Status Signals: A Sociological Study of Market Competition by Joel M. Podolny

Author:Joel M. Podolny
Language: eng
Format: mobi, pdf
Published: 2010-11-18T10:31:00+00:00

where r(t) is the transition rate or hazard of niche entry, h(t) is an unspecified baseline rate for the transition, X is a matrix of time-constant covariates, Y(t) is a matrix of time-varying covariates, and B and S are vectors of unknown regression parameters. Because h(t) is an unspecified step function, the Cox model offers an extremely flexible means for modeling time dependence. While the Cox model accounts for interarrival time dependence (i.e., the time from when the patent is granted to the time at first entry and then the time between subsequent entries) with an unspecified baseline rate for the transition, the time since last arrival is not the only form of time dependence that is likely to affect the rate of citation. In addition to modeling interarrival times, it seems appropriate to include two additional clocks that can accelerate the baseline rate. The first is a variable (updated monthly) to denote the calendar time. The second is a variable reflecting the age of the niche (updated monthly). Assuming that firms are more likely to enter a niche after they become aware of it, the composite baseline should increase as a function of the time since the patent at the center of the niche was introduced. However, because the relevance of a patent should decrease with the time since it was introduced, it is also important to include an age-squared variable as well to allow for nonmonotonicity.

Earlier in this chapter, the theoretical discussion of the properties of a technological niche emphasized three attributes: the quality of the focal invention in a niche, the structure of relations among the inventions in the niche, and the status of the actors associated with the niche. I shall now describe the operationalization of these attributes.

The quality of the focal invention. As has been the case in the analyses of the investment banking and wine industries, a critical analytical question is how to control for underlying quality. And, akin to the situation in the investment banking industry, the difficulty that industry observers have in inferring the comparative quality of inventions almost necessarily implies the absence of a completely adequate means to control for the quality of an invention. There are three possible ways to respond to this difficulty. One is to forgo any attempt to control for quality, under the assumption that what is not observable cannot have an effect on the likelihood of niche entry. A second approach is to treat quality differences across patents as unobserved heterogeneity and then to devise some method for controlling for this unobserved heterogeneity. A third possibility is to rely on one of the measures of quality that has been put forth in economic analyses of patents as indicators of innovative activity. In our analysis, we have chosen the second and third approaches, which both suggest the same control variable for quality.

In their discussion of unobserved heterogeneity, Heckman and Borjas (1980) noted that unobserved differences across units are likely to result in occurrence dependence. A frequently cited example of this type of unobserved heterogeneity arises in the context of research on job mobility.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.