The notion that employment is affected by the level of aggregate demand has been a major part of Keynes' contribution to economic theory and policy. In the post-war period a new set of economic policies relying on demand management and a growing role of the state helped sustain unprecedented growth rates in industrial countries. Since the mid-1970s, confronted with high inflation and structural problems, most countries and international economic organizations - the ILO being an exception - have abandoned the view that demand policy can reduce unemployment; a new orthodoxy of anti-inflationary monetary policy, fiscal consolidation and reduction of state intervention in the economy has dominated the economic opinion and policy in both advanced and developing countries.

The unprecedented levels of unemployment in Europe and in many other parts of the world, however, point out the need to pay attention again to the dynamics of demand and the role it may play in an active employment policy.

The economic policies based on the theory of Keynes (1942) reacted to the experience of dramatic mass unemployment of the 1930s in Europe and in the US with an explicit objective to reach "full employment" and the determination to use demand policy, and public expenditure in particular, as a means to expand production and employment up to the point where joblessness was confined to a frictional unemployment of 2-3% and to voluntary unemployed (Beveridge, 1944).

The dynamics of growth in such an approach relied on the multiplier effect that autonomous demand (be it private investment, net government spending or net exports, all variables which are independent from current income levels) has on the level of output and incomes. Conversely, an inadequate demand may lead to a viciuous circle of low growth, higher unemployment, even lower demand and generally negative expectations, resulting in a depression.

The positive effect on economic and employment growth, however, is constrained by several factors. Output growth is limited by a country's propensity to save and import, and by the rate of taxation.

The effect on employment depends on the growth of productivity, as the number of jobs increases only if production increases faster than productivity. As production grows, productivity tends to increase, due to greater scale of production, learning by doing and efficiency gains summarized in the "Kaldor-Verdoorn law". In turn productivity and employment growth are affected by a variety of "supply side" factors including the nature of investment, the rate and nature of technological change, the forms of organization of production, etc.

Let us now consider in turn the different components of demand and their impact on employment.

2. The components of demand

Consumption

Private consumption remains the largest destination of product and income, at about 60% for the total of the world economy (sse table 2 in the section on the Empirical evidence). Consumption depends on the level of disposable income; as incomes increase the propensity to consume falls, as consumption patterns generally follow the "Engel's curve" logistic shape. A progressive redistribution of income would therefore increase the economy's aggregate propensity to consume and (in absence of external imbalances) may lead to faster growth. The current trends towards greater income polarization in most countries have the opposite effect of reducing the multiplier effect; moreover, higher unemployment and negative expectations on growth and income security lead to lower consumption and greater cautionary savings also from wage-earners, slowing down further the multiplier effect and contributing to the depressive mechanisms pointed out above.

Investment

The role of investment is crucial in generating growth. In the Keynesian model investment is part of autonomous demand and is not constrained ex ante by available savings; rather it depends on entrepreneurial opportunities, expectations and interest rates. In all growth theories, moreover, investment generates the new productive capacity needed for growth and employment.

The main constraint on investment comes from interest rates, which in the last decade have reached unprecedented high levels, slowing down investment in most countries (see table 2 in the Empirical evidence section). This can be considered as a major cause for the slower economic and employment growth, and the evidence of the positive association between the investment intensity of economies and their growth of incomes and productivity is summarized also in Figure 2 in the Empirical evidence section below.

The positive impact of high investment on aggregate growth, increased incomes and productivity has been widely documented. However, since the 1980s a large part of investment has been targeted at the restructuring and rationalization of production with "intensive" investment usually incorporating technological change, with a labour-saving nature. The expansive effects of the investments which are "extensive" of productive capacity, and therefore require greater use of labour, have been offset in many countries, especially in industry, by the restructuring effects. Therefore, the degree to which investment may support employment creation should not be taken for granted and needs examination in the specific contexts of national economies, business cycle phase, sectoral patterns and business strategies.

Export

Exports are the element of demand which depends on other countries' income. Since the 1980s great hopes have been put on exports in a context of sluggish domestic growth. The faster growth of world trade relatively to domestic demand has led many countries to rely on export growth as the main source for economic progress. The experience of world areas shown below in the section on the Empirical evidence suggests that only East Asian countries have been successful in associating high export shares in their GDP to faster growth of both GDP and per capita incomes. Their experience can hardly be generalized to all developing countries in spite of the insistence of such a policy prescription by international organizations such as the IMF and the World Bank. And for advanced countries the insistence on "export competitiveness" as a major policy priority has hardly been able to support faster growth and job creation (see the chapter on "Unemployment and international competitiveness").

Government expenditure

The long-term growth of government expenditure has been a major characteristic of rapid post-war economic growth in all countries. Government spending has expanded new sectors of economic activitiy, in services such as health and education, has supported innovation and industrial growth, has created the conditions for growing private activities and for redistribution across generations, social classes and regional areas, assuring social cohesion.

The expansion of government expenditure in the form of acquisition of goods and services directly stimulated production and employment growth; in the form of transfer payments it usually stimulated private consumption because of redistributional patterns; in the form of public investment it contributed to the expansion of autonomous demand at the core of the Keynesian growth model. Insofar as the economic and social returns to public investments are greater than market interest rates, there is no reason for worrying on the negative consequences of deficits and debt.

Since the 1980s, however, the high levels of public spending, deficits and debts have been viewed as a major culprit for high inflation, high interest rates, instability and slow growth. The new orthodoxy has emphasized the need for policies targeted at reducing all three aspects of government intervention. The arguments used have included the inflationary effects of spending, the pushing up of interest rates, the worsening of external imbalances in small open economies, the inefficiency of government-run activities, the need to reduce tax rates. In fact, the high taxation needed to finance high levels of government spending in advanced countries has been criticized for several reasons, and may also have the effect of depressing private consumption and investment.

A particularly active role has been played by international organizations such as the IMF, the World Bank, the OECD and European Commission in making most countries implement such policy prescriptions. The result has been a general, albeit small, decrease in government expenditure as a share of GDP in industrial countries of Europe, North America and East Asia (see the Empirical evidence section), together with a widespread wave of privatization of state-owned firms.

Policies for reducing government deficits have became institutionalized in several countries; in the US Congress has acted on the issue; in the European Union the terms of the Maastricht Treaty for creating the European Economic and Monetary Union included a requirement that deficits should not exceed 3% of GDP; in many developing countries, from South Africa to East Asia, the conditions set (or suggested) by international organizations such as the IMF have included similar restrictions.

As the cumulation of government deficits raises public sector debt, similar policies have been directed at reducing its size relatively to GDP. The argument here has been the need to reduce the pressure on interest rates (whose high level have been at the root of slower growth in recent years) and avoid the "crowding out" of private investment.

Such policies of cutting down state intervention in the economy have been hitting some of the most important sources of employment growth. In fact, the expansion of public sector employment has been by far the major source of employment in the 1970s and 1980s in most European countries. Even the OECD Job Study had to recognize that "in Europe (EC and EFTA countries combined) under 4 million new private jobs were generated over the last two decades compared to almost 7 million in the public sector, whereas in North America the private sector accounted for over 30 million new jobs and the public sector for only about 5 million" (OECD, 1994, p.60).

Naturally, demand is shaped also by the distribution of national income between wages, which have a much higher propensity to consume, and profits, which may affect firms' expectations and investment decisions.

3. The relevance of demand in policies against unemployment

The criticism of Keynesian views and policies developed since the 1970s has now turned into a new orthodoxy where a "natural rate" of unemployment is accepted, set by the structure of the economy and labour supply (Friedman, 1968) and a "non-accelerating inflation rate of unemployment" (NAIRU) is defined as the rate after which a reduction in unemployment has the effect to increase prices. The focus is therefore turned on the functioning of the labour markets; the fear that an active demand policy might start off inflationary pressures is the main reason for abandoning traditional Keynesian policies.

Obviously, labour market mechanisms and demand factors, as well as technological change and the structure of the economy, all contribute to the explanation of the current unemployment problems. However there is no conclusive evidence on whether unemployment is mainly linked to labour market conditions or is resulting from insufficient demand (for econometric evidence arguing for the importance of demand see Coen and Hickman, 1988).

The weakening of demand in most economies, however, suggests that some of the current unemployment may be of "Keynesian" nature, and might be reduced by expansionary policies.

Is this mechanism still able to work? The ILO World Employment Report (ILO, 1996, p.89) summarizes the evidence from the world econometric models of the National Institute of Economic and Social Research in Britain and of the Observatoire Francais des Conjonctures Economiques in France. The findings are that a coordinated fiscal expansion in Europe of 1.25% of GDP would lead to a 3.5% growth of output after three years and to a 1.5-2% fall of unemployment after four years, with a 1.5% additional inflation in a country such as Germany, after five years. Moreover, a coordinated stimulus in all EU countries would not lead to external imbalances as the degree of economic integration of the EU is extremely high. The growth of production and incomes would generate enough additional tax revenue to greatly reduce the initial government budget deficit, and measures to smooth the expansion of supply may further reduce the inflationary pressure.

However, reliance on the "automatic" operation of the multiplier effect of demand policy should now be qualified with a careful consideration of the economic context and structural conditions of national economies. In particular, attention should be paid to the changes in the growth model of advanced economies which have emerged in the last decade.

Rapid growth and successful demand policies in the post-war decades were associated to the diffusion of a ‘Fordist’ growth model which had combined a technological paradigm based on mechanical and chemical innovations, with the organization of mass production and the expansion of mass consumption made possible by a significant redistribution of productivity gains to wage earners. Such a combination of developments has required not only an active aggregate demand policy, but also a complex regulation system which included a specific set of institutions, active public policies on both the demand and the supply side, and major social arrangements which have led to the emergence of welfare states. These mechanisms have been common to all advanced countries, although national specificities have obviously emerged in the development of institutions and modes of regulation.

In the last decade many fundamental aspects of such a growth model have changed. In particular, a new technological paradigm based on Information and Communication Technologies (ICTs) has emerged, the forms of organization of production have become more decentralized, economic and financial activities take place at an increasingly global scale, the nature of new jobs, their location and forms of employment have moved away from previous patterns.

All these factors are bound to affect the way in which an expansion of aggregate demand can be turned into new employment, and therefore a return to active demand policy has to be accompanied by consideration of the specific macroeconomic, structural, social and institutional context of national and regional economies where it may be implemented. In particular, the following issues need to be explicitly addressed in order to make a return to demand policy effective against unemployment:

• an international coordination of expansion is essential, at least at a broad regional level (e.g. the European Union, or East Asia), in order to avoid external imbalances;

• new arrangements in labour contracts and institutions are needed, overcoming some of the rigidities of the previous model and avoiding the social uncertainty and economic hardship resulting from the current spread of part-time, temporary, low wage work.

• the expansion of the supply of new goods and services meeting new economic demands and social needs, especially those based on ICTs, needs to be facilitated, in order to match the process of technological and structural change with the evolution of demand patterns. This issue is developed in the next section.

4. The role of demand in emerging industries and markets

The emergence of the new technological paradigm based on ICTs and the changes in the structure of the economy (see the entries on "Unemployment and technological change" and on "Unemployment and the structural composition of the economy") point out the need for specific consideration of the role of demand in emerging industries and markets.

Demand growth, price elasticities, and income elasticities are important for industry-specific employment trends. The higher the rate of demand growth in an industry, the higher employment growth will be. However, the positive demand effects may be offset by productivity growth. Indeed, there is empirical evidence that the positive correlation between industry-specific productivity growth and employment has turned into a negative one in more recent years (Appelbaum and Schettkat, 1995), as most manufacturing industries have been dominated by the restructuring and rationalization strategies of firms, with a strong labour-saving bias.

Where then new industries offering the opportunity of high employment growth can be found? If we examine the development of activities based on the new technologies, a dramatic mismatch can be found between the high potential of new products and consumption patterns offered by ICTs, with more various and ‘personalized’ goods and services, and the lack of emergence of new large markets with strong demand. The slow learning processes in consumption, the need for social innovations (particularly in the use of time) required to ‘match’ the opportunities of technological innovations, the lack of appropriate institutions and public policies managing such problems are all factors which may explain such a mismatch. But a more direct economic factor is also important, associated to the current distribution of incomes with the reduction of the wage share. The strong polarization pattern, most extreme in the US, but clearly present also in Europe, has reduced the aggregate demand effects and has prevented the emergence of a large demand for new ICT-based products from wage-earners (Petit, 1995, Appelbaum and Schettkat, 1996).

In turn, without the opportunities offered by new markets and growing demand, the transformations on the supply side, driven by technological change, firms’ restructuring and the search for competitiveness, have reduced, rather than expanded, the production and employment base.

Therefore, an active demand policy is needed also for implementing the structural and technological changes currently under way The coherence between the objectives of aggregate economic growth and of industrial and innovation policy is essential. Without a more expansionary policy, the impact of supply side measures is likely to emphasise the losses of old activities and the cost-cutting effects of actions for competitiveness, rather than the gains in new activities, competences and employment. This has clearly been a factor which can explain the weak growth and employment performance of many countries, especially in the EU, in recent years.

It is in this context that some scope for demand policy has been recognized also by international organizations such as the OECD which have long been critical of expansionary policies. In the recommendations of the OECD study on Technology, productivity and job creation (OECD, 1996, p.227) a role for demand is identified in the limited field of "potential developments in network-based multimedia as an example of the role of governments in new demand articulation in a potential growth area". The focus is on government role in the following fields:

• the "shaping of emerging markets for network-based services", offering them a "vision of how new activities will emerge" in order to reduce uncertainty in the markets;

• "government as a leading edge user" of the information infrastructures and applications

• "promoting public understanding" in order to lower social barriers to the diffusion of ICTs in workplaces and households.

A new generation of demand-led expansionary policies is required to reduce current unemployment in all countries. It needs to be qualified and adapted to the current economic conditions, paying attention to the increasing globalization of production, to the structural changes of national economies, and to the evolution of labour markets. As the ILO World Employment Report argued, "full employment is still feasible and highly desirable", and this objective maintains its policy relevance "in spite of the rise in non-standard forms of work and other recent changes in the labour market" (ILO, 1996, p.xvi).

Table 1 shows the distribution of GDP in the main elements of demand in all world countries in 1995 (the total does not sum up to 100 because of the national imbalances). Table 2 summarizes this evidence for major regional areas, showing the evolution of the components of demand since 1980 and comparing it to growth performances of GDP and GDP per capita. All data are taken from the World Bank World Development Report, 1997. 

The structure of demand shows some important regularities and substantial variations across countries.

Consumption is always the major destination of GDP (increasing from 59 to 63% at the world level from 1980 to 1995) and with much higher shares in low income countries. The lowest values are found in some oil-exporting countries and in fast-growing Asian economies relying with a high investment effort or a strong export orientation.

Among rich countries, a share of consumption generally between 55 and 60% contrasts with the position of the US with 68% of GDP devoted to private consumption.

Investment is a component of demand with great variation, ranging from the extremely low accumulation of the poorest countries (below 15% of GDP), to a share around 20% typical of high income countries (the US again is an outlier with just 16% of GDP going to investment), to shares above 25% in Japan, East Asian economies and other fast growing countries. Over time the share of investment has declined slightly at the world level and in rich countries, with major falls in Sub Saharian Africa and Latin America, and a major rise in East Asia.

Exports as a share of GDP have remained stable at the world level and in high income countries between 1980 and 1995 at 22%, but have almost doubled in East Asia (reaching 29%) and South Asia (reaching 14%), and fallen in other regions. The export orientation varies widely from small oil-exporting countries or export platforms (147% of GDP in Hong Kong) to large, domestically oriented economies (12% in India). Even in poor countries it is often high because of the need of the countries to earn foreign currencies needed for debt service, putting a heavy burden on their prospects for development.

Government consumption as a share of world GDP has declined from 16 to 15% over the period considered, with higher reductions in rich countries, and generally lower levels in Asia and Japan, and higher shares in Europe.

The impact of demand structure on growth rates is pointed out in Table 2 for the main world regions. Two relationships stand out and are shown in the Figures.

Figure 1 shows the positive association between high shares of investment in GDP and higher growth rates of GDP per capita (but the same holds also for GDP growth). East Asia is an extreme performer in this regard, but a systematic positive link is found also for the other regions.

Conversely, the role of export orientation in sustaining GDP per capita (and GDP) growth is much less straightforward, as Figure 2 shows. East Asia and Sub Saharian Africa have very similar shares of export in GDP, but have the highest and lowest GDP per capita performance. In parallel, South Asian economies have reached high growth with the lowest export orientation.

The lessons are that while investment is an essential factor for long term growth, export can play a different role in particular regions, and the current emphasis on export orientation or competitiveness as key policies for growth and employment appears unwarranted by the actual growth performance of world countries.

A more specific evidence on the employment patterns associated to such performances is provided in the Empirical evidence section of the entry on "Unemployment and development: urban areas".

References

* The theoretical and policy approaches which have identified demand as key factor for growth and reduction of unemployment include the following:

Beveridge, W. (1944) Full employment in a free society, Allen and Unwin, London, 1944.

Boyer, R. (1988), 'Formalizing growth regimes'. In Dosi G. - Freeman, C. -Nelson, R. - Silverberg, G. - Soete, L. (eds). Technical Change and Economic Theory, Pinter, London, pp. 603-35.

Keynes, J. M. (1936), The general theory of employment, interest and money, MacMillan, London, 1942.

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

* Overviews, empirical studies and policy arguments of employment patterns associated to demand dynamics and economic change include:

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

E. Appelbaum and R. Schettkat (1996), Product demand, productivity and labour demand in a structural model, Paper for the TSER conference "Technology, economic integration and social cohesion", Cepremap, Paris, 22-23 November 1996.

Caracostas, P. - Muldur, U. (1995), Long cycles, technology and employment: current obstacles and outlook, STI Review, n.15.

Coen, R.M. and Hickman, B.G. 1988. Is European Unemployment Classical or Keynesian?, American Economic Review, vol.78, 188-93

Grieve Smith, J. (1997) Full employment: a pledge betrayed, Macmillan, London.

Michie, J. and Grieve Smith, J. (eds.) (1994) Unemployment in Europe, Academic Press, London.

Pianta, M. (1995), "Technology and growth in OECD countries, 1970-1990", Cambridge Journal of Economics, vol.19, n.1, pp.175-187.

M. Pianta, R. Evangelista, G. Perani (1996), The dynamics of innovation and employment: an international comparison, Science, Technology Industry Review, n.18, 1996.

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.

Rowthorn, R. (1995) Capital formation and unemployment, Oxford Review of Economic Policy, n n.1.

* The employment consequences of the combination of demand factors with changes in the supply structure and technologies have been investigated by a large literature, including:

Fagerberg, J. (1994), "Technology and international differences in growth rates". Journal of Economic Literature, vol. 32, pp. 1147-1175.

C. Freeman, J. Clark e L. Soete (1982), Unemployment and technical innovation, London, Pinter.

C. Freeman e L. Soete (1987) (eds.) Technical change and full employment, Oxford, Blackwell.

C. Freeman e L. Soete (1994), Work for all or mass unemployment?, London, Pinter.

Petit, P. (1995b), Technology and employment: key questions in a context of high unemployment, STI Review, n.15.

Science, Technology, Industry Review (1995), special issue on Technology, productivity and employment, n.18, Paris, OECD.

* The neoclassical views of unemployment and the criticism of the role of demand in reducing current unemployment can be found in:

Bean, C. European unemployment: a survey, Journal of Economic Literature, 1994, N.2.

Drčze, J.H and Bean, C.R. (eds.). 1990. Europe’s Unemployment Problem, Cambridge (Mass.) MIT press

Friedman, M., (1978), The role of monetary policy, American Economic Review, March.

Layard, R. - Nickell, S. - Jackman, R. 1991. Unemployment: Macroeconomic Performance and the Labour Market, Oxford, Oxford University Press

Layard, R. - Nickell, S. - Jackman, R. 1994. The Unemployment Crisis, Oxford, Oxford University Press

Malinvaud, E. 1990. What Do we Mean by Explaining High Unemployment?, Structural Change and Economic Dynamics, vol.1, 15-26.

Malinvaud, E. 1994. Diagnosis Unemployment, Cambridge, Cambridge University Press

Phelps, E. 1992. A Review of Unemployment, Journal of Economic Literature, vol.30, 1476-90

* The policy debate has been influenced by several reports, including the following:

Centre for Economic Policy Research 1995. European Unemployment: Is There a Solution?, London, CEPR

ILO, World Employment 1996/97, National policies in a global context, Geneva, ILO, 1996.

OECD (1994), The OECD Jobs Study, 2 voll. OECD, Paris.

OECD (1996), Technology, productivity and job creation, 2 voll., Paris, OECD.