| This site is designed by |
 | Organizations Evolving |
The immense learning of Aldrich's book shows that there is a lot of promise in evolutionary theory applied to organizations. Let me start this review by contrasting our problem here with the problem of investigating evolutionary theory in biology.
The logic of biological evolutionary theory has itself evolved over the century and a half or so since Darwin. This means that its form is shaped by what works in that scientific niche. For example, the separation of "reproduction" into genes, mixing of genes in sexual reproduction, structural shaping of the immediate descendants' bodies by the mix, the physiological system dependent on that structure, and the behavioral possibilities that result in the reproductive success or failure of the descendants, have little direct analogy in the development of General Motors out of a buggy manufacturer.
But the logic of investigation in biology is adapted to that chain of separations, because they provide exogeneity, the knowledge of where a particular cause came from. The first step justifies the human genome project, which would make little sense in organizational behavior. I suppose the nearest analogy is the sort of "descent trees of patents" that Podolny and Stuart (1995) work on. But we do not believe in the high level of stability of reproduction of bits of genetic information in the buggy manufacturer, nor that we could identify those genes in the Chevrolet division.
The same applies to biological niches, to which genetic forms adapt. We can sort out the grand ecological niches of desert, tropical rainforest, tundra, prairie, and so on. When we go to a desert, we look for water conserving behaviors, so water conserving structures, and so eventually water-conserving genes. Sunlight is not scarce on a desert because vegetation is sparse; there is not enough water to support dense vegetation. Therefore we do not look for how a cactus can produce an efficient panoply of leaves to get as much sunlight as possible for the least investment of wood growth, or to get wood strong enough to support so many leaves in the wind with a small weight.
At the two ends of the reproductive chain, then, the genes and the niches, we can look for exogenous causes. Variations in water and competition for sunlight are exogenous at the niche end. Variations in the genes of parents and in the pollen that gets to a cactus flower are exogenous at the genetic end. The delicate unfolding of causal structures-that makes Darwin able to untangle the tangled bank, to explore the neat fit of the flower to the body of the insect and the insect to the shape of the flower-relies on this ease of locating exogenous causes, so difficult to find in organizational evolution.
Sometimes in Aldrich we run across an elegant example where the logic unfolds neatly. He reports on Han's study of the bifurcation of certified public accountant firms into the (then) big six versus multiple smaller ones (pp. 246-247). The big clients choose the big CPA firms because only big firms can handle the massive data review necessary. Middle sized corporations choose the particular big firms that the big clients in their industry choose, showing that the imitation that constitutes replication follows within-industry descent lines. The same selective forces apply within an industry's niche in the capital market, for that is what auditing firms are oriented toward; different industries are different niches. The smallest firms, whatever their industry, choose smaller auditing firms, indicating that they are in a different niche in the capital market, though perhaps (by being in the same industry) the same one in the product market.
But sometimes crucial logical links are missing. An example is his analysis on p. 219 of Fligstein's (1990) findings about the spread of multi-divisional firm organization from the 1920s to the 1980s. He notes the arguments that the selective forces toward the multi-divisional form are stronger if a corporation operates in several product markets, and has higher executives oriented to either sales or to finance (the capital market). Selective forces for multidivisional forms are evidently stronger if a corporation is in an industry dominated by that form. That is, distinctive niches in either the sales or capital markets evidently select for this form.
Thus we should, by evolutionary logic, expect that there would be niches with different forms, niches with the multi-divisional form and niches with others forms. But Aldrich's conclusion is, "Thus, a type of transformation that began in the 1920s, when two large firms [Du Pont and General Motors, both in multiple markets and dominated by executives and board members strongly oriented to the capital market, ALS] adopted it, became a standard feature of large firms' structures by the 1980s. From a cohort-specific practice, MDF evolved into community-wide organizational knowledge." But this means that the selective situation must have changed as the form became dominant in its niches, so that the whole spectrum of industrial corporations dependent on the capital market adopted the form, whether or not they were in multiple markets and dominated by sales executives.
Fligstein in fact made an argument that the multi-divisional form became institutionalized in the capital market for industrial corporations toward the end of the period (and also that finance became more dominant in higher executive positions in such firms), and thus that the selective pressures did indeed change. For the capital market, at least, and for large industrial firms, there was a different niche requiring a different form of corporation in the 70s and 80s than in the 20s through the 60s.
The problem with Fligstein's (1990) argument (and so presumably the reason Aldrich missed a crucial part) is that there are no easy exogenous causes, which we know come from somewhere else, that change the nature of the niche in the capital market occupied by the moderate to huge sized industrial corporations. And we do not know why retail chain department stores, with 40 or so divisions headed by buyers and allocated separate floor space in the stores (or separate pages in the catalogs), were under different pressures earlier. Further from the main time of origin of department stores, all such stores were selected for using multiple autonomous divisions. The departments in the tiny department store in Mount Pleasant, Michigan, where I worked in my teens, had a multidivisional form with quite autonomous buyers and separate accounting schemes for each and the whole bit. In Fligstein's argument there is no way to measure the rainfall and the sunlight, and to know where they come from, to sort out the causal autonomy of niche forces. So the problem for us in organizational evolutionary studies seems to me to be get deeply into the physiology of the organizations in our studies, into the flows of information, goods, and activities modifying those flows.
An early study, that we have forgotten, is a wonderful example of what I think we need: Benjamin Zablocki's Alienation and Charisma: A Study of Contemporary American Communes. We didn't notice it much when it was published, because the population of communes had already disappeared, and because business schools do not want to know what makes communes disappear, on average, even faster than restaurants. There is extensive discussion of the contexts, the main varieties of alternative services provided for people of different needs, and the like, hinting at selective pressures. But there are also detailed studies of the networks among people within the communes, the degree of individualism of the motives for joining, the degree to which one is converted in different types of communes into someone who cares for the fate of the whole, what makes communes split into warring factions, what varieties of leaders can hold them together under different circumstances, and so on, that are shown to affect the rates of individuals quitting and communes disappearing.
That is, there is detailed (and quantitative) study of the physiology of the communes as structures, the main varieties of physiologies, and the different vulnerabilities of different forms to different sorts of physiological disruption. It turns out, for example, that the problem of turnover is quite different from the problem of the survival of the commune. For example, if a person is well-loved within a commune, he or she is considerably less likely to quit. But if everybody loves everybody, the commune as a unit is more likely to dissolve or split. Zablocki seems never to be in trouble about what is exogenous, what endogenous, for a given analysis. You can almost see the bee adapting to the flower, and the flower to the bee, and never getting lost in the tangled bank of communes growing every which way, to completely fill their little and disappearing niche.
Thus it seems to me that what organizational evolutionary studies need is an exemplar, as well as an encyclopedia, in order to become a dominant paradigm. Aldrich has provided the encyclopedia, and a fine encyclopedia it is. I propose Zablocki as the exemplar, to show us how to do network analysis, turnover analysis, survival determinants, varieties of niches in a given "industry," and the like; a wonderful example of what fine work looks like. It took him about a decade to do it right, and then unfortunately nobody noticed. I propose that, after giving Aldrich a careful reading, we turn to Zablocki's book and apply the standards and elegance of the exemplar to the important outstanding research questions that Aldrich raises in the concluding chapter of Organizations Evolving.
References
Fligstein, Neil (1990). The Transformation of Corporate Control. Cambridge, MA, Harvard University Press.
Podolny, Joel and Toby Stuart (1995). "A Role-Based Ecology of Technological Change." American Journal of Sociology 100(5): 1224-1260.
Zablocki, Benjamin David (1980). Alienation and Charisma: A Study of Contemporary American Communes. New York, Free Press.