Steven Lukes (1974) Power: A radical view - Summary
Summary: In the 1960s Dahl and the other pluralists proposed a view of power, wherein a person has power over another person, if the first person can get the second person to do something she or he would not do otherwise. They generally focused on the exercise of actual power, at who brought up and supported which alternatives, at the actual disagreement, and at which alternatives were finally adopted. Their prototypical situations are decision-making processes in groups. This view of power, the 1D view, is a liberal view, it focuses on issues, on how those issues are decided, and on which observable conflict of interests exists. In 1962, Bachrach and Baratz criticised this view as superficial and self-serving, as America celebrating its pluralism. There is another side to power, the power to prevent issues from being discussed or decided upon. In political organisations, some issues are foregrounded, others are hidden from view, benefiting specific groups in a systematic way. In this sense, non-decisions, that is suppressing challenges to the interests of those making decisions, are themselves decisions. This view, the 2D view, is a reformist view, it focuses on potential issues, determined by observable conflicts of interest, and how they can be prevented from becoming actual issues. But I believe this is still not enough: Agenda-setting and exclusion of potential issues, ultimately influencing what another person thinks, wants, and desires may be the height of power. In this sense, socialisation, mass media, and controlling the information flow are forms of thought control: The most effective form of power is preventing conflict. If agenda-setting is successful, the political public is prevented from even hearing minority opinions, which thus remain minority opinions. This view, the 3D view, is a radical view, it focuses on agenda-setting, and includes covert and latent conflicts over real interests, which may differ from self-perceived interests. Talk about real interests always opens one up to charges of paternalism. But Gramsci realised one can learn from how people behave in abnormal situations, where the apparatus of power is less strict. In such situations, the words and effective actions of persons may not fit, and one can learn something about their real interests, about what they would do in the absence of power. This is a really hard problem to overcome, but it is not inherently impossible.
Source: Steven Lukes (1974) Power: A radical view. First published as a small 64page-volume. New York: MacMillan, 1974. Second edition includes the full original text, supplemented by two new chapters. New York: MacMillan, 2005.
(Full text [PDF, 160kb] at chula.ac.th, English)
This summary is licensed CC:BY-SA.
Detailed Summary
[1] The way of identifying power I am going to propose is theoretically and politically radical.
This way is necessarily evaluative. It is always contested.
It can be applied empirically.
In describing power, I will deal with methodological individualism, behaviourism, and the roles values play in explanation, false consciousness, and the bias of pluralism.
In the 1960s Dahl and the other pluralists championed a view of power that traces its roots back to Weber.
In 1962, Bachrach and Baratz criticised this view. It was both superficial and it was self-serving, America celebrating its pluralism.
The pluralists attacked Bachrach and Baratz in return.
I believe Bachrach and Baratz were essentially right, but they did not go far enough.
In the following, I will present three views.
I will call the view of the pluralists the one-dimensional view of power.
I will call the view of Bachrach and Baratz the two-dimensional view of power.
I will present a third view, which I will call the three-dimensional view of power.
[2] I believe calling Dahl, Polsby, Wolfinger, and their colleagues pluralists is mistaken.
They do want to show that power is pluralist, but this is just the conclusion they are aiming for.
Their methods and approaches can be employed for other projects, and can lead to other non-pluralist conclusion.
Dahl: A person has power over another person, if the first person can get the second person to do something she or he would not do otherwise.
Dahl mentions both ability and success, potential and actual power. But the pluralists generally focus on the exercise of actual power.
Dahl looks at who brought up and supported which alternatives and which alternatives were finally adopted.
The observable or reconstructed behaviour of participants is the object, from which power can be measured. Their prototypical situations are decision-making processes in groups.
Power, influence, control, and other related terms are often used as synonyms by the pluralists.
The pluralists believe open conflicts to be best-suited, as it is easy to see who backs which alternative, and which alternative wins out in the end.
For Dahl, "actual disagreement in preferences" is a necessary condition for power research.
Dahl, in one case, mentions the possibility of a proposal encountering no opposition.
However, the framework of the pluralist can not handle this possibility.
Without an observable conflict between conscious preferences manifested in behaviour, they see no power.
For pluralists interests are preferences, conflicts of interests are conflicts of preferences.
Pluralists reject the idea that people can be unconscious or mistaken about their interests.
Polsby: If the researcher can know the objective interests of a class, and the class disagrees, the researcher can call this "false class consciousness". This method can never fail, it is metaphysical, not empirical.
The 1D view of power focuses on how those issues are decided, on which observable conflict of interests exists.
[3] Bachrach and Baratz, in contrast, believe there are two sides to power.
The first side is analogous to the 1D view of power.
The second side considers the power to prevent issues from being discussed or decided upon.
Schattschneider: "Organisation is the mobilisation of bias." In political organisation, some issues are foregrounded, others are hidden from view.
B&B: Values and institutional procedures benefit specific groups, often but not always an elite minority, in a systematic way.
Power is control over another person's behaviour.
Coercion is power by means of threats.
Influence is power without any threat.
Authority is power by means of values held by the controlled.
Force is power by removing the choice of non-compliance.
Manipulation is force, without the controlled noticing control.
According to B&B, the 1D view of power is too focused on conscious behaviour.
Non-decisions (conscious or unconscious) are themselves decisions.
A decision is "a choice among alternative modes of action".
A non-decision is a decision suppressing challenges to the interests of those making decisions.
Dahl: The attention of the political class is necessary for something being a political issue.
There are potential issues, which are not actual issues due to non-decisions.
B&B: A key issue is an issue challenging the power, authority, or the decision-making process.
B&B also focus on observable conflict, though it may be overt.
B&B: Without conflict, the presumption should be consensus. Political science is unable to determine how this consensus came about.
Conflicts are found by observing grievances, overt or covert. Thereby the interests of those outside the political system can be considered.
The 2D view of power is a critique of the 1D view's focus on behaviour.
Decisions on potential issues, determined by observable conflicts of interest, can be prevented.
[4] The 2D view is much better than the 1D view, as it includes agenda-setting into its considerations.
However, the 2D view of power is still too limited.
The 2D view is still too behaviour-focused.
The bias of the political system is not simply the result of individual decisions and actions.
Agenda-setting and exclusion of potential issues are often the result of group behaviour.
Weber: Power is "individuals realising their wills despite the resistance of others".
A group can act in a certain way, and the action is not the action or decision of particular individuals.
The specific form of organisation of a group may itself have effects.
Marx: "Men make their own history but they do not make it just as they please; they do not make it under circumstances chosen by themselves, but under circumstances directly encountered, given and transmitted from the past."
The 2D view is still too focused on observable conflicts.
Even some of B&B's own concepts (manipulation and authority) can do their work without conflict.
Influencing what another person thinks, wants, and desires may be the height of power.
Socialisation, mass media, and controlling the information flow are forms of thought control.
Dahl: "leaders also shape preferences"
The most effective form of power is preventing conflict.
The 2D view seems to believe that an absence of grievances points to an absence of contrary interests.
It is not quite clear what a grievance actually is.
It is possible to shape thoughts and preferences to make the status quo appear natural, beneficial, unalterable, or without alternative.
There may be latent conflicts between the interests of those making decisions and the "real interests" of the excluded, of which they may or may not be conscious.
1D view: behaviour, decision-making, issues, conflict, interests. participation
2D view: critique of behaviour focus, (non-)decision-making, (potential) issues, (covert) conflicts, interests, grievances
3D view: critique of behaviour focus, decision-making, agenda-setting, (potential) issues, (covert) (latent) conflicts, (real) interests
[5] Power is "ineradicably value-dependent".
How to look at power is always disputed, and this dispute is already politics.
The most basic idea is someone affecting someone else in a significant way.
We thus need to think about what makes a way of affecting significant.
All three views share the idea that "A exercises power over B when A affects B in a manner contrary to B's interests."
Parsons connects power to institutionalised authority and consensus and disconnects it from force and coercion.
Parsons: Power is "a facility for the performance of function in and on behalf of the society"
Arendt: Power is a feature of groups, of people's coordinated action. Power is consensual, it is "the very condition enabling a group of people to think and act in terms of the means-ends category."
Madison: "All government rests on opinion."
In both cases, violence, conflicts, and struggles for power are re-defined as not being power at all.
Arendt: "Power and violence are opposites; where the one rules absolutely, the other is absent."
Both see power as an ability, not a relationship. Both definitions exclude what is most interesting to students of power.
Everything that Arendt or Parson can express about power, can also be expressed by the 3D view of power.
In my opinion, cases of persons or groups affecting one another without conflict of interests qualify as influence, but not as power.
I am unsure whether rational persuasion counts as power and/or influence. One person significantly affects another, but the resultant behaviour is caused by the autonomous decision of the second person to accept the reasons of the first. This may well be related to the Kantian antinomy of causality and reason.
Is it possible for one person to exercise power over another, thereby advancing the other person's real interest?
We could think of short-term power as self-destroying once the second person recognises their real interests. This can be abused in the form of paternalism.
We could also think of the second person's autonomy as their supreme interest, which means a violation of autonomy can never be in their interest. This will lead to most cases of influence becoming cases of power.
If we can find an empirical way of finding a person's real interests, I prefer the first view.
[6] Interest is also inherently evaluative.
Liberals tend to identify interests with what people actually prefer, as seen by political participation.
Reformists agree, but allow for indirect and concealed preferences.
Radicals believe preferences can result from a system contrary to people's interests. Thus real interests are what people "would want and prefer, were they able to make the choice."
The 1D view is broadly liberal, the 2D view broadly reformist, and the 3D view a broadly radical view.
[7] The pluralists, in their 1D view of power, "studied actual behavior, stressed operational definitions, and turned up evidence".
Their studies mirror the biases of their object of study.
Dahl shows us diversity, different people making decisions regarding different issues.
Dahl: Voters use elections as an indirect kind of influence on politician's decisions.
Dahl: The dissatisfied will find another political representative.
If power can really set the agenda, this diversity is an illusion.
Pluralism in decision-making is consistent with unity in agenda-setting.
The 1D view is unable to recognise this possibility.
Dahl only studies successful interventions by the dissatisfied, and concludes the dissatisfied can intervene.
If a powerful group can not accept an issue, there might be indirect cases of agenda-setting.
The 2D view can show cases of systematic bias, but only in cases where observable grievances are prevented from becoming issues.
B&B's study on poverty and race in Baltimore is superficial, because it focuses on individual decisions and actions.
The real exclusionary forces at play are inaction and institutional inertia.
The 3D view can give a sociological explanation of how demands are prevented from being voiced or from becoming dangerous.
How can we study what does not happen?
Polsby: For any event, there are an infinity of alternative non-events. Which are significant and which are not?
Polsby: We should only accept answers which refer to the desires of community members.
Wolfinger: If we apply an external theory of expected behaviour to such non-events, we can not distinguish between actual exercises of power and errors in the theory.
These counterarguments claim difficulties to actually be impossibilities.
In everyday understanding, an exercise of power is a conscious, intentional act of individual persons.
I believe we can speak of "exercise of power" even in the case of groups, or if it happens unconsciously.
The operative sense of "exercising power": If two agents are both exercising power, they both affect another person simultaneously, and the person really does change behaviour, this behaviour is overdetermined. Both exercised power, but neither of them individually made a difference.
The effective sense of "exercising power": If there is no other force intervening, one person exercises power over the other, if the other person really does change their behaviour as a result. The exercise of power makes a difference.
A person may change the behaviour of another person in many different ways, with only some of them being what the first person actually wants. Only in such cases is an exercise of power successful.
How can we find an exercise of power?
An exercise of power is conceptually reliant on comparison with the counterfactual situation of what would have happened without the exercise of power.
In observable conflicts, where the alternatives are publicly spelled out, the counterfactual is obvious.
If there is no observable conflict, we need to justify the counterfactual in another way. This will not be easy, but it is not, in principle, impossible.
A good example is Matthew Crenson (1971) The Un-Politics of Air Pollution: A Study of Non-Decisionmaking in the Cities. This book operates in between the 2D and the 3D view of power.
Crenson assumes "the proper object of investigation is not political activity but political inactivity." He compares two communities in Indiana, similar geographically, demographically, and pollution-wise.
Crenson shows that the town with a single company and a strong party took 13 years longer to act on air pollution. The company, US Steel, without ever becoming an overtly political actor, at first successfully prevented pollution from becoming an issue, and then influenced the resulting decisions.
The company did nothing, and in verbally agreeing that air pollution was a problem, they even prevented a political conflict from arising.
In comparative studies, Crenson shows that an industry reputation for power, combined with their silence and inaction on the problem, greatly decreases the likelihood of air pollution becoming a political issue.
Clean air is a diffused common good, with no direct benificient. However, the costs of maintaining clean air mostly fall on industry. There is thus only weak and diffuse support, but strong opposition.
Crenson further shows that issues do not arise independently, but are connected: If one collective issue arises, other collective issues are much more likely to also arise.
Decision-making is directed by prior non-decision-making.
The Crenson study basically takes a 2D approach.
It surpasses B&B's 2D view by looking at inaction, by looking at institutional power, and by looking on how raising an issue can be prevented.
Crenson: "there is something like an inarticulate ideology in political institutions [... promoting] the selective perception and articulation of social problems and conflicts"
If agenda-setting is successful, the political public is prevented from even hearing minority opinions, which thus remain minority opinions.
Crenson succeeds in presenting the counterfactual: We can assume that people would rather not be poisoned by the air they breathe, even if they do not publicly state so.
Crenson also shows how the inaction of institutions prevents the matter from becoming a political issue.
[8] Identifying the right counterfactual is a problem specific to the 3D view.
The 3D view also needs to show how power is exercised in such cases of inaction.
It seems hard to decide whether an injustice is accepted due to an exercise of power or whether it actually reflects a consensus based on a different value system than ours.
But in some cases, we can find evidence for a consensus being the result of power silencing dissent.
Gramsci notes that sometimes the words and effective actions of a person do not fit.
Gramsci: If the masses show such a mismatch between words and actions, this is the expression of a social contrast. The masses see the world one way, and sometimes, in special circumstances, act accordingly. But the masses also see the world another way, the way of their oppressors, and they talk and normally act according to this other way.
One can learn from how people behave in abnormal situations, where the apparatus of power is less strict.
Gramsci: The church is constantly influencing its adherents. If this connection is interrupted, e.g. during the French Revolution, the church loses its influence.
When, in Czechoslovakia in 1968, the apparatus of power was relaxed the people acted quite differently than before.
The reaction of people to perceived opportunities of advancement in hierarchies can also tell us a lot.
Some people consider the Indian caste system to be consensually accepted, even by the lower castes.
Srinivas: While it is theoretically forbidden for a lower caste to emulate Brahmins, this has happened frequently. If a lower caste succeeds in adopting a vegetarian diet and Sanskritizing its rituals, it can change its position in the hierarchy.
If a caste is economically successful, it often tends to do this, in order to increase its status.
If there is a possibility to rise in the caste system, it is often taken.
In theory, the positions of the castes in the hierarchy are unchangeable. But in effective action, castes attempt to change their position.
The introduction of universal suffrage changed the acceptance of the caste hierarchy by the lower castes.
The lowest caste, the Untouchables, have often used conversion to other religions as a way to escape the caste system.
This evidence is always, by its very nature, indirect and thus non-conclusive. But using such evidence, exercises of power, and the counterfactuals needed, may be identified.
According to the 3D view, an exercise of power can be a case of inaction, it can be unconscious, and it can be exercised by groups. All of these present problems for the 3D view in identifying such exercises of power.
If inaction leads to a potential issue not arising, this is a double non-event. But a non-event may still leave traces. Those who refrain from acting may still have considered acting and its consequences.
In a Freudian fashion, people may be unconscious as to their motives for action. This is not specific to power analysis, and it is a widely discussed problem.
Alternatively, people may be unconscious of the perception and interpretation of their actions by others. This, however, does not obscure identification of an exercise of power.
Still alternatively, people may be unconscious of the results of their actions. This is the real problem for analysis of power.
If people are unable to know what the results of their actions on others will be, it seems wrong to classify these as exercises of power.
However, if people do not know the effects of their actions because they did not try to find out, these might well be exercises of power.
If people could have taken steps to learn about the effects of their actions, but did not, and are thus ignorant about the effects of their actions, they might well exercise power.
It seems also quite hard to determine whether an institution is exercising power or whether observed effects are due to structural determination.
In Marxism, this is the fight between voluntarism and determinism.
For Althusser and Balibar the capitalist totality and its structure determines its elements.
Poulantzas: Miliband sees class, the state, etc not as objective structures, but as reducible to interpersonal relations. He is looking at subjects as social actors, he is, in the end, simply doing sociology.
Miliband: Poulantzas sees the structure as so strict, individual persons, even heads of state, are doing nothing but executing what is already determined by the system. He is thus unable to understand the dialectical relation between the state and "the system".
There are other possibilities besides structural determinism and methodological individualism.
Social research has to look at the relations between individual actors and structures.
Individuals act as part of groups, and not just due to their individual motivations.
Individuals act within the limits of structures, but they have at least some autonomy, they can at least somewhat act differently from how they actually do.
"The future, though it is not entirely open, is not entirely closed either (and, indeed, the degree of its openness is itself structurally determined)."
A system that is totally determined by structural relations is a system without power.
One could redefine power alternatively, as Poulantzas did.
Poulantzas: Power is "the capacity of a social class to realize its specific objective interests"
Power then becomes an effect of structure on the practice of class struggle.
I believe this to be misleading.
If we call something an exercise of power, we believe the person exercising power could have acted differently, or the group could have organised differently and thus acted differently.
Saying something is an exercise of power is also saying the person/group exercising it is at least partially responsible.
Responsibility is the result of the action or inaction of a person or group.
This is not the place to discuss the boundaries between structural determination and exercises of power.
C.W. Mills: Sociological fate refers to events that can not be controlled by an identifiable group with the power to decide and able to predict the consequences of their decisions.
Those who can change the world to the benefit of (large parts of) society, but do not, exercise power and can be held accountable for their (in)actions.
[9] The 1D view of power reproduces the bias of the system it studies.
The 2D view of power can critically reflect this bias, but it sees it too narrowly.
The 3D view of power understands that latent conflicts can be suppressed. It has a number of serious difficulties, but it can overcome them.
We can understand power in a deeper way.












