AN ORGANIZATIONAL STRATEGY EVOLVING

The aim of this final part is to show that the concepts introduced can be used to describe the selective evolution of a strategy. A strategy can be seen as a memetic entity evolving by selective processes including evolutionary learning.

The aim is not so much to explain the evolution, but to show how such evolution can be described by an approach that is based upon the concepts and logic introduced above. I cannot do this exhaustively, but will show that many parallels exist between evolutionary approaches involving memes, and the evolution of strategies. I am not aware of selective evolutionary approaches in policy development as a means of analysis, while such approaches are common in understanding conceptual evolution in science. This second part serves as an introduction of such an approach to policy science.
The strategy I focus on is an explicitly stated strategy of an organization. Below I shall describe what I mean by such a strategy. After that I shall describe how such a strategy can be said to evolve. Finally I shall focus on different parts of a strategy, and different selective events that have a causal effect on the evolution.

The strategy of an organization

To begin with, I shall give a definition of what I refer to by strategy and stress the connection of such a strategy with an organization [8] . I define a strategy as 'a program of actions to be taken in order to achieve agreed upon goals'. The words 'agreed upon' and 'actions' are usually meant to refer to the members of the organization that holds the strategy. Thus the members of the organization that are in charge agree upon the strategy, and those members of the organization that take actions.

By strategy I mean a written or otherwise explicitly stated strategy. This is not to be confused with strategies that an outside-observer might recognize in a system, or personal strategies, consciously held or not. Thus I am referring to a self-description that a particular organization has made about actions to be taken with regard to goals.
A strategy as described above is by definition an intentional plan to change things. It describes particular problems, by defining goals to be solved. But planning actions is not the same as implementing them. It is important to see that the actions planned in a strategy and the actions actually taken are not the same thing. In memetic terms the strategy is a meme-complex (in which different memetic lineages can be recognized) that (partly) instructs for actions. The meme-complex can thus be seen as analogous to genes, and the actions taken, together with a lot of other things, are analogous to the phenotype.
According to the definition, a strategy must include both goals and actions to achieve those. The actions stated are connected to the goals by a body of logic, or knowledge. This knowledge can be included in a written strategy, but often this will not be the case. However when actions are connected to goals, a body of knowledge is necessarily drawn upon.
A strategy prescribes actions to achieve changes, which are believed to contribute to the achievement of goals. These goals will usually refer to states of things outside the system (like we want to be number one in our market), but can also refer to changes in the organization (we need to become more efficient). A strategy belongs to a specific organization.

The evolution of a strategy

As stated above my aim is to describe the evolution of a strategy. To do this I assume a situation where there are more 'generations' of such a strategy. If an organization makes a strategy regularly, we can call the different versions of it generations, like the generations of a species.

The differences between the different generations can be seen as evolution in the weakest sense of the word, meaning nothing more than simply change in time (for different meanings of the word evolution, see Hodgson, 1993 ; Modelski and Poznanski, 1995 ). This is the first step of the approach.
The second step is to divide the changes into new variation, deletions, and substitutions. Introduction of new variation, deletions and substitutions can occur with regard to goals, actions and (parts of) the body of knowledge. Deletions can be taken to be the result of a selective event of some kind. A Substitution is at the same time a deletion and an introduction of new variation of an element of the strategy with a similar function. For example the substitution of a prescribed action that is meant to contribute to the achievement of a goal is equivalent to a selective event where the two actions are interacting in view of the criterium of which one will contribute (the most) to the achievement of the goals, given a great number of particular circumstances of different kinds. When there is a deletion, there has been a weeding out, meaning that a part of the strategy is rejected, but that there is no direct substitution.

If we include the second step of the approach we not only take evolution to refer to change, but we also look at evolution as a selective process. This selection is however only reflected in the change of the strategy, while the real selective processes are taking place in the organization.

In memetic terms we can say that the strategy is a (shared) part of the memepool, held by individuals of the unit-pool, which is largely equivalent to individuals in the organization. For this analogy to work it must be added that only a part of the memepool is focused upon. This part is structured around functional aspects of (strategic) issues that are taken to be problems that need to be solved. In fact such an approach is also not uncommon in biology. There is often a strong bias for focusing on genes that spread through the unit-pool, having selective value (or are functional in surviving the selection process), or in other words, being solutions for the problems that we denote to as the selection pressures (Campbell, 1974 ).
The generations of strategies can be seen as single photographs taken at a different moments. In reality a continuous movie [9] is playing. The selective events, as well as the inclusion of new variation, that we can detect when comparing two generations, have taken place somewhere between the generations. Again, in comparison with biology, genefrequencies in genepools are measured at certain points in time.

Step three in the approach involves the interpretation of the found changes. In case of new variation, we can firstly ask where did it come from (from outside the organization or was it thought up in the organization), and if there was more variation abundant, which was not introduced. That would mean that there have been selective events in the acceptance of new variation, and that other variation that was detected has been rejected. The next question to ask then would be where in the organization did this selection take place, and by what criteria.

In the case of deletions we can ask who has taken the decision to delete, and by what criteria. In examples of public policy, a distinction is possible between criteria concerning A) the way a problem can be solved regarding the causes of the problem (to reduce acid rain, you have to reduce the emission of substances that cause it), B) the causal relations and normative aspects of instruments you can use to solve the problem (to shoot the people that are involved in the emission is illegal, and morally unacceptable, while advising them to reduce emissions simply will not work), as well as C) criteria regarding the political situation in which decisions are taken. A good solution might be selected out because the involved politicians thought that such a decision might be bad for their career, for the electorate situation of the party, or doesn't fit into the memetic lineages the party sticks to (concerning ideology).
The evolution of a strategy is a process in which many selective events take place where different sets of criteria, or different selection pressures can be distinguished. These events cannot be seen apart from the different 'players' that are involved in different selective events.
Hoppe ( 1993 ) for instance, in an essay on policy argumentation and belief systems refers to a distinction between high, middle and low games or levels of political judgement. At the highest level political party elites, think tanks, prestigious political commentators and ideologues, and some top-notch public managers debate issues on levels of rational choice and system vindication. At the middle level the legislature, the upper and middle levels of the executive branch, interest groups, lobbies, journalists and spokesmen of various social and political movements debate issues on policy design/adoption and (the initial stages of) policy implementation. At the low levels middle and lower-level members of the executive branch, with its score of professionals, technicians and experts, and courts debate issues of policy implementation and evaluation processes including technical verification. The specific levels with particular players and criteria are of no concern to this essay, but they illustrate different selective systems operating on parts of a strategy in different parts of an organization.
Hoppe seems to assume that the criteria used in decision making at specific levels are separable. In other words, the levels are not only distinguishable by the identity of different players, but also by different criteria used to judge policy issues. This means that players at the lowest level do not use criteria of what Hoppe refers to as 'rational choice' to judge policy issues.
However, players at the middle level may anticipate the decisions they believe will be selected out by political criteria in the political process. In this situation the evaluating actors are politicized, or infected by memes consisting political criteria. In addition, politicians can use evaluative, and instrumental criteria for the weeding out of policy issues they do not want to get accepted because this does not suit their political aspirations.
In a memetic perspective it matters which people use which criteria in selective events, because it can have effects on the evolutionary learning potential of an organization with regard to achieving its goals. For instance if new variation in solutions is weeded out by the political process, there will be less chance that new successful solutions will be encountered.

An evolving strategy and memetic selection

I would like to recall some memetic insights and concepts in our approach for the evolution of a strategy.

First the planned actions in a strategy are instructions that can steer the organization toward reaching its goals. In this way the class of planned actions is analogous in its instructional nature to the genotype of an organism, while the class of actually executed actions is analogous to the characteristics of the phenotype (Hull, 1988a ). The goals in the strategy are based upon problems to be solved, being a selective system (Campbell, 1974) for the organization. Thus a strategy is seen as instructional toward dealing with selective forces.
Secondly, a strategy is followed through time as an evolving memetic entity. Without loosing sight of the causal actors that are involved in all kinds of processes that influence the evolution of the strategy, the main focus of the approach is on a strategy as an entity of its own. This entity reflects agreed upon or social knowledge about what problems are, what actions should be taken, and the causal connections between those two. A third point is that the different parts of a strategy can change (more or less) independently in time, involving new variation, and deletions by selective events. When new variation appears we can ask where the new memes came from, why they, and not others, were selected, why they were not selected before if they were already present for a longer time, etc. When deletions occur we can ask where and by whom they were weeded out, and by what criteria. These questions about variation and selection of parts of the strategy are connected to memetic questions in two ways. First there are questions about the particular routes memes have taken through a unit-pool, or from unit-pool to unit-pool. Secondly there are questions about what selection criteria have been used, and by whom.
The last memetic focus is about evolutionary learning occurring between generations of the strategy. As mentioned before, selective events as such do not guarantee that there is any kind of evolutionary learning.
In a strategy that has relatively stable goals (compared to the speed with which new knowledge can be acquired) a kind of evolutionary learning can take place. Any problem which a strategy aims to solve, has causal relations that determine what things must be changed to solve it (the emission of acid substances must be changed to solve the acid rain problem). In the case of acid rain, a considerable amount of scientific research has been done to be able to establish which causal relations can, and which can't be influenced. This research can be said to contain a large amount of trial and error learning, in the quality of scientific experiments. This acquired knowledge is used by the makers of acid rain policy to determine and explain what approaches are and are not viable. This example shows that over time, knowledge about the problem can accumulate, and be shared by memetic transmission. In the same way knowledge accumulates about what policy instruments are useful in what conditions. Here policymakers in environmental policy can learn from the trials with those instruments in other policy fields.
In this sense we can say that the body of knowledge that connects the actions with the goals in a strategy can evolve by selective evolution, including evolutionary learning transmitted by memetic processes. However other selective events can actually counteract evolutionary learning. For instance the weeding out of possible variation in the political process can counteract evolutionary learning towards the goals in strategy.

next part

Back to hypertext essay on memes

Back to Memes Back to Evolution Back to Hans-Cees