Recent European economic growth: why can't it be like the golden age?

Crafts, Nicholas

This paper evaluates claims that European productivity growth has been undermined by weak competition and excessive regulation and taxation. The argument has some validity in the context of the information and communications technology (ICT) revolution which has placed a greater premium on flexibility. The UK has been well-placed to benefit from ICT but experiences a productivity gap reflecting historically-low investment combined with high employment of low-productivity workers. Rapid Irish growth has benefited greatly from low corporate taxation but the massive contribution of ICT production to productivity growth suggests it is a special case not a role model.

Keywords: Economic growth; ICT; Product market regulation; TFP growth

JEL Classifications: N14; O47

**********

From the early 1950s to the mid-1990s, labour productivity grew more rapidly in western Europe than in the United States. During the Golden Age which ended in the early 1970s, a large initial American productivity lead was quickly reduced by rapid European catch-up growth. From the Golden Age to the mid-1990s growth on either side of the Atlantic was much slower but catch-up continued, albeit at a much slower pace. Since the mid-1990s, American productivity growth has revived while European productivity growth has fallen further and catch-up has generally disappeared. These developments are summarised in tables 1 and 2. (1) A standard American perspective on recent European growth performance is that it has been impaired by too much taxation, too much regulation and too little competition (Baily and Kirkegaard, 2004). Handicapped thus, European countries have failed to take advantage of the opportunities for faster growth that the ICT (information and communications technologies) age presents.

An argument of this kind might be based on the micro foundations of endogenous growth and can also be understood in terms of the concept of 'social capability' introduced into the analysis of catch-up growth by Abramovitz (1986). In the model represented in figure 1, which is adapted from Carlin and Soskice (2006), the downward-sloping relationship captures the relationship between technological progress and capital intensity in the Solow model for a given savings rate. Technological progress is taken to be endogenous and to depend on the expected returns to innovative effort which are greater with higher capital (and output) per worker as depicted in the upward-sloping Schumpeter relationship.

[FIGURE 1 OMITTED]

The Schumpeter relationship is shifted to a more favourable position when the incentives to innovate are strengthened at any given capital per worker. In standard endogenous innovation models this could reflect factors such as a lower marginal cost of R & D or lower direct taxes. In more recent versions of these models, increases in competition (and deregulation which is pro-competitive) can be important in promoting favourable shifts in the Schumpeter relationship. This may be based on the difference between pre-innovation and post-innovation rents in the presence of entry threats (Aghion et al., 2001) or greater competition may speed up the adoption of new technologies which require managerial effort in firms with agency problems (Aghion et al., 1997).

The theoretical arguments seem persuasive yet, according to the OECD, the trend in Europe since the 1980s has been towards deregulation while competition has been intensified by globalisation and, in particular, greater integration of European markets. Until recently, stricter regulation did not preclude faster productivity growth in Europe than in the United States. Moreover, if these arguments are as powerful as American critics of Europe believe, is it not paradoxical that the UK, which has pursued the policies that they advocate more vigorously than other European countries and which has had a sizeable productivity gap to make good, has not experienced outstanding productivity growth (Cotis, 2006)? Perhaps, however, the 'Celtic Tiger' episode of dramatic catch-up by Ireland does vindicate the American diagnosis since it is widely held to be explained in significant part by a very low corporate tax rate and the FDI that it has attracted.

In the light of this discussion this paper will review the evidence relating to the American perspective on current European productivity performance provided by recent research. The following questions will be addressed:

* Why has the pre-1995 productivity catch-up by the European countries ceased?

* What explains the paradoxical combination in the UK of seemingly well-designed policy reform, considerable scope for catch-up but unexciting productivity growth ?

* What are the lessons from recent Irish experience for the European productivity agenda?

The analysis will take a long-run view of recent productivity growth by placing it firmly in the context both of the Golden Age and the subsequent growth slowdown.

The Golden Age of European growth

The period between 1950 and 1973 is conventionally known as the Golden Age of European economic growth. Table 1 reports rates of growth of real GDP per hour worked of over 4 per cent per year for most western European countries and, as the regression in table 2 confirms, countries with the lowest initial productivity levels grew much faster. The initial productivity gap was substantial in most cases - in all but two countries labour productivity in 1950 was less than 60 per cent of the American level. There was strong beta and sigma convergence in western European productivity. That said, some countries, notably West Germany, did better than their initial productivity gap would have indicated and others, notably Ireland and the UK, did substantially worse. While the UK had less scope for catch-up in the Golden Age than almost all its European rivals, it was not merely caught up but overtaken by many of them and various econometric analyses confirm that there was an avoidable productivity growth failure in the UK of somewhere between 0.5 and 1 percentage point per year (Broadberry and Crafts, 2003).

During the Golden Age there was a substantial acceleration in technology transfer from the United States to Europe aided by increased codification of knowledge, the proliferation of multinational firms, investments in human capital and larger markets resulting from European integration (Nelson and Wright, 1992). This played an important part in rapid catch-up growth but in many countries this was also based in the 1950s and 1960s on recovery from war damage and the transfer of labour out of agriculture (Temin, 2002).

A striking feature of the European Golden Age was that investment as a share of GDP was higher than before or since. A very influential hypothesis to explain this was put forward by Eichengreen (1996), who stressed the crucial role of corporatist postwar settlements in many European countries. The key was a cooperative equilibrium between employers and workers' wage restraint in return for high investment under conditions of highly-centralised wage bargaining and institutional arrangements designed to enhance the credibility of such deals. In terms of figure 1, this promoted an outward shift of the Solow relationship. Strikingly, neither the UK nor Ireland achieved this corporatist result- in both economies collective bargaining was strong but decentralised.

Productivity growth outcomes in the Golden Age were influenced by social capability in ways which the American perspective on more recent European growth would recognise. This can be illustrated by comparisons between the UK and West Germany. At the time, the British tax system was notorious for its high marginal (though not average) rates and a comparative study of OECD countries found that it was the least conducive to growth (Tanzi, 1969). Two further substantial differences whose importance has emerged subsequently were in corporate governance and industrial relations.

In the UK, firms were more frequently characterised by the absence of a dominant external shareholder while competition was generally weak, as reflected in relatively high price-cost margins in manufacturing (Crafts and Mills, 2005), and this combination had seriously adverse effects on productivity growth (Nickell et al., 1997). The one major improvement in anti-trust policy during the Golden Age, the Restrictive Practices Act of 1956, which led to the break-up of many of the cartels which had proliferated in the previous 30 years, did generate a marked improvement in productivity performance (Symeonidis, 2004), which underlines just how ill-judged was the protectionism and lenient competition policy of the period.

With regard to industrial relations, where legislation was seriously defective, many UK but virtually no German firms were confronted by multiple unionism whose opportunism inhibited sunk-cost investments, a classic hold-up problem. Multiple unionism exacted a significant penalty in terms of foregone investment and productivity growth (Bean and Crafts, 1996; Denny and Nickell, 1992) and precluded the 'wage moderation for high investment' deals that characterised successful postwar settlements in other countries. Moreover, reform of industrial relations, along with many other desirable supply-side reforms such as abolition of punitive marginal tax rates, privatisation and a shift from industrial to competition policy, were all subject to the veto of organised labour while governments sought some kind of incomes policy in pursuit of full employment.

Slowdown after the Golden Age

After the early 1970s, European growth slowed down quite markedly. In fact, underlying productivity growth fell by more than the raw data of table 1 suggest. European responses to the difficult macroeconomic environment of the 1970s and 1980s entailed reductions of labour inputs promoted both by collective bargaining and by government policies which disproportionately reduced employment of low productivity (especially the young and the elderly) workers. Bourles and Cette (2005 and 2006) developed an econometric methodology to normalise for the productivity impact of different labour market structures and their 'structural productivity' estimates which are for a subset of European countries and indicate the productivity to be expected if employment and hours by age were at United States levels, are reported in table 3. The general result is to trim labour productivity growth for the period 1973 to 1995 quite appreciably, although by amounts that vary considerably for individual countries.

The end of the Golden Age had a number of root causes, including the exhaustion of transitory components of fast growth such as recovery from the shocks of the war and interwar depression, diminishing returns to investment as the postwar boom went on, and significant reduction in the scope for catch-up as the productivity gap with the United States narrowed--by 1973 the estimates in table 1 show that nine countries had labour productivity at least 70 per cent of the US level. The 'wage restraint/ high investment' cooperative equilibrium, which game-theoretic analysis suggests was in any case quite fragile, came under severe pressure as the rewards for patience seemed to evaporate in a world of much greater capital mobility, slower productivity growth and floating exchange rates (Cameron and Wallace, 2002).

At the same time, the corporatist model of economic growth was less appropriate in economies which now needed to become more innovative and less imitative in achieving productivity growth, as Eichengreen himself (2006) has recently stressed. Thus, aspects of institutional arrangements that sustained long-termist factor accumulation, such as capital markets geared to bank rather than equity finance and strong employers' associations that promoted vocational training while college education was relatively underdeveloped, were less well designed for sustaining growth in an era when it became more important to promote favourable shifts in the Schumpeter relationship of figure 1. Nevertheless, the European model continued to underpin strong investment in physical and human capital, such that by the 1990s leading European economies operated at a greater capital intensity and had many fewer workers with low qualifications than the United States (O'Mahony and de Boer, 2002).

However, the corporatist postwar settlements, through their strong commitment to welfare spending and social protection, entailed a legacy that informs the American perspective on recent European growth. This is set out in tables 4, 5 and 6. In Europe, spending on social transfers as a share of GDP grew rapidly from the 1960s to the 1980s (Lindert, 2004) and was financed to a considerable extent by 'distortionary' taxation. (2) The estimates in Kneller et al. (1999) suggest that the increases in taxation reported in table 4, which typically were much greater than in the United States, reduced the growth rate in the average European country by about 1 percentage point per year in 1995 compared with 1965.

Table 5 reports indices of product market regulation compiled at the OECD. These are designed to reflect the extent to which the regulatory environment is conducive to product market competition. The trend in the table to lower scores indicates that in every country there has been pro-competitive deregulation since the 1980s. However, this was later and less complete in European countries than in the United States. Table 6 reports an index of employment protection which is designed to reflect the cost implications of employment adjustment arising from regulations on severance pay, dismissal procedures etc. This shows that a common European policy response to the difficulties of the 1970s was to increase employment protection but since the 1980s policy changes have tended to be in the opposite direction. Employment protection in European countries was higher than in the United States throughout the past 40 years.

These indicators are broadly consistent with a diagnosis that the typical European country was exposed to more regulation, weaker competition and more distortionary taxation than the United States both before and after the mid-1990s. If these aspects of incentive structures had become less favourable after the Golden Age, in many countries much of the damage had already been done by the early 1980s. Similarly, it was at this point that the United States was most ahead of Europe in terms of product market deregulation. Yet, even on the productivity growth estimates corrected for differences in labour market structure this did not preclude catch-up in the pre-1995 period.

ICT and productivity growth since 1995

The estimates in table 1 show a revival in American productivity growth after 1995 at a time when productivity growth in most European countries slowed down. There has been, of course, a wide range of post-1995 productivity growth rates in Europe from Ireland at the top to Spain at the bottom. And it should also be recognised that efforts made in Europe in the recent past to reform their labour markets and reduce unemployment have made observed labour productivity growth lower than the underlying performance with unchanging labour market structure. Table 3 reflects this in most countries for which estimates of this type are available. In all, taking into account both tables 1 and 3, in 1995-2004 seven European countries (Finland, Ireland, Greece, Netherlands, Sweden, Switzerland and the UK) exhibited stronger productivity growth than in 1973-95 and six had faster productivity growth than the United States in 1995-2004.

The contribution that ICT made to labour productivity growth can be estimated using growth accounting techniques. In principle, this approach would identify contributions from ICT capital-deepening, TFP growth in ICT production and (unremunerated) TFP spillovers from the use of ICT capital. In practice, this last aspect has proved somewhat elusive. Growth accounting estimates are reported in table 7. They confirm that ICT contributed more to productivity growth in the United States than in the EU both before and after 1995 and that it played a significant role in the acceleration of American productivity growth after the mid-1990s.

Within Europe, in the late 1990s three countries, namely, Finland, Ireland and Sweden, had a greater contribution to productivity growth from ICT production than the United States. In each case, gross output in ICT production was much larger relative to GDP-22.5 per cent in Ireland compared with 3 per cent in the United States (Timmer et al., 2003)--and each of these countries had higher labour productivity growth than the United States in 1995-2004. That said, it is also clear that, especially in the big continental European economies, there was a substantial weakening of non-ICT related sources of productivity growth in the EU after 1995.

Another take on the role of ICT in recent growth can be obtained by considering the part that particular sectors have played. Here the big story is the contribution made by market services that are intensive in the use of ICT, in general, and the distributive trades, which account for close to 20 per cent of employment in the typical OECD country, in particular. This may reflect the impact of TFP spillovers and it certainly involves so-called 'soft savings' achieved through the extra information that ICT provides that permits better organisation of inventories, logistics etc. Indeed, estimates for the period 1995-2001 suggest that the differential contribution made by these sectors in the EU compared with the United States exceeded the total difference in productivity growth between the two (Inklaar et al., 2003).

Table 8 reports estimates of labour productivity growth in distribution. The post-1995 acceleration in the United States is notable and has not been matched across the EU, although some countries, including the UK, have seen an increase in productivity growth. The speeding up of productivity growth in the United States has mostly been based on faster TFP growth in the sector (McGuckin et al., 2005). Table 8 also reports estimates of the contribution to overall productivity growth made by ICT-using market services in 1996 to 2001. The big advantage that accrued to the United States obviously looms large relative to the gap in productivity growth with European countries.

Competition, regulation and productivity performance

A considerable amount of empirical analysis has been carried out recently that permits an assessment of the claim that disappointing European productivity growth since the mid-1990s has been a result of insufficient competition and excessive regulation. Tables 5 and 6 reported that in general European countries have had higher levels of employment protection and product market regulation than has been the case in the United States, although OECD measures also show a trend towards deregulation in Europe, especially since 1990.

Estimates of price-cost margins indicate that they have been higher in EU countries than in the United States, especially in market services, as table 9 reports. This relative weakness of competition in Europe is associated with a greater prevalence of poor management practices which translates into lower levels of TFP (Bloom and van Reenen, 2006). Estimates using a calibrated general equilibrium model suggest that, if price-cost margins in the Euro Area were reduced to American levels, real GDP would rise by 8.6 per cent and hours worked by 4.5 per cent, so that real GDP per hour worked would be about 4 per cent higher (Bayoumi et al., 2004). In other words, the European catch-up of the United States would have been more complete had greater competition been encouraged.

The links between product-market regulation and productivity performance can be revealed by a strategy of using the OECD or similar measures to examine the impact of regulation on entry rates and price-cost mark ups and then of these intermediate variables on innovation, investment and productivity. There is quite convincing evidence that tighter regulation leads both to less entry (Brandt, 2004; Cincera and Galgau, 2005) and higher mark-ups (Griffith and Harrison, 2004) and that the deregulation introduced by the European Single Market reduced mark-ups (Griffith et al., 2006). Cincera and Galgau (2005) found that increases in entry and exit rates in an industry had a positive impact on labour productivity growth and Brandt (2004) found a strong correlation between entry rates and subsequent productivity growth rates in service sectors. Griffith and Harrison (2004) found that lower mark-ups raise investment while Griffith et al. (2006) found that lower mark-ups promote innovation. Taken together, these results suggest that product market regulation tends to inhibit productivity performance through raising barriers to entry and inhibiting competition.

This is consistent with, and helps to explain, the results that have been found using growth regressions looking at catch-up growth at the industrial-sector level by Nicoletti and Scarpetta (2003) which inform the results on potential productivity gains from further deregulation reported in table 10. They suggest that slowness to deregulate product markets may explain some, though by no means all, of the shortfall in recent European productivity growth.

A key finding of the literature on regulation and productivity performance is that the effects are greatest in the face of new technological opportunities and that the adverse impact on ICT-using services is especially strong. Restrictive regulations have deterred investment in ICT capital directly (Conway et al., 2006) and the indirect effect of regulation through raising costs has been relatively pronounced in ICT-using sectors (Conway and Nicoletti, 2006). Overall, Nicoletti and Scarpetta (2005, table 10) found a strong correlation between product-market regulation and the contribution of ICT-using services to aggregate productivity growth.

Similar findings exist for employment protection. Caballero et al. (2004) found that, in countries where regulation is enforced, high levels of employment protection considerably slow down the process of creative destruction and the labour force adjustment that it entails. Their results imply that the difference in job security between France and the United States could have been responsible for a difference of about 0.5 percentage points per year in labour productivity growth in the recent past. Here too, however, the cross-country regression evidence is that employment-protection regulation has had a relatively strong negative effect on investment in ICT equipment (Gust and Marquez, 2004). The rationale is that higher firing costs are an obstacle to the upgrading of the labour force and reorganisation of working practices that are central to realising the productivity potential of ICT.

These findings help to explain the apparently paradoxical relationship between trends in regulation and disappointing recent European productivity performance. It is not that regulation has become more stringent, but rather that existing regulation has become more costly in the context of a new technological era based on ICT. Abramovitz famously argued that the European catch-up of the United States in the postwar period depended on 'social capability' to assimilate new technologies effectively but, as he remarked, 'social capability [involves] adjusting steadily to the requirements of technological opportunities' (1986, p. 402).

A UK productivity paradox?

The UK can be seen as a test of the central hypotheses of the American perspective on European growth failures. It has a policy mix closer to what these critics advocate than other European countries. This is decisively the case according to the regulation indices in tables 5 and 6, where the UK is seen to have had much lower employment protection than elsewhere in Europe, while the UK has not only had less product market regulation than its European peer group but also, in the recent past, than the United States. However, table 1 shows a substantial labour productivity gap with the European leaders and the rate of growth of UK labour productivity was only 2.20 per cent per year during 1995-2004--hardly tigerish!

This might be seen as surprising and contrary to the American perspective, especially as the Government has prioritised its productivity agenda since 1997 on the basis of 'New Labour' policies. There has been no return to 'Old Labour' policies; on the contrary, Thatcherite supply-side reforms that addressed many of the policy errors of the Golden Age have basically remained in place and have been notably strengthened with regard to competition policy in terms of the Competition Act and the Enterprise Act which came into effect in 2000 and 2003, respectively.

There are three possibilities which are not mutually exclusive and all of which probably play a part, namely, that the UK productivity performance has been better than is sometimes supposed, that UK regulation is less pro-competitive than the OECD indices suggest, and that the American perspective overstates the case for competition and deregulation.

First, it should be recognised that it is Irish productivity performance that is extraordinary and not matching it does not represent a failure given that ICT production is a relatively small share of gross output in the UK (2.6 vs 22.5 per cent). The labour productivity growth is actually quite similar to what the 'naive' catch-up models of table 2 estimated for 1950-73 or 1973-95 predict (2.4 and 2.1 per cent, respectively) for a country with an initial productivity level in 1995 at 86.1 per cent of the American level. It is also important to take into account the estimates relating to 'structural' productivity in table 3. These show relative UK labour productivity levels in 2004 in a much more favourable light compared with other European countries, notably Germany, Italy, Netherlands and even France, although in that case there is still a big gap after the adjustment proposed by Bourles and Cette (2006) has been made.

Second, there clearly are some doubts about the accuracy of the OECD measure of product market regulation as applied to the UK. In particular, the implications of the UK's extremely strict planning restrictions are not taken into account (Baily and Kirkegaard, 2004). Moreover, while the UK scores well on the overall product market regulation measure, the OECD sectoral measure for the retail trade, which has played such a large part in American productivity growth, showed that in 1996 the UK had a relatively high score of 3.4 which fell to 2.0 in 2003 when five European countries had lower scores.

Third, as tables 7 and 8 reported, the UK has experienced a relatively strong contribution to productivity growth from the regulation-sensitive ICT-using services sector and ICT capital deepening has been above the EU average. The estimates of Nicoletti and Scarpetta (2005) on the impact of different regulatory policy stances reported in table 10 suggest that in the 1990s the UK could have expected a TFP growth advantage from this source of about 0.5 percentage points per year and might have expected to remove most of the earlier TFP gap with those countries by 2000. That is precisely what table 11 suggests did happen. At the same time, there is nothing in the literature to suggest that the UK's greater commitment to deregulation would of itself redeem the long-standing shortfalls in physical and human capital formation resulting from the advantages of postwar corporatist institutions in other European countries (Broadberry and O'Mahony, 2004) which account for the remaining labour productivity gap.

Lessons from Ireland ?

Ireland seems at first sight to epitomise the claims of American critics of European growth performance. Since 1973, Ireland has moved from virtually bottom to nearly top of the league in terms of the level of real GDP per hour worked, as table 1 reports. Ireland is a lightly regulated country and has relatively low distortionary taxes. Famously, Ireland has attracted a huge amount of FDI by maintaining much lower corporate taxes than other European countries and has become the prime location inside the EU for American multinationals who export the vast majority of their output (Gropp and Kostial, 2000).

Irish growth during the 'Celtic Tiger' phase has been sustained both by strong employment growth and strong productivity growth. Fast growth represented a belated catch-up as policy errors from the Golden Age such as protectionism and neglect of human capital formation were corrected and macroeconomic stabilisation was achieved but was augmented by a very elastic supply of labour in a regional economy starting with high unemployment where labour market reform dramatically reduced the NAIRU (Barry, 2002; Nickell, 2003).

Table 12 reports growth rates for the period since 1987 when the Celtic Tiger phase began. The experience has been remarkable but less stellar than is sometimes thought for reasons that are implied in the preceding description. Transfer pricing by multinationals has resulted in a big excess of GDP over GNP (about 16 per cent in 2004) such that growth and productivity performance are better measured in terms of the latter (Cassidy, 2004). Because of the big increase in employment, labour productivity has grown less rapidly than output per head. Thus, table 12 reports that while real GDP per head grew at 5.8 per cent per year between 1987 and 2004, the growth rate of real GNP per hour worked was 3.9 per cent. Exports are roughly equal to GNP and the terms of trade have fallen by over 10 per cent since 1987 as prices fell for ICT goods in particular. The implication is that real income has been growing less fast than real product and consumption per person has grown at 4 per cent per year.

The key aspect of recent Irish productivity performance is the huge contribution of ICT production, as reported in table 7. If this is taken out of labour productivity growth, then the other components have not been exceptional. Clearly, the large ICT production sector in Ireland is a result of supply-side policies that have attracted FDI and have developed a successful cluster. These include interventions that have improved the infrastructure and labour supply available to the multinationals (Crafts, 2005). However, the success of the low tax, light regulation approach would have been much less spectacular if the composition of manufacturing output had happened to be less favourable.

In sum, this suggests that Ireland is something of a special case rather than an experience that can easily be replicated across the European economies as a whole. Specialisation in ICT exports is not possible for everyone and low corporate taxes are most effective in attracting a large share of FDI if other countries do not follow suit. Scope for catch-up is lower in most countries than it was in 1980s Ireland. Ireland's example is perfectly consistent with the American perspective but should not be oversold.

Conclusions

American critics of European growth performance have stressed that European economies are handicapped by excessive taxation and regulation and by weak competition. They argue that this has been reflected in superior productivity growth in the United States since the mid-1990s. This paper has evaluated these claims by addressing three specific questions posed in the introduction with the following results.

European countries generally have not exploited the opportunities of the ICT age as well as the United States. This is particularly apparent in their relative lack of success in obtaining a large contribution to productivity growth from services that are intensive in the use of ICT and this may well be related to problems of competition and regulation. It is important to recognise that this has occurred in an era when European countries have been reducing product market regulation and it implies that existing regulations, notably in retailing, have become more costly in terms of foregone productivity growth.

As a lightly-regulated economy, the UK has been better positioned than the other big European economies to prosper in the ICT era. This has been reflected in TFP growth and relatively strong contributions to productivity growth from both ICT-using services and ICT capital deepening. Standard comparisons of productivity levels between the UK and its European peer group exaggerate the gaps since the greater flexibility of the UK labour market allows more low productivity workers to be employed. Nevertheless, productivity performance in the UK still shows the impact of institutions that have been less successful than those elsewhere in Europe in achieving high rates of investment in human and physical capital and this reveals the limitations of the American perspective on Europe.

Ireland has also gained from light regulation and, especially, low corporate taxation which has promoted very rapid growth based on FDI. A striking feature of the Irish economy is the very high share of gross output in ICT production made by the multinationals. The phenomenal technological progress that this sector has enjoyed in the recent past has been primarily responsible for a very high rate of TFP growth for Ireland. But, precisely for this reason, Ireland is more a special case than a model that can easily be replicated.

Overall, recent research has clearly established that weak competition typically has a significant cost in terms of slower productivity growth. Nevertheless, strong productivity performance also requires institutions and policies that sustain capital accumulation and this is an aspect in which Europe has been relatively successful. As figure 1 depicts, endogenous growth theory suggests that investment and innovation both matter for long-run growth outcomes.

REFERENCES

Abramovitz, M. (1986), 'Catching up, forging ahead, and falling behind', Journal of Economic History, 36, pp. 395-406.

Aghion, P., Dewatrimont, M. and Rey, P. (1997), 'Corporate governance, competition policy and industrial policy', European Economic Review, 41, pp. 797-805.

Aghion, P., Harris, C., Howitt, P. and Vickers, J. (2001), 'Competition, innovation and growth with step-by-step innovation', Review of Economic Studies, 68, pp. 467-92.

Alesina, A., Glaeser, E. and Sacerdote, B. (2005), 'Work and leisure in the US and Europe: why so different?', CEPR Discussion Paper No. 5140.

Baily, M. and Kirkegaard (2004), Transforming the European Economy, Institute for International Economics.

Barry, F. (2002), 'The Celtic Tiger era: delayed convergence or regional boom', ESRI Quarterly Economic Commentary, Summer, pp. 84-91.

Bayoumi, T., Laxton, D. and Pesenti, P. (2004), 'Benefits and spillovers of greater competition in Europe: a macroeconomic assessment', NBER Working Paper No. 10416.

Bean, C. and Crafts, N. (1996), 'British economic growth since 1945: relative economic decline ... and renaissance?', in Crafts, N. and Toniolo, G. (eds), Economic Growth in Europe since 1945, Cambridge University Press.

Bloom, N. and van Reenen, J. (2006), 'Measuring and explaining management practices across firms and countries', NBER Working Paper No. 12216.

Bourles, R. and Cette, G. (2005), 'A comparison of structural productivity levels in the major industrialised countries', OECD Economic Studies, 41(2), pp. 75-108.

--(2006), 'Les evolutions de la productivite 'structurelle' du travail dans les principaux pays industrielles', Bulletin de la Banque de France, 150, pp. 23-30.

Brandt, N. (2004), 'Business dynamics, regulation and performance', OECD Science, Technology and Industry Working Paper No. 2004/3.

Broadberry, S. and Crafts, N. (2003), 'UK productivity performance from 1950 to 1979: a restatement of the Broadberry-Crafts view', Economic History Review, 56, pp. 718-35.

Broadberry, S. and O'Mahony, M. (2004), 'Britain's productivity gap with the United States and Europe: a historical perspective', National Institute Economic Review 189, pp. 72-85.

--(2007), 'Britain's twentieth-century productivity performance in international perspective', in Crafts, N., Gazeley, I. and Newell, A. (eds.), Work and Pay in Twentieth-Century Britain, Oxford University Press.

Caballero, R., Cowan, K., Engel, E. and Micco, A. (2004), 'Effective labor regulation and microeconomic flexibility', NBER Working Paper No. 10744.

Cameron, G. and Wallace, C. (2002), 'Macroeconomic performance in the Bretton Woods era and after', Oxford Review of Economic Policy, 18, pp. 479-94.

Carlin, W. and Soskice, D. (2006), Macroeconomics: Imperfections, Institutions and Policies, Oxford University Press.

Cassidy, M. (2004), 'Productivity in Ireland: trends and lessons', Central Bank of Ireland Quarterly Bulletin, Spring, pp. 82-105.

Central Statistical Office (2006), Statistical Yearbook of Ireland.

Cincera, M. and Galgau, G. (2005), 'Impact of market entry and exit on EU productivity and growth performance', European Commission DG for Economic and Financial Affairs Economic Papers No. 222.

Conway, P., de Rosa, D., Nicoletti, G. and Steiner, F. (2006), 'Regulation, competition and productivity convergence', OECD Economics Department Working Paper No. 509.

Conway, P., Janod, V. and Nicoletti, G. (2005), 'Product market regulation in OECD countries: 1998-2005', OECD Economics Department Working Paper No. 419.

Conway, P. and Nicoletti, G. (2006), 'Product market regulation in the non-manufacturing sectors of OECD countries: measurement and highlights', OECD Economics Department Working Paper No. 530.

Cotis, J.-P. (2006), 'Economic growth and productivity', paper presented to Annual Conference of the Government Economic Service.

Crafts, N. (2005), 'Interpreting Ireland's economic growth', background paper for UNIDO, Industrial Development Report.

Crafts, N. and Mills, T.C. (2005), 'TFP growth in British and German manufacturing, 1950-1996', Economic Journal, 115, pp. 649-70.

Davis, S. and Henrekson, M. (2004), 'Tax effects on work activity, industry mix and shadoweconomy size; evidence from rich-country comparisons', NBER Working Paper No. 10509.

Denny, K. and Nickell, S. (1992), 'Unions and investment in British industry', Economic Journal, 102, pp. 874-87.

Dew-Becker, I. and Gordon, R.J. (2006), 'The slowdown in European productivity growth: a tale of tigers, tortoises and textbook labor economics', paper presented to NBER Summer Institute Macroeconomics and Productivity Workshop.

Eichengreen, B. (1996), 'Institutions and economic growth: Europe after World War II', in Crafts, N. and Toniolo, G. (eds), Economic Growth in Europe since 1945, Cambridge University Press.

--(2006), The European Economy since 1945: Co-ordinated Capitalism and Beyond, Princeton University Press.

Faggio, G. and Nickell, S. (2006), 'Patterns of work across the OECD', London School of Economics CEP Discussion Paper No. 730.

Griffith, R. and Harrison, R. (2004), 'The link between product market regulation and macroeconomic performance', European Commission DG for Economic and Financial Affairs Economic Papers No. 209.

Griffith, R., Harrison, R. and Simpson, H. (2006), 'Product market reform and innovation in the EU', IFS Working Paper No 06/ 17.

Groningen Growth and Development Centre (2006), Total Economy Database.

Gropp, R. and Kostial, K. (2000), 'The disappearing tax base: is FDI eroding corporate income taxes?', IMF Working Paper No. 00/173.

Gust, C. and Marquez, J. (2004), 'International comparisons of productivity growth: the role of information technology and regulatory practices', Labour Economics, 11, pp. 33-58.

Inklaar, R., O'Mahony, M. and Timmer, M. (2003), 'ICT and Europe's productivity performance: industry level comparisons with the United States', Groningen Growth and Development Centre Research Memorandum GD-68.

Kneller, R., Bleaney, M. and Gemmell, N. (1999), 'Fiscal policy and growth: evidence from OECD countries', Journal of Public Economics, 74, pp. 171-90.

McGuckin, R., Spiegelman, M. and van Ark, B. (2005), The Retail Revolution: Can Europe Match US Productivity Performance?, The Conference Board.

Nelson, R. and Wright, G. (1992), 'The rise and fall of American technological leadership', Journal of Economic Literature, 30, pp. 1930-64.

Nicoletti, G. and Scarpetta, S. (2003), 'Regulation, productivity and growth', Economic Policy, 36, pp. 9-72.

--(2005), 'Regulation and economic performance: product market reforms and productivity in the OECD', OECD Economics Department Working Paper No. 460.

Nickell, S. (2003), 'A picture of European unemployment: success and failure', London School of Economics CEP Discussion Paper No. 577

--(2005), 'What has happened to unemployment in the OECD since the 1980s?', unpublished presentation to Work and Pensions Economics Group, HM Treasury.

Nickell, S., Nicolitsas, D., and Dryden, N. (1997), 'What makes firms perform well?', European Economic Review, 41, pp. 783-96.

OECD (1981), Long-Term Trends in Tax Revenues in OECD Member Countries.

--(2006), Revenue Statistics, 1965-2005.

Oliveira-Martins, J., Scarpetta, S. and Pilat, D. (1996), 'Mark-up pricing, market structure and the business cycle', OECD Economic Studies, 27(2), pp. 71-105.

O'Mahony, M. and de Boer, W. (2002), Britain's Relative Productivity performance: updates to 1999, NIESR.

Prescott, E. (2004), 'Why do Americans work so much more than Europeans?', Federal Reserve Bank of Minneapolis Quarterly Review, 28(1), pp. 2-13.

Symeonidis, G. (2004), 'The effect of competition on wages and productivity: evidence from the UK', mimeo, University of Essex.

Tanzi, V. (1969), The Individual Income Tax and Economic Growth, Johns Hopkins University Press.

Temin, P. (2002), 'The golden age of European growth reconsidered', European Review of Economic History, 6, pp. 3-22.

Timmer, M. and Ypma, G. (2006), 'Productivity levels in distributive trades: a new ICOP dataset for OECD countries', Groningen Growth and Development Centre Research Memorandum GD-83.

Timmer, M., Ypma, G. and van Ark, B. (2003), 'IT in the European Union: driving productivity divergence?', Groningen Growth and Development Centre Research Memorandum GD-67.

NOTES

(1) The analysis here is based entirely on real GDP per hour worked. It is well-known that catch-up in terms of real GDP per person ceased in the early 1970s since when employment and annual hours worked per person employed have fallen in Europe relative to the United States. The reasons for these developments are controversial but beyond the scope of this paper; for competing recent analyses of why Americans work more than Europeans see Alesina et al. (2005), Davis and Henrekson (2004), Faggio and Nickell (2006) and Prescott (2004).

(2) 'Distortionary' is the term used by Kneller et al. (1999). This is basically direct taxation which in many new growth models has adverse effects on growth, whereas indirect taxes do not.

Nicholas Crafts *

* University of Warwick. e-mail: N.Crafts@warwick.ac.uk. I have benefited from helpful discussions with Steve Broadberry and am grateful to Steve Nickell for providing me with unpublished data. The usual disclaimer applies. Table 1. Levels and rates of growth of real GDP/hour worked a) Levels (1990GK$) 1950 1973 Switzerland 8.68 Switzerland 18.88 UK 7.35 Netherlands 18.08 Sweden 6.82 France 17.18 Netherlands 6.77 Sweden 17.15 Denmark 6.72 Belgium 16.95 Norway 6.09 Italy 16.16 Belgium 5.83 W. Germany 16.05 France 5.46 Denmark 15.88 Italy 5.10 Norway 15.70 W. Germany 4.36 UK 14.05 Finland 4.03 Austria 13.39 Austria 3.57 Finland 11.61 Ireland 3.45 Greece 10.15 Spain 2.52 Spain 9.47 Greece 2.43 Ireland 9.45 Portugal 2.15 Portugal 9.21 USA 11.89 USA 21.29 1995 2004 Norway 31.67 Norway 39.20 France 31.24 France 37.04 W. Germany 30.08 Belgium 33.76 Belgium 29.21 Netherlands 31.72 Italy 27.61 Austria 31.20 Netherlands 27.57 Denmark 30.02 Denmark 27.01 UK 29.26 Austria 24.76 Sweden 29.23 Germany 24.50 Italy 28.81 UK 24.06 Germany 28.50 Sweden 23.47 Ireland 28.22 Switzerland 23.33 Finland 27.10 Spain 22.24 Switzerland 26.20 Finland 22.05 Spain 21.85 Ireland 19.60 Greece 18.31 Greece 14.70 Portugal 17.05 Portugal 14.31 USA 27.95 USA 34.54 b) Rates of growth (% per year) 1950-73 1973-95 1995-2004 Switzerland 3.43 Switzerland 0.97 Norway 2.40 UK 2.86 Netherlands 1.94 France 1.92 Sweden 4.09 France 2.76 Belgium 1.63 Netherlands 4.36 Sweden 1.44 Italy 0.48 Denmark 3.80 Belgium 2.51 Netherlands 1.58 Norway 4.20 Italy 2.47 Denmark 1.18 Belgium 4.74 W. Germany 2.90 Austria 2.60 France 5.11 Denmark 2.44 Germany 1.70 Italy 5.14 Norway 3.24 UK 2.20 W. Germany 5.85 UK 2.48 Sweden 2.47 Finland 4.70 Austria 2.83 Switzerland 1.30 Austria 5.93 Finland 2.95 Spain -0.20 Ireland 4.46 Greece 1.91 Finland 2.32 Spain 5.93 Spain 3.94 Ireland 4.13 Greece 6.42 Ireland 3.37 Greece 2.47 Portugal 6.54 Portugal 2.02 Portugal 1.96 USA 2.57 USA 1.25 USA 2.38 Source: Groningen Growth and Development Centre (2006). Note: Figures for Ireland from 1973 onwards are for GNP/HW, see text; GDP/HW in Ireland was 9.43 in 1973, 22.17 in 1995 and 33.49 in 2004 and the rate of growth of GDP/HW was 3.94 per cent in 1973-95 and 4.68 per cent in 1995-2004. 1990 GK$ total GDP, in millions of 1990 US$ (converted at Geary Khamis Purchasing Power Parities). Table 2. Unconditional convergence regressions 1950-73 1973-95 1995-2004 Constant 7.300 3.932 3.035 (10.017) (4.942) (2.461) Initial GDP/HW -0.057 -0.021 -0.013 (6.817) (-1.847) (0.952) [R.sup.2] 0.752 0.139 0.006 Note: Regressions based on data for table 1; t-statistics in parentheses. Table 3. Observed and structural real GDP/hour worked a) Levels (1990GK$ and (%US)) 1973 1995 Observed Structural Observed Structural Finland 11.61 12.39 22.05 20.07 (54.5) (58.2) (78.9) (71.8) France 17.18 17.44 31.24 28.21 (80.7) (81.9) (111.8) (100.9) Germany 16.05 n/a 24.50 21.31 (75.4) (87.7) (76.2) Ireland 9.45 10.41 19.60 17.36 (44.4) (48.9) (70.1) (62.1) Italy 16.16 14.75 27.61 23.42 (75.9) (69.3) (98.8) (83.8) Netherlands 18.08 17.76 27.57 23.35 (84.9) (83.4) (98.6) (83.5) Norway 15.70 14.95 31.67 28.90 (73.7) (70.2) (113.8) (103.4) Spain 9.47 9.96 22.24 18.41 (44.5) (46.8) (79.6) (65.9) Sweden 17.15 16.95 23.47 21.88 (80.6) (79.6) (84.0) (78.3) UK 14.05 14.05 24.06 22.77 (66.0) (66.0) (86.1) (81.5) 2004 Observed Structural Finland 27.10 25.79 (78.5) (74.7) France 27.04 33.28 (107.2) (96.4) Germany 28.50 26.60 (82.2) (71.2) Ireland 28.22 26.27 (81.7) (76.1) Italy 28.81 24.84 (83.4) (71.9) Netherlands 31.92 28.09 (91.8) (81.3) Norway 39.20 36.16 (113.5) (104.7) Spain 21.85 19.85 (63.3) (57.5) Sweden 29.23 27.64 (84.6) (80.0) UK 29.26 28.22 (84.7) (81.7) b) Rates of Growth (% per year) 1973-95 1995-2004 Observed Structural Observed Structural Finland 2.95 2.22 2.32 2.83 France 2.76 2.21 1.92 1.86 Germany 2.90 n/a 1.70 1.62 Ireland 3.37 2.35 4.13 4.70 Italy 2.47 2.12 0.48 0.67 Netherlands 1.94 1.26 1.58 2.08 Norway 3.24 3.04 2.40 2.52 Spain 3.94 2.83 -0.20 0.86 Sweden 1.44 1.17 2.47 2.63 UK 2.48 2.22 2.20 2.41 Source: Derived using Bourles and Cette (2006). Note: 1990GK$ as table 1. Table 4. Distortionary tax revenues (% GDP) 1955 1965 1980 1995 2004 Austria 19.5 21.2 26.7 29.6 30.6 Belgium 15.3 19.5 30.0 32.4 33.7 Denmark 15.0 17.8 27.0 33.1 32.8 Finland 15.2 17.3 23.2 31.8 30.2 France n/a 21.3 30.0 31.2 32.3 Germany 20.2 21.2 27.3 26.8 24.6 Greece n/a 10.0 13.9 18.6 22.0 Ireland 11.8 11.8 17.4 17.3 18.7 Italy 17.1 15.4 21.8 29.2 30.3 Netherlands 17.9 22.4 31.3 31.3 25.5 Norway 15.8 17.4 27.5 25.2 30.9 Portugal 9.3 8.8 12.6 19.2 21.2 Spain n/a 8.7 17.9 22.9 25.0 Sweden 18.2 24.1 35.7 34.8 37.4 Switzerland 12.9 11.5 19.5 21.7 22.3 UK 19.6 20.3 24.9 22.7 24.5 USA 19.1 19.1 21.7 22.9 20.8 Source: OECD (1981; 2006). Table 5. Product market regulation (0-10) 1998 1998 1975 1980 1985 1990 (1) (2) 2003 Austria 8.67 8.50 8.17 7.50 6.50 3.00 2.33 Belgium 9.17 9.17 9.17 8.83 5.67 3.50 2.33 Denmark 9.17 9.17 9.17 7.83 5.00 2.50 1.83 Finland 9.17 9.00 8.50 7.67 4.50 3.50 2.17 France 10.00 10.00 10.00 8.67 7.17 4.17 2.83 Germany 8.67 8.67 8.50 7.67 4.67 3.17 2.33 Greece 9.50 9.50 9.50 9.50 8.83 4.67 3.00 Ireland 9.50 9.50 9.50 8.33 7.33 2.50 1.83 Italy 9.67 9.67 9.67 9.67 7.83 4.67 3.17 Netherlands 9.33 9.33 9.33 9.33 4.83 3.00 2.33 Norway 9.17 9.17 8.33 7.50 5.83 3.00 2.50 Portugal 9.83 9.83 9.83 8.83 7.33 3.50 2.67 Spain 8.50 8.33 8.33 7.83 5.83 3.83 2.67 Sweden 7.50 7.50 7.50 7.33 4.00 3.00 2.00 Switzerland 6.83 7.00 7.00 7.00 6.17 3.67 2.83 UK 8.00 8.00 7.17 6.00 2.33 1.83 1.50 USA 6.17 4.50 4.00 3.67 2.67 2.17 1.67 Sources: 1978 to 1998 (I) REGREF from Conway and Nicoletti (2006); 1998 (2) and 2003 PMR from Conway et al. (2005). Table 6. Employment protection (0-10) 1960- 1965- 1973- 1980- 1988- 4 72 9 7 95 1998 2003 Austria 3.25 3.25 4.20 6.35 6.50 5.50 4.85 Belgium 3.60 6.30 7.75 7.75 6.75 5.00 5.00 Denmark 4.50 4.90 5.50 5.50 4.50 3.50 3.50 Finland 6.00 6.00 6.00 6.00 5.65 5.00 5.00 France 1.85 4.40 6.05 6.50 7.05 7.00 7.00 Germany 2.25 5.25 8.25 8.25 7.60 6.50 5.60 Greece 8.00 7.75 7.00 Ireland 0.10 0.95 2.25 2.50 2.60 2.50 2.80 Italy 9.60 9.95 10.00 10.00 9.45 7.50 4.85 Netherlands 6.95 6.95 6.95 6.95 6.40 5.50 5.50 Norway 7.75 7.75 7.75 7.75 7.30 6.50 6.50 Portugal 0.00 2.15 7.95 9.70 9.65 8.50 8.00 Spain 10.00 10.00 9.95 9.55 8.70 7.00 7.50 Sweden 0.00 1.05 7.30 9.00 7.65 5.50 5.50 Switzerland 2.75 2.75 2.75 2.75 2.75 2.75 2.75 UK 0.80 1.05 1.65 1.75 1.75 1.75 1.75 USA 0.50 0.50 0.50 0.50 0.50 0.50 0.50 Source: Nickell (2005). Table 7. Contributions to labour productivity growth (percentage points per year) a) ICT ICT capital TFP in ICT deepening production 1980-95 1995-2004 1990-95 1995-2001 Austria 0.40 0.47 0.11 0.20 Belgium 0.69 0.70 0.07 0.12 Denmark 0.64 0.74 0.06 0.10 Finland 0.47 0.56 0.19 0.69 France 0.29 0.32 0.20 0.31 Germany 0.47 0.40 0.16 0.24 Greece 0.10 0.38 0.00 0.01 Ireland 0.17 0.55 1.36 3.62 Italy 0.35 0.41 0.15 0.21 Netherlands 0.44 0.39 0.06 0.11 Portugal 0.24 0.47 0.24 0.31 Spain 0.32 0.27 0.10 0.15 Sweden 0.49 0.75 0.18 0.57 UK 0.45 0.62 0.24 0.38 EU 0.42 0.46 0.15 0.27 USA 0.66 0.79 0.24 0.44 b) Non-ICT Non-ICT capital Non-ICT deepening TFP 1980-95 1995-2004 1990-95 1995-2001 Austria 0.69 0.79 0.39 1.12 Belgium 0.46 0.09 0.89 1.02 Denmark 0.59 0.51 1.42 0.19 Finland 0.75 -0.18 1.00 1.98 France 1.18 0.75 -0.17 0.55 Germany 0.62 0.39 1.77 0.63 Greece 0.32 0.67 -0.22 1.66 Ireland 0.66 1.31 1.57 -0.01 Italy 0.73 0.50 0.89 -0.07 Netherlands 0.35 -0.04 0.40 0.26 Portugal 0.55 0.75 1.10 0.19 Spain 0.85 0.14 0.77 -0.77 Sweden 0.51 0.25 0.75 0.11 UK 0.70 0.27 1.21 0.09 EU 0.76 0.36 0.99 0.19 USA 0.16 0.34 0.37 0.36 Sources: Dew-Becker and Gordon (2006) and Timmer et al. (2003). Table 8. Contribution of ICT-using services to labour productivity growth, 1996-2001 (percentage points per year) a) Labour productivity growth in distributive trades (% per year) 1980-95 1995-2002 Austria 1.3 2.5 Belgium 0.3 0.7 Denmark 3.1 2.3 Finland 2.5 1.6 France 3.5 0.7 Germany 1.2 1.4 Greece -1.5 3.2 Ireland 0.6 1.5 Italy 1.3 0.5 Netherlands 1.4 1.8 Norway 2.9 5.9 Portugal 1.0 0.6 Spain 0.0 0.3 Sweden 2.8 2.5 UK 2.8 3.7 EU 1.7 1.5 USA 2.7 5.7 b) Contribution of ICT-using services to labour productivity growth, 1996-2001 (percentage points per year) Austria 0.5 Belgium 0.4 Denmark 0.4 Finland 0.3 France -0.1 Germany 0.2 Ireland 0.7 Italy 0.1 Netherlands 0.3 Norway 0.4 Spain 0.1 Sweden 0.6 UK 0.8 USA 1.3 Sources: Nicoletti and Scarpetta (2005) and Timmer and Ypma (2006). Table 9. Price-cost margins, 1980-92 Manufacturing Market services Belgium 1.17 1.80 Denmark 1.15 1.64 Finland 1.24 1.24 France 1.16 1.48 Germany 1.21 1.34 Italy 1.18 1.96 Netherlands 1.21 1.45 Norway 1.18 n/a Sweden 1.16 1.16 UK 1.15 1.37 USA 1.15 1.25 Source: Oliveira-Martins et al. (1996). Table 10. Change in TFP growth over 10 years from adopting regulatory best practice (percentage points) Austria 0.42 Belgium 0.60 Denmark 0.37 Finland 0.59 France 0.62 Germany 0.70 Greece 1.12 Italy 0.70 Netherlands 0.44 Norway 0.38 Portugal 0.65 Spain 0.41 Sweden 0.51 UK 0.11 Source: Nicoletti and Scarpetta (2005). Table 11. A decomposition of the UK's labour productivity gap (percentage points) France/UK Germany/UK 1979 2000 1979 2000 Labour productivity gap 31 21 30 17 Labour quality 6 4 5 4 Physical capital 17 17 9 12 TFP 8 0 16 1 Source: Broadberry and O'Mahony (2007). Note: In 1979 Germany is West Germany only. Table 12. Irish economic growth, 1987-2004 (% per year) Real GDP/head 5.8 Real GDP/head 5.3 Labour input 2.1 Real GDP/hour worked 4.4 Real GNP/hour worked 3.9 Real personal consumption/head 4.0 Sources: Derived from CSO (2006) and Groningen Growth and Development Centre (2006).