~~~~~~~~~~U w- -u A joint publication by the Development Research Gr-oup and the Poverty Reduction Group of the World Bank No. 8 September 1997 Investing in Infrastructure: . In the ce of _nigation untargeted investmen program A Growth Strategy that would produce gains for the poor; bui; a program targeting house- Favors the Poor? holds with the smallest per capita land holiffngs would produce the mation fromn a nationally-representative most pro-poo distriution of H ow Large are gains from basic sample of 4,800 households. The house- benefits. nifrastructure investraents, and to hold survey collects data on access to Cniin fBscIfa - 2 s 2 1 E -X w-_ E w_ Conditios of 5 Bsi Inr- whtextent are they beneficial to infrastructural fadilities (electricity, potable the poor? Researchers variously argue water, etc.). A cmmunty question- structure in Viet Nam that such investments can reduce poverty. naue-conducted, only in rural areas- Viet Namn's current physical infr~astruc- It may be thiat the poor have least access to comiplements the household-level data wit ture is uinapeate. In rural areas, nearly a finfatructure and so will benefit most information on shared resources at thi lrd of the population live in communures from new" investments; alteratively the "commune" level, such as schiools, health wWou a passable road. Nearly half do poor may be concentrated in seomrs of the clincs, and irrigation services, not bave access to passenger transport, econlomy wher rates of return to mnfra- The studly first describes the pattem of an moed o have electnicity. Barely structure investmets are highest. access to basic infrastructure in Viet Nant afo ma rpLv sirgtd n -~~~~~~~~~~~~~~~~~~~~hl fana cro lnd i* _ s iriatd and_ -M A recent World Bank study-drawing It then focuses on one category of mfa 95% of the rural population live mn com- on the 1992-93 Viet Nam Living Stan- structural investment-irrigafion services mumes wWitout access topipe water. In dards Suirvey (VNLSS)-examnines the -and assesses its potential mlpact on geea,tepo en ohv os c extent to which empirical evidence from poverty. Two main policy conclusions gcera to epo tm tiiendn to ae nonseoor, Viet Nanm supports these argunmets. The emnerge: bu nrrlaesacess to mifwrcmtany types pofr study is part of a wider effort in the World * Because basic infrastructure serv inftastruct reais badcs for both(sy tabl at Bank's Development Research Group to ices are inadequatte for al income Bloa uNderstard how pubic speding polcies groups in Vet Nam, generalized in- Gen- d i affect well-being at the household level, vesent in public infrastructure Gv popcore wtee arefi in arness particullrly among the, poor. projects will not automawticly tar- thiiastrpublic r hesore direce towgardnes The 1992-93 VNLSS provides infor- get thaepoor. swomcubepasrionuces i tefe thab rds th-buted, much less well-targeted to the Access to Basic Infrastiructure In Rural Viet Nam pr Rural North Rural SouthW lI eing in Irodga- -Infratuctre Ty-Pmeo Non-Poor Poor Non-Poor Poor Passable Road 89.6 70.4 56.5 60.0 tion Sevices BenAefit the Passeniger Transport 53.4 44.1 59.7 63.2 Poor? Electiiiry 68.1 49.8 21.7 18.8 In andtich ways does this geaietozed lack Pipe-borne Water 6.2 2.2 8.4 7.8 Of infrastructure affect the distribution of Lower Secondary School 93.2 89.3 81.0 84.9 gains from furture investmentsa Not sur- Upper Secondary School 9.6 9.2 12.0 9.1 psingl, the answer depends on how the Clinic 97.1 92.3 89.3 95.0 marginao benefits from individual infra- Fio es rewivpercen ofralpopulation tiheniTi commnunes nAh ace" to tlis eifra,ure, s r are distributed For example, m te case of irrigation, if a Te simulated results indcate that em housded's land is alrady filly irrigated, an undfferniated expansion of inigation Tagtn, z.r., un th clearly the maal beefits to that (sunulation 1) would be redistributive- prgt hus household of exanding the total igt havin hgher proportonate gans to: .. . ara are zero. If it were true that the rich poorerhouseholds. With t ss had f*Uy irgated land while the poor did However, tgetng the irrigation expan- not, then clearly the greatest benei of an sion to households with the smallest per ..- exanon m irgation services would go capita lnd holdings (sinuation 4) pro- p t to the poor. duces the most progressive mcidcene of rg i distrib.tion ln Viet Nan, however, the VNLSS sur- gains, as well as the Larget absolute vey resuls inicate that the non-poor have benfits to the poor. Under this s ifglina more of both inigated and non-irried tion, the poorest households receive a net land. At first glance it would seerm, then, annual gai of 4.5% of total household Determining how investent benefits that an galtt irr on expan- expeditures, whie the gam to the wealhi- will be distributed depends on controlling sion would not be an inportant estis only 0.1% ofheirtotal expendiures. for fictors thatinfemarginal returns. redistnibutive instrument the poor would And, the magntude of the gain to the Simulated results for te specfic case of benefit, but probably less so than the non- poorest households is larger than that for irtion indicate that even an across-te- poor. any other group of households. board increase in irrigation servmces would However, a nunber of oher importat All siwlions show the highest totl produce income gains for the poor. But a fctors may influence the mrginal bene- impact on net crop incomes would occur progran specifically targeted to house- fits of inigation. It is often argued, for in Viet Nam's two poorest regions (the holds with low per capta landholdings example, that small fans are more pro- Northern Uplands and the North Coast) woud have the mnost progressive results. ducive; since the poor tend to have small where the outcmes also show the most And, fine-tning targeang to highly ii- Thus, their nmginal benefits may be pro-poor distnrbution. poveiished regions of the cony would higher. Education level, fhinily size, and The reults also indicate that education fiuther increase the prgese distribu- land quality mnay also determine produc- conrbutes significantly to agnculural tion ofbenefits. tivity, and hence crop income. productivity inm VietNam, interaction ef- Finally, important conplmarities Using a model of household crop in- fects between education variables ancl land exist betwen difrent infiastructural in- comes, the study simulates the impact of a are generally positive, notably so for pri- vestents in terns of their capacity to 10 percent expansion in irigated land area mary schooling, reduce poverty. Given the postive inter- for four aenative targeting strategies. action beween education and agricultural Simulation 1 simply extnds imation to Summary productvty found in the case of Viet alli Tn households that bave non-irrigatd Re . e fr V N . . Nam, the marg impact of iigation land Simulation 2 limits the expansion to cte thtdeficit i aces t basi projects is liely to be enhancmd by simul- frrms curety without access to aey irri- . . taneous investments in education. gatd lnd.Simlaton3 trges irigtio ifrastructure are so widespread that new on the basis of low total household la investmets will not automatically be wel smw_ van de Walle, Dcininique. 1995. Infwa. holdtngs. Sirnulation 4 dhoulb d ,,nga tageted to the poor In fact in Viet NTam sbrue and Powny in rid Non. Pubhic *lio Sisuof l pericpta - there is amnple scope for the non-poor to EC(CS Dfhvi= Poliy Rese:h Depamet hodi on e basLs of low per capzta land capture the lion's share of direct gains WodBak Wasingkxm,DC.46pp. from such investmrents. Poverl Lines is prepared by Gillette Hall with assistance from Yim Shafer This issue was edited by Brian tCulane. For information orfora subsaiption, contacts LSMS Administrator (AttU Poverty Lines) VLst our World Wide Web sites: World Bank, N8-058 Poverty Group Home Page: httpiho.worldbankofglhcvppoverty/centmnts.fl 1818 H Street, N.W. LSMS Home Page: hbtpiJwwworldbk.orgatlprdrdphMsmsllsnshome.html Washington, D.C. 20433 E-mail addrs: Ismsrwoirdbank.org Fax: (202) 522-1153