Sie sind auf Seite 1von 19
The Empirical Side of the Power Elite Debate: An Assessment and Critique of Recent Research
The Empirical Side of the Power Elite Debate: An Assessment and Critique of Recent Research

The Empirical Side of the Power Elite Debate: An Assessment and Critique of Recent Research Author(s): Harold R. Kerbo and L. Richard Della Fave Source: The Sociological Quarterly, Vol. 20, No. 1 (Winter, 1979), pp. 5-22 Published by: Blackwell Publishing on behalf of the Midwest Sociological Society Stable URL: http://www.jstor.org/stable/4106384

Accessed: 26/03/2010 16:44

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=black.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

information about JSTOR, please contact support@jstor.org. Blackwell Publishing and Midwest Sociological Society are

Blackwell Publishing and Midwest Sociological Society are collaborating with JSTOR to digitize, preserve and extend access to The Sociological Quarterly.

http://www.jstor.org

The Sociological Quarterly 20 (Winter 1979):5-22

The Empirical Side of thePowerEliteDebate:

an Assessmentand Critique of RecentResearch*

Harold R. Kerbo, California Polytechnic State University

L. Richard Della Fave, North Carolina State University

From the earliest writings in social science there have been lively debates over the extent to which societies are dominated by elites. Recently, empirical data have been considered for elite backgrounds, elite interlock, elite unity, and elite influence on

public policy, but interpretation of the data continue to be problematic. The findings are often confusing and conflicting mainly because of differingmethodologies, defi-

nitions and indicators of elite status. Focusing on the four areas of search listed, we compare the findings in an attempt to explain some

When possible, we have prepared summaries of the consistent findings, which tend

to show, with respect to these issues, greater support for elite theories as opposed to pluralist theory. Finally we discuss some of the major questions in the debate that

current research is

quantitative re- of the conflicts.

unable to answer, and outline future researchneeds.

One of the longest running debates in the social sciences is over the nature of

power stratification, a

Marx, Mosca, Pareto, and others. Since the

however, the debate has become even more heated.

terest since the 1960s the

complex. But, contrary to what we might

of power stratification, until recently, has not brought

of systematic empirical research.

modest volume of research: first, the

sible (Mills, 1956:363); and second, the

debate that can be traced back at least to the works of

writings of C. Wright

Mills (1956),

And with the increased in-

in the debate have become more

opposing positions

expect, the new interest in the nature

with it an increased level least two reasons for the

study is often simply inacces-

question is so politically significant

We can suggest

reality under

at

values than by

Dye,

1976; Mankoff, 1970).

The result

a shift in the power elite con-

theory and empirical

the 1950s,

that theorists too often have been influenced more by political

empirical research (Bachrach, 1967;

has

decade, however, we have witnessed

been ambiguous theories that make empirical research difficult.

In the

past

troverSy from a debate dominated almost completely by

theory to one striking a research. Like the shift

allegedly due to changing value

assumptions (see Bachrach, 1967; Walker, 1966), the changing assumptions in

the late 1960s and early

Lowi, 1973:vii-ix).

ever, we find

eters to be used in

of national elites as we do with respect to

definitions, and param-

more equitable balance between

toward pluralism which occurred in

1970s have led to an increased interest in elites (see

With an increase in the volume of empirical research, how-

nearly as much disagreement on indicators,

empirical studies

01979 by The Sociological Quarterly. All rights reserved. 0038-0253/79/1600-0005$00.75

*Revision

San

their

comments on this work, and Diane Goldman for her editing and typing of the manuscript. Harold

San Luis

of

a paper presented at the

1978. We

thank

Department

of

annual meeting of

Hermann

Strasser,

Social

Sciences,

the American

Richard

Sociological

Kathy

Association,

Kerbo

for

Francisco,

helpful

R. Kerbo's

Obispo, California 93407.

September,

address is:

Shaffer, and

Polytechnic

California

State University,

6

THE SOCIOLOGICAL QUARTERLY

theory. As the debate becomes more focused on

problem of

more evident. The present work is

quantitative research on the

four key areas: elite

on public policy. In each of

to explain

empirical questions, the old

differing elite concepts and indicators (Zuckerman, 1977) becomes

primarily

devoted to a critical examination of recent

of national elites focusing on elite unity, and elite influence

compare the findings, attempting

existence and power

backgrounds, elite interlock,

these areas we will

future research needs.

differences and contradictions. Finally, we will summarize the overall

suggested by the several sets of data, discuss the theoretical im-

picture of elites

plications, and outline

Research on Elite Backgrounds

Social class ties and other

racial, sex, regional,

and educational backgrounds of elites (for example,

1976; Zweigenhaft, extent to which the

non-elite backgrounds,

dominates elite

per-

fect, research has search worthwhile

the first to take up the re-

search tradition left

found in his "sociology of leader-

ship method," which attempts to show that one of the most important means of

upper class dominance is

ciety. Thus, much of his work involves

dividuals in such positions. Domhoff specifies

membership (Domhoff

listing member ber of the

ber of the family belonging to an elite club or attending

any female mem-

an elite prep school;

is assumed if the "father was a million-

corporation executive or corporation law-

yer" and the person attended an elite prep school or belonged to an exclusive

and clubs. Identification with any one

of these five categories places an individual on Domhoff's list of upper class

background terest. Much research has been concerned with

1975),

characteristics of elites have been of in-

religious,

Prewitt and McAllister,

have involved the

but the most important questions

or closed to those from

top is open

and the extent to which one

segment of the society (such as an upper class)

the

correspondence between background, on

the other, is far from

positions. Though

the one hand, and attitudes, values, and behavior, on

G. William Domhoff (1967,

(1967:143)

by the death of C.

model of a "governing class" is

through

shown sufficiently strong correlations to make this line of re-

(Edinger and Searing, 1967; Dye, 1976:149).

1970)

was one of

Wright

Mills. The basis of Domhoff's

in the so-

holding key institutional positions

identifying

the class

also

backgrounds of in-

upper class

five major indicators of

1970:21-27;

see Domhoff 1967:33-37): (1) a

in

one of the various blue books or the Social Register; (2) any male

of the family attending

family belonging

one of the elite

prep

schools; (3) any male mem-

to one of the elite social clubs; (4)

(5)

and finally, upper class membership

aire entrepreneur pr $100,000-a-year

club on an extended list of these schools

membership.

Much of Domhoff's work is not systematic. Only with

respect to a few posi- of individuals from the

(1967:51)

tions has he been able to determine overall percentages

upper class. Table 1 shows that among the top 15 to

20 financial, industrial,

transportation, utility, and merchandising corporations, Domhoff

found an overall majority of the directors of these corporations to be members of the upper class as defined by his indicators. For the government sector, he

Domhoff

examined "key" cabinet posts. Covering a period from 1932-1964,

Empirical Side of thePowerEliteDebate

7

(1967:97-99)

eight of thirteenSecretariesof Defense to be membersof the upper

found five of eight Secretariesof State and

and Freitag (1975)

the Treasury, and class.

an impressive

Recent work by Mintz (1975)

shows

Table 1. Findings on Corporate andGovernmentElite Backgrounds

 

%

Business

Members

%

Upper

%

Elite or

of

Previous

Class

Business

Upper

Exclusive

Corporate

Study

Membership

Elite

Class

Clubs

Elite***

1. Domhoff

Corporate Elite

Directors of top

 

20

industrials

54

15

banks

62

15

insurance

44

15

transportation

53

15

utilities

30

GovernmentElite (1932-1964)

Secretaries of

State

63*

Secretaries of

Defense

62

Secretariesof Treasury

63

2. Mintz and Freitag Cabinet Secretaries (from

1897-1973)

all cabinet

66.0

Democrats

60.4

Republicans

71.3

all cabinet (with business elite

 

63.4

only before cabinet) all cabinet (with business elite

76.1

before or after) Democrats

73.6

Republicans

78.1

all cabinet (this includes business elite before or after cabinet)

 

90.0

all cabinet

(this includes business

54.6

before and after)

 

3. Dye

Government Elite**

25

Corporate Elite

30

Government Elite

 

6

Corporate Elite

Government Elite

44

26.7

*Listing in Social Register indicating upper class membership. **More expanded definition of government and corporate elite than Domhoff.

***Dye's

corporate

lawyers.

elite

excludes

corporate lawyer,

Mintz and Freitag's business elite includes

corporate

8

THE SOCIOLOGICAL QUARTERLY

amountof

systematic data concerning the

corporate andsocialclasstiesof indi-

government. They compiled infor-

viduals in the executive branch of the Federal

mation on the background characteristicsof all cabinet members serving thirteen

Presidents from

terms of their indicators of business elite

as a director or

(such as

a bank, insurance company, utility, or transportation company) in the United

States"; or an individual was considered a member of the business elite if "(1)

the person

Hubbell law directory; (2)

having been a member of such a

rative

indicators of up-

per class

"An individual was defined as a member of the social elite if he was listed in the

1897 to 1973 (a total of 205 individuals and 358 positions). In

backgrounds, "an individual was con-

position

sidered a member of the business elite if he/she held a

officer of a major industrial corporation or non-industrial corporation

was listed as a member of a corporate law firm in the Martindale-

the

person was listed in a biographical source as

corporate law firm; or (3) a biographical nar-

a major

stated that the individual had served as a lawyer representing

Mintz's (1975:133)

U.S. corporation" (Freitag, 1975:150).

membership

were similar to, but broader than, those used by Domhoff:

Social Register; had attended one of an

tory academies; had attended Harvard, Yale, or Princeton; was a member

of the 105 social clubs listed in the front of the Social Register."

Freitag and Mintz report (see Table 1) that "Nearly

ficials who held office in the

cial or business elite includes those

Further analysis shows that 54.6

were members of the business elite before the cabinet

were members of the business elite before or after, and 66

bers of the upper class before

extremely

small set of exclusive prepara-

of one

90% of all cabinet of-

were

members of either the so-

It should be noted that this figure

period

1897-1973

." (Mintz, 1975:135).

belonging

to the business elite before or after the cabinet position.

percent

were members of both, 63.4 percent

position, 76.1 percent

percent were mem-

joining

the cabinet (Mintz, 1975:135; Freitag,

lawyer

is excluded, the

if the criterion of corporate

percent (Freitag,

backgrounds

1975:141). These studies show

(before and after) for cabinet

shows all to

from a high of 100 percent

1975:151).

In

addition,

number in the cabinet with business elite backgrounds before and after the cabi-

net position only drops to 62

a steady increase in business elite

administra-

tions

have had a

to a low of 53.8 Thomas Dye's

Pickering, 1973)

vides

must begin

measures of institutional elites. Included

tors

half of

tion and

with industrial corporations, banks, insurance companies,

selects "the

president and vice-president; secretaries, under-secretaries, and assistant secre-

taries of all executive departments; White House presidential advisors and am- bassadors-at-large; congressional committee chairpersons and ranking minority

tions, and transportation. For the government elite, Dye

members and little difference between Democrat and Republican

(Freitag, 1975:142-44).

And a breakdown by

cabinet posts

majority from the business elite-ranging

percent for Secretaries of Labor (Freitag, 1975:147).

(1976;

massive

Dye and

study

Pickering,

1974; and

Dye,

DeClercq, and

of institutional elites in the United States pro-

the works of Domhoff, Freitag, and Mintz. We

indicators and elite are the direc-

and

Pickering, 1974:901-5)

among the corporate

an interesting contrast to

with

Dye's

(see

Dye

and presidents (3,572 individuals) of the top corporations that control one-

the corporate assets in the nation (i.e., beginning

working

with

the top corpora-

down until one-half of the assets are included). This is done

utilities, communica-

(1976:12)

Empirical Side of the Power Elite Debate

9

committee members in the House and Senate; House and Senate majority and

Court Justices; members of the Fed-

Council of Economic Advisors." In addition Dye

elite for most of

his analysis (total

includes

eral Reserve Board and the

minority party leaders and

whips; Supreme

top

military officers and secretaries in the

government

of 286 individuals in all, with 59 of these

from the military).

occu-

Using the above indicators, Dye presents the following findings: 26.7 percent

corporate elite

of the members of the government elite have held positions in the

(Dye, 1976:136);

16.6

percent of the government

elite

(Dye,

elite

1976:159);

list their principle

83.5 percent

of

government

in

top

pation as being in the corporate

ernment elite have held a

of the corporate elite have held previous positions

another 56.1

Using indicators of upper class similar to those of Domhoff, Dye (1976:152)

found 30 percent of

ernment elite to have upper class

sive

the gov-

law firms.

previous position in the government elite; 39.6 percent

elite; and

in the

percent of the government elite have held positions

the

class clubs from a list similar to

corporate elite and 25 percent of the non-military gov-

origins. Looking only to membership in exclu-

Domhoff's, Dye (1976:164) found

percent of the corporate elite and 6 percent of the non-military government

upper

In

44

elite to be members.

studies, their comparability must be considered. Begin-

ning with the indicators of upper class membership,

broadly similar. The main difference is that Mintz includes attendance at Har-

vard, Yale, and Princeton and an expanded list of exclusive social clubs (see

Mintz, 1975:133). Dye's indicators

claims to use

includes parents' elite institutional position

as an indicator of

In the use of exclu-

upper class membership. In another respect, however, Dye

has a more exclusive

sive social clubs as an indicator, Domhoff includes four clubs not used by Dye

Finally, Domhoff's listing of elite prep

sent greater problems

Domhoff's and Mintz's are

comparing these

of

upper class membership, however, pre-

(1976:151)

of comparability. While Dye

upper

class

1970:23).

Domhoff's basic indicators, he also

definition of

(see Dye, 1976:164; Domhoff,

1970:22-33;

Dye, 1976:154).

membership.

schools is slightly more inclusive

It is

tance of Dye's differing indicators of

do

sions.

than Dye's listing-37

versus 33 (Domhoff,

impossible to estimate precisely the impor-

but the differences fairly firm conclu-

upper class membership,

not appear large enough to prevent our arriving at some

Turning to indicators of corporate elite status,

corporations; 33 of the top

backgrounds

of

Dye

and Domhoff differ in that

Domhoff looks only to the 15 or 20

20) looks to all of those controlling half of the assets in the

of the top industrial

transportation,

communications, and utilities; the top 50 banks;

panies). This is not a serious problem, however, and in fact

those at different "levels" of corporate

elite status. Another difference is with Dye's inclusion of corporation directors

and presidents in the business elite because Domhoff includes only the directors

of the corporations. The most serious obstacles to comparability, however, are

with the work of Mintz and of Freitag who include all top corporations (though

they never say exactly how many

consist of the directors and "officers" of these

corporations (again, they do not mention how far they go in including officers-

wealthiest corporations while Dye (1976:

particular area (100 in

18 top insurance com-

corporations

and the

gives us an oppor-

tunity to compare the

they include in the top-see

Freitag, 1975:

140)

and consider the elite to

10

THE SOCIOLOGICAL QUARTERLY

see

the business elite while

corporate

determined

We can see

dropped only

from 76.1 percent to 62 percent when corporate lawyers are excluded (Freitag,

however, when we note that the cabinet ties to the business elite

Freitag,

Also, Mintz and Freitag include corporate lawyers in and Domhoff do not. The effects of the inclusion of

Dye lawyers by Mintz and Freitag, and their exclusion by Dye, cannot be

1975:140).

(mainly

that this has

because of Dye's

only

larger definition of the government elite).

a minimal effect on Mintz and Freitag's findings,

1975:141).

Finally, we must consider the differing indicators of membership in the gov-

at a few cabinet positions (from 1932

ernment elite. While Domhoff looks

to 1964),

while Dye has

they nificant problem in our

the

There

tary as compared to

the

have

only

Freitag and Mintz include the entire cabinet (from 1897 to 1973),

an inclusive definition. These differences should not concern us

refer to

differing levels within the government elite. The only sig-

estimation, is that Dye's

indicator is

because

too broad, including

military elite as part of the government elite throughout most of his analysis.

background of mili-

154, 159). Thus, because

percent of his total government elite when possible we

are clear differences in the recruitment patterns and

civilian leaders (Dye,

1976:152,

from

the figures presented here.

we can draw the following

military makes up 21

dropped the military

Despite these problems,

by Mintz's study; (2)

conclusions about the Domhoff's estimation of

is rein-

backgrounds of corporate and government elites: (1)

a high proportion of cabinet members having

forced

terms, the higher the percentage recruited from

was suggested, but not followed up by

example, see Keller, 1963:209-10,

upper class background

As the corporate elite is

earlier studies

319);

(3)

defined in more restrictive

upper class backgrounds (this

of elite backgrounds-for

Viewing the President's Cabinet

in

high percentage of in-

as the top of the government

dividuals from the upper class (66

government elite).

dicators of a government elite are used; (4)

backgrounds among members of the

appear more prevalent

is (63.4 percent to 26.7

more inclusive the definition of the corporate elite is (i.e., Freitag's definition).

elite, Mintz's data show a

percent to 25 percent

Dye's expanded

increase as more limited in-

Thus, upper class backgrounds

government

the more restrictive the

When looking to corporate elite

elite corporate backgrounds

definition of the government elite

and the

percent with Dye's larger government elite),

Elite Interlock

Elite interlock can be described as the situation in which an individual simul-

taneously holds two or more elite

elite

the previous characteristics of, or

ual (a

with the overlap between elite

examination of elite interlocks is a

tional power distribution because the more numerous the interlocks, the greater

the number of key positions held by the same individuals; thus, the more power is concentrated in the hands of a few individuals or positions. Dye (1976:134) examined the extent of interlock between the government

positions.

This is different from the study of

backgrounds because in the case of

dynamic view);

elite backgrounds we are dealing with

institutional positions held by, an elite individ-

but with the study of elite interlock we are concerned

in time (a static view). An concern for students of na-

positions at one point

subject of major

Empirical Side of the Power Elite Debate

11

and other elite

positions and greatly emphasized the fact that only

elite positions in his

percent) are in

positions which will result in

paper.

19.4 percent

study what he calls the public interest sector. This is

are so interlocked. Most of these

of the government

positions (91.2

hardly surprising when we consider that

holding

between government and corporate elites, one must, therefore, examine elite

some of the research-

backgrounds (or dynamic interlocks) as has been done by

For this reason we will restrict

government officials are prevented from

conflicts of interest. To study ties

other

ers discussed in the previous section of this

our concern in this

section to interlocks within the economic sector.

As Sonquist and Koening

(1975:196)

put it, "It is important

to focus atten-

of the rela- dynamics of

tion on the specific nature of interlocks because an understanding

tionships between

power and control in

critical when we consider the extent to which the economy is dominated increas-

ingly by a smaller number of

20) data show, in 1970, more than 50

greater importance

of interlocks with top corporations is also

that the corporations on top today, contrary

more likely to stay question of economic

rate investors are found in the

corporations (Subcommittee of Governmental Affairs, 1978:1). Thus, the num-

ber of interlocks

recent research, we note that

were interlocked with other top positions"

involving incumbents of top corporate

with other

sectors. Though they would be interesting, no figures are given by Dye on how

many know is that while 40

lic interest sectors) account for these

elite positions held by corporate

Dye's data include

the corporate elite are interlocked with corporations

tion, it would be interesting

directors

interlocks

below the top 100. In addi-

only the top 100 corporations, he cannot say how many of

interlocking firms can tell us much about the

the U.S. political economic system." This is especially

giant corporations. For example, as Dye's (1976:

percent

of the industrial assets are con-

corporations. The

1969:536).

indicated by research which shows

to those on top in the past, are

Also important in this

that just 21 corpo-

trolled by 100 of the more than 200,000

on top (Mermelstein,

concentration is the new data showing

these

(1973,

top 5 stockholders in over half of the top 122

top corporations becomes increasingly important.

1974,

1976)

percent) rather than

What we do

among Turning first to Dye's

"44% of all (see Table 2).

top corporate positions

Most of the interlocks

positions were within the corporate sector (72

individuals account for this 44

percent interlock in positions.

percent of all positions

(government, corporate, and pub-

individuals

are interlocked, only 20 percent of the 4,101

interlocked positions; and there is a lifetime average of 11.1

elite individuals (Dye, 1976:130, 135). Because

to know the amount

of interlock among corporate

only (excluding corporate presidents), and the pattern and types of

among

the

various corporations.

(1974)

answers a few of these questions. Allen

examined the amount of interlock among the directors of the 200

financial institutions and found

and financial institutions had an aver-

was also divided to show that the average

was 16.92 for the financial institutions, 9.62 for the industrial corporations, and

7.41 for the remaining nonindustrial corporations. Also of interest is that the

amount of assets held was strongly correlated with the number of interlocks

even when the differing size of the directorate is controlled (.49, larger

age of 10.41 interlocks in 1970. This

A recent study by Allen

(1974:399)

largest nonfinancial

(see our Table 2) that these

corporations and 50 largest

corporations

(.57),

12

THESOCIOLOGICAL QUARTERLY

Table 2. Economic Elite Interlock (Static View)

Study

Findings

1. DYE

Total Corporate Interlocks* (positions with interlocks) Of Total Interlock

44.0%

percent to

percent to

percent to

other corporate positions

72.7%

public interest positions

25.9%

government positions

0.2%

2. ALLEN (1970 Data)**

 

Average

interlocks of

corporations

10.41

Average

interlocks of

financial institutions

16.92

Average

interlocks

of

industrial corporations

9.62

Average

interlocks

of

non-industrial corporations

7.41

3. DOOLEY (1965 Data)**

 

Average

interlocks of

corporations

9.9

Average

interlocks of

financial institutions

15.2

Average

interlocks

of

industrial corporations

9.1

Average

interlocks

of

non-industrial corporations

8.6

Average interlocks to size of corporation*** less than .5 billion assets

6.0

1.0

to

1.4 billion assets

6.8

1.5

to

1.9 billion assets

9.2

3.0

to

3.9 billion assets

16.4

5.0

and over billion assets

23.7

*Top 100 industrial, 50 banks, 18 insurance, 33 transportation, communication, and utilities in 1970 (board of directors and presidents).

**Top 200 non-financial and top 50 financial corporations

***Selected

categories from Dooley's

table.

(board of directors).

corporations having more interlocks). (See Allen, 1978 which was in press simultaneously with this paper.)

who collected data on the

(200 non-

with a similar set

financial and 50

interlocks among the boards of directors in the top 250

Another interesting study

is

Dooley (1969:314)

corporations

1965

financial). By comparing his data

from

for 1935,

Dooley

in 1935 and 223

found that the number of interlocks were about the same (225

we find that larger corporations have

Dooley

(1969:316)

larger corporations

we find more interlocks.

Also of

corporation, the average

is greatly increased.

number of interlocks with remain-

Soref's (1976) findings confirm that

through ownership, rather than executive posi-

likely to have upper class backgrounds, are

in 1965, Dooley, 1969:315). In comparing Dooley's data with

Allen's (for 1970)

lists a progressively in-

(see our Table

2). Thus, much like the findings on the upper-class backgrounds of elites, as we

move closer to the top

interest is Dooley's

that by excluding board members who also hold executive

positions within

ing board members

more interlocks than

smaller ones (see our Table 2).

creasing number of interlocks from smaller to

(1969:317)

finding

the

these men whose power comes

tion in the corporation, are more

Empirical Side of the Power Elite Debate

13

more concerned with major decisions, and are able to spread a web of influence

the

to a larger number of corporations.

debate between those who argue that the board of directors have more power

(see Soref, 1976:360) and those who

more power (see Pfeffer, 1972; also see Zald, 1969; Seider, 1977).

findings show the number of interlocks in-

than for in-

that the executive officers hold

We are, however, nowhere near

argue

higher

1970

for financial (15.2)

solving

Finally, Dooley's

(1969:316)

volving financial institutions are much

dustrial (9.1).

cial interlocks (16.9

average). This

stitutions in the

Allen

terms of assets) of sphere of influence is

number of interlocks) in his 1965

have a

tions. And

interlocks" (Bunting, 1976a,b),

not directly but through interlocks to a common third corporation. These are

found

finds these groups

identifying 15 "tight-knit" groups (in terms of

financial institutions. But, the fact remains that their

Allen's (1974:399)

data show even more extensive finan-

average), with the 1935 financial interlocks least frequent

seems to indicate the growing importance of financial in-

(14.8

economy (see Subcommittee on Governmental Affairs, 1978).

argues,

however, that this is related to the greater size (in

the

greater.

In

data, Dooley (1969:320)

at the centers or hubs are found top financial institu-

on what can be called "indirect

are tied,

(1974:399)

recurring pattern:

also of interest is the recent

findings a situation in

which two corporations

to be increasing recently, and with a

majority of the host corporations

(third corporation or midpoint) in the financial sector (Bunting, 1976a:34).

In related research, Sonquist

top

and Koening (1975:204)

797 corporations in 1969

number to 401 corporations by

Using Levine's (1972)

including only

sociometric

central cliques (Sonquist

began their work with

Fortune's

those with at

least double interlocks.

of corporations were identified: isolates, trivial dyad

and

Of the last, 32 major

central

contained in their appendix). Of special interest are the financial cliques. These

tend to expand outward toward satellite cliques, while the non-financial cliques

were more self-contained. One of the most

creasing

stitutions have influence

also

financial in-

on the in-

later reduced this

(including 11,290 directorships), but

method, four types

members, satellite cliques,

(details on all of these are

and Koening, 1975:206).

further analysis

cliques were identified for

important recent findings

economy shows that

power of financial institutions in the

not only through interlocks with other corporations but

through the control of voting shares of stock in major corporations (Sub-

Affairs, 1978).

companies held 34.8

study,

committee on Governmental

The proportion of stock ownership

corporations who often

In 1976, banks, investment companies, and

percent of the voting rights of all corporate stock

This

of course brings up an interesting question of who controls the voting stock in

the Subcommittee on Governmental Affairs

(1978:260)

banks."

in the United States (Subcommittee on Governmental Affairs, 1978:14).

insurance

hold voting rights with this stock.

by institutions is increasing rapidly, and it is financial

the major banks. In their

found, "the principal stock voters in large banks are-large

to be Morgan

top

banks

and Company which is the number

The most important seems

in the nation (Bankamerica, Citicorp,

Manufacturers Hanover, Chemical N.Y. Corporation, and Bankers Trust N.Y.). In comparing these studies we find few problems with incompatible indicators such as those encountered with the elite background research (with the excep- tion of Dye's research which includes corporate executives with directors). The

one stock voter in five of the

14

THE SOCIOLOGICAL QUARTERLY

nature of corporate interlocks include

directors is fairly

high, and has remained so through this century; (2)

executives in the corporation are more likely

interlocks; (4) A few powerful and

important cliques can be identified within this mass of interlocks, with large

The average

most significant findings relating to the

the following: (1) The extent of interlock

among corporate

Directors who are not also

to be involved in these interlocks;

(3)

The larger the corporation the more

financial institutions often at the center of these cliques; and (5)

number of interlocks leading from financial institutions has been growing stead-

ily, at least since 1935.

Elite Unity

The existence of elite

unity or cohesiveness is of central importance for those

this impor-

tance we would expect much research on the question,

We might conclude that some basis for

showing

agree that common background alone is not enough. With the exception of a few

example Baltzell, 1958), little has been done empirically on the recent work of Domhoff.

Domhoff's (1974) most noted research in this area, in which he uses qualita-

this question until

case studies (for

when recalling the findings

who argue from a ruling class or "governing class" position.

unity

exists

the upper class backgrounds of many

Given

but that is not the case.

elite individuals. But both sides

tive and quantitative methods, is contained in his study of San Francisco's Bo-

hemian Club and

this organization. Through an

cent to be members of Union Club. He

upper class. In addition, Domhoff (1974:30)

resident regular members" listed in other upper class "blue

Domhoff

show "that at least one officer or director from 40

corporations in America was

year,

25 life insurance

top

sented' by at least one officer or director" (Domhoff, 1974:32).

Retreat. He attempts

argues

that 38

interprets

percent

as a

to demonstrate the upper class nature of

examination of the membership he finds 27 per-

the "most exclusive club" in San Francisco, the Pacific

of the resident members belong