The macro rebalancing in the Eurozone
Price and, even more so, cost competitiveness has improved markedly since real effective exchange rates (REER) peaked in the Eurozone member states, mainly at some point in in 2008-09. Figure 1 shows, however, that the process of external adjustment has been associated with lower investment and, more generally, lower output growth in the stressed economies.
Figure 1. Relative prices and activity in selected Eurozone countries (change between the year of the ULCT-deflated REER peak and 2014 projected)
What happened? Our suggested narrative is that the deterioration of competitiveness during the credit boom left the imbalanced countries, when the financial cycle turned, with no alternative but to pursue internal devaluation. Given labour and product market rigidities, however, the adjustment was initially driven more by compression of demand than by a reduction of costs relative to the other Eurozone countries and the rest of the world. The ensuing shortfall of investment, coupled with labour market hysteresis, produced a significant contraction in potential output.
The evidence looks consistent with this narrative. Figure 2 refers to the subsample of Eurozone economies used in this column, for which we have both micro1 and macro information. For the selected ‘stressed’ economies2 (Italy, Slovenia and Spain) the process of rebalancing has been associated with domestic demand relative to foreign demand (‘relative demand’) contracting more quickly than relative costs in the initial years of adjustment. In comparison, for ‘non-stressed’ economies (Belgium, Finland, France and Germany), on both accounts adjustments were much more moderate and for three of them the relative demand is recovering.
Figure 2. Changes in cost competitiveness and domestic demand relative to foreign demand: Stressed (left) and non-stressed (right) economies (index, 2009=100)
Gauging structural reform effectiveness and needs
Looking further ahead, growth perspectives largely depend on productivity and, in particular, total factor productivity (TFP), the performance of which has been disappointing in the Eurozone both before and during the crisis. This is shown in Figure 3, where the stressed economies are again Italy, Spain and Slovenia, and the non-stressed are Belgium, Finland, France and Germany.
Figure 3. TFP growth in selected Eurozone countries and the US (Annual average GDP growth in various sub-periods; 2000-2014 projected)
However, while the call for productivity-enhancing structural reforms is strong, the particular policies, their transmission channels and their eventual assessment are often left undefined. In the following, drawing from recent theoretical frameworks (e.g. Restuccia and Rogerson 2008), we illustrate how newly collected information from firm-level sources can help provide empirical underpinnings to devise and assess such policies. In particular, we show how micro information can be used to assess whether structural policies enhance resource reallocation towards the most effective use, thereby fuelling sustainable growth.
Growth in aggregate productivity can originate from two distinct channels (see Bartelsman et al. 2013):
- First, the boost can come from general improvements in productivity at the firm level, for example through better (human) capital or technology.
- Second, increases in aggregate productivity can originate from resources (labour and capital) being reallocated to more effective use, i.e. through the selection and allocation of resources to the most productive firms.
In a recent paper using the new CompNet database, we provide empirical evidence on this second channel (Lopez-Garcia et al. 2015). In particular, the database includes 'joint moments' that describe the joint distribution of firm-level productivity and a number of relevant firm-level covariates/determinants, including firm size (see the Appendix for a complete list). Using this information, we can track for each country whether, for example, resources flow from less to more productive firms. Such indicators can aid in disentangling how different types of structural reforms affect the functioning of markets, and offer hints on possible impeding factors. In what follows, we provide two concrete examples of such indicators.
For policy purposes, it would be useful to have an indicator of the extent to which credit is efficiently allocated. CompNet has estimated a firm-level “indicator of credit constraints” (ICC) by using data drawn from the balance sheets and profit and loss accounts of firms (see Box 4 in Lopez-Garcia et al. 2015).
By tracking the joint distribution of productivity with the ICC, one gains insight into the efficiency of credit allocation over time and across countries. The main idea is that bank credit is often provided to firms on the basis of the availability of collateral (or, worse, improper relationships) rather than on the basis of investment quality (proxied by productivity). This may happen for a number of reasons, including agency problems, and can result in credit constraints on highly productive firms with insufficient collateral, rather than on firms/projects below a productivity threshold. When economic uncertainty rises, especially in countries with a distressed banking sector, the inefficiency is likely to increase.
To explore this issue in the case of the Eurozone, Figure 4 shows the share of credit-constrained firms in each decile of the productivity distribution. We compare developments before (2008 or earlier) and during (2009-2012) the crisis for the same two subsets of Eurozone countries.
Figure 4. Share of credit constrained firms by labour productivity decile, stressed and non-stressed economies (Spanish data only available from 2008 on)
For both sets of countries in the two periods, the share of credit-constrained firms falls with firm productivity (i.e. the most productive firms are the least constrained), but even the most productive firms face some constraints. Comparing the pre-Crisis to the Crisis period, while in non-stressed countries credit constraints appear to remain broadly unchanged, they deteriorate substantially in stressed countries. This holds in particular for the least productive firms. It seems that banks let the burden of worsening credit conditions fall disproportionately on those firms least deserving of credit. In a sense, one can say that despite tight credit, the efficiency of capital allocation across firms has improved in countries that are under stress. This calls, for instance, for persevering with structural measures aimed at improving the access of high-potential young firms to finance, especially in the stressed economies.
Labour market reallocation
Turning now to evidence on labour reallocation, we focus on a comparison of the same two sets of countries: Belgium, Germany, Finland and France (non-stressed) and Spain, Italy and Slovenia (stressed). We look how employment has changed in those countries for the most productive versus the least productive firms.
Figure 5. Percentage change in total labour for firms above/below labour productivity median, for stressed (red) and non-stressed (blue) countries
Figure 5 shows the employment growth of firms above or below the median productivity in two different periods. The left charts show the employment growth of firms between 2000-2003 and 2004-2008, that is, the change over a ‘growth’ period, and the right charts show the same information for 2004-2008 and 2009-2012, which reflects the changes triggered by the crisis.
Focusing on the stressed countries (red), employment growth broadly turns negative compared to the boom years prior to the crisis, when overall employment was growing. But what matters here is that low-productivity firms lost proportionally more jobs than high-productivity firms. Hence, the crisis seems to have had a cleansing effect in the stressed countries. Going forward, the data will allow tracking whether recent structural reforms indeed facilitate the reabsorption of labour into the most productive activities as demand picks up.
For non-stressed countries (blue) the reverse pattern is seen. Further exploration by country and industry is needed to fully explain this. One channel may be that some large firms hoarded labour in the crisis, and also shifted to the lower productivity group. Another possibility is that high-productivity exporting firms were especially impacted by the 2009 trade collapse, thus harming overall employment for the highest performing firms. In this case, future data updates will allow tracking of the rebound of employment in exporting firms.
In this column, we provide two illustrations of how we can use indicators built up from firm-level data to assess the impacts of structural reforms, particularly those impinging on efficient resource allocation, in the credit and labor markets, respectively. Using simple joint distributions between firms’ productivity and selected covariates, we show that for stressed EU countries there is evidence that credit and labour is actually being reallocated towards the most productive use following the crisis. This would be in line with the postulate that the crisis and ensuing structural policies may be generating ‘cleansing effects’.
It is obvious that the evidence presented is still preliminary and only suggestive of the potential use of the dataset. One use of the data, for instance, will be to run regressions of observed country/industry/time differences in terms of labour allocative efficiency on institutional factors, sector regulations or credit market conditions. Further, the evolution of the (joint) moments of firms within industries and countries can be used in the calibration or indirect inference of structural models. In so doing, one could disentangle the factors underlying resource reallocation and thus design appropriate policies built on solid empirical evidence. While the current evidence shows cleansing taking place in the downturn in stressed economies, time will tell whether the recent structural reforms will allow the reallocation channel to turn an uptick in activity into strong productivity and output growth.
Authors’ note: The opinions expressed in this column are those of the authors and do not necessarily reflect the views of the European Central Bank. The authors would like to thank Matteo Bugamelli and Paloma Lopez-Garcia for their useful comments, and Benjamin Bluhm and Matteo Iudice for their research assistance.
Bartelsman, E, J Haltiwanger and S Scarpetta (2013), “Cross-Country Differences in Productivity: The Role of Allocation and Selection," The American Economic Review 103(1): 305-34.
Haltiwanger, J (2011), "Firm dynamics and productivity growth", EIB Papers 5/2011, European Investment Bank.
Lopez-Garcia P, F di Mauro and the CompNet Task Force (2015), “Assessing EU competitiveness: The new CompNet micro-based database”, ECB Working Paper Series no. 1764.
Restuccia, D and R Rogerson (2008), “Policy Distortions and Aggregate Productivity with Heterogeneous Establishments”, Review of Economic Dynamics 11(4): 707–20.
1 The micro sample includes firms with more than 20 employees in order to ensure better comparability.
2 “Stressed economies” refers to the Eurozone countries that either are/were participating in a financial assistance programme (e.g. Spain) or, as in the case of Italy and Slovenia, have macroeconomic imbalances that the European Commission has labelled as “excessive”. These countries also experienced significant market turbulence from 2010 until at least the summer of 2012. According to this definition, the other Eurozone stressed economies, on which this column does not focus because of lack of micro data, are Cyprus, Greece, Ireland, Latvia and Portugal.