The research and writing regarding producer services have many purposes, with different processes under focus:  macroeconomic, microeconomic, territorial, and social.  Each purpose has led researchers to distinct ways of conceptualizing and categorizing the economic and social changes manifested in the growth of intermediate service activities.  Ultimately, each purpose derives from a desire to attribute the origins of wealth across the factors of land, labor, capital, and intellectual resources, or to understand the distribution of wealth among the specific owners of those resources, distinguishing them by corporate or social affiliation and by location.  This brief paper reviews these purposes, placing them in mutual context, and relates each purpose to a set of corresponding concepts and categorizations.  As a result, we are able to identify which categorizations are best suited for different purposes, and which categorizations and underlying concepts show promise for further development.  Some of these categorizations would move us beyond a subject of “producer services.”

 The table below encapsulates these relationships among categorization schemes and the processes that are our objects of research.  The cell entries present the rationale for using each scheme as a basis for conceptualization and empirical research into each of the processes.  This matrix provides an unusual perspective on research themes and questions.  In some cases, the cells suggest new ways of thinking about the economic and social changes subsumed under producer-service-industry growth.  Shaded cells indicate those combinations of categories and research objects that hold greatest promise for further work, given the process concerned.  In many cases, the underlying concepts point our attention away from the sector-or activity-driven categories of producer or intermediate services, toward other ways of viewing economic activity.
 
 

THE  ECONOMIC  PROJECT
 These paragraphs quickly review the study of services as motivated by economic concerns for the sources and distribution of wealth across factors and divisions of the economy.

Macroeconomic change
 Sectors.  Much of the original concern for defining a “services” sector grew out of the mere recognition that employment in agricultural, mining, and manufacturing establishments has become a steadily decreasing proportion of total employment in U.S., Canadian, and most Western European economies [Clark 1940;  Fisher 1935;  Greenfield 1966].  Therefore, sectors, defined according to the material relationships between inputs and outputs in a discrete production process, are the category underpinning this view of macroeconomic change.  Most scholars now agree that the linearity of this model of economic change is too limiting, and that the reliance on output sectors as proxies for productivity or export propensity is unwarranted.  Nonetheless, the rapid employment growth in non-manufacturing sectors has raised political and academic attention.  Studies of the reasons for this growth (e.g., Tschetter [1987]) often combine sectoral, functional, and occupational categories.

 A primary function and challenge of macroeconomic study is understanding the sources of economic growth.  This understanding has been alternatively based on:  intensification of returns from territorial resources;  increased productivity in the use of territorial resources;  and extra-territorial demand.  Industrial sectors are relevant to growth models to the extent that returns to the use of resources vary across sectors or that the propensity to export varies across sectors.  These issues have been among the ways that service sectors have proven controversial.  However, it is (or, at least, has become) counter-intuitive that sectoral categories should be the basis for productivity differentials or export propensities.  Thus, the importance of this categorization scheme has been reduced in understanding economic growth.

 Activities.  Some researchers have focused attention on activities rather than sectors as the categories that differ in their contribution to economic productivity.  There is increasing recognition that productivity in actual fabrication processes is very high, and has increased largely because of other, engineering and planning, activities.  While the activities surrounding actual production are sources of increased productivity in production, their own productivity has not increased at as rapid a pace, certainly not before the implementation of non-hierarchical, networked data access and manipulation [note 2].  Therefore, activities associated with production and consumption — design, finance, planning, logistics, marketing, and distribution — have become important categories of economic research.  When these activities are located in separate establishments or companies, they are measured as service sectors.  Regardless, however, these activities are important generators and determinants of productivity increase.

 Carnevale and Rose [1998: 8] developed a hybrid categorization of economic activities by aggregating work functions across industrial sectors.  Their five major categories are thus defined by activity, occupation, and sector, and are given names descriptive of the typical venue for the setting of the activity:

• extractive production (“the farm”) — direct labor in agriculture, mining, fishing, logging, and forestry;
• industrial production (“the factory”) — direct labor in manufacturing, public utilities, construction, and freight transport;
• low-skilled services (“the counter”) — direct labor in consumer services that require quickly-acquired training (personal services, retail services, catering);
• high-skilled services (“the hospital and classroom”) — direct labor in consumer services that require specialized training (health care, education, public safety);
• administration and coordination (“the office”) — all labor engaged in managing, administering, and coordinating, regardless of industrial sector [note 3].

 Networks.  Institutionalist comparisons of national economies often refer to the relative presence and tightness of inter-corporate networks of firms that supply each other with inputs, technology, capital or credit, and even management in repeated, reciprocal relationships.  Networks can be based on ownership, banking, familial, territorial, or ethnic commonalities among core members [Castells 1996].  The operation of networks has been credited with international (and even interregional) differences in rates of growth, degree of stability, and technology diffusion among firms in a region, nation, or world region.  Service activities — financial institutions, trading companies, information systems — are clearly a part of such networks [Coffey and Bailly 1991;  1992].  The ways in which service activities and service firms exist within and contribute to identifiable networks remain important areas of research.

Microeconomic outcomes
 What determines the level of profit seen by a particular business enterprise?  This is important for obvious reasons of asset allocation by lenders and equity investors, and from a macroeconomic concern about the rates of return to investment in the economy as a whole or in territorial subdivisions.

 Sectors.  To the extent that capital is mobile across industrial sectors, there is no reason to expect profit differentials across sectors:  sectors with high returns to invested capital should see increased capital investments.  However, rates of growth and return on investment do vary by sector, perhaps reflecting the presence of sunk costs in slower-growth sectors, or the increased risk in rapidly changing sectors.  Therefore, sectoral divisions matter.  Are there systematic differences in profit levels across service sectors?  Are higher performing services more likely to exist in specialized firms, rather than adding value within more integrated companies?

 Firms.  Profit rates vary across firms, and high profits are assumed to accrue to form-specific assets of location, technology, management, or reputation.  Reputation is especially important in sectors whose output is not immediately observable and comparable, such as in some services.  These bases for monopolistic competition, and thus the categorization of activities according to firms, make some substantive sense.  It is the individual firm, whatever the sector, that makes decisions to jettison or add products, functions, or markets in an attempt to focus its efforts on those activities through which it can derive maximum advantage from its core resources [Milgrom and Roberts 1990;  Quinn et al. 1990;  Roach 1991].  In this framework, the firm is the object of study, categorized according to the activities in which it engages.

 Strategic groups.  Competitive strategy entails seeking the greatest return for firm-specific assets.  Despite the specificity of the particular assets, this often leads firms to compete, not broadly across an entire product market, but most intensively with firms deploying their assets in similar ways.  Typical strategic options are focus or narrowness (in product line or in geographic markets) versus breadth or comprehensiveness, and cost-leadership versus quality leadership.  Corporate behavior and patterns of investment should vary by strategic group within a sector:  capital intensification, employment size and growth, market extent and growth, and capital allocation to vertical integration.  While adherence to any clearly defined competitive strategy brings the potential for real (or economic) profit in a model of monopolistic competition, profit rates and the use of capital may vary among groups [Porter 1980;  O’Farrell et al. 1993;  Lindahl and Beyers 1998].  Strategic groups may be composed of enterprises within or across sectors.

 Production chains.  The corollary of product, functional, or market focus by firms (or other organizations) is the use of production chains of organizations in the creation of value from conception, production, and marketing.  Unfortunately, the very flexibility that production chains or networks afford their participants make them very difficult to study empirically.
 

THE  SOCIAL  PROJECT
 Besides the economic-growth rationales for the study of service sectors and service activities within all sectors, there are social outcomes of the changing economy.  These can be encapsulated by the questions “Who benefits from the changing structure of the economy?”  and “Who is hired as employees, and under what circumstances?”  The answers to these questions move us toward an understanding to the way that economic change affects the way people live (geographic questions adding an important dimension, as well).  How do our categorizations of economic structure shape the way we ask these questions?

Distribution of returns
 Sectors.  The common distinction between low-wage, labor-intensive sectors and those that pay higher wages for a less intensive use of labor to staff substantial capital equipment suggests that industrial sector (again, broadly construed to include agriculture, mining, and services) is a key variable in the division of economic returns among capital, labor, and entrepreneurship.  The distinction is assumed to lie in differences in the economically viable technologies across sectors.  This distinction is widely recognized and used in the analysis of international investment and trade flows.  However, technological variations within sectors, and the observation of widely differing wages and profit rates across firms within the same sector, question the general utility of this distinction.

 Strategic groups.  The competitive advantages pursued and exploited by different firms within the same sector typically rely differentially on capital equipment, working capital for logistics or distribution systems, technical skills, or production and marketing labor.    These differentials hypothetically influence the returns that each of these sources of advantage can claim.

 Labor occupation.  Much academic and applied interest in the changing nature of the economy, including the increased proportion of service industries in the economy, is based on the changing employment of people by occupation.  The wage, benefit, and tenure differences across occupations are key measures of the division of gains.  Empirical findings abound concerning the temporally increasing returns to occupations that generally require university education, across and within sectors [Carnevale and Rose 1998]. To the extent that differential returns by occupation reflect the balance of supply and demand of occupation-specific capabilities, these differentials should exist only as deeply or as long as the imbalances.  However, the occupational division of labor reflects actual skills required to perform tasks, as well as attitudes toward work and social attributes that are signaled by educational attainment.

 Service industries and activities have been vilified for their concentrations of low-skill, low-wage occupations relative to the moderate pay of production jobs in capital-intensive manufacturing [Bluestone and Harrison 1982;  Walker 1985].  However, the consequent bifurcation of wage levels in service industries and activities has not been uniformly found [Quinn 1988;  Castells 1996;  Beyers and Lindahl 1997a].  In fact, the changing proportions and characteristics of labor occupations themselves bear further investigation, rather than using sectors or activities as proxies for occupation.

 Among the important characteristics of occupations are the nature of labor markets across occupations, and the ethnic and gender compositions of the labor supply (by location).  The operation of labor markets varies substantially across occupations, perhaps more substantially than across industries, activities, firms, or locations (certainly, than across locations within a given country context).  Important dimensions include the presence of employment ladders within an occupation, labor organization into trade unions, the importance of employment mobility, and the modes of employer and employee search.

 Given the importance of wage employment for most people and households in industrialized settings, the relationships among ethnicity, gender, and occupation are important determinants of  groups’ economic and social welfare.  Service activities and sectors, construed broadly, present substantial occupational stratification [Kutscher 1988].  In almost every sector, men and women, majority and minority ethnics, nationals and immigrants, and class (and age) cohorts routinely claim different occupational roles.  These roles are mediated through education and employment networks, and are often permeable [Carnevale and Rose 1998;  Wootton 1997].  But they persist:  how do they affect the impact of the growth of services employment on people, households, neighborhoods, and regions?

 Production chains.  The social divisions of labor and the distinctions in the returns to labor by activity and social characteristics are among the motivations for breaking production chains into separate corporate (and non-corporate) entities.  Sharp divisions are easier to maintain across corporate boundaries.  How do social divisions align with the contracting, self-employment, and employment-agency relationships in the provision of a given service?  How do these relationships vary according to the degree of social divisions (size of recent-immigrant community, degree of ethnic heterogeneity), or institutions such as minimum-wage or -benefit packages?

Individual labor employment
 In addition to the social divisions in the returns to labor and capital, the growth in service activities and occupations as a proportion of the economy has obvious implications for individuals’ employment:  doing what, where, under what terms and tenure, using or developing what knowledge or capabilities?

 Sectors.  Differential growth among sectors is important for individuals’ employment outcomes to the extent that skills, training, and work conditions are functions of sectoral affiliation.  While this is quite true for direct-production jobs across the broad divisions of agriculture, mining, manufacturing, and services, it is less true for indirect or management jobs, or across finer sectoral lines.

 Activities.  Labor employment, its demand, and its returns can be conceptualized in terms of the work to be performed and its contribution to the overall economy.  Carnevale and Rose [1998] focused on the changing functions to be performed in the American economy, and the implications for labor employment by function, skill level, and venue.  They note that workers in different functions are paid differently, reflecting both skill level and (perhaps more importantly) the changing importance of labor functions for overall organizational productivity.  However, among the fastest-growing category of “office workers,” most jobs (including remunerative, status-laden, and secure jobs) require few occupation-specific skills or training:  the function is more important than the occupation.  In their view, the individual’s general educational background and professional experience are the major determinants of who takes which functional roles.

 Castells [1997: 243-5] elaborated a three-dimensional division of labor in which to situate individual workers:  the fundamental tasks to be performed;  the level of interaction and communication required by workers, which also describes the degree of autonomy with which a worker operates relative to other workers or organizations;  and the location of the worker and task within a decision-making hierarchy, which describes the degree of autonomy with which the worker defines task execution.

 This perspective leads to questions regarding the changing mix of functions called for and the design of the work processes to fulfill them, rather than questions regarding labor-demand growth by sector or occupation.  However, empirical investigation requires new standardized data-collection protocols, primary research tailored to each project, or heroic assumptions regarding the distribution of functions across occupations and sectors.

 Occupation.  Labor-market study focused on the individual most often relies on supply and demand by occupation.  Growth in service activities in general and producer-services in particular has been reflected in the growth of professional, managerial, technical, and clerical occupations.  At this broad scale, these occupational categories probably do reflect the immediate limitations of individuals in their job searches.  In many of the producer services, however, fine occupational distinctions may not be as important as in manufacturing, health services, or education.

 Production chains.  The conceptual relevance of production chains and inter-organizational networks for individual labor employment is an open question.  It depends on the presence of employment chains or connections within the production chain or across organizations in a network, or the possibility that successful new-firm creation is enhanced by prior experience within some part of a production chain that one then serves entrepreneurially.  Do individuals become embedded in and mobile within particular production chains and inter-organization networks, or is employment change more related to sector, function, or occupation?
 

THE  GEOGRAPHIC  PROJECT
 What I will call “the geographic project” is primarily economic, but is also concerned explicitly with the location of these economic activities and their economic impact on localized regions.  Indeed, the questions of activity location and regional development are separable, and appear separately in Table 1.  In this section, I will discuss them as two parts of the same, fundamentally geographic question.

 Sector.  Traditional industrial location theory is based on a sector-driven set of input, market, and competitive factors that determine the least-cost, maximum-revenue, or maximum-profit location, where costs of procuring factors, applying labor and capital, and satisfying the market vary with location.  Lentnek et al. [1992; 1995] outlined location models for services to manufacturing as a function of the production-opportunity costs faced by clients as distance to the service increases, and the minimum market size required for economic provision of the service.

 Models derived from these theoretical relationships are difficult to apply in some service sectors when the dominant input is labor and product distribution entails little or no physical movement of clients or products.  In fact, one can distinguish a set of sectors on this basis, which would exclude retail and wholesale trade, basic education and health care, and services to facilities, and would include much of the producer services.  For these large and growing sectors, sector-specific location factors are fairly irrelevant.  Specific activities’ reliance on sources of labor, information, or communications infrastructure are key, and these needs vary by the nature, size, and organizational status of the function rather than the industrial sector of the activity.

 Models and explanations of territorial growth and development (at the national and sub-national scale) rely heavily on sector-based analyses of growth, linkages, and productivity.  The usual assumptions are that localized factors or markets create an advantage for certain kinds of activities, distinguishable by sector;  and that the growth, linkage patterns, and productivity trends of regional activity is reasonably decomposed by sector.  However, sectoral differences are not the only basis from which service activities respond to regional differences;  patterns of growth, linkages, and productivity increase have great intra-sectoral variation.  In these ways, sector-based models of regional growth and development fall short of their purposes.

 Economic function.  Massey [1984] brought the recognition of geographic fragmentation of functions within corporations to the forefront of economic geography.  The emphasis on economic function (e.g., corporate control, product and process planning, component production, product assembly, and client servicing) represented a dimension oblique to industrial sector (though the product-level activities are assumed to lie within product-focused sector).  The fragmentation was generally called a spatial division of labor (or, at its international scale, the new international division of labor ([Fröbel et al. 1980]), to emphasize the mobility of capital, technology, and control across the branches of large firms, compared to the relative immobility of labor and labor regulation.  However, the empirical manifestations of interest were the different functions within or organized by companies operating in multiple locations.  The presence of such discrete functions and locations makes it easier to identify the business services that occur within manufacturing, retailing, or other sectors, and provide some empirical data for the study of business-service location.

 The most common approaches to analyzing location of different functions are the respective labor-skill requirements of functions, and the relative need to locate in proximity to other functions within or outside the firm.  This has proven helpful in researching agglomeration of types of functions in “world cities” [Sassen-Koob 1984], the suburban metropolitan location of other functions [Moss and Dunau 1986;  Nelson 1986], and the long-distance relationships employed for other functions [Fröbel et al. 1981].  An alternative conceptualization relies on the information content of different economic functions, under the assumption that certain types of information inputs and analysis requires proximity to information sources or clients.

 With respect to regional development, these conceptualizations of the distinct locational tendencies for specific economic functions provide one key interpretation for patterns of localized economic growth or decline.  High-wage locations that provide the kinds of labor, infrastructure, and proximity beneficial to control, planning, or specialized-information gathering functions prosper with those activities.  High-wage locations not as favored by such activities lose relative advantage.  Selected low-wage, low-regulation, politically stable locations become attractive for activities with less need for agglomeration, information, or proximity [Martinelli 1991].

 There is an additional implication of the spatial division of functions for regional development.  The activities that require infrastructure and information-rich environments themselves add to the localized ease of obtaining specialized information and assistance, creating external benefits to other activities in the region — activities that may exist in many regions, but may grow more productive and competitive in certain regions.  Again, this possibility is rationale for studying the activity-mix as well as the industry-mix of regions.

 Production venue.  In many ways, the precursor of the producer-services studies since the mid-1980s were the studies of office location and growth in major cities [Armstrong 1972 and 1979;  Daniels 1975 and 1977].  These studies were prompted by the burgeoning real estate and employment markets in metropolitan central business districts.  As a result, they studied measures such as net increases in commercial office space and employment by zone within the metropolitan area. The study of suburban commercial centers [Daniels 1985;  Garreau 1991;  Harrington and Campbell 1997] continues in this tradition.  Recognition that these trends were in turn based on changes in the organization of the economy and of companies within it led to the sector-based study of services.

 The localized distribution of economic activities has implications for employment, household-income, property-value, and transport-demand variations across large metropolitan areas.  For example, producer-service and other public and private administrative services have been referred to as the employment base driving inner-city residential gentrification in the 1970s.  These activities were the major tenants of the massive office developments of the 1980s, in central cities and selected suburbs of major North American cities.  These are all important considerations in local development and infrastructure planning.  These distributional characteristics are made more important by of the separation of  planning, tax-base, and public-service provision among the jurisdictions of most large metropolitan areas.  If the locational distribution of employment and earnings shifts from one part of a metropolitan area to another, the associated streams of income and property taxes may shift from one municipality, or even one state, to another.

 Labor occupation.  Relatively few academic studies approach location or regional development primarily from occupational data.  However, this approach is highly relevant for at least two important research themes.  The first concerns the spatial divisions of labor mentioned above.  While there are social and regulatory aspects to the international and interregional division of labor within and across sectors, a key feature is the employment of workers (through employment arrangements, self- or contracted employment, or employment agencies) of different occupations in different locations.  The effects of globalization on occupations is like that of any other resource under the Storper-Samuelson theorem:  increased demand for and wages to plentiful occupations in a given country, and a reduced premium paid to relatively scarce occupations.  Given the occupational specificity of many individuals, trends in labor demand and supply by occupation and location should be at the forefront of our concerns — perhaps in front of the related concern for labor demand and supply by sector and location.

 The second important research theme that can be addressed with occupational categories by location is the intra-metropolitan distribution of occupation-specific residence and employment opportunities.  This influences the demand for urban transportation for all employed residents.  However, for low-income residents, the journey to work can become an expensive barrier, given the name “spatial mismatch.”  Spatial mismatch has generally been studied by comparing rates of employment, unemployment, wage earnings, or commuting distance or time for working-age adults in central cities versus suburbs, usually by race and/or gender.  The underlying phenomenon of concern is the trend toward suburbanization of employment in most metropolitan areas, while (in the U.S.) poorer, less-educated, and racial-minority households remained concentrated in central cities.  Surely the relative location of occupation-specific supply and demand is relevant to this concern — at least as relevant as the sector-specific demand that is the focus of the original studies [Kain 1968;  c.f. Wylie 1996].

 With regards to understanding regional growth, occupational differences in wages and working hours affect the ways that employment affect regional economies.  The nature of household economic multipliers will obviously vary with the wage level (again, where “wages” reflect income earned through direct or contracted employment).  In addition, the kinds of household expenditures may vary by the (occupation-related?) conventions of gendered employment, working hours, and arrangements (affecting household expenditures for child-care, clothing, cleaning, and catering) [Hanson and Pratt 1992;  Massey 1995].

 Firms.  Ultimately, location decisions are functions of individuals who control economic assets.  Without rehearsing the history of behavioral study of industrial location, we can ask whether variations in firms’ assets, organization, or size systematically affects location decisions across the service sectors.  Beyers and Lindahl [1997b] surveyed producer-service firms of varying sizes, including sole proprietorships, finding successful firms in widely varying locations, serving local and non-local markets in each subsector. The individual producer services contractor or entrepreneur can engage in extreme locational flexibility, maintaining information connectivity based on a professional nexus and small-scale information technology [note 4].  But large-scale, corporate location requires labor, network connections, and physical infrastructure at such a scale that requires concerted corporate and public-sector action, as investment in one component in isolation is insufficient for information-intensive development.  Smaller firms may have other locational requirements for successful operations, partially substituting localization within an agglomeration for internal economies of scope.

 From a regional-development perspective, this translates to hypotheses regarding differential linkages and performance (growth, market areas, profit levels) of firms by size and organizational characteristics, especially in peripheral areas [O’Farrell et al. 1992;  Beyers and Lindahl 1996].

 Strategic groups.  There are several reasons to hypothesize relationships between the competitive strategy pursued by a service provider and its intra-metropolitan and interregional location(s):  types and range of labor requirements of different strategies;  variations in the use of spatial divisions of labor within the organization;  the ability of sales and profit levels to support expensive, central locations;  the ability of localized market characteristics to support different strategies.  In the regional dimension, the strategic orientations of dominant service firms influences labor occupations and compensation, and localized demand for external support services — even as small firms’ strategy development must recognize the localized environment of external support services.

 Production chains and networks.  How are production chains localized?  The concept of chains of linked activities adding value until final consumption is traditionally given locational specificity through the study of intra- and inter-regional input-output relationships.  The spatial division of labor is recognized as an additional point of connection between the economic and the geographic.  Storper [1995: 209] suggested an additional point of connection:  the localized knowledge and relationships that increase the efficiency of inter-organizational production chains.  Service activities within larger organizations and those hired by organizations play important roles in the establishment and maintenance of production chains.  The locational dimension of these activities within the production chain relies on the same three geographic processes:  input-output linkages, localized variation in labor characteristics, and localized knowledge.  Researching the location of service activities and linkages within production chains is difficult empirical work, but should be studied via these three processes.  The benefit would be a conceptual and empirical integration of “services” within broader economic interactions, in a more relational way than merely noting the location-specific growth of service activities.

 From a regional-development perspective, while it is clear that activities within a region benefit from participation in extra-regional production chains and information networks, it is not clear what localized characteristics encourage such participation or influence its benefit.  Empirical research has shown that the benefits from inter-organizational networking are greatest in connection-rich contexts with large networks.  Capello and Nijkamp [1996], Mackun and MacPherson [1997], and MacPherson [1997] among others, have found that “context” includes local context.  Again, how localized are successful production chains and inter-organizational networks?  Do other conditions of the inter-organizational linkages (final-product characteristics, corporate or social linkages among participants, communications technology) affect the localization of successful linkages?
 

UNDER-INVESTIGATED  RELATIONSHIPS  AND  UNDER-DEVELOPED  CONCEPTS
 This review suggests that certain categorizations of economic activity need to be de-emphasized, and others emphasized, in our attempts to understand the changing role of intermediate or business-related services in the economy and in regional development.  Four conclusions emerge:

1. The industrial-sector categorization scheme has conceptual limitations for some of the purposes that motivate current producer-service study.
2. Labor occupations still shape some of the individual and social-group outcomes of economic activity, though fine distinctions among occupations may be decreasing in importance.
3. The increased business use of inter-organization networks and the increased academic recognition of the way production is oriented around variable chains of production (or value added) should increase our use of networks and production chains as organizing frameworks.
4. New frameworks for conceptualization and categorization are difficult to implement in the absence of standardized data.

 To the extent that occupational differences mediate the relationship between wage earners and the economy, occupational distributions are central to the study of how individuals and social groups fare over time and by location.  To the extent that the social outcomes of economic activity motivate the geographic study of producer services, that study should relate economic change to occupational designations, requirements, institutional structures, and distributions.  How is the production process translated into sets of labor tasks, and how are these structured by occupation?

 The roles that producer-service activities play in the economy actually encourage us to shift our emphasis to the chains of planning, control, production, and marketing activities within and among organizations.  However, a severe empirical problem arises from the flexibility and variability that are the hallmarks of these relationships.  A geographic problem arises from the spatial inconstancy of these relationships.  In this as in any of the processes that seem to be de-linked from geography by means of inexpensive and ubiquitous communication and transportation, the key is the characteristics that remain localized:  physical infrastructure, most households, labor regulation, and long-term relationships fed by proximity.  The object of geographic study becomes the relationships between the mobile and the localized.

 Not only are new frameworks difficult to implement, the shift we are undergoing from an emphasis on sectors to more varied approaches has led us into a period of uncertainty about our purposes.  This paper attempts to clarify some of the reasons for the current heterodoxy.  Given the heterogeneity of approaches, research on “service activities” will reflect the researcher’s conceptualization of the economy.
 

NOTES

1.  Prepared for the annual meeting of the Association of American Geographers, Boston MA, 26-29 March 1998.  The author thanks Jeffrey D. Garneau for assistance.

2.  In sector-based empirical overviews, Hackett [1990] and Roach [1991] noted the negligible impact that service-companies’ very large expenditures on information technology had on total-factor or labor productivity during the 1980s.  They attributed this to a tendency to automate inefficient and non-responsive routines using hierarchically controlled, inflexible information and control systems.  They suggested more thorough revision of service production and marketing processes.  McConnell [1996] suggested that substantial improvements in computerization’s increase in labor productivity awaited the flexibility of non-hierarchical, networked computing and information management in the 1990s.

3.  Carnevale and Rose included all employment in the finance, insurance, and real estate sector within this broad category, despite the obvious relationship between some of the direct (and largely personal) services component of this sector and the “counter” workers, presumably because of the physical setting of FIRE activities and the up-skilling that networked technologies are allowing in the lower ranks of FIRE occupations.

4.  Beyers and Lindahl [1997b: 904] characterized 20 percent of the 418 producer services firms they surveyed in a range of sectors and urban/rural contexts as “high flyers,” rapidly growing, typically single-establishment, and generally small (fewer than 8 employees).  “A number of these firms include those run by migrants to rural areas, or to cities, who have brought or started a business built around their specialized expertise.”
 
 

REFERENCES
Armstrong, R.B.  1972.  The Office Industry: Patterns of Growth and Location.  Cambridge, Mass.:  MIT Press.

Armstrong, R.B.  1979.  National trends in office construction, employment and headquarter location in US metropolitan areas.  P.W. Daniels, ed.  Spatial Patterns of Office Growth and Location, pp. 61-94.  London: Bell.

Baran, B.  1985.  Technological Innovation and Deregulation:  The Transformation of the Labor Process in the Insurance Industry.  Working Paper no. 9, Berkeley Roundtable on the International Economy.  Berkeley, California: University of California.

Beyers, W.B. and Lindahl, D.P.  1996.  Comparing the competitiveness of rural and urban producer services.  Presented at the Western Regional Science Association meetings, Napa CA, February.

Beyers, W.B. and Lindahl, D.P.  1997a.  Flexibility, information technologies, and the changing workplace in the producer services.  Presented at the North American Meetings of the Regional Science Association International, Buffalo NY, November.

Beyers, W.B. and Lindahl, D.P.  1997b.  Strategic behavior and development sequences in producer service businesses.  Environment and Planning A 29: 887-912.

Bluestone, B. and Harrison, B.  1982.  The Deindustrialization of America.  New York: Free Press.

Capello, R. and Nijkamp, P.  1996.  Telecommunications technologies and regional development: theoretical considerations and empirical evidence.  Annals of Regional Science 30(1): 7-30.

Carnevale, A.P. and Rose, S.J.  1998.  Education for What?  The New Office Economy.  Princeton: Educational Testing Service.

Castells, M.  1996.  The Rise of the Network Society.  Oxford: Blackwell.

Clark, C.  1940.  The Conditions of Economic Progress.  London: Macmillan.

Coffey, W.J. and Bailly, A.  1991.  Producer services and flexible production: an exploratory analysis.  Growth and Change 22: 95-117.

Coffey, W.J. and Bailly, A.  1992.  Producer services and systems of flexible production.  Urban Studies 29: 857-68.

Daniels, P.W.  1975.  Office Location.  London: Bell.

Daniels, P.W.  1977.  Office location in the British conurbations: trends and strategies.  Urban Studies 14: 261-74.

Daniels, P.W.  1985.  Locational dynamics of producer and consumer services.  Ch. 8 in Service Industries: A Geographic Appraisal.  London: Methuen.

Fisher, A.G.B.   1935.  The Clash of Progress and Society.  London:  Macmillan.

Fröbel, F., Heinrichs, J. and Kreye, O.  1980.  The New International Division of Labour.  Cambridge:  Cambridge U. Press.
Garreau, J.  1991.  Edge City: Life on the New Frontier.  New York: Doubleday.

Greenfield, H.I.  1966.  Manpower and the Growth of Producer Services.  New York:  Columbia U. Press.

Hackett, G.P.  1990.  Investment in technology: the service sector sinkhole?  Sloan Management Review 31: 97-103.

Hanson, S. and Pratt, G.  1992.  Dynamic dependencies: a geographic investigation of local labor markets.  Economic Geography 68(4): 373-405.

Harrington, J.W. and Campbell, H.S.  1997.  The suburbanization of producer service employment.  Growth and Change 28 (3): 335-59.

Kain, J.  1968.  Housing segregation, Negro employment, and metropolitan decentralization.  Quarterly J. of Economics 82: 175-97.

Kutscher, R.E.  1988.  Growth of services employment in the U.S.   B.R. Guile and J.B. Quinn, eds.  Technology in Services, pp.47-75.  Washington: National Academy Press.

Lentnek, B.J., MacPherson, A. and Phillips, D.  1992.  Optimum producer service location.  Environment and Planning A 24: 467-79.

Lentnek, B.J., MacPherson, A., and Phillips, D.  1995.  A market coverage model for producer services.  Papers in Regional Science 74: 389-99.

Lindahl, D.P. and Beyers, W.B.  1998.  The creation of competitive advantage by producer service firms.  Economic Geography, forthcoming.

Mackun, P. and MacPherson, A.  1997.  Externally-assisted product innovation in the manufacturing sector: the role of location, in-house R&D and outside technical support.  Regional Studies 31(7): 659-68.

MacPherson, A.  1997.  The role of producer service outsourcing in the innovation performance of New York State manufacturing firms.  Annals of the Assoc. of Amer. Geogr. 87: 52-71.

Martinelli, F.  1991.  Producer services’ location and regional development.  P.W. Daniels and F. Moulaert, eds.  The Changing Geography of Advanced Producer Services, pp. 70-89.  London: Belhaven.

Massey, D.  1984  Spatial Divisions of Labour.   London:  Macmillan.

Massey, D.  1995.  Masculinity, dualisms and high technology.  Transactions of the IBG 20: 487-99.

McConnell, S.  1996.  Computers and employment: an overview.  Monthly Labor Review (Aug.): 1-ff.

Milgrom, P. and Roberts, J.  1990.  The economics of modern manufacturing: technology, strategy, and organization.  American Economic Review 80 (3): 511-28.

Moss, M.L. and Dunau, A.  1986.  The Location of the Back Offices: Emerging Trends and Development Patterns.  New York: Sylvan Lawrence Research and Data Center, New York University.

Nelson, K.  1986.  Labor demand, labor supply and the suburbanization of low-wage office work.  A.J. Scott and M. Storper, eds.  Production, Work, Territory.  London:  Allen and Unwin.

O’Farrell, P.,  Hitchens, D, and Moffat, L.  1992.  The competitiveness of business service firms: a matched comparison between Scotland and the South East of England.  Regional Studies 26: 519-33.

O’Farrell, P.,  Hitchens, D. and Moffat, L.  1993.  The competitive advantage of business service firms:  a matched pairs analysis of the relationship between generic strategy and performance.  Service Industries J. 12: 1-16.

Porat, M.  1975.  Defining an Information Sector in the U.S. Economy.  Report no. 15, Program in Information Technology and Telecommunications.  Stanford, California:  Stanford University.

Porat, M.  1977.  The Information Economy: Definition and Measurement.  Washington DC: U.S. Department of Commerce, Office of Telecommunications, publication 77-12.

Porter, M.  1980.  Competitive Strategy.  New York: Free Press.

Quinn, J.B.  1988.  Technology in services.  B.R. Guile and J.B. Quinn, eds.  Technology in Services, pp.16-46.  Washington: National Academy Press.

Roach, S.S.  1991.  Services under siege: the restructuring imperative.  Harvard Business Review (Sep.-Oct.): 82-91.

Romanoff, E. and Levine, S.H.  1993.  Information, interindustry dynamics, and the service industries. Environment and Planning A 25: 305-16.

Sassen-Koob, S.  1984.  The new labor demand in global cities.  M.Smith, ed.  Cities in Transformation.  Beverly Hills: Sage.

Singelmann, J.  1978.  From Agriculture to Services: The Transformation of Industrial Employment.  Beverly Hills: Sage.

Storper, M.  1995.  The resurgence of regional economies, ten years later:  the region as a nexus of untraded interdependencies.  European Urban and Regional Studies 2(3): 191-221.

Tschetter, J.  1987.  Producer services industries: why are they growing so rapidly?  Monthly Labor Review (Dec.): 31-40.

Walker, R.A.  1985.  Is there a service economy? The changing capitalist division of labor.  Science & Society 49(1): 42-83.

Wootton, B.H.  1997.  Gender differences in occupational employment.  Monthly Labor Review (April): 15-24.

Wyly, E.K.  1996.  Race, gender, and spatial segmentation in the Twin Cities.  Professional Geographer 48(4): 431-44.