2014年-IMF国际货币组织全球_Estimating_Sri_Lanka’s_Potential_Output_16页_506kb

Summary of "Estimating Sri Lanka's Potential Output"

Core Content

This IMF Working Paper explores various methodologies to estimate Sri Lanka's potential output and output gap. It evaluates univariate statistical filters, multivariate filtering models, the production function approach, and the structural vector autoregression (SVAR) method. The paper emphasizes that model-based approaches, which incorporate multiple observable economic indicators, offer a richer and more accurate interpretation of potential output compared to univariate methods that rely on a single time series.

Main Viewpoints

• Potential Output Estimation is essential for assessing economic performance and guiding macroeconomic policy.
• The univariate filters (Hodrick-Prescott, piece-wise linear trend, Baxter-King, and Christiano-Fitzgerald) provide similar estimates of potential growth but are limited in their ability to attribute changes in growth to specific economic factors.
• Model-based approaches are more informative as they integrate data on inflation, unemployment, and capacity utilization, leading to a more nuanced understanding of the economy's cyclical state.
• The multivariate filter (MV) estimates suggest a slight increase in potential output after the end of the civil conflict in 2009, with potential growth rising from about 5.5% in 2009 to 7% in 2012. This increase is more gradual than that estimated by univariate filters.
• The production function approach aligns with the MV results, estimating potential growth of around 7% by the end of the sample period, and shows a positive output gap in 2012.
• The SVAR approach confirms the general trend but shows slightly more volatility, with a more pronounced slowdown in potential growth during the 2009 global crisis and a faster recovery in 2010-11. It also suggests that potential growth in 2012 was slightly below 7%.

Key Information

I. Introduction

• Estimating potential output is crucial for understanding economic performance and setting macroeconomic policies.
• Sri Lanka's post-conflict growth in 2009-2012 raises questions about whether it reflects a permanent shift in potential output or a temporary demand-driven increase.
• The paper aims to evaluate different estimation techniques to provide a comprehensive view of the country's potential output and output gap.

II. Estimation Methodologies

A. Univariate Statistical Filtering Methods

• Four methods are used: Hodrick-Prescott (HP), piece-wise linear trend (LT), Baxter-King (BK), and Christiano-Fitzgerald (CF).
• These methods are simple and widely used but have limitations, including the inability to incorporate information from multiple variables and the potential for spurious cycles in integrated data.

B. The Multivariate Approach

• Based on the model from Benes et al. (2010), this approach uses multiple economic indicators to estimate potential output.
• It incorporates a Phillips curve and Okun's law to relate inflation, unemployment, and output gaps.
• The capacity utilization gap is also used to improve potential output estimates.
• The model allows for dynamic adjustments and hysteresis effects, where changes in equilibrium unemployment can influence potential output over time.

C. The Production Function Approach

• Uses the Cobb-Douglas production function to estimate potential output based on capital and labor utilization.
• Potential output is calculated using the equilibrium capacity utilization rates and the unemployment rate gap.
• The results are consistent with the multivariate filter, showing potential growth of around 7% by the end of the sample period.

D. Structural Vector Autoregression (SVAR) Approach

• Based on Blanchard and Quah (1989), it distinguishes between permanent and transitory components of output using long-run restrictions.
• The model includes real GDP growth, real credit growth, and inflation as variables.
• It assumes that supply shocks have a permanent impact, while demand shocks are temporary.
• The SVAR results suggest a slightly more volatile potential growth path compared to the multivariate filter.

III. The Data

• The sample period is 1997Q1–2012Q4, using quarterly GDP, unemployment, and inflation data.
• Real GDP growth averaged 6.25% between 2003–2012, with two distinct phases: 5.75% before 2008 and 7.5% after 2009.
• Post-conflict growth was partly driven by increased domestic activity and the availability of factor inputs.
• Capacity utilization data are limited to annual manufacturing data, so a quarterly capacity utilization index is constructed, incorporating both industrial and service sectors.

IV. Estimation Results

A. Univariate Filters

• Potential GDP growth rates are similar across methods.
• A break is detected in 2002 and 2008 for the piece-wise linear trend method.
• Potential growth increased to nearly 7% after the conflict, with the HP filter showing the smoothest trend.
• The output gap was sharply negative in 2009 but turned positive by 2011, reflecting excess demand.

B. Multivariate Filter

• Estimates suggest a gradual increase in potential output after 2009, from 5.5% to 7%.
• The output gap showed a larger and earlier build-up of excess demand compared to univariate results.
• The model incorporates Okun's law, Phillips curve, and capacity utilization to provide a more comprehensive view.

C. Production Function

• Results are similar to the multivariate filter, with potential growth estimated at 7% by the end of the sample.
• The output gap is also positive in 2012, consistent with the multivariate approach.

D. Structural VAR

• Potential growth estimates are similar to the multivariate filter but show more volatility.
• The model indicates a more intense slowdown in potential growth during the 2009 crisis and a faster recovery in 2010–2011.
• It suggests that potential growth in 2012 was slightly below 7%, and the output gap began to decline in mid-2012.

V. Conclusions and Policy Implications

• All methods point to similar business cycle patterns in Sri Lanka's recent economic history.
• The post-conflict growth appears to reflect a permanent upward shift in potential output, supported by the moderation in inflation and the increase in capacity utilization.
• The paper highlights the importance of model-based estimation in capturing the cyclical state of the economy and informing macroeconomic policy decisions.
• The output gap is estimated to be positive in 2012, indicating the economy may have been operating above potential, which could lead to inflationary pressures if sustained.