EXCERPTS (Sections 1-5; Bibliography)
Paper published by IPPR, March 1998
Frank Vandenbroucke
Download Word 6.0 Version
1.1. A first conceptual distinction is introduced by Glyn and Sutcliffe (1992, pp. 76-77). The globalisation of capitalism can mean one of two things : (i) the spread of capitalist relations of production, or (ii) an increase in the international interdependence of the world economic system.
Consider the wave of denationalizations since the beginning of the 1980s, the collapse of central planning in Central- and Eastern Europe, and the introduction of market mechanisms in China. These changes instantiate a "spread concept" of globalisation : capitalist relations of production are spreading when an increasing proportion of the economy is based upon them. It is to be contrasted with the "interdependence concept". "Interdependence" applies to the interaction between national economies : national economy A and national economy B are more interdependent, when the impact of developments in A upon developments in B (and vice versa) is growing.
"Spreading" and "growing interdependence" are conceptually distinct processes, but they can be going on at the same time. Consider the process which has been most prominent in triggering the debate about globalisation, i.e. the replacement of "State Led Industrialization" by "Export Led Growth". This means that countries once weakly linked to international trade (because of protective barriers to trade and investment) now get strongly linked. This process involved both the spreading of markets and increased interdependence in the world economy.
It should be stressed that, so far, we are talking about interdependence of economies, not about the capacity of nation states to act. The latter is a related, but different issue, to be taken up in (1.3), when I introduce a distinction between "internationalization" and "globalisation".
1.2. Glyn and Sutcliffe introduce a second conceptual distinction : during the last two hundred years the fundamental nature of the world economy has changed on several occasions in respect both of its degree of globalisation and of the extent to which one nation is hegemonic in Kindleberger’s sense (Kindleberger, 1973). Such a dual characterization generates in principle four polar cases : hegemonic systems, with relatively open or closed economies; leaderless ones with relatively open or closed economies. "Very schematically, the gold standard represented a hegemonic system (UK dominance) but with market forces imposing discipline on relatively open economies, the centre of gravity being represented by the policies and performance of the UK. The 1920s and 1930s represented a non-hegemonic system with relatively closed economies pursuing independent policies. In the 1950s and 1960s the USA was the hegemonic power presiding over relatively closed economies and with discipline exerted through IMF and other non-market channels of US power as well as through markets. Finally the decline of US domination left the world economy leaderless in the 1970s and 1980s with increasingly open economies dominated by market forces, but without a unique centre of gravity." (Glyn and Sutcliffe, 1992, pp. 78-79).
Hirst and Thompson (1996) don’t apply this dual characterization in any systematic way, but make a related point. They warn us not to confuse conjunctural changes (floating exchange rates, monetary turbulence...), related to the weakening of US hegemony, with a structural shift towards a global economy. "The immediate conjuncture of the weakening of US hegemony led to conditions in which it might appear that a globalized economy would emerge. (...) The point is that these changes were conjunctural, although important in their effects and large scale, and that they were at least in part policy driven. The period of extreme volatility and turbulence did not last for long. (...) The old post-1945 multilateral order was not restored, but a drift into uncontrolled market forces on the one hand, or negative competition between the major emerging trading blocs on the other, was prevented." (Hirst and Thompson, 1992, pp. 14-15). Whether or not the deregulation of the international monetary and financial system is "conjunctural" or "structurally linked" with the decline of US hegemony is a question to which I have no answer. But that does not diminish the relevance of the dual characterization of states of the world, referring to both dimensions of the pair "global"/"non-hegemonic".
Note, however, that a non-hegemonic "leaderless" world may be detrimental to national governments’ capacity to pursue domestic social strategies. Hegemonic power can substitute for something which is otherwise difficult to achieve : voluntary cooperation among nation states. Cooperation, stability and predictability are important features, shaping the context for domestic action. West-European social democracy developed in the immediate post-war years under the umbrella of American economic hegemony. Hence, to understand the predicament of West-European social democracy today, it is important to make the distinction between a leaderless world and a global world. A leaderless world may need more vigorous efforts for economic and financial cooperation - to be implemented, for instance, within the European Union.
1.3. Whether or not increased economic interdependence inhibits national governments in their pursuit of specific economic and social policies is yet a different question. The distinction made by Hirst and Thompson between a "global economy" and "an international economy" is useful in that context. They propose the following general description of what is meant by "globalisation". Though most of the proponents in the debate would not fully endorse it (Giddens may not, for example), I think it captures well the essence of much of the debate :
"It is widely asserted that we live in an era in which the greater part of social life is determined by global processes, in which national cultures, national economies and national borders are dissolving. Central to this perception is the notion of a rapid and recent process of economic globalisation. A truly global economy is claimed to have emerged or to be in the process of emerging, in which distinct national economies and, therefore, domestic strategies of national economic management are increasingly irrelevant. The world economy has internationalized in its basic dynamics, it is dominated by uncontrollable market forces, and it has as its principal economic actors and major agents of change truly transnational corporations, that owe allegiance to no nation state and locate wherever in the globe market advantage dictates." (Hirst and Thompson, 1996, p. 1).
Those who hold this view need "a relatively clear and rigorous model of what a global economy would be like and how it represents both a new phase in the international economy and an entirely changed environment for national economic actors." (Hirst and Thompson, 1996, p. 7). Hence Hirst and Thompson develop two basic contrasting ideal types of international economy, a fully globalized economy and an open international economy :
"An inter-national economy is one in which the principal entities are national economies. Trade and investment produce growing interconnection between these still national economies. (...) The form of interdependence between nations remains, however, of the ‘strategic’ kind. That is, it implies the continued relative separation of the domestic and the international frameworks for policy-making and the management of economic affairs, and also a relative separation in terms of economic effects. (p. 8) "In (...) a global system distinct national economies are subsumed and rearticulated into the system by international processes and transactions. (...) The international economic system becomes autonomized and socially disembedded, as markets and production become truly global. Domestic policies, whether of private corporations or public regulators, now have routinely to take account of the predominantly international determinants of their sphere of operations." (p. 10).
Hence, globalisation must encompass the idea of "macro-economic unification" which, in the words of Glyn and Sutcliffe, "represents the idea that the world is now really a single economy in the macro-economic sense. That means that the main determinants of income and employment can now only be understood at a global and no longer a national level." (Glyn and Sutcliffe, 1992, p. 77).
On the basis of descriptive and quantitative data Hirst and Thompson conclude that globalisation, if the concept has to signify something distinctly different from internationalization, is a myth. As I see it, the distinction between the two concepts is most relevant as a distinction between two states of the world, rather than a distinction between processes : the expression "globalisation", in this framework, means that we are close to a truly global state of the world. I agree with Hirst and Thompson’s ideal-type distinction, so conceived, and with their conclusion. I would however stress that the related concept of "macro-economic unification" is a matter of degree. Intensifying internationalization (via increased trade, investment and financial flows) leads to higher levels of macro-economic interdependence, e.g. via increased import-leaks in Keynesian demand policies or integration of financial markets. (Hirst and Thompson make related points.) So understood, there is a continuum of states of the world, characterized by various degrees of macro-economic unification, affecting in a diverse way the impact of domestic economic policy instruments on domestic economic results.
1.4. Finally, it is possible that the processes described in the previous paragraphs do not take place at a worldwide level, but within a more confined area (e.g. within Europe, North-America, Asia and the Pacific...). For that reason a distinction is sometimes made between between regionalisation and globalisation. In order to be consistent with the conceptual framework developed in (1.3), it is better to draw the following distinctions : (i) "globalisation" on a regional level; (ii) "globalisation" on a worldwide level; (iii) "internationalization" on a regional level; (iv) "internationalization" on a worldwide level.
It is quite clear that, from the point of view of trade, the European economies went through a process of internationalization on a regional level, whilst the openness of the European economy as a whole vis-à-vis world markets has not increased significantly (in contrast to the US economy, which became more open to world trade over the last 30 years). One could even argue that within Europe a process of "regional globalisation" is developing, as European nation states lose their capacity to use certain traditional instruments of economic policy, whilst "worldwide globalisation" is much less relevant with regard to discussions on Europe’s travails. To make the picture even more complicated, one could argue that from the point of view of smaller economies, such as Belgium, Europe already presents itself as a globalized policy entity with regard to some aspects of traditional demand and exchange rate management, whilst that is not the case from the point of view of, say, Germany. Just as "interdependence", "globalisation" need not be a symmetric concept. I mention this variety of possible descriptions of the European situation, not in order to rush to conclusions, but to emphasize the complexity of the issues involved, and the need to look closely at the facts with which different national economies in different parts of the world are confronted.
2.1. Consider the facts captured in the following quotes : "In 1978 one third of the global work force lived in centrally planned economies. At least one third also lived in countries weakly linked to international trade because of protective barriers to trade and investment. In contrast most of the world’s population now lives in countries either deeply integrated into the world economy or rapidly becoming so. China, India and the former USSR jointly make up nearly half of the global labour force." (Wes, 1996, p. 1-2). "The proportion of the world’s workers who are shielded from the influence of international economic forces likely will drop from around two-thirds in the late-1970s to about one-tenth by the end of the 1990s." (Diwan and Walton, 1997, p. 2).
These quotes refer both to increasingly spreading capitalist relations of production and to growing interdependence. The spreading of capitalist markets is certainly a hard fact and a revolutionary change. In itself, the mere "spread" aspect of globalisation is important, first of all from an ideological point of view (collapse of state planning, collapse of state led industrialization in the Third World), and, secondly, because of its possible economic impact in the future. Even if we are sceptical about the claims of globalists concerning today’s level of "interdependence" globalisation, today's "spread" globalisation might tell us something about the interdependence globalisation that can be expected 10 or 20 years from now.
2.2. As many have pointed out, such a development is not without precedent. Sachs and Warner summarize both sides of the picture neatly. On the one hand : "The years between 1970 and 1995, and especially the last decade have witnessed the most remarkable institutional harmonization and economic integration among nations in world history. While economic integration was increasing throughout the 1970s and 1980s, the extent of integration has come sharply into focus only since the collapse of communism in 1989. In 1995 one dominant global economic system is emerging." (p. 1). But they also point out : "It is often forgotten today, in the flush of the communist collapse after 1989, that global capitalism has emerged twice, at the end of the nineteenth century as well as the end of the twentieth century. The earlier global capitalist system peaked around 1910 but subsequently disintegrated in the first half of the twentieth century, between the outbreak of World War I and the end of Word War II. The reemergence of a global, capitalist market economy since 1950, and especially since the mid-1980s, in an important sense reestablishes the global market economy that had existed one hundred years earlier." (Sachs and Warner, 1995, p. 5).
2.3. With regard to interdependence the following issues are important :
2.3.1. Trade
* Just before the onset of World War I world merchandise exports (which exclude services) accounted for 11.9% of GDP of OECD countries. By 1950 this fraction had fallen to only 7.1%. International trade as a share of world GDP recovered its 1913 level around the mid-1970s. This share has been increasing more or less steadily since then, to reach 17.1% today.
* Exports from low-wage Newly Industrializing Economies (NIEs) have undergone significant changes in their composition, and are growing at an unprecedented rate. Their manufacturing exports entail a major change in the international division of labour since the 1960s, when developing countries primarily exported raw materials.
Much of the confusion about the importance of globalisation follows from two statements, which are both true : (i) there has been an explosion of exports from NIEs, a phenomenon that was essentially non-existent twenty-five years ago, and (ii) they are still not very large as a proportion of total OECD spending. Exports of manufactured goods from NIEs amounted to 0.24% of GDP of all industrial economies in 1970; in 1990 they had increased to 1.61% of GDP of all industrial economies (Krugman, 1995, Table 4).
* Globalisation is, in a sense, better suited to describe the recent American experience than the European experience. The relative importance of world exports for the European economy as a whole (i.e. excluding intracommunity trade) increased only slightly over the last thirty years, while it nearly doubled for the US. US imports rose even more rapidly. This holds for imports in general, and for non-OECD imports.
The proportion of non-OECD imports to manufacturing GDP shot up in the US from 1.2% in 1970 to 5.4% in 1990. In France the increase was from 1.8% to 3.7%; in Germany from 1.8% to 3.7%; in the UK from 2.6% to 4.4% (Desjonqueres et al., 1997, Table 1, p. 21).
* In Europe, intra-European trade surged. In general, one could say that the dominant pattern in the world economy has been regionalisation, of which the prominent example is the "europeanisation" of the economies of the European Economic Community.
* Though the aggregate volume of world trade is, today, not at a level which marks a qualitative difference from previous experience, there are, according to Krugman (1995, p. 331- 337), at least four new aspects of modern world trade - new in the sense that they did not have counterparts in the previous golden age of the global economy. These are the rise of intra-trade, trade in similar goods between similar countries; the ability of producers to slice up the value chain, breaking up a production process into many geographically separated steps; the resulting emergence of supertraders, countries with extremely high ratios of trade to GDP; and the emergence of large exports of manufactured goods from low-wage to high-wage nations.
Aggregate trade figures may understate the competitive pressure from trade on the developed countries’ manufacturing sector, since manufacturing now weighs more heavily in imports than 50 or 100 years ago. And they may understate the increasing vulnerability of employment in some sectors in the OECD economies to developments in other parts of the world. For instance, slicing up the value chain makes outsourcing easier. Hence the impact of changes in relative wages on the demand for labour in the sectors concerned increases, which makes employment more vulnerable to wage changes; or, the threat of relocation becomes more credible. This is a central argument in Rodrik (1997a), which I discuss in Section 3.3.
2.3.2. Financial flows
* During the 1980s there has been a surge in foreign direct investment, both between developed countries and - in the second half of the 1980s and the beginning of the 1990s - from the developed world to a number of nations in the developing world. However, in a long term historical perspective, including the period before the first World War, and on an aggregate level, the figures are not without precedence. Moreover, Krugman (1994, p. 119) estimates that the entire net outflow of investment has reduced the total capital stock of developed countries by only 0.5% of what it would otherwise have been.
* With regard to international financial markets, three developments are to be distinguished : (i) their growth; (ii) their deregulation; (iii) their integration. The explosive growth and deregulation of international financial markets are well-documented : in the 1970s and 1980s controls on international flows of capital were gradually dismantled, and ever larger sums of money are now moving across borders. Whether we really live in a financially fully integrated world, is a much debated issue, which I will not discuss here. Do the undisputed changes make external financial credibility more important for governments today than 30 years ago ? Probably yes, because the Bretton Woods arrangements that were still in place 30 years ago made it easier to avoid or fight runs on currencies, at least in the short term. However, states have never been able in the long run to defend exchange rates that are unsupported by real economic performance, as the UK history of the 1960s and 1970s shows (cfr. Hirst, 1997, pp. 420-421). In Section 4 I argue that the abolition of capital controls rendered peculiar "non-textbook" mixtures of fiscal and monetary demand management harder to sustain, but capital mobility has not made demand management impossible, contrary to a widespread view.
2.3.3. Labour markets
There is a global labour market for some specific highly skilled workers; an atomic physicist or an airline pilot or a surgeon can migrate where they want (Glyn and Sutcliffe, 1992, p. 86). But the market for medium-skilled and unskilled labour is not at all globalized; it is on the contrary more and more closed on a national basis. This is an important distinction with earlier periods in our history (Glyn and Sutcliffe, 1992, p. 86; Hirst and Thompson, 1996, pp. 22-26).
With regard to labour markets the crucial question is whether increased international trade substitutes for labour mobility across borders, i.e. whether it produces similar effects on relative wage levels. According to the classical Hecksher-Ohlin and Stolper-Samuelson models of economic theory it can well have such an effect. If goods can move around freely, and some other conditions are satisfied, those models predict that labour may come to face the same wages everywhere, even if labour itself is not mobile. This is what people like Reich (1991) must imply when they talk about "global labour markets". I tackle this question in Section 3.
Some conclusions on the basis of (2.3) :
* From the point of view of the developed countries there has been growing interdependence, via trade, foreign direct investment and financial markets, since the 1950s.
* Although modern world trade is different from trade in the previous golden age of the global economy, we have, so far, no obvious arguments to qualify the changes in patterns of trade and the growing interdependence as reflecting the emergence of a "global economy", radically different from our standard notion of the "international economy". Nation states are confronted with a process of gradually increasing economic interdependence, but one cannot conclude from the data that they must have lost their capacity to act and react strategically as nation states and to pursue domestic economic and social strategies. I write "so far" in the previous sentence, because I will come back to the issue of constraints on government action, focusing on a more precise question, in Section 4.
To some extent "regionalisation" is even more apt as a descriptive concept than "internationalization", most of all with regard to trade. Within Europe, "europeanisation" of the economies is an important fact. From the point of view of the small supertraders, like Belgium, Ireland and maybe The Netherlands, one could argue that this is a process of "European-level globalisation", in that they face within Europe the situation that the globalizers claim to be true for the world as a whole. If, for some reason, one wants to use the notion of "globalisation" (rather than "internationalization") to point to an increase in truly worldwide trade interdependence, it is by all means better suited to describe recent American experience than the European experience.
* The current level of interdependence is not unprecedented, though the degree of economic integration that has been attained now within the European Union is unprecedented.
* However, global interdependence is growing, as trade is growing. Without unexpected political or economic shocks to the world economy, it will gradually reach unprecendented levels.
2.4. One of the most significant changes of the last twenty years is not economic globalisation, but globalisation in the area of mass media and mass communication. It is the unprecedented possibilities for ultra-fast and worldwide dissemination of cultural, political and commercial information (of which financial information is but a part) which creates a new setting for political action and ethical reflection. Ethics and politics are closely linked to human communication. Global mass communication raises the possibility of an emerging "global civil society" (it would be far-fetched to say that it exists already). But it may also fuel unnecessary anxiety about our own position in the world and our own security, and thus contribute to the revival of local nationalisms or xenophobia.
I believe Giddens is right in saying that what he considers as a general process of globalisation also concerns "the transformation of local, and even personal contexts of social experience." (1994, p. 5). It is however not clear which policy prescriptions follow from this general intuition. Certainly, the institutions and social phenomena which are now spreading across the world are not confined to competitive markets and capitalism : witness the dissemination of democratic institutions, cultural practices,.... In this sense the conceptual dichotomy introduced in Section 1.1. is conceived too narrowly : there are not only economic institutions which can be spreading and/or creating interdependencies; more dimensions of modernity are spreading in today’s world (cfr. Giddens, 1990). However, for the purpose of our inquiry the question then is whether causal links exist between those dimensions of "globalisation" and constraints on social democracy in developed welfare states. In Section 4 I emphasize values and beliefs as constraints on social democracy. If cultural dimensions of globalisation cause changes in prevailing values and beliefs - conducive, for instance, to the "marketisation of reward" - then there is such a causal link. The evidence, however, does not admit more than speculation on this issue.
2.5. Awareness of global ecological problems
The increasing awareness of global ecological problems among the public at large is to be considered as an important fact in its own right. If there is a legitimation crisis of the nation state, it is also a crisis of faith in the nation state’s ability to address global ecological problems (Dunn, 1994, p. 5-6). People - fortunately - see connections which they formerly did not see.
High-skilled workers are more likely to be winners on OECD labour markets, whilst low-skilled workers are more likely to be losers, both in terms of income and job-opportunities. Nearly everybody agrees that the disadvantage of the low-skilled - both in terms of earnings and employment prospects - has increased in nearly all developed countries, most dramatically in the UK and the USA. There is less agreement about the explanation for this persistent tendency : is it (a) trade with newly industrializing countries of the Third World, (b) technological change, biased in one way or another against unskilled workers, or (c) changes in institutions (decline of trade unions, changes in regulation, e.g. with regard to minimum wages, changes in the bargaining process) ? Most of the debate has focused on the choice between trade explanations and technology explanations. Rival institutional explanations of growing inequality are sometimes neglected in the "trade versus technology" discussions and the related surveys.
I am not able to adjudicate in this complex and ongoing discussion, but under (3.1.) I formulate five "propositions" which emerge as more-or-less consensus opinions from the literature on trade and technology, and I add some comments.
3.1. Trade, technology and the financial return to skill : five propositions
1.1. Biased development in demand for skills
It is generally agreed that the increasing disadvantage of the unskilled, measured in terms of employment and income, follows from a fall in demand for unskilled labour over the last couple of decades, relative to the demand for skilled labour, in most developed countries. These changes in demand are to be analysed in relation to supply. Both "trade" and "technology" explanations take this as a starting point.
Two caveats are appropriate here. First, Nickell and Bell (1995) show that one should be careful in specifying the link between relative demand shocks and relative unemployment rates for skill groups. A greater absolute rise in the unemployment rate of the least-qualified can be the result of a complex chain of labour-market adjustments to a general fall in demand which is not, in the first instance, biased against them. Hence, disadvantage on the labour market and biased demand changes are not necessarily synonymous. Second, much of the increasing inequality has taken place within skill categories and not between categories. I take up this important finding in Section 3.2.
3.1.2. Trade-off between inequality and employment
Economists disagree about the causes of the shift in demand against the less-skilled : is it to be explained by growing trade with low-wage countries, which employ abundant reserves of unskilled labour, or is it to be explained by changes in technology, biased against the less-skilled ? According to Wood, most participants in the debate on the role of trade versus technology would however agree that "this shift in demand (against the less-skilled, F.V.) has increased wage inequality, or, where labour market institutions have propped up unskilled wages, as in Europe, raised unemployment among the unskilled" (Wood, 1995, p. 58). The existence of a trade-off between inequality and employment is indeed largely accepted in (most of) the economics profession, and increasingly so in the political world.
In reality this trade-off has not been evident over the last 15 years. The UK had a poor employment record over most of the period, with growing relative unskilled unemployment, and increasing wage inequality; conversely, the Netherlands had a good record in terms of wage equality and employment growth for a significant part of the period, to take but two examples. At least, one has to bring in other factors : macro-economic policy, which is important for economic growth; the development of the average wage level, which is important for the employment-intensity of growth and for its sustainability...The general argument that there is no visible trade-off between equality and employment is forcefully made by Franzmeyer, Lindlar, Trabold (1996) in a study commissioned in the framework of the preparation for the last Dutch EU Presidency. They summarize their account of economic performance in the OECD countries in the 1980s and the beginning of the 1990s with the following stylised comparison (p. 143) :
A trade-off between equality and unemployment since the early 1980s ?
A stylised comparison
More inequality Constant inequality
Higher unemployment United Kingdom Continental Europe
Constant unemployment USA Japan
Divergent macro-economic fiscal and monetary policies and the development of the average wage level account to a large extent for the results displayed in this table. Note that this comparison does not stretch further than the beginning of the 1990s. But changes since then (e.g. the decrease in UK unemployment vis-à-vis persistent high unemployment in France and Germany) may again be explained to a large extent by macro-economic developments.
A trade-off between pre-tax wage equality and employment for the unskilled can be defended as a reasonable assumption on a strict ceteris paribus basis, i.e. if there are no other influences on the demand for labour, such as changes in the average wage level, changes in macro-economic policy, changes in active labour market policies, training and education policies.... Rowthorn (1995) presents a small model on trade, employment and inequality which is based on such assumptions, and in which the trade-off between pre-tax wage inequality and employment for the low-skilled is a crucial chain in the argument. The model also shows how wage subsidies and taxes can restore post-tax equality.
Hence, accepting that there is, ceteris paribus, a trade-off between pre-tax wage equality and employment for the low-skilled, does not mean that social democrats are caught in an insoluble dilemma between post-tax income inequality and employment in a non-ceteris-paribus world, wherein macro-economic policies, active labour market policies, wage subsidies, training and education... can all play their role.
3.1.3. Revival of the Hecksher-Ohlin trade model
If one accepts the idea that there must have been biased developments in demand for unskilled labour, how might these be linked to trade ? Here a venerable economic theory, the Hecksher-Ohlin (HO) trade model, comes in.
The HO trade model explains patterns of trade between countries by the relative supply of factors of production within countries. Their endowment with factors of production shapes their "comparative advantage" : countries export those products which are relatively intensive in factors of which they have a relatively larger supply. The HO model has made a remarkable come-back as an explanation for patterns of trade between developed and developing countries, on the basis of the skill-intensity of products. The assumption that capital is internationally mobile means that countries are distinguished chiefly by how skilled their labour forces are.
Wood (1994, Chapter 3) and Sachs and Shatz (1994) use different definitions of skill, but show that trade between developed and developing countries closely conforms to the HO trade model. That is not sufficient to prove a causal link between trade and wage inequality or unemployment. An analytical link is provided by the Stolper-Samuelson (SS) theorem. Stolper and Samuelson showed theoretically that, under a list of restrictive assumptions, freeing trade caused an unambiguous reduction in the real wage of the relatively scarce factor (in our case unskilled labour). The skill version of HO-SS creates a framework for thinking about the "winners" and "losers" and about the "gains from trade" with compelling political consequences. As Wood (1994, p. 5) puts it : "[If North-South] trade is based on the availability of skills, rather than of capital, then expansion of trade with the South does not tend to drive down the average real wage in the North, nor to pull up the share of profits at the expense of labour, nor to push workers in general into unemployment. What it does, instead, is to widen the economic gap between skilled and unskilled workers." In other words, the key distributional conflict in contemporary society is between skilled and unskilled labour.
Economists disagree about the theoretical and empirical relevance of SS to the analysis of inequality on our labour markets; some of the crucial predictions of SS don’t pass empirical tests. However, it is not necessary to stick to the classroom version of HO-SS to argue that trade had some influence on labour markets, as Cline (1997) persuasively argues in a detailed survey of the vast literature. Cline discusses more than 30 studies, focusing mainly on US wage inequality. He reports for each study its estimate of the fraction of the increase in the skilled/unskilled wage ratio caused by trade, immigration, and/or globalisation. There is a concentration in the distribution of estimates somewhere in the range of 10 to 15 per cent. However, there are important extremes as well. The most obvious extreme is that of Wood (1994) who places 100 per cent of the causation on trade over the past two decades. At the other extreme, Berman, Bound and Griliches (1994), Bound and Johnson (1992) and Lawrence and Slaughter (1993) attribute little if any role to trade, and an overwhelming role to technological change (Cline, 1997, p. 139 and Table 2.3). Cline considers a reasonable estimate based solely on the literature he reviewed to be that international influences contributed about 20 per cent of the rising US wage inequality in the 1980s (Cline, 1997, p. 144), although his own estimate is significantly higher. However, Cline also emphasizes that there was a massive equalizing effect at work on the supply side of the US labour market : a persistent rise in the availability of skilled relative to unskilled labour. Hence, if the intent is to detect the relative contribution of trade (and immigration) to the full set of forces causing a widening of wage inequality, this contribution should be gauged against the gross unequalizing forces rather than just the net outcome. In Cline’s own estimate, measured against the full set of unequalizing influences (including erosion of the real minium wage, the decline of unionization, skill-biased technological change, and other unexplained forces) trade and immigration only accounted for about one-tenth of the gross unequalizing pressures over the past two decades.
One of the major studies on the US is Sachs and Shatz (1994), which enjoys the privilege of being cited with approval both by the relativists (Krugman, 1995, p. 344, note 13) and by some of those who believe in a rather important effect of trade. Their evidence with regard to trade (they look at relative product prices of the traded goods) cannot account for a significant widening of wage inequalities, though it points in that direction (1994, p. 40). In general, they conclude : "The overall changes in employment and in wage inequalities are too large to be explained by the changing trade and price patterns of the past 15 years. It is likely that technological change is playing a role independent of internationalization. Yet we cannot precisely measure the relative importance of these two factors - trade and technology - mainly because one cannot observe and measure technological change with any precision. We do not believe in choosing one or the other explanation, however : both are at work..." (1994, p. 4).
3.1.4. Global and local gains from trade; no protectionism
Since nearly everybody agrees that both the North and the South gain on average, protectionism is ruled out as a response (e.g. Cline, 1997, p. 271; Wood, 1994, p. 393). Rodrik sees a trade-off between free trade and social cohesion and proposes amendments to international institutions and trade agreements, but he warns against protectionism (Rodrik, 1997a, p. 75f).
Rowthorn (1995) also considers the introduction of "labour standards", conditioning the access of Southern goods to Northern markets. They are modelled by an increase of the relative wage of unskilled workers in the South by 50 per cent. Global income declines in these simulations, while the result for the South depends on specific assumptions, but is always a mixed blessing. A globally efficient solution to achieve more equality in the South and between unskilled in the South and unskilled in the North requires fiscal redistribution within the South (1995, p. 16; direct redistribution between North and South is not considered in this model).
3.1.5. Policy conclusions are indifferent to choice between trade or technology explanations; and the willingness to redistribute is crucial
Wood’s estimates of the impact of trade are considered to be exaggerated. His framework for analysing policy proposals remains however generally applicable and interesting (Wood, 1994, Chapter 10).
In order to cope with unemployment and inequality, Wood proposes a combination of (i) investment in education and training to expand the skilled labour supply in the North, which would raise economic efficiency as well as narrow the gap in wages and unemployment between skilled and unskilled workers; (ii) creating more demand for unskilled labour, by subsidies to private employers or public employment schemes; and (iii) redistribution of income from skilled to unskilled workers. Given the fact that (i) is a policy raising both efficiency and equity, it is understandable that it is cheered by nearly all governments in the North. Wood, however, analyses its technical and economic limits, and stresses that education and training are unlikely - though being the best option - on their own to be able to solve the problems caused by declining demand for unskilled labour (Wood, 1994, chapter 10; 1995, p. 78).
He also examines the consequences of applying these policy proposals when the underlying diagnosis would be wrong and concludes that it is better to apply them, even if we are not absolutely certain about the diagnosis. If the choice of diagnosis is only between trade and technology, there is in fact not much difference. As Burtless says (1995, p. 815), these policies will be equally helpful whether the sagging demand for unskilled workers is caused by overseas competition or biased technical change. Wood’s final judgement is instructive : "Since (once protection is ruled out) the appropriate policy responses to a falling demand for unskilled labour are the same, whether it is caused by trade or new technology, the opponents in the debate over causation often agree on solutions. But economists can always find something to disagree about : in this case, some argue that subsidies, by easing the plight of the unskilled, will slow the supply response, and thus protract the problem. However, low wages and unemployment are an obstacle as well as an incentive to skill acquisition : the net effect of the subsidies, particularly on the education of the children of the unskilled, is quite likely to be positive. A more serious risk is political : it is the reluctance among the skilled to pay the taxes needed to finance the subsidies to the unskilled." (Wood, 1995, pp. 78-79).
Rowthorn (1995) provides some numerical insights into the magnitudes involved. He compares policy scenarios with a base simulation featuring dramatic consequences from the opening up of trade with the South for the unskilled in the North (a 13.6 percentage point rise in unemployment for unskilled workers in the North, when relative wages are rigid; or, alternatively, a 13 per cent reduction in the pre-tax real-wage rate of the unskilled to preserve full employment). One policy scenario implies redistributing the average gains for the North (redistributing an 8.1 per cent increase in the North’s National Income in the flexible wage case). The redistributive simulation supposes that wages are flexible and that the tax system is used to compensate unskilled workers and restore their relative income to its pre-trade level. To compensate for the "base" case, the average rate of tax on skilled workers must be increased by 8.7 per cent, while the unskilled need to receive a subsidy of 23.2 per cent of their income. A remarkable feature of Rowthorn’s result is that taxation entails a Pareto-improvement : "These figures imply very large adjustments in marginal tax rates. Since taxes are in practice already progressive in their incidence, the result would be a steeply graded system with very [high marginal] tax rates for skilled workers and, possibly, negative tax rates for the unskilled. Such a situation might be politically unacceptable to skilled workers. In our simulations, employment subsidies require higher tax rates than unemployment benefits, but they also raise both national income and the pre-tax wages of skilled workers. The latter effect is more powerful, so that skilled workers are ultimately better off with employment subsidies despite the higher tax rates involved. However, this is a subtle calculation in which the direct cost of higher tax rates for skilled workers must be weighed against the indirect benefit arising from a stronger demand for their labour. Many of them will be keenly aware of the extra taxes needed to finance employment subsidies, but less convinced about the resulting benefits. In practice, this may lead them to oppose a tax and subsidy policy, which may well be to their ultimate benefit." (Rowthorn, 1995, p. 13). We have reason to be sceptical about the spell of such calculations. Certainly, they offer no robust proof that virtuous circles of redistribution, whereby everybody gains, can be created by a simple change in tax parameters.
Rowthorn indicates that such policy scenarios are equally valid for large shifts arising from new technology (1995, p. 18). The American economist Edmund Phelps recently published a book, in which he persuasively argues in favour of systematic, massive and permanent subsidies for firms employing low-wage workers (Phelps, 1997). Just as Rowthorn’s, his argument for low-wage employment subsidies admits of various economic forces which may have contributed to the widening wage gap in the US, but all point to the same policy prescription. Phelps also emphasizes that his plan, which replaces the Earned Income Tax Credit, would finance itself. Some of his calculations, however, amount to heroic assumptions and "shots in the dark" (as he admits himself, p. 127), to such an extent that I deem his overall budgetary account more a gamble than a watertight argument. If this self-financing property has to be the decisive political argument in favour of the scheme, as it stands it cannot be convincing. In fact, it is improbable that a society would engage in Phelps’s scenario unless it is prepared to redistribute resources to the advantage of the low-skilled without a matching return for the high-skilled. At least, there must be a willingness to take the "risk", from the point of view of the high-skilled, that such social investment is not fully self-financing, not even in the longer term. Rowthorn rightly adds a similar caveat to his numerical results.
Those who argue for redistribution are not necessarily old leftists nor ardent adherents of globalisation theories : in his exchange on globalisation with Kapstein, Krugman also insists on the willingness to redistribute : "The important point - on which I agree with Kapstein - is that it should be possible to make a large difference to the incomes of low-wage workers (in the United States) or to their prospects for employment (in Europe) without devastating effects on the budget. In the United States, the crucial thing to remember is just how poor the poor are and how rich the country is. If the United States were willing to devote, say, two percent of GDP to income supplements for the working poor, the effect would be dramatic. (...) A better target for Kapstein’s ire might be influential figures who insist that the only way to help the poor in America is to cut taxes on the rich." (Krugman, 1996, p. 167).
Reich is adamant with regard to the necessity to redistribute, when he discusses the policies needed to cope with globalisation. The former US Minister of Labour offers sobering reading for those who think that education and training are costless policies : "The greatest challenge comes in summoning the political determination to embark upon them. Good education, training, health care, and public infrastructure - available to all Americans - will be costly. (...) Thus a central question concerns the extent to which America’s fortunate citizens (...) are willing to bear these burdens." (Reich, 1991, p. 250)
3.2. The "marketisation" of reward : inequality within skill groups and the role of institutions
3.2.1. Both "trade" and "technology" explanations of growing inequality take it as a starting point that the increasing disadvantage of the unskilled follows from a fall in demand for unskilled labour relative to the demand for skilled labour. They focus on relative changes between skill groups. But is the overall growth of income inequality completely explained by changing inequality between skill groups ? Gosling et al. illustrate the increasing dispersion of earnings within the lower skilled occupations and the lower education groups in the UK (Hills, 1996, Figure 6.6 p. 149). In general it seems that, however narrowly one defines groups, one still finds an increase in income dispersion. Greater instability in earnings may be part of the explanation why inequality among people with the same characteristics increases : if short-term fluctuations in income become larger, then inequality measured on the basis of yearly earnings will also increase (Gottschalk, 1997, pp. 32-33, referring to US data). But how can we then explain the residual increase in within-group inequality ? Some researchers hold that this too is a matter of skill differences, albeit statistically invisible ones; others conclude that the increased dispersion within narrowly-defined occupational or educational groups is a reason for seeking additional explanations (Atkinson, 1997, pp. 308-309). According to Dinardo, Fortin and Lemieux (1995, pp. 34-35) this finding strenghtens the case of those who look for institutional explanations. Unlike supply and demand considerations, the two institutional explanations they consider - minimum wages and unionization - have clear and measurable impacts on both the dispersion of wages between and within different groups of workers in the United States. Atkinson (1997, pp. 309-311) mentions not only institutional factors such as declining unionization and wage regulation, but also social custom and norms : "It may, for exogenous reasons, have become socially acceptable to have larger wage differentials within the workplace." Although the empirical importance of rising within-group inequality does not diminish the significance of education for social policy, it challenges much of the received wisdom on skills and inequality : increasing financial returns for education and skills are part of a broader phenomenon, a process of "marketisation" of reward for work. Under (3.2) I discuss explanations for inequality focusing on formal institutions - leaving aside the intangible though important determinants of income best described as social custom and norms, and the underlying social values.
3.2.1. The role of formal institutions
How can we account for the different social impact of global economic developments on countries which are economically so close as the United Kingdom and the Netherlands, or Canada and the United States, if the global impact is really so overwhelming ?
The Dutch success in maintaining equality and fighting unemployment can to a large extent be explained by the "Wassenaar Agreement" of 1982 between the peak trade union and employers organizations, and the remarkable moderation of average wage increases that prevailed for the following 15 years. The difference with the UK has to be explained by both the "hardware" and the "software" of Dutch society, an issue to which I return in Section 4.
Canada and the United States have much in common : similar cultures, a similar standard of living, and similar economic institutions. It can be shown that during the 1980s the same underlying economic forces were at play in both countries (DiNardo and Lemieux, 1997, pp. 632-635). Moreover, with regard to labour market institutions Canada and the US are not so different from each other (in comparison to, say, the US versus continental European countries). Nevertheless, summary measures of inequality of disposable income from the Luxembourg Income Study show virtually no change for Canada from 1981 to 1987, whilst inequality increased markedly in the US from 1979 to 1986 (Atkinson et al., 1995, p. 49; cfr. also pp. 64-65). Relatively small differences in institutions must be powerful enough to explain the difference.
DiNardo and Lemieux (1997) compare the evolution of hourly wages of male workers in Canada and the US over the period 1981-1988 by computing counterfactuals such as the distribution of wages that would prevail if all workers were paid according to the observed non-union wage schedule, or if the minimum wage had remained the same. They find that the much more severe decline in both the unionization rate and the minimum wage rate in the United States, compared to Canada, accounts for two-thirds of the differential growth in wage inequality between the two countries.
The same methodology has been applied in earlier studies focusing on developments over time in single countries. DiNardo, Fortin and Lemieux (1995) use data on the distribution of hourly wages in the United States from 1973 to 1992, and estimate a decomposition of the changes in the distribution between 1979 and 1988 into five components : the effect of the minimum wage, unionization, other personal attributes, supply and demand and residual factors. They find that changing unionization rates, changes in the composition of the work force, and changes in supply and demand are important explanations of the change in wage inequality over this period. The decline in unionization had a substantial effect on the distribution of men’s wages, but a negligible effect on the distribution of women’s wages. The supply and demand conditions affected workers in ways that are consistent with the increasing returns to skill that are well documented in many studies on the US. Unlike much previous research, they find that the decline in the real value of the minimum wage over the period 1979-1988 had a large impact on the distribution of wages. It explains up to 25% of the change in the standard deviation of men’s log wages and up to 30% of the change in the standard deviation of women’s log wages. Its role is important to changes in the standard deviation both between and within different groups of workers (cfr. supra). They conclude that labour market institutions are as important as supply and demand considerations in explaining changes in the U.S. distribution of wages from 1979 to 1988. They also find that the increase in unionization and in the real value of the minimum wage between 1973 and 1979 contributed to the decrease in wage inequality over that period.
Bell and Pitt (1995) use the same methodology to assess the impact of changes in union density on the male wage structure in the UK over the period 1982-1993. They find that approximately 20% of the increase in the variance of log wages over the period can be attributed to changes in unionization.
These empirical studies are limited by the same methodological problem as most of the factor-content-of-trade studies we mentioned earlier : they ignore general equilibrium effects. In fact they only establish the difference between the actual distribution of wages and the distribution that would have prevailed "if individual attributes had remained at their 1979 level and workers had been paid according to the wage schedule observed in 1988" (DiNardo, Fortin, Lemieux, 1995, p. 11). They cannot take account of the impact of changes in the distribution of individual attributes (e.g. the degree of unionization, and thus the power of unions) on the structure of wages in general equilibrium, or on the number of jobs available. I believe (though this is not more than a conjecture at this stage) that this means that they underestimate the impact of changes in unionization and wage bargaining. True, the argument in a general equilibrium reasoning can go either way (cfr. the disputed impact of minimum wages on the number of jobs). But my conjecture is based on the fact that distributional developments in individual nations have diverged largely (e.g. USA/Canada; UK/Netherlands), with figures which imply that the bulk of the divergence and of the labour market phenomena themselves have to be explained by differences between nations - which are predominantly institutional.
Dinardo et al. and Bell and Pitt studied changes over time; Blau and Kahn (1996) explain different levels of inequality between 10 OECD countries in a highly interesting study, which provides a more sophisticated account of institutional determinants of (male) wage formation and inequality. They find that market forces - relative differences in the supply and demand of three broad skill groups between the US and the other OECD countries - do not appear to be a viable explanation for the international differences in inequality. High-skill workers are considerably scarcer relative to low-skill workers in other countries relative to the United States, yet the Blau and Kahn-data show that, with one exception (Switzerland), they earn more relative to low-skill workers in the United States than elsewhere. The most persuasive explanation for differences in male wage inequality is provided by labour market insititutions, chiefly the relatively decentralized wage-setting mechanisms in the US. The higher level of inequality in the US reflects considerably more compression at the bottom of the distribution in the other OECD countries relative to the United States, but a much smaller difference in the degree of wage inequality at the top of the distribution. The Blau and Kahn results imply that union pay policies that bring up the bottom of the union wage distribution are common to all countries, including the US. However, unions in other countries appear to be more succesful than US unions both in reducing inequality in the union sector and in extending such policies to the non-union sector. I believe this result does not contradict the results of DiNardo et al. (1995) concerning the impact of the legal minimum wage over time.
3.3. Trade, factor mobility and bargaining power
Rodrik (1997a) develops an argument which, in a sense, straddles concerns with the market for skills (as discussed in section 3.1) and concerns with the shape of institutions (section 3.2). He sees two channels through which globalisation affects labour markets. The first of these (and the one that has been extensively examined in the literature; cfr. section 3.1.) is the effect on relative demands for skilled and unskilled workers : trade is supposed to result in an inward shift in the demand curve for low-skilled labour in the advanced countries. This means that for any given wage level for low-skilled workers there will be less demand for their labour, in comparison with demand for their labour in a "non-globalized" world. The second channel, which Rodrik emphasizes, has to do with the greater ease with which domestic workers, particularly the low-skills type, can be substituted by other workers across national borders, either through trade (outsourcing) or through foreign direct investment. If it gets easier to replace worker A by similarly skilled workers B, C, D, ..., then A’s employment becomes more vulnerable to changes in his wage level relative to the wages of B, C, D.... So conceived, trade increases the elasticity of demand for labour and flattens the demand curve for labour at home. This means that the impact of changes in the wage level of low-skilled workers on the demand for their labour becomes greater, in comparison with the impact of similar changes in a "non-globalized" world.
Rodrik discusses three consequences of increased "substitutability", which undermine the postwar social bargain between workers and employers. First, workers now have to pay a larger share of the cost of improvements in work conditions and benefits. Second, they have to incur greater instability in earnings and hours worked in response to shocks to labour demand or labour productivity; that is, volatility and insecurity increase. (As I indicated earlier, instability may be part of the explanation for the rising within-group inequality of measured incomes in the US. As such it is one factor, among others, resulting in increased "marketisation" of rewards.) Third, their bargaining power erodes, so they receive lower wages and benefits whenever bargaining is an element in setting the terms of employment. The relevant measures of openness in this context are not the volumes of trade or investment, but the ease with which international transactions can be carried out.
Apart from the foregoing trade- and investment-related points, Rodrik also argues that once economic globalisation moves beyond a certain point, governments can no longer finance the income transfers required for social security in open economies, because the tax base becomes too footloose.
Rodrik’s arguments are cogently put, though he recognizes that we cannot be certain about the quantitative impact they imply, since the basic research on these questions has yet to be undertaken. Apart from the practical conclusions with regard to social policy and trade policy which he puts forward in the concluding chapter, I believe his theoretical argument, if validated by empirical research, would be highly significant from an European point of view. The focus on trade with low-wage countries ignores the fact that less-skilled workers in Germany or France are also in competition with similar workers in the UK or the US, markets with which the former countries are considerably more tightly integrated than they are with India or China. And while North-North trade may have little perceptible impact on the relative demand for unskilled labour, it makes this demand more elastic in all countries involved (Rodrik, 1997a, p. 26). In other words, the increase in the elasticity of demand for labour is a more general phenomenon, which is as much linked to "europeanisation" as to "globalisation". It is a direct consequence of international economic integration, regardless of the economic structure and the identity of trading partners. The same holds for Rodrik arguments with regard to taxation of capital : if deemed relevant, it must certainly be relevant within Europe.
Globalists, like Giddens and Gray, assert that the new context in which governments are functioning makes it impossible to implement some of the traditional social-democratic conceptions of redistribution, let alone Keynesian demand management. Hirst and Thompson fight both the enthusiastic and pessimistic globalisation theorists but take it as obvious that "radical expansionary and redistributive strategies of national economic management (are) no longer possible in the face of a variety of domestic and international constraints" (Hirst and Thompson, 1996, p. 2). Nevertheless they want to stress the scope for national action : "But this should not lead us to dismiss or ignore the forms of control and social improvement that could be achieved relatively rapidly with a modest change in attitudes on the part of key elites." (p. 7).
In this section I want to discuss the nature of the external and internal constraints facing governments who desire both more employment opportunities and a fair distribution of post-tax wages. For twenty years economic textbooks, journalistic and political statements have abounded with analyses of and references to the external constraints governments face with regard to macro-economic management. External constraints refer to (i) balance-of-payments constraints; (ii) trust in currencies as expressed by financial markets; and (iii) limits to taxation and social security contributions, because of cost-competitiveness and because of the mobility of some of the factors which are taxed, e.g. capital and top-management. It is impossible to survey this literature. I limit myself to making some conceptual points, which are useful in the debate in which social democracy is engaged.
First of all, we have to address a hidden assumption, which lies behind the idea that "social-democratic economic policies" have become impossible because of the insurmountable external constraints imposed by globalisation. The hidden assumption is that from a social-democratic point of view the appropriate policy when confronted with unemployment is in all circumstances some mixture of fiscal and monetary demand expansion organized by governments. This hidden assumption is simply untrue. In his important work, Crisis and Choice in European Social Democracy, Scharpf neatly expounds various "problem constellations" for which fiscal or monetary demand management offers no solution : stagflation, fuelled by cost-push inflation, is one example. Incomes policies were the key additional instrument needed to tackle stagflation. Hence, close cooperation between governments and unions was necessary (Scharpf, 1991, pp. 25-37). And even when confronted with the usual swings of the business cycle successful economic policy depended on the voluntary cooperation of unions and employers. The policy instruments directly available to governments are not sufficient to cope with all problem constellations, not even in a closed economy.
I stress this point, not because external constraints do not exist; they do. I stress it in order to eliminate the bizarre idea that "in the golden era" - before globalisation, so to speak - social democracy could always successfully rely on the single track of fiscal and monetary demand management by governments to fight unemployment, whatever the problem constellation causing it.
4.1. External constraints on demand management and redistribution
4.1.1. External constraints on Keynesian demand management
Demand management in open economies is externally constrained. But on this topic confusion prevails. It is, for instance, wrong to claim that global capital markets have made domestic demand management impossible. Different economic regimes call for different instruments of demand management, but do not rule it out. In a closed economy governments can rely on both fiscal and monetary policy to boost demand. In an open economy, the rules of macro-economic policy change. Broadly speaking, if the government chooses to fix the exchange rate, it will need to rely less on monetary policy and more on fiscal policy, as compared with the closed-economy case. If it chooses to let the currency float, the opposite is true : it will need to rely less on fiscal policy and more on monetary policy. In a world of mobile capital and weak capital controls governments run into trouble, when they try to maintain a semi-fixed exchange rate regime, without surrendering monetary policy exclusively to that purpose (i.e. without relying only on fiscal policy for demand stabilization).
Now, the fact that demand management is effective, does not mean that demand expansion is unconstrained by balance-of-payments considerations. Consider, for instance, the case of demand expansion in a regime of semi-fixed exchange rates and weak capital controls. In order to assess the external character of constraints on demand expansion, the following conceptual distinctions are useful :
(i) At the most basic level of analysis the impact of government aggregate demand management on domestic production and employment is expressed by the keynesian multiplier. This impact is reduced by "import leakages". To the extent that trade becomes more important, the value of the multiplier declines for individual countries, who would pursue a "go-it-alone" demand expansion. A reduced impact is not the same thing as a constraint however.
(ii) Rising imports, induced by demand expansion, lead to an external constraint when an unsustainable trade deficit emerges. Again, this is the typical problem of the bold go-it-alone expansion strategy in highly integrated economies like most European economies (cfr. the Mitterand experiment in 1981). It explains why many social democrats have argued that expansionist "national Keynesianism" has become very difficult within individual European countries, but can be applied at the level of trade blocs nearly as succesfully as it would have been applied thirty or forty years ago.
(iii) The impact of an emerging - or even an expected trade deficit - may be amplified by reactions in financial markets. Amplification effects of this kind are enhanced by (i) the increased quantitative importance of international money markets, (ii) the absence of a regulated international monetary system and (iii) the lack of cooperation between nations in fighting exchange rate crises. These factors, together with the frightening experience of several periods of monetary turbulence in the 1970s and the 1980s, may explain why governments have become increasingly cautious with regard to reactions in money markets, and substituted concern for the micro-economic competitiveness of their companies for traditional concepts of macro-economic management.
Apart from some specific nations, it is however not self-evident that the reason why central bankers today are generally "cautious" with regard to demand management is predominantly to be found in external factors. Fear of inflation is a domestic issue for the German Bundesbank (cfr. footnote ). And the development of public debt has become an increasing concern in its own right in many European countries and in the US, motivating restrictive fiscal policies. Moreover, it is not true that macro-economic policy has been so restrictive all over the place since we entered into the era of financial deregulation and non-cooperation. Monetary policy in the US has been expansionary over much of the last decade, and is seen by many as a crucial factor in the American economic success story of the last few years. (E.g. Franzmeyer, Lindlar and Trabold (1996) who emphasize differences in macro-economic policy between Europe and the USA, in their account of economic developments over the last decade.) Taking all these elements into consideration, it appears that external constraints, as understood in (ii) and (iii), are not a sufficient explanation for the general restrictiveness of demand management in Europe, in comparison with the US.
(iv) Muet (1997) provides a more subtle explanation for the cautiousness of European central bankers and governments. The combination of the semi-fixed exchange rate mechanism and non-cooperation with regard to monetary and fiscal management made it impossible to apply to Europe the mix of monetary and fiscal policies which was so successful to the United States in the 1990s. In Muet’s analysis, lack of cooperation (which presents itself as an "external problem") makes a combination of "internal" constraints (fear of inflation and mounting public debt) insurmountable. Muet’s macro-economic study corroborates the thesis which Jacques Delors has been pleading for years : "L’incapacité des nations européennes à mettre en oeuvre des politiques de relance pour sortir rapidement des récessions est inhérente à un ensemble de nations indépendantes, mais fortement interdépendantes. L’intégration des marchés de biens et de capitaux, ne rend pas impuissante les politiques économiques, mais elle crée une profonde asymétrie entre les différentes politiques. En l’absence de coordination, les politiques de flexibilité et de compétitivité par les coûts, qui sont d’autant plus efficaces qu’elles sont menées à l’échelle nationale, sont toujours développées de façon excessive. (...) Ce biais restrictif des politiques européennes est en partie le coût de la non-Europe politique. Les nations européennes sont presque aussi interdépendantes que les Etats américains, mais à la différence des Etats-Unis d’Amérique, l’Europe ne dispose pas des institutions de politique économique adaptée à cet espace économique." (p. 3). Delors’ proposition, recently reiterated by the French government, is to complement EMU with some form of economic governance for Europe.
So far, the concept of cooperative (and in fact very cautious) growth policies has been blocked by a joint effort of European conservative forces. Although it currently implies an uphill political battle, European social democracy should endorse Delors’ concept of European cooperation. Delors has reiterated his proposal for economic cooperation within EMU, making it more precise, in a recent paper (Delors, 1997b; cfr. also Delors, 1998). However, I should also stress that Muet’s economic argument is most convincing in terms of the efficiency of short-term adjustment policies ("sortir rapidement des récessions"), important though that is. I believe the more fundamental long-term constraints on successful demand management are of an internal nature, as they reflect distributional conflicts within countries. I discuss them in (4.2.) below. It is (fear of) these domestic constraints which make most central bankers so fundamentally cautious with regard to expansionist policies in Europe.
4.1.2. External constraints on redistribution via taxation and social security contributions.
Two issues are involved here : cost-competitiveness on the one hand, and the risks associated with the international mobility of specific factors of production at the other hand.
Social security contributions and taxation add to wage costs and capital costs, and thus run up against constraints of cost-competitiveness. However, it is misconceived to state that distributive efforts as such run up against external constraints : it is the combination of distributive efforts, as organized by governments, and uncompromising claims by economic agents (workers, self-employed, investors) to receive target levels of net compensation, which meets with external constraints. Conceptually, I consider this "external" constraint as a variant of the internal constraints discussed in (4.2.) below.
Mobility of factors of production is considered to be a problem for redistributive governments with regard to capital and top management. Notwithstanding the fact that I am strongly in favour of a European Union level of taxation policy (guidelines for avoiding predatory tax competition), I believe the argument that mobility of management personnel is an important constraint with regard to personal income taxation is overstated in those countries where changes are most needed : it is surely untrue with regard to the UK tax debate (Glyn, 1997a).
4.2. Internal constraints on demand management
For most individual European nation states external constraints, as described under (4.1.iv), are a reality in short-run policy decisions. In a somewhat longer-term perspective "external constraints" cannot explain why governments should abstain from demand management of the Keynesian type, except maybe for some of the small European supertraders. With cooperative trade unions and employers it is in principle always possible to pursue a mix of policies which stimulates demand and so employment, but creates neither a wage-price spiral, nor a public debt spiral. That is a fortiori true when we take the European Union as a whole. However, employment-stimulating demand management requires long-term average wage moderation (hence, concerted incomes discipline in one way or another) and willingness on the part of the government to adjust taxes as a function of macro-economic needs, i.e. also upwards, when necessary, and willingness on the part of workers to pay them without shifting them to firms.
For a deeper understanding of the issues involved, I again refer to Scharpf (1991; see especially Chapter X on the limits to government action). Glyn also stresses the importance of resolving conflicts on income distribution, as a key condition for successful egalitarian full-employment policies : "Above all, the internationalisation of financial markets has accelerated the speed, drama, and costs of financial markets’ retribution on governments whose policies are not deemed credible. But it is important not to identify this problem of credibility solely with narrow class prejudice. Sustainable egalitarian packages face very fundamental domestic obstacles. The costs of these packages have to be voted for; the conflict likely to be released by a sustained expansion of employment has somehow to be contained and channelled into directions consistent with maintaining the momentum of expansion." (Glyn 1997b, p. 8). Glyn’s analysis refers to Kalecki’s seminal paper on the political aspects of full-employment. Viable policies for expanding employment entail costs which must be explicitly counted and willingly shouldered by wage and salary earners. Unless social democracy can gain support for such an alternative, mass unemployment is set to continue as the mechanism by which distributional conflict is contained (Glyn, 1995a, p. 55).
4.3. Internal constraints on egalitarian employment policies
4.3.1. Conditions for successful employment policies
In the introduction I explained why it is necessary to qualify the idea of "employment" as an objective of social justice. I therefore introduced the notion of "egalitarian employment policies". Let us now bring together the analysis under (3.1.5.) and (4.2.), and assume that successful employment policies, so conceived, require both appropriate demand management and targeted employment policies for the low-skilled. Targeted employment policies for low-skilled workers can be organized by means of wage subsidies which make greater pre-tax wage flexibility socially feasible and/or by public employment schemes for the low-skilled, combined with more investment in education and training. It is not necessary to settle here on the extent to which the former (aggregate demand) or the latter (selective policies) are more important, nor on the precise rationale for the latter (be it technology or trade). Whatever the precise balance between demand management and selectively targeted employment policies, the internal constraint for successful egalitarian employment policies hinges upon :
(a) willingness to redistribute from rich (often high-skilled) to poor (often low-skilled)...
(i) to finance targeted employment policies,
(ii) to fund investment in education and training, and
(iii) to remedy unacceptable income inequalities which cannot be eliminated by such policies,
and
(b) willingness to accept some discipline with regard to the growth of the average wage level in both slack and tight labour markets (and, consequently, the acceptance that the evolution of wage and overall income differentials is a legitimate concern for policy too, with a view to the social and economic sustainability of an incomes discipline).
The expression "willingness to redistribute" requires more explanation, firstly with regard to the notion of "redistribution", and secondly, with regard to the use of the word "willingness".
The policies entering under (a-i) and (a-ii) are generally described as "social investment". Since those policies are not funding cash benefits and consumption, they are often sharply distinguished from "social spending", i.e. spending on social benefits, which is implied by (a-iii). The investment/spending distinction is useful, as it highlights that an intelligent welfare state should not only engage in paying benefits, but also in investment in human and social capital (a point stressed by the Commission on Social Justice, 1994). It is, in addition, often suggested that spending on benefits instantiates "redistribution", while social investment is something entirely different. There is an intuitive and politically attractive argument supporting such a dividing line : it holds that social investment makes everybody better off, while spending on social benefits makes some people better off at the cost of others; hence, the latter is "redistributive", and the former is not. Although it may be politically convenient, and in many cases appropriate, analytically it is difficult to maintain that social spending always embodies redistribution, so conceived, whilst social investment never does. First, social investment doesn’t come cheap, certainly not in the short run. And even in the longer term, some people may be permanently worse off, as a consequence of reallocating the resources invested in society. Or, to put the same idea in other words : even if social investment raises economic efficiency sufficiently to generate a macro-economic return, there is no guarantee that each individual citizen will be a net beneficiary. (Diminishing state funding for Oxford and Cambridge, or installing a "graduate tax", and increasing spending in other parts of education, are, probably, redistributive policies in the longer term too.) Secondly, some key policies are difficult to classify using the foregoing taxonomy : wage subsidies entail both an investment (in human capital, because people get work experience, thanks to the wage subsidy) and a direct form of redistribution (because some people’s taxes are funding an increase in other people’s net wages). Certainly, Phelps (1997) is not alone in arguing that wage subsidies for the low-skilled become self-financing and make everyone better off in the longer term. But the argument is clearly not watertight, and cannot be considered as "decisive". It thus seems very improbable that the high-skilled will be ready to engage in a policy of wage subsidies for the low-skilled, if they are absolutely not prepared to take the risk that such a policy is not self-financing in the longer term. The reality of things is that both much of "social spending" and much of what we call "social investment" appeal to our willingness to redistribute resources, often from well-identifiable high-skilled, high-income people, to well-identifiable, low-skilled, low-income people. And, in setting priorities, we are facing an inevitable trade-off between those two tracks of redistribution, more precisely, a trade-off between increasing benefits for poor people who cannot improve their situation via employment (for instance, because they are retired) and investing in jobs and skills for those who can escape poverty by means of employment.
The expression "willingness to" must not be understood as signifying a "moralistic" approach to social policy, neglecting the role of institutions. It captures a dispositional property of a complex social system shaped by institutions (e.g. wage-bargaining structures; tax-and-benefit systems) and by social culture, or, more precisely, patterns of choice within the social institutions, based upon prevailing beliefs and values. By using the word "willingness" I want to go beyond an account of formal institutions only and to emphasize the importance of beliefs and values in society. The design of formal institutions is, of course, of fundamental importance to generate the "dispositional property" to give employment priority on wage increases and to redistribute resources. But, in itself, institutional engineering doesn’t seem sufficient.
An example may illustrate this. In the schemes produced by the literature on collective bargaining structures, Belgium and the Netherlands are often similarly situated. However, over the last 15 years collective bargaining and employment in the Netherlands have developed in a significantly different way to that in Belgium. The Netherlands do not outstrip Belgium in employment volume as defined by Eurostat, i.e. in total hours worked per inhabitant of working age. Contrary to received wisdom, in this respect Belgium performs slightly better than its northern neighbour. But employment volume improved steadily over the last ten years in the Netherlands, faster than in Belgium or most other European countries, and it keeps improving. A consistent policy of wage moderation, agreed between employers organizations and trade unions, stretching from 1982 till today, explains much of this success. The Belgian record, in comparison, is poor : the 80s and 90s feature an unhappy history of mistrust between employers and trade unions, intermittent authoritarian intervention by governments, bouts of relatively large wage increases alternated with periods of forced wage freezes or cuts... Second, and most important, the hours worked are distributed over many more people in the Netherlands, by means of part-time employment, again in the framework of a broad social consensus. The Dutch experience proves that, in order to create jobs, one need not destroy social dialogue and essential labour market institutions, nor is it necessary to create huge inequalities : social consensus can deliver jobs. We need not copy everything from the Netherlands. The Dutch don’t have a magic formula; a significant job-deficit for the low-skilled persists. But, clearly, consensus between employers and trade unions has been very important in improving the overall employment situation, and to revitalize Dutch "corporatism" as an attractive model for European social policy, certainly more attractive than the Anglo-Saxon model.
It is not so easy to attribute these divergent experiences in two otherwise rather similar countries, Belgium and the Netherlands, to differences in the "hardware" of collective bargaining institutions only. It seems that part of the explanation has to be found in the "software" of collective bargaining, as it developed in the Netherlands over the last 15 years : the software consists of beliefs and ideologies, which supported mutual trust, at least at the level of peak organizations; and values which motivate economic and social actors. Marquand stresses essentially the same point in his account of the British experience, as opposed to the Rhenish model : social democratic governance needs a cultural bias, cooperative practices "depending on tacit understandings and uncodified assumptions, which are almost incomprehensible to those brought up in the Anglo-Saxon tradition." (Marquand, 1997, p. 183; cfr. p. 40).
Using a formal model of wage bargaining, Rowthorn (1992) illustrates the crucial importance of the "subjective mood" of the parties involved in collective bargaining. His simulations suggest that moderately fragmented bargaining structures (as in the European continental core) are highly vulnerable to shifts in the mood of play, and perform very badly when cooperation breaks down so that bargaining behaviour becomes dominated by considerations of short-run loss or gain. Conversely, if long-term self-interest or genuine altruism leads unions to display a modest degree of concern for others, the bargaining structure may lose much of its importance as a causal factor in macro-economic performance. Of course, beliefs and values of trade union negotiators are not created in a vacuum : they reflect the beliefs and values of society at large, and respond, more specifically, to the presence or absence of a willingness to cooperate creatively on the part of employers’ organizations. More generally, beliefs cannot be treated as exogenous; they are influenced by social experience, and in this sense partially endogenous in the social process.
The notion of "willingness to redistribute", as a constraint for successful employment policies, needs yet another qualification. The reality of social stratification in modern societies does not allow us to conceive of redistribution as a transfer from the "top" to the "bottom" only. Thompson (1997) describes the result of conservative economic policies and other changes embedded in the Anglo-Saxon economies, as the "30-50-20 percent" economy. Thirty percent or so of the population is confined to a bottom layer made up of those either permanently unemployed, or in low wage, low skill, part-time employment, or at the margins of employment with uncertain job prospects. On the other hand, there are the top twenty percent whose incomes have risen dramatically. In the middle stand the fifty percent who constitute the main contributors to aggregate demand. They still maintain good job prospects, and have more or less successfully adjusted to the newly emergent skill demands of the economy. The political problem, according to Thompson (and many others) is the growing tax averseness of the middle 50%, which makes flexible demand management of a Keynesian type and redistribution difficult. What is needed, according to Thompson, is a new "social settlement" for national solidarity. "The problem is to forge a ‘consensus’ amongst the 30-50-20 society, one in which the centrally important middle 50% get something out of that settlement as well as contributing to it." (Thompson, 1997, p. 170.) I do not believe that the nature of this challenge is illuminated by a dogmatic opposition of "universalist" versus "selective" principles for welfare. The fundamental challenge is to obtain majority support for an overall distribution of benefits (in cash and in kind) and burdens (social security contributions and taxes) which is considered as "fair".
People’s willingness to redistribute resources has to be underpinned by principles of reciprocity concerning the balance between rights and obligations within employment programmes and in society at large, and by administrative efficiency in the implementation of targeted employment programmes and social policies in general. Although reciprocity is a key value in egalitarian justice, I will not elaborate on it here. Instead, I will return to the main issue of this paper.
4.3.2. The impact of "globalisation"
The question now is whether there is anything in the nature of "globalisation" which has made the conditions for success discussed in the preceding paragraphs - willingness to accept some income discipline, willingness to redistribute, underpinned by reciprocity and efficiency in implementation - impossible to fulfil (recall Gray’s reference to an "insoluble contradiction", quoted in the Introduction). I can think of sociological developments, linked to developments which are in turn linked to the internationalization as we described it in Sections 2 and 3, which make it harder to fulfil those conditions. But I cannot see any reasons why the facts described in the preceding sections make it impossible to overcome the domestic constraints.
Let me give one example of a development which perhaps has made it harder to redistribute incomes and coordinate wage bargaining :
* To the extent that skill has become more important as an explanatory factor of quite visible wage inequalities, such inequalities come to have a more "biographical" character : they seem to be more related to personal history and qualifications than to class as traditionally understood (Giddens, 1994 makes this point, on pp. 143-144; it certainly is a compelling intuition; cfr. Rosanvallon, 1995, p. 209; Van der Zwan, 1996, pp. 263-264).
This has two consequences, one on the level of prudential reasoning, the other on the level of dominant conceptions of justice.
On the level of prudential reasoning, skilled people may become less inclined to pay for a collective insurance system, e.g. with regard to unemployment, because their risk of being unemployed is visibly less than the risk of the low-skilled. Atkinson makes a similar point with regard to unemployment insurance and the risk of getting unemployed in general, without reference to levels of skill : "The idea that voters determine their policy choices behind a genuine ‘veil of ignorance’ (...) may provide an explanation as to why political support for unemployment compensation declined with the onset of recession. During the years of full employment of the 1950s and 1960s, support for the Welfare State persisted since, when unemployment was low, people remained uncertain whether they would be affected if we returned to unemployment of the level of the 1930s. By the time that the rise in unemployment in the 1980s had levelled off, however, people had a much better idea as to whether or not they were likely to be at risk and what was the probability of finding another job. The veil had been lifted. The majority found that they were not at risk, and they ceased to give as much weight to the risk of unemployment in their objective function." (Atkinson, 1997, p. 317).
On the level of conceptions of justice, the increasingly biographical nature of wage inequalities may have reinforced "desert" notions of inequality, whereby acquired skill is considered to be completely deserved. This is the point made by Van der Zwan. Reich (1991, pp. 291-292) covers more or less the same ground when he strongly criticises the notion of "just desert" with regard to the high-skilled "symbolic-analysts" who are getting increasingly rich.
All this sounds plausible, but remains highly speculative. The yearly British survey of Social Attitudes does not indicate that it is true, on the contrary. The proportion of respondents who say that the gap between high and low incomes is too large has been growing systematically between 1983 and 1995. Moreover, in the 1995 responses there is remarkable unanimity across all social groups that the gap between high and low incomes is too large. However the numbers are ‘cut’ - whether they are displayed by economic activtiy, social group or educational attainment - a similar percentage of respondents say that the gap is too large. Yet this remarkable consensus has not always existed. Earlier in the survey’s history, those in higher income groups or in social groups I and II were substantially less likely than average to say that the income gap was too large (Jowell et al., 1996, p. 86). Bryson (1997) discusses fluctuations in public support for welfare in the UK. Although the unemployed are more pro-welfare than those in work, one of the fixed points is that highly educated people are more pro-welfare than those with no qualification. The time series are presumably too short to make a definitive judgment, but, clearly, the hypothesis that the impact of "globalisation" destroys solidarity is not supported here.
* One might conjecture that the same "biographical" shift in wage inequality has helped in undermining the legitimacy of centralized, corporatist bargaining, although I surmise that the decline of traditional manufacturing industries and the rise of small and medium enterprises are more important explanatory factors in that respect - if one searches for explanations linked to "globalisation". Similarly, export-led restructuring in industry can create strains in established collective bargaining structures.
This leaves us with some speculation and a research programme on reasons why it may have become harder to redistribute resources and to organize explicit incomes policies in a "globalized" world (if that is the case), but it does not provide us with any argument to say that redistribution and some form of concerted incomes discipline have become impossible.
Rodrik notes that "the Netherlands, Austria, Norway, Switzerland are all small countries that are, if anything, more exposed to the forces of the global economy than the UK is. If globalisation were a significant counterweight to egalitarian strategies, we would surely expect to see the consequences in these countries. To be sure, each one has had to grapple with the consequences of openness, and in some cases - the Netherlands being the chief example - significant reforms in labor markets and welfare systems had to be undertaken. But the important point is that each of these countries has managed to retain its distributive social institutions. For all its reforms, no one would confuse the Netherlands for the US or even the UK." (Rodrik, 1997b, pp. 15-16).
On the basis of the facts discernable in the real world, I find it hard to substantiate the claims made by Giddens on the demise of social democracy (see the quotes in the introduction of the paper published by IPPR). First, I believe Giddens attributes to Keynesianism "in the golden era" a cybernetic efficiency which it never had - whatever the degree of globalisation or "reflexivity". Second, I see no clear link between his notions of "intensified social reflexivity" or "the post-traditional order" and the problems which beset economic policy in the 1980s and the 1990s. Of course, if these Giddensian notions boil down, in the context of economic policy, to "lack of social cohesion among workers", or "lack of cohesion within trade unions", or "lack of cohesion within employers’ organizations"... as explanatory factors for the increasing difficulties of solidaristic wage restraint, then we have a clear link and a clear theory. But in that case Giddens’s terminology obscures rather than clarifies the problems of social democracy (whatever its possible merits in another intellectual context). It obscures the political and moral choices that condition the future of social democracy.
Other aspects of Giddens’s account of the demise of Keynesian social regulation seem to me too vague to be assessed empirically in any meaningful way. Take, for instance, the idea that "Keynesianism, like some forms of policy which helped structure the welfare state, presumes a citizenry with more stable lifestyle habits than are characteristic of a globalized universe of high reflexivity." (Giddens, 1994, p. 42). Two aspects of this idea have to be distinguished. There is an evident link between welfare policy and changes in lifestyles, discussed at length in the literature on welfare reform. The link, however, between the success of Keynesianism and "stable lifestyle habits", is very hard to grasp as long as criteria for "stability" of lifestyles are lacking. As the argument stands here, one could as well argue that Keynesian demand expansion was successful in the golden era of social democracy, precisely because Western Europe went through a period of rapid and profound change in lifestyles. The heyday of Keynesianism in Europe was characterized by the acquisition of all sorts of new consumer durables for domestic use and cars, which revolutionized households, and, according to many scholars of the golden era, thus sustained a virtuous circle of demand as long as those particular markets were not saturated. Rostow (1978, pp. 260-267) has documented the unique surge in the leading sectors fueling postwar growth in Western Europe (passenger cars, television receivers, and the sectors with which these were linked). Consumption patterns and life-styles already developed in the interwar and immediate postwar years in the US, diffused rapidly in Western Europe and Japan. Rostow shows how macro-economic policies were successful in terms of economic growth, partly because of rapidly changing consumption patterns. The change in lifestyle in Europe came not only briskly; it was certainly inspired by something "global", i.e. mass consumption of consumer durables as it had developed earlier in the United States. I do not want to settle the argument here; I want to indicate that Giddens’s thesis, as it stands, is too vague to be tested by historical evidence.
The preceding section focused on constraints on governments. But what can we say about the values that should guide their action ?
In the introduction to this survey I put an egalitarian conception of employment policy at the heart of welfare reform. In Section 4 I concluded that successful egalitarian employment policies depend, amongst other conditions, on the willingness to redistribute resources in society. I did not propose a precise yardstick for redistribution : I did not say, for instance, how the unfairness or fairness of an actual distribution of resources has to be measured. However, it is clear that the analysis in Section 4 calls for an overall conception of fairness in the distribution of burdens and benefits in society. Inevitably, such a conception has to refer to some specific notion of equality. More precisely, it has to specify those dimensions of people’s lives in which we think equality is a worthwile political goal. (This idea is discussed, form conflicting perspectives, in Franklin, 1997).
One of the most surprising aspects of some of the political literature on the future of social democracy which uses the concept of globalisation, is the idea that a new context has emerged making this endeavour pointless and misconceived from the start. Gray maintains that no "overarching principle of distribution" is plausible, and opts for "complex fairness". He thus implies that "social democratic egalitarian goals" are not only unattainable but also essentially misguided (Gray, 1996, p. 45-46). There is no ground for the latter conclusion, on the contrary. Rather than signalling the end of egalitarian values, the empirical literature on the impact of global trade signals their utmost relevance - as do most of the studies examining the causes of inequality over the last two decades. The real social phenomena we, rightly or wrongly, associate with globalisation not only resuscitate and justify calls for distributive fairness; with regard to foundational values they call for precisely those concepts that were developed in modern egalitarian philosophy.
5.1. Equality, choice and responsibility
If fortunes are now changing in a skill-driven world, to use Wood’s expression, some of the abstract problems and models which have been discussed in the framework of egalitarian philosophy over the last 20 years are becoming even more relevant today than when these discussions started. Just as with the Hecksher-Ohlin models of trade, one could say : these are textbook discussions whose time has come.
LeGrand captures a core idea of modern egalitarian philosophy as follows : "(O)ur judgements concerning the degree of inequity inherent in a given distribution depend on the extent to which we see that distribution as the outcome of individual choice. If one individual receives less than another owing to her own choice, then the disparity is not considered inequitable; if it arises for reasons beyond her control, then it is inequitable." (LeGrand, 1991, p. 87). More formally : "(A) distribution is equitable if it is the outcome of informed individuals choosing over equal choice sets." (idem). Equality, so conceived, is not uniformity, for instance uniformity of income, independent of people’s personal choices and personal effort (for the simple reason that this would constitute inequality in social advantage). It is however more demanding than "equality of opportunity" as the latter is conventionally used : "Individuals’ choice sets are determined not only by the social and individual barriers they face but by their initial resources or endowments, which include their natural abilities and the resources that they acquire through inheritance, gifts, family background, education prior to the age of majority, etc. Equalization of choice sets thus may require judicious manipulation of economic and other barriers in order to advantage the less well endowed. Or it may require compensating those with little natural ability by other resources, such as education, so as to bring their range of choices as close as possible to those naturally endowed." (LeGrand, 1991, pp. 91-92). One may add to the last example the possibility that people who perform poorly, independent of their will, on the labour market, have to be compensated financially via redistributive taxation (or wage subsidies or mechanisms like EITC). Personal responsibility, based upon choice, is thus a key concept in modern egalitarian philosophy.
Personal skills are the combined result of individual talents and effort to develop those talents. Hence, the market value of your skills is the combined result of luck and choice : (i) luck, first of all with regard to your original talents, as determined by the genetic endowments with which you are born and your early education, during which you did not make choices yourself; and secondly, luck with regard to the market for your skills, i.e. the interaction between the demand for your skills, which is influenced by other people’s preferences, and the competing supply of the same skills by other workers; (ii) personal choice, with regard to the kind of skills you choose to develop on the basis of your talents, and the effort you put into it. This is the domain characteristic of modern egalitarian philosophy. That these issues are intrinsically difficult does not mean that they can be assumed away. Political discussions about the relevance of personal skills and about the extent to which individuals are responsible for the skills they develop and the position they consequently have on the labour market, are an extension of the fundamental ethical discussion about "talent", "choice" and "just desert" which lies at the heart of egalitarian justice.
Though heavily influenced by social class, "skill", once acquired, is a strictly individual characteristic. It is indeed "biographical" in nature. There is, moreover, a large variety of marketable skills, and a continuum of market prices for skills, prices which, according to the empirical literature, are increasingly diverging. Hence, skills are cross-cutting social classes as they were traditionally perceived by social democracy. This social fact, which seems to be growing in importance, reinforces the case for a definition of social democracy in terms of values, and, more precisely, in terms of the values that govern relations and interaction between people. Cohen (1995, Chapter 6) gives normative arguments why socialists should be concerned with "exactly what a commitment to equality requires, and with exactly what sort of obligations productive and talented people have to people who are relatively unproductive, or handicapped, or in special need". But he also explains the actual occurrence of a shift of attention in former Marxist circles to this normative question by "profound changes in the class structure of Western capitalist societies, changes which raise normative problems which did not exist before, or, rather, which previously had little political significance." (Cohen, 1995, pp. 144-145). One of the changes he stresses is the fact that the working class cannot be identified with the needy people in society. This then forces a choice between the principle of self-ownership of labour power, hence of skills and talents, embedded in the traditional socialist doctrine of exploitation, and a principle of equality of benefits and burdens which negates that self-ownership principle. What the recent empirical literature on inequality adds to this polar problem of "needy" versus "productive", is the fact that within the working class there is a continuum of people, ranging from unskilled to very highly skilled, confronted with increasingly unequal market prices for their skills.
If it is true that this large variety of skills and associated market incomes cutting across the working class undermines class-based feelings of community and erodes political support for egalitarian ideas, then this is not a reason to abandon the moral concept of an egalitarian community. However, that is precisely the type of reasoning Gray’s rejection of social democracy implies (cfr. Gray, 1996, pp. 32-33) : it conflates problems of constraints with problems of values.
5.5. Values, constraints and justified pragmatism
Political action requires values and articulate thinking about values. In the preceding section I summarily indicated a modern and coherent conception of equality, which can be derived from a deep-felt value such as "equal concern for all members of society". Clearly, that conception is demanding : it calls for the prevention, correction or compensation of any and all "brute luck inequality" arising independently of people’s choices. Economic, social and cultural constraints make that unattainable in today’s society. A pragmatic strategy may aim at a threshold version of equality, as White argues : what matters most urgently from the moral point of view is that everyone has an equal chance of leading a decent, minimally fulfilling life, and, accordingly, that we prevent, correct or compensate for brute luck disadvantage in strategic goods, insofar as this pushes people below certain decency thresholds in their prospects for well-being and agency (White, 1997, p. 62). Constraints justify pragmatism; but we should be clear about the nature of the constraints, lest confusion and lack of direction result. The survey of the literature in this paper suggests that shifts in social and cultural norms with regard to fairness and reward for work and with regard to the actual functioning of labour market institutions can be as "constraining" for social-democratic ambitions as economic shifts in demand for labour, induced by technology or trade. Appropriate responses to the former type of constraint cannot be the same as responses to the latter type of constraint. Political movements should not overestimate their capacity to influence social norms (witness the sobering findings with regard to the values of "Thatcher’s children" in Jowell et al., 1997). But they should not underestimate the repercussions of their actions and their arguments on social norms in the longer-term. Social pragmatism has to be a conscious strategy, based upon hard, empirically validated facts and on coherent thinking about values and constraints.
ATKINSON Anthony B. (1990), "Income Maintenance for the Unemployed in Britain and the Response to High Unemployment", Ethics 100, April, pp. 569-585.
ATKINSON Anthony B., RAINWATER Lee and SMEEDING Timothy M. (1995), Income Distribution in OECD Countries. Evidence from the Luxemburg Income Study, OECD, Paris.
ATKINSON Anthony B. (1997), "Bringing Income Distribution in from the Cold", The Economic Journal, 107, March, pp. 297-321.
BAIROCH Paul (1996), "Globalisation Myths and Realities: One Century of External Trade and Foreign Investment", in BOYER and DRACHE (1996), pp. 173-192.
BELL Brian D. and PITT Michael K. (1995), Trade Union Decline and the Distribution of Wages in the UK : Evidence from Kernel Density Estimation, Nuffield College Discussion Paper N° 107, November.
BERMAN Eli, BOUND John and GRILICHES Zvi (1994), "Changes in the Demand for Skilled Labor within US Manufacturing: Evidence from Annual Survey of Manufacturers", Quarterly Journal of Economics 109, N° 2, May, pp. 367-397.
BLAIR Tony (1997), Speech to the Party of European Socialists' Congress, Malmö, 6 June.
BLAU Francine and KAHN Lawrence (1996), "International Differences in Male Wage Inequality : Institutions versus Market Forces", Journal of Political Economy, Vol. 104, N° 4, August, pp. 791-837.
BORJAS George J., FREEMAN Richard B. and KATZ Lawrence F. (1997), "How Much Do Immigration and Trade Affect Labour Market Outcomes", Brookings Papers on Economic Activity, 1, pp. 1-90.
BOUND John and JOHNSON George (1992), "Changes in the Structure of Wages in the 1980s: An Evaluation of Alternative Explanations", American Economic Review, Vol. 82, N° 3 June, pp. 371-392.
BOYER Robert and DRACHE Daniel (eds.) (1996), States Against Markets. The limits of globalisation, Routledge, London.
BRYSON Caroline (1997), "Benefit claimants: villains or victims ?", in JOWELL Roger, CURTICE John, PARK Alison, BROOK Lindsay, THOMSON Katarina and BRYSON Caroline (eds.), British Social Attitudes. The 14th report. The end of Conservative values ?, Ashgate, Aldershot, pp. 73-88.
BURTLESS Gary, "International Trade and the Rise in Earnings Inequality" (1995), Journal of Economic Literature, Vol. XXXIII, June, pp. 800-816.
CARLIN Wendy and SOSKICE David (1990), Macroeconomics and the Wage Bargain. A Modern Approach to Employment, Inflation and the Exchange Rate, Oxford University Press, Oxford.
CLINE William R. (1997), Trade and Income Distribution, Institute for International Economics, Washington.
COHEN Gerald A. (1995), Self-ownership, freedom and equality, Cambridge University Press, Cambridge.
COMMISSION ON SOCIAL JUSTICE (1994), Social Justice. Strategies for National Renewal, Vintage, London.
CORRY Dan (1997), Should we continue to work with the de-regulated labour market ?", Renewal, Vol 5, n° 1, Febr. 1997, pp. 35-49.
CORRY Dan and HOLTHAM Gerald (1995/6), Growth with Stability. Progressive macroeconomic policy, IPPR, London.
CROOK Clive (1997), "The Future of the State", Supplement to The Economist, 20 September.
DE BEER Paul (1997), "Het wonderlijke poldermodel", Socialisme & Democratie, Vol. 54, N° 9, pp. 375-382.
DELORS Jacques (1997a), Réflexions et propositions pour un nouveau modèle de développement, May, mimeo.
DELORS Jacques (1997b), Pacte pour la coordination des politiques économiques, Groupement d’Etudes et de Recherches "Notre Europe", August 6.
DELORS Jacques (1998), Interview by GRANT Charles "The former Commission president remains supremely confident of the EU's future : economic union, political democracy and twice the size.", New Statesman, 2 January, pp. 14-16.
DESJONQUERES Thibaut, MACHIN Stephen and VAN REENEN John (1997), Another Nail in the Coffin ? Or Can the Trade Based Explanation of Changing Skill Structures Be Resurrected ?, paper presented at the LOWER Conference, 12-13 December, London.
DE SWERT Gilbert and JANSSEN Ronald (1996), "De Hollandse Kwiekte", Gids op Maatschappelijk Gebied, 1996, Vol. 87, N° 12, pp. 1027-1038.
DILNOT Andrew (1997), "Crisis in Welfare", in Peter ASKONAS and Stephen F. FROWEN (eds), Welfare and Values. Challenging the culture of Unconcern., Macmillan, London.
DINARDO John, FORTIN Nicole M. and LEMIEUX Thomas (1995), Labour Market Institutions and the Distribution of Wages 1973-1992 : A Semiparametric Approach, National Bureau of Economic Research Working Paper N° 5093, Cambridge Mass., April.
DINARDO John and LEMIEUX Thomas (1997), "Diverging Male Wage Inequality in the United States and Canada, 1981-1988 : Do Institutions Explain the Difference ?", Industrial and Labor Relations Review, Vol. 50, N° 4, July, pp. 629-651.
DIWAN Ishac and WALTON Michael (1997), "How International Exchange, Technology, and Institutions Affect Workers : An Introduction", The World Bank Economic Review, Volume 11, Number 1, January, pp. 1-15.
DUNN, John (1994), "Foreword" to "Contemporary Crisis of the Nation State ?", Political Studies, Volume 42, Special Issue.
FORTIN Nicole M. and LEMIEUX Thomas (1997), "Institutional Changes and Rising Wage Inequality: Is There a Linkage?", Journal of Economic Perspectives, Vol. 11, Number 2, Spring, pp. 75-96.
FRANZMEYER Fritz, LINDLAR Ludger and TRABOLD Harald (1996), Employment and Social Policies under International Constraints. A Study for the Ministerie van Sociale Zaken en Werkgelegenheid (SZW) of the Netherlands by the Deutsches Institut für Wirtschaftsforschung in Berlin, November.
FRANKLIN Jane (ed.) (1997), Equality, IPPR, London.
FREEMAN Richard B. (1995), "Are Your Wages Set in Beijing ?", Journal of Economic Perspectives, Volume 9, Number 3, Summer, pp. 15-32.
GARRETT Geoffrey (1995), "Capital mobility, trade, and the domestic politics of economic policy", International Organization 49, 4, Autumn, pp. 657-87.
GARRETT Geoffrey and LANGE Peter (1995), "Internationalization, institutions, and political change", International Organization 49, 4, Autumn, pp. 627-55.
GARRETT Geoffrey (1997), "A Virtuous Global Circle", Boston Review, Volume XXII, Number 6, December 1997/January 1998, pp. 11-12.
GEELHOED L.A. (1997), "1997 : Een delta in Europa", ESB, 1 januari 1997, blz. 4-8.
GIDDENS Anthony (1990), The Consequences of Modernity, Polity Press, Cambridge.
GIDDENS Anthony (1994), Beyond Left and Right. The Future of Radical Politics, Polity Press, Cambridge.
GIDDENS Anthony (1995), "Government’s last gasp ?", The Observer, Sunday 9 July 1995, p. 25.
GIDDENS Anthony (1997), "Dialogue with G. Soros", New Statesman, 31 October, pp. 24-26.
GLYN Andrew (1995a), "Social Democracy and Full Employment", New Left Review, May-June 1995, pp. 13-55.
GLYN Andrew (1995b), "The Assessment : Unemployment and Inequality", Oxford Review of Economic Policy, Vol. 11, N° 1, Spring, pp. 1-25.
GLYN Andrew (1997a), "No escape from tax", New Economy, Summer, Vol. 4, N° 2, pp. 74-77.
GLYN Andrew (1997b), "Egalitarianism in a Global Economy", Boston Review, Volume XXII, Number 6, December 1997/January 1998, pp. 4-8. (An earlier version of this paper has been published as "Internal and External Constraints on Egalitarian Policies in Europe" in Politik und Gesellschaft, 2/1997.)
GLYN Andrew and SUTCLIFFE Bob (1992), "Global but Leaderless : The New Capitalist Order", R. MILIBAND (ed), The Socialist Register 1992, pp. 76-95.
GOTTSCHALK Peter (1997), "Inequality, Income Growth, and Mobility: The Basic Facts",Journal of Economic Perspectives, Vol. 11, Number 2, Spring, pp. 21-40.
GOTTSCHALK Peter and SMEEDING Timothy M. (1997), "Cross-National Comparisons of Earnings and Income Inequality", Journal of Economic Literature, Vol. XXXV, June, pp. 633-687.
GOWAN Peter and ANDERSON Perry (eds.) (1997), The Question of Europe, Verso, London.
GRAY John (1996), After Social Democracy : Politics, capitalism and the common life, Demos, London.
HEYLEN Freddy (1992), Labour Market Chararcteristics and Macroeconomic Performance, Doctoral dissertation, Faculteit Toegepaste Economische Wetenschappen, UFSIA/Universiteit Antwerpen.
HILLS John (ed.) (1996), New Inequalities, Cambridge University Press, Cambridge.
HIRST Paul and THOMPSON Graham (1996), Globalisation in Question, Polity Press, Cambridge.
HIRST Paul (1997), "The global economy - myths and realities", International Affairs, 73, 3 pp. 409-425.
JOWELL Roger, CURTICE John, PARK Alison, BROOK Lindsay and THOMSON Katarina (1996), British Social Attitudes. The 13th Report, Darmouth Publishing Company Ltd, Aldershot.
JOWELL Roger, CURTICE John, PARK Alison, BROOK Lindsay, THOMSON Katarina and BRYSON Caroline (1997), British Social Attitudes. The 14th report. The end of Conservative values ?, Ashgate, Aldershot.
KINDLEBERGER Charles (1973), The World in Depression 1929-39, Allen Lane, London.
KRUGMAN Paul (1994), "Does Third World Growth Hurt First World Prosperity ?", Harvard Business Review, July-August, pp. 113-121.
KRUGMAN Paul (1995), "Growing World Trade : Causes and Consequences", Brookings Papers on Economic Activity, 1, pp. 327-377.
KRUGMAN Paul (1996), "First, Do No Harm", Foreign Affairs, Volume 75, N° 4, July-August, pp. 164-170.
KRUGMAN Paul and LAWRENCE Robert (1993), Trade, Jobs, and Wages, National Bureau of Economic Research Working Paper N° 4478, September
KRUGMAN Paul and VENABLES Anthony J. (1995), Globalisation and the Inequality of Nations, National Bureau of Economic Research Working Paper N° 5098, Cambridge Mass., April.
LAWRENCE Robert Z. and SLAUGHTER Matthew J. (1993), "Trade and US Wages: Great Sucking Sound or Small Hiccup ?", Brookings Papers on Economic Activity, 2, pp. 161-226.
LE GRAND Julian (1991), Equity and Choice. An Essay in Economics and Applied Philosophy, Harper Collins Academic, London.
MADDISON Angus (1991), Dynamic Forces in Capitalist Development, Oxford University Press.
MADDISON Angus (1995), M