A corporation’s culture as an impetus for spinoffs and a driving force of industry evolution

Abstract

We offer a theory of spinoffs that explains some salient aspects of these important market entrants. In infant industries, a great share of new market opportunities is depleted by firms that spinoff from incumbents. A model emphasizing the relation between incumbents’ evolving corporate cultures and the generation of spinoffs explains this regularity in industry evolution. By doing so, we capture different patterns in firm development that finally will help explain the evolutionary paths that industries may follow. We show that organizations reach a critical size that entails the collapse of a cooperative culture and triggers the exodus of personnel founding own firms. Thereby, organizations with a cooperative culture active in a dynamic business environment provide ideal training grounds for potential founders. Moreover, we argue that cooperative firm cultures and processes of “entrepreneurial imprinting” are important sources of spinoffs’ superior capabilities concerning their later market performance. We relate our findings to empirical evidence on developmental patterns in industries, such as genealogies and performance of spinoffs.

Appendices

Appendix A

The general form of logistic models is the following:

$$n^{\prime} =n+xn\left( 1-\frac{\sum\limits_{i} {n_{i} } }{K} \right),$$

(11)

where the growth rate of n is denoted by x. We can derive this growth rate form n′ given by Eq. 1 in the text:

$$\begin{array}{@{}rcl@{}} x&=&\frac{n^{\prime} -n}{n}=\frac{1}{n}\left({pn\left({1+r_{c} }\right)+n( {1-p} )\left({1+r_{o}( n )} \right)-n}\right), \notag \\ &=&p({1+r_{c} })+({1-p})\left({1+r_{o}(n)}\right)-1. \end{array}$$

(12)

which gives the growth rate used in Eq. 3.

Appendix B

In order to account for the effects of new personnel joining the firm and the necessary “renewal” of the socialization of existing employees, who are then considered as if they were personnel just joining the firm, we suppose that in each time step a cohort of n employees “retires” and is replaced by n new employees who are socialized by all n old employees, plus the entrepreneur. In addition, we assume that all new employees show neutral behavior when they join the firm. Moreover, these new firm members encounter other employees at random. With the help of the cultural transmission table below (Table 1), we specify the probability that a particular set of role models with different weights makes an individual acquire the behavioral variant c or o, given a changing group size.

Table 1 The probability of agents acquiring behavior c or o given a particular set of models (Entrepreneur/Leader, Peers) that have different total weights (A _E , A _P

The variable p measures the frequency of the c type in an infinite meta population of firms of size n. That is, for illustrative simplicity we are here modeling only the deterministic effect of evolutionary processes. In any given firm, stochastic effects will be important. However, in an infinite population of firms with particular characteristics, p will perfectly describe the average frequency of the cooperative behavior and ( 1 − p) the opportunistic behavior. Therefore, the average pairing probability of role models in the transmission table will have A _E + pnA _P probability of transmitting c to each new member of a cohort and probability ( 1 − p)nA _p of transmitting the behavioral variant o. Thus, in an infinite population of firms of size n, the partial recursion for the socialization phase with the frequency of c after transmission, p′, given that is was p before transmission, is expressed by Eq. 6 in the text.

Appendix C

We assume employees to choose an individual at random from the total number of a firm’s employees. Due to the direct bias μ _co alone, which captures the attractiveness of opportunistic behavior, agents are probabilistically more likely to adopt behavior o (0 ≤ μ _co ≤ 1) when they encounter it or to stick to it if they already behave opportunistically and meet a cooperative colleague. In addition, the conformist component, 2 p′ − 1, which depends on the frequency of behavior c in the firm, modifies the adoption probabilities as described in the text. Then, the probabilities of switching are given by Table 2.

Table 2 The probability of employees acquiring c or o given the behavior encountered

Using the probabilities of each possible pairing of “Self” and “Other”, we can calculate the frequency of behavior c after this kind of transmission process by multiplying the former by the different probabilities of switching to behavior c. We get the following recursion:

$$p^{\prime\prime} = {p^{\prime}}^{2} [1] + 2p^{\prime} (1 - p^{\prime}) \left[ \frac{1}{2} \left\{ 1 + (\eta (2p^{\prime} - 1) - (1- \eta) \mu_{co}) \right\} \right] + (1 - p^{\prime})^{2} [0]\,.$$

(13)

Simplifying gives (7) in the text.