Sage Journals: Discover world-class research

Abstract

Creative destruction originates in successful development and launch of a radically different product/process/market configuration, or radical innovation, by a company. A radical innovation can materialize in a business firm only if autonomy is extended to organizational personnel, towards experimentation with heterogeneous knowledge from outside the organization. In this article, I inquire into conditions that enable the materialization of radical innovation—an important precursor to creative destruction—while affording top management the wherewithal to stay in control. An extended version of March’s computational model suggests that when a radical innovation project initiates with a low level of collective human capital with respect to the new process–product–market domain being considered and the project personnel are provided a certain extent of autonomy to experiment in the new domain, top management can detect a change to organizational outcomes by changing the rate of exploitation. Thereby top management obtains a feeling of being in control of proceedings. This situation is, therefore, conducive to the materialization of creative destruction. The indirect method of control to nurture radical innovation—demonstrated in this research—constitutes an improvement over the autonomy-averse ‘agency’ and ‘transaction cost economics’ approaches that are sub-optimal in open systems characterizing radical innovation efforts in corporate venturing.

Keywords

Autonomy collective human capital computational simulation control creative destruction environmental dynamism exploration and exploitation far-from-equilibrium conditions order creation

The fundamental impulse that sets and keeps the capitalist engine in motion comes from the new consumers, goods, the new methods of production or transportation, the new markets, the new forms of industrial organization that capitalist enterprise creates … (I)n (the) capitalist reality… [the] competition which counts [is]… from the new commodity, the new technology, the new source of supply, the new type of organization…competition which commands a decisive cost or quality advantage and which strikes not at the margins of the profits and the outputs of the existing firms but at their foundations and their very lives”. [This process of creative destruction] “…incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one.

—Schumpeter (1942/1994)

Introduction

Schumpeter (1942/1994) observes that the real challenge to the survival of a business firm rarely comes from the current set of competitors operating on today’s technology and serving the current needs of consumers. Rather, organizations risk mortality if they are unable to fashion radically new products and processes to serve the future needs of consumers. Schumpeter coined the term ‘creative destruction’ to designate the process of replacing old products, processes and ways of competing by new ones. We note that creative destruction is the outcome of successful launch of a ‘radical innovation’ by a business firm. Success is attributed when the innovative product (or service) attracts large numbers of consumers, perhaps at the expense of incumbent firms.

A radical innovation is a type of innovation that induces fundamental changes—constituting a clear departure from existing practices in the organization (Dominguez-Escrig et al., 2019)—having a high degree of novelty for the company that develops them (Crossan & Apaydin, 2010). Bringing about a radical innovation involves the development of entirely new goods or services that require substantial shifts in product lines, or major changes to the production process (Hall & Soskice, 2001; Witt & Jackson, 2016).

A radical innovation project involves a lot of experimentation and learning-by-doing (Govindarajan & Trimble, 2005; Nonaka, 2007; Nonaka & Takeuchi, 1995 ). Many failures—that are honest mistakes—are inevitable. For this reason, several scholars (e.g., Burgelman, 1983; Chanda & McKelvey, 2020; Levinthal & March, 1993) recommend providing a safety net guaranteeing non-punishment upon failure of innovation-related experimentation. The rationale is that, absent such safeguards, employees and middle managers would be loath to stick their necks out to work on radical innovation—leaving the firm in danger of demise in the wake of other firms’ successful innovations.

Executive tolerance of missteps in autonomous learning and experimentation activities indeed affords a company an opportunity to put up a strong show in the creative destruction race. But this arrangement effectively constitutes top management relinquishing control of the uses of a certain proportion of resources of the company. A credible guarantee is lacking that these resources will always be deployed for useful purposes and not squandered. Top management is unlikely to let autonomous activities continue for long, unless they possess some lever for control of proceedings of the innovation project.

In fact, company management is known to play a less passive role. From time to time, top management exercises its authority by shutting down some projects embodying autonomous strategic activity, and allowing others to continue (Burgelman & Grove 1996). A classic illustration is provided by the fortunes of a couple of corporate entrepreneurship projects in Intel, described in Burgelman (2002) and Burgelman and Grove (2007). In one case—concerning permitting continued manufacture of microprocessors at a time when Intel was predominantly a memory manufacturing company—the top management came out strongly in support of the entrepreneurial engineers behind the microprocessor project. Eventually, the company strategy was updated, and Intel became a microprocessor company in the early 1980s.

In a second case in the late eighties—concerning permitting continued development of processors with reduced instruction set computing (RISC) architecture—the top management shut down the activity. Intel chose to stay on course with its complex instruction set computing (CISC) architecture and went on to dominate the microprocessor market over the next three decades.

The instances above suggest that, in certain cases, senior managers have some reason to feel that they are in control of ongoing activity attempting to birth a radical innovation. These instances constitute fertile grounds for the successful materialization of radical innovation, and by extension, creative destruction. In other cases—where senior management lacks a feeling of being in control and withdraws autonomy to project personnel—creative destruction is likely to get stymied. In this research, I seek to unearth such conditions. It seems worthwhile to inquire: What would be an appropriate mechanism for control of radical innovation in a business firm? The question is important on two grounds.

First, the longer-term survival of a business is intricately linked to its ability to bring about radical innovations. In other words, a firm that fails to periodically bring about radical innovation stands the risk of demise when other firms launch successful radical innovation and take away its market share and revenue. Students of strategic management seek to forestall this negative outcome. This stems from the fact that, for the discipline of strategic management, an important imperative is finding ways and means to enable longer-term survival of the firm (Levinthal & March, 1993, p. 110).

Second, many business firms are handicapped in delivering radical innovation under dominant governance approaches such as agency theory (AT) (Eisenhardt, 1989; Jensen & Meckling, 1976) and transaction cost economics theory (TCE) (Williamson, 1981) that mandate strict monitoring and fulfilment of preset goals derived from past knowledge, through a regimen of sanctions and rewards. Delivering radical innovation becomes difficult or impossible because TCE and AT approaches fail to distinguish between honest mistakes and shirking or intentionally harmful behaviour (Chanda & McKelvey, 2020; Shimizu, 2012) and (therefore) recommend prompt punishment for any deviation from pre-agreed objectives. In contrast, in a radical innovation project, objectives evolve over time as management and personnel gain more knowledge; administering the organization on objectives set on the basis of expired past knowledge is inappropriate (Chatterjee et al., 2018). Hence, businesses can benefit if a mechanism can be found whereby senior managers may exercise control by means other than directly monitoring the actions of organizational members for alignment with preset goals. Thereby autonomy can get extended, enabling far-ranging innovation necessary for creative destruction. Effectively, the research question noted above endeavours to find enablers to the process of creative destruction by seeking enablers to radical innovation.

In this study, I set out to explain under what circumstances ‘creative destruction’ gets supported and otherwise. The theory is based on the computational simulation model in March (1991). March (1991) studied the variation in the level of organizational knowledge (i.e., knowledge that is stored in physical artefacts such as databases, routines, forms and procedures in an organization) attained by varying the rate of learning by organizational members. The model shows the interplay of exploitation (refinement, modelled by rate of member learning) and exploration (experimenting with internal and/or external variety, modelled by enabling the persistence of diverse knowledge in members). In all his experiments, the level of initial collective human capital (CHC) (i.e., knowledge residing in the heads of organizational personnel at the time of commencement of simulation experiments) was the same. CHC refers to the human capital residing in organizational members (Chanda et al., 2018; von Nordenflycht, 2011). Human capital comprises ‘… the knowledge, skills, and abilities a firm’s workers acquire as a result of their learning, education, training, and experience’ (Becker, 1964, quoted in Lee et al., 2015, p. 796).

In a subsequent development, Chanda and McKelvey (2020) show that when an innovation task initiates with no significant misalignment of the CHC to the task at hand (i.e., involve the same moderate level of initial CHC as in the experiments in March, 1991), top management finds it hard to detect changes to organizational knowledge upon changing the rate of exploitation, under conditions where the lower ranks have autonomy. In conditions where autonomy is absent, top management is in a position to readily detect a change in organizational knowledge upon a change in the rate of exploitation. For this reason, top management is likely to be reluctant to extend autonomy. This suggests that radical innovation having the potential to bring about creative destruction is unlikely to materialize under these circumstances.

In this study, I alter March’s model to allow variation of the initial level of CHC in the innovation organization. Specifically, I present outcomes for a situation where a radical innovation task initiates with CHC far out of alignment with the knowledge required for radical innovation, i.e., under conditions of lower CHC (Chanda et al., 2018) than that in the experiments in March (1991). I find that when a certain degree of autonomy is extended to middle- and lower-level employees, relatively large changes in organizational knowledge can be obtained simply by changing the rate of exploitation. Because of this, top management is able to use the rate of exploitation as a lever for control. This finding suggests that management should foster conditions of autonomy. Successful materialization of radical innovation—and, by extension, creative destruction—is feasible under these circumstances.

The elaboration of March’s agent-based simulation model enriches the discussion about the role of exploration and exploitation in organizational outcomes by drawing attention to the crucial roles of autonomy to organizational personnel and control by the top management in the assimilation of heterogeneous new knowledge distinct from that embodied in the CHC of the organization. The simulation findings suggest that ‘creative destruction’ has a greater chance of effective materialization when an innovation organization’s initial human capital is insufficient to cope with demands of the innovation task. Conversely, heightened top management control—in conjunction with denial of autonomy for lower-level employees—is more likely when an innovation organization’s CHC is well-aligned to an innovation task, thwarting creative destruction.

Theoretical Foundations

Radical Innovation as a Precursor to Creative Destruction

A radical innovation by one company changes the technology of a process or product in a way that imposes requirements that the existing resources, skills and knowledge of other companies satisfy poorly or not at all (Abernathy & Clark, 1985). A radical innovation may involve fundamental changes that represent revolutionary modifications in technology (Dewar & Dutton 1986) and serve as the basis for further technical developments (Datta & Jessup 2013). Thereby, radical innovation engenders a redefinition of what is required to achieve a competitive advantage. In cases where the disruption is both deep and extensive, such innovation creates new industries (Abernathy & Clark, 1985). In sum, radical innovation is a necessary precursor to creative destruction. In this research, I focus on what it takes to bring about the materialization of a radical innovation in a business firm.

Difficulties in Bringing about a Radical Innovation

Birthing of a radical innovation in a business firm is a notoriously difficult task. On the one hand, creative people have to be provided autonomy to experiment. On the other hand, top management cannot simply stay put, firm in the belief that such autonomy will be put to good use. Joseph Goldstein, a former vice-president of the RAND Corporation, notes that: ‘The administrator has not only the difficult task of administering research, but also the equally difficult and important job of refraining from administering’ (Goldstein, 1961, p. 4, quoted in Augier et al., 2015, p. 1147). The issue becomes intriguingly difficult even if the flowering of creativity from intellectuals can somehow be fostered: ‘… the flaring of intellectual outliers involves the stimulation of intellectual variety, but successful variation inhibits the further generation of variety. Successful intellectual outliers become the basis of a new intellectual orthodoxy’ (Augier et al., 2015, p. 1152). An endeavour to find a way through these difficulties leads us to inquire into the interplay of autonomy and control.

Control and Autonomy

Drawing from Child (1973) and Tannenbaum (1968), Brenner and Ambos (2013) suggest that ‘control’ may be defined as any process (mechanism, instrument or strategy) applied by an organization to ensure the execution of organizational goals and plans. Three broad classes of control mechanisms used by firms comprise outcome, process and social control (Cardinal et al., 2010; Harzing, 1999; Kirsch, 1996, 1997; Snell, 1992; Turner & Makhija, 2006). When managers use ‘outcome controls’, they merely specify the level of performance expected; they do not lay down rigorous processes regarding how outcomes are to be achieved. ‘Process controls’ are ‘mechanisms that clearly specify the appropriate behaviors and processes in which employees must engage’ (Turner & Makhija, 2006, p. 207). Process controls are ideal when the level of outcome that is attained is not fully in control of the employee; rather, external entities and factors have a big role in determining the final level of the outcome. Finally, ‘social controls’ operate through normative pressure to ensure social obligation and facilitate the sharing of values among organizational members (Lange, 2008).

Several streams of literature discuss autonomy and control as a dichotomous condition (see, e.g., Crilly & Sloan, 2014). Upon inspection of this point of view, it appears that any measure of autonomy gets diminished when some form of control is imposed from on high. However, when nuances regarding a particular kind of control are considered, it may be possible to envisage providing autonomy to employees in aspects orthogonal to the ambit of the matter controlled. Below we take a brief tour of some flavours of control found in the neighbourhood of management literature. Thereafter, I discuss the kinds of autonomy and control suited for materialization of radical innovation that potentially leads to the onset of creative destruction.

The ‘Accounting’ literature prescribes constant monitoring of variances in budget in order to detect deviations from plan and to initiate corrective action. This approach, however, places an unconscionable degree of confidence on estimates concerning potential return on investment, and horizons of inflows and outflows of cash. Estimated numbers are often guesstimates, fashioned from vastly incomplete information. Radical innovations of the nature envisaged for creative destruction rarely entail cash flow characteristics that can be foreseen with the level of accuracy demanded by accounting systems. Therefore, accounting controls accomplish precious little in helping top management obtain a sense of being in control when they allow autonomous functioning encompassing experimentation with heterogeneous new knowledge from outside the organization, for the creation of new knowledge.

Literature originating in ‘Economics’—for example, ‘agency theory’ and ‘transaction cost economics’ theory—recommends that employees be committed to goals ahead of time in as many aspects as can be metered in a cost-effective manner (Jensen & Meckling, 1976; Williamson, 1981). Additionally, constant surveillance of employees and credible punishment upon unfavourable performance variance is deemed necessary, such that situations do not go out of control. These approaches assume that past knowledge can reliably inform on prospect of future success. They are inappropriate for governing situations involving learning by experimentation for the purpose of the creation of knowledge (Chatterjee et al., 2018).

The literature on ‘corporate entrepreneurship’ (e.g., Burgelman, 1983) advises senior managers to tolerate a certain extent of autonomous strategic activity championed by middle managers. To cultivate autonomous strategic activity, the senior management of a company allows a certain proportion of resources to be deployed in projects that do not flow from mandates from the top. An oft-quoted example in this regard is Google Inc. allowing its technical personnel to spend a certain proportion of time on any idiosyncratic project (Edelman & Eisenmann, 2011). Top management extends tolerance because it is aware of the potential of beneficial discoveries in improving the fortunes of the company—viz., getting ahead in the creative destruction race. When discoveries meet with market success, top management updates the company strategy. A question remains though: What provides the top management a sense of being in control, while simultaneously being open to extending autonomy to the lower ranks for experimentation with new knowledge, and what makes top management decide to pull the plug on an innovation effort?

Why Are Radical Innovation Activities Less Amenable to Conventional Ways of Control?

In order to develop new knowledge necessary for bringing about radical innovation, organizational personnel need to experiment repeatedly. Prior to experimentation, it is often not possible to predict whether an experiment will be fruitful and worthwhile for an organization, or otherwise. Several experiments will yield no tangible gain to the organization. Yet they must be carried out. In this context, Thomas Alva Edison (reportedly) remarked that it is the learning from a large number of failures that contributed to his discovery of a functioning filament of the incandescent lamp.

The matter is somewhat more complicated in an organizational situation. Single-actor models as in Edison’s story do not pose problems of monitoring and control, since rewards are contingent to the appropriateness of the quality and degree of effort exerted. In contrast, in a business firm, there is a division of labour. Certain personnel carry out experiments in a bid to fashion innovative products and services. Other personnel—top management—evaluate their contribution.

In any experiment to develop knowledge, failure can originate from two important causal mechanisms. In one case, competent organizational personnel expend honest effort. Yet, a useful solution or breakthrough remains elusive. In another situation, failure is an outcome of either incompetence or shirking or both, by the experimenter(s). The twin possibilities of the origin of failures in learning and knowledge development efforts create an evaluation problem for the top management of an organization.

Top management’s job is to sanction failures arising from incompetence and/or shirking. In the role of a nurturer of a company, they should hold-off from punishing failures that are unavoidable when learning through experimentation. Burgelman (1983, p. 1362) and Levinthal and March (1993, p. 107) advise that top management should provide a ‘safety net’ such that failures in autonomous experimentation are not sanctioned. Otherwise, organizational personnel shall cease to experiment with new knowledge, effectively eliminating the company from the creative destruction race. Thus, in order to nurture efforts involving radical innovation, top management needs to extend autonomy to organizational personnel. In sum, an important imperative in organizational life concerns finding ways and means by which top management can retain a sense of control, even when autonomy is extended to employees and middle managers to experiment with new knowledge, in a bid to bring about a radical innovation.

Towards a Model for Management Control of Radical Innovation

At its essence, a radical innovation takes shape when the CHC of members of an organization, augmented with the learning of heterogeneous knowledge from outside the organization, facilitates the development of organizational knowledge necessary for the materialization of the innovation. CHC corresponds to the sum or average of individual expertise (Barrick et al., 1998; Chan, 1998; Chanda et al., 2018; DeShon et al., 2004, referred to in Pil & Leana, 2009, pp. 1108–1109) of members involved in a radical innovation project. ‘Organizational knowledge’ comprises databases, user manuals, rules, forms, procedures, etc., created by the project team. It is stored in a repository, the ‘organizational code’ (March, 1991). Some examples of organizational knowledge are knowhow regarding sourcing, manufacturing, storage and distribution, standard operating procedure between sub-teams, recipes for inducting and training members, etc. In the computational simulation model (Chanda, 2017; Chanda & McKelvey, 2020; Chanda & Miller, 2019; Chanda & Ray, 2015; Chanda et al., 2018; Gong et al., 2021a, 2021b; March, 1991; Wu et al. 2021a, 2021b), the stock of organizational knowledge is augmented by drawing from the stock of CHC of members. Conversely, the stock of CHC of members is altered when members learn from the organizational code. Importantly, when members are provided autonomy to obtain knowledge from outside the organization, there is an additional flow of heterogeneous knowledge into the stock of CHC of members.

An innovation project either involves developing knowledge that is quite similar to or well-aligned with the initial CHC of members, or it may involve knowledge that is substantially non-aligned with the initial CHC of members. Dierickx and Cool (1989) posit that joint consideration of stocks and flows is necessary to develop adequate theory regarding phenomena. Thus, when an important flow—the rate of members learning from the organizational code (i.e., the rate of ‘exploitation’, as elaborated below)—is varied, differences in the initial level of the stock of CHC may lead to distinctive outcomes. This, in turn, may influence sustained support or abandonment of a radical innovation project.

A key contention of this article is that the top management can use the rate of exploitation as a lever of control to assess change in the extent of new organizational knowledge created by a team engaging in radical innovation. Top management is quite likely to let certain arrangements involving autonomy continue, if they are able to detect significant changes in organizational knowledge levels upon changing the rate of exploitation. Below, I describe the connotations of changing the rate of exploitation and observing the nature of outcomes involved, in various phases of an internal corporate venture.

Increasing the rate of exploitation in an innovation project effectively translates to enhancing the use of existing knowledge and tools to assimilate heterogeneous external knowledge in a bid to fashion novel products or services. Actions towards upping the exploitation rate and checking for enhancement of organizational knowledge give momentum to an idea to move from an abstract concept to a prototype, thence to a model with demonstrable functionality and thereafter to a potentially manufacturable product (or deployable service) for which there is interest from prospective customers, suppliers and distribution partners.

In the early stages of an innovation project, additions to organizational knowledge can be inferred when, at each milestone meeting, assessments are carried out regarding how many assumptions made earlier were found to be valid, and how many were wrong, necessitating framing of fresh assumptions (Govindarajan & Trimble, 2005), and updating of timelines. At the intermediate stages of the project, the extent of newness of functionality and features (with respect to those in the firm’s current offering) provides indication regarding augmentation to organizational knowledge. At advanced stages of the project, evidence regarding the extent of interest from external parties—e.g., prospective customers, potential upstream (supply-side) and downstream (distribution-side) partners, etc.—and evidence regarding manufacturability to scale can indicate further significant augmentation to organizational knowledge.

Top management is quite likely to let an existing organizational arrangement continue, if they find a significant extent of change in organizational knowledge levels upon modulating the rate of exploitation. We note, though, that evaluations for significant changes in organizational knowledge make sense only after a project has been in progress for a certain amount of time, i.e., organizations need to ‘… buffer internal ventures from an immediate market test’ (Kogut & Zander, 1992, p. 393). Hence, top management will need to exhibit a certain amount of ‘strategic patience’ (Iyer & Davenport, 2008; König et al., 2013)—by holding off from evaluating the project for a certain length of time—before making a definitive decision about letting the project continue or stop it.

We note that using the efficacy of varying the rate of exploitation for evaluation represents a departure from measuring progress primarily by financial metrics. Financial measures may be unavailable altogether, since products are mere ideas, manufacturing costs are indeterminable pending further learning, and there is considerable uncertainty regarding prospective customers and their willingness to pay (Govindarajan & Trimble, 2005). Using the rate of exploitation as a strategic control tool helps develop additional insight into the progress of an innovation project, because goals change and new assumptions replace prior ones in successive milestone meetings, a process Argyris (1976) cites as ‘double-loop learning’. We note that replacing financial metrics with metrics suited to double-loop learning requires setting aside ‘strategic rigidity’—connoting a preference for the status quo style and approach (Shimizu & Hitt, 2004).

Simulation Model

The simulation model follows the conceptual specifications (Chanda & Miler, 2019) provided by March (1991). In all, I make two modifications.

For a subset of experiments, I lower the level of the initial CHC in the organization (relative to the level of the initial CHC applicable in March’s experiments) following the method provided by Chanda et al. (2018).

Second, in order to denote the presence of autonomy whereby organizational members are allowed to experiment using heterogeneous knowledge from outside the organization, I allow randomization of a subset of belief dimensions of a subset of the organizational members, following a specification suggested in Chanda and McKelvey (2020). In March (1991), the inflow of heterogeneous knowledge from outside the organization was modelled by allowing randomization of the entire belief strings of a subset of members in any given time step. I use the specification from Chanda and McKelvey (2020) in order to make the model of learning from external sources a bit realistic, viz., by invoking a lower rate of infusion of heterogeneous knowledge from outside the organization. In the paragraphs that follow, I provide details of the simulation model.

The Environment and the Project Organization Attempting Radical Innovation

In the simulation model, the external reality (R), representing the environment, is an M-bit string. At the beginning of a simulation experiment, each bit (or dimension) of R is given a random value from the set S1 = {‘−1’, ‘+1’}. The elements of the set S1 are code numbers for different applicable values of knowledge for a given dimension of knowledge. As a hypothetical example, in the case of a new product development scenario, desirability of establishing company-owned distribution channels is signified by a value of ‘+1’ in a particular knowledge dimension, and desirability of a franchisee model is signified by a value of ‘−1’ in the same knowledge dimension of R. The reality string R is generated at the beginning of a simulation experiment, such that each element of S1 has a 50% chance of appearing in a given bit position.

A simulation experiment is carried out for T periods (time steps). In a stable environment, the values in R remain the same throughout the T periods of a simulation experiment. If a simulation experiment is conducted in a dynamic environment, in any time step of the simulation experiment each bit of R has a certain probability for flipping (from ‘+1’ to ‘−1’ or vice-versa), given by a parameter p₄.

The project organization attempting radical innovation is constituted of N members and one entity, the organizational code OC. OC is a repository of organizational knowledge developed during the course of a radical innovation project. OC is an M-bit string that has all values set to ‘0’ at the beginning of each simulation experiment. The element value ‘0’ signifies ‘no opinion’ regarding the value of the corresponding knowledge dimension. Thus, OC has zero knowledge in the beginning.

Each organizational member possesses an M-bit belief string. To construct a Marchian population, that is, a population with moderate CHC, the belief string of each member is given a random value from the set S2 = {‘−1’, ‘+1’, ‘0’}, at the beginning of each simulation experiment. Each element of set S2 has a one-third probability of appearing in any member’s belief string. Separate draws are conducted to generate belief strings for N members of the organization. Thus, on average, in a population with moderate CHC, one-third of the belief dimensions of a member contain correct knowledge, and another one-third contain incorrect knowledge, that is, the ratio of incorrect: correct beliefs is 1:1.

To construct a sub-Marchian population (i.e., a population with a low level of CHC with respect to a task at hand) with deficiency D per cent, first a Marchian population is generated. Thereafter, D per cent of member bits are randomly selected. The values in the selected bit positions are overwritten with values opposite that of the value in the corresponding bit positions of the reality string R. For example, if the seventh bit position of a member is chosen for update, the value in this position will be overwritten with a value of ‘+1’ if the seventh bit in R is ‘−1’, else it is overwritten with a value of ‘−1’. Further, if we use D = 15%, in a population with a low level of CHC with respect to the radical innovation attempted, the proportion of incorrect beliefs:correct beliefs is 1.5:1.

Learning by the Organizational Code

The organizational code accumulates knowledge by learning from knowledgeable members at a rate given by a parameter p₂. The organizational code is assumed to have the ability to identify members who are more knowledgeable about the external reality R than itself. In each period, the organizational code identifies the set of such elite members. For every bit position, the organizational code takes a poll among the elites, to determine how many recommend the value to be ‘+1’ and how many recommend the value to be ‘−1’. If more elites recommend a non-zero value that is different from the value present in the organizational code, the probability that the organizational code value remains unchanged is given by (1 – p₂)^k, where k (> 0) is the difference between the number of elite members who recommend a change, and the number of elite members who do not recommend a change. The elite members who have ‘0’ value for a given bit position do not contribute to the decision-making of the organizational code, for that particular bit position. In the computer program, learning by the organizational code takes place based on the values of members’ beliefs in the immediate prior period. It follows that organizational code knowledge values commence undergoing change from the second period onwards.

Learning by Organizational Members

In each time step, members learn from the instance of the organizational code that existed at the immediate prior period. For each bit position in the belief string of a member, the corresponding value in the organizational code is read. If the organizational code has a non-zero value that is different from the value in the member’s belief string, the member updates his/her value to the value from the organizational code with a probability p₁. The parameter p₁ represents the rate of member learning. Alternately, I denote an organization as carrying out exploitation at a high rate when p₁ is set to a value 0.90. Otherwise, when p₁ is set to a value of 0.10, I denote the organization as carrying out exploitation at a low rate.

Autonomy

In situations where autonomy is permitted, there is one additional parameter, p₅, signifying the proportion of belief dimensions of a subset (about one-fourth) of organizational members that are randomized as a result of acquiring heterogeneous knowledge from outside the organization. Randomization of a belief-bit is carried out by assigning a value from the set S2 = {‘−1’, ‘+1’, ‘0’} with each element of S2 having a one-third chance of materializing. Absence of autonomy is implemented by setting p₅ to a value of zero.

Model Parameters

A given simulation experiment is repeated 10,000 times and the outcomes reported are averages of the experimental runs. On each iteration, the values in the reality string R and the values in the knowledge strings of organizational members are regenerated by means of fresh random draws. In Table 1, I present the parameter values used in experiments. These values are by and large the same as used in March (1991). Robustness checks with some neighbouring values for each parameter confirm that the qualitative results are consistent.

Table 1.

Parameters in the Experiments.

Parameter	Description	Value in Experiments	Robustness Check
M	The number of bits of (i) reality string, (ii) organizational code string and (iii) member belief string	30	25, 35
N	The number of members in the organization attempting radical innovation	50	40, 60
p ₁	Member learning rate, low & high rate of exploitation	0.10 0.90	0.15, 0.25 0.75, 0.85
p ₂	The rate of learning by the organizational code	0.50	0.40, 0.60
p ₄	The probability that a bit of reality flips at any time step. A non-zero (positive) value signifies environmental turbulence	0.0025 to 0.02	N/A
p ₅	The proportion of belief dimensions that are randomized, for a subset (25%) of organizational members, when autonomy is granted to imbibe heterogeneous knowledge from outside the organization	0.10	0.08, 0.12
Collective Human Capital (CHC)	CHC is given by the set of beliefs of organizational members. CHC is moderate (Marchian CHC) if all beliefs are random draw from {−1, 0, +1}, i.e., deficiency D = 0, at the beginning of a simulation experiment. CHC is low (sub-Marchian CHC) if certain percentage of member beliefs— originally created by random draw—are overwritten with values opposite that of values in corresponding bit positions of the initial reality string, i.e., D > 0, at the beginning of a simulation experiment	0 0.15	N/A 0.14, 0.16

Results

In all the graphical results, the outcome variable comprises the absolute difference in organizational knowledge developed by low versus high exploitation rates. Further, I adopt a convention that if this difference is greater than 10%, top management can readily detect change in organizational outcomes upon changing the rate of exploitation and thereby obtain a feeling of being in control of proceedings in the innovation organization. The qualifying limit of 10% is shown by a horizontal arrow in the graphs presented. When the difference is less than 10%, top management will not be in a position to use varying the rate of exploitation as a lever for assessing control on the innovation project. Absent a feeling of being in control, top management is quite likely to intervene in an adverse manner, for example by withdrawing autonomy or by outrightly shutting down the project.

Further, an innovation organization working with ‘Marchian CHC’ connotes that the innovation project calls for CHC that is close to the existing knowledge base of a company. The instances where the innovation project requires CHC that is vastly different from the CHC currently present in a company are designated as working with ‘sub-Marchian CHC’. I first inquire into the probability of materialization of creative destruction in a stable environment (Figures 1 and 2). Thereafter I consider environments of varying levels of turbulence over the medium and long term (Figures 3 and 4).

Prospects for Materialization of Radical Innovation, Stable Environment

Figure 1 corresponds to a situation where the environment is stable (p₄ = 0) and autonomy is not provided to the lower ranks (p₅ = 0). We observe that the 10% mark for discernibility of outcomes of low versus high exploitation is met for innovation organizations with Marchian CHC. In this case, top management will, quite likely, let existing arrangements continue. In other words, while it is unlikely that top management will approve proposals for more autonomy to the ranks, it is also unlikely that they will shut down an initiative abruptly. Innovation delivered is likely to be incremental in nature, since the innovation organization works on a project that is aligned with the kind of CHC that is already existing in the company. Thus, we have:

Proposition 1. Incremental innovation, rather than radical innovation, is the most likely outcome when an innovation project involves CHC close to that already existing in the organization, and autonomy is not extended to the lower ranks.

Figure 1 also shows that innovation projects working with sub-Marchian CHC fail to afford a sense of being in control to top management, given that the 10% mark is not attained. These innovation projects, therefore, are unlikely to survive top management scrutiny for long, i.e., they tend to get shut down without producing a tangible outcome. Thus, we have:

Proposition 2. Radical innovation is unlikely to materialize from an innovation project involving CHC vastly misaligned to the CHC already existing in the organization, where the lower ranks are not extended autonomy.

Figure 1.

Parameters. M = 30, N = 50, p₂ = 0.50, p₁ = 0.90 for high exploitation, p₁ = 0.10 for low exploitation, p₅ = 0, p₄ = 0.

Figure 2 corresponds to a situation where the environment is stable (p₄ = 0) and autonomy is provided to the lower ranks (p₅ > 0). We observe that projects operating with Marchian CHC fail to attain the 10% mark necessary for discernibility by top management for getting a sense of being in control. These projects are, therefore, likely to be shut down. Thus,

Proposition 3. Radical innovation is unlikely to materialize when an innovation project involves CHC close to that already existing in the organization, and autonomy is extended to the lower ranks.

Figure 2.

Parameters. M = 30, N = 50, p₂ = 0.50, p₁ = 0.90 for high exploitation, p₁ = 0.10 for low exploitation, p₅ = 0.10, p₄ = 0.

Figure 2 further shows that the curve sub-Marchian CHC breaches the 10% mark for discernibility after the initial 20–25 periods and stays above the line for the entire duration of observation (100 periods). Hence, top management can readily discern change in organizational knowledge by varying the rate of exploitation and thereby obtain a sense of being in control. In such cases, top management is less likely to withdraw autonomy or shut down the innovation project abruptly. Thus, we have:

Proposition 4. Radical innovation has a good chance of materializing, when innovation project involves CHC vastly misaligned to the CHC already existing in the organization and the lower ranks are extended autonomy.

Prospects for Materialization of Radical Innovation, Turbulent Environment

Figure 3 shows outcomes corresponding to absolute differences in organizational knowledge developed by low versus high exploitation, under conditions of varying environmental turbulence (p₄ > 0), over the medium term (50 time steps), for the four archetypes consideration: innovation projects involving Marchian versus sub-Marchian CHC, in the presence and absence of autonomy. Figure 4 shows the corresponding outcomes over the longer term (100 time steps). One time step may be considered to signify one month or one quarter, depending on the context of a problem. A low value of environmental turbulence (p₄ = 0.0025) signifies that over a longer-term observation period of 100 time steps, the external reality—corresponding to the radical innovation attempted—changes to the extent of 25%. Moderately high environmental turbulence (p₄ = 0.0075) signifies that the external reality changes to the extent of 75% in a similar period. Very high turbulence (p₄ = 0.02) signifies that the external reality changes to the extent of 200% in 100 time steps. In practice, one rarely encounters this high level of turbulence for an extended period. Over a medium term (50 time steps) the corresponding values of the extent of change of the external reality are about 50% lower (i.e., 12.5%, 37.5% and 100% extent of change in the external reality, for low, moderately high and very high turbulence, respectively).

Figure 3.

Parameters. M = 30, N = 50, T = 50, p₂ = 0.50, p₁ = 0.90 for high exploitation, p₁ = 0.10 for low exploitation, p₅ = 0 for absence of autonomy, p₅ = 0.10 for presence of autonomy.

Figure 4.

Parameters. M = 30, N = 50, T = 100, p₂ = 0.50, p₁ = 0.90 for high exploitation, p₁ = 0.10 for low exploitation, p₅ = 0 for absence of autonomy, p₅ = 0.10 for presence of autonomy.

In Figure 3, we observe that, for innovation projects with sub-Marchian CHC functioning in the presence of autonomy, the 10% mark for discernibility of outcomes of low versus high exploitation is met for environmental turbulence varying from low (p₄ = 0.0025) to very high (p₄ = 0.02) in the medium term (T = 50 periods). Over the longer term (T = 100 periods) though, we see from Figure 4 that the discernibility criterion is not met beyond a moderately turbulent environment (p₄ = 0.0075). Thus, we have:

Proposition 5. Radical innovation has a good chance of materializing, when an innovation project involves CHC vastly misaligned to the CHC already existing in the organization, and the lower ranks are extended autonomy, under conditions of low to moderately high environmental turbulence.

Summary of Results

In Table 2, I provide a summary of the propositions developed, in the form of a 2 × 2 matrix. These pertain to a situation where a company is functioning in a stable environment, or in an environment characterized by low to moderately high turbulence.

Table 2.

Prospect of Materialization of Radical Innovation, Stable Environment and Environments Characterized by Low to Moderately High Turbulence.

	Innovation project involves CHC close to that existing in the company	Innovation project involves CHC that is vastly misaligned to the CHC existing in the company
Lower ranks are provided autonomy	Radical innovation is unlikely to materialize	Materialization of radical innovation is probable
Autonomy is denied to the lower ranks	Radical innovation is unlikely to materialize	Radical innovation is unlikely to materialize

Discussion

The study demonstrates that, when top management extends autonomy to organizational personnel, it is afforded a lever for control in the form of modulating the rate of exploitation only when the innovation project involves CHC vastly different from the CHC already existing in the organization, and not otherwise. Top management, due to its preference for being in control, should extend autonomy in the former case and withhold autonomy in the latter case. The circumstances under which autonomy is feasible constitute conditions suitable for the fructification of radical innovation, a key precursor to materialization of ‘creative destruction’. Likewise, the circumstances under which autonomy is not feasible will, very likely, not see materialization of creative destruction.

We noted that extending autonomy comes at a price—the prospect of opportunistic behaviour by beneficiaries (Shimizu, 2012). Hence, it is likely that the top management of an organization will be reluctant to extend autonomy, unless it has some means for control. In this regard, the top management faces a dilemma in choosing which judgment criteria to use. Criteria fashioned on extant knowledge in the organization may be of little use when an organization sets out to explore domains far from those served by its current capabilities. Irrelevant or idiosyncratic criteria are of no use in effecting control either. In the parsimonious model of March (1991), we are left with just one measure—varying the rate of exploitation—to serve as a lever of control. Indeed, we find that changes upon operation of the lever are readily detectible under autonomy under one condition—when an organization sets out to explore with a low level of human capital with regard to the needs of the innovation project. When an organization explores a domain near to that served by its current capabilities, autonomy has a much lesser chance of being supported or encouraged, given that top management is not afforded a lever of control.

In effect, an intriguing insight from the modelling effort develops as follows. Attempting to do something for which a company has meagre CHC is inherently risky. Exploration of domains far from an organization’s current suite of capabilities is likely to engender a high probability of failure. This is because an organization may, inappropriately, attempt to apply knowledge templates governing existing businesses to the new situation. Such need, in turn, arises from the necessity of top management to exert control on the fledgling initiative. However, if a company can set aside strategic myopia—unwillingness to examine the possibility of extending the corporate strategy to domains far removed from that involved in a company’s current strategy (Tripsas & Gavetti, 2000)—and set out on a path of radical innovation in a domain where personnel have low CHC relative to the needs of the task at hand, conditions can be supportive: top management can have a sense of being in control (given that they can observe change in organizational outcomes upon change in the rate of exploitation), even when they extend a certain extent of autonomy to the lower ranks.

We also note that an organization has a high probability of obtaining success by exploring the neighbourhood of their current capabilities. This is because organizational personnel are in a position to quickly identify similarities and differences between what they know and what is needed for the new effort. This reduces the complexity involved in assimilating new knowledge. Yet, autonomy to the lower ranks, if any, is likely to be short-lived, because, otherwise, if autonomous strategic behaviour is permitted, the outcomes from low and high rates of exploitation are not distinguishable, depriving senior management of a lever for control. The senior management will have a better hang of the strategic significance of their current business efforts and the dangers of drifting away or replacing their current business with another business. They will, most likely, shutter the innovation project, or withdraw autonomy altogether.

Limitations of the Study

This study inquires into conditions within a business organization that are likely to be favourable for giving birth to radical innovation—an important precursor to creative destruction. However, external conditions frequently have a role to play in determining whether the business environment is supportive of fostering radical innovations. The latter is out of scope of this study. Interested readers may consult Stam (2018) to appreciate the impact of industrial policy in fashioning a supportive environment for the materialization of creative destruction. Second, developing sufficient organizational knowledge to bring about a radically innovative product is necessary, but not sufficient for coming into being of creative destruction. The ultimate commercial performance of incumbents and new entrants is determined by the balance and interaction among: (i) investment in developing the new technology, (ii) technical capabilities and (iii) the ability to appropriate the benefits of technological innovation through specialized complementary assets (Tripsas, 1997).

Third, it has been assumed that top management has the wherewithal to alter the rate of exploitation in an innovation organization without incurring prohibitive costs and without derailing the innovation project (as can happen, say, when key personnel leave out of disappointment from insufficient top management support). Fourth, it is assumed that top management can measure in near real-time, the outcomes of its actions (actions pertaining to varying the rate of exploitation), i.e., delays in feedback on action are not vastly significant. Future research may benefit from considering and controlling for these contingencies.

Some Implications

Governance of a business firm by TCE or AT approaches fails to support risky experimentation necessary for bringing out radical innovations. The research presented here suggests an alternative governance approach—whereby top management can extend autonomy to middle- and lower-level personnel and still be in control of the progress of the innovation effort—and highlights the boundary conditions of validity of the approach.

In an internal corporate venturing project, millions of dollars may get spent upfront, with weak or questionable contribution to the progress of the strategic initiative (Chanda & Ray, 2016, 2021; Keil & Mähring, 2010; Montealegre & Keil, 2000). Often, this happens due to paucity of adequate oversight by top management, on account of their lack of familiarity with the novel developments. The research presented here suggests a mechanism whereby top management can remain in control of a corporate venture, even when the extent of novelty is high. Application of the insights developed in this research has significant potential to lower the failure rate of strategic projects involving radical innovation.

The study further suggests that when managers consider shutting down an initiative, they should introspect whether they are doing it because they are uncomfortable with not having control, or because of its poor merit. It is not easy to gauge the merit of a fledgling initiative. Errors can crop up if the judgment criteria—fashioned in some prior context—do not fully apply to the present situation. Organizationally too, top management may wish to avoid infringing existing commitments—to products, to technology, as well as to arrangements with suppliers and customers. Thus, discomfort with a radical innovation initiative may arise from sources other than the lack of a lever for control, namely, political actions by internal and external stakeholders to protect their turf. Departmental heads may, for instance, protest that they are held to demanding standards for financial returns, while the innovation initiative gets all the funding and attention without generating a comparable level or return. The advice from Govindarajan and Trimble (2005) is salient here: A new venture initiative should not be controlled by financial measures (e.g., return on investment); rather, the focus should be on how soon assumptions made at the start of the initiative get verified or replaced by newer assumptions based on organizational learning.

Finally, analysts from financial markets nowadays demand disclosure and explanation of every company action in terms of impact on the next quarter’s earnings (Chanda & Ray, 2015; Rappaport, 2006). This may make managers reluctant to extend autonomy to fledgling efforts, stymieing innovation. Further, in developed financial markets, top managers of business firms are instructed by investors to desist from diversifying risk, on the ground that investors can diversify risk at lower cost by building portfolios according to their individual risk appetite (Porter, 1987). This view alternately manifests as an aversion to allowing unrelated diversification. To investors focused on short-term benefits, a few firms dying on account of not diversifying risk is not a matter of concern, since their portfolio construction activity anticipates such demise(s) (hence the demand for minutia regarding the impact of the company’s action on next quarter’s earnings). One reckons that long term, ‘patient’ investors would prefer to have firm survival emphasized. However, they appear to have less voice. Even regulatory pressure—for example, pressure to comply with the minimum public shareholding regulation in India and associated rules mandating that promoters reduce their shareholding and decision-rights over time—may choke innovation. For example, the founder of Kotak Mahindra Bank in India lamented that regulations forced him to reduce his (promoter) stake to around 20%, effectively stymieing his enthusiasm to nurture the organization to greater heights (ENS Economic Bureau, 2020; Manali, 2020; PTI, 2020; Surabhi, 2021). In sharp contrast, in the United States, the steady flow of major innovation in companies like Google/Alphabet (Arrieche & Drozdovica, 2022), Facebook/Meta (Goldfarb, 2022) and SpaceX (Maidenberg, 2022) appears to be related to concentrated decision-rights the respective promoters are allowed to hold (more than 50% in all three companies cited). Thus, even as this article focuses on managerial control, issues stemming from influence actions by external entities may serve to distort efforts towards radical innovation.

Directions for Future Research

The study suggests some topics for future research. Though success with effecting creative destruction has been long known to contribute to organizational longevity, there has been little research to date on what makes conditions more conducive for managerial control of radical innovation in a business firm. This research adds to the body of literature by a formal modelling demonstration of a means by which top management can extend autonomy to middle and lower-level personnel while retaining a sense of being in control. It follows that research is needed to understand what actions managers take to vary the rate of exploitation and what information signals they pay attention to, in order to assess the direction of movement of outcomes.

Moreover, given that the infusion of heterogeneous external knowledge has a crucial role to play in organizational success, particularly when CHC is low, it would be interesting to investigate the process by which new knowledge overcomes the scepticism of organizational members, so as to eventually become validated organizational knowledge. The recourse probably lies in principles of learning-by-doing, for instance by means of low-cost probes into the future (Brown & Eisenhardt, 1997) or by variation of organizational routines (Feldman & Pentland, 2003), etc. Further research is needed in this area.

We may also note a problem with extant research that considers control and autonomy as a dichotomy. This view equates the act of senior management extending autonomy to organizational personnel with senior management relinquishing control. The study demonstrates that such characterization constitutes an incomplete view. My research highlights specific conditions under which managers can be in control, and concurrently extend a certain degree of autonomy to organizational personnel. Future research on innovation themes should look for avenues that steer clear from considering autonomy and control as an either/or construct pair and invoke an orthogonal conception instead.

Concluding Remarks

In the extant management literature, in order to get around the fact that distinguishing honest mistakes in learning efforts from shirking or incompetence is not easy, AT and transaction cost approaches demand unlimited foresight in senior management. Thereby, swift punishment gets meted out for any failure that is detected in the organization. Such ham-handed prescriptions tend to drive out innovation activities that require risky experimentation by organizational personnel.

Autonomous functioning and risky experimentation materialize only when middle and operational levels are provided leeway to try out their ideas without having to anticipate harsh punishment should certain experimentation not bear fruit. A certain degree of autonomy to middle and operational levels is a pre-requisite for giving birth to radical innovation having the potential to unleash creative destruction. Managers should opt to cultivate external heterogeneity guided by the idea that diversity is the basis for continued order (Burgelman, 1983, 2015; Prigogine, 1980). Success with creative destruction, in turn, is linked to long-term organizational survival. I demonstrate that in situations where senior managers may exercise control by means other than by directly monitoring actions of organizational members for alignment with preset goals, autonomy can get extended, enabling far-ranging innovation necessary for creative destruction.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author received no financial support for the research, authorship and/or publication of this article.

References

Abernathy

W. J.

, & Clark

K. B.

(1985). Innovation: Mapping the winds of creative destruction. Research Policy, 14, 3–22.

Argyris

(1976). Single-loop and double-loop models in research on decision making. Administrative Science Quarterly, 21(3), 363–375.

Arrieche

, & Drozdovica

(2022). Alphabet shareholders: Who owns the most GOOGL stock? Capital.com, November 29. https://capital.com/alphabet-shareholder-who-owns-most-googl

Augier

, March

J. G.

, & Marshall

A. W.

(2015). The flaring of intellectual outliers: An organizational interpretation of the generation of novelty in the RAND Corporation. Organization Science, 26(4),1140–1161.

Barrick

, Stewart

, Neubert

, & Mount

(1998). Relating member ability and personality to work-team processes and team effectiveness. Journal of Applied Psychology, 55, 267–306.

Becker

G. S.

(1964). Human capital: A theoretical and empirical analysis, with special reference to education. Columbia University Press.

Brenner

, & Ambos

(2013). A question of legitimacy? A dynamic perspective on multinational firm control. Organization Science, 24(3), 773–795.

Brown

S. L.

, & Eisenhardt

K. M.

(1997) The art of continuous change: Linking complexity theory and time-paced evolution in relentlessly shifting organizations. Administrative Science Quarterly, 42(1), 1–34.

Burgelman

R. A.

(1983). Corporate entrepreneurship and strategic management: Insights from a process study. Management Science, 29, 1349–1364.

10.

Burgelman

R. A.

, & Grove

A. S.

(1996). Strategic dissonance. California Management Review, 38/2 (Winter), 8–28.

11.

Burgelman

R. A.

(2002). Strategy as vector and the inertia of coevolutionary lock-in. Administrative Science Quarterly, 47(2), 325–357.

12.

Burgelman

R. A.

, & Grove

A. S.

(2007). Let chaos reign, then rein in chaos—repeatedly: Managing strategic dynamics for corporate longevity. Strategic Management Journal, 28(10), 965–979.

13.

Burgelman

R. A.

(2015). Prigogine’s theory of the dynamics of far-from-equilibrium systems informs the role of strategic entrepreneurship and innovation in organizational evolution. In Shalley

C. E.

, Hitt

M. A.

, & Zhou

, (Eds.), Oxford handbook of creativity, innovation, and entrepreneurship: Multilevel linkages (pp. 433–444). Oxford University Press.

14.

Cardinal

L. B.

, Sitkin

S. B.

, & Long

C. P.

(2010). A configurational theory of control. Sitkin

S. B.

, Cardinal

L. B.

, Bijlsma-Frankema

K. M.

(Eds.) Organizational control (pp. 51–79). Cambridge University Press.

15.

Chan

(1998). Functional relations among constructs in the same content domain at different levels of analysis: A typology of composition models. Journal of Applied Psychology, 83, 234–246.

16.

Chanda

S. S.

(2017). Inferring final organizational outcomes from intermediate outcomes of exploration and exploitation: The complexity link. Computational and Mathematical Organization Theory, 23, 61–93.

17.

Chanda

S. S.

, & McKelvey

(2020). Back to the basics: Reconciling the continuum and orthogonal conceptions of exploration and exploitation. Computational and Mathematical Organization Theory, 26(2), 175–206.

18.

Chanda

S. S.

, & Miller

K. D.

(2019). Replicating agent-based models: Revisiting March’s exploration-exploitation study. Strategic Organization, 17(4), 425–449.

19.

Chanda

S. S.

, & Ray

(2015). Optimal exploration and exploitation: The managerial intentionality perspective. Computational and Mathematical Organization Theory, 21(3), 247–273.

20.

Chanda

S. S.

, & Ray

(2016). Learning from project failure: Globalization lessons for an MNC. Thunderbird International Business Review, 58(6), 575–585.

21.

Chanda

S. S.

, & Ray

(2021, April 29). Why do strategic projects fail? SSRN. 10.2139/ssrn.3836325

22.

Chanda

S. S.

, Ray

, & McKelvey

(2018). The continuum conception of exploration and exploitation: An update to March’s theory. M@n@gement, 21(3), 1050–1079.

23.

Chatterjee

, Chanda

S. S.

, & Ray

(2018) Administration of an organization undergoing change: Some limitations of the transaction cost economics approach. International Journal of Organizational Analysis, 26(4), 691–708. IJOA-07-2017-1202

24.

Child

(1973). Strategies of control and organizational behavior. Administrative Science Quarterly, 18(1), 1–17.

25.

Crilly

, & Sloan

(2014). Autonomy or control? Organizational architecture and corporate attention to stakeholders. Organization Science, 25(2), 339–355.

26.

Crossan

M. M.

, & Apaydin

(2010) A multi-dimensional framework of organizational innovation: A systematic review of the literature. Journal of Management Studies, 47(6), 1154–1191.

27.

Datta

, & Jessup

L. M.

(2013). Looking beyond the focal industry and existing technologies for radical innovations. Technovation, 33(10), 355–367.

28.

DeShon

, Kozlowski

, & Schmidt

, Milner

, & Wiechmann

(2004). A multiple-goal, multilevel model of feedback effects on the regulation of individual and team performance. Journal of Applied Psychology, 89, 1035–1056.

29.

Dewar

R. D.

, & Dutton

J. E.

(1986). The adoption of radical and incremental innovations: An empirical analysis. Management Science, 32(11), 1422–1433.

30.

Dierickx

, & Cool

(1989). Asset accumulation and sustainability of competitive advantage. Management Science, 35, 1504–1511.

31.

Dominguez-Escrig

, Mallén-Broch

F. F.

, Lapiedra-Alcami

, & Chiva-Gomez

(2019). The influence of leaders’ stewardship behavior on innovation success: The mediating effect of radical innovation. Journal of Business Ethics, 159(3), 849–862.

32.

Edelman

, & Eisenmann

T. R.

(2011). Google Inc. Harvard Business School case, 9-910-036.

33.

Eisenhardt

K. M.

(1989). Agency theory: An assessment and review. Academy of Management Review, 14(1), 57–74.

34.

ENS Economic Bureau. (2020). Kotak Mahindra, RBI settle promoter stake tussle, voting rights to be capped at 15%. The Indian Express, January 31. https://indianexpress.com/article/business/banking-and-finance/kotak-mahindra-rbi-settle-promoter-stake-tussle-voting-rights-to-be-capped-at-15-6243459/

35.

Feldman

M. S.

, & Pentland

B. T.

(2003). Reconceptualizing organizational routines as a source of flexibility and change. Administrative Science Quarterly, 48(1), 94–118.

36.

Goldfarb

(2022). Zuckerberg motivates supervoting stock resistance. Reuters, October 27. https://www.reuters.com/breakingviews/zuckerberg-motivates-supervoting-stock-resistance-2022-10-27/

37.

Goldstein

J. R.

(1961). RAND: The history, operations, and goals of a nonprofit corporation. RAND Paper P-2236-1, RAND Corporation.

38.

Gong

, Le

, Zhang

, Chen

, & Zeng

(2021a). Organizational adaptability influenced by practice strategy, environmental dynamism, and absorptive capacity. Complexity, 4241485.

39.

Gong

, Le

, Zhang

, & Chen

(2021b). Impacts of practice combinations on organizational knowledge: Based on March’s exploration-exploitation model. Complexity, 5618287.

40.

Govindarajan

, & Trimble

(2005). Organizational DNA for strategic innovation. California Management Review, 47(3), 47–76.

41.

Hall

P. A.

, & Soskice

(2001). An introduction to varieties of capitalism. In Hall

P. A.

, & Soskice

(Eds.), Varieties of capitalism: The institutional foundations of comparative advantage (pp. 1–68). Oxford University Press.

42.

Harzing

A.-W.

(1999). Managing the multinationals: An international study of control mechanisms. Edward Elgar.

43.

Iyer

, & Davenport

T. H.

(2008). Reverse engineering Google’s innovation machine. Harvard Business Review, (April) 59–68.

44.

Jensen

M. C.

, & Meckling

W. H.

(1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305–360.

45.

Keil

, & Mähring

(2010). Is your project turning into a black hole? California Management Review, 53(1), 6–31.

46.

Kirsch

L. J.

(1996). The management of complex tasks in organizations: Controlling the systems development process. Organization Science, 7(1), 1–21.

47.

Kirsch

L. J.

(1997). Portfolios of control modes and IS project management. Information Systems Research, 8(3), 215–239.

48.

Kogut

, & Zander

(1992). Knowledge of the firm, combinative capabilities, and the replication of technology. Organization Science, 3(3), 383–397.

49.

König

, Kammerlander

, & Enders

(2013). The family innovator’s dilemma: How family influence affects the adoption of discontinuous technologies by incumbent firms. Academy of Management Review, 38, 418–441.

50.

Lange

(2008). A multidimensional conceptualization of organizational corruption control. Academy of Management Review, 33(3), 710–729.

51.

Lee

J. Y.

, Bachrach

D. G.

, & Rousseau

D. M.

(2015). Internal labor markets, firm-specific human capital, and heterogeneity antecedents of employee idiosyncratic deal requests. Organization Science, 26(3), 794–810.

52.

Levinthal

D. A.

, & March

J. G.

(1993). The myopia of learning. Strategic Management Journal, 14(S2), 95–112.

53.

Maidenberg

(2022). Elon Musk’s SpaceX stake moves lower, but executive retains control. Wall Street Journal. December 23. https://www.wsj.com/articles/elon-musks-spacex-stake-moves-lower-but-executive-retains-control-11671761294

54.

Manali. (2020). Uday Kotak ends long standoff with RBI; sells 2.8% stake in Kotak Mahindra Bank. Business Today, June 2. https://www.businesstoday.in/industry/banks/story/uday-kotak-ends-long-stand-off-with-rbi-sells-stake-in-kotak-mahindra-bank-260011-2020-06-02

55.

March

J. G.

(1991). Exploration and exploitation in organizational learning. Organization Science, 2(1), 71–87.

56.

Montealegre

, & Keil

(2000). De-escalating information technology projects: Lessons from the Denver international airport. MIS Quarterly, 24(3), 417–447.

57.

Nonaka

(2007). The knowledge-creating company. Harvard Business Review, July–August https://hbr.org/2007/07/the-knowledge-creating-company.

58.

Nonaka

, & Takeuchi

(1995). The knowledge-creating company: How Japanese companies create the dynamics of innovation. Oxford University Press.

59.

Pil

F. K.

, & Leana

(2009). Applying organizational research to public school reform: The effects of teacher human and social capital on student performance. Academy of Management Journal, 52, 1101–1124.

60.

Porter

M. E.

(1987). From competitive advantage to corporate strategy. Harvard Business Review, 65(3), 43–59.

61.

Prigogine

(1980). From being to becoming. WM Freeman.

62.

PTI. (2020). In compliance with promoter stake reduction requirement: Kotak Bank. Business Standard, August 11. https://www.business-standard.com/article/finance/in-compliance-with-promoter-stake-reduction-requirement-kotak-bank-120081100047_1.html

63.

Rappaport

(2006). Ten ways to create shareholder value. Harvard Business Review. (September issue).

64.

Schumpeter

J. A.

(1942/1994). Capitalism, socialism and democracy. Routledge.

65.

Shimizu

(2012). Risks of corporate entrepreneurship: autonomy and agency issues. Organization Science, 23(1), 194–206.

66.

Shimizu

, & Hitt

(2004). Strategic flexibility: Organizational preparedness to reverse ineffective strategic decisions. Academy of Management Executive, 18, 44–59.

67.

Snell

(1992). Control theory in strategic human resource management: The mediating effect of administrative information. Academy of Management Journal, 35(2), 292–327.

68.

Stam

(2018). Enabling creative destruction: An entrepreneurial ecosystem approach to industrial policy. Utrecht University, Working paper series 18-05. https://www.uu.nl/sites/default/files/rebo_use_wp_2018_1805.pdf

69.

Surabhi. (2021). Lowering promoter stake: No relief for Kotak Bank: Bombay High Court refuses to stay RBI’s Dec 31 deadline. The Hindu Businessline, December 6. https://www.thehindubusinessline.com/money-and-banking/hc-refuses-to-stay-rbi-deadline-for-kotak-mahindra-bank-promoter-stake-dilution/article25762612.ece

70.

Tannenbaum

A. S.

(1968). Control in organizations. McGraw-Hill.

71.

Tripsas

(1997). Unraveling the process of creative destruction: Complementary assets and incumbent survival in the Typesetter industry. Strategic Management Journal, 18, 119–142.

72.

Tripsas

, & Gavetti

(2000). Capabilities, cognition, and inertia: Evidence from digital imaging. Strategic Management Journal, 21, 1147–1161.

73.

Turner

K. L.

, & Makhija

M. V.

(2006). The role of organizational controls in managing knowledge. Academy of Management Review, 31(1), 197–217.

74.

von Nordenflycht

(2011). Firm size and industry structure under human capital intensity: Insights from the evolution of the global advertising industry. Organization Science, 22(1), 141–157.

75.

Williamson

O. E.

(1981). The economics of organization: The transaction cost approach. American Journal of Sociology, 87(3), 548–577.

76.

Witt

M. A.

, & Jackson

(2016). Varieties of capitalism and institutional comparative advantage: A test and reinterpretation. Journal of International Business Studies, 47(7), 778–806.

77.

, Ma

, Su

, & Wu

C. H.

(2021a). Double-layer learning, leaders’ forgetting, and knowledge performance in online work community organizations. Journal of Organizational and End User Computing, 33(1), 92–117.

78.

C. H.

, Wu

, Ma

, Su

, & Tsai

S. B.

(2021b). Organisational leadership style, network structure, and knowledge performance in online knowledge community organisations. Enterprise Information Systems, 15, 868–887.

Enabling Creative Destruction: A View from the Perspective of Managerial Control

Abstract

Keywords

Introduction

Theoretical Foundations

Radical Innovation as a Precursor to Creative Destruction

Difficulties in Bringing about a Radical Innovation

Control and Autonomy

Why Are Radical Innovation Activities Less Amenable to Conventional Ways of Control?

Towards a Model for Management Control of Radical Innovation

Simulation Model

The Environment and the Project Organization Attempting Radical Innovation

Learning by the Organizational Code

Learning by Organizational Members

Autonomy

Model Parameters

Results

Prospects for Materialization of Radical Innovation, Stable Environment

Prospects for Materialization of Radical Innovation, Turbulent Environment

Summary of Results

Discussion

Limitations of the Study

Some Implications

Directions for Future Research

Concluding Remarks

Footnotes

Declaration of Conflicting Interests

Funding

References