My impatience with analytical Marxism largely derives from my impatience with analytic philosophy more generally. I have immense difficulty believing that condensing every political and social issue into a logical proof or game theoretical model helps us engage them in any substantive way. But the exceptionally frustrating thing about analytical Marxism is the way in which it endeavors to improve (even salvage) the Marxist approach by getting rid of its most valuable component: the dialectical method.
Consider the problem of defining class. Herbert Kitschelt (not an analytical Marxist, but a helpful elucidator in this case) highlights a very real issue in the introduction to The Transformation of European Social Democracy:
"Marxist debates have often engaged in a sterile confrontation between objectivist and voluntarist class conceptions. Objectivists claim that property relations of social categories ('class in itself') eventually give rise to conscious class mobilization under the impact of increasingly bitter distributive struggles with capital and a general decline of the economy. ... But such economic determinism has failed to predict empirical class action adequately. Conversely, political voluntarism renders class theory tautological; whenever political actors define themselves as classes, social classes exist and make a difference" (p. 13).
The problem, empirically, is as follows. Defining and operationalizing the working class in strictly objectivist terms does not yield the conclusion that it has, as Marx and Engels predicted, come to comprise the majority of society, or consistently served as a principal actor in deciding policy outcomes. If one tries to sidestep this problem — as, say, Adam Przeworski does — by defending a strictly voluntarist conception of the working class, one gets caught in a tautology that explains nothing.
This opposition between objectivist and voluntarist conceptions of class misses the mark, however, by insisting that we must choose between class-in-itself and class-for-itself. This is a false choice that misses the point of these categories to begin with. (Would Kant impel us to choose between the noumenal and the phenomenal?) This is why the work of Przeworski is particularly dispiriting. He understands that "[c]lasses are not a datum prior to the history of concrete struggles" (Capitalism and Social Democracy, p. 69), but only in a one-sided way. That the working class is a datum formed out of concrete struggle means, for Przeworski, that it functions solely as a site of identification and mobilization.
The working class is not, however, formed only by the "subjective" political actions of its members, but also by its "objective" location within a matrix of property, production, and distribution relations. What Przeworski ignores is that the constitution of the capitalist system, which makes the working class a comprehensible entity, is itself a product of struggle; continual bourgeois domination, strike breaking, the enforcement of property laws, etc. cannot be understood except as tools of the capitalists in class struggle. What we have is an oscillation between, and a mutual constitution of, the objective and the subjective. Class struggle does not only occur when classes act in conscious concert, as Przeworski seems to suppose. It is ever-present, sustained by the logics of everyday life within the capitalist system.
The choice between "voluntarist" and "objectivist" conceptions of class is, again, a false one, with which we are presented only when we lose sight of the dialectic and suppose that everything must be flattened into a sterile typology of static categories. And unsurprisingly, choosing either option over the other leaves us with a notion of class that is unworkable.
Analytical Marxism's rejection of the dialectic amounts to a rejection of Marx's method, which is deeply unfortunate. Marx's methodological insights are incredibly valuable and underutilized in the social sciences. Consider this passage from "The Method of Political Economy" in the introduction to the Grundrisse:
"It seems to be correct to deal with the real and the concrete, with the real precondition, thus to begin, in economics, with e.g., the population, which is the foundation and the subject of the enter social act of production. However, on closer examination, this proves false. The population is an abstraction if I leave out, for example, the classes of which it is composed. These classes in turn are an empty phrase if I am not familiar with the elements on which they rest, e.g. wage labor, capital, etc. These latter in turn presuppose exchange, division of labor, prices, etc. ... Thus, if I were to begin with the population, this would be a chaotic conception of the whole, and I would then, by means of further determination, move analytically towards ever more simple concepts, from the imagined concrete towards ever thinner abstractions until I had arrived at the simplest determinations. From there the journey would have to be retraced until I finally arrived at the population again, but this time not as the chaotic conception of the whole, but as a rich totality of many determinations and relations" (p. 100).
Contrast this with the far more reductive approach laid out by Jon Elster in his "Case for Methodological Individualism:"
"By methodological individualism I mean the doctrine that all social phenomena (their structure and their change) are in principle explicable only in terms of individuals — their properties, goals, and beliefs. This doctrine is not incompatible with any of the following true statements: (a) Individuals often have goals that involve the welfare of other individuals. (b) They often have beliefs about supra-individual entities that are not reducible to beliefs about individuals. 'The capitalists fear the working class' cannot be reduced to the feelings of capitalists concerning individual workers. By contrast, 'The capitalists' profit is threatened by the working class' can be reduced to a complex statement about the consequences of the actions taken by individual workers. (c) Many properties of individuals, such as 'powerful,' are irreducibly relational, so that accurate description of one individual may require reference to other individuals" (p. 453).
Whereas Marx proposes an oscillation from the abstract to the concrete and back, so that the simplest determinations may be articulated and totalities may be understood as richly as possible, Elster proposes a nosedive into the concrete, after which we will be so bloodied and battered that no ascension back toward the totality is possible. Instead, we writhe on the cold ground, desperately flopping toward relations between multiple individuals, but never able to conceive of anything on a grander scale than that.
Yet following Elster's lead, we still use comparatively abstract categories of "worker" and "capitalist," seeing classes as incentive structures directing rational individuals. But we are compelled to suppose these categories, not to mention the nebulous model of individual "rationality" that purportedly persists within them, without justifying or articulating them in any real sense. When we dispense with the dialectical movement between abstract and concrete as Elster does, then, we get the worst of both worlds. We wind up incapable of conceiving a rich totality, and using chaotic abstractions anyway.
When you lose the dialectic, you lose Marxism's central critical and analytical tool. You reduce Marxist categories to the flattest possible caricatures, and then proceed to use them regardless. If you see no value in the dialectic, why bother with Marxism at all?