Merton on unintended consequences

Robert K. Merton, in his classical piece, The Unanticipated Consequences of Pur- posive Social Action, clearly states what is included in the ‘consequences’ of an action:

Rigorously speaking, the consequences of purposive action are limited to those elements in the resulting situation which are exclusively the outcome of the action, i.e., those elements which would not have occurred had the action not taken place.

(Merton 1936: 895)

Thus, unless the action of an individual (or individuals) is, at least partially, causally responsible for the ‘consequence’, it is not the consequence of that ac- tion. If there is an unintended consequence of an action, it is plausible to think of other unseen or neglected (disturbing) causal factors which prevented the action from bringing about the intended end. However, for us to consider this as an unintended consequence of an action it is also necessary that if the action had not taken place, the unintended consequence would not have occurred. That is, the action in question is a necessary condition for the unintended consequence, ceteris paribus.

When unseen or neglected factors interfere with one’s action, unintended con- sequences may be brought about. It is because of these unseen or neglected causal factors that Merton uses ‘unanticipated consequences’ interchangeably with ‘un- intended consequences’.1 It seems natural to think that all of the unanticipated consequences are unintended. However, one can think about cases where the con- sequence was intended but unanticipated. For example, people buy lottery tickets with the intention to win the lottery, but most of them do not anticipate that they will win the lottery, for if they did we could hardly be able to explain the surprise of the winners. Moreover, anticipated consequences may be unintended. I may an- ticipate that if things go wrong my action may bring about certain consequences; however, it is not my intention to bring about those consequences as a part of my action. Now, for the sake of the argument, let us assume, like Merton, that in most cases an unanticipated consequence is unintended, or vice versa. Later, this distinction will prove to be useful for understanding invisible-hand arguments (especially in Chapter 5).

Merton classifies unanticipated (unintended) consequences according to the factors that help their coming about. Merton (1936: 898–901) argues that ‘the most obvious limitation to a correct anticipation of consequences of action is pro- vided by the existing state of knowledge’. He then lists some of the different kinds of factors, such as ignorance, error and the ‘imperious immediacy of interests’, that may cause unintended consequences.

We might think of different cases to conceive of Merton’s distinctions. For example, let us assume that agent A intends to achieve X and he believes that by doing Y he can achieve X. Thus, A thinks that Y causes X and acts accordingly. Of course, A can be wrong in supposing that Y causes X. It might be the case that Y causes V. If this is the case, when A does Y to achieve X, V will happen. V is an unintended consequence of A’s action and this consequence can be accounted with A’s lack of knowledge of the causal determinants of X.

Another case might be that A is right about the causal determinants of X but he or she is ignorant of the other factors which may change the course of events that will follow his or her action. Therefore, we can say that A is ignorant of the fact that Y causes X ceteris paribus. If this is the case, when A does Y to achieve X, because of other interfering factors some other event, say Z, will happen. Z is an unintended consequence of A’s action and it can be accounted with A’s ignorance about the other possible interfering factors.

Of course, A can be right in assuming that Y causes X but he might be unable to execute Y successfully. So, A might fail to do Y or might do something different than Y (as in the example where I used the wrong key to open the lock), say T, which in turn causes V. If this happens to be the case, we can consider V as an unintended consequence of A’s action and account it with A’s error.

If A is stimulated by his or her ‘imperious immediacy of interests’, he or she might want to achieve X by doing Y without thinking about the other further consequences of his or her action. If it is the case that Y causes X, V and Z, when A does Y to achieve X, he or she will bring about X, V and Z all together. Because V and Z were no part of A’s intentions, they are the unintended consequences of A’s action and can be accounted with A’s imperious immediacy of interests, or his or her short-sightedness.

These four cases lie beneath Merton’s account of unanticipated consequences of purposive human action.2 Merton focuses on the reasons why agents might be unable to anticipate the consequences. Accordingly, it suffices for him to discuss the agents’ lack of knowledge, ignorance or ‘imperious immediacy of interests’. Yet, even if we consider these as causal factors that may bring about unintended consequences, we have to realise the fact that they are only partially responsible for the generation of unintended consequences. In each and every case other fac- tors help unintended consequences in coming about. For every case we may think of many possible (and different) causal processes that may lead to unintended consequences – in combination with agent’s ignorance or lack of knowledge, etc. Thus, we may not explain exactly why the unintended (or unanticipated) event had taken place by merely referring to the factors specified by Merton. Such an explanation should be able to give an account of the course of events leading to the unintended consequence – in addition to the specification of the agent’s inabil- ity to anticipate them. The type of classification presented by Merton only alerts us to those different relations between the agents and other causal factors that are responsible for the unintended consequence, but it does not tell us anything about the ‘other causal factors’. In other words, this is a partial ‘causal classifica- tion’ and a full causal classification would require that all types of causal mecha- nisms that agents could not foresee, or do not know about, are spelled out. Such a classification would also require a specification of different types of actions (e.g. individual, collective, etc.), different types of consequences (e.g. physical, individual, social, etc.) and different ways in which these consequences affect the individuals.

In this line of inquiry, Merton attempts to classify unanticipated consequences according to the sum-total consequences of action:

These sum-total or concrete consequences may be differentiated into (a) consequences to the actor(s), (b) consequences to other persons mediated through (1) the social structure, (2) the culture and (3) the civilization.

(Merton 1936: 895)

The actions, he tells us, may be differentiated ‘into two types: (a) unorganized and (b) formally organized’ (Merton 1936: 896).3 This distinction implies that there are differences between consequences of individual action and what we might call consequences of social action.4 Similarly, Raymond Boudon points to different configurations of the consequences of social action:

The number of possible recombinations of the following criteria does, there-fore, define the number of possible configurations: 1. No participant (1a), some participants (1b), all participants (1c) attain their individual objectives; 2. Producing, at the same time, benefits (2a), or problems (2b), or else collec- tive benefits and problems (2c); 3. Each of these applying only to some (3a), or to all the participants (3b).

(Boudon 1982: 6)

These configurations, together with Merton’s classifications, suggest that there is a vast variety of unintended consequences and that there might be different causal processes behind the unintended consequences of organised action and unorgan- ised action, or individual and social action. Although it may not be possible to determine the different types of causal mechanisms behind different unintended consequences, it is still important to understand the possible range of unintended consequences. One possible way to classify unintended consequences is from the perspective of their places of materialisation relative to the target of agent’s in- tentions. The next section discusses this classification by introducing a table of possibilities, which combines Merton and Boudon’s tentative classifications to give a broader picture.

Source: Aydinonat N. Emrah (2008), The Invisible Hand in Economics: How Economists Explain Unintended Social Consequences, Routledge; 1st edition.

Leave a Reply

Your email address will not be published. Required fields are marked *