Abstract

Much of the current interest in processes of restructuring evolves around issues of changing occupational and social structures. While there is broad agreement that economic restructuring has produced a new occupational structure in the Western World, there is a great deal of controversy characterizing the changing structure. This is to be expected since we leave behind us a well-known social structure and are entering another which is still in the process of being shaped.

It is suggested in this article that if one is to comprehend the complexity involved when discussing these issues a fruitful starting point would be different welfare state regimes. Furthermore, it is suggested that different types of welfare states produce different occupational cupationalstructures and to varying degrees, and for various reasons, do they lead to social polarisation. Thus, differences within welfare state regimes can explain why we are experiencing divergent trajectories in employment and social stratification.

Keywords

Polarisation, inequality, welfare states, economic and social restructuring.

Helle Nørgaard: Institute of Geography, University of Copenhagen,
Øster Voldgade 10, 1350 Copenhagen K., Denmark.

Danish Journal of Geography 96: 103-109, 1996.

The Question in Case

What seemed conventional wisdom about contemporary social change in international Western cities namely that polarisation of the occupational and income structures is growing and leading to increased social polarisation (Sassen, 1991) has recently been questioned (Hamnett, 1994a & 1994b). Hamnett (1994a & 1994b) challenges the socalled 'global city thesis' presented by Sassen (1991), in which she claims that the processes of economic change are leading to growing polarisation of the occupational and income structures as well as increased social polarisation. Sassen (1991) finds that this is particularly the case in the global cities of New York, London and Tokyo but extends the thesis to other cities as well. Rather, Hamnett (1994a & 1994b) puts forward the notion that professionalisation in both London and Randstad appears to be dominant, for which reason he also rejects the notion of increased social polarisation.

In this regard the key issues are: (1) whether there is a growing polarisation of the occupational and income structures, and (2) whether social polarisation is increasing. Consequently, the focal point of the debate is the division of labor. While concentrating on the city scale, the debate is on changes in the distribution of the population of a city. Marcuse (1989) describes this clearly:

"The best image...is perhaps that of the egg and the hour glass: the population of a city is normally distributed like an egg, widest in the middle and tapering off at both ends; when it becomes polarized the middle is squeezed and the ends expand till it looks like an hour glass. The middle of the egg may be defined as intermediate social strata" (p. 699).

The question in case is then whether the processes of economic change have led to increased occupational and income polarisation (the hour glass) or rather to professionalisation (the egg). In order to answer this question a broad approach is suggested.

A Broad Approach

First of all, it should be made clear that profound economic
changes have structurally changed the economy and the

labor markets in all cities in the Western world. Cities on both sides of the Atlantic have to varying degrees undergonea process of deindustrialisation since the 19505, which has lowered employment in manufacturing considerably.This has in all cases been accompanied by a growth of the service sector. Thus, most cities have transformedfrom a manufacturing to a service-oriented economy.But the way in which the restructuring has altered occupational structures has been quite different and to a large degree dependent on different welfare state regimes.

Different Welfare States Regimes

The variety of institutional and ideological arrangements is represented by different types of welfare states seeking to realize differing social outcomes. Types of welfare states range from 'strong interventionist' to market-oriented' states but are, in all cases, exposed to similar economic, political and social pressures. While the processes of restructuring have broken down previously existing links between social balance and economic growth in all welfare states, what is considered acceptable living conditions does vary. Thus, the characteristics of the different welfare state models are most relevant to understanding the variations in social and socio-spatial polarisation. With reference to Pierson (1991) these fall into three categories: the Liberal (US, Canada and Australia), the Conservative (France, Germany, Italy) and the Social Democratic (Sweden, Norway) welfare states. One could also follow the categories of Therborn (1987) operating with The Strong Interventionist (Sweden, Norway), the Soft Compensatory (Belgium, Denmark, Netherlands), the Full Employment oriented (Switzerland, Japan) and the Market-oriented (Australia, Canada, US and UK) welfare states. Whichever category one chooses the point is that they indicate differences in ideology and institutional arrangements within different types of welfare states. In the following section some differences in relation to occupational structures and social polarisation will be pointed out.

Factors Influencing the Occupational Structure

Some of the reasons for differing outcomes can be explained
by the following. Since it is a rather complicated task to
determine exact relationships between these elements, the

following list merely suggests issues which, directly or indirectly,
impact on both occupational and social structures.

Some important elements are:

• the degree of economic restructuring and prior occupational

• redistributive programs,
• unions and minimum wages,

• the role of immigrants and established ethnic minorities,
and

• the constellation of political interests and actions of the
participants.

The degree of economic restructuring and prior occupational direction will influence current and future occupational structures. For cities which had their main economic base in manufacturing the impact of economic restructuring and the shift to services have been far more severe than those with a tradition in trade, services, etc. Both London and New York have been globalised and have experienced a transition of their economic base to finance and real-estate. Their loss of employment in manufacturing has been quite similar with losses of more than 50% over a 20 year period from the late 1960s to late 1990s (Fainstein et al., 1992, Markusen & Gwiasda, 1994, and Sassen, 1991.) Contrary to this "the process of economic restructuring has been slower in the Netherlands than elsewhere, partly because the economy of the Randstad has always been more oriented towards services and trade rather than the industrial sector" (Hamnett, 1994a, p. 415).

Although redistributive programs only "ameliorate some of the worst consequences of the inequalities in our societies" (Marcuse, 1989, p. 701) Social Democratic/strong interventionist welfare states do assure greater economic equality than Liberal/market-oriented welfare states. The former being a universal welfare system, everyone has the right to social services including immigrants, guest-workers, etc. Thus, various services; unemployment benefits, schooling, training, housing subsidies etc. have been provided to those who have been displaced by deindustrialisation. This has rarely been the case in Liberal welfare states, which strongly impact occupational and income structures.

Unions have traditionally played a significant part in negotiatingwages in Social Democratic/strong interventionist welfare states. Consequently, the level of minimum wages and benefits in these nations is relatively high. While higher minimum wages narrow the gap between those at

the top and at the bottom of the occupational and income structures, higher minimum wages, in some cases, make automatisation of certain work tasks more profitable. The higher minimum wages can therefore lead to unemploymentand unemployment rates are in fact growing in Social Democratic welfare states such as Denmark and Sweden. And more so than in Liberal welfare states such as the United States, where employing low-wage workers is more profitable. While there may be an increase in the number of jobs in the United States because of the low wages, it is often necessary for, e.g. a formerly skilled manufacturing worker to hold two jobs in order to maintain the same living standard. Also, in the United States it is seemingly politically acceptable that large numbers of illegal immigrants work in low-wage jobs. In the residual welfare state holding a job is more of a necessity compared with the extended welfare state, where unemployment appearsmore politically acceptable and is thus being compensatedfor by unemployment benefits and social services.However, although such benefits may limit economicdifferences between employed and unemployed populations,they do not hinder social stigmatization. Thus, even extensive social benefits cannot eliminate social polarisation,exclusion, etc. Therefore, social divisions within the strong interventionist welfare state are to be found between those employed and those unemployed, whereas social divisions in the market-oriented welfare state are primarily to be found between those employed with high wages and low-skilled, low-wage/unemployed populations.

The steady and large influx of immigrants (legal and illegal) to the United States obviously has a significant impact on occupational and income structures. New York is a city with "extremely high and continuing levels of immigration and a large low-wage immigrant labour supply" (Sassen, 1984, p. 140). The immigration, although considerable, is much smaller in the Netherlands and even in the UK. Thus, the situation is, as stressed by Hamnett (1994a) "very different in scale from that in the US" (p. 422) and, according to Hamnett, has had no effect of labour markets and social polarisation. But increasing immigration and migration are to be expected with the elimination of national borders within the European Union which will present a real challenge to these welfare states.

As a final comment it should be emphasized that constellations of political interests and actions of the participants, to a great extent, determine specific economic and social outcomes. Indeed, several studies have shown that

the impact of the restructuring process is mediated, which is specific to national and local circumstances (Andersen & Jørgensen: 1994, Beauregard: 1989, Fainstein & Fainstein: 1986 & 1987, Fox-Przeworski et al: 1991, Judd & Parkinson: 1990, Logan & Swanstrom: 1990, Matthiessen & Andersson: 1993, Pickvance & Preteceille: 1991 and Stone & Sanders: 1987).

It is not possible in this article to conclude which factors most strongly influence occupational and income structures. It seems obvious, however, that processes of inclusion and exclusion in the labour market are a major determinant. But only a detailed study of each of the cities will make this clear. This paper is not in a position to differentiate occupational structures in the three cities. However, in using statistical data on changes in the occupational structure for New York City one finds a trend similar to the one shown by Hamnett (1994a & 1994b) for London and Randstad, namely that of professionalisation. This is hardly surprising since qualification requirements have increased in most occupational categories over the past twenty years or so. Thus, even working as a sales clerk requires some degree of skills and is therefore not available to everyone.

This data alone is evidence of how occupational categories, even within one nation, are changed from one decade to another. Still, it seems fair to conclude that the overall trend in occupational structures in New York City in the period from 1970-1990 is one of professionalisation.

Data on income distribution in New York City shows that the number of households with an income of less than $10,000 has decreased rapidly and that the number of households earning more than $35,000 has increased even more rapidly.

Obviously, one has to consider the differential values of
these incomes over the time period considered. The trend
is nevertheless one of increased incomes for more house-

holds. Thus, if one merely relies on statistical data on changes in occupation and income the trend for New York City, as for London and the Randstad, is one of professionalisation and rising incomes. But, the fact that the work-force has become relatively better educated and earning better incomes does not exclude social polarisation which is perhaps even rising. For social polarisation is not a measure of occupation and income alone.

First of all, since it is unclear what exactly is meant by social polarisation (Pahl, 1988 and Marcuse, 1989 & 1993), it becomes difficult to determine whether the pattern is one of growing polarisation. The central question is then how to define as well as measure social polarisation.

How to Measure Social Polarisation?

The first issue to consider is what exactly is meant by social polarisation. As Pahl (1988) points out, social polarisation can mean very different things to different people. Hamnett (1994a) also emphasizes that "social polarisation remains a most unclear and ill-defined con cept There is uncertainty over the units of measurement, the variables to be incorporated and the definition itself (p. 405). So much more unfortunate is it when Hamnett (1994a & 1994b) makes use of very narrow variables such as national statistical data on occupation and income to conclude on the extent to which social polarisation exists in both London and Randstad. And it is no better that Sassen (1991) makes her conclusions upon impressions. The phenomenon of polarisation cannot be measured from personal feelings nor can it be generalized through statistical data. In the following section the issues of how to measure social polarisation is considered. While definitions of social polarisation can become so broad that they become unmanageable, it is nevertheless crucial to consider how to best measure social polarisation.

Typically the evidence for the existence of social polarisation is found in statistics about occupation, income distribution, educational attainment, or some other indicator of prosperity, wealth or personal achievement. But what exactly does that tell? Statistical data can inform about changes in absolute numbers of various job categories and their relative changes over time. Thus, to determine whether the occupational pattern is one of professionalisation or polarisation a calculation of the absolute numbers of high, middle and low-strata jobs and whether they have increased in proportion of each other over a period of time is required. A relative increase of low and high-wage jobs and a decrease of middle-wage jobs suggests occupational and income polarisation whereas a relative increase of middle-wage jobs suggests professionalisation of the occupational structure. This implies a categorization of specific job types into respectively high-, middle- and low strata.

Since national statistical data on occupational and income distribution operate with different job categories and different definitions of these, it becomes problematic to compare. In the US for example, the use of 'manager' as a job title is widespread because (1) as a compensation for a higher salary, and (2) managers are not paid for working overtime. Large numbers of store clerks with limited

responsibilities thus carry the job title as 'manager', which has obvious implications for national statistical data on occupational structures (Fainstein, 1994). Also, since it is the choice of the person carrying out the study to determine whether a specific type of job should be grouped in one category or another (high, middle or low), the outcome may vary. And as Hamnett (1994a) points out "the occupational data for the Netherlands provided by the Labour Force Survey are very ambiguous and unsatisfactory for the purpose of occupational class analysis consisting as it does of a rather uneasy mixture of occupational and sectional categories" (p. 409). Another important aspect of using statistical data is the unit of measurement. As explained by Pahl (1988):

"an emphasis on the household as the salient social and economic unit, rather than the individual, has important implications for contemporary social structure. Whilst the occupational structure based on individual jobs typically produces a hierarchically structured pyramid of advantage, these new tendencies allow a sizable proportion of households incorporating low or modest incomes to a higher position in the hierarchy as a result of their collective households work practices" (Pahl, p. 252).

The accuracy of conclusions as regards occupational patterns, which are merely based on statistical data, are thus questionable. However, more serious problems arise when the 'evidenced' occupational and income patterns are considered the sole basis on which social polarisation is measured.

The reality is that social polarisation is not merely an extension of low-wage employment. Rather there is "an increasing tendency to polarisation along many dimensions in our societies" (Marcuse, 1989, p. 700). While social polarisation in the US is strongly related to low and decreasing wages of a growing number of low-skilled populations, in many cases, social polarisation has a different reasoning e.g. in Western Europe. Here social polarisation is more related to exclusion from the labour market than to differences between high- and low-waged populations. Thus, whereas social structures in the US is based on disparities between those at the top and those at the bottom of the occupational structure, the division in most of Western Europe is one between employed and unemployed populations.

Social polarisation is also related to informal or unrecordedwork.
Pahl (1988) raises questions about the role
and function of various forms of work outside employment

and how it may affect the system of stratification in society as a whole. He finds that such alternative work strategies have not compensated those who are forced or expelled from the labour market. Rather "the indications are now that households based on 'core' workers are permanently advantaged throughout the life-cycle and the cumulation of further component wages and involvement in all forms of work leads to increasing household privilege and advantage". Informal work "reinforces, rather than reduces or reflects, contemporary patterns of inequality". Thus "certain households are becoming increasingly more fortunate, whereas others are becoming increasingly more deprived" (Pahl, 1988, pp. 249-252). For this reason alone the tendency may, in some cases, be one of very strongly increasing polarisation which will not be revealed by statistics or elsewhere.

Still, personal resources are not simply an outcome of occupational (formal or informal) and income position. Cultural and social resources, e.g. lifestyle, value system, age, gender, ethnicity, are equally important. The individuals' collective resources, and constraints, should ideally form part of a model of social inequality and polarisation; a model following the various divisions in society: relations of production, consumption, race, income, ethnicity or color, gender, household composition, age and housing tenure (Marcuse, 1989). Furthermore, a model of changing class divisions should ask "is the new and growing middleclass/middle strata composed of elements of the old skilled working class and the new white-collar service class? How stable are the new, more privileged categories? Can they hand on their position to their children? What do the children inherit - property, cultural capital, training, certain attitudes to work and employ ment....or what?" (Pahl, 1988, p. 260).

Although most data indicate that certain households are becoming increasingly more fortunate, whereas others are becoming increasingly more deprived, it remains crucial to broaden narrow definitions and measurement of social polarisation, which merely stress changes in the distribution of income and wealth.

Concluding Remarks - is Social Polarisation on the Rise?

Based on the discussion above it is suggested that a definition
of social polarisation should take into account more

than just statistical data on occupational structures and visual impressions. But how exactly to define social polarisation and determine how to measure it is a task which lays ahead. But these issues are of such importance that they cannot be left to simplistic methods of measurement and evaluation.

Although there will be no answer in this article to the question of rising social polarisation in New York, London or Randstad it has hopefully made some contribution in suggesting a method of analysis on the issues of changing occupational and social structures. One can only regret that on the whole there have been relatively few attempts "to bring together the mass of separately conceived and executed social, political, economic and other studies of London and New York in order to systematically explore the causes and consequences of economic restructuring and their urban impacts" (Fainstein et al., 1993, p. 13). Instead it is suggested in this article that in order to answer these important questions comprehensively, an extensive study of each of the cities is required. A study that should take specific welfare state regimes as its point of departure.

Rather than questioning whether and to what extent certain patterns prove true in particular cities, in this article emphasis has been on the underlying forces creating distinct patterns. Thus, in addressing general patterns of urban development it is suggested that trends of occupational and income polarisation can co-exist with trends of professionalisation. It is proposed that divergence in occupational and income structures is largely an outcome of the cities' particular history, their economic base and social and political composition which is reflected in different types of welfare states. Thus, it can be concluded that welfare states do have a direct causal impact on how employment structures and, as a result, how new axes of social conflict, evolve.

References

Andersen, H. T. & J. Jørgensen (1994): Urban Competition and Regeneration Strategies - Examined Inside Out. The Danish Ministry of the Environment, Spatial Planning Department: Spatial Development in Europe. Research Colloquium 17-18 March, 1994. Copenhagen.

Beauregard, R.(ed.) (1989): Economic Restructuring and
Political Response. Urban Affairs Annual Reviews, 34.

Census Tracts: New York, N.Y. Census of Population and
Housing 1970. Standard Metropolitan Area.

City of New York: Socio-economic Profile 1980. A Portrait of
New York City. From the 1980 Census. Department of City
Planning.

City of New York: Socio-economic Profile 1990. A Portrait of
New York City. From the 1990 Census. Department of City
Planning.

Fainstein, S. et al. (eds.) (1992): Divided Cities. New York &
London in the Contemporary World. Blackwell. Oxford, UK
& Cambridge, USA.

Fainstein, S. (1994): Conversations at Rutgers University,
School of Planning and Public Policy. December, 1994.

Fainstein, N. & S. Fainstein (1987): Economic Restructuring and
the Politics of Land Use Planning in New York City. Journal
of the American Planning Association, 53, 2: 237-49.

Fainstein, S. & N. Fainstein (eds.) (1986): Restructuring the
City. The Political Economy of Urban Development. Longman.
New York and London.

Fox-Przeworski, J. etal. (eds.) (1991): Urban Regeneration in a
Changing Economy. Claredon Press, Oxford.

Hamnett, C. (1994a): Social Polarisation in Global Cities:
Theory and Evidence. Urban Studies, 31,3: 401-24.

Hamnett, C. (1994b): Socio-economic Change in London: Professionalization
not Polarization. Built Environment, 2, 3:
192-203.

Judd, D. og M. Parkinson (eds.) (1990): Leadership and Urban
Regeneration. Urban Affairs Annual Reviews, 37.

Logan, J. & T. Swanstrom (eds.) (1990): Beyond City Limits.
Urban Policy and Economic Restructuring in Comparative
Perspective. Temple University Press, Philadelphia.

Marcuse, P. (1989): Dual City: a Muddy Metaphor for a
Quartered City. International Journal of Urban and Regional
Research, 14, 4: 697-708.

Markusen, A. & V. Gwiasda (1994): Multipolarity and the Layering of Functions in World Cities: New York City's Struggle to Stay on Top. International Journal of Urban and Regional Research, 18, 2: 167-94.

Matthiessen, C. W. & Å.E. Andersson (1993): Øresundsregionen.
Kreativitet. Integration. Vækst. Munksgaard. København.

Musterd, S. (ed) (1994): A Rising European Underclass? Built
Environment. 20, 3. Alexandrine Press. Oxford, UK.

Pahl, R.E. (1988): Some Remarks on Informal Work, Social
Polarisation and the Social Structure. International Journal of
Urban and Regional Research, 12, 2: 247-68.

Pickvance, C. & E. Preteceille (eds.) (1991): State Restructuring
and Local Power. A Comparative Perspective. Pinter Publishers,
London and New York.

Pierson, C. (1991): Beyond the Welfare State? The New
Political Economy of Welfare. The Pennsylvania State University
Press, Pennsylvania.

Sassen, S. (1991): The Global City. New York, London, Tokvo.
Princeton University Press. Princeton.

Sassen-Koob, S. (1984): The New Labor Demand in Global Cities in M. P. Smith (ed.) - Cities in Transformation. Urban Affairs Annual Reviews, 26: 139-71. Sage, Beverly Hills.

Stone, C. & H. Sanders (eds.) (1987): The Politics of Urban
Development. University Press of Kansas.

Therborn, G. (1987): Welfare States and Capitalist Markets,
Acta Sociologia, 30 (3/4): 237-54.