Eurozone productivity: the long awaited pick-up

Transcription

1 Economic & Financial Analysis Economics 1 May 218 Eurozone Eurozone productivity: the long awaited pick-up But can it last? As the Eurozone recovery matures, a focus on medium-term growth is returning. With population ageing, productivity gains seem to be the key to improved growth potential. Even though the short-term outlook for productivity is improving, a more permanent pick-up is not yet assured, as long as lagging companies fail to catch up. In that regard, structural reforms have yet to produce noticeable effects in key sectors. It could therefore take some time before the Eurozone breaks out of its structural productivity rut even though the trough seems to have passed. From now to 225, we expect 1.3% productivity growth annually. For the ECB, this is a classic Catch-22. It is damned if it continues to keep easy financing conditions in place for keeping unproductive firms alive, and damned if it doesn t as it kills off catch-up opportunities for laggards. The post-crisis period has seen dismal growth both in labour productivity and total factor productivity (TFP) in the Eurozone, even though innovation and technological progress seem to be moving at lightening pace. To paraphrase Robert Solow s famous quote from 1987: you can see the digital age everywhere but in the productivity statistics. While many streams of thought exist in the literature, it seems hard to point to one clear reason for this productivity puzzle. There are indications that the digital age doesn t lend itself very well to GDP statistics, meaning that some of the decline in productivity is simply due to measurement errors, a problem we won t address in detail in this article. Specifically looking at the supply side of the economy, recent evidence shows that productivity growth at the frontier remains strong, but that a large tail of laggards puts a drag on aggregate productivity growth. The businesses at the frontier are typically larger organisations in terms of revenue, younger and more capital intensive. Technologies are not diffusing as quickly to laggard firms, making it more difficult for a long, unproductive tail of companies to improve. Structural reforms could help in this regard, but the pace of reform across Eurozone economies remains weak. At the same time, the financial crisis has had a negative impact on productivity growth as credit constraints have made necessary investments difficult. While this problem has now been addressed, paradoxically the low interest rate environment has also kept unproductive firms alive, thereby weighing on aggregate productivity. Make no mistake, productivity is now definitely improving, but a lot of it is cyclical, due to the better utilisation of both labour and capital. Some composition effects also play a role, as the more productive manufacturing sector is growing more rapidly in this stage of the recovery. Of course the formidable advances in technology over the past decade will take some time to be adopted. But for the productivity gains to be more permanent on a macro level, the investment pick-up will have to be broader and lasting. Bert Colijn Senior Economist Eurozone Amsterdam bert.colijn@ing.com Peter Vanden Houte Chief Economist Eurozone Brussels peter.vandenhoute@ing.com ING.com/ THINK This leaves the ECB in a Catch-22 and perfectly explains ECB President Draghi s constant cry for structural reforms at every ECB press conference. Monetary stimulus helps for the long tail of unproductive firms to catch up with the technological frontier, while at the same time it keeps unproductive firms alive. In the medium term, it therefore leaves the productivity upturn rather modest. We expect overall Eurozone labour productivity growth to come out around 1.3% in the next five to ten years, a pick-up from the previous decade, but still leaving it below its average in the nineties. SEE THE DISCLOSURES APPENDIX FOR IMPORTANT DISCLOSURES & ANALYST CERTIFICATION 1

2 Current productivity gains might be short-lived Since the economic crisis started a decade ago, measured productivity growth has been negative for several years in many countries. This does not mean that technological progress has been reversed, it merely reflects the underutilisation of resources. The lack of demand causes inefficient use of labour and capital and this influences both labour and total factor productivity. When looking at the correlation between productivity growth and the number of businesses indicating a lack of demand as a factor hindering their business, we find a strong inverted correlation between the two in recent years, suggesting that most of the developments in productivity growth since the start of the crisis in 28 can be explained by the varying degree of underutilisation of resources. Fig 1 Eurozone TFP and lack of demand Euro Area, Total Factor Productivity (rhs) "Factors Limiting the Production, Demand", inverted (lhs) Source: The Conference Board Total Economy Database, European Commission, ING Research -4 Given that stronger demand in 218 is likely to lead to a further shrinking of the negative output gap, by definition the utilisation of resources will improve. That means that we can expect improvements in productivity growth in the countries and sectors struggling most with below potential output. Countries such as Spain, Finland and France would appear to have the most to gain in the short term, while countries such as Germany and Austria could benefit less in terms of productivity gains from the economic recovery. Fig 2 Sectoral developments in labour productivity 4.% 3.% 2.% 1.%.% -1.% -2.% -3.% Source: EUKLEMS, ING Research 2

3 From a sectoral perspective, it is mainly the service sectors that have seen an outright labour productivity decline in recent years. That said, productivity gains in the ICT sector, which had been very fast over the 2-8 period, have slowed significantly as well. Underutilisation of resources also seems to play a role here given the lack of demand that businesses in the sectors indicate. Mainly in ICT sectors, such as telecommunication services, we find that the lack of demand is still severe, while also in professional services this seems to be the case. A number of manufacturing subsectors also have a significant amount of slack in their markets, which suggests that the sectoral productivity gains from improved utilisation of labour and capital is likely to be divided between industry and service sectors. Fig 3 Share of manufacturing in GDP and productivity growth Q Q Q Q Q Q Q3 2 Q2 21 Q1 22 Q4 22 Q3 23 Q2 24 Q1 25 Q4 25 Q3 26 Q2 27 Q1 28 Q4 28 Q3 29 Q2 21 Q1 211 Q4 211 Q3 212 Q2 213 Q1 214 Q4 214 Q3 215 Q2 216 Q1 217 Q Labour productivity growth (%YOY) Share of manufacturing in GDP (pp change YOY, RHS) Source: Thomson Datastream That said, there are also (mainly service) sectors that indicate little cyclical demand issues but experience negative labour productivity growth, structurally dragging down productivity growth for the whole economy. This phenomenon is not new of course and relates to the work of William Baumol, with productivity gains thought to be less pronounced in the services sector than in manufacturing. This means that in developed countries, where the weight of the service sector in the economy increases structurally, productivity gains tend to slow. However, this doesn t exclude periods during which the more cyclical manufacturing sector grows more rapidly than the economy at large, temporarily boosting productivity gains because of the compositional effect. We find indeed that labour productivity growth for the whole Eurozone economy has been correlated with the variation of the share of the manufacturing sector in overall GDP. To be sure, this is unlikely to be a permanent effect as the share of manufacturing in the economy is structurally declining. Of course, digitalisation and artificial intelligence have the potential to increase productivity in services (eg, in retail, e-commerce is believed to be two times more productive than store sales), but it might take some time to see the new technologies being adopted and the full productivity gains realised. Structural improvements seem harder, despite technological progress As current productivity improvements seem to rely on better use of factors, it remains to be seen how long the momentum can be sustained. Although difficult to measure, it seems that output gaps are almost closed at the moment, meaning that structural factors influencing productivity need to take over to continue on a path of strongerthan-current productivity growth. There are roughly four streams of thought about why productivity growth has been structurally weak (Brynjolfsson et al, 217: and 3

4 each one seems to add parts of the explanation. The first stream of thought is that the technological progress is actually not that impressive at all. The proponents of innovation pessimism reckon that the gains to be made from the digital revolution are actually limited in scope and see this current period of development as much less impressive than previous developments of general purpose technologies. Second, technologies are successfully implemented by the happy few of this world, but have not become available to all businesses. The third idea is that lags between innovation and implementation simply means that it will take longer for the economic benefits of innovation to be noticed. The fourth explanation is that we are miscalculating productivity. Recent developments are difficult to capture in the framework of the national accounts, a framework designed in the 193s when most production was industrial. This risks understating economic activity and, with that, gains in productivity. This is a potentially important explanation of productivity loss, especially in the US where the digital revolution is further advanced, and could have implications for the entire framework in which we assess economic output. While that is definitely relevant from a wealth perspective, we focus on the first three more measurable factors in this note. A mix of the explanations provides a reasonable attempt at solving the puzzle. The techno-pessimists could be right that the exceptional productivity growth of the postwar period may not return, while some of the improvements in economic activity are probably not accurately captured by GDP. The second and third explanations seem most relevant for the Eurozone recovery of productivity growth in the medium term. Even though the digital revolution might not be as significant for GDP growth as previous revolutions have been, the full effects have yet to benefit the total economy. According to the McKinsey Global Institute, new technologies still take between eight and 28 years from commercial activity to 9% adoption rates. In that regard the productivity gains from the 199s internet boom have started to wane, while the latest wave of technological progress has not yet showed up in the productivity figures. Not taking into account countervailing forces, McKinsey sees the potential of a catch-up to 2% annual productivity growth over the next decade on the back of digitisation, though the impact may not be the same everywhere. Fig 4 Smaller companies have weaker productivity Fig 5 Digital innovations reach smaller businesses at a slower pace Source: Eurostat Source: Eurostat As the OECD (215) has argued, productivity growth at the frontier of technological progress has continued to be fast paced. Dispersion to the rest of the economy seems to be the most challenging. The large number of (not always but often) smaller companies 4

5 that have yet to successfully implement new technologies is weighing down on aggregate productivity improvements. The long tail of companies that are not very productive has been a drag on aggregate data for quite some time and is a more persistent problem in the service sector than in manufacturing, where economies of scale have been more relevant for a longer period of time. As can be seen from figures on cloud computing for example, small businesses have far lower levels of adoption than larger firms and also have not experienced any growth between 214 and 216. The pace of adoption seems to increase with the size of an organisation, as illustrated by the struggles that smaller businesses have in implementing digital technologies. Post crisis financing effects have had an ambivalent impact Another reason productivity growth has yet to pick up is that the crisis has not only impacted productivity through underutilisation of factors, but also through a lack of investment in the post-28 period. Bear in mind that labour productivity growth is determined both by Total Factor Productivity growth (mainly due to technological change and efficiency gains) and capital deepening (the amount of capital input per worker). The hardest hit Eurozone economies have seen a significant tightening of credit conditions for a prolonged period of time. This has hurt investments, generally linked to improved innovation and implementation of new technologies. This phenomenon was especially true for small and medium sized companies, which rely most on bank credit. Thanks to the unconventional monetary policy measures put in place by the ECB, these financial constraints have been ebbing away since 215. Even though growth in investment has been weak over recent years, investment in ICT and intellectual property, which is more related to productivity growth, has been growing faster than in other categories. That may sound positive, but investment growth in these categories has still been weaker than before the crisis, adding to the weakness in productivity growth over recent years. As the investment environment is showing modest improvements and demand for investment increases, it looks like productivity growth stands to gain as the current upturn matures. However, easy financing conditions are not univocally positive for productivity growth. The low interest rate environment helps the weaker, less productive firms survive, in turn keeping the growth environment more anaemic, thereby perpetuating the low interest environment (see, for example, Fig 6 Investment in research and ICT is weaker than pre-crisis, but above average GFCF growth, % 35% 3% 25% 2% 15% 1% 5% % -5% Total Construction ICT equipment Intellectual property Source: Eurostat, ING Research 5

6 The BIS has done work on the number of zombie firms 1 in the Eurozone economy and shows that over the recent period of negative interest rates, the number has surged. In 215, for which the latest data has been calculated, firms that have existed for ten years or more with a ratio of EBIT to expenses below one is around 1%. This low ratio of earnings to expenses suggests that these businesses are a significant drag on aggregate productivity in the Eurozone economy. As interest rates are only expected to increase slowly, it seems that the drag on productivity will remain. Fig 7 Looser financial conditions generally have benefited investment Fig 8 Zombie firms have surged in the period of negative rates Q1 24 Q4 24 Q3 25 Q2 26 Q1 27 Q4 27 Q3 28 Q2 29 Q1 21 Q4 21 Q3 211 Q2 212 Q1 213 Q4 213 Q3 214 Q2 215 Q1 216 Q4 216 Q3 217 Q2 218 Investment non-fin corps (% YOY) Net change in credit standards (2Q lead, RHS) Share of zombie firms in Eurozone 3-month euribor Source: Eurostat, ING Research Source: BIS, Macrobond So while the end of QE could make the life of zombie firms tougher and weed out unproductive activity, it will also have a negative impact on financing conditions of SMEs, which have improved over the years of stimulus. This is especially the case in the periphery of the Eurozone, where investments have picked up far more slowly than elsewhere. The end of the asset purchase programme could therefore have an adverse effect on the pick-up of productivity growth as it could make investments in relevant technologies more difficult. In a sense, the end of QE could provide the first insight into the aggregate effect of low rates on productivity gains. Structural reforms may boost productivity but have slowed A key factor for productivity improvements is the regulatory framework in which Eurozone businesses operate. As the ECB has mentioned time and again, Eurozone countries need to make structural reforms to their business environment to support stronger productivity growth. Creating more competition will weed out zombie firms and reallocate resources to more productive companies. Recent improvements in regulation in labour- and product markets will probably be beneficial to the productivity outlook for the medium term, but the question is how much of that effect has already worked its way into current productivity numbers. Cross-country analysis of improvements in services regulations suggests a weak relationship with productivity growth, but limited data makes the timing of the impact on productivity rather difficult to determine. Some countries have indeed seen product market regulation improve for quite a few sectors. Among the larger Eurozone economies, Italy seems to have made the most significant progress. The appetite for reforms has waned as the economic recovery progressed though, making the outlook for further developments towards more efficient markets subdued at best. This also seems to be the case at the European level. Some studies have shown that older populations are, in general, less in favour of structural reform, making it harder to push through in Europe given the demographic outlook. Efforts aimed at creating a single services market have been going on for some time, but 1 Zombie firms are defined as as listed firms with a ratio of EBIT to expenses below one, with the firm aged 1 years or more. 6

7 even the most optimistic timetable would not foresee a functioning single market in the coming decade. This means that the upside potential for productivity improvements remains high, but that it is to materialise only at a snail s pace over the coming years. While we identify a number of factors that could increase productivity growth over the coming decade in the Eurozone, one must not be blind to a number of adverse factors potentially limiting the projected productivity gains. According to Aiyar, Ebeke and Shao ( the ageing of the workforce in the euro area could lower TFP growth by about.2 percentage points each year between 214 and 235. Fig 9 Product market regulation changes have been favourable during crisis years , decline indicates competition promoting measures Transportation Telecoms Energy Professional services Transportation Telecoms Energy Professional services Transportation Telecoms Energy Professional services Transportation Telecoms Energy Professional services France Germany Italy Spain Source: OECD, ING Research calculations ECB is caught in a Catch 22 So while the economic recovery is currently boosting productivity growth, it seems that this will be rather short-lived given the close to full use of capital and labour. Although technological improvements and innovations can be seen all around, it remains a struggle for a large share of firms to benefit from these changes for a variety of reasons. The precarious relationship between expansionary monetary policy and productivity growth implies a serious conundrum for the ECB. Definitely one that explains Draghi s cry for structural reforms in Eurozone countries at every ECB press conference. As the impact of monetary policy on productivity is hard to determine, it looks like low interest rates and low productivity growth have been caught in a deadlock that only a rapid take-up of technologies by laggard firms can break. This can be boosted by improving market regulation, but as the pace of reform has stagnated, not much is to be expected here. Though we believe that the coming decade is likely to be better in terms of productivity gains than the last decade, it seems that the positive impact of the digital wave will be limited, given the slow adoption in Europe. With an expectation of 1.3% labour productivity growth on average for the period, productivity gains remain significantly below those seen in the nineties, not to mention the post-war period. 7

8 Fig 1 Labour productivity is likely to bounce back just marginally from lows 7% 6% 5% 4% 3% 2% 1% % Source: The Conference Board Total Economy Database, ING Research 8

9 Disclaimer This publication has been prepared by the Economic and Financial Analysis Division of ING Bank NV ( ING ) solely for information purposes without regard to any particular user's investment objectives, financial situation, or means. ING forms part of ING Group (being for this purpose ING Group NV and its subsidiary and affiliated companies). The information in the publication is not an investment recommendation and it is not investment, legal or tax advice or an offer or solicitation to purchase or sell any financial instrument. Reasonable care has been taken to ensure that this publication is not untrue or misleading when published, but ING does not represent that it is accurate or complete. ING does not accept any liability for any direct, indirect or consequential loss arising from any use of this publication. Unless otherwise stated, any views, forecasts, or estimates are solely those of the author(s), as of the date of the publication and are subject to change without notice. The distribution of this publication may be restricted by law or regulation in different jurisdictions and persons into whose possession this publication comes should inform themselves about, and observe, such restrictions. Copyright and database rights protection exists in this report and it may not be reproduced, distributed or published by any person for any purpose without the prior express consent of ING. All rights are reserved. The producing legal entity ING Bank NV is authorised by the Dutch Central Bank and supervised by the European Central Bank (ECB), the Dutch Central Bank (DNB) and the Dutch Authority for the Financial Markets (AFM). ING Bank NV is incorporated in the Netherlands (Trade Register no Amsterdam). In the United Kingdom this information is approved and/or communicated by ING Bank NV, London Branch. ING Bank NV, London Branch is subject to limited regulation by the Financial Conduct Authority (FCA). ING Bank NV, London branch is registered in England (Registration number BR341) at 8-1 Moorgate, London EC2 6DA. For US Investors: Any person wishing to discuss this report or effect transactions in any security discussed herein should contact ING Financial Markets LLC, which is a member of the NYSE, FINRA and SIPC and part of ING, and which has accepted responsibility for the distribution of this report in the United States under applicable requirements. 9