P.Montiel (1969)s 'ContractLonary DevaluatLon in Developing CountrLes: An AnalytLcal Overview,' I.M.F. Staff Papers. Lopez, R. (1990): 'Economic growth, CapLtal Accumulation, and Trade Policy ln LDCs,' mimeo, The World lank. Mayer, C. (1989): OFLnancLal Systems, Corporate FLnance and 7conomic Development,' mimeo, City UniversLty Business School. McDonald, R. and D. SLegel (1986): 'The Value of WaLtLng to Invest,' Quarterly Journal of Zconomics 101, pp. 707-28. Morales, J.A. (1990) ' The TransLtLon From StabllisatLon to Sustained Growth in Bolivia,' in Bruno, H. et Al. Musalem, A. (1989): 'Prlvate Investment ln HMelco: an Emplrlcal Analysis,' World lank PPR WorkLng Paper 0 183. Ortlz, G. (1990)s 'Mexico Beyond the Debt Crises. Towards Sustalnable Growth wlth Prlce Stability,' ln Bruno et. al. PLndyck, R. (1986): 'Capital Risk and Models of Investment BehavLor,' MIT WorkLng Paper, 1819-86. - 62 - ------- (1988a)t 'Options, Flexibility and Investment Decisions,' MIT Center for Znergy Policy Research Working Paper: 88-018. ------ (1988b): 'Irreversible Investment, Capacity Choice, and the Value of the Firm,' American Economic Review 78, pp. 969-85. ------- (1989): 'Irreverslbility, Uncertainty, and Investment,' mimeo. Rama, M. (1987): 'Inverei6n Privada: Teoria Q y Enfoque de Desequilibrio,' 7th Latin American Meeting of Econometric Society: Abstracts and Papers, Vol. 4, ------- (1989)s Empirical Investment Equations in LDC's, mimeo. The World Bank. Risager, 0. (1988): 'Devaluation, Profitability and Investment: A Model with Anticipated Wage Adjustment,' Scandinavian Journal of Economics 90. Rodrick, D. (1989): 'Policy Uncertainty and Private Investment in Developing Countries,' NBEE Working Paper 2999. Rodrik, D. (1990): 'How should Structural Adjustment Programs be Designed?' World Develoiment 18. Rodriguez, C. A. (1989): 'The Argentine Debt Crises,' in S. Edwards and F. Larrain Debt and Adiustment. The Latin American Experience. Basil Blackwell. Sachs, J. (1988): 'The Debt Overhang of Developing Countries,' in Debt, Stabilization, and Development: Essays in Memory of Carlos Diaz Alejandro. Oxford: Basil Blackwell. Serven, L. (1990): 'Anticipated Real Exchange Rate Changes and the Dynamics of Investment,' mimeo, The World Bank. Serven, L. and A. Solimano (1990): 'Private Investment and Macroeconomic Adjustment: A Survey,' mimeo, The World Bank. Shafik, N. (1990): 'Modelling Investment Behavior in Developing Countris: an Application to Egypt,' mimeo, The World Bank. Shapiro, H. (1966): 'Investment, Output and the Cost of Capital,' Brookings Papers on Economic Activity, 1. Shloifer, A. and K. Vishny (1989): 'The Efficiency of Investment in the Presence of Aggregate Demand Spillovers,' Journal of Political Economy 96. Soliaano, A. (1986)s 'Contractionary Devaluation in the Southern Cones: The case of Chile,' Journal of Developing Economics 23. - 63 - ----- (1988): "La Deuda Externa y el problema de la Inversion en America Latina: Situacion actual y opciones," mimeo, Santiago-Chile. ------ (1990a): "Inflation and the Costs of Stabilization. Historical and Recent Experiences and Policy Lessons," in the World Bank Research Observer, July, Vol. 5 # 2. ------ (1990b): "How Private Investment Reacts to Changing Macroeconomic Conditions. The Chilean Experience in the 1980s." WPS # 212, The Work Bank. Sundararajan, A. (1987): "The Debt-equity Ratio of Firms and the Effectiveness of Interest Rate Policy," IMF Staff Papers. Tobin, J. (1982): "Money and Finance in the Macroeconomic Process," Journal of Money, Credit and Banking, Nobel Lecture, Vol. 14, 2. Tornell, A. (1989): "Real vs. Financial Investment: Towards an Explanation of 'Short-Termism'," mimeo, Columbia University. van Wijnbergen, S. (1982): "Stagflationary Effects of Monetary Stabilization Policies: A Quantitative Analysis," Journal of Development Economics. ------ (1983a): "Credit Policy, Inflation and Growth in a Financially Repressed Economy," Journal of Development Economics, 13. ------ (1983b): "Interest Rate Management in LDC's" Journal of Monetary Economics, 12. ------ (1985): "Trade Reform, Aggregate Investment and Capital flight. On credibility and the value of information," Economic Letters 17. World Bank (1988): World Development Report 1988. pRF Working Pager Saries Contact flw Au tor 1pape WPS582 Adjustment Programs and Bank Vittorio Corbo January 1991 L. Oropesa Support: Rationale and Main Stanley Fischer 39075 Results WPS583 World Bank Lending for Education Marlaine E. Lockheed January 1991 C. Cristobal Research, 1982-89 Alastair G. Rodd 33640 WPS584 Whiher Hungary and the European Alfrod Tovias January 1991 G. lbgon Communities? Sam Laird 33732 WPS585 Financial Innovation and Money Patricio Arrau January 1991 S. King-Watson Demand: Theory and Empirical Jos6 De Gregorio 31047 Implementation WPS586 The Challenging Arithmetic of Poverty Martin Ravallion February 1991 WDR Office in Bangladesh 31393 WPS587 Ouantifying the Magnitude and Martin Ravallion February 1991 WDR Office Severity of Absolute Poverty in the Gaurav Datt 31393 Developing World in the Mid-1980s Dominique van de Walle Elaine Chan WPS588 Obstacles to Developing Small and Brian Levy February 1991 E. Madrona Medium-Sized Enterprises: An 37496 Empirical Assessment WPS589 To Prescribe or Not to Prescribe: Jeffrey S. Hammer February 1991 0. Nadora On the Regulation of Pharmaceuticals 31091 in Developing Countries WPS590 The Domestic Financial Market and Premachandra Athukorala February 1991 M. Kienzle the Trade Liberalization Outcome: Sarath Rajapatirana 30733 The Evidence from Sri Lanka WPS591 Global Indicators of Nutritional Risk Rae Galloway February 1991 0. Nadora 31091 WPS592 Official Credits to Developing Asli DemirgOcg-Kunt February 1991 S. King-Watson Countries: Implicit Transfers to the Harry Huizinga 33730 Banks WPS593 Risk Management in Sub-Saharan Stijn Claessens February 1991 J. Carroll Africa Ying Oian 33715 WPS594 Size Rationalization and Trade Mark J. Roberts February 1991 S. Fallon Exposure in Developing Countries James R. Tybout 37947 WPS595 Hungary: Financial Sector Reform Mario I. Blejer February 1991 Z. Sequis in a Socialist Economy Silvia B. Sagari 37665 PRF Working PaFgr Sarias Contact ili AuthorD for pager WPS596 The Mexican Sugar Industry: Brent Borrell February 1991 P. Kokila Problems and Prospects 33716 WPS597 Rent Sharing in the Multi-Fibre Refik Erzan February 1991 G. Ilogon Arrangement: Theory and Evidence Kala Krishna 33732 from U.S. Apparel Imports from Ling Hui Tan Hong Kong WPS598 Africa Region Population Projections: Patience W. Stephens February 1991 0. Nadora 1990-91 Edition Eduard Bos 31091 My T. Vu RodoHfo A. Bulatao WPS599 Asia Region Population Projections: Eduard Bos February 1991 0. Nadora 1990-91 Editon Patience W. Stephens 31091 My T. Vu Rodolfo A. Bulatao WPS600 Latin Arnerica and the Caribbean My T. Vu February 1991 0. Nadora Region Population Projections: Eduard Bos 091 1990-91 Edition Patience W. Stephens Rodolfo A. Bulatao WPS601 Europe, Middle East, and North Eduard Bos February 1991 0. Nadora Africa Region Population Projections: Patience W. Stephens 31091 1990-91 Edition My T. Vu RodoHfo A. Bulatao WPS602 Firm Output Adjustment to Trade Mark A. Dutz February 1991 0. Nadora Liberalization: Theory with 38009 Application to the Moroccan Experience WPS603 The Role of Officially Supported Asli DemirgOg-Kunt February 1991 G. logon Export Credits in Sub-Saharan Refik Erzan 33732 Africa's External Financing WPS604 Foreign Trade and Its Relation to Faezeh Foroutan February 1991 S. Fallon Competition and Productivity in 37942 Turkish Industry WPS605 Overview of Contractual Savings Dimitri Vittas March 1991 W. Pitayato- Savings Institutions Michael Skully nakarn 37666 WPS606 Adjustment Policies and Investment Luis Serven March 1991 E. Khine Performance in Developing Andr6s Solimano 39361 Countries: Theory, Country Experiences, and Policy Implications