Abstract
Selected Issues
Total Factor Productivity Growth Slowdown1
After the global financial crisis, Poland’s total factor productivity (TFP) growth has decelerated significantly, falling to less than a half of what it was before the crisis. A combination of external headwinds (weak external demand and FDI) and domestic structural bottlenecks have slowed technology diffusion and exacerbated the drag from allocative inefficiencies in the Polish economy. Given the medium-term global outlook, the TFP growth in Poland will likely remain well below its pre-crisis level in the absence of structural reforms.
A. Recent Trends and Challenges
1. The TFP growth slowdown has been a global phenomenon, especially after the global financial crisis (GFC). A marked and persistent TFP growth slowdown (or even TFP loss in some cases) has contributed to the post-GFC recessions across all income groups. Its negative contribution to growth has been particularly large in emerging markets (EMs) (Adler et al. (2017)).
2. Poland experienced a deceleration in TFP growth as well (Figure 1). Over the last decade or so, Poland enjoyed a higher average TFP than many other EMs. However, its TFP growth has slowed significantly following the GFC, from an annual average of 2.4 percent over 2003–07 to barely 1 percent over 2013–16, although there was a temporary recovery during 2010–12. While the extent of slowdown is broadly in line with that observed in other EMs, it represents a larger drop compared to the EU average.
Poland’s Long Term TFP Developments
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: Penn World Table 9.0, Conference Board, and IMF staff calculations.Poland’s Long Term TFP Developments
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: Penn World Table 9.0, Conference Board, and IMF staff calculations.Poland’s Long Term TFP Developments
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: Penn World Table 9.0, Conference Board, and IMF staff calculations.3. Since the early 2000s, Poland’s TFP growth has been on a steeper declining path than in advanced Europe (Figure 1). Poland’s trend (i.e., HP filtered) TFP growth has declined from above 2 percent in the early 2000s to about zero in 2015, while the EU average trend TFP growth has largely returned to its pre-crisis rate. Hence, the positive TFP growth differential that Poland managed to maintain against the EU average and advanced Europe (EU-152) over the past decade and a half has disappeared. This suggests that the TFP growth slowdown in Poland is not just a natural consequence of income convergence (in which case, the decline would have been smaller than the EU or advanced Europe average), but likely an outcome from a combination of domestic (structural) and external factors.
4. The global TFP growth slowdown began before the GFC, but was compounded by crisis legacies, resulting in a persistent and prolonged global slump in productivity growth. Several long-term structural forces are viewed as the key drivers of this slowdown before the GFC: (i) a slowing pace of innovation at the global technological frontier and reduced productivity spillovers to the rest of the world;3 (ii) population aging across the globe, adversely affecting productivity; and (iii) fading structural reform efforts during the 2000s, especially in accumulating human capital (e.g., via education reforms) in many emerging markets and developing countries. After the onset of the GFC, the TFP growth slowdown has been exacerbated by several factors: (i) policy uncertainties and tight credit conditions during the GFC, combined with pre-GFC corporate balance-sheet vulnerabilities, have hampered firms’ access to credit and led to less investment in intangible and physical assets; (ii) resource misallocations within and/or across sectors, which started before the crisis, may have worsened during the GFC; and (iii) global trade slowdown following the GFC may have limited further integration into the global value chains (GVCs) and cross-country technology diffusion.
5. This chapter explores the factors behind Poland’s productivity growth slowdown and their likely impact on the future TFP growth. The analysis starts by identifying which sectors and types of firms are the main contributors to Poland’s TFP growth slowdown. It then separates the role of allocative efficiency (i.e., how efficiently resources are allocated across sectors) and technical efficiency (i.e., how advanced the technology is within a sector) in contributing to the aggregate productivity growth slowdown. In this context, the chapter also examines the role of external conditions, mainly the GVC participation and external financial conditions, as well as the role of domestic bottlenecks that could affect aggregate TFP growth. Finally, the findings are used to inform the baseline projections of TFP growth over the medium term.
B. Understanding Poland’s Productivity Slowdown
6. This section uses firm-level and sector-level data to examine TFP growth from a micro perspective. Publicly available data on Polish firm- and sector-level TFP are limited. Our firm-level TFP analysis is largely based on the ORBIS and BACH databases, which cover only 3–4 percent of the total number of firms, accounting for around 50 percent of employment in Poland (see the Appendix for details). Therefore, the results should be interpreted with caution. Some robustness checks are presented in the Appendix.
7. Since the early 2000s, the TFP level has been declining in agriculture and other industries (excl. manufacturing), but has been largely flat in manufacturing and market services (Figure 2).4 Agriculture and other industries (excl. manufacturing) account for a relatively small share of the total value-added in Poland, so their declining TFP levels have had a limited impact on the economy-wide productivity.5 In contrast, manufacturing and market services are relatively large both in terms of value-added and employment, and hence, the sluggish TFP growth in those sectors could be a significant drag on the aggregate TFP growth. Other sectors, including basic services, construction, and trade, have seen moderate TFP growth. The sectoral patterns in TFP are broadly similar to those in labor productivity.
Poland: TFP Trends by Sector
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: Eurostat, ORBIS, and IMF staff calculations.Poland: TFP Trends by Sector
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: Eurostat, ORBIS, and IMF staff calculations.Poland: TFP Trends by Sector
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: Eurostat, ORBIS, and IMF staff calculations.8. Small and medium-size enterprises (SMEs) have experienced sharper declines in TFP during and after the GFC than large firms (Figure 3). Based on the available firm-level data, SMEs in Poland appear to be more productive, on average, than larger firms. This may be an evidence of allocative inefficiencies, as more productive firms should be able to expand and increase their share in total employment. During the post-GFC period, however, large firms appear to have been more successful in maintaining their productivity and continuing capital deepening (which is evident in an increase in labor productivity), while SMEs experienced sizeable and persistent TFP losses.
Poland: TFP Trends by Firm Size
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: ORBIS and IMF staff calculations.Poland: TFP Trends by Firm Size
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: ORBIS and IMF staff calculations.Poland: TFP Trends by Firm Size
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: ORBIS and IMF staff calculations.9. More leveraged firms had better TFP performance before the GFC, but performed worse than less leveraged firms after the GFC (Figure 4). In all size groups, the median firms with higher leverage ratios had higher TFP before the crisis. After the crisis, however, those with higher leverage ratios tended to have lower TFP. Moreover, the TFP gap between high- and low-leverage firms has been widening during the post-crisis period. A possible explanation is that highly leveraged firms faced tighter financing constraints during and after the GFC. Therefore, they had to cut investment more than their less-leveraged peers, which has constrained TFP growth in high-leverage firms.
Poland: TFP Trends by Leverage Ratio and Firm Size 1/
(Log total factor productivity; median firms in each group)
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: ORBIS and IMF staff calculations.1/ Leverage ratio is the ratio of long-term debt and short-term liabilities on total assets. In each firm-size group, firms that have leverage ratio ranked in the first quartile are classified as firms with low leverage ratio and firms that have leverage ratio ranked in the fourth quartile are classified as firms with high leverage ratio.Poland: TFP Trends by Leverage Ratio and Firm Size 1/
(Log total factor productivity; median firms in each group)
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: ORBIS and IMF staff calculations.1/ Leverage ratio is the ratio of long-term debt and short-term liabilities on total assets. In each firm-size group, firms that have leverage ratio ranked in the first quartile are classified as firms with low leverage ratio and firms that have leverage ratio ranked in the fourth quartile are classified as firms with high leverage ratio.Poland: TFP Trends by Leverage Ratio and Firm Size 1/
(Log total factor productivity; median firms in each group)
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: ORBIS and IMF staff calculations.1/ Leverage ratio is the ratio of long-term debt and short-term liabilities on total assets. In each firm-size group, firms that have leverage ratio ranked in the first quartile are classified as firms with low leverage ratio and firms that have leverage ratio ranked in the fourth quartile are classified as firms with high leverage ratio.10. In what follows, we use the OECD framework to disentangle different factors that could be behind the TFP growth slowdown. Figure 5 presents a simplified version of the conceptual framework in OECD (2015) “
A Conceptual Framework to Assess Productivity Gaps
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Source: OECD (2015).A Conceptual Framework to Assess Productivity Gaps
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Source: OECD (2015).A Conceptual Framework to Assess Productivity Gaps
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Source: OECD (2015).11. A decomposition of Poland’s aggregate TFP growth suggests that allocative inefficiencies across sectors may have been the main drag after the crisis (Figure 6). The decomposition of the TFP growth, based on the McMillan and Rodrik (2011) approach using the sector-level data,6 points to continued improvement in the “within” component, which measures within-sector technical efficiency, but a sharp deterioration in the “between” component (measuring the allocative efficiency across sectors) after the crisis. Before the GFC, both technical and allocative efficiencies had similar positive contributions to the total TFP growth, but after the GFC, the contribution of allocative efficiency became negative, almost completely offsetting the positive contribution from technical efficiency. Rising allocative inefficiencies across sectors may reflect several factors, including labor hoarding (perhaps, related to limited labor mobility across sectors) as well as the tendency to invest in low-risk-and-low-return (and hence, less productive) sectors in the environment of high macroeconomic and policy uncertainty.
Poland’s TFP Decomposition: Allocative Efficiency vs. Technical Efficiency
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: ORBIS, Regional Economic Issues, Central, Eastern, and Southeastern Europe (May 2016), and IMF staff calculations.Poland’s TFP Decomposition: Allocative Efficiency vs. Technical Efficiency
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: ORBIS, Regional Economic Issues, Central, Eastern, and Southeastern Europe (May 2016), and IMF staff calculations.Poland’s TFP Decomposition: Allocative Efficiency vs. Technical Efficiency
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: ORBIS, Regional Economic Issues, Central, Eastern, and Southeastern Europe (May 2016), and IMF staff calculations.12. Improvement in technical efficiency continued after the GFC, but at a slower pace (Figure 6). Following Hsieh and Klenow (2009), resource misallocation within each sector can be measured as gaps in the marginal products of labor and capital across firms within the same sector. In Poland, this type of resource misallocation has diminished following the GFC in most sectors, except in agriculture and market services. Continued improvements in resource allocation across firms within the same sector seem to have played an important role in raising aggregate technical efficiency over time, as evidenced by Poland’s steady progress toward the global technological frontier in the last two decades. 7, 8 More recently, however, the improvement in technical efficiency has slowed.
C. The Role of External Conditions
13. External conditions can affect technical efficiency in Poland through trade and FDI channels. This section examines how Poland’s participation in the GVC and capital inflows have influenced its TFP growth. The goal is to use this analysis to inform our baseline TFP projections under the WEO forecasts of the global environment over the medium term.
14. Historically, Poland’s TFP growth has been strongly correlated with its GVC participation, which in turn, is influenced by external demand (Figure 7, chart 1). During 1995–2011, Poland has rapidly integrated into the German supply chain, notably in manufacturing and services. The Polish exports of computers and electronics, machinery and equipment, and motor vehicles have foreign value added of more than 40 percent. After the GFC, the growth rate of Poland’s GVC participation declined amid the global trade slowdown, which coincided with a significant TFP growth slowdown.
Poland: External Conditions and TFP Growth
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
1/ Defined as a share of intermediate goods produced by Poland as inputs into production of other countries.2/ Defined as foreign value-added in Polish exports.3/ Include constant, time trend, fixed effects and model errors.Sources: Penn World Table 9.0, OECD TiVA database, World Economic Outlook (April 2017), and IMF staff calculation.Poland: External Conditions and TFP Growth
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
1/ Defined as a share of intermediate goods produced by Poland as inputs into production of other countries.2/ Defined as foreign value-added in Polish exports.3/ Include constant, time trend, fixed effects and model errors.Sources: Penn World Table 9.0, OECD TiVA database, World Economic Outlook (April 2017), and IMF staff calculation.Poland: External Conditions and TFP Growth
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
1/ Defined as a share of intermediate goods produced by Poland as inputs into production of other countries.2/ Defined as foreign value-added in Polish exports.3/ Include constant, time trend, fixed effects and model errors.Sources: Penn World Table 9.0, OECD TiVA database, World Economic Outlook (April 2017), and IMF staff calculation.15. Staff’s estimates suggest that both external demand and capital flows may provide some support to Poland’s TFP growth going forward. Following the April 2017 WEO approach, country-specific external factors are constructed and a cross-country panel regression (covering 86 advanced and emerging market economies) is used to estimate the contributions of external factors to the TFP growth in Poland (Figure 7, chart 2). External demand (as a proxy for the GVC participation) and capital flows had the largest contributions to TFP growth in Poland in the past. Going forward, with the projected recovery in the EU and continued accommodative global financial conditions as advanced economies normalize their monetary policies only gradually, the external environment will provide some support to Poland’s TFP growth over the medium term, but such support will be limited.
D. The Role of Domestic Factors
16. Domestic structural and institutional characteristics affect the technology diffusion and hence the TFP growth as well. The literature has identified a range of structural and institutional characteristics that could affect allocative and technical efficiencies:9
Factors affecting both allocative and technical efficiencies: structure of the economy (i.e., relative shares of agriculture, manufacturing and service sector), flexibility of the labor market, government efficiency, and restrictiveness of regulation.
Factors affecting allocative efficiency: affordability of financial services, and business climate.
Factors affecting technical efficiency: quality of institutions (e.g., judicial independence, impartial courts, and protection of property rights) and infrastructure gaps. Evidence on the role of research and development (R&D) spending is more mixed—it seems that what matters is not only the level of spending but also the nature of spending. For example, for emerging and developing countries, complementary R&D spending to facilitate the adoption of global advanced technologies by domestic firms appears to be most effective.
17. Poland has made much progress on structural and institutional reforms over the past 25 years, but gaps remain:
Based on a wide range of indicators from the World Bank (WB), OECD and World Economic Forum (WEF), the areas where Poland is lagging the most compared to its peers (OECD countries) include infrastructure, business regulation, labor market efficiency, and R&D/innovation. The Poland specific studies by international institutions further identify shortcomings in human capital development and institutions/government efficiency. A still relatively high share of agriculture and relatively low share of services (compared to peers) suggest that there may be scope to re-allocate resources towards higher-productivity sectors.10
But there are also areas where Poland scores well relative to peers: it has market-friendly institutions; low barriers to trade and investment; less regulatory complexity and less regulatory protection of incumbents; as well as relatively high quality of human capital. These strengths should be preserved and better leveraged.
Chapter 4 provides an in-depth analysis of the key structural bottlenecks and quantifies the impact of reforms in these areas on long-term potential growth.
Poland: Structural Reform Gaps
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: OECD, World Bank, WEF, IEF, EFW, and IMF staff calculations.Poland: Structural Reform Gaps
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: OECD, World Bank, WEF, IEF, EFW, and IMF staff calculations.Poland: Structural Reform Gaps
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: OECD, World Bank, WEF, IEF, EFW, and IMF staff calculations.E. Baseline TFP Growth in the Medium Term
18. This section provides the baseline projections of Poland’s TFP growth in the medium term. Under the baseline, the key boost to the TFP growth will come from continued improvement in technical efficiency, supported by the projected improvement in the external environment. Meanwhile, the allocative inefficiencies would remain a drag in the absence of further structural reforms. Domestic demographics and projected investment path would also affect TFP growth.
19. Under the baseline, Poland’s TFP growth will recover moderately in the near term but will remain flat at around 1 percent over the medium term (Figure 9). Key considerations are:
External demand from Poland’s key export destinations is expected to recover moderately in the near term, but the recovery will be more gradual over the medium term (based on the latest WEO projections). Therefore, external demand is projected to boost TFP growth only moderately. If Poland’s manufacturing sector manages to increase its GVC participation by more than suggested by the external demand recovery (e.g., if manufacturing sector’s TFP growth could return to its peak TFP growth before the crisis), then the medium-term TFP growth could be closer to 1½ percent, but this would likely require additional reform efforts and more FDI.
Investment is expected to increase only gradually, especially private investment (see Chapter 2). Therefore, the boost to TFP growth from investment is also projected to be limited.11
Aging work force is expected to have a substantial negative impact on TFP after 2030. Several recent studies find statistically significant impact of an aging work force on TFP growth (Aiyar et al. (2016) and Adler et al. (2017)). Poland’s demographics has already been worsening and the share of senior workers (age 55 and above) in the total work force is expected to rise significantly since 2030. While the negative impact from an aging work force will not affect the TFP growth in the medium term, it may be substantial beyond 2030.12
Poland: Baseline TFP Growth Projections
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: WEO and IMF staff calculations.Poland: Baseline TFP Growth Projections
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: WEO and IMF staff calculations.Poland: Baseline TFP Growth Projections
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: WEO and IMF staff calculations.20. The baseline TFP projections are in line with past experiences, but a much higher TFP growth would be needed to achieve income convergence with the EU average by 2030. Figure 9 provides a comparison with the experience of Korea, which managed to sustain a rapid pace of income convergence to advanced economies for over 3 decades. The baseline TFP projections for Poland are in line with Korea’s experience when Korea had a similar level of income per capita (1999) as in Poland (2015). However, the baseline TFP growth falls short of Korea’s TFP performance during its fastest income convergence episode (starting from around 1982) when Korea’s income per capita was much lower than Poland’s currently. Even if Poland could achieve Korea’s TFP growth of the 1980s and 1990s, its income convergence to the EU average would occur sometime beyond 2030.13
F. Concluding Remarks
21. Reinvigorating TFP growth in Poland requires closing structural gaps and better leveraging comparative strengths. Poland’s long-term average TFP level is relatively high compared to its EM peers. Its recent TFP growth slowdown after the GFC also appears to be in line with the experiences in many EMs. However, a protracted TFP growth slowdown would hinder Poland’s continued income convergence toward the EU living standards. With the existing domestic structural bottlenecks and the expected limited support from the external environment, Poland’s TFP growth will likely remain lower than its pre-crisis level over the medium term. Boosting TFP growth requires addressing structural bottlenecks, most notably in infrastructure, business regulation, labor market efficiency, and R&D/innovation. At the same time, Poland also has some comparative strengths, including market-friendly institutions, low barriers to trade and investment, less regulatory complexity and less regulatory protection of incumbents than its peers and a relatively high quality of human capital. Preserving these strengths by maintaining strong policies is critical for sustained strong growth. Chapter 4 discusses possible reform scenarios and the quantification of their impact on potential growth.
References
Adler, G., R. Duval, D. Furceri, S. Celik, K. Koloskova, and M. Poplawski-Ribeiro, April 2017, “Gone with the Headwinds: Global Productivity” and “Technical Appendices I-VIII”, International Monetary Fund Staff Discussion Note 17/04.
Aiyar, S., C. Ebeke, and X. Shao, 2016, “The Impact of Workforce Aging on European productivity”, Working Paper WP/16/238 (International Monetary Fund: Washington, DC).
Albinowski, M., J. Hagemejer, S. Lovo, and G. Varela, 2015, “Sustaining Micro Competitiveness to Ensure Convergence and Macro Resilience of the Polish Economy”, Republic of Poland, Ministry of Finance Working Paper No 21.
Andrle, M., 2017, “Structural and Fiscal Measures to Increase Potential Output in Austria”, Austria Selected Issues, IMF Country Report No. 17/27 (International Monetary Fund: Washington, DC).
Budina, N., 2017, “Spain: Drivers of Firm Productivity Growth”, Spain Selected Issues, IMF Country Report No. 17/24 (International Monetary Fund: Washington, DC).
Ebeke, C., K. Krogulski, and R. Sierhej, 2015, “Raising Productivity Growth in Poland: The Role of Structural Transformation”, Republic of Poland Selected Issues, IMF Country Report No. 15/183.
Egert, B., and P. Gal, December 2016, “The Quantification of Structural Reforms in OECD Countries: A New Framework”, OECD Economics Department Working Papers No. 1354.
Gal, P. N., 2013, “Measuring Total Factor Productivity at the Firm Level Using OECD-ORBIS,” OECD Economics Department Working Paper, No. 1049.
Gopinath, G., Kalemli-Ozcan, S., Karabarbounis L., and Villegas-Sanchez C., forthcoming, “Capital Allocation and Productivity in South Europe,” Quarterly Journal of Economics.
Hsieh, C., and P. J. Klenow, 2009, “Misallocation and Manufacturing TFP in China and India”, the Quarterly Journal of Economics, Vol. CXXIV, Issue 4.
International Monetary Fund, 2013, “German-Central European Supply Chain—Cluster Report”, Country Report No. 13/263 (International Monetary Fund: Washington, DC).
International Monetary Fund, 2015, “Where are We Headed? Perspectives on Potential Output?”, World Economic Outlook Chapter 3.
International Monetary Fund, 2016, “Time for a Supply-side Boost? Macroeconomic Effects of Labor and Product Market Reforms in Advanced Economies”, World Economic Outlook Chapter 3.
International Monetary Fund, 2016, “Central, Eastern and Southeastern: How to Get Back on the Fast Track?”, Regional Economic Issues.
International Monetary Fund, 2017, “Roads Less Traveled: Growth in Emerging Market and Developing Economies in a Complicated External Environment”, World Economic Outlook Chapter 2.
Kalemli-Ozcan, S., and others, “How to Construct Nationally Representative Firm-Level Data From the ORBIS Global Database,” NBER Working Paper, No. 21558.
McMillan, M., and D. Rodrik, 2011, “Globalization, Structural Change and Productivity Growth”, NBER Working paper 17143.
OECD, 2015, “The Future of Productivity”, OECD.
Pragyan, D., 2016, “Post-Crisis Adjustment in Latvia: Evidence from Firm Level Data”, Republic of Latvia Selected Issues, Country Report No. 16/172 (International Monetary Fund: Washington, DC).
Appendix I. Data Sample and Comparison of Data Sources
A. Data Sample
Firm-level data is based on the ORBIS database provided by Bureau van Dijk.1 The original dataset has over five million firm-year observations for 2000–15. By applying some filters and data imputations (see Box A.1. for details), our dataset includes 128,845 firm-year observations over 2003–13 (Table A.1.). The data filtering methodology follows the literature (see Gal, 2013; Kalemli-Ozcan et al., 2015; and Gopinath et al., forthcoming) and includes only years with sufficient observations. Data imputations to extend the coverage follow Gal (2013).
Data Coverage
Data Coverage
| Year | Agriculture | Manufacturing | Construction | Trade | Market Services | Basic Services | Other Industries | Total |
|---|---|---|---|---|---|---|---|---|
| 2003 | 22 | 548 | 156 | 671 | 294 | 98 | 71 | 1860 |
| 2004 | 54 | 942 | 319 | 1221 | 567 | 166 | 159 | 3428 |
| 2005 | 216 | 2796 | 830 | 3853 | 1644 | 504 | 540 | 10383 |
| 2006 | 312 | 3654 | 1163 | 5337 | 2294 | 724 | 674 | 14158 |
| 2007 | 444 | 4363 | 1480 | 6689 | 3311 | 1003 | 736 | 18026 |
| 2008 | 434 | 4493 | 1553 | 6888 | 3335 | 1048 | 745 | 18496 |
| 2009 | 432 | 4594 | 1590 | 7028 | 3389 | 1097 | 765 | 18895 |
| 2010 | 192 | 3255 | 1041 | 4814 | 1737 | 606 | 579 | 12224 |
| 2011 | 180 | 2540 | 803 | 3797 | 1523 | 527 | 486 | 9856 |
| 2012 | 346 | 2390 | 1014 | 4072 | 2934 | 730 | 412 | 11898 |
| 2013 | 313 | 1765 | 749 | 3197 | 2683 | 651 | 263 | 9621 |
| Total | 2,945 | 31,340 | 10,698 | 47,567 | 23,711 | 7,154 | 5,430 | 128,845 |
Data Coverage
| Year | Agriculture | Manufacturing | Construction | Trade | Market Services | Basic Services | Other Industries | Total |
|---|---|---|---|---|---|---|---|---|
| 2003 | 22 | 548 | 156 | 671 | 294 | 98 | 71 | 1860 |
| 2004 | 54 | 942 | 319 | 1221 | 567 | 166 | 159 | 3428 |
| 2005 | 216 | 2796 | 830 | 3853 | 1644 | 504 | 540 | 10383 |
| 2006 | 312 | 3654 | 1163 | 5337 | 2294 | 724 | 674 | 14158 |
| 2007 | 444 | 4363 | 1480 | 6689 | 3311 | 1003 | 736 | 18026 |
| 2008 | 434 | 4493 | 1553 | 6888 | 3335 | 1048 | 745 | 18496 |
| 2009 | 432 | 4594 | 1590 | 7028 | 3389 | 1097 | 765 | 18895 |
| 2010 | 192 | 3255 | 1041 | 4814 | 1737 | 606 | 579 | 12224 |
| 2011 | 180 | 2540 | 803 | 3797 | 1523 | 527 | 486 | 9856 |
| 2012 | 346 | 2390 | 1014 | 4072 | 2934 | 730 | 412 | 11898 |
| 2013 | 313 | 1765 | 749 | 3197 | 2683 | 651 | 263 | 9621 |
| Total | 2,945 | 31,340 | 10,698 | 47,567 | 23,711 | 7,154 | 5,430 | 128,845 |
For presentation purposes, the sector-level TFP data (based on NACE Rev 2) is aggregated into 7 broad sectors, weighted by the number of employees in each sector:
Agriculture: A-Agriculture, forestry, and fishing
Manufacturing: C-Manufacturing
Other Industries: B-Mining and quarrying, D-Electricity, gas, steam, and air conditioning supply, E-Water supply, sewerage, waste management and remediation activities
Construction: F-Construction
Trade: G-Wholesale and retail trade; repair of motor vehicles and motorcycles, H-Transportation and storage, I-Accommodation and food service activities
Market services: J-Information and communication, K-Financial and insurance activities, L-Real estate activities, M-Professional, scientific and technical activities, N-Administrative and support service activities
Basic Services: O-Public administration and defense, compulsory social security, P-Education, Q-Human health and social work activities, R-Arts, entertainment and recreation, S-Other service activities.
Rules for Data Filtering and Imputation
✓ Consolidation of the accounts: To prevent double counting, drop firms with consolidated accounts (C1 and C2) if they are also classified with unconsolidated accounts.
✓ Minimum number of employees: Drop firms with less than 3 employees.
✓ Negative values for tangible fixed assets and value added: Drop firms that have negative values of tangible fixed assets or value added in any year.
✓ Data imputation for value added: Fill in the gaps by summing up “Cost of employees” and “EBITDA”.
✓ Data imputation for tangible fixed assets: Fill in the gaps with “Total Fixed Assets”.
✓ Outliers: Drop firms if their capital, labor, or value added (at least one of them) has a growth rate in the top and bottom 0.1 percent of the growth distribution in the respective sector group at least once during the sample period.
✓ Continuity of firm data: Keep firms that have data available for value added, capital stock, and labor for at least 3 consecutive years. This rule is relaxed only for 2012–13 to ensure sufficient observations.
B. Comparison of Data Sources
For the sector-level TFP analysis, Bank for Accounts of Companies Harmonized (BACH) database provided by the Banque de France can be used as an alternative data source.2 The BACH database gathers the data for Poland from the Statistical Survey of the Central Statistical Office (GUS) of Poland. The data provided in BACH are sectoral aggregates of the firm-level survey data and the aggregation is done for general purposes, not specifically for measuring total factor productivity. The number of firms aggregated in each sector is reported in the database.
The coverage of Polish firms in the ORBIS database and that in the BACH database are broadly comparable (see Table A2). The BACH database includes firms with unconsolidated accounts, whereas the ORBIS includes firms with both consolidated and unconsolidated accounts. For 2005–15, both datasets have around 500,000 year-firm observations with unconsolidated accounts, which suggests that in terms of the number of firms the two databases have similar coverage. However, the total number of employees covered in the BACH database in 2013 is three times larger than the total number of employees covered in the ORBIS database when only firms with the unconsolidated accounts are included. As we also include firms with consolidated accounts if they do not have unconsolidated accounts, the coverage of employment of our adjusted sample is broadly comparable to that in the BACH database.
Comparison of BACH and ORBIS Databases
Comparison of BACH and ORBIS Databases
| BACH | ORBIS | |
|---|---|---|
| Number of firms (2005–15) | (firms with unconsolidated accounts only) | 5,155,047 (including both consolidated and unconsolidated) |
| Number of employees (2013) | (unconsolidated firms only) | 10,403,750 (including both consolidated and unconsolidated firms) |
Comparison of BACH and ORBIS Databases
| BACH | ORBIS | |
|---|---|---|
| Number of firms (2005–15) | (firms with unconsolidated accounts only) | 5,155,047 (including both consolidated and unconsolidated) |
| Number of employees (2013) | (unconsolidated firms only) | 10,403,750 (including both consolidated and unconsolidated firms) |
A comparison of the sector-level TFP calculated using our dataset based on ORBIS with that from the BACH database suggest the following (Figure A.1.):3
The TFP levels have been relatively flat in manufacturing, trade, and market services sectors based on data from both databases.
The TFP levels of the basic services sector calculated from both datasets follow similar paths until 2009, however, the paths diverge afterwards.
The trends are very different in agriculture and construction. According to the BACH database, the TFP level of agriculture has increased the most over time, and almost reached the TFP level of manufacturing by 2013. However, based on the ORBIS database, productivity in agriculture has been declining, notably after the GFC, and it has always been the lowest among all sectors. According to the BACH database, the construction sector has experienced a large increase in TFP growth before the GFC, followed by a large decline. However, the ORBIS database suggests that productivity in the construction sector was resilient during 2008–09, and has remained largely flat afterwards.
Poland: A Comparison of Sectoral TFP based on the BACH and the ORBIS Databases
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: BACH, ORBIS, and IMF staff calculations.Poland: A Comparison of Sectoral TFP based on the BACH and the ORBIS Databases
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: BACH, ORBIS, and IMF staff calculations.Poland: A Comparison of Sectoral TFP based on the BACH and the ORBIS Databases
Citation: IMF Staff Country Reports 2017, 221; 10.5089/9781484310311.002.A003
Sources: BACH, ORBIS, and IMF staff calculations.The TFP trends from our dataset are comparable to other Poland-specific studies. In particular, Albinowski et al. (2015) uses a firm-level dataset from the Statistical Survey of the Central Statistical Office of Poland to analyze the productivity trends in manufacturing and its sub-sectors. They find that the manufacturing sector has experienced an average annual TFP growth rate of 5 percent over 2006–13. Using our data, we find an average annual TFP growth of the manufacturing sector of around 4 percent for the same period, close to the estimates in Albinowski et al. (2015) based on data from the national source.
Prepared by Ran Bi, Ezgi Ozturk and Yevgeniya Korniyenko.
The EU-15 comprised the following 15 advanced European countries: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, and the United Kingdom.
The evidence of a global technological frontier slowdown is somewhat mixed, depending on how the frontier is defined. IMF (2016b) finds that the global frontier (defined using the country-level data) has stopped expanding after the GFC, while OECD (2015) finds that the global frontier (defined not at the country level, but as the top 10 productive firms globally in each industry) has returned to its pre-crisis trajectory but the diffusion to the rest of the firms has slowed significantly. In any case, there seems to be evidence that technology spillovers from advanced economies are likely to slow down.
Our sectoral classification follows NACE Rev 2, but for presentational purposes, we group narrow sectors into broader categories in the sectoral charts (see Appendix). Our “manufacturing” remains the same as in NACE Rev 2, and “other industries” covers mining and quarrying, electricity, gas, steam and air conditioning supply, and water supply, sewerage, waste management and remediation. Thus, the classification of manufacturing is comparable with Albinowski et. al (2015).
A comparison of the shares of different sectors in the total valued added and in total employment of sample firms (drawn from the ORBIS database) with the corresponding shares based on the Eurostat sectoral data (Figure 2) shows that our sample represents the true population reasonably well. Agriculture is the most underrepresented sector, likely because the ORBIS database is missing many small and micro firms.
Ebeke et al. (2015) examines labor productivity across European countries using a similar decomposition approach.
Given the limitations in sectoral data, the stochastic frontier analysis cannot be done at the sectoral level for Poland.
This draws on the stochastic frontier analysis based on country-level data in IMF (2016b). The global frontier represents the maximum amount of output that can be obtained from given inputs. Then, relative technical inefficiency of a country is measured by its distance from the frontier.
This is a comprehensive list of factors based on the literature. However, not all factors included here are found to be statistically significant in all studies, possibly due to sample and measurement issues.
See the Article 2015 Selected Issues Paper on structural transformation.
Adler et al. (2017) finds that a 1 percentage point increase in the investment-to-capital stock ratio could boost annual TFP growth by 0.506 percentage points.
According to Eurostat projections, a five-year cumulative increase in the share of senior workers is expected to be 2 percentage points or more beyond 2030, which could translate into a five-year cumulative decrease in TFP growth by 1.5 percentage points or more based on Adler et al. (2017).
Using the baseline assumptions on the evolution of the working-age population and public investment, the TFP growth required to achieve convergence with the EU average per capita income by 2030 should be closer to 4 percent per year. These estimates are based on the simulations using the IMF’s Flexible System of Global Models (FSGM), where private investment responds endogenously to higher TFP growth (see Chapter 4), as well as the standard growth accounting approach.
See https://www.bvdinfo.com/en-gb/our-products/company-information/international-products/orbis for details.
See https://www.bach.banque-france.fr/?lang=en for details.
For presentation purposes, we aggregated the sectors in the BACH database into the same 7 sectors as explained above. For Poland, the BACH database does not have data for two sectors: K-Financial and insurance activities; and O-Public administration and defense, compulsory social security.
