How is "work" socially constituted?

How does "work" take shape in an advanced manufacturing and service economy? Is the division of labor a natural outcome of technology, or is it the result of a concrete set of social processes involving the strategies and interests of several groups? What kinds of social processes determine the bundle of skills, knowledge, and training that go together to represent a job classification? How is the suite of tasks and activities assigned to a given job arrived at?

Here is a brief description from the Bureau of Labor Statistics of one specific skilled job function within a factory environment, the millwright (link):

Millwrights install, dismantle, repair, reassemble, and move machinery in factories, power plants, and construction sites. Millwrights typically go through a formal apprenticeship program that lasts about 4 years. Programs are usually a combination of technical instruction and on-the-job training. Others learn their trade through a 2-year associate’s degree program in industrial maintenance. Employment of millwrights is expected to decline 5 percent from 2010 to 2020. Despite declining employment, job opportunities should be good for those with a broad set of skills in machine maintenance.

Here is a description of a low-skill job function, the hand laborer and material mover (link):

Hand laborers and material movers transport objects without using machines. Some workers move freight, stock, or other materials around storage facilities; others clean vehicles; some pick up unwanted household goods; and still others pack materials for moving. Generally, hand laborers and material movers need no work experience or minimum level of education. Employers require only that applicants be physically able to do the work. Employment of hand laborers and material movers is projected to grow 14 percent from 2010 to 2020, as fast as the average for all occupations. Job prospects should be good because the need to replace workers who leave the occupations should create a large number of job openings.

And here is a highly skilled industrial job, the mechanical engineer (link):

Mechanical engineering is one of the broadest engineering disciplines. Mechanical engineers design, develop, build, and test mechanical devices, including tools, engines, and machines. Mechanical engineers need a bachelor’s degree. A graduate degree is typically needed for promotion into managerial positions. Mechanical engineers who sell services publicly must be licensed in all states and the District of Columbia. Employment of mechanical engineers is expected to grow 9 percent from 2010 to 2020, slower than the average for all occupations. Job prospects may be best for those who stay abreast of the most recent advances in technology.

So how is it determined which tasks are assigned to which classifications? And what social processes determine which individuals receive the right kind of training to qualify for any of these jobs?

One researcher who has contributed a lot to this issue is Charles Sabel in Work and Politics: The Division of Labour in Industry and The Second Industrial Divide: Possibilities For Prosperity (with Michael Piore).  Sabel's work is particularly insightful in the light it sheds on the sociological processes and structures through which the bundles of skills needed in individual workers are determined and then transmitted through a training regime. Here is how Sabel frames the problem I'm interested in here:

This is an essay about the reasons industrialists create different kinds of factory jobs, about why workers put up with these jobs when they do, and about what they want when they do not. It shows how workers' ideas of self-interest, born of the principles of honor and dignity they bring to the factory, can be transformed by workplace struggles. And it shows how these struggles, colliding or combining with conflicts in the larger society and between nations, can reshape technologies, markets, and factory hierarchies. (Work and Politics, 1)

Notice the most fundamental idea being advanced here: that the natures of work, jobs, and implementation of technology are all the result of active back-and-forth negotiations over time between workers and owners.  The structure of work doesn't follow from a certain level or kind of technology; instead it is the result of an extended "game" in which the players seek out definitions of work and technology that suit them. Workplace struggles "reshape" technologies and jobs, and the division of labor is a socially mediated and historically conditioned reality.

This chapter argues that the capitalist organization of production creates clusters of jobs offering workers systematically different opportunities for the use and acquisition of skills, and for regular employment. The capitalists create jobs of various types, and the worker tries to find one suited to his ambition. (31)

Essentially the labor process is defined by several constraints, imperatives, and interests.  Products can be made for use and for sale in a market through a variety of combinations of tools, skills, and labor.  By subdividing tasks, economists and managers since Adam Smith have recognized that it is possible to increase efficiency or quality or both. And profit-oriented owners and managers have recognized that skilled workers can demand more for their services than unskilled workers.  So pursuing a combination of specialized tools and machines with repetitive, unskilled labor has been a profit-maximizing strategy since Henry Ford began experimenting with the assembly line.  "Fordism" is a specific economic and technological system, involving factory production for a mass market, specialized tools and machines, and mass unskilled labor.

There are many alternatives that exist for defining skill regimes that would get the job done for a given level of technology and organization. The Fordist regime separates workers into a low-skill group where only a minimal amount of training is required and the worker's activities are limited to a repetitive and simple set of actions, a smaller group of high-skill technicians, and a group of supervisors who oversee the activities of both groups. In this Fordist regime the low-skill workers are not expected to exercise independent judgment or to function as creative problem-solvers. The high-skill technicians are assigned a greater degree of independence and scope for problem-solving.

But, as Sabel observed in an important 1985 article with Jonathan Zeitlin ("Historical Alternatives to Mass Production"; link), the Fordist regime is not the only solution possible for advanced science-based production. There were alternatives. One such alternative is a high-skill, high-independence paradigm where ordinary production workers employ much more sophisticated skills and knowledge to carry out the production process. Here is how they describe their central argument:

Mass production -- the combination of single-purpose machines and unskilled labour to produce standard goods -- has been throughout this century the undisputed emblem of industrial efficiency. No more. Powerful currents of technological change are stirring up this sedimented lesson of the past. The development of numerically controlled machine tools which can be programmed to perform many different tasks automatically; the spreading use of such machines in highly competitive small firms in industries as diverse as engineering and textiles and regions as distinct as Baden-Wurtemberg and Emilia-Romagna; the increasing capacity of some large firms in Japan, West Germany, and the United states to switch production rapidly from product to product; the creation of new jobs which blur the distinction between skilled and unskilled work -- all these churn up established understandings of modern production methods. Engineers and managers take increasingly seriously the possibility that economic success in the future may depend on the flexible use of multi-purpose or universal machines and skilled labour to make an ever changing assortment of semi-customized products: a system that reverses the principles of mass production. (133)

The Volvo production system illustrates this regime. This approach bundles the market advantages associated with flexible production with a skills regime that focuses on teams of producers with advanced and flexible skills and an extensive degree of production independence.

These questions are pressing today for two important reasons. First, many of the "good" jobs that existed in the United States in manufacturing have either disappeared or have suffered major reductions in compensation. Moreover, the manufacturing jobs that are returning to the US are coming back at much lower rates of pay.  And jobs have disappeared for reasons that are very consistent with Sabel's argument: companies have made deliberate, strategic decisions to de-skill labor within their operations and to offshore some parts of the production process. Or in other words, companies have restructured work to increase profits. That is the logic of a capitalist economy.

Second, most observers accept the idea that worker productivity in the future -- and therefore worker compensation -- will be determined by the level of knowledge and skill the worker possesses. High skill and high knowledge production adds a lot of value to the product, and is compensated accordingly. But companies have a choice to make, whether they adopt production processes that depend on a smaller number of high-skill and high-pay workers, or processes that depend on a larger number of low-skill, low-pay workers.  And the company is of course aware of the terms of bargaining that exist with regard to these two groups; the high-skill group is more able to exercise influence in the bargaining process and thereby increase its compensation more rapidly.  Deskilling of the production process is therefore a tactical choice on the part of the employer.

This line of thought seems to have disjunctive consequences, neither of which is promising for the future of the American middle class: either companies will restructure their activities to incorporate a higher mix of high-skill workers in substantially smaller numbers; or they will continue to expand activities around a low-skill model while exercising substantial downward pressure on wages.  And this implies that employment growth will be slow, or else there will be more robust jobs growth in low-pay jobs leading to a falling standard of living for the majority of workers.

(Here is a post on Stanley Aronowitz and William DiFazio's version of this kind of analysis in The Jobless Future: Second Edition.)