Politics Philosophy Economics 2002 Barry 155 84

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

1/31

http://ppe.sagepub.com/Politics, Philosophy & Economics

http://ppe.sagepub.com/content/1/2/155Theonline version of this article can be found at:

DOI: 10.1177/1470594X02001002001

2002 1: 155Politics Philosophy EconomicsBrian Barry

Capitalists Rule Ok? Some Puzzles About Power

Published by:

http://www.sagepublications.com

On behalf of:

The Murphy Institute of Political Economy

can be found at:Politics, Philosophy & EconomicsAdditional services and information for

http://ppe.sagepub.com/cgi/alertsEmail Alerts:

http://ppe.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

What is This?

- Jun 1, 2002Version of Record>>

by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from by Daniel Silva on November 1, 2013ppe.sagepub.comDownloaded from

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

2/31

Capitalists rule OK? Somepuzzles about power

Brian BarryColumbia University, USA

abstract Even if we do not observe those who own or manage capital doing anything,

are there nevertheless good reasons for saying that they have power over

government? My thesis is that, on any analysis of power over others that

enables us to say that voters have power over those elected and that consumers

have power over producers, we also have to say that those who own or control

capital have power over government. Conversely, the reasons that can be given

(and have been given) for denying that owners of capital have power over

governments would be equally good reasons for denying that voters have

power over governments and that consumers have power over producers.

keywords power, capitalism, democracy

I) The question

Political philosophy and political economy are both well-established disciplines,

but tend to be carried on in separate compartments. What do we get if we try to

put them together? No doubt there are many possibilities, but the question that I

have chosen involves the relation between political power and economic

position: even if we do not observe those who own or manage capital doing any-thing, are there nevertheless good reasons for saying that they have power over

government? Philosophy enters in (and is, in fact, the main element) in as far as

this is a purely conceptual analysis. To anticipate my conclusion, the answer to

the question turns on the interpretation of facts that scarcely anybody would wish

to deny.

Why should we care about this? For an answer, I suggest that we think about

the ideological underpinnings of the current politico-economic dispensation in

the western countries conventionally described as liberal democracies. The key

politics,philosophy & economics article

Brian Barry is Arnold A. Saltzman Professor of Political Science at the Political Science Department,

Columbia University, 7th Floor International Affairs Building, 420 West 118th Street, New York,

NY 10027, USA [email: [email protected]] 155

SAGE Publications Ltd

London

Thousand Oaks, CA

and New Delhi

1470-594X

200206 1(2) 155184

023541

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

3/31

features of this are a political system in which governments and legislators derive

their tenure from the results of a popular vote, an economic system in which

goods and services are allocated through the mechanism of a market, and private

ownership of capital, which carries with it the right to make decisions that affect

the lives of others in fundamental ways. Apologists for this system need to beable to say that, despite appearances to the contrary, the electoral system gives

citizens power over those elected and the market gives consumers power over

firms. But they are liable to be uncomfortable about the notion that private

ownership of capital gives owners of capital power over governments. My thesis

is that, on any analysis of power over others that enables us to say that voters

have power over those elected and that consumers have power over producers,

we also have to say that those who own or control capital have power over

government. Conversely, the reasons that can be given (and have been given) for

denying that owners of capital have power over governments would be equally

good reasons for denying that voters have power over governments and that

consumers have power over producers.

I have said that appearances are against electoral politics and the market as

instruments for providing citizens and consumers with power. The reason for this

is that there is no getting round the fact that, in both the polity and the economy,

decision-making power is concentrated in very few hands, relative to the size

of the population. Consider first politics. Over a long career, Robert Dahl has

indefatigably promoted the notion that, while political systems such as that of the

USA and cities within it cannot properly be described as democracies because

they do not meet the demanding conditions that would be required for such anattribution, they can be described as polyarchies. As he wrote in an encyclopaedia

article: The term polyarchy refers to the processes and institutions of large-scale,

representative democracy of the type developed in the twentieth century. The

term was rarely used before 1953 when it was deliberately reintroduced into

the vocabulary of political science by Robert A. Dahl and Charles E. Lindblom

inPolitics, Economics and Welfare.1 I do not know why, despite Dahls enor-

mously influential position within the discipline, the concept has been so uni-

formly shunned. However, it is fortunate that it has failed to catch on, because it

puts the emphasis in exactly the wrong place.

Polyarchy is, obviously, modelled on oligarchy: in an oligarchy, few rule(in Greek, arche means rule); in a polyarchy, many rule. But it is not true that

what distinguishes the USA from (say) Myanmar is that many rule in the former

while few rule in the latter. Few rule in either, and the same goes for any state

that has ever existed. Perhaps more people rule (in the sense of having a share in

the right to make decisions binding on the polity) in the USA than in Myanmar,

but that is neither here nor there. The significant difference between them is the

way in which the title to rule is acquired: in one case, it arises from a competi-

tion for the right to rule in which voters determine who wins and, in the other

case, it does not.

politics, philosophy & economics 1(2)

156

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

4/31

What difference this makes is, of course, precisely the question of the relation

between having institutions providing for competition between would-be rulers

for votes and having democracy. But the starting point must be that, give or

take an abrogative referendum here and a recall there, it would be hard to say

that rulers (the executive plus the legislature) in the western liberal democracieshave significantly less scope for discretionary decision-making than the rulers in

any country with a different form of government, such as rule by a supreme

council made up of the army officers who staged a coup against the previous

government. All that can be done is to try to mitigate it by saying that, thanks to

the way in which the competition for office works out, the people have power

where politics is competitive in this sense, whereas in other systems the people

do not have power. This claim, it will be noted, takes us back to the meaning of

the Greek words making up democracy: that the people (demos) have strength

(kratos). The claim has to be, in other words, that, although they do not rule, the

people nevertheless have power over the rulers, and thus, indirectly, have power

to determine the content of the decisions taken by the rulers.

This still leaves all the hard work to be done. What exactly is it to have power

over somebody, and how exactly does competition result in voters having power

over rulers, if it does? These questions will be taken up later. But I think it is worth

making a couple of remarks here on what has been said so far. First, it is clear that

we can settle the question by stipulation if we simply define any country with

competitive politics as a democracy. But even if we conclude that a country with

competitive politics is a democracy, that should be because we believe that

competitive politics gives power to the people. The connection between the insti-tutions and the distribution of power should still be a contingent one. Thus, Dahl

was unquestionably on the right track in wanting to make democracy turn on

power relations while using some other expression for the political institutions

characteristic of contemporary western countries. Moreover, Dahls definition of

polyarchy (in its finally developed form) gives us exactly what we want: citizen-

ship is extended to most adults, and the rights of citizenship include the opportu-

nity to oppose and vote out the highest officials in the government of the state.2

I have described this informally as competitive politics, leaving implicit the

stipulation that the competition between different elements in the political elite

is for support among voters rather than, for example, for support among themilitary. The established usage that comes closest to doing the job is, I suppose,

representative government, and this is fine as long as we add the caveat that the

term representation has to be purged of any built-in assumptions about what

the electoral connection actually does: representatives, in other words, are taken

to represent in virtue of being elected, regardless of the way in which this may or

may not affect their behaviour. Also, of course, it is hardly worth asking what the

relation is between representative government and democracy (understood as the

power of the people) unless the people have voting rights, so we should stipu-

late that we are assuming universal suffrage or something close to it.

Barry: Capitalists rule OK?

157

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

5/31

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

6/31

There are, as far as I am aware, no countries with a market economy in which

a substantial proportion of the output is produced by consumer cooperatives.

Consumers, therefore, have no direct say in the decisions made by firms.

Governments intervene (but less and less) to regulate such things as public utili-

ties, and this may (or may not) be done in such a way as to protect the interestsof consumers. By and large, however, decisions about what to make and how to

price it are taken by firms, which have an enormous amount of discretion under

the law. In the light of this undeniable fact about the legal order in which the

market is embedded, there is something breathtakingly audacious about the

commonly made claim that markets guarantee consumer sovereignty. This

expression is, obviously, a variant on popular sovereignty, considered as an

attribute of a political system. Both are ideological constructs which obfuscate,

rather than illuminate, the reality, which is that citizens do not run the govern-

ment and consumers do not run the firms. The claim that has to be made is that

consumers derive power from competition among firms. By exercising their right

to choose how to spend their money, consumers can exercise power over firms,

because they can lower their profits by ceasing to buy the products of a firm if

those products are less attractive than the ones offered by its rivals.

In the extreme version of this story, popularized by Milton Friedman, firms

have no discretion about what to make or how to price it, because they will go

out of business unless they follow the optimal path. In this sense, firms have no

power whatever. It may be noticed, though, that even a firm with a monopoly

(for example, a drug company with a lucrative patent) may have no power in the

same sense. For it may be said that, unless those running it exploit its monopolyposition to the utmost, they will face removal by the major shareholders or a

hostile takeover bid that the shareholders will find attractive. The issue of dis-

cretion therefore appears to be something of a red herring, in as far as even the

most brass-necked apologists for the market would hardly wish to maintain that

the lack of power on the part of a monopolist has as its counterpart power over it

by its victims. The key assertion must be, therefore, that in a competitive market

consumers can get firms to sell them things they want at prices that allow only

for a normal profit, because a firm that does not do so will lose out to other firms.

How far this constitutes a defence of actual market economies is a question I do

not need to raise here. I am concerned with the form that the defence takes, andthe way in which it invokes the notion that a competitive market gives consumers

power.

II) Some puzzles about power

Up until now, I have been reporting claims about power rather than asking

exactly what they mean and how they might be substantiated. Any further

progress in the argument depends upon getting a grip on the concept of power.

Needless to say, I shall concentrate on developing the ideas that I need for the

Barry: Capitalists rule OK?

159

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

7/31

purpose of analysing the claims I have introduced. But even to do that I need to

raise a question that is fundamental to any attempt to talk about power at all. This

is: should we think of power as the possession of an ability or as the possession

of the means giving rise to an ability? Both conceptions, it will be noted, make

reference to an ability, and this is essential to the definition of power, as maybe seen by consulting any dictionary. (I note in passing that this stipulation

eliminates a large proportion of the proposed definitions in the political science

literature, which make having power equivalent to actually causing things to

happen.) The word power is derived from potere, the late vulgar Latin infini-

tive corresponding to our to be able, the infinitive form of can, and the con-

nection is retained in French and Italian, where the nounspouvoirandpotre are

the same as the infinitive form of can.

Let me offer a definition of power that is intended to capture the notion of

power as ability: power is the ability to bring about desired states of the world by

acting. This is, obviously, a very broad definition, which leaves completely open

the means that might be employed to get something: they could include doing it

yourself, asking other people for what you want (the Clinton approach to house

furnishing), persuading somebody that it is a good idea for them to do something,

calling on legitimate authority (so that being told to do something is taken as a

reason for doing it), offering a reward for doing it, and threatening sanctions for

not doing it. There has been a lot of discussion about the propriety of bringing

these heterogeneous means under the concept of power. But it seems to me that

there is no problem in doing so, as long as we are clear that power is then being

used in a very general sense. What causes confusion is the widespread tendencyto define power more narrowly for certain purposes (as I shall be doing in due

course) so that only certain means are included, but then to use in connection

with it other terms whose domain is now excluded because they relate only to

other means. (I shall be able to explain this more clearly below when I have an

example to hand.)

So far, so good. But we still have to hear from the alternative way of con-

ceptualizing power, which identifies it with the possession of the means of bring-

ing about desired outcomes. I shall argue for its rejection, but I think it deserves

serious consideration because it makes the possession of power much easier to

establish. This makes me reluctant to drop it, and this reluctance is reinforced bythree further considerations: first, it involves departing from the definition

advanced by my favourite political philosopher; second, it involves dissenting

from what I regard as by far the best book on the subject; and, third, it involves

recanting my own previous published position on the subject, though since that

was put forward a quarter of a century ago, I can feel a certain detachment in

relation to it.

In chapter 10 ofLeviathan, Hobbes gives a definition of power that is as broad

as the one I put forward, but different in what it makes power out to be. The

POWER of a man, (to take it Universally), is his present means, to obtain some

politics, philosophy & economics 1(2)

160

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

8/31

future apparent Good.3 Hobbes means this literally: thus, for example, Riches

joyned with liberality is Power; because it procureth friends, and servants and,

more generally, what quality soever maketh a man beloved, or feared of many;

or the reputation of such quality, is Power; because it is a means to have the assis-

tance, and service of many.4 Hobbes was no ordinary language philosopher,and would undoubtedly have given short shrift to the objection that it sounds odd

to say that riches or personal attributes are power. He would quite likely have

riposted that an abstract ability has no place in a scientific treatise, and is little

better than the Entities, andEssences of Aristotle.5 This kind of reductionism

has a natural appeal to philosophers captivated by science: Peter Morriss, in

attacking the tendency to equate power with the means of power, quotes W.V.O.

Quine as claiming that, when we say sugar is soluble, we are saying that it has

a structure suitable for dissolving.6 No doubt it does, but this is not what we are

saying. All the same, we could decide to swallow our linguistic qualms to gain

the advantage of a more operational concept.

The trouble is that this gain in our capacity to recognize the existence of power

comes at a price. For it means that we close the logical gap between having

the means and having the ability. To see how this can be a real disadvantage

analytically, let me move (and not before time) to the type of power with which

I shall be primarily concerned in the rest of this article. This is the ability to bring

about desired states of the world by acting in such a way as to overcome the

resistance of others. What this is intended to capture is the idea of a power

relationship as one that incorporates a clash of wills. Thus, getting something by

asking for it is not a clash of wills, because the person who gives it does so freely.But I also wish to exclude persuasion and invocation of legitimate authority,

because at the end of the process there is no resistance to be overcome.

I shall follow a common usage in calling the ability to overcome resistance

social power, though there is obviously something arbitrary about this, since all

power, except the ability to do things oneself, involves social relationships.

Social power is commonly thought of as an especially efficacious form of power.

But it is clearly less reliable (as Max Weber emphasized) than being able to get

people to do what you tell them to do because they believe they have a duty to.

Similarly, if you can persuade people that they really want to do something for

its own sake, you can be more confident of its happening than if you fail and haveto try to overcome their resistance. Be that as it may, however, the three cases of

power that I introduced in the first section all appear to fall within the category

of social power.

Social power has as its instrument the ability to change another partys incen-

tives. This takes two forms: making the action one wishes the other to perform

more attractive and making the alternatives to it less attractive. Taking up the

second of these forms, let me define power over others as follows: A has power

over B if A has the ability to bring about desired actions on the part of B by

exploiting Bs belief that A can make B worse off contingently on Bs behaviour.

Barry: Capitalists rule OK?

161

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

9/31

This is a definition in terms of power of the ability to change behaviour. The

corresponding definition of power over others along Hobbesian lines makes

power actually consist in the ability to make others worse off contingently upon

their actions.

This second way of doing things has the analytical convenience that, if weonce specify the amount that A can make B worse off, we can read off As power.

I followed this line in the article on power to which I referred and Keith Dowding

does the same in his bookRational Choice and Political Power.7 But the objec-

tion to it is that it leaves no room for a question about the relation between what

A can do to B and As ability to get B to do things. Dowding, it is interesting to

notice, actually concedes this, but does not draw the conclusion that the defini-

tion has to be amended. Let us follow his exposition.

Consider a case in which A can at will lower Bs utility by more than the

difference between the utility B gets from the action he would choose to do if left

to his own devices (u1) and the utility B gets from the action demanded by A

(u2). Then, Dowding says, the difference u1u2 that A can make to Bs welfare

is a measure of As power.8 The rationale of this is that if t1 is the measure of

the disutility to B of the sanction A threatens, then the threat will be successful if

t1 > u1u2.9 Notice, however, that these two sentences are actually about two

different things. The first concerns As actual ability to lower Bs welfare. The

second talks about a threat made by A. But A may threaten less of a reduction in

Bs welfare than he could actually carry out or he may threaten more of a reduc-

tion than he could actually carry out. Even an empty threat may be effective if it

is believed: a notice in front of your house threatening Armed Response orSavage Dogs may encourage would-be burglars to try their luck elsewhere,

even if there is no agency or dog to back it up.

Let us suppose that A does have the ability to implement the threat and B

believes that he does. Even with these assumptions, which Dowding appears to

be making, it still does not follow that the threat will be successful when the

conditions are met. The fact that B would be better off complying with As

demand than refusing to do so and suffering the sanction for non-compliance

threatened by A does not actually ensure that B will comply.

Dowding himself recognizes this, writing: Modern non-co-operative bargain-

ing theory . . . shows that A does not necessarily have the amount of power sug-gested by this account. B may rationally refuse to comply with As threats despite

t1 > u1u2.10 According to Dowding, this is because there is a long-term advan-

tage in gaining a reputation for stubbornness, thus making it less likely that one

will be subjected to threats in future. But even if B does not look beyond the pres-

ent case, it may be rational to refuse to comply. For on the assumption (which

Dowding takes to hold normally) that it costs A something to carry out a threat,

A will be worse off carrying it out if B refuses to comply than he would be for-

getting about it. Only a sense of the advantage of acquiring a reputation for puni-

tiveness will provide a reason for carrying out a failed threat. This further empha-

politics, philosophy & economics 1(2)

162

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

10/31

sizes the indeterminacy of the relation between As having the ability to lower

Bs utility and As being able to get things he would not otherwise get.

Notice, however, what is happening to the concept of power in Dowdings

analysis. If we stick to the original notion that power is to be identified with the

ability to lower anothers utility, then that is what power is, and it makes no senseto say that A does not necessarily have the amount of power suggested by this

account.11 Clearly, this statement is implicitly calling on a different conception

of power, which must be that of power as the ability to get people to do things.

Indeed, Dowding explicitly says that, if one group has more resources than

another, the greater resources of the first group may be the means by which that

group is able to have power over the second, but it is not the same as that

power.12 But this is, manifestly, an abandonment of the original definition,

according to which the possession of the means was the power. The conclusion

to which Dowding is driven, almost despite himself, seems to me inescapable. I

shall, therefore, follow the line that equates As power over B with the ability to

change Bs behaviour and then specifies the means: the belief on the part of B

that A can make him worse off.

So far, I have been discussing the case in which A can make the course of

action B would otherwise follow less attractive. The other case is that in which

A can make a course of action that B would not otherwise follow more attractive.

We can say here that resistance is still overcome, in that B would not perform the

action desired by A in the absence of a belief that A will make him better off if

he does it. However, it has to be conceded that calling this an instance of social

power is liable to seem strange if our paradigm of social power is power overothers. Thus, for example, David Baldwin argues that the Louisiana Purchase

should be regarded as an exercise of power by the USA.13 Now if we were think-

ing of power as power over someone, we could tell a story that would make the

acquisition of the Louisiana territory into an exercise of power by stipulating that

the French government anticipated an attack by the USA which it either would

not be able to resist or would not find it worth resisting, given other demands on

its military capacities in 1803. We might then say that France took the money

simply as a more attractive option than the alternative of being forced to cede the

territory. But in the absence of such a story, there is nothing except scale to dis-

tinguish the Louisiana Purchase from my buying a hot dog from a street vendor.It is true, as Baldwin says, that Jefferson was able to get France to do some-

thing it would not otherwise do, i.e. transfer title in this land to the United States

and that this way of describing the Louisiana Purchase is compatible with

standard notions of power employed by social power theorists.14 Thus, Dowding

accepts the implication that, if social power is the ability to manipulate incen-

tives, the ability to make offers is a form of social power; and John Harsanyis

theory of social power brings threats and offers within a single analytical frame-

work.15 Even so, I think that our reluctance to call the Louisiana Purchase an

exercise of power (in the absence of implicitly coercive background conditions

Barry: Capitalists rule OK?

163

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

11/31

such as I sketched) stems from the notion that, in a relationship of social power,

one party should stand to gain and the other to lose. In this case, however, both

parties can keep the status quo if they choose, so both gain from the exchange.

Formally, the answer is (as Baldwin says) that both parties have power in rela-

tion to each other. Thus, I can get the hot-dog vendor to do what he would nototherwise do (that is, hand me a hot dog) and he can get me to do something I

would not otherwise do (that is, hand over some money). Unless we are very

careful to distinguish this sort of power from power over others, we shall finish

up by making this case look like the case of mutual nuclear deterrence, which is

reciprocal power over. This would be highly unilluminating.

It is thus clear that, if we are going to treat the perceived ability to make

others better off contingently on their actions as a form of social power, we have

to distinguish it sharply from power over others. One obvious difference is that

one would prefer not to be subjected to power over, whereas one can only gain

from being the target of being offered a reward for doing something. (Even if A

lowers Bs welfare unconditionally so as to make B more receptive to an offer,

B cannot be worse off with the offer of a reward than without it, though he may

well regret having something that A wants.) At the same time, though, there are

real advantages to bringing the two phenomena together. To see this, consider

threats of punishment and offers of rewards. These are characteristic ways of

trying to create beliefs about ones ability to make others worse or better off, and

the contingencies under which one will do so. (It is, however, a crucial part of my

analysis that social power can exist in the absence of specific threats and offers.)

The point is, then, that the case for treating threats and offers within a singleanalytical framework is rather compelling.

To put the case at its strongest, there are some cases in which we can describe

what is essentially the same situation as either one in which a threat is issued or

one in which an offer is made. Thus, if B has expectations about benefiting

under As will, we are inclined to talk about a threat if A says he will not leave

B anything unless he pursues a certain course of action. If the expectations were

weaker or non-existent, we would be more inclined to talk about an offer by A to

B to leave him money if he does pursue that course of action. But in both cases,

Bs future benefit depends on his doing what A demands. Similarly, if a firm that

has announced plans to expand its operations in Britain puts out a statementsaying that it will cancel them unless Britain joins the Euro, that is a threat. If a

firm puts out a statement saying that it will expand if Britain joins the Euro, but

not otherwise, that is a contingent offer. But this difference arises only because

the baseline from which we measure gains and losses is different in the two cases:

in the first case, we count the expansion as an element in a sort of hypothetical

status quo, whereas in the second we take the actual status quo as our baseline.

Even where the relevant status quo is clear, it may still be artificial to treat threats

and offers as if they operated in a different currency. If firms come and go all the

time, a government might reasonably treat a threat by some firm to close unless

politics, philosophy & economics 1(2)

164

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

12/31

it doesxand an offer by another form of the same size to open if it does not do

xas cancelling out.

Bribery is a corrupt form of exchange in which the quid pro quo for money or

other material benefits is a decision by an official, a vote by a legislator, and so

on. What should we say about campaign contributions and personal benefits suchas gifts, holidays, lucrative speaking engagements, and so on? These are typically

described as providing influence, though this is by no means the only source of

influence: a trusted confidante may well have the most influence over a politi-

cian. The core conception of influence in this context is the ability to shape other

peoples judgements or beliefs. An influential theatre critic, for example, is one

whose views are taken seriously by a lot of people in deciding what plays to

go to see. A lobbyist has influence, similarly, if he is able to get politicians or

officials to take serious account of what he says. Those in the trade tell us that the

most important part of the job is cultivating good relations with the target and,

above all, ensuring access when needed, so as to be able to put the clients case.

What money buys, on this account, is the ability to obtain a sympathetic hearing

for your point of view, and that is precisely the kind of thing that influence

covers. No doubt there is a good deal of humbug and hypocrisy around, and it

would be naive to think that this is the only way in which money talks in

politics. Some members of Congress are, notoriously, nothing but the paid

mouthpieces of corporations (the Senator from Boeing, Monsanto, and so on)

and, in such cases, to talk about influence is merely a polite evasion: what we

have is an established pattern of bribery, as against one-off bribery. Let me make

it clear that it is open to us to think that buying privileged access is just as obnox-ious as bribing. My point is simply that we need to distinguish two different

modes of obtaining favourable decisions or votes.

Influence of the form I have been discussing is one form of power in its most

general sense. For it is an ability to act in a way that brings about desired states

of the world. But it is only one form of power, and there is no overlap between

influence and social power. Yet most political scientists treat influence as if it

were synonymous with power, even where (as is usual) they are talking about

social power. This is the most significant illustration of the point made earlier

that we can get into a lot of trouble by narrowing down the concept of power and

then treating inappropriate terms as interchangeable. I do not attempt to influencethe hot-dog vendor by offering him my US$1.25, nor does he attempt to influ-

ence me by offering to sell it at that price. Nor would I be attempting to influence

him if I put a gun to his head and demanded a hot dog in return for not pulling

the trigger. He might attempt to influence me by displaying an endorsement of

his wares by a rabbi, and I might try to influence him by suggesting that he would

please the customers more by substituting a different mustard. The common

practice of using influence as the verb form of power can only lead to con-

fusion.

Although social power and influence are disjoint concepts, it is worth noticing

Barry: Capitalists rule OK?

165

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

13/31

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

14/31

ing a market, so the power of voters over governments depends on a set of rules

defining a constitution. In both cases, we may speak of these rules as giving rise

to specific powers, for example, the power to set prices or the power to vote.

Moreover, in a voting system, there has to be a set of rules stipulating what out-

come (policy, candidate, and so on) arises from any given pattern of votes. Wecan then ask how much power each voter has within a given set of rules, in the

sense of power to make the outcome correspond to the one he favours. Call this

outcome power. It is important to notice that, although this is power within the

general definition of power, it is not social power. For the process of aggregating

votes is one in which only the votes cast enter: the outcome simply emerges from

the pattern of votes. Of course, all kinds of social power may enter into an indi-

vidual voters decision about the way in which to vote: a member of a parliament

may vote in a certain way because he has been bribed to do so or threatened with

a scandal if he does not do so, or because he hopes to obtain ministerial office by

following the party line or fears expulsion if he does not. But a measure of the

power of a vote ignores all that and simply asks how that vote can factor into an

outcome under different configurations of votes. Because voting rules are pre-

cise, they lend themselves to precise indexes of voting power, and the resulting

literature (to which I confess to having contributed) has had a disproportionate

impact on the study of power. The crucial point to make about it is that, in the

nature of the case, outcome power can tell us nothing about social power. The

temptation to think that we can use outcome power as a model for social power

must be firmly resisted.

Thankfully, I have no need to enter into the analysis of outcome powerhere. All we need is to recognize that the outcome power of the voters (taken

together) must be the foundation of whatever power the voters (taken together)

have over the government. In a system of representative government, the

composition of the legislature and (where it is directly elected) the executive is a

mapping from the way in which the electorate votes. It is true that, in a mass

electorate, no individual voter ever makes a difference to the result. (The margin

of error is always more than one vote, whatever the voting system.)18 But to

deduce from this that the voters do not together have outcome power would be

like saying that, because a heap of sand is still a heap if you withdraw one grain,

no amount of sand can make the difference between something that is a heapand something that is not. If enough electors vote differently, the result of the

election will definitely change, in just the same way as the heap will cease to be

a heap if you take away enough sand. We do not imagine that an individual con-

sumer can change the pricing policy or the product mix of a firm by making a

purchase or not making a purchase. But if enough consumers start buying or stop

buying, we can have little doubt that it will make a difference. (I take this to be

common ground, regardless of what might be said about power.) Similarly, any

electoral outcome can be changed to another if enough voters vote that way.

None of this, however, tells us anything about the power of citizens over rulers

Barry: Capitalists rule OK?

167

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

15/31

in a system of representative government. Once elected, as I pointed out in the

first section, governments have an immense amount of discretionary power, that

is to say, outcome power as provided for under the constitution. Can the outcome

power of the electorate that enables them to decide who attains office be turned

into power over those who are elected? It is easy enough to set out the way inwhich this may be thought to happen. We suppose that governments value office,

and are therefore motivated to get re-elected. We need not imagine that they are

motivated (entirely, mainly, or at all) by the trappings of office or the possibility

of benefiting from corrupt deals. Politicians may have policies that they want to

see carried out, and recognize that their best chance of seeing them carried out is

to do it themselves by attaining or retaining office. If a government wants to be

re-elected, it has to somehow persuade enough voters to support it at the next

election. And if voters decide how to vote on the basis of the performance of

the government in the previous electoral period, this gives the government an

incentive to behave in ways that will lead to its being preferred by a majority of

voters to the opposition. The voters together, therefore, have the means of power

over the government, because they can make it worse off (that is, make it lose

office) contingently on what it does. Although possession of the means of power

does not itself constitute the ability to change behaviour, the suggestion we are

to explore is that we have here a mechanism that can be expected to give voters

power over governments. (Formally, we have to add that the government has to

believe that voters can make it worse off, as firms have to believe that consumers

can make them worse off. But since it is evident that voters and consumers have

this ability, I shall not bother to mention the belief in what follows and will talkas if the fact were enough.)

It is worth noticing here an analogy with the power of consumers over firms.

If we took the relevant baseline for calculating gains and losses at each election

to be one in which no party got any votes, there would not be any way in which

the government could be made worse off by an election. We could then analyse

an election as an exchange in which the currency is performance in office and

voters. The government goes first, and supplies a certain record for the period

since the previous election. The voters then decide whether to reward it with

votes or not. We could do this, but it would obviously be very artificial. As with

markets, we can make more sense of what is going on by starting from a positionin which there is a government and then treating its loss of office as making it

worse off. This is the basis on which we can say that the voters together have the

means of power over the government.

The institutional setting I have sketched is the simplest possible to get the

discussion launched. Thus, it does not take account of the possibility of divided

government (as where the executive is directly elected), and it assumes the exis-

tence of a government and an opposition as two entities that could trade places

while retaining their identities. We do not thereby limit the scope of the analysis

to two-party systems, but we do presuppose a stable coalition structure, so that

politics, philosophy & economics 1(2)

168

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

16/31

all the parties forming the government leave office if they cease to have a major-

ity and a completely different set of parties takes over. Since I am interested in

the logic of the electoral sanction, I can best display it by taking up the kind of

case that best lends itself to its operation.

There is something, on the face of it, paradoxical about the analysis justoffered. We assume that voters are interested in bringing about future behaviour

by the government: they want the virtually unlimited de jure discretion of the

government after the election is over to be constrained de facto by fear of elec-

toral retribution at the following election. But if they are to maximize their power

over the government, they must base their vote at the coming election entirely on

its performance since the previous election. If they allow any thoughts about the

future to enter into their calculus, they weaken the electoral sanction, which (in

the nature of the case) can only ever operate on the basis of past performance. To

the extent that voters believe the promises of a government that has done badly

to do better in future, they give governments in future less reason to guide their

actions by asking how they can most effectively please a majority of the elec-

torate. Using the vote to choose a better government prospectively is costly to

voters in terms of their control over the incumbent.19

For example, when the Conservatives dumped Margaret Thatcher as their

leader, they did so in the hope that this human sacrifice would wash away their

sins. They invited the voters to, in effect, draw a line under their performance

prior to the assumption of John Major to the post of prime minister. This appears

to have worked, in as far as the Conservatives succeeded in winning a narrow

majority at the next election despite their previous unpopularity. Clearly, how-ever, if the voters respond to a deathbed repentance of this kind, it has the likely

result that, in future, governments will feel that they can get away with a lot

more than if they were always going to be judged on their whole record since the

previous election.

The crucial point is that whatever power the voters are able to exert over the

government comes about in virtue of the governments expectations about how

its performance will affect its prospects at the next election. Once the election

comes round, there is nothing the voters can do to change the governments

performance: the past is past. Why then bother to vote according to its record?

The answer has to be that this is the only way of keeping the electoral sanctionplausible for the future. Ultimately, the whole business turns on mutual trust: the

voters install a government in the hope that it will act in ways that are calculated

to appeal to voters, and the government has to trust that it will not be thrown out

at the next election as long as it does act in such ways.

It has been argued that any system of term limits makes this kind of power

over those elected impossible. The argument holds, if it is valid, for any fixed

maximum number of terms that can be served, but its logic can be displayed by

looking at the case corresponding to that of the American presidency in which

the number is two. The idea is that, when a president comes up for re-election,

Barry: Capitalists rule OK?

169

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

17/31

voters will know that . . . [he] will have no incentives to seek re-election and

will vote against him.20 But then, of course, since he will have only one term

whatever he does, he will not have incentives to behave well during the [first]

term and voters will not elect him.21 Since all candidates are equally liable to

term limits, the implication (which the authors do not themselves draw) is that thevoters should not vote for anybody.

If we keep clearly in mind the logic of retrospective voting, however, we can

see that this argument is fallacious. Unless we assume that voters want to reward

or punish governments as an end in itself (which falls outside the model), they

are always forward looking when they vote retrospectively. Their object is to

keep governments worried about the electoral consequences of poor perform-

ance, and they can do this only by discriminating at each election on the basis of

past performance. Term limits, obviously, limit the opportunities that voters have

to keep governments on their toes in this way, since a president in his second

term loses office automatically at the end of it, and cannot, therefore, be made

worse off by electoral means. But voters can still deploy the electoral sanction

in relation to presidents who are seeking a second term. If, as the argument I

have quoted suggests, they vote against a candidate in this position, they are

gratuitously throwing away the electoral sanction.

It is, of course, true that the electoral sanction will not be available to constrain

the president if he is in his second term. But, as I have emphasized, voting for the

government or throwing it out on the basis of its performance is always intended

to affect the incentives of governments in the future. All we have to add in this

case is that we have to assume sufficient sophistication on the part of votersto enable them to see that what they are doing is making an investment in the

efficacy of the electoral sanction for the next time it will be applicable, which is

to say any time there is a president in his first term.

It is worth noticing that the notion (which I took over) that a president in his

second term is immune to electoral sanctions depends on the assumption that

presidents care only about their own re-election and not about being succeeded

by a president of the same party. Assume that voters will vote for the candidate

of the outgoing presidents party if they approve of the presidents performance

and will otherwise vote for the other major partys candidate. Assume also that

presidents attach as much value to being succeeded by a member of the sameparty. Then term limits will make no difference even a limit of one term.

Clearly, this model depends on the existence of unitary parties whose candidates

are interchangeable and a perception on the part of the voters that this is so. But

it is probably closer to the truth about American politics than the alternative in

which all candidates are simply regarded purely as individuals by the voters and

regard themselves purely as individuals. Political scientists thought, for example,

that the favourable economic conditions would help Gore in the 2000 election,

and they might have done more for him if he had not perversely distanced him-

self from the administration. Similarly, there are reasons for thinking that Clinton

politics, philosophy & economics 1(2)

170

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

18/31

took some decisions primarily to help Gore: for example, releasing oil from the

strategic reserves to hold the price down in the run-up to the election.

Nevertheless, we may conclude from all this that the logic of retrospective

voting is rather arcane. If three distinguished political scientists could get it

wrong, we would surely be justified in wondering how far it actually motivatesvoters. Our doubts about it are liable to be increased, I suggest, when we ask how

voters are supposed to decide whether to vote for the incumbent government or

the would-be government. The simplest decision rule for voters would be to ask

how good things have been during the governments term of office (on various

commonly accepted criteria of goodness), voting for it if things have been good

and against it if they have not. Bad things happen to incumbents who preside

over recessions, scandals, international humiliations, domestic turmoil, and the

like.22 In its unadorned form, this leaves no room for any judgements about the

governments responsibility for such events or any assessment of the likelihood

that the opposition would have done any better in the circumstances. Indeed, it

has been suggested that a government tends to get a boost in the polls from any-

thing that makes people feel better, such as the countrys doing well in the World

Cup or a spell of unusually fine weather. Clearly, this is pretty hit and miss as a

way of exerting power over governments, since it means that they may be re-

elected or thrown out on the basis of factors over which they have had little or no

control. Nevertheless, it still gives governments an incentive to do as well as they

can, since it remains true that they will improve their prospects in that way. But

the weaker the link between the governments actions and the conditions on the

basis of which voters decide how to vote, the more the government may betempted to regard the outcome as a lottery and pursue its own ends whether

idealistic or corrupt.

Suppose that voters turn against the government whenever things go badly,

regardless of what the government may be able to say in its own defence. Then,

it is important to observe, anyone who can lower the subjective welfare of

voters has a power resource that it may well be able to exploit, to get the govern-

ment to do things it would not otherwise do. Any body (whether domestic or

foreign) that can threaten, say, economic disruption or international embarrass-

ment will have a means of power over the government. Even if the government

could reasonably say that, taking the long view, the threatened sanction is worthputting up with, rather than caving in to the demand that is backed by the

sanction, voting according to actual results is liable to make governments highly

vulnerable to threats to disrupt the lives of citizens.

A good example of this process at work was the campaign by truck drivers in

a number of Western European countries in autumn 2000 to disrupt the flow of

oil to consumers in order to bring pressure to bear on the government to lower

fuel taxes. The point of this was not to inconvenience ministers personally: no

doubt their ministerial vehicles would still have been supplied with fuel, and, if

anything, the absence of other vehicles on the road would make it easier for them

Barry: Capitalists rule OK?

171

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

19/31

to get about. The point was, rather, to create chaos so that the government would

lose electoral popularity unless it acceded to the truckers demands. This

manoeuvre was successful to the extent that the French government caved in, and

the other governments that stood firm did lose support. The British government,

for example, lost its lead in the public opinion polls for the first time since theprevious election.

I am inclined to think that it is a mistake to analyse the link between subjective

welfare and voting in strategic terms at all. Rather than thinking of it as a half-

baked way of trying to exercise power over the government, we might do better

to think of it in purely causal terms. The weather, success in the World Cup, or

lines at the pumps may all influence the dispositions of voters. (Bear in mind that

influence is a very broad term: a painter can be influenced by a landscape or an

author by a long-dead writer.) Even if voters act purely reflexively, however, this

does not mean that strategy is absent from the situation altogether. For, as I have

pointed out, anyone who can lower the subjective welfare of the voters has a

means of power over the government.

We can tell a more sophisticated story about the ways in which voters make

their decisions that will breathe more life into the idea that voting on the basis of

the governments record can be a way of exercising power over it. According to

this, voters decide on a subjective level of welfare that the government has to

have met before they will return it to office, and this can take account of the con-

ditions under which the government is operating, so that the standard can be

set lower if the environment is unfavourable.23 We must always bear in mind,

however, that all this means is that each individual voter makes a decision onthese lines when the election comes around. In as far as voters exert power, it is

purely via the governments guesses about the way in which what it does will

relate to what the voters will do. Thus, to talk about a performance criterion set

by the electorate is to talk about something that does not exist and cannot exist.24

For this suggests that some performance criterion is set in advance and then

implemented at the election: otherwise, it would not make sense to say that

the electorate can choose the performance criterion to motivate the incumbent

optimally and have it be credible that this criterion will in fact be employed.25

This is, transparently, an attempt to make the case fit the model of threats or

promises made in advance and then implemented according to the complianceor non-compliance of the target. But there is, clearly, no institution by which a

mass electorate can, collectively, choose the performance criterion. This would

require some kind of voting system that does not anywhere exist and could

not exist, since it would have to aggregate the diverse demands of millions of

people. This does not mean that the electorate cannot exert power over the

government, but it can do so only in the indirect way (looking backward to affect

the future) that I have described. In as far as aggregation of demands takes place,

it occurs in the interaction between the governments guesses about the voters

and the voters responses.

politics, philosophy & economics 1(2)

172

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

20/31

How are voters to decide how well the government has done? Let us, for the

moment, continue to think of this as a question of how competent it has been in

pursuing goals that would be widely agreed to be desirable. How competent does

the government have to be to deserve re-election? Clearly, if the criterion is set

too high, governments in future will be liable to conclude that they will probablybe thrown out whatever they do, and if it is set too low, they will be liable to con-

clude that they will probably be re-elected whatever they do. In either case the

electoral sanction will form a very weak incentive. But what would be a realistic

criterion? The only sensible one that I can see is that the government should be

re-elected if the opposition would have done worse in its place, and otherwise

thrown out. It seems pointless to compare the governments performance with

that of some hypothetical alternative when there is (ex hypothesi) only one

actual alternative. Moreover, by focusing on a comparison between the govern-

ment and the opposition, we stay in touch with the idea that the power of the

voters stems from competition for their votes.

The point becomes far clearer if we abandon the assumption that voters are

concerned only with the governments competence in bringing about generally

desired conditions. Although those political scientists who talk in terms of

estimates of competence usually mention that voters are also concerned with

actual policies, it is clear that the whole terminology is inappropriate to issues

such as the death penalty, gun control, abortion, blood sports, anti-drug policy

and hundreds of others, where voters have views about policies as such, rather

than regarding them merely as means to valued states of the world. To say that a

government is competent if it pursues policies of which I approve is a verycurious way of talking, since there may be no special skill required to enact or

administer them. Moreover, the conclusion that rational voters have to compare

the actual government and the actual opposition becomes crystal clear here.

Suppose that I regard the policies of the current Labour government on civil

liberties and asylum issues as extremely obnoxious. Should this lead me to vote

for the opposition at the next election? That would be crazy, because I have

excellent reasons for believing, on the basis of the directions from which they

have attacked the government, that they would have been even worse.

We can now see how there really is something paradoxical about pure retro-

spective voting as a way of exerting power over governments. For it can workonly if voters make a comparison between what the government does (or

achieves) and an estimate of what the opposition would have done (or achieved)

if it had been in office. This estimate, will, obviously, have to be drawn in large

part from what the opposition says it would have done instead, though it can be

improved by bringing in other information. For example, if the opposition now

says it would have done something that it passed up the opportunity to do the last

time it was the government, it may be asked how sincere its change of heart is,

or it may be thought that its largest financial contributors are prepared for the

opposition to say it would have done something, but would have balked at its

Barry: Capitalists rule OK?

173

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

21/31

actually doing so if it had been the government. The question that now arises is

this: if voters are prepared to throw out the government on the strength of an

estimate of what the opposition would have done instead, why not vote on the

basis of an estimate of what the opposition would do if it won the next election?

And then, of course, this would need to be compared with an estimate of what thegovernment would do if it were re-elected.

Recall that we got into all this by asking what can be done about the fact that,

once elected, a government has virtually unfettered legal power. Voting on the

basis of the governments past performance was put forward as a way in which

voters could exert power over governments. If I am right, however, rational

retrospective voting requires the construction of a hypothetical alternative world

in which the opposition is the government. This involves some ability to figure

out the dynamics of policy formation within the opposition party the interplay

between, among other things, the policy preferences of the key politicians and

those of the rank and file in parliament, the views of large contributors, and the

sentiments of party members. While it is not actually incoherent to maintain that

this can be done for the past but not for the future, it is difficult to see why this

should be believed. If there is a problem of induction here, it seems no more

than a particular application of the general one that the fundamental laws of

nature might change tomorrow.

After this rather circuitous journey, then, we arrive at the conclusion that

voters may as well cast their votes on the basis of their expectations for the com-

peting parties conduct in the future. Of course, it is quite compatible with this to

say that voters use the past as a signal of future policy outcomes.26

If, like theauthors of that quotation, we call this retrospective voting, there will always be a

retrospective element in voting. But it seems to me much clearer to confine the

term retrospective voting to the kind of thing I have been analysing under that

name: voting on the basis of the past in order to keep the electoral sanction

maximally credible in the future. If the voters make their decisions on the basis

of predictions about how parties will behave in future (their policies, their com-

petence, their honesty, and so on), they will inevitably give a large amount of

weight to the evidence provided by past performance. But this does not change

the crucial point, which is that the voters outcome power is being used to select

the party believed to be best in the future, and only incidentally in order to exertpower over the government through electoral sanctions.

The upshot is, I believe, as follows. First, there is a lot of evidence that voters

are more likely, other things being equal, to support the government when things

go well and are liable to turn against it when they do not, where the criteria for

going well are widely shared values. Although this could be conceived of as an

exceedingly crude attempt to exercise power over government, my suggestion is

that it is more reasonably thought of as reflexive rather than rationally strategic.

And second, although the notion of pure retrospective voting, as a way of maxi-

mizing the power of voters over governments, is not precisely incoherent, it

politics, philosophy & economics 1(2)

174

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

22/31

becomes extremely implausible once we see what it involves. It is very hard to

see why, if voters were prepared to make the informational assumptions required

for retrospective voting to work, they should not switch to prospective voting. I

am strongly inclined to doubt if retrospective voting exists at all, once we distin-

guish reflexive responses from it. However, a certain amount of power overgovernments arises as a by-product of prospective voting in as far as one of the

elements entering into these expectations is the performance of the government

in comparison with the hypothetical performance of the opposition. Governments

thus still have an incentive to act in ways that voters approve of.

IV) The power of capitalists

I am going to assume that owners and managers of capital (directors of large

companies, for example) have power over governments in a variety of ways.

Especially in the USA, where candidates have to raise their own funds to a large

degree, financial contributions give influence (in the way discussed in section II)

and also social power in as far as politicians do things for contributors not

because they are persuaded by them, but as part of a corrupt quid pro quo. When

[party] funds come from special interests, they are exchanged for favors. Pre-

sumably, if Philip Morris Co. Inc. contributed in 1996 over US$2.5 million to

the Republican National Committee . . . , it must have expected at least US$2.5

million in favors; otherwise its management should have been thrown out by the

stockholders.27 In some cases, it may be appropriate to talk about power over

politicians in this context: if a politicians future depends on maintaining contri-butions from a few large contributors, it would be reasonable to set the baseline

for computing gains and losses at the present level of contributions, so that a

threat to stop contributing would count as an attempt to exert power over the

politician.

The argument about Philip Morris goes a little too fast, however. For the direc-

tors of the company could have said that the Republican position was in any case

more favourable to their interests than the Democratic one, and that the US$2.5

million was simply an investment in improving the Republicans chances. Philip

Morris would then, in terms of the analysis here, be giving the Republicans the

means to influence the electorate. On the present hypothesis, however, the com-pany would not be hoping to get power or even influence over the Republicans,

because the party would do the same thing anyway. It might be hard to convince

shareholders, though, that US$2.5 million (a drop in the bucket in relation to total

expenditure) would make enough difference to the probability of the Republicans

calling the shots in Washington to constitute a profitable use of the funds. In any

case, the ability of business to advance the chances of friendly candidates and,

especially, incumbents who offer the security of a well-known voting record

gives it disproportionate influence over voters, even if we leave aside the influ-

ence or power over the candidates that it might yield.28

In addition, of course,

Barry: Capitalists rule OK?

175

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

23/31

owners of capital or their representatives can purchase advertisements them-

selves. If their interests are challenged, they may do this on a huge scale, as with

the [US]$100 million spent by opponents of the Clinton health plan in 1993

4.29 (The amount spent on advertisements in favour was presumably close to

zero.) The result is to increase their opportunities to influence public opinion,and . . . to influence policymakers who treat these advertisements as indicators

of potential public opinion.30 Indeed, in terms of my categories, the ability to

influence the voters can be a means of power over governments to the extent that

they take account of this ability when deciding what policies to pursue.

Taking all these forms of power and influence as given, what I am concerned

with here is the account to be given of what Charles Lindblom, in his classic

Politics and Markets, called the privileged position of business. The foundation

of this is, as Lindblom pointed out, that government officials cannot be indiffer-

ent to how well business performs its functions because depression, inflation,

or other economic distress can bring down a government.31 As we saw in the

previous section, this could come about because voters reflexively respond to bad

times by turning against it or because they believe (rightly or wrongly) that the

opposition would have done a better job in the same circumstances and that this

is a reason for giving it a chance to try its hand at the next election. At the same

time, as Lindblom again points out, the fundamental rule of a capitalist economy

is that the government cannot force a firm to invest or produce if those who own

it, or their agents, choose not to: although governments can forbid certain kinds

of activity, they cannot command business to perform.32 Even if a company is

quite profitable, the government cannot stop it from closing down a plant ifits owners believe that they could make more money by relocating to another

country with lower wages, poorer protection of workers and worse environmental

standards, for example.

It is not at all apparent that Lindblom would be happy saying that this relation

between business and government gives business power over government. We

require, in order to understand it, he says, no conspiracy theory of politics, no

theory of common social origins uniting governments and business officials,

no crude allegation of a power elite established by clandestine forces. Business

simply needs inducement, hence a privileged position in government and politics

if it is to do its job.33 This might be read as suggesting that business is simply inthe fortunate position that what government does to further its own ends (keep

the voters happy with the state of the economy) involves its taking very serious

account of the interests of business. Along these lines, Keith Dowding argues that

in a capitalist society, capitalists are systematically lucky because the welfare of

everyone is dependent upon the state of the economy and capitalism is the motor

of the economy.34 More specifically, politicians will not be re-elected if the

economy plunges into recession, so they must ensure that their policies help the

economy which means helping the interests of capital. This does not itself

make capital powerful. Rather, it is lucky, for capital does not set the constraints

politics, philosophy & economics 1(2)

176

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

24/31

upon what is feasible; it simply benefits from what is.35 At this point, Dowding

invokes an argument from Jon Elster to the effect that, if an outcome arises from

market forces (for example, the tax rate that maximizes revenue), it cannot be

described as one embodying power.36 Similarly, Donald Wittman says that the

Marxists [who say capitalists have power] have it about 98 per cent right, butthe remaining 2 per cent makes all the difference to the analysis. It is true that

democratic governments are severely constrained in their policy choices by eco-

nomic forces. But the Marxists are wrong in attributing these forces to capitalists

or to capitalism itself. Capitalists dont control, markets do.37 It wasnt me,

mum, honest!

It should be added that Dowding says capitalists may have power as well as

luck. Capitalists . . . have no need to intervene partly because they are lucky, and

partly because the politicians may be afraid to act in ways contrary to the inter-

ests of business lest businessmen do intervene.38 In an earlier formulation,

Dowding wrote that capitalists are lucky because they are capitalists in a

capitalist system with a competitive party structure . . . They may be powerful as

well, but there is an empirical difference between the two. If they are systemati-

cally lucky and not powerful, then when their interests are challenged they will

not be able to respond; if they are also powerful then they can respond.39

Unfortunately, Dowding does not explain what he means by the two key terms in

those quotations, intervene and respond. However, I assume that it should

refer to something more exciting than capitalists quietly going about their busi-

ness taking decisions about investment or disinvestment, increasing or decreas-

ing production, and so on. It sounds as if we should be looking for some kind ofconcerted effort involving threats to do horrid things to the economy unless their

demands are met. My argument will be that, if there is any sense in which

consumers have power over firms and voters have power over governments,

capitalists have power over governments merely by acting as individual profit-

maximizing agents.

Why might this be denied? We already have some idea of what may be said.

In the remainder of this section, I shall take up all the reasons I can think of for

saying that capitalists do not have power if they simply behave in the way I

describe.

1) First objection: to say that capitalists have power, in the absence of anyevidence, is to give vent to an anti-capitalist ideology, and not a serious contribu-

tion to social science. This is the burden of the pluralist attack on the notion of

a power elite. As an argument against the claim that there is a unified group in

every city and every country pulling the strings, so that governments are merely

puppets, this is fair enough if diligent attempts to discover such string pulling

have failed to discover it. But the thesis to be assessed here is that capitalists

have power over governments in virtue of anticipation by governments of loss of

popular support if they act contrary to the interests of capitalists. Thus, capital-

ists are able to deter governments, and deterrence is an exercise of social power.

Barry: Capitalists rule OK?

177

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

25/31

Formally, we need to expand the definition of power over to include deterrence,

but this should be uncontroversial. Thus, we can add to the original definition that

A has power over B if A can prevent B from doing something that A does not

want B to do by making B worse off if he does it than he would be if he kept to

the action preferred by A.The thesis I am proposing can still be attacked on the basis of lack of evidence.

Thus, Nelson Polsby wrote: How can one tell . . . whether or not an actor is

powerful unless some sequence of events, competently observed, attests to his

power? . . . If these events do not occur, then what grounds have we to suppose

that the actor is powerful. There appear to be no scientific grounds for such a

supposition.40 The objection is that there is no observable set of events asso-

ciated with successful deterrence. We are probably safe in thinking that those

who find this line of argument congenial against claims about the power of

capital are keen on the idea that mutual nuclear deterrence prevented war and

challenges to the vital interests of either side during the cold war. But what set of

observable events would demonstrate (or even provide evidence for the proposi-

tion) that deterrence was happening? If the USA and the Soviet Union had

started lobbing nuclear missiles at one another, we would have been able to

observe something (though only briefly, if we lived in one of those countries), but

what we would have been seeing was failed deterrence. (Similarly, if we

observed capitalists closing down plants or moving to other jurisdictions, we

would be witnessing failed deterrence, if the conditions that have led to their

doing so could have been avoided by different government policies.)

The fallacy lies, of course, in the behaviouralist dogma that what is observedmust be a set of events. There is plenty of information to support talk about

nuclear deterrence. We know what nuclear bombs can do to a population, we

know roughly how many each side has, and we have some idea of the way in

which they are targeted. There is nothing metaphysical about saying that these

are the means of deterrence, even though we cannot observe actual deterrence

going on. Similarly, we know what capitalists can do to a population, and

therefore indirectly to a governments re-election prospects. Why should we

not believe that governments are deterred from carrying out a whole variety of

otherwise attractive policies?

Moreover, although we do not (and could not in the nature of the case) observeadverse actions by capitalists in countries with business-friendly governments

such as those of Britain and the USA, we can observe what happens when

governments fail to provide capitalists with a favourable enough environment.

The textbook example, which doubtless traumatized a whole generation of politi-

cians in western countries, was the capital flight that brought the new socialist

government in France to its knees in 1981. The governments reversion to more

capital-friendly tax and social benefit policies was as good an illustration of the

successful exercise of power as one could hope to find.

As I was writing this, theNew York Times carried a story saying that over the

politics, philosophy & economics 1(2)

178

8/13/2019 Politics Philosophy Economics 2002 Barry 155 84

26/31

last year . . . confidence in the government [of South Africa] has lagged among

business people, who are overwhelmingly white. Many have felt alienated by Mr.

Mbeki, who has focused more on rooting out racism and racial disparities than

on promoting reconciliation.41 This lack of confidence had the predictable con-

sequence: In 2000, [Mbeki] saw his popularity plummet as the economy shedjobs.42 Notice that capitalists in South Africa were able to hold the government

hostage not only on the basis of its economic policies, but also its own political

agenda of rooting out racism and racial disparities. President Mbeki, therefore,

devoted his state of the nation speech to encouraging confidence by pushing

reconciliation at the expense of more radical goals and promising trade liberal-

ization. As officials are quoted as saying, persuading local businesses to

reinvest in the economy is the priority, in the hope that strong signs of business

confidence . . . will attract foreign dollars and, ultimately reduce joblessness.43 I

rest my case.

2) This is all very well, it may next be said, but power over needs threats, and

we do not see much in the way of threats from capitalists. The first thing to say

is that we do see threats all the time: firms claim (as I mentioned in section II)

that they will leave or not make planned investments if Britain does not join the

Euro. Similarly, firms put out statements about dire consequences for employ-

ment of, say, raising the minimum wage or refusing to raise the permitted axle

weights of trucks on the roads. The second thing to say is that power over was

not defined in a way that mentioned threats. A has power over B if A can get B

to do something he would not otherwise do in virtue of Bs belief that A could

make him worse off if he does not do it. A threat by a firm will be ineffective ifit is not believed by the government; conversely, if the government believes that

a certain action will have adverse consequences because of the action of a firm,

that firm certainly has the means of power over it and may well have power over

it.

The third point to make, which is especially relevant here, is that the power of

consumers over firms and the power of voters over governments do not typically

manifest themselves in the form of threats. Consumers may complain about

faulty products to get their money back, but they rarely write disinterested letters

to firms telling them that their products are no good and that they are planning to

switch suppliers. Similarly, only a tiny proportion of the electorate writes lettersto the government threatening to vote against it unless it does something or stops

doing it. In all three cases, the means of power is the possibility of exit, and the

presence or absence of voice is essentially irrelevant.

It is noteworthy that Robert Dahl made this point in relation to voting in 1961,

the heyday of his behaviouralist phase. Why, he asks, has the Mayor of New

Haven not tried to increase taxes? It was not, I think, because someone said

Mayor Lee dont you dare raise taxes! . . . He anticipatedwhat might happen

to him in the next election if he should raise taxes.44 Dahl calls this indirect

influence, but, since he is one of those who uses power and influence as if

Barry: Capitalists rule OK?

179

8/13/201