Cognitive Maps of Class and Racial Inequalities in England and the United States

Author(s): Wendell Bell and Robert V. Robinson

Source: American Journal of Sociology, Vol. 86, No. 2 (Sep., 1980), pp. 320-349
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/2778667 .
Accessed: 18/07/2013 13:01

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
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 [email protected].

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to
American Journal of Sociology.

http://www.jstor.org

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
CognitiveMaps of Class and Racial Inequalitiesin
Englandand theUnitedStates1
WendellBell
Yale University

RobertV. Robinson
Indiana University

A 1975 exploratory studyis used in thisinvestigation of the cogni-

tivemaps of class and racial inequalitiesheld by 113 Americanand
101 Englishrespondents. An Index of PerceivedInequalityis con-
structedfromnineitemsdealingwithinequalities, bothamongclasses
and betweenraces, of education,occupation,income,respect,and
treatment by the police and courts.The indexappears reliableand
valid,particularly forthe UnitedStates.For theUnitedStates,fac-
tor analysesshow that perceptionsof class inequalitiescannotbe
separatedempiricallyfromperceptionsof racial inequalitieson the
basis of theirinterrelationships.
Thus a singledimensioncan account
forthe commonvariationof the nineitems.But in Englandpercep-
tionsof class and racial inequalitiestend to separateinto two dis-
tinct,thoughcorrelated,clusters.Additionally,for both countries
thereis supportforgroupingthe itemsinto thosedealingwithin-
equalityof opportunity and thosedealingwithinequalityof treat-
mentand condition.Finally,comparedwiththe English,Americans
perceivedmoreinequality,were morelikelyto see a growingeco-
nomicgap betweenthe richand the poor,saw a largernumberof
socialclasses,and weremorelikelyto say thatmoneyis thedefining
criterionof class.

The idea that inequalitymustbe perceivedbeforeit can becomea basis

mobilizationand action is containedin Marx's well-
for self-interested
knowndistinction betweenKlasse an sick and Klasse fursick,thatis, be-
tweenmembersof a class who are not yet consciousof theircommonpo-
1 We gratefullyacknowledgefinancialassistancefromtheNationalInstituteof Men-
tal Health(grantno. 5-T01-MH12133), whichsupported theYale Comparative Soci-
ologyTraining Program, 1969-77,of whichthisstudyis part,and fromtheConcilium
on International and Area Studies,Yale University.In addition,we thankJanet
Grigsby, CandaceM. Kruttschnitt,and H. WesleyPerkinsfortheirimportant contri-
butions;MichaelMann of the London School of Economicsand PoliticalScience,
England,forhis adviceon designing the study;FrancisJ. Anscombeforstatistical
RobertH. Ross,who made thelargestcontribution
counsel;and, especially, to for-
mulatingthe itemsreportedhereto measureperceivedinequality. We also wish to
acknowledge thehelpful of SusanOlzak,HenryA. Walker,and twoanony-
suggestions
mousAJSreviewers aboutan earlierversionof thispaper.
? 1980by The University
of Chicago.0002-9602/81/8602-0005$01.50

320 AJS Volume86 Number2

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

sition,interests,and historicrole and thosewho are. Explicitlyacknowl-

edginghis indebtednessto Marx, Centers(1949) so dramaticallydocu-
mentedthe importanceof class consciousnessand interestsin dividing
membersof societythat class identification or subjectivesocial class has
becomea standarditemin muchsociologicalresearch.2 Additionally,per-
ceptionsof the whole class systemand of the criteriaunderlying them
have receivedattentionat least since the classic Yankee City studies
(Warnerand Lunt 1941). Attentionhas continuedthroughthe workof
Popitz et al. (1957) in Germany,Willener(1957) in Switzerland,Bott
(1957), and, morerecently,Goldthorpeet al. (1969) in Britain,up to
therecentstudyof social standingin Americaby Colemanand Rainwater
(1978). Moreover,subjectiverankingsof theprestigeof occupationshave
now become an importantpart of sociologicalknowledge,3 and Broom
et al. (1978) remindus that even the burgeoning studyof mobilityhas
been largelydependenton sons' proxyreportsof parentaloccupational
statusesin whichperceptions-orpossiblymisperceptions-may play a
part.
Recently,researchershave givenrenewedattentionto othersubjective
reactionsto societalinequality,such as the ideologyof class differences
and social judgmentsof the justiceor fairnessof varioussocial equalities
and inequalities(Alves and Rossi 1978; Bell 1977; Bell and Robinson
1978; Jasso and Rossi 1977; Laumann and Senter 1976; Mann 1970;
Robinsonand Bell 1978; Tallmanand Ihinger-Tallman 1979). Such stud-
ies aim to measurethe evaluativemaps peoplehave of theinequalitiesof
thestratification systemin theirsociety.Moreover,theylinkthedistribu-
tionalfactsof inequalitieswitha senseof injusticeand the accompanying
moraloutragethancan lead, undersomeconditions, to protestand revolt
(Gurr 1970; Moore 1978; Nagel 1974).
Yet, obviously,one cannotfeeloutrageover the injusticeof inequality
unless one believes that inequalityexists.Althoughsome writers(Bell
1962; Form and Rytina 1969; Huber and Form 1973; McClosky1964;
Mau, Hill, and Bell 1961; Seeman 1960; and Seemanand Morris,n.d.)
have proposedmeasuresof perceptionsof inequality,to date theirwork
has notbeen followedup adequately.
In thispaper we endeavor,first,to advance such workby constructing
and testingin a preliminaryway a newindexof thecognitivemapsor per-
ceptionspeople have of class and racial inequalitiesin theirsociety,an

2Workinga few years later than Centers,Martin (1954) demonstratedthe importance

of class identification
in Britain.
3 Treiman (1977) counts more than 100 studies of occupational prestigein 60 differ-
ent countriesup to 1972. Other noteworthystudies that include images of the class
structureare Bulmer (1975), Curtisand Jackson (1977), McKenzie and Silver (1968).
Moorhouse (1976), Nordlinger(1967), and Runciman (1966).

321

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

Index of PerceivedInequality.No value judgmentsas to the fairnessor

unfairness of theseinequalitiesare involvedin thismeasurement. Instead,
whatis measuredis howpeopledepictor reconstruct thesocial realitiesof
equalitiesand inequalitiesaroundthem.
Second,we attemptto determine empiricallywhetherperceptions of in-
fromperceptionsof inequal-
equalitiesof social class can be disentangled
ities of race. Otherresearchers(Kluegel 1978; Wright1978), of course,
have asked questionsconcerningthe effectsof class relativeto race on
somedependentvariable.Additionally, manyresearchers have foundthat
race and class are so oftencorrelatedas to requireno special reference.
We ask different questions:To what degreeare inequalitiesof class and
race (1) perceivedby membersof societyas a singleset of covariantin-
equalitiesor (2) viewedas separatedimensionsof stratification, varying
independently?
Third,we enlargethe cognitivemaps of inequalityas measuredby the
Index of PerceivedInequalityby reportingadditionalaspects of class
imagery.For example,we ask abouttheperceivedchangesin theeconomic
differences amongclasses,thenumberof social classespeopleperceive,the
criteriapeopleuse to definesocial classes,and respondents'generalimages
of theclass structure.
Fourth,and finally,we compareperceivedinequalitiesin two societies,
Englandand the UnitedStates.These twocountriesare well suitedto our
purposesin thatmuchhas beenwrittenaboutthenatureand extentof in-
equalityin them.Moreover,despiteimportant theirhistorical
similarities,
conditionsand culturaltraditions in significant
have differed waysregard-
ing inequality.Since Tocqueville'sDemocracyin Americawas published
over a centuryago, it has becomecommonplaceto contrast"classbound
and inegalitarian"Britainwith"classlessand egalitarian"America.Marx
([1867] 1962, p. 449), for example,referredto Britainas the "classic
ground"of capitalismand based his abstractmodelof class conflict largely
on the historicalBritainand the Britainof his day (Giddens 1973). In
contrast,he saw the United States as lackinga strongconsciousnessof
class,sinceclassesthere"have notyetbecomefixedbut continually change
and interchange theirelements"([18521 1972, p. 444). More recently,
Alford(1963, p. 109) commented thatBritishworkersare moreclass con-
scious than Americanworkers,and Lipset (1963, p. 517) observedthat
"the UnitedStatesemphasizesachievement, equalitarianism,universalism,
and specificity," while Great Britainemphasizesthe oppositevalues of
ascription,elitism,particularism, Recent researchin the
and diffuseness.
two countrieson objectivestratification (Treimanand Terrell1975) and
class consciousness and voting(Alford1963; Robinsonand Kelley 1979;

322

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

Rose and Urwin 1969) generallysupportstheseaccounts.4None of the

to in thisparagraph,however,consideredperceptions
authorsreferred of
inequality.

DATA

The data on whichthisstudyis based are interviews averagingabout an

hourin lengthwithpersons18 years of age or over in England and the
UnitedStates.Interviewing in bothcountrieswas doneduringthesummer
of 1975. In the UnitedStates,interviews wereconductedin the citiesof
East Haven, New Haven, West Haven, and Woodbridge,Connecticut,
whichwerechosento providea sampleof individualsrepresenting a wide
range of occupational,ethnic,and racial backgrounds.Using the New
Haven City Directory,1974 as our samplingframe,we selectedevery
Nth name startingwitha randomnumberfrom1 to N. We interviewed
113 personsfully-73.4% of thepotentialrespondents afterpersonswho
had changedaddressleavingan unoccupieddwellingunitor weretoo old
or ill to be interviewed
had been removed.
In London,using statisticsprovidedby the PopulationStudiesOffice
of the GreaterLondon Council,we chose seven wards whichwere most
comparablein occupationand race with the four GreaterNew Haven
cities. As our GreaterLondon samplingframewe used the Registerof
Electorsin each appropriate boroughhall. For each ward,theRegisterin-
cluded all registeredvotersage 18 and over and was up to date as of
October10, 1974. It is reasonablyaccurate.Voterregistration cards are
hand deliveredto all householdsin mostLondonboroughsand containa
messagethatfailureto fillin the formis punishableby law. Immigrants
fromCommonwealth countriesare allowedto registerto voteimmediately
upon arrival.As in the UnitedStates,we selectedeveryNth name with
a randomstart.
The wardswereKensal Rise,Kenton,Sudbury,and Tokyngtonin Brent
Borough;East Barnetand GardenSuburbin BarnetBorough;and Rus-
kin in SouthwarkBorough.We interviewed a totalof 101 respondents in
theLondonarea fora 73.2% responserate.
The sample areas in GreaterNew Haven contained18.3% nonwhites

4We should acknowledge that some writershave found less inequality in England
than in the United States. For example, Bornschierand Ballmer-Cao (1979, p. 503),
reportingdata for the early 1970s, give Gini indexes of inequality of 40.90 for the
United States and 34.61, showingless inequalityin relativeterms,for the United King-
dom. These Gini indexes compare with 60.20 for Honduras, the highestin the sample
of 50 countries,and 29.08 for Taiwan, the lowest. Also, data on sectoral income dis-
tributionsin the mid-sixtiesshowed less inequalityin the United Kingdom than in the
United States (Taylor and Hudson 1972, p. 264).

323

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

and 23.8% professionals, managers,and kindredworkers(PMOs) accord-

ing to the 1970 census.In the sampleareas of GreaterLondon the com-
parablepercentagesfor1971 were 14.0% and 20.6%. The UnitedStates
sampleareas werecloserto the nationalaveragesthan the Englishones:
of the 1970 totalpopulationin the UnltedStatesnonwhites were 12.5%
and PMOs 23.2%, while forEnglandand Wales only 2.5% of the total
populationwas nonwhitein 1971 and 9.9% was classlfiedas PMOs.
Thus, forthecross-national analyses,we cannotassumethatour results
correspond to an overallunbiasedcomparisonof Englandand the United
States,since neithersample,especiallythe Englishone, is representative
of theentirecountry,Nonetheless, comparisons can be meaningful in that
we have minimizedthe effectsof racial and occupationaldifferences be-
tweenthe twocountries.Furthermore, we add statisticalcontrolsforoccu-
pation,race,and sex in assessingdifferencesbetweencountries, thuselimi-
natingpossiblebiases whichmightarisefromthefailureto matchsamples
perfectly.
We stressthatthisis an exploratory study,based on smallsamples.Yet
thesmallsamplesizesare balancedby in-depthinterviews w'thopen-ended
questionsand probesspaced throughout whichenabled us to pay close
attentionto the respondents' framesof reference,Also,mostof the inter-
viewershad helped designthe studyand were to code and analyze the
data.5

AN INDEX OF PERCEIVEDINEQUALITY
To measurerespondents' perceptions of inequalityor equalitywe propose
the Index of PerceivedInequality(IPI). Table I givesthenineitemscon-
stitutingthe IPI.6 The itemsare constructedso as to tap respondents'
perceptions of inequalitiesof both class and race on a wide rangeof di-
mensions.In fiveitemsrespondents are asked to comparepeopleof d'iffer-
ent social classes accordingto theirequalityor inequalityof (no. 1) re-
spect,(no. 2) treatment by thepoliceand courts,(no. 3) good jobs, (no.
6) a collegeor universityeducation,and (no, 9) a good income,
The remainingfouritemsdeal with respondents'comparisonsof the
equalityor inequalitybetweenracialgroups,as definedby color,repeating
fourof the distributed characteristics:(no. 4) good jobs, (no. 5) treat-
mentby the policeand courts,(no. 7) a collegeor university education,
and (no. 8) income.7
5 A fulldescriptionof the data is ava'ilable on request (Bell and Robinson 1976).
6 The correlationsbetweenall pairs of itemsare given in AppendixA.
T Althoughfor the sake of symmetrywe would have liked to include a RACE-RE-
SPECT item,we decided that some respondentswould thinkit inappropriate-perhaps
a person'srace.
insulting-to ask about the amount of respectthat derivessolely froma

324

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 TABLE 1
PERCENTAGE DISTRIBUTIONS OF RESPONDENTS ACCORDING TO ITEMS
COMPOSING THE INDEX OF PERCEIVED INEQUALITy, ENGLAND
AND THE UNITED STATES, 1975

United Mean Difference

England States (UnitedStates
theIndexofPerceivedInequality
ItemsComposing (%) (%) minusEngland)

peopleofhighersocialclassaregivenmuchmore
1. " In America[Britain],
respectthanthoseofa lowersocialclass" (CLASS-RESPECT):
agree(perceivesmostinequality).
Strongly 32 53
Somewhatagree., 49 34
Undecided. 3 5
Somewhatdisagree.12 6
disagree(perceivesleastinequality)
Strongly 4 2

Total.. 100 100

Cases (). .101 113

Mean* , 2.93 3.30 +?37
2. " Generally,people ofIhighersocial classes in this countryget easier
treatment by thepoliceand thecourtsthanpeopleoflowerclassesdo"
(CLASS-LAW):
Strongly agree(perceivesmostinequality). 25 51
Somewhatagree.31 22
Undecided .13 12
Somewhatdisagree .14 9
Strongly disagree(perceivesleastinequality) 17 6

Total..100 100

Cases (N) . . 101 113

Mean. 2.34 3.04 +.70
have
3. "In actual practice,peopleof lowersocial class background the
same chanceto get good jobs as peopleof higherclass backgrounds"
(CLASS-OCCUPATION):
Strongly disagree(perceivesmostinequality) 32 35
Somewhatdisagree .25 27
Undecided . 4 9
Somewhatagree.21 13
Stronglyagree(perceivesleastinequality). 18 16
Total...... . 100 100

Cases (N) .... 101 112

Mean. 2.33 2.51 +.18
as
4. "These days blacks [colouredpeople]have the same opportunity
whitesto getgoodjobs in America[Britain]"(RACE-OCCUPATION):
Strongly mostinequality)
disagree(perceives .. 15 20
Somewhatdisagree . 23 20
Undecided. 7 10
Somewhatagree.. . 28 18
Strongly ....
agree(perceivesleastinequality) 27 32

Total.... , , . .. I ..... 100 100

Cases (N) . 100 113

Mean.. 1.71 1.80 +.09
5. "How about race?tOverall,whitesget easiertreatment by the police
and thecourtsthanblacks[coloured people]do" (RACE-LAW):
Strongly agree(perceivesmostinequality) 4 28
Somewhatagree.,.. 22 13
Undecided .17 15
Somewhatdisagree . 27 25
Strongly disagree(perceivesleastinequality) 30 19
Total. 100 100

Cases (N) . .99 111

Mean.. I 1.44 2.05 +.61
* The meansofeachitemarebasedon scoring from
theresponses 0 to 4, withtheperceptionofthemost
thehighest
receiving
inequality score.
questions
following
schedule
t Becausetheseitemsappearedin theinterview on class,theyhavea lead
phraseaddedtomaintain logicalcontinuity.

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 TABLE 1-(Continued)

United Mean Difference

England States (UnitedStates
Items Composingthe Index ofPerceivedInequality (%) (%) minusEngland)

6. "In fact,in America[Britain]all persons,regardlessofsocial class back-

ground,have an equal opportunityto get a collegeor universityeduca-
tion" (CLASS-EDUCATION):
Stronglydisagree(perceivesmostinequality) 15 23
Somewhatdisagree........................ 15 10
Undecided............................... 4 3
Somewhatagree........... 19 27
Stronglyagree (perceivesleast inequality).... 47 37

Total .................................... 10G, 100

Cases (N) .................................. 101 113

Mean .................................... 1.31 1.54 +.23
7. "And what about different races?tBlacks [coloureds]have thesame op-
portunityas whites to get a college or universityeducation" (RACE-
EDUCATION):
Stronglydisagree(perceivesmostinequality) 9 20
Somewhatdisagree ..................... 15 13
Undecided............................... 9 2
Somewhatagree.......................... 26 30
Stronglyagree (perceivesleast inequality) .. 41 35

Total .................................... 100 100

Cases (N) .................................. 101 113

Mean .................................... 1.24 1.51 +.27
8. " In America[Britain],a whitepersonusuallymakesmoremoneythan a
black [coloured]person for doing the same job" (RACE-INCOME):
Stronglyagree (perceivesmostinequality).... 5 19
Somewhatagree.......................... 19 23
Undecided............................... 9 12
Somewhatdisagree............. ............ 24 26
Stronglydisagree(perceivesleast inequality) 43 20

Total .................................... 100 100

Cases (N) .................................. 99 112

Mean .................................... 1.20 1.94 +.74
"
9. Regardlessofsocial class, anyonewillingand able to workhas an equal
chance to earn a good income in this country" (CLASS-INCOME):
Stronglydisagree(perceivesmostinequality) 5 18
Somewhatdisagree........................ 11 13
Undecided............................... 3 3
Somewhatagree.......................... 27 28
Stronglyagree (perceivesleast inequality).... 54 38

Total .. . .100 ..................... 100

Cases (N) .................................. 100 112

Mean .................................... .86 1.45 +.59

t Because theseitemsappearedin theinterviewschedulefollowingquestionson class,theyhave a lead

phraseaddedto maintainlogicalcontinuity.

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

We will use in the textthe identifying labels appearingin table 1. In

each item,"CLASS" or "RACE," designatingthe social categoriesbeing
compared,is combinedwith "RESPECT," "LAW," "OCCUPATION,"
"EDUCATION," or "INCOME," designatingthe "good" beingdistrib-
uted.
The itemsare furtherdifferentiated by the aspect of equalityor in-
equalitywithwhichtheydeal. Five items(nos. 3, 4, 6, 7, and 9) deal with
therespondents' perceptions of equalityor inequalityof opportunity or the
chanceto obtainthe "good" specified.Two (nos. 2 and 5) deal withthe
treatment receivedby members of differentclassesor races.And two (nos.
1 and 8) deal withcondition,that is, the currently existingdistribution
of the specified"good." Note thatRACE-INCOME (no. 8) is distinctive
in thattheperceivedinequalityor equalityof incomeby race is for"doing
thesamejob," thatis, withoccupationcontrolled.
In orderto minimizeacquiescencebias,we wordedtheitemsso thatfive
statementselicit rejectionand fourelicit acceptancewhen a respondent
perceivesinequality.The responsesto each itemare scoredfrom0 to 4,
fromleast to mostperceivedinequality.
In table 1, note theorderof theitemsaccordingto themean response.
In bothcountriesrespondents weremostlikelyto see inequalityof respect
by social class,81% in Englandand 87% in the UnitedStatessomewhat
or stronglyagreeingthat such inequalityexists.Unequal treatment ac-
cordedto different social classes by the police and courtswas next,with
56% of the Englishand 73% of the Americansseeinginequality.This
was followedby the chance forpersonsof different social classes to get
a good job, with 57% of the Englishand 62% of the Americansbeing
awareof inequality.
In Englandthe nextfiveitemsrankedfrommostto least perceivedin-
equalityas follows:RACE-OCCUPATION, RACE-LAW, CLASS-EDU-
CATION, RACE-EDUCATION, and RACE-INCOME. In the United
States,however,theywererankedsomewhatdifferently, the RACE-LAW
and, especially,the RACE-INCOME meansbeingrelativelyhigher.
Respondentsin both countriesperceivedleast inequalityin CLASS-
INCOME, withonly 16% of theEnglishand 31% of theAmericansdis-
agreeingwiththestatementthat"regardlessof social class,anyonewilling
and able to workhas an equal chanceto earn a good incomein thiscoun-
try."'
Comparingthe class withthe racial inequalitieson the fourdimensions
forwhichcomparison is possible(CLASS-LAW withRACE-LAW,CLASS-
OCCUPATION withRACE-OCCUPATION, CLASS-EDUCATION with
RACE-EDUCATION, CLASS-INCOME withRACE-INCOME), we see
that respondents in both countriesperceivedmore class than racial in-

327

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

equalitywith respectto law, occupation,and educationbut moreracial

thanclass inequalitywithrespectto income.
Thereis littleevidencethatgreaterinequalityof conditionor treatment
thanofopportunity is perceivedin eithercountry,despitethefactthatthe
valuesystems ofbothcountries-especially theUnitedStates-favorequal-
ityofopportunity butpermitsomeinequalitiesof conditionand treatment.
Overall,the degreeto whichinequalitiesof varioussortsare perceived
is remarkably similarin the twocountries.The Spearmanrank-order cor-
relationis .87.
Comparingthe two countriesitemby itemin table 2, one can see that
in everycase the mean perceivedinequalityis greaterin the American
thanin the Englishsample.Of the fourvariablesused in bothclass and
race items,incomeand law showlarge and statistically differ-
significant
encesbetweenthe countrieson bothclass and race. The othertwo,occu-
pation and education,both of whichdeal with equalityof opportunity,
show statisticallynonsignificant differenceson both class and race. The
largestdifferencebetweenthecountriesis in thecase of RACE-INCOME,
a conditionwithoccupationcontrolled.
Thus, the Englishwereconsiderably morelikelythanthe Americansto
perceivetheirsocietyas egalitarianboth in incomeand in treatment by
the police and courtswith regardto both social classes and races. Also,
somewhatmoreof theEnglishperceivedtheirsocietyas egalitarianin the
respectgiventhe social classes thandid Americans.However,concerning
educationaland occupationalopportunity-both forracesand forclasses-
Englishand Americanperceptions of inequalityweremorealike and the
differencesbetweenthe countriesare not statisticallysignificant.
In table 3 we combinethe nine itemsinto an Index of PerceivedIn-

TABLE 2
MEAN DIFFERENCES IN PERCEIVED INEQUALITY BETWEEN
THE UNITED STATES AND ENGLAND AND STATIS-
TICAL SIGNIFICANCE BY ITEM, 1975

Mean Country
Differencein
Perceived
Inequality
(U.S. minus
Items England) P<

RACE-INCOME ........... .74 .001

CLASS-LAW ............... .70 .001
RACE-LAW ................ .61 .01
CLASS-INCOME ........... .59 .01
CLASS-RESPECT .......... .37 .01
RACE-EDUCATION ....... .27 N.S.
CLASS-EDUCATION ....... .23 N.S.
CLASS-OCCUPATION ...... .18 N.S.
RACE-OCCUPATION ...... .09 N.S.

328

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities
TABLE 3
PERCENTAGE DISTRIBUTION OF TOTAL SCORES FOR THE INDEX
OF PERCEIVED INEQUALITY, ENGLAND AND
THE UNITED STATES, 1975

United
Total Scores on the Index England States
ofPerceivedInequality* (%) (%)
31 or more(perceivesmostinequality).... .9 15.2
26-30.............................. 5.9 7.9
21-25.............................. 11.0 16.9
16-20.............................. 26.9 21.2
11-15.............................. 31.6 20.3
6-10 .............................. 17.8 15.9
5 or less (perceivesleast inequality) ..... 5.9 2.6
Total . .............. 100.0 100.0

Cases (N) ........................... 101 113

Highest score possible................. 36 36

Highestscore achieved........ ........ 33 36
Lowest scorepossible................. 0 0
Lowest score achieved......... ........ 0 4
* Personswhodid notanswerall oftheitemscomposing
theIPI weregiventotal
scoresbasedon theitemswhichtheydid answer.

equality (IPI) and give its distributionof total scores.Thus the cumu-
lativeeffectsof theitem-by-item differences
betweenthetwocountries can
be seen.The meanscoreon theIPI in England,15.3,is significantly lower
than the mean score of 19.1 in the United States (P < .0001). These
scoresappear in AppendixB, as do variousmeasuresof reliabilityand
validity(Cronbach 1951; Heise and Bohrnstedt1970). The resultsare
acceptable,especiallyfortheUnitedStates.
Becauseourefforts to matchtheEnglishand Americansampleswerenot
entirelysuccessful, betweencountriesby control-
we testedthe differences
ling forgender,race, and occupationindividuallyand simultaneously. In
everycase of subgroupcomparisonthe Americanrespondents, regardless
of gender,race, and occupation,perceivedmoreinequalitythan did the
Englishrespondents. This was trueof theIPI as a wholeas wellas forthe
race itemsalone and the class itemsalone. It is importantto note,how-
ever,that therewas considerablevariationamongsubgroupsin the be-
tween-country differencesin mean IPI scores.Americannonwhitemales
saw moreinequalitythanEnglishnonwhite malesby a meandifference of
10.6,Americannonwhitefemalessaw 8.3 morethanEnglishnonwhitefe-
males,Americanwhitefemalessaw 3.7 morethanEnglishwhitefemales,
and Americanwhitemalessaw only 1.2 morethanEnglishwhitemales.
Furthermore, thecorrelation
of theIPI withcountry, scoringtheUnited
States as 1 and Englandas 0, was .24. It remainsthe same or nearlyso
withgender(.24), race (.23), and occupation(.24) controlledindivid-

329

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

ually; withgenderand race together(.23); and withgender,race, and

occupationcontrolledsimultaneously (.24). Thus, we concludethat the
finding thatAmericansperceivedmoreinequalitythanthe Englishis gen-
uineand notsimplytheresultof differences in gender,race,or occupation
betweenthetwosamples.
Nonetheless,perceptionsof inequalityvaried considerably withineach
country,with some people seeinginequalitywhereotherssaw equality.
This is especiallytruein theUnitedStates,wherethe varianceof theIPI
is 76.3, nearlytwiceas large as the Englishvarianceof 39.6. Moreover,
fromanotheranalysisof thesedata (Robinsonand Bell, forthcoming), we
foundthata path modelincludingvariousmeasuresof objectivesocialpo-
sitionexplainedless than 10% of thevariancein perceptions of inequality
in Englandbut nearly45% in theUnitedStates.Not onlydid Americans
perceivemoreinequalitybut also theirperceptions weremorecloselyre-
lated to objectivesocialposition.Hence,Americanswerefarmoresystem-
aticallypolarizedon thisquestionthanweretheEnglish.
Althoughit is beyondthe scope of this paper, our findingsraise the
questionwhetherperceptions reflectreal differencesin inequalitybetween
England and the United States. People may perceivemore inequality
where,in fact,thereis more,or theremay be littlerelationbetweenper-
ceivedand actualinequality.To takejust one variablethathas beenwide-
ly studied,Taylor and Hudson (1972) forthe mid-1960sand Bornschier
and Ballmer-Cao(1979) fortheearly1970sreportrelatively moreincome
inequalityin theUnitedStatesthanin theUnitedKingdom.Furthermore,
thereis considerableevidencethat economicdifferences separatingthe
classeshave decreasedin recentyearsin Britain(Thompson1975, 1976).
The evidence,however,is difficult to evaluate,especiallyin comparative
perspective, because of the effectsof differingtax structures,government
provisionof variousbenefitsrangingfromhealthto social security,and
the natureof the data available. A numberof writersin Britainhave
questionedthe adequacy of the data showinga trendtowardequality
(Abel-Smithand Townsend1965; Titmuss1962; Townsend1962; Wed-
derburn1962). Westergaardand Resler (1975) even doubt the general
assumptionthatwelfarestatemeasuresundertheLabour Partyin Britain
wereeverreallydirectedtowardlarge-scaleincomeredistribution.
In the UnitedStates,Paglin (1975), controlling forage and a number
of transfer payments(e.g., publichousing,rentsupplements, foodstamps,
foodassistance,Medicaid,and social services),foundless incomeinequal-
ity in the UnitedStates thanhas been foundusingmethodsthat do not
take such thingsinto account.Furthermore, he found,with controls,a
declinein incomeinequalityfrom1947 to 1972.
An assessment of theactualdifferences betweenEnglandand theUnited
Statesin inequalityalong variousdimensionsis beyondthe scope of this

330

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

paper and invitesfurtherresearch,sensitizedto problemsof measurement.

In any event,evenif Englishsocietyin 1975 was moreequal thanAmeri-
can societyon the variableswe have specified,thatfactwouldbe only a
partialexplanationof our findings.It does not tell us whysome English
respondents perceivedmoreinequalitythan some Americanrespondents.
Nor does it tell us why,withineach country, some respondents perceived
moreinequalitythanothers.Perceptionsof inequalitymay not reflectac-
curatelytheamountof actual inequalityin thewholesociety,bothbecause
people,beingdifferentially
placed socially,experience different
partsof so-
ciety,and because groupinterests, values, and societalmythscolorper-
ceptions.

THREE ALTERNATIVE UNDERLYING STRUCTURES

A SingleFactor: Socio-Economic-Racial
Inequality?
We factoranalyzed the nine items composingthe IPI in orderto test
threehypotheses,the firstbeing,"The nine items can be considereda
single dimension."The resultsconcerningthis hypothesisare shownin
table 4, wherewe give the principalfactormatricesforboth countries.
For England,58% of the commonvarianceof the itemsis explainedby
theprincipalfactor,smallerthanone wouldlike but a clearlyevidentfac-
torwithan eigenvalueof 1.99. For theUnitedStates,a hefty88% of the
commonvarianceis explainedby the principalfactorwithan eigenvalue

TABLE 4
PRINCIPAL FACTOR MATRICES OF ITEMS COMPOSING THE INDEX OF PERCEIVED
INEQUALITY) ENGLAND AND THE UNITED STATES, 1975

ENGLAND UNITED STATES

Factors Factors

1 (Prin- Variance 1 (Prin- Variance

cipal Explained cipal Explained
ITEMS Factor) 2 3 (%)* Factor) 2 (%)*
CLASS-EDUCATION . 35 .66 -.04 56 .65 .02 43
CLASS-OCCUPATION .54 .27 -.02 36 .62 -.04 38
CLASS-INCOME . 41 .09 -.02 18 .72 -.14 55
CLASS-RESPECT . 22 .09 .29 14 .50 .41 41
CLASS-LAW . 51 -.10 .52 53 .44 .37 33
RACE-EDUCATION . 69 .07 -.29 56 .80 -.29 73
RACE-OCCUPATION . 42 -.29 -.28 34 .65 -.25 49
RACE-INCOME . 17 -.37 -.04 17 .56 .14 34
RACE-LAW . 66 -.37 .05 57 .65 .11 44
% common
varianceexplained
by each factor.58 27 15 ... 88 12 ...
Eigenvalue.1.99 .91 .52 ... 3.59 .50
* The percentageof the total varianceof each variableaccountedforby the combination
of all factors
shown(thatis, 100X hi2).

331

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

of 3.59. Thus, thereis evidencethat perceivedinequalitiesof class and

race can be accountedforby a singledimensionin bothcountries, particu-
larlyin theUnitedStates.
In England,however,a secondfactormakesmorethana faintappear-
ance, explaining27% of the commonvariancewithan eigenvalueof just
under1.00. The eigenvaluesof a thirdpotentialfactorin England (0.52)
and a secondpotentialfactorin the UnitedStates (0.50) are too smallto
warrantmuchattention.

A Two-FactorOblique Solution:Class and Race?

Studiestoo numerousto mentionhave asked questionsconcerning the ef-
fectsof class relativeto race on some dependentvariableor concerning
the extentto whichclass and race are associated.Here we ask different
questions:Are inequalitiesof class and race perceivedby membersof so-
cietyas a singleset of covariantinequalities?If so, to what degree?Or
are theyviewedas separatedimensions of stratification,
varyingindepen-
dently?Thus, a secondhypothesis we testis, "Perceivedinequalitiescan
be separatedempirically intoperceivedinequalitiesofclass on theonehand
and perceivedinequalitiesofrace on theother."
Holzingerand Harman(1941) suggestthecomputation of B-coefficients
to testfortheexistenceof clusterings of subsetsof itemsfroma totalset.
A B-coefficient is definedas "100 timesthe ratio of the averageof the
intercorrelationsof a subsetor groupof variablesto theiraveragecorre-
lationwithall remaining variables"(p. 24). If a B-coefficient
equals 100,
thevariablesof a subsetdo notclustertogether;thatis, theyare,on aver-
age, no morehighlycorrelatedwitheach otherthanwiththe otheritems
in the totalset. Holzingerand Harmansuggestthat,as a tentativestan-
dard, a B-coefficientshouldbe at least 130 to concludethat variablesof
a subset"belongtogether"apartfromothervariablesin the totalset.
For the Englishdata, the B-coefficients are 225 forthe race itemsand
158 forthe class items,suggesting that the perceivedinequalitiescan be
separatedintotwosets. For theUnitedStates,the resultsare mixed.The
B-coefficientof 135 forthe race items,thoughbarelyexceedingthe mini-
mumcriterionof 130, showsthat thereis a tendencyforthe race items
to clusterseparately.The B-coefficient of 105 fortheclass items,however,
showsthatperceivedinequalitiesof class are empirically inseparablefrom
thoseof race.
Much the same picturecan be seen in table 5, wherewe have graphi-
cally rotatedthe matricesto an oblique solution.We rotatedto test the
hypothesisof subsetclusteringby race and class so as to maximizethe
item-factor correlationsof the race itemswitha race-onlyfactor,if there
is one,and thoseof theclass itemswitha class-onlyfactor,if thereis one.

332

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

TABLE 5
HYPOTHESIZED MATRIX AND GRAPHICALLY ROTATED OBLIQUE PATTERN
MATRICES MAXIMIZING POSSIBLE CLASS AND RACE FACTORS,
ENGLAND AND THE UNITED STATES, 1975

HYPOTRESTS ENGLAND UNITED STATES

ITEMS Race Class Race Class Race Class

Class items:
CLASS-EDUCATION ..... 0 + -.54 .93 .38 .30
CLASS-OCCUPATION .... 0 + .00 .61 .55 .09
CLASS-INCOME ......... 0 + .13 .33 .91 -.20
CLASS-RESPECT ........ 0 + .03 .22 -.99 1.49
CLASS-LAW ............. 0 + .40 .19 -.91 1.37
Race items:
RACE-EDUCATION..... + 0 .31 .46 1.41 -.65
RACE-OCCUPATION.... + 0 .55 -.09 1.20 -.59
RACE-INCOME .?...... . + 0 .50 -.31 -.09 .60
RACE-LAW .............. + 0 .78 -.04 .09 .60
Correlationbetweenthefactors ... .60 .95

NOTE.-Factor loadings hypothesized to be small are shown by "O," and those hypothesized to be large are
shown by "+."

in thetable are patternfactorloadings,whichare used to

The coefficients
determine the clustersof variablesdefinedby oblique factors.They mea-
surethedependenceof theitemson thetwofactorsand in thissensemay
be interpreted as regressioncoefficients
of the variableson the factors.As
is true of regressioncoefficients, theirabsolute value may exceed 1.00
(Rummel1970,pp. 398, 405).
For England,fourof the fiveclass items,as hypothesized, have higher
patternloadingswitha class factorthanwitha race factor,the exception
being CLASS-LAW. And threeof the fourrace items,as hypothesized,
have higherloadingswitha race factorthanwitha class factor.The ex-
ceptionthis time is RACE-EDUCATION. Thus, as in the case of the
B-coefficients,thereis some supportforthe hypothesisthatperceivedin-
equalitiesof class are separable fromperceivedinequalitiesof race in
England.
Yet note that this"best possible"solutionforrace and class factorsis
farfromperfect.Only twoitemloadingson each factorare effectively re-
ducedto zeroon the"wrong"factor,CLASS-OCCUPATION and CLASS-
RESPECT for the race factorand RACE-OCCUPATION and RACE-
LAW forthe class factor.Moreover,the correlation betweenthe factors
themselves is .60, so that,even if each factorhad been moreclearlyde-
lineated,thefactorsthemselves are quite highlycorrelated.
For the UnitedStates,the resultsof the oblique rotationpresentedin
table 5 showlittleevidenceof a race factorseparatefromclass. In fact,
threeof the class itemshave higherpatternloadingson the race factor
thanon theclass factor.Onlytwoof thefourrace itemshave higherload-
333

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

ings on the race than the class factor.Furthermore, the correlationbe-

tweenthe factorsof this"best possible"fitof thedata to a race and class
two-factorsolutionis .95. That is, evenif we couldconcludefromthefac-
tor analysisthatperceivedinequalitiesof race and class wereempirically
separatedimensions, as we cannotdo forthe United States,the factors
themselves would be so highlycorrelatedthat one mightstill argue that
empirically theycould be consideredbasicallythesame thing.
Thus, forEnglandthereis evidencebothfora singlegeneralfactorand
for somewhatdistinct,thoughcorrelated,factorsdealingwith race and
class. For the UnitedStates the evidenceoverwhelmingly favorsa single
factor.An investigator would,therefore, be justifiedin usingthe IPI in
eithercountry, combining all nineitemsintoa singlescore.8
For the UnitedStates,withlittleempiricalsupportforsubdividing the
itemsinto two groups,the decisionto separaterace and class mustrest
on reasoningother than the evidencewe have given. In any event-
whetherin Englandor the UnitedStates-the decisionshoulddependon
the purposeof the investigator and the natureof the theoryhe or she is
using.
Sociologistsmightagreethatsocial class and race are different phenom-
ena theoretically,no matterhow intercorrelated theymay be underpar-
ticularsociocultural, economic,or politicalconditions.9 They may be in-
separable empiricallyunder extremeconditionsof racial conquestand
slavery,or foreigndominationand colonialrule. Underotherconditions,
at othertimesor in otherplaces, theymay be muchless correlated.Yet,
evenwhereperceptions of racial and class inequalities-andtheactual in-
equalitiesin social realityon whichtheyare based-are highlycorrelated,
a researcher may wishto keep themanalyticallydistinctand make their
interrelationshiptheoretically problematic, possiblyfocusingresearchon
thedegreeof therelationship itself.

Is Therean AdditionalUnderlying
Factor Structure?
The resultsof a factoranalysis,of course,usuallypermita varietyof rep-
A numberof alternativeclusterings
resentations. of variablesintofactors
may equallywell summarizethe underlying intercorrelations
betweenthe
8 Each of the nine items, of course, can be weighted by its factor score coefficient
when computinga total score.
9 Despite the empiricallyoverlappingperceptionsof class and race inequalities that
we found, respondentsin both countrieshave a good grasp of the analytic distinct-
ness of the two concepts.At least in responseto an open-endedquestion in which we
asked the respondentsto name the social classes they had counted,very few of them
used racial descriptions.In fact,none did in England, although2% referredto "immi-
grants,"by which they may well have meant "coloured people." Among the U.S. re-
spondents,4%o referredto racial characteristicswhen asked to describe social classes
in the United States, with an additional 2% mentioningnationality.In theirstudies

334

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

items.This is one reasontheresultsof factoranalysesare sometimes con-

fusingwhep theyare performed withoutthe guidanceof theory:a pro-
fusionof different findingsmay result.Yet we cannotleave the factor
analysiswithoutconsideringthe null hypothesis:"Except for a single-
factorsolutionof all itemsand the possibilityof a two-factor (class and
race) solution,thereis no evidenceof an alternative
underlying structure."
In table 6 we give oblique rotationsof the originalmatricesusingthe
directobliminmethodwithdelta set at zero (Nie et al. 1975, p. 486).
Note thatforbothcountriesthefiveitemsdealingwithperceivedinequal-
ity of socioeconomicopportunity clustertogetherwith high loadingson
factor1. This factoris reasonablyneat,althoughRACE-LAW and RACE-
INCOME in the UnitedStatesalso have noticeable,althoughlower,load-
ingson it. But the firstfiveitemsare morehighlyrelatedto this factor
thanare any of theothervariablesand can be safelyconsidered perceived
inequalityof opportunity. This factorcuts across race-classdivisions:it
containsthreeclass and two race items.Thus, thereis evidenceof still
anotherunderlying structure-anopportunity factor.
If thepurposeof theresearcher led himto wantto investigate perceived
inequalityof opportunity, the fiveitemsdealingwithopportunity could
be used to composean Index of PerceivedInequality of Opportunity

TABLE 6
ROTATED PATTERN MATRICES WITH DELTA EQUAL TO ZERO FOR
DIRECT OBLIMIN LOADINGS WITH KAISER NORMALIZATION,
ENGLAND AND THE UNITED STATES, 1975

ENGLAND UNITED STATES

ITEMS 1 2 3 1 2

Opportunityitems:
CLASS-EDUCATION ........... .67 .43 .04 .50 .22
RACE-EDUCATION ........... .67 -.30 -.07 .91 -. 10
CLASS-OCCUPATION .......... .54 .02 .14 .53 .14
RACE-OCCUPATION . 26 -.51 -.12 .76 -.11
CLASS-INCOME ............... .34 -.09 .12 .71 .05
Conditionitems:
RACE-INCOME . .............. -.09 -.41 .04 .31 .34
CLASS-RESPECT .............. .04 .07 .36 -.00 .64
Treatmentitems:
CLASS-LAW ........... ........ -.01 -.16 .70 -.00 .58
RACE-LAW .................... .19 -.59 .29 .41 .33
Correlationsbetweenfactors:
1 .............................. ... -.14 .34 ... .60
2 .............................. -.12

in Boston and Kansas City, Coleman and Rainwater (1978, p. 92) found that some
respondents"identifiedcertain ethnic groups as classes unto themselves;most, how-
ever,did not."

335

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

(IPIO).10 Further,if one wereinterested

in measuring perceivedinequal-
ityof treatmentand conditionseparately,additionalitemsshouldbe writ-
tendealingappropriately withtheseaspectsof inequality.

EXPANDING THE COGNITIVE MAPS OF INEQUALITY

of Changesin Class Inequality

Perceptions
Our finding thattheEnglishperceivedless inequalitythanAmericansruns
directlycounterto somepriorcommentary and theory.Part of thediscrep-
ancymaybe explainedaway,however,if recentdifferential social changes
in England and the United States have resultedin England'sbecoming
moreegalitarianthanthe UnitedStates.Such may be the case, if the re-
spondents'perceptionsare accurate.
We asked severalopen-endedquestionsabout the respondents' percep-
tionsof changesin the class structure.
One dealt withchanges in the eco-
nomicgap betweenthe social classes: "Now I'd like to ask you about
trendsor changesoverthe past generationin the thingswe've been talk-
ing about. . What about economicdifferences betweensocial classes?
Have thesedifferencesincreased,decreasedor stayedabout thesamedur-
ing thepast generation?"The resultsare givenin table 7.
A muchlargerpercentageof theEnglishthanof theAmericans(56.5%
vs. 25.5%) perceivedtheeconomicdifferences separatingthesocialclasses
as havingdecreasedduringthepast generation. Conversely,relativelyfew

TABLE 7
PERCEPTIONS OF CHANGES IN ECONOMIC DIFFERENCES
AMONG THE CLASSES OVER THE PAST GENERATION,
ENGLAND AND THE UNITED STATES, 1975

United
Economic Differences among England States
the Classes Have: (%) (%)

Increased a lot....................... 5.0 23.6

Increased some....................... 6.9 11.8
Stayed about the same................ 25.7 33.6
Undecided or don't know.............. 5.9 5.5
Decreased some...................... 20.8 16.4
Decreased a lot ......... ............. 35.7 9.1

Total ............................. 100.0 100.0

Cases (N)........................... 101 110

10 See AppendixB for the reliabilityand validityof IPIO.

336

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

of the English,only 11.9%, believedthatthe economicgap had widened,

comparedwithovera third,35.4%, of theAmericans.
In England,some respondents viewedthe narrowing gap betweenthe
workingand middleclassesas theriseof theworkingclass intothemiddle
class,theimprovement of theconditionof blue-collarworkersto themid-
dle-classlevel.The tradeunionswereseenby manyrespondents as having
playeda powerfulrolein causingthischange.Othersviewedit as thefall
of thewhite-collarmiddleclass,notan absolutefall,perhaps,but a failure
of middle-classconditionsto improveover recentyearsor even a decline,
makingproblematicthe continuation of the stylesof life enjoyedin the
past.
In theUnitedStates,thepictureof theclass systemand howit has been
changingwas morediversethan in England.IndividualAmericanswere
as certainas the Englishabout whathad been happening,but as a group
theydisagreedmoreoverexactlywhatit was. In general,manyAmericans
saw therelativelossesof themiddleclass and relativegainsof theworking
class,as did the English.Americans,too,wereaware of programsof gov-
ernmentalassistanceand the "welfare"and dependentclasses that have
been createdand in some cases liftedfromabject poverty.And theyun-
derstoodthat thereis moreequal distribution thaneverbeforeof certain
materialpossessionsthroughout theclass structure.
Yet, Americans, unlike
the English,were sometimespainfullyaware of the existenceof great
wealthand its influence in the society.Many Americansperceiveda vast
gulfbetweenthe veryrichand the "average"American,a gulf,in their
view,thatcontinuesto widenand to mocktheirrelativelymodestmone-
taryachievements.

The Numberof Social Classes Perceived

Early in the interviewwe asked: "Some people say that it's possibleto
classifyAmericans[people in Britain] into social classes. In your own
judgmenthow manysocial classes wouldyou say thereare in thiscoun-
try?"We givetheresponsesin table8. The Englishperceivedfewerclasses
on average than Americans(3.1 vs. 3.9). Moreover,Americananswers
variedmorethanthe English,from"no social classes" to "nine,""many,"
"plenty,"or "quite a few."Englishanswersvariedonly from"none" to
"six." Finally,the Englishwere morethan twiceas likelyas Americans
(23% vs. 10%) to reporta dichotomous modelof the class structure.1'

11 In a study of workers in Barking, England, Moorhouse (1976, p. 480, table 4)

found that 29%ohad a dichotomousimage of the class structure.However, he draws
the erroneous conclusion that "the majority view of the class system in Britain is

337

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

TABLE 8
PERCEPTIONS OF THE NUMBER OF SOCIAL CLASSES,
ENGLAND AND THE UNITED STATES, 1975

United
England States
N Social Classes Perceived (%) (%)

Many, plenty,quite a few...... ......... 0 7

9................................. 0 1
8. .............................. 0 1
7. .............................. 0 5
6. .............................. 2 5
5. .............................. 6 9
4. .............................. 12 13
3. .............................. 52 45
2. .............................. 23 10
1. .............................. 1 2
None, thereare no classes...... ......... 4 2
Total ............................... 100 100
Cases (N) ............................. 94 104

Our findings supportColemanand Rainwater's(1978, pp. 24-25) con-

tentionthatcognitivemaps of inequalitylend themselvesto bothcontin-
uous and categoricalrepresentations. For example,in bothcountriespar-
ticularrespondents mightnamea givennumberof social classesin answer
to our question,but in elaboratingtheysometimesreferred to additional
classesand made finer,and stillfiner,distinctions,and manyrespondents
-especially in the UnitedStates-specifiedsome continuousvariableon
whichclassesdiffered. Yet Colemanand Rainwatermayoverstatethecase
fortheperceptionof class as a continuum. Althoughin Englandand espe-
cially in the UnitedStates therewas a tendencyforrespondents to dis-
play an underlying continuum in theirimageryof theclass structure,
there
also was considerable consensuson thesignificant breaksin thecontinuum
that separatedifferent classes,even thoughrespondents sometimesused
different labels forthem.
Giventhe importance of class imagery-especiallythe polarizationinto
twoclasses-in Marxisttheory,we have organizedourdata to testa lead-
ing hypothesis.Ossowski(1963, pp. 19-57), for example,theorizesthat
a dichotomousview of class is consistentwitha perceptionof greatin-

very much a dichotomizedone" (1976, p. 493), even thoughhis own data show that
52%o of the workers perceived a three-classmodel. Our data also contradictMoor-
house's statement,but not his data. Althoughabout one-quarterof the English per-
ceived a dichotomizedclass structure,and although far more of them perceived this
than did Americans,the dominant perception,held by 52%o of our English respon-
dents,was of a trichotomousclass system.

338

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

equalityand lack of mobility, whiletheperception of a greaternumberof

classes coincideswithan imageof the class structureas beingrelatively
open and boundaryfree.A dichotomous view,of course,is consistent with
mostconflictmodels,such as Marx's based on relationship to the means
of productionand Dahrendorf's based on authorityrelations.12 The more
continuousviews,however,are consistentwith a functionalist model of
social class, emphasizing relativelyopen recruitment intooccupationsand
universalism and achievement as distributivecriteria(Parsons 1954; Blau
and Duncan 1967).
In table 9 we showtheamountof inequalityperceived(mean scoreson
the IPI) by the numberof classesrespondents saw. In England,contrary
to theOssowskithesis,personsholdinga dichotomous class modelare less
likelyto perceiveinequalitythanare personswhosee threeor moreclasses.
In fact,the relationship betweenthenumberof classesand perceptions of
inequalityis linearand positivein Englandand the correlation(r .15)
betweenthetwovariablesis significant (P < .05), supporting thealterna-
tivehypothesis:thegreaternumberof classesperceived,themoreinequal-
ityis perceived.
In theUnitedStates,respondents holdinga dichotomous class imagedid
perceivethe greatestinequality(mean IPI score 23.9). They do not
differsignificantly, however,fromrespondentsperceivingfour or more
classes (mean IPI score 22.2). Moreover,the linearcorrelation
- in the
U.S. data is also positive.It is .22, whichis significantat greaterthanthe
P < .01 level. However,the nonlinearcorrelationcoefficient, eta, is .40,
revealinga significant departurefromlinearity. Thus, for the United States

TABLE 9
MEAN SCORES ON THE INDEX OF PERCEIVED INEQUALITY
BY THE NUMBER OF CLASSES PERCEIVED, ENGLAND
AND THE UNITED STATES, 1975

MEAN SCORE ON THE INDEX

OF PERCEIVED INEQUALITY

N CLASSES PERCEIVED England United States

Four or more.................... 17.8 (19) 22.2 (42)

Three.......................... 15.0 (49) 16.8 (47)
Two .......................... 14.8 (21) 23.9 (11)
Oneornone..................... 13.6(5) 9.8 (4)
NOTE.-N cases on which means are based is given in parentheses.

12Westergaard(1975, p. 253), however,challengesthe assumptionthat a power and

conflict"image of society must necessarilybe associated with a dichotomousview of
society."

339

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

the Ossowskithesiscannotbe acceptedand the alternativeis not clearly

borneout either.
We can also test a second Marxisthypothesis.Ossowskisays further
that a dichotomous modelof the class structure,"us" versus"them,"is
held by disadvantagedgroupsand enablesthemto rationalizetheirrela-
tivelylow positionsbecause of the clear obstacleto theiradvancement.
A more graduated,pluralisticmodel of classes is held by advantaged
groups,accordingto Ossowski,and allows themto rationalizetheirprivi-
leges by the claims that no clear boundariesor obstaclesseparatefrom
themselves peoplewhoare less welloffand thattheclass structure
is based
upon equalityof opportunity.
If Ossowskiis correct,we would expectcertainmeasuresof objective
privilege-education,occupationalprestige,ownershipof themeansofpro-
duction,and familyincome-to be positivelycorrelatedwiththe number
of classesperceived.13
Whenthe numberperceivedwas regressedon these
fourvariables,we obtainedtheresultsshownin table 10.14

TABLE 10
UNSTANDARDIZED AND STANDARDIZED REGRESSION COEFFI-
CIENTS SHOWING THE EFFECTS OF EDUCATION, PRESTIGE,
OWNERSHIP, AND FAMILY INCOME ON PERCEIVED NUMBER
OF CLASSES, ENGLAND AND THE UNITED STATES, 1975

ENGLAND UNITED STATES

VARIABLE B B

Education ............... .09 .12 .28* .30*

Prestige ................. -.00 -.00 .01 .08
Ownership .......... ..... .85* .25* .17 .04
Family income ........... .00 . 05 -. 00 -.17
Intercept ................ 2.65 ... 2.63 ...
2 ...................... 11.8% 12.8%

* Significant
at the.05 level.

13 See Vanneman and Pampel (1977) for an analysis of the Ossowski thesis that is
somewhatdifferentfromthat undertakenhere.
14 Measurementof the independentvariables is as follows: Education is coded in com-
parable categoriesin England and the United States with the advice of Dr. Michael
Mann of the London School of Economics and Political Science. Categoriesare (with
the United States in parentheses): graduate degrees (graduate or professionaltrain-
ing); college degree (college degree); G.C.E. Advanced Level, Higher Level Diplomas
from furthereducational institutions,C.S.E. H Level, accountingqualifications(par-
tial college-at least one year-or specialized training); G.C.E. OrdinaryLevel, Ordi-
nary Diplomas, City and Guilds qualifications,C.S.E. 0 Level (high school graduate);
no qualifications,leftschool age 15 or over (10-11 years of school); no qualifications,
left school 12-14 (7-9 years of school); and no qualifications,left school below age
12 (less than 7 years of school). Occupational prestigeis coded in Treiman's (1977)

340

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

Objectivepositionexplainsonlya modestpercentageof the variancein

thenumberof classesperceivedin England( 11.8% ) and theUnitedStates
(12.8%). In England,onlyownershipof the meansof productionis sig-
nificantlyrelatedto thenumberof classesperceived.In the UnitedStates
onlyeducationis a significantdeterminant, whilefamilyincomeis nearly
in theoppositedirectionfromthatpredictedby Ossowski.Thus,
significant
theOssowskithesisreceivesonlymodestsupportfromthistest.
Lockwood (1966) offersa descriptionof "workingclass imagesof so-
ciety"whichmayhelpexplainwhythecorrelation betweendisadvantaged
positionand thetendency to hold a dichotomized imageof theclass struc-
tureis not higher. He distinguishes three ideal typesof worker,each of
whichmaintainsa different of
image society. "Proletarian"workershave
the dichotomous"us" versus"them" class model based on power that
Ossowskiwritesabout. "Deferential"workers,however, have a hierarchi-
cal or continuousmodel based on prestigeor status, and "privatised"
workershave a continuousmodelbased on moneyor wealth.The last two
typesof workershave class imagessimilarto thatwhichOssowskiattrib-
utes to advantagedgroups.

The Criteriaof Social Classes

Whenwe asked respondents to name the classes theyperceived,theyde-
scribedthemmainlyin termsof money,that is, incomeor wealth.The
predominance of moneyas a criteriondefining class was notedby Gold-
thorpeet al. (1969, p. 147) and Moorhouse(1976) in Britishstudiesand
by Colemanand Rainwater(1978) in theirstudiesof Bostonand Kansas
City. However,Americanswere nearlytwiceas likelyas the Englishto
specifyincomeor wealthas bases of class membership. In America,79%
of respondents gave economiccriteriaas bases for class differentiation,
whilein Englandonly46% did (P < .0001).
Anotherdifference betweenthetwocountries, fullyexpectedgiventheir
differenthistoricalbases,was in theresponse,"aristocracy"or "the aristo-
craticclasses." Only 2% of Americansused thesetermsor anythinglike
them,comparedwith 16% of the English.Thus, the Englishwere less

prestigeunits, which are specificallydesigned for cross-nationalpurposes and have

been previouslyused in a comparison of England and the United States (Treiman
and Terrell 1975). Housewives not employedoutside the home are assigned theirhus-
band's occupation and young people listingtheirparentsas chiefbreadwinnersin the
household are given their father'soccupation. Ownershipof the means of production
is coded so that persons derivingincome from their own business are classifiedas
owners and all othersare classifiedas nonowners (Robnison 1979, chap. 3). Family
income is the respondent'sbest estimateof the income fromall sources earned or re-
ceived by familymembersbeforetaxes.

341

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

likelythan Americansto differentiate classes on the basis of inherently

continuousvariables,such as incomeor wealth,and somewhatmorelikely
thanAmericansto name at least one class thatcorresponds to moreclas-
of the class structure
sical interpretations in termsof discreteand clearly
boundedgroupsand to hold a dichotomous This
imageof class structure.
withRobinsonand Kelley's (1979) finding
is consistent thatoccupational
status-a continuousdimension-is moreassociatedwithAmericanthan
withBritishpoliticalattitudesand behavior,whilecontrolof themeansof
productionand authority-dichotomous class variables-are moresalient
to Britishthanto Americanpoliticalorientations and actions.

SUMMARY AND CONCLUSIONS

In a study of 113 randomlyselectedrespondentsin the GreaterNew

Haven area and a demographically similarrandomsample of 101 re-
spondentsin GreaterLondon,England,we exploredthe cognitivemaps
of class and racial inequalitiesand testedseveralhypothesesconcerning
theimageryof social class.
First,we proposedand testedan instrument to measureperceptionsof
inequality,an Index of PerceivedInequality(IPI). It containsnineitems,
fivedealing with inequalitiesof social class and fourdealingwith in-
equalitiesof race. The itemsspecifyinequalitiesamongclasses and be-
tweenracesin education,occupation,income,and treatment by thepolice
and courts.An additionalitem dealing with inequalityof respectwas
specifiedforsocial class.
We foundthat Americansperceivedsignificantly moreinequalitythan
the English,the largestdifferences being in inequalitiesof incomeand
treatment by thepoliceand courtsconcerning bothracesand classes,and
in inequalitiesof respectconcerning classes.Differences betweenthecoun-
triesin perceivedinequalityof educationaland occupationalopportunity
weresmallerand notstatistically significant.The meandifference between
countriesin the IPI, thoughinfluenced in size by gender,race,and occu-
pation,held up whencontrolsforthemwereintroduced.
Despite the differences betweencountriesin perceivedinequality,the
nineitemswererankedquite similarly by bothAmericansand theEnglish.
In both countriesinequalitiesof social class in respect,administration of
justice,and occupation were perceived as larger than any of
inequalities
race. The opportunity to earn a good income,regardlessof social class,
was perceivedby both sets of respondentsas the source of the least
inequality.15
15We call the reader's attentionto another comparativestudy of perceptionsof class
in the United States and Great Britain (Vanneman 1980), which was publishedafter
we went to press.

342

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

Fromfurther examinationof the IPI we concludedthat the indexwas

both reliableand valid in both countries,althoughthe findings wereless
strongin England.
Second,we performed threedifferent factoranalyses.The firsttested
the hypothesisthat a single dimensioncould account for the common
variationamongthe nine items.For the UnitedStates,the resultswere
supportive:88% of the commonvariancewas explainedby a principal
factor.For England,however,the resultswere again less satisfactory:
only58% of thecommonvariancecould be explainedby a singlefactor.
Yet, evenin England,no otherfactorachievedan eigenvaluegreaterthan
1.0. Thus, the IPI can be used as a single compositeindex in both
countries.
The secondfactoranalysistestedthehypothesis thattherace and class
itemsare separable,that all the race itemscould be loaded only on a
race factorand all the class itemsonly on a class factor.We performed
graphicallyan oblique rotationpassingthe axis of one factorthroughthe
averageof the race itemsand the axis of a second factorthroughthe
averageof theclass items.For theUnitedStates,the resultsemphatically
contradictthe hypothesis:Americanperceptionsof race and class in-
equalitieswere not empiricallydistinct.That is, respondentswho per-
ceivedinequalitiesof race saw inequalitiesof class and vice versa.
The hypothesis, however,was tentativelyaccepted for the English-
"tentatively"because some of the items have sizable loadings on the
"wrong"factorand because the eigenvalueof the second factoris less
than 1.0. Yet perceptions of inequalitiesof race wereto some extentem-
piricallyseparable fromthose of social class in England. Even there,
though,thecorrelation betweentherace and class factorswas quite large,
and no greatviolencewould be done to the factsif one lumpedall the
itemsinto the IPI and dealt withthemas a singlescore.
We performed a thirdfactoranalysisin orderto uncoverany additional
underlying structure in the patternsof interrelationshipsamongthe nine
items.Such a structuredid emerge.Five of the itemsdealingwithper-
ceptionsof inequalityof opportunity tend to clusterin both countries,
particularlyin England. Thus, dependingon the purpose of one's re-
search,thereis anotherway of constructing an indexwith these items,
an Index of PerceivedInequalityof Opportunity(IPIO). Its reliability
and validityare acceptable.
Third,we enlargedthe cognitivemaps of inequalityin severalways.
We reportedthatEnglishrespondents wereless likelythanAmericanones
to perceivea growingeconomicgap betweensocial classes.And theyper-
ceivedfewerclasses totally,beingtwiceas likelyas Americansto hold a
dichotomous imageof class structure. In England,havinga dichotomous
viewof class structuredid notcoincidewithperceiving greaterinequality;

343

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

in the UnitedStates the resultsweremixed.Thus, Ossowski'shypothesis

concerning the relationship betweena dichotomousimageof societyand
inequalitywas not supported.
A second hypothesisput forwardby Ossowski,that disadvantaged
groupsare morelikelythan privilegedones to have a dichotomous view
of class, receivedmodestsupport.Ownershipin England and education
in the United States were positivelyrelated to the numberof classes
perceived.
In both countriesmoney-incomeor wealth-was the mostfrequently
specifiedcriterion of social class, althoughAmericansweretwiceas likely
to mentionmoneyas the English.Practicallyno Americansmentioned
aristocracy as a criterion of class,but 16% of theEnglishdid. Americans
were,thus,morelikelythanthe Englishto mentiona continuousvariable
-money-as the primary basis of stratification, whilethe Englishwere
more likelyto name at least one class corresponding to more classical
interpretations of class in termsof clearlyboundedgroupsand weremore
likelyto hold a dichotomous imageof the class structure.
In conclusion,the cognitivemaps of inequality,measurements of which
we have proposedhere,combinedwithknowledgeof people's evaluative
mapsof inequality,shouldbe powerfultoolsforgreaterunderstanding of
discontent, social protest,and demandsforsocial change-or, of course,
theopposite:contentment, acceptance,and hopesforpreservingwhatnow
exists.

344

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 O ~ ~~~~~~~ "C
"Cs oOnm *- -o ote

~q
??R???
$
*'
e+
m e ON

oP s ?? ed,+?? sd-,
ce,WiW $.0 .:

p
< n ,C 0 $ 4 *q * -41 "14

?
> H , * * * * * * ~
4~~~~~~o
?Q> * r* e or*
?n nN:NN s en t

S WN
X-? 'H :en*Yemene O? oo C~~~~~~~~~~~~~.
dS Ne - =%~~~~~~~~~~~~~~c
O
Z;:::::;
~ ~~~~~~~~
cq .0 0
:X

H ;.4 e10
W -14 s |1W X Z; t '

t ???X?X?is=Pi>2 z%*WX~~~~~~~~~~~~~C-

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

APPENDIX B
RELIABILITY AND VALIDITY COEFFICIENTS, MEANS, AND STANDARD DEVIATIONS
FOR THE INDEXES OF PERCEIVED INEQUALITY (IPI) AND PERCEIVED INEQUAL-
ITY OF OPPORTUNITY (IPIO), ENGLAND AND THE UNITED STATES, 1975

ENGLAND UNITED STATES

STATISTIC IPI IPIO IPI IPIO

Cronbach'sa ......................... .65 .69 .85 .83

HeiseandBohrnstedt's . ............. .75 .69 .87 .84
Averageitem-totalscorecorrelation* ... .33 .37 .57 .64
Heise and Bohrnstedt'sp ...... ........ .86 .83 .93 .91
Heise andBohrnstedt's
V&2............. .01 .00 .00 .00
Mean score.......................... 15.3 7.5 19.1 8.8
Standard deviation................... 6.3 4.5 8.7 6.0

*The item being correlatedeach timewas removedfromthe total score.

REFERENCES

Abel-Smith,Brian, and Peter Townsend. 1965. The Poor and the Poorest: A New
Analysis of the Ministryof Labour's Family ExpenditureSurveys of 1953-54 and
1960. London: Bell.
Alford,Robert R. 1963. Party and Society: The Anglo-AmericanDemocracies. Chi-
cago: Rand McNally.
Alves, Wayne M., and Peter H. Rossi. 1978. "Who Should Get What? Fairness Judg-
mentsof the Distributionof Earnings." AmericanJournal of Sociology 84 (Novem-
ber): 541-64.
Bell, Wendell. 1962. "Equality and Attitudesof Elites in Jamaica." Social and Eco-
nomic Studies 11 (December): 409-32.
. 1977. "Inequality in IndependentJamaica: A PreliminaryAppraisal of Elite
Performance."Revista Interamericana7 (Summer): 294-308.
Bell, Wendell,and Robert V. Robinson. 1976. "Description of the Basic Data: Lon-
don-New Haven Survey 1975." Mimeographed.New Haven, Conn.: Yale Univer-
sity.
. 1978. "An Index of Evaluated Equality: Measuring Conceptions of Social
Justicein England and the United States." Pp. 235-70 in Comparative Studies in
Sociology,edited by Richard F. Tomasson. Vol. 1. Greenwich,Conn.: JAI.
Blau, Peter M., and Otis Dudley Duncan. 1967. The AmericanOccupationalStructure.
New York: Wiley.
Bornschier,Volker, and Thanh-Huyen Ballmer-Cao. 1979. "Income Inequality: A
Cross-nationalStudy of the Relationshipbetween MNC-Penetration,Dimensionsof
the Power Structureand Income Distribution."American Sociological Review 44
(June): 487-506.
Bott, Elizabeth. 1957. Family and Social Network.London: Tavistock.
Broom, Leonard, F. Lancaster Jones, Patrick McDonnell, and Paul Duncan-Jones.
1978. "Is It True What They Say about Daddy?" AmericanJournal of Sociology
84 (September): 417-26.
Bulmer, Martin, ed. 1975. Working-ClassImages of Society. London: Routledge &
Kegan Paul.
Centers,Richard. 1949. The Psychology of Social Classes: A Study of Class Con-
sciousness.New York: Russell & Russell.
Coleman, Richard P., and Lee Rainwater. 1978. Social Standing in America: New
Dimensions of Class. New York: Basic.

346

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

Cronbach, Lee J. 1951. "CoefficientAlpha and the Internal Structure of Tests."

Psychometrika16 (September): 297-334.
Curtis,Richard F., and Elton F. Jackson. 1977. Inequality in AmericanCommunities.
New York: AcademicPress.
Form, William H., and Joan Rytina. 1969. "Ideological Beliefs on the Distributionof
Power in the United States." AmericanSociological Review 34 (February): 19-31.
Giddens,Anthony. 1973. The Class Structureof the Advanced Societies. New York:
Barnes & Noble.
Goldthorpe,John H., David Lockwood, Frank Bechhofer,and JenniferPlatt. 1969.
The AffluentWorker in the Class Structure. Cambridge: Cambridge University
Press.
Gurr,T. R. 1970. Why Men Rebel. Princeton,N.J.: PrincetonUniversityPress.
Heise, David R., and George W. Bohrnstedt.1970. "Validity,Invalidity,and Reliabil-
ity." Pp. 104-29 in Sociological Methodology, edited by Edgar F. Borgatta and
George W. Bohrnstedt.San Francisco: Jossey-Bass.
Holzinger, Karl J., and Harry H. Harman. 1941. Factor Analysis: A Synthesis of
Factorial Methods. Chicago: Universityof Chicago Press.
Huber, Joan, and William H. Form. 1973. Income and Ideology: An Analysisof the
AmericanPolitical Formula. New York: Free Press.
Jasso, Guillermina,and Peter H. Rossi. 1977. "Distributive Justiceand Earned In-
come." AmericanSociological Review 42 (August): 639-51.
Kluegel, James R. 1978. "The Causes and Cost of Racial Exclusion fromJob Author-
ity." AmericanSociological Review 43 (June): 285-301.
Laumann, Edward O., and Richard Senter. 1976. "Subjective Social Distance, Occu-
pational Stratification,and Forms of Status and Class Consciousness: A Cross-na-
tional Replicationand Extension."AmericanJournal of Sociology 81 (May): 1304-
38.
Lipset, Seymour Martin. 1963. "The Value Patterns of Democracy: A Case Study in
ComparativeAnalysis."AmericanSociological Review 28 (August): 515-31.
Lockwood, D. 1966. "Sources of Variation in Working Class Images of Society."
Sociological Review 14 (November): 249-67.
McClosky, Herbert. 1964. "Consensus and Ideology in AmericanPolitics." American
Political Science Review 58 (June): 361-83.
McKenzie, Robert, and Allan Silver. 1968. Angels in Marble: Working Class Con-
servativesin Urban England. Chicago: Universityof Chicago Press.
Mann, Michael. 1970. "The Social Cohesion of Liberal Democracies." AmericanSocio-
logical Review 35 (June): 423-39.
Martin, F. M. 1954. "Some Subjective Aspects of Social Stratification."Pp. 51-75 in
Social Mobility in Britain,edited by D. V. Glass. Glencoe,Ill.: Free Press.
Marx, Karl. [1867] 1962. "Preface to the First German Edition of Capital." Pp. 448-
52 in Selected Works,by Karl Marx and FriedrichEngels. Vol. 1. Moscow: Foreign
Languages PublishingHouse.
. [1852] 1972. "The EighteenthBrumaireof Louis Bonaparte." Pp. 436-525 in
The Marx-Engels Reader, edited by R. Tucher. New York: Norton.
Mau, James A., Richard J. Hill, and Wendell Bell. 1961. "Scale Analysis of Status
Perception and Status Attitudein Jamaica and the United States." Pacific Socio-
logical Review 4 (Spring): 33-40.
Moore, Barrington,Jr. 1978. Injustice: The Social Bases of Obedience and Revolt.
White Plains, N.Y.: Sharpe.
Moorhouse, H. F. 1976. "Attitudesto Class and Class Relationshipsin Britain." Soci-
ology 10 (September): 469-96.
Nagel, J. 1974. "Inequality and Discontent: A Nonlinear Hypothesis." World Politics
26 (July): 453-72.
Nie, Norman H., C. Hadlai Hull, Jean G. Jenkins,Karen Steinbrenner,and Dale H.
Bent. 1975. StatisticalPackage for the Social Sciences. 2d ed. New York: McGraw-
Hill.

347

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
AmericanJournalof Sociology

Nordlinger,Eric A. 1967. The Working-ClassTories. Berkeley: Universityof Cali-

forniaPress.
Ossowski, Stanislaw. 1963. Class Structurein the Social Consciousness.New York:
Free Press.
Paglin, Morton. 1975. "The Measurementand Trend of Inequality: A Basic Revision."
AmericanEconomic Review 65 (September): 598-609.
Parsons,Talcott. 1954. "A Revised AnalyticApproach to the Theory of Social Stratifi-
cation." Pp. 386-439 in Essays in Sociological Theory.New York: Free Press.
Popitz, Heinrich,Hans Paul Bahrdt, Ernst August Jures,and Hanno Kesting. 1957.
Das Gesellschaftsbild des Arbeiters.Tubingen: Mohr.
Robinson,Robert V. 1979. "Ownership,Authority,and Occupational Prestige: A Syn-
thesisand Cross-nationalStudy." Ph.D. dissertation,Yale University.
Robinson, Robert V., and Wendell Bell. 1978. "Equality, Success, and Social Justice
in England and the United States." AmericanSociological Review 43 (April): 125-
43.
.Forthcoming. "Cognitive and Evaluative Maps of Inequality in England and
the United States: An ExplanatoryModel."
Robinson, Robert V., and Jonathan Kelley. 1979. "Class as Conceived by Marx and
Dahrendorf: Effectson Income Inequality and Politics in the United States and
Great Britain." AmericanSociological Review 44 (February): 38-58.
Rose, Richard,and Derek Urwin. 1969. "Social Cohesion,Political Parties and Strains
in Regimes." ComparativePolitical Studies 2 (April): 7-67.
Rummel,R. J. 1970. Applied Factor Analysis.Evanston, Ill.: NorthwesternUniversity
Press.
Runciman,W. G. 1966. Relative Deprivation and Social Justice: A Study of Attitudes
to Social Inequality in TwentiethCenturyEngland. Berkeley: Universityof Cali-
forniaPress.
Seeman, Melvin. 1960. Social Status and Leadership: The Case of the School Board
Executive. Columbus: Bureau of Educational Research and Service,Ohio State Uni-
versity.
Seeman, Melvin, with the assistance of Richard T. Morris. N.d. A Status Factor Ap-
proach to Leadership. Columbus: PersonnelResearch Board, Ohio State University.
Tallman, Irving, and Marilyn Ihinger-Tallman. 1979. "Values, Distributive Justice
and Social Change." AmericanSociological Review 44 (April): 216-35.
Taylor, Charles Lewis, and Michael C. Hudson. 1972. World Handbook of Political
and Social Indicators. 2d ed. New Haven, Conn.: Yale UniversityPress.
Thompson,Eric J., ed. 1975. Social Trends,No. 6. London: Her Majesty's Stationery
Office,Central StatisticalOffice.
. 1976. Social Trends,No. 7. London: Her Majesty's StationeryOffice,Central
StatisticalOffice.
Titmuss,Richard M. 1962. Income Distributionand Social Change: A Study in Criti-
cism.London: Allen & Unwin.
Townsend, Peter. 1962. "The Meaning of Poverty." British Journal of Sociology 13
(September): 210-17.
Treiman, Donald J. 1977. Occupational Prestige in Comparative Perspective. New
York: AcademicPress.
Treiman,Donald J., and Kermit Terrell. 1975. "The Process of Status Attainmentin
the United States and Great Britain." AmericanJournal of Sociology 81 (Novem-
ber): 563-83.
Vanneman,Reeve D. 1980. "U.S. and BritishPerceptionsof Class." AmericanJournal
of Sociology 85 (January): 769-90.
Vanneman,Reeve D., and Fred C. Pampel. 1977. "The AmericanPerceptionof Class
and Status." AmericanSociological Review 42 (June): 422-37.
Warner,W. Lloyd, and Paul S. Lunt. 1941. The Social Life of a Modern Community.
Yankee City Series,vol. 1. New Haven, Conn.: Yale UniversityPress.
Wedderburn,Dorothy Cole. 1962. "Poverty in Britain Today-the Evidence." Socio-
logical Review 10 (November): 257-82.

348

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions
 CognitiveMaps of Inequalities

Westergaard,J. H. 1975. "Radical Class Consciousness: A Comment."Pp. 251-56 in

WorkingClass Images of Society, edited by Martin Bulmer. London: Routledge &
Kegan Paul.
Westergaard,J. H., and HenriettaResler. 1975. Class in a CapitalistSociety: A Study
of ContemporaryBritain. London: Heinemann.
Willener,Alfred.1957. Images de la societe et classes sociales: Une Etude de la per-
ception et des repre'sentations
des diffeirences
sociales. Berne: Staempfli.
Wright,Erik Olin. 1978. "Race, Class, and Income Inequality." AmericanJournal of
Sociology 83 (May): 1368-97.

349

This content downloaded from 200.52.255.1 on Thu, 18 Jul 2013 13:01:18 PM

All use subject to JSTOR Terms and Conditions