Abstract
1. Measuring potential output involves a high degree of uncertainty, especially in economies in the midst of transition such as Romania. The output gap serves as an important indicator of a sustainable, non-inflationary growth path and allows assessing macroeconomic policies against it. Potential output and the output gap are unobserved variables and are thus estimated with considerable uncertainty. This uncertainty is especially large in Romania, a country in the midst of transition. The Romanian economy underwent substantial structural changes since the beginning of transition. Political and economic reforms spurred an uptake in economic growth, fueled by foreign investment and technology transfer. At the same time, the declining manufacturing sector pushed large numbers of workers out of the labor force, thus depressing employment rates.
1. Measuring potential output involves a high degree of uncertainty, especially in economies in the midst of transition such as Romania. The output gap serves as an important indicator of a sustainable, non-inflationary growth path and allows assessing macroeconomic policies against it. Potential output and the output gap are unobserved variables and are thus estimated with considerable uncertainty. This uncertainty is especially large in Romania, a country in the midst of transition. The Romanian economy underwent substantial structural changes since the beginning of transition. Political and economic reforms spurred an uptake in economic growth, fueled by foreign investment and technology transfer. At the same time, the declining manufacturing sector pushed large numbers of workers out of the labor force, thus depressing employment rates.
2. Economic activity boomed in Romania in the run-up to the global crisis but contracted sharply thereafter. Decompositions suggest that Romania’s growth acceleration since 2000 was fueled mainly by improvements in TFP and, in the run-up to the global crisis, increasingly by a foreign financed credit and investment boom.2 With the onset of the crisis, the boom came to a halt and unemployment increased, accompanied by a sharp contraction in TFP. Going forward, staff expects growth to be driven by a recovery of domestic investment, fueled by inflows of EU funds, and a resumption of productivity growth. However, medium term growth projections have been revised down from 4 percent to 3.5 percent due to delays in structural reforms and limited capacity to absorb EU funds.
Romania: Growth Decomposition 2000–2017
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff calculations.Romania: Growth Decomposition 2000–2017
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff calculations.Romania: Growth Decomposition 2000–2017
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff calculations.3. This note uses a variety of techniques to determine the extent to which Romania’s growth potential was affected by the crisis and how fast it can be expected to recover. Based on a sample of 88 banking crises over the past four decades, the October 2009 World Economic Outlook concludes that output typically does not return to its old trend path following a financial crisis. The reason is that balance sheet effects and significant declines in production factors leave lasting scars: first, falling employment rates translate into lower labor force participation or lasting increases in structural unemployment through hysteresis effects (Ball, 2009); second, limited access to finance hampers investment needed to maintain and upgrade the existing capital stock; and third, business failures and skill depreciation among the unemployed render obsolete parts of productive capacity.3 This note uses data for the period 2000:Q1–2017:Q4 to estimate potential output and the output gap based on three techniques, a univariate Hodrick-Prescott (HP) filter, a production function (PF) approach and the multivariate (MV) filter described in Benes et al (2010).
4. The multivariate filter proposed in Benes et al (2010) has important advantages over more conventional approaches and is thus chosen as a benchmark model. The HP filter and the PF approach are perhaps the most popular techniques used to estimate potential output and the output gap, partly due to their simplicity. However, estimates of potential that are based on HP filtering—either of output itself or of its production factors—have known deficiencies related to end-point estimates. Moreover, these filters ignore relevant information from structural economic relationships. For instance, a period in which inflation is low and stable would likely be one in which output is close to potential, while falling inflation rates suggest a negative output gap. Estimates of trend GDP that ignore the decline in inflation may thus understate potential. The MV filter proposed in Benes et al (2010) could serve to avoid such pitfalls.4 It uses Bayesian techniques to simultaneously estimate key macroeconomic relationships. The output gap is inferred as a common factor that drives variables sensitive to the cycle, such as inflation, unemployment and capacity utilization.
5. Both the HP filter and the PF approach suggest that the global crisis resulted in a large drop in potential growth. The HP filter and the PF approach provide similar estimates of potential output, suggesting that it grew at a pace of around 5–6 percent in pre-crisis years in spite of a shrinking labor force. The initially negative output gap turned positive in 2006, reaching a maximum of about 6–7 percent of potential in 2008. When the crisis hit, actual output dropped below potential with a low point of -3 to -4 percent of in 2010. Going forward, potential growth is projected to recover only gradually and is outpaced by projected growth for some years as the Romanian economy bounces back from the crisis and inflows of EU funds boost investment. Both techniques project the output gap to close around 2014/15 and medium term potential growth to be around 3.5 percent.
Potential GDP Levels
(BN 2005 LEI, SA)
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.Potential GDP Levels
(BN 2005 LEI, SA)
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.Potential GDP Levels
(BN 2005 LEI, SA)
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.6. The MV filter predicts a larger output gap in the medium term. The estimates imply a smaller drop in potential growth on impact of the crisis than suggested by the HP filter and the PF approach. Likewise, the output gap is larger and more persistent throughout the medium term. After reaching a low point of -4 percent of potential in 2010, it shrinks only marginally until 2012 and remains negative throughout our projection horizon. Compared to previous estimates, however, these estimates suggest a somewhat less persistent output gap in the medium term. Estimates presented as part the fifth review of the current SBA predicted an output gap of almost 5 percent of potential for 2012 and projected it to remain negative throughout the medium term, reaching -2.5 percent of potential in 2017. The smaller output gap under current estimates also has implications for the assessment of the fiscal policy stance. In particular, the MV filter predicts a somewhat smaller cyclically adjusted fiscal retrenchment following the crisis. The structural deficit is projected to stabilize around 1 percent in the medium term.
Potential GDP Growth
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.Potential GDP Growth
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.Potential GDP Growth
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.
Output Gap
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.Output Gap
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.Output Gap
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.7. The estimates suggest that the crisis left lasting scars in the economy and potential growth will take time to recover. The MV approach predicts potential growth to remain below 2 percent until 2014 and to increase only gradually to 3.1 percent by 2017. This slow recovery suggests that the crisis left lasting scars in the Romanian economy. In particular, the unexpectedly strong growth rate in 2011 likely reflected temporary factors such as the exceptional harvest rather than a recovery of output potential. Pre-crisis potential growth rates of 5 or 6 percent per year will be difficult to attain in the absence of major reforms that bring more people to work and attract more investment.
Conventional Structural Fiscal Balance
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.Conventional Structural Fiscal Balance
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.Conventional Structural Fiscal Balance
Citation: IMF Staff Country Reports 2012, 291; 10.5089/9781475560381.002.A001
Sources: WEO; Haver; and IMF staff estimates.New And Old Estimates Of Potential Growth And The Output Gap
New And Old Estimates Of Potential Growth And The Output Gap
| 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Output Gap | |||||||||||||||||
| New (MV approach) | −2.4 | −3.0 | −1.1 | −1.6 | 0.9 | 3.0 | 6.2 | −1.1 | −4.1 | −3.8 | −4.0 | −3.5 | −2.8 | −2.3 | −1.8 | −1.5 | |
| Old (5th Review) | 0.6 | 0.9 | 4.2 | 3.4 | 6.2 | 7.6 | 10.0 | −0.3 | −3.6 | −3.6 | −4.8 | −4.6 | −4.1 | −3.5 | −3.0 | −2.5 | |
| Potential Growth | |||||||||||||||||
| New (MV approach) | 4.7 | 5.9 | 6.0 | 5.5 | 4.8 | 4.3 | 3.4 | 1.3 | 1.2 | 1.8 | 1.4 | 1.8 | 2.2 | 2.7 | 2.9 | 3.1 | |
| Old (5th Review) | 4.5 | 5.0 | 5.0 | 5.0 | 5.0 | 5.0 | 5.0 | 3.0 | 1.5 | 2.5 | 2.0 | 2.9 | 3.2 | 3.4 | 3.5 | 3.5 | |
| Conventional Structural Fiscal Balance | |||||||||||||||||
| New (MV approach) | −1.8 | −1.3 | −3.0 | −0.2 | −1.6 | −4.2 | −7.1 | −6.8 | −4.9 | −2.8 | −0.8 | −0.6 | −0.7 | −0.8 | −1.0 | −1.1 | |
| Old (5th Review) | −2.8 | −2.5 | −4.8 | −1.8 | −3.4 | −5.8 | −8.5 | −7.2 | −5.1 | −2.8 | −0.6 | −0.2 | −0.3 | −0.4 | −0.6 | −0.8 | |
New And Old Estimates Of Potential Growth And The Output Gap
| 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Output Gap | |||||||||||||||||
| New (MV approach) | −2.4 | −3.0 | −1.1 | −1.6 | 0.9 | 3.0 | 6.2 | −1.1 | −4.1 | −3.8 | −4.0 | −3.5 | −2.8 | −2.3 | −1.8 | −1.5 | |
| Old (5th Review) | 0.6 | 0.9 | 4.2 | 3.4 | 6.2 | 7.6 | 10.0 | −0.3 | −3.6 | −3.6 | −4.8 | −4.6 | −4.1 | −3.5 | −3.0 | −2.5 | |
| Potential Growth | |||||||||||||||||
| New (MV approach) | 4.7 | 5.9 | 6.0 | 5.5 | 4.8 | 4.3 | 3.4 | 1.3 | 1.2 | 1.8 | 1.4 | 1.8 | 2.2 | 2.7 | 2.9 | 3.1 | |
| Old (5th Review) | 4.5 | 5.0 | 5.0 | 5.0 | 5.0 | 5.0 | 5.0 | 3.0 | 1.5 | 2.5 | 2.0 | 2.9 | 3.2 | 3.4 | 3.5 | 3.5 | |
| Conventional Structural Fiscal Balance | |||||||||||||||||
| New (MV approach) | −1.8 | −1.3 | −3.0 | −0.2 | −1.6 | −4.2 | −7.1 | −6.8 | −4.9 | −2.8 | −0.8 | −0.6 | −0.7 | −0.8 | −1.0 | −1.1 | |
| Old (5th Review) | −2.8 | −2.5 | −4.8 | −1.8 | −3.4 | −5.8 | −8.5 | −7.2 | −5.1 | −2.8 | −0.6 | −0.2 | −0.3 | −0.4 | −0.6 | −0.8 | |
8. Returning to pre-crisis rates of potential growth will require decisive implementation of an ambitious structural reform agenda. Since the last Article IV consultation, Romania has fallen behind many of its peers in terms of perceptions of corruption, ease of doing business, and competitiveness: the healthcare and education systems are underfunded and of relatively low quality; the business climate could be more welcoming, and the energy and transport sectors are still dominated by inefficient state-owned enterprises. Achieving higher rates of potential growth will require decisive implementation of an ambitious reform agenda. Over the past two years, pension and labor market reforms were put in place as well as measures to improve the regulatory and pricing framework for the energy sector. However, significant reform gaps remain and progress has been slow. Going forward, the reform agenda should focus on improving institutional and regulatory quality, modernizing the healthcare system as well as reforming the energy and transport sectors, and state-owned enterprises. Measures to protect the most vulnerable are also important. Furthermore, the authorities should step up the absorption of EU funds to unlock a higher growth potential.
Statistical Appendix
Hodrick-Prescott (HP) filter: The smoothing parameter is set equal to 1600 as is standard for quarterly data.
Production Function: We assume a Cobb-Douglas form for the production function where output is given by γ = A × Lα × K1-α where A is TFP, L is employment and K is the capital stock. We use the perpetual inventory method to determine the initial capital stock and set the depreciation rate to 0.05. The existing capital stock is assumed to equal potential. The labor share is assumed to be α = 0.7, and the HP filter is used to provide estimates of potential employment and TFP. The smoothness parameter for HP filtering is set to 1600.
Multivariate filter: We use quarterly data for growth, core inflation, inflation expectations, capacity utilization and the unemployment rate to estimate the model described in Benes et al (2010) by Regularized Maximum Likelihood. The model was calibrated to ensure reasonable smoothness of potential output growth, and assumes a steady-state growth of 3.7 percent, a steady state unemployment rate of 6 percent, and a labor share in output of 70 percent.
References
Ball, L. (2009). “Hysteresis in Unemployment: Old and New Evidence,” NBER Working Paper 14818.
Benes, J., K. Clinton, R. Garcia-Saltos, M. Johnson, D. Laxton, P. Manchev and T. Matheson (2010). “Estimating Potential Output with a Multivariate Filter,” IMF Working Paper 10/285.
Prepared by Christian Saborowski.
The growth decomposition is based on the production function approach discussed below.
In standard production function approaches, the latter effect would remain unexplained and thus be reflected in total factor productivity (TFP).
Indeed, for much of the period under consideration, the Romanian central bank was successfully fighting inflation, suggesting that the link between inflation and the output gap should be taken into account when estimating potential growth. However, this link is likely to be less stable in the context of a transition economy in which inflation is volatile and the economy experiences major structural transformations.
