Report No: ACS22437
.
Republic of Mozambique
Maputo Urban Poverty and Inclusive Growth
GREATER MAPUTO:
URBAN POVERTY AND INCLUSIVE GROWTH
.
June 2017
.
GSU13
AFRICA
.
.
Standard Disclaimer:
.
This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The
findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors
of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in
this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any
judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such
boundaries.
.
Copyright Statement:
.
The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may
be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages
dissemination of its work and will normally grant permission to reproduce portions of the work promptly.
For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright
Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470,
http://www.copyright.com/.
All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World
Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org.
ii
GREATER MAPUTO:
URBAN POVERTY AND INCLUSIVE GROWTH
June 2017
Mozambique – Country Management Unit
Africa Urban and Resilience Unit
Global Social, Urban, Rural and Resilience
iii
Copyrights
iv
Contents
Acknowledgements .......................................................................................................... vi
Index of Tables and Figures ........................................................................................... vii
List of Acronyms .............................................................................................................. ix
Executive Summary .......................................................................................................... 1
I. Introduction ............................................................................................................... 2
II. Putting the Urban Poor on the Map ........................................................................ 4
Poverty Mapping Methodology ...................................................................................... 4
Poverty Headcount Mapping........................................................................................... 5
Poverty Intensity and Gap Analyses ............................................................................... 6
Robustness analysis ......................................................................................................... 8
III. Access to Basic Urban Services ................................................................................ 9
Location and Density Matter ........................................................................................... 9
Correlation between Poverty and Access ...................................................................... 12
Exposure to Natural Risks ............................................................................................. 15
IV. Access to Affordable Land and Housing ............................................................... 17
Land and Housing Market Assessment ......................................................................... 17
An Increasingly Unaffordable Housing Market ............................................................ 18
V. Leveraging Land Based Infrastructure Financing ............................................... 23
Unmet Revenue from Land Based Financing ............................................................... 23
Options for Increasing Land Based Financing .............................................................. 26
VI. Recommendations .................................................................................................... 30
At National Level .......................................................................................................... 30
At Municipal Level ....................................................................................................... 31
References ........................................................................................................................ 34
End Notes ......................................................................................................................... 38
v
Acknowledgements
This study was prepared by World Bank team led by André Herzog (Senior Urban Specialist –
Task Team Leader), and included Somik Lall (Lead Urban Economist), Javier Baez (Senior
Economist), Pedro Olinto (Senior Economist), Kenneth Simler (Senior Economist), Shohei
Nakamura (Economist), Bontje Marie Zangerling (Urban Specialist), Hannah Kim (Young
Professional), Dany Jones (GIS Expert), Benjamin P. Stewart (Geographer), Henry Cherkezian
(Land and Housing Expert/Consultant), Lycia Lima (Poverty Economist/Consultant), Bernardo
Weaver Barros (Public Finance Expert/Consultant), Tito Yepes (Urban Economist/Consultant),
Julia Oberreiter (Institutional Development Expert/Consultant), and Adrienne Acioly (Urban
Expert/Consultant).
The concept note and the draft report was significantly enriched thanks to peer review comments
provided by Victor Vergara (Lead Urban Specialist); Uri Raich (Senior Urban Specialist); and
Vasco Molini (Senior Economist).
The report was prepared under the overall guidance of Sameh Wahba (Director for Urban and
Territorial Development) and Bernice Von Bronkhorst (Practice Manager for Africa Urban and
DRM). The team also received valuable guidance from Mark Lundell (Country Director for
Mozambique, Comoros, Madagascar, Mauritius e Seychelles) and Mark Austin (Program Leader).
The Greater Maputo Urban Poverty and Inclusive Growth benefited immensely from the
collaboration with representatives from the Maputo Municipal Council, Matola Municipal
Council, and the Spatial Development Planning Unit under the Ministry of Transport and
Communication. The team is also thankful to the development partners of the Decentralization
Partners Coordination Group that provided inputs during the preparation of the study.
The final report was divided into two volumes: Volume I - Main Report; and Volume II –
Background Papers. The Main Report (Volume 1) was written by André Herzog, Hannah Kim,
and Adrienne Acioly. Volume II comprises of five background papers and will be published
separately. Lycia Lima, with inputs from Javier Baez, Pedro Olinto, Kenneth Simler prepared the
Poverty Mapping paper. Shohei Nakamura with inputs from Bontje Marie Zangerling, Tito Yepes,
and Julia Oberreiter prepared the ‘Spatial Patterns of Housing, Infrastructure, and Poverty’ paper.
Bernardo Weaver Barros with inputs from Remígio Chilaule, Shohei Nakamura and Henry
Cherkezian prepared the ‘Land Value Analysis’ and ‘Estimating Property Tax Potential’ papers.
Henry Cherkezian with inputs from Julia Oberreiter prepared the ‘Assessing Land Value Capture
Instruments’ paper. The team is thankful to Bruce Ross-Larson and Mike Crumplar from
Communication Development, who edited the report.
vi
Index of Tables and Figures
Table 1. Mozambique’s System of Cities
Table 2. Poverty rates explained by access to infrastructure services and density, controlling for
location
Table 3. Examples of Land Based Finance for Local Governments
Figure 1. Changes in poverty headcount in Maputo and Matola municipalities (1997-2007)
Figure 2. Poverty headcount, Maputo City and Matola (1997 and 2007)
Figure 3: Poverty Gap Results for Maputo City and Matola (1997-2007)
Figure 4: Poverty Intensity Results for Maputo City and Matola (1997-2007)
Figure 5. Headcount results for Maputo City and Matola with 95 percent confidence intervals
Figure 6. Districts and distance from the CBD
Figure 7. Access to amenities by poverty and housing status in the GMA (2008/9)
Figure 8. Changes in access to infrastructure in Maputo by distance (1997, 2007, and 2013)
Figure 9. Density and infrastructure access in the GMA, 2007
Figure 10. Locations of educational facilities and poverty rates
Figure 11. Health facilities and poverty rates (2007)
Figure 12. Households per health facility within 2 km (2007)
Figure 13. Flood-prone areas and poverty in the Greater Maputo Area (2007)
Figure 14. Where can a median income family buy nondurable housing
Figure 15. Average home value per m2, 2016
Figure 16. Share of leapfrog development in African cities
Figure 17. Density by distance from the CBD
Figure 18. Poverty rate and distance to the CBD in the GMA, 1997 and 2007
Figure 19. Growth in tax revenue by main source in Maputo (2010–15)
Figure 20. Growth in tax revenue by main source in Matola (2010–15)
Figure 21. Policy options to increase IPRA collection in Maputo
Figure 22. Maputo and Matola IPRA baseline revenue forecast (2017–2021)
vii
Figure 23. Maputo construction costs according to market estimates (2016)
Figure 24. Matola construction cost according to market estimates (2016)
Figure 25. Comparing revenue increases from various IPRA collection policy decisions
viii
List of Acronyms
GMA Greater Maputo Area
IAF Inquérito aos Agregados Familiares IAF
INE Instituto Nacional de Estatísticas
IOF Inquérito sobre Orçamento Familiar
IPRA Imposto Predial Autárquico
IASISA Imposto Autárquico de SISA
MAEFP Ministério da Administração Estatal e Função Publica
MEF Ministério de Economia e Finanças
PDE Programa de Desenvolvimento Espacial
TAE Taxa por Atividade Econômica
ix
Executive Summary
The overall objective of the Maputo Urban Poverty and Inclusive Growth Study is to inform the
municipal governments of Maputo and Matola in the Greater Maputo Area (GMA) on how to
better target and finance urban poverty reduction programs by building critical knowledge on
urban poverty and land-based municipal revenue enhancement. The study carried out an in-depth
poverty and vulnerability mapping and analysis in order to better understand the spatial distribution
of poverty in the GMA. It showed that poverty has significantly reduced in most of the
neighborhoods in GMA from 1997 to 2007. The study also showed a strong correlation between
poverty reduction and access to basic services and urban infrastructure, though it was not possible
to identify a correlation between poverty incidence of natural hazards based on availability data.
The study also carried out for the first time in the GMA a comprehensive housing market
assessment in order to collect primary data on real estate market value, which was used as basis
for estimating the net present value of different policy options to increase property tax. It suggested
that municipal revenue for property taxation could significantly increase if municipalities would
improve tax compliance, increase tax rate from actual 0.4 to 1.0 percentage, and use market value
to assess property tax. In this case, Maputo City Council would generate Mt 44 billion and Matola
Mt 29 billion in NPV terms. Finally, study provided specific policy and institutional development
recommendations at national and municipal levels with the ultimate objective to increased urban
infrastructure finance, and better target pro-poor interventions not only in GMA, but also
benefiting other cities in Mozambique.
1
I. Introduction
As the prime urban agglomeration in Mozambique, the Greater Maputo Area presents key
challenges and opportunities for the country’s goals in promoting inclusive economic growth.
Despite strong economic growth over the past two decades, Mozambique continues to be one of
the poorest countries in the world. GDP growth reached an average of 8.4 percent between 1993
and 2013, contributing a reduction in poverty rate from 69.4 percent in 1996 to 54.7 percent in
2009. However, more than half the population is still below poor line, placing Mozambique in
the13th place among the poorest country in the world in 2013. High levels of underemployment
and poor human development indicators suggest that economic growth has not been inclusive.
While the majority of the population is economically engaged, there are high levels of
underemployment as most people live from subsistence agriculture and informal urban activities.
In fact, when one considers unpaid family workers as unemployed, the unemployment rates jump
to almost 40 percent in 2009. Human development levels are also low, as indicated by
Mozambique ranking 178 out of 187 countries in the UN’s Human Development Index. Moreover,
progress has been geographically uneven, with the poorest concentrated in the rural areas and in
the Central and Northern regions.
Cities are performing better than rural areas in terms of poverty reduction. Indeed, poverty
levels declined faster in urban areas than rural areas. Between 1996/7 and 2008/9, poverty declined
by 24 percentage points in urban areas, compared to 10 percentage points in rural areas 1. As in
most other countries, urbanization and economic growth seem to go hand in hand. However, urban
poverty remains high at 49 percent.
Most of the urban growth has happened in the Greater Maputo Area. With about 32 percent
of the total urban population, the Greater Maputo Area (GMA) is Mozambique’s largest urban
agglomeration. GMA reached 1.7 million inhabitants in 2007, distributed in two municipalities:
Municipal Councils of Maputo (1.09 million inhabitants), and Matola Municipal Council (675
thousand inhabitants). The municipality of Maputo houses most of the population of the GMA.
But Matola is growing four times faster than Maputo and is quickly becoming a dormitory city for
the new urban poor, and the middle class that increasingly cannot afford to live in good locations
close to Maputo city center (see table 1).
Table 1. Urban system of cities
1997 2007 Average
City size growth
Pop % Pop % 1997-2007
GMA 1,396,700 34.5 1,716,632 32.0 2.1
Maputo 966,000 23.8 1,068,607 19.9 1.0
Matola 430,700 10.6 648,025 12.1 4.2
Intermediate Cities 1,840,348 45.4 2,350,395 43.8 2.5
Small 815,226 20.1 1,294,792 24.1 4.7
Total 4,052,274 100.0 5,361,819 100.0 2.8
Source: INE Census 1997 and 2007. Note: Small intermediate cities have population
between 500 thousand and 1 million. Small cities are municipalities with less than 500
thousand inhabitant. The smallest municipality hosted 12,564 inhabitants in 2007.
2
Most of the population in the Greater Maputo Area live in underserved neighborhoods.
These areas have only basic infrastructure, roads are not paved, sanitation network has not reached,
drainage systems are inexistent, and most houses are built with nondurable material. Incidence of
airborne diseases such as diarrhea and malaria are more common in underserved neighborhood
due to recurrent flooding and poor sanitation conditions. Most people in these areas work in the
informal economy, and those that have a job end up spending two to three hours to community to
the center of Maputo where most formal jobs are located. Therefore, full urbanization of these
areas can significantly improve the lives millions living in these informal neighbourhoods of Greater
Maputo, reducing their vulnerability to commonly preventable diseases and to the exposure to natural
hazards that often destroy their income and assets, and increasing access to economic opportunities.
However, improving urban infrastructure and services requires both finance to extend the coverage and
quality of urban services and infrastructure to the whole GMA, as well as good targeting of urban reduction
interventions.
Taking this into consideration, this study aims at mapping and analyzing urban poverty in
GMA while unveiling the potential to increase urban infrastructure finance through land-
based revenue instruments. Using available population census and household welfare surveys,
the study mapped out the incidence of poverty in GMA in order to better understand the spatial distribution
of poverty, its trends, and its correlation with availability of infrastructure, access to basic services, quality
of housing, as well as the incidence of natural hazards. In addition, the study undertook primary data
collection on the value of land and housing in GMA in order to estimate the potential of increasing property
tax through different policy and institutional reforms. Analyzing both urban poverty and the potential for
land-based infrastructure finance, this study will help Maputo and Matola municipalities to better target and
finance urban poverty reduction programs and interventions. The study will be also relevant to the ongoing
decentralization and national urban policy debate2 led by the Ministry of State Administration and Public
Function (MAEFP) and the Ministry of Economy and Finance (MEF) by highlighting the importance of
land-based finance in contributing to national goals to accelerate poverty reduction and economic growth.
3
II. Putting the Urban Poor on the Map
Poverty Mapping Methodology
In order to understand the spatial patterns and the trends of poverty over time within the
Greater Maputo Area, two poverty-mapping exercises were produced with the most recent
statistics date available. This study created two poverty maps using most recent available data
including the household welfare surveys of 1996/97 (Inquérito aos Agregados Familiares IAF),
and 2008/09 (Inquérito sobre Orçamento Familiar - IOF), and the national population census of
1997 and 2007. The poverty maps follow Elbers, Lanjouw & Lanjouw (2003) methodology that
combines household’s welfare data with census population data that is more desegregated in order
to create a measurement of poverty at the smallest geographic level that have statistic relevance.
While household survey data is often richer as it collects information on a wide range of variables,
the data is obtained from interviews of a sample of the population, which limits its statistical
representativeness, for instance of poverty rates, at high levels of disaggregation. In contrast,
census data includes all the households but due to its large geographic coverage, it gathers a more
limited amount of information. In a nutshell, the household survey data is used to estimate the
parameters of a set of explanatory variables (also available in the census) used to model household
consumption for the area under study (country, province, etc.). These estimates are used to impute
consumption in the census and thus poverty indicators at geographic units. Because the last
population census was carried out in 2007, it was not possible to produce a more up-to-data map
of poverty in GMA 3.
In order to achieve reasonable precision in the estimates through the small area estimation
methodology, geographic units must be large enough - minimum of 10.000 households. If
units are too small, confidence intervals may be too wide, producing not very informative poverty
estimates. In the construction of such geographic units, three basic criteria were established: i)
geographic adjacency, ii) a minimum of 10 thousand households in each unit according to the 2007
Mozambican national census and iii) all bairros should be grouped within the same urban
administrative division 4. These criteria are important for reaching more homogenous geographic
units and, more importantly, for creating units which respect the existing urban district,
administrative postos and bairro boundaries. Maputo municipality is divided into 7 districts: five
contiguous and two isolated 5. Since the two isolated units together account for approximately
5,000 households and it is not possible to contiguously combine them, they are each used as a
single unit. Using the five contiguous districts, there are 55 bairros, which are combined into 17
geographic units. Matola municipality encompasses three administrative postos in which there are
41 bairros grouped into 11 geographic units.
4
Poverty Headcount Mapping
The Greater Maputo Area has been one of the drivers of poverty reduction in Mozambique.
The results of the poverty mapping exercises show that while national poverty levels have seen a
modest decline and geographically uneven poverty reduction in spite of strong economic growth, 6
poverty has reduced faster in the GMA. Maputo Municipality reduced poverty from 37 percent to
10 percent between 1996/7 and 2008/9. The poverty maps produced in this study (figure 1) show
that all geographic units within Maputo Municipality reduced poverty between 1997 and 2007. In
Matola, half of its geographic units managed to reduce poverty, though at a more modest rate than
in Maputo Municipality.
Figure 1: Changes in poverty headcount in Maputo and Matola municipalities (1997-2007)
Source: INE IAF 1996/7, IOF 2008/9, Population Census 1997 and 2007
Poverty reduction is not uniform in the GMA. Analyzing the spatial distribution of poverty, it
is evident that prevalence of poverty is higher in Matola for both years considered, although the
difference in poverty incidence between Maputo and Matola becomes much higher in 2007 (figure
2). This is due to the fact that the poverty numbers in Matola did not change much over time, while
Maputo experienced a sharp reduction in poverty in all units between 1997 and 2007. Although
exhibiting overall lower poverty than Matola, Maputo also shows in both years a high prevalence
of poverty, with the exception of two units 7, which already in 1997 exhibited relatively low poverty
rates and registered in 2007 rates as low as 3 and 5% in poverty headcount. These units comprise
the higher income Maputo neighborhoods of Polana Cimento and Sommershield, as well as the
neighborhoods Central, Alto Mae, Coop and Malhangalene.
Maputo Municipality has outperformance Matola in terms of poverty reduction. The analysis
shows that in fact all 17 geographic units in Maputo reduced poverty between 1997 and 2007,
while the results in Matola are less positive since only nearly half of its geographic units managed
to reduce poverty, and at a much more modest rate than the geographic units defined for Maputo.
All the five geographic units that registered an increase in terms of poverty headcount are located
in Matola, including one that recorded an increase in the poverty headcount of 5%. When looking
at the point estimates, the results show a decline in poverty rates in 23 out of the 28 geographic
5
areas between 1997 and 2007, with 14 of those 23 areas registering a decline greater than 15 p.p.
These 14 areas are all located in Maputo, out of which 4 particularly stand out by having reduced
poverty by more than 27 p.p. These four areas that displayed these extraordinary results in Maputo
include the neighborhoods of Zimpeto, Magoanine, Malhazine, Mahotas, Costal do Sol, Albasine,
Ferroviario and Laulane.
Figure 2: Poverty headcount, Maputo City and Matola (1997-2007)
Source: INE IAF 1996/7, IOF 2008/9, Population Census 1997 and 2007
Poverty Intensity and Gap Analyses
Poverty intensity has also decreased in most of the Greater Maputo Area. To measure
intensity of poverty, two indicators were used. One is the poverty gap (figure 3), which estimates
the depth of poverty as measured by how far, on average, the poor are from the poverty line. The
second is the poverty severity gap (figure 4), which measures inequality within the poor, and gives
higher weights to the poor the farther away from the poverty line they are. In 20 of the 28
geographic units, the intensity of poverty decreased. Among the 8 geographic units that saw an
increase in poverty, the highest increase was 4 percent in the poverty gap and 3 percent in poverty
severity. 8 Both the poverty gap and poverty severity indices are better measures than the headcount
ratio, because they measure depth of poverty and inequality within the poor instead of simply
counting the number of people below the poverty line and considering them equally poor.
6
Figure 3: Poverty Gap Results for Maputo City and Matola (1997-2007)
Source: INE IAF 1996/7, IOF 2008/9, Population Census 1997 and 2007
Figure 4: Poverty Intensity Results for Maputo City and Matola (1997-2007)
Source: INE IAF 1996/7, IOF 2008/9, Population Census 1997 and 2007
The spatial patterns of poverty in the Greater Maputo Area show that poverty rates are
higher the farther people are from the central business district. While the nationalization of
land gave the opportunity to many poor families to build the houses close to city center, since then
market forces have pushed the poor to outer areas and brought better-off households to move in.
7
Poverty rates used to be much higher in the neighborhoods immediately surrounding the city center
located in Maputo Municipality. But as the poor households sold their properties and moved to
outer areas looking for affordable housing, particularly in Matola, poverty rates decreased closer
to Maputo central district and increased in some bairros of Matola. Rural-urban migration also
concentrated in outer areas where land is much more affordable. Between 1997 and 2007, the
number of nonpoor households increased in every distance range, whereas the number of poor
households increased mainly in the areas beyond 10 km from the CBD.
Robustness analysis
The above comparisons have to be interpreted with caution. Since the IAF and IOF surveys
are representative of the Maputo City Province, it is possible to estimate the poverty rates for this
geographic level with the survey dataset. Using the poverty mapping procedure, we can also
estimate poverty rates with the census data and compare them with the survey outputs as a
robustness check. Such analysis was carried out for both the 1997 and 2007 exercises using the
poverty headcount and poverty gap for the comparisons with 95 percent confidence intervals. The
results show also that, since the confidence intervals for survey and census calculations intersect,
it is not possible to reject the hypothesis that these estimations are not statistically distinguishable
from each other. It is also important to take into consideration that, given that the poverty rates
were imputed on fairly small geographic units (around 10,000 households), the standard errors and
the confidence intervals are relatively large. Therefore, even if the estimated points indicate a
decrease in the headcount ratios, the actual variation of the poverty numbers depend on a more
careful analysis of the confidence intervals.
Figure 5: Headcount results for Maputo City and Matola with 95 percent confidence intervals
1997 2007
Source: INE IAF 1996/7, IOF 2008/9, Population Census 1997 and 2007
8
III. Access to Basic Urban Services
The study analyzed the correlation of poverty with the availability of urban infrastructure,
access to basic services, and incidence of natural hazards in GMA. Building upon the poverty
mapping, the analysis focused on 96 neighborhoods (or bairros in Portuguese) in the five
administrative districts of the Maputo Municipal Council, and in the three administrative posts in
Matola Municipal Council (Figure 6). Most residents of Greater Maputo Area live in bairros within
25 kilometers from the central business district (CBD), which for the purpose of the analysis are
defined as the bairro Central C, and it is located in District 1. A key variable underpinning the
spatial analysis is the Euclidian distance between the centroid of each bairro and that of Central C.
Figure 6. Districts and distance from the CBD
Source: authors
Location and Density Matter
Residents living in informality face greater challenges in accessing basic urban
infrastructure and services. Indeed, the distance or travel time to various basic services differs
depending on poverty status (poor or nonpoor) and housing materials (durable or nondurable) in
2008/9 (Figure 7). Among the poor, mean values of distance or travel time to most amenities are
higher for households in nondurable housing than in durable housing. For instance, compared with
poor households in durable housing, poor households in nondurable housing spend about 7 more
minutes to get to a water source; 8 more minutes to get to the nearest bus stop; 10 more minutes
to get to the nearest health facility; and 11 more minutes to the nearest police station. Similarly,
among nonpoor households, those in durable housing enjoy greater access than other households
9
in nondurable housing. Among households in nondurable housing, poor households spend more
time accessing amenities than nonpoor households.
Figure 7. Access to amenities by poverty and housing status in the GMA (2008/9)
Source: Authors’ calculations based on the INE IOF 2008/9
Note: ND refers to nondurable housing, D to durable housing.
Although access to key urban services improved in recent years, there is a major gap in
access for those living in outer areas. As illustrated in Figure 8, access to piped water and
electricity improved considerably from 1997 to 2013. However, access to sanitation – particularly
a septic tank connection – has only improved marginally. Across all these services, there are
significant disparities in access based on distance from the central business district. Maputo is
divided into five districts, with the central business district located in district 1. Access to piped
water inside housing is limited mainly to district 1. Outside district 1, only 10 percent (or less) of
households have access to piped water inside their dwelling. A majority of residents instead rely
on water for drinking either from piped water outside housing or standpipes, private or public.
Residents of Matola also have very limited access to piped water inside their housing, though those
10
who live within 15 km of the CBD in Matola have better access to piped water outside the home
than Maputo residents. As distance to the CBD increases, the main use of a sanitation facility
changes from toilets connected to a septic tank to unimproved facilities. 9 For electricity, virtually
no one had access in Matola beyond 20 km (at least in 2007). 10
Figure 8. Changes in access to infrastructure in Maputo by distance, 1997, 2007, and 2013
Source: Authors’ calculations based on the 1997 and 2007 censuses, and
World Bank/WSP Household Survey 2013 11..
Note: Distance is measured from bairro Central C in district 1.
Smooth lines are based on a locally weighted regression.
Moreover, residents living in more consolidated areas tend to enjoy better access to
infrastructure. As shown in figure 9, the bairros in Maputo within 3 km of the CBD enjoy high
access rates regardless of density. Bairros more than 10 km away tend to have low density and a
lower share of households with access to piped water and other services. The other bairros (3–10
km from the CBD) show no clear pattern between density and infrastructure access. By contrast,
density seems to be positively correlated with the share of households with access to infrastructure.
Higher density could be correlated with better access to services because it may be physically
easier and less expensive for government to provide infrastructure and services to residents in more
consolidated places than in sparsely inhabited areas. People could also move to areas with better
access to services, which can make the area even denser.
11
Figure 9. Density and infrastructure access in the GMA, 2007
Source: Authors’ calculations based on the 2007 census
Note: Distance is measured from bairro Central C in district 1.
Correlation between Poverty and Access
More importantly, there is a positive correlation between access to basic services and poverty
reduction. Results from a series of regression models based on bairros in 1997 and 2007 point to
the link between improvements in infrastructure access and the reduction in poverty rates across
bairros. 12A 10 percentage-point increase in the share of households with access to piped water
was associated with a 2.7 percentage point reduction in poverty rates; standpipes, 3.5 points; septic
tanks, 5.9 points; improved toilets, 6.0 points; and electricity, 7.3 points. In addition, a change in
density was also positively correlated with changes in the share of households with infrastructure
access. 13
However, access to infrastructure services does not alone explain poverty. Distance from the
central business district and density are also important considerations. For every level of access to
services, there are multiple levels of poverty, except for the central business district, where poverty
is very low and access to services is very high. A more complete picture to correlate poverty with
access to services should include location and density. A series of regression models looked at the
extent to which poverty rates are explained by access to infrastructure services and density,
controlling for location, by including bairro fixed effects, with a constant for each bairro to help
segregate variance by location. Three services – access to a septic tank, improved toilets, and
electricity – observe significant correlations controlling for density in a nonlinear form. Low
access to services provision is correlated with high poverty in periurban areas that are farther away
from the central business district. But after controlling for distance and density, services are
significantly correlated with poverty.
12
Table 2. Poverty rates explained by access to infrastructure services and density, controlling for
location
(1) (2) (3) (4) (5) (6) (7)
Household density 22.27** 16.95** 19.74** 10.51 18.82*** 14.36** 23.06***
(8.321) (7.374) (7.380) (8.277) (6.660) (6.749) (6.151)
Density squared -3.786* -2.557 -3.389* -1.287 1.161 1.220 -2.588*
(1.998) (1.731) (1.764) (1.881) (1.751) (1.662) (1.428)
Piped water (%) -0.440** -0.267
(0.182) (0.172)
Standpipes (%) -0.426*** -0.354**
(0.123) (0.128)
Septic tank (%) -1.012*** -0.587*
(0.359) (0.312)
Improved toilets (%) -0.712*** -0.598***
(0.159) (0.163)
Electricity access (%) -0.730***
(0.137)
Bairro FE Yes Yes Yes Yes Yes Yes Yes
Year FE Yes Yes Yes Yes Yes Yes Yes
R2 0.703 0.756 0.780 0.724 0.801 0.829 0.834
Obs. 54 54 54 54 54 54 54
Source: Authors’ calculations based on the 2007 census.
Note: Dependent variables are poverty rates (%). Standard errors in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01.
Constant terms are not shown.
While poor areas in Maputo and Matola benefited from greater access to primary school
facilities, progress has been slow for secondary schools. Primary school facilities expanded
from 1997 to 2007 in areas with high poverty rates. Data from the World Inequality Database on
Education (WIDE) indicate that in the GMA, 27 percent of children of primary school age did not
attend school in 1997, but only 3 percent in 2007/08. 14 Figure 10 confirms this general trend and
shows that areas underserved in 1997—in particular the bairros George Dimitrov (24), Inhagoia,
Bagamoyo (23), Hulene (18, 20), Chamanculo (13, 14), Luis Cabral and Jardim (22)—saw a
greater number of primary school facilities. 15 At secondary level, however, education
infrastructure expanded on a much smaller scale. While Maputo and Matola cities benefited from
an overall upgrading of secondary school facilities, some high-poverty areas remain almost
unchanged. These include the western bairros of Tsalala, Mahlampswen, Ndlavela (9, 10); and
northern bairros of Matola such as Mukatine, Mali, Muhalaze (6); but also the eastern bairros of
Maputo such as Davide, Albazine (25), Triunfo, and Costa do Sol area (21) as well as Katembe
area (27).
More worrisome, education outcomes continue to be low for the urban poor. Graduation rates
stayed low through 2007/08—62 percent from primary school and a mere 10 percent from
secondary school. 16 Both maps confirm—within the areas with high poverty counts—the
13
correlation between the poor children and those the farthest from primary and secondary facilities.
The cost of school supplies is another burden for poor families, nationally. 17 Around 29 percent of
the country’s poorest children cannot read or write, and are almost ten times more likely not to set
foot in a classroom than rich children (3 percent). 18 On gaps in mean school years, young
Mozambicans in the GMA ages 20–24 spent 4.3 years in formal education in 1997 but 6.1 years
in 2008. Being poor exacerbates inequalities, with the poorest children spending less than 2.33
years in school.
Figure 10. Locations of educational facilities and poverty rates
Source: World Bank poverty maps with overlay of primary and secondary education georeferenced data from the
Programa de Desenvolvimento Espacial (PDE) 19.
Note: The figure shows an overlay of the 1997 and 2007 poverty headcount ratio, with private and public education
facilities for primary (blue) and secondary (green) education.
Access to health services remains inadequate, particularly for the poor, as they need to travel
long distances to reach the nearest health facility. No georeferenced data on health care facilities
could be found for 1997, so figure 11 shows data from 2007. In areas of high poverty in Matola—
particularly western bairros such as Tsalala, Mahlampswen, Ndlavela (9, 10); northern bairros of
Matola like Mukatine, Mali, Muhalaze (6), and Matola city bairros including Matola A-J and
Trevo (0, 1)—distance to the nearest health facility is a common problem, and the poorest are less
likely than the rich to seek medical attention. 20 bairros (21 percent) have no health facility within
2 km of the center; 37 bairros (39 percent) have only one health facility; and 19 bairros
(20 percent) have two. The number of poor households with no access to a health facility within 2
14
km is three times as high as for nonpoor households (figure 12). This may explain why traditional
medicine is still commonplace at the bairro level. A Government report of 2012estimates that
more than 70 percent of Mozambicans rely on it; for many, it is the only health service available
(with a ratio of 1 healer to 200 people). 20 In 2007, Mozambique had a density of just 0.0026
medical doctors (physicians) per 1,000 people and is classified as one of the countries with a
critical shortage of human resources for health. 21 The World Health Report ranked Mozambique’s
overall health system performance in 1997 as 184 of 191 countries. 22
Figure 11. Health facilities and poverty rates Figure 12. Households per health facility
(2007) within 2 km (2007)
Source: World Bank Poverty Maps with overlay of Source: Authors’ calculation based on the 2007 census
health facilities georeferenced data from the Programa and Programa de Desenvolvimento Espacial.
de Desenvolvimento Espacial (PDE).
Note: The height of a bar indicates the number of
households. X-axis indicates the number of health
facilities within 2 km from the center of each bairro.
Exposure to Natural Risks
Access to basic infrastructure such as paving and drainage in underserved informal
settlements is also likely to further increase the exposure to floods in the Greater Maputo
Area. Although the GMA is prone to recurrent flooding and erosion, existing data 23 do not indicate
that the poor are disproportionately more affected by it. Figure 13 shows that no significant
correlations were found between level of poverty and share of area exposed to flooding. While it
was not possible to correlate the current incidence of flooding with poverty levels, the continuous
informal urbanization of peri-urban areas will likely increase the exposure of the urban poor to
15
flooding and erosion. As discussed, most of the urban infrastructure network is concentrated in the
core areas (cidade cimento), while the informal settlements that continue to growth in peri-urban
areas lack the most basic infrastructure such as paving and drainage. Moreover, the peri-urban
areas are mostly flat which makes particularly difficult for rainwater to dissipate.
Figure 13. Flood-prone areas and poverty in the Greater Maputo Area (2007)
Source: World Bank map based on data from the 2007 census, and 2008 household survey, combined with the
Sanitation and Drainage Master Plan for Greater Maputo/AIAS.
Note: The areas in gray-scale are not covered by the analysis. The analysis combines available data on flood-prone
areas with the spatial distribution of poverty. The flood-prone areas identified are scattered across the entire GMA,
reflecting both the more homogenous topography and the lack of drainage infrastructure in most neighborhoods.
16
IV. Access to Affordable Land and Housing
Land and Housing Market Assessment
Housing markets play a very important role for cities, both in terms of access to affordable
shelter, and as a revenue potential through land value capture. This chapter drills into
secondary and primary data to analyze, for the first time, the land and housing market of the GMA.
The objective is to better understand an important real estate market segment not formally
recognized by legal nor institutional frameworks. Since land is legally considered as a public asset,
it cannot be commercialized nor taxed based on market value. Ultimately, this analysis may set
forth policy and institutional reforms which can enhance municipalities’ ability to capture urban
land value to finance much needed urban infrastructure. Additionally, overlapping housing price
data with poverty map may bring light to unofficial housing market and its affordability aspects.
These aspects may affect particularly shelter options for the urban poor.
The study collected data on housing market by location and type of construction material in
the GMA. Following Dowall’s land value assessment methodology (Dowall 1991), the team
conducted interviews with local formal and informal real estate brokers knowledge of real estate
transactions in the different neighborhood. Through this interviews, it was possible to classified
housing prices by different categories: (i) small, medium, or large, (ii) durable and non-durable
material homes, and (iii) location. These were the five steps taken to collect and analyze GMA
housing data. For each data point, Dowall recommends to gather three interviews. But he also
explains that this would create an unnecessary repetition burden turning the whole exercise
inviable. Hence the solution is to extrapolate the original sample. One interviewed expert with
experience beyond his original area provided information about other areas. That reduced the
number of interviews needed for the survey. Another important methodological aspect is the notion
that the team needs to gather the median values, not the average. So, instead of adding three
answers about the same question and dividing the result by 3, the team has to discard the highest
and the lowest and consider only the median value. That way, it is possible to avoid outliers
impacting the final value.
Besides primary data collection, the team has also based some of its efforts on secondary data
as well. Secondary data sources include: the 2007 population census, and real estate market prices
available on 2012 government surveys. These figures help understand market prices evolution
through time. Having historical series data boost projections’ accuracy for future estimates on real
estate prices. Finally, it is important to note that the data was collected during the recent GDP
slowdown driven by the end of the commodity super cycle that in the past fueled real estate prices
in GMA and concerns with debt sustainability. Anecdotal evidences show that the housing market
has cooled down in the GMA, and prices deflated. But since economic growth is expected to
bounce back in the coming years, and the constraints in the supply side of the land and housing
delivery systems have not been addressed, the findings of these chapter remains valid.
17
An Increasingly Unaffordable Housing Market
Housing has become increasingly unaffordable in Maputo. Housing, even in informal areas, is
particularly unaffordable in good location in Maputo . The housing market in GMA has two
submarkets, one for durable housing and one for nondurable housing. Durable dwellings have
walls made of either cement or bricks and a roof made of either cement or tiles. Durable housing
costs between Mt 10,000 and Mt 145,000 per m2 while nondurable housing costs between
Mt 3,200 to Mt 41,300 per m2. At an income-to-price ratio of 3, a large part of the nondurable
housing market in Maputo therefore becomes unaffordable, including a one-bedroom unit (see
figure 14). At a ratio of 5, one-bedroom unit made of nondurable material is within the reach of
households with median income. As a result, an increasing proportion of nonpoor households live
in nondurable housing, farther away from the central business district. Assuming that people can
afford to buy a house that costs less than three times their annual income—or five times at most—
the affordability lines for a median-income family are Mt 309,960 and Mt 516,600, respectively.
Figure 14 Where can a median income family buy nondurable housing?
Source: Primary data collection of housing prices 2016
Note: Bairros in yellow indicate where average price of nondurable housing is within five (three) times of median
family annual income in the GMA
18
Distance from the city center is a mayor price driver, but land values are also an important
factor. Housing is most expensive in district 1 and 2, but new real estate development is
gentrifying district 3 and also increasing home prices while other residents are moving to districts
4 and 5 in search of bigger plots and more affordable housing (see figure 15). The demand for
housing is growing in all segments of the market due to overall population growth, smaller family
structures, but also larger flows of money into real estate from the main players in the natural
resource industry. As supply of land and housing has not accompanied growth in demand due to
policy and intuitional bottlenecks, a result, prices have skyrocket in GMA.
Figure 15 Average home value per m2 (2016)
District 8 MT
19,217
District 5
District 4
District 9 MT 16,156
MT 21,343
MT 5,903
District 3
District 7 District 2
MT 11,236
MT 24,186 MT 18,203
District 1
MT 65,859
Source: Source: Primary data collection of housing prices 2016
GMA has one of the most inefficient land use in the continent, According to a recent study
(Baruch 2015) that used night light data to compare African cities, Maputo has the second highest
percentage of leapfrog development. In other words, 50% of new development is taking place in
low density areas, disconnected from existing consolidated urban areas (figure 16). There has been
19
also a sharp acceleration of leapfrog development in the last decade, from 25% during the period
of 1990-2000 to 50% in 2000-2010 period. This coincides with the acceleration of urban
population growth as well as the boom in the housing market fueled by the commodity super cycle.
As a result, the GMA is characterized by “islands” of high density surrounded by a low density
urban fabric. Household density declines farther from the CBD, as plot area per house increases
with distance. While the CBD has average densities, with up to 4,000 households per km2 within
3-4 km, density decreases significantly at 10-15 km from the CBD, until it meets the CBD of
Matola (figure 17).
Figure 16 Share of leapfrog development in African Figure 17. Density by distance from the CBD
cities
Source: World Bank staff estimates based on the 2007
Source: Baruah (2015) census
Leapfrog development undermines economies of scale and makes the provision of basic
services more difficult and expensive. Several factors underline the problem with inefficient land
use in GMA. First, and foremost, land was nationalized after Independence, but since then people
have been trading the land without a proper legal and institutional framework that recognize land
market forces and ensure transparency and accountability. The current flaws in the land policies
created a highly dysfunctional informal land market that inhibit land to be traded more
competitively in an open market. Secondly, property tax as the study analyzed is very low, and not
charged for vacant land. If property taxes would be higher, and charged if land is kept vacant, this
would create pressure to landlords trading land that is not bringing enough economic return or
social benefit. Finally, since land was declared a state asset, many well-connected individuals and
firms acquire formally or informally vast areas and are keeping idle waiting for prices to go even
higher.
Inefficient land use not only contributes to inflate prices artificially, but also increase the cost
and complexity in providing urban infrastructure and services. The problem is that lack of an
urban agglomeration economy makes it more difficult and expensive to provide basic services later
on because those areas are beyond the boundaries of urban infrastructure network and service
provision, and cut across very low density areas. Instead of facing marginal costs, investment
requirements need to consider more expensive structural expansion costs shared by fewer users in
order to provide services to households that are located farther away. Vacant land that is not for
20
sale reduce the supply of land, and in a city where demand has been growing substantially, this
artificially increase the overall price of land. If all the stock of vacant land in GMA would be
brought into the market, prices would decrease and equalize based on real supply and demand.
High housing prices are pushing the poor and middle class far from jobs and good
infrastructure. As figure 18 shows, household income decreases away from the central business
district of Maputo. From 1997 to 2007 this situation only got worse. While it is common to find
socioeconomic differences as prices closer to city center is often higher in most cities in the world,
it is contradictory that this phenomenon is taking place in GMA where land is public, and legally
now for sale. The risk of this informal land market is that few are profiting from their privileged
access to information and finance, while the majority of the poor population are subjected to
stressed land sales as they do not have enough information to negotiate with buyers.
Figure 18 Poverty rate and distance to the CBD in the GMA, 1997 and 2007
Source: World Bank staff estimates
Note: Distance is measured from bairro Central C. Smooth lines are based on a locally weighted regression.
This inefficient land use and dysfunction housing market push the poor and middle-class to
the urban fringes, which in turn increase the needs for investing in transport. Currently, it
takes up to two hours for the residents to get to the central business district of Maputo, limiting
their access to job opportunities and increasing their expenses with transport. Recent road
infrastructure development has taken place but the existing road traffic capacity is not meeting the
travel demand of daily travelers to CBD. Residential areas are expanding in a fragmented way in
the suburbs, leading to increased commuting to the central business district between the East and
West axes of the Greater Maputo, and the North and South axes within Maputo (JICA, 2014). This
is causing increased levels of traffic congestions at a travel speed of less than 20km/hour during
peak hours between homes, jobs and businesses, making commuting slow and costly for the urban
poor. While road projects such as the Ring Road Project, a four lane arterial 74 km road, the works
21
on Avenida Marginal, Maputo´s beach road and the Katembe Bridge Project, are aiming to link
the districts to Maputo, there is limited arterial roads to complement the national roads and other
main roads to the suburbs and districts.
In sum, access to affordable housing is deteriorating rapidly in GMA, weighting both in the
pockets of ordinary citizens and local authorities. Since land policies are decided at national
level, it is urgent that the Government of Mozambique addresses the flaws in the current land legal
and institutional framework which makes land so unaffordable in GMA. At the same time, Maputo
and Matola municipalities can use their instruments at their disposal, such as territorial planning,
land management, and land-based finance to incentivize a more efficient urban development.
22
V. Leveraging Land Based Infrastructure Financing
Rapidly growing cities face huge challenges in financing urban infrastructure and services.
As urban areas grow faster than the capacity to finance the requirements in infrastructure and
services, local governments face mounting pressure to address both the current urban deficit in
underserviced areas as well as to lay out the primary networks to keep pace with future growth.
This is particularly true in metropolitan areas as in the case of Maputo, which concentrates the
largest share of the urban population and urban economy in the country. As such, the two main
municipalities, Maputo and Matola, are expected to be less dependent on intergovernmental fiscal
transfers, and more fiscally self-sufficient than the smaller municipalities and rural areas.
International experience shows that larger urban agglomeration can leverage land based
financing as one of the most important source of urban infrastructure finance. Land based
finance instruments are designed to capture both the future increase in real estate from public
finance, and the already internalized value of real estate. Property tax is the most common
instrument at disposal of municipalities. Other land-based finance instruments include betterment
levy, transfer of land use rights. Land value capture financing generate significant resources to
finance infrastructure, complementing or as an alternative from borrowing (table 3).
Table 3. Examples of Land Based Finance for Local Governments
Instrument Objectives
Recurring Land and Building Taxes • Tax is assessed based on the current value of the
property and payed regularly (i.e.; annually)
Betterment Charges and Special Assessments • Generally one-time charges for cost-recovery of future
public infrastructure investments
Developer Extractions • One time charge for the approval of additional
development or issuing of building permits
Land Value Increment Taxes • Allow to capture part of the land value added due to
infrastructure improvements or change in land use for
often goes beyond cost-recovery
Sale of Development Rights • Government sells or action to developers and investors
the right to build more than the limits set by land use
regulations
Land Leases and Sale • Government mobilize resources from selling or leasing
public land
Transfer Taxes and Stamp Duties • Government captures part of the property value when
the title is transfer to another party
Source: Leveraging Land Based Finance for Local Governments: A Reader. UNHabitat 2016.
Unmet Revenue from Land Based Financing
However, municipalities in the Greater Maputo Area are not fully taking advantage of its
land based financing instruments. The country’s decentralization laws empowered
23
municipalities with a variety of land based financing instruments, including property tax (IPRA),
transfer tax (IASISA), economic activity tax (TAE), which is assessed based on the value of the
land (TAE), and betterment levy. But these land based financing instruments are highly
underutilized.
Compared to other cities around the world, property tax revenue has a great potential to
boost OSR. This study focused primarily on the potential to increase municipal revenue from
IPRA due to the importance of property taxation in shoring up municipal finance as seen in many
cities around the work. Moreover, betterment levy has yet to be implemented in Mozambique
though it is provided by law. IASISA is in place, but most underreport the actual value of sales.
According to the Organization for Economic Co-operation and Development (OECD), property
taxes are on average responsible for half of local revenues, 24 but only recently reached 33.5 percent
in Maputo and 15.5 percent in Matola. In 2015, the average IPRA value in Maputo was five times
smaller than the IPRA tax levied by cities like Campinas (2015, São Paulo, Brazil), which is
US$446.70 25 and Durban (2013, eThekwini, South Africa) which is US$353.31. 26 For Matola,
IPRA was fifteen times smaller.
Tax revenues from land based instruments have substantially increased over the past years
in Maputo. The Municipal Council of Maputo (CMM) has undertaken a significant effort to
modernize its property tax system with the support of the World Bank funded Maputo Municipal
Development Project – Phase II (ProMaputo II). Under ProMaputo II, land based tax revenues
increased from Mt 89 million in 2010 to Mt 518 million in 2015 (figure 14). In Matola, tax
revenues also grew from 2011 to 2015 but from a lower base (figure 15). Tax revenues increased
from Mt 22 million in 2011 to Mt 100 million in 2015 (the equivalent of Mt 104.16 per capita). In
2015, the shares of total tax revenue across the various land value capture instruments were
43 percent in property transaction tax (SISA), 33 percent in property tax (IPRA), and 6 percent for
economic activity (TAE).
Figure 19. Growth in tax revenue by main source in Maputo (2010–15)
Maputo
600 518.402
476.135
500
397.708
400 331.784
300 229.025 223.514
211.713
176.065
200 142.729
173.663
89.207 76.833 137.752
100 86.045
110.613
9.446 67.725
27.426 26.465 31.534 30.839 34.085 32.170
0 11.544
2010 2011 2012 2013 2014 2015
Fiscal Revenue IPRA IASISA TAE
Source: Department of Finance of Maputo Municipal Council (April 2016).
Note: figures in MT thousands
24
In Matola, tax revenues also grew from 2011 to 2015 but from a lower base. Tax revenues
increased from Mt 22 million in 2011 to Mt 100 million in 2015 (the equivalent of Mt 104.16 per
capita). The main source of land value capture is SISA, at 45 percent in 2015 (similar to Maputo).
However, IPRA accounted for only 15 percent and TAE 4 percent of collections in 2015. As the
city of Matola only began collecting SISA in 2013, data relate only to 2014 and 2015. Very low
projections in 2014 —the first year of actual tax collection—explain the wide variance with
execution.
Figure 20. Growth in tax revenue by main source in Matola (2011–15)
Matola
120
100.152
100 82.724
80 66.435
60
44,934
40 22.495 25.504
22,298 15.495
20
4.390 4.572 8.442
3.781
0 642 598
8
1,235 2,302 4,348
2011 2012 2013 2014 2015
Fiscal Revenue IPRA IASISA TAE
Source: Department of Finance of Matola Municipal Council (May 2016).
Note: figures in MT thousands
However, the average value of IPRA for both municipalities is extremely low, and its growth
in recent years does not adequately reflect the rapid urbanization in both cities. In 2015, the
average value of IPRA was Mt 4,421 for Maputo and in Mt 1,529 for Matola. The average IPRA
value in Maputo is one-fifth the IPRA tax levied by cities like Campinas (2015, São Paulo, Brazil),
which is US$446.70 27 and Durban (2013, eThekwini, South Africa) which is US$353.31.28
Currently neither municipality currently has an index to systematically record increases in property
values. The Municipal Commission of Urban Building Assessments (CAAPU), which by law is in
charge of assessing real estate values for tax purposes, is required to provide an annual review of
basic values for IPRA collection, incorporating real estate valuations made in the current year. 29
When this study was conducted, Maputo has its CAAPU already operating, albeit with delays in
its assessments. Matola had not approved regulations for establishing a CAAPU.
The number of property units listed in the IPRA cadaster is extremely low. In Maputo,
comparing its data with the 200,000 estimated properties by the Administration of the Real Estate
Housing Stock of the State (APIE), 30 38 percent (76,616 housing units) were registered, but only
20 percent (39,279) were in the actual collection cadaster in 2015. Matola’s data are even patchier,
with only 5 percent (10,137 housing units) in its cadaster. Moreover, supervision system for tax
administration is fragmented among several departments. There are different inspection teams for
each municipal town council and different teams for each public sector responsible for the various
25
land value capture instruments, including property tax (IPRA), property transaction tax (SISA),
economic activity (TAE), and land use right titles (DUAT).
Options for Increasing Land Based Financing
Maputo and Matola municipalities have been increasing their revenue from property tax,
but there are alternative policy options that can boost IPRA even further. As such, the first
option is continuing with the same directions. That would be the baseline for comparing with
alternative options which have the potential to increase revenues even more. After all, IPRA tax
revenue is indeed growing. Hence, if no policy change is done, the growth in IPRA revenue will
continue at the same rate. The second policy option is to enhance enforcement of IPRA collection.
If enforcement is raised to the higher levels as it is currently enforced in District 1, then revenue
can raise to the levels explained on that model. The third policy option is to update construction
costs based on market value. This model would not require any change in the IPRA formula, but
only to move the main factor for assessing property price closer to real market value. The fourth
policy option requires revising the IPRA formula to start charging 1% tax rate instead of the actual
0.4%. The fifth policy option would use the market value of real estate assets instead of the current
formula which mainly assess the construction costs and disregard the actual market value of the
land.
Each policy is valued through the traditional net present value (NPV) valuation method,
which estimates the value of each policy decision. The projections take into consideration the
attrition rates based on the Maputo City Council report on Polana Caniço Urbanization to ensure
that these higher projections are adjusted to forecast the more likely attainable revenue. This
attrition rate was estimated by adding the baseline to the tax collection efficiency rate multiplied
by the spread between the IPRA potential and the baseline. This is done throughout the years,
using IPRA projections. Even accruing for attrition, 2021 collection (orange column) is still
expected to reach Mt 1.5 billion versus Mt 0.8 billion for the baseline for that year.
Scenario 1: Business as Usual
The first policy option for Maputo and Matola to continue with the current approaches to
IPRA collection the way it has for the past decade. Therefore, this first policy option would not
generate any increase in NPV. However, as one can see from the figure 22, there is currently a
wide discrepancy between Maputo and Matola in terms of IPRA collection.
Figure 22. Maputo and Matola IPRA baseline revenue forecast (2017–2021)
26
$900 822
MT millions
$800
$700 639
$600
494
$500
383
$400
298
$300
$200
$100 17 22 28 36 46
$0
2017 2018 2019 2020 2021
Maputo Matola
Source: Maputo and Maputo municipalities. 2015. Revenue Forecast Report; author’s calculations
Notes: in Mt millions
Scenario 2: Potential IPRA collection with improved enforcement
The second policy option is to collect IPRA from all taxable properties in Maputo and
Matola. This policy option does not imply legal changes. It only requires better enforcement of
tax collection throughout all durable homes. This would require more officials and a better tax
registry. Expanding collection enforcement to all durable homes in Maputo, regardless of district,
would generate Mt 4 billion in NPV. The same policy option would generate Mt 2 billion for
Matola in NPV.
Scenario 3: Potential IPRA collection using market based value of construction material
The third policy options require the municipalities to update the value of construction
material using actual market value. The third option also do not require legislative changes as
construction costs are calculated at the discretion of municipalities. Therefore, it would only
require an administrative effort to assess the real market construction cost. That would yield Mt
11 billion for Maputo and Mt 11 billion for Matola in net present value (NPV) terms. It would be
important also to segment the value of construction cost at least in three levels: low, medium and
high-end construction. This would change the current way of using construction cost based on just
one reference value (average) where those below have their property overestimate, and those above
underestimated.
Maputo’s arbitrated construction cost is much closer to nondurable material construction,
rather than durable. As figure 23 shows, this is what causes the city to collect much less than it
would, if it simply charged based on what construction cost residents really bore, out of building
their own homes, rather than some arbitrated cost which does not reflect the reality of durable
material homes costs. By setting the reference value of construction material low, the municipality
is in fact giving an incredible subside to all those households living in more valuable properties.
Figure 23. Construction costs according to market estimates (2016)
27
Source: Authors’ calculation
Matola has less expensive, albeit not much less, construction cost per m2. This obviously has
little relation to any changes in terms of goods and services inputs, albeit some local labor
components might be less expensive than in its peer city Maputo. The difference relies much more
on the nature of the construction projects, which on the high-end finishing materials, end up being
less sophisticated, and cater to a local, rather than to an international market. As one can see from
the figure above, Matola municipality’s arbitrated value is also comparable to nondurable material
construction, rather than the durable material housing which usually, if not preponderantly pays
IPRA in Matola.
Figure 24. Matola construction cost according to market estimates (2016)
Source: Authors’ calculation
28
Scenario 4: Potential IPRA increase through changing tax rate to 1.0 percentage.
The four policy option would require the national government to change the IPRA formula
in order to increase the tax rate from 0.4 to 1.0 percentage of the property value. The value
of this policy change would yield Mt 23 billion for Maputo, and Mt 21 billion for Matola in NPV.
This simple revision in the IPRA formula at national level would increase IPRA revenue for both
municipalities significantly. Since an increase in local tax has a political dimension, the tax hike
could be phased in several years, and first applied to the more valuable properties since this stock
of prime real estate is currently highly subsidized. Progressive IPRA tax rates could be also applied
to vacant land as a deterrent strategy to speculation. In any case, it would be important to change
the legislation and allow municipalities to set their own IPRA tax rate, only setting minimum
levels.
Scenario 5: Potential IPRA collection using property market value
The optimal scenario would combine all previous policies options, but would estimate the
property value based on actual market value. This is clear winner by far for both municipalities.
Changing tax revenue collection to 1 percent of a home’s market value, the Maputo City Council
would generate Mt 44 billion in NPV, and Matola Mt 29 billion in NPV terms. This last policy
option would be an important source of funds to support the needed infrastructure investment in
GMA. Similarly, to make these increases in IPRA tax more acceptable, the municipalities could
spread the tax hikes in 3 to 5 years, focusing first on the upper market properties and vacant land.
That would revert the current regressive nature of the IPRA formula, and how it has been applied,
and thus it would create a broader popular support for increasing IPRA collection.
Figure 25. Comparing revenue increases from various IPRA collection policy decisions
Maputo Matola
Source: Authors’ calculation
29
VI. Recommendations
Sheltering the largest share of the urban population and an important share of the country’s
economy, the Greater Maputo Area has a strategic role in promoting inclusive growth.
Indeed, as in most other countries, urbanization, economic growth, and poverty reduction seem to
go hand in hand. As this study showed, poverty has significantly declined in most of the Greater
Maputo Area. Investments in improving urban infrastructure and services have been also
associated with decline in urban poverty. However, the ‘urban deficit’ is still considerable, as the
two main municipalities in the GMR, Maputo and Matola, have growth considerable faster than
their capacity to provide basic services and urban infrastructure. With almost two million
inhabitants, most living in underserviced neighborhoods, the needs for infrastructure investment
will only increase as the GRM population is expected to double in the next 25-30 years.In order to
address these challenges, but also take advantage of GRM primacy to promote inclusive grow, it
will be important to implement a combination of local and national government actions. Some
actions are urgent as prerequisites for other actions, while others can be implemented more
gradually. The findings of this study helped to identify several recommendations targeting both
national and municipal government:
At National Level
Review the IPRA formula to better capture the real value of urban land. Reformulating the
IPRA formula is critical and urgent. Instead of estimating property value based on construction
costs, the property tax should guide municipalities to use market assessments to estimate property
value. In addition, the property tax code should remove the reference on the value of the square
meter of land, fixed at 5 percent of the value of the construction square meter. Indexes of property
location should also be reviewed, with properties in more valuable neighborhoods valued
accordingly. Without a regulatory framework that recognizes land markets, municipalities use
different methods to estimate the value of land and property. For instance, based on the IPRA
formula, a square meter of urban land is worth Mt 563 on average, while the land value assessment
carried out for this study estimated that values range from Mt 14,402 to 21,603. This shows an
inconsistency between the tax basis of IPRA and IASISA, which should be both based on an asset
value that does not deviate from the market value.
Remove caps for municipalities to adjust the IPRA formula to increase revenue. The IPRA
code needs to become more flexible to address the diversity of urban realities both across
municipalities and within. To implement this change, the code should have only minimum
parameters, and not impose a ceiling such as in the case of the 0.4 percentage property value rate,
and the set rates for depreciation rate and location. By only imposing minimum parameters, the
IPRA code could give autonomy for the municipalities that need urgent investments in urban
infrastructure and services to strike a better balance in terms of local taxes and quality of living.
This would be particularly important for Maputo and Matola in as citizens’ expectations in a
vibrant metropolitan area is much higher than in smaller urban centers.
30
Create a national information system with land, construction, infrastructure market prices
references. The central government should establish basic guidelines for all municipalities to
collect taxes using market prices. It should provide technical and financial support to the
municipalities, helping them develop, deploy, upgrade, and maintain a permanent database.
Market value data can provide key data on the asset value of transferred properties, as well as
surface area and type of use. Analysis of the cadastral data can highlight the dynamics of urban
growth, the pace of real estate speculation, and the potential for increasing land based infrastructure
financing.
Review and incentivize municipalities to apply betterment levies. The Law 01/2008 that
regulates betterment levis is limited and does not consider all kinds of public improvements for
the purpose of cost recovery. Currently it is not applied for road pavements or curbs and gutters
placements on existing or new roads. The law should be reviewed to improve contributions in
cases where the public improvement directly increases the value of private property. These could
include levies on new roads, enlargement, lighting, rehabilitation of squares and public roads;
construction and expansion of parks and gardens; and beautification works in general. The central
government could also provide more clear guidance and incentives through co-financing for
municipalities to use betterment levies to recover all or part of the cost from infrastructure
investment, particularly when these investment takes place in rich areas of the city.
Create new land based financing instruments. While decree 1/2008 includes the possibility of
applying a betterment levy when public investment in infrastructure will in the near future result
in increased property values, currently only property tax and transfer tax are being applied by
municipalities in Mozambique. Betterment levies, while part of the own source revenue
instruments, are not yet applied in Maputo and Matola. Likewise, new legislation would be
required to create and regulate other land based financing instruments, such as developer
extraction, land value increment taxes, sale of development rights, and land leases and sales.
Create a multiuse addressing cadaster unifying the cadasters from utilities and national,
provincial, and municipal governments. Mozambique lacks a postal code system. The lack of a
shared territorial base system for the central government, municipalities, and public utilities, makes
the collection of user fees and property taxes very complex, costly and inefficient.
Monitor and evaluate municipal finances. It is important for the national government to
periodically assess the performance of municipalities in collecting own-source revenue and create
incentives (such as performance grants) to increase fiscal autonomy. Audit reports on municipal
finances should be publicly available to create transparency and increase accountability.
At Municipal Level
Expand and periodically update the IPRA cadaster to include information on land and
building characteristics. The real estate cadaster is the basis for the municipal information
management system, where processes related to the registration and re-registration of land and
construction are monitored and recorded. However, the system has important gaps and should
include the features of the land, building, occupied area, use, and construction in order to help
make appropriate tax assessments and apply a formula for IPRA that reflects market values and
international best practice.
31
Integrate IPRA and municipal land registries to avoid duplications and save resources in
expanding them. An information system must integrate IPRA, DUAT and IASISA registries as
well as the Land Registry Office (Conservatoria de Registo Predial). This would help streamline
procedures of all involved organs agencies, optimizing resources and increasing control and
accuracy of records. The issuance of IPRA payment receipts is not directly linked to the actual
“regularized” ownership of land. DUAT is predominantly handled in physical records while IPRA
in electronic records. Therefore, when Maputo or Matola charges IPRA to all registered properties
in its database, it might fail to capture all the DUAT holders. 31 The implementation of the
Integrated System of Municipal Management (Sistema Integrado de Gestão Municipal—SIGEM)
should help integrate municipal data and enable a more efficient analysis, including data
comparisons and triangulation. 32
Establish and operationalize the CAAPU with management autonomy and with the
responsibility to evaluate properties, values, and indexes to establish the values of the IPRA
and IASISA. Maps containing the market value of the properties in each of the districts and areas
of the city could be prepared. This information would be disclosed to all citizens to guarantee
access to information, especially for the most vulnerable families, whose property may be subject
to stress-sales. The IPRA regulatory proposal, prepared by the Municipal Council of Maputo in
2009, could be used to professionalize CAAPU’s activities. In order to provide greater
transparency and information dissemination, CAAPU’s composition should be expanded to more
stakeholders, particularly representatives of the construction industries and universities.
Review urban structural plans to incorporate the new land value capture instruments.
Municipalities should establish a new posture on the use and occupation of land, setting maximum
urban indexes and coefficients for specific neighborhoods or regions, considering the local
characteristics and needs of the community. When registering the property before a PPU or
detailed plan, holders of vacant and underused land should obtain permits from the municipality.
In case of noncompliance, they should be required to make social use of the property, under penalty
of losing possession and returning it to the municipality.
Establish higher minimum penalties for nonpayment of property tax. Higher fines could help
increase property tax collection. But penalties should be preceded or accompanied by a wide
campaign to encourage payment of the overdue property taxes. The current dual-payment system
(bank and Finance Department) is archaic, prone to errors and fraud. IPRA invoices are issued and
delivered, and payment can be made at bank branches or directly to the Department of Finance of
Maputo or Matola. Although payment at the Department of Finance is directly registered in the
IPRA cadaster, those at banks are considered paid only when the actual payment slips are
physically presented to the Department of Finance (and subsequently recorded in the IPRA
cadaster). Most taxpayers seem therefore prefer paying the IPRA directly at the Department of
Finance.
Set a limited period (1 year) for individuals and firms to pay overdue IPRA before increasing
penalties. Preceding the amendments, fines and collection procedures, a campaign must be
implemented for the settlement of debts in arrears, to encourage payment, before the
implementation of new procedures.
32
Use legal procedures to recover assets of individuals and firms that owe IPRA. Procedures set
out in legislation, such as public auctions, expropriations, pre-emptions, or judicial charges, should
be used to recover resources from citizens and enterprises who have defaulted on their IPRA
payments and give greater credibility and agility to the tax collection system. Clear and objective
local regulations for IPRA should be set out and relevant information shared publicly, as stipulated
by Decree 61/2010. Because the amount is set annually, it is critical to inform the population of
these amounts by neighborhood/district and property type. It is also important to institute fines and
penalties (such as judicial collection, under penalty of the property being sold, for example) for
nonpayment or deliberately false information on the property value for tax purposes, as provided
for IASISA in Decree 46/2004.
Improve Urban Poverty Targeting through Spatial Analysis of Poverty. The present exercise
has shed light on how poverty has evolved between 1997 and 2007 in Maputo and Matola, looking
at highest possible geographic disaggregation that would allow for reliable poverty estimates. The
study showed that urban poverty has been on sharp decline in most neighborhoods of the Greater
Maputo Area. But the findings also suggest that prevalence of poverty is higher in Matola. While
in Maputo all units show a decline in poverty, in half of Matola’s units there has actually been an
increase in poverty rates. Poverty has also been associated with access to infrastructure and
services, location, density and quality of housing. There has been also a higher influx of poor
people in Maputo than Maputo as land and housing costs in the city center of Maputo are
increasingly unaffordable for most of the GMA residents. Poverty maps are therefore a useful tool
to identify priority areas for interventions, prioritize pro-poor infrastructure investments, as well
as better target affordable housing policies.
33
References
Artur, L., and D. Hilhorst. 2012. “Everyday Realities of Climate Change Adaptation in
Mozambique.” Global Environmental Change 22(2): 529–536.
http://doi.org/10.1016/j.gloenvcha.2011.11.013.
Baruah, Neeraj. 2015. “Splintered and Segmented? Fragmentation of African Cities’ Footprints.”
Presentation at the Spatial Development of African Cities Workshop, World Bank,
Washington, DC, December 16–17.https://collaboration.worldbank.org/projects/spatial-
development-research/blog/2016/03/18/spatial-development-of-african-cities-program-
pilots-innovative-new-research.
Bedi, T.A Coudouel and K. Simler (2007) “More than a pretty picture: using poverty maps to
design better policies and interventions”. The World Bank Group.
Constitution of the Republic of Mozambique - 16/11/2004.
Centre for affordable housing in Africa (CAHF), 2016. “housing finance in Africa: a review of
some of Africa’s housing markets”. Centre for Affordable Housing in Africa.
Decree nr. 23/2008 - 01/07/2008 - approving the Regulation on Land Use Management.
Decree 26/2011, of 15/06/2011 - Establishes the Organic Statute of environmental fund
(FUNAB).
Decree n º 38 088, of 12/12/1950 - Code of Tax Executions.
Decree In. 46/2004 - 14/09/2004 - Code of IASISA.
Decree In. 60/2006 - 24/10/2006 - Regulation on Urban Land Use.
Decree In. 61/2010 - approves mechanisms of determination and correction of the asset value of
urban properties.
Decree nr. 63/2008 - 30/12/2008 – Municipal Tax Code.
Decree In. 65/2010 - 21/12/2010 - Alteration of the Legal regime of the State Housing
Development Fund (FFH).
Decree In. 66/98, of 08/12/1998 - Regulation of the Land Law.
Dowall, D.E. 1991. The land market assessment: a new tool for urban management. Washington,
DC: Urban Management Program, The World Bank.
Elbers, C., J. O. Lanjouw, and P. Lanjouw. 2003. “Micro-Level Estimation of Poverty and
Inequality.” Econometrica 71 (1) 355–64.
ESRI Portugal. 2015. Report First Phase of SIGEM Expansion.
Haughton, Jonathon, and Shahidur R. Khandker. 2009. Handbook on Poverty and Inequality.
Washington, DC: World Bank.
34
Hawkins, P., and O. Muxímpua. 2015. Developing Business Models for Fecal Sludge
Management in Maputo. Washington, DC: World Bank.
Japan International Cooperation Agency. 2014. Comprehensive Urban Transport Master Plan
for the Greater Maputo. Retrieved from: http://open_jicareport.jica.go.jp/pdf/12152609.pdf.
INE (National Institute of Statistics). 1998. Inquérito Nacional aos Agregados Familiares
1996/1997. Maputo, Mozambique: INE.
INE. 2010. Inquérito sobre Orçamento Familiar 2008/2009. Maputo, Mozambique: INE.
INE Statistics (database). Maputo, Mozambique: INE. Available at:
http://www.ine.gov.mz/estatisticas.
INE. 2010. Inquérito sobre Orçamento Familiar 2008/2009. Maputo, Mozambique: INE.
Law In. 02/97, of 28/05/1997 – Basic Law of municipalities.
Law In. 19/2007, of 18/07/2007 - Territorial Planning Law.
Law In. 01/2008 - 16/01/2008 – Defining the Financial, Budgetary and Asset Values of the
municipalities and municipal tax system.
Lusaka, December 2000. Retrieved from: http://www.paris21.org/sites/default/files/949.pdf.
MAEFP. 2013. Report of the basic diagnostic of municipal finances. Cities and Climate Change
Project.
MAEFP. 2014. Baseline Study for the Program of Massive Land Regularization (DUATs).
Climate Change Project.
Maputo Chapa project retrieved from https://chapasproject.wordpress.com/.
Maputo City Council. 2009. Regulation of Property Land Tax (IPRA) Proposal.
Maputo City Council. 2009. Plano of Urban Structure of the City of Maputo, Seminary on Urban
Poverty, Maputo.
Maputo City Council. 2012. Evaluation of the environmental impact of the General Urbanization
Plan of Katembe.
Maputo City Council. 2015. Requalification of Polana Caniço A and B – Economic and
Financial Feasibility study, 29 of December 2015.
Maputo City Council. 2015. Model of Technical and Financial Sustainability for the Massive
Regularization Program (DUATs).
Mid Term Fiscal scenario for the period 2016-2018, Maputo, July 2015.
Maputo Transport network information retrieved from http://www.emtpm.co.mz/index.html.
Matola City Council. 2016. Manual for the Evaluation of urban properties.
Ministry of Public Works and Housing (2011) Housing Policy and Strategy for Mozambique.
Paranacidade. 2014. Action Plan for the implementation of IPRA.
PARIS21 Regional Workshop Paper. 2000. The Poverty Reduction Action Plan in Mozambique.
Regulation of Municipal Property Tax (IPRA) - Proposal - City Council of Maputo – 2009.
35
Resolution In. 53/AM/2011 - 16/03/2011 - Municipal Assembly of Maputo.
Resolution 19/2011 - of 08/06/2011 - Policy and Strategy of Housing
Republic of Mozambique. 1975. Constitution of the Republic of Mozambique from 25th of June
1975.
Republic of Mozambique. 2012. Relatório da Revisão do sector de saúde, September 2012.
Republic of Mozambique. 2007–2015. Final Report of the Citizen Satisfaction Report Card.
Maputo Municipal Council, Maputo.
Terra. 2017. “ Preco do metro quadrado em Sao Paulo”. Terra.com.br. Confira o m2 para venda
mais caro de Sao Paulo. 5/2/2017 from terra economia infograficos
The World Bank. 2013. “Doing Business in Bogota, Colombia”. Doingbusiness.org. The World
Bank Group. 5/2/2017. Data from Subnational Doing Business 2013.
The World Bank. Doing Business (database). Washington, DC: World Bank. Available at:
http://www.doingbusiness.org/data.
UNESCO. 1997, 2007. World Inequality Database on Education (WIDE). Retrieved from:
http://www.educationinequalities.org/countries/mozambique/indicators/edu0#?dimension=al
l&group=all&age_group=|edu0_prim&year=|2008.
UNHABITAT. 2016. Leveraging Land Based Finance for Local Governments: A Reader.
UNU-EHS. 2014. World Risk Report 2014. United Nations University Institute for Environment
and Human Security (UNU-EHS).
United States Department of Labor. 2014. Findings on the Worst Forms of Child Labor.
Mozambique country profile. Retrieved from
https://www.dol.gov/agencies/ilab/resources/reports/child-labor/mozambique.
WHO. 1997, 2007. World Health Report Statistician Annex. Retrieved from
http://www.who.int/whr/annexes/en/.
WHO. 1997, 2007. Global Health Observatory data repository. Retrieved from
http://apps.who.int/gho/data/node.main.A1444?lang=en&showonly=HWF.
World Bank. 2009. Mozambique Municipal Development: Lessons from the First Decade.
World Bank. 2010. Analysis of Municipal Revenue Potential in Mozambique, Summary Report.
World Bank. 2015. Country Partnership Framework for the Republic of Mozambique. For the
period FY16-FY20.
World Bank. 2015. Stocktaking of the Housing Sector in Sub-Saharan Africa: Challenges and
Opportunities. Washington, DC: World Bank.
World Bank. 2015. Strategical Plan for the development of the fiscal and financial management
for Matola City Council.
World Bank. 2016. Africa’s Pulse. Vol. 13. Retrieved from
http://documents.worldbank.org/curated/en/2016/04/26197314/africas-pulse-april-2016.
36
World Bank. 2016. Accelerating Poverty Reduction in Mozambique: Challenges and
Opportunities. Washington, DC: World Bank.
http://documents.worldbank.org/curated/en/383501481706241435/pdf/110868-ENGLISH-
PUBLIC-Final-Report-for-Publication-English.pdf.
World Bank. 2017 (upcoming). Mozambique Urbanization Review. Washington, DC. World
Bank.
UN-Habitat. 2003. The Challenge of Slums: Global Report on Human Settlements 2003.
London: Earthscan.
37
End Notes
1
Source: INE Household Welfare Surveys (IOF 1997, 2007).
2
The World Bank, and other development partners have been supporting and participating in the ongoing
decentralization and urban development policy debate under the framework of current operations (Cities and
Climate Change Project (P123201), and National Decentralized Planning and Finance Program (P107311)). After
the announcement of the National Five Year Plan in early 2015, MAE and MEF are preparing studies and strategies
to operationalize national support to municipalities and districts as a framework to new national urban and municipal
development programs.
3
At the time this study was conducted, the INE was undertaking the last household survey (IOF 2017), and planned
to carry out the next population census in the end of 2017. Once these both datasets are available it will be possible to
produce a more up-to-date poverty map of GMA.
4
For Maputo Municipality, the smallest urban administrative division available is the district (there are no
administrative posts in Maputo Municipality), whereas in Matola Municipality it is the administrative post.
5
The two isolated districts are Ka Tembe and Ka Nyaka.
6
World Bank (2016).
7
Units 12 and 13
8
Here, as in the headcount analysis, it is also important to consider the confidence intervals for a more precise
assessment.
9
It is the household’s responsibility to build a sanitation facility in Maputo. Installation of a septic tank costs
US$220, and an improved latrine US$54 (Hawkins and Muxímpua 2015).
10
Toilets connected to a septic tank in the 2007 census does not necessarily mean the toilets are connected to
sewerage. In Maputo in 2013, only 9 percent of residents had sewer connections (Hawkins and Muxímpua 2015).
11
The share of each type of infrastructure access for each bairro is calculated from census data for 1997 and 2007
and household survey collected by the World Bank Water and Sanitation Program (WSP) in 2013.The WSP surveys
were collected from 1,273 households in 47 bairros in Maputo City. See Hawkins and Muxímpua (2015) for more
detail.
12
A series of fixed-effect regression models for bairros in 1997 and 2007 were estimated. Some bairros are
aggregated to larger bairros as in the preparation of poverty mapping. The models control for time-invariant bairro
characteristics, such as distance from the CBD and area, although this alone does not guarantee causality.
13
Estimating the same models by excluding bairros with relatively lower poverty rates, the results did not change
much except that for septic tank, whose linkage with poverty rates become negligible.
14
Calculations based on the WIDE data base numbers. Retrieved from http://www.education-
inequalities.org/countries/mozambique/indicators/edu0#?dimension=all&group=all&age_group=|edu0_prim&year=|
2008.
15
Mozambique’s education system is organized in three main stages: primary from 7 to 10; secondary from 11 to
18; and higher education. Primary education is free.
16
Idem.
17
United States Department of Labor: 2014 Findings on the Worst Forms of Child Labor, Mozambique country
profile. Retrieved from https://www.dol.gov/agencies/ilab/resources/reports/child-labor/mozambique.
18
Calculations based on WIDE database numbers. Retrieved from http://www.education-
inequalities.org/countries/mozambique/indicators/edu0#?dimension=all&group=all&age_group=|edu0_prim&year=|
2008.
19
Programa de Desenvolvimento Espacial https://mozgis.gov.mz
20
Republic of Mozambique: Relatório da Revisão do sector de saúde, September 2012.
21
WHO Global Health Observatory data repository. Retrieved from
http://apps.who.int/gho/data/node.main.A1444?lang=en&showonly=HWF.
22
WHO World Health Report statistical annex. Retrieved from http://www.who.int/whr/annexes/en/.
23
Despite data limitations, multiple sources of existing data was used to determine whether the poor are more likely
to be affected by flooding in the Greater Maputo Area. No historical, georeferenced data of flood events are
38
adequate for spatial analysis of flood hazards at the local level (for instance, high-resolution satellite images from
the floods in 2000 and 2013). To determine the possible extent of pluvial flooding in the GMA—areas at risk of
flooding after heavy precipitation—the analysis uses spatial data from the Drainage and Sanitation Master Plan for
Greater Maputo in 2016.23 As part of the preparation of this Master Plan, “problematic” (flood-prone) areas in the
GMA were identified by crossing the following sources of information: analysis of a digital terrain model to identify
low-lying areas that were validated with aerial photography on Google Earth; information provided by the technical
staff of the Municipal Councils of Maputo and Matola; and information provided by field technicians after site
visits. The dataset covers the entire GMA except Katembe. The dataset distinguishes between “natural” or “reserve”
flood-prone areas (which should in principle be uninhabited) and those occupied by urban settlements. The analysis
is based on overlaying the available flood hazard data from the Master Plan with poverty and population density
figures for 2007 (calculated by the World Bank based on the household survey 2008/09 and population census 2007)
as well as spatial data on built-up area from the Global Human Settlement Layer (GHSL) for 1990 and 2014, as well
as satellite imagery from GoogleEarth.
24
P-13 report, Component C - Municipal Finance - Swiss Agency for Development and Cooperation, Austrian
Development Agency, World Bank, April 2011.
25
Campinas City Council, Department of Finance, 2014.
26
African Tax Institute – University of Pretoria, McCluskey and Franzsen, 2013.
27
Campinas City Council, Department of Finance, 2014.
28
African Tax Institute—University of Pretoria, McCluskey and Franzsen, 2013.
29
Decree 61/20100
30
Report of the Support Program of 13 Municipalities within the Center and North of Mozambique
(P-13)—Component C—Municipal Finances.
31
The data on which was gathered either at the time of granting the Property License Agreement; or at the
inspection before the preparation of a detailed plan of each district; or through a municipal regularization program.
32
According to the 2007 UN-Habitat document "Profile of the Urban Sector of Mozambican", in cases where
municipal cadaster services exist and the area is covered by an official plan of land use, access for an individual
DUAT was achieved after around 103 bureaucratic steps and three years.
39