Journal Issue
Share
Article

Euro Area Policies: Selected Issues

Author(s):
International Monetary Fund. European Dept.
Published Date:
July 2017
Share
  • ShareShare
Show Summary Details

Can Structural Reforms Foster Real Convergence in the Euro Area?1

A lack of productivity catch-up—lower productivity growth in countries with lower initial per capita income and productivity—explains much of the lack of income convergence within the euro area. This paper shows that countries with lower initial productivity levels can have larger productivity gains from labor and product market reforms than countries with higher initial productivity. This points to structural reforms as critical to reducing productivity gaps to foster real income convergence.

A. Context

1. There has been a secular decline in productivity across all advanced economies. This decline occurred even before the global financial crisis because of several factors. First, aggregate productivity declined due to the reallocation of resources toward sectors where productivity growth was slower, and declining productivity growth within sectors which accounted for the bulk of employment and economic activity (Dabla Norris and others, 2015). Second, structural headwinds—from an aging workforce, slower human capital accumulation and slowing trade integration—also contributed to the decline in productivity (Aiyar and others, 2016; IMF 2016a). Finally, in the aftermath of the global financial crisis, some euro area countries have displayed persistent productivity losses stemming from weak corporate and bank balance sheets; adverse feedback loop of weak aggregate demand, investment and capital-embodied technological change; and elevated economic and policy uncertainty (Adler and others, 2017).

2. In addition, there have been significant and persistent productivity differentials across countries in the euro area. These productivity differences pre-dated, but were aggravated by, the global financial crisis and took place against the background of the wider long-term slowdown of productivity in all advanced economies (as noted above).

  • There has been little catch-up in labor productivity. Productivity gaps between euro area countries have persisted over the years (Figure 1). Countries with higher levels of labor productivity in 1999 (i.e., before the euro was introduced) have continued to witness a rise in labor productivity since then, in sharp contrast with the stagnant labor productivity levels in countries which already had low productivity in 1999 (left panel, Figure 1). This led to a widening gap in labor productivity within the euro area over time. As a result, the dispersion of labor productivity levels between EA-12 countries has remained broadly unchanged over the past 20 years, displaying a lack of o-convergence (right panel, Figure 1).2

  • Productivity gaps exist even at the sectoral level (Figure 2). The median labor productivity level in the manufacturing sector in Spain and Italy remains well below that of other euro area countries even though it is on a rising trend. The gaps are narrower in the service sector. But it is noteworthy that, even here, the productivity levels have stagnated over the last two decades in Italy and Spain, whereas productivity levels in other euro area countries are on a rising trend.

  • Moreover, while TFP growth has slowed across the board it has declined faster for countries with lower initial productivity levels (text figure). TFP growth—derived from decomposing income per capita growth into capital, labor and TFP contributions—has been particularly weak in the four countries that had the lowest labor productivity in 1999 (Greece, Italy, Portugal and Spain). Compared to their EA-12 peers, TFP growth in these countries accounts for a smaller share of total output growth, both before and after 1999, pointing to structural features as part of the underlying explanation. In the aftermath of the crisis, TFP even shrunk due to the failure to resolve crisis legacies (Adler and others, 2017).

Figure 1.Cross-Country Labor Productivity Gaps in the Euro Area

Figure 2.Cross-Country Sectoral Labor Productivity Gaps in the Euro Area

Source: EU KLEMS (2016) and IMF staff calculations.

Note: The green bar is the median productivity level in manufacturing and services sector for the listed countries. The data are not PPP adjusted.

TFP Dynamics

Average annual per capita growth rates in percent, unweighted

Note: Productivity groups defined on the basis of labor productivity in 1999. The cutoff for the low and high initial productive countries is based on the median of the observed productivity value in 1999. Low initial productivity is the average of countries below the median in 1999 and high initial productivity is average of countries above the median in 1999. No 1990s data available for Austria.

Sources: AMECO, Haver Analytics, and IMF staff calculations.

3. The persistent differences in productivity led to the lack of real income convergence within the euro area (based on Schoelermann (forthcoming), ECB 2015, ECB (forthcoming), Franks and others, (forthcoming)).3 Before the global financial crisis, there was steady income convergence in the period leading up to the introduction of the euro. After the euro was introduced, convergence stalled, followed by divergence starting with the global financial crisis. At the same time, countries that joined the euro area from 2007 onwards experienced continued convergence in the run-up to their accession.

4. This paper examines the role of structural reforms in reducing the divergence of productivity within the euro area. There is an extensive literature which shows that labor and product market reforms can improve resource allocation and enhance productivity gains. To the extent productivity gains are higher for countries with low initial levels of productivity than they are for countries with high initial levels of productivity, they can also narrow productivity differentials over time, thereby facilitating real income convergence.

B. Methodological Framework

5. The analysis focuses on three model specifications to assess how productivity at the country and sector level responds to country- and sector-specific reforms.

Model specification

6. Model 1: Country-level reform, country-specific productivity effects. Following IMF (2016b), and Adler and others (2017), the analysis uses a dynamic approach to identify the effect of labor and product market reforms on real productivity over time, using a sample of 20 European countries over the period 1980 to 2014.4 The main innovation in this paper is that it models the effect of reforms in a non-linear way by conditioning the effect of reforms on countries’ initial productivity levels. Hence, the model allows us to examine whether the effect of structural reforms is stronger in countries with initially larger productivity gaps.5 Since reforms have lagged effects on productivity, the non-linear impact of reforms on productivity growth is estimated at different time horizons using the local projections method (Jordà, 2005). Specifically, the model takes the following form:

where pi,t+h is the natural logarithm of real hourly labor productivity (defined as real value added divided by total hours worked from the Penn World Tables dataset) in the country i in year t+h. The model 1 is estimated at each yearly horizon h = 0, 1, …, 5. The variable Ri,t is a binary variable which captures the occurrence of a structural reform at time t. The reform can be a product market (PMR) or an employment protection legislation (EPL) reform. Control variables X include the lagged hourly productivity level (in natural logarithm), and contemporaneous and past crisis dummies (defined as annual growth below −3 percent). The crisis dummies help control for the cyclicality of productivity changes within countries as they recover from a crisis. These variables also help ensure that the estimated effect of reforms on productivity growth is not contaminated by the correlation between reforms and cyclical developments as illustrated in IMF 2016b. The model also controls for country and year fixed effects (ui, γt) to account for country-specific unobservable factors and common shocks to all countries, respectively.

7. Model 2: Country-level reform, sector-specific productivity effects. The above analysis is replicated at the sector level to identify the effects of reforms at the country level on productivity growth at the country-specific sectoral level. The identification strategy relies on the assumption that productivity growth at the sectoral level will not directly affect reform adoption and implementation in the country as a whole. To estimate the effect of labor and product market reforms on sectoral productivity growth, an amended version of equation [1] above is estimated at different time horizons using the local projection method:

where the subscript s denotes the sector, c the country, and t the year.6psc,t+h is the natural logarithm of real hourly labor productivity in constant euros observed in sector s (defined as real euro value added divided by the total number of hours worked by employees engaged in sector s in country c in year t+h). The model controls for a rich set of sector-cum-country fixed effects to account for time-invariant factors that are specific to sectors belonging to different countries such as the size of sectoral employment, sector*country and year fixed effects.

8. Model 3: Sector-specific reform, sector-specific productivity effects. The availability of disaggregated sectoral data makes it possible to identify the impact of sector-specific reforms on sector-specific productivity growth. For example, given the availability of the sectoral EUKLEMS data, it is possible to assess the extent to which a reform shock at the sectoral level (such as, the reduction in the number of licenses needed to engage in retail trade, or the reduction in the regulations pertaining to the professional service sector) influences the productivity growth of the retail trade and professional services sectors. The main caveat surrounding this analysis would be the high risk of endogeneity associated with the sector-specific reform as this reform is more likely to be triggered by expected productivity gains in the sector. The model takes the following form where the unit of observation becomes the country and the reform impact is estimated for a given sector. The model controls for the log of employment in the specific sector and country and year fixed effects.

Data

9. Reform measures. Equations [1] and [2] are estimated using a new narrative-based dataset of major labor and product market reforms that was used in IMF 2016b and Adler and others (2017), and will be published in Duval and others (forthcoming). An important advantage of this database is that it identifies the precise nature and timing of major legislative and regulatory actions taken by advanced economies since the early 1970s in key labor and product market policy areas.78 Equation [3] is estimated using sector-specific reforms drawn from the OECD database. Sector-specific reform measures are only available for product market reforms and not labor market reforms.

10. Labor productivity. Equation [1] is estimated using country-level data on labor productivity drawn from Penn World Tables. This data is in international PPP terms, therefore allowing cross-country comparisons in the levels of productivity. Equations [2] and [3] are estimated using the 2016 release of EUKLEMS sectoral data by the European Commission which provides a rich source of information on labor productivity but also employment and other macro variables that are critical to assessing sectoral performance. The sectoral-level data are available for a set of 11 countries (Austria, Belgium, Finland, France, Germany, Italy, Spain, Sweden and the United Kingdom) for 1995–2014. Unlike the country-level productivity data from the Penn World Tables, the EUKLEMS data is not PPP adjusted. While this may bias the results somewhat, the bias is limited by controlling for sector-country fixed effects. This is because the gap between PPP and non-PPP adjusted labor productivity at the country level is relatively constant over time for a specific country, suggesting that the gaps between PPP and non-PPP adjusted sectoral labor productivity might also be constant over time for a given country, and thereby absorbed by the fixed effects.

Choice of Reforms

11. While the structural reform priorities in euro area countries are wide ranging, the analysis below focuses on employment protection reform and product market reforms. These reforms boost flexibility in the labor market and facilitate the efficiency and the entry exit of firms respectively. The role of product and labor market deregulation in fostering output and productivity growth is well documented (e.g., Adler and others, 2017; Bouis, Duval and Eugster, 2016; Nicoletti and Scarpetta, 2005). The productivity boost arises because such reforms can facilitate the diffusion of technology and innovation across companies, increase the incentives to innovate, and improve resource allocation by weeding out less productive firms and workers (OECD, 2015). The analysis can, in principle, be expanded to include additional reforms.

C. Results of Empirical Analysis

12. There have been persistent differences in the degree of labor market flexibility amongst euro area countries. Countries with the most protected labor markets in 2000 continue to remain more protected than the most flexible euro area labor markets despite some progress since the global financial crisis (text figure). Empirical analysis suggests that, should countries with low initial productivity levels further reduce excessively high levels of employment protection on regular work contracts, this could be associated with a significant boost in labor productivity (Figure 3, left panel). The impact of further reforms on countries with high initial productivity levels is likely to be more muted (Figure 3, right panel). The differential productivity implications suggest that reforms could eventually foster real convergence. Similarly, at the sectoral level, the gains in labor productivity tend to be larger for those sectors which had low levels of productivity to begin with compared to the high initial productivity sectors (Figure 4, right panel versus left panel, respectively).

Figure 3.Estimated Impact of Employment Protection Reform on Aggregate Labor Productivity

Source: IMF staff estimates.

Note: See footnote 7 of the paper for the definition of the reforms. The shock occurs at t = 0. The x-axis measures years after the reform effects. The dependent variable is cumulative labor productivity growth at each horizon. The estimation method is the local projection method (Jorda 2005) –equation [1]– and the reform dummy enters additively, and in interaction with lagged log level of productivity. The models control for lagged log of productivity level, past crisis dummies, country and year fixed effects . The dashed lines denote 90 percent confidence bands.

Figure 4.Estimated Impact of Employment Protection Reform on Sectoral Labor Productivity

Source: IMF staff estimates.

Note: See footnote 7 of the paper for the definition of the reforms. The shock occurs at t = 0. The x-axis measures years after the reform effects. The dependent variable is cumulative labor productivity growth at each horizon. The estimation method is the local projection method (Jorda 2005) –equation [1]– and the reform dummy enters additively, and in interaction with lagged log level of productivity. The models control for the log of sectoral employment, sector*country and year fixed effects. The dashed lines denote 90 percent confidence bands.

Evolution of Employment Protection Legislation

(Index)

Source: OECD/IDB Employment Protection Database.

Note: The indicator measures the strictness of regulation of individual dismissal of employees on regular/indefinite contracts. Calculated across 21 aspects of employment protection regulation and then converted into a score of measured on a 0–6 scale, with higher values representing stricter regulation. The cutoff for the most and least regulated countries is based on the median of the observed EPL value in 2000. Least protected index is the average of countries below the median in 2000 (i.e. least regulated) and most protected is average of countries above the median in 2000 (i.e. most regulated).

13. Differences in product market efficiency and flexibility have persisted across euro area countries. The gaps in the overall OECD index of product market flexibility (PMR) have closed significantly over time (text figure), and especially after the crisis, in part due to the important role of EU-legislated reforms in network industries. However, at a more granular level, the gaps in the professional services sector regulations and business climate have marginally narrowed but not closed (Figure 5, right panel), whereas the gaps in the retail trade sector have widened as reforms appear to have occurred in the better performing countries (Figure 5, left panel).

Figure 5.Evolution of Sector-Specific Measures of Product Market Flexibility

Product Market Reform

(0–6, with 6 being most regulated)

Source: OECD, Product Market Regulation Database.

Note: Product market reform indicators are normalized on a 0–6 scale, where lower values reflect a more competition-friendly regulatory stance. The cutoff for the most and least regulated countries is based on the median of the observed product market reform value in 2003. Least protected index is the average of countries below the median in 2003 (i.e. least regulated) and most protected is average of countries above the median in 2003 (i.e. most regulated).

14. Product market reforms can bridge productivity gaps. Were countries with low initial levels of aggregate labor productivity to implement product market reforms, they would receive a more significant boost to productivity than those with high levels of initial productivity (Figure 6, top panel). These results also hold at the sectoral level across countries as well as at the sector specific level. Thus, the cumulative impact of product market reforms on sectoral productivity is larger over time for sectors with low productivity levels versus high productivity sectors (Figure 6, middle panel). Likewise, at the sector-specific level, reforms that lower entry barriers, for example, in the professional services sector, have a greater impact on productivity in countries where the initial productivity in this sector was low (Figure 6, last panel).

Figure 6.Effect of Product Market Reforms on Labor Productivity in Europe

(Percent)

Source: IMF staff estimates.

Note. See footnote 7 of the paper for the definition of reforms. The shock occurs at t = 0. The x-axis measures years after the reform effects. The dependent variable is cumulative labor productivity growth at each horizon. The estimation method is the local projection method (Jorda 2005) and the reform dummy enters additively, and in interaction with lagged log level of productivity. The models control for lagged log of productivity level, past crisis dummies, country and year fixed effects. The dashed lines denote 90 percent confidence bands.

D. Conclusions

15. The marginal return from reforms is higher for countries or sectors that are further away from the productivity frontier. The intuition for these results is as follows. Since low productivity countries tend to also be those that exhibit significant resource misallocation, there is greater scope for reforms to move factors of production toward the most productive sectors. Moreover, reforms which facilitate the entry of productive workers and firms and the exit of less productive workers and firms may be disproportionally beneficial in low-productivity countries. This is because such reforms would allow the most productive firms to attract the most productive workers, while providing incentives for boosting innovation—including through greater investment in technology and human capital—in the face of stronger competition. Finally, empirical evidence suggests that countries with large labor market rigidities (and therefore low productivity) tend to have a higher probability and intensity of labor market reform when compared to countries with low labor market rigidities (Ebeke, 2017). All these factors may help explain why reforms are likely to be stronger in countries in need of a bigger supply-side boost.

16. Policy implications. Labor and product market reforms need to be implemented in all euro area countries to reverse the long-term secular decline in productivity and facilitate adjustment to challenges from technological innovation. These reforms are especially important in countries with low initial productivity levels such as Greece, Italy, Portugal and Spain because the productivity effects of reform are likely to be greater in these countries, allowing them to catch up with the higher productivity euro area countries. By helping to narrow productivity differences, reforms will help reduce productivity and competitiveness gaps, reduce current account imbalances and foster real convergence. Reforms should also be accompanied by measures to improve labor force participation and address the crisis legacies of weak balance sheets and economic and policy uncertainty (Adler and others, 2017).

References

    Adler, G., R.Duval, D.Furceri, S. K.Çelik, K.Koloskova, and M.Poplawski-Riberio. 2017. “Gone with the Headwinds: Global Productivity.” IMF Staff Discussion Note 17/04, April (International Monetary Fund: Washington).

    Aiyar, S., C.Ebeke, and X.Shao. 2016. “The Impact of Workforce Aging on European Productivity.” IMF Working Paper No. 16/238. (International Monetary Fund: Washington).

    Bouis, R., R.Duval, and Eugster. 2016. “Product Market Deregulation and Growth: New Country-Industry-Level Evidence.” IMF Working Paper No. 16/114 (International Monetary Fund: Washington).

    Dabla-Norris, E., S.Guo, V.Haksar, M.Kim, K.Kochhar, K.Wiseman, and A.Zdzienicka. 2015. “The New Normal: A Sector-Level Perspective on Productivity Trends in Advanced Economies.” IMF Staff Discussion Note 15/03, March (International Monetary Fund: Washington).

    Duval, R., D.Furceri, J.Jalles, and H.Nhuyen, forthcoming. “A New Narrative Database of Product and Labor Market Reforms in Advanced Economies.” IMF Working Paper (International Monetary Fund: Washington).

    Ebeke, C. 2017. “Who Dares, Wins: Labor Market Reforms and Sovereign Yields.” IMF Working Paper No. 17/141 (International Monetary Fund: Washington).

    ECB. 2015. “Real Convergence in the Euro Area: Evidence, Theory and Policy Implications.” Economic Bulletin, Issue 5.

    ECB, forthcoming,Convergence in the Euro Area: A Long-Term Perspective” Occasional Paper Series.

    Franks, J., B.Barkbu, R.Blavy, W.Oman, and H.Schoelermann, forthcoming, “Economic Convergence in the Euro Area: Coming Together or Drifting Apart?IMF Working Paper, No. 2017/XX (International Monetary Fund: Washington).

    International Monetary Fund. 2016a. “Global Trade: What’s Behind the Slowdown?” World Economic Outlook, Chapter 2, October.

    International Monetary Fund. 2016b. “Time for a Supply-side Boost? Macroeconomic Effects of Labor and Product Market Reforms in Advanced Economies.” World Economic Outlook, Chapter 3.

    Jordà, Ò.2005. “Estimation and Inference of Impulse Responses by Local Projections.” American Economic Review95 (1): 16182.

    Nicoletti, G., and S.Scarpetta. 2005. “Product Market Reforms and Productivity in the OECD.” OECD Economics Department Working Papers, No. 460, OECD Publishing, Paris. http://dx.doi.org/10.1787/726517007575

    OECD. 2015. “The Future of Productivity,” Joint Economics Department and the Directorate for Science, Technology and Innovation Policy Note, July, OECD publishing.

    Schoelermann, H. (forthcoming). “Real Income Convergence in the Euro Area.” IMF Selected Issues Paper, 2017.

Prepared by Angana Banerji, Christian Ebeke, Hanni Schoelermann and Jesse Siminitz (all EUR) as well as Ksenia Koloskova (RES).

See Schoelermann, H. (forthcoming) for a detailed analysis of real income convergence in the euro area.

Real convergence is defined in these papers as the process whereby the real per capita GDP levels of lower-income economies catches up with those of higher-income economies on a durable basis.

The sample is restricted to European countries and includes Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Slovak Republic, Spain, Sweden, Switzerland, and the United Kingdom observed over the period 1970 to 2013.

In the absence of data on TFP levels, the following analysis focuses on labor productivity developments.

In this study, sectors are based on the highest level of aggregation provided in NACE Rev. 2 classification and do not include the financial sector (identified by the letter K).

The database identifies all legislative and regulatory actions related to product market regulation and employment protection legislation mentioned in OECD Economic Surveys. For any of these actions to qualify as a major reform or “counter-reform”—namely a major policy change in the opposite direction—one of the following criteria has to be met: (1) the OECD Economic Survey uses strong normative language to define the action (e.g., “major reform”); (2) the policy action is mentioned repeatedly across different editions of the OECD Economic Survey; or (3) the OECD indicator of the regulatory stance displays a very large change (in the 5th percentile of the distribution of the change in the indicator). Consequently, the EPL reform dummy takes values {1; 0; −1}, denoting a deregulating reform, absence of a reform, or a reversal of a reform. For PMR, the analysis focuses on major deregulation in seven network industries and the PMR reform dummy takes value of 1 if there were at least two reforms over three years in one or more of the seven network industries (airlines, gas, electricity, postal services, rail, road transportation, and telecom) and 0 otherwise.

The reform dummy captures only reforms that are sufficiently large. As a robustness check, a forthcoming working paper on this topic will assess the impact of the actual size of reforms. Ebeke (2017) shows that countries with high initial levels of employment protection—and thereby low initial levels of productivity—tend to have a higher probability of big labor market reforms.

Other Resources Citing This Publication