Europe s Growth Crisis

Transcription

1 Europe s Growth Crisis Kieran McQuinn Central Bank of Ireland Karl Whelan University College Dublin Very Preliminary Draft November 5, 2013 Abstract Even before the financial crisis of 2007/08, there were significant questions about Europe s long-term growth prospects. After a long period of catching up with US levels of labour productivity, Euro area productivity growth had, from the mid-1990s onwards, fallen significantly behind. Using data for the period 1970 to 2006, McQuinn and Whelan (2008) identified declining rates of total factor productivity (TFP) growth and weaker capital accumulation as areas for concern in an European context. In updating this earlier analysis, we find that the growth prospects of the euro area have deteriorated further. TFP growth since 2000 has averaged only 0.2 percent and weak investment is now having a large negative supply-side effect. Europe s demographics are now also contributing to a decline in the workforce. We also provide a medium-term projection for GDP based on unchanged policies and discuss the possible impacts of structural reforms. JEL classification: O40, O47, O16. Keywords: Growth, Capital Deepening, Structural Reforms kieran.mcquinn@centralbank.ie. The views expressed in this paper are our own, and do not necessarily reflect the views of the Central Bank of Ireland or the European System of Central Banks. karl.whelan@ucd.ie.

2 1 Introduction There are many different aspects to the ongoing slump that is affecting euro area economies. Perhaps the most commonly discussed aspect is the series of problems associated with high debt burdens. The euro area as a whole has a level of public debt to GDP that is high by modern historical standards but not particularly high when compared with other economies such as the US, the UK or Japan. However, many euro area countries, such as Italy, Portugal, Greece, Ireland and Spain have high levels of public debt relative to GDP while many economies also have serious problems with the sustainability of various forms of private debt. These problems with public and private debt have lead to substantial fiscal adjustment in many euro area countries while private sectors in many countries are going through a sharp deleveraging process. However, unlike the famous Swabian housewife whose prudent saving has no effect on her neighbours, the widespread belt-tightening across Europe has had a substantial negative effect on GDP. The euro area s slump, in which real GDP is still well short of its pre-crisis peak, has meant that limited progress has been made in improving debt ratios. The euro area s current problems raise the question of whether history provides useful examples of countries that solved substantial debt problems via belt-tightening and debt reduction as opposed to allowing debt to grow while debt ratios fell due to nominal GDP growth. Reinhart and Rogoff s (2009) evidence on the resolution of debt crises points towards nominal GDP growth as a key factor in resolving many debt crises. However, few doubt that the ECB is determined to keep Eurozone inflation below 2 percent. Because many of the countries with significant debt problems need to recover competitiveness relative to the Eurozone s core countries, this suggests inflation rates in these countries are likely to be less than two percent in the coming years which will do little to ease debt burdens. With low inflation in the euro area taken as a given, a return to steady real GDP growth is likely to be the most effective mechanism for restoring Europe s highly indebted governments, businesses and households to sustainable situations. This paper examines prospects for growth in the Eurozone. It follows up on an earlier paper of ours (McQuinn and Whelan, 2008) that focused on trends up to mid While the period of growth in Europe prior to the global crisis of 2008 is sometimes referred to as the boom, our paper had noted that long-run trends in both productivity and per capita hours worked were deteroriating to the point where potential output 1

3 growth in the euro area was at a historical low point and apparently on a negative trend. After a long period of catching up with US levels of labour productivity, Euro area productivity growth had, from the mid-1990s onwards, fallen significantly behind. In this paper, we update our calculations from our 2008 paper, provide projections of Eurozone growth out to 2060 based on recovery scenarios and long-term demographic trends and discuss the potential impact of structural reforms. Because of data restrictions, we restrict our analysis to the twelve countries that participated in the euro area prior to the most recent accessions. 1 Overall, our findings are sobering for those expecting economic growth to deal with the euro area s debt problems over the next decade. Among the results we report are the following. Total factor productivity (TFP) growth in the twelve country euro area group has declined in each decade since the 1970s. Over the years , a mix of years that includes multiple slow-downs and expansions, TFP growth has averaged only 0.2 percent per year. The slump in investment due to the downturn is now having significant negative supply-side effects. We estimate capital stock growth, which averaged 2.4 percent over the past two decades, is now down to about 1 percent. Using a capital elasticity of one-third, this is currently reducing the supply-side growth potential of the euro area economy by about 0.5 percentage points per year. While Europe s demographic ageing is often presented as a longer-term problem that will cause problems relating to pension systems in the future, the ageing process is affecting Europe s growth potential right now. Eurostat projections show the work-age population of the euro area declining from 2014 onwards. Combined with a downward trend in the average workweek and a likely flattening of participation rates, we project a small decline in hours worked in the euro area over the next decade followed by larger declines in subsequent decades. With a medium-term projection of TFP growth of 0.2 percent and gradual declines in the work-age population and average workweek, we project a baseline real GDP growth rate 1 These are Ireland, Belgium, the Netherlands, Luxembourg, France, Spain, Portugal, Germany, Finland, Austria, Greece and Italy. We apologise in advance for using the short-hand euro area to describe out analysis, though these countries do account for the vast majority of euro area GDP. 2

4 in the euro area of less than one percent over the next decade even if the unemployment rate and investment share of GDP return to their pre-crisis levels by The most obvious response to these gloomy projections is that Europe can implement various structural reforms to product and labour markets that can boost labour force participation and productivity. We discuss one such set of reforms, a wide-ranging package of measures recently discussed in the OECD report of Johansson et al (2013). While these reforms are to be welcomed and would have a positive effect on growth, we estimate that a package of reforms to ease product market regulations, delay the age at which people can retire and reform labour markets, would boost euro area GDP growth by only 0.4 percent per year. While this cumulates to large effects, when coupled with our baseline result, it would still not be enough to restore the euro area economy to growth rates over one percent. We conclude that Europe s policy makers need to focus on addressing the current shortfall in aggregate demand as well as on improving aggregate supply and that a long-term policy of encouraging immigration may need to be an important element in keeping the European economy growing. 2 Accounting for Growth in Europe In this section, we review the historical sources of growth in the euro area and compare these euro area s performance with that of the US. 2.1 Growth Accounting We start from a standard assumption that output is produced according to a Cobb-Douglas production function Y t = A t K α t L 1 α t (1) where Y t is real GDP, K t is capital input, L t is labour input (defined in this paper as total hours worked), and A t is total factor productivity. Output growth can then be written as Y t A = t K +α t L +(1 α) t (2) Y t A t K t L t 3

5 With data on output growth, capital growth, and labour growth in hand, this equation can be used to calculate TFP growth. European statistical agencies do not provide a measure of the productive capital stock, so we construct this series ourselves. We assume that the initial stock of capital in 1970 equals the steady-state value implied by the Solow growth model in this year (this is discussed in further detail below) based on prevailing trends at that point for GDP growth, the investment share of GDP and the growth rate of labour input. We then derive the rest of the series from a perpetual inventory method according to the definition K t = (1 δ)k t 1 +I t 1 (3) with a depreciation rate of six percent per year. While these are very specific assumptions, we do not find that our calculations about the recent performance of the euro area economy are significantly affected by reasonable changes to these assumptions. The issue of the elasticity of output with respect to capital, α, is a tricky one. Traditionally, a value of about one-third has often been used based on the observation that for countries with good income-side national accounts such as the United States, the labour share of income has traditionally been around two-thirds. However, this share has declined since the 1980s in Europe and has also been declining in recent years in the US. While there seems to be an absence of any single clear explanation for this patter (see Elsby, Hobijn and Sahin, 2013, and Lawless and Whelan, 2011) the changes in this share seem to have little to do with structural changes in the substitutability of capital and labour. In the absence of clear guidance from income data, our approach here is to keep things simple and use the standard value of α = 1 3 for all cases. The data in our paper are annual and mainly cover the period 1970 to Output, employment levels, unemployment rates, total population and investment levels are all sourced for each country from AMECO - the annual macro-economic database of the European Commission s Directorate General for Economic and Financial Affairs (DG ECFIN). 2 Information on the average work week is taken from the Groningen Growth and Development Centre (GGDC). 3 Table?? presents results for the euro a rea and the US of the growth accounting exercise which allocates output growth according to its three components while Figure 1 charts the com- 2 Available online at: 3 Available online at 4

6 ponents of the decomposition for the euro area. A number of trends are evident from Figure 1. While there are significant cyclical movements in the growth rate of total hours worked in the euro area, there is an underlying downward trend. Over the period , growth in total hours worked in the euro area averaged only 0.1 percent per year. The growth rate of TFP has also gradually declined over time. We provide estimates for each of the ten year intervals preceding TFP growth was running at 2.7 percent in the first half of the 1970s, fell to 1.7 percent over , to 1.4 percent over and to 0.8 percent over The period from has seen TFP decline at an average rate 0.3 percent per year, mainly due to a 3 percent decline in Capital stock growth in the Euro area was over 4 percent per year in the early 1970s but then fluctuated between 2 and 3 percent between the late 1970s and Recent years, however, have seen a significant decline in the investment share of GDP and the euro area capital stock is now growing at an annual rate of about 1 percent. The logic of neoclassical growth models tells us that TFP growth will be the key determinant of labour productivity growth over the next few decades. This raises an important question: What do the these calculations suggest is going to be the likely trend rate of growth for TFP in the coming years? We believe that the average growth rate over the period of 0.2 percent per year may well represent a reasonable value for the medium-term trend growth rate for TFP in the euro area. This period incorporated a number of years of expansion and falling unemployment as well as two periods of rising unemployment so the total cyclical effect on this average is probably quite low. An alternative way of extracting an underlying trend from these figures is to use a fliter. Applying a Hodrick-Prescott (HP) filter (with a value of λ = because of our use of annual data) we estimate an underlying annual trend growth rate for TFP of only 0.09 per percent. A final reason to project a low rate of TFP growth in the coming years is the evidence for the US. Applying our methodology to the US data, we estimate that TFP growth has also slowed significantly. Indeed, we estimate an average TFP growth rate of only 0.3 percent over the period Productivity performance in the US in the current expansion has been relatively poor, particularly, when considering that the early phases of expansions usually see strong productivity 5

7 growth due to cyclical events. Robert Gordon (2012) has detailed a number of reasons why productivity growth is likely to be low in the coming years in the US. He points out that, from a very long-term perspective, US productivity growth has been falling since the 1950s and that the current round of innovations in areas such as healthcare and information technology are less transformative than previous waves of innovation. Taken together, these considerations suggest that, in the absence of important changes to economic policies, there is little reason to expect a substantial improvement in TFP growth in the euro area in the coming years. Table?? presents an alternative accounting breakdown of the growth performance of the euro area and the US. Using the following identity: Y t L t = Y t L t A t A t +α ( K t K t ) L t L t (4) labour productivity growth can characterised as a function of TFP growth and capital deepening (growth in capital per unit of labour). Our calculations show that capital deepening has fallen by a smaller amount than TFP growth. For the period 2000 to 2012, our calculations ascribe 0.7 percent of the 0.9 percent average growth rate of productivity to the capital deepening effect. This pattern, in which capital deepening displays a different pattern over time from TFP growth, can persist for some time. However, as we discuss below, with a stable investment share of GDP ultimately a slow rate of TFP growth will also translate into a slow rate of capital deepening. In fact, the steady-state growth rate of a Solow model economy is g 1 α where g is the growth rate of TFP; with a value ofαof one-third, this equates to1.5g meaning that along a steady growth path, only one-third of the growth in output per hour is due to capital deepening. 6

8 Table 1: Decomposition of Euro Area and US Output Growth Rates (%) Euro Area United States Period y a k l y a k l Note: The table shows the contribution of growth in labour inputs, capital inputs and TFP to total output growth. 7

9 Table 2: Decomposition of Euro Area and US Output per Hour Growth Rates (%) Euro Area United States Period ( y l) a ( k l) ( y l) a ( k l) Note: The table shows the contribution of capital deepening and TFP to the growth rate of labour productivity. 8

10 Table?? presents the different components of growth for the sub-periods and for all of the countries in the euro area 12 grouping as well as the UK and Sweden to provide some non-euro-area comprisons. Over the period 2007 to 2012, only Germany, Belgium, and Austria within the Euro area had a positive average rate of TFP growth. Over the longer perspective of , the average rate of TFP growth in the twelve euro area countries in our sample ranged from 0.9 percent in Ireland to minus 0.7 percent for Luxembourg. In general, however, there is a striking pattern of low rates of TFP growth. For example, Germany and Finland had the highest average TFP growth rates in the euro area over this period apart from Ireland but their average growth rates of 0.6 percent per year were still lower than achieved by the euro area as a whole in each of the ten-year periods prior to Table?? repeats the decomposition of growth in output per hour since 2000 into TFP growth and a capital deepening effect. Over the period , ten of the fourteen countries examined here had capital deepening effects contribute more to growth than TFP improvements and the other four had contributions from capital deepening that were larger than would be sustainable along a steady growth path. Worryingly, many of the worst performers in relation to TFP growth since 2000 are those with the large stocks of public debt: Italy, Greece, Portugal and Belgium all recorded negative TFP growth over the period

11 Table 3: Decomposition of Output Growth Rates for European Countries (%) Country y a k l y a k l Belgium Germany France Greece Ireland Italy Spain United Kingdom Sweden Finland Lux Portugal Austria Netherlands

12 Table 4: Decomposition of Output Per Worker Growth for European Countries(%) Period ( y l) a ( k l) ( y l) a ( k l) Belgium Germany France Greece Ireland Italy Spain United Kingdom Sweden Finland Luxembourg Portugal Austria Netherlands

13 2.2 Factors Determining Labour Input Labour productivity and hours worked together determine the path of GDP in any economy. Tables 2 and 4 provided a decomposition of the determinants of labour productivity growth, Table?? provides a decomposition of the percentage change in total hours worked as a function of four different components: Changes in population, the participation rate, the employment rate and the average workweek. Figure 2 provides graphical evidence on these four variables. It shows that in the period since 1970, population growth has fluctuated around a relatively low average value of about one half of one percent. Growing female labour force participation has driven a significant increase in the fraction of the population that is available for work. Offsetting this factor, the average workweek has declined steadily and unemployment rates have risen to a much higher average value than those recorded in the early 1970s. Rather than focus in detail on the factors that have driven past fluctuations in labour input, we want to focus on the outlook for the future. A number of patterns are now in place that point against significant increases in total hours worked in the euro area economy in the next few decades. The population of the euro area 12 group of countries is expected to grow very slowly over the next few decades. Eurostat projections show 0.2 percent per year growth in population between 2013 and 2023 followed by a gradual reduction in population growth until population begins declining in 2048 (see Figure 3). More seriously, the euro area s population in the normal working-age bracket of 15 to 64 years old is set to peak at current levels and then begin declining. Eurostat project that the workage population will fall at an average annual rate of 0.13 percent in the decade finishing in 2023 and then fall at a rate of 0.4 percent per year in the subsequent decade. Among those are in the working-age bracket of 15 to 64 years old, the increase in participation rates has gradually tailed off. This represents female labour force participation reaching a plateau in many countries but also the effects of population ageing. Participation rates fall off as people get closer to official retirement age so the ageing of Europe s population is likely to limit further gains in the participation rate. This is not an issue that is limited to Europe. Indeed, there is evidence that declining participation and population ageing are 12

14 already having a significant restraining effect on employment growth in the US. 4 There is also no evidence, as of yet, that the trend decline in the average workweek is about to end. Some of this decline likely reflects increase female labour force participation in parttime employment so, as this source of increased participation flattens out, the trend decline in the average workweek may also cease. Unlike the previous pattern of weak TFP growth being more of a factor with high-debt countries, population ageing is an issue that appears likely to affect most euro area countries equally. Figure 3a shows that Eurostat projections anticipate substantial declines in the fraction of the population aged between 15 and 64 in all twelve of the member states analysed here. 4 See Fallick, Fleischman and Pingle (2010) and Stock and Watson (2012) for a discussion of the effect of ageing and labour force participation on the US labour market. 13

15 Table 5: Decomposition of Growth in Hours Worked (%) Euro Area Period Total Pop. P. Rate Emp. Rate Workweek United States Period Total Pop. P. Rate Emp. Rate Workweek Note: Pop. refers to population, P. is the participation rate, Emp. is employment and Workweek is average hours worked by employees. 14

16 Table 6: Decomposition of Growth in Hours Worked (%) Period Total Pop. P. Rate Emp. Rate Workweek Belgium Germany France Greece Ireland Italy Spain United Kingdom Sweden Finland Lux Portugal Austria Netherlands Period Total Pop. P. Rate Emp. Rate Workweek Belgium Germany France Greece Ireland Italy Spain United Kingdom Sweden Finland Lux Portugal Austria Netherlands Note: Pop. refers to population, P. is the participation rate, Emp. is employment and Workweek is average hours worked by employees. 15

17 3 Longer-Run Outlook We have outlined a number of medium-term trends that provide reasons to be downbeat about growth in the euro area in the coming years: Trend growth in TFP appears to be very low and ongoing demographic patterns are going to have a negative impact on the growth rate of employment in the euro area. Still, there are a number of aspects of the euro area economy that are very likely to improve from their current position. Even with a slow recovery constrained by fiscal adjustment, private balance sheet problems and tight bank credit, unemployment should fall from its current high level of over 12 percent. Similarly, the negative supply-side effects of the current low rate of investment are likely to be unwound over time as the economy moves more firmly out of its current slump. How are these positive and negative factors likely to shape economic growth in the coming decades? Here, we report results from a simulation of a simple supply-side model that projects growth using recent trends for TFP, uses the demographic projections just described and also assumes an unwinding of the cyclical problems with low investment and high unemployment. The model s ingredients are described as follows: Y t = A t K α t L 1 α t (5) K t = (1 δ)k t 1 +I t 1 (6) L t = (1 u t )(p t Pop t ) H t (7) I t = s t Y t (8) loga t = g (9) The evolution of the capital stock depends on last period s rate of investment which we project as a time-varying ratio of total real GDP. Labour input is is modelled as a product of the employment rate (1 u t ), the participation rate for those in the workage age population, (p t ), the working age population, (Pop t ) and the average length of the workweek (H t ). The assumptions underlying the simulation are as follows. TFP growth is assumed to continue to grow at its average of 0.2 percent. The ratio of investment to GDP is projected to recover by 2020 to a rate of 21 per cent and 16

18 is constant thereafter. The rate of unemployment falls to a rate of 8 per cent by 2020 and is also constant thereafter. The participation rates of the workage population flattens out at 75 per cent. The average work week continues to decline over the period 2012 and 2020 at the same rate as the 2000 to 2012 average and then stays flat at 29 hours. The workage population follows the Eurostat projection described in Figure 3. Figures 4 and 5 provide graphic illustrations of these assumptions. Before describing the results of this simulation, it may be useful to provide some observations on the analytics of the model. This simulation is essentially a practical application of the famous Solow model of economic growth. One useful way of thinking about this model comes from an alternative decomposition for output introduced by Hall and Jones (1997). Defining the capitaloutput ratio as X t = K t Y t (10) Output per hour can now be expressed as Y t L t = A 1 1 α t α 1 α Xt (11) This decomposition has been used in a number of previous studies, most notably by Hall and Jones (1997). DeLong (2003) shows that the capital-output ratio in this model follows a so-called error-correction equation of the form X t = λ(x X t ) (12) such that it adjusts towards a long-run or steady-state level determined by where the adjustment speed is X = g 1 α s (13) +n+δ. g λ = (1 α)( +n+δ). (14) 1 α 17

19 McQuinn and Whelan (2007) use data from the Penn World Tables to show that convergence speeds for the capital-output ratio tend to conform closely to the Solow model s predictions. Relative to the more familiar decomposition of output per hour into TFP and capital-per-hour terms, this decomposition has an important advantage. The long-run capital-output ratio can be shown to be independent of the level of A t, something which is not true of capital-per-hour. Hence, this formulation completely captures the effects of A t on long-run output, while the more traditional decomposition features a capital deepening term that depends indirectly on the level of technology. These calculations show that, over the long-run with constant values for g and n, the capitaloutput ratio converges to its steady-state. Thus, equation (??) tells us that all growth in output 1 α per hour ends up being due to the At. This term grows at rate g 1 1 α. Thus, in our example with TFP growth of g = per year and a value of α = 1 3, we end up with a long-run steady-state growth rate of g 1 α = or only 0.3 percent per year. We estimate, however, that the current value of the euro area s capital-output ratio is about 2.7 while its long-run steady-state estimate, based on a projection of 0.2 percent per year TFP growth and a slightly negative growth rate of labour input, is about 3.4. This means that the model will generate growth in output per hour that is greater than 0.3 percent along a transition path. The pace of convergence, λ is also slow when the pace of economic growth is as weak as we project. With our depreciation rate of six percent per year, the pace of this convergence is about four percent per year so the convergence dynamics of the capital-output ratio are projected to last for a long time, meaning a very gradual transition to the steady-state growth rate of output per hour of 0.3 percent per year. Figure 6 shows the growth rates in output and output per hour generated by this model. The two positive cyclical factors play some role in boosting growth in the years up to Rising investment leads to a gradual increase in the growth rate of the capital stock from 0.75 percent in 2013 to a peak of 1.44 percent in 2021 before gradually declining again. The reduction in the unemployment rate leads to a temporary increase in total employment up to 2020 before demographic patterns reassert themselves and produce a gradual decline in employment. (See Figure 7.) Our projected decline in the average workweek, however, undoes the positive effect of rising employment so that total hours worked declines by about 0.06 percent in the decade ending in

20 Putting these figures together, we project that output per hour will grow at an average rate of 0.61 percent over the period while GDP will grow at an average rate of only 0.55 percent. After this decade, output per hour is projected to grow in the subsequent decades at average rates of 0.67 percent in (declining employment leads to a temporary boost via a capital deepening effect), 0.54 percent in , 0.42 percent in as this rate gradually eases towards its steady-state value of 0.3 percent. (Figure 8 illustrates the gradual transition of the capital-output ratio to its long-run steady-state level.) With falling hours worked, euro area GDP grows at 0.31 percent per year in before growth almost ceases in the subsequent decades, with falling labour input almost offsetting small increases in labour productivity. 4 The Potential Impact of Structural Reforms The scenario we have just painted is a fairly grim one. However, it is based on our best estimate of the current medium-term trend for TFP growth in the euro area as well as projections of demographic trends that are reasonable extrapolations of where Europe is heading without significant changes in areas of labour policy such as retirement ages or migration. Indeed, it could be argued that our assumptions about a gradual return to normal cyclical conditions in the coming years is, perhaps, a bit optimistic. High unemployment rates may persist for longer than we have projected because of hysteresis effects while the investment rate may not recover as far as we have projected because of diminished opportunities for investment in a slow growing economy. It is reasonable to ask, however, whether Europe s governments will permit an outcome as poor as the one we project here to occur. We noted in our previous paper on this topic that there were many reforms of product and labour markets that could potentially help to boost labour productivity as well as hours worked. While we produced scenarios based on reforms achieving certain positive outcomes, our analysis was short on details of which specific reforms could be undertaken. In the years since that paper, there has been significant progress in the collection and provision of information on labour market and product market structures by organisations such as the OECD and the European Commission as well as new research detailing the potential growth effects of specific structural reforms. Here, we focus on one such set of estimates, from OECD researchers Johansson et al (2013). These researchers consider the combined impact of three different kinds of reforms. 19

21 Product Market Reforms: They consider a case in which all the OECD s measure of the restrictiveness of product market regulation is adjusted to equal the average level of regulations in the five best practice countries in 2011 (i.e. the United States, the United Kingdom, Ireland, Canada and the Netherlands). Labour Market Reforms: They consider a case in which the replacement rate of unemployment benefits is reduced by 10 percentage points, the tax wedge is cut by 4 percentage points and the OECD s measure of the volume of active labour market policies is increased by one standard deviation. Retirement Reforms: They consider a case in which the average duration of individual active life slowly converging in all countries towards the standard observed in Switzerland. Figure 9 shows the combined estimated effects of the introduction of these policies by 2060 on the level of GDP in each of our twelve euro area countries. 5 The estimated effects differ widely from 3.7 percent in Ireland to over 30 percent in Italy. These larger estimates, however, are projected to take place over a very long period of time: On average, these reforms boost Italian GDP by 0.6 percentage points per year. Weighting the annual effect on growth of these reforms by 2012 GDP produces an average effect of 0.4 percentage points per year. Figure 10 puts this effect in context by showing how this additional growth would change our baseline scenario. While such additional growth would be welcome, it would still see euro area GDP growth average below one percent over the next two decades. Many caveats should be applied to this analysis. This is one specific set of estimates of the effects of one specific set of structural reforms. There are likely to be many concrete measures that have positive effects on both productivity and employment and many could have stronger effects than those estimated here. However, these calculations are a useful reminder that structural reforms are not necessarily going to be an elixir that restores Europe to a fast pace of economic growth. 5 The OECD reported estimates of the effect of each country introducing these reforms on their own as well as the effect of all countries introducing these reforms. The latter effects are generally smaller because they estimate an increase in global interest rates as a result. Here we are reporting the more partial equilibrium calculations that do not include this global interest rate effect. 20

22 5 Conclusions Despite the need for the euro area to return to faster rates of economic growth to deal with severe balance sheet problems affecting many of its members, the current set of supply-side trends are not at all encouraging for the growth prospects over the medium- and long-term. TFP growth is very poor, investment is low and demographic problems are going to restrain the growth of labour input in the coming years. Combined with a series of demand-side problems, such as tight fiscal policy and restrictions on bank credit, the prospects for the euro area economy over the next decade do not look so good. Indeed, even allowing for a return of investment and unemployment to pre-crisis levels, we project growth in the euro area that is well below one percent over the next decade. In this sense, despite the intense focus on debt levels, the euro area is facing a growth crisis as much as it is facing a debt crisis, with the latter perhaps more a symptom of the former. The standard response to the current slump from Europe s policy-making community is that the way to restore growth is via a series of reforms of labour and product markets. There can be no doubt that many of the institutions that govern these markets are in need of reform. However, as illustrated by the OECD calculations reported here, we have little concrete evidence that these reforms can offset the negative factors likely to affect long-term growth trends. In relation to the future direction of policy-making in Europe, our results certainly support the need to introduce the kinds of structural reforms of product and labour markets recommended by the OECD and the European Commission. However, it is likely that these reforms will be easier to implement when Europe s economy has more fully recovered from its current slump and benefits will probably take a long time to have a positive impact. Two other areas of policy are worth emphasising, one short-term and one long-term. In the short-term, the focus of policy-makers on supply-side reforms should not distract from the fact that, at an unemployment rate of over 12 percent, the euro area s current level of GDP is well short of its current supply potential. Morever, as we document here, the current cyclical slump in investment is having a substantial negative effect on the growth in supply-side capacity. The euro area s current ratio of public debt to GDP is high by modern historical standards at 95 percent but many of its member states are able to borrow at very low rates and the pricing of ESM-issued securities shows that there are few concerns about the solvency of the euro area as 21

23 a whole. There is thus a strong economic case for a large investment program aimed at reducing unemployment and raising the supply capacity of the economy, funded by the euro area as a whole. Given that public capital funding is usually the first item slashed when governments cut back on spending in a crisis, it is likely that many of the projects funded by such an initiative involve spending that will need to be undertaken at some time in the future anyway so a program of this sort may have limited long-run effects on debt levels. Unfortunately, Europe s political constraints are likely to rule out such a programme for the foreseeable future. Over the longer-term, Europe needs a plan for dealing with a pattern of population ageing that is set to have enormous effects. Policy initiatives to delay retirement ages and to encourage labour force participation are undoubtedly part of the solution to the problems posed by ageing. However, these initiatives are likely to be very unpopular politically and may have negative implications for productivity. A policy of large planned increases in the amount of immigration into the EU, while also politically challenging, may turn out to be the only way to keep the European economy expanding in the future. 22

24 References [1] DeLong, J. Bradford (2003). Macroeconomics, McGraw-Hill. [2] Fallick, Bruce, Charles Fleischman, and Jonathan Pingle (2010). The Effect of Population Aging on the Aggregate Labor Market. in Labor in the New Economy, edited by Katharine G. Abraham, James R. Spletzer, and Michael Harper. University of Chicago Press for the National Bureau of Economic Research. [3] Gordon, Robert J. (2012). Is U.S. Economic Growth Over? Faltering Innovation Confronts the Six Headwinds, NBER Working Paper [4] Michael Elsby, Bart Hobijn, Aysegul Sahin (2013). The Decline of the U.S. Labor Share, Federal Reserve Bank of San Francisco Working Paper [5] Hall, Robert and Charles I. Jones (1997). Why Do Some Countries Produce So Much More per Worker than Others?, Quarterly Journal of Economics, 114, [6] Johansson, Asa, Johansson, Yvan Guillemette, Fabrice Murtin, David Turner, Giuseppe Nicoletti, Christine de la Maisonneuve, Philip Bagnoli, Guillaume Bousquet, Francesca Spinelli (2013). Long- Term Growth Scenarios, OECD Economics Department Working Papers, No. 1000, OECD Publishing. [7] Lawless, Martina and Karl Whelan (2011). Understanding the Dynamics of In?ation and Labor Shares, Journal of Macroeconomics, Volume 33, pages [8] McQuinn, Kieran. and Karl Whelan (2007). Conditional Convergence and the Dynamics of the Capital-Output Ratio, Journal of Economic Growth, Volume 12, [9] McQuinn, Kieran. and Karl Whelan (2008). Prospects for Growth in the Euro Area, CESifo Economic Studies, Vol 54(4), pp [10] OECD (2013). Economic Policy Reforms, Going for Growth. [11] Reinhart, Carmen and Kenneth Rogoff (2009). This Time is Different: Eight Centuries of Financial Folly. Princeton University Press. [12] Stock, James and Mark Watson (2012). Disentangling the Channels of the 2007Ű09 Recession, Brookings Papers on Economic Activity, Spring, pages

25 Figure 1 Euro Area Output and Output Constitutants Growth (%): GDP Growth 5.0 Capital Stock Growth Total Hours Growth 5 TFP Growth

26 Figure 2 Euro Area Labour Market Outcomes: Population Growth 0.50 Fraction of Population in Labour Force Unemployment Rate 40 Average Weekly Hours Per Employee

27 Figure 3 Euro Area Future Demographic and Employment Trends 340 Total Population (Millions) 215 Population Aged (Millions) 0.68 Fraction of Population Aged

28 Figure 3a Proportion of Total Population in 15 to 64 Age Bracket France 65.0 Belgium Portugal Lux % 60.0 Austria % 60.0 Ireland 57.5 Finland 57.5 Netherlands Italy 55.0 Germany 55.0 Spain Greece

29 Figure 4 Baseline Future Euro Area Output Components 21.5 Investment Rate 3 TFP Growth % %

30 Figure 5 Baseline Labour Market Assumptions 11.5 Unemployment Rate 33.0 Average Work Week % hours Participation Rate %

31 Figure 6 Baseline Future Euro Area Growth Rates 4 Output 4 Output Per Hour % -1 %

32 Figure 7 Baseline Future Euro Area Labour Supply (Millions) 150 Total Employment 235 Total Hours

33 Figure 8 Baseline Future Euro Area Capital Variables 3.4 Capital Output Ratio 5 Capital Stock Growth Ratio 2.6 %

34 Figure 9: The Effect of Ambitious Structural Reforms on GDP Per cent

35 4 Figure 10 Baseline and OECD Scenario Output Growth % Baseline OECD Reform