ArticlePDF Available

Housing Affordability: The Land Use Regulation link to Informal Tenure in Developing Countries

Authors:

Abstract

This paper provides empirical evidence on the causal association between land use regulation and housing affordability in cities from Latin America, where informal residential tenure condition of households is widespread. We collected a nationwide survey of local land use regulation from planning professionals in Argentina's municipalities comprised in the big urban metropolitan areas, and filling the gap of the lack of a source of comparable and systematic knowledge on the topic. A set of land use indicators are then created allowing the analysis of the regulatory environment according to some of the main issues (e.g., existence of land use plans; authorities involved in zoning changes and residential projects approval processes; existence of building restrictions, infrastructure provision, the presence of access to land regulatory elements, and the cost related to project approvals). Then, using data from the National Households Survey and the National Census, we estimate the effect of land regulation on households' formal/informal tenure condition. Between other findings, we document that those municipalities that have incorporated more land planning regulatory measures into their legal and regulatory frameworks also face the cost of larger informal land sectors. We also find negative effects on formality for higher residential approval costs, tighter regulation (in the form of more authorities involved in housing projects approvals), and positive effects on formal tenure housing driven by the existence of inclusionary policies.
Housing Affordability:
The Land Use Regulation link to Informal Tenure
in Developing Countries
Cynthia Goytia
1
and Ricardo Pasquini
2
Torcuato Di Tella University
February, 2016
Buenos Aires, Argentina
1
Corresponding author: Cynthia Goytia. Head of the Master Program on Urban Economics at Universidad
Torcuato Di Tella (Buenos Aires, Argentina) and Chair of the Research Center on Urban Policies and
Housing (CIPUV). cgoytia@utdt.edu
2
Senior researcher at Research Center on Urban Policies and Housing (CIPUV) at Universidad Torcuato Di
Tella (Buenos Aires, Argentina).
2
Abstract
This paper provides empirical evidence on the causal association between land use
regulation and housing affordability in cities from Latin America, where informal
residential tenure condition of households is widespread. We collected a nationwide survey
of local land use regulation from planning professionals in Argentina’s municipalities
comprised in the big urban metropolitan areas, and filling the gap of the lack of a source of
comparable and systematic knowledge on the topic. A set of land use indicators are then
created allowing the analysis of the regulatory environment according to some of the main
issues (e.g., existence of land use plans; authorities involved in zoning changes and
residential projects approval processes; existence of building restrictions, infrastructure
provision, the presence of access to land regulatory elements, and the cost related to project
approvals). Then, using data from the National Households Survey and the National
Census, we estimate the effect of land regulation on households’ formal/informal tenure
condition. Between other findings, we document that those municipalities that have
incorporated more land planning regulatory measures into their legal and regulatory
frameworks also face the cost of larger informal land sectors. We also find negative effects
on formality for higher residential approval costs, tighter regulation (in the form of more
authorities involved in housing projects approvals), and positive effects on formal tenure
housing driven by the existence of inclusionary policies.
Key words: housing affordability, exclusionary land use regulation, land markets, housing
informality, land use regulation index, Latin America.
JEL Codes: R14, R52, O54
Acknowledgments
We gratefully acknowledge the financial support of the Lincoln Institute of Land Policy.
This research project has been made in agreement and close collaboration with the
Municipal Affairs Secretariat at the National Government. We are also especially grateful
to all participants at Lincoln Institute of Land Policy International Research Seminar (Costa
Rica) and the International Congress on Land Markets and Informality in Argentina. The
remaining errors are the sole responsibility of the authors.
3
Table of Contents
I. Introduction 4
II. Brief Literature Review 6
III. Methodology 13
i. The Regulation Survey to Municipalities 13
ii. Main Regulatory Issues and the Definition of Land Regulation Indicators 14
iii. Estimating a Tenure Choice Model using a Households Survey Database (2007) 19
Database and Explanatory Variables 19
Tenure Condition and Informality Definition 19
Indicators at the Urban Agglomerate Level 20
Econometric Approach 21
iv. Estimating the Percentage of Tenure Informality using Municipal-level Census Data
(2001) 22
IV. Results 23
i. Regulation Survey Preliminary Results 23
ii. Land Regulation Indicators 27
iii. Households Tenure Condition in Argentina 29
iv. Tenure Choice Model Econometric Results 30
v. Percentage of Population with Formal Tenure Model: Econometric Results 33
V. Conclusions 34
References 36
Appendix A: Regulation Indicators Tables 40
Appendix B: Tenure Choice Tables 43
4
Introduction
Given the fact that land regulation is a complex issue with little data available, this study
fills a gap in urban policy in Argentina by contributing with applied research that lays the
foundation for policy intervention. This research project has the specific aim of assisting in
the understanding of land policy and of the specific issues that may be affecting land tenure
informality.
The first thing to be remarked about land use policy in Argentina is the great variability
among provinces and municipalities, where there is no legal framework guiding urban
development and land use emanating from the National State. The existing legislation
forms a disperse set of rules laws, decrees and ordinances stemming from provincial
and municipal governments.
In general, land regulation in Argentina comprises land-planning laws at a provincial level
and rules and ordinances at a municipal level. For example, one of the issues in the land
regulatory structure is whether municipalities are empowered with independence to
establish their local regulations about land. Not all provinces incorporate laws in this
direction. The pattern for many provinces is the existence of a group of laws, typically
outdated, that tackle only some of the major land issues.
Land use regulations have been studied by the literature of developed countries as a major
factor accounting for why housing seems to be inelastic in many cities. The literature also
studied the potential effects of regulation, both on house prices and on the amount of
building activity (Gyourko et al., 2007). Land use regulation can affect building in a
number of ways because not only does it set up minimum consumption levels (i.e., the
amount of housing that can be built and other quantitative regulations, setting minimum lot
sizes, heights as well as the allocation of open space) but it also affects costs indirectly
through expensive or sometimes long lasting permitting procedures which raise the final
cost of housing units in the locality. However, where the informal sector is relevant for
providing access to land for low income households, the scope of its development might be
indirectly regulated by not servicing it with certain public infrastructure investments such
as connections to sewerage, water, and road systems (Henderson and Feler, 2008) or by
other fiscal and redistributive instruments that form part of the regulatory framework for
land use.
It is interesting to note that there is not much empirical literature that analyzes the key
effects of land use regulations in developing countries. However, a stream of research has
recently focused on Latin American countries (Lall et al., 2007; Biderman, 2007;
Henderson, 2007; Henderson and Feler, 2008) in order to analyze the way in which
regulations which might increase prices in the formal market could promote more
untitled-informal housing development.
Unfortunately, in Argentina, there is relatively little knowledge of the nature of regulatory
frameworks for land use so it is not possible to account for their potential effects. Not
surprisingly, this means that we do not fully understand the way in which a regulatory
environment might constrain the housing supply or affect market prices. This in turn might
affect the tenure choice of households, thus driving informality. Even though informal
tenure had became prevalent in various agglomerates of Argentina, in general, the
regulation only deals with the formal part of the area while no special regulatory
instruments are devised for dealing with areas of informality, such as de ZEIST (Zones of
Especial Interest) in Brazil.
5
In order to help remedy these major shortcomings of information about land use regulation
in Argentina, we conducted a nationwide survey of local land use regulation covering the
municipalities located in the 28 (major) urban agglomerates, covering almost all the urban
land universe in Argentina
3
. As a first outcome of this survey, a set of indicators are built
up, summarizing the main issues related with the regulatory environment for residential
purposes. An approach similar to the one of Gyourko, Saiz, and Summers (2006), on their
measure of the Wharton Residential Land Use Regulation Index in U.S., is followed here.
While some of their indicators, as for example, when analyzing the actors and pressures
involved in the definition of the regulation, are replicated in this study, several other topics
have been added that might provide a more suitable explanation for the analysis of the
tenure condition in developing countries, where informal markets are significant. As an
example, the process of infrastructure expansion, the presence of redistributive and access
to land elements in the regulation, and fiscal policies, are focused on here.
4
Does our measured regulation have an effect on the actual tenure condition of households
in Argentina? The second stage of the research had the objective of exploring the actual
relationship between the existing land regulation and the actual patterns of tenure condition.
In Argentina, like in most Latin American countries, the tenure condition of households
presents various formal (i.e. owner, renter) and informal (i.e. owner of the house but not of
the land, occupant, renter in an informal settlement) modes. In this case, as a first approach
on the issue, we focused on those factors that determine informal (in opposition to formal)
types of tenure. A particular attention was placed on the definition of an informal tenure
condition, using alternative criteria in order to define it with different conceptual
assumptions. The definitions explored alternatives that took into account a mix of tenure
status, physical setting and lack of basic infrastructure and services.
Two empirical approaches were followed in order to explore the relationship between
regulation and tenure condition. First, using the Permanent Households Survey (Encuesta
Permanente de Hogares, EPH) from the National Institute of Statistics and Census
(INDEC), a households’ tenure choice model was estimated in the cross-section of
households. The database allowed assigning each household a set of indicators of the
(average)
5
existing land regulation in the urban agglomerate in which lives, as well as a
number of characteristics (e.g., life-cycle, socioeconomic), in order to predict its
3
The definition belongs to the National Institute of Statistics and Census (INDEC) of
Argentina. In particular, we follow this classification since the INDEC’s households
survey (Encuesta Permanente de Hogares (EPH)) is the source of the households’ tenure
choice data that we analyze in the second stage of the Project (see the progress on this stage
on the chapter below).
See the Methodological Report for the definition of the agglomerates, as well as the names of the
municipalities comprised by them.
4
We built up indicators and classified them according to their main topics: (i) Land Use
Plan and Regulation Indicator (LPI), (ii) Zoning and Residential Projects Approval Processes
Indicator (ZRPI), (iii) Building Restrictions Indicator (BRI), (iv) Infrastructure Provision
Indicator (IPI), (v) Access to land Regulation Indicator (ALRI), (vi) Municipality Fiscal
Indicator (MFI), and (vii) Projects Approval Costs Indicator (ACI).
5
See more details on the methodological section.
6
formal/informal tenure choice. The analysis allows recognizing the effects of certain
characteristics of the regulation into the tenure condition. In addition, for example, age, sex,
or the immigrant condition also affects the propensity of having a formal or an informal
tenure. In order to check for the robustness of the results on regulation indicators, an
alternative source of data was explored. Using the Argentinean Census of 2001, a model of
the percentage of population with formal/informal tenure across municipalities’
jurisdictions was estimated.
The paper is organized as follows: the following section provides a brief literature review
on the relationship between regulation and access to land. Section III describes the
methodology of the paper, including the characteristics of our survey, the construction of
our regulation thematic indicators, a discussion of the alternative measures of informality
and tenure conditions, and the econometric models to be estimated using each of the
available databases. Section IV discusses the results, including descriptive statistics of the
state of the regulation in Argentina, the patterns of tenure and the econometric results.
Brief Literature Review
Land Use regulation
In this section we refer to the context of the literature on land use regulation, summarizing
and drawing some conclusions from existing research while interpreting the evidence
currently available. As we have already stressed in the introduction, there is relatively little
published research on land market regulation in developing countries. Therefore, the
analysis of the effects of land use regulation in those countries is, compared to research on
regulation in developed countries, incipient. In this review we want to identify issues that
set the basis for our research about the effects on informality.
In the last five years, the economics literature has renewed its interest on the effects of land
use regulation, a research mostly led by Glaeser, Gyourko and Saks (2006)
6
, who based
their conclusions on earlier research by Malpezzi (1996). Malpezzi showed
The effect that land use regulation has on households from different income groups is not
addressed in the literature from developed countries. While most of the literature that is
based on US data considers a generic consumer, some evidence about the effects of land
use regulation on low-income households comes from Malaysia, thanks to Bertaud and
Malpezzi (2001). They suggest that restrictions are more costly for lower income families
because the land costs of housing are raised due to a “forced consumption”. This means
that, if standards are lowered, the ratio of profitability for the provision of low-income
housing is raised, therefore increasing the incentives for developers to supply housing for
this segment of the market. They calculate that if restrictions on construction and roads
were loosened so that 55 percent of the developable land may be saleable, instead of the
actual 40-45 percent,
7
this would double a developer’s profitability ratio, hence shifting his
interest from middle-income groups to low-income housing supply.
6
“Urban Growth and Housing Supply,” Journal of Economic Geography, 6, 71-89.
7
In European cities, 65 percent of the land under development is saleable.
7
There is evidence of the same problem in China (Wu, 2004; Zhu, 2005) and in India
(Sivam, 2002), where land and housing restrictions might affect in a disproportionate way
the supply for low-income consumers by increasing the costs in the formal markets. In
those places, the main effect is to develop an informal housing market rather than to
increase the housing prices paid by low-income consumers. As a result, land is elastically
supplied at lower costs in the informal sector. Therefore, in most developing countries
(Latin America and the Caribbean, as well as Sub-Saharan Africa and China), rapid
urbanization has been consistently forcing a significant proportion of the urban population
to live in informality. These people are located in different kinds of informal places, such as
squatter settlements due to land invasion or informal commercial subdivisions. The
main characteristics of these settlements are their differing degrees of tenure insecurity as
well as the lack of basic infrastructure services, such as water and sanitation.
The argument about land use regulation as a potential cause of informality is associated
with what happened in Argentina, in the province of Buenos Aires, when the Decree-Law
8912, regulating urban land use and setting minimum lot sizes, was enacted by the State in
1977. The new requirements for a minimum lot size an area equivalent to 300 square
meters, forcing developers to finance a complete infrastructure, were more than low-
income consumers could afford (World Bank, 2005). As a result, the low-income
submarket, that had helped a great part of the low-income population to get access to
housing, was practically eliminated. Now developers had incentives to devote their
attention to the higher income segments because new land use restrictions left poorer
households out of the market (World Bank, 2005; Goytía and Lanfranchi, 2009 in Lall et
al., 2009).
Within the empirical literature that evaluates the effects of land use regulation, two
methodologies are applied. First, direct calculations, in which the costs of regulation are
treated as a residual; that is to say, regulations constitute an increase in prices beyond
construction costs and land valuation. Second, empirical exercises that attempt to estimate
econometrically the effects of land use regulation on housing prices and the elasticity of
supply by specifying models where regulation affects supply elasticity or shifts the supply
curve. What is expected is that more regulated communities should have lower elasticity or
smaller supply responses to increases in housing prices.
For the first group of papers, which use hedonic models to assess the costs of building
height regulation in Manhattan, Glaeser et al. (2005b) provide empirical evidence of large
increases due to scarcity caused by this type of regulation. Another paper, by Glaeser and
Gyourko (2003), found large regulation costs in several cities based on hedonic regressions,
city by city, to estimate the cost of regulation. The cost is estimated as the difference among
the reported house value, the sum of construction (replacement) costs and the shadow
valuation of land.
In the second empirical approach the effect of regulation is identified as altering the price
elasticity of supply (Glaeser et al., 2006). Most of this literature builds up on Malpezzi
(1996) and uses a regulation index the Wharton Land Use Regulation Index that is
based on different measures of land use restrictions, such as zoning applications approved,
8
or the time it takes to process approvals
8
. However, this literature does not explore in depth
the particular effects of regulations or the types of measures that most affect housing
supply. In most studies, indices are used or several regulation measures are selected to
estimate their particular effects.
It is also important to note that attempts to compare regulation measures or to estimate the
most costly ones have been unsuccessful (Henderson, 2009 in Lall et al., 2009). The models
applied assessed the effects of direct land use regulation, such as height restrictions,
building standards or timing to get permits, all of which might affect production costs and
development fixed fees that increase prices without affecting the input costs of housing
production.
In this group of studies, the effects of a number of different regulation measures are
assessed for communities in the state of California by Quigley and Raphael (2005). Such
measures include restrictions on the number of building permits, infrastructure
requirements, open space zoning, density and height restrictions and community
participation in zoning approvals, among others. The elasticity of supply in the two
formulations, both when regulation affects supply elasticity and when considering its
effects on shifting the supply curve, are estimated. When the effect of changes in the price
of new starts according to the construction permits issued is considered, price supply
elasticity is found to be lower in more regulated communities. Therefore, an increase in the
number of regulations established in the community reduces the number of building permits
that are issued.
One of the difficulties of such empirical analysis on the effects of land use regulation is the
problem of endogeneity. Most of the empirical literature from the US considers land use
regulations as “exogenous”, rather than as endogenous policy decisions that are made in
response to local market conditions. This approach is inconsistent with a decentralized
institutional framework where local governments establish their own regulations. The
literature argues that land use regulation is a policy decision made by each community to
try to smooth the effects of rapid population growth. It is said that when a locality faces a
positive demand shock, such as rapid migration growth, it may impose tight regulation in
the form of increased minimum lot size or a greater number of review processes/permits
required for new constructions. In order to support this idea, new research shows that cities
that are more regulated in land use are likely to be the ones that are growing rapidly
(Gyourko, Saiz, and Summers, 2006).
Mayer and Somerville (2000) attempt to overcome this issue, considering regulation as
endogenous by using some community measures, such as presidential voting patterns, as
instruments for regulation. They find particular restrictive effects steaming from the
number of building permits issued, which are limited by the extended amount of time taken
for approvals, cities’ referenda on growth proposals, and user’s development fees. Since
they had quarterly data from cities, they showed that regulation affects supply elasticity in
8
A set of State regulation is also included in his estimations.
9
the long run because, in more regulated environments, developers tend to anticipate any
increase in demand by having a larger stock of approved lots for development.
Another aspect of regulation, introduced by Pollakowski and Wachter (1990) are
externalities, which means that an increase in restrictiveness in one community forces
people to migrate to other communities; this supports the idea of interactions between
different localities. They create an index of density restrictions to show that housing prices
are higher in more regulated communities.
Finally, the emergence of the effects of regulation in the literature for developed countries
is supported by more recent analysis on the exclusionary policies of local jurisdictions, such
as the one about the Tiebout literature reviewed in Epple and Nechyba (2004), as well as
the study on superstar cities in the US, by Gyourko, Mayer and Sinai (2006). In this
economic literature, stratification of the population by income within different communities
is based on the consumption of local public goods (Epple and Nechyba, 2004 and Hesley,
2004). The result is that richer communities will want to exclude lower income entrants
because of fiscal externalities. The rationale that underpins this notion is that lower income
residents are a tax burden for existing residents because they consume local public goods
and services but pay less than the average amount in local taxes. Land use regulation
exclusionary measures can be based on a number of restrictions: restricting the number of
new housing units that can be built, thus limiting total population; setting minimum
consumption levels that exceed what low-income households can afford; or making
housing more expensive through costly permitting procedures, all of which finally raise
construction costs for new housing, preventing new migrants from settling in those places
because they cannot afford them.
Another main effect of increased migration, which exclusionary regulation intends to avoid,
is congestion, which means higher living costs, dissipating the benefits of agglomeration
and diminishing the quality of life.
9
In order to avoid that, the literature suggests that
(some) cities impose tight land use regulation, hindering further residential housing supply.
Gyourko, Mayer and Sinai (2006) show that this argument is consistent with a model of
high-amenity cities that impose regulations to limit entry and skew the population towards
the highest income households, so as to enhance the welfare of the resident populations.
Most of the literature reviewed is based on developed countries, particularly the US, and
focuses on the analysis of the direct effects of land use regulation on housing prices. It
shows that price variations across cities in the US may reflect different degrees of local
regulation (Glaeser et al., 2005; Gyourko et al, 2006). Price increases could just reflect land
scarcity in cities where the amount of developable land may be exhausted. The general
perception is that this scarcity is caused by regulation that does not allow any increase in
density or building height, thus limiting capacity.
Land use regulation and informality
This body of research from developed countries provides some arguments that have to be
revised if they are to be applied to developing countries because countries where this
Some basic characteristic of this superstar cities mentioned in the literature are slower population growth and
an increased share of high-income households over time.
10
literature is based have no informal markets, and institutions are relatively strong
(Henderson, 2007). While in developed countries market regulation restricts housing supply
and limit population growth, in developing countries formal market restrictions lead to the
development of an informal land and housing sector which ignores land-use regulation. A
large informal sector develops, which represents 10 to 45 percent of the land and housing
market in some of the most important Latin American cities (World Bank, 2005).
Two types of informal land development may be distinguished. One of them is the squatter
settlement, villa miseria, and favela or barrio de ranchos, among other names given in
Latin America to settlements originated through invasions or unauthorized occupations
on public or private land.
10
A second mechanism to access land is provided by informal
commercial urbanizations, where private plots in the urban peripheries are developed and
sold on a market basis, disregarding one or more planning/land use regulations. The
illegality in most of these settlements is not conforming to land use regulations or to
servicing requirements for land subdivisions. The irregularity of neighborhoods also has to
do with the lack of provision of public goods, such as paved streets, public lighting, waste
collection, security, among other attributes that usually characterize these types of
settlements.
11
There is not much empirical literature that analyzes the key effects of land use regulations
in developing countries. Although there is such literature for developed countries, a stream
of research has recently been applied to Latin American ones (Lall et al., 2007; Biderman,
2007; Henderson, 2007; Henderson and Feler, 2008) in order to analyze how regulations,
which might increase prices in the formal market, may promote more untitled-informal
housing development.
Henderson (2007) and Henderson and Feler (2008) examine some of its implications in the
context of Brazil, where informal markets are a major source of housing for low-income
migrants. In order to do this, they outline a conceptual framework concerning the political
economy of indirect land use regulation in developing countries. Henderson (2007) argues
that in developing countries exclusion of low-income migrants by direct regulation may not
be possible. The institutional framework ambiguous rules, lack of enforcement and
informal constraints leaves the operation of this informal segment of the market outside
regulatory procedures.
12
For Henderson, this literature on direct regulation is deficient, and
does not cover issues of indirect regulation that arise in many developing countries. This
10
Despite the initial invasion of vacant public or private land for the creation of this type of settlement, land
purchase is the rule for most of the dwellers (Gilbert and Ward, 1981), and it even includes some type of
payment to the organizers, or other agents, who provide coordination and guarantee a certain level of security
(Lanjouw and Levy, 2002, 987). The arrival of new residents who may buy or rent can make the settlements
become densely populated (Gilbert and Ward, 1981, 98).
11
The generalized use of these practices and the lack of other valid alternatives to house the poor has made
many scholars to claim that it may be better to use terms such as ‘informality’ or ‘irregularity’ (Gilbert, 2002)
rather than ‘illegality’ when referring to those settlements because the basic rules that guide ownership are
followed, unlike the case of land invasions.
12
In developing countries, if regulation is excessive people have to operate outside the
formal market, where such regulation is overlooked. This issue, introduced in the academic
literature by Turner (1972), was stressed by de Soto (1989) in his book The other path.
11
means that, while the informal sector avoids direct land use regulations, the local
government may attempt to indirectly regulate the scope of its development by not
servicing it with certain public infrastructure investments, such as connections to sewerage,
water, and road systems, or by threatening the tenure security of residents (Henderson and
Feler, 2008). In this sense, the under provision of infrastructure becomes central to the idea
of urban exclusion.
This group of studies focused on Brazil, from 1970 to 2000. The first study, by Henderson
and Feler (2008), tracks the evolution of service provision within a constant sample of
localities that, in 1970, was conditioned to be 50 percent urbanized, at least. The evidence
provided suggests that within a decade, from 1970 to 1980, some localities rapidly
expanded their service provision while others remained under-serviced, even in 1991 and
2000. This issue might be considered a strategic element for un-servicing the urban poor.
Another interesting study from Brazil (Henderson, 2009 in Lall et al., 2009) argued that the
formal sector housing was made unaffordable for low-income households when the
national law from 1979 required a minimum lot size for any housing construction.
Furthermore, this gave origin to stricter local minimum size requirements imposed by
several localities. A higher demand for housing within urban areas led to informal suburban
development. At the same time, localities denied the provision of basic services to stop
informal development.
13
While two types of informal settlements developed in Brazil
(favelas and loteamentos), it is argued that the latter were mainly caused by the effects of
the above mentioned law.
The effects of land use zoning and density regulations on formal housing supply and slum
formation across Brazilian cities between 1980 and 2000 are examined in Lall et al. (2007).
The performance of cities that have lowered land subdivision standards (minimum lot size),
below the 125 square meters that are normally required, according to the law, is assessed.
In order to do so, they created a model of formal housing supply and slum formation, where
population growth was endogenous and decisions of household migration were influenced
by inter-city variations in land use regulation. It was interesting to note that the elasticity of
formal housing supply in Brazil is very low and can be compared to that of Malaysia,
which has a tightly regulated housing market. Regarding formal housing supply, this limits
adjustments in response to increases in demand, and therefore promotes informality.
Relaxing land use regulation, zoning, and land use planning are found to improve housing
market performance by stimulating a response in the formal sector housing. For example, a
reduction in the minimum lot size is found to increase housing supply and cause higher
population growth. Therefore, pro-poor minimum lot size regulations increase both
migration and the number of formal house residents. Lall et al. (2007) observed that urban
zoning regulations increase the growth of the formal housing market and of the city
population at the same rate, and therefore no net effect on informality is found. However, a
faster growth in population than the formal housing supply response may be one of the
reasons for the increase in slum formation.
13
The main assumption is that during the 80s, exclusionary policies in terms of under provision of servicing
were possible even though most localities had democratically elected majors. Dominating elites could
legitimately deny services to the informal sector while after the policy reforms of the 90s it was not so easy to
implement this kind of strategic behavior.
12
It is important to notice that land use regulation measures which manage densities,
particularly minimum lot size regulations, have important effects in terms of housing
supply and slum formation. Contrary to conventional wisdom, lowering minimum lot size
regulations does not lead to a decrease in slum formation. If city population growth were
exogenous and people did not consider local regulations and residential location decisions
at the time of migrating, then lowering minimum lot sizes would allow cities to
accommodate more residents into formal housing developments and would no doubt
reduce slum formation. However, regulations are a part of household migration and
residential choice decisions, and hence, the exact effect of lowering regulatory standards is
not so obvious. Basically, their model suggests that the net effect of land regulations
depends on the extent to which new formal housing supply absorbs new demand, both from
current informal sectors and from migrants attracted by more flexible regulations.
This means that the cities that lowered minimum lot size regulations not only experienced a
higher growth in the formal housing stock but also in the number of migrants. The resulting
city population growth exceeded that of the formal housing supply, exacerbating the slum
formation problem. Therefore, local measures intended to increase access to land for the
poor such as flexible land subdivisions enhance welfare if housing with different
specifications for different sub-segments of the housing market is also supplied, thus
allowing low-income residents to buy those units which they can afford and which meet
their preferences.
Lall et al. (2007) found that cities that offer improved access to land compared to others
that do not are likely to disproportionately attract (poor) migrants. Informality may grow if
this induced population growth is higher than the adjustment of formal housing supply.
They argued for the importance of clarifying what are the sources of land and housing
supply distortions -that reduce the elasticity of formal housing supply- in future research.
Finally, they turned their attention to policies that aim at reducing barriers for the access to
land, and which have to be accompanied by instruments that relax preexisting distortions in
the land market so as not to exacerbate informal development.
Another empirical study, by Biderman (2008), analyzes the connections among informality,
urban land use and building regulations in Brazil, relating the elasticity of demand in the
informal and formal sectors of the housing market by using the theoretical framework
wherein untitled housing is stimulated by inappropriate regulations that raise the prices in
the formal market.
Four measures of urban regulation are examined, and zoning is the one that has the biggest
impact on informal settlement development. The notion that formal and informal markets
are completely independent is strongly refuted since the findings reinforce the idea of the
existence of two sub-markets with different standards, where land use and building
regulations have an impact on informality.
Following this argument, it is worth noticing that local regulations currently vary greatly
across municipal governments, even in the same region or state. Some local governments
have introduced measures that tend to favor access to land and to urban services for the
most poor. Therefore, an important issue that should be taken into consideration in our
study is the potential effect that those instruments might have in determining formal or
informal tenure choice.
Definition of informality in empirical studies
A necessary step to consider the problem of informality in empirical research is how
informality should be defined in order to be appropriately measured. So, the question here
13
is: how are households living in the informal sector identified in these studies? Three
alternative criteria could be applied to define informal settlers by using data from the
national household’s survey, which will have different conceptual implications (Biderman,
2008).
First, considering the main concept of informality as illegality, housing may be defined as
informal or irregular if it does not comply with legal aspects of regulation, especially
formal tenure of the land. While in this option informality is mainly defined by ownership
rights, the perception of security alters the definition of this indicator. This measure of
informality is under-reported in the national census due to the fact that most households
living in informal settlements, and which have paid for the land they occupied, consider
themselves homeowners, although no formal title has been granted.
Second, the physical condition of the settlement has to be taken into account, according to
the answer provided to the question from the census as to whether households lived in
irregular settlements. The answer allows the identification of those settlements that have
unpaved roads or no street number. However, the concept of irregularity of the settlement
widely differs from that of informality (only 5 percent of households) because many other
factors that characterized this last one were not considered.
Third, the definition based on the lack of public services, particularly water connection to
the public network or lack of full servicing must be considered. This parameter should be
controlled in the case of land tenure/housing when it comes to meeting land use regulations
because considering only infrastructure might bias informality to wealthier households.
Methodology
I. i. The Regulation Survey to Municipalities
Our survey of land regulation in Argentina covered a selection of the main issues related
with the land regulation for residential purposes
14
and was targeted to those municipalities
in the 28 big urban agglomerates in Argentina (a total of 118 municipalities). According to
the 2001 Census these agglomerates cover 67 percent of the total population in Argentina.
The definition of big urban agglomerates is given by the National Institute of Statistics and
Census (INDEC) of Argentina. In particular, in this study we are forced to follow this
classification since the INDEC’s households survey (Encuesta Permanente de Hogares
(EPH)) is the (only) source of the household-level tenure choice data.
We distributed the survey across all municipalities. For that purpose, we established a
collaboration agreement with Secretaría de Asuntos Municipales (Municipal Affairs
Secretariat SAM) at the Ministerio del Interior de la Nación Argentina. The Secretariat
helped us in generating a contacts’ database including the information of the key people in
land regulation in each municipality. We contacted the Planning Director in each
municipality. Where none existed, we contacted a planning officer, specially designated by
the Mayor in each locality to answer the survey.
In order to minimize the non-response, respondents were contacted and followed up by
telephone. In order to collect the answers we also built up a web page, which facilitated the
task for respondents.
14
The survey covers some of the main issues related with the land regulation for
residential purposes. The selection of issues has been made by the researchers and
benefited by the comments of several experts.
14
The final response rate we achieved was 75 percent. One main issue of concern is related
to the sampling procedure and identification of sample selection bias in the response to our
questionnaire.
15
Table A.1 shows that the sample is well represented considering
municipalities classified by region or according to their total population. The response rates
were above 70 percent for all regions (with the exception of the North-West Argentina
(NOA) region where the response was 63 percent) and above 67 percent for all population
quantiles. These results suggest to us that our effort in collecting responses was very
valuable and that we do not expect any significant selection bias.
16
II. Main Regulatory Issues and the Definition of Land Regulation
Indicators
In order to examine land regulation we started by following an approach similar to the
one of Gyourko, Saiz, and Summers (2006) on their measure of the Wharton Residential
Land Use Regulation Index (WRLURI) for the United States. We replicated some of their
indicators, as for example, when analyzing the actors and pressures involved in the
definition of the regulation, or on the projects’ approval processes. However, this
approach, as well as others covered by studies for developed land markets, contains
elements that do not fit the reality of land use in a developing country as Argentina.
17
As
an example, annual building limits, while particular important to understand the land use
development in US markets, lack importance for most municipalities in Argentina. But
more importantly, this approach lacks enough elements for the analysis of the dynamics
of the tenure condition in informal areas. Our survey adds several topics that might
provide an explanation for this central issue. As an example, we focus on the process of
infrastructure expansion, on the presence of redistributive and access to land elements in
the regulation, and on fiscal policies.
We have built up indicators and classified them according to their main topics:
15
As already explained, the survey instrument was sent to the urban planning director of
each one of the municipalities that from part of the urban agglomerates across the country.
The list was obtained by contacting each major helped by the Sub-secretary of Municipal
Affairs of the Presidency of the Nation; subsequently those authorities were contacted by
phone and email, to complete the questionnaire. The decision to answer the survey could
not be random if certain types of municipalities had different response rates to our survey.
16
In order to check for sources of biases in the response rate we did also estimate response
models using the available Census data including characteristics of the population (e.g.,
proportion of population with less than 14 years or more than 65 years, proportion of
immigrants), education and socioeconomic status (e.g. using unmet basic needs indicators)
and available infrastructure (e.g., sewerage infrastructure, water network, natural gas
network, electricity). We did not find any significant correlation between these variables
and the response rate.
17
See more on this on the Literate Review Section.
15
i. Land Use Plan and Regulation Indicator (LPI)
ii. Zoning and Residential Projects Approval Processes Indicator (ZRPI)
iii. Building Restrictions Indicator (BRI)
iv. Infrastructure Provision Indicator (IPI)
v. Access to land Regulation Indicator (ALRI)
vi. Municipality Fiscal Indicator (MFI)
vii. Projects Approval Costs Indicator (ACI)
Land Use Plan and Regulation Indicator (LPI)
The aim here is to capture the extent in which a plan for the use of land exists and
whether it has been formally established in the legal and regulatory framework. At both
the provincial and the municipal level, two indicators (provincial and municipal) reflect the
existence of land use plans and whether these has been promulgated as laws or decrees
(at the provincial level) or as regulations at the municipality level (e.g. ordenanzas, urban
planning codes). These indicators take the value of one in the case a plan for the use of
land exists and it has already been incorporated in the respective legal or regulatory
framework; one-half in the case the plan exists but it hasn’t been promulgated, and zero
otherwise.

 




Zoning and Residential Projects Approval Processes Indicator (ZRPI)
This indicator is aimed to capture the involvement of different governmental authorities,
and the community organizations in the approval of residential projects. The indicator
considers the approval of projects that require zoning changes and those regular projects
that do not require zoning separately.
The Zoning Change Approval Indicator (ZAI) was adapted from Gyourko, Saiz, and
Summers (2006) and reflects the degree of difficulty of a certain project to obtain a zoning
change approval. Our survey asked which authorities are involved in zoning change
approvals. The listed organizations are: i) The executive power at the municipal or
communal level, ii) The Planning Commission, iii) The Zoning Board or Council, iv) The
Local (Municipal) Council, v) Provincial level governmental officials, and vi) The
Environmental Evaluation Committee. The index adds the value of 1 for each organization
involved. Finally, the indicator also adds a value of 1 if residential projects requiring
changes in current zoning must be presented, debated or approved in local assemblies
(public hearings) or meetings with the community, and equals zero otherwise.
16


The Regular Project Approval Indicator (RPAI) is analogous to the previous indicator. It
considers the authorities involved in the approval of projects which do not require changes
in zoning. The authorities considered are: i) Planning Commission, ii) Local Council/ local
officials, iii) Environmental Revision, iv) Design Revision Office (e.g. cadastre office) and
iv) Other authority reported. The index adds one for each authority involved.

The Zoning Change Approval Indicator (ZAI), and the Regular Project Approval Indicator
(RPAI) are combined in a single indicator by averaging the value of both indicators. That is,
we give equal weight to the two dimensions of the indicator when we build the Zoning
and Residential Projects Approval Processes Indicator (ZRPI):

Building Restrictions Indicator (BRI)
The following concepts are related with restrictions in the supply of residential buildings,
and then summarized in an aggregate indicator. These are: i) Lot size restriction; ii)
Maximum Land Use and iii) Maximum Total Building.
First, our survey asked whether there is a minimum residential lot size restriction, and the
size of the requirement in case it exists. The indicator will take a higher value for a larger
minimum lot size, indicating a higher restriction to the access to land. The indicator
considers minimum size lot restrictions in low and high densities areas separately, and
adds both dimensions in the aggregate indicator.
Second, the indicator also incorporates the existence of Maximum Land Use and
Maximum Total Building Restrictions, and the perception reported by specialists of these
as actually being active restrictions for new residential developments in the jurisdiction.
These restrictions are combined in the Building Restrictions Indicator (BRI) as follows:
 


Where dlotsizehigh is a dummy variable that takes the value of one if a minimum lot
restriction is incorporated in the municipality regulation, lotsizehigh is the size of the
minimum lot size restriction in high density areas. dlotsizelow and lotsizelow are the
analogous variables for low densities areas. dmaxlanduse and dmaxtotbuild are dummy
17
variables taking the value of one if a maximum land use restriction or maximum building
restrictions are in place. landuseopinion and totbuildopinion are subjective variables that
range from 1 to 5, and take a higher value reflecting the degree in which the respondent
believes that these are active restrictions for the supply of residential buildings.
Infrastructure Provision (IPI)
In this indicator we consider how basic infrastructure and public services are provided in
sub-urban areas or in areas where these services lack.
We consider two major issues. First, we ask if the municipality has defined an urban
perimeter where it guarantees the provision of basic services to new residential
developments. We define a sub-indicator that, for those municipalities that have defined a
perimeter, adds one for each service that is guaranteed. The “Urban Perimeter Infrastructure
Provision (UPIP)” sub-indicator is defined as:
    

Where upx is a dummy variable that stands for the provision of service x within the urban
perimeter.
The second issue is how infrastructure is financed in those regions that lack complete
access to basic services. We consider here if the municipality and the public services
related firms finance the service extension to these areas. If neither the municipality nor the
respective public service firm provides finance, then the cost is completely born by the
developers or new users. Two sub-indicators (IPMUN and IPPUBSERV) are constructed in
other to capture the role of the municipality and the public services firms respectively:
IPMUN = STD (munfinelectricity + munfinsewerage + munfinwater + munfingas +
munfinpavement + munfinsidewalk + munfin streetlig htingposts)
 
  

Where munfinx is a dummy variable that stands for the municipality financing the extension
of the service x and pubservfirmx the analogous for the respective public service firm.
Finally the three sub-indicators are added in the Infrastructure Provision Indicator (IPI). A
higher value for this indicator is expected to reflect a more active role of the municipality in
the provision of infrastructure.

Access to land Regulation Indicator (ALRI)
This is a measure of the presence of redistributive and access to land related elements in
the regulation of the use of land. The index adds one for each of the following elements
18
incorporated in the regulation: i) Recovery of the added value (appreciation) of land, ii)
Obligatory use of the urban land, iii) Regularization of occupied land (e.g., establishing
that occupied land, after a certain period of time, and if there is no opposition, might be
regularized in favor of the occupant), iv) Building permits reserve for social projects., v)
Obligatory donation of land for social projects, vi) Obligatory donation of land for public
equipment (e.g., schools, green areas), vii) Possibility for the municipality to acquire land
for social purposes, viii) Fiscal Incentives for zones that are desired to be developed.




Municipality Fiscal Indicator (MFI)
This indicator is aimed to reflect the power of the municipality in obtaining local
resources. The following issues are incorporated: i) The total tax collection per capita,
which is aimed to reflect the available economic resources for the municipality, ii) In
relation to the effectiveness in tax collection, we will analyze the effective tax revenue as
a measure of total tax billing. This measure should reflect the efficiency of the
municipality in its taxes collecting function. iii) We incorporate two other measures
related to the building registry for fiscal purposes. First, we analyze a subjective dummy
variable taking the value of one if respondents consider that the building registry or
cadastre (i.e., catastro) has been recently updated. Second, an objective measure
accounts if updating has been made in the last two years. The mentioned aspects are
collected in the Municipality Fiscal Indicator (MFI):


Projects Approval Costs Indicator (ACI)
This indicator is aimed to reflect costs related to residential projects registration
procedures. It considers time and monetary costs.
Approval time (AT) is a measure of the average time the revision of a project takes
between presentation and approval. This is a subjective indicator, since there are low
chances of respondents having a precise estimation of the average delay. We asked
separately the average time for single-unit and multiple-units residential building projects.
The AT variable is then defined as the average time for the two procedures.
19

The survey also asked the monetary value that is charged for a property registration. In
practice, many buyers of land or properties do not have formal land tenure because they
avoid the costs related with this registration. We will incorporate this cost as a relevant cost
in our comparative analysis. A dummy variable will take the value of one in the case the
municipality displays a cost of property above a threshold to be determined in the sample
(e.g., the 66th percentile in the sample).

IV. Estimating a Tenure Choice Model using a Households Survey
Database (2007)
Our first approach to explore the effect of land regulation on formal tenure condition is
estimating a tenure choice model with cross-section data from the Permanent Households
Survey (Encuesta Permanente de Hogares, EPH) of the National Institute of Statistics and
Census (INDEC). In this section we provide more details on the data used in this exercise,
the alternative definitions employed in order to explore the formality condition, how we
incorporated regulation into the estimation, and the details on the econometric approach.
i. Database and Explanatory Variables
Using INDEC’s data we have constructed a database covering 28 urban agglomerates and
more than 69,700 representative households. This represents information on nearly 250,000
people. The database is a cross-section for the first quarter of 2007. The database is very
rich in information regarding socioeconomic and demographic characteristics, as well as
the effects of other strategic policy related-variables, such as the nationality and
immigration condition of the members of the households (see more below)
18
.
ii. Tenure Condition and Informality Definition
As a first approach, we grouped tenure conditions into formal” and “informal” groups.
The groups are translated into a dummy variable that will be the center of the analysis. The
formal group comprises formal owners (of the land and the dwelling they occupy) and
renters. The informal group comprises owners of the dwelling but not of the land, self-
claimed owners because of the payment of property taxes, or other type of occupants
(without approval). This approach can be seen as the standard approach found in the
literature (see for instance Cruz and Morais (2008). We call this measure Formal Tenure
Number 1.
This first definition of formal and informal types of tenure is mostly based on legal
ownership rights and aims to capture the lack of well defined property rights, by defining as
informal households those who are owners of the house rather than the plot. However,
there are reasons to believe that the National Survey data under-reports the measure of
owner informality (see Goytia and Lanfranchi, 2009 in Lall et al, 2009). In particular, the
literature points to settlements which have been originated by informal commercial
subdivisions and where households have already paid for the land they occupy. As a result,
18
In order to be able to compare results, we replicated several of the variables in the study of Cruz and Morais
(2008), for the case of urban agglomerates in Brazil.
20
many households consider themselves to be homeowners, although no formal title has been
granted. In addition, households that fear risk of eviction are more prone to declare
ownership of the plot.
We employ a second criterion to define informality, focused in the physical conditions at
the settlement. In a second variable (formal measure number 2) we exclude from the formal
tenure group (and add correspondingly to the informal group) those households that declare
to be owner or renters, and are located in what the INDEC survey defines as an informal
settlement (i.e.: a slum). The definition of emergency settlements provided by the INDEC
Survey considers non-compliance with building codes and related urban regulation. In
general, these are substandard areas, encompassing a group of 50 dwelling units or more,
occupied land without authorization, privately or publicly owned, laid out in a scattered and
dense manner, lacking essential public infrastructure services, also known regionally as
villas miseria, asentamientos or tomas. These settlements are characterized by the
illegality of tenure due to irregularities of the settlement, or great risk on the formal (or the
partial) tenure condition in case it exists.
However, this alternative concept might still be an incomplete measure of formality. The
reason is that the Housing Surveys might not capture many other places that are not
explicitly considered informal settlements, but with similar physical characteristics that
may affect the tenure condition. We then consider a third alternative definition of
informality (measure 3), which considers informal ownership to include those households
located in areas that lack certain basic public services. In particular, we will consider that a
dwelling is informal if the source of its water is through a manual pump (i.e., not connected
to a water network, or to a suitable substitute of good bacteriological quality), if it has no
connection to water inside the house, or if it has no sewer installation (no connection to
sewer network, no septic tank or cesspool).
iii. Indicators at the Urban Agglomerate Level
So far, our regulation indicators have been defined for the municipal jurisdiction level.
19
Those indicators allow a comparative analysis of the regulation across municipal
jurisdictions throughout the country. However, at this stage we also need to know how
regulation characterizes the use of land for each urban agglomerate level as a whole
(recall that an urban agglomerate might be comprised of more than one municipal
jurisdiction), since the objective now is to examine the patterns of household tenure
condition in relation to the existing regulation in their location. The reason for not
exploring the relationship at the municipal level is essentially practical. The EPH Survey
has been built in order to be representative at the urban agglomerate level, and we
cannot know in which municipalities the households interviewed are located. Since we
only know the urban agglomerate in which each household is located (and its respective
weight into the overall agglomerate population) we are forced to generate regulation
indicators for the urban agglomerate level.
In order to generate regulation indicators characterizing the urban agglomerate level we
need to aggregate the indicators for the municipalities that comprise each one of them.
We therefore need appropriate weightings to generate weighted averages of the
19
See more on the definition of indicators in the previous section.
21
regulation indicators. Our first criterion in weighting municipalities’ indicators is the total
population. The resulting formula is the following:
  



Where k indexes each of the indicators described in this section, j indexes each one of the
28 urban agglomerates, i indexes each municipality, and I is the total number of
municipalities in the urban agglomerate j.
20
A source of concern may arise however, when using the population variable as the
weighting variable. The reason is that, if there are interaction effects between
municipalities, and then the population is an endogenous variable reflecting the
regulation characteristics across municipalities. Using the population variable as a weight
of indicators in this case would yield the result of giving less weight to regulation
indicators in those municipalities with stricter regulations of access to land, and more
weight to the indicators of municipalities with softer regulations.
iv. Econometric Approach
We analyze the relationship between households’ tenure condition and the explanatory
variables with the estimation of a panel data econometric model. Equation 1 below
illustrates an example of an econometric specification to be analyzed.



 
(1)
20
The resulting indicators might be interpreted as the “average regulation indicator at the
urban agglomerate” (e.g., the average presence of access to land elements in the
regulation at the urban agglomerate level) and, as explained above, this simplification is
made in order to match “average tenure choice indicators at the urban agglomerate”. It
should be emphasized that, in the case there is free mobility of people at the urban
agglomerate level, a stricter access-to-land regulation in a certain municipality will
probably externalize the tenure choice condition in the surrounding others. An average
approach to the regulation at the urban agglomerate level will then be appropriate in the
particular case that mobility costs are low enough to encourage within-urban agglomerate
migration but high enough to constrain between-agglomerates migration. It also follows
from this reasoning that the econometric methodology will have to test for a possible lack
of independence between urban agglomerate level observations.
22
Where i is the sub-index for the household tenure, the dependent variable, is one of the
three definitions of the formal-informal tenure dummies explained above. Results for the
three models are reported in this paper.
k1, k2, k3, k4 and k5 stands for the number of regulation indicators, and the number of life
cycle, wealth, social vulnerability, and location-related variables respectively. The
definition of variables is given in Table B1. The basic statistics of the variables are reported
in Table B.5.
regulationindicator stands for each of the regulation indicators that were defined in the
previous section (i.e., LPI, ZRAI, BRI, IPI, ALRI, and ACI) and where averaged for each urban
agglomerate. The group of demographic variables include: size of the family, age of the
head of household and marital status. The group of income and wealth-related variables
include: the household income, head of household level of education, and other income
proxies. The social vulnerability-related variables incorporate: gender (of the head of
household), the immigrant condition, the economic dependence, status in the job market.
Location variables incorporate: urban agglomerate. Notice that in our household-level
model of tenure choice it is reasonable to assume strict exogeneity of the regulation
indicators, since they are defined in the urban agglomerate dimension, and therefore not
affected by individual decisions.
The model is estimated using a Probit regression with weighed data, which includes all
demographic, socioeconomic and location controls variables, and the regulation indicators.
V. Estimating the Percentage of Tenure Informality using Municipal-level
Census Data (2001)
Methodologically, one of the most problematic issues in the previous tenure choice analysis
is the aggregation of regulation indicators into the urban agglomerate level a limitation
forced by the National Survey data-. We exploit a second source of data in order to avoid
the problem and compare results. The 2001 National Census counts with tenure information
available for each municipality for which we have regulatory information. The data then
allows to test the regulation indicators at the municipal level.
The second source of information has its limitations, however. The data is not available at
the household level (only aggregate figures are publicly available), and therefore we cannot
replicate the exact model. The second, and most important, limitation is because our
regulation indicators have been created according to 2009 information, and therefore we
might have certain biases due to the temporal mismatch of eight years
21
.
In this case, we estimate a cross-section model in a database where each observation covers
the jurisdiction of a municipality. The aim here will be simply to explore the relationship
between regulation indicators and the tenure measures
22
. The econometric model to be
estimated is defined as follows:
21
Nevertheless, we do not expect significant changes in regulation in this period, since in most jurisdictions
the regulation has been reported to be outdated.
22
Causality will not be possible to identify since regulation is theoretically expected to be determined by the
tenure condition of the municipality population.
23



 
(2)
Where j is the subindex of the municipality jurisdiction; h1, h2, h3 and h4 stand for the
number of regulation indicators and demographic, wealth, and location variables
respectively.
The dependent variable, pformaltenure, is the percentage of households declaring a formal
tenure of the dwelling in which they live. The formal tenure measure used here is similar to
the first definition employed in the previous tenure choice model. Formal ownership is
defined as those head of households who state they are owners of the house and the land,
renters or legal occupants (i.e., with authorization)
23
.
As in equation (2) regulationindicatorj stands for the regulation indicators defined in the
previous section. Also, the following socio-demographic controls percentages were
incorporated: population below 14 years old, senior population (i.e., above 65), male
population; population not born in Argentina and the number of years of education. As
regards wealth-related controls, the following indicators were included: i) percentage of
population with material resources needs, according to a deprivation index (i.e., Índice de
Privación Material
24
); ii) the percentage of population with at least one unmet basic need,
according to the unmet basic needs indicator (i.e., Índice de Necesidades Básicas
Insatisfechas
25
). It should also be noticed that we do not incorporate an income measure.
This is because within the census questionnaire there is not a question about household
income. Finally, we incorporated as a control the percentage of population that has
immigrated from other localities, provinces or countries in the previous five
26
years, and the
standard dummies for different regions in Argentina.
The model is estimated using a standard Ordinary Least Squares (OLS) method, and in
order to take into account some geographic interactions that might be in place, we proceed
to adjust for errors in the regression allowing for possible error correlation within each
urban agglomerate.
27
VI. Results
III. i. Regulation Survey Preliminary Results
The responses from a nationwide survey of residential land use regulation in nearly a
hundred municipalities across Argentina are used in this study to develop a series of
indicators to capture the stringency and main features of local regulatory environments at
23
See more details on variable definitions in the Appendix.
24
The Índice de Privación Material is an indicator available in the 2001 Census database which establishes a
criterion to measure the lack of material resources among the population. Several observable variables are
used with the aim of recognizing population with deprivation of current (i.e., short term) material resources,
those with deprivation of patrimonial (i.e., long term) material resources or the ones with deprivation of both
(defined as convergent).
25
Since these indicators are highly correlated we decided to use only one of them for alternative
specifications.
26
We take this period of time since this is (only) the period which is asked in the immigration question of the
Census.
27
Recall the discussion above of the interaction between several municipal jurisdictions within a same
agglomerate.
24
municipal and provincial level. But first, we describe what the average land use regulatory
environment looks like. The Survey of land regulation has been divided into
several categories, providing information on the general characteristics of land management
processes, detailed aspects of land use regulation for residential uses, infrastructure, fiscal
issues, land market generalities and legal processes for land access and registration.
The first set of questions elicited information on the levels of government and the main
normative instruments, which have an effect on each jurisdiction. In nearly 30 percent of
the municipalities there are some higher level of government norms, such as provincial
laws or/and plans for land use, that provides basic guidelines for land use at the local level.
(e.g. Buenos Aires, Chaco, San Juan) (Tables C.1, C.2 and C.3). As well, several provinces
granted total autonomy to municipal bodies for setting all land use regulatory requirements
in their jurisdictions, such as Cordoba, Catamarca, Neuquén, La Pampa, San Luis, Salta and
Santa Cruz, among others. At an aggregate level, it seems that a considerable proportion of
local jurisdictions have municipal plans for land use (70 percent) , while most of them have
a set of ordinances to regulate land use (93,5 percent). Some jurisdictions have municipal
ordinances as the main -and single- body for land use regulation. As well, 28 percent of the
jurisdictions have another set of complementary norms and plans as part of their main
regulatory environment for land use. For example, Rosario has a particular set of
comprehensive complementary plans (i.e, preservation of historic areas, metropolitan
plans, among others) while others have strategic plans, sometimes not linked to current
ordinances for land use.
What is interesting as well is the degree to which regulations are up to date. The question
asking for the date of last review of municipal or provincial plans and laws shows a great
disparity among replies. Few jurisdictions have recently updated their plans (both
municipal or provincial) and many of them have long standing plans and laws regulating
land use, dating from a maximum of 47 years, for municipal plans (average of 11.8 years)
to 15 years for provincial plans, on average with maximums of 32 years, such as Buenos
Aires (table C.3 and C.4) While few municipal ordinances have been recently updated, on
average there is 12.1 years since last updated date and 36 years maximum since last
adaptation for the oldest ones.
A second set of questions inquired on the general characteristics of the regulatory process,
which dealt with who is involved in the process (e.g., states, localities, councils,
legislatures, courts, etc.) and who has to approve or reject zoningor rezoning requests.
(Graph C.1) We also asked for other factors which guide the normative framework for the
regulatory process in each municipality. The question listed six different entities/groups
ranging from a local planning commission to an environmental review board. The more
groups with approval rights, the more potential veto points for any given development
proposal, which can also be interpreted as reflecting a more stringent, bureaucratic and less
laissez faire local regulatory environment. Any project requiring a zoning change is mainly
approved by the legislative local council or the Municipal Executive body. Other bodies,
such as planning commissions or planning offices, still have a relevant role in many
jurisdictions for changes in zoning approvals (35 percent). Involvement of provincial
bodies is still significant (30 percent), mainly in municipalities where Provincial laws and
plans are in place, like those municipalities in Buenos Aires Province which still have less
autonomy in this kind of decisions. Environmental review boards are less involved in
granting zoning changes, however in several localities their participation is mandatory (10
percent of the jurisdictions). Several jurisdictions have another set of complementary
25
requirements for granting zoning changes, such as the legislative power of the provincial
government, or more specific commissions formed ad hoc within the municipal
administration.
At the time to inquire about the way in which institutional mechanisms are currently used
for enforcement or consultation, few participation of legislative judicial bodies (in only 3
percent of jurisdictions) is shown, while citizen participation is exercised in 60 percent
of the jurisdictions. (Graphs C.2 and C.3) This last issue can be considered as a measure
of direct democracy and captures whether there is any kind of community meeting or
assembly before which any zoning or rezoning request must be presented and voted up or
down. It can also be taken as a measure of more tight restrictions set by the community to
avoid zoning changes and densification.
Project approvals which do not require any zoning change are mainly handled by Cadastre
commissions, public works and planning offices. Only in less than 30 percent of the
jurisdictions other bodies, such as environmental commissions or other public sector
officials are involved for granting approval of permissions for new projects. (Graph C.4)
A third set of questions pertained to the rules of local residential land use regulation. These
included queries as to whether there are any permits on new constructions, such as FOS
(factor of plot occupation ) and FOT (factor of total occupation), as well as information on
the presence of minimum lot size requirements, donations or collaborations for affordable
housing requirements, open space dedications and requirements to pay for infrastructure.
The wide variety of jurisdictions encompassed in the urban agglomerates show, on average,
that 25 percent of zoned land is designated for low density residential use, while 12 percent
is allocated for high density and 20 percent for mixed uses. However, several jurisdictions
devote up to 80 percent of its zoned area for low density residential use. Industrial uses
involved 10 percent of local land with maximums of up to 50 percent) while rural use
is provided for 31 percent of usable land on average. (Table 5)
Some type of density control is exercised in most municipalities. However, it is quite
interesting to note that not all jurisdictions have minimum lot size requirements for low
density areas (72 percent) ,and 67 percent have this kind of restriction for high density
areas. The average sizes for both minimum lot size requirements are 495 and 393 square
meters, respectively. However, the variation across jurisdictions is high. Stringency is as
low as a 100 square meters minimum plot through 1000 square meters for residential high
density use. (Table 6)
Other density requirements are FOS and FOT, which are present in 90 percent and 84
percent of municipalities, respectively. This question tries to capture whether there were
any statutory limits on the number of square meters for building permits which are
authorized for construction in any given plot. They are considered to be constraining
housing supply by 66 percent and 64 percent, as expressed by respondents perceptions,
when computing as affirmative responses in our analysis, those replies that together get 4
and 5 points from the 1 to 5 scale. (Table C.6 and Graphs C.5)
The determination of an urban perimeter or boundary within which urbanization can take
place, is observed in 48 percent of the municipalities. (Graph C.6) The basic infrastructure
and services that the municipality provides within this perimeter varied considerably among
jurisdictions. There is great heterogeneity among jurisdictions, street sweeping, cleaning,
and lighting, and pavement are essentially provided by most municipalities, while some
jurisdictions provide sanitation and water services, as well. (Graph C.7) Another potentially
important facet of the local regulatory environment involves requiring developers to pay a
26
share or total costs of any infrastructure ( or its improvement) associated with new
development. In a third of localities, mainly those of Buenos Aires Province, the developer
has to provide basic infrastructure services in order for subdivisions to be approved, at the
same time that municipalities are responsible for paved streets and public lighting, as well
as sanitation and sidewalks, to a lesser extent. Private companies are in charge of providing
electricity, gas and water in many jurisdictions (41 percent, 32.5 percent and 34.5 percent
respectively), while sanitation provision has altered between private firms and public
sector. (Table C.7)
Another set of questions are mainly related to several factors for land access, some of
which can be highly significant for low income housing. Provinces and localities have
substantial influence over land systems that conditioned supply , not only land-use planning
and enforcement , such as subdivision requirements or indicators of land occupation, but
also the real property tax and the public deeds registry, among others.
First, respondents were asked about social housing provision. This question was included
considering that social housing, mainly the construction of finished housing units financed
completely by the public sector, is still the main policy intervention for housing the very
poor. (World Bank, 2006) National and provincial governments are the main providers of
housing to address the needs of the lower income households. Although this strategy is not
able to cope with overall housing needs, land availability for public programs still severely
limits overall policy intervention. On average 636 and 578 units have been provided in each
jurisdiction with maximums of up to 6.200 and 4.000 units. However, on average, the
number of units provided is far less than needed.
28
Municipal bodies and NGOs
complement the construction of social housing by providing 85 and 50 units on average per
jurisdiction, respectively. Informal comments with public sector authorities point out that
land availability is the most severe constrain for implementing social housing programs in
many localities. The next set of questions address the instruments available in legislation
and land use regulation that may be used by local/provincial governments to overcome this
situation. (Table C.8)
There is great heterogeneity among municipal jurisdictions on the elements that form part
of their regulatory framework. Rural land preservation, public purchase of land for social
uses or legal regularization of informal settlements are present in almost 50 percent of the
municipalities’ regulatory frameworks for land use regulation. However, land reserves for
social housing, fiscal instruments for added value capture, and donations for social housing
uses are present in 10 percent of the jurisdictions. As well, obligations for selling land for
public infrastructure is widespread (87 percent of the jurisdictions) Other instruments, such
as fiscal ones that allowed for differential fees to be applied for mobilizing vacant land, are
present in 2 percent of the municipalities. (Graph C.8) Fiscal incentives, such as reduced
fees for construction projects which are localized in areas where revitalization programs are
implemented, are established in 33 percent of the jurisdictions
Cadastre registers are updated in 52.5 percent of the municipalities (standard deviation,
7.76). While 24 percent of the cadastre registers were updated in 2009, more than 50
percent of the updating was done in the last 5 years. However, there are still many
jurisdictions where last updating was done before 1980 (5 percent). Some jurisdictions have
not had their registers updated since 29 years ago. (Table C.9)
28
Further estimations about the ratio of social housing to population with unsatisfied basic needs or living
below the poverty line are been included in next report.
27
The survey inquires about Municipal revenues (as a way in which the municipal budget
might allow for capital investments that will favor the poor) On average, jurisdictions
collect 63 percent of payments issued, but efficiency in collecting shows great variation
among jurisdictions. (Table C.10 and C.11)
Vacant land is a main issue in Argentina, as well as in many other Latin American cities
(Clichevski, 1999). Amounts of vacant and under-utilized land lie within cities, in general,
and Greater Buenos Aires, in particular. Many academics have pointed out that accessing
this vacant land offers one of the most effective levers at hand for affordable housing
development of both the public and private sector. Privately owned parcels can be
developed in a straightforward way, through incentive mechanisms to place them in the
market. A first step consists of understanding the legal status of vacant lots. Although much
of this land is privately-owned, (60 percent of vacant plots, on average in each jurisdiction),
another 16.5 percent of plots have complex ownership problems that are difficult to solve,
particularly in Greater Buenos Aires. As well, substantial tracts are also owned by a number
of government agencies which are estimated to cover up to 21 percent of vacant land
available in each jurisdiction (7 percent municipal, 6 percent .provincial and 8 percent
national) (Table C.13)
Several questions elicit information about technical aspects of regulation and fiscal policy.
In particular, we want to inquire about urban growth strategies and whether they are
implicit in regulatory or fiscal instruments. It is interesting to show that 45 percent of
jurisdictions encourage further urban development through completing existing
urbanization, rather than supporting urban extension and sprawl (18 percent). Densification
is encouraged as main urban growth strategy in 22 percent of the municipalities, while 12
percent have not defined a particular urban growth strategy. (Graph C.9)
The perceptions about what would be the most severe limitations that hindered access to
land point out in the direction towards high land costs (52 percent) high cost of
infrastructure (43 percent) and the low income of the population that need to be supported
by some kind of public policy (80.5 percent), as evidence of a considerable increase in the
house to income ratio that reduces households affordability. Updating of land use
regulation is considered important for 27 percent of the respondents, while concentrated
land ownership is problematic for 13.5 percent of jurisdictions, on average. (Table C.14)
Finally, other set of questions on technical issues focused on the average time for a
residential project from its initial presentation to final approval. It takes 31 days for a single
family dwelling (from as low as 2 to 180 days) and 52 days, on average for a multi family
project, with a maximum of 240 days, showing a great heterogeneity between jurisdictions.
The average cost is $ 500, with maximums of $ 1.200. (Tables C.15 and C.16)
IV. ii. Land Regulation Indicators
Tables A2 and A3 provide some insights on what our regulation indicators can tell about
land regulatory environments across Argentinean jurisdictions.
29
In Table A2 we report the average value of indicators according to the country geographic
regions
30
and to the size (measured by total population) of jurisdictions, Recall that
indicators have been standardized to have a mean of zero and standard deviation equal to
29
See the Methodological Section for the definition of the Regulation Indicators.
30
The regions are Great Buenos Aires (GBA), North-West Argentina (NOA), North-East Argentina (NEA),
Patagonia, Pampeana and Cuyo.
28
one. It follows that the sign of an average value in Table A2 will point out whether the
jurisdictions involved are above the whole sample average jurisdiction positive sign- or
below the average -negative sign-.Also, an average value above one for a category will
point out a group of jurisdictions which average deviate from the mean by more than the
average deviation.
There are some regions that display higher values in most indicators than others. For
example, the Pampeana region displays the maximum regional average for the LPI, ZRAI
and ALRI indicators, and the second maximum for the IPI and ACI. Conversely, Patagonia
for example, presents the minimum values for four of the six indicators (LPI, BRI ALRI
and ACI). This pattern might not be surprising, since municipalities located in the
Pampeana regions are typically more populated than those in Patagonia (or NOA). But still,
when considering all regions and indicators, there is no clear pattern of correlation among
them. Taking for example the case of the North-East Argentina (NEA) region, Table A2
shows that the region is one where there is less presence of land use regulation plans (-
0.09), it is the region where there is least difficulty for obtaining a zoning or a regular
project approval (as measured by the ZRP Indicator, -0.31), but is the region with the
maximum value for the IP Indicator (0.3), suggesting a stronger participation of the
municipality in the provision of infrastructure.
31
Table A2 also provides evidence about the pattern of regulation according to the size of the
municipalities
32
. The LP Indicator increases monotonically with population. This is not
surprising since literature from developed countries have already shown that more
populated areas are the ones more highly regulated. This is confirmed for the Argentina
case, where more populated jurisdictions have higher values of the indicator of land use
plans (LPI) as well as higher values for the ZRA Indicator.
Both the indicators for regulation related with building restrictions (BRI) and to access to
land (ALRI) also increase with population, although in the case of the BRI the highest
value for the indicator is found for the second from the top- quintile of municipalities with
more than 153.000 and less than 288.000 inhabitants (0.41), and in both cases (BRI and
ALRI) the minimum value for the indicator is found in the second from the bottom-
quintile of population (-0.38 and -0.15 respectively).
The indicator that results somewhat surprising or seems no to be correlated with population
is the IPI (Infrastructure Provision). The IPI shows a higher value in the smallest
jurisdictions, which indicates a higher degree of public financing of the new infrastructure.
As these are urban agglomerates it may not be surprising to find that some complete
urbanized municipalities do not finance infrastructure provision. But still, the relationship
found between the IPI and population is still not clear.
31
The results for the Local Assembly, for example, show that Pampeana and Patagonia have higher values for
this sub index, in terms of presence of elements of direct democracy required for changes in current zoning.
This indicates that in these jurisdictions, changes in current zoning must be presented and debated or
approved in community meetings or local assemblies. This type of requirement is more concentrated in these
regions, as well as in cities with more than 288.000 inhabitants. It would be expected to be easier to block
projects in order to avoid congestion or increased densification using this instrument.
32
The categories distinguish municipalities that have less than 24.500 inhabitants, more
than 24.500 and less than 51.500, more than 51.500 and less than 153.000, more than
153.000 and less than 288.000 and more than 288.000 inhabitants.
29
Finally, although the maximum approval costs are found in the largest jurisdictions, the
figures for the ACI indicator do not suggest a clear relationship between these costs and the
population.
Correlations across the Regulation Indicators
Table A.3 reports simple correlations across the indicators. Nearly all of them are positive,
although only a few of them are statistically significant.
As was suggested by the previous correlation with population, the LPI, and the ZRAI are
positively and significantly correlated. Once again, those jurisdictions that have
incorporated local or provincial level land plans or have incorporated them into their
regulatory framework tend to need the permission of more authorities for zoning changes or
regular projects. The LPI indicator also appears positively correlated with the infrastructure
provision indicator (IPI). The association is significant at conventional levels. Plans may
establish how basic infrastructure and public services are to be provided in urban and
suburban areas, i.e. by defining an urban perimeter where new residential developments are
granted basic infrastructure services provision and defining whether the private public
sector firms or the public sector is going to finance its extension to these areas.
Notwithstanding this, both mentioned positive correlations are still below 0.5, (0.29 and
0.39), which indicates that still both indicators maintain a degree of independence.
Building restrictions exhibit a positive and significant correlation with infrastructure
requirements (IPI), and this correlation is statistically significant at conventional
significance levels. This result suggest that municipalities that impose lot size restrictions,
as well as tight maximum total building and maximum land usage restrictions, tend to
guarantee the provision of public services inside a defined urban perimeter, or alternatively,
in those jurisdictions where the municipality has responsibility to finance part of the
services, requirements to the developers on the supply of land are also set. Also, those
municipalities that guarantee the access to infrastructure within its boundaries may also
decide to constraint the generation of new lots in suburban areas (e.g. with minimum size
lots) in order to restrict the expansion of infrastructure at a rate higher than what is possible
to finance.
Finally, it is interesting to note that the indicator for approvals costs (ACI) is also positively
and statistically significant correlated with the IPI. This might be another indicator in the
same mentioned direction, that municipalities that regulates the access to infrastructure or
take an active role in its provision might delay the time between the application and the
approval of standard projects, and therefore increase the costs that are related to residential
projects registration procedures.
iii. Households Tenure Condition in Argentina
The National Households Survey allows a description of the tenure condition of the
Argentinean population. According to Survey responses, nearly 91 percent of households
have a formal type of tenure of the dwelling they occupy (Formal Measure I in Table B2).
Formal ownership, which is defined as the legal ownership of the land and the dwelling,
accounts for 67 percent of households. Formal renters equal another 16 percent. Another 6
percent of households are legal occupants. The remaining 9 percent of households maintain
an "informal" form of tenure. This group comprises owners of the dwelling but not of the
land, self-claimed owners which pay property taxes, and other illegal occupants. Table B3
allows checking the degree of variability of formality across urban agglomerates in
Argentina. Agglomerates such as Gran Resistencia or Gran Tucumán display the largest
30
percentages of informality with 81 and 84 percent of households under formal tenure
conditions. On the other extreme, Rawson-Trelew reaches 98 percent of formality.
As previously mentioned, there is a risk of over-estimating formality measures, since there
are reasons to believe that households living in informal settlements which have paid for
the land they occupy considered themselves to be homeowners, although no formal title has
been granted.
Slums are characterized by the illegality of tenure, or some risk on the formal (or the
partial) tenure condition, as well as the lack of basic infrastructure. Then, the definition of
informal settlement becomes crucial for our analysis. According to the Survey (Table B2),
2 percent of the households live in these kinds of slums. In some urban agglomerates, as the
case of Posadas and Gran Resistencia, the figure climbs to 8 percent and 9 percent
respectively
33
. When incorporating households living in emergency settlements into the
informality group, the measure of formality for the whole country displays nearly no
changes (90 percent, Formal Measure II, Table B2).
We next test a different definition of an informal settlement, incorporating the access to
basic infrastructure. As mentioned above, the lack of water and sewerage basic
infrastructure is considered in order to approximate true informality. Table B2 shows that, 5
percent of households in Argentina obtain their water from a manual pump. In addition,
nearly 8 percent of households do not have a water connection inside their dwellings. In the
case of urban agglomerates such as Posadas and Gran Resistencia, the figure climbs to 14
percent. In other regions, such as Concordia and Formosa, which according to the definition
of the Survey do not have households living in emergency settlements, the figure of
households without water connection inside their dwelling reaches 10 percent and 15
percent, respectively.
34
We also consider the basic sewerage infrastructure. We incorporate
in the informal group those households without a sewerage network connection, nor septic
cam or cesspool. In Argentina 1 percent of households live in this category.
When we incorporate in the analysis the physical conditions of the settlement, the
associated Formal Measure tenure condition decreases to 86 percent (Formal Measure III,
Table B2). Concluding, the incorporation of the most basic infrastructure might be useful to
check for the true formality percentage, extending the existent measured informality
percentage from 9 percent to 14 percent.
V. iv. Tenure Choice Model Econometric Results
Results on Regulation Indicators
The econometric results on the effect of regulation on the tenure condition of households
suggest that in those jurisdictions that have incorporated land plans into their regulatory or
legal frameworks (as measured by the LPI) it is less likely that households will obtain a
formal tenure condition. Higher residential project approval costs (measured in monetary
and time terms) seem also to have a negative effect on the probability of obtaining formal
tenure. The rest of our regulation indicators are found to display no clear relationship with
the formal tenure condition.
33
These figures are not shown.
34
These figures are not reported in tables and are available upon request.
31
Table B.5 displays the results of the econometric model estimation for the regulation
indicators. The three columns allow the comparison of results across the three alternative
measures of the formal tenure condition.
35
In the case of the LPI, a negative and statistically significant relationship with the formal
tenure condition holds across the three alternative definitions of the dependent variable.
The results suggest that in those jurisdictions where plans for the use of land have been
formally established in the legal or regulatory framework, households find it more difficult
to access formal tenure, or in other words, a household will have a higher probability of
being informal. An increase in 1 point in the indicator reduces the probability by 13
percent.
36
The effect increases (to 17 percent) when considering Measure II (incorporating
those households located in emergency settlements), but decreases in the case of Measure
III (-11 percent) suggesting that LPI provides a poorer explanation of those households
without access to basic infrastructure. The negative effect of the LPI in the formal tenure
condition might suggest that those jurisdictions that have incorporated land planning into
their legal or regulatory frameworks have tended to limit the increase of the formal
residential sector, in concordance with theories of exclusionary regulation (See for instance
Biderman, 2008).
37
38
.
A significant effect of regulation on the tenure condition is also found for the Approval
Cost Indicator (ACI), displaying a negative influence on formality. An increase in a value
of 1 in the indicator is associated with a decrease in the probability of tenure of 9 percent.
The result might therefore reflect that time and costs related to residential projects
approval/registration procedures lower the probability of households becoming formal
owners/renters. As explained above, the intuition of the result is straightforward, and might
be evidence in favor of the supporters of the simplification of approval processes. The
result might also reflect the existence of an implicit exclusionary policy in certain
agglomerates.
In the case of the Building Restrictions Indicator (BRI), results are less clear. A positive
effect on the formal tenure condition appears for Measures I and II, suggesting that in those
urban agglomerates with, on average, tighter building restrictions, households have a larger
probability of having a formal tenure condition. This result would contradict our a priori
hypothesis, since restrictions limiting intensity of land use might affect low-income
households, who have a higher demand of high density developments than consumers who
are better off. (Henderson, 2009 in Lall et al., 2009) Nonetheless, the effect disappears
when incorporating into the informality group those households with the least access to
infrastructure (Measure III), and so further analysis is needed in order to draw definitive
conclusions in relation to this indicator.
The ZRAI, IPI and ALRI Indicators appear consistently to reflect no statistically significant
effect on the tenure choice condition.
39
35
See the discussion on these definitions in the Methodological Section above.
36
Recall that indicators have been standardized to have a standard deviation of 1.
37
Notice that both provincial plans and municipal regulation have been jointly analyzed in this stage. Further
research should evaluate the different levels of regulation separately.
38
As explained in the literature section, many provincial-level plans might have replicated the Buenos Aires
Law of 1977, which had the objective of limiting the uncontrolled development of land, and defining the
conditions for formality and excluding the lowest segment of the demand.
39
In particular the result for the ZRAI was unexpected; since our ex-ante hypothesis was that the ZRAI would
proxy the degree of planning (and in line with the exclusionary regulation, might replicate the negative effect
32
Demographic, Socioeconomic and Location controls
As expected, most socioeconomic and wealth related characteristics of households result in
statistically significant explanatory variables of theirformal or informaltenure
condition. Significant variables include household size (negative effect), the quantity of
children (negative), household income (positive), and the formal employment condition
(positive). Other vulnerability conditions such as the sex of the head of household, or the
migratory condition exhibit no significant relationship.
Table B.7 reports the estimated marginal effects.
40
Once again, each column represents a
specification of the model for each of our formal tenure measures. In relation to
socioeconomic variables, for example, the model predicts that education increases the
probability of having a formal tenure: 22 percent to 37 percent increase for a household
with complete primary, 17 to 26 percent for complete secondary, and 15 to 19 percent for
complete university.
41
The age of the head of household (and its square) resulted in a not
significant effect (or an effect close to 0). The household size and the number of children
(under five years old) in the house are variables negatively related to the tenure condition.
An extra child decreases the probability of formal tenure by 12 percent (calculated at the
mean of 0.3 children) and an extra person in the house reduces the probability by 3 percent
(at the mean of 3.7 people), which can be explained by the fact that poorer households have
relatively more children.
As expected, there is a positive relationship between income measures and the formal
tenure condition. The logarithm of the total household income variable is significant and
positive, predicting an increase of 16 percent to 33 percent in the formal tenure probability
for an increase of a 1 percent in incomeat the mean of 1,652 pesos. Another important
variable is the formal work condition. The variable is found to have a positive effectof 10
to 22 percent according to the specificationon the formal tenure condition. Finally the
economic dependence variable does not show a significant effect, which might explained
by the fact that several income related variables have already been incorporated as controls.
In general no effects are found on the tenure condition for the social vulnerability related
variables. No effect is found for head of household employed in domestic service, no clear
effect for the sex of the head of household, or for the migrant status of the head of
household (considering both domestic and international migrants).
Finally, notice that explanatory power of the model is low. The pseudo R-square measure is
7 percent in Model I and increases to 13 percent in Model III. The incorporation of the
population with lack of basic infrastructure in the informal group improves the explanatory
power of the model.
of LPI). In the case of the IPI indicator, no clear effect was expected nor in the case of the ALRI. Notice that
the existence of access to land elements in the regulation might also be an endogenous factor, taking place
when an important percentage of the electorate is under an informal condition. The expected effect was not
clear.
40
Calculated on the average value of the continuous explanatory variables, and representing the discrete
increase from 0 to 1 in the case of the dummy variables.
41
Notice that since the three education variables are jointly incorporated in the model, and given the fact that
informality affects the segment of the population with the lowest educational level, it is not strange to find a
lower coefficient for a higher level of education.
33
VI. v. Percentage of Population with Formal Tenure Model:
Econometric Results
The estimation of the model of the percentage of population with a formal tenure using
municipal-level data allows the comparison of results with those from the previous tenure
choice model. Table B6 reports the results for the estimation of equation (2).
42
Columns
(1) (2) and (3) report different specifications of the model where some of the
socioeconomic variables were alternatively added. The model is jointly significant and has
been finally estimated with 62 observations
43
Regulation Indicators
In general the comparison with the tenure choice model reported three indicators with
coincident results (LPI, BRI and IPI), two of them could not be corroborated (ACI and
ZRAI), and one indicator presents no concluding results (ALRI).
In the case of the LPI, results are corroborated, displaying a negative and significant (at 5
percent) coefficient. The negative relationship found here between the LPI and the formal
tenure percentage corroborates the negative effect that has been previously found for the
existence of land plans in the presence of informality. The estimated coefficient suggests
that an increase in a value of 1 in the indicator
44
is related to a decrease in formal tenure of
nearly 1 percentage point.
Also consistent with the results of the tenure choice model, the estimated coefficients
suggest no relationships between the formal tenure percentage and the Building
Restrictions Indicator (BRI), nor the Infrastructure Provision Indicator (IPI). The respective
non-significance cannot be rejected at standard significance rates in any of the
specifications.
The results for the ACI and ZRAI are not coincidental from what was found in the previous
model. In the case of the Approval Costs Indicator (ACI), which in the previous model
presented a negative effect on tenure choice, is found to display no significant relationship
with the percentage of population with formal tenure.
Zoning and Approval Regulation Indicator (ZRAI), which in the previous model presented
no relationship with the tenure choice, shows here a negative relationship with the
percentage of formality across municipalities. This indicator, which captures the authorities
involved in the approval of zoning and regular projects, was a priori expected to display a
negative sign because of its relationship with exclusionary regulation. Notice also that the
relationship emerges once the percentage of population with unmet basic needs was
controlled for. However, more evidence is needed in order to achieve definitive results in
relation to this indicator.
Another indicator which displays the a priori hypothesis here is the Access to Land
Regulation Indicator (ALRI), displaying a positive coefficient. The positive coefficient
would validate the idea that the inclusion of regulation fostering the access to land has
indeed an effect on actual formality. In this case, the estimated coefficient suggests that an
increase in 1 point in the ALRI indicator can be related to an increase in the percentage of
households with a formal tenure of 0.5 percent. Notice also here that the relationship
42
See Section III.iv
43
From the total of 89 municipalities in our regulation database, several cases are lost due to non response in
certain regulation questions.
44
All indicators have been standardized in order to get a standard deviation equal to one, which implies that an
increase of 1 point is equivalent to an increase in the average deviation of the indicator.
34
emerges once the percentage of population with unmet basic needs was controlled for.
Nevertheless, our results for the tenure choice model cannot confirm this result and still in
this case more evidence is needed in order to arrive to definitive conclusions
Demographic, Socioeconomic and Location Controls
Socioeconomic variables present some interesting results. Recall that because of the
correlation among them, two of the variables that capture the lowest income segment of the
population (i.e., percentage of population with material resource needs according to a
deprivation index and the percentage of population with at least one unmet basic need)
were added to the regression in individual specifications. In the case of the proportion of
population with at least an unmet basic need, the expected negative coefficient was found,
predicting 2.6 percent of decrease of informality for a reduction of 10 points in the
percentage of population with unmet basic needs. On average, 14 percent of the population
in the municipalities that were analyzed displays at least one unmet basic need, and this
percentage reached 34 percent in the poorest municipality. In the case of the percentage of
population with material resources needs (according to the Índice de Privación Material)
however, the hypothesis of no relationship cannot be rejected at standard significance
levels. An interesting result arises for the average number of years of education variable;
since it reports a negative coefficient, but only when incorporating the unmet basic needs
variable as control.
45
Once controlling for the percentage of population with most
probability of becoming informal, the education variable might approximate the presence of
other exclusionary mechanisms that increase informality in relatively more wealthyor
human capital concentrationmunicipalities. More precisely, once controlling for the
percentage of population with less income, a higher average educational level might
capture greater inequality for a given municipality, therefore giving place for exclusionary
mechanisms and leaving the lowest income individuals in the informal sector.
Demographic controls presented no relationship with the percentage of formality. No clear
relationship with the tenure condition was found for the demographic variables that
captured the percentage of youngest and oldest within the population. In addition, no
relationship was found for the total population as a control.
The percentage of population that migrated in the five years previous to the census was
found to be positively related to the formal tenure condition. In other words, those
municipalities that received greater immigration in previous years were those that on
average presented a larger percentage of formal tenure. This result is statistically
significant, with a confidence level of 1 percent. The result might suggest that population
migrates to municipalities where there is easier access to a formal tenure condition.
Finally, the only regional dummy that was found significant in this model is the one
corresponding to the NEA region. The result suggests that this region holds a larger degree
of tenure informality (nearly 4 percent) that cannot be explained by the other incorporated
variables.
VII. Conclusions
This paper has presented a set of empirical findings in relation to the complex topic of the
regulation of land for residential use, and its effects on households’ access to a formal
tenure of residential land. The topic has yet been scarcely accounted in the literature; in
particular there is little evidence for developing countries. This study, based on a survey of
45
The same result also appears for the percentage of population with secondary education, not reported in the
Table.
35
land regulation across municipalities in Argentina provides a valuable source of
comparable and systematic information of land regulation.
Our survey confirmed that regulation of residential land in Argentina is of a quite
heterogeneous nature, and only in some provinces have been guided by land plans that have
been incorporated in their respective legal frameworks. Also, in general land plans have
been found to be quite outdated, with average ages of more than 10 years in the case of
provincial-level or municipal-level land plans. There is great heterogeneity not only in the
nature of regulation but also in terms of its contents (e.g. the existence of zoning, the
authorities involved in approval processes, approval costs). Heterogeneity is also huge in
terms of the guarantee of provision of services (e.g. less than half of the jurisdictions
guarantee the provision of basic services such as public water supply, street lighting or
electricity within their urban perimeters), or the establishment of inclusionary elements in
the regulation (e.g. less than 50 percent of the municipalities consider the regularization of
lots with illegal occupation and less than 20 percent incorporate measure to recoup value
added).
Do our regulation measures have an effect on the actual tenure condition of households in
Argentina? The regulation data obtained from the survey was synthesized into thematic
indicators, and then analyzed in relation to the residential tenure patterns across the country
in order to provide insights into this question.
The research provided a first set of insights. The most consistent result across our models
and specifications is the negative effect found for the existence land plans, together with the
incorporation of these plans in the respective legal or regulatory frameworks, on the formal
tenure condition (i.e., the results for the LPI Indicator in the two estimated models). This
finding might be consistent with the hypothesis of a “minimum consumption” regulation
(as suggested for example by Gyourko et al., 2007), that is translated into a relatively larger
informal land market for the lowest segment of the population. Although more results
specifically related with the contents of the regulation are needed in order to validate this
hypothesis, the present results call attention to this issue.
Other regulation thematic indicators have shown some results in line with our a priori
hypothesis, but given the fact that results have not appeared consistently in all our
measurements, more research is still needed in order to establish definitive conclusions.
This is the case for example of the Zoning and Regular Projects Approval Indicator
(ZRAI), which exhibit a negative effect on formality in most estimates (e.g. in the case of
the percentage of population with formal tenure model). Since more authorities involved in
project and zoning approvals are also expected to be correlated with tighter regulation, a
negative effect on formality was expected. Another interesting result was found in the case
of the Access to Land Regulation Indicator (ALRI), since as expected, exhibits a positive
relationship with tenure formality for those municipalities that report inclusionary policies.
Finally, in the case of the Approval Costs Indicator (ACI), a negative effect on the
probability of households of becoming formal owners or renters was found for those
agglomerates with relatively higher approval costs. Even though these results are
promising; more research is needed in relation to these dimensions of regulation.
Other indicators have shown no relationship with the tenure condition across all of our
estimations. This is the case of the provision of infrastructure Indicator (IPI), and the
building restrictions Indicator (BRI). Recall that while building restrictions explicitly limit
the supply of land, the literature has pointed out that the provision of infrastructure might
function as well as an indirect restriction of the supply (i.e. by not servicing land with
36
certain public infrastructure investments). At this point our evidence provides low support
for these hypotheses.
Finally, it is worth mentioning that the analysis has also brought to light several interesting
patterns regarding the relationship of socioeconomic variables with the tenure condition.
For example, the paper shows how controlling the tenure choice pattern with the percentage
of population with most probability of becoming informal (i.e. the population with basic
unmet needs) the average education level of a population appears to be associated with
more informality. In this case, the results might also suggest the presence of other
exclusionary mechanisms associated with a higher average level of human capital and
relatively higher inequality. (See for example: Henderson, 2007, or Henderson and Feller,
2008). Socioeconomic patterns in relation to tenure choice will also be closely examined in
a forthcoming paper.
There are several ways to deepen our understanding of this topic. First, the analysis of the
contents of regulation is still at a very early stage. For example, to separately analyze the
effect of specific components of the regulation, such as zoning or certain building
restrictions will help to better understand the effects of regulation. Another important point
to be added to the tenure choice analysis is the existence of vacant land. Forthcoming
analysis should measure the extent in which land is available across jurisdictions and
analyze how it affects the relative size of formal and informal residential sectors. Also,
there is a need to evaluate the existence of measures of regulation enforcement across
jurisdictions. Notice that this concept might be a crucial variable omitted from our analysis.
The incorporation of variables related with these concepts then becomes a priority in this
line of research.
Forthcoming analysis should also be focused in deepening the analysis of the determinants
of regulation: exploring the observable patterns of municipalities that are related with
regulation, and the changes that have originated regulation such as, for example, changes in
immigration patterns, or the effects of rapid population growth. Some initial findings in this
paper suggest that regulation reflects observable characteristics such as population, density
or other regional characteristics. Research is needed to better understand how regulatory
patterns are related to many observable characteristics of municipalities. In particular how
municipalities interact between them in the definition of regulation, a topic not sufficiently
covered in this paper. Finally, the present analysis should also be complemented with a
study of the effect of regulation on land prices, a topic not covered in our paper.
References
Bertaud, A. and S. Malpezzi. 2001. “Measuring the Costs and Benefits of Urban Land Use
Regulation: A Simple Model with an Application to Malaysia,” Journal of Housing
Economics, 10, pp. 393-418.
Biderman, C. 2008. Informality in Brazil: Does Urban Land Use and Building Regulation
Matter?, Working Paper, Cambridge, MA:Lincoln Institute of Land Policy
Capitani, Danilo. 2008. "Proyecto de Ley Ordenamiento Territorial y Usos del Suelo
Provincia de Santa Fe," Paper presented at the Conference "Reforma Urbana en
Argentina: Desafíos Legislativos y de Gobierno para la Gestión del Suelo," October 29,
2008
37
Cruz, B. de Oliveira, and Morais, M. da Piedade. 2008. "Housing Demand, Tenure Choice
and Housing Policy in Brazil," Housing & Urban Development Corporation Ltd, Vol. 11 No.
1 April.
Cairncross, Sandy. 1990. ‘Water supply and the urban poor’, in Jorge E. Hardoy, Sandy
Cairncross and David Satterthwaite (eds.), The Poor Die Young: Housing and Health in
Third World Cities, Earthscan Publications, London, pages 109-126.
Daniere, A.. 1992. “Determinants of tenure choice in the third world: An empirical Study
of Cairo and Manila,Journal of Housing Economics. 2 (1992), pp. 159-184.
de Soto, Hernando. 1989. The Other Path, I .B. Tauris Publishers, London
Edwards, Michael.1990. "Rental Housing and the Urban Poor: Africa and Latin America
Compared", In Housing Africa's Urban Poor. Philip Amis and Peter Lloyd (eds.), 253-272.
Manchester: Manchester University Press.
Epple, D. and T. Nechyba. 2004. “Fiscal Decentralization” in Handbook of Regional and
Urban Economics, Vol. 4, J.V. Henderson and J-F Thisse (eds.), North Holland.
Estache, A., V. Foster y Q. Wodon. 2002. Accounting for Poverty in Infrastructure Reform-
Learning from Latin America’s Experience. Studies in Development Series. Washington,
DC: World Bank Institute
Galiani, S., P. Gertler, and E. Schargrodsky. 2005. “Water for Life: The Impact of the
Privatization of Water Supply on Child Mortality, Journal of Political Economy 113, 83-
120.
Galiani, S., González-Rozada, M., Schargrodsky, E. 2008. Water expansions in
shantytowns: Health and savings,” Economica. Reprinted in Chong, Alberto (ed.)
Privatization for the Public Good? The Social Effects of Privatizations in Latin America,
forthcoming, Washington D.C.: InterAmercian Development Bank.
Gilbert, Alan. 1983. "The Tenants of Self-help Housing: Choice and Constraint in the
Housing Markets of less Developed Countries", Development and Change, Vol. 14, 449-
447.
Gilbert, Alan. 1993. In Search of a Home: Rental Housing in Latin America. London: UCL
Press Limited.
Glaeser, E.L., J. Gyourko and R. Saks. 2006. “Urban Growth and Housing Supply,” Journal
of Economic Geography, 6, 71-89.
38
Goodman, A. C. 1988. "An econometric model of housing price, permanent income,
tenure choice and housing demand", Journal of Urban Economics 23(1):327-353
Gyourko, J., C.J. Mayer and T. Sinai. 2006. “Superstar Cities” NBER Working Paper No.
12355
Gyourko, J., Saiz A., Summers A. A. 2008. A New Measure of the Local Regulatory
Environment for Housing Markets: The Wharton Residential Land Use Regulatory Index,
Urban Studies 45(3), 693-729.
Goytia C. and Lanfranchi,G. 2010. “Informal neighbourhoods in Buenos Aires Metropolitan
Region: understanding the effects of land regulation on the welfare of the poor.” in Lall,
S.V et al (eds), Urban Land Markets Improving Land Management for Successful
Urbanization, Spinger : XXXVII, 408 p.
Hardoy, J.E., Mitlin, D. & Satterthwaite, D. 2001. Environmental Problems in an Urbanizing
World: Finding Solutions for Cities in Africa, Asia and Latin America. Earthscan, London
Helsley R.W. 2004. “Urban Political Economics” in Handbook of Regional and Urban
Economics, Vol. 4, J.V. Henderson and J-F Thisse (eds.), North Holland.
Henderson, J.V. and E.Y. Ioannides. 1983. "A model of housing tenure choice," American
Economic Review, 73(1):98-113.
Henderson, J. V. 2007. The effect of residential land market regulations on urban
welfare. Urban Research Symposium. Washington, DC: The World Bank.
Henderson V and Feler L. 2008. “Exclusionary Policies in Urban Development, How under-
servicing of migrant households affects the growth and composition of Brazilian cities.
NBER Working Papers and LACEA.
Iwarere, L.J. and J.E. Williams. 1991. "A Micro-Market Analysis of Tenure Choice using the
Logit Model", The Journal of Real Estate Research, 6(3):327-339.
Lall, S. V., H. G. Wang, and D. da Mata. 2006. Do urban land regulations influence slum
formation? Evidence from Brazilian cities. Mimeo.
Lim, Gill Chin. 1987. Housing Policies for the Urban Poor in Developing Countries,”
Journal of the American Planning Association, Vol. 53, No. 2, 176-185.
Malpezzi, S. 1996. Housing prices, externalities, and regulation in U.S. metropolitan
areas. Journal of Housing Research 7(2):209241
39
Mayer, C.J and C.T. Somerville (2000) “Land use regulation and new construction,”
Regional Science and Urban Economics, 30(6), pp. 639-662.
McDonald, J. F. 1979. Economic Analysis of an Urban Housing Market. New York:
Academic Press.
Mehta, M. and D. Mehta. 1989. Metropolitan Housing Market: a Study of Ahmedabad,
New Delhi: Sage Publications.
Mills, W.E and W. E. Oates. 1975. Fiscal Zoning and Land Use Controls, Lexington Books
Pollakowski, H.O. and S.M. Wachter. 1990. The Effects of Land-Use Constraints on
Housing Prices, Land Economics, 66(3), pp. 315-324.
Rosen, H. 1979. Housing decisions and the U.S. income tax: An econometric analysis,”
Journal of Public Economics, 11:1-23.
Royston, Patrick. 2005. “Multiple imputation of missing values: Update of Ice.” Stata
Journal 5(4): 527-36.
Turner, John F. C. and R. Fichter. 1972. Freedom to build: Dweller control of the housing
process. New York: The Macmillan Company.
Water and Sanitation in the World’s Cities. 2003. United Nations Human Settlements
Programme, Earthscan, London.
World Bank. 2006. Review of Argentina’s Housing Sector: Options for Affordable Housing
Policy. Argentina Country Management Unit Finance, Private Sector and Infrastructure
Management Unit, Latin America and the Caribbean Region
40
Appendix A: Regulation Indicators Tables
Response
Rate %
Sample
Population:
Municipalities in
Big Urban
Agglomerates
Region:
No.
%
No.
%
Cuyo
73
11
12.4
15
12.6
GBA
83
29
32.6
35
29.4
NEA
75
12
13.5
16
13.4
NOA
63
12
13.5
19
16
Pampeana
70
16
18
23
19.3
Patagonia
89
8
9
9
7.6
Population (inhabitants):
less than 24.500
67
16
18
24
20.2
more than 24.500 y less than
51,500
67
16
18
24
20.2
more than 51,500 y less than
153,000
83
19
21.3
23
19.3
more than 153,000 y less than
288,000
80
20
22.5
25
21
more than 288,000
78
18
20.2
23
19.3
Total
75
89
100
119
100
41
Table A2: Regulation Indicators. Averages by Region, Population, and Density
LPI
ZRAI
BRI
IPI
ALRI
ACI
Region
Cuyo
-0.09
0.17
0.09
-0.63
0.23
-0.15
GBA
0.56
0.26
0.21
0.00
-0.03
0.19
NEA
-0.09
-0.31
-0.17
0.30
0.15
-0.04
NOA
-0.34
0.17
-0.02
0.00
0.51
0.04
Pampeana
0.63
0.68
-0.35
0.19
0.91
0.06
Patagonia
-0.42
0.27
-0.36
0.16
-0.13
-0.32
Population (Inhabitants)
menos de 24.500
-0.35
-0.30
-0.27
0.30
-0.12
-0.06
más de 24.500 y menos de
51,500
-0.04
-0.38
-0.43
0.01
-0.15
-0.17
más de 51,500 y menos de
153,000
-0.38
0.04
0.02
-0.08
-0.09
0.04
más de 153,000 y menos de
288,000
0.23
0.25
0.41
-0.29
0.11
-0.28
más de 288,000
0.50
0.38
0.11
0.14
0.23
0.51
42
Table A3: Indicators Correlation Matrix: All Municipalities
LPI
ZRAI
BRI
IPI
ALRI
ACI
LPI
1
89
ZRAI
0.3944
1
0.0002
85
85
BRI
0.1442
0.1618
1
0.1934
0.1465
83
82
83
IPI
0.2989
0.0946
0.188
1
0.0044
0.3893
0.0887
89
85
83
89
ALRI
0.1236
0.1389
0.1308
0.1426
1
0.2654
0.2162
0.2477
0.1984
83
81
80
83
83
ACI
0.0059
0.0048
-0.0187
0.2101
-0.0203
1
0.9624
0.9696
0.8822
0.0903
0.8727
66
65
65
66
65
66
Note: p value in italics for null of zero correlation
43
Appendix B: Tenure Choice Tables
Table B1: Variables and Definitions
Explanatory Variables
Life Cycle and Households Characteristics
Age
Age of the head of household
Head of Household level of
Education:
complete_ primary
Equals one if head of household has completed primary
school, equals zero otherwise.
complete_secondary
Equals one if head of household has completed secondary
school, equals zero otherwise.
complete_university
Equals one if head of household has completed secondary
school, equals zero otherwise.
household_size2
Household size
qchild_under5
Number of children under 5 years living in the same house
Wealth and Income Characteristics
income_household_total
Total income per household
income_per_capita
Total household income per capita
ind_econo_dependency
Individual income/total household income
head_income
Head´s total income
economic_dependency
Head of household income as percentage of total
household income
employer
Equals one if the head is an employer, equals zero
otherwise.
employee
Equals one if the head is an employee, equals zero
otherwise.
domestic_service
Equals one if the head works in domestic service, equals
zero otherwise.
formal_work
Equals one if the head has a formal job, equals zero
otherwise.
Basic Needs and Social Vulnerability
Unmet Basic Need Indicators
ubn
Equals one if Unmet Basic Needs (U. B. N.), equals zero
otherwise.
ubn_house_density
Equals one if U.B.N. people per room, equals zero
otherwise.
ubn_house
Equals one if U.B.N. house, equals zero otherwise.
Table B1: Variables and Definitions (Cont.)
ubn_sanit
Equals one if U.B.N. Sanitary Conditions, equals zero
otherwise.
ubn_edu
Equals one if U.B.N. Education, equals zero otherwise.
ubn_survival
Equals one if U.B.N. Survival, equals zero otherwise.
ubn_portion
Portion of U.B.N (one is worse)
sex
Equals one if head of household is male.
44
dummy_marital
Equals one if head of household is married or lives
together with a couple, equals zero if it is single, separated,
divorced or widow.
international_migrant
Equals one if the househoold is an international migrant,
equals zero otherwise.
domestic_migrant
Equals one if the household is a domestic migrant, equals
zero otherwise.
Locational Variables
more500
Metropolitan area, with more than 500.000 inhabitatns
dummy_gba
Dummy for Gran Buenos Aires region
dummy_noa
Dummy for NOA region
dummy_nea
Dummy for NEA region
dummy_cuyo
Dummy for CUYO region
dummy_pampa
Dummy for PAMPA region
dummy_patagonia
Dummy for PATAGONIA region
45
Table B2: Formal Tenure Definitions and Related Variables*
Variable
Mean
Standard
Error
Min
Max
Formal Tenure Definitions
Formal Measure I
0.909
0.0001
0
1
Formal Measure II
0.903
0.0001
0
1
Formal Measure III
0.857
0.0001
0
1
Owners
0.672
0.0002
0
1
Renters
0.162
0.0001
0
1
Other formal conditions
0.060
0.0001
0
1
Located in Emergency
Settlements
0.020
0.000
0
1
Water Source
Network
0.881
0.0001
0
1
Automatic Pump
0.114
0.0001
0
1
Manual Pump
0.005
0.0000
0
1
Water Connection
Inside dwelling
0.920
0.0001
0
1
Outside dwelling inside land
0.071
0.0001
0
1
Outside land
0.009
0.0000
0
1
Sewerage Connection
Network
0.637
0.0001
0
1
Septic Cam
0.251
0.0001
0
1
Cesspool
0.103
0.0001
0
1
Pit in the ground
0.009
0.0000
0
1
*EPH sampling weights have been used for calculations.
46
Table B3: Formal Tenure Condition by Urban Agglomerate. In
percentage
Formal
Measure I
Formal
Measure II
Formal
Measure
III
Gran Resistencia
81
80
76
Gran Tucumán - T. Viejo
84
84
78
Posadas
85
84
82
Jujuy - Palpalá
86
86
78
Ushuaia - Río Grande
89
89
89
Partidos del GBA
89
89
81
Gran La Plata
91
91
88
Formosa
91
91
76
Cdro. Rivadavia - R.Tilly
92
92
90
Concordia
92
92
88
Salta
92
92
83
Ciudad de Bs As
92
90
90
Viedma Carmen de
Patagones
92
92
88
Gran Rosario
93
92
90
Gran Santa Fé
93
93
90
Gran Mendoza
93
91
90
Corrientes
93
93
89
Gran Córdoba
93
92
90
Neuquén Plottier**
93
92
90
Río Gallegos
93
93
92
La Rioja
93
93
87
Gran San Juan
93
93
85
Mar del Plata - Batán
93
93
90
Gran Paraná
94
94
93
Bahía Blanca - Cerri
96
96
95
S.del Estero - La Banda
96
96
85
Río Cuarto
96
96
92
San Nicolás Villa
Constitución
96
95
93
Gran Catamarca
97
97
88
San Luis - El Chorrillo
97
97
93
Santa Rosa - Toay
97
97
96
Rawson Trelew
98
97
92
Total
91
90
86
47
Table B4: Explanatory Variables Basic Statistics
Mean
Mean Std.
Error
Min
Max
Lyfe Cycle and Households
Characteristics
age
50.521
0.00
14
99
Head of Household level of Education
complete_ primary
0.871
0.00
0
1
complete_secondary
0.426
0.00
0
1
complete_university
0.145
0.00
0
1
household_size2
3.337
0.00
1
23
qchild_under5
0.254
0.00
0
5
Wealth and Income Characteristics
income_household_total
1059.902
0.18
0
463100
income_per_capita
428.876
0.08
0
154367
head_income
750.536
0.15
0
460000
economic_dependency
0.560
0.00
0
1
employer
0.054
0.00
0
1
Employee
0.703
0.00
0
1
domestic_service
1.336
0.00
0
2
formal_work
0.665
0.00
0
1
Basic Needs and Social Vulnerability
Ubn
1.000
0.00
1
1
ubn_house_density
0.086
0.00
0
1
ubn_house
1.000
0.00
0
1
ubn_sanit
0.018
0.00
0
1
ubn_edu
0.004
0.00
0
1
ubn_survival
0.062
0.00
0
1
ubn_portion
0.231
0.00
0
1
sex
0.684
0.00
0
1
dummy_marital
0.622
0.00
0
1
international_migrant
0.082
0.00
0
1
domestic_migrant
0.233
0.00
0
1
Locational Variables
more500
0.791
0.00
0
1
dummy_gba
0.554
0.00
0
1
dummy_noa
0.081
0.00
0
1
dummy_nea
0.044
0.00
0
1
dummy_cuyo
0.059
0.00
0
1
dummy_pampa
0.237
0.00
0
1
dummy_patagonia
0.025
0.00
0
1
48
Table B5: Econometric Results. Marginal Effects for the formal/informal
Tenure Choice Model*
Dependent variable:
Formal
Tenure
Measure I
Formal
Tenure
Measure II
Formal
Tenure
Measure III
LPI_2
-0.139***
-0.169***
-0.106***
(0.040)
(0.039)
(0.037)
ZRAI_2
-0.051
-0.037
-0.030
(0.043)
(0.043)
(0.040)
BRI_2
0.066**
0.078***
0.035
(0.028)
(0.028)
(0.026)
IPI_2
0.032
0.039
-0.009
(0.027)
(0.027)
(0.025)
ALRI_2
0.019
0.022
0.008
(0.031)
(0.030)
(0.028)
ACI_2
-0.090***
-0.089***
-0.074***
(0.028)
(0.027)
(0.025)
Observations
7483
7483
7384
Predicted Probability
0.922
0.919
0.886
Log likelihood first iteration
-2304
-2379
-3023
Log likelihood second
iteration
-2150
-2204
-2656
Degrees of Freedom
22
22
22
chi2
308.3
350.1
734.9
Pseudo R2
0.0669
0.0736
0.122
*Only coefficients corresponding to regulation indicators are reported in
this table. The remaining control variables, although included in the
model, are not reported here.
Note: Robust Standard Error in parenthesis.
49
Table B6: Econometric Results. Marginal Effects for the Formal Tenure Model*
Dependent Variable: Percentage of
Population with formal tenure
(1)
(2)
(3)
(4)
(5)
LPI
-0.946**
-1.237**
-0.956**
-1.258***
-1.258***
(0.455)
(0.595)
(0.449)
(0.423)
(0.423)
ZRAI
-0.561*
-0.602**
-0.592**
-0.658
-0.658
(0.285)
(0.284)
(0.287)
(0.469)
(0.469)
BRI
-0.146
-0.115
-0.118
-0.107
-0.107
(0.290)
(0.323)
(0.285)
(0.333)
(0.333)
IPI
0.098
0.015
0.130
0.015
0.015
(0.280)
(0.257)
(0.312)
(0.355)
(0.355)
ALRI
0.511**
0.394
0.549**
0.394
0.394
(0.238)
(0.259)
(0.208)
(0.331)
(0.331)
ACI
0.480
0.227
0.483
0.190
0.190
(0.333)
(0.283)
(0.328)
(0.317)
(0.317)
propmen14
-0.269
-0.388**
-0.248
-0.384*
-0.384*
(0.178)
(0.147)
(0.184)
(0.224)
(0.224)
propmay65
-0.013
-0.103
-0.017
-0.104
-0.104
(0.165)
(0.167)
(0.173)
(0.312)
(0.312)
anios_educacion_prom
-1.204**
-0.126
-1.504***
-0.240
-0.240
(0.570)
(0.375)
(0.534)
(0.615)
(0.615)
pmigrantes
0.296***
0.295***
0.304***
0.299***
0.299***
(0.091)
(0.087)
(0.090)
(0.095)
(0.095)
dregion_NEA
-3.680***
-3.962**
-3.786***
-3.961***
-3.961***
(1.193)
(1.434)
(1.185)
(1.030)
(1.030)
propatleast1nbi
-0.261**
-0.295**
(0.112)
(0.108)
por_ipmh_sinreccorrientes
-0.017
(0.118)
por_ipmh_solopatrimonial
-0.017
(0.118)
ipmh_convergente
-0.000
(0.000)
totalpob
0.000
0.000
0.000
(0.000)
(0.000)
(0.000)
Constant
111.830***
104.583***
113.711***
105.364***
105.364***
(9.149)
(6.996)
(9.153)
(12.381)
(12.381)
Observations
62
62
62
62
62
R-squared
0.723
0.682
0.732
0.682
0.682
Adjusted R-squared
0.656
0.604
0.659
0.595
0.595
F
18.90
12.99
15.03
7.907
7.907
50
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
51
Table B7: Econometric Results. Marginal Effects for the Formal/informal
Tenure Choice Model*
Formal Tenure
Measure I
Formal Tenure
Measure II
Formal
Tenure
Measure III
(1)
(2)
(3)
age
-0.015
-0.013
-0.004
(0.014)
(0.013)
(0.012)
age2
0.000**
0.000**
0.000
(0.000)
(0.000)
(0.000)
complete_primary
0.214***
0.252***
0.368***
(0.074)
(0.072)
(0.065)
complete_secondary
0.168***
0.180***
0.259***
(0.052)
(0.051)
(0.048)
complete_university
0.176**
0.188**
0.146**
(0.077)
(0.077)
(0.074)
household_size2
-0.025*
-0.024*
-0.046***
(0.014)
(0.014)
(0.012)
qchild_under5
-0.119***
-0.112***
-0.123***
(0.039)
(0.038)
(0.036)
logincome_household
0.159***
0.159***
0.329***
(0.035)
(0.035)
(0.032)
economic_dependency
-0.073
-0.063
0.104
(0.092)
(0.091)
(0.084)
domestic_service
-0.024
-0.018
-0.039
(0.092)
(0.090)
(0.082)
formal_work
0.103*
0.140***
0.222***
(0.054)
(0.054)
(0.049)
sex
0.063
0.056
-0.110*
(0.071)
(0.071)
(0.066)
dummy_marital
-0.183***
-0.168**
-0.023
(0.068)
(0.067)
(0.061)
international_migrant
-0.009
-0.094
-0.110
(0.099)
(0.095)
(0.087)
domestic_migrant
0.071
0.045
0.025
(0.052)
(0.052)
(0.047)
Constant
0.070
-0.112
-1.935***
(0.387)
(0.382)
(0.349)
Observations
7483
7483
7384
Predicted Probability
0.922
0.919
0.886
Log likelihood first
iteration
-2304
-2379
-3023
Log likelihood second
-2150
-2204
-2656
52
iteration
Degrees of Freedom
22
22