Unemployment, Structural Change and Globalization

M. Pianta and M. Vivarelli

1. The structural composition effect. Sectors have different growth rates of production and demand, and different employment intensities. Institutional arrangements, regulations and policies affect in different ways economic activities and their employment prospects. As a result, the specific sectoral composition of national economies may lead a variety of possible employment performances.

2. The capital intensity effect. Sectors have different investment requirements and employment growth tends to be lower where capital intensity is higher. These sectors also tend to be characterized by a prevalence of labour-saving investments and innovations, leading to worse employment outcomes.

3. The technology effect. Sectors are characterized by markedly different patterns of technological change. In some manufacturing and service industries innovation is largely produced endogenously, through R&D, design, software and engineering activities. In some traditional industries, as well as in agriculture and most services, innovations are generally supplied by other sectors, and incorporated in new machinery and equipment, or intermediate inputs. Other things being equal (namely demand, capital and skill requirements, etc.) sectors capable of producing their own technology tend to introduce more product innovations which may create new jobs, while sectors acquiring innovations from outside are dominated by process innovations whose employment impact is generally negative (Vivarelli, 1995; Evangelista 1999).

4. The productivity effect. The capital and innovation intensity of sectors interact with other factors, such as the skills of the workforce, the learning processes, organizational models, infrastructural conditions, localized external economies, etc, leading to different productivity growth rates. The impact of productivity on employment, however, is a complex issue; depending on the sources of productivity growth and on demand patterns, productivity growth may parallel employment growth (as in the 1960s and 1970s), or may be associated to job losses (as in the 1980s and 1990s). Sectoral specificities lead to a wide divergence of productivity performances, which affect the aggregate growth pattern of national economies (Appelbaum and Schettkat, 1995)

5. The demand effect. The growth of demand is not the same across the economy. The income elasticity and price elasticity of demand for individual sectors affects the growth rates of industries, which is also affected by institutional and policy factors, such as the organization and development of new markets and the role of public demand.

Countries which are more active in the industries with a faster growing demand tend to increase faster production, export and employment, at the expense of countries specialized in slow-growing sectors.

Changes in demand patterns tend also to take place slowly, especially when a technological paradigm shift is under way, as is currently the case with the emergence of ICT-based technologies. (Pasinetti, 1981, Freeman and Soete, 1994).

6. The trade effect. Sectors and markets widely differ in the importance of international trade. In markets more open to international competition, such as most manufacturing industries, the employment outcome of demand growth can be reduced by import penetration, and by foreign outsourcing of domestic firms. Conversely, competitive, export-oriented sectors, and industries with high national self-reliance may expand jobs alongside production. Services generally are less open to international competition and this has strongly contributed to their faster employment growth. (Wood, 1994).

2. Industry and services

Obviously, the first major distinction is that between industry and services. It is well known that in advanced countries employment growth in recent decades has concentrated in the latter.

The OECD data on employment growth in industry and services show that in the European Union (including 13 countries) the number of jobs in industry has fallen at average annual rates close to 1 per cent in the 1970s and 1980s, and at above 3 per cent in the 1990-94 period. In the US and Canada the 1970s saw a growth above 1 per cent, the 1980s a stability, and the 1990s a 1 per cent decrease. In Japan employment has steadily grown by 0.7-0.8 a year since 1970 up to 1994. In services Europe has also had an employment performance worse than its competitors, with increases close to 2 per cent a year in the 1970s and 1980s and at 0.6 per cent in the 1990s. In North America service jobs have grown at above 3 per cent in the 1970s and at 1.7 per cent in the 1990s, with a pattern slightly higher than in Japan (OECD, 1996A, p.12).

In the section devoted to the empirical evidence, a detailed comparison of employment performances in industry and services is provided for the US, Japan and European Union from 1975 to 1996. In the US total employment has increased over the period by 45%, resulting from an increase above 10% in industry and above 60% in services. In Japan total jobs have grown by 25%, with an increase below 20% in industry and above 40% in services. In Europe total employment has stagnated, with a 6% increase over the period, while industrial jobs have fallen by almost 20% and service jobs have increased by 40%.

Services account for about 70% of total US employment, close to 60% in Japan and between 50 and 60% in European countries; in spite of the high heterogeneity of the activilities falling in this field, including some very high and some very low skilled and paid jobs, services are the most dynamic job creating sector. The much larger weight of services in the US is at the root of its better aggregate employment performance.

Moreover, the US and, to a lesser extent, Japan show overall employment increases which are sustained by still positive job creation capacities in industry. On the other hand, European countries are characterized by an overall employment stagnation which is mainly due to a dramatic labour-saving tendency within the industrial sectors. In other words, in Europe the service sector hardly compensates job losses within industrial sectors, while in the US and Japan the positive employment performances are determined by both the service expansion and by the continuing positive contribution of industrial sectors.

The importance of the composition of the industrial structure is highlighted by a few simple data. In the section devoted to the empirical evidence, it is shown that Japan and the US have close to half of their manufacturing value added in industries that have increased employment at the OECD level, while Europe lags behind and, conversely, has much higher activities in the sectors where job reductions have been greater.

The combination of generally lower aggregate growth and higher shares of industrial employment in sectors dominated by labour-saving restructuring is at the root of Europe’s unemployment problems. Structural factors, rather than labour market functioning, are a key explanation for the job losses in Europe’s industry.

Specific cross-sectoral studies have been carried out on the factors shaping employment changes in advanced countries.

Pianta, Evangelista and Perani (1996) have shown that, over the 1980s, the performances in terms of growth of value added and employment strongly differ across countries. Looking at the impact that technological and structural change - proxied by patenting and gross investment, as well as growth of value added - have on employment growth across sectors, the study found that for the aggregate of G-6 economies, the sectors showing the highest rates of investment and innovation have experienced in the 1980s greater growth of output and employment.

However, such a process becomes more uneven when the individual countries - especially the European ones - are considered, because the benefits of the compensation effects are distributed as a result of a competitive process, which is affected by the sectoral specialization of individual countries. Countries with greater activity in the fields where new demand and employment grow faster are better positioned to capture the compensation effects, while economies dominated by declining industries are major losers in the employment effects of technological change. In Europe, and particularly in Italy, the "virtuous circle" between technology, growth and employment is much weaker as investment and innovation have focused on the restructuring of traditional sectors, and are associated to large labour-saving effects.

A cross-sectoral analysis is carried out in another study (Pianta 1998) on five European countries (Denmark, the Netherlands and Norway as well as Germany and Italy: all countries with data from the new Community Innovation Survey) and employment changes in 1989-93 are shown to be positively affected by changes in demand and output (value added and exports) and by the prevalence of product innovations. On the other hand, innovative expenditure per employee (or R&D expenditure per employee or per sales) and the share of new products in exports turn out with a negative employment impact.

In explaining the changes (in most cases the decrease) of employment in European industries, demand factors and product innovations have a positive impact. However, a greater innovation (or R&D) expenditure is associated to worse employment outcomes, suggesting a dominant pattern of labour-replacing technological change. An interesting finding relates also to the export performances of European countries. The sectors where new products are more relevant in exports have worse employment outcomes, suggesting that in the industries which are more innovative and internationally open European competitiveness has been weak.

The evidence emerging from these relationships is that, in the cycle of growth which includes the recession of the early 1990s, demand factors have been crucial in sustaining employment, while techological factors show contrasting effects for product innovations (positive) and general technological efforts (negative).

A generalization for European countries, the US and Japan has also been carried out on the basis of the relevance of industries dominated by the employment-friendly model of high demand growth and product-oriented innovations, or by the labour-displacing pattern of low demand and process innovations. Again, European countries emerge with the worst positions both in terms of their industrial structure (with a large weight of the latter group) and in terms of their relative growth performances (with slower growth of value added and employment in the former group).

4. Policy issues

Strong differences emerge, as we have seen, between industry and services and across manufacturing sectors in terms of their growth rates, demand patterns, export orientation and nature and intensity of investment and technological change, all key factors shaping countries' employment outcomes. The structure of the economy has therefore to be considered closely in all analyses of unemployment. This is even more relevant in the case of developing countries, where sectoral differences, institutional, social and geographical factors often lead to a marked dualism in the economy. The characteristics of the economic structure therefore represent a key issue to be considered in the analysis and policy on unemployment.

The relevance of sectoral specificities suggests a policy approach which targets the industries with high growth and employment potential, combining public and private efforts in the complex process of structural change. This approach had been at the root of decades of highly successful industrial policies in advanced countries, and has led the rapid growth of new industrial countries in East Asia and elsewhere.

However, in the last decade, the neo-liberal ideology has emphasized the reliance on market-based selection mechanisms also as a tool for developing the sectoral specializations and the pattern of structural change for individual countries. Industrial policies have been considered as a "distortion of the market", as they provide incentives and support which can be specific to particular sectors and groups of firms. Therefore, the scope for policy has generally been limited to creating framework conditions (equal across sectors and firms) and to assuring the efficient working of specific markets (OECD, 1994, 1996).

As a result the polarization of countries' performances has increased, with those stronger in the fast-growing industries capturing most of the employment gains. This approach needs be reversed, and some "targeting" of sectors with high growth and employment potential, high externalities and high capacity for learning is required in all countries interested in strengthening their activities in these fields.

The rationale for a new wave of industrial policies accompanying structural change is made stronger by the current transition to a technological paradigm dominated by Information and Communication Technologies (ICTs) and by the emergence of a range of industrial and service activities largely based on ICTs. The policy aim in this field has to be the emergence of new economic activities able to meet economic and social needs, to develop new markets and offer new growth and employment opportunities. However, so far the evolution of most ICT activities has been driven by the design of suppliers rather than by the requirements of the users, resulting often in a limited expansion of new activites and in a unrealised potential of the new technologies. The "technology push" which in past decades has created countless innovations in ICTs appears now as a straitjacket for the expansion of economic activities based on ICTs, as what is lacking now are, on the one hand, the coordination and coherence of organisational, institutional and social innovations and, on the other hand, the operation of a "demand pull" able to launch the growth of new large markets for new goods and services (High level expert group, 1995).

However, focusing the analysis and policy on the supply structure alone leads to a one-sided view of the complex process of structural change. The development of new sectors of activity, offering important employment prospects, requires action on the demand side. In recent years demand policy has been severely constrained in most countries by continuing anti-inflationary monetary and fiscal policies, reducing demand, depressing growth and increasing unemployment.

What is needed for fostering the development of sectors with a high employment potential - either in ICTs or in the rest of the economy - is the ability to provide an effective "demand pull", relying not so much on old-style public procurement, but rather on new schemes "empowering the users", which might accelerate the development of markets for new appropriate goods and services, favouring in this way the process of structural change and the transition in specialisation patterns.

The success in such a transition has to be assessed considering both gains and losses in value added and employment. In general terms, a "positive" transition to a new pattern of specialisation is one in which the net gains are more than what is needed to "compensate" the losers through the various mechanism of redistribution, in order to assure social cohesion.

Finally, the coherence between the objectives of industrial and innovation policy and those of aggregate economic growth is essential. Without an active industrial policy, the evolution of countries' economic structure is likely to consolidate their traditional strengths and weaknesses, as a result of the operation of comparative advantages. This pattern is unlikely to provide positive employment prospects, as the pressure is on competitive performances and cost reductions. On the other hand, a policy aimed at reducing unemployment, developing new economic activities and productive competences requires a long-term effort at changing a country's economic structure, combining action on the supply side (developing new infrastructures, investment, innovation, competences and labour skills) and on demand side, orienting private consumption and public expenditure to the selected new fields where the new employment potential is greater.

Industry and services

The empirical evidence is based on data for the US, Japan and the 15 EU countries over the period 1975-1996. The link between growth in value added and employment will be investigated both in aggregate and separately for industry and services.

As it can be seen from figure 1, over the period 1975-95 the EU employment performance has been stagnating while the US shows an overall increase of about 45% and Japan of about 25%. In other words, over the long run, Europe has been characterized by almost jobless growth , while the US and Japan have exhibited a clearly superior job-creation capacity. It is important to notice that these differences would probably result even larger if total working time evolution - and not the number of employees - is taken into account (for instance, Padalino and Vivarelli (1997) found out a negative working time trend in West Germany and France over the period 1960-1994.

These revealed differences turn out to be even more obvious if attention is turned towards the industrial sector alone. As it can be seen in figure 2, while the US and Japan are characterized by (moderately) positive trends (about +15%), Europe exhibits a clear downward pattern (about -18%).

Even in services, although all the three areas show steadily increasing trends, the US is characterized by the best employment performance (about +63%), followed by Japan (about +46%) and the EU (about +42%).

The divergence in the employment performances of industry and services is remarkable, and can be found in all three areas, regardless of their different performances, especially in the case of European countries. Of course, the weak employment dynamics of Europe might be considered as the result of a stronger productivity performance, which might improve competitiveness and lead to faster growth. Yet, this does not seem the case in the EU, whose GDP growth has indeed been lower than in the US and Japan. So, in terms of employment, job losses due to productivity gains are not compensated through a competitiveness effect.

The data used for investigating the sectoral differences within manufacturing industry come from the OECD STAN database (1996) for 31 manufacturing sectors (ISIC classes) and include most OECD countries - the US, Japan and all countries of the European Union (with the exception of Ireland and Luxembourg, and with the inclusion of Norway) are considered. The variables considered include value added (in constant prices) and employment (number engaged), measured as average annual rates of change from 1975 to 1994.

The context for this analysis at the sectoral level is provided by Figure 4, showing for the OECD aggregate the rates of change of value added and employment of the 31 manufacturing sectors, illustrating the overall pattern of structural change which has taken place over the two decades investigated in the previous section. Across industries a positive relation is evident, and the 31 sectors can be associated to three groups.

1. Growth sectors, where value added and employment have both increased, including plastic, printing, other chemicals, electrical machinery, food, aircraft and motor vehicles.

2. Declining sectors, with an opposite pattern of fall in both value added and employment, including leather, footwear, tobacco and shipbuilding.

3. Sectors in restructuring, with growth in production and fall in employment, include all remaining sectors and the values for total manufacturing.

For total manufacturing, the rates of change of value added and employment differ significantly across the three areas. Employment has increased by an annual rate of 0.43% in Japan and 0.11% in the US, and has fallen by -1.32% in Europe. Value added has increased by 3.86% in Japan, 2.94% in the US, and 1.62% in Europe.

If we look at the group of the European Union the dynamics of value added and employment show a rather different distribution, presented in Figure 5. A systematic gap is evident between the overall OECD pattern and the much lower performances of Europe, which has most of its industries in the restructuring quadrant. Japan is the country with the highest number of sectors in the growth quadrant, while the US has an intermediate position.

If we focus on the most recent years, on the 1989-94 period, Europe's performance appears even more negative, with only one sector - plastic - left in the growth quadrant, while all the textile and leather sectors, all transport industries, non electrical machinery, metal products, iron and steel, pottery and rubber moving into the declining quadrant. Europe's distance from the US and Japan has further increased in the 1990s.

The way the sectoral structure of national economies affect the aggregate employment outcomes shown above can be highlighted by the shares of the industries showing employment growth, moderate decline or strong decline in the 1975-94 period.

Table 1 ranks the 31 industries according to their employment growth at the OECD level and shows their importance in the industrial structure (measured by the shares of manufacturing value added) of Europe, the US and Japan in 1975 and in 1994. Again, the "burden" on Europe’s performance resulting from a larger presence in sectors with lower growth clearly emerges.

Table 1: Employment change and the sectoral structure of manufacturing industry

The industries with positive employment growth in the OECD group in 1975 accounted for about 37-38% of manufacturing value added in Japan and the US, and just 32 per cent in Europe; at the end of the period all countries have increased their shares, with Japan at 49%, the US at 44 and Europe at 40%. The stronger and increasing presence of the US and Japan in the sectors with a positive employment dynamics is a major factor behind their better aggregate employment performances highlighted above.

In the industries with moderate employment decline no clear cut picture emerges, while Europe is strongly over-represented in the job losing industries, with 14 per cent of its value added in 1994 (down from 18 per cent in 1975), against the 9-11 per cent of the US and Japan, which have been able to reduce much more rapidly their reliance on manufacturing activities with a marked job destruction pattern.

* The issue of structural change, and its impact on the evolution of employment is examined in the following works. The fundamental theoretical contribution is that of Pasinetti (1991).

E. Appelbaum and R. Schettkat 1995, Employment and productivity in industrialized economies, International Labour Review, vol.134, n.4-5, pp: 605-623.

Pasinetti, L. 1981. Structural Change and Economic Growth, Cambridge, Cambridge University Press

Pasinetti, L. 1990, "Structural Change and Unemployment", in Structural Change and Economic Dynamics, vol. 1, n. 1.

Pianta, M. - Evangelista, R. - Perani, G. 1996, The Dynamics of Innovation and Employment: An International Comparison, Science Technology Industry Review, Paris, OECD, 67-93

Pini P. 1995, "Economic growth, technological change and employment: empirical evidence for a cumulative growth model with external causation for nine OECD countries, 1960-1990", Structural Change and Economic Dynamics, n.6, pp.185-213.

Sakurai, N. 1995, "Structural change and employment: empirical evidence for 8 OECD countries", STI Review, 15.

Scherer, F. M. 1992, International High-Technology Competition, Cambridge, MA, Harvard University Press.

* The specific issues raised by the emergence of ICTs are addressed in:

Freeman, C. and Soete, L. (1994), Work for All or Mass Unemployment? Computerised Technical Change into the Twenty-first Century, London, Pinter

High level expert group, 1996, Building the Information Society for us all, Commissione europea, DG V, Bruxelles.

* The role of technology and trade across sectors and in the process of structural change and employment creation is examined in the following works:

Evangelista, R., 1999, Knowledge and investment. The sources of innovation in industry. Aldershot, Elgar.

Padalino, S. and Vivarelli, M. 1997. The Employment Intensity of Economic Growth in the G-7 Countries, International Labour Review, vol. 136, 191-213.

Pianta, M. 1998, Innovation, demand and employment, paper presented at the TSER project Conference: "Technology, Economic Integration and Social Cohesion", CEPREMAP,, April 15-16, 1998, Paris

Vivarelli, M. 1995. The Economics of Technology and Employment: Theory and Empirical Evidence, Aldershot, Elgar

Wood, A. 1994, North South trade, employment and inequality: changing fortunes in a skill-driven world. Clarendon Press, Oxford.

* Empirical analyses and policy recommendations are found in:

OECD 1994 The OECD Jobs Study. Part I: Labour market trends and underlying forces of change, OECD, Paris.

OECD 1996A. Technology, Productivity and Job Creation, Highlights. Paris, OECD

OECD 1996B. Technology, Productivity and Job Creation, Analytical Report, Paris, OECD