The distinction between productive and unproductive labor is one of the more controversial ideas in economics. It can be traced to economic thought in antiquity,1 has been part of modern economic thought since at least the 17th century, and features prominently in Physiocratic and classical political economy (Marx 1969, 176–82; Boss 1990). While this distinction is rejected by utilitarian economics, it remains a fundamental concept in Marxian economics despite sometimes heated debates within Marxism itself.2 It is now associated most prominently with Marxist economics, though quite similar ideas can be found in other heterodox approaches, as in Veblen (1901) for example.
The literature on this topic demonstrates its continuing importance for Marxian economics. The productive/unproductive distinction can be clearly defined, is theoretically consistent, and provides a rigorous basis for empirical research. This distinction is also integral to Marxian theories of class and exploitation, and as such is an important aspect of Marxism’s ethical critique of capitalism. A longstanding and growing literature documents the tremendous growth in unproductive activities in developed economies over the last century. But there is no consensus about what the consequences of this are for firm performance or aggregate dynamics, and this question represents one of the frontiers of Marxian economics.
Marx develops a conception of productive and unproductive labor first briefly in Grundrisse (Marx 1993, 272–3, 305–6), and then subsequently through an extensive criticism of others in Theories of Surplus Value (1969, 152–300). Taking inspiration from Adam Smith he offers this definition:
Productive labor, in its meaning for capitalist production, is wage-labor which, exchanged against the variable part of capital . . . , reproduces not only this part of the capital (or the value of its own labor-power), but in addition produces surplus-value for the capitalist. It is only thereby that commodity or money is transformed into capital, is produced as capital. Only that wage-labor is productive which produces capital.
(Marx 1969, 152)
Marx retains this conception of productive labor in Capital with very little change.3 When first analyzing the labor process he remarks that only labor engaged in the creation of a product is “productive labor,” but also notes that in capitalism this criterion alone is not sufficient to determine whether labor is productive (1967a, 181). He returns specifically to this question again and gives this definition:
Capitalist production is not merely the production of commodities, it is essentially the production of surplus-value. The laborer produces, not for himself, but for capital. It no longer suffices, therefore, that he should simply produce. He must produce surplus-value. That laborer alone is productive, who produces surplus-value for the capitalist, and thus works for the self-expansion of capital. . . . Hence the notion of a productive laborer implies not merely a relation between work and useful effect, between laborer and product of labor, but also a specific, social relation of production, a relation that has sprung up historically and stamps the laborer as the direct means of creating surplus-value.
(Marx 1967a, 508)
These descriptions give two specific criteria for a Marxist definition of productive labor in capitalism. Assuming that production is viable and results in a surplus, productive labor satisfies two conditions:
1 It is production labor.
2 It is paid labor employed by capital.
Labor that does not meet both of these conditions is, by definition, unproductive.
These two conditions provide a theoretically rigorous and empirically useful definition, and together provide a common basis for most Marxist definitions of productive labor (compare, for example, Mohun 2003, 45; Shaikh and Tonak 1994, ch. 2; Resnick and Wolff 1987, 132–41; Foley 1986, 118–20; Howard and King 1985, 128–32). This definition involves two dimensions, a distinction between production and nonproduction activity, which is common to all economic theories, and a distinction between capitalist and non-capitalist production relations, which is of particular concern for Marxism. It is important to note that Marx’s distinction does not attempt to assess whether a type of activity is useful or useless (by whatever criteria), or whether it is a “good” instead of a “bad,” and it should also be noted that the difference between material and immaterial labor (or the distinction between goods and services) plays no part in this definition (Marx 1969, 156–60).4 Rather the salient characteristics are whether an activity is production and whether it is in the employ of capital. These two criteria provide a relatively simple definition of productive labor, but they also require further specification of what production is, and what it means to be employed by capital.
Much of the early part of volume one of Capital is devoted to developing a conception of capital. Marx refers to it as “self-expanding value” and notes that when money (value) is advanced for the purpose of making more money it becomes capital (Marx 1967a, 153). Marx’s famous circuits of capital are designed to illuminate precisely this process, and, according to condition (2) above, only labor that enters into a circuit of capital is potentially productive labor. This potential is realized when condition (1) is also satisfied.
Employers advance capital to purchase labor power, and labor power exchanged for a wage from capital becomes (variable) capital. But not all producers sell their labor power, and not all wages are advances of capital. Peasants, artisans, and other self-employed persons produce and sell commodities, but they neither sell their labor power for a wage nor do they produce primarily by hiring others to do it for them. In Marx’s assessment the productive/unproductive distinction does not apply to this type of producer because they are outside of capitalist social relations.
But what is the situation with independent handicraftsmen or with peasants who do not employ any workers, hence do not produce as capitalists? . . . they are producers of commodities, as always in the case of peasants . . . and I buy the commodities from them, . . . In this relation they meet me as sellers of commodities, not as sellers of labor, and this relation therefore has nothing to do with the exchange between capital and labor, hence it also has nothing to do with the distinction between productive and unproductive labor, which depends merely on whether the labor is exchanged for money as money or for money as capital. They therefore belong neither to the category of productive workers nor to that of unproductive workers, although they are producers of commodities. Their production is not subsumed under the capitalist mode of production.
(Marx 1994a, 141)
The case of independent production for use, as in household labor, would similarly fall outside of capitalism and be neither productive nor unproductive. In contrast, Marx considers wage labor paid out of “revenue”—a term for the income of a capitalist used for personal consumption (Marx 1967a, 591; Marx 1969, 291)—unproductive because it includes the sale of labor power but this does not involve an advance of capital. In various places Marx works through numerous examples contrasting workers paid from revenue to perform personal services for a capitalist with workers paid from capital to produce these same services as a commodity for sale. The general principle that he derives from these examples is that unless labor is exchanged against capital it does not enter into a circuit of capital, creates no surplus value for a capitalist, and cannot be considered productive.
The employees of state agencies are another important case of workers who receive wage and salary income that is not paid from capital. State agencies perform government functions associated with what Shaikh and Tonak call the “maintenance and reproduction of the social order” (Shaikh and Tonak 1994, 60).5 This includes functions of government at the national, state, and local level. These activities are primarily funded by revenues raised from taxation and the issuance of debt, and usually are not sold as commodities in markets. Marx’s discussion of government officials and employees clearly indicates that, following Smith, he views them as unproductive, but his analysis is very limited.6 In some cases he includes them with landlords, usurers, and rentiers, but in other cases he includes them with “ideological classes” (Marx 1967a, 446) who, directly or indirectly, perform services for the industrial capitalist. In both cases these groups are paid out of the surplus value of the capitalist, which does not return directly to them in the form of a marketable product, and hence these payments are not advances of capital. Marx uses the term “faux frais” (literally “false costs,” but interpreted variously as “overhead costs” or “incidental expenses”) to refer to costs like this, and includes direct business taxes in this category of expenses (Marx 1967a, 220; 1967b, 421, 457). Unlike state agencies, Marx sees state-owned enterprises as potentially both engaging in production and operating with a capital stock (Marx 1975, 200; 1967b, 97). Hence they may be capitalist, and wage laborers at these enterprises may perform productive labor and create surplus value. This treatment carries over into the modern Marxist literature (Resnick and Wolff 1987, 231–74; Shaikh and Tonak 1994, 59).
In Capital Marx refers to production as the labor process and describes it as “human action with a view to the production of use-values, appropriation of natural substances to human requirements” (Marx 1967a, 183–4). This presents production as deliberate human action to create goods or services able to satisfy human wants. But production takes a specific form under capitalist social relations. In this case it must not only satisfy some want or need, but must also be profitable. Profit gives the owner of capital an incentive to engage in production; usefulness or serviceability of the product is necessary but it is not a sufficient condition to induce capitalist production.
Production is a subset of human activities, shaped by but analytically distinct from nonproduction activities. This distinction is not uniquely Marxist. Stone and Stone (1961, 32) use the concept of a “production boundary” to define a set of production activities and differentiate them from nonproduction activities in the economy.7 For Stone and Stone nonproduction activity consist of consumption, accumulation and international trade, but their taxonomy is only one of the many possible. The boundary between production and nonproduction activities and the classification of nonproduction activities are very basic theoretical issues with competing perspectives. The question of what is and is not production has a long history, and different positions in these debates help define competing paradigms in economics.8 Indeed, much of the disagreement over productive and unproductive labor is fundamentally paradigmatic dispute over the location of the production boundary.
Marx’s views are clearly grounded in classical political economy, and hence differ significantly from the current orthodoxy, which is strongly shaped by utilitarianism. Alfred Marshall, for example, that great popularizer of the utilitarian approach to economics, expresses this rather plainly when defining production:
Man cannot create material things. When he is said to produce material things he really only produces utilities. In the mental and moral world indeed he may produce new ideas. But in the physical world, all the he can do is either to re-arrange matter so as to make it more useful, as when he makes a log of wood into a table; or to put it in the way of being made more useful by nature, as when he puts seed where the forces of nature will make it burst out into life. It is sometimes said that traders do not produce: that while the cabinet maker produces furniture, the furniture dealer merely sells what is already produced. But there is no scientific foundation for this distinction. They both produce utilities, and neither of them can do more: the furniture-dealer moves and re-arranges matter so as to make it more serviceable than it was before, and the carpenter does nothing more.
(Marshall 1890, 116)
For Marshall, like Jevons before him (Brennan 2006), what economic activity produces is utility, and any activity that creates utility is production. He seems not to recognize that this pushes the production boundary back so far that it would include a utility-producing smile exchanged between passers-by on the street. A more recent contribution that also advocates an expansive production boundary is that of Becker, who is well aware that he pushes the production boundary back so far that it all but disappears and production comes to encompass the majority of human action (Boss 1990, 247–51). Becker (1965, 503–4) uses the concept of “productive consumption” to refer to activity typically interpreted as consumption (such as eating, sleeping, and “play”) that he classifies as production. This leads him to conclude that “Not only is it difficult to distinguish leisure from other non-work but also even work from non-work” (Becker 1965, 504).
In contrast to these expansive views, very early in his economic work Marx (1973, 83–111) adopts a production boundary that defines distribution, circulation and consumption as nonproduction activities. This production boundary is made explicit by Marx’s famous circuit of money capital, M – C . . . P . . . C' – M', where M and M' represent quantities of money capital, C and C' quantities of commodity capital, either as input into production or output, and P is the transformation of capital inputs in production. While this is familiar, what is often overlooked is that production is only one moment in the circuit. The moments M – C and C' – M' represent circulation (the conversion of money into inputs and production output into money) and as such are nonproduction activities outside the production boundary.
In Capital Marx discusses nonproduction activities in some detail, with particular attention to circulation (1967b, chs 5, 6; 1967c, part IV), finance (1967c, part V), and land rental (1967c, part VI). In his analysis of circulation Marx notes that nonproduction activities are indeed necessary for capitalist production to take place, but this does not make them forms of production:9
In the production of commodities, circulation is as necessary as production itself, so that circulation agents are just as much needed as production agents. The process of reproduction includes both functions of capital, therefore it includes the necessity of having representatives of these functions, either in the person of the capitalist himself or of wage laborers, his agents. But this furnishes no ground for confusing the agents of circulation with those of production, any more than it furnishes ground for confusing the functions of commodity capital and money-capital with those of productive capital. The agents of circulation must be paid by the agents of production. But the capitalists, who sell to and buy from one another, create neither values nor products by these acts, this state of affairs is not changed if they are enabled or compelled by the volume of their business to shift this function on to others.
(Marx 1967b, 126–7)
It is worth considering precisely what circulation entails because, as Marx was well aware, production and circulation typically occur together and are not as easily separated as the circuit of money capital makes it appear. In practice circulation is typically the business of wholesale and retail trade enterprises, and involves several distinct processes: transportation, warehousing, marketing and the exchange of legal title to a commodity. In part these different activities are reflected in the final purchasers’ price of a good, which can be decomposed into the producers’ price, the wholesale and retail trade margins, and the transportation cost (Lawson, Bersani, Fahim-Nader and Guo 2002, 22). For Marx both transportation and warehousing are, within limits, aspects of the production process continued in circulation (Marx 1967b, 52–5, 136–52; see also Savran and Tonak 1999 and Tregenna 2009), and thus the practice of separating trade margins from transport costs is consistent with a Marxian decomposition of the activities of the trade sectors into production and nonproduction activities. The nonproduction activities of the enterprises in these trade sectors consist of marketing and the exchange of title, which in themselves create no good or service that enters into consumption, but rather are costly activities associated with the realization of produced commodities.10 Market exchange is only one of the many ways, along with planning, lottery, potlatch, tradition, kinship networks, right of citizenship, etc., societies use to allocate produced goods and services to final consumers, and these are all alike in the sense that none of them are forms of production.
Of course one could reject this argument by simply assuming that trading activities are per se the production of services, and this is the convention taken in modern national accounts. But this is clearly a decision based on paradigmatic grounds, not on anything obvious or intrinsic about these activities. Consider, for example, an increase in the retail trade sector mark-up ceteris paribus. Current conventions in national income accounting would report this as an increase in real national income and product because they interpret the retail trade sector as producing a service, the output of which is measured by its total trade margin (Bureau of Labor Statistics 2012). Consequently, an increase in the retail trade margin is recorded as an increase in the production and sale of retail services, even if the total volume of goods passing to consumers remains unchanged. This is axiomatic for a conception of the economy that assumes all remunerated activities are production, but it cannot be defended on utilitarian grounds or any grounds other than a paradigmatic commitment to treating all remunerated activities symmetrically as production. This is a theoretical choice, nothing more. Similar examples abound. The private, market-based U.S. healthcare system spends significantly more on activities, like billing and coding, risk assessment and risk management, finance, advertising, etc., than government, non-market systems. These circulation activities are entirely separate from patient care but significantly increase the cost of providing it. And yet these costs are counted as part of the “output” of health care sector under current conventions in economic accounting.
Looking beyond circulation activities, for Marxism most supervision in the workplace is, like circulation, nonproduction activity, but it is nonproduction for a very different reason. Supervision becomes necessary when the interests of workers and their employers are contradictory or antagonistic. Marx notes that when labor is a “combined social process” involving a number of different laborers working together, coordination is necessary and “this is a productive job (Marx 1967c, 383).” However, he also notes that whenever production is based on antagonism between the direct producer and the owner of the means of production (as under slavery and capitalism) management must also involve supervision and coercion to ensure labor productivity (Marx 1967a, 330–2).
The antagonism of interests between workers and their employers is easy to identify. A worker’s effort level is variable and the quantity of surplus they produce depends directly on work intensity. Work intensity is costly for the worker, requiring the exertion of energy and attention, but the profitability of the enterprise is directly a function of their effort level. Therefore, the effort level chosen by an unsupervised worker will certainly be less than what their employer would choose for them (see Chapter 5, “Labor and Labor Power,” in this Handbook). Supervisory labor is then a consequence of this antagonistic or contradictory relationship, necessary in order to impose the employer’s objectives on employees.
Supervision, in this case, is not necessary to produce, it is necessary to produce under capitalism because the conflicting incentives of worker and employer require it.11 It makes little sense to think of coercive supervision when a worker works independently, as do peasants, the self-employed, etc., and, in Marx’s estimation, it would not be necessary under communism either (Ollman 1977, 30). What is left then is that supervisory labor (as distinct from the labor of coordination) is nonproduction activity because it is necessitated by antagonistic social relations of production and the profit making imperative of capitalist production rather than anything intrinsic to a specific production activity. Instead of a productive input, coercive supervision consumes a portion of the surplus created in production (Marx 1972, 355). This contrasts with modern conventions, which recognize that principal-agent relations give rise to agency problems that require costly solutions, but, taking an expansive view of production, counts these solutions as production.
Space limitations prevent significant discussion here of the other two types of nonproduction activity discussed at length in Capital, finance and the rental of land. Marx never seriously entertains the question of whether these are production and simply accepts that they are not. The financial sector generates its income by taking deposits and lending money at interest, which is why Marx’s circuit of interest bearing capital, M – M', contains neither production nor circulation (Marx 1967c, 391).12 Land rental provides access to land and existing structures. Even though without them a market economy would be constrained, it is highly dubious to claim that these sectors produce a product (measured by sectoral income) rather than simply a claim on the output of the productive sectors.
The empirical Marxist literature estimating unproductive labor stretches back at least to the 1920s, usually done as part of estimating of value aggregates for national economies.13 This includes work on a number of different countries published in many languages, but the English-language literature focuses primarily on the U. S. and the discussion here reflects that orientation. Notable contributions to the development of Marxist empirical methods include Gillman (1958), Wolff (1977), and Moseley (1991). Shaikh and Tonak (1994) construct the methodology that is now standard for estimating productive and unproductive labor. This is described, analyzed in detail, and significantly updated by Mohun (2005, 2014) and Paitaridis and Tsoulfidis (2012). Olsen (2011) uses Shaikh and Tonak’s basic theoretical approach to develop a Marxian social accounting matrix and emphasizes the class aspects of this work.
Shaikh and Tonak estimate productive and unproductive shares of total employment for the 1948–89 period (1994, Table F). They find that 85 percent of total employment growth in the U.S. economy over this period was unproductive, and that the unproductive share of total employment grew from 43 percent to 64 percent. Furthermore, they find that while a majority of unproductive labor is in nonproduction sectors, the increase in unproductive labor increased disproportionately because of growth in nonproduction employment in production sectors (trade, government, etc.).14 Mohun (2005, 2006, 2014) and Paitaridis and Tsoulfidis (2012) present more recent estimates. Drawing general conclusions from these various estimates is difficult because they are not easily comparable. All of these estimates are ultimately based on the U.S. Input-Output and National Income and Product Accounts, supplemented by Bureau of Labor Statistics surveys, and these change significantly as a result of the transition from the older Standard Industry Codes (SIC) system of industrial classification to the current North American Industrial Classification (NAICS) system in the late 1990s and early 2000s. These changes are significant enough that data using the SIC system are not directly comparable with that using the NAICS system, and likewise the Marxian estimates using data collected with these different classification systems are not comparable. Differences in methodology, revisions to data sources, and the inclusion or exclusion of general government also gives rise to significant differences in results.
Despite their differences, these studies indicate that the long-term trend of a growing unproductive share of total employment, which Gillman (1958) first identified as beginning in the late nineteen-teens, persisted through the 1980s and then levelled off, but the growing share of wage and salary income going to unproductive workers continued unabated. The twentieth century U.S. economy is characterized by persistent growth in unproductive labor and a concomitant increase in the unproductive share of wage and salary income, with these trends diverging after 1990. Easy characterization of the U.S. experience is further complicated by Mohun (2014, 363–5), who shows that over the 1964–2010 period the increasing share of wage and salary income captured by unproductive labor went entirely to supervisory and managerial employees, while the share going to the “unproductive working class” remained constant. Mohun’s analysis shows the importance of a Marxian class analysis that goes beyond a simple distinction between productive and unproductive labor and considers the composition of these broad class groupings. Characterizing unproductive labor as an undifferentiated group obscures important within-group differences.
The Marxian classification scheme for productive and unproductive labor may be theoretically consistent and provide a sufficient basis for empirical research, but the important question remains whether this distinction yields important insights for economic and social analysis. The importance of the finding that the U.S. economy added large numbers of unproductive laborers in the 20th century is called into question by the fact that it also grew robustly over much of this period. What precisely is the substantive research program motivated by the stylized fact of a growing class of unproductive laborers?
One consistent argument made by Marxist authors is that unproductive activities are simply a drag on economic performance because they consume surplus value and reduce profits, which interferes with accumulation and growth (Gillman 1958, 85; Moseley 1991, 153; Shaikh and Tonak 1994, 213; Paitaridis and Tsoulfidis 2012, 217–18). Baran (1957) and Baran and Sweezy (1966) take a somewhat different perspective and argue that unproductive activities are “waste” within the capitalist economic system, but this waste has an unintended benefit. They propose a tendency for the surplus to rise in the era of monopoly capitalism (Baran and Sweezy 1966, chs 3–4), potentially giving rise to a realization problem from a lack of effective demand. For them, growth of unproductive activities provides a necessary outlet for the expanding surplus and thereby supports the effective reproduction of the system.
There is support in Marx’s writings for the idea that some unproductive activities are ‘wasteful’ in the sense of arising simply out of the antagonism inherent in capitalist social relations, or the ideological, policing, and public welfare functions necessary to maintain stability in an exploitative, class-divided society. But it is much more difficult to sustain the argument that Marx viewed all unproductive activities as simply burdening capitalist reproduction and growth. Marx clearly argues that some unproductive activities may increase surplus value in capitalist production, thereby increasing profitability, and indeed in many cases this is precisely why an industrial capitalist is willing to pay for them.16 For example, in analyzing circulation Marx (1967b; chs 5–6; 1967c, 279–80) argues that merchant’s capital reduces the time of circulation M – C and C' – M', which reduces the turnover time of industrial capital and increases surplus value. Supervisory labor increases work intensity of productive laborers, and in so doing increases the rate of surplus value. Similar arguments can be made for other forms of nonproduction activity.
Some Marxists have pursued research along these lines. Gordon (1990; 1996, ch. 3) studies the growth of corporate bureaucracies in the postwar period, with particular emphasis on supervisory labor, and argues that U.S. corporations increasingly relied on a “stick” strategy of conflictual labor relations, in which workers receive low wages and little job security and must therefore be intensely supervised in order to produce efficiently.17 Duménil and Lévy (2003, 2011) emphasize that the purpose of unproductive labor within enterprises is to maximize the profit rate, and focus attention on the growth of unproductive labor as a means that corporations use to increase profit beyond simply inducing productive workers to increase labor intensity.
Resnick and Wolff (1987, 187–91) were perhaps the first to stress the complex and contradictory role of unproductive labor in Marxian economic dynamics. They emphasize that capital accumulation is inversely related to growth in unproductive labor only if this growth leaves the Marxian rate of profit unchanged or declining. Since the effect of any expenditure of surplus value on an enterprise’s rate of profit is highly contingent, so too is its effect on capital accumulation. Increasing expenditure on the internal bureaucracy of a firm, for example, could reduce total cost by reducing the cost of finance, purchasing, etc., thereby increasing profit, and potentially capital accumulation. But if this expenditure simply increases the compensation of employees without reducing total cost, profit and accumulation would be reduced rather than increased. This same contingency exists for almost any use of surplus by the enterprise.
Several authors have sought to establish the impact of unproductive labor by introducing it into models of economic growth. Wolff (1987, ch. 4) made the first sustained effort along these lines, but his work owes more to Baumol’s two-sector (manufacturing, service) model of unbalanced growth and Solow’s neoclassical growth model than it does to Marxism. In effect, Wolff adapts Baumol’s theory of the differential rates of productivity growth between goods and service sectors to model the effect of unproductive activity on a capitalist economy. Dutt (1991) takes inspiration from Wolff’s work and endeavors to develop an avowedly “neo-Marxist” two-sector model. But again the long-period dynamics are driven by differential rates of “productivity” growth in the two sectors (positive in the productive sector, zero in the unproductive sector), and the long-period dynamics are ultimately driven by something akin to “Baumol’s disease.” There is certainly something to be gained in Marxism by recognizing these Baumol-type effects, but it is just as certain that the effect of unproductive activity on Marxian dynamics cannot be limited simply to this.
More recent work introduces unproductive labor into Marxian models in a more complex way than has been done previously. Olsen (2015) and Vasudevan (2016) integrate unproductive labor into dynamic models of a capitalist economy, but unlike the Wolff-Dutt approach allow for these activities to potentially increase work intensity, reduce the turnover time of capital, or both. Olsen also introduces endogenous technological change resulting from ongoing transformation of the labor process and finds that, like Resnick and Wolff’s firm-level analysis, at the aggregate level the effect of unproductive activities on economic performance is contingent. It may squeeze profits and reduce economy-wide growth or increase work intensity, reduce turnover time, and develop productivity-enhancing technological change, thereby increasing profitability and the rate of growth. Olsen’s analysis of Shaikh and Tonak’s (1994) data shows that both outcomes are observed at different points in the U.S. during the postwar period.
This emerging literature indicates that the Marxian perspective on unproductive labor is changing, and it raises important questions about precisely why it is called “unproductive.” One answer is simply that Marx inherited this terminology from classical political economy. But this ignores another less obvious but more profound explanation: the productive/unproductive distinction is the basis for Marx’s theory of class. In his mature economic works Marx very clearly analyzes the relationship between different occupations and social groups by referencing their position relative to the production, appropriation and distribution of a surplus.18 This requires a clear explanation of where surplus is created and how it subsequently circulates through the constellation of positions that constitute the class structure. In such a taxonomic system some produce surplus value, and hence are productive workers, while others receive the surplus value produced by others. Marx’s analysis of the class structure of capitalism is then intimately linked to the productive/unproductive distinction because it is an integral part of his theory of the origin of surplus value. The Marxian ethical and moral indictment of capitalism as a historically-specific class system that exists and reproduces itself through the exploitation of workers is also based on the idea that workers who create surplus value by their labor in production neither receive nor control it (Resnick and Wolff 2005, 34; Henry 1975). But if all wage and salary employment is production, then surplus value production is ubiquitous and both Marxian class theory and its ethical implications are greatly circumscribed.
What is clear is that the Marxist productive/unproductive distinction has sufficient theoretical and empirical bases, and provides unique and provocative insights into a capitalist economy. Research using this distinction has only recently reached the point where it is possible to employ many of the methods common in social science research. Whether this emerging line of research will yield insights that transcends paradigmatic and political boundaries remains to be seen.
1 Aristotle (1988, book I, part X, 15) groups retail trade with usury as ways of making money from money, which he considers unnatural because money, in itself, creates nothing.
2 See Hunt (1979, 309, note 4) for a list of this literature within Marxism, but Hunt’s treatment of the productive/unproductive distinction itself is not recommended. A more recent example is the exchange between Mohun (1996, 2002), Houston (1997) and Laibman (1999).
3 Gough (1972) provides significant detail of Marx’s treatment of productive and unproductive labor in Capital and Theories of Surplus Value. See also Marx’s Economic Manuscript of 1861–63 (Marx 1994a, 121–46) and “Results of the Direct Production Process” (Marx 1994b, 442–52), both of which contain significant analyses of this topic. Gough did not have access to either of these texts.
4 This point is unambiguous (see, for example, Marx 1969, 158) but is an important source of confusion in the literature. Marx uses a variety of occupations throughout his work to illustrate the conditions under which they are productive or unproductive depending on the social circumstances of the labor process, including writer, schoolteacher, piano player, actor, musician, prostitute, clown, and singer, so it is difficult to understand how there could be any controversy over this issue.
5 See also Resnick and Wolff (1997, ch. 5; 2002) for a Marxist analysis of the state.
6 Marxists have generally accepted as a settled issue that government activities are unproductive, but few have analyzed this in detail. Among the few that do are Shaikh and Tonak (1994, 60) and Savran and Tonak (1999, 138), both of which argue that government functions are nonproduction activity. There are sufficient grounds in Marx’s work to conclude that this is indeed his position (1969, 303–4).
7 Unlike the usage here, “production boundary” is often used to distinguish types of production included in national accounts from those, like illegal activities and household production for use, which are not. Following Stone and Stone (1961) this entry uses this term to distinguish between production and nonproduction.
8 Brennan (2006, 404) provides a very insightful analysis of the evolution of this idea from a feminist perspective and argues that the production boundary is “one of the most historically debated aspects of economic thought.” Studenski (1958, 11–26) traces a four-century-long history of the development of concepts of production and national income in economic thought, beginning in the 16th century. Studenski’s work exhibits obvious erudition, but his treatment of Marx is badly misleading and borders on polemical. Studenski seems to mistakenly interpret Soviet national accounting, with its exclusive focus on material production, as an accurate expression of Marx’s views. Boss (1990) also covers some of the more recent literature and, while no more sympathetic towards Marx than Studenski, does a significantly better job interpreting his work.
9 Boss (1990, 6–7) calls reasoning of this type an “input-output error” guilty of an “unproductiveness fallacy”. She coins these terms but fails to provide convincing theoretical foundations, instead seeming to assume that anything “economically necessary” simply must be considered production, and hence to argue the converse is to advance a fallacy. But the Marxian position is that many nonproduction activities, like trading, civil governance, military, police, etc., are forms of social consumption, not production (Shaikh and Tonak 1994, 1–5).
10 Baran and Sweezy (1966, ch. 5) analyze circulation from a Marxist perspective with particular emphasis on marketing.
11 See Bowles (1985; 2004, 269–78) for a rational choice microeconomic analysis of the capitalist employment relation that derives the existence of nonproductive supervisory labor in competitive equilibrium.
12 Marx’s views are, of course, significantly more sophisticated than this (Marx 1967c, part V). See Lapavitsas (2013, esp. ch. 6) for a modern analysis of the profits of the financial sector.
13 Shaikh and Tonak (1994, 161–72) provide a comprehensive survey this early empirical Marxian literature.
14 Mohun (2014, Tables 2 and 3) provides a comprehensive breakdown of productive and unproductive sectors by SIC and NAICS sectors. An appendix also contains a very useful description of data sources and in-depth discussion of methods.
15 This section draws from Olsen (2015).
16 It may seem contradictory to propose that unproductive activities may increase the production of surplus value and yet be considered unproductive. But these activities remain nonproduction, and hence unproductive, because the production boundary is established by function, not effect. Marx’s position on this is discussed by Olsen (2015, 39).
17 Goldstein (2012) extends Gordon’s analysis through the first decade of the 2000s.
18 See Chapter 3, “Marxian Class Analysis,” in this Handbook, and Resnick and Wolff (1987, ch. 3). Wright (2015, viii, 9–11, 44–7) discusses varieties of Marxian class theory with particular emphasis on the issue of exploitation and class.
Aristotle. 1988. The Politics. Edited by Stephen Everson. Cambridge: Cambridge University Press.
Baran, P. A. 1957. The Political Economy of Growth. New York: Monthly Review Press.
Baran, P. A. and P. M. Sweezy. 1966. Monopoly Capital. New York: Monthly Review Press.
Becker, G. 1965. “A Theory of the Allocation of Time.” The Economic Journal 75(299): 493–517.
Boss, H. 1990. Theories of Surplus and Transfer. Boston: Unwin Hyman.
Bowles, S. 1985. “The Production Process in a Competitive Economy: Walrasian, Neo-Hobbesian, and Marxian Models.” American Economic Review 75(1): 16–36.
Bowles, S. 2004. Microeconomics: Behavior, Institutions and Evolution. Princeton: Princeton University Press.
Bowles, S. and A. Jayadev. 2007. “Garrison America.” The Economists’ Voice 4(2): Article 3.
Brennan, D. M. 2006. “Defending the Indefensible? Culture’s Role in the Productive/Unproductive Dichotomy.” Feminist Economics 12(3): 403–425.
Bureau of Labor Statistics (U.S. Department of Labor). 2012. “Wholesale and retail Producer Price Indexes: margin prices.” Beyond the Numbers 1(8): 1–4. Available at: http://www.bls.gov/opub/btn/volume-1/pdf/wholesale-and-retail-producer-price-indexes-margin-prices.pdf (accessed July15, 2016).
Duménil, G. and D. Lévy. 2003. “Production and Management: Marx’s Dual Theory of Labor.” In Value and the World Economy Today, R. Westra and A. Zuege, eds. New York and Basingstoke: Palgrave Macmillan.
Duménil, G. 2011. “Unproductive Labor as Profit-Rate-Maximizing Labor.” Rethinking Marxism 23(2): 216–225.
Dutt, A. 1991. “Unproductive Activity and Stagnation: A neo-Marxian Model.” Review of Radical Political Economics 23(1 and 2): 95–103.
Foley, D. 1986. Understanding Capital: Marx’s Economic Theory. Cambridge, MA: Harvard University Press.
Gillman, J. M. 1958. The Falling Rate of Profit. New York: Cameron Associates.
Goldstein, A. 2012. “Revenge of the Managers: Labor Cost Cutting and the Paradoxical Resurgence of Managerialism in the Shareholder Value Era, 1984–2001.” American Sociological Review 77(2): 268–294.
Gordon, D. 1990. “Who Bosses Whom? The Intensity of Supervision and the Discipline of Labor.” The American Economic Review 80(2): 28–32.
Gordon, D. 1996. Fat and Mean. New York: The Free Press.
Gough, I. 1972. “Marx’s Theory of Productive and Unproductive Labour.” New Left Review I/76: 47–72.
Henry, J. F. 1975. “Productive Labour, Exploitation and Oppression—A Perspective.” Australian Economic Papers 14(24): 35–40.
Houston, D. 1997. “Productive-Unproductive Labor: Rest in Peace.” Review of Radical Political Economics 29(1): 131–147.
Howard, M. C. and J. E. King. 1985. The Political Economy of Marx, 2nd edition. New York: New York University Press.
Hunt, E. K. 1979. “The Categories of Productive and Unproductive Labor in Marxist Economic Theory.” Science and Society 43(3): 303–325.
Jayadev, A. and S. Bowles. 2006. “Guard Labor.” Journal of Development Economics 79(2): 328–348.
Laibman, D. 1999. “Productive and Unproductive Labor: A Comment.” Review of Radical Political Economics 31(2): 61–73.
Lapavitsas, C. 2013. Profiting Without Producing. London and New York: Verso.
Lawson, A. M., K. S. Bersani, M. Fahim-Nader, and J. Guo. 2002. “Benchmark Input-Output Accounts of the United States, 1997.” Survey of Current Business 82: 19–43.
Marshall, A. 1890. Principles of Economics (1st edition). London and New York: Macmillan and Co.
Marx, K. 1967a. Capital, Volume 1. New York: International Publishers.
Marx, K. 1967b. Capital, Volume 2. New York: International Publishers.
Marx, K. 1967c. Capital, Volume 3. New York: International Publishers.
Marx, K. 1969. Theories of Surplus Value, Part 1. London: Lawrence & Wishart.
Marx, K. 1972. Theories of Surplus Value, Part 3. London: Lawrence & Wishart.
Marx, K. 1973. Grundrisse. London and New York: Penguin Books.
Marx, K. 1975. “Notes on Adolph Wagner.” In Karl Marx, Texts on Method. Trans. and edited by T. Carver, 159–219. Oxford: Basil Blackwell.
Marx, K. 1993. Grundrisse. New York and London: Penguin Books.
Marx, K. 1994a. Economic Manuscript of 1861–64, in K. Marx and F. Engels, Collected Works, vol. 34, 7–247. New York: International Publishers.
Marx, K. 1994b. “Chapter 6: Results of the Direct Production Process.” In K. Marx and F. Engels, Collected Works, vol. 34, 355–461. New York: International Publishers.
Mohun, S. 1996. “Productive and Unproductive Labor in the Labor Theory of Value.” Review of Radical Political Economics 28(4): 30–54.
Mohun, S. 2002. “Productive and Unproductive Labor: A Reply to Houston and Laibman.” Review of Radical Political Economics 34(2): 203–220.
Mohun, S. 2003. “Does All Labour Create Value?” In Anti-Capitalism: A Marxist Introduction, A. Saad-Filho, ed., 42–58. London: Pluto Press.
Mohun, S. 2005. “On Measuring the Wealth of Nations: The US Economy, 1964–2001.” Cambridge Journal of Economics 29(5): 799–815.
Mohun, S. 2006. “Distributive Shares in the US Economy, 1964–2001.” Cambridge Journal of Economics 30(3): 347–370.
Mohun, S. 2014. “Unproductive Labor in the U.S. Economy 1964–2010.” Review of Radical Political Economics 46(3): 355–379.
Moseley, F. 1991. The Falling Rate of Profit in the Postwar United States Economy. New York: St. Martin’s Press.
Ollman, B. 1977. “Marx’s Vision of Communism: A Reconstruction.” Critique 8(1): 4–41.
Olsen, E. K. 2011. “Modeling the Economic Surplus in a SAM Framework.” American Journal of Economics and Sociology 70(5): 1175–1207.
Olsen, E. K. 2015. “Unproductive Activity and Endogenous Technological Change in a Marxian Model of Economic Reproduction and Growth.” Review of Radical Political Economics 47(1): 34–55.
Paitaridis, D. and L. Tsoulfidis. 2012. “The Growth of Unproductive Activities, the Rate of Profit, and the Phase-Change of the U.S. Economy.” Review of Radical Political Economics 44(2): 213–233.
Resnick, S. A. and R. D. Wolff. 1987. Knowledge and Class. Chicago: University of Chicago Press.
Resnick, S. A. and R. D. Wolff. 2002. Class Theory and History. London and New York: Routledge.
Resnick, S. A. and R. D. Wolff. 2005. “The Point and Purpose of Marx’s Notion of Class.” Rethinking Marxism 17(1): 33–7.
Rubin, I. I. 1973. “Productive Labor.” In Essays on Marx’s Theory of Value, I. I. Rubin, 259–275. Montreal and New York: Black Rose Books.
Savran, S. and E. Ahmet Tonak. 1999. “Productive and Unproductive Labour: An Attempt at Clarification and Classification.” Capital & Class 22(68): 113–152.
Shaikh, A. M. and E. Ahmet Tonak. 1994. Measuring the Wealth of Nations: The Political Economy of National Accounts. Cambridge: Cambridge University Press.
Stone, R. and G. Stone. 1961. National Income and Expenditure. London: Bowes and Bowes.
Studenski, P. 1958. The Income of Nations. New York: New York University Press.
Tregenna, F. 2009. “Services’ in Marxian Economic Thought.” Cambridge Working Papers in Economics (0935), Faculty of Economics, University of Cambridge. Available at: http://www.econ.cam.ac.uk/research/repec/cam/pdf/cwpe0935.pdf (accessed July 15, 2016).
Vasudevan, R. 2016. “Financialization, Distribution and Accumulation: A Circuit of Capital Model with a Managerial Class.” Metroeconomica 67(2): 397–428.
Veblen, T. 1901. “Industrial and Pecuniary Employments.” Publications of the American Economic Association, third series, 2(1): 190–235.
Wolff, E. N. 1977. “Unproductive Labor and the Rate of Surplus Value in the United States, 1947–67.” In Research in Political Economy, Volume 1, P. Zarembka, ed., 87–115. Greenwich: JAI Press.
Wolff, E. N. 1987. Growth, Accumulation, and Unproductive Activity. Cambridge: Cambridge University Press.
Wright, E. O. 2015. Understanding Class. London and New York: Verso.