To get at your questions, Max, I need a defintion of productive labor that is a bit more complete than than simple one I've been using. To wit: productive labor is that labor which: (1) produces surplus value for capital, and in that process (2) creates, modifies, or conserves use value, or is technically necesasary for the distributuion of use value (commodities). Products must have use value (be saleable as commodities) for surplus value to be realized as value to capital.
Whether labor is productive is determined by the social relations of its use; e.g., whether capital can extract surplus value from the product or service labor provides. But productive labor comes in several forms. It can produce services as well as goods. Productive labor can be expended in circulation, as well as production. Most things need to be packacaged, transported, handled, stored, and distributed. Performed for a capitalist, this labor is productive because it is necessary for sales, i.e., to the realization of the use value of the commodity, and to the realization of the surplus value embodied in the commodity. For example the above labor can prevent product deterioration or make consumption more convenient.
Supervisory personnel, as well as production workers, can be productive because some coordination and unification of the labor process is necessary. But the vast bulk of management in today's corporations, of course, are unproductive and are paid out of surplus value. Their work "arises from the antagonistic contradictions between capital and labor, from the class antagonistic character of capitalist production". [Marx]
One more important distinction. To be considered productive, labor must add to the use value, and thus the surplus value, of a product of service. For example, what Marx called "pure" buying and selling, undertaken because production takes the commodity form (e.g., most advertising) add nothing to use value and are therefore unproductive. Even if that labor is employed by a capitalist. The surplus value that an advertising capitalist extracts from its labor must come from the surplus value created when the product being advertized is made, since no surplus value has been added in circulation.
With that basic outline in place, I can tackle your easier questions:
> That aside, does a secretary in an advertising agency
> produce surplus for the owners? Retail clerk? Why not?
Different answers for each. While there is nothing inherently unproductive about secretarial work, the secretary in the advertising agency is unproductive if that advertising adds no use value to the products it hawks. Virtually always the case, I think. The cost of advertising, including the secretary's wage, comes out of the surplus value created when the product being advertized is made.
The retail clerk, however, is likely to be necessary to the distribution of products for sale. As such, use value and surplus value are enhanced. The clerk is productive, and when the store owner extracts surplus value from him, it adds to the mass of surplus value.
But your second set of questions is more difficult, and in fact, are a different kind of question than the first:
> If a publicly-financed bridge contributes to business
> output and surplus value, what is the purpose of
> distinguishing those who built it from 'productive' labor?
Very good question. You're getting here at a further dimension in the definition of productive labor I have not yet discussed. Marx defined productive labor as I have set it out above in order to explain the existence of surplus value and the origin of capitalism as exploitation. But capitalism is not just about the creation of surplus value and exploitation of labor by capital, but also about the accumulation of surplus value on an expanding scale. So we need to add a theory of accumulation and growth (the laws of motion) to the theory of exploitation.
To understand accumulation, you must look not only at whether labor produces surplus value, but also whether the resulting product is *productively consumed*, paving the way for the next round of production on an expanding scale. To cut to the chase, a product is productively consumed if it enters into the means of production (capital goods) owned by capitalists for the next round, or is part of the subsistence consumption basket of productive labor. In other words, a product is productively consumed if it contributes to the reproduction of capital goods or productive labor. All other goods and services are called luxuries, in the basic sense of not being necessary for capital accumulation. Luxuries are not just yachts (though of course yachts would be a part, but all consumption by capitalists and unproductive labor, government services not part of labor's subsistence (schooling provided by government through taxes, I have argued, is part of labor's subsistence needs; military spending is an example of government provided luxury).
Thus some marxists argue that to be considered productive labor, a worker must create surplus value for capital, and his product must enter the reproduction of capital goods or productive labor. Some argue further that if the concern is a theory of growth, the first part about creating surplus value for a specific capitalist doesn't even matter--they want to sever the link between productive labor and the creation of surplus value, and, in this case, define labor as productive solely on the basis of whether it is part of the reproduction cost of capital goods or productive labor.
Which takes me back to your question. But first I need to make another distinction. There are actually three kinds of labor: Productive and unproductive labor in the employ of capital, and noncapitalist labor. Public school teachers, e.g., are noncapitalist labor, as are all government workers, including those that built the bridge in your question.
Now you and I, Max, could argue all day about whether that publicly financed bridge, which you say plays a part in wealth creation, is to be counted as part of capital's means of production. Certainly bourgeios stats do not treat it that way. Profit rates are calculated on the basis of privately owned fixed capital investment only. All goverment enters the GDP at its employee compensation only (no profits, of course, but no governemt owned investment either). If, however, you could find a way to argue that the bridge is part of capital's MOP, there are some marxists (e.g., James Becker, Justin Blake) who, if they accepted your argument, would be with you in calling the workers that built the bridge productive labor. They would be noncapitalist labor as productive labor. Just don't count me among them.
> More broadly, seems to me that financial crises or their
> absence depend no less on the finances of "non-productive"
> enterprises as of productive ones. Anything that generates
> profits can generate bankruptcy and ensuing social
> instability. So whyfore the surplus concept? This
> is elementary to you and others, I imagine, but it
> might be of some interest to the less theoretically
> experienced among us.
The short answer to the first part is that there does tend to be a differnce in the type of contradiction typically associated with factors that create surplus value compared to those that use it up. The function advertising firms perform in soaking up surplus value can be replaced in many ways, should those firms go under. More difficult adjustments are needed when productive labor is laid off.
But when you ask, whyfor the surplus concept, I think we are back to questions about the importance of productive labor, the source of surplus value. Why concern ourselves with productive labor distinctions at all? There are several kinds of answers to that. I'll offer two. First, it helps to understand the origins of capitalism. Capitalism was created by the transformation of self-employed labor into wage labor producing surplus value for capital. At the time of transformation, most of labor was turned into productive labor for capital, but not all of it. You can't understand the origins of capitalist factories without an understanding of productive/unprioductive/noncapitalist distinction. As the creation of surplus value has proceeded, capitalism has evolved in many ways. It is useful to see the role each kind of labor has played in that process.
Second, capitalism is *about* exploitation. Exploitation is defined as the difference between what labor produces for capital and what it receives from capital in the form of a wage apprioximating its social subsistence needs. Quantitavely, you can't know what surplus value is until you know what the subsistence needs of productive labor are. Capitalism is also about the accumulation of surplus value on an expanding scale. To understand capitalist growth, to see the contradictions, antagonistic and nonantagonistic, that develop as the laws of motion play out, you must begin with a concept of productive labor as the source of surplus value.
> On a different note, I agree that the poverty rate
> leaves something to be desired, but its inadequacies
> have nothing to do with state of the debate on poverty,
> IMO. It could be five points lower or higher in every
> year and few consequences would ensue.
Yes. The nature of the discussion (too pathetically limited to call it a debate) of poverty typically has nothing to do with the number of people considered to be poor.
> On the other hand,
> the availability of a consensus measure -- even a bad one --
> is quite important in policy debates. It stimulates
> discussion, and it provides an entry point for some
> kind of policy response.
Agree. So spread the word around Washington that the official poverty numbers are based on a food budget meant only to keep people alive for two weeks (it was called the emergency food budget) and does not even come close to addressing minimumm physical, let alone social, subsistence. It's a joke left behind by Lyndon Johnson. Poverty is supposed to have something to do with what Marx meant by social subsistence, like the family budget numbers the BLS used to publish.
And the next time Doug uses the official poverty numbers as if they mean someting, jump on him, willya.
Roger