Have populations of labor unions, semiconductor firms, and newspapers been subject to similar processes of density-dependent growth? In a general sense the answer appears to be yes. For each population, rates of mortality depended on density in essentially the same way. In all three studies the mortality rate fell with growing density to a point and then rose with further increases in density. Populations of dissimilar kinds of organizations have been subjected to similar processes of legitimation and competition, as can be seen by comparing Figures 11.2, 11.4, and 11.8.
Despite the importance of the consistency among the findings about density dependence in the three populations, differences in the strength of the effects deserve note. The effect of legitimation conveyed by growing numbers was quite similar for the populations of newspaper publishers semiconductor firms, and labor unions: the mortality rate dropped sharply with increasing density for all three populations. At the minimum, the implied mortality rate was 0.33 times as large as the zero density rate for newspaper publishers, 0.41 times as large as the zero density rate for unions, and 0.03 times as large as the zero density rate for semiconductor firms. Given the numerous differences among populations and the definitions of ending events, this strikes us as an important consistency.
On the other hand, the effect of competition on the mortality rate differed considerably among the three populations. Competitive pressures were apparently strongest for unions. When density grew from the point of minimum mortality, the implied mortality rate more than doubled in the population of unions. At this point it was about 60 percent as large as the zero density rate. For semiconductor firms the rate grew about seven-fold from its minimum when density reached its observed maximum. At this point it was only 22 percent of its zero-density level. But for the population of newspaper publishers, the rate grew only 13 percent as density grew from the point of minimum mortality to the historical maximum. Thus the competitive process seems to have been much stronger for labor unions and semiconductor manufacturing firms than for newspaper publishing firms. It is interesting to note that these findings do not agree with the view that nonmarket organizations do not face competitive pressures.
In analyzing the mortality of labor unions and exits of semiconductor firms, we also explored competition between subpopulations. We found strongly asymmetric competition in each case. For national labor unions, the density of craft unions had a strong positive effect on the disbanding rate of industrial unions, but the reverse was not the case. For semiconductor merchant producers, the density of subsidiary firms had a strong positive effect on the exit rate of independent firms, but the reverse was also not true. These analyses show some of the promise of our approach for analyzing competitive links between populations empirically.
Source: Hannan Michael T., Freeman John (1993), Organizational Ecology, Harvard University Press; Reprint edition.