Blog Name

Brazil's Potential Output: A Case of Cognitive Dissonance?



The International Monetary Fund's most recent World Economic Outlook (WEO) report (IMF 2015) confirms many economists' continued low expectations for Brazil's potential output and growth.

Although estimating potential growth is subject to well-known pitfalls, the importance of such estimates in the policymaking process requires them, however imperfect. As the recent WEO pointedly underscores, the economic definition of potential output requires a framework based on economic theory, not historical averages or econometric filters that characterize various measures of trend output. Typically, estimates of potential output are based on specifications for the country's production function, including the physical capital stock, education and measures of human capital, demographics, the total labor force, and productivity, among other factors.

The economic definition of potential output also differs from the widely used concept of "sustainable" output. Potential output measures how much an economy can grow without igniting undue inflationary or deflationary pressures, while most uses of "sustainable" output refer to an economy's capacity to grow while maintaining macroeconomic stability, more broadly defined. If a country is growing at or close to its potential rate due to rapid expansion in public sector credit and other stimulus policies while experiencing large capital inflows that appreciate the currency and keep inflation at bay, its growth may not be sustainable; once capital flows subside, macroeconomic imbalances will force a painful adjustment process. This is a simple but fairly accurate description of Brazil's growth trajectory during much of the recent period (2004–15).

Although potential output estimates are subject to statistical and model uncertainty and may vary according to the applied methodology, they generally deliver qualitatively similar results when the time horizon used is broadly the same.

Table 1 Variations in Potential Growth Estimates for Brazil (percent)
Barbosa-Filho 2011 (2) 3.5
Bacha and Bonelli 2012 (1) 4.0
IMF 2013 (3) 3.5
OECD 2013 (1) 3.7
Bacha and Bonelli 2014 (1) 3.3
IMF 2015 (4) 2.3
  1. Unspecified time horizon
  2. Estimates for the next ten years (2012–22)
  3. Estimates for 2010 to 2030
  4. Current WEO database estimates for 2015 to 2019

Sources: Bacha and Bonelli (2012, 2014), Barbosa-Filho 2011, IMF (2013, 2014), OECD 2013.

Table 1 shows that, in line with the WEO's findings, different methodologies and specifications for the production function yield broadly similar results. However, the table also suggests that the starting point matters: If the analysis is undertaken during or shortly following a period of robust growth, estimates tend to be more favorable, whereas if it is done during harsher economic times, potential growth estimates may well suffer.1 Although the Brazilian economy was already decelerating substantially between 2011 and 2013, average growth during the period was still above 3 percent. During the second half of 2014, and into the first quarter of 2015, growth stagnated on average—in the first three months of 2015, the economy contracted 0.9 percent on a cumulative basis.2 These developments may have driven the IMF (2015) potential growth projection of 2.3 percent, in sharp contrast to the recent literature and to its own estimates made only two years ago.

Interestingly, potential output estimates by market participants suffer a similar bias. Figure 1 shows the evolution of median forecasts of real GDP growth four years ahead from the Central Bank of Brazil's weekly Focus Survey. Provided that analysts expect the output gap to close within a four-year time horizon, this measure may serve as a reasonable proxy for potential output projections. While projections for four years ahead remained close to 3 percent even late into 2014, they have since plummeted to only 2.3 percent in 2015, underscoring the impact of current conditions on medium-term market forecasts. In addition, the evolution of estimates during Brazil's boom years—from 2003 to 2010—highlight the optimism bias identified by Ho and Mauro (2014): Estimates of potential output during this period reflected not the deep structural problems that have plagued the economy for years but a misguided perception that these problems would be resolved as long as growth could be sustained by substantial external windfalls.


Where does this leave us? Global growth is currently mired in uncertainty. As recent IMF research shows, the financial crisis had a substantial and lasting effect on potential growth across the globe. While it may be reasonable to assume that potential growth is destined to disappoint both in advanced and emerging economies, the psychological discomfort generated by any incoming conflicting information should not sway estimates. The case of Brazil shows just how much the constant struggle to reduce mental stress from clashes between beliefs and information can set policies on a very painful path.


Bacha, E. B, and R. Bonelli.2012. Accounting for the Rise and Fall of Brazil's Post WWII Growth. Discussion Paper no. 25. Rio de Janeiro: Institute for Economic Policy Studies (

Bacha, E. B and R. Bonelli.2014. Post-scriptum to "Accounting for the Rise and Fall of Brazil's Post WWII Growth." Discussion Paper no. 25. Rio de Janeiro: Institute for Economic Policy Studies (

Barbosa-Filho, F. 2011. Uma Estimativa do Produto Potencial no Brasil, Economia e Tecnologia 27: 5–16.

Ho, Giang, and Paulo Mauro. 2014. Rapid Growth in Emerging Markets and Developing Economies: Now and Forever? Policy Brief 14-26. Washington: Peterson Institute for International Economics.

IMF (International Monetary Fund). 2013. How Fast Can Brazil Grow? Selected Issues, Brazil 2013 Article IV Consultation.Washington.

IMF (International Monetary Fund). 2015. Where Are We Headed? Perspectives on Potential Output. World Economic Outlook. Washington.

OECD (Organization for Economic Cooperation and Development). 2013. Economic Outlook Report for 2013. Paris.

1. This resonates with recent research by Ho and Mauro (2014) (see PIIE Policy Brief 14-26 Rapid Growth in Emerging Markets and Developing Economies: Now and Forever? ), who find that economic projections tend to have an optimism bias. The bias that we identify in Brazil's potential output projections reflects current sentiment and economic conditions, which may still render them overly optimistic.

2. That is, growth of GDP measured as the average of four quarters over the average of the previous four quarters.

More From

Related Topics