5. ENERGY AND ECONOMIC DEVELOPMENT
5.2. Evolution of the final energy use pattern of the Brazilian economy
5.2.1. Decomposition analysis of changes in the final energy use of the Brazilian economy
Figure 5.4 shows the evolution of Brazil’s GDP at producer prices and at market prices
Bangladesh Overall primary energy intensity (toe/thousand US $ PPP-1995)
FIG. 5.3. Overall primary energy intensity versus per capita GDP in selected countries and regions of the world in 2000 [5.3].
1970 1974 1978 1982 1986 1990 1994 1998
GDP (producer prices) Taxes and subsidies over products Billion US $ PPP, at constant 2000 prices
FIG. 5.4. Evolution of GDP at producer prices and at market prices (producer prices plus taxes and subsidies over products) [5.11–5.14].
(producer prices plus taxes and subsidies over products).
Figure 5.5 presents indices of final energy use, GDP at producer prices and final energy intensity of the Brazilian economy over the past three decades (unless otherwise stated, final energy figures in this section exclude final energy use by households). Both the final energy use and the GDP at producer prices more than tripled during the 1970–2000 period. Furthermore, final energy use and the GDP at producer prices were highly correlated throughout the 30 year period. As a consequence, the final energy intensity of the economy is fairly stable for the whole period.
These data do not, however, provide sufficient insights about underlying factors and possible future changes [5.16]. Therefore, a more disaggre-gated analysis is useful for assessing issues such as the sectoral contributions to final energy use and to the overall final energy intensity, and the factors affecting sectoral activities and energy efficiencies.
Among the questions that need to be addressed are the following: What circumstances are hidden by the aggregate final energy figures of Brazil? Why is a desirable decoupling trend between final energy use and GDP not occurring in Brazil? What key driving forces would enable Brazil to leapfrog some steps and decouple final energy use and GDP? Is Brazil adopting less efficient energy technologies?
What is happening to the economic structure in Brazil? Are Brazil’s choices, hidden in the country’s final energy intensity path, leading to a sustainable energy future? To answer these questions, a more systematic and disaggregated approach is necessary.
The decomposition technique applied here to the
country’s sectoral final energy data allows energy use changes to be separated into their basic effects (activity, structure and intensity), which helps to identify the major drivers that affect final energy use variations in an economy.
The decomposition of the Brazilian economy’s final energy use changes is performed at two different levels. First, the technique is applied to the final energy use of the economy as a whole, breaking the Brazilian economy down into four sectors: agriculture, industry, services and energy.
Then, a closer examination is made of the industrial and services sectors. The industrial sector is further divided into nine branches: mining and quarrying, non-metallic minerals, iron and steel, non-ferrous metals and metalworking (hereinafter, non-ferrous metals), chemicals, food and beverages, textiles, pulp and paper, and other industries. The services sector is divided into three branches: public (including public administration, and sewage and water), transport and other services (including commerce).
Figure 5.6 presents the evolution of the final energy use of the Brazilian economy by sector. The average growth rate of final energy use was 4.3%
per year during the 1970–2000 period. Short term declines in the general trends are observed for the 1980–1983 and 1989–1992 periods and for the change in pace for the 1998–2000 period. During the whole period analysed, final energy used by the energy sector increased at an average rate of 7.3%
per year, while the industrial and services sectors registered average annual growth rates of 4.3% and 4.7%, respectively. The agricultural sector grew at the average rate of only 1.1% per year in this period.
0 50 100 150 200 250 300 350 400
1970 1974 1978 1982 1986 1990 1994 1998
Final energy use GDP (producer prices) Overall final energy intensity
1970 = 100
FIG. 5.5. Indices of final energy use, GDP at producer prices (US $ PPP-2000) and final energy intensity of the Brazilian economy [5.11–5.15].
Note: Final energy figures do not include final energy use by households.
1000 2000 3000 4000 5000 6000 7000
1970 1974 1978 1982 1986 1990 1994 1998
Agriculture Industry Services Energy
PJ
FIG. 5.6. Final energy use of the Brazilian economy by sector [5.15].
Note: Transport is included in services. Data exclude final energy use by households.
The dynamics of final energy use by sector presented in Fig. 5.6 would be better understood with a simultaneous analysis of the evolution of Brazil’s GDP and its sectoral composition, and of the evolution of the sectoral final energy intensity over the period. However, there are different and even contradictory sectoral trends (structural changes and energy intensity coefficients) that make a direct empirical inspection difficult to perform, resulting in a somewhat confused analysis.
In this sense, it is better to present the findings of the decomposition analysis first, and then discuss the effects on the GDP, sectoral composition and sectoral final energy intensities.
Figure 5.7 shows the results of the final energy use decomposition analysis for the economy from 1970 to 2000 divided into five-year periods. The figure shows the basic effects (activity, structure and intensity) and the term of interaction.2
The activity effect was the most important effect behind energy use changes in the economy throughout the 1970–2000 period, except for the 1985–1990 period. The same holds true for the more detailed decomposition analysis performed for this study: overall economic growth was the most important driver of final energy use changes in Brazil during the 1970–2000 period. However, structural changes and energy intensity improve-ments had smaller but significant effects on energy use during this period — effects that, if understood with respect to their causality and consequences, could guide future decisions about strategies to enhance sustainable energy development in Brazil.
These strategies must consider not only the impact of economic growth on energy use, but also the influence that structural changes and sectoral final energy intensity might have on total energy use.
5.2.1.1. Activity effect
Figure 5.8, which shows the evolution of value added (at producer prices) by sector from 1970 to 2000, helps to explain the results for the activity effect, including the minor influence of such an effect on the total energy use changes in the 1980-1985 and 1985–1990 periods.
The economic performance of Brazil in the 1980–1985 and 1985–1990 periods was strongly impacted by tight economic policies implemented by the Government to deal with the balance of payments crises in 1981–1983 and the exacerbated inflation in the late 1980s, respectively [5.17, 5.18].
Such circumstances strongly influenced the magnitude of the activity effect in the 1980s (a negative environment for investments and business
2 The term of interaction accounts for changes not explained by a single effect alone; that is, it reflects the simultaneous movement of more than one basic factor (activity, structure and intensity). In the divisia index approach used here, it can be either calculated itself by varying two or more factors simultaneously or reckoned as a residue. In any case, unlike other decomposition tech-niques, the term of interaction is not an error residue, but a real effect that cannot be cleanly attributed to one variable or the other. A new aluminium plant in an economy, for instance, would, ceteris paribus, simultane-ously increase the activity level, shift the economic structure towards energy intensive branches and affect the overall final energy intensity of the economy (both by introducing a more efficient technology in the aluminium branch and by changing the sectoral weights that influence the overall final energy intensity). For details, see Ref. [5.10].
1970–1975 1975–1980 1980–1985 1985–1990 1990–1995 1995–2000
PJ
Total change Activity Structure Intensity Term of interaction
FIG. 5.7. Decomposition of energy use changes in the Brazilian economy [5.11–5.15].
1970 1974 1978 1982 1986 1990 1994 1998
Agriculture Industry Services Energy Billion US $ PPP, at constant 2000 prices
FIG. 5.8. Evolution of value added at producer prices by sector [5.11–5.14].
Note: Transport is included in services.
uncertainties) in comparison with other sub-periods. However, the average growth rate of GDP for the entire period was 4.3% per year (the same as the growth rate of final energy use in the Brazilian economy). It is important to note the difference in the economic growth rates in the 1970s, 1980s and 1990s. During the 1970s, the average annual growth rate of GDP was 10.1% in the early part of the decade and fell to around 7.4% by the end of the period. The 1980s and 1990s brought economic crises and diminishing expectations; annual economic growth in these two decades fluctuated between 1.3% and 3.1%, averaging around 2%.
Such differences in economic performance are related not only to international events, such as the oil shocks of 1973 and 1979 and international debt crises (beginning with Mexico’s default in 1982), but also to choices made to deal with the modifications in the market conditions (both international and domestic) and the economic distortions that resulted from those choices.
The 1970s were the time of the ‘Brazilian economic miracle’, an economic boom lasting from 1968 to 1973, and the Second National Development Plan, or II PND, launched in 1974.
During the ‘economic miracle’, Brazil registered impressive economic performance, based first on the easing of a previously tight economic policy (to stimulate economic activity in the face of high idle capacity) and then on high private (until 1975) and public investment rates (see Fig. 5.9). In contrast, the II PND aimed at implementing and/or expanding basic industrial branches that were constraining economic growth and generating relevant deficits in the trade accounts, counting on foreign capital and public debt. In addition, during its implementation, the II PND was supposed to
(and in fact did) contribute to maintaining the high economic performance by offsetting the decline in private investments (due to the new market conditions and the incipient deterioration of the macroeconomic fundamentals) with higher public investment in basic industrial branches and infra-structure, such as roads, power plants and buildings.
[5.19–5.21].
In the 1980s, Brazil’s macroeconomic performance was strongly impacted by the balance of payments crises in 1981–1983, worsening inflation and the tight economic policies implemented by the Government to deal with these two issues [5.17, 5.18]. Delays in dealing with balance of payments problems, changes in US monetary policy in the 1980s, the new international financial environment worsened by Mexico’s default (1982) and delays in changing domestic economic policies led to Brazil’s debt crisis and to the so-called ‘lost decade’ of the 1980s, where a series of financial adjustments and monetary reform plans (Cruzado in 1986, Cruzado II in 1986–1987, Bresser in 1987, Verão in 1989 and Collor in 1990) based on stop-and-go policies failed to re-establish economic stability and exacerbated inflation.
Short term necessity rather than consistent long term development strategy guided public policies in Brazil after the 1980s [5.22]. Even relatively successful policies, such as the consoli-dation of capital and the development of indigenous energy technology and markets, were considerably distorted by arrangements set to address short term needs (underpricing of energy and material resources to control inflation, excessive subsidies, fiscal and domestic industrial protection schemes, etc.).
During the 1990s, the economic adjustments continued with ambitious State reform and a market liberalization strategy, which included the privatization of several State owned companies and market deregulation [5.23]. The privatization and market liberalization programme continued throughout the 1990s, and the Government con-cluded the national debt renegotiations, improved the public accounts and further stabilized inflation with the so-called Real Plan [5.24, 5.25]. Such a new environment was fundamental for raising the economic expectations of both national and foreign investors and for increasing economic growth in Brazil during the 1990s compared with the 1980s.
However, the economic performance of the 1970s was not achieved again.
10 12 14 16 18 20 22 24 26 28
1970 1975 1980 1985 1990 1995 2000
Investments/GDP, at market prices (%)
FIG. 5.9. Investment as a percentage of GDP at market prices (US $ PPP-2000) (based on Ref. [5.12]).
Note: Investment here excludes inventory changes, meaning it considers only gross fixed capital formation.
The privatization and market liberalization programme was deepened and consolidated during the 1998–2000 period [5.23, 5.26]. Although some institutional arrangements in the power system were not successful [5.27, 5.28], several market distortions were reduced and new sectoral regulatory regimes were set (see Chapter 2 for a discussion of energy regulatory reform). Favour-able international market conditions during this time made it possible to attract foreign investment (see Fig. 5.10) and to boost economic growth in Brazil.
However, delays in promoting structural adjustments in the country’s public and foreign accounts left the Brazilian economy vulnerable to external shocks (see Fig. 5.11). Overvaluation of the Brazilian real and successive financial crises worldwide (in Mexico, East Asia and the Russian Federation, and with Brazil’s own devaluation crisis in 1999) contributed to the decline in economic growth throughout the 1990s [5.23].
The bottom line of the analysis of the contri-bution of the activity effect to energy use changes in Brazil is that higher economic growth has led to higher energy use. Nevertheless, structure and intensity effects have either magnified or offset the activity effect. Thus, a fundamental task in deciphering the energy profile of a country is to further analyse how structure and intensity effects evolve.
5.2.1.2. Structure effect
The decomposition in Fig. 5.7 shows that the structure effect for the Brazilian economy as a whole was negligible during the 1970s and in the 1985–1990 period, and modest during the 1995–2000 period. However, the structure effect was relevant during the 1980–1985 and 1990–1995 periods.
Table 5.1 presents the contribution of each sector in
0 1 2 3 4 5 6
1970 1974 1978 1982 1986 1990 1994 1998 2002
FDI/GDP, at market prices (%)
FIG. 5.10. Evolution of the ratio of foreign direct investment (FDI) to GDP at market prices (based on Ref. [5.12]).
-2 -1 0 1 2 3 4 5 6 7 8
1974 1978 1982 1986 1990 1994 1998 2002
0 10 20 30 40 50 60 70
Current account deficit (left axis) Net public debt (right axis)
Percentage of GDP Percentage of GDP
FIG. 5.11. Evolution of net public debt and deficit in current account (based on Refs [5.12, 5.14]). Net public debt = gross public debt – public assets; current account = merchandise account (goods exports and imports) + service account + unilateral (current) transfers. Negative figures mean surpluses in current account in the relevant year.
TABLE 5.1. COMPOSITION OF GDP BY SECTORa (%) [5.11–5.14]
Sector 1970 1975 1980 1985 1990 1995 2000
Agriculture 12.3 11.5 11.0 11.9 8.1 9.0 7.7
Industry 34.6 39.3 39.7 38.9 32.8 31.5 28.9
Servicesb 56.4 52.7 53.0 56.0 70.6 61.0 59.2
Energy 3.5 3.6 4.0 5.9 5.6 4.9 8.3
Financial dummy (FISIM)c –6.9 –7.2 –7.8 –12.7 –17.1 –6.4 –4.1
Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0
a GDP at producer prices based on US $ PPP, at constant 2000 prices.
b Transport is included in the services sector.
c Financial dummy — financial intermediation services indirectly measured (FISIM), an adjustment introduced to prevent double counting (see footnote 3). For detailed information, see Ref. [5.29].
terms of value added to GDP at producer prices.
The trends in energy use in each of these sectors are depicted in Fig. 5.6.
Note that the sum of the sectoral shares exceeds 100% in Table 5.1. This is due to a term identified as the financial dummy, which is a negative share considered in the calculations.3
Table 5.1 shows that the value added share of the industrial sector in GDP increased during the 1970s and decreased during the 1980s and 1990s.
Good overall economic performance, mainly due to the ‘Brazilian economic miracle’ and the II PND, was behind the shift towards the higher industrial share in the economic structure of Brazil in the 1970s. Poor economic growth, price controls on basic materials produced by State owned companies (mainly steel and petrochemicals), industrial com-modities price collapse and changes in relative prices due to inflation in favour of some services (financial services and dwelling rents, for instance) explain the reduction of the industrial sector’s share of GDP in the late 1980s and 1990s.
The share of the agricultural sector registered a downward trend through the period as a whole (except for temporary increases in 1985 and 1995).
The long term decline in agricultural commodity prices, a policy bias towards industry in the 1970s, changes in relative prices in favour of services due to inflation in the 1980s and 1990s (including strict price controls on agricultural products as part of stabilization plans to combat inflation) and the lack of consistent long term policies for agriculture explain the shift away from agriculture in the 1970-2000 period. The exceptions in 1985 and 1995 were mainly motivated by short term policies to increase domestic supply (through financing schemes and fiscal incentives or compensations for price controls to help control inflation) and exports (to deal with imbalances in the balance of payments).
Despite large investments in energy supply (see Fig. 5.12), the contribution of the energy sector
to GDP presented a modest upward trend until 1985, when Government interventions in energy prices became more frequent and generalized debts among power corporations (generators and distrib-utors) eroded profitability (see Chapter 2). The institutional restructuring from 1995 to 2000, including privatization and market liberalization, improved the energy sector’s profitability and value added. Also, during the late 1990s, the increase of petroleum production in Brazil by Petrobras and the company’s strategy of shifting the oil product mix towards higher value added products also contributed to increasing the share of the energy sector in the GDP at producer prices [5.34].
The II PND, launched in 1974, set as a priority target an increase in the production capacity of capital goods, basic materials, energy, transport and communication. The construction of Itaipu (12.6 GW) and Tucuruí (4.0 GW), Brazil’s two largest hydropower plants, and the offshore upstream investment programme of Petrobras started at that time. The investments in capital goods, basic materials and energy promoted by the II PND were consolidated in the early 1980s, explaining the relevant structure effect in the 1980-1985 period. The fall in the industrial and energy sector shares was behind the negative structure effect in energy use in the 1990–1995 period. The energy sector’s share of GDP increased from 4.9% in 1995 to 8.3% in 2000, explaining the positive structure effect in 2000. In summary, structural changes resulting from policies and decisions made as well as economic circumstances have affected the overall energy intensity and energy use of the Brazilian economy.
After a decrease in the 1970s, the contribution of services to GDP increased sharply in the late
3 ‘Financial dummy’ is how the System of National Accounts of Brazil denominates the so-called financial intermediation services indirectly measured (FISIM), which is the value of the services provided by financial intermediaries for which no explicit charges are made [5.29]. The apparent differences between GDP at producer prices in Figs 5.4 and 5.8 arise because the financial dummy is not made explicit in the latter (total value added at producer prices + financial dummy = GDP at producer prices).
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
1970 1974 1978 1982 1986 1990 1994 1998
%
Energy (electricity + oil + alcohol) Electricity Oil Alcohol
FIG. 5.12. Investment in the energy sector as a percentage of GDP at market prices (US $ PPP-2000) (based on Refs [5.12, 5.28, 5.30–5.33]).
1980s and early 1990s (the financial services and dwelling rental branches led to this rise as a result of high inflation). This contribution then decreased as inflation was controlled (the financial dummy is highly correlated to inflation4).
5.2.1.3. Intensity effect
The decomposition of overall energy use presented in Fig. 5.7 shows that the intensity effect was negative during the 1970s and positive through the 1980s and 1990s. In the 1970s, the intensity effect partially offset the activity effect, while during the 1980s and 1990s the intensity effect reinforced the activity effect.
Figure 5.13 shows the evolution of final energy intensity by sector (agriculture, industry, services and energy) and for the economy as a whole. The relative stability of the overall final energy intensity of the Brazilian economy results from complex movements of sectoral growth and many factors that compensate one another. The differences in sectoral final energy intensity trajectories reflect the impact of several factors, including the sectors’ own activity, structural and efficiency changes and conflicting intra-sectoral trends.
Since the industrial and services sectors will be further analysed in later sections of this chapter, only the energy intensities of agriculture and energy are discussed here. The trajectory of the final energy intensity of agriculture can be split into two major
trends [5.35]. From 1970 to 1985, the substitution of modern energy carriers for fuelwood and the intro-duction of more efficient technologies set into motion a downward trend in the agricultural energy intensity indicator. From 1985 to 2000, increased mechanization in the countryside induced an upward trend in the agricultural energy intensity.
Changes in the agriculture product mix and in the
Changes in the agriculture product mix and in the