Content uploaded by Luiz Carlos De Santana Ribeiro
Author content
All content in this area was uploaded by Luiz Carlos De Santana Ribeiro on Nov 12, 2017
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://rsa.tandfonline.com/action/journalInformation?journalCode=cres20
Download by: [CAPES] Date: 05 July 2017, At: 12:51
Regional Studies
ISSN: 0034-3404 (Print) 1360-0591 (Online) Journal homepage: http://rsa.tandfonline.com/loi/cres20
Structuring investment and regional inequalities in
the Brazilian Northeast
Luiz Carlos De Santana Ribeiro, Edson Paulo Domingues, Fernando Salgueiro
Perobelli & Geoffrey John Dennis Hewings
To cite this article: Luiz Carlos De Santana Ribeiro, Edson Paulo Domingues, Fernando Salgueiro
Perobelli & Geoffrey John Dennis Hewings (2017): Structuring investment and regional inequalities
in the Brazilian Northeast, Regional Studies
To link to this article: http://dx.doi.org/10.1080/00343404.2017.1327709
View supplementary material
Published online: 05 Jul 2017.
Submit your article to this journal
View Crossmark data
Structuring investment and regional inequalities in the Brazilian
Northeast
Luiz Carlos De Santana Ribeiro
a
, Edson Paulo Domingues
b
,
Fernando Salgueiro Perobelli
c
and Geoffrey John Dennis Hewings
d
ABSTRACT
Structuring investment and regional inequalities in the Brazilian Northeast. Regional Studies. This paper evaluates the long-
run regional impacts of the construction of three oil refineries proposed for Brazil’s Northeast (NE) region. A dynamic,
interregional computable general equilibrium (CGE) model was developed, with 28 regions in the NE and the rest of
Brazil and 30 sectors. The database construction methodology could be applied to any other country. The main results
of the refinery investment simulations in the NE indicate positive impacts in all regions. However, the construction and
operation of these three refineries would generate an increase in intraregional inequalities. On the other hand, in the
long run, these investments could contribute to the NE gaining an increased relative share of Brazilian gross domestic
product (GDP).
KEYWORDS
structuring investment; regional inequalities; computable general equilibrium (CGE); Brazilian Northeast
摘要
结构投资和巴西东北部的区域不均。Regional Studies.本文评估巴西东北部（NE）的三座炼油厂规划兴建将带来的长
期区域影响。本研究对巴西东北部及其他地方的二十八个区域与三十个部门，建立一个动态、跨区域的可计算均衡
模型（CGE）。该数据集的建立方法，能够应用至任何其他国家。东北部炼油厂投资模拟的主要结果，显示对所有区
域皆产生正面影响。但这三座炼油厂的兴建和运作，将会增加区域之间的不均。另一方面，长期而言，这些投资将
有助于东北部增加其在巴西国内生产毛额（GDP）中的相对比例。
关键词
结构投资;区域不均;可计算均衡模型（CGE）;巴西东北部
RÉSUMÉ
La structuration de l’investissement et les inégalites régionales dans le Nord-Est du Brésil. Regional Studies. Ce présent
article évalue les impacts régionaux à long terme de la construction de trois raffineries de pétrole proposées pour la
région du Nord-Est (NE) du Brésil. À partir de 28 régions situées dans le NE, ainsi que dans le reste du Brésil, et de 30
secteurs, on a construit un modèle dynamique d’équilibre général calculable (Megc) interrégional. On pourrait appliquer
àn’importe quel autre pays la méthodologie relative à la construction de la base de données. Les principaux résultats
des simulations de l’investissement dans les raffineries du NE laissent voir des impacts positifs dans toutes les régions.
Toujours est-il que la construction et l’exploitation de ces trois raffineries finiraient par un creusement des inégalités
intrarégionales. D’autre part, à long terme, ces investissements-là pourraient aider le NE à réaliser un accroissement de
sa part relative du produit intérieur brut (Pib) brésilien.
© 2017 Regional Studies Association
CONTACT
a
(Corresponding author) ribeiro.luiz84@gmail.com
Economics Department, Federal University of Sergipe, Aracaju, SE, Brazil.
b
domingues.edson@gmail.com
Economics Department, Federal University of Minas Gerais, Belo Horizonte, MG, Brazil.
c
fernandosalgueiro.perobelli@gmail.com
Economics Department, Federal University of Juiz de Fora, Juiz de Fora, MG, Brazil.
d
hewings@illinois.edu
Agricultural & Consumer Economics, Economics, Geography and Urban & Regional Planning, Regional Economics Applications Laboratory,
Urbana, IL, USA.
REGIONAL STUDIES, 2017
https://doi.org/10.1080/00343404.2017.1327709
MOTS-CLÉS
structuration de l’investissement; inégalités régionales; équilibre général calculable (Egc); Nord-Est du Brésil
ZUSAMMENFASSUNG
Strukturelle Investitionen und regionales Ungleichgewicht im Nordosten Brasiliens. Regional Studies. In diesem Beitrag
werden die langfristigen regionalen Auswirkungen des geplanten Baus von drei Erdölraffinerien im Nordosten Brasiliens
untersucht. Hierfür wird ein dynamisches interregionales berechenbares allgemeines Gleichgewichtsmodell mit 28
Regionen im Nordosten und in anderen Regionen Brasiliens sowie mit 30 Sektoren entwickelt. Die Methode zum
Aufbau der Datenbank ließe sich auf jedes andere Land anwenden. Die Hauptergebnisse der Simulationen der
Investitionen in Raffinerien im Nordosten lassen auf positive Auswirkungen in sämtlichen Regionen schließen. Allerdings
würden der Bau und Betrieb dieser drei Raffinerien das Ungleichgewicht zwischen den Regionen erhöhen. Andererseits
könnten diese Investitionen langfristig dazu beitragen, dass der Nordosten einen höheren relativen Anteil am
brasilianischen Bruttoinlandsprodukt erzielt.
SCHLÜSSELWÖRTER
strukturelle Investitionen; regionales Ungleichgewicht; berechenbares allgemeines Gleichgewicht (CGE); Nordosten Brasiliens
RESUMEN
Inversión estructural y desigualdades regionales en el noreste de Brasil. Regional Studies. En este artículo analizamos cómo
repercutiría a largo plazo la construcción prevista de tres refinerías de petróleo en la región noreste de Brasil. Para ello se
desarrolló un modelo dinámico de equilibrio general computable interregional, con 28 regiones al noreste y otras regiones
de Brasil, así como con 30 sectores. El método para crear la base de datos podría aplicarse en cualquier otro país. Los
principales resultados de las simulaciones de inversión en las refinerías al noreste del país indican impactos positivos en
todas las regiones. Sin embargo, la construcción y operación de estas tres refinerías provocarían un aumento de las
desigualdades intrarregionales. Por otra parte, estas inversiones podrían contribuir a largo plazo a un aumento de la
participación relativa de la región noreste en el producto interno bruto (PIB).
PALABRAS CLAVES
inversión estructural; desigualdades regionales; equilibrio general computable; noreste de Brasil
JEL C68, R15
HISTORY Received 11 August 2016; in revised form 29 April 2017
INTRODUCTION
In 2007, large oil reserves in deep water, the so-called pre-
salt,
1
were discovered in Brazil. Even with the emergence
of alternative energy sources, oil and its derivatives remain
the main constituents of the Brazilian energy matrix,
accounting for 39.4% of total energy sources, according
to the National Energy Balance published in 2015 by the
Ministry of Mines and Energy (Ministério de Minas e
Energia (MME), Brasil, 2010).
For this reason, the Brazilian oil industry is undergoing
change. There has been major investment in the sector,
especially in new oil refineries, ports and petrochemical
plants. The Abreu e Lima refinery is located in the state
of Pernambuco. The refineries Premium I and II have
been announced for the states of Maranhão and Ceará,
respectively, all located in the Northeast region. The
main reason for constructing these refineries is that the
pre-salt reserve produces light oil, which requires new
refining technologies. This paper explores the potential
impact of these investments on the Northeast, with a par-
ticular focus on their impact on the already high levels of
intra-regional inequalities (Melo & Simões, 2011).
One of the issues to be addressed regarding capital
investment of this order of magnitude (around R$84 bil-
lion) is whether, following Hirschman (1958) and Perroux
(1967), this investment will start an essentially polarized
process, at least initially before spreading to the rest of
the economy.
In the Brazilian context, the main cause of regional
inequality has been linked to industrialization (Diniz,
2006; Guimarães Neto, 1997; Haddad, 1999; Melo &
Simões, 2011; Simões & Freitas, 2014). Baer and Geiger
(1978) argue that industrialization increased regional
inequalities between 1960 and 1970, favouring the South-
east region. Economies of scale gains for industry in the
Southeast were so high that even with transportation
costs, products reached the Northeast with competitive
prices, limiting further the emergence of industries in the
Northeast region.
This argument was also used by Diniz (2006). Between
the mid-19th century and 1960, there were processes of
industrial and urban growth as well as production diversi-
fication. However, these phenomena generated an intense
population and economic concentration in the Southeast,
especially in the metropolitan areas of São Paulo and Rio
2Luiz Carlos de Santana Ribeiro et al.
REGIONAL STUDIES
de Janeiro, leading to higher levels of socio-economic
inequalities among Brazilian regions.
Although the Brazilian literature on regional inequal-
ities is extensive, few studies have assessed the evolution
of this phenomenon in recent years (Silveira-Neto &
Azzoni, 2012). Ferreira (2004) argues that traditional
regional development policies did not contribute to redu-
cing regional inequalities. Carvalho, Somik, and Timmins
(2006) indicate that these policies had positive impacts in
attracting industries to poorer regions, but the impact on
regional inequalities has been limited.
The levels of regional inequality in Brazil are among the
greatest in the world (Baer, 2007; Shankar & Shah, 2001);
however, it has decreased in recent decades, as shown by
Azzoni (2001), Diniz (1993,2006), and Silveira Neto
and Azzoni (2011, 2012). It is important to highlight
that according to Mújica et al. (2014, p. 405):
Four of the BRICS countries [Russia, India, China and
South Africa] showed increases in both income level and
income inequality between 1990 and 2010. The exception
was Brazil, where income inequality decreased over the
same period.
In Brazil, the deconcentration of industrial production,
the control of inflation in the mid-1990s, the minimum
wage-appreciation policy and government income-transfer
programmes were the main factors contributing to the pro-
cess of regional deconcentration (Diniz, 1993; Silveira
Neto & Azzoni, 2011). Despite this reduction in regional
inequalities, the Northeast remains the most unequal
region in intra-regional terms.
In addition to structural changes in economic activity
since the 1960s, the Brazilian Northeast has shown econ-
omic growth rates that in some periods were above the
national average (de Araújo, 2004; Melo & Simões,
2011; Silveira Neto & Azzoni, 2011). However, changes
in the composition of economic activity and overall econ-
omic growth failed to reduce the intraregional inequality.
Economic growth generated by industrial change has
had almost no impact on the internal regional dynamics,
reflected by the absence of social transformation that has
inhibited the development process of the region (Furtado,
1972). Furthermore, industrial-oriented investment has
resulted in increased heterogeneity (de Araújo, 2004;
Lima, 1994) in the Northeast, creating simultaneously
regions based on dynamic and modern structures and stag-
nant regions (de Araújo, 2004; Guimarães Neto, 1997). As
a result, the impacts of modernization have been quite
limited.
The heterogeneity that characterizes the region has
given rise to the expression ‘various Northeasts’, as high-
lighted by de Araújo (2004). From her perspective, there
is the mining and metallurgical and agro-Maranhão
Northeast, the Northeast of western Bahia and sugar
cane Northeast extending along the coast between Rio
Grande do Norte and Alagoas states, cocoa Northeast in
the south of Bahia, the agro-industrial Northeast of sub-
medium São Francisco and the semiarid Northeast.
Melo and Simões (2011) concluded that the gross dom-
estic product (GDP) per-capita growth rates of Northeast-
ern municipalities between 2000 and 2006 were not
affected by their neighbours. In other words, they did not
detect spatial spillovers, implying that these areas are poorly
integrated economically and that the growth process has
not contributed to the spread of capital throughout the
region through spatial spillovers.
The present paper aims to assess the long-term regional
impacts of structural investment, focusing on the possible
construction of three new oil refineries in the Brazilian
Northeast. Computable general equilibrium (CGE)
models can be an appropriate tool to achieve this goal
because the phenomenon being studied involves different
impacts between regions (north-eastern states and the
rest of Brazil) and economic sectors. Partridge and Rick-
man (2010) state that the use of CGE models for regional
policy analysis has increased significantly in recent years. To
achieve our aim, a dynamic, interregional CGE model with
spatial detail for Northeast regions, called B-NORIM
(Brazilian Northeast Interregional Model) was created.
Several studies have shown that CGE models are suit-
able instruments to simulate infrastructure investment
impacts on certain regions because they take into account
the structural and interregional characteristics of the Bra-
zilian economy in an integrated and consistent manner,
and they can therefore evaluate the impact of different
types of infrastructure investments located sectorally and
geographically (Domingues, Betarelli, & Magalhães,
2011;Domingues,Magalhães,&Faria,2009; Haddad
& Haddad, 2010; Haddad, Hewings, Perobelli, & dos
Santos, 2010; Haddad, Perobelli, Domingues, & Aguiar,
2011;V
iana,Domingues,&Diniz,2014). This enables
analysis to be made for the creation of more efficient pub-
lic policies aimed to support development and regional
planning.
This paper will provide results that perhaps could be
used by scholars and policy-makers from other countries
with similar characteristics to Brazil, such as population
and territory size, regional inequality, level of income, and
the challenge of identifying sectors with significant interna-
tionally competitive advantages. Furthermore, the data-
base’s construction could be applied for any other economy.
It is also important to highlight that not only for Brazil
but also elsewhere the analysis of investment in infrastruc-
ture is of interest for two reasons: (1) infrastructure invest-
ment plans are often guided by regional policy aims with
the expectation of benefiting lagging regions; and (2)
measuring the benefits by regions is essential not only for
planning but also for appropriate assignment of costs in
the case where there is collaborative financing between fed-
eral and state government agencies.
We can add that there is often considerable uncertainty
about the regional economic effects of infrastructure invest-
ment. Thus, this kind of exercise is an asset to be used in
Brazil, but this well-established methodology could be
applied in other countries.
The paper is structured as follows. The next section pre-
sents the method used for the database description
2
and
Structuring investment and regional inequalities in the Brazilian Northeast 3
REGIONAL STUDIES
simulation strategy. The following section describes the
results and their evaluation, while the final section has a
summary of the analysis and considerations for future work.
METHOD
B-NORIM is an interregional CGE model with recursive
dynamic mechanisms
3
developed for 28 spatial units in the
Northeast and the rest of Brazil. It is a bottom-up
4
model
such that national results are obtained from the aggregation
of the regional results. B-NORIM is based on the theoreti-
cal structure of TERM (The Enormous Regional Model)
(Wittwer & Horridge, 2010). All these models are based
on the Australian tradition of CGE modelling of the Johan-
sen (1960) type, originally derived from the ORANI model
(Dixon, Parmanter, Sutton, & Vincent, 1982).
One of the main advantages of TERM and B-NORIM
compared with the normal CGE models is their ability to
handle a large number of regions or sectors. This results
from a more compact data structure due to the adoption
of simplifying assumptions. This model treats each region
as a separate economy. In addition, TERM was developed
to enable fast simulations with many regions so as to con-
struct models for large economies such as the United
States, China and Brazil (Horridge, 2012).
In B-NORIM each sector produces only one commod-
ity using intermediate inputs, of domestic and imported
origin, during the production process, as well as primary
factors (capital, labour and land). A standard multilevel
production function is assumed. At the first level, a con-
stant elasticity substitution (CES) is assumed for the choice
of aggregate intermediate inputs and primary factors. At
the next level, a Leontief production function is assumed
for the choice of specific intermediate inputs and factors.
Household demand structures establish optimal consump-
tion compositions by vector selection of goods that maxi-
mizes a specific utility function under a budget
constraint. Hence, for each region of the model, there is
a representative household that consumes domestic and
imported goods. The household demand specification is
based on a combined system of preferences CES/‘linear
expenditure system’.
This model operates with equilibrium equations in the
goods market (domestic and/or imported) consumed locally
as well as in the market factors (labour, capital and land).
Purchase prices of all users (producers, investors, households
and government) are given by the sum of the basic values
with the direct and indirect taxes on sales and margins.
Large-scale multiregional CGE models, such as B-
NORIM, have become an important tool for policy analy-
sis. Such models incorporate innovations in theory, con-
struction and model application (Giesecke & Madden,
2013). The distinguished and specific calibration procedure
in our model, explained in the next session, ensures that it
captures all the specificities in the regional economies in the
model. Therefore, even the standard specification of the
model (one representative household in each region or
competitive equilibrium in all markets) has a specific and
detailed data.
Simulation strategy
Simulations were conducted to evaluate the economic
impact of the oil-refining investments. Two simulations
were designed. The first represents a baseline scenario of
the Northeast/Brazilian economic growth, without consid-
ering any additional investment over the period 2006–27.
5
This scenario is divided into five sub-periods: 2006–11,
2012–15, 2016–19, 2020–23 and 2024–27. The first
period (2006–11) and most of the second (2012, 2013
and 2014) are based on observed macroeconomic data of
the Brazilian economy, as shown in Table 1.
For the other periods, 2016 to 2027, the simulation
assumed adopted growth of real GDP, investment and
exports, of 2.5% per year. Government and household
Table 1. Variables of the baseline scenario, 2006–15 (% variation).
Period
Gross domestic
product (GDP)
Household
Consumption
Government
consumption
Gross fixed capital
formation Export
2006 4.00 5.38 3.55 8.32 4.84
2007 6.01 6.33 4.08 15.56 6.18
2008 5.02 6.44 2.06 13.78 0.41
2009 –0.23 4.19 2.93 –14.11 –9.25
2010 7.57 6.41 3.94 28.21 11.72
2011 3.92 4.76 2.22 5.61 4.79
Accumulated 29.12 38.52 20.30 65.62 18.75
2012 1.76 3.90 3.18 –0.58 0.55
2013 2.74 2.88 2.21 6.12 2.09
2014 0.15 0.89 1.35 –4.40 –1.07
2015
a
–2.70 –2.40 (–2.4)
b
–12.30 8.00
Accumulated 1.88 5.26 4.32 –11.55 9.67
Notes:
a
Estimates from the Central Bank of Brazil (Inflation Report, September 2015) (Banco Central do Brasil, 2015).
b
It is assumed that government spending suffers the same variation of household consumption.
Source: Authors’own elaboration based on data from the Instituto Brasileiro de Geografia e Estatística (IBGE), Fundação Centro de Estudos do Comércio
Exterior (FUNCEX) and IPEADATA.
4Luiz Carlos de Santana Ribeiro et al.
REGIONAL STUDIES
consumption are endogenous in the model. As the goal is
to estimate the impact of oil refinery investments on this
trajectory, the value of economic growth in the baseline
scenario becomes less relevant. The model generates the
growth trajectories of each region and sector, consistent
with the adopted scenario.
The other simulation exercise refers to policy simu-
lation, i.e., the increase in sectorial investment related to
the oil-refining industry located in the regions of Suape-
PE, Rosario-MA and Fortaleza-CE. These investments
produce deviations from the baseline scenario for all vari-
ables in the model.
Figure 1 shows the simplified structure of B-NORIM
from simulations and a choice of endogenous and exogen-
ous variables. As previously discussed, the 2005 database
was the starting point for building the model. The baseline
and policy scenarios provide inputs to the model through
shocks to exogenous variables and B-NORIM generates
different results in the form of regional and sectorial indi-
cators (variables). The deviation resulting from the policy
scenario in relation to the baseline scenario provides the
basis for estimating the impact of the investments.
RESULTS AND DISCUSSION
The results are reported as the cumulative difference
between 2006 and 2027 in relation to the model’s baseline
scenario. According to the model’s mechanisms, it is
expected that these investments will not only generate the
growth of the sectors themselves and associated sectors
but also lead to rising incomes, consumption and employ-
ment in selected regions directly. Indirectly, there may be
competitive effects that reduce the growth of other regions,
mainly due to the displacement effects of employment,
capital and investment.
Table 2 shows the long-term impacts on the main
macroeconomic indicators according to the regions and
north-eastern states in the model. The main final demand
component that could contribute to the positive impact on
GDP would be investment. Changes in investment as
accumulated deviation between 2006 and 2027 in relation
to the baseline scenario of the Rosario-MA, Fortaleza-CE
and Suape-PE would be 76.45%, 0.6% and 4.47%, respect-
ively. This is mainly the direct effect of the shock in the oil-
refining sector of these regions.
In general, all the regions would show positive impacts
on real GDP. The five regions with the largest impact on
GDP would be: Rosário-MA (5.79%), Suape-PE
(1.42%), Itapecuru Mirim-MA (0.17%), Rest of Pernam-
buco (0.16%) and Aglomeração Urbana de São Luís-MA
(0.15%). Among the top five, three would occur in the
sub-regions of Maranhão state. This result can be explained
partly because of the investment value of the Premium I
refinery which is much greater compared with the other refi-
neries and the region is smaller. Therefore, this investment
is relatively greater, as we can see in the column ‘Investment’.
The impacts on Itapecuru Mirim-MA and Aglomera-
ção Urbana de São Luís-MA can be interpreted as spillover
effects from investment in Rosário-MA. This effect would
occur through regional purchases, as 14% of Rosário’s total
domestic purchase comes from sub-regions of Maranhão.
In other words, as highlighted by Hirschman (1958),
these sub-regions would benefit from Rosário-MA
trickle-down effects, while apparently polarization effects
would be minimal because no region would be negatively
affected by the construction of the refineries. There is a
spatial concentration of growth around the regions where
progress begins. These results together show that Maran-
hão would be positively affected the most among the
Northeastern states with a GDP growth of 0.35%.
The impacts on Fortaleza-CE, the location of Premium
II, and on the other regions of Ceará state would be smaller
compared with the other regions receiving investment. It is
important to mention, however, that despite this invest-
ment being higher than in the case of Abreu and Lima,
the construction of Premium II was simulated in 2019–
Figure 1. Simplified structure of the simulations in B-NORIM.
Source: Authors’own elaboration.
Structuring investment and regional inequalities in the Brazilian Northeast 5
REGIONAL STUDIES
23 and, as a result of value updating of state and regional
investments calculated by the model, it generated a much
lower shock value when compared with the other refineries.
In addition, the metropolitan area of Fortaleza is one of the
largest economies in the Northeast, and it is less sensitive to
exogenous impacts. In other words, the refinery would
cause a production capacity increase in the following period
and, therefore, this investment plus the sector’s gross fixed
capital would generate a higher return on the capital. How-
ever, the gross fixed capital was already high in this region,
which would generate a relatively minor impact.
Although Bahia, Ceará and Pernambuco states have the
most dynamic economies and also the greatest relative
importance in terms of intersectoral relationships in the
Northeast, this was not reflected when, for instance, the
impacts on GDP of these states are analyzed. Excluding
Bahia, where no refinery would be built, the oil-refining
sector of Pernambuco has no strong sectorial linkages
(Guilhoto, Azzoni, Ichihara, Kadota, & Haddad, 2010).
On the other hand, in Ceará state, although this sector
has linkages above the average, the value of the Premium
II investment would be small, as explained above.
Table 2. Macroeconomic and regional results: accumulated deviation 2006–27 compared with the baseline (%).
Regions/variables
Real gross domestic
product (GDP)
Household
consumption Investment Export Import Employment
Litoral Ocidental
Maranhense
0.02 0.01 0.07 –0.01 0.03 0.02
Aglomeração Urbana de
São Luís
0.15 0.16 1.09 –0.01 0.18 0.17
Rosário 5.79 0.19 76.45 –0.02 7.25 0.20
Lençóis Maranhenses 0.07 0.06 0.54 0.01 0.07 0.07
Baixada Maranhense 0.04 0.03 0.21 0 0.04 0.04
Itapecuru Mirim 0.17 0.13 0.54 0.02 0.13 0.15
Chapadinha 0.04 0.03 0.16 0 0.04 0.04
Rest of Maranhão 0.08 0.07 0.34 –0.01 0.06 0.08
Maranhão 0.35 0.11 9.40 –0.01 0.84 0.12
Piauí 0.03 0.03 0.22 –0.01 0.04 0.04
Baixo Curu 0.04 0.04 0.25 –0.01 0.06 0.05
Médio Curu 0.01 0.01 0.03 0 0.02 0.02
Canindé 0.02 0.01 0.10 0 0.03 0.02
Baturité 0.02 0.01 0.13 0 0.03 0.02
Cascavel 0.02 0.02 0.09 –0.01 0.03 0.03
Fortaleza 0.05 0.03 0.60 –0.01 0.11 0.04
Pacajus 0.01 0.01 0.05 –0.01 0.02 0.02
Rest of Ceará 0.02 0.01 0.13 –0.01 0.03 0.02
Ceará 0.04 0.03 0.47 –0.01 0.08 0.04
Rio Grande do Norte 0.05 0.04 0.28 0 0.05 0.05
Paraíba 0.04 0.02 0.24 –0.01 0.04 0.03
Vitória de Santo Antão 0.05 0.03 0.23 –0.01 0.05 0.04
Mata Meridional
Pernambucana
0.09 0.07 0.44 0 0.10 0.08
Recife 0.10 0.06 0.35 –0.04 0.08 0.07
Suape 1.42 0.52 4.47 0.03 1.66 0.56
Rest of Pernambuco 0.16 0.10 0.54 –0.01 0.13 0.12
Pernambuco 0.27 0.14 1.20 –0.02 0.39 0.13
Alagoas 0.03 0.02 0.16 –0.01 0.05 0.03
Sergipe 0.08 0.07 0.20 –0.01 0.09 0.09
Salvador 0.03 0.02 0.11 0.02 0 0.03
Rest of Bahia 0.02 0.01 0.08 –0.02 0.02 0.02
Bahia 0.02 0.01 0.10 0.01 0.01 0.02
Rest of Brazil 0.01 0 0.04 –0.01 0.02 0.01
Brazil 0.01 0.01 0.06 –0.01 0.03 0.01
Source: Authors’own elaboration based on B-NORIM (Brazilian Northeast Interregional Model) simulations.
6Luiz Carlos de Santana Ribeiro et al.
REGIONAL STUDIES
The impact on the rest of Brazil’s GDP would be
0.01%, suggesting that there would be a very small comp-
lementary effect instead of a substitution effect. In other
words, the rest of Brazil would take advantage to sell
more to the Northeast. Moreover, the construction of
three new refineries in the Northeast would increase its par-
ticipation in Brazilian GDP in 2027. Without the refinery
investments, this share would increase about 0.37% in
2027, while with the construction of three refineries this
share would rise by 1.01% in the same year.
The construction of these three refineries would mean
higher production costs in all sectors of the regions that
would receive these investments due to the increased
demand for productive factors (capital and labour). This
increase in production costs is passed on to final consumers
via price increases, which makes local goods relatively more
expensive than imported goods. Given the model’s mech-
anism of substitution effect, this would stimulate imports
(positive changes in all regions of the model, especially in
the regions where the refineries are located) and discourage
exports (negative changes in almost all regions). Further-
more, the increase in production also generates increased
imports.
Although some regions usually face high unemployment
rates, the investments and their indirect effects tend to
attract more skilled labour, which tend to reallocate from
other sectors. Low productive activities, such as agriculture,
usually experience a decrease in labour force and production
as workers move to service sectors to take advantage of
higher wages. Therefore, the projected increase in prices
and costs in the model is very suitable to the situation.
As expected these investments in infrastructure would
have a positive impact on aggregate employment in all
regions,
6
especially in regions where the refineries are
located. The increase in employment, in turn, would lead
to higher incomes and consumption. The impact on house-
hold consumption in the regions of Rosario-MA, Forta-
leza-CE and Suape-PE, for example, would be 0.2%,
0.04% and 0.56%, respectively.
It is worth noting that in the model regional propensity
to consume from labour income is fixed, and regional
income is derived from labour and capital remunerations.
Therefore, changes in capital income become changes in
regional savings. This is a conservative hypothesis because
there is little information about the disposition of capital
income property in both the national and regional dimen-
sions in Brazil. Probably, in the case or refining invest-
ments, the capital income accrues to all households in
Brazil, and even the federal government, as Petrobras is a
controlled state company with market stocks. On the
supply side, the increase in capital participation is the
most important factor that would explain the regional
GDP growth, while on the demand side this role would
be played by investment expenditure.
According to the New Economic Geography, the con-
centration of industries in a given region would mean
cheaper goods and services via the reduction of transpor-
tation costs and, hence, would contribute to the emergence
of agglomeration economies (Baldwin, Forslid, Martin, &
Robert-Nicoud, 2003; Fujita, Krugman, & Venables,
1999; Fujita & Thisse, 2002). Interestingly, the results
obtained through simulations with a perfect competition
model suggest this process for Maranhão and Pernambuco
states because only the sub-regions of these states would see
a decline in the general price index of goods and services.
In order to intensify the process of emergence of
agglomeration economies in the Northeast, it is worth
mentioning the importance of the creation of consumer
markets at the regional level. Otherwise, instead of redu-
cing the transportation cost with goods derived from oil
refining, would raise this cost as the main Brazilian consu-
mer centre is in the Southeast. Furthermore, the simulation
results indicate the need to strengthen the Northeastern
consumer market because the impacts on employment
and consumption would not be very significant.
Added to this, it is essential the Northeastern pro-
duction chains densification in order to minimize income
and employment spillovers through the purchase of inputs,
goods and services which are essential to the functioning of
possible growth poles led by three new refineries.
This set of factors should create a favourable economic
environment for attracting other industries and qualified
labour. The result of this process, according to Marshall
(1890), is the agglomeration of people and industrial activi-
ties in a given geographical area and consequently urban-
industrial development. For Jacobs (1969), the diversity
of economic activities developed in cities is the largest
and the main source of positive externalities.
7
Sectorial impacts
Regarding the B-NORIM economic sectors, Table 3
shows the impact on output as the accumulated difference
between 2006 and 2027 relative to the baseline scenario.
We highlighted in light grey the sectors that lose out
with the construction of the refineries, i.e., they are the seg-
ments there would be a relative decrease in output. More-
over, other sectors would be winners as there would be
positive changes in their respective sectorial production.
It is very important to assess the magnitude of the signal
(qualitative bias) than simply analyze only the ‘numbers’
that CGE models provide (quantitative bias). From this
sample, for instance, policy-makers may perceive sectorial
and regionally sectors that gain/lose with the construction
of new refineries in the Northeast.
In the oil refining industry, 17 regions would lose out
with this policy (oil refineries construction). This is because
of the price effect, that is, with the construction of new refi-
neries these regions would lose market. Analyzing the
results, it is clear that there is fall in prices in the regions
where refineries are located, while there are higher prices
in other regions. Interestingly, in 27 regions, Salvador-
BA, one of the losing regions, is the petrochemical complex
of Camaçari and Landulpho Alves Refinery (RLAM).
The most significant impacts on the output of the oil-
refining sector would take place in the regions that receive
direct investment, namely numbers 3 –Rosario-MA
(7.5%), 15 –Fortaleza-CE (0.4%) and 23 –Suape-PE
(4.6%). This result is expected because Guilhoto et al.
Structuring investment and regional inequalities in the Brazilian Northeast 7
REGIONAL STUDIES
Table 3. Impacts on sectorial output: accumulated deviation 2006–27 compared with the baseline (%).
Sectors 1 2 3
a
456789101112131415
a
16 17 18 19 20 21 22 23
a
24 25 26 27 28 29
Agriculture and livestock 1.0 0.1 1.4 0.1 0.1 0.3 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.4 0.2 0.1 0.0 0.0 0.0 0.0
Oil and natural gas 0.5 1.6 0.2 1.1 0.6 2.0 0.6 1.1 0.5 0.2 0.2 0.2 0.2 0.2 0.3 0.2 0.2 0.2 0.3 0.2 0.5 0.7 0.0 0.4 0.2 0.1 0.1 0.0 0.1
Other mining and quarrying 0.0 –0.2 0.2 –0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.3 0.2 0.7 0.3 0.1 0.0 0.0 0.0 0.0
Food and beverage 0.1 0.0 0.2 0.0 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.2 0.1 0.0 0.0 0.0 0.0 0.0
Textiles and vestments 0.0 –0.2 0.1 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.3 0.1 0.0 0.0 0.0 0.0 0.0
Wood products –excluding furniture 0.0 0.0 0.3 0.0 0.0 0.1 0.0 0.1 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.2 0.1 0.0 0.0 0.0 0.0
Pulp and paper products 0.0 –0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0
Oil refining and coke 0.0 0.2 7.5 0.0 0.0 0.1 0.0 0.0 0.0 –0.1 0.0 0.0 0.0 0.0 0.4 0.0 0.0 0.0 0.0 0.0 0.1 0.0 4.6 0.2 0.1 0.0 0.0 0.0 0.0
Alcohol 0.2 0.5 0.1 0.4 0.3 0.8 0.2 0.4 0.2 0.1 0.1 0.2 0.1 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.3 0.2 0.5 0.6 0.1 0.1 0.0 0.0 0.0
Chemicals 0.0 0.0 0.6 0.0 0.1 0.2 0.1 0.1 0.0 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.3 0.1 0.5 0.2 0.0 0.0 0.0 0.0 0.0
Rubber and plastic goods 0.0 0.0 0.3 0.1 0.0 0.1 0.0 0.1 0.0 0.0 0.1 0.1 0.0 0.1 0.1 0.0 0.1 0.1 0.1 0.1 0.2 0.1 0.2 0.2 0.1 0.0 0.0 0.0 0.0
Cement 0.1 0.4 0.4 0.2 0.2 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.2 0.2 0.2 0.1 0.2 0.1 0.2 0.6 0.4 1.7 0.5 0.2 0.1 0.1 0.1 0.0
Manufacture of steel and derivatives 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.2 0.2 0.3 0.2 0.1 0.0 0.0 0.0 0.0
Metallurgy of non-ferrous metals –0.1 –0.3 0.1 –0.1 0.0 –0.1 0.0 0.0 –0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.3 0.1 0.0 0.0 0.0 0.0 0.0
Metal products –excluding machinery
and equipment
0.2 1.1 0.7 0.6 0.3 1.5 0.2 0.6 0.3 0.1 0.2 0.2 0.2 0.2 0.2 0.1 0.2 0.2 0.3 0.3 0.5 0.4 0.4 0.5 0.2 0.1 0.1 0.1 0.0
Machinery and equipment, including
maintenance and repairs
0.3 1.3 0.5 0.6 0.3 1.5 0.2 0.6 0.3 0.3 0.3 0.3 0.2 0.2 0.2 0.3 0.2 0.3 0.4 0.4 0.7 0.5 0.4 0.8 0.3 0.2 0.1 0.1 0.0
Electrical appliances 0.1 0.0 0.2 0.0 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.0 0.0 0.0 0.0
Machinery for office and computer
equipment
0.4 2.2 0.4 1.0 0.4 2.4 0.4 0.9 0.5 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.4 0.5 0.4 0.8 0.6 0.4 0.9 0.3 0.2 0.1 0.1 0.1
Electrical machinery, equipment and
materials
0.2 1.1 0.3 0.5 0.2 1.5 0.2 0.5 0.2 0.1 0.2 0.2 0.1 0.2 0.2 0.1 0.2 0.2 0.3 0.2 0.4 0.3 0.6 0.4 0.2 0.1 0.1 0.1 0.0
Electronic material and
communication equipment
0.2 0.9 0.5 0.4 0.2 0.9 0.2 0.4 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.4 0.3 0.4 0.4 0.2 0.1 0.1 0.1 0.0
Medical and hospital equipment,
measurement and optical
0.1 0.4 0.1 0.2 0.2 0.6 0.1 0.3 0.1 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.3 0.2 0.1 0.3 0.1 0.1 0.1 0.1 0.0
Automobile industry 0.1 0.5 0.1 0.2 0.1 0.5 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.2 0.1 0.0 0.1 0.0 0.0
Other transport equipment 0.0 0.1 0.3 0.1 0.0 0.3 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.2 0.2 0.3 0.2 0.1 0.0 0.0 0.0 0.0
(Continued)
8Luiz Carlos de Santana Ribeiro et al.
REGIONAL STUDIES
(2010), when using pure linkages indexes, showed that the
oil-refining segment is incipient in the Northeast, especially
in Alagoas, Maranhão, Paraíba, Pernambuco and Piauí.
To exploit better the results of Table 3, we adopted the
following criteria: 10% of the largest impacts were selected
taking into account all the regions (highlighted in dark
grey), that is, those sectors that would present the greatest
variation in output, such as oil and gas, alcohol, cement,
and metal products –except machinery and equipment,
machines and equipment, including maintenance and
repairs, office machinery and computer equipment,
machinery, electrical appliances and materials, electronic
material and communications equipment and construction.
It is notable that the sectors that would be impacted
more are those that have direct or indirect relations during
the phases of construction and operation of the new refi-
neries. In other words, these are the sectors that would pro-
vide inputs for the construction as well as for oil processing.
It is worth noting that in the period in which the shock
occurs, that is, in which rises the gross fixed capital of the
industry, its demand will have to source capital-intensive
sectors, i.e., construction and segments related to machin-
ery and equipment, for example.
Furthermore, according to this criteria, there is a
regional concentration in relation to major impacts, since
15 regions would present sectors that would fulfil this cri-
teria. Following the same logic of the impact on output,
Table 4 shows the impact on sectorial employment.
The highest impact would be in the oil and natural gas
industry, which would be the supplier of the main input for
the oil refineries. We can see an interesting spillover effect
in the sub-regions of Maranhão because the greatest impact
would not be in Rosario-MA, the location of Premium I,
but in numbers 2 –Algomeração Urbana de São Luís
(0.7%) and 6 –Itapecuru Mirim-MA (0.4%).
Tables 3 and 4 show that the services segment in gen-
eral would present positive impacts in almost all regions,
except for only five regions, considering the output. This
result, especially for the regions where the refineries are
located, suggests that industrial development would tend
to be accompanied by the complexity of the services seg-
ment. According to Ribeiro, Lopes, Simões, and Moreira
(2013), this would contribute to urban diversification,
which is associated with the production modernization
and increased scale of economic activities in these regions
(Perobelli, 2004). A similar result had already been pointed
out by Ribeiro et al. (2013), who suggest diversification and
complexity in the services segment in Suape-PE due to the
construction of the Abreu and Lima refinery.
There is a discussion between development services
segment and urban hierarchy (size of municipalities). Per-
obelli, Cardoso, Vale, and Rodrigues (2015) argue that this
is a critical point regarding the Northeast where, except for
the metropolitan areas, other municipalities are small, with
a small scale for the development of more complex services.
Markusen (2004) points out that the higher the inten-
sity in trade relations in the service segment, the greater the
potentialization effect of the growth transmission of an
industrial pole for peripheral regions.
Table 3. Continued.
Sectors 1 2 3
a
456789101112131415
a
16 17 18 19 20 21 22 23
a
24 25 26 27 28 29
Furniture and products from diverse
industries
0.2 0.3 0.2 0.2 0.1 0.8 0.1 0.2 0.1 0.0 0.1 0.1 0.1 0.1 0.1 0.0 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.2 0.1 0.1 0.1 0.1 0.0
Industrial services of public utility 0.0 –0.3 0.1 –0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.2 0.1 0.0 0.0 0.0 0.0 0.0
Construction 0.2 1.7 0.9 0.7 0.3 3.2 0.2 0.7 0.3 0.2 0.2 0.2 0.1 0.1 0.3 0.2 0.1 0.2 0.1 0.1 0.1 0.3 5.4 0.6 0.1 0.1 0.1 0.1 0.0
Trade 0.2 0.8 0.5 0.4 0.2 0.8 0.2 0.4 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.3 0.2 0.4 0.4 0.2 0.1 0.1 0.1 0.0
Transport, storage and postal mail 0.1 0.2 1.1 0.2 0.1 0.5 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.3 0.2 1.0 0.4 0.1 0.1 0.1 0.0 0.0
Other private services 0.0 –0.1 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.2 0.1 0.1 0.0 0.0 0.0 0.0
Public services 0.0 0.0 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.0 0.0 0.0 0.0 0.0 0.0
Note:
a
Oil refineries’location numbers: 3, Rosário-MA; 15, Fortaleza-CE; and 23, Suape-PE.
Source: Authors’own elaboration based on B-NORIM (Brazilian Northeast Interregional Model) simulations.
Structuring investment and regional inequalities in the Brazilian Northeast 9
REGIONAL STUDIES
Table 4. Impacts on sectorial employment: accumulated deviation 2006–27 compared with the baseline (%).
Sectors 1 2 3
a
456789101112131415
a
16 17 18 19 20 21 22 23
a
24 25 26 27 28 29
Agriculture and livestock 0.1 0.2 1.2 0.1 0.1 0.3 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.1 0.1 0.1 0.1 0.1 0.4 0.2 0.1 0.0 0.0 0.0 0.0
Oil and natural gas 1.3 5.5 0.8 3.3 1.6 5.5 1.6 3.1 1.4 0.6 0.6 0.7 0.6 0.6 0.7 0.6 0.5 0.4 0.6 0.4 0.9 1.1 0.3 0.6 0.3 0.2 0.2 0.1 0.1
Other mining and quarrying 0.0 0.2 0.5 0.1 0.0 0.1 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.3 0.2 0.6 0.3 0.1 0.0 0.0 0.0 0.0
Food and beverage 0.1 0.2 0.4 0.1 0.1 0.2 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0
Textiles and vestments 0.1 0.0 0.2 0.1 0.1 0.1 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.3 0.1 0.0 0.0 0.0 0.0 0.0
Wood products –excluding furniture 0.1 0.2 0.6 0.1 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.3 0.2 0.4 0.3 0.1 0.1 0.1 0.1 0.0
Pulp and paper products 0.0 0.2 0.3 0.1 0.0 0.1 0.0 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.1 0.1 0.0 0.1 0.1 0.1 0.1 0.1 0.3 0.1 0.1 0.0 0.0 0.0 0.0
Oil refining and coke 0.0 0.7 0.0 0.2 0.1 0.4 0.1 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 –0.1 0.1 0.2 0.0 0.2 0.0 0.0 0.0 0.0
Alcohol 0.5 1.8 0.3 1.0 0.5 1.7 0.5 0.9 0.4 0.4 0.3 0.3 0.3 0.3 0.3 0.3 0.2 0.1 0.2 0.1 0.2 0.2 0.3 0.5 0.1 0.1 0.1 0.1 0.0
Chemicals 0.1 0.4 1.0 0.2 0.1 0.3 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.1 0.1 0.1 0.1 0.3 0.1 0.4 0.2 0.0 0.1 0.0 0.0 0.0
Rubber and plastic goods 0.1 0.2 0.5 0.1 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.4 0.3 0.1 0.1 0.1 0.1 0.0
Cement 0.1 0.5 0.7 0.3 0.2 0.2 0.1 0.2 0.1 0.2 0.2 0.1 0.1 0.2 0.2 0.2 0.1 0.2 0.1 0.2 0.5 0.4 1.6 0.5 0.2 0.1 0.1 0.1 0.0
Manufacture of steel and derivatives 0.1 0.3 0.2 0.1 0.0 0.1 0.0 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.1 0.1 0.1 0.1 –0.1 0.1 0.1 0.0 0.0 0.0 0.0
Metallurgy of non-ferrous metals 0.0 0.0 0.4 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.3 0.1 0.0 0.0 0.0 0.0 0.0
Metal products –excluding machinery
and equipment
0.2 0.8 0.6 0.4 0.2 1.0 0.2 0.4 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.3 0.2 0.4 0.3 0.4 0.5 0.2 0.1 0.1 0.1 0.0
Machinery and equipment, including
maintenance and repairs
0.3 1.5 0.7 0.7 0.3 1.6 0.3 0.6 0.4 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.5 0.4 0.8 0.6 0.6 0.9 0.3 0.2 0.1 0.1 0.0
Electrical appliances 0.1 0.2 0.3 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0
Machinery for office and computer
equipment
0.2 0.5 0.4 0.3 0.2 0.5 0.2 0.3 0.2 0.3 0.3 0.2 0.2 0.2 0.2 0.3 0.2 0.2 0.3 0.3 0.5 0.4 0.4 0.6 0.2 0.2 0.1 0.1 0.1
Electrical machinery, equipment and
materials
0.2 0.9 0.4 0.4 0.2 1.2 0.2 0.4 0.2 0.1 0.1 0.2 0.1 0.1 0.2 0.1 0.1 0.2 0.3 0.2 0.4 0.3 0.5 0.4 0.2 0.1 0.1 0.1 0.0
Electronic material and communication
equipment
0.1 0.5 0.5 0.3 0.1 0.5 0.1 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.3 0.3 0.3 0.4 0.2 0.1 0.1 0.1 0.0
Medical and hospital equipment,
measurement and optical
0.1 0.4 0.2 0.2 0.1 0.4 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.3 0.1 0.1 0.1 0.1 0.0
Automobile industry 0.0 –0.5 0.1 –0.2 –0.1 –0.7 –0.1 –0.2 0.0 0.1 0.1 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 –0.1 0.2 0.1 0.0 0.0 0.0 0.0
Other transport equipment 0.0 0.2 0.3 0.1 0.0 0.2 0.0 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.1 0.0 0.0 0.1 0.1 0.1 0.2 0.1 0.2 0.2 0.1 0.0 0.0 0.0 0.0
(Continued)
10 Luiz Carlos de Santana Ribeiro et al.
REGIONAL STUDIES
Regional inequalities
What impact would the construction of the three North-
east refineries have on regional inequalities? We attempted
to answer this question as follows: (1) intra-regionally con-
sidering the 28 regions of the Northeast (excluding the rest
of Brazil); and (2) intra-state considering the sub-regions
of the states that will receive the refineries. The results
used for this are the regional GDP changes (simulation
results) and the region shares (database).
In general, the addressed topic about regional inequal-
ities could be evaluated among regions or among people. In
this paper, the first aspect is considered. In essence, the idea
is just to look at the regional GDP distribution in 2027
with and without the oil refineries. Therefore, we do not
take into account any income change among households
since, as noted above, B-NORIM has a single representa-
tive household agent for each region.
To analyze the impact that the construction of three
refineries could have in the Northeast, the method proposed
by Domingues, Magalhães, and Faria (2009) was adopted,
that is, we calculated the Gini index from the regional GDP
at current prices of the baseline and policy (taking into
account the refineries construction) scenarios in 2027.
The idea is to see if there would be a positive (concentration)
or negative (deconcentration) variation of the Gini index.
Table 5 shows the Gini index calculated in the baseline
scenario and impacted by refineries and their relative vari-
ation for different spatial scales used in the B-NORIM
model. Remember that these results consider the GDP
generated by simulations of the baseline and policy scen-
arios in 2027.
All areas considered in the analysis would see a positive
variation of the Gini index, indicating that the construction
of new refineries in the Northeast would increase inequal-
ities, albeit marginally, in both intra-regional and intra-
state terms.
A similar result was found by Domingues et al. (2009).
According to these authors, the Growth Acceleration Pro-
gram (PAC) in Minas Gerais generated positive impacts
on the state’s GDP. However, they contributed to the
increase the intra-state inequalities.
Maranhão state would have the greatest Gini index
variation, 0.34%, whereas Ceará state would have the smal-
lest variation, 0.006%. It is interesting to note that the
Table 4. Continued.
Sectors 1 2 3
a
456789101112131415
a
16 17 18 19 20 21 22 23
a
24 25 26 27 28 29
Furniture and products from diverse
industries
0.1 0.3 0.4 0.2 0.1 0.4 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.3 0.2 0.1 0.1 0.1 0.1 0.0
Industrial services of public utility 0.1 0.0 0.4 0.0 0.1 0.2 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.4 0.1 0.1 0.0 0.0 0.0 0.0
Construction 0.2 1.7 1.1 0.8 0.3 3.2 0.3 0.8 0.3 0.3 0.2 0.2 0.2 0.2 0.4 0.2 0.1 0.2 0.1 0.1 0.1 0.2 3.0 0.4 0.1 0.1 0.1 0.1 0.1
Trade 0.2 1.0 0.5 0.4 0.2 0.8 0.2 0.4 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.3 0.2 0.2 0.3 0.1 0.1 0.1 0.1 0.0
Transport, storage and postal mail 0.2 0.6 1.7 0.4 0.2 0.7 0.2 0.4 0.2 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.2 0.2 0.4 0.2 1.1 0.4 0.1 0.1 0.1 0.1 0.0
Other private services 0.0 0.0 0.2 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.3 0.2 0.1 0.0 0.0 0.0 0.0
Public services 0.0 0.0 0.3 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.0 0.0 0.0 0.0 0.0 0.0
Note:
a
Oil refineries’location numbers: 3, Rosário-MA; 15, Fortaleza-CE; and 23, Suape-PE.
Source: Authors’own elaboration based on B-NORIM (Brazilian Northeast Interregional Model) simulations.
Table 5. Impacts on regional inequality: Gini indexes of
regional gross domestic product (GDP) in 2027 in the
baseline and policy scenarios.
Region
Gini index
Baseline
scenario
Policy
scenario %
Northeast 0.6541 0.6544 0.0477
Maranhão 0.6503 0.6526 0.3439
Ceará 0.7352 0.7353 0.0060
Pernambuco 0.5026 0.5032 0.1185
Source: Authors’own elaboration based on B-NORIM (Brazilian Northeast
Interregional Model) simulations.
Structuring investment and regional inequalities in the Brazilian Northeast 11
REGIONAL STUDIES
magnitude of the Gini index variation is correlated to the
cost of each refinery. In other words, the higher the invest-
ment in terms of refinery construction cost, the greater the
impact on regional inequality.
These results are consistent with those in the literature.
According to Diniz (1993), for instance, while infrastruc-
ture investment generates positive economic outcomes, in
regional terms such investment produces heterogeneous
impacts and tends to increase regional inequalities.
CONCLUSIONS
The objective of this paper was to evaluate the long-term
economic impact that the construction of three new oil refi-
neries would have in the Brazilian Northeast and also to
assess the effect of such investment on regional inequalities.
The topic addressed is very important because despite the
reduction in Brazilian regional inequalities in recent
years, the Northeast remains the most unequal region in
the country.
The main results indicate that the construction of the
refineries in the long-run would increase the Brazilian
Northeast’s share in national GDP. On the other hand,
such investment would increase the intra-regional and
intra-states inequalities, albeit marginally.
These results, although related specifically to Brazil,
they may provide important insights for other countries
facing the same problems. In other words, it is reasonable
to suppose that countries like Brazil, i.e., with high regional
inequalities, tend to produce similar results when develop-
ing infrastructure investment. The robustness of the results
was achieved through systematic sensitivity analysis. All the
results can be considered robust in relation to the main
model parameters, i.e., the elasticity of substitution
between domestic regions and the ratio between invest-
ment and capital. In other words, all the results were not
sensitive to these parameters.
The results have shown a typical problem of equity ver-
sus economic growth. It is worth noting that we discussed
regional inequality and not income inequality. For this
reason, we cannot suggest, for instance, policies based on
income transfer. However, the institutional framework of
sectorial public policies seems to be an appropriate mech-
anism to try to minimize this problem of equity. Further-
more, to increase its effectiveness, sectors with greater
linkages and with major impacts on the northeastern pro-
ductive structure could also be encouraged.
New investment in refining in the Brazilian Northeast
would require skilled labour at both the construction and
operational stages. It would therefore be important for
the local labour market to be able to meet this demand.
In this regard, through partnerships with various agencies,
the government could develop coordinated public policies
focused on training the local labour force, especially
through the provision of professional technical courses
and providing scholarships.
Another aspect associated to public policies is to encou-
rage the development and/or strengthening of consumer
markets at the regional level, mainly because the results
indicate a weak regional market as the impact on employ-
ment and household consumption were relatively small.
Insofar as the consumer market is structured in the North-
east, the impacts of large industrial investment could be
better absorbed by the region. Furthermore, it is essential
that the Northeastern production chains are strengthened
in order to minimize the employment and income spil-
lovers that occur in part due to the purchase of inputs
from other regions.
To sum up, from the results, one can draw some impor-
tant conclusions for the preparation and driving of public
policies: equity; professional training; the creation of
regional consumer markets; and the strengthening of
Northeastern production chains.
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the
authors.
SUPPLEMENTAL DATA
Supplemental data for this article can be accessed at http://
dx.doi.org/10.1080/00343404.2017.1327709
NOTES
1. Located in the coast of Southeast region, between the
states of Espírito Santo and Santa Catarina.
2. It appears as Appendix A in the supplemental data
online.
3. The investment and capital stock accumulation
following mechanisms and intersectorial shift according to
pre-established rules based on depreciation and return rates.
4. For more details, see Wittwer and Horridge (2010).
5. The baseline scenario simulations endogenously gener-
ate investment volumes needed for the sectors growth,
causing the capital stock to grow throughout the scenario.
However, new investments, which escape of the logic of
the model’s economic structure, are not produced
without the specific shocks being placed on the scenario,
such as the refineries.
6. Here a hypothesis in the model closure was adopted.
Since the sectorial employment of the oil refining had a
very large drop in relation to the capital stock, it was con-
sidered constant in the regions of refineries’location from
the period in which they suffered the shocks.
7. It is important to highlight that the model cannot
demonstrate the realization of these economies. The idea
was just an attempt to make a link between this kind of
result (generated by a CGE model) with the theory of
agglomeration economies.
REFERENCES
Azzoni, C. R. (2001). Economic growth and regional income
inequality in Brazil. Annals of Regional Science,35, 133–152.
doi:10.1007/s001680000038
12 Luiz Carlos de Santana Ribeiro et al.
REGIONAL STUDIES
Baer, W. (2007). Brazilian economy: Growth and development (6th
Revd Ed.). Boulder: Lynne Rienner.
Baer, W., & Geiger, P. P. (1978). Industrialização, urbanização e a
persistência das desigualdades regionais no Brasil. In W. Baer,
P. P. Geiger, & P. R. Haddad (Eds.), Dimensões do desenvolvi-
mento brasileiro. Rio de Janeiro: Campus.
Baldwin, R., Forslid, R., Martin, P. G. O., & Robert-Nicoud, F.
(2003). Economic geography and public policy. Princeton:
Princeton University Press.
Banco Central do Brasil. (2015). Relatório de Inflação,17(3), 1–125.
Carvalho, A., Somik, L., & Timmins, C. (2006). Regional subsidies
and industrial prospects of lagging regions. Washington, DC:
World Bank. http://doi.org/10.1596/1813-9450-3843
de Araújo, T. B. (2004). Northeast, northeasts: What northeast?
Latin American Perspectives,31,16–41. doi:10.1177/
0094582X03261184
Diniz, C. C. (1993). Desenvolvimento poligonal no Brasil: Nem
desconcentração nem contínua polarização. Nova Economia,
31(1), 35–64.
Diniz, C. C. (2006). A busca de um projeto de nação: O papel do
território e das políticas regional e urbana. Revista Economia,
7(4), 1–18.
Dixon, P. B., Parmanter, B. R., Sutton, J., & Vincent, D. P. (1982).
ORANI, a multisectoral model of the Australian economy.
Amsterdam: North-Holland.
Domingues, E. P., Betarelli, A. A., & Magalhães,A. S. (2011). Quanto
vale o show? Impactos econômicos dos investimentos da copa do
mundo de 2014 no Brasil. Estudos Econômicos (São Paulo),41(2),
409–439. doi:10.1590/S0101-41612011000200008
Domingues, E. P., Betarelli, A. A., Magalhães, A. S., Oliveira, H. C.,
& Valladares, L. M. (2009). Calibragem do modelo ORANIG para
os dados da matriz de insumo-produto nacional (2005) (Mimeo).
Belo Horizonte: CEDEPLAR/UFMG.
Domingues, E. P., Magalhães, A. S., & Faria, W. R. (2009).
Infraestrutura, crescimento e desigualdade regional: Projeção
dos impactos dos investimentos do Programa de Aceleração do
Crescimento (PAC) em Minas Gerais. Pesquisa e Planejamento
Econômico,39(1), 121–158.
Ferreira, P. (2004). Regional policy in Brazil: A review. Retrieved
February 3, 2016, from http://www.fgv.br/professor/ferreira/
RegionalPolicyFerreira.pdf
Fujita, M., Krugman, P. R., & Venables,A. (1999). The spatial economy:
Cities, regions and international trade. Cambridge, MA: MIT Press.
Fujita, M., & Thisse, J.-F. (2002). The economics of agglomeration.
Cambridge, MA: Cambridge University Press.
Furtado, C. (1972). Análise do modelo brasileiro. Rio de Janeiro:
Civilização brasileira.
Giesecke, J. A., & Madden, J. R. (2013). Regional computable gen-
eral equilibrium modeling. In P. B. Dixon & D. W. Jorgenson
(Eds.), Handbook of computable general equilibrium modeling,
Vol. 1A (pp. 379–475). Amsterdam: Elsevier.
Guilhoto, J. J. M., Azzoni, C. R., Ichihara, S. M., Kadota, D. K., &
Haddad, E. A. (2010). Matriz de insumo-produto do Nordeste e
estados: Metodologia e resultados. Fortaleza: Banco do Nordeste
do Brasil.
Guimarães Neto, L. (1997). Desigualdades e políticas regionais no
Brasil: Caminhos e descaminhos. Planejamento e Políticas
Públicas,35,41–99.
Haddad, E. A. (1999). Regional inequality and structural changes:
Lessons from the Brazilian experience. Aldershot: Ashgate.
Haddad, E. A., & Haddad, P. R. (2010). Major sport events and
regional development: The case of the Rio de Janeiro 2016
Olympic Games. Regional Science Policy,2(1), 79–95. doi:10.
1111/j.1757-7802.2010.01017.x
Haddad,E.A.,Hewings,G.J.D.,Perobelli,F.S.,&dosSantos,
R. A. C. (2010). Regional effects of port infrastructure: A
spatial CGE application to Brazil. International Regional
Science Review,33(3), 239–263. http://doi.org/10.1177/
0160017610368690
Haddad,E. A., Perobelli, F. S., Domingues,E. P., & Aguiar, M.(2011).
Assessing the ex ante economic impacts of transportation infrastruc-
ture policies in Brazil. Journal of Development Effectiveness,3(1), 44–
61.http://doi.org/10.1080/19439342.2010.545891
Hirschman, A. O. (1958). The strategy of economic development. New
Haven and London: Yale University Press.
Horridge, J. M. (2012). The TERM model and its database. In G.
Wittwer (Ed.), Economic modeling of water (pp. 13–35).
London: Springer.
Jacobs, J. (1969). The economy of cities. New York: Random House.
Johansen, L. (1960). A multisectoral model of economic growth.
Amsterdam: North Holland.
Lima, J. P. (1994). Economia do Nordeste: Tendências das áreas
dinâmicas. Análise Econômica,12(22), 55–73.
Markusen, J. R. (2004). Multinational firms and the theory of inter-
national trade. Cambridge, MA: MIT Press.
Marshall, A. (1890). Principles of economics. London: Macmillan.
Melo, L. M. C. d., & Simões, R. F. (2011). Desigualdade econômica
regional e spillovers espaciais: evidências para o nordeste do
Brasil. Revista Econômica do Nordeste,42(1), 9–24. Retrieved
from http://ideas.repec.org/p/cdp/texdis/td364.html
Ministério de Minas e Energia (MME), Brazil. (2010). Balanço
energético nacional. Retrieved from https://ben.epe.gov.br
Mújica, O. J., Vázquez, E., Duarte, E. C., Cortez-Escalante, J. J.,
Molina, J., & Barbosa da Silva, J. (2014). Socioeconomic inequal-
ities and mortality trends in BRICS, 1990–2010. Bulletin of the
World Health Organization,92(6), 405–412. http://doi.org/10.
2471/BLT.13.127977
Partridge, M. D. M. D., & Rickman, D. S. D. S. (2010).
Computable general equilibrium (CGE) modelling for regional
economic development analysis. Regional Studies,44(10), 1311–
1328. http://doi.org/10.1080/00343400701654236
Perobelli, F. S. (2004). Análise das interações econômicas entre os estados
brasileiros. 2004 (Tese de Doutorado). São Paulo: IPE/USP.
Perobelli, F. S., Cardoso, V. L., Vale, V. A., & Rodrigues, L. C.
(2015). Localização do setor de serviços e sua relação com questões espa-
ciais no Brasil: Uma análise a partir do Censo Demográfico de 2010.
Perroux, F. (1967). Economia do século XX. Porto: Herder.
Ribeiro, L. C. S., Lopes, T. H. C. R., Simões, R., & Moreira, T. M.
(2013). Suape: Novo polo de crescimento? Novos Cadernos
NAEA,16(1), 29–60.
Shankar, R., & Shah, A. (2001). Bridging the economic divide within
nations: A scorecard on the performance of regional development policies
in reducing regional income disparities. Retrieved from http://
elibrary.worldbank.org/doi/pdf/10.1596/1813-9450-2717
Silveira Neto, R. D. M., & Azzoni, C. R. (2011). Non-spatial gov-
ernment policies and regional income inequality in Brazil.
Regional Studies,45(4), 453–461. http://doi.org/10.1080/
00343400903241485
Silveira-Neto, R. M., & Azzoni, C. R. (2012). Social policy as
regional policy: Market and nonmarket factors determining
regional inequality. Journal of Regional Science,52(3), 433–450.
http://doi.org/10.1111/j.1467-9787.2011.00747.x
Simões, R., & Freitas, E. (2014). Urban attributes and
regional differences in productivity: Evidence from the external
economics of Brazilian micro-regions from 2000–2010. Journal
of Economic and Financial Studies,2(1), 30–44. doi:10.18533/
jefs.v2i01.51
Viana,F. D. F., Domingues,E. P., & Diniz, C. C. (2014).Infraestrutura
turistica no Nordeste: Uma análise de projeção de impactos do
Programa de Desenvolvimento do Turismo do Nordeste –
PRODETUR/NE II. Revista Econômica do Nordeste,45(2), 54–69.
Wittwer, G., & Horridge, M. (2010). Bringing regional detail to a
CGE model using census data. Spatial Economic Analysis,5(2),
229–255. http://doi.org/10.1080/17421771003730695
Structuring investment and regional inequalities in the Brazilian Northeast 13
REGIONAL STUDIES
