orgtheory.net

modifcations of org. ecology

Fabio

One of my favorite hobbies is thinking about critiques, alternatives or modifications to organizational ecology theory. In an article I wrote on trends in economic sociology, I argued that in some cases, it might be better to drop the U-shaped mortality hypothesis and maybe think about the “L shape,” especially if you believe in standard industrial organization theory. Why? Low density = zero competition. Thus, part of the curve should be flat. Here are two other alternatives to current arguments about density and mortality:

  • Split population model – In a recent comment, Omar suggested that maybe there are two or more populations: more mature orgs that never die because they have all kinds of supports and less established groups. Example: colleges. Non-profits have tons of protections and never die (see research by Zajac and Kraatz) and for-profit colleges die all the time.
  • Dyanmic density effect models – Maybe the mortality curve changes over time. I can imagine that a “sunset” industry has a high mortality rate for nearly all types of firms, so the density doesn’t matter. Similarly, I could imagine during a boom, the density effects might be mitigated. You see a little of this in ecological papers that include period  dummy variables, but few attempts to tease of changing functional forms. Ideally, there might be some sort of differential equation version of the theory relating some feature of the market to the form of the mortality curve.

When I raise these ideas with some folks, they seem confused because they think ecological dynamics are a setteled issue. Perhaps, but if you read the fine print of ecology articles, you get tests vs. the null hypothesis, not tests among multiple models. Might be a good summer project for some graduate student. Add your own alternative density effects models in the comments.

Advertisements

Written by fabiorojas

April 25, 2007 at 4:36 am

Posted in fabio, mere empirics

3 Responses

Subscribe to comments with RSS.

  1. Martin Ruef’s “tolling bells” paper in Industrial and Corporate Change attempts to partially address the second issue (temporal heterogeneity of the effects of density) by incorporating population age directly into the density dependent model as a multiplier of density (p. 67, eqs 5, 6 and 7). Ruef finds clear evidence for his population of medical schools that the temporal heterogeneity model outperforms the standard density dependent model. The key question is whether a “general” temporal heterogeneity model can be specified that is as elegant and easy to formalize as the stylized fact of density dependence or whether the inclusion of time means that ecological theory has to become a “historical” theory tied to the peculiarities of each of the organizational populations that it studies.

    Like

    Omar

    April 25, 2007 at 2:01 pm

  2. Omar: “The key question is whether a “general” temporal heterogeneity model can be specified that is as elegant and easy to formalize as the stylized fact of density dependence or whether the inclusion of time means that ecological theory has to become a “historical” theory tied to the peculiarities of each of the organizational populations that it studies.”

    If so, then that would really take the wind out of the sails of the hard core ecology crowd. It’s trying oh-so hard to be a-historical, yet, history creeps in. Maybe we should arrange a meeting of ’65 Stinchcombe and Hannan ’00.

    Like

    Fabio Rojas

    April 25, 2007 at 2:48 pm

  3. […] Apr 25th, 2007 by Ryan Lanham modifcations of org. ecology […]

    Like


Comments are closed.

%d bloggers like this: