POLICY BRIEF
IMPACTS OF THE CONSTRUCTION
OF HYDROPOWER PLANTS ON
DEFORESTATION IN THE AMAZON
NEW STUDY SHOWS THAT BUILDING PLANTS
DOES NOT ALWAYS STIMULATE DEFORESTATION
Brazil faces an ongoing debate about the construction of hydropower plants in the Amazon,
which often environmentally and socially disrupts areas where they are built, resulting in
deforestation, migration, and the flooding of forests and properties. On the other hand, the
power generated by hydropower plants provides an essential part of Brazil’s energy matrix,
producing about 60% of the electricity currently used in the country.
For over a decade, Brazil has concentrated policy efforts on curbing deforestation. Thus, the
development of infrastructure projects in the Amazon raises an important question: does
the arrival of hydropower plants drive deforestation? Not necessarily. Analysis conducted
by researchers at Climate Policy Initiative (CPI)1, with INPUT and in cooperation with the
Brazilian Development Bank (BNDES), shows diverse results for 10 plants built between
2003 and 2011 in the Legal Amazon. This region has been on the forefront of the policies
to combat and prevent deforestation2 (Figure 1). While some hydropower plant projects
stimulate deforestation, others help to avoid it.
The study results show how important it is for policymakers to understand the details
of specific plants, as well as how behaviors and circumstances lead to such different
consequences. This understanding is crucial to formulating public policies that ally economic
development and environmental protection3.
1 For the complete study, access: http://www.inputbrasil.org/publicacoes/estudo-efeitos-locais-de-hidreletricas-no-brasil/?lang=en
2 Two other criteria guided the analysis: the plant must have received BNDES financing and must be located more than 100 km away
of any other plant.
3 CPI researchers also analyzed socioeconomic local effects of construction of hydropower plants in Brazil: http://www.inputbrasil.org/
publicacoes/efeitos-socioeconomicos-locais-da-construcao-de-hidreletricas-no-brasil/?lang=en
1
Figure 1: Hydropower plants built in the Amazon between 2003 and 2011
Ferreira Gomes
Sto. A. do Jari
Belo Monte
Estreito
Jirau
Santo Antônio
Dardanelos
Teles Pires
Colíder
São Salvador
Group 1 - Hydropower plants that stimulated deforestation
Group 2 - Hydropower plants that avoided deforestation
Belo Monte dam
Accumulated deforestation between 2001 and 2014
Note: The figure shows all hydropower plants studied, by those that caused and avoided deforestation, respectively. The
figure also shows the accumulated deforestation between 2001 and 2014. It can be seen that hydropower plants are
generally located in areas with high deforestation rates, along the so-called “Arc of deforestation” but they are not necessarily
one of its drivers. As such, for a better understanding of the impact caused by these plants, it is necessary to understand
deforestation trends before construction.
2
MAIN
RESULTS
• Between the beginning of construction of the
• Results on the impact of the plants are very uneven.
hydropower plants in the study’s sample and
While in some cases construction led to more
2013, just over one million hectares of forest were
deforestation, in other regions, forest clearing would
cleared in a 100 km radius around the 10 plants.
have been higher without the construction. For
Of these, the plants directly contributed to the
instance, around the Santo Antônio and Jirau dams,
clearing of 222,000 hectares. This means that the
the loss of forest was two times higher than what
contribution of the 10 plants to the total amount of
would have occurred had they not been built. Yet,
deforestation is lower than the combination of all
near the Teles Pires dam, deforestation was 40%
other factors: only 22% of deforested area around
lower than what would have occurred in the dam’s
all the plants together can be attributed to their
absence.
construction.
ANALYSIS FOR
POLICY MAKERS
The CPI study identified two different patterns around the construction of hydropower plants
in the Amazon: while some plants stimulate deforestation, others prevent deforestation that
would have occurred in the areas where they were built. The wide variation shown in the
cases studied illustrates that discussions about hydropower plants cannot be reduced to
broad evaluations of the impact of these developments on deforestation.
The different degrees of deforestation that occurred because of the dam construction suggest
that nuances need to be incorporated in the licensing procedures of future projects and,
more generally, in related public policies. Only through a deeper understanding of conditions
associated with the construction of hydropower plants can policymakers begin to mitigate the
environmental risks and impact associated with the development of these projects.
The results of the two different patterns are summarized below. The deforestation
calculations used in the analysis include forest clearing needed for construction that were
authorized by the government and illegal clearings.
3
Group 1: Santo Antônio and Jirau, Estreito, São Salvador and Colíder
The analysis of the Santo Antônio and Jirau, Estreito, São Salvador and Colíder plants shows
that over 175,000 hectares of forest were cleared in a radius of 100 km of these dams
because of construction.
Of the 10 plants studied, Santo Antônio and Jirau plants triggered the most deforestation.
The analysis shows that forest clearing around these hydropower plants was twice what
would have occurred if they had not been built. Total clearing in a 100 km radius was
208,000 hectares in the first six years after construction began. Of those, 96,000 can be
attributed to the plant.
Researchers at CPI suspect that deforestation in those cases might be due to the dams’
proximity to a large city (Porto Velho), resulting in large migration movements. However, the
data do not allow for verification of this hypothesis.
In Estreito, of the 119,000 hectares cleared, more than 30% are attributed to the construction
of the dam. The São Salvador plant was also responsible for more than 30% of the total 92,000
hectares cleared in the area. Finally, about 11,000 hectares of forest were cut down because of
the construction of Colíder dam, representing 12% of total deforestation in the area.
Group 2: Teles Pires, Ferreira Gomes, Santo Antônio do Jari and Dardanelos
Contrary to what happened with the plants of the previous group, the analysis of Teles Pires,
Ferreira Gomes, Santo Antônio do Jari, and Dardanelos hydropower plants reveals that an
additional 51,000 hectares in a radius of 100 km would have been cleared in the absence of
their construction.
In this group, UHE de Teles Pires dam stands out: an area significantly larger would have
been cleared had the plant not been constructed. As a consequence of the construction, it
is estimated that over 30,000 hectares retained their native vegetation. That means that the
Teles Pires dam reduced the clearing that would have happened by 40%.
All other plants in this group also had a positive balance on deforestation. Ferreira Gomes
avoided 12,000 hectares of forest clearing, a reduction of more than 26% of what would have
occurred; Santo Antônio do Jari avoided over 2,000 hectares (7% reduction); and Dardanelos,
around 3,000 hectares (2% reduction).
This positive balance in the hydropower plants from Group 2 might have happened because
of increased control and monitoring by public institutions of the areas around the plants. This
official presence might have restrained the action of land grabbers and squatters in those
areas who would have likely contributed to much higher forest destruction totals. Yet, only a
case-by-case analysis of these projects could confirm this hypothesis.
4
Figure 2, below, compares the results for Groups 1 and 2, showing the year average of
estimated and observed forest clearing in a radius of 100 km around each plant.
Figure 2: Estimated and observed forest clearing in a radius of 100 km around each plant
Hydropower Plants
Negative numbers indicate that the
presence of the hydropower plant
helped avoid forest clearing and thus
deforestation would have been greater
had the plant not been constructed
Contribution
to observed
deforestation (%)
Santo Antônio e Jirau
54
Estreito
34
São Salvador
32
Colíder
12
Dardanelos
-2
Santo Antônio do Jari
-7
Ferreira Gomes
-26
Teles Pires
-40
0
5
10
15
20
25
30
Thousand hectares
Numbers reflect the yearly average of
deforestation observed and estimated
after the beginning of construction
estimated deforestation in the absence of the plant
observed deforestation
5
THE BELO MONTE DAM
Forest clearing around the Belo Monte
plant occurred in a very particular way
in the period preceding construction
in 2011. Specifically, between 2007
and 2009 there was a large increase
in deforestation in the area around
the place where the plant would be
constructed that is not comparable
with any other in the region during
the same period. Consequently, there
is no appropriate comparison group,
which limited the application of the
methodology to this particular plant.
Promotional photo/ May 2014
CONCLUSION
CPI’s results provide some of the first evidence
and the expansion of agriculture stimulated by the
that isolates the impact that dam construction
development of waterways. Noncompliance with
has on deforestation in the Amazon. This evidence
environmental legislation during the construction
allows researchers and policymakers to take a more
and differences in the levels of compensation
nuanced view of what is happening and calls for an
measures may also explain why some plants
understanding of why the outcomes vary.
contributed to more forest loss.
Part of the associated deforestation from dam
As Brazil plans to expand it hydroelectric power
construction comes from what government officials
capacity, further analysis that examines these
authorize in the licensing procedures. However,
factors and isolates specific causes is necessary for
numerous other causes often make the impact
policymakers to proactively mitigate environmental
more acute, including migration, proximity to urban
damage.
areas and settlement projects, road construction,
6
ABOUT
THE DATA
Methodology
• The study uses a rigorous economics method that allows for the comparison of different
areas and the construction of contrafactuals. The method is known as synthetic control
for compared studies (Abadie, 2010 and 2013). For each area around a hydropower plant,
the researchers used an algorithm that generates a comparison area based on other
locations with hydropower potential where dams have not been built. The comparison
area is designed in a way that mirrors the trajectory of deforestation of the affected area
in the period before construction of the plant. The effect of the plant is calculated as the
difference between the affected area and its respective comparison area.
Limitations
• CPI’s analysis allowed for successful measurement of the impact of new hydropower
plants on deforestation in the Amazon. However, although the researchers have
hypotheses, the study does not allow for an understanding of the mechanisms behind the
deforestation caused by the plants. The analysis is blind to certain local realities that could
only be captured in case studies about each specific plant.
• Due to data availability, the study analyzes only the effects of hydropower plants on forest
clearing. There are other relevant environmental indicators, such as biodiversity and
quality and quantity of water downstream. However, systematic data that would permit an
analysis of the impact of dams on those other indicators, similar to the approach used for
forest coverage, do not exist.
Sources
• Remotely sensed data from Hansen et al. (2013)
• Geographic information system from the Energy Sector (SIGEL), ANEELL
7
AUTHORS
Juliano Assunção
Climate Policy Initiative (CPI) & Núcleo de Avaliação
de Políticas Climáticas da PUC-Rio (NAPC/PUC-Rio),
Department of Economics, PUC-Rio
[email protected]
Dimitri Szerman
Climate Policy Initiative (CPI) & Núcleo de Avaliação de
Políticas Climáticas da PUC-Rio (NAPC/PUC-Rio)
[email protected]
Francisco Costa
Escola Brasileira de Economia e Finanças da
Fundação Getúlio Vargas (EPGE/FGV)
[email protected]
Media Contact
Mariana Campos
[email protected]
www.inputbrasil.org
January/ 2017
The Land Use Initiative (INPUT – Iniciativa para o Uso da Terra) is a dedicated team of specialists who work at the forefront of how to increase
environmental protection and food production. INPUT engages stakeholders in Brazil’s public and private sectors and maps the challenges for
a better management of its natural resources. Research conducted under INPUT is generously supported by the Children’s Investment Fund
Foundation (CIFF) through a grant to the Climate Policy Initiative. www.inputbrasil.org. This study is also the result of a partnership between CPI and
BNDES that contributed with data on hydropower plants as well as technical support and comments.
8