This is a very confused subject. Here are some observations that might help
sort out some of the issues.
1. Marx took the distinction between productive and unproductive labor from
Smith. Smith's own discussion is characteristically ambiguous and
inconsistent. His main motivation is to consider the problem of capital
accumulation: productive labor in this light is supposed to add to the
physical "funds", the consumption fund and capital funds, that constitute
the economy's capital, while unproductive labor merely draws down the
consumption fund. (Productive labor also draws down the consumption fund,
but is supposed to add more to the consumption or other funds so that there
is a net addition.) In this respect the distinction between productive and
unproductive labor is parallel to the modern distinction between investment
and consumption. Smith contrasts the decision of a wealthy individual to
employ liveried servants on an estate (unproductive) with the decision to
employ the same people as hands in a factory (productive). But Smith
muddies the waters by conflating this definition with another one: that
productive labor fixes value in a tangible, vendible object (the
"physicalist" interpretation). As Juriaan and many others have pointed out,
this definition is increasingly implausible as the basis for analyzing
contemporary capitalist economies with enormous and rapidly growing service
sectors. Finally, Smith argues that the employment of productive labor
increases the wealth of the employer, while the employment of unproductive
labor decreases the wealth of the employer. This raises yet another problem
in cases where the employment of labor is privately profitable but adds
nothing to the social capital funds, for example, in the case of
advertising.
2. It has always seemed to me that Marx's discussion of this problem starts
out as verging on a dialectical joke. His main point is that the
distinction between unproductive and productive labor, like other concepts
in bourgeois political economy, has an ideological content: the bourgeois
economist is bound to view labor that produces surplus value as productive
and labor that does not view surplus value as unproductive. Thus part of
Marx's critical point is to show that the other aspects of Smith's
discussion (the "tangible commodity" and the "social funds") are in fact
subsidiary to this class animus.
3. The problem of how to account for the wages of unproductive labor (and
the costs of unproductive capital!) gets tangled up with the differences
between analyzing the reproduction of capital and the reproduction of
society. In the first two Books of Capital I Marx assumes that there is no
difference between these two questions: necessary labor is identical to
paid labor, and surplus labor to surplus value, and he abstracts from labor
expended outside the wage relation altogether. This becomes a problem in
the discussion of productive and unproductive labor, which is partly
motivated by the idea that socialist society would be less wasteful than
capitalist society (a point of view that developed in Baran and Sweezy's
work, for example), a question that can be conceptualized only by
separating the level of social reproduction and the level of capitalist
reproduction.
4. The problem of accounting for unproductive labor raises the issue of ex
ante and ex post conceptualizations. The overall point of Marx's analysis
of the labor theory of value, it seems to me, is to show that the division
of value added between wages and surplus value is the way in which the
capitalists appropriate the social surplus product, thereby establishing
that capitalist society is a class society like feudal or slave societies.
This is, however, an ex post way of looking at the matter. In the analysis
of the circuit of capital Marx is at pains to point out that wages are not
paid out of the realized value added, but advanced out of accumulated
capital (the ex ante viewpoint). After the wage is paid the worker has no
claim to any part of the value of the product (as opposed to a
share-cropper), and the capitalist takes the entire risk of realizing the
value in the product. Part of the difficulty that arises in accounting for
unproductive labor arises because we want a single accounting system to
represent both the ex ante and ex post viewpoints, which is impossible.
Thus it is not inconsistent to hold that the wages of unproductive labor
are advanced from accumulated capital and that the consumption of
unproductive labor is a deduction from the surplus value, representing the
surplus product of the society. The solution to these puzzles lies, in my
opinion, in being clearer on the aims of particular accounting systems,
and, if necessary, creating different accounting systems to answer
different questions.
Duncan
Duncan K. Foley
Department of Economics
Graduate Faculty
New School University
65 Fifth Avenue
New York, NY 10003
(212)-229-5906
messages: (212)-229-5717
fax: (212)-229-5724
e-mail: foleyd@cepa.newschool.edu
alternate: foleyd@newschool.edu
webpage: http://cepa.newschool.edu/~foleyd