Report No. 37324-BR Brazil São Paulo: Inputs for a Sustainable Competitive City Strategy (In Two Volumes) Volume II: Background Report March 10, 2007 Finance, Private Sector and Infrastructure Management Unit Latin America and the Caribbean Region Document of the World Bank ACKNOWLEDGEMENTS The "SBo Paulo: Inputs for a Sustainable Competitive City Strategy" report was prepared under the Task Management of Jose Guilherme Reis and Maria Emilia Freire based on contributions from a team of Bank staffandconsultants, namedbelow. Volume I 0 Jose GuilhermeReis. Volume I1 0 Chapter 1: Tito Yepes; 0 Chapter 2: FernandoBlanco; 0 Chapter 3: Jose GuilhermeReis, MariamDayoub andMonicaRomis; 0 Chapter 4: David le Blanc; 0 Chapter 5: Tito Yepes, Andre Herzogand MonicaRomis; 0 Chapter 6: Jose GuilhermeReis. Natalie Palugyai contributed to the editing of the report. The team would like to thank the peer reviewers Ellen Hamilton, Christine Fallert Kessides, and Jose Luis Guasch for providing important and insightful comments. The report also benefitedfrom helpful inputs from Makhtar Diop, Jack Stein, Susan Goldmark, Marianne Fay, Ethan Weisman, Jennifer Sara, Martin Gambrill, BerniceK.Van Bronkhorst,Jorge Rebelo, and DanielHoornweg. The backgroundpapers preparedfor this report includeUrani, A. "Novas Estruturas Institucionais para a GovernanGa da RegiBo Metropolitana de SBo Paulo"; and, Biderman, C., and T. Yepes "Returns to Scale and Locationof Economic Activities in Brazil." The team acknowledges the important contributionsreceived from government officials in Sgo Paulo, especially from FranciscoLuna, Planning Secretary, and his successor Manuelito Pereira Magalhges Jhior and from Gerald0 Biasoto, Vice-president of EMURB. Javier Toro, from the PlanningSecretariat, providedinvaluablesupport. The team also benefitedfrom insightsprovided in an international meeting organized by the World Bank in SBo Paulo in April 2006 and from comments received in a meetingorganized by the PlanningSecretariat in SBo Paulo in December 2006. The findings and views expressed here are exclusively those of the World Bank and neither representsthe views ofthe government of SBo Paulo nor ofthe federal government. S i 0 Paulo: Inputs for a SustainableCompetitiveCity Strategy Table of Contents Introduction: Settingthe Stage for the MetropolitanRegionof Slo Paulo's RecoveryStrategy...................6 1.UnderlyingFactors of the EconomicTransitioninthe MRSP.................................................. 10 1.1. RecentTrends ............................................................................................................... 10 Creationinthe MRSP ............................................................................................................... 1.2. Factors Underlyingthe DisappointingPerformanceof EconomicGrowthandJob 14 2. ImprovingFiscal Performance andCreditworthiness.......................................................................... 22 2.1. The OverallFiscal Situationofthe MRSP................................................................... 22 2.2. The Fiscal Situation inthe Municipality of Ssio Paulo................................................. 23 3.3. FiscalAdjustment Scenarios: ProjectionExercises ..................................................... 26 2.4. Policy Challenges: Sustainabilityand Quality ofthe FiscalAdjustment ..................... 28 2.5. The Agenda for the Long-Term:A Note on FiscalRules, Debt, andAsymmetry.......30 3. ImprovingCompetitiveness and the Investment Climate.......................................................... 32 3.1, The Business Environmentin the MRSP..................................................................... 32 3.2. Some Aspects of the Investment Climate: ComparingMRSP's Competitivenesswith Other Regions ............................................................................................................................ 34 3.3. LocationFactors: Survey Results................................................................................. 39 3.4. .............................................................................. 3.5. Conclusions and Recommendations............................................................................. Determinants of Firm Performance 47 48 4 ImprovingServiceDelivery inLow IncomeHousing.............................................................................. . 50 4.1, UrbanGrowthinthe MRSP and its Implicationson Housing..................................... 50 4.2 The Main Actors Responsible for HousingPrograms inthe MSP ............................... 52 4.3. AddressingHousingDeficits: Understandingthe Actors, InterventionsandDifferences across Municipal Zones ............................................................................................................. 53 Concerns and the Needfor PreventionPolicies........................................................................ 4.4. InstitutionalChallenges Arising from this Scheme: Lack of Coordination, Financing 61 4.5. Targeting and Affordability of HousingSolutions....................................................... 64 4.6. Conclusions and Recommendations: Institutionaland Instruments............................. 65 5. ImprovingInstitutionalPartnerships ......................................................................................... 69 5.1 . Cities as Partners .......................................................................................................... 69 5.2. DevelopingCoordination: Evidence from the Transport Sector.................................. 71 5.3. CurrentChallenges and Opportunitiesfor ImprovingGovernance inthe MRSP ........73 6 PolicyAgenda: Recommendations for the Short- andMedium-Term . ...................................... 76 List ofAnnexes Annex 2: Importanceof LocationFactors:MeansTest Results.................................................... Annex 1: Survey Methodology...................................................................................................... 90 94 Annex 3: Location Determinants of Productivity .......................................................................... 95 Listof Tables Table 2. 1 FRL and financial indicators.2002 and 2004 (percent) . .............................................. 23 Table 2.2. SBo Paulo: Decompositionofthe increase ofnet consolidateddebt to net current Table 2.4. Si30 Paulo municipalgovernment: results of fiscal adjustment scenario Table 2. 3. SBo Paulo municipalgovernment: evolutionof expenditures structure, 2000-2005 ...25 revenues, 2001-2004 (percent) ...................................................................................................... 26 Table 3 Table 3.2. Means firms deal with crime and violence, 2005 (percent of firms per region)..........39 . 1.Doingbusiness ..................... 27 in Brazil: where is easiest? .................................................................. 33 Table 3 3 Top five locationfactors, by region . . ............................................................................. Table 3.4. Correlationbetween locationand its drivers:MRSPcompared with alternate Brazilian 42 regions ........................................................................................................................................... Table 3. 5 43 44 Table 3 .6. Fivemainadvantagesof . Top five locationfactors, by technology sector ........................................................... ........................................... Table 3 8.Top seven locationfactors: manufacturingand services sectors inthe MRSP Table 3. 7.Fivemain five disadvantagesof firm location, by region firm location,by region....................................................... 44 . ...........45 Table 3. 9 Locationdeterminants of total productivity factor . ...................................................... 47 Table 4 1.Slumupgrading:Unitsdelivered . ................................................................................. 48 Table 4. 2. Strengthening propertyrights (beneficiaries) 54 57 Table 4 3 MSP: Direct provisionof housing, 2001-2004 .. . .............................................................. ......................................................................... ........................................................... 59 Table 6. 1. Investmentclimate: Selectedrecommendations for improvingthe business Table 4 4. Support to the privatehousingmarket 60 environmentofthe MRSP .............................................................................................................. 83 Table 6.2. Service Delivery Strategy: Housing ............................................................................ 86 List ofFigures Figure 1. Brazil and selectedmetropolitanregions: unemploymentrate. 1992-2004(percent)......6 Figure2 Selectedmetropolitanregionsin Brazil: poor populationas a share of total population. . 1992-2004(percent) ........................................................................................................................ ............10 6 Figure 1. 2 Figure 1. 1. SBo Paulo state and MRSP: share of industry in nationalGDP, 1940-2000 .Brazil, Si30 Paulo state ....................................................................................................................................................... and MRSP: index of industrialGDP, 1970-2000(1970=100) Figure 1 3.Change in manufacturingGDP by city size, 1970-2000(percent) . ............................ 10 Figure 1. 4.Change in non-manufacturingGDP by city size, 1970-2000(percent) 12 12 Figure 1 .5. Average annual GDP per capitagrowth, 1970-2000(percent) .................................... ................................. ..................... 13 Figure 1.7. Populationby householdmonthly income, 1997and 2002 (million) Figure 1.6. Brazil and MRSP: unemploymentrate, 1992-2004(percent) ........................ 13 Figure 1. 8. Brazil: GDP and GDP per capitagrowth, 1900-2000(in percent) ............................. 14 17 Figure 1.9. Industry sector: changein shares invalue-added by skill level, 1991-2000(percent)l9 ...............19 Figure 1. 11. Services sector: change in shares in value-added by skill level, 1991-2000(percent) Figure 1. 10. Industry sector: change injob shares by skill level, 1991-2000(percent) ....................................................................................................................................................... 20 Figure2. 1. MRSP: aggregatedfiscal balances, 2000-2004(percent of currentrevenues).........23 Figure 1. 12. Services sector: change injob shares by skill level, 1991-2000(percent)..............20 Figure2.2. Si30 Paulo: fiscal balances, 2000-2005 (percent of net currentrevenues)..................25 revenues, 2005-2015 (percent) ...................................................................................................... Figure 2.3. SBo Paulo municipalgovernment: projectionson net consolidateddebthet current ............................................... 27 Figure3. 1.Timeto start a businessper state (days)...................................................................... Figure2.4. MRSPmunicipalities:tax collectionefficiency, 2002 29 Figure3. 2.Procedurecomplexity indexper state ........................................................................ 34 35 3 Figure3 3 Cost to enforcea contract per state (percent ofthe debt) ........................................... Figure3. 4 Average sales losses due to breakages duringtransportationand electricity outages. . .. 35 2002 (percent oftotal sales) Figure3. 5. Average share of workers with college degree, 2002 (percent of firm's labor force) 37 ........................................................................................................... 36 Figure3.6. Average share of workers who use computers, 2002 (percent of firm's labor force) 37 Figure3. 7. Average share of managerialtime spent dealing with government regulations, 2002 (percent) ......................................................................................................................................... 38 Figure3. 8. Government's efficiency inthe provisionof public services, 2002 (percent of firms) ....................................................................................................................................................... 38 Figure3.9.Average paymentsrelatedto security as a shareoftotal sales, 2002 (percent) .........38 Figure 3. 10.Average share of sales lost due to crime andviolence, 2002 and2005 (oftotal sales) ....................................................................................................................................................... Figure 3. 11.Selected metropolitanregions:Average share of sales lostdue to crime and 38 violence, 2002 and 2005 (of total sales) ........................................................................................ Figure 3. 12.Rankingof locationfactors: Percentageof firms that rank the factor as "extremely 39 ..................................................................................................... 40 Figure 3. 13.Investment climate versus Locationfactors important" or "very important" ............................................................. 41 Figure3. 14.Brazilianstates with best businessenvironment (percent) ....................................... Figure3. 15 Braziliancitieswith best business environment (percent) . ....................................... 46 Figure3. 16. Service sector survey: percentage of services firms that rankedfactor as "important" 46 ........................................................................................................................ Figure4. 1.Copingwith urban growththrough low-incomehousing or "very important" 46 .......................................... Figure5. 1.MRSP:Institutionalscheme for transport planning, operation, and overseeing........73 54 ListofBoxes ............................................................... Box 2. The FinancialWeakness of COHAB Box 1. The Raiseofthe BrazilianDevelopment Engine 13 ................................................................................. 53 Box 3 Estimatingthe Cost of UpgradingInformal Settlements................................................... . 57 Box 6.Cons6rcios Box 5.Inter-MunicipalityCooperation:A BrazilianExperience.................................................. Box 4.CDHU's Interventionsinthe MSP and SEio Paulo's Contributionsto the ICMS..............60 .......................................................................................................................... 74 70 ....................................................................................................... Box 8. The Case of Chicago Box 7. The Case of Barcelona .......................................................................................................... 77 79 List of Maps 15 Map 1. 3.Municipality of SEio Paulo:Inventoryof developments 1997-2002 Map 1. 2. Municipality of SEio Paulo:assessed landvalues, 2002 ................................................ Map 1. 1.Distributionofemployment .......................................................................................... 16 .............................. Map 4 1.Annual growth rate by censustrack inthe MRSP, 1991-2000(percent per year) . .......51 16 4 ABBREVIATIONS AND ACRONYMS BNH BancoNacional de HabitaqBo National HousingBank CDHU Companhiade DesenvolvimentoHabitacional e Housingand Urban DevelopmentCompany Urbano of the State of SBo Paulo CEF Caixa EconGmicaFederal Federal Savings and Loan Bank CEPAC Certificado de Potencial Construtivo Adicional Certificate of Potential Additional Construction COHAB CompanhiaMetropolitana de HabitaqBode SBo Paulo Metropolitan Housing Company of SBo Paulo CPMF ContribuiqLo Provis6ria sobre MovimentaqBo ou ProvisionalContributionon Financial Transmisslo de Valores e de CrCditos e Direitos de Transactions MovimentaqBoFinanceira EMPLASA Empresa Paulistade Planejamento Metropolitan0S.A. Slo Paulo's MetropolitanPlanning Company EMURB EmpresaMunicipal de UrbanizaqBo Municipal Urbanization Company FDI Investimento EstrangeiroDireto ForeignDirect Investment FGTS Fundo de Garantiapor Tempo de Serviqo Private Sector Severance Fund FMH Fundo Municipal de HabitaqBo Municipal Housing Fund FRL Lei de Responsabilidade Fiscal Fiscal Responsibility Law FUNDURB Fundo de Desenvolvimento Urbano Urban Development Fund GDP Produto Intemo Bruto Gross Domestic Product GIS Systemade InformaqBo Geografica GeographicInformation System IBGE Instituto Brasileiro de Geografiae Estatfstica Brazilian Institute for Geography and Statistics ICA Pesquisasobre o Clima de Investimentos Investiment Climate Assessment ICMS Imposto sobre CirculaqBo de Mercadoriase Serviqos State Sales Tax IDB Banco Interamericanode Desenvolvimento Inter-AmericanDevelopment Bank IETS Instituto de Estudos do Trabalho e Sociedade Labor and Social Studies Institute IFC CorporaqBo FinanceiraInternacional International Finance Corporation IPCA Indice de Preqos ao Consumidor Amplo Consumer Price Index IPEA !nstituto de PesquisaEconGmicaAplicada Institutefor Applied Economic Research IGP-DI !ndice Geral de Preqos - Disponibilidade Intema General Price Index -Domestic Availability IGP-M !ndice Geral de Preqos do Mercado General Market Price Index IPCA IndiceNacional de Preqos ao Consumidor Amplo Widen Consumer National Price Index IPTU Imposto Predial e Territorial Urbano Urban Property Tax ISS Imposto Sobre Servigos Tax on Services LDO Leide Diretrizes Orqamentfirias Budgetary Guideline Law LOA Lei Orqamenthria Anual Annual Budget Laws MRSP RegiBo Metropolitanade SBo Paul0 Metropolitan Region of Silo Paulo MSP Municipio de SBo Paulo Municipality of SBo Paulo M W Salfirio Minimo Minimum Wage PAR Programade Arrendamento Residencial Housing LeasingProgram PDE Plan0 Diretor EstratCgicodo Municipio de SBo Paulo Municipal Strategic Development Plan PMHP Programa Municipal de HabitaqBoPopular Municipal Popular HousingProgram PNAD PesquisaNacional por Amostra de Domicilios Brazilian Household Survey PPA Planos Plurianual Multi-Year Plan PPP Parceria Publico-Privada Public-Private Partnership PRIH Perfmetros de ReabilitaqaoIntegrada do Habitat Area of Habitat IntegratedRehabilitation PROVER Programa de Urbanizaqho e RegularizaqBo de Favelas Program for Slum Urbanization and Regularization PSH Programa de Subsidio ti HabitaqBo de Interesse Social Program for Subsidies to Social Interest Housing R&D Pesquisae Desenvolvimento Research and Development SBPE Sistema Brasileiro de Poupanqae EmprCstimo Brazilian Savings and Loan System SEADE FundaqBo Sistema Estadualde Analise de Dados Foundation State System for DataAnalysis SEHAB Secretaria da HabitaqBo e Desenvolvimento Urbano SBo Paulo Municipal Housing and Urban Development Department SEMPLA SecretariaMunicipal de Planejamento SBo Paulo Municipal Planning Department SMF Secretaria Municipal de Finanqas Municipal Finance Secretariat S T N Secretaria do Tesouro Nacional National Treasury Secretariat ZEIS Zonas Especiais de Interesse Social Zones of Special Social Interest WBG wss Grupo Banco Mundial World Bank Group ProvisBo de hguae saneamento basico Water supply and sanitation 5 1. INTRODUCTION: SETTINGTHE STAGE FORTHE METROPOLITANREGION OF SA0 PAULO'S RECOVERY STRATEGY 1. Slo Paulo, the largest metropolitan area in South America, faces the challenge of unleashing a processof recovery under severe fiscal constraints. As many other large cities in the world, the Metropolitan Region of Si30 Paulo (MRSP) has been going through a process of change in their economic structure, characterized by de-industrialization,partially compensated by a rise in the services sector. There are two particular aspects in the ongoing process in Municipality of Si30 Paulo (MSP). First, the decline in economic activity and the poor results obtained in some of the most important social indicators seemed to be particularly intense in the last 25 years. Second, the gravity of the fiscal problem is acute in the case of the MSP, which seems to impose severe limitsto the possibilitiesandalternatives of public policy intervention. 2. The problems faced by the MRSP are not unique in Brazil, since metropolitan regions have been the loci in which the impacts of slow growth and aggravation of social problems have been more acute. Metropolitanregions such as the MRSP were at the heart of earlier development strategies and have been suffering from mounting social problems in a context of limited financial resources. Brazil as a country is not going through a de- industrializationprocess.However, its metropolitanregions are changing their original vocations. This is particularly evident in the metropolitan regions of Si30 Paulo and Rio de Janeiro. Similarly, Brazil as a country is not loosing employment, but its metropolitan regions are, particularly the MRSP (Figure 1). Furthermore, while Brazil as a whole has seen a decline in the poverty rates, the trend in metropolitanregions has been an increase in the number of the poor; again, this was particularlyevident inthe MRSP (Figure 2). Figure1.Brazilandselectedmetropolitanregions: Figure2. SelectedmetropolitanregionsinBrazil: unemploymentrate, 1992-2004(percent) poorpopulationas a share oftotal population, 1992- 2004 (percent) l81 [+Bras11+Riode Janeiro+SboPaulo----Cuntiba1 1 351-I+ Belo + Rio de Janeiro +Sbo Paulo--Cuntiba 4 1 I I 10 -I I 1990 1992 1994 1996 1998 2000 2002 2004 2006 1990 1992 1994 19% 1998 2000 2002 2004 200 Source: Basedon PNAD/IBGE. Source: Basedon PNADABGE. 3. Slo Paulo's indicators have been more negative than other metropolitan regions within Brazil, which indicates that it is suffering from an array of agglomeration externalities.Rapid urbanizationcombined with sluggish economic opportunitiesand rising unemployment rates resulted in a metropolitan area plagued with social problems, criminality, and a diminished capacity to compete with other regions in the attraction of 6 private investments. Currently, SBo Paulo is the largest and the slowest growing city in Brazil (1.4 percent average per year of GDP growth).' Compared with the rest of metropolitan Brazil, SBo Paulo has witnessed above average unemployment rates. SBo Paulo's economic decline occurredin the last thirty years with major economic recessionsthat hit the manufacturingsector in the early 1980s and the early 1990s. More recently, the shrinking of SBo Paulo's economic base was particularly significant: accordingto Funduq6o Seade, the municipality of SBo Paulo's share in the state's GDP declined from 33 percentto 30 percent between 1999 and 2003, a period marked by the average low growth of the Brazilian economy. Currently, economic and quality of life indicatorsfor Sgo Paulo are consistently lower than those of other large cities in Brazil. Out of the 49 metropolitanregions in Brazil, the MRSP has the highest population density (2,245 inhabitantsper km2),the highest housingrental prices, the fourth highest share of the population livingin slums (9 percent), and the fifth highest share ofthe populationliving in informal housing (16 percent) (Cities Alliance, 2005). While some of the MRSP's problems fall outside the scope of municipalcontrol, there is muchroom for city-level improvements. 4. The municipality of Si30 Paulo fiscal crisis, aggravated in the 1980s and 1990s, has furthermore limited the ability of the administration to cope with municipal problems. SBo Paulo's successive budget deficits led to a large debt exposure at the end of 2005, when the net consolidateddebt reached223 percent of net currentrevenues, far above the Fiscal Responsibility Law (FRL) ceiling of 174 percentn2This deterioratingfiscal situation was driven by expenditure increases, specifically current expenditures. The concern is that the lack of fiscal capacity has hindered the quality of basic infrastructure, which is an essential factor in attracting private investments and competitive activities. Other arguments consider that the lack of economic growth has eroded the fiscal capacityof the municipality, impedingits ability to provide essential services and address the most pressing demands of the poor. Evidence suggests that some of the concerns regardingthe lack of adequate maintenanceof infrastructures are indeed an obstacle for growth and competitiveness.Inaddition, the lack ofjob opportunitieshas certainly contributedto the increase in the incidenceof crime and violence, magnifying the difficulties of establishinga businessenvironmentconduciveto growth inthe region. 5. I t is incorrect, however, to perceive SHo Paulo as a degraded area; SHo Paulo benefits from many advantages. Furthermore, it is a transforming metropolitan area that, with the right policies, can confront the shocks caused by economic transition. SBo Paulo has many assets: a favorable geographic position, it is a naturalhub for air transportation,it is close to the mainport ofthe country, and it has proximity to agriculturaland manufacturingproduction.It also has some of the most prestigious universities in the country, such as the University of SBo Paulo, which supplied the municipality with the highly qualified professionalsneeded to support the area's rapid growth period. The attractiveness of the city and the region is recognized by entrepreneurs, which regard the city of SBo Paulo as having the best business environment in Brazil and value its endowments as key factors for businessl~cation.~ 6. Through an analysis of selected topics, this study aims to offer inputs for a successful recovery strategy for the city and the Metropolitan Region of SHo Paulo. The study first presents an analysis of the underlying factors of the economic transition in the ' Ten million people in 2005 in the municipality of Slo Paulo, over 18 million people in the metropolitan area, one ofthe five largest cities inthe world, and where one out of every 10Brazilianslives. Complementary Law No. 101/2000. 3 Sfio Paulo was recently appointed as the best city in Latin America to do business (America Economia, 2006) and is frequently mentioned as a successfulcase of transformationfor other metropolitanareas (e.g., the recent study about Rio de Janeiro by McKinsey). 7 MRSP, highlightingthe factors behindthe recentperformanceof the MRSP in terms ofjob creation and growth. Then, four inputs that would lead to a `recovery strategy' for the MRSP were discussedin detail: `Improving FiscalPerformanceand Creditworthiness' assessed the current fiscal situation, revising the fiscal scenario, and the links of economic performance and fiscal spending. `Improving Competitiveness and the Investment Climate' analyzes the business environmentin the MRSP, usingbothdirect surveys and an econometric modelwith the aim of uncoveringthe main strengths and weaknesses of the region in terms of attracting new investments. `Improving Service Delivery in Low-Income Housing' carefully examines housing policies and programs, which, given the MRSP's current economic transition and fiscal constraints, are key services responsible for improving the livelihood and opportunities of the poor. Finally, `Improving InstitutionalPartnerships' discusses current and potential activitiesthat improve partnerships among municipalities in the MRSP as well as within the different levels of government andthe privatesector. 7. The key area in the recovery strategy is `Improving Fiscal Performance and Creditworthiness.' Chapter 2 provides an assessment of the current fiscal situation, revising the fiscal scenario, and the links between economic performance and fiscal spending. The focus of the chapter is the municipality of Siio Paulo, who has been suffering a long period of fiscal distress that led to a very high level of indebtedness. Projectionexercises for the evolution of fiscal accounts for the period 2005-2015 are presented. The main message of the chapter is that, notwithstandingthe challenges for the consolidation of the fiscal adjustment that is being implementedby the current administration,there are positive signs on the fiscal front, suggesting that the unavoidable adjustment to comply with the FiscalResponsibilityLaw is feasible and can be implementedwhile maintainingminimumlevels of investment expenditures. 8. A second area of focus within the recovery strategy is `Improving Competitiveness and the Investment Climate.' Chapter 3 assesses the business environment in the MRSP, using both direct surveys and econometric models, to describe the factors that influence perceptions of the business climate, factors that influence firm location within Brazil, and the differences between the manufacturing and services sectors with regards to firm location. Specifically, the study uses responses from a survey on 447 manufacturing firms, selected among the 1,642 firms interviewed for the Brazil Investment Climate Survey (ICs) carried out by the World Bank in 2003 and interviewedover the phone using a computer-assisted telephone interview system (the methodology is reported in Annex l), in order to assess those areas in which municipal governments must improve in order to raise competitiveness in the MRSP. 9. A third area of focus under the MRSP recovery strategy is `Improving Institutional Partnerships.' Chapter 4 explains that a confusing institutional framework, overlapping responsibilities,the lackof coordinationacross governmentlevels, and little interactionwith the private sector have hamperedthe MRSP's ability to react to economic and fiscal shocks. The municipality of Sgo Paulo shares basic problems and challenges with surrounding municipalities, including traffic congestion, densification, deficiencies in the provision of basic services, slow job creation, and serious shelter and housing shortages. However, policy and urbanplanningcoordinationdeficienciesbetweenthe Municipality of Sgo Paulo and the other 38 municipalities of the metropolitanregion, commonplace throughoutmetropolitanareas in Brazil, limit the MRSP's ability to solve common issue^.^ This chapter highlights the example of inter and intra-municipal transport, which due to inefficiencies is currently resulting in significant 4 See Chapter 5 on housingpolicies. 8 traffic congestion: over 100 km of congested roads at peak hours, low occupancy rates of 1.5 passengersper car, and 3,200,000 cars circulatingeachday (Jacobi, Segura, and Kjellen, 1999). 10. Finally, `Improving Service Delivery' is the fourth focus area within the MRSP recovery strategy. The challenges of service delivery are illustrated in chapter 5 through careful examination of housing policies and programs in the Municipality of SHo Paulo (MSP). The housing sector was chosen because of its importance in the municipal investment budget and the coordinationchallenges involved, as well as being an area in which policies can have high social impacts. In addition, housing as an example of municipal service delivery is particularly relevant for the MSP given the context described in chapter 2: loss of employment caused by de-industrialization, increases in the overall population and more specifically the proportion of the poor population, and fiscal restraints that limit the municipality's ability to provide all the public services needed under these circumstances. The chapter provides recommendations to improve the efficiency and targeting of public housing programs in the municipality of SBo Paulo (MSP). The proposed improvements are based on a consolidatedview of interventions of different government levels (Le., federal, state, and municipal) undertaken through variousprograms. 11. SHo Paulo is a global city and in an increasingly competitive world, Brazil needs prosperouscities to compete. Brazilian cities contributeto economic output, host the majority of the country's population (82 percent), and are the main recipients of foreign direct investment (FDI). Worldwide examples attest to the impact of good city management in enabling the emergence of new economic activities and expansion of innovation. In an increasingly competitiveworld, Brazil needsto harness the potentialof its urban and regional competitiveness and promote its cities to world class agents of growth and innovation.The difficulties faced by the SBo Paulo metropolitan region are complex, but through effective policies such as those emphasized in this report, negativeexternalities can be mitigated, institutional capacityenhanced, mobility increased, competitiveness improved,and adjustmentsto shocks facilitated. 9 1.1, RECE~TTRE?JDS 4 .... - 0 . . 2 ,o. _._I .... f FA--Q--I> 0 ..-.. ...Q. .Gr0At , -f - .?a.. - . . . . . . . . 0 .$C0 ..... .. ......... <' . . . . .- . . . . 0 " . . 19. The MR'RSP a trade-off between ~~ ~ ~and~ socialce x~~ ei n~d ~et ~Instead, the situation involves i u r ~ s . ~ ~ ~ r ~ ~ hat cannot firid stable jobs, therefore r ~ ~ ~ i ~ r j ~ ~ CFSS in order to keep j itants, '111~last ccns C f 1.1 14 FOR CUR CPW REG S& PA BH RJ 99 Source: Rased on IPEAdctta has heea weakening in I" rcasc in local dem kens the natural pro tors expected from the home market, 13 n-D n 22, There are ~ ~ ~c h ~nl l ein gIn~~urban sewicun~ ~ ~ ~ r~ o given the c~ ~ s n ~c ~ ~~ ~ r ~ t ~ ~ ~ of jobs, The MKSP has ti circular $ ~ ~ c ~uithr cits center located at the heart ofthe m ~ ~ ~ ~ ~ c ~ p a ~ i ~ of SBa Patilo. k4ap 1. I shows the ~ ~ ~ i n i ~ i pofaSBu~Paulti uut~jnedin red with the est -job u l i y density located in its ar pa tided center and ~ u ~ * u n dbyeareas of low job ~ j e ~ s iThe .Southcrn ~ t ~ part of tfic citj does riot represent a concern, sirico its popul~~~ondensity is low. However, the ern, Northern, and thc U'estcrn areas are densely p ~ despift.~the luu~a ~ ~ ~uf ~~ ~ ~ ~ ~~ jobs, The. p~)p~~lation these areas o o t ~ ~ tu ello~citysccnter, r e ~ r ~ s ~ natsi ~ ~ sgt a ~ ~ ~ ~ a l from ~ ~ e u c h a l l e ~for ~the ~ r ~ ~ s psystem ~ ~ ~ ~ i ~arid ~highero~n ~ a e ~x p~~ ns ~pfor ~~pcople~libing in ~ ~ ~ ~r s those arcas. In fact, the 2002 arigin-desrinatiati matrix for the MRSP shows that inore than 30 percent of the ~ ~ t ~daily trips that or~~~natedin p e ~ ~ ~ ht~iu~~icipal~ties MRSP tiad r j ~ e ~ ~ r a i of the ariotticr ~ u n ~ c i ~asatheir~ ~ ~ s ~ ~ t i a twhifet ~on15 10 percent of the trips ~ene~ated the l i ~ o s , in m~~nicipali~yof SBo Paul0 tiad dcsti~a~iun~outside of the city. Map 2.1 also shows that other urban areas, such as Campinas, which go beyond chc MRSP's p ~ r ~ p ~ e~r a ie~~~,thighnjob ~ a i densities. 23. Urban ad price dynamicsthr purposes in S%uParrlu's Cieq Hall. Ideally it could hate been ~~t~~~~ with census data in order tomodel tery reduced pricesand a Eo increases in land prices transla~einto increased walth for ho~~seh~lds m e r a time period who ~ r a ~ sthatoland into iricome ~enerat~oi~.The ~ ~ n i c i p ~ofl Sao ~ r ~ ~ $ yPaufo seems to have a very c ~ n ~ e n t ~ alande price dynamic, ~ ~ ~u n dceh~ ~ ~pos$~bil~ties~the city to ~cneratesveafth ~ d i i n e ~ f r and, therefore, s$ren~r~~ene ~ ~ i ~These o ~ s ~ ~ ~seemr~onsis~en~ Ghat is ~ ~ i s ~ ~ s ~ e ~ i its d n d . ~ i o ~ ~ with s above it1 Figure I.7. In a context of reduccdecono~icgrowth the conccn~ra~i~nwcatth in few of areas could explain the low income dynamic, P u b l ~policy cF~al~en~esare j ~ c r c givensthe ~ ~ ~ ~ ~ ~ ~ ~ ~ n c e n ~of~lowi oinconle ~ ~ p u ~ a ~(5%oiths increased social service. needs) rind the r ~ ~ i n c o s ~ s ~ q ~si e~na~ ~per~~capita fiscaf revenues. Tax revenues per capita grew w11y 0.8 percent ~i ~~ ~ n ~ per ~ n nin~real i ~ terms between I995 and 2004. 'This low growth was iiot c o ~ ~ ~ p eby~other~ ~ d ~ s a fiscaf revenues, s i x e total revenues grew 0.6 percent per a n ~ in real tcrrris oicr thc silnie u ~ period. Source: Bascdon City Hall Data. I S 23. The 1970s marked the ~ n ~ paint~far~the~process of c ~ n ~ ~ n t ~of ~economic e n a i o n activity in the MRSP, which s t a ~ t ~indthe late XIX century, Vast f~ter~~ure anal~~ingexists thc rcasons behind the large d~~pariti~sa r n o ~~ r~a~~ j lregionsnthat resufted in a rich South and a i ~ ~ poor North, Most e~~fana~~ons MRSP8s the center of iveafth concer~tra~ion,F i ~ ~ h e ~ i ~ ~ o r e . site the ~~~erencesacross regions have not only been a subject of analyses but the cause of s#~s~an~ia~ citj o f Manaus in order to rcd 2003). Although it t~oufdIIOE be e n: and. (iii) lower transpo once h'82i195 economic growth s lato XlX century. S i r n ~ ~ i ~ ~ nthe~ ~ ~ai s~li~i,aGDP per capita grew more than 3 percent per e r l n annL~r~during the six decades between tfie 1920s and tfie 1970s. These two s i ~ ~ i u ~ t ~trends n e o u ~ arc ~ a ~ ~ ~ cs~~rpr~singbecause after 1 ~ 8 thc~exact reverse process took pface. After 19813, ~ l a r l y 0 e c ~ i ~ o ~ n ~ c in the MRSP s r o ~ ~ e d r activity ~ oand~its sharei in~the B ~~ ~ ~ ~~oln iobegan ~n ~ ~ e ~~~~ to i drop. During this same tirnc ~ c r j ~ r~ad~~i CDPaper capita grot5-th st~gn~tedless than I ~ ~ n at percent per year. Gitcn that tower ~ ~ i ~o ~h ~ei r~rafcs of~rreturn~son in~estiiien~s, ~~ ~ ~ rcduccd profits coufd have forcud firms to improve effi~i~ncy,f a1 the while ~ n ~ r ~~ cc ~~ ~a ~~ r ~ L ~ l a ~ ~ ~ ~~eakene~.Thosc firms tiiat located outside the MRSP became more ~onipeti~~t~e and gaincd a largcr share of thc market, This ~ ~ e ~ docs8notne x~~ ~~why the ~ h l i~n ~ lostSpa~icipationin P the national GDP. It does tiowebet i~i~Is~rathat the positive e ~ t ~ ~ a f iofiae city do not t e ~ s ~ u f ~ ~~ioe~ ~np~c i~~ ~reducedprofits in order to n ~ a i n focal~activity growth. fors a t ~ ~ ~ n Note: Atisrage Source. Basedon osc is rtat tivc but iilti road map 17 26. The hypothesis of size: is the MRSP oversized?The MRSP already hosts 18 million inhabitants. Size is both a potential representation of congestion and a large home market. Congestion affectseconomic activity through the increasedcompetitionamong capitalists for the use of production factors, which imposes higher costs (e.g., higher land prices in business districts). This increased competition can also increase transaction costs, such as those related to the logistics of transport. The home market effect refers to the buying capacity of the local demand, including the number of inhabitants and their average real income. This force is expected to act positively since, in a large city, demand is also expectedto be large. However, as discussed in Section 1.1, the MRSP observed concentration of profits from land valorization in a reduced income group, as well as high unemployment; both phenomena weakened the local demand or home market.The MRSPtherefore is mostly facingthe negative effects of size. 27. Analyzing Brazil as a whole, there seems to be a negative relationship between city size and the concentration of economic activity. Large areas lost participation in the manufacturing sector and this activity was taken on by small population agglomerates. After reachinga certain maximum populationsize, large cities could not sustain high economic growth rates. In fact, SBo Paulo's economic growth rates were only 0.6 percent per year during the last three decades while small areas observed economic growth rates above 4 percent, However, this hypothesis is difficult to test given that there are only two `large city' cases in Brazil - Rio de Janeiro and SBo Paulo. The argument is further weakened when considering the different geographic agglomerations or the varying definitions used for concepts such as `metropolitan,' `municipalities' or `counties.' A `large city', for example, is usually defined as a conglomerate of urban areas. It follows that every `large city' would lead to different conclusions based on the specifics of the varying urban areas that comprise it.The MRSP is large, butthe 39 municipalities comprising it are not. A simple solutioncould be to divide MRSP into smaller cities, but then, the opposite problemwould arise: while size would not be the factor affectingcompetitiveness, there would be a need for a city's institutionsto behave consistentlywith its size. In fact, as discussed in chapter 6, coordinationproblems among metropolitanmunicipalities are substantial. 28. Transport infrastructure that links firms to large markets has the most important external impact on firm-level performance. Transportation costs have been decreasing consistently since the late sixties, apparently affecting the capacity of the MRSP to maintain economic activity. Using firm level data from the 2001 Annual Industrial survey along with regional data at the microregional level, Lall, Fundenbrug and Yepes (2003) found that firms located in areas that require relatively lower transport costs to reach large markets benefit from considerable cost savings." Although their results suggest that firms benefit from being in locations with good access to markets, they do not suggest that improving interregional connectivity would necessarily assist lagging regions in the short-run. Improving inter-regional connectivityimplicitly reduces a naturaltariff barrier so that firms currently serving large markets and benefiting from economies of scale can more easily expand into new markets and compete with local producers. This would be detrimental for local productionand employment in lagging regions but would favor producers inthe MRSP. 1.2.3 What is Happening at the Sector level? 29. Global experienceshows that firms kept their headquarters or high-skilledworkers @&D) in their original cities rather than relocating them into new industrial centers. " See "Transport Costs, FiscalIncentivesand IndustrialPerformancein LaggingRegions:Evidencefrom Brazil" Submittedto RegionalStudies. 18 Relocation of firms towards those new urban centers occurred in highly competitive sectors,such as clothing, furniture, toys, and cars; mid- and low-tech specific-inputs as well as land and/or labor intensive sectors left traditional large urban areas. Industries the remained are technology and skill-intensive and those that depend on proximity to research facilities, teaching institutions, business services, and financial markets (Yusuf, 2006). Potential sectors to be strengthened in traditional large cities are business services (e.g., financial, accounting), leisure, health, personalservices, logistics, education, and research. 30. From the sector analysis perspective, one alternative explanation to the MRSP's structural transformation is that by 1970 it had some "wrong sectors" (i.e., the sectors that under-performed in the 1990s). Figure 1.9 shows the decrease in the share in value-added between 1991 and 2000 for the manufacturing sector by skill level. In the MRSP, sectors requiring medium-skilled workers witnessed a decreased in market participation that is much greater than the decrease that occurred within the low and high-skilled sectors. Looking at the . share in employment, one noticesthat the decrease was less severe in every class, except for the medium-skilledmanufacturingindustry (Figure 1.10). This signifies that productivity (as value added per worker) was converging toward national averages in both the low- and high-skilled manufacturingindustries(shares inboththe value-added andthe labor force dropped by 5 percent and 2 percent, respectively). For the medium-skilledsector the share in value added dropped by 10 percent and the share in jobs dropped by 8 percent. Since the MRSP started with high productivity levels relative to Brazil as a whole, it is possible that some convergence either in productivity or in wages (which, in equilibrium, should be the same) within Brazil has been occurring. Figure 1. 9. Industry sector: change in shares in Figure 1. 10, Industrysector: change injob shares by value-addedby skill level, 1991-2000 (percent) skill level, 1991-2000 (percent) 6% T- -- - - - -. _ _ 6% -- __ ._ _ . _ - 4 % 14 fm 0% `-2% 1 , ........................ Lu +Medium ud% +Medium - .......................................................... -6% 4-H h ... ..... .......... -8% ~ J < 1 2-3 CUR REC PA RJ . f 3,060 3.265 '1,864 I.966 1@1. However, the ~ r n t ~ ~ e ~for r Special Zones af Socia! Interest (ZEIS) is not Q k conducive tu private sector ~ ~ t ~ ~ ~becnusc ~ r complex t l ~ ~ it is n e ~ ~ and ~estricti~~e. Thcrc arc many different types of areas where zoning and b u i l d i i ~r ~~ ~ ~ apply.~One of the ~~ ri osb ~ e 1 ~ ~ ~ i ~ i ~ is that a same plot can be ~ n\ssrious ~eypcs oflrestricted areas: ~S ~ p Zones of~~Social~ ~ ~ ~ ~ ~ EIS I?2>3, or 3)- t3r habi1itation~~~~~~~~~~~# de Rr 60 110. In addition, zones classified in ZEIS are subjected to strict redevelopment guidelines. For example, in areas regulatedas ZEIS-3, which are located inthe center of the city, redevelopmentmust be such that at least 40 percent of the developed area is composed of Social Interest Housing (Habitaq6o de Interesse Social, HIS) of less than 50 square meters, and 40 percent is composed of Popular Market Housing (Habitaq6o do Mercado Popular, HMP). The other 20 percent is not subject to constraints and can be commercial real estate or high-value housing. The mandated proportion of social units does not promote financial viability. Thus, attracting private developers to develop in these areas has been a challenge. Subsidies will most likely be needed, along with a mixing of ZEIS and non-ZEIS areas in order to allow cross- subsidies from high-endhousingor commercialdevelopments to low-incomehousing. 111. Another concern with the redevelopment of ZEIS relates to the redistribution that will take place as a result of the local improvements. Mechanisms based on expropriation and resale of consolidated parcelsto developers will most likely result in windfall gains to developers or to the new residents, not to the current low-incomeresidents, whose interests were meant to be protected through the creation of ZEIS. An alternativeto the expropriation scheme could be the use of techniques of land readjustment, which have been widely used in Asian countries, such as Japan and Korea. This type of arrangement allows the initial owners to have a stake of the land improvement benefits. Inthe case of Si30 Paulo, this can be consideredparticularlydesirable.62 4.4. INSTITUTIONAL CHALLENGES ARISING FROM THIS SCHEME: LACK OF COORDINATION,FINANCING CONCERNSAND THE NEED PREVENTIONPOLICIES FOR 4.4.1 Coordination Challenges 112. Municipalitieswithin the MRSP have adopted highly idiosyncratic strategies to deal with their respective low-income housing problems. The differences relate to the content of slum upgrading interventions (limited to urban upgrading versus "integral" urban, social, and economic approaches), and to the features of municipal administration^.^^ While it undoubtedly reflects on varying responsesto differences in demographic and housingconditions such as local income distribution,the precariousnessof informalsettlements, andthe degree oftheir integration inthe formal city, it is also a direct consequence of the absence of institutionalcoordinationat the metropolitan level. 113. The absence of a coordinationstructure for the housing sector at the MRSP level is problematic, given that, in many aspects, the relevant scale of housing problems is the MRSP itself, not individual municipalities. This issue of coordination also concerns housing and land invasion issues related to the provision of infrastructure services at the metropolitan level. It seems that, on such specific issues, well functioning institutional mechanisms of coordination between the municipalities and the state are needed.64 Important collaborationwas nevertheless achieved under the Bank-financedGuarapiranga project, and is currently being furtheredduring the preparationof the proposed follow-up Mananciais project, in which a dozen municipal governments are discussing with the state authorities a combined effort for urban 62Two staff members from the planning Secretariat in MSP have been sent to Japan to study land readjustmenttechniques. 63For example, municipalbudget for housing,number of staff, and capacities in the Housing Secretariats, andcooperationbetweenSecretariatswithin the municipality. 64The FederalConstitutionstipulatesthat all metropolitanregions shouldhave an agency, a council, and a development fund. Legally, a reorganizationof MRSP is expectedto occur in the near future, followinga draft law (PLNo. 06/2005), which is under considerationby the State HouseofRepresentatives. 61 upgrading (including housing improvements), social inclusion, and water pollution control in areas of mananciais. 114. There seems to be a scarcity in information sharing, exchanges in knowledge and practices (e.g., monitoring systems), as well as a scarcity of technical cooperation between municipalities in the MRSP. Presently, the formalized exchange of information between the MRSP's municipalities(and with the state of Sgo Paulo) is limited, and is compounded by a lack of cooperation between the respective municipal housing Secretariats. The exception to these barriers beingthe coordinationestablished under the water pollution control and urban upgrading programs focusingon the Billings and Guarapiranga reservoir areas -this coordinationevolving from a need to foster links between state and municipal government activities regardingMRSP water pollution control policies rather than being based on existing coordinated housing and urban upgradinginitiatives.Due to their varying economic, political, and social characteristics, it is very difficult for municipalities to coordinate actions on their own. A more informal but practical solutionthat has progressivelybeen achieved by ABC cities, consists of the creation of consdrcios with specific mandates for environmentalprograms.65However, these success stories are partly attributed to the similar profiles and political alignments of the participating municipalities. 115. At the local level, there is very limited coordination with regards to the different housing programs found within different government levels. For example, CDHU runs conjuntos and mutirlo programs with financial conditions and housingproducts that are similar to those found under COHAB's administration. Other federal government programs that involve associations and adopt the mutirlo model also exist. One would expect that the demand to register in those programs on behalf of those associations would be shared by the different players. However, this is not the case, and the result is a menu of "housing shops" and programs, with different eligibility criteria and financial terms, which households and associations have to visit one-by-one. Furthermore, given the complexity of the programs, it is difficult for potential beneficiariesto get informationabout all the programsthat exist. 4.4.2 FinancingConcerns 116. The MSP has no direct financial control over most of the housing programs being run within its boundaries (Table 5.1). 66Recently, due to its difficult financial situation, COHAB did not finance construction programs other than mutir6es. The IDB, via the financing of interventionprograms in cortiqos and Locaqlo Social, PROVER, and PROCAV, has been an important source of financial resources to the housingsector of the MSP. The state government, via CDHU, was the main player in the MSP, with a direct delivery of more than 3,500 housing anits per year. Finally, the PAR representedan importantcomponent of public programs. Inmost cases, the MSP does retain shared or total control of project management, as is the case with projectsrunwith the support of internationaldonors. However, this is not the case for some ofthe 65 A =ABC is composed of the followingseven municipalities:Santo Andre, B = Slo Bemardodo Campo, C = Slo Caetano do SUI, Diadema,Mauti, Riberlo Pires, and RioGrande da Serra. 66 CDHU has no legal obligation either to spend money in housingprograms in the city or to subordinate its interventions to a city strategy that would be designed by the Municipality. Subsidized credits for housing distributed by CEF are not under the control of the municipality,except for certain credit lines (e.& the new 0peruqo"es Coletivus), which involve a demand selection by municipalities. The PSH subsidy program is subject to a biddingprocess. The PAR involves negotiationson a case-by-case basis with CEF. Finally, the contribution of internationalis marginal interms of financial resources, inaddition to beingcostly interms o ftime andmobilizedpersonnel. 62 programs run by CDHU, which explains why programs similar to rnutirces are run by both CDHU and COHAB. These programs have differentcharacteristicsand different beneficiaries. 117. The current allocation of financial resources to public housing programs is skewed toward finished housing units. Simulations of the resources required to upgrade the existing informal housing stock, taking into account the flows of new households, approximateresources inthe realmof R$300and R$780millionper year. The variationdepends on the speed at which the current stock is beingupgraded (10 to 20 years) and on the proportionof households which have to be relocated.This estimate is higher than the current municipalresources being invested in housing, which averagedR$200millionduring2001-2004, and beyondthe share of the budget that is spent on upgrading.However, if the resources available for housing in the MSP from the differentgovernment levelswere combined, these estimatedresourceswould be met. 118. Paradoxically, the municipality cannot significantly improve the situation of informal settlements by itself, but there are already resources in place that could change trends substantially. Partnerships with other government levels and internationalorganizations have to be formed. Scenario analyses show that financial resources devoted toward physical upgradingshould be largerthan those directed toward the productionof new units for relocation (currently not the situation). The number of finished units necessary to absorb relocationneeds is inferior to the number of units actually delivered in the MSP when considering those finished units produced by all programs together (PAR, CDHU, and COHAB). It would be beneficialto reduce supply and reassign resources to quality improvements (Le. slum upgrading). In other words, coordination among government levels is necessary in order to rebalance housing interventions. Given current conditions, it would be more sensible to decrease quantity oriented programs (productionof new units) in favor of quality orientedprograms(Le,, upgrading). 119. Sufficient financial resources are not being directed toward families that are not creditworthy. Housing programs and subsidized credit lines rely mainly on mortgages, at the expense of other solutions such as progressive housing and rental housing, which may be more sustainable or affordableto low-incomehouseholds. Regardingthe Federal FGTScredit lines, the average loans are R$55,000 for collective buildings (CrPdito Associativo); R$37,000 for individual loans; and, only R$5,500 for construction material^.^' Micro-credit programs, which could constitute a powerful alternative to classical mortgages for the low-income residents, are still experimental.Two pilot micro-creditschemeswere launched in September, 2004.68 4.4.3 The Need for PreventionPolicies 120. Lastly, the lack of prevention policies ultimately results in invasions of public and environmentally protectedland. Figuresfrom the two last Censuses(1990 and 2001) show that significant flows of new households are formed every year in peripheral and reservoir areas. While some of them struggle with increasing densification, newer invasions of land on environmentallysensitive areas continueto occur and are not addressedat all by public policies.69 61Proportion of subsidized lending resources to loans for construction materials: about 3 percent in the municipalities of S8o Paulo and Santo Andrk; and, 11 percent in Guarulhos (it is equal to one-sixth of lendingfor housepurchases). The municipality initiatedpartnerships with Bradesco and Banco Popular do Brasil aimed at providing families with credit below R$ 5,000 and R$ 10,000, respectively, to acquire constructionmaterials. The two pilot projects concernedtwo small MSP areas with 350 and 240 families, respectively. 69In addition, the municipal administrationhas followed a policy of upgrading in priority old settlements (e.g., with more than 20 years of existence) to avoid giving incentives to new invasions. Overall, public policies in slum areas have been laggingbehindthe dynamics of settlements that occurred on the ground. 63 The enforcement of land-use restrictions lies outside of the municipal authorities' reach. However, this will change with the implementationofthe recently-approvedGuarapirgangaland- use/water pollution control law. Moreover, no alternativesolutions that deal with the invasion of public land exist elsewhere in the city (e.g., small serviced, semi-serviced or surveyed plots accessible to low-incomehouseholds). 4.5. TARGETINGAND AFFORDABILITY OFHOUSING SOLUTIONS 121. Targeting and affordability conditions of housing solutions are limited. Finished housing units produced by the MSP or CDHU do not constitute a sustainable solution for households with income below three MW. Most of the MSP's housingprograms aimed at the direct provisionof subsidized housingunits are unaffordablefor the targeted ho~seholds.~~These programs are clearly unsustainable on a large scale and could only serve a small fraction of the population below three MW in the city. This holds true even in a scenario in which the entire municipalhousingbudget is directedto them. Unfortunately,conjuntos and mutiriio programs run by CDHU are not more financially sustainable than other programs and entail approximatelythe same subsidy rates. 122. The new public rental housing program (Loca@o Social) and recently built housing compounds have been mostly populated with families that have relocated from slums and at-risk areas; their earnings are often below three MW. As a result, these programs have encountered payment and rent collection problems, implying that these programs are closer to a lottery scheme than a sustainable subsidy programs for low-incomeresidents. 123. The PAR has been relatively successful, but suffers from an imbalance in its financial structure. It may be the most successfulexample of cooperationbetweenthe MSP and actors from other government levels.PAR units have representeda significant part of the overall provision of housingunits to middle and low-incomefamilies. Thus, the MSP has been putting a great emphasis on the program. However, the PAR could not constitute a universal response to low-income housing provision by itself, especially with respect to families that relocate from informal settlements or at-risk areas. Due to the selection of beneficiary households by CEF (the municipality "organizes" demand and provides lists of applicants to CEF), on average, the households selected by the program were those with the highest incomes. This left other programs such as LocacZo Social to deal with the rest of the households. While this has probably permitted CEF to limit the non-performing contracts on PAR units, from the municipality's standpoint, the solutionfor the neediest households has to be found elsewhere. Concerns with the future of PAR in the city have more to do with its financial imbalances than with the achieved adaptation ofthe federal programwithin the MSP. 124. Attempts to involve the private sector, such as through the private rental market, have not proved to be efficient. Rent subsidies (Bolsa Aluguel), a form of personal rental allowance, was created to allow selected low-income families to find an accommodation in the private rental market.'l Theoretically, this system was attractive because it involvedthe private sector's supply of rental units. However, there are two main problems with its parameters: (i) the maximum rental permitted (via maximum rent-to-incomeratios) are too low and correspond at 70The ex-ante subsidy rates from the consumer's point of view are 70-90 percent for units in conjuntos programs, mutirtio programs,or locaqtio social. Ex-post,subsidiesare even higher because late installment paymentsor rents prevail. " The subsidy comprises a monthly allowancethat covers part or all of the rent plus a system of guarantees for the landlord. 64 best, to a small portion of the formal rental supply; and, (ii)subsidy rates reach 80 percent or higher for households with monthly incomes under three MW, which makes scaling-up unfea~ible.~~Moreover, in practice, the program has been used only as a temporary solution to house families displaced fromfavelas, at-risk areas, or vacated derelict high-risebuildings, and not as an "open" subsidy program, such as a voucher programor an entitlement. 125. Interventions in MSP's slums seem to have been mostly focused on physical upgrading, at the expense of social and economic programs. In the Mananciais area, the lack of economic opportunities and jobs is a pervasive problem. From anecdotal evidence, it also appearsthat the municipality lacks social staff to adequately cover all of its upgrading programs. This results in small social teams on each project, constrainingthe amount of social activitiesthat canbeundertaken. 4.6. CONCLUSIONSAND RECOMMENDATIONS:INSTITUTIONALAND INSTRUMENTS 126. The Municipality of SPo Paulo's size and the complexity of MRSP's housing challenges justify coordination of housing interventions both between different government levels and among municipalitiesat the metropolitan level. Currently, with the exception ofthe initiatives under the Guarapiranga and the Billings programs, no such coordination is taking place. The lack of a national interventionframework for the housing sector, in particular, has impededthe inclusionof nationaland state programs within the MSP local housing intervention strategy. Furthermore, the absence of an appropriate governance structure that defines and oversees a coordinated program has hindered the development of coherent interventionsand the most efficient use of resources. The actions proposed to achieve better implementation of municipal housing policies can be divided into two groups. The first group concerns the institutional setting within the MSP's municipal administration. The second concerns the instrumentsusedto enableand implementhousingpoliciesand programs. 4.6.1 Institutional Recommendations 127. The first goal of a comprehensive housingstrategy for MRSP should be to integrate the programs found across the different government levels and create a framework that coordinates the allocation of public resources for housing from across the different government levels. The long-term objective should be that all public interventions in the municipalityfollow a housing strategy based on observations of housing needs and a multi-year implementationplan that takes national and state financing into account. Direct implicationsof this new approach would be: (i)the rationalization of programs and interventions among institutionalplayers, such that redundant programs would be eliminated- this would be a first step toward the harmonizationof subsidies across the board, allowing for the reallocation of municipal fwnds to segments of the low-income population that are not currently reached by public programs; (ii)the conceptionand implementationof a city-widepreventionstrategy aimed at providingnewcomers with alternatives to the invasionof new, protected, or at-risk areas; (iii) the coordinated managementof the demand from householdsand resident associations, through a shared information and monitoring system; (iv) the specialization of the different players, according to their comparative advantages, insomuch that municipal resources could focus on mandated activities (e.g., regularization and urbanizationprograms); and, (v) a decrease in the volatility and uncertainty of financial resources beyond the municipal budget, which currently constitutesa serious obstacle to the scaling-up of housinginterventions. ''Nowadays, the scale of the program is negligible, with only 1,500 beneficiaries. 65 128. Second, the MSP should improve the coordination amongst the various housing interventionsfinanced and implementedby the municipality. The different Secretariatsof the MSPshould cooperate more closely in terms of coordinatinghousing policies, resources, and information systems. The Municipality should use efficient information, project management and monitoring systems that allow for a control of the costs of the municipal programs. The technical capacities of COHAB staff should be increased to allow COHAB to efficiently implement municipal housingprograms. General recommendations are (i)integration of informationsystems, within COHAB and SEHAB departments, between HABI and COHAB (short-term), and SEMPLA (medi~m-term);~~ (ii)development of a monitoringsystem for the various programs for better control of the use of the municipalities' resources by various subcontractors; (iii)systematization of ex-postevaluations of housingprograms, which should be included in the monitoring system to be put in place under the partnershipwith Cities Alliance; (iv) increases in the long-termvisibility of public interventions(Cities Alliance strategic planning activity); and, (v) continued strengthening of COHAB through improvements of its internal capacities and adoption of more sound commercialpractices. Inthe long-term, housingprograms and interventions implemented at the municipal level should be consistent with a metropolitan strategy for housing; externalities should be addressed between municipalities. In the current context, this vision may seem ambitious and runs intovarious institutionaland political obstacles. The initiation of new large-scale upgrading actions in the Billings and Guarapiranga reservoir areas, under the proposed Mananciais project, constitute an ideal laboratory for improved coordinationof objectives and interventions 129. Short-term recommendations for the MSP's comprehensive housing strategy include: (i)arranging long-termpartnerships with CDHU for the provisionof relocationunitsfor those households located in at-risk areas and favelas; (ii)harmonizing the characteristics of similar programs and sharing of informationon the demand for such programsby individualsand association^;^^ and, (iii) creating a one-stop shop (Poupatempo de Habitaqclo) that gathers housing producers, financing providers, and providers of subsidies for low-income populations. There is an opportunity for increased informationsharing and the exchange of practices between municipalities. The Housing Secretariats of different municipalities should organize regular informationsharing meetings. Topics of exchange could include: (i) content and implementation of slum upgrading interventions; (ii)monitoring and evaluation systems for housing programs; (iii)useoflegalandregulatoryinstrumentsenabledbytheLawofCities(e.g., taxversusland);75 and, (iv) evaluationof PPPs in urban operations and similar interventions.Again, the Mananciais project presentsan opportunityfor developingthese areas of collaboration 4.6.2 TechnicalRecommendations 130. While the optimization of municipal subsidy programs requires technical and in- depth research of the municipal programs,the following summarizes the general direction that these programsshould take: (i) decreasing the depth of subsidies on householdfinancing for mutirGes and ~onjuntos;~~ (ii) continue strengthening the financial performance of COHAB, 73Interms of differentgovernmentlevel, integrationof informationsystems, inthe long-term,shouldoccur between CDHUandCEF. 74For example, mutirdes programs done by CDHU and the MSP could use a shared system of demand identificationand selection,and offer similar subsidies. "LawNo.10.257/2001. 76The reflection on the adequate changes to be operated should ideally be a shared exercisewith CDHU, whose programs (Pro-Lar, Nucleos Habitacionais, and Pro-Lar MutirZo) have slightly different characteristicsbut essentially face the same problems. 66 by fighting non-payment and improving rent collection in Locaqlo Social programs;77 (iii) revising the parameters of the Bolsa Aluguel program so that the maximum rent-to-income ratio increases to allow a larger portion of the housing stock to be eligible in the program, especially for the lower income brackets; this should coincide with a reduction of the subsidy amount; (iv) adding social and economic components to those urbanization programs carried out in the city's slum areas; (v) since households that are unable to take a mortgage are disadvantaged interms of the generosity of public-programsand subsidies (e.g., cortiqos and PSH), it would be appropriate to reviewthe subsidy schedule of housing programs in order to improve equity between them and the other household^;^^ (vi) since it is currently very difficult to have ready-made information about the state of individual projects, their costs, and a view of the strategy followed more globally by COHAB and SEHAB, better project management information systems accessible to all municipal employees are needed. This would allow for continuous time monitoring of upgrading projects as well as construction projects financed by COHAB and SEHAB; and (vii) given that often there is no evaluation of municipal interventions, resource allocation should be dictated by the relative performance of the various programs. 131. A long-term goal should be ensuring that within the consolidatedhousing strategy, that segments of the low-income populationthat are unable to afford mortgages are better supported by public programs -this would be an alternative to the provision of finished housing units.Actions to realign the allocation of public resources toward housingneeds include introducing new programs that target underservedportions of the low-income population. Actions recommended include: (i)devising a strategy for the provision of small, low-cost serviced or semi-serviced land where low-income households can settle as an alternative to the invasion of public land;79(ii)implementinghousingmicro-credit programs and allocating sufficient resources for the scaling-up of such programs, these programs should be based on the outcomes encouraged by Cities Alliance within the implementation and evaluation of the pilot programs;" (iii) undertaking a study of the rental sector to assess the feasibility of incentive programs in the supply of affordable rental housing for low-income households;81and, (iv) allowing collective rental units to satisfy living standards that are less than the actual minimal standards, in order to directly compete with the cortiqos. ~~ 77 Ifthe municipal administrationintends to pursuethe Locup70Social program, there would be a need to revise the policy of dweller selection. Essentially, there is a gap between the cultural background and the financial capacities of the populations relocated from slums and at-risk areas, and the requirements of life in vertical units. Even if the municipality does not intendto scale up the programs of Locuq60 Social, the problems arising in rent collection and payment o f collective charges could translate into liabilities to the FMH and future liabilities due to under maintenance and the range of social problems that have been observed in the ownership compounds built by COHAB. 78 The degree of freedom of the municipality itself may be small in this area (e.g, parameters of action the program in cortiCos depend on the federal government and CDHU). Thus, this is more a recommendation for the state and federal governments. 79 This should constitute the top priority, giventhe importance of the problemandthe projectedincrease in new households living in environmentallyprotectedareas. Cities Alliance is supporting the MSP in the institutional and financial strengthening of SEHAB and while pursuingthe Strategic HousingPlan, enabling it to create and maintaincitywide informationsystems for monitoring, evaluation, and follow-up of public housing policies so that good practices can be shared and replicated at a national and global level, reflecting the endeavor to achieve the millennium goal of "cities without slums." Ifinterest rates continue to decrease inBrazil, within a couple ofyears they couldbe at a levelsuch that rental becomes a good investment. Thus, it shouldbe possibleto devise incentives to bridgethe interestrate differential between rental and other types of investment. Small subsidies could, then, provide sufficient incentivesto investors to providenew rentalsupply for middle and low-incomehouseholds. 67 4.6.2.i The Headwater-ReservoirMananciaisAreas and Informal Settlements 132. A critical element of an MRSP housing and upgrading strategy should be a comprehensive approach for the mananciais areas, including the implementation of the recently-approved land-use/water pollution control law for Guarapiranga basin (Lei Especifica de Guarapiranga) and the development and implementationof a similar law for the Billings basin. The new law contemplates instruments and incentivesto promote appropriate land-use and the upgradingof existing informal settlements within the basin, such that pollution of the reservoir is reversed. The land-use/urban informality/environmental protection nexus encountered in the mananciais basins is one of MRSP's critical development challenges requiring integrated interventions through vertical and horizontal cooperation on macro infrastructure provision, urban upgradingand housing solutions. These initiatives, pioneered under the World Bank-financedGuarapiranga project, will be further developed and rolled-outunder the proposed follow-up Mananciais projectwhich the state and municipalgovernments are currentlypreparing. 4.6.2.ii Attracting Private Sector Participation 133. Given the magnitude of housing problems in the city and the relative scarcity of public resources that can be mobilized for housing interventions, actively seeking out partnerships with the private sector should be a priority. Zoning regulations in the MSP should be one area of focus. In particular, the criteria for any redevelopment being planned in zones classified as ZEIS seem too stringent to arouse the interest of the private sector. Moreover, it seems that there are too many ZEIS, in the sense that public interventionscan only work with a small fraction of them. A first recommendation is to revisethe classificationof ZEIS zones inthe PDE, using the revision of the PDE that will occur in 2007. The aim should be to reduce the number of ZEIS and concentrating on the most important or strategic zones. A related recommendation is to revise the parametersthat define the minimal proportions of HIS and HMP within ZEIS perimeters. A second recommendation, made in the context of the considerationof the interventionin the Nova Luz perimeter, is to study the possibility of usingtechniques of land readjustment as an alternativeto expropriation. 134. The search for innovative PPPs should not be limited to ZEIS. The municipal strategy should focus on other areas, including: (i)introducing market incentives for the production of middle and low-incomeunits in the renovationof the city center; (ii) maximizing the use of existing legal and regulatory instruments (e.g., certificates of development rights as proposed for Paraisdpolis and Helidpolis, and exchange of land against forgiving of local tax debts), while evaluating instruments that have already been used (e.g., CEPAC, Urban Operations); (iii)undertaking an inventory of public vacant land in the MSP and devising a strategy and action plan for the use of this land;'* and, (iv) finding alternativesto mobilizingthe vacant housing stock through studies of the vacant housing stock and of rental housing as an inve~tment.'~ 82 Presently, there seems to have been some studies by individual entities (e.g., COHAB) on their land holdings. However, no consolidated vision seems to exist, even within the MSP, regarding the nature, extent, and localizationof public landin MSP.A study dedicatedat producingsuch an inventory would be needed,as a first step towards a coherent plan for usingpublic landholdingsinthe long-term. 83Studies should aim at assessingthe nature, quality, and location of the vacant housing stock to separate units in categories relatedto the economic motivesfor which the unitswere left vacant. 68 5. IMPROVINGPARTNERSHIPS INSTITUTIONAL 5.1. CITIESAS PARTNERS 135. Under Brazil's fiscal federalism, institutionalcoordination appears to be as important as revenue sharing for policy provision. Dividing the government's tasks and budget into three levels does not signify clearly defined geographical limits of responsibilities. Because policy outcomes are only visible at the local level, many conceptual problems arise when defining the scope of action for the local, state or federal governments. These conceptual difficulties arise from two sources, which are related in a trade-off: on one hand lies the political economy of competitionamongpoliticians; on the other hand lies the economics ofjoint policy provision. 136. Interms ofthe political economyofcompetitionamongpoliticians,electedgovernments need to provide visible policies for their constituencies in order to continue in their current elected positions or advance higher along the ladder of power. Constituencies, however, are located in one geographical space, which is defined simultaneously by the municipal, state and federal levels to which they belong. Therefore, government levels compete to differentiate their outcomes from one another. In terms of the economics ofjoint policy provision, delivering the right policy outcomes often requires sufficient financial resources and implementationcapacity. For example, geographic subdivisions will not require health facilities that include all medical specialties. Instead, the aim is to match the types of services provided by the health facilities within a geographical subdivision will the health conditions and location specifications of that subdivision's given population. Transportation and housing are other areas of public service provisionthat require a systemic vision to determine the manner in whichjobs are distributed in the metropolitanspace and the implications of this distribution on commuting and land prices. Finding the right scale of geographical provision signifies being consistent with the political economy; this can be done through an advance stage of coordination: partnership.Althoughthere is competition among cities, there are policy challenges worth pursuingthrough partnerships. 137. The MRSP, and particularly its core, the Municipality of Slo Paulo, has not been able to consistently make use of policy and institutional instruments that would guide metropolitangovernance and the implementationof basic municipal development policies, plans, and interventions.This was exemplified in paragraphs 113-116 within Chapter 5 and the discussion of limited housing policy coordination. In addition, constant political changes have signifiedinterruptionsto implementationby prior administrations.As a result, private actors have lost trust in sustained and long-term approaches to structural local problems, reinforcing individualist and short-termbehavior, of which the goal is to maximize contingent opportunities rather than sustainedinvestments. 138. Although many challenges lie ahead for the metropolitangovernance of the MRSP, it is worth recognizing achievements in institutionalcoordination. Government levels have taken steps that put them on the right path but those advances could also be threatened given underlying incentives. Therefore, examining the sequence on how current results were achieved and strengthening the mechanisms that feed policy-making with comprehensive analyses ofthe metropolitanregionremains at the center ofthe institutionalchallenge. 139. In the context of fiscal constraint and increasing public service demand pressures, coordination with other public agents is crucial. A confusing institutional framework with 69 overlapping responsibilities and the lack of coordination across government levels adds to the difficulties in municipal areas. There is a need of coordination both vertically (across levels of government: federal, state, municipal) and horizontally (across municipalities), particularly in the education, health, housing, and transportation sectors, where municipalities play a key role in service provision. 140. The MRSP has made limited progress in metropolitan governance. Metropolises around the world are increasingly aware of the necessityto coordinatethe municipalitieswithin a metropolitan region and are taking action (see Box 5). In the MRSP however, the general perception is that of an institutional vacuum and lack of organizationat the metropolitan level, which generates uncoordinated public policies, lack of initiatives to develop intra-institutional cooperation, and certain apathy from the private sector and civil society with regards to metropolitanissues. 141. The political and economic hegemony of the capital city in relation to the surrounding municipalities, which compose the metropolitan area, is one of the most difficult obstacles to establishinginter-municipal cooperation. The metropolitanmodel of the state of Si30 Paulo, however, does not recognize the different political and economic shares that the Municipality of Stio Paulo has vis-a-vis the other municipalities in the region. The voting power in the metropolitan region is equally divided between the state government and the 39 municipalities, and each one of the municipalities has equal power. This model region does not offer a representative political basis for capital cities, such as SBo Paulo, which represents 30 percent of the state's GDP. Currently, the state government of SBo Paulo is discussing mechanisms of shared metropolitan management, encompassing the creation of a development council, a metropolitanfund, and a metropolitanagency. Box 5. Inter-MunicipalityCooperation: A BrazilianExperience Probably,the most famous case of innovativemetropolitangovernance took placeinthe ABC regionofthe MRSP, comprising seven municipalities (Santo AndrB, S8o Bemardo do Campo, S8o Caetano do Sul, Diadema,MauB, RiberloPires, andRio Grande da Serra). Initially,the seven municipalities establishedan inter-municipalcouncil to address the failures of the 1970's metropolitanplans, particularly in relation to waste disposal.In 1997, the inter-municipalchamber of the extended ABC (Crimarado GrandeABC) was created to boost sub-regional integration by mobilizing forces fkom both the government and the non- government sectors - addin'g up to more than 100 entities. In 2003, the federal governmentjoined this chamber - an important step, since 11priorities (out of 15) of the chamber needfederal support (e.g.,the federalpublic universityandthe guarantqeqofPetrobrasto expandthe petrochemicalcomplex). The chamber structure is based on a deliberative council, an executive coordinatingunit, and inter-sector and inter-institutional technical groups. The decisions are basedon consensus among the members of the deliberative council, which is composed by the State Governor, the state Secretary of Science and Technology, and Economic Development, the state Secretary of Employment, the seven mayors, the chairmen of municipal councils, and presidents of the municipal councils, the state and federal representativesinthe legislaturefor the region, five representativefromthe GreatABC Citizenship Forum, five workers' representatives,and five representativesof economic sector organizations.The deliberative council meets every four months to approveaction proposalsmadebythe 15 thematicgroups. A new institutional mechanism, the ABC Economic Agency, was recently created to give more entrepreneurship to actions intendedto promote economic development in the region, which have faced profound economic restructuring. The ABC Council is a minority shareholder (49 percent), while the remaining51percentis divided amongnon-stateorganizations that support business developmentandjob creationinthe region, includingSEBRAE, CIESP, Chamber of Commerce, unionsandcompanies fkomthe region's petrochemicalcomplex. 70 142. Another major obstacle to coordination is the lack of incentives between the Municipalityof SPo Paulo, the state of Si50 Paulo, and the federalgovernment.The economic and political weight of the state of Sgo Paulo and the Municipality of Si50 Paulo in the country's macro-scenario often creates more incentives for competition than for collaboration. In many occasions, the three most important executive positions within the country have been divided among rival political parties, which perceived vertical inter-governmentalcooperation as a zero- sum game. In many cases, whenever a problem of metropolitan scale becomes a major public issue, the three levels of government prefer blaming each other to joining forces, which is understandable giventhe political visibility ofthese positionswithin nationalpolitics. 143. This national and international role of the MRSP affects both the capacity of political actors and private actors to cooperate at the regional level. The MRSP, and particularly the MSP, has become the best location for transnational corporate headquarters as well as for the representation of big national corporations. Although these firms are fiscally located in the metropolitan region, their interests are spread nationally and globally. Consequently, it has been difficult to convince decision-makingbodies of large corporationsto engage in solutions for the metropolitanarea. These companies quite often search for individual solutionsto the problems of urban agglomeration, such as congestion and urbanviolence. It is not difficult to understand why Sgo Paulo spends enormous amounts in private security and ranks secondafter New York City in the amount of helicoptersused. 5.2. DEVELOPING COORDINATION:EVIDENCE FROMTHE TRANSPORT SECTOR 144. The history of institutional coordination among entities running different components of the public transportation system is as old as the MRSP. First it was the problem of coordination of public companies within the same mode. For example, train lines were built by different public companies with different gauges, missing links, and many other coordinationproblems. Passenger lines were built in the last quarter of the nineteen century, but only after many versions were finally merged under the CPTM (Companhia Paulista de Trens Metropolitanos) in late 1990s. Historically, there was also dispersion of government agencies responsible for overseeing and planning, which led to tensions between the Municipal and State agencies. 145. Today's system is still complex, as expected within an urban area such as the MRSP. The system, which includestrains, subway, and buses, is operated by public and private actors under a wide variety of contractual arrangements. As presented in Figure 6.1., there are seven main entities taking part in the planning, investment and operation of the system, besides the thousands of operators and contractors in place. Both State and Municipality hold responsibility, which is undertaken by their respective Secretariats. On the side of the State, service is divided by mode: CPTM operates the train system, MetrB builds and operates the subway, and EMTU/SP (Empresas Metropolitanas de Transportes Urbanos) is the regulator of inter-municipal bus c~ncessions.~~ On the Municipality side SPTrans oversees intra-urbanzone concessionsfor bus operation and CET (Companhiade Engenhariade Trhfego) serves as planner and controllerofthe road network operation. 146. Nonethelesshuge progresshas been made in terms of effective coordination. In May 2006, all institutions involved signed an agreement to create an executive board to head the 84Agency created under the 1970s' Federal initiative to promote public-private partnerships for the operation of metropolitan buses. It regulates zone concessions and the BRT between Stio Mateus and Jabaquara,a 33kmtrunk operatedthe public-privatepartnershipMetrB. 71 metropolitan transportation system. Though in the past there was already some degree of coordination, this agreement defines the new IntegratedTransport Administration Committee - CDTI- as: (a) a tool to strengthen partnerships between the State and Municipal Transport Secretariats, (b) aligner of transport planning, administration, and overseeing, (c) promoter of efficiency by setting operation standards and investment priorities, and (d) promoter of an integratedvision of passengers' accessibilitythroughunified analysis andtariffs. The scope ofthe agreement is ambitiousbecause it establishesthe creationof legalmechanismsthat give the CDTI authority for decision making, along with legal definitions to secure investment backups from State and Municipal governments. It avoids creating a separate public entity, as usual in institutional reforms of this sector in Brazil, by deeply involving both Secretariats. Basically CDTI will work as an executive or corporativeboardofthe metropolitansystem oftransport. 147. This first step is crucial because it gives a broader scope to interventionsand creates political and financial leverage. Besides the conceptual definitions, the agreement went further by giving CDTI authority to establish Executive Agendas. The short-run agenda contains five componentsalready beingundertaken throughseparate activities and implying large investments: (a) new subway lines, (b) Tiradentes highway, (c) Pedro D. passengers' hub, (d) integrated fare, and (e) Diadema-Broklin Trunk. In the first meeting of the CDTI, the short-run agenda was expanded to include analytical and planning work, energy rationalization, and the Guarulhos-Tucuruvi trunk. The Agreement also defined a 90-day deadline to produce an integrated medium-runagenda. 148. I n practice,subsequent steps are demonstratingthe need to feed policy-makingwith stronger and more unified technical analysis. A review of minutes from CDTI meetings shows that heads of involved institutionsare becomingfully aware of this situation. Despite the willingnesstojoin efforts and overcome the aforementioned realitiesofthe political economy, an integrated and broad vision of the metropolitan system of transport is not easy to achieve. Despite recognition of its systemic characteristics, more technical work is required in order to integrate State and Municipal visions. Some salient issues already appearing are: consistency of investment plans in a context of financialtrade-offs as seem through the example of non-paying passengers or parallel services like train and buses trunks; consistency between State and Municipal investment agendas; consistency between transportationmodes and user responses to tariffs and supply; and the legal consistency of the geographical scope of interventions; among others. The approach of prioritizing short-run coordination objectives is therefore appropriate although it is also important to quickly develop a scheme that will delegate how the technical levels will work together, while preservingthe executive board spirit of the CDTI. A step in this direction was already taken with the creation of a technical committee to undertake the Origin Destinationsurvey. 72 Figure 5. 1. MRSP: Institutionalscheme for transportplanning,operation, and overseeing State Transport I Municipal Transport Secretariat Buses Traffic Controller Source: basedon state andmunicipality governments' websites. 5.3. CURRENT CHALLENGES AND OPPORTUNITIES FOR IMPROVING GOVERNANCE THE IN MRSP 149. Based on the evidence, it seems clear that addressing the social, economic, and environmental problemsof the MRSP will be intrinsically linked to the capacity of its own society to overcome deep political mistrust, negative economic competition, and social paralysis.The traditional public sector models of metropolitangovernance basedon regionalism or political fragmentation have demonstrated their limits vis-his the particular context of the MRSP. Moving beyond this model inthe near future will not be easy because it requires a change inthe political culture, which has deep historicalroots. 150. Local actors can reinforce and create new mechanisms of metropolitan governance that achieve increased coordinationwithin the metropolitan area. A wide array of public policies as well as legal and institutionalinnovations is available to promote inter-governmental cooperation, public-private partnerships, and more accountable and transparent governments. The public consortium is probably one of the most relevant institutional innovations for metropolitan regions as it facilitates partnerships among the different federative entities, increasingthe capacity of the associative groups to solve common problems without undermining their political autonomy (see Box 6). Through public consortia, the different entities of a metropolitanregioncan associateautonomously to implementboth specific actions and longterm programs. 151. The MRSP should explore mechanisms of public consortia to find solutions that require joint political action, investments, and human resource from different federative entities without circumscribing individual municipal boundaries. Public services can be provided in partnership with neighboringmunicipalities and with the state government. These partnerships can reduce fixed costs and spread capital investments over a broader spectrum of users; this ultimately cuts the unit costs of the delivery of public goods and services. The health sector in particularwould benefitfrom public consortia, as it requires large investmentsorganized by the size of demand and the complexity of the services. For instance, in the State of Minas Gerais, the inter-municipal consortium of the Alto Silo Franscisco observed the operations of different health units, from basic to specialized centers. This resultedin increases in the regional capacity to offer services as well as cuts in operational costs and transfers for treatment in the healthsystem inthe capitalcity of Belo Horizonte. 73 Box 6. Consdrcios The mechanism of municipal consortium was institutionalized in 2005 by Law No. 11,107, providingthe enabling regulationfor the "Union, States, FederalDistrict, and Municipalitiesto create public consortia and cooperation agreements with different federative entities, authorizing associative management of public services as well as the total or partialtransfer of charges, services, peopleor goods essentialfor the continuity of transferred services." The public consortia, thus, provide the conditions for joint actions (horizontalandvertical association),withthe aimofproducingbetterresultsthanifeachmunicipalitywere to act individually.Attending particularly to the pressures of State governments, the Public Consortium Law guaranteedthe presence of the State inany public consortiumwhich involves the Federalgovernment andmunicipalities,inorder to preservethe hierarchyofthe FederativePact. The public consortia, furthermore, can manage their own patrimonial andbudgetary resources, as well as establishmanagement structures to execute their activities. Their resources can be generated boththrough incomeobtainedfrom their activities, or throughcontributions fromthe different entities, approvedintheir respectivebudget law. It is importantto highlight, however, that regulation of public consortia follow the same fiscal responsibilityprinciples.Inthis respect, the public consortiaare subjectto operationalauditing bythe Tribunal de Contas. 152. Another important institutional and policy innovation is the federal law of the Statute of Cities. This law representeda historic movement to promote urban reforms started in the early 1960s, but that became effective only with the approval of the Law No. 10,257/2001. The Statute of Cities provides the general policy instruments and objectives for municipalities, enabling local governments to implementprogressive legislationbased on more socially inclusive land use as well as more transparent and participatorypublic management and decision making. Although the constitutionalprovisions are delineated in very generic terms, the introduction of concepts of democratic management, cooperationamong governments, civil society, and private sector, and transparency are very innovative.Among them, specific issues can be highlightedto improve governance: a) Direct participation of citizens and organized civil society in the formulation, execution, monitoringand evaluationof development programs, plans, andprojects; b) The concept of partnerships in the process of urbanizationthat address social interests; c) Public hearings between the local government and citizens during the implementation of interventions which potentially create negative impacts on the environment or to the safety and livelihoodof the population; d) Local development planning focused on the territorial distribution of population and economic activities, aimed at avoidingnegative impacts of urbangrowth; e) Planningofmetropolitanregions, urban agglomerations, and micro-regions; f ) Social control and citizenparticipationin municipalpublic expenditure management; g) Sector-basedcommittees in each level of the federation(national, state, and local); h) Conferencesat each levelofthe federationto debate urban policies; i) Provisionsfor popularinitiativesof law proposals, localdevelopment programs, plans, and projects. Although most of these constitutional provisions are not mandatory and were not specifically formulated for metropolitanregions, in principlethis policy and institutionalframework could be legally incorporated in metropolitan governance. However, there is very little evidence from Metropolitan Regions that these principles and incentives for good governance have been translated into publicpoliciesand concrete benefitsfor the population. 74 153. Public-Private Partnerships (PPPs) are particularly relevant considering fiscal constraints of the different federative entities, lack of incentives for private sector investments, and a pressingneed to invest in infrastructure. The PPP Law (No. 11,079/2004) i s also a governmental attempt to attract the privatesector for much neededinvestments, ensuring economic returns for those investments in public infrastructureand increasing transparency and trust in the normativeframework of privateinvestment in infrastructure.PPPs can be particularly relevant to the MRSP because o f the fiscal constraints of the MSP, which heavily limit its investmentcapacity. 154. Strategic actions should focus on progressivelyengaging the key political, economic, and social actors in a consultativeprocess, which could lead to a metropolitan development pact. It is not possible to expect quick results and success from this process, as it will take time for these actors who have historic differences and interests to build consensus. Anyway, recent efforts on the reorganization of the metropolitanregionalgovernance represent a step in the right direction. 75 6. POLICY AGENDA:RECOMMENDATIONS FORTHE SHORT- AND MEDIUM-TERM 155. Through an analysis of selected topics, this study aimed at offering inputs for a successful recovery strategy for the city and the Metropolitan Region of Si30 Paulo. The study first presented an analysis of the underlyingfactors of the economic transition in the MRSP, highlightingthe factors behind the recent performanceof the MRSP in terms of job creation and growth. Then, four inputs that would lead to a "recovery strategy" for the MRSPwere discussedindetail: `Improving FiscalPerformanceand Creditworthiness' assessed the current fiscal situation, revisingthe fiscal scenarioandthe linksof economic performance and fiscal spending. `Improving Competitiveness and the InvestmentClimate' analyzed the business environment in the MRSP, using both econometric models and direct surveys, with the aim of uncovering the main strengths and weaknesses of the region in terms of attracting new investments. `Improving Service Delivery in Low-Income Housing' carefully examined housing policies and programs, which, given the MRSP's current economic transition and fiscal constraints, are key services responsible for improving the livelihood and opportunities for the poor. Finally, `Improving InstitutionalPartnerships' discussed current and potentialactivitiesthat improve partnerships among municipalities in the MRSP as well as within different levels of government andthe privatesector. 156. Based on the conclusions of this analytical work, the major policy challenges and recommendations found under the four `Recovery Strategy' inputs are described below. These recommendations are divided according to the two conceptual differences found among the `Recovery Strategy' inputs. `Improving Fiscal Performance and Creditworthiness'and `Improving InstitutionalPartnerships'representtools that should be undertaken in order for the administrationto succeed in implementingrecovery strategies and policies. `Improving Competitiveness and the Investment Climate' and `Improving Service Delivery in Low-Income Housing' are targeted strategies that will help the MRSP bounceback from the challenging context of economictransitionpresentedin Chapter 1. 6.1 Tools that will help the MRSP successfully implement targeted recovery strategies and policies: `ImprovingFiscalPerformanceand Creditworthiness'and `Improving InstitutionalPartnerships' 6.1.1 `Improving Fiscal Performance and Creditworthiness:' The Municipality of SBo Paulo needs to implementa solid fiscal adjustment. This should and can be done preserving minimum levels of public services and investment, essential to help to unleash economic growthin the region. 157. As discussed, the main problem for the SBo Paulo municipal government's fiscal accountsis its highlevel of indebtedness.The net consolidated debt was equal to 223 percent of net current revenue at the end of 2005, far above the Fiscal ResponsibilityLaw (FRL) ceiling of 174 percent, which should be reduced to 120 percent of net current revenue by 2016. More than its high level, the problem is the continuous increase in the level of indebtedness over the last years. The change in this trajectory in 2005 is a positive and mostly welcome move, but its sustainability is yet to be confirmed. 158. Solving the fiscal crisis is a necessary step in launching a new growth strategy for the city and the metropolitan area. Most successful cases of cities that reinventedthemselves 76 were cases in which the cities went through a genuine fiscal adjustment process. Organizing the municipality finances help to provide predictability to the private sector - and the private sector, not the state, is the driver o f local economic growth. A number o f famously distressed cities suffering the strains o f economic restructuring have managed their way to fiscal stability and economic recovery. Examples from North America (District o f Columbia, Nassau County, New Haven, New York City, and Philadelphia) and Europe (Barcelona) support this view (see Box 7). All these cases are anchored in a permanent, sustainable fiscal stance, avoiding the use o f fiscal "one-shots" to balance their budgetsg5Cities need to get to the root causes o f their financial crises and adopt principles o f budgeting, multi-year financial planning, and transparent financial reporting. Once cities like District o f Columbia, New York, and Philadelphia achieved financial stability, they made themselves more competitive by creating an environment conducive for economic growth by cutting taxes and/or expanding investment in basic services. Box 7. The Case of Barcelona The fmancial situation of Barcelona in 1991 was characterized by a significant deficit in the public accounts: primary savings were earmarked to cover financial expenditures, while investments and debt amortization had to be financed through capital markets, and the total debt had reached an equivalent to 157 percent of the total current revenues.Given these circumstances, the city adopted a policy to stabilize the level of debt (1992-95), and increase investments. However, these results were to be achieved given that two conditions were respected:(a) not increasingtaxes, and (b) reducing public debt by 10 percentage points. The framework adopted to convert the deficit of the current account into surplus was based on a plan to increase revenues, while reducing expenditures and enhancing asset management. During this period, Barcelona reduced its public sector staff by 2,561 jobs, and improved public sector efficiency through reorganization, enhanced public information system and capacity building programs. As a result, the total debt reducedfrom 153 percentoftotal current revenues in 1993to 59 percent in2005. Inaddition, the enhancedfinancial solvency prompted Barcelonato improve its credit ratings (e.g., Fitch: AA- in 1997 to AA+ in2006, and S&P: A- in 1994to AA in2003). 159. In 2005, the new municipaladministration promoteda strong fiscal adjustment. For the first time since 2002, Si40 Paulowas able to achieve positive fiscal balances, which were large enough to generate negative borrowing requirements. The adjustment was based on revenue increases and strong investment expenditure cuts, while current expenditures continued to increase. Additionally, the fall o f the price index used to correct the debt and the favorable revenue performance in 2005 led to a considerable fall o f debthet current revenues from 251 percent in 2004 to 223 percent in 2005, a stronger decrease than that projected in the tightest adjustment scenario depicted in the projection exercise. Information through December 2006 indicates that the ratio was at 197 percent. This realized value is lower than the estimated in the fourth projection scenario for 2006 that was 202 percent; therefore, the municipality is doing better than expected even in the tightest scenario. 160. The high degree of indebtedness of the municipality demands the continuity of current fiscal adjustment efforts. This continuity will depend on the ability o f the municipal government to enhance its revenue collection ability and to control the expansion o f current expenditures. However, the fiscal adjustment effort required to return to a sustainable path for municipal accounts would impose less o f a sacrifice than initially perceived, given the high level o f current expenditures reached in the past. *'Examples of fiscal "one-shots" are long-term debt used to fill in one-year operating budget shortfalls, putting off capital investment, maintenance and technology, growing transfers of funds from city-owned utilities, sales of large governmentassets (e.g., land, facilities), and deferral ofpay to city employees. 77 161. The city of SHo Paulo has the appropriate instruments to implement this fiscal adjustment, as the level of spending has been rising in recent years. Moreover, some transitory factors that have aggravated the fiscal crisis, such as the index usedto correct the debt stock, have recently, at least partially, bounced back. Regarding the revenues side, estimates suggest that Siio Paulo is adequately exploringthe possibilities, showing a higher efficiency than other municipalities. The current positive scenario for continued prolonged economic growth in the country is, thus, key to the fiscal accounts ofthe municipality. 162. Maintaining investments in infrastructureis essential and the analysis shows that it i s feasible. There are many studies showing that, at the national level, higher levels of infrastructureinvestmentswould leadto higher growth. Inthe case of Siio Paulo, infrastructureis perceived by firms as one of its main assets and, at the same time, one of its weaknesses. Maintainingand improving its favorable endowment of infrastructure is thus essential. This can be achieved through private sector participation, but, as it is widely recognized, public investments still play a decisive role, especially in areas such as housingandtransport. 163. Public expenditures in infrastructure should focus on quality, rather than on quantity. Creatingfiscal space for public investments is very important, but, given the short-run limits imposed by taxation and current expenditure levels inthe context of budgetary rigidity, the focus should be placed on improving the quality of public expenditures in infrastructure. Supportingthe "right" project, one of the conditions for efficient public expenditures, requires appropriate project selection and design - Si30 Paulo has had a very negative experience with the creation of "white elephants." Adequate project selection involves identifying and ranking projects by economic return, all alternative uses of available resources. Enhancing municipal expenditure management, through the strengthening of the municipalplanningsystem is essential for this whole process. The steps taken by the administration are correct and positive, but more needsto be done. 6.1.2 `Improving Institutional Partnerships:' The MRSP needs continuity of programs and policies - engaging the private sector in the MSP's city planning, reinventing institutions and working through partnerships with other public agents can help. Furthermore, policy coordination between the federal and state levels and within the metropolitanarea plays a critical role. 164. Lack of continuity of policies and programs has been identified as a major problem in the MSP. The changes in the political command of the municipaladministration- actually all major political parties in Brazil have governed the city in the last 20 years - has been reflected in continuous changes in policies and programs. It is not surprising thus that the quality of service deliverywas affected; havingthe housingsector as the best example. 165. The question is not political alternation, but the fact that there is a need to go beyond political divisions and to ensure a broader consensus around issues and priorities for the city of SHo Paulo. Engagingmultiple stakeholders in the process can help to guarantee program continuity and hold public agencies accountable, in spite of changes in their political affiliation. Laying down the institutionalbasisfor establishingcommon groundpolicies, basedon shared values and visions about the strategic routes for the city is also essential. In this regard, it i s important to assure private sector/civil society forms of participation and engagement in the discussion of the cities and the metropolitanregion's future. Experiences from other parts of the world (see Box 8) suggest that support from civic, political, and business leaders is necessary to cater the interests of all parts involved and to improve the socio-economic environment in a metropolitanregion. 78 Box 8. The Case of Chicago , In 1980, The Economist described Chicago as a city whose "faqade of downtown prosperity" disguised a creaking political machine, the erosion of its economic base and severe racial problems. For a decade, Chicago suffered from industrial decline, political instability, people and business leaving, and racial tensions increasing. Currently, the same magazinedescribes Chicago as "a city buzzing with life, humming with prosperity, sparkling with new buildings, new sculptures, new parks, and generally exuding vitality." How did Chicago recover from its period of deindustrialization? There is no single answer, but the strong involvement ofthe private sector andthe civil society surely helped. As The Economist put it, "if you are a boss of a big business anywhere in the Chicago area, you are expected to take an active part in the civic life o f the city." Business names appear everywhere in the boards of universities hospitals, museums, local charities, and opera companies, and they take part in any public endeavor. This involvement is somehow institutionalized through the "Commercial Club," an organization of leading business and civic leaders established in 1877 to "advance the public welfare and commercial interests of metropolitan Chicago by cooperative effort, social intercourse and a free interchangeof views." The controversy over the expansion of Chicago's O'Hare airport provides a good example of the private sector involvement. Chicago's businessmen, who travel constantly, knew better than anyonethat the airport hit its capacity inthe 1990s. Therefore, who better than them could to have advocated for its expansion? The civic committee of the Commercial Club commissioneda study on the economic impact of the project and, after encountering opposition from some politicians, lobbied for three to four years until it convinced them to expand the airport. The expansionwas expectedto increase traffic, reduce delays, create between 50,000 and 195,000 jobs, and cost $6.6 billion, raised almost entirely through passenger, gate and landing fees -not through taxes. In 1996,the Commercial Club undertookthe Metropolis Project, an effort to ensure the preeminenceofthe Chicago MetropolitanRegion inthe 21st century. The Metropolis Project was based on the assumptionthat the issues facing the region were interdependent. These issues relate to unlimited, low density sprawl; concentrationof poor minorities; the spatialmismatch betweenjobs, affordable housing andtransportation; and disparate degrees of access to quality education. More than 200 members of the Commercial Club examined those issues in special committees organized around six areas: education, economic development, taxation, governance, transportation and land use and housing. The committees consulted with experts and met with regional community, civic and government representatives.The culmination of their two-year effort was areport titled: "Chicago Metropolis 2020: PreparingMetropolitan Chicagofor the 21st Century." A new organization, Chicago Metropolis 2020, continues the process started by the Commercial Club and is implementing the ideas contained in the report. Chicago Metropolis 2020 has focused on creating collaborations with other organizations and is developing a new kind of "civic entrepreneurship" inthe region. 166. In a context of fiscal constraint and increasingdemand pressures,coordinationwith other public agents is crucial. An ambiguous institutional framework with overlapping responsibilities and the lack of coordination across government levels adds to the difficulties in municipal areas. There is a need for coordination that is both vertical (across levels o f government: federal, state, municipal) and horizontal (across municipalities), especially in education, health, housing, and transportation, sectors in which municipalities play a key role in service provision. 167. The MRSP has made only limited progress in metropolitan governance. Metropolises around the world are increasingly aware o f the necessity to coordinate across municipalities within a metropolitan region. In Sgo Paulo, in contrast, there is a general perception o f an institutional vacuum and insufficient efforts at the metropolitan level, which generate uncoordinated public policies, lack o f initiatives to form intra-institutional cooperation, 79 and certain apathy from the private sector and civil society toward metropolitan issues. The political and economic hegemony ofthe city in relationto the surroundings municipalitiesthat are part of the metropolitan area is one of the most difficult obstacles to overcome in order to establish inter-municipalcooperation. The municipalityof SBo Paulo has more incentivesto build relations with the State and the National governments than with the surrounding metropolitan cities due to its prominenteconomic role in the country's economy. 168. This national and international role of the MRSP affects the capacity of both political and private actors to cooperate at a regional level. The MRSP, particularly the Municipality of SBo Paulo, has become the best locationfor the headquarters and representation of big national and transnational cooperation. These firms can be fiscally located in the metropolitan region, but its interests are spread nationally and globally. It has been difficult to convince the decision-making bodies of large corporations to engage in solutions for the metropolitanarea. These companies quite often search for individual solutionsto the problems of urban agglomeration, such as congestion and urban violence (Sgo Paulo has the second largest number of helicoptersinthe world,just losingto NewYork City). 169. Local actors can reinforceand create new mechanisms of metropolitan governance in order to achieve increasedcoordinationinside the metropolitanarea.There is a number of institutional innovations available to promote inter-governmental cooperation, public-private partnerships, and more accountable and transparent governments. The public consortium is probably one of the most relevant institutional innovations for metropolitan regions, promoting partnerships among the different federative entities, which increase the ability of the associative group to solve common problems without underminingtheir political autonomy. Through public consortiums, the different entities of a metropolitan region can autonomously associate to implementspecific actions in the pursuitof long-term programs. Public services can be provided in partnership with neighboringmunicipalities and associated with State governments, reducing fix costs and spreading capital investments over a broader basis of users, which ultimately cut unit costs of delivery of public goods and services. 170. The balance of water supply and demand in MRSP is a critical issue for the city's competitivenessand economic growth. MRSP's lowper capitawater availability is comparable to that prevailing in the driest areas of the Brazilian Northeast. Furthermore, half of the city's potable water supply is imported from neighboringriver basins, which is contentious given the demands of other large urban conurbations vying for the same water. The remainder comes from headwater-reservoir systems (manancias)within the MRSP itself. The GuarapirangaandBillings reservoirs make crucial contributions,together providingpotable water for 20 percent of MRSP's population(around 4 million people). Recent forecasts for the metropolitanregion indicate that, by 2010, there is serious risk of demand outstripping supply - with such projections assuming that MRSP's currentlyoperationalmananciais (Guarapiranga, Billings and other smaller systems) will remain fully utilized or be further expanded. Should GuarapirangaandBillings be lost as raw water bodies for the city supply, then the next-nearest sources to replace them are at great distance and could only be brought to MRSP at high cost and after successful negotiation with other stakeholders. A concerted effort to recuperate reservoirs is needed involving both state and municipalactors. 171, The land-uselurbaninformality/environmentalprotection nexus encountered in the mananciais basins is one of MRSP's criticaldevelopmentchallengesthat requiresintegrated interventions through vertical and horizontal cooperation. Despite these manancias areas having been subjected to severe land-use restrictions through decades-old water resources management and environmental legislation, illegal settlements have nevertheless proliferated, 80 causing direct pollution of the reservoirs through wastewater and garbage discharge and storm run-off and silting, and thus threatening their future use as raw water bodies for potable supplies. The state and municipal governments should prioritize the implementation of the recently- approved land-use/water pollution control law for Guarapiranga basin (Lei Especijka de Guarapiranga)and the development and implementationof a similar law for the Billings basin. The new Guarapiranga law contemplates instruments and incentives to promote the appropriate land-use and the upgradingof existing informal settlements within the basin, such that pollution ofthe reservoiris reversed. 6.2. Targeted Strategies for the MRSP's Recovery: `ImprovingCompetitiveness and the Investment Climate' and `ImprovingService Deliveryin Low-IncomeHousing' 6.2.1 `Improving Competitiveness and the Investment Climate:' With the adoption of appropriatepolicies, ,950 Paulo is in good shape to attract private investments; the city has many advantagesto foster privatesector-led growth. 172. Both the city and the MRSP have many assets to attract private investments. The MRSP has a very favorable geographical position, being a natural hub for air transportation, and being close to the main port of the country, as well as to agricultural and manufacturing production. Infrastructureis also listed as an asset both by firms that are inside and outside the metropolitanregion. This is recognized by entrepreneurs, which regard the state and the city of Silo Paulo as the best place to do business in Brazil, as shown by the ICs complementary survey with manufacturingfirms. 173. I f this is true, then why is the MRSP performing below its potential?One of the possible answers is that the regionneedsto exploit its advantagesand promoteitselfto attract and retain businesses. Several studies show that there is a need to switch from policies that seek to reduce costs (of capital, labor, and tax) to policies that exploit and strengthen "invisible" cost advantages. A study on the Montachusett region (in New England, United States) shows that firms usually do not have time and money to conduct exhaustive searches for their location and tend to rely on the information provided to them (Doeringer et al., 1987). Therefore, it seems essential for cities and regionsto be highly visible and find good matches between their strengths and companies' needs. Inthe cited study, firms that hadrecently started up or expanded indicated that their location decision was minimally influenced by visible costs. On the other hand, they reported that contacts with a well-informed economic development official were key in influencing their location choices. Cost factors were incidental as long as they fell within a reasonablerange.86 174. The analysis points out to some short- and medium-run actions concerning business regulation. Based on evidence collected by the Doing Business in Brazil report and the Investment Climate Assessment, three areas for improvement in business regulation:complexity of contract enforcement, time and cost to start a business, and complexity of payingtaxes. 86 Businesses also react to local capabilities to attract investment.City governments' efforts in promoting themselves are usually rewarded. In a survey conducted with 500 top European companies, entrepreneurs reported that Barcelona and London are the two cities that best promote themselves to attract investment (Cushman & Wakefield Healey & Baker, 2004). As a result, the survey reports that London ranks in the first place as best city for business location, while Barcelonajumped from the 11" position in 1990 to the 6" positionin 2004. This is not surprisinggiventhe great effort that Barcelonahas put inre-inventingitself anddevelopinga soundstrategic planning. 81 175. Land seems to be an important bottleneckto any strategy of poverty reduction and economic growth in the MRSP. The findings from the housing sector analysis and the firm survey clearly point out that landunavailabilityis one ofthe mainweaknessesof the MRSP. The supply of formal housinghas been insufficient in the municipalityof Silo Paulo and low-income households have turned into the informal sector to match their housing needs. Land is also essential for firms, as it provides the physical space for production.Indeed, a clear land titling is perceivedas a top priority for firms surveyedinthis study. Also, landavailability-that is the one of the main drivers of locationdecision - appears to be an advantage for all regions except the MRSP. Land unavailability is the third main disadvantage of the MRSP and one of the main reasons for firms leavingthe region.Therefore, it seems importantfor the MRSPto secure a clear property rights system. This is particularly important given that high-technologyfirms consider landtitling system as the most importantlocationfactor. 176. The MRSP has a comparativeadvantagewhen skills and technology are considered. The share of workers with college education in firms located inthe MRSP is significantly higher than that among firms not located in this region. The same is true for the share of workers who ' frequently use computers to perform their duties. An econometric exercise showed that the quality of the labor force along with proximityto suppliers and urban infrastructure are the main driversof firm's productivityinthe MRSP. 177. Infrastructure is the factor that most affects firm location decisions and the MRSP provides infrastructure of higher quality when compared to the other surveyed regions. When the quality of infrastructureservices was analyzed (as the share of total sales lost due to infrastructure services outages), the MRSP clearly provides services with higher quality than other Brazilianregions.When askedabout the factors that constrain growth, as is done inthe ICs, infrastructureis not rankedas a significantinhibitorto growth. 82 e e . e . . . m 00 e . . . e e . . . s 6.2.2 ImprovingServiceDeliveryinLow-IncomeHousing 178. Through the adoption of both institutional and technical policies, the MSP can maximizeits provision of low-income housing. Improved coordinationwith housingprograms under different government levels, more accurate populationtargeting, and engaging the private sector lie at the heart ofthis section's policy recommendations. 179. The first goal of the MSP's comprehensive housing strategy should be to integrate the programs found across the different government levels and create a framework that coordinates the allocation of public resources for housing across the different levels of government. The long-termobjective should be that all public interventions in the municipality follow a housing strategy based on the observation of housing needs and a multi-year implementation plan that takes national and state financing into account. Direct implications of this new approach would be: (i)the rationalization of programs and interventions among institutional players, such that redundant programs would be eliminated- this would be a first step toward the harmonization of subsidies across the board, allowing for the reallocation of municipal funds to segments of the low-income population that are not currently reached by public programs; (ii)the conceptionand implementationof a city-wide preventionstrategy aimed at providing newcomers with alternatives to the invasion of new, protected, or at-risk areas; (iii) the coordinated management of the demand from households and resident associations, through a shared information and monitoring system; (iv) the specialization of the different players, according to their comparative advantages, insomuch that municipal resources could focus on mandated activities (e.g., regularization and urbanization programs); and, (v) a decrease in the volatility and uncertainty of financial resources beyond the municipal budget, which currently constitute serious obstacles to the scaling-up of housinginterventions. 180. Second, the MSP should improve the coordination amongst the various housing interventions financed and implemented by the municipality. The different Secretariats of the MSP should cooperatemore closely in terms of coordinatinghousingpolicies,resources, and information systems. The Municipality should use efficient information, project management and monitoring systems that allow for a control of the costs of the municipal programs.The technicalcapacities of COHAB staffshould be increasedto allow the institutionto efficiently implement municipal housing programs. General recommendations are (i)the integrationof informationsystems within COHAB and SEHAB departments, between HABIand COHAB (short-term), SEMPLA (medium-term), and CDHU and CEF (long-term); (ii) the development of a monitoring system for the various programs for better control of the use of the municipalities' resources by various subcontractors; (iii)the systematization of ex-post evaluations of housingprograms, which should be included in the monitoring system to be put in place under the partnership with Cities Alliance; (iv) increases in the long-term visibility of public interventions (Cities Alliance strategic planning activity); and, (v) the continued strengthening of COHAB through improvements of its internal capacities and adoption of more soundcommercialpractices. 181, While the optimization of municipal subsidy programs requires technical and in- depth research of the municipal programs, the following summarizes the general direction that these programs should take: (i)decreasingthe depth of subsidies on household financing for mutirGes and conjuntos; (ii)continue strengthening the financial performance of COHAB, by fighting non-paymentand improvingrent collectioninLocaq6o Social programs; (iii) revisingthe parameters of the Bolsa Aluguel program so that the maximum rent-to-incomeratio increases to allow a larger portion of the housingstock to be eligible in the program, especially for the lower income brackets, which should coincide with a reduction of the subsidy amount; (iv) adding social and economic components to urbanizationprograms carriedout inthe city's slum areas; (v) 84 since households that are unable to take a mortgage are disadvantaged in terms of the generosity of public programs and subsidies (e.g., cortiqos and PSH), it would be appropriate to review the subsidy schedule of housing programs in order to improve equity between them and the other households; (vi) since it is currentlyvery difficult to have ready-made informationabout the state of individual projects, their costs, and a view of the strategy followed more globally by COHAB and SEHAB, better project management information systems accessible to all municipal employees are needed; and, (vii) given that often there is no evaluation of municipal interventions, resource allocation should be made based on the relative performance of the various programs. 182. A long-term goal should be ensuringthat, within the consolidated housingstrategy, the segments of the low-income populationthat are unable to afford mortgages are better supported by public programs -this would be an alternative to the provision of finished housingunits.Actions to realignthe allocationof public resources towardhousingneeds include introducingnew programs that target underservedportionsofthe low-incomepopulation.Actions recommended include: (i)devising a strategy for the provision of small, low-cost serviced or semi-serviced land where low-income households can settle as an alternative to the invasion of public land; (ii)implementinghousingmicro-creditprograms and allocating sufficient resources for the scaling-upof such programs, these programs should be based on the outcomes encouraged by Cities Alliance within the implementationand evaluationof pilot programs; (iii) undertakinga study ofthe rentalsector to assess the feasibility of incentive programs inthe supply of affordable rental housing for low-income households; and, (iv) allowing collective rental units to satisfy living standards that are less than the actual minimal standards, in order to directly compete with the cortiqos. 183. Given the magnitude of housing problems in the city and the relative scarcity of public resources that can be mobilized for housing interventions, actively seeking out partnershipswith the private sector should be a priority. Zoning regulations in the MSP should be one area of focus. In particular, the criteria for any redevelopment being planned in zones classifiedas ZEIS seem too stringent to arouse the interest of the private sector. Moreover, it seems that there are too many ZEIS, in the sense that public interventionscan only work with a small fraction of them. A first recommendation is to revisethe classificationof ZEIS zones in the PDE, using the revision of the PDE that will occur in 2007. The aim should be at reducing the number of ZEIS and focusing on the most important or strategic zones. A related recommendation is to revise the parameters that definethe minimal proportions of HIS and HMP within ZEIS perimeters. A second recommendation, made in the context of the considerationof the intervention in the Nova Luz perimeter, is to study the possibility of usingtechniques of land readjustment as an alternativeto expropriation. 184. The search for innovative PPPs should not be limited to ZEIS. The municipal strategy should focus on other areas, including: (i)introducing market incentives for the productionof middle and low-income units in the renovation of the city center; (ii)maximizing the use of existing legal and regulatory instruments (e.g., certificates of development rights as proposed for Paraisopolis and Heliopolis, and exchange of land against forgiving of local tax debts), while evaluating instruments that have already been used (e.g., CEPAC, Urban Operations); (iii) undertaking an inventory of public vacant land in the MSP and devising a strategy and actionplan for its use; and, (iv) finding alternativesto mobilizing the vacant housing stock through studies on this matter and on rentalhousingas inve~tment.~' 87 Studies should aim at assessingthe nature, quality, and location of the vacant housingstock to separate units in categories relatedto the economic motives for whichthe unitswere left vacant. 85 d 3$8 I-' lu E '2.- C 0 C se, .-2 -'3 I-' .e I-' a i 'Fa BE 5 0 i .-C g2a I-' E5k0 B L & 1 ! 3* Es I- a i s vi I I- E$ I-' T :8 m i 8 I- > I 1 i 1I ! z2 ! ! 1 I ! E Ei I i E4 3 : ? 7 Y c ! >L I c L .-C REFERENCES Barnes,W., and Brooks, J. (1999). "World-Class local Economies," in World bank, Business Briefing: World Urban Economic Development. London: World Markets research Cities Alliance (2004). "Integrating the poor: UrbanUpgradingandLandTenure - Regularization Centre, 32-6. inthe City of Sgo Paulo," December2004. Cohen, N. (2000) "Business locationdecision-makingand the cities: Bringing companies back." The BrookingsInstitutionCenter on Urbanand MetropolitanPolicy Working Paper. Cushman & WakefieldHealey & Baker (2004) European Cities Monitor 2004. Mimeo.Available at www.cushmanwakefield.com. Diamond, D., and Rottmann, E. (2004). "Housing Finance in Brazil." Consultant report for the World Bank, January. Doeringer et al. (1987) "Invisible Factors in Local Economic Development." New York and Oxford: Oxford UniversityPress. Doeringer, Peter and David Terkla (1990). "Turning Around Local Economies: Managerial Strategies and Community Assets." Journal of Policy Analysis and Management 9(4 Fall):487-506. Doeringer, Peter and David Terkla (1992). "Japanese Direct Investment and Economic Development Policy." EconomicDevelopmentQuarterly 6(3 August):255-272. Dollar, D., S. Wang, L.C. Xu, A. Shi (2004) "Improving City Competitiveness through the Investment Climate: Ranking 23 Chinese Cities." Available at SSRN: http://ssrn.com/abstract=530744. Duranton, G. and D. Puga (2004) "Micro-Foundations of Urban Agglomeration Economies." In Henderson, J.V. and J.-F. Thisse (eds.), Handbook of Regional and Urban Economics, vo1.4, Elsevier:North-Holland, pp. 2063-2117. Glaeser, E.L. (2005a) "Urban colossus: Why is New York America's largest city?" NBER Glaeser, E.L. (2005b) "Reinventing Boston:1630-2003." Journal of Economic Geography 5, pp Working Paper Series 11398. 119-153. Hamilton, J., J. Thisse and Y. Zenou (2000) "Wage Competition with Heterogeneous Workers and Firms." Journal of Labor Economics 18, pp, 453-472. Imparato, I.and J. Ruster (2003) Slum Upgrading and Participation: Lessons fFom Latin America. World Bank Directionsin DevelopmentPolicy. Jacobi, P., D.B. Segura, and M. Kjellen (1999) "Governmental Responses to Air Pollution: Summary of a Study of the Implementationof Rodizio in Sgo Paulo." Environmentand Urbanization11(1), pp. 79-88. Lall, S.V., R. Funderburg, and T. Yepes (2003) "Location, concentration, and performance of economic activity in Brazil." World Bank Policy ResearchWorkingPaper No. 3268. L e Blanc, D. and E.Rottmann(2006) "Analysis of housingprograms in Sgo Paulo." Mimeo. Marques andTorres (Ed.) (2005) " S ~ OPaulo: segregaggo, pobreza e desigualidades sociais." Sgo Paulo:Senac. Mora, M., and F. Giambiagi (2005) "Federalismo e endividamento subnacional: Uma discussgo sobre a sustentabilidadeda divida estadual e municipal." PEA Texto para Discussgo No. OECD (2004) Territorial Reviews - Montreal Canada.Paris, France: OECD. 1142. Rosenthal, S. and W. C. Strange (2004) "Evidence on the Nature and Sources of Agglomeration Economies." In Henderson, J.V. and J.-F. Thisse (eds.), Handbook o f Regional and UrbanEconomics, vol. 4, Elsevier:North-Holland, pp. 2119-2171. Schumenner, R.W.(1982) Making business location decisions.NJ: Prentice-Hall. 88 Schumenner, R.W. (1991) "Geography and the character and performance of factories." In: Herzog, H.W., and A.M. Scholttmann (ed.) Industry location and public policy. Universityof TennesseePress. SEHAB (2004) "Balanco qualitativede gestsio, 2001-2004." Municipality of Ssio Paulo. Spink, P., M.A. Texeira. R. Clemente (2005) "Metropolitan Governance in Brazil: institutional push or organization pull." Initial country study prepared for the Metropolitan Governanceinthe Americas Study coordinated by Robert T Wilson and Peter Ward: LDJ School of Public Affairs, UniversityofTexas, Austin. Tendler, J. (2000) "The Economic Wars between the States." MIT-Department of Urban Studiesmank of the Northeast Project. Presented at the OECD/State Government of Ceara Meeting on Foreign Direct Investment and Regional Development, Fortaleza, 12 December2002. Torres, H., H. Alves, and M.A. Oliveira (2005) "Ssio Paulo peri-urban dynamics: some social causes and environmental consequences." Paper presented in the XXV International PopulationConference, Tours-France, July 18-23. World Bank (2002) "Brazil National Housing Policy - Maintaining Momentum for reforms." Policy notes on housing Finance, Subsidies and Land Policies, Draft internal version, June 26. World Bank (2004) "Water Quality and PollutionControl Project." Report No. 28962. World Bank (2005a) "Brazil: Program Document for a Programmatic Loan for Sustainable and EquitableGrowth: HousingSector Reform." May 11. World Bank (2005b) Doing Business Report. Washington D.C.: The World Bank. World Bank (200%) "Brazil InvestmentClimate Assessment, Vol. 11." Mimeo. World Bank (2006a) DoingBusiness in Brazil. Washington D.C.: The World Bank. World Bank (2006b) "Crime, Violence and Economic Development in Brazil: Elements for EffectivePublic Policy." Mimeo. Yusuf, S., and K. Nabeshima (2006) "Two decades of reform: the changing organization dynamics of Chinese industrialfirms." World Bank Policy Research Working Paper No. 3806. 89 Annex 1: Survey Methodology The 447 firms were selected among the 1,642 firms interviewed for the Brazil Investment Climate Assessment that the World Bank carried out in 2003. Firms were selected according to two methods. All establishments located in the Ssio Paulo MetropolitanRegion and included in the ICA sample were included in this sample as well. Establishments located in the rest of Brazilian Southeast(Minas Gerais, Rio de Janeiro, and Ssio Paulo), South (Parana, Rio Grande do Sul, and Santa Catarina), and Northeast (Bahia, Maranhao, Ceara, and Paraiba) were randomly selected. We selected these regions because they provide a comparison among the industrial Southeast, a growingregion inthe South, andthe less developedNortheast. The survey was carried on over the periodMarch-May2006. The interviewswere conducted in Portuguese over the phone using a Computer Assisted Telephone Interview (CATI) system. The questionnaire consisted of 30 questions. The questionnaire was pre-tested with a group of 12 Brazilianfirms, and as a result, a number of changeswere made. Some of the data gathered with the questionnaire include: Locationcharacteristics (nature ofthe space, where is located, when operations beganthere); Why the firm decided to locate where it did, and main advantages and disadvantages of the location; Location process (how long, how the government responded, alternatives taken seriously into consideration). Different surveys were targeted to plants that had recently moved from MRSP. Along with the more general questions appropriate to all types of plants, mover plants were asked pertinent questions about previous location. By interviewing plants that moved out MRSP, we tried to identify the factors that inthe MRSP may push away manufacturingplants. Survey responses Establishments in the sample are mainly small firms (48 percent of establishments have 99 employees or less), low-tech (64 percent of establishments are in a low-technology sector), and they rarely relocate(only a quarter of the sample relocated over time). The charts below show survey responsesby region, size, sector, andrelocation. 90 Was thfs e ~ # b i f ~ haiwsys~beenlocated in this general ~ e ~ vicintty? High Technolo 36% Y 91 Table A.1 Plant characteristicsby region Size Micro 104 24 22 21 28 Small 215 50 43 50 50 Medium 101 20 31 22 17 Large 25 6 4 7 5 Sector High- 162 50 41 36 16 Low- I 285 50 59 64 84 Technology Urban Location Downtown 56 13 12 13 12 Near downtown 167 41 41 43 32 Near highway 162 29 34 35 35 Other city in 53 11 11 7 21 Other 1 9 2 3 3 1 MRSP firms are older and more "locally" oriented than plants locatedelsewhere since they are more likely to buy and sell product within the metropolitan region. We compared firms located in MRSP to firms located in the other regions by looking at the means of some variables for each of them. Table A.2 presents the means of age of the plant, input market, output market, and relocation for each region. First, we can observe that firms in MRSP are older than establishments located elsewhere, their average age being 22 year. Firms in the South and Northeast are statistically significant younger than firms in MRSP. Second, firms in MRSP tend to buy their inputs in the same metropolitan area (59%) or within the state (45%). These proportions are statistically significant larger than the proportions in the other regions. By the same token, a large majority o f firms in the South, Southeast, and Northeast buy their inputs in other states, and their proportions are statistically significant larger than the share of MRSP firms that buys inputs in other states (at 1% level). In terms of final markets, the majority of firms in MRSP sell their products within the metropolitan area (59%), other states in Brazil (53%), or other locations within the state (40%). With respect to relocation, 30% of firms that relocated are in the Southeast, a proportion that is statistically significant larger than the one for MRSP (at the 10% level). In MRSP, an equal share of firms that relocated where previously established either in the same city or within the metropolitan area. In the other regions, a vast majority of firms where previously located in the same city. This indicates that firms tend to move locally and support the idea that local governments can make a difference inattracting economic activities. 92 Table A.2 Differences inplant characteristicsfor regions Other locations I 288 39.47 I 71.43*** I 72.50*** I 75.25 *** in Brazil Other locations 41 13.16 8.93 11-67 1.98*** in the world Output Market Same 222 58.77 37.50*** 60.00 40.59*** metropolitan Other locations I 149 40.35 27.68** 46.67 I 15.84*** within state Other locations II 287 52.63 69.17*** I 55.45 Same region (different state) -- 0 3 3 0 Out of the region _ _ 4 7 3 5 I I I I I I Notes: Sample size: 447. Percentages refer to n, divided by the total number of firms in region i. Sub- samples: RMSP (114 f m s ) , South (112 firms), Southeast (120 fms), Northeast (101 f m s ) . * Proportion i s statistically different from RMSP proportion's at the 10% level. **Proportion i s statistically different from RMSP proportion's at the 5% level. ***Proportion is statistically different from RMSP proportion's at the 1%level. 93 W 2 2 I- 9 N N ** 2m rr) **I- 2 C! rr) F 00 I- N m ** 2r- - rr) d I- W 2 N m c? N N m m m m I- d W 'c! N d N 2 3 2N 0 d Annex 3: LocationDeterminantsof Productivity In chapter 4 we do an estimationof total,productivity factor (TFP) usingboth Cobb-Douglas and Translog functional forms. Data is from the Pesquisa de Atividade Econ6micaPaulista (PAEP). It is a firm level survey of active and registeredfirms operating in2001 of State of the Sgo Paulo. Research undertook by Fundaqgo Sistema Estadual de Analise de Dados (SEADE). Responses refer to 2001 between July 2002 and June 2003. It is composed of slightly different questionnaires for Industry, construction, commerce, services, and banks. In commerce and servicethere are two versions ofthe questionnaire: complete for firms with 20 or more employees and simplified for the rest. It also has a different questionnaire for headquarters and local establishments. Here analyses were done for establishments in manufacturingand services using the complete questionnaire data. The Cobb-Douglas specificationis as follows: logy = a: +PIlogK, p2logMI+p3logL,+U, + While the Translogspecificationis: log = a: p1logK,+p2 logM,+p3 1% L,+ + The specifictopic of interest was the impact of locationdeterminants (LD)ofTFP. By transformingequations (1) and (2) into (4) and(5), respectivelywe are able to capture that: logy; = a: p,logK, p2 logM, +p3 logL,+ + + Location determinants where treated as dummies. In PAEP questionnaire local units where inquired about importance o f specific location determinants and grade them from 1 to 4 in importance, being category 4 equivalent to highly important. Here we recode grades 1 and 2 to 0 and grades 3 and 4 to 1. Inmanufacturinglocation determinants are: ur034 d I low labor costs ur035 d I qualification and training o f employees ur036 d . proximity to suppliers ur037 d proximity to consumers ur039 d access to transport system ur040 d telecommunication infrastructure ur041 d urban infrastructure ur042 d life quality for employees ur043 d proximity to research and technological diffusion centers ur044 d infrastructure of technical services to support industry ur045 d federal and state fiscal incentives ur046 d municipal fiscal incentives ur047 d low land prices Inservices the followingwere asked: ur038 d low labor costs ur048 d IIDroximitv to sumliers ur049 d 1 proximity to consumers ur050 d I proximity to other unitso fthe firm ur063 d existence o f specialized support services ur064 d hotels, convention centers and restaurant infrastructure ur065 d proximity of central offices o f other firms Iur066 d proximity of public administration entities and public firms ur067 d 1 qualification and training o f employees For some determinants, a region categorical variable was included as categorical dummy as well as an interacting term with other variables. The regions and their value are: 96 1 1 SBo Paulo, without ABC (Santo Andre, SBo Bernardo and SBo Caetano) 2 ABC (Santo Andre, SBo Bernardo and SBo Caetano) 3 Metropolitan Region of Baixada Santista and Registro 4 SBo Jose dos CamDos 5 Sorocaba 6 Campinas 7 North of Estado 8 West of Estado Local units were grouped for regressions analysis according to two different classifications for Industry analysis and only one for Services. These categorizations come from groupings of major CNAE (National Classification of Economic Activities) classes. One categorization for Industrywas relatedto category of use (catuso). They were: 0 Not Classified 1 Non-durable consumption goods 2 Intermediate Goods 3 Capital goods and durable consumption goods The other categorization for Industryalso usedfor Services was a simplified CNAE: 0 Agriculture and Extractive 1 Medium qualification Industry 2 Low qualification Industry 3 High qualification Industry 4 Public utility Services 5 Construction 6 Commerce in General 7 Commerce and low qualification services 8 Firm-oriented service 9 Public Service 10 Associative, Recreational activities and international organizations 99 I Bad defined activities and others Determinantsof Employmentby Sector. As we saw, SBo Paulo is losing employment and production to other regions. However, some sectors are staying. We have discussed some alternative hypothesis explaining why the sector may be leaving the city and illustratedthe discussion with some stylized factors. However, we do not know which factors are actually playing a significant role in the sectors decision. To analyze the behavior of the sector let us departure from the simplified dynamics proposed in the neo- classical growth mode1,88 developed in tandem by Solow (1956) and Swan (1956), i.e that the variation of capital stock over time is given by: Kl =I,-S,K, =s,Y, -6,K, 88For hrtherdetails, see Barro and Sala-i-Martin (1995). 97 where Ki is the stock of capital in sector i,Ii is total investment, Siis the rate of capital depreciation in sector i,si is the savings rate, Y is the total economic production and the variation in stock is considered investment.89 The notation of a point above the variable represents its change with respect to time as usual. If we assume that the production function has constant returns to scale, we can write the production function in terms of the per capita ratio between capital and labor: yi = fi(ki), where the lowercase variables denote that the particular variable is in "per capita" terms90, in other words, yi YiLi and ki = Ki/Li. Since we will treat each sector separately we can drop the industry index for now. We will return to this point later. Based on the definition of ki (hereafter k) and yi (hereafter y) it is easy to verify that: k = - d(K/L)- I? --- nk dt L (7) Ifwe divide equation (6) byLand substitute it into equation (7),we havethat: k = sf(k) -(n +6 ) k (8) The stationary state can be defined as the point in time in which the change in capital stock is constant. In particular, the stationary state inthe Solow-Swan model o occurs when the change in capital stock is zero, in other words: sf (k*) = ( n S)k + * (9) where the variables with asterisks denote the steady-state values. Dividing both sides of equation (3) by k we have that: yk = k / k = sf(k)/ k - ( n +S) (10) where Yk is the percentagevariation in capital stock across time. Ifwe substitute the savings rate implicit in equation (9) into equation (10) we have that: Equation (6) implies that Yk will be zero when k = k*, in other words, in the steady-state. Ifwe assume a Cobb-Douglasgl production function, J' = ,the variation in capital stock can be written as: 89 As i s well known, giventhis hypothesis, I ISYis simply an accounting identity. 90 To be more precise, the aggregate variables defined with respect to the labor force may be different to those in terms of population and this difference may not be constant across time depending on the demographic stage. 9' This result is valid for any production function exhibiting constant returns to scale. Using a Cobb- Douglasproductionfunction gives us an analytic solutionthat facilitates the interpretationof results. 98 Adoptinga log-linear approximationof equation (12) giventhe stationary state we havethat: where p = For the Cobb-Douglas function, we can estimate that ln(y/y*) = p ln(k/k*). Afterwards, we can derive this same approximationfor the percentage change in the employment across time, Y L : yL 1-/3 ln(L/ L*) (14) Equation(14) is adifferentialequation92 whose solutionis: From equation (15), it is possible to conclude that p is a good measure of the speed of Convergence.If p is very high, the second term on the right hand side of equation (10) will tend towards zero and Lt = L*. In reality, if p