7.3. An Information Feedback View of Management and Policy
The market growth model embodies the essence of the information feedback view of the firm. There is more to this view than feedback loops, stocks and flows – important though they are. There is also an underlying philosophy about management and policy (in all areas of business and society), which I want to reflect on now before developing the model any further. A good place to start is with Forrester's original comments about the craft of management. In his seminal book Industrial Dynamics (Forrester, 1961, p 93) he describes management as the process of converting information into action.
If management is the process of converting information into action, then management success depends primarily on what information is chosen and how the conversion is executed. Moreover the difference between a good manager and a poor manager lies right at this point between information and action.
The viewpoint is cognitive but also action oriented. As Forrester (1994, p 51) goes on to explain:
A manager sets the stage for action by choosing which information sources to take seriously and which to ignore. A manager's success depends on both selecting the most relevant information and on using that information effectively. How quickly or slowly is information converted into action? What is the relative weight given to different information sources in light of desired objectives? How are these desired objectives created from available information?
Here is a parsimonious and stylised portrayal of management, entirely consistent with the principles of information feedback systems, yet capable of capturing a very wide range of managerial behaviour. Leadership, charisma and other important attributes of management (vital to business performance and explicitly addressed in the organisational and strategy literature) are represented implicitly and subtly in system dynamics through their influence on the information sources deemed important enough to justify action.
This information-processing view of management leads directly and naturally to the fundamental representation scheme in system dynamics. Again quoting from Forrester's early work:
A simulation model is based on explicit statements of policies (or rules) that govern decision making. The decision making process consists of three parts: the formulation of a set of concepts indicating the conditions that are desired, the observation of what appears to be the actual conditions, and corrective action to bring apparent conditions toward desired conditions.
(Jay Forrester, 1961, Industrial Dynamics, Pegasus Communications, Waltham, MA. Quotes are from Chapter 10 in the book. Reproduced by permission of Jay W. Forrester.)
The verbal description above translates into Figure 7.11. It represents how managers make adjustments to organisational asset stocks or resources through operating policy. The policy is represented mathematically as: corrective action = (desired resource – apparent resource)/time to correct gap. In a simple formulation, the desired resource is constant and the apparent resource is identically equal to the actual resource.
Figure 7.11. Converting information into action – an information feedback view of management
More sophisticated asset stock adjustment can be portrayed too, capturing the ambiguity of resource management found in practice. For example, the apparent resource may differ from the actual resource because of reporting delays, error or bias. In extreme cases, the condition of the actual resource may not be known at all and so cannot be managed or controlled. The desired resource need not be constant but instead varies over time, through a process of goal formation, depending on conditions elsewhere in the business. Hence, this stock adjustment formulation though compact is very versatile. Moreover, it raises lots of interesting questions about managers' ability to build and to balance organisational resources by focusing attention on: (1) whether, for each resource, there is a clear communicable goal as the basis for corrective action; (2) whether resources are accurately known or not; and (3) whether, having sensed a gap between desired and apparent resource, management react quickly or slowly to close the gap.
7.3.1. Information Available to Decision Makers and Bounded Rationality
According to the feedback view of management, information drives the corrective actions of organisations as they seek to build and maintain a balanced set of resources. Which information sources, among those available, are actually used by managers? This question lies behind the Baker criterion and the formulation guidelines mentioned earlier in the chapter. In principle, the state of every single resource in an organisation is relevant to every point of decision making in the organisation so that stock adjustments are fully informed. Consider, for example, a supply chain comprising retailers, wholesalers, distributors and a factory. What information is relevant for the factory's production planning? Obviously the factory should take account of distributors' orders because the factory supplies the distributors. The factory should also take account of its own inventory and backlog condition and much more besides. Clearly there is a lot of information in the supply chain that, theoretically at least, is relevant to production planning. For example, there is customer demand and the amount of downstream inventory and backlog at every stage of the supply chain. Compiling and making sense of all this data, however, is a huge task, beyond the abilities of normally competent people, even in today's information society
In reality, things are much simpler. Factory managers normally pay most attention to tangible information that is available locally such as distributors' orders and factory inventory. The more general point is that decision makers typically use much less information than the total available to them. Moreover, the available information is less than is commonly presumed. One way to think about this selection process is to picture operating policy surrounded by information filters, as shown in Figure 7.12. The figure shows five possible filters. The first and most basic filter stems from people's cognitive limitations. It doesn't take much information for us to feel overwhelmed and so we pick the signals that seem most relevant and ignore the rest. The torrent of email messages is a daily reminder of our information limits. The formal name for cognitive limitations as they affect decision making is 'bounded rationality', a term first proposed by Nobel laureate Herbert Simon. The concept was developed in the influential literature of the Carnegie School as a theory of firm behaviour and an alternative to conventional microeconomic theory (Simon, 1979; 1982).
Figure 7.12. The policy function, information filters and bounded rationality – behavioural decision making
The other filters are created by the organisation itself as it parcels out responsibilities across functional areas and departments. In a sense, the organisation compensates for individuals' cognitive limitations by placing them in a departmental structure that means they can get on with their own job without worrying about what is happening in every other department. As Simon (1976) and Barnard (1938) have pointed out, the organisation provides a psychological environment that conditions (sometimes very powerfully) how people act and it is the function of the executive to 'design' this environment so departmental efforts are coordinated. The CEO is a designer with various design tools at his or her disposal (Forrester, 1996). Most obvious are operating goals, rewards and incentives (the second filter) that direct people's attention and effort to organisational objectives. Such inducements reduce the complexity of decision making by prescribing what needs to be achieved by different departments. The potential downside is a functional mentality, though this syndrome is not necessarily a problem in a well-designed organisation, it just means people are focused (and that could be an advantage).
The next filter represents the effect of information, measurement and communication systems on decision making. Undoubtedly, computers and the internet have massively boosted the amount of information flowing around and between organisations. Again supply chains are a good example. These days it is possible for factories to hold data on retail orders at point of sale and distributors' inventory for use in production planning. However, such data are not necessarily as convenient or trusted as information gleaned from informal channels such as casual conversations, ad hoc meetings and walkabouts. Moreover, informal 'intelligence' is often more persuasive and timely than information available on the official database.
The fourth filter represents the effect of organisational and geographical structure. Organisations frequently set up or acquire a new business unit at arms length from the existing business. A good example is the fledgling low-cost airline Go, set up by BA in the 1990s, to compete with easyJet and Ryanair in the European short-haul business. Go was deliberately made independent of BA, operating its own planes with its own staff and information systems. The whole point was to design an enterprise that could decide and act in its own right, free from the influence of parent BA. The amount of information flowing between the two organisations was much less than if they were seamlessly merged in a single airline. Another example is the MINI car division of BMW, created as an independent business unit, able to take its own decisions (under corporate guidance) on product development and capital investment in order to develop a new and distinct brand in the highly competitive global small-car market.
The fifth filter is the most subtle yet, and also the most powerful. It captures the attenuating and amplifying effect on information of tradition, culture, folklore and leadership. From a cognitive view these intangible traits shape the psychological environment in which people take decisions and act. They define what the organisation stands for and therefore what really needs attention. Consider, for example, Google and its co-founders Larry Page and Sergey Brin. Commentators say they are intellectually obsessed with an omniscient and omnipotent algorithm for mining the world's knowledge. This belief is now part of Google's culture, permeating the minds of thousands of employees and helping to coordinate their actions. Another example is MIT with its culture of technological excellence that pervades all departments including the humanities and management, shaping decisions on faculty recruitment, the curriculum and choice of students.
7.3.2. Nature of Decision Making and the Decision Process
An important conclusion from the discussion so far is that the feedback view of organisations incorporates behavioural decision making and assumes bounded rationality (Morecroft, 1983; Sterman, 1989). This perspective on decision making distinguishes system dynamics sharply from traditional microeconomics in which 'economic man' makes objectively rational decisions, weighing up all available sources of information to arrive at an optimal (profit maximising) configuration of resources. Figure 7.13 captures the essential philosophical stance to decision making in system dynamics. Any operating policy sits amid its filters, bombarded by information originating from all asset stocks in the system. Consistent with the Baker criterion, however, only a handful of information flows penetrate to the policy core leading to action and stock accumulation.
A corollary is that decision making is conceived as a continuous process for converting varying information flows into signals that determine action (rates of flow). In system dynamics, a decision function does not portray a choice among alternatives, as found in a decision tree with its various discrete nodes and branches. Neither is it the familiar logic of 'if-then-else' that applies to individual decisions like whether to take a taxi or catch the bus. The crucial point is that we are viewing decision processes from a distance where discrete choices disappear, leaving only broad organisational pressures that shape action.
Choosing the proper distance from which to view and model a system is crucial. As Forrester has noted, we are not as close as a psychologist, concerned with the mechanisms of human thought, delving into the nature and sources of cognition, personality and motivation. We are not even close enough to see each separate decision, but instead observe a modulating stream of decisions. We may not be close enough to care whether one person or a group creates the decision stream. On the other hand, we are not as remote as a public stockholder who is so far from the corporation as to be unaware of the internal structure, social pressures and decision points. An intermediate position that is best is similar to the perspective of the Board of Directors, top management team, strategy consultant, or investment banker.
Figure 7.13. Behavioural decision making leading to stock accumulation and feedback
