Professional Documents
Culture Documents
Bedt Prsctices in Slum Improvement Brazil
Bedt Prsctices in Slum Improvement Brazil
August 2008
August 2008
This report was written for Development Innovations Group (DIG), by David A. Smith, of the Affordable Housing Institute, with substantial assistance from: Dr. Janaki Blum, Director of the Affordable Housing Institute Motlalepula Mmesi, Research Associate, AHI Ashleigh Flesher of the Development Innovations Group (DIG) participated along with David Smith in the field work for this case study. In addition, the authors benefited from extensive commentary, dialog, insight, and courtesy Jessica Budds, the Open University in London; Rodrigo Moraes, Diagonal Urbana, So Paulo; and Anaclaudia Rossbach, Interaco, So Paulo. Any errors are the authors'.
August 2008
TABLE OF CONTENTS
Case Study Purpose.....................................................................................................................4
So Paulo: an urban profile.............................................................................................4 a. The nation, state, and city ....................................................................................... 4 b. History and its legacy of multi-ethnicity ................................................................ 5 c. The city today .......................................................................................................... 5 d. So Paulo's economy and its implications for the housing market ...................... 6 II. The problem of informal settlements in So Paulo....................................................6 a. Macroeconomics, the modern era, and Lula ......................................................... 6 b. Formal housing ....................................................................................................... 7 c. Formal housing finance ......................................................................................... 7 d. Favelas: already the largest form of housing, and still growing .......................... 8 e. The economics of So Paulo's slum population .................................................... 9 III. Financing slum upgrading in So Paulo.................................................................10 a. The 'invisibility' of slums ...................................................................................... 10 b. Eviction-relocation efforts: So Paulo and Brazil through 1990 (Cingapura) .. 11 c. Financing as the outcome of macroeconomic policy .......................................... 13 d. Cingapura: in situ demolition/ rebuilding new construction walkup flats ......... 15 e. The Guarapiranga approach: water and sanitation infrastructure retrofitted .. 20 f. Regularization, resolo, cortios, and more........................................................... 22 IV. Results, Impacts and Opportunities for Replication...........................................26 a. Workarounds and their consequences ................................................................. 26 b. The cognitive dissonance of informal settlements: illegal but un-evictable ....... 27 c. Cingapura: a kinder, gentler, inadvertent eviction?............................................ 28 d. Guarapiranga: the Hippocratic-oath ('do no harm') urbanization .................... 29 e. Financing slum upgrading: government pays non-recoverable cost.................. 29 f. Markets always clear; creating economic value from 'dead capital' .................. 29 g. Whither housing finance, slum upgrading, and microfinance in So Paulo? .. 30 V. Insights and takeaways from the So Paulo experience........................................30 a. Slum upgrading represents predominantly non-recoverable costs ..................... 30 b. When upgrading slums, spend public resources on public spaces, not private homes ............................................................................................................................ 31 c. Upgrading requires compromising formal standards with informal reality....... 32 d. Completing the buildings is not enough; legal and social follow up is essential 32 e. Financing schemes need to be protected against a shift in government ............ 33 f. Slum upgrading is itself an agency that makes political change ........................ 33 g. Slum upgrading requires specialized lenders, including government lenders ... 34 h. Personal slum dweller finance is a space that will be occupied by someone. 34 i. Upgraded communities are ripe for housing microfinance ................................ 34 VI. Conclusion.....................................................................................................................35 a. Cingapura and Guarapiranga are successes ....................................................... 35 b. Slum upgrading prepares the economic conditions the climate for housing microfinance................................................................................................................. 35 I.
Best Practices in Slum Improvement: The Case of Sao Paulo, Brazil Case Study Purpose 1
August 2008
On the face of it, So Paulo would appear a market ripe and over-ripe for housing microfinance: A booming economy creating massive rich-poor gaps in housing supply. Massive and continuing immigration to the city leading to an explosion of favelas. Approaching a decade of stable macroeconomic policy creating viable long-term lending for the first time. A pro-active municipal government and a legacy of housing affordability experiments.
Yet housing microfinance is absent, or at least in its infancy. When something that ought to be prevalent is missing, that is a curious incident, and its absence is extremely significant. To understand why, and what principles the So Paulo experience offers for other situations, requires understanding So Paulo's challenges of urbanization and slum upgrading, including:
So Paulo's position within Brazil and Latin America. Brazil's history of hyperinflation and its consequences. The expansion of So Paulo as it occurred contemporaneous with the hyperinflationary period. So Paulo's legacy of municipal programs.
I. a.
The largest nation in South America both in area (8,500,000 km) and population (182,000,000) Brazil is the continent's natural economic and demographic powerhouse. It is the smallest of four nations (China, India and Russia being the others, collectively BRIC) that make up the quarter most observers believe will be the world's rising powers in the 21st century. Under its Federal government, Brazil is divided into 27 states (including the DF), of which So Paulo state is the largest (40,000,000 people), strongest economically (more
1
This case study was prepared by David A. Smith, using information we gathered from studies, reports, and extensive field interviews that were conducted 11-15 May, 2008, in So Paulo. (A list of principal sources is provided in Sources ) As we have not independently verified any of the statements made by our sources, there is the likelihood that other sources would see events differently, and may have facts that contradict those set forth herein.
August 2008
than 18% of the nation's GDP flows through metro So Paulo), richest (per capita), and most politically important. Within SP state, metropolitan So Paulo city represents almost half the population (city population of 11,000,000, metro-area population close to 20,000,000), making it one of the largest metropolitan areas in the world. Much of this growth has been recent, with So Paulo doubling in population between 1960 and 1985. In the last twenty years, the growth has been peripheral, as the urbanizing immigrating population has moved beyond the city's original footprint, stressing infrastructure, housing, political systems, and natural resources. b. History and its legacy of multi-ethnicity
Founded as a plantation colony, So Paulo was originally entirely agricultural, with vast slave populations who labored on the coffee plantations. After slavery was abolished in 1888, Brazil encouraged large waves of European immigration (Portugal, Spain, Italy, Germany), leading to today's multi-ethnic population. While there are numerous exceptions, generally speaking race and economic class correlate: upper-income groups more European in heritage, lower-income groups more African. c. The city today
Founded at the confluence of two rivers 70 kilometers from the sea (the Tiete and the Pinheiros), So Paulo sits at 800 meters' elevation. Climate is temperate year-round freezes are virtually unknown, it never snows, and summer temperatures seldom exceed 30 C. The climate is a blessing for inhabitants but creates an urbanization challenge, as ramshackle insecure self-built housing that simply shelters people from the rain is at least notionally habitable. Rapid urbanization in a geographically challenging setting has stressed the ecology, leading to:
Water pollution. Vast infrastructure projects have been undertaken to channel and to clean up the Tiete and the Pinheiros. Water supply. Like many fast-growing cities, So Paulo has had to construct an extensive network of reservoirs and aqueducts, leading among other initiatives to the Guarapiranga cleanup program described below. Transportation challenges. While So Paulo has 60 kilometers of metro and 260 kilometers of railway, most transportation is by auto, and traffic jams are endemic. This places a huge premium, especially for the poor, in living close to where they can earn their income (formally or informally). 'Sprawl.' While So Paulo does not 'sprawl' in the sense of low-density use, its growth has been haphazard, with a lack of adequate infrastructure, leading to congestion and wildly varying land prices based on locations in submarkets. High crime, especially in favelas and lower-income neighborhoods.
August 2008
High levels of informal employment and domestic workers, with service work plentiful if low-paying. So Paulo's economy and its implications for the housing market
d.
Today's So Paulo economy has shifted from agriculture to manufacturing and finance, with So Paulo the business hub of Latin America. The Brazilian stock market, Bovespa, is headquartered in So Paulo, as are most of Brazil's largest corporations and many multinationals. In terms of per-capita income, So Paulo is the richest large city in Brazil, with a GDP equal to 1one-eighth of Brazil's national GDP. It continuously attracts immigrants from the poorer and more rural northeast of Brazil, including many 'geographical bachelors,' married men or unmarried young men who move to the city to earn cash money which they remit to their family back in 'the country.' The continuous immigration, coupled with the increasing value of certain pockets of downtown or Central Business District So Paulo, mean that some favela pockets become so valuable it is worthwhile paying their residents to relocate farther out or 'anywhere but here,' leading to periodic waves of outward relocation of the slum population, with implications for transportation and the water supply, as seen in the Guarapiranga and Billings reservoir upgrading programs.
II. The problem of informal settlements in So Paulo For representative photographs of life in So Paulo's favelas, see Exhibit 1, pictures. a. Macroeconomics, the modern era, and Lula
Hyperinflation and modern macroeconomic policy. Like many emerging nations, Brazil has experienced periods of hyperinflation, whose effect has been to squelch or suppress development of long-term stable interest rates essential to the emergence of a proper housing finance system. As an economist reflecting on it commented2 : Imagine that your rent doubled every 10 weeks. That your credit card charged 25% a month interest. That food and clothes went up 40% a month. That the value of your savings declined 2000% in a year! This was Brazil for ten years, from 1987 to 1997. During those ten years 40% of GNP was eaten up by inflation, and everyone got rid of cash as fast as possible, because it lost value in your pocket. No one saved money. And the majority of people were reduced to buying only the essentials of life, which devastated whole industries that produced all kinds of optional goods and services. Brazil only began taming inflation in 1994, with the launching of Plano Real, the economic stabilization process. Brazil's emergence as a financial market and housing market dates from this interval.
2
August 2008
The election and administration of Lula da Silva. In 2001, there occurred the most remarkable event in Brazil's modern history: the election as President of Luiz Incio Lula da Silva. In many ways Brazil's Lech Walesa, Lula had little formal education, quit school after the fourth grad, and worked as a So Paulo shoeshine boy and street vendor. In 1978, at 33, he was elected president of the Steel Workers union, rising to be president of the Central nica dos Trabalhadores (CUT), which he co-founded in 1983. In 1980 he and several others founded the PT or Workers' party, left-wing and progressive. With Brazil's gradual emergence from de facto military dictatorship (highlighted by direct presidential elections in 1989), Lula rose politically, becoming a Congressman (1986) from So Paulo, presidential candidate (1989, 1994, and 1998), his power base being low-income neighborhoods and favelas. In 2002, he was elected Brazil's president. Lula's election worried many observers, including Brazil's upper classes and outside economic advisors, for his background (no formal education) and orientation (strong populism) suggested Lula might reverse the monetary policy that was improving Brazil's economy. In fact, though still a social populist, Lula maintained the macroeconomic approach, submitting conservative budgets and avoiding deficit spending. As a consequence, Brazil's economy has continued to boom. If there is a lesson from Brazil's last forty years, first of dictatorship and then transitory and full democracy, it is that politicians committed to true democracy are much more likely to develop sound macroeconomic policies that militarists. Lula's 2006 re-election and his commitment to retire in 2011 after his second term in accordance with Brazil's constitution have undergirded capital market confidence in Brazil as a place to invest. Inflation today runs in the range of 6-10% and most market analysts are bullish on Brazil. b. Formal housing
Hyperinflationary environments are death to savings, death to financial innovation, and death to housing as a financial asset. Brazil's recovery from addiction to hyperinflation is so recent, in real estate terms, that its market displays a curious underdevelopment of expected real estate forms. Despite a 74% homeownership rate, the mortgage market is only 2% of GDP (the US is historically 75% of GDP, and with the subprime run-up has topped 80% of GDP). Fewer than 40% of formally sold homes nationwide (fewer than 25% in So Paulo) use mortgage finance; the rest are bought with cash. In So Paulo, less than one-quarter of formal sales come from mortgage financing. c. Formal housing finance
Formal finance is available but not widely used (see previous section). The Housing Finance System (SFH) was established in 1964, and the Caixa Econmica Federal (CEF) emerged as a mortgage origination source in the mid-1980s (when hyperinflation rendered this largely a futile exercise). Still, with mortgage financing in its infancy (as of 2005 there were only six mortgage companies operating), mortgage lending is dominated
August 2008
not by private firms but by Caixa Econmica Federal the second largest bank in Brazil which originates 70% of all mortgages in Brazil. Most Brazilian borrowing or savings rates are quoted as a spread above the TR, which was introduced in the mid-1990s (that is, during hyperinflation) as a means of inflationindexing returns. Interest rates for mortgage products remain very high in real terms (anywhere from 8% to 16% above the TR), although Caixa and various other state-influenced organizations can offer much lower rates for target customer groups. The result is a crazy quilt of financing alternatives (see Exhibit 2 attached, and Section 5A below) that disincentivize even formal buyers of quality homes from tapping mortgage finance. Further, lack of up-market acceptance of mortgage financing means that the banks have ample expansion areas in the Class A and B socioeconomic groups and have no need, yet, to move down-market into Classes C, D, and E, who remain largely outside the formal financing system.
d. Favelas: already the largest form of housing, and still growing As shown by the municipal map at right, So Paulo's favelas are mainly on its periphery; around the reservoirs south of the city, and in the eastern or northern rolling hill areas. In relative terms compared with large peers So Paulo's favelas are higher-quality housing than many slum areas in the global south. Over-simplifying for convenience favelas vary dramatically in quality, especially between those established for longer periods and newer settlements the favelas show a physical configuration and space allowance that compares:
Perhaps slightly better than Alexandra in Johannesburg (South Africa) Generally superior to the accommodations in Dharavi, Mumbai (India) Vastly superior to Kibera, Nairobi (Kenya)
August 2008
New construction properties under the Cingapura program (see Section 4D) are equivalent in configuration to the Mumbai co-op high-rises sponsored by India's National Slum Dwellers Federation, and larger 48 m versus 22 m. In terms of locations, So Paulo's favelas are generally more remote to work and the central business district than Dharavi (which is exceptional), comparable to Johannesburg and Cape Town, and further out than Kibera. Also, they are less geographically isolated, blending seamlessly into lower-class more formalized neighborhoods in some areas, although the Cingapura properties have an enclave feel at least at the site level. Commuting from favela to jobs is more difficult than in similar cities like Johannesburg or Cairo, although the sheer scale of So Paulo and its difficult, hilly terrain, are contributing factors as much as the lack of transportation infrastructure. e. The economics of So Paulo's slum population Income Classes A-B-C-D-E. For statistical purposes 3 , Brazil classifies households economically by income relative to the minimum wage (currently about US $250 per month 4 ), as follows: Social class Class A Class B Class C Class D Class E Monthly income Over 25 times the minimum wage Between 10 and 25 times the minimum wage Between 4 and 10 times the minimum wage Between 2 and 4 times the minimum wage Up to 2 times the minimum wage
More than half the population is in Class E, and most of those are informally or irregularly employed. Markets will generally produce as 'affordable' housing what can be afforded by middleclass households. This is neither elitism nor discrimination; rather, it reflects that the value of urban land equals the difference between what a developed home is worth, and what it costs to develop that home. Class E households and what they can afford. As a general rule, people around the world can afford to pay 35% to 40% of their income for housing (including utilities). Translating this into affordable levels for Brazilians yields the following affordable amounts (at the time the Cingapura properties were undertaken):
Set by the Brazilian Association of Research Companies (ABEP - Associao Brasileira de Empresas de Pequisas). Source: http://www.pch.gc.ca/progs/ac-ca/progs/rc-tr/market/publications/music_bresil/1_e.cfm. So Paulo's minimum wages are about 20% higher than Brazil's national average.
4
August 2008
Income: multiple of minwage Monthly income (R$) Affordable at 25% Affordable at 40% Utilities per month Condo fee (maintenance) Costs of occupancy Minimum for rent/ loan repay Maximum for rent/ loan repay
1 170 45 70 45 50 95
2 340 85 135 45 50 95
(50) (25)
(10) 40
35 110
120 245
205 380
290 515
Class D customers those between 4 and 6 minimum wages can afford some occupancy payment; those in Class E between 1 and 3 minimum wages can afford little if any payment for occupancy (as distinct from utilities or maintenance). The result is that Class D and particularly Class E residents cannot afford market-quality housing, so they live where they can, most of them in favelas of varying quality. Some, the more mature neighborhoods, are good quality smaller flats and houses whose main deficiency is a lack of municipal services; others, often the newest settlements and the most recent land invasions, live in newly constructed makeshift housing unsuitable for human habitation.
III. Financing slum upgrading in So Paulo Like many things we find abstractly unpleasant, slums are often willfully overlooked by local communities and their municipal or state governments until their sheer scale and negative externalities (traffic, pollution, lack of sanitation, crime) make them impossible to ignore. a. The 'invisibility' of slums On the municipal map of Nairobi, the Kibera dam shows a lake and green space. Google Earth reveals a sprawling dense neighborhood that is home to 1 in 5 Nairobians. Thus it is with slums; for most of formal society the world of payroll stubs, checking accounts, birth certificates and postal addresses the slum is a place not to be seen. Initially the formal city ignores the favelas even as those who live in the favelas 'ignore the city' as a political organism. They may be the people who keep the city going, through the low-paid and mundane jobs that they do domestic staff, street cleaners, factory workers, transport drivers but they receive few if any city services and claim no rights.
August 2008
As favelas expand, however, they become too large to ignore, intruding (in Mumbai) on railway lines, riverbanks, power cables, cemeteries, parks, and monuments. As a more general rule, slums sprout in land of least-risk-of-eviction, which is usually a mixture of least-alternate-value and least-motivated owner. Least-alternate-value. Land that is low-lying, hard to develop, hilly, unstable, or noxious. Least-motivated-owner. Private entities land-banking against a possible future development, or government bodies with procedural or administrative inhibitions against enforcement.
Then the city adopts one of three postures: Expulsion. The city seeks to make slum dwellers disappear, not just from their current location but from the city entirely. Demolition/rebuilding. Under the rubric of urban renewal, the city demolishes the entire slum neighborhood, replacing it with clean new apartments usually concrete, usually walkup flats. Upgrading and infrastructure. Existing homes are left in place, but the neighborhood is 'urbanized' streets are aligned and widened, drainage is improved, homes are connected with the water and sanitation grids. This last approach usually also parallels efforts to secure land tenure and regularize or legalize ownership and title, bringing records into harmony with physical reality.
It's important to see these three actions as stemming from three distinct perspectives regarding slum dwellers: Expulsion. "You are not good people and we want you out of our city." Demolition/ rebuilding. "Your homes and your informal community are not good and we will replace them with a nice new one." Infrastructure. "Your home and informal community are good, it's our city that is deficient, and we will bring you the same benefits other citizens receive."
In 1968, the population of So Paulo was roughly 5,500,000; in the forty years since then, it has grown at a 3% annual rate, to a current population of over 18,000,000, one of the world's five largest metropolises. In that forty-year period, So Paulo has pursued all three strategies, in sequence, with varying and instructive results. b. Eviction-relocation efforts: So Paulo and Brazil through 1990 (Cingapura) During the twenty years between the mid-1960's and the mid-1980's, Brazil had a 'benevolent' military dictatorship and recurring bouts of hyperinflation. Whether or not these two are related, they contributed to a highly schizophrenic, start-and-stop-and-startagain approach to macroeconomic fiscal policy, housing policy, the problems of slums.
August 2008
COHAB: Direct new construction of public housing. Under programs implemented by the Companie Metropolitana de Habitacao, COHAB, the municipality of So Paulo constructed roughly 150,000 flats, mid-rise walkups. All were built on public land, by private builders. Once constructed, they were owned and operated by the municipality, with results that are virtually universal around the world namely inadequate maintenance, uneven enforcement of the rules, patronage political and otherwise, gradual physical deterioration, and a slow reversion to what in So Paulo has become known as the 'verticalized slum.' Mutiro: build-it-yourself formalized housing. A romantic-proletarian concept, Mutiro authorized and financed individual slum dwellers to build their own homes. While the idea was laudable, the program was beset by practical difficulties. Land that was expropriated to assemble parcels for Mutiro sites was intended to be compensated at a market price, making the resulting homes expensive even with self-built labor. (Further, government was often slow to pay landowners.) Families were not trained in construction; money went missing, not stolen but lost. Costs mounted. Many homes were never completed. Changes in municipal administration also led to abrupt discontinuation of some properties. Relocation payments. As a general rule, slum upgrading requires either a reduction in density (families per hectare) or dramatic verticalization (one-story shacks replaced by five-story walkups), or both. Achieving lower density, either permanent or merely temporary, further requires clearing families off-site. While in some later programs relocation housing was developed, in COHAB and Mutiro (as well as numerous other schemes), some municipal governments adopted the least-aggravation approach of paying slum dwellers to relocate. For R$ 1,000 to R$ 5,000, the municipality would buy the existing dwellings on condition that the residents then relocated anywhere else. This short-sighted strategy had predictable consequences. Since the people being relocated had come to the city to make more money, when they moved off-site, it was not to return to the countryside, but rather to pick another greenfield area to invade. Relocation therefore did not eliminate the favela problem; rather it displaced it from better neighborhoods to worse ones. In fairness to the municipality, with the rapid urban immigration occurring throughout this period, it would have been all but impossible to keep up with the growth and to prevent the spread of favelas; however the municipality never got in front of the problem by designating new areas for low-income development and laying suitable sites-andservices infrastructure. Instead it played catch-up, and inconsistently at that. Making new slums: which land gets invaded? If some slum dwellers are merely being moved about ("blowing your leaves into your neighbor's yard," as one observer put it), as their former favelas were formalized or improved, where would they relocate?
August 2008
Fundamentally, to the line-of-least-resistance land, which is usually (a) publicly owned, and (b) of poor quality. Public more often than private. To create a new favela, one needs unoccupied and undeveloped land. Public land, particularly if remote from the city center, is a convenient place to 'encourage' residents to relocate when inner-city land has become valuable enough to develop. Green-space laws conservation easements, wetlands, highway or railway rights of way can be meaningless to slum dwellers, who need a place to live at low cost, and who can relocate if they are forced to. Laws intend to beautify the city by reserving land for green space thus can have the perverse effect of assuring it will be a favela. Conversely, the uncertainty of tenure makes investing in the home risky (either for home owners or for slum landlords). Poor quality. Slums are built on low-value land; since markets are efficient, this means land that is low-lying, prone to water incursion, contaminated, unstable, or downwind of something unpleasant (like a pulp plant). In So Paulo, municipal administrations tacitly encouraged relocation by picking an area for 'temporary' resettlement. Heliopolis began as a temporary home of several hundred families; today it is one of So Paulo's largest favelas, with 100,000 people. c. Financing as the outcome of macroeconomic policy 1. Junta and hyper-inflation: a legacy of capital-finance avoidance. Today's Brazilian and So Paulo financing environment is the legacy of two main phenomena: (a) bouts of rampant hyperinflation, which created a profound distrust of long-term financing of any kind, and (b) periodic financial edicts that have created capital flows theoretically available for housing. To begin with, nearly all deposits and loans are denominated in rates plus inflation. The inflation indices are imprecise, and some appear to be downwardly biased (that is, less than true inflation), making cash savings a losing proposition. Similarly, until hyperinflation calmed down with the Lula da Silva government, banks made excellent returns providing capital to expanding industrial corporations, or simply buying inflationindexed government securities, and regarded finance of housing as a complicated and low-profitability business. Under Brazilian law, formal employers are required to contribute 8% of an employee's salary into a social security fund, the FGTS. This housing security account, akin to a retirement account (in the US, a 401k or an IRA), is earmarked for the depositor employee, and available for the employee on defined events, like retirement, severe illness, purchase of a home, or departure from the company (when the employer is supposed to transfer or pay it out). The effect is to create a vast pool of investment capital on deposit, where it is susceptible to government edicts regarding its use. One such is a law requiring banks to put 1% of their deposits into microfinance and similar loans. In point of fact, few do.
August 2008
2. The dog that didn't bark: absence of residential mortgage finance. Classical residential lending, as distinct from consumer finance, is in its infancy (less than 2% of GDP). High-end home buyers generally pay cash (another legacy of the history of hyperinflation) or finance on a very short-term basis without regard to a mortgage. Mortgage finance development is further inhibited by a 'homestead' law that provides that a foreclosed owner may not be dispossessed of their sole residence. The effect, naturally, is to make it all but impossible for banks to loan on classical residential mortgage finance, so the marketplace as adopted a lease-to-own work-around. Under the lease-to-own concept, the parties sign a contract under which: The prospective buyer becomes a renter. Rent payments are the same monthly amount as would arise in an amortizing loan. When the final rent payment is made, title transfers to the buyer.
The effect of this work-around is to give home buyers fewer rights than they would have under a traditional mortgage loan system; thus a well-intended law has precisely the opposite result. 3. Municipal improvements: leaning on multilaterals. There is no national housing agency; the national housing bank was shut down in 1986, a scapegoat for failed public-housing development policies (COHAB, see below). (The recently created Ministry of Cities is working on a National Housing Plan that is due to be announced in May, 2008.) There is no national financing for slum upgrading or favela urbanization; instead the problem is owned by states (Brazil has 26) or municipalities (of which So Paulo is by far the largest). The Caixa Economic Federal (National Savings Bank) pays 6% above inflation on deposits, and is a distribution source for loans from the Federal government to municipalities. The result is another work-around, using international capital providers as the municipal financing source. Projects that are more expensive than can be funded with a normal annual operating budget in other words, most of them have to tap an international capital provider, such as the World Bank or the Inter-American Development Bank (IDB, in Portuguese BID). As discussed below, the introduction of foreign providers overlays new requirements and conflicting priorities that complicate programs (and are directly related to the perceived failure of the Cingapura program, described below). 4. The explosion of consumer finance. As a half-dozen countries Korea, Thailand, Turkey, Brazil, Mexico discovered during the 1990s, hyperinflation isn't just ruinous for the growth of a nation's productionoriented economy, it also stunts the consumer-oriented economy because people cannot go to normal lenders and borrow money.
August 2008
When banks will not lend money, vendors will. Wal-Mart is one of Brazil's fastestgrowing vendors. Credit cards are plentiful, and though the rates are high, they remain popular. (Plasma TVs and cell phones are status symbols.) Most stores of larger consumer goods cars, furniture, televisions, computers will sell them on credit, with long repayment period (several years). Naturally, the implied interest rates are very high (topping over 100% annually), but lacking better alternatives consumers choose them. (For a recent example, see Exhibit 1, a story of Banco Popular's home furnishings consumer loan program.) If the vendors are uncomfortable with the collection risk, they can sell the loans at a (high) discount to a factoring agent, who undertakes collection. Factoring is a large business in So Paulo. With the decline of hyperinflation the major banks have discovered vastly increased buying power in the middle and lower-middle classes (what Brazilians classify as Class C and Class D, equivalent to 2 to 10 minimum wages). To tap this market without entangling themselves in regulations, many of the banks have set up leasing companies, and also financiera subsidiaries or brand affiliates (akin to 'housing finance companies' of a bygone era in the US). They are omnipresent: in downtown So Paulo it is almost impossible to walk half a kilometer without passing a financiera storefront. To compete with this, and to undercut the usurious interest rates, Caixa has a lowpaperwork home improvement loan program that does not require the borrower to prove income, or to have title to property. It simply finances up to R$ 8,000 with repayment over eight years (96 months). Loan turnaround time is a few days. The program has been wildly popular: in 2005 it financed 400,000 families, and in 2006 roughly 170,000 families. 5. Microfinance: faint stirrings? Microfinance exists informally in favelas, as community members lend to start bakeries, hairdressers, and other neighborhood services. Banco Real, a subsidiary of ABN Amro, is experimenting with microfinance. The expectation is that as the C/D market becomes better served, the banks will want to move further down-market but in the meantime, with vast profits to be made in the higher-end and more formal sectors, those are gaining the lion's share of attention. Further discussion of housing microfinance as it may merge in Brazil, or may already be present by another name, is provided in Section 5G below. d. Cingapura: in situ demolition/ rebuilding new construction walkup flats Begun under Mayor Maluf (1992-96), Cingapura is the colloquial name for a program of demolition-and-rebuilding on site, without displacing current residents. The development concept is simple: rationalize a favela by creating new-construction publicly funded housing, five or six-story walkup flats, which are then sold to the residents in a condominium or cooperative format.
Best Practices in Slum Improvement: The Case of Sao Paulo, Brazil (At right, a typical Cingapura property, Vila Nilo.) Class E customers and what they can afford. Like slum dwellers the world over, most of those in Class E are illiterate, largely unskilled, and with a high percentage (estimates ranged from 40% to 60%) of women as heads of household, which includes four children.
August 2008
All-in, a typical Cingapura property cost 5 somewhere between R$ 40,000 and R$ 50,000 apiece. Assuming for convenience 100% financing, at a market financing rate roughly 15% annually, 25-year term, which is more favorable than most of these customers could obtain a typical loan would require payments of between R$ 500 and R$ 650 monthly, which is affordable (at 30% of income for housing cost excluding utilities) for families with monthly incomes between R$ 1,280 and R$ 1,600, or about 7-9 minimum wages. In short, Cingapura flats would be affordable, absent subsidy, only by upper Class C households, not the Class E households occupying the favelas that the Cingapura properties were to displace. As we saw above, Class E customers can afford only minimal amounts for housing. Cingapura would therefore be premised on a nominal occupancy payment, initially set at R$ 57 per month, which undoubtedly seemed affordable based on the income profiles reported in the early Cingapura projects. With a market monthly capital recovery cost between R$ 500 and R$ 650 monthly, the residents' share would represent only 9% to 11% of the total, leaving 89% to 91% to be repaid from the public sector. Choosing sites: eminent domain for land assembly. Cingapura was in part a civic beautification scheme. Sites were chosen, with relative disregard for property rights, based on visibility, including: Alongside major avenues, highways and arterials. So Paulo currently has no completed beltway ringing the city, so all traffic flows into and out of the urban core. Riverbanks. So Paulo's two main rivers, the Tiete and Pinheiros, were rechanneled (as part of a massive effort to reduce their pollution. Land on which
These are considered high costs, in part because the typical Cingapura property, built to a height of 20 meters/ 60 feet and on reclaimed or geologically soft land, had to drill footing anywhere from 8 to 15 meters (25 to 50 feet) below the surface to stabilize the buildings. Some sources also suggested that because the program was municipally financed subject to government procurements, costs were inflated along the way.
August 2008
favelas sat became river channels; river bends became newly dry land for relocated favelas. Land was either deeded over from the applicable government, or taken by a compensated expropriation proceeding 6 . The development sequence. Cingapura's basic operational sequence was (and is) as follows: 1. 2. 3. 4. Neighborhood identification. Enumeration of residents and households. Determination of eligibility. Beginning of resident education in high-rise living. Demolition of existing favelas and residents' relocation to nearby temporary housing. 5. New construction of five-six story walkup flats. 6. Occupancy by returning households. 7. Post-occupancy social services. This whole sequence took a very long time. From neighborhood identification through reoccupancy often lasted four or five years; meanwhile, communities that, however informal, were socially functional, found themselves uprooted, leading to a loss of social cohesion. Those who moved back to the new site were often quite different from those that moved away. How it was financed: IDB loan. With the restoration of democracy in the late 1980's, Brazil's government entered a period characterized by two trends: Strong municipal government, weaker central government. Partisan politics with fragile coalitions.
As it relates to slum upgrading, these conditions meant that the national government had little or no activity in affordable housing. (The national Housing Bank had been shut down in 1986, some said a scapegoat for the dictatorship's previous policies.) Further, the costs of slum upgrading are very high, beyond the reach of a normal municipal budget. Municipalities thus faced the challenge of borrowing long term without reference to a domestic housing agency, and of obligating future civic administrations to repayment of the future debt service. The solution a municipal finance work-around was to involve the Inter-American Development bank (IDB, or BID in Portuguese) as the long-term lender, with payments obligated out of a call on the general municipal budget. In effect, slum upgrading was treated financially as a component in municipal infrastructure no different from riverbank improvement, new power plants, or highways and transportation.
6
Land assembly by compensated expropriation or eminent domain was effective in building terms the land was acquired, and built upon but less effective in equitable government. The municipality often took land first and paid later. Price disputes were common and protracted.
August 2008
As a small serendipitous byproduct, involvement with the IDB represented a kind of training wheels or practice run for future municipal finance by the municipality of So Paulo. While this has yet to happen, it is likely in the next half-decade, and the slum upgrading experience will be relevant in demonstrating the municipality's payment history. The residents' post-occupancy status. Residents were initially promised that the flats they occupied would have a proper registered title, but this process was not made a preoccupancy priority; the focus, understandably if short-sightedly, was on completing and occupying the units. Residents thus received a document granting them permission to occupy their flat, and anticipated that title would be forthcoming. The residents' post-occupancy economics. When Cingapura was established, the initial payment, entitled a 'permission fee,' was R$ 57 monthly, leading to these basic resident economics: Permission fee Condo fee (maintenance) Utilities (typical family) Total monthly cost to occupy Implied minimum income at 30% Income measured in minimum wages R$ 57 R$ 50 R$ 45 R$ 157 R$ 510 3.0 x
Thus the program was designed to be 'affordable' to households at 3x minimum wages and above, leaving those Class E households whose income was lower to find a way of raising their income or devoting a larger percentage of it to housing and occupancy costs. What happened afterwards. Cingapura did not merely transform neighborhoods; it powerfully changed people's lives. Many of the changes were for the better; some were not.
Some households had difficulty adjusting to high-rise life. Some residents, unable or unwilling to pay for utilities, resorted to rural methods such as starting cooking fires on their apartment floors with predictable consequences in apartment damage. Other residents who had operated home-based businesses (auto repair, food vending) could not do so from flats in the air. Hence the emphasis, in later phases, on walkup and low-rise structures. Social programs are necessary, and underfunded. While attendance may be irregular, the introduction of municipality-funded social programs into Cingapura properties -- job training, adult literacy, family counseling creates a safety net that these residents have never had before. At the same time, the municipality budgeted for too few staff. "One to two girls trying to keep this whole place
August 2008
clean? They should have had more post-construction social management work. Otherwise, within a couple of years, the houses will look bad again."
Residents increase their consumption and get into debt. Many residents, upon moving into their new apartments, set out to furnish them in a manner suitable to their new surroundings. Buying furniture and tapping the plentiful if usurious consumer credit (see 4C4 above), they dug themselves swiftly into a debt hole. As one interviewee put it, "It's something I find curious: every time you have urbanization, a mountain of rubbish leaves the favela. Old furniture, refrigerators that are huge and extremely old, they throw all this away to buy new appliances that fit within the new unit." Newly created green space was invaded, then fenced. Originally the Cingapura high-rise properties had open play areas and small flower gardens. These were swiftly invaded by new households who built makeshift housing. Only by putting up strong metal fencing all around could the municipality keep the open space open and green. The community changes, first losing cohesion and only slowly regaining it. Many of the original pre-development families left their neighborhoods. Others, finding themselves with a much larger apartment than they have ever had before, and possibly facing higher occupancy costs than they expected, opted to sell their occupancy rights. Gradually the income profile with the Cingapura project rose from Class E (1-3x minimum wage) to Class D/C (4-6x). The 'permission fee' never changed. Although Brazil has experienced significant inflation since Cingapura was completed, the original BRL 57 monthly permission fee remains unchanged. A market price for 'permission fee' occupancy emerged. People now buy and sell occupancy rights in Cingapura properties. Thus, although they lack title which acts as a material drag on financeability, and hence on property values there is at least a functioning economic market, which allows some household mobility and encourages home maintenance and home improvement.
Cingapura properties today. Cingapura properties today are dotted with For Sale signs (Vende-se esta casa). Prices cluster around BRL10,000 to BRL 15,000 apiece corresponding roughly with the levels affordable to Class D/C households, not Class E. Even as it signals a socioeconomic shift, the emergence of a trading market is a hugely positive sign --- because it shows that post-Cingapura properties are now part of the legitimate urban fabric of So Paulo. Consider:
Home resales create a formal and visible economic environment where none existed before. People buying and selling homes at market prices means People who know they can sell their homes for cash are more likely to improve them.
August 2008
The emergence of observable prices makes the homes collateral for lenders, and should decrease the cost of borrowing.
Further observations. Before Cingapura, common wisdom in So Paulo held that favelas were unfixable. Cingapura shows that it's possible to force opponents of slum upgrading to change their reason from "it's impossible" to "we can't afford it." That is a significant shift, since a municipality can afford what it believes important enough to fix. At the same time, the program has had its share of challenges.
Dropping high-rises onto sites independent to the topology of the area was expensive and disruptive. Neighborhood-level communities could not survive the dramatic and extended uprooting and then replanting. Politics intruded heavily in the choice of sites to be redeveloped. Choosing low-lying travertine land increased costs and created ongoing geological risks.
Even though Class E's were the target customer, many could not afford their new apartments, and resold to higher-income newcomers. Class D's thus displaced Class E's. As one of our interviewees put it, "the lower the family's income, and the better the apartment they live in, the faster they're going to leave, either evicted or through resale. They're more likely not to make the payments and to get into trouble." e. The Guarapiranga approach: water and sanitation infrastructure retrofitted In terms of social premises, the Guarapiranga program is the flip side of Cingapura. Where Cingapura preserved the streetscape but demolished and rebuilt housing, in Guarapiranga as few houses as possible were demolished and rebuilt, and it was the streetscape that was wholly transformed. Settlement of Guarapiranga: genesis of the slum upgrading program. Guarapiranga is a large artificial lake formed by damming the river, in the extreme southern part of So Paulo, to create a vast reservoir that today supplies roughly one-quarter of So Paulo's drinking water. Originally remote, the area was settled informally by residents relocated from better-located favelas in central So Paulo. Initially the small number of settlers could be safely ignored by municipal government, but as So Paulo's population swelled, so too did the Guarapiranga favelas, and with them, the effluent flowing into the water supply. "If nothing had been done," said one source, "we would lose the reservoir, and have to bring water from very far away. So we had a self-interested economic reason to clean the reservoir." Further, cleaning the reservoir had to involve the residents, who by now were so numerous they physically blocked any potential expansion or infrastructure improvement, who had anti-eviction legal claims, and who represented a significant political force.
August 2008
Cleaning up the water by cleaning up the slums. In self-defense not to improve the favelas, but rather simply to clean up the water supply, the water company, SABESP, in concert with the city, undertook a massive upgrading. For a total mid-1990s cost exceeding $350 million, the water company (SABESP), with World Bank financing, comprehensively redid all of the open infrastructure, including:
390 kilometers of new water collector, back-bone collector and outfall networks 264 kilometers of sewer network to serve 80% of the 580,000 inhabitants 26,700 new sewer domicile connections serving 125,000 residents 8,050 domicile connections in existing networks 13 square kilometers' worth of new drainage, paving, and restoration 20 sewage pumping stations with automated, centralized effluent removal systems
The slum upgrading was purely incidental it had to be done to stabilize the hillsides, because it was the people whose effluent was polluting the water, and it was their houses that were risking mudslides. The works package: urbanizing the neighborhood. From the neighborhood's perspective, the upgrading package was comprehensive, including: Pavement Collection sewerage. Drainage Retaining wall. Channeling, storm drains. Runoff, families are at risk. Landscaping. Removal, de-densification. Need to be removed. Storm drains or impinging regularized streets.
The city also identified or name alleyways and streets, with numbered addresses, and thus linked into the city's formal registration system. The customer's contribution; Cingapura redux. Like Cingapura, the Guarapiranga underwriting was predicated on residents paying only the minimal amount they could afford, with the balance something greater than 90% of the total project cost being funded out of the municipal budget (directly or indirectly, being routed through the water rates paid by residents of So Paulo). Because the Guarapiranga properties are townhouses rather than high-rises, each homeowner does his or her own maintenance, with the result that Guarapiranga homes are more affordable in money terms: Cinagpura Permission fee Condo fee (maintenance) Utilities (typical family) R$ 57 R$ 50 R$ 45 Guarapiranga Discussion R$ 57 R$ 0 R$ 45 Required only in high-rises
Best Practices in Slum Improvement: The Case of Sao Paulo, Brazil Total monthly cost to occupy R$ 152 Implied minimum income at 30% R$ 510 Income measured in minimum 3.0 x wages R$ 102 R$ 340 2.0 x
August 2008
Guarapiranga today. Despite isolation from the central city, the Guarapiranga neighborhoods are cheerful places. As in Cingapura, for-sale signs are plentiful, and the reported prices are in the same R$ 10,000 to R$ 15,000 range in the Cingapura properties, suggesting that affordability remains the controlling variable. In Guarapiranga, the residents maintain (or improve) their homes themselves, but common areas are swept and trash is picked up by a groundskeeper, who lives in the neighborhood and is paid one minimum wage monthly for his services. As in Cingapura, there is an ongoing effort to provide social management skills, and the on-site manager seeks to instill neighborhood pride: "I don't accept that people will throw garbage into the drains. The community worked hard to end the garbage, now some people toss things into the trash. Why throw it in the drain? We fought to see this urbanization; the city left it beautiful. If we have trash, put it in the bin, put it in a plastic bag." Most observers also believe that moving in to an urbanized neighborhood, or living in one that has been urbanized, improves people's sense of self. "People dress better after urbanization. They feel like real people." Further observations. While socially and physically Guarapiranga and Cingapura are at opposite ends of the spectrum, in financial terms they are basically identical:
Targeted at Class E households. Resident contribution has wound up at the affordable level (roughly 30% of household income). Balance is paid through municipal funds. No domestic lending agency available; municipality is the prime mover. Borrowing uses an outside agency (the IDB, the World Bank) to provide a workaround to the economic schizophrenia of successive municipal administrations.
f. Regularization, resolo, cortios, and more Development and finance of formal affordable housing within So Paulo has been hindered by a tangled structure of land irregularity and enforcement failures. Additionally, So Paulo is replete with initiatives: one local expert counts a total of forty directed at low-income Class C and below. While the resolution of land ownership and land use issues is well beyond either the capacity of any private entity or the time horizons against which microfinanciers and banks make decisions, the factors are relevant because they influence customers' ability to acquire, improve, finance, and sell their informal homes. 1. Land regularization.
Best Practices in Slum Improvement: The Case of Sao Paulo, Brazil So Paulo's favela-type housing can be illegal for any of three reasons:
August 2008
Located on public land, and hence illegally developed. Most of the Guarapiranga favelas are in this situation, as are some of the Cingapura properties. In both cases, the municipality has gravitated to a tacit acceptance in the form of a 'use permission'. Located on invaded private land. Private land, such as that intended for high-end development or adjacent to large infrastructure facilities (e.g. power plants), can be invaded and occupied. On improperly subdivided land. Some private land was illegally subdivided, by the owner or his agent, without government permission and without a properly approved cadastre of identified plots. Families bought in good faith, thinking they properly owned the land on which their home sits, only to discover they did not.
In all three cases, the physical configuration differs from the legal records, with the result that sooner or later, the homeowner's economics are overturned. Several initiatives seek to regularize land ownership. Medida Provisria. An emergency law enacted in 2001, this enables people to regularize their land, with special concessions for housing occupants. People who bought houses before 30 Jun 01 have to prove home ownership for five years and can secure a retroactive regularization of their land status. 7 Adverse possession land transfer. The city has a law whereby, if a house holder has occupied a privately owned property for five consecutive years, with no dispute to that occupancy being raised by the land owner, then the land title is vacated and transferred (at no cost) to the homeowner. This is a potential path to regularize illegal subdivisions and illegal occupations of private land. Sluggish regularization. Regularization has never been given high judicial or administrative priority. Each incoming administration seeks to make its mark with new programs and new homes. As a result, previous generations of housing, such as Cingapura properties, languish in an endless process of uncompleted regularization, even despite efforts to regularize under Mayor Suplicy, when 45,000 titles were regularized. 2. Resolo and the ZEIS regions. So Paulo has established 1,075 Special Social Interest Zones (ZEIS) throughout the citymost of them in favela neighborhoods. For each ZEIS, the goals are (a) legalizing and regularizing land tenure, (b) enforcing development or vacant or under-used land, and (c) including a portion of affordable housing within each development.
A different statute requires that for each 100 square meters whose title is regularized, the homeowner has to donate to the municipality 200 square meters of land somewhere else. The only time these economics work is when the regularized land is hugely valuable and the newly donated land cheap in short, circumstances that lead to removal of favelas, not their regularization,
August 2008
Theoretically, designation as a ZEIS leads to a cooperative arrangement where empowered residents work with an enlightened municipal government to develop an overall redevelopment and regularization plan. In practice, the process takes 18 month just to approve the enumeration and neighborhood profile, and the resulting plans are so enormous (e.g. over 2,000 pages, including exhibits) that they are incomprehensible to most residents. Additionally, the city has the resolution department (Resolo), whose purpose is to bring irregular or illegally subdivided lots into compliance. As with the ZEIS regions, the process is much better on paper than in reality. 3. Cortios Corticos one-room accommodations of the group-living type that crop up on lowerincome rapidly urbanizing cities, 8 typically in dilapidated buildings in the city center. Each resident occupies a single room some are double-bedded, with a day-shift worker sharing the space with a night-shift worker. Bathrooms are communal and down the hall. There are no kitchens (residents often have a hot plate or similar simple heater). A cortio is a place to store personal possessions (not valuable ones, as they are too readily stolen), to spend a few hours in the evening, and to sleep. Residents are individuals, usually men, usually workers. They are in town to make money and to remit it somewhere, usually backed to their geographically separated family in the countryside. More recently, with So Paulo's traffic congestion reaching epidemic levels, some workers who have houses in the favelas also use a cortio to cut four hours a day off their weekday commute, returning home for the weekends. Catering to a worker population, for a cortio, location is everything. They thus tend to be much older buildings, owned by very distant absentee owners, and maintained by a superintendent or building chief who acts economically like a net lessor: he collects the rent in cash and remits a stipulated monthly amount to the absentee landlord. With such an arrangement, the superintendent has no incentive to maintain or improve the property, and they became festering hovels. (This is a consistent economic trajectory; in the US, rooming houses were a respectable form of urban single residence in the 1880's, but by the 1930's they had become 'flophouses,' synonymous with alcoholism, homelessness, crime, and squalor. As So Paulo has urbanized, cortios have expanded. A total of roughly 600 cortios house 38,000 individuals and families. In 2006, after twenty years of debate, the municipality finally enacted a minimum property standard for cortios, with real enforcement teeth. Owners can be ordered to improve their property, under threat of fines, whereupon one of three results occurs:
In the US they were called rooming houses; in countries like South Africa they are sometimes called hostels or dormitories. The modern US version is called Single Room Occupancy (SRO), and often used to house a disabled population that requires additional living or health services.
August 2008
Owner does renovations, comes into compliance. Owner removes the property entirely from rental stock, converting it into another use. Owner closes down, and the property is condemned (or confiscated in lieu of unpaid fines).
Fines and the threat of property confiscation have the effect of forcing into the open absentee landlords that had heretofore been able to shield themselves. Remarkably, nearly all the owners have chosen to do the renovations and comply. This is a testament mainly to the enormous rental values that can be commanded for even microscopic sleeping accommodations in downtown So Paulo. Within 1 years of enactment, 200 cortios have been renovated to code. At the current rate, within three more years, all the cortios will be brought into compliance. 4. Slum inventories and enumerations. Before undertaking a slum upgrading under any of the current programs the municipality surveys the slum area. So Paulo has invested heavily in electronic property and land records, with satellite photograph, GPS location, and computerized property records. The harder element is human. What are the boundaries of individual dwellings, and who lives in each one? Surveys are done by professionals, but they are outsiders to the neighborhoods. Not knowing the families or the neighborhood, they do not necessarily go into every street or house. Further, they are susceptible to manipulation by the residents. When a municipally-authorized surveyor comes to the neighborhood, the family expects eventually to receive a house. Perhaps, thinks the resident, if I list my extended family, they will give me not just one house, but also one for my aunt in the north and my cousin who's living in the next slum. The result is an inflated number of beneficiaries. 5. Other programs. So Paulo has additional important programs, omitted here in the interests of economy in the case history. The best general summary is in Ensuring the right to the city: pro-poor housing, urban development, and tenure legalization in So Paulo, Jessica Budds, Paulo Teixeira and SEHAB, listed in the Sources section. 6. Implications. So Paulo has too many programs and not enough consistency. This appears to be an outcome of (a) over-emphasis on program origination at the municipal level rather than being supplemented with state or national programs, and (b) an absence of long-term finance or efficient wholesale-type capital raising ventures (e.g. a housing finance agency or municipal housing borrowing entity) that emphasizes quick visible change rather than more effective neighborhood redevelopment. New municipal governments, upon arrival in house, want to make a visible impact. This makes tempting the prospect of suspending of existing programs, even the ones that are working well, and inaugurate new programs, all focused on counting new units rather
August 2008
than improving existing ones (e.g. by regularization). The new programs often cannibalize demand for older programs and lead to a crazy quilt of resources that is hard for practitioners to navigate.
IV. Results, Impacts and Opportunities for Replication Given the extraordinary growth in So Paulo, and the equally extraordinary inventiveness and effort made by successive administrations over the decades, what is the outcome? What can those of us elsewhere in the world take away from the So Paulo experience, and how can we apply it? Further, what is the potential in So Paulo for microfinance, housing microfinance, and slum upgrading? Where are the niches where private for-profit and non-profit mission entrepreneurial entities can prosper even as they provide valuable services and homes to residents? a. Workarounds and their consequences
1. Workarounds Markets emerge because economic imperatives trump either judicial obstacles or administrative indifference. Work-around Workaround. A system that serves a particular business niche or solves a particular market problem in a non-standard way to replicate the benefits of a conspicuously missing element, or to avoid a counterproductive legal/ administrative obstacle. Workarounds represent the triumph of ingenuity and market economics over government or administrative blockages. The knee is a workaround to make primates bipedal. Classical microfinance is a workaround to enable financing in utterly informal communities 2. Favela expansion outstrips formal development and urbanization. Even allowing for a relative slowdown in the growth of So Paulo compared with other Brazilian cities, it's evident that housing demand will continue vastly to outstrip formal supply. Further, millions of existing homes sit on land that they have no legal and formal title to occupy. Formal regularization will continue to proceed slowly, but there appears to be slowly emerging informal acknowledgment, as evidenced by 'use permissions' and resale prices of existing homes. 3. Financing workarounds: microfinance but only for consumer goods?
August 2008
In Brazil, as we have seen above, nothing in the formal financial sector was conducive to the development of housing microfinance. Hyperinflation made money assets unreliable and encouraged banks to earn their yield without risk simply by buying inflation-indexed government securities. Into this void stepped the retailers and consumer finance companies, who created financial forms and innovated in partnership with financeiras and banks (see Exhibit 3). The resulting legacy is a crazy quilt of interest rates (see Exhibit 2, repeated below for reference):
The table makes clear that there is a huge market for affordable microfinance. Personal credit costs upwards of 50% annually, even as consumer deposits pay no more than 12%, more typically 6%. b. The cognitive dissonance of informal settlements: illegal but un-evictable
The slums are so vast, their population so pervasive throughout So Paulo, that even as their existence is formally rejected, it is informally tolerated. As one source put it, "these families are sure that the public sector is not going to repossess the housing. Cars can get
August 2008
repossessed much more easily than houses. As a result, the last thing they pay is the rent." That tolerated non-compliance is one of two pressures reinforcing the status quo. The other is the obstacle course necessary to formalize property in the judicial system, its significant drain on judicial resources, and its low priority in municipal development. The favelas thus become, as slums often do, a twilight zone in which people can live quite happily so long as they make few demands on the formal system. If the violence is contained with the favelas, then to formal So Paulo it is acceptable, and the favelas can even become convenient disappearance zones. "People get killed; if I'm a corrupt policemen being bothered by you, if I take you to the slum, I can make you disappear permanently from my life. Slums are the dumping ground of violence and slum dwellers are the victims of violence." While slums are far from the only violent places high crime is a problem in So Paulo it is easy to tag the 'crime-zone' label on slums. c. Cingapura: a kinder, gentler, inadvertent eviction? Cingapura represents a cruel irony: housing that is 'too good' for the Class E households who are its intended beneficiaries. These Class E residents have trouble living in Cingapura, for several reasons: More costly than their previous residence. As we saw in Section 4E above, because they are high-rises not flats, Cingapura properties are more expensive to live in than Guarapiranga's self-built walkups. Larger than families 'need'. Aside from the increase in quality, the housing is objectively much more space than the families have heretofore had. They adjust their consumption to match. Beyond the level of 'effective demand.' In economics, effective demand means what I will actually pay, on a sustainable basis, for a commodity. It's often a consumption level lower than my objective need (set by planners) and much lower than my anticipated demand (based on my hopeful imagining). People who move in to Cingapura properties find they are difficult to afford. Poorly suited for self-employed informal workers. Lacking any ground-level access or any yard, Cingapura's high-rises are poorly suited to individual informally employed workers, be they ragpickers, laborers, vegetable vendors, or specialists such as electricians. Yet the Cingapura apartments are good housing for people whose income is a little higher: Class D and the lower end of Class C. Thus when a Class E household finds it difficult to live in a Cingapura high-rise, it has an alternative sell the use permission for a market price (R$ 15,000) and move elsewhere. Over time, the neighborhood changes from a Class E community to Class D/C, as some Class E households are voluntarily, slowly, economically displaced so even as the
August 2008
neighborhood gentrifies, it does so to benefit those other than the people for whom the Cingapura property was built. Is this a bad thing? That depends on whether one sees the goal as improvement of a particular place, or the housing of a particular group of people. d. Guarapiranga: the Hippocratic-oath ('do no harm') urbanization
The Guarapiranga exercise is slum upgrading as serendipitous byproduct of municipal improvement. Thus it does as little as possible in improving individual houses, although it's unclear whether this is a conscious social repudiation of the Cingapura experience, or simply an economic practicality driven by the need to keep water-quality-improvement costs in line. Probably it is both in a happy marriage of convenience. The result works well, at least in its objectives. Class E residents are not economically displaced rather they are woven into the city's fabric in a more formal way. The existing social community is not dislocated or disrupted. Residents are given visible evidence it is worth improving their housing. e. Financing slum upgrading: government pays non-recoverable cost
Upgrading a slum consists of both private and public investment: private investment in improving homes, and public investment in civic infrastructure everything from mailboxes and streetlights to water and sanitation. In both Cingapura and Guarapiranga, although the capital was spent on vastly different uses, the economics are strikingly similar: Slum upgrading: the So Paulo financial model (Applicable to both Cingapura and Guarapiranga)
Capital cost is financed by a massive borrowing Municipality is the borrower, capital is provided by an international lending entity (workaround against municipal policy inconsistency brought about by changes of government). On an ongoing basis, residents pay what they can afford ('effective demand'). Remaining debt service cost is funded by the municipality, directly (general revenues) or indirectly (water rates via SABESP).
Significant for this discussion is the role of the outside entity World Bank or IDB as, in effect, a financier who accepts the political risk of policy shifts brought about by new elections, and who thus provides a 'political credit enhancement' of the infrastructure project. f. Markets always clear; creating economic value from 'dead capital'
August 2008
Both Cingapura and Guarapiranga are economic successes: Existing homes have value and trade freely in the marketplace. Values are rising. For these residents, this is a sea change. Their communities are now recognized and 'on the map' in a literal sense, which creates the probability of a continuum of growth, investment, and appreciation. That is a huge change. g. Whither housing finance, slum upgrading, and microfinance in So Paulo?
The profusion of programs actually works against So Paulo, creating boundaries of inconsistency in treatment of different households, difficulty in combining resources, and lack of knowledge among program participants. What is missing are all the lending programs at any scale: whether small home improvement loans (that is, BRL 500 or so), housing upgrade loans (BRL 5,000), purchase of an informal home (BRL 15,000), or neighborhood-level improvements (say BRL 500,000). These gaps are more striking given the strong evidence of demand: consumer finance companies occupying the space have proven that money can be made lending to very poor people.
V.
So Paulo shows all the signs of a market ready for a boom in affordable housing finance and housing microfinance, yet there are no such entities offering themselves into this niche. Since the economics of slum formation and slum improvement are universal, the absence of self-identified microfinanciers is significant, as is the experience of So Paulo in creating its slum upgrading programs to date. a. Slum upgrading represents predominantly non-recoverable costs
As shown in Section 3E, slum dwellers not only cannot afford the full cost of their renovated homes, they can afford only a tiny fraction of that cost. Efforts to make them pay more succumb to the Law of Economic Gravity. As one source said, "People have a cap on what they can pay. They will pay what they can pay. They won't pay more than they can pay." Further, efforts to force consumers to pay more than historically affordable levels backfire, as did So Paulo's attempts to charge slum dwellers for basic water consumption 9 . In view of this, all the excess cost is non-recoverable from the customers, and must be financed via an outside source typically, the direct or indirect taxing power of a municipality.
SABESP meters all the houses, and allows households to use water free up to a set maximum of liters per day consistent with pre-upgrading consumption. Only the overages are charged.
August 2008
The basic financing model of slum upgrading, shown in Section 5E, appears virtually an economic axiom. b. When upgrading slums, spend public resources on public spaces, not private homes
At their simplest essence, Cingapura and Guarapiranga represent contrasting theories of slum upgrading:
Cingapura: fix the homes, let the infrastructure fix itself. Guarapiranga: fix the infrastructure, let the homeowners improve themselves.
The differences in physical and upgrading models can be summarized in the following table: What doesn't work
Benefit the whole area, not just a few. House-based improvements divide a community into winners-and-losers; public-space improvements give everyone a benefit and are much more likely to be maintained afterwards. Giving people formal access to a free or cheap resource.
Laying new pipes without redoing the streets. The resulting tangled nest of piping cannot be maintained costeffectively. Car-centric street layouts. Favelas are transited largely on foot. Upgrading means pathways, stairs and steps. There is no room for sidewalks; instead cars (if any) thread their way through pedestrian-dominated streets. Ignoring the community. Slums are not homogenous; each has its own social cohesion and community. Upgrading oblivious to the community will be resisted and will be destructive. Imposing formal building standards on existing favela homes. See Section 6C.
Pedestrian and dense streets. The urban model much more resembles the medieval or renaissance hilltop town than the postwar arterial or grid.
Bringing in genuine community participation. Remarkably, if you want to know what people want and will value, ask them. If the residents want public lighting, it will not be vandalized after installation.
August 2008
Relocating everybody off-site for an extended construction period. Many of those who move back are different from those who moved away and in the meantime, the social fabric is destroyed.
Finding a modus vivendi with the drug dealers. Drug dealers are a major revenue source in many favelas; if they are rooted out, they have many ways of resisting. They are forced out of communities only by communities themselves, and only after the community has become more formal and less poor, not before.
c.
Upgraded slums are still substandard housing at least by market standards because it costs far too much, in both financial and human terms, to demolish and rebuild anew. (Cingapura demonstrated this the hard way.) Improving existing homes creates a tension what is 'good enough' to preserve, and what is 'bad enough' to require replacement? The challenge is greater because many slums are built on geologically risky sites, with inadequate foundations, the potential for mudslides or subsurface subsidence, and vulnerability to rains or flooding. Formal laws will apply to municipally funded programs unless special exemptions are made. To get the exemptions requires architects with some practical flexibility, and a government committed to achieving results even if that means overriding entrenched bureaucracies. d. Completing the buildings is not enough; legal and social follow up is essential
Time and again, planners and program designers assume that if the slum family is placed safely into a new higher-quality apartment, everything else will follow. It does not, for reasons relating both to families and government:
Families are not used to living in formal or high-rise accommodations. Training and reinforcement are required. Experience at Cingapura indicates that four years after occupancy may be the minimum reasonable interval for social support. Government is an unreliable counterparty 10 . Anything promised to be done after the ribbon-cutting is at risk of never being done, either through a change in administration or simply a lack of priority resources. This is particularly true if the important is intangible like formal title or contentious like compelling landowners to regularize parcels.
In 1995, price of a monthly installment of an apartment at Cingapura R$ 57. Still R$ 57 now, with all the inflation. And the default rate is very high.
10 Another, smaller example: in Cingapura the occupancy fee (to repay some of the debt service) was set at R$ 57 in 1994, and expected to increase annually with inflation. Fifteen years later, it is still R$ 57, or effectively about onethird as much, a slow windfall to the residents that has gone unnoticed.
August 2008
e.
Municipalities are political bodies prone to sudden shifts of both policy (who are we serving?) and programs (how are we going to do it?). Yet slum upgrading, being capital intensive, obligates a municipality to make payments over many future administrations (15-35 years). The World Bank and IDB involvement, which otherwise seems incongruous and unnecessary, can thus be seen as a two-fold workaround:
Against shifts in municipal policy. Since no private body would lend to a municipality, only a public-spirited body will; and both banks have a claim against national governments, that are in turn in a position to enforce compliance out of the municipality. Against the absence of a specialist housing finance entity. The same roles could be played by a domestic state or national housing agency yet Brazil has had done for more than 20 years, since the National Housing Bank was shut down. Slum upgrading is itself an agency that makes political change
f.
Many people decline to tackle slum upgrading because it seems an enormous, insoluble problem. When it is tackled, slum upgrading attracts supporters from all walks of civic life.
Academics have an important public-policy activity to study, locally and in real time. Technical specialists can make a living providing services to slum upgrading programs. Contractors and workers learn specialized disciplines like masonry, electrical upgrading, and so on. Public administrators can gravitate into the housing discipline. Judges become more reluctant to compel eviction or relocation where the government is acknowledging the slum dwellers' political legitimacy. Elected officials gain constituents and an issue they constituents support. Government administrators gain regulatory experience and a programmatic history. Residents gain new employment opportunities (grounds-keeping, social work in the communities), which adds money, skills, and role models into the communities. Residents gain hope, and a sense of being in the community. "People dress better. They feel like real people." How much can that mean for children?
August 2008
In short, slum upgrading is a beacon that attracts those who would see it. It changes the political dynamics. g. Slum upgrading requires specialized lenders, including government lenders
Every developed nation has at least one national housing bank, with Mexico and Thailand having the most successful. Very large nations such as the US have multiple such lenders: the US has fourteen government-sponsored banks (Fannie Mae, Freddie Mac, and the twelve Federal home Loan Banks), plus 52 state housing finance agencies (one per state plus DC and Puerto Rico). Brazil has minimal formal mortgage finance. Remarkably, Class A and B homes are often bought with cash, not financing, a legacy of the period of hyperinflation 11 . Banks are only slowly learning how to underwrite traditional mortgage finance, and it will be some time before they have become efficient enough at the transaction level to go downmarket into Class C and D borrowers. Brazil has no housing bank. Although the Caixa Economic Federal (National Economic Bank) is offering a housing program, R$ 35 billion from the pension fund, its products are aimed at the middle market, and it has no competitors in the banking space. Specialized lending or a conduit for affordability are absent. The recent national reorganization, creating the Ministry of Cities, offers the prospect of creating a national housing capacity at the national level. As the state of So Paulo is Brazil's richest, this would be a good place for the Ministry of Cities to start. h. Personal slum dweller finance is a space that will be occupied by someone
Within slums, income generation is critical so that the residents can gain the money needed to improve their homes. That consumer demand is being satisfied now by retailers and financeiras catering to the consumer retail trade. Small-volume high-rate lending is not absent from Brazil; rather, what is absent are lenders that are mission-oriented or willing o finance any customer expenditure, not simply high-end consumer goods. i. Upgraded communities are ripe for housing microfinance
"Once you upgrade a slum, people invest in their houses." That investment signals both the physical infrastructure and the formal recognition of status.
The interesting analog is South Arica, where consolidation in the financial sector led to the building societies (realestate-based savings institutions akin to US Savings and Loans, S&Ls) were absorbed into the big four banks, and in the process lost all institutional knowledge of how to underwrite mortgage finance as distinct from consumer. The result is a knowledge gap in the 21st century of capability present in South Africa's 20th century.
11
August 2008
Even better, from the perspective of potential microfinancers, is that Brazil's civic investment, in Cingapura and Gurapiranga projects, is creating a large, well-identified, geographically concentrated group of consumers who want to better their surroundings and will have some financial wherewithal to do it. Meanwhile, the consumer finance companies charge exorbitant rates. Formal banks are intrigued by the potential to down-market but lack the ability to originate high-quality small-scale loans at low cost. So Paulo, in short, is ripe for the emergence of specialized housing microfinanciers working in post-Cingapura and post-Guarapiranga neighborhoods.
In the midst of all the challenges and problems, we should not lose sight of the most basic fact: Measured in the simplest possible way, both Cingapura and Guarapiranga are successes:
Social policy. The places are much better than they were before, and better than their neighboring areas. Economic. Existing homes have value and trade freely in the marketplace. Values are rising.
Intellectual. There is rich learning from the experiences. So Paulo has a large number of people who are highly committed to urban improvement, whether they reside in municipal government, at the urban planning department of the University of So Paulo, and in private companies that specialize in slum upgrading social work. Nexus for services. Post-upgraded properties create a place for and a culture of social work available to the residents. While individual households slip back, this magnet for public assistance in a focused and continuing way is itself valuable.
b. Slum upgrading prepares the economic conditions the climate for housing microfinance Housing microfinance emerges when the conditions are right:
August 2008
A metropolitan environment with economic growth and a rising overall GDP per capita. A rising slum dweller class willing to invest in their homes. Civic acknowledgment of favelas and civic legitimization of slum dwellers. Civic investment in slum upgrading and regularization.
Scale is a necessary determinant: the market has to become large enough to support multiple entrants and actors. So is focus housing microfinance is not classical microfinance. Meanwhile, consumer finance is the enemy of housing microcredit, because it consumes the customers' share of wallet without helping them improve their station in life. So Paulo's slum upgrading efforts have created an environment where housing microfinance entities ought to emerge rapidly. At the same time, the prevalence of consumer finance represents a serious risk of distorting the market to make it simultaneously harder for housing microfinance companies to emerge and for householders to have the disposable income to afford to improve their homes.
Best Practices in Slum Improvement: The Case of Sao Paulo, Brazil Exhibit 1: Interest rates in Brazil Courtesy of Henry Cherkezian
August 2008
August 2008
August 2008
Sources So Paulo has a thick literature of sources on slum upgrading, and a large number of very capable and committed people working on it. Our thanks go to all of them. Interviews Interviews were conducted over the telephone (1-10 May) and in person (11-15 May) from the following principal sources, all of whom gave very generously of their time and expertise:
City of So Paulo o Elisabete Franca, Director of Social Housing Development; o Maria Teresa Diniz, Coordinator of Paraispolis Project
Diagonal Urbana: o Katia Mello, Partner in Charge o Rodrigo M. Moraes, Commercial and Foreign Relations Manager; o Maria Ines D'Alessio Reis, Contract Specialist; o Conceio Ondei, Contract Specialist o Elza Maria Braga de Carvalho, Arquitect - Contract Specialist o Eliana Cruz, Technical Coordinator
Henry Cherkezian, Urban planner and consultant. Interao: o Anaclaudia Rossbach, Director and President Inter-American Development Bank: o Jose Brakarz o Hector Salazar Sanchez
Lingo Brazil: o Alex Barros, Interpreter Standard and Poors: o Regina Nunes, Managing Director o Lisa Schineller, Director of Sovereign Ratings
August 2008
Reference materials In the course of this assignment, we read or reviewed numerous materials from the voluminous reporting on So Paulo's experience. While in no way relying on any of the sources, we acknowledge our debt to these researchers and writers who provide such helpful background, and particularly the following sources:
1. 2. 3. 4. 5. 6. 7. 8. 9.
African Roundtable on Upgrading Low-Income Settlements, World Bank Conference, October 3-5, 2000, Johannesburg Basic Costs of Slum Upgrading in Brazil, Alex Abiko, Luiz Reynaldo de Azevedo Cardoso, Ricardo Rinaldelli, and Heitor Cesar Riogi Haga Brazil's Urban Land and Housing Markets: How Well Are They Working? David E. Dowall Ensuring the right to the city: pro-poor housing, urban development, and tenure legalization in So Paulo, Jessica Budds, Paulo Teixeira and SEHAB Favela-Bairro Scaled-up Urban Development in Brazil, IDB case study, Jose Brakarz and Wanda Engel Aduan Gurapiranga: Urban and Environmental Rehabilitation in the City of So Paulo, Elisabete Franca and others. Integrating the Poor: Urban Upgrading and Land Tenure Regularization in the City of So Paulo, Cities Alliance Land Regularization, Ministry of Cities Legalizing Informal Settlements in So Paulo, Ana Lucia Ancona, City of So Paulo
10. Micro-credit in Brazil, Power Point deck, Henry Cherkezian 11. Understanding Slums, the Case of So Paulo, Mariana Fix, Pedro Arantes, and
Giselle Tanaka
12. Urbanization of Favelas, the So Paulo Experience, City of So Paulo.
August 2008
Reviewers In addition to the sources listed in the preceding section, this report benefited from reviews and commentary by the following individuals. All mistakes, however, are the author's. 1. Jessica Budds, Open University 2. Henry Cherkezian, Consultant 3. Billy Cobbett, Cities Alliance 4. Rodrigo M. Moraes, Diagonal Urbana 5. Regina Nunes, Standard and Poors 6. Anaclaudia Rossbach, Interacao 7. Lisa Schineller, Standard and Poors