Reforming Delivery of Urban Services in Developing Countries

Reforming Delivery of Urban Services

in Developing Countries

Evidence from a Case Study in India

Economic and Political WeeklyAugust 18, 20073405case study, first, the demography and the economic base ofLudhiana are described in Section IV, since they have implica-tions for consumption of various public services. Section Vcontains an analysis of whether there is a need for reformingservice delivery in the city. Finances and its relationship withservice delivery are also discussed here. Section VI explores anypotential bottlenecks to the reform process. Section VII analysespossible triggers of the reform process and any reform actions.The final Section VIII summarises the findings, implications forother cities and contains concluding remarks.IILiterature ReviewWhile the literature on urban service delivery is vast, onlyrelevant studies are summarised here. Given the importance ofwater supply, we find that a majority of studies on urban publicservices focus on water supply since it is of most importanceat the local level.A study by the University of Birmingham (1999) based on 35urban centres in India finds that private sector participation isunlikely to have a significant impact on delivery of public servicessuch as water supply in the medium term because of too manyvested interests in the existing institutional patterns. This researchsuggests that until there is demand for institutional developmentfrom municipalities, which in turn, is generated by demand forbetter service from customers, there can be no sustainable advancesin service delivery.Expansion and improvement of public services is essential forimproving the quality of life and productivity in all developingcountries. Fox and Edmiston (2000) find how some Africancountries have been diligent in expanding the infrastructurenecessary to provide public services but unfortunately, most havenot done a very good job of paying for them. They present acase study of water supply services in Egypt, which highlightsthe importance of user charges to enable expansion of coverageof water supply services, since taxes or intergovernmental rev-enues are less likely to provide a consistent funding source ascompeting uses can diminish available resources. Based on surveydata from Lagos, Reedy (1986) found that many participants didnot have access to public water supply because the supplyauthority could not afford to expand service delivery.There are several studies of urban service delivery in cities ofIndia. Sridhar, Mathur and Nandy (2006) examine whether thesolution to better public service delivery in India’s cities shouldbe to close cities to migration or correct the under-pricing ofservices. They estimate the marginal costs of providing a kilolitre of water supply in several of India’s cities to be in the rangeof Rs 2.77 to Rs 5.08 and find that a few large (million-plus)cities are indeed under-pricing their water, so closing the citiesto migration to firms or households might not be the solution.Rao and Agarwal (1991) estimate unit costs of five publicservices – administrative services, police services, primaryeducation, secondary education, and public health – and theirexpenditure needs in the Indian states. Based on their state-levelanalysis they find that in order to ensure even average levels ofthe services they examine, a sizeable increase in transfers topoorer states would be required.Zérah (2002) reports findings from a household survey inVijayawada (a city with 2001 population of 8,00,000 in AndhraPradesh). That survey indicated that 77 per cent of the city’shouseholds considered water to be cheap or very cheap. In theproject zone, household willingness to pay was, in fact, morethan one and a half the average times the existing charges,testifying for the sustainability of water charges as a means offinancing investment.Raju, Praveen and Anand (2004) study the resources andmanagement constraints in providing adequate and safe drinkingwater supplies in a medium sized city, Kolar in Karnataka. Theyfind that (a) though 66 per cent of the city’s households had pipedwater supply, 68 per cent of them had unauthorised connections,thus depriving the exchequer of its revenue; (b) the per capitasupply was only one-third of the urban water supply norms; and(c) the city supplies were based on groundwater, which was quitecontaminated because of which the households had to dependon more expensive (Rs 100-500 per month) sources, while thepublic water fee was just Rs 45 per house per month. That studydemonstrates how local government apathy often creates a non-optimal situation for the public in the case of water.For the purposes of this paper, studies that detail efforts toimprove service delivery in urban areas, and the possible reasonsfor their success or failure, focusing on the role of the availabilityof finance, would be relevant. Savage and Dasgupta (2006) isone such study, which demonstrates how more revenue andinfrastructure do not mean better services, in the case of Ban-galore. Their data from the Water and Sanitation Programme(2005) show significant revenue increases (accounting for in-flation and population) during 1995-2003 for various serviceproviders (Bangalore Mahanagara Palike, Bangalore Water Supplyand Sewerage Board (BWSSB), and Bangalore DevelopmentAuthority (BDA)). However even with this, service outcomeshave declined. They point out that there is a need for a city-wideset of reforms that not only entail revenue but also ensure effectiveexpenditure management, new management approaches and theneed to focus on specific outcomes.Summarising, the literature reviewed here shows the need forgovernment involvement and citizen participation in overseeingthe provision of basic services, and that privatisation is not apanacea. The existing studies also confirm that user chargesenable the provision of a sustainable level of public services suchas water supply, taking into account the investment needs re-quired in developing countries to expand coverage.This case study adds to this literature by focusing on the issueof reforming local public services, and examines the bottlenecksand triggers of such reform, which have not been explored muchin the literature, despite the wave of decentralisation that hasswept across the globe. The answers to these questions haveimplications that can serve as input for service delivery reformin cities in many developing countries that are reeling under thenegative impact of poor public services and are similar to Ludhianain many substantive senses.IIIStudy HypothesesThere are certain hypotheses with respect to service deliveryand finances that we might a priori test.The first hypothesis follows from the literature we have re-viewed. Cities in which user charges reflect the unit or marginalcosts of providing services are able to deliver a more adequatelevel of the service, when compared to cities where such costsare not covered. This hypothesis assumes that finances (both

Economic and Political WeeklyAugust 18, 20073406capital and revenue expenditures) are important in service de-livery outcomes.Second, irrespective of whether finances are important inservice delivery outcomes, it is likely that other factors such asinstitutional arrangements are likely to impact service delivery(as both bottlenecks and triggers to reform), in the context ofcountries with multiple tiers of government.An examination of these hypotheses will enable us to dispelmyths about service delivery and finances. If it were found thatfinances do not affect service delivery, further examinationshouldenable us to focus on other aspects of service delivery.Alternatively, even if finances are important in determiningservice delivery outcomes, the hypotheses will be able tothrowlighton other factors that might be equally importantinservice delivery. Thus, the case study makes an attempt tooffera holistic view of reforms in the delivery of importanturbanservices, in the context of growing and emergingeconomies.Since there is usually a multiplicity of meanings attached tothe word “reform”, here, it is defined as a change for the better,in the form of improvements by correction of defects and errorsin local public service delivery.IVCase Study of LudhianaIn the case study, the economic base and demography of thecity are described since they have direct implications for theconsumption and needs of various public services.Demography and Economic BaseLudhiana, the largest city in Punjab is located 300 kilometresnorth-west of New Delhi. It is strategically located in the middleof Punjab, and is a corridor between Punjab and the rest of thecountry for the transport of agricultural and industrial products.The city is well known for its industrial growth and is calledthe Manchester of India. There is considerable industrial activityin the city consisting of cycles, machine tools including sewingmachinery, auto spare parts, and hosiery units. Ludhiana producesa large number of bicycles with Hero and Avon being the mostpopular brands. In hosiery, the city is exporter to Russia, Europe,and other parts of the world.Ludhiana is one of the three municipal corporations in Punjabwith its 2001 census population of 13,98,467. In 1991, theLudhiana Municipal Corporation (LMC) covered a total area of135 square kilometres yielding a 1991 density of 7,743 per squarekilometre, and a 2001 population density of 8,775 per squarekilometre (for an area of 159 square kilometres). The populationdensity of the MC increased from roughly 5,519 per squarekilometre in 1981 to 7,743 in 1991. Along with the rapid increasein its population and in-migration, the city has also witnessedtremendous expansion in its area recently. Increases in spatialexpansion, as with increases in population, mean that the pro-vision of various services such as water supply, sewerage andsanitation have to be stepped up, even for maintaining a constantlevel of these services.The population density of Ludhiana (district) for 2000-01, 804persons per square kilometre, was also the highest of all districtsin the state of Punjab, when compared to the state average of482.3 per square kilometre. This shows the need for LMC tooptimise city efficiencies with respect to the delivery of all localpublic services.The annual growth rate of Ludhiana (the city)’s population over1991-2001 was 2.9 per cent, higher than India’s national averagefor the growth rate of urban population over this period, 2.7 percent. This somewhat supports the notion that we are looking ata relatively well-performing city, since population is usuallyattracted by what it perceives to be better economic opportunityin any given area.According to the 1991 census of India, Ludhiana’s occupa-tional structure has a strong manufacturing base. Table 1summarises Ludhiana (district)’s occupational structure from the1991 Census. This shows nearly one-fourth of Ludhiana’s labourforce concentrated in non-household based manufacturing fol-lowed by cultivation and agricultural labour. Of course, itsoccupational structure may have changed now but more recentcorresponding census data are not yet available.Consistent with the census picture, we find other recent datawhich show a large number of small, medium and large firmsin Ludhiana (district)1 that contribute to total employment andoutput. Table 2 shows this for the more recent years – 1996 and2003. While the data shown in Table 2 are cumulative, theyindicate that small firms in the district accounted for nearly all(99 per cent of) the firms established during 1996 and 2003. Thesesmall units accounted for 85 per cent of all employment createdin 1996, and 78 per cent of jobs in 2003. They accounted,however, for a little greater than 40 per cent of total investmentand production in 1996 and only 22 per cent of investment in2003, consistent with the well known fact that small units arelabour-intensive rather than capital-intensive.The city’s economic base has implications for various civicservices. While the hosiery industry is water-intensive (dyeing),it also is quite polluting. Further, basic knowledge of the inputsand processes of the bicycle, sewing machine and auto spare partsindustry demonstrate the need for disposal of waste rubber andscrap metal.VNeed for Reform in Service DeliveryThe need, if any, for reforming service delivery in Ludhianawas assessed on the basis of (a) the actual state of service deliveryfor important services such as water supply and sewerage, com-pared with the recommended standards for these services;(b)systemic inefficiencies such as the high levels of unaccountedfor water; and (c) long-term irreversible consequences of businessas usual, such as contamination of groundwater.Service DeliveryWater SupplyThe main source of water supply in Ludhiana is undergroundwater, which is drawn through shallow and deep tube wells. Theinstitutional arrangement is that while the Punjab Water Supplyand Sewerage Board (PWSSB), which is a parastatal agency,undertakes the planning, development and construction of majorwater and sewer networks, the operation, maintenance and theexecution of minor developmental works is provided by the LMC.The service level statistics for water supply in Ludhiana aresummarised in Table 3. The city has the infrastructure capacity

Economic and Political WeeklyAugust 18, 20073407to provide water to the entire population but, overall, only 80per cent have access to piped water supply (Table3). As of1998,only 60 per cent of the city’s population was covered bymunicipalwater supply.2None of the private houses have metered water connections,which has implications for water usage. As there is no volumetricconsumptive tariff regime, the customers in these locales aregenerally not concerned about water loss.When the access of households to water supply in urban Punjaband urban India is compared with that of Ludhiana, the city hasabove average access, called for by its higher population density.The data in Table 4 for urban Punjab, urban India and LMC,from India’s census, 2001, must have been actual data from 1998or 1999. Data from the PWSSB in Table 3 are local, and recent,from 2003. So while census data show a 69.5 per cent tap watersupply access in the LMC, more recent data from PWSSB showan 80 per cent access of population to piped water supply in theLMC. The actual coverage of the city’s population with pipedwater supply is likely to be in the range of 65-70 per cent.Table 5 shows the compilation of facts as they pertain to watersupply. According to the PWSSB, actual average daily produc-tion of water in LMC is 284 MLD (million litres daily), out ofwhich only 140.5 MLD is charged. Only 137 MLD is consumedthrough house connections (which are not metered) in the city.3The137 MLD accounts for a mere 98 LPCD (litres per capita daily),which implies water supply of only about 8-10 hours (Table 5)a day. The lack of continuous supply is due to the fact that groundwater has to be pumped up through tube-wells and power is aconstraint for this. Further, there could be seasonal variationsin the supply, for which data were not available from the PWSSB.While being above average when compared to the state andthat for the country as a whole, the level of water supply isinadequate in the LMC, when compared to desirable benchmarks.The ideal norms for water supply vary greatly across cities andare based on considerations of industrial use, public use, firedemand and losses in transmission, according to the oldestcommittee on urban services, the Zakaria Committee (1963). Itrecommends a service level target of 150 LPCD for piped watersupply with sewerage (70 LPCD without sewerage), which includesroughly 20 per cent wastage of water.The actual supply of water consumed in LMC being only 98LPCD is highly inadequate, taking into account the national idealsupply standard of 135 LPCD or 175 LPCD, accounting fordistribution losses, recommended by India’s National Commis-sion on Urbanisation or the 150 LPCD recommended by theZakaria Committee for cities with piped water supply systemswith sewerage.SewerageWhile the Zakaria Committee recommends that for large urbancentres (such as LMC), the service level target should be fullcoverage by sewerage with treatment, a formal sewerage systemexists in only half the city’s area. Sewage treatment plants arenon-existent in the city, with the collected sewage being dis-charged into the local stream, or the nearest water body. Thisleads to contamination of the water supply network and alsopollutes the aquifer, which serves as a source of the city’s water.4Table 6 summarises the level of service for sewerage. It maybe noted that only 55 per cent of the city’s population have accessto a sewerage connection, much less than the coverage for watersupply. At present only 10 per cent of the city is covered by stormwater drainage facilities. Given the Zakaria Committee’s recom-mendation of 100 per cent coverage of city area, the extent ofsewerage access in the LMC is also highly inadequate.FinancesWhile the state of the various services emphasises the needforreform, we test the various hypotheses raised at the beginningof the case study, based on a study of Ludhiana’s finances.Table 3: Service Level Statistics for Water Supply,Ludhiana, 2003Description of ServiceNumberor PercentagePopulation as per 20011440000Population in 20031517000Total number of houses (2003)316042Average number of people per household4.8Population having access to piped water supply1213600In percentage80 per centPopulation served by household connections658000Population served by stand-posts18000Population served by independent institutional set up(Engineering colleges, universities, commercial andindustrial establishments)72000Total population served748000Total number of private houses with metered connection0Total number of houses with unmetered connection137071 (21 per centof population withhouseholdconnections,658,000)Total number of house connections137071Total number of commercial connections25663Number of stand posts350Number of tube-wells347Number of overhead reservoirs and capacity in MLD63 (38 MLD)Total length of distribution line (mains andsub-mains) in kms1467Source:Punjab Water Supply and Sewerage Board (PWSSB), 2003.Table 2: Industry in Ludhiana (District)Small Scale Units*Medium and Large Units1996200319962003Number of units26,44042,70488168Investment (million Rs)4,919.5211,168.646,958.2439,045.92Employment1,93,2202,73,59335,00075,185Production (million Rs)26,414.728,2213.634,968.48NANote:*According to the ministry of small-scale industries, government ofIndia, a small-scale industrial unit is an industrial undertaking in whichinvestment in fixed assets in plant and machinery does not exceedRs10 million.Source:District Industries Centre, Ludhiana.Table 1: Ludhiana (District)’s Occupational StructureOccupation CategoryEmployment (in Per Cent)Cultivation20.00Agricultural labour16.67Livestock1.44Mining0.00Household-based Manufacturing0.48Manufacturing, non-household-based25.29Construction3.80Trade and commerce13.63Transport and Communications5.27Other services13.41Source:1991 Census of India.

1212

Economic and Political WeeklyAugust 18, 20073409water receipts covered only 50 per cent a few years ago (in 2001-02)but has been declining since. The coverage is much better withrevenue expenditure since there have been instances when waterreceipts (from user charges) have covered more than 90 per cent,but since 2000-01, that has also been declining. The water tariffsfor unmetered water connections from 2003-04 were revised(Table 9), and revenues started increasing. The water tariffsapplicable for 1999-2003 are shown in Table 10.For the financial year 2002-03, the total production cost ofwater (in Rs per kilolitre) excluding capital expenditure wasRs2.06 (including distribution losses), with revenue income perkilolitre being only Rs 0.92. As of 2002-03, revenue incomecovered roughly only 45 per cent of the total production costof water (with or without distribution losses).The story with regard to sewerage receipts is even worse, sinceas of 2003-04, receipts from the service covered only one-thirdof total expenditure, and a little over half of the revenue expen-diture (Table 11). LMC’s sewer tariffs are shown in Table 12.Water and sewerage being basic services, their capital expen-ditures cannot be covered through user charges but for financiallysustainable service delivery, at least the O&M expendituresshould be covered. While our hypothesis implies that higherexpenditures should lead to better service delivery, in thisparticularcase, the decline in service levels of water supply and sewerageand a decline in expenditures seem to be moving together.Table 13 summarises the service level for sewerage in LMC.It shows declining levels of population access to sewerage since1999-2000.5A review of the LMC’s financial position thus indicates thatthere is no room for complacency. It shows that the need forservice delivery reform in Ludhiana is financial as well, withthe government of Punjab having abolished octroi in the state(effective from September 2006). While the city’s committedexpenditures are fairly low, and discretionary expenditures arehigh as a proportion of total expenditure, they do not indicatecommensurate increase in expenditure on developmental worksto increase the quantum or quality of public services.VIBottlenecks to ReformThe question arises as to why the financial and service deliveryperformance in Ludhiana has been less than satisfactory, whenjudged against national benchmarks set by India’s NationalCommission on Urbanisation or the Zakaria Committee, and whatare the bottlenecks to better service delivery and reform in thisand other similar cities in developing countries.Institutional Arrangements for Waterand Sewerage ServicesDespite the importance of financial resources for reformingservice delivery, consider the final hypothesis that it is likely thatother factors such as institutional arrangements impact servicedelivery. As a study by the Times Research Foundation (1997)found,there are 17 state-level pieces of legislation and 23 piecesof central legislation that affect the functioning of municipal areasin Punjab. Among many other things, we identified a statutoryoverlap of functions for planning, preparation and execution ofwater supply and sewerage schemes with the PWSSB (PWSSBAct, 1976) that also vests with the LMC (as per the PunjabMunicipal Corporation Act, 1976), as discussed earlier. Thisdistinction between “major” and “minor” works does not, how-ever, exist in the statute, making the classification of a projectas “major” or “minor” arbitrary causing the service to take a“passing the buck” attitude.Apart from the issue of statutory overlap, sometimes it isdesirable to have local autonomy in certain functional areas. Itis obvious that the LMC would be in a better position to judge,which localities are served better, and which ones would needinfrastructure upgradation or replacement.Lack of CapacityThe LMC, however, does not have the financial resources or,more importantly, the technical expertise to execute water supplyand sewerage projects independently (which is presumably whythe PWSSB is currently entrusted with undertaking work of acapital nature). The technical expertise is lacking, as the pro-portion of the city’s area covered by regular employment isTable 6: Service Level Statistics for Sewerage System,Ludhiana,2003Description of ServiceProportion/NumberCoveredPopulation access to sewerage in numbers834350In percentage55Population served through sewer house connectionin (numbers)585928Population served through soakage pits/septictanks (in numbers)849072Population served by independent institutional sewage(own systems in universities, collegesand so forth) (in numbers)72000Population served through conservancy system (in numbers)10000Actual number of sewer connections121985Number of institutions having independent sewerage system22Total average daily sewage flow in MLD292Length of outfall sewer (in kilometres)8Length of intercepting sewer (in kilometres)96Length of lateral/branch sewer (in kilometres)964Total length of sewer (in kilometres)1098Source: Punjab Water Supply and Sewerage Board, 2003.Table 7: Trends in Water Supply, Ludhiana Municipal CorporationDescription1996-971997-981998-991999-20002000-012001-022002-03Actual average daily production of water (MLD)193.05210.1234245247258284Total average daily water discharge capacity (MLD)360360390452484508454Daily water consumption through house connections (MLD)122.39133.2172184185131137Daily water consumption through handposts (MLD)12.7912.756.5633.53.5Total average daily water charged (MLD)135.14145.95178.5178182135140.5Percentage of water unaccounted for (per cent)3030.502425254750.50Daily water supply duration (in hours)12131213121010Source: Punjab Water Supply and Sewerage Board.

121212

Economic and Political WeeklyAugust 18, 20073411approachhas been adopted in city planning with the result thatservice quality is not monitored.Lack of Public ParticipationWhere accountability or public participation exists, there,servicedelivery is likely to be successful. The question is: domechanisms or forum exist for grievance redressal and informa-tion sharing in the city? There are no NGOs or consumer organi-sations that deal with public service delivery issues in the city.Compare this with drinking water as an area where Self-employedWomen’s Association (SEWA) women have taken theleadinGujarat. The outsourcing of tube-well operations in LMC is anexample of attempts to involve citizens in the delivery of publicservices. Although public services and their delivery/effective-ness are debated among the elitist groups of the city,thereis nosystematic discourse on the delivery of various services.The media can do much to disseminate information and improvelocal government performance in public service delivery. Despitethe city’s above average literacy, it was found that the totalnewspaper circulation and readership in Ludhiana (of 65,700)covered only one-fourth of the city’s households in 2001. Giventhe apparent lack of information to a substantial portion of thecity’s households,it is not clear what the incentives are for thepolicymaker cum service provider (the LMC/PWSSB) to beaccountable to the public.Currently, a major problem with the LMC functioning is thelack of public participation or discourse in budgeting. Certainly,better information to the public and their participation in theprovision of various services is likely to act as a rapid triggerfor reform. Some initiatives here could be to enable the web-presence of LMC budgets, its priorities and public discourseregarding the various problems associated with service delivery.VIITriggers for Reform and Reform ActionsCrises trigger reforms, as in the case of the balance of paymentscrisis of 1991 that made India take to the path of economic andpolitical reform. There are some triggers that could make thereform happen in the LMC.Abolition of Octroi and Financial ResourcesPre-existing financial resources are a crucial factor that deter-mine the internal readiness for reforming public service deliveryin most cities. A city with a large financial base with buoyantrevenue sources (such as an appropriately designed property tax)need not reform its service delivery, as it is presumably in aposition to offer an acceptable/adequate level of services, judgedagainst benchmarks.Heavily as the LMC’s finances are dependent on octroi, thegovernment of Punjab abolished octroi in the state, effectiveSeptember 2006. Given the fact that octroi accounts for morethan 70 per cent of the LMC’s own source revenues, surely theabolition will come as a trigger to persuade LMC to be morefinancially responsible. Some ways of making the LMC morefinancially responsible would be to bring into the tax net thoseresidential properties that have remained exempt.6 Further,discussions indicated that there are frequent under-assessmentsin the city of taxed (commercial or business) property that rangeanywhere from 2-20 times of their reported value.While not possible completely, one way to compensate foroctroi abolition could be to increase water charges, which couldbe used to finance the setting up of a sewage treatment plant(in addition to the initiatives envisaged under the SAP), whichcould appeal to the public because of the questionable waterquality in the LMC area.With the abolition of octroi, the state government of Punjabwill most likely come up with a compensation package for thelocal government. While our financial analysis indicates thattransfers have not been important in the LMC context, withoctroiabolition, they could well become very important in thetransition period.If this were to happen, octroi abolition mightnot act as a trigger to make LMC more financially accountableand reform service delivery.Once the octroi abolition trigger occurs, the reform actions thatneed to be taken could be summarised as follows:Privatisation in Service DeliveryAs is clear, the LMC (or the PWSSB, respectively beingpublicand para-public entities) is the lead service provider ofpublic services in most instances. The Ludhiana case demon-strates a different variant of the client-policymaker-serviceproviderrelationship highlighted by World DevelopmentTable 13: Sewerage Access Over Time, LMCDescription1996-971997-981998-991999-20002000-012001-022002-03Population access to sewerage (in numbers)73100077500010000001050000720000740000834350Population served through sewer house connections (in numbers)437500464000680000724500649000584860585928Population served through soakage pits/septic tanks (in numbers)130000135000438000306000715000717000849072Population served by independent institutional sewerage (in numbers)600006500070000150000700007200072000Population served through conservancy system (in numbers)100000100000250000100000100001000010000Actual number of sewer connections700007100075000105000110000116972121985Number of institutions having independent sewerage system20222222222222Total average daily sewerage flow in MLD300310351400400267292Length of intercepting sewer (in kilometres)73.773.775779310196Length of lateral/branch sewer (in kilometres)569.28710724743797903964Total length of sewer (in kilometres)642.98783.779982089010041098Source: Punjab Water Supply and Sewerage Board.Table 12: Sewerage Charges, LMCDiameter ofDomestic (in Rs PerCommercial (in Rs PerSewer PipeConnection Per Month)Connection Per Month)4 inches501006 inches112.502258 inches200400Source: Punjab Department of Local Government.

[li