Peter Praet: Structural reforms and long-run growth in the euro area
Intervention by Mr Peter Praet, Member of the Executive Board of the European Central Bank, on panel "Long-run growth, monetary policy and financing of the economy" at the 43rd Economics Conference of the Central Bank of the Republic of Austria, Vienna, 15 June 2015.
As the ECB's accommodative monetary policy is playing its part in the euro area recovery, structural reforms are the domain where there is more still to do to create the conditions for sustainable long-run growth, which is critical to the integrity of our monetary union. There is however no "one size fits all" model for how countries should go about tackling structural challenges. While there are principles that apply across countries, each economy is different and reforms have to be tailored to national conditions. As a central bank, our interest is not in how countries implement reforms, but whether they succeed in doing so.
There are two main channels through which structural reforms can support long-run growth in the euro area, namely through increasing the adjustment capacity of the economy and through raising its potential growth rate. Put differently, reforms can raise both the trend of long-run growth and reduce the fluctuations around that trend. Both aspects are particularly important in a monetary union, which makes structural reforms commensurately more pressing. The environment for introducing structural reforms is better today than for several years: all the conditions are in place for governments in the euro area, individually and collectively, to begin addressing their long-term challenges.
* * *
Ladies and gentlemen,
Thank you very much for inviting me to participate in this panel on "long-run growth, monetary policy and financing of the economy".
For some this may seem like an unusual combination of topics, as it is often argued that monetary policy has no long-run effects. Monetary policy can however affect long-run growth in several ways. First, by increasing certainty over the future price level it can support investment and efficient resource allocation. Second, by reducing fluctuations in the business cycle it can help forestall hysteresis effects. Third, if one takes a "leaning against the wind" view of monetary policy, it can play a role in tempering the financial cycle and misallocation of resources that often comes with it.
But monetary policy is already playing its part in the euro area, so I do not want to dwell on this topic today. I will focus instead on an area where there is more still to do to create the conditions for sustainable long-run growth - that is, structural reforms in the euro area.
The main point I would like to make is as follows. For many euro area countries structural reforms are central to higher long-run growth. And that each economy achieves this is in turn critical to an efficient implementation of monetary policy and, over time, to the integrity of monetary union. But there is no "one size fits all" model for how countries should go about tackling structural challenges. While there are principles that apply across countries, each economy is different and reforms have to be tailored to national conditions.
As a central banker, my interest is therefore not in how countries go about strengthening their economies. This is for governments individually and collectively to decide who know their national challenges best. My interest is in whether they succeed in doing so, due to the impact this has on price stability and on the cohesion of the union as a whole.
In the remainder of my remarks I will elaborate on these points, focusing on the two main channels through which structural reforms can support long-run growth in the euro area.
Increasing adjustment capacity
The first is through increasing the adjustment capacity of the economy. This means two things: first, that output is less affected by shocks because relative prices adjust quickly; and second, that the recovery in output is faster because the economy is able to reallocate resources in a more efficient way. Adjustment capacity is important in itself, but it also supports long-term growth insofar as it reduces the depth of recessions and the duration of unemployment, lessening labour and capital hysteresis.
When countries do not have an independent monetary policy and exchange rate, as in the euro area, the ability to adjust to shocks in this way becomes commensurately more important. To avoid large increases in unemployment it is crucial that national institutions can facilitate a smooth adjustment process. In other words, the need for a higher degree of "resilience" comes automatically with membership of a monetary union.
While this is first and foremost a national responsibility, it is also of common interest in a monetary union. If some members consistently rebound more slowly from shocks than others, it complicates the achievement of price stability at the euro area level. The risk also increases that differences in structural unemployment become entrenched across the Union. And insofar as this weakens the political rationale for monetary union - that all members are better off over time inside the Union than they would be outside - it also weakens its long-term cohesion. That has negative spillovers for all member economies.
It is in this context that several of my colleagues on the ECB Executive Board have recently called for stronger common governance at the European level over economic policies.1 The ability of all countries to adjust well to shocks is not only key for the delivery of our mandate, it is vital for the integrity of the currency. This is an area of legitimate interest for the central bank which is the guardian of that currency.
However, it is important to stress that the ECB's interest in structural reforms should not be misinterpreted as a call for centralisation along a "one size fits all" model. We do not need all euro area countries to adopt identical structural reforms. What we need is a framework that takes into account both how countries differ based on their national conditions, and how they are similar by virtue of being in a monetary union. Within those parameters there are various combinations of country-specific institutions that can produce smooth adjustment.
Let me illustrate what I mean by focusing on one example: labour markets.
Theory suggests that, in a monetary union, there are certain principles that should apply to labour markets across countries. For instance, without the option of exchange rate devaluation, maintaining wages in line with productivity over time is thought to be central to sustaining competitiveness and avoiding painful internal devaluations. Ensuring that wages can respond to changes in labour demand or supply is also seen as a key element in limiting the employment cost of shocks. If firms have the scope to cut costs on the intensive margin - i.e. wages, bonuses and working time - they are less likely to cut costs through the extensive margin, that is, labour shedding.
Wage-productivity links have indeed proven to be an important factor in countries' external positions. France, for example, has experienced large losses in export market shares since the start of monetary union in 1999, and part of the reason for this is that labour market institutions have allowed wages and productivity to de-link across firms. Micro-level data show that wages have grown almost as much in the least productive French firms over the last decade as they have in the most productive. By contrast, in Germany there is clear wage dispersion according to firm productivity.
Moreover, while the strength of the crisis shock has inevitably led to high job destruction, the employment cost does appear to have been influenced by the margins of adjustment available to firms. On the whole those that had scope to adjust on the intensive margin have cut jobs less. For example, new micro level research from the Eurosystem finds that firms with flexibility at the plant-level have reduced employment less during the crisis than those bound by centralised wage bargaining agreements, partly because they have been more able to adjust wages and working time to economic conditions.2
Labour market policies that reflect such principles may therefore be beneficial in most euro area countries. But it does not follow from this that there is an optimal model for the labour market that all must emulate. Economies are complex, and how labour markets function depends on manifold interactions at the national level. Designing structural reforms thus requires a nuanced and country-specific approach.
For example, how the labour market adjusts to shocks depends not only wage formation, but on the overall constellation of labour market institutions within national economies. That is, how the adjustment margins interact with other features such as unemployment insurance, employment protection and active labour market policies - the latter being particularly important in the current context of a "cleansing" recession requiring substantial reallocation and retraining of workers. And we know from international experience that different combinations of labour market institutions can produce similar employment outcomes.
If one compares, for example, the largest euro area economy, Germany, with the US and Denmark, we can see that they have different levels of public expenditure on labour market protection (0.1% in the US, 0.7% in Germany and 2.1% in Denmark), different levels of union density (US 11%, DE 19%, DK 69%) and a different percentage of employees covered by wage bargaining agreements (US 13%, DE 62%, DK 80%). Unemployment rates in all three countries have nonetheless been on a downward trend since 2009.
Moreover, how labour market institutions affect adjustment also depends critically on interactions with other policy areas. In Greece, for example, it took around two years for lower labour costs to translate into lower prices, largely because product markets were highly protected and did not react. Due to structural factors the responsiveness of the economy to price changes has also been relatively low: Greece ranks below most other small economies in terms of both share of foreign trade to GDP and elasticity of exports to price competitiveness. Policy objectives like raising competitiveness therefore have to take broad set of national conditions into account: competition in product markets, the quality of judicial systems and public administration, the ease of doing business, to name but a few.
There are of course other factors that are relevant as well, in particular the role of demand policies. But the key point is about diversity. It is not enough to give one-dimensional prescriptions such as that the all labour markets must become more flexible. What matters is that the combination of policies and institutions within each country produces an outcome that is satisfactory for its citizens and sustainable for the euro area as a whole.
Underscoring this message is important, otherwise it can wrongly seem as though monetary union deprives citizens of democratic choice. One could get the idea that which political orientation a country opts for is unimportant, as it will have to implement to same structural reforms anyway. There are some minimum requirements that come with being part of a monetary union. But there are various ways of meeting them. This is perhaps a notion that, in the future, we could do a better job of conveying.
Reinforcing supply capacity
Alongside adjustment capacity, structural reforms contribute to long-run growth through a second channel: by increasing the supply capacity of the economy, or its potential growth rate. Well-designed structural policies not only increase the quantity of inputs to the production process - i.e. higher labour supply and capital growth - but they can also foster the more efficient use of those inputs across and within sectors, that is, higher total factor productivity (TFP).
The supply capacity channel is also especially important in a monetary union. In the absence of large-scale fiscal transfers and with limited labour mobility across countries, all member economies need to be able to sustain high levels of growth and employment for the Union to be cohesive over the long-term.
Potential growth is however weak in many euro area countries. This is in part because the crisis has lowered both capital growth, through a steep fall in investment, and labour supply, through higher structural unemployment. But it also reflects weak long-term trends in productivity. For example, between 20004 TFP increased cumulatively by only 1.5% in the euro area, while in the US it rose by 10.9% in the same period. This not a situation over which we can be complacent, not least given the substantial damage that still remains from the crisis.
Structural reforms can play an important role, in different ways, in addressing the challenges related to each production factor.
Labour supply in the euro area will inevitably be affected by the impact of ageing societies. Yet with high structural unemployment there is clear scope to increase quantities, especially through labour market policies targeted at reactivating the long-term unemployed. In Portugal, for example, ECB internal estimates find that active labour market policies can explain about one-third of the improvement in employment since the trough. The 2012 Spanish labour market reform is also estimated to have a saved the destruction of about 60,000 jobs in the short term, and had a medium-term impact of some 300,000 fewer jobs destroyed.3 These are admittedly relatively small gains relative to the scale of the challenge, but they provide an indication that determined actions in this area can produce results.
Though the investment-to-GDP ratio in the euro area is currently still 3.5 pp below its pre-crisis level, capital growth is projected to bounce back as the economy strengthens and accelerator effects take hold. However, there are two key risk factors to this outlook, both of which structural reforms can help mitigate.
The first is that pessimism among firms about future growth prospects continues to weaken the business case for higher investment. 5 years ahead growth expectations among forecasters have been falling continuously since 2001, from around 2.7% then to 1.4% today, which may have filtered through into "animal spirits". In this context, structural reforms that raise expectations over the path of potential growth can have an important psychological impact, insofar as they reduce uncertainty and dislodge negative sentiment.
The second risk factor is that the persistence of a debt overhang in parts of the euro area continues to act as a major drag on firm and household spending. Here structural reforms can be supportive both through their positive effect on GDP - the denominator effect - and through facilitating nominal debt reductions - the numerator effect. In the latter case, this comes down to issues such as improving the efficiency of insolvency regimes, out-of-court restructuring frameworks and judicial systems.4
However, we know that it is not just raising the quantity of investment that matters for long-term growth; the quality of investments, and how they contribute to TFP growth, is just as significant. Indeed, it is telling that since 2000 total investment has been slightly higher as a percentage of GDP in the euro area than in the US, while producing a much inferior TFP performance. Two factors can help explain this discrepancy: first, the relatively higher share that US firms have invested ICT capital, and second, the efficiency with which they have turned that investment into productivity gains. And it is structural factors which, at least in part, account for these two differences.
First, while there is variance within the euro area, on the whole there is a relatively higher proportion of micro and small firms in Europe than in the US. European firms are also more static, in the sense that they do not grow or shrink over time. For example, the average size of a manufacturing sector start-up in the US and Italy is roughly the same within its first two years - 5-10 employees. After ten years, however, the average US firm has grown to around 75 employees, while the average Italian still has below 15 employees.5
This matters because the size and growth rate of firms tends to have strong impact on ICT diffusion. Small firms are generally characterised by a relatively lower accumulation of ICT capital due to the higher fixed costs they face.6 They also tend towards higher risk aversion and encounter greater difficulties in collecting resources to finance more innovative projects.7 An economy populated by small firms that do not grow is therefore likely to be one with lower investment in ICT, and therefore lower TFP growth.
The reasons why euro area firms grow more slowly are complex, but structural policies certainly play a role. In Italy, for example, regulations that kick-in at the 15 employees threshold appear to have encouraged firms to stay small (although these are de facto now no longer in force).8 A similar phenomenon has been observed in France, where size-contingent regulations appear to cause firms to cluster below 50 employees.9
Wider product and labour market regulations may also affect firm growth by hindering firm entry and exit and hence discouraging reallocation. For example, differences in product market protection are associated with differences in ICT adoption and diffusion in the service sector.10 Recent evidence suggests that the quality of institutions can also affect reallocation: enhancing the efficiency of civil justice can lead to higher rates of firms' market entry and attract foreign direct investment.11 In this sense, the incentives for within-firm innovation may be closely linked to the capacity for between-firm reallocation.12
However - and this brings me to second point - even in euro area countries that have kept pace with the US in terms of ICT capital, firms have been less able to exploit the productivity potential of their investments. For example, between 2001-07 the average annual contribution of ICT capital to GDP growth was identical in the US and Belgium (0.38 pp), yet TFP contributed 0.53 pp more to GDP growth in the US. Structural and institutional factors can in part explain these differences too.
One key element is differences in the quality of human capital. The impact of ICT on productivity is crucially mediated by the quality of management and its ability to implement the necessary organisational changes required by new technologies. While the US ranks highly in terms of management quality, with some such as Germany close behind, most euro area countries are estimated to have average or below-average management practices.13 This clearly weighs on productivity growth. Indeed, one study finds that management practices account for about one quarter of cross-country and within country TFP variations.14
Addressing this human capital gap above all requires structural policies linked to education and training. But it also has a broader dimension linked to ownership structures and meritocracy. For instance, rigid family ownership structures have been found to be associated with lower management quality, as they limit the talent pool from which firms can draw.15 Such structures also tend to reinforce firm smallness.16
In sum, structural reforms that address the nexus between firm size, organisation and ICT are central to raising TFP growth in the euro area. But let me stress that what I am talking about is not an agenda to promote digital technology firms or build "Silicon Valleys", as important as that might be. What is most crucial for TFP growth is the diffusion of new technology into the ICT-using sector - namely services - where the euro area lags most behind the US. As the largest part of the euro area economy, exploiting ICT in this sector is critical for the euro area to significantly boost its aggregate productivity.
Raising productivity is not a challenge that we can take lightly. It is not only central for a cohesive monetary union based on real economic convergence. It is also a necessary condition of supporting ageing societies. As the European Commission's new Ageing Report shows, the economic age dependency ratio - the ratio between the inactive elderly (65+) and total employment - is projected expected to rise from 44.6% in 2013 towards 66.4% in 2060.17 Only with higher productivity growth can so few sustainably support so many.
Let me conclude.
What I have argued today is that structural reforms can raise long-run growth in two ways: by raising the trend of long-term growth, and by reducing the fluctuations around that trend. Both these aspects are particularly important for economies in a monetary union. This makes structural reforms commensurately more pressing.
This is not to say that all the euro area's problems are structural. Demand policies remain crucial to close a still-large output gap and to secure a strong cyclical recovery. And it is possible that some issues that are currently considered to be structural, such a high long-term unemployment, could reverse in a stronger demand environment. That is to say, if hysteresis operates in the downswing, it may also reverse in the upswing.
Yet according to all estimates potential growth in the euro area is weak, and has been on a declining trend for at least 15 years. A strong and sustained recovery cannot therefore come from demand policies alone. It has to entail reforms that improve the allocative efficiency of the economy and unlock its supply capacity.
The environment for introducing structural reforms is better today than for several years. Monetary policy is extremely accommodative. Activity is recovering. And credit supply constraints are falling, allowing finance to flow quickly to the new investment opportunities that reforms create. All the conditions are therefore in place for governments in the euro area, individually and collectively, to begin addressing their long-run challenges.
1 For instance, see Draghi, Mario (2015). "Structural reforms, inflation, and monetary policy." Introductory speech at the ECB Forum on Central Banking, 22 May 2015.
2 di Mauro, F., and Ronchi, M. (2015). "Centralisation of wage bargaining and firms' adjustment to the great recession: A micro-based analysis." CompNet Policy Brief, No. 8, May 2015.
3 BBVA (2013). Revista Situación España, 2013 Q2.
4 For more on this point see speech by Peter Praet, "Repairing the bank lending channel: the next steps", 17 November 2014.
5 Criscuolo et al. "The dynamics of employment growth: new evidence from 18 countries." CEP Discussion Papers, No. 1274, June 2014.
6 Pellegrino, B., and Zingales, L. "Diagnosing the Italian disease." Unpublished manuscript, September 2014.
7 Amatori, F., Bugamelli M. and A. Colli (2011), "Italian Firms in History: Size, Technology and Entrepreneurship", Banca d'Italia Economic History Working Papers, No. 13, October 2011. See also Bugamelli, M., Cannari L., Lotti F. and S. Magri (2011), "Il Gap Innovativo del Sistema Produttivo Italiano: Radici e Possibili Rimedi", Banca d'Italia Occasional Papers No. 121, April 2012.
8 Schivardi, F., and Torrini, R. (2004). "Firm size distribution and employment protection legislation in Italy." Banca d'Italia Economic Working Paper, No. 504, June 2004.
9 Garicano L., LeLarge C. and J. Van Reenen (2012) "Firm size distortions and the productivity distribution: Evidence from France." NBER Working Papers, No.18841, March 2012.
10 Dabla-Norris, S., 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 discussion Note
11 Lorenzani, D. and Lucidi, F. (2014), "The Economic Impact of Civil Justice Reforms" EC Economic Papers No. 530.
12 Andrews, D. and Criscuolo, C. "Knowledge-based capital, innovation, and resource allocation." OECD Economics Department Working Papers, No. 1046, May 2013.
13 Bloom, N., Sadun, R., and J. Van Reenen. "Americans Do IT Better: US Multinationals and the Productivity Miracle." American Economic Review, Vol. 102, No. 1, February 2012. See also Pellegrino, B., and Zingales, L. "Diagnosing the Italian disease." Unpublished manuscript, September 2014.
15 Bloom, N., and Van Reenen, J. "Measuring and Explaining Management Practices Across Firms and Countries." The Quarterly Journal of Economics Vol. 122, No. 4, November 2007.
16 Bugamelli, M., Cannari L., Lotti F. and S. Magri (2011), "Il Gap Innovativo del Sistema Produttivo Italiano: Radici e Possibili Rimedi", Banca d'Italia Occasional Papers No. 121, April 2012.
17 European Commission (2015), The 2015 Ageing Report.