Social Indicators

Author(s): Kenneth C. Land

Source: Annual Review of Sociology, Vol. 9 (1983), pp. 1-26
Published by: Annual Reviews
Stable URL: http://www.jstor.org/stable/2946054
Accessed: 31/10/2010 11:43

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=annrevs.

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.

Annual Reviews is collaborating with JSTOR to digitize, preserve and extend access to Annual Review of
Sociology.

http://www.jstor.org
Ann.Rev. Sociol. 1983. 9:1-26
Copyright( 1983 byAnnual ReviewsInc. All rightsreserved

SOCIAL INDICATORS
KennethC. Land
ofSociology,
ResearchCenterandDepartment
Population The University
ofTexasat Austin, Texas78712
Austin,

ABSTRACT
This paperreviewsthedevelopment ofthefieldofsocial indicatorsfromits
originsin the 1960s to the present.Three classes of social indicatorsare
identified:
normative welfareindicators, whichfocuson directmeasuresof
welfareand are subjectto theinterpretation thatiftheychangein theright
directionwhile otherthingsremainequal thingshave gottenbetteror
peopleare betteroff;satisfaction indicators,whichmeasurepsychological
happiness,and lifefulfillment
satisfaction, by usingsurveyresearchinstru-
mentsthatascertainthesubjectiverealityin whichpeoplelive;and themost
inclusivecategory,descriptive social indicators,whichare indexesofsocial
conditions(i.e. contextsof humanexistence)and changesthereinforvari-
ous segmentsof a population.Correspondingly, two conceptionsof how
social indicatorsare to be interpreted and used are discussed:One, which
emphasizesthepolicy-analytic uses of social indicators,presumesthatthe
properrelationship of social indicatorsto social policyoccursat the level
of operatingor managingorganizations; the other,whichemphasizesthe
uses of social indicatorsin social reporting, presumesthattheproperrole
ofsocialindicatorsis publicenlightenment and theformation ofgeneralas
opposedto operationalpolicy.Threesociologicalcontributions to the de-
scriptivesocialindicators/enlightenment approachare described:thedevel-
opmentofreplication and longitudinal studies,theproductionofanalytical
studiesof social change and social reports,and the creationof formal
modelsfortheanalysisofdata on socialchange.Currentresearchproblems
thatareidentified includethedevelopment ofsocialaccountingsystems, the
construction of indicatorsof institutional and the
values and structures,
productionof improvedsocial forecastsand forecasting techniques.It is
concludedthat,whileissuesof publicconcernmay changefromtimeto
time,the criticalpublic and privatesectorscontinueto need statistical
information about currentsocial conditionsand trends.

0360-0572/83/0815-0555$02.00
2 LAND

INTRODUCTION
The termsocial indicatorswas born and givenits initialmeaningin an
attempt,undertaken by the AmericanAcademyof Artsand Sciencesfor
the NationalAeronauticsand Space Administration, to detect(evaluate)
and anticipate(assess)thenatureand magnitudeofthesecond-order conse-
quencesof thespace programforAmericansociety.The Academy'sman-
date was to investigatethe social, economic,and technologicalconse-
quencesthatwerenotimpliedeitherbythestatedobjectivesoftheprogram
(unintendedconsequences)or by the technologicaldiscoveriesemerging
immediately fromthe program(indirectconsequences).Severalscholars
involvedin thisprojectsoonbecamefrustrated bythelack ofsufficient data
to detectsuch effects. Furthermore, both a systematicconceptualframe-
workand a methodology forsophisticatedanalysisweremissing.Conse-
quently,some of those involvedin the Academy projectattemptedto
developa systemofsocialindicators-statistics, statisticalseries,and other
formsofevidence-withwhichto detectand anticipatesocial change.The
resultsof this generalizedversionof the projectwere published(Bauer
1966) in a volumethat inauguratedthe contemporary period of social-
indicatorsresearch.
The appearanceof this volumewas not an isolatedevent.Interestin
social indicatorswas furtherbuttressed by publicationof thereportof the
NationalCommissionon Technology,Automationand EconomicProgress
(1966). In theprocessofassessingtheeffects of factoryautomationon the
economyand society,the Commission(1966: 95-97) commentedon the
lack ofa systemforchartingsocial changesand advocatedthattheUnited
States Governmentestablisha "systemof social accounts" that would
facilitatea cost-benefit
analysisofmorethanthemarket-related aspectsof
societyalreadyindexedbytheNationalIncomeand ProductAccounts.At
aboutthesametime,Sheldon& Moore (1968: 3; see also Moore & Sheldon
1965)tooknoteoffar-reaching changesinthestructure ofAmericansociety
and assembleda massivevolumeof essayson indicatorsof social change.
Space limitationsprevent mentionofall theproposalsfor,and contributions
to,researchin social indicatorstouchedoffby thisinitialroundofpublica-
tions(an annotatedbibliography publishedby Wilcox et al in 1972 con-
tainedover 1,000items).
This sharpimpulseofinterest in social indicatorshas historically impor-
tantsocial-scienceprecursors withlinksto theeventsof the 1960s.Gener-
ally speaking,it grew out of the movementtowards collection and
organization ofnationalsocial,economic,and demographic data thatbegan
in Westernsocietiesduringthe 17thand 18thcenturiesand acceleratedin
thetwentieth century(Gross 1966;Carley1981:14-15).A moreproximate
 SOCIAL INDICATORS 3

and sociologicallygermaneprecedent is theworkofWilliamF. Ogburnand

his collaboratorsat theUniversity of Chicago in the 1930s and 1940s on
thetheoryand measurement of social change(Land 1975b).As chairman
of the President'sResearchCommitteeon Social Trends,Ogburnsuper-
visedproductionof thetwo-volume RecentSocial Trends(1933), a prece-
dent-settingcontribution to social reporting.Ogburn's ideas about the
measurement of social changeinfluenced severalof his students-notably
AlbertD. Biderman,Otis Dudley Duncan, AlbertJ. Reiss, and Eleanor
BernertSheldon-who haveplayedmajorrolesintheemergence and devel-
opmentof the fieldof social indicatorsduringthe past two decades.

DEFINITIONS OF SOCIAL INDICATORS

Shortlyafterthe appearanceof the publicationscitedabove therearose a
chorusofproposalsconcerning thedevelopment of social indicators.Dun-
can (1969: 1) dubbedit a social movement. This enthusiastic responsewas
due both to the successesand to the increasingly apparentlimitationsof
economicindicators.In theearly1960s,economists, usingindicatorsfrom
theNationalIncomeand ProductAccountsalong witheconometric mod-
els, were able to suggesteconomicpolicies (e.g. tax cuts) that had the
intendedresultofexpandingthegrossnationalproduct(GNP) byaboutthe
expectedamount(de Neufville1975). These successessuggestedto social
scientiststhatan analogousset of "social" indicatorsmightbe similarly
used in the manipulationof social policy.
But themoresuccessfuleconomicindicatorsbecame,themoreobvious
weretheirlimitations as measuresof generalsocial welfareor well-being.
They failedto reflectthe qualityof life,the dimensionof equity,or such
side-effectsof economic prosperityas environmental pollution(Carley
1981: 17). Some of the more specificlimitationsof economicindicators
identifiedby critics(e.g. Gross 1966) are that:(a) such measuresas the
GNP cannotbe equatedwithpsychologicalsatisfaction, happiness,or life
(b) marketvaluationsofgoodsand services(i.e. theirobserved
fulfillment;
orimputedprices)arenotnecessarily relatedto theircontributionsto social
well-being;(c) nonmarket activitiesand impactsof marketactivitiestend
tobe excludedfromconsideration; (d) distributional considerationstendto
be obscuredin the GNP averages;and (e) GNP indicatorsassume the
existenceof a social rule system,incorporating values,norms,and struc-
turesgoverning humanbehavior,buttheydo notmeasurethis,nordo they
measuretheinterdependence betweenchangesin thissystemand theecon-
omy.
Such criticismshave producedtwo generalreactions:attemptsto alter
theNationalIncomeand ProductAccountsto reflect well-beingissues,and
4 LAND

to developindicatorsand accountsthatcomplement
efforts theuseful(and
inescapable)economicindicators.These two approachesstillcharacterize
muchcurrentsocial-indicators research.
NormativeWelfareIndicators
Takingas its premisethe propositionthatsocial indicatorsshouldrelate
an earlydefinition
considerations,
directlyto social-policy-making by the
economistMancurOlson characterizeda social indicatoras a
statistic
ofdirectnormative whichfacilitates
interest concise,comprehensive andbal-
ancedjudgments abouttheconditionofmajoraspectsofa society.It is,in all cases,a
directmeasure ofwelfare thatifitchangesin the
andis subjectto theinterpretation
'right' whileotherthings
direction, remain equal,thingshavegottenbetter, or people
arebetteroff. onthenumbers
Thus,statistics ofdoctorsorpolicemencouldnotbesocial
whereas
indicators, onhealth
figures orcrimeratescouldbe.(US Department ofHealth,
Education andWelfare 1969:97).
In thelanguageofpolicyanalysis(see e.g. Fox 1974: 120-23),socialindica-
torsare "target"or "output"variables,towardschangesin whichsome
publicpolicy(program,project)is directed.Such a use ofsocial indicators
requiresthat(a) the societyagreesabout whatneeds improving; (b) it is
possibleto decideunambiguously what "gettingbetter"means;and (c) a
highdegreeofaggregation in theindicatorsexiststo facilitate
national-level
analyses.
Experiencedsocialanalystsquicklyperceiveddifficulties inmeetingthese
requirements. 0. D. Duncan (1969: 3-4) observedthatthe emphasison
aggregation maybe premature. Sheldon& Freeman(1970: 98) arguedthat
what is of "directnormativeinterest"today may not be so next year;
furthermore, the requirement thatindicatorsbe "directmeasuresof wel-
fare"rulesout manyvariablesthatmaybe relevantto an understanding of
theindicator.This requirement presupposesthat"outputs"can be unam-
biguouslyidentified and ignoresthe fact that the outputof one social
processmay be the inputof another.
Indicators
Satisfaction
An alternative approachto the definition of social indicatorshas its'roots
in TheHuman MeaningofSocial Change(Campbell& Converse1972). A
companionto theSheldon& Moore (1968) volume,thiscollectionofessays
arguedthatthe directmonitoring of keysocial-psychological states(atti-
tudes,expectations, feelings,aspirations,and values) in the populationis
necessaryto an understanding of social change.In a subsequentvolume,
Campbellet al (1976: 4) connectedthisproposition to thenotionofquality
oflife:"The researchwithwhichthisbook is concernedderivesfromthe
convictionthatthe relationship betweenobjectiveconditionsand psycho-
logicalstatesis veryimperfect and thatin orderto knowthequalityoflife
 SOCIAL INDICATORS 5

experienceit will be necessaryto go directlyto the individualhimselffor

hisdescription ofhowhislifefeelsto him."Whiletheterm"qualityoflife"
is seldompreciselydefined, it typicallyinvolves"a senseofachievement in
one's work,an appreciationof beautyin natureand the arts,a feelingof
identification withone's community, a senseof fulfillmentof one's poten-
tial" (Campbellet al 1976: 1).
In thisapproach,socialindicatorsseekto measurepsychological satisfac-
tion,happiness,and lifefulfillment by directlyascertainingthe"subjective
reality"in whichpeoplelive.Towardsthisend,twomajormethodological
studies(Andrews& Withey1976; Campbellet al 1976) have exploredthe
utilityof varioussurveyand analytictechniquesin mappingindividuals'
feelingsof satisfaction withvariousaspects ("domains") of theirexperi-
ences. These studiesexamineddomainsrangingfromthe highlyspecific
(house,family, etc)to theglobal(life-as-a-whole). Otherapplicationsofthe
techniquesproducedby thesestudieshave appearedin the recentjournal
literature (e.g. Andrews& Inglehart1979; Michalos 1980; Headey 1981).
The linkbetweenchangesin objectiveconditionsand psychological states
is bothindeterminate and paradoxical.For instance,Easterlin(1973) and
0. D. Duncan (1975a) foundthat,whilehigherincomelevelsare associated
withhigherlevelsof happinessor satisfaction forindividualsin a given
countryat a giventime,higherper capitalevelsofnationalincomedo not
producehigheraveragelevels of nationalsatisfaction over timeor cross
sectionally. Becauseofthisand similarparadoxes,subjectiveindicatorswill
likelybe a permanent fixture in social-indicatorsresearch.However,from
thestandpointof a generaltheoryof social change,it is not yetclear that
themostdesirablesubjectiveindicatorsare satisfaction indexesas opposed
to indicatorsofexpectations, aspirations, attitudes,or values.Furthermore,
itis notclearthatsatisfaction indexesprovidean unambiguouscriterion for
theformulation ofpublicpolicy.For instance,theremaybe circumstances
in which it is not possibleto maximizeboth individualand collective
satisfactions (see thediscussionin Land & Juster1981: 18). A case in point
is whenthesatisfactions ofsomeare reducedby integrated neighborhoods.
In addition,researchto date (e.g. Andrewsand Withey1976) showsthat
theareas of lifemostcrucialto an individual'soveralllifesatisfaction are
thoseleast amenableto publicpolicyactions.

Social Indicators
Descriptive
A thirdapproachto the definitionof social indicators,buildingupon the
Ogburnlegacyofresearchon social trends,focuseson socialmeasurements
and analysesdesignedto improveour understanding of what the main
featuresof societyare, how theyinterrelate,and how thesefeaturesand
theirrelationshipschange(Sheldonand Parke 1975: 696). 0. D. Duncan
6 LAND

(quotedin Sheldon1971:430) emphasizesthatthebedrockofsocialindica-

torsconsistsof social measurements,
upon whicha varietyof edificesmay
be constructed:

Whatwemusthave,minimally, arequantitativestatements
aboutsocialconditionsand
socialprocesses,
repeatedlyavailable
through time,thereliability
andvalidity ofwhich
arecompetentlyassessedandmeetminimal standards. Ifsuchstatements-"socialmea-
surements"-can beorganizedintoaccounts... somuchthebetter.Ifsomecombination
ofmeasurements or quantities
derived fromelementary magnitudescan be shownto
servea clearinterpretive
purposeas "indicators,"
so muchthebetter. As accounting
schemes,modelsofsocialprocesses,andindicators aredevelopedandtested, ouridea
ofwhattomeasure will,ofcourse,
change.Butthatdoesnotaltertheprinciple thatthe
basicingredients
arethemeasurements themselves.We aretalkingaboutinformation,
theprocessingofinformation,andthereporting ofprocessedinformation.

In thisview,social indicatorsare simplyindexesof social conditionsand

changesthereinovertimeforvarioussegmentsof a population.By social
conditionsaremeantboththeexternal(physicaland social)and theinternal
(subjectiveand perceptional) contextsofhumanexistencein a givensociety
(Land 1975b:14). Earlycontributions to thisperspectiveincludeBiderman
(1966),Sheldon& Moore(1968),0. D. Duncan (1969), Ferriss(1969, 1970,
1971), Sheldon& Freeman(1970), and Campbell& Converse(1972).
Because it does not requirethat social indicatorsbe definedsolelyin
termsof objectiveor subjectivewell-being, thisdefinition is moregeneral
thanthosecitedabove. It producesdescriptivesocial indicators-indexes
ofthestateofa societyand thechangestakingplace withinit. Illustrations
can be foundin Social Indicators1976 and Social IndicatorsIII (US De-
partment ofCommerce1978,1980).Thesevolumesreportdozensofstatis-
ticalserieson such topicsas health,housing,publicsafety,schooling,and
social participation.Althoughdescriptive social indicatorsmaybe moreor
less directlyrelatedto thewell-beinggoals of government policiesor pro-
grams,theyare not limitedto such uses.
As statistics,
descriptive social indicatorsusuallytakeone offourforms,
or some combinationthereof:(a) state-occupancy rates-indexes of the
prevalenceofsomesocialstatein thepopulation(e.g. percentofthepopula-
tionthatis disabled);(b) transition rates-ratesoftransition betweensocial
states(e.g. the probabilityof survivingfromage x to age x + 1); (c)
occurrence/exposure rates(e.g. thenumberofbirthsper 1000femalesaged
15-44) or ratios(e.g. theproportion ofdeathswithina givenyearattribut-
able to a particularcause); and (d) indexessummarizingpopulationor
subpopulationdata in formsa, b, or c (e.g. median familyincome in
constantdollars)(cf. Land 1979: 219-28). These can be orderedby degree
ofabstraction fromthosethatrequireonlyone or twodata seriesand little
processing(e.g. theage-specific deathratesfora sequenceof age groupsin
 SOCIAL INDICATORS 7

successiveintervalsof time)to those that involvethe more complicated

processingof severaldata seriesinto a singlesummaryindex(e.g. a life
expectancyindex).Descriptivesocial indicatorscan be formulatedat any
oftheselevelsofabstraction(cf.UnitedNationsStatisticalOffice1975:27).
This is analogous to the practice in descriptiveeconomic indicators,
wherein"tonsofsalablecoal shipped"maybe used as a simpleindexofthe
outputof the coal-miningindustrywhereas"gross nationalproducein
constantdollars"is a highlyprocessedindicatorof economicactivity.

THEORETICAL APPROACHES TO SOCIAL

INDICATORS
Severalotherdefinitionaldistinctions,pertainingto the relationsof social
indicatorsto publicpolicies,government programs,and analyticalmodels
ofsocialchange,haveappearedintheliterature (see e.g. Land 1971;Carlisle
1972;Sheldon& Land 1972;Land 1975a,b). However,in termsoftheuses
to whichtheindicatorsare put,thethreecategoriesdescribedabove com-
prisethe main contemporary To understandthese uses re-
distinctions.
quiresa moredetaileddescription ofthetheoretical approachesunderlying
the definitions.

ExtensionsofPolicyAnalysisand Microeconomics
From the outset,normativewelfareindicatorshave been surroundedby
policy-analytic/microeconomic theoriesand concepts.Initially,the idea
was that social indicatorscould be linkedto public programsvia input-
outputtablescalled "policyaccounts":

The factthatrationalpolicynecessitates to program

linkingsocial indicators inputsmeans
thatsocial indicatorsalone do not provideall of the quantitative informationneeded
foreffective decisionmaking.Ultimately, we mustintegrate our social indicatorsinto
policyaccountswhichwould allow us to estimatethechangesin a social indicator that
couldbe expectedto resultfromalternative levelsof expenditure on relevantpublicpro-
grams.

Of course,it was recognizedthatsocial conditionsare affected

by things
otherthanpublicprograms:

Thecondition ofan aspectofa nationdepends,notonlyon a particular

publicprogram,but
also on manyotherthings.Healthand lifeexpectancy, forexample,dependnotonlyon
publichealthprograms, thestandard
butalso on privatemedicalexpenditures, ofliving,the
8 LAND

qualityofnutrition,
theexposuretocontagious diseases,andthelike.Thus,todeterminethe
outputofa publicprogram we normally haveto solvesomething likewhattheeconometri-
cianwouldcall the"specification
problem";we havetoidentify ordistinguish
thosechanges
inthesocialindicatordue tothechangedlevelsofexpenditure on thepublicprogram.This
is oftennota tractabletask,butit could contributemuchto trulyrationaldecisionmak-
ing.(Bothquotations takenfromUS Department of Health,EducationandWelfare1969:
101).

Researchin thistraditionhas recentlytakentwo main forms.

The firsttravelsundertherubricof nationalgoals accounting(Terleckyj
1975).It attempts to giveexplicitdefinition to nationalgoals and priorities,
to estimateprogresstowardtheachievement ofthesegoals,and to measure
the cost of furtherachievement. This is done by (a) identifying national
social concernsand indicatorsto measureconditionsrelevantto those
concerns(morewillbe said aboutthisin thenextsubsection);(b) projecting
baselinetrendsin theseconditionsand in resourcespotentially availablefor
alteringthem;(c) estimating thecostsand effects of discretionarygovern-
mentprogrammatic activitiesaffecting social conditions;and (d) calculat-
ingthemaximumfeasibleimpactofcombinations ofdiscretionary activities
that can be undertaken, givenresourceconstraints, takinginto account
multipleconsequencesofprograms, nonadditivity ofeffects,and othercom-
plications.
One of the weakestlinksin this programis the projectionof baseline
trendsin social conditions.Terleckyjdoes thiscrudely(1975). Similarly,
because social conditions(and changestherein)depend on manythings
otherthangovernment programs,thereare severeproblemsin estimating
thechangesin socialconditionsthatcan be expectedfromvariouslevelsof
government expenditure.
Partlyin recognitionof this,recentresearchon resourceaccounting
systemshas adapted the householdproductionand consumptionframe-
workof"thenewhomeeconomics"to takeintoaccountthewaysin which
"households"combinecommodities purchasedin themarket(e.g. grocer-
ies) withcapitalstocks(e.g. refrigerators, ranges,cookware,and utensils)
and timeexpenditures in activities(e.g. cooking) to producehousehold
commodities(e.g. meals) that produce feelingsof satisfaction(see e.g.
Juster,Courant& Dow 1981;Terleckyj1981). In thisway,it is anticipated
thatthe effectsof changesin levels of living,consumptionpatterns,life
styles,etc can be tracedthroughhouseholdproductionand consumption
activitiesto theirimpactson the "ultimatewell-beingoutcomes" (e.g.
lengthof healthylife). Furthermore, by postulatingthe dependencyof
individualor householdsatisfaction levels on the latter,it is possibleto
incorporatethe social-psychological traditionof researchon satisfaction
withinthisframework.
 SOCIAL INDICATORS 9

Such a systemrequiresvoluminousdata on householdtimeexpenditures

on a largearrayof"production"activities(see below). Moreover,thevalid-
ityofthesystemrestson twoassumptions. The firstis thatit is meaningful
to applyeconomicconceptsofproduction and consumption, and therelated
theoretical apparatus,to nonmarketactivities-i.e.that nonmarket activi-
tiesand relatedsocial structures
organizationcan be adequatelyprotrayed
byeconomicconceptsand models.Second,thisapproachassumesthatthe
effects of the"manyotherthings"thatimpacton social conditionscan be
aggregatedto obtainsocietal-leveleffectson social conditions-i.e. that
there is no "fallacyof composition"in extrapolatingfromindividual
householdbehaviorsto societal-levelconditions.

and Social Reporting

Enlightenment
Severalothertheoreticalapproachesto social indicatorshave been built
upona microeconomics foundation.For instance,Fox (1974; Fox & Ghosh
1981) appliedutilitymaximization theoryto developa "monetaryindexof
totalincome"as the sum of an individual'swage (and otherreal-dollar)
incomeplus an imputeddollarvalue forthepsychicsatisfaction obtained
fromvariousactivities(e.g. gettinga haircut)in varioussettings(e.g. a
barbershop),and Liu (1976) used microeconomicconceptsto definea
"qualityof lifeproductionfunction"fromwhich he computeda global
"qualityof life"indexforUS metropolitan communities. All of theseap-
proachesrestupon thepresumption that the properrelationship of social
indicatorsto social policyoccursat a levelofoperatingor managingorga-
nizations.
An alternative perspective assumesthattheproperroleof social indica-
torsis enlightenment-i.e. theycontribute to publicunderstanding ofsocial
conditionsand social changeand to theformulation ofgeneralas opposed
to operationalpolicy(Biderman1970; Sheldon& Parke 1975). This ap-
proachemphasizestheuse ofsocial reporting (a) to describesocial trends;
(b) to explainwhy an indicatorseriesbehaves as it does and how this
knowledgeaffectsinterpretation; (c) to highlightimportantrelationships
amongseries;and (d) to suggesttheconsequencesreasonablyattributable
to changesin a series(Parke & Seidman 1978: 15).
Social reportinghas motivatedthe developmentof descriptivesocial
indicatorssinceOgburn'stime.The contemporary era of social reporting
datesfromthepublication ofIndicatorsofSocialChange(Sheldon& Moore
1969) and Towarda Social Report(1969). Since then,thefederalgovern-
menthas periodicallypublishedcompendiaof stastisticalserieson major
social trendsin the United States (US Departmentof Commerce1974,
1977,1980).Thesehavebeensupplemented at thefederallevelbynumerous
specialreportsandbysocialreportsat thestateand local levels.Manyother
10 LAND

nationalgovernments and severalinternationalagenciesalso have pub-

lishedsocial reports(forsurveysofthese,see Zapf 1976and Glatzer1981).
Althoughall of these reportspresentstatisticalinformation depicting
important aspectsofcurrentsocial conditionsand theirunderlying histori-
cal trendsand conditions,theyvarygreatlyin the amountsof attention
givento the explanatoryand interpretive goals of social reporting.For
instance,theUS Government reportscitedabovetendto interpretonlydata
characteristicsand limitations.Special volumesof interpretative essays,
writtenby experiencedresearchersin varioussubstantiveareas, have ac-
companiedthetwomostrecententriesin thisseries(Taeuber 1978,1981).
Drawingon largebodiesof empiricalresearch,analyticmodels,concepts,
and theories,theseessaysdrivehometheinescapableheterogeneity of the
"social" in social indicators.

Dimensionsof Theoryand theDefinition

of Social Indicators
Space limitationspreventfurtherelaborationon these two different ap-
proachesto social indicators.Nonetheless,even thesebriefcharacteriza-
tionsprovidea backgroundforunderstanding thegenesisofthedefinitions
ofsocialindicatorsdescribedabove.For,ifone holdsthesocial-engineering
objectivesofthepolicy-analytic/microeconomics approachto socialindica-
tors,thenone's interestis in normativewelfareindicators.Furthermore, if
one embedsthisapproachin a householdproductiontheoryin whichthe
"finaloutputs"ofindividualandjointproductionare consumedto provide
utility,thenit followsthatsubjectivesatisfactionindexescan be absorbed
withinthesameframework. On theotherhand,ifone assumestheenlight-
enmentgoalsofsocialreporting, thenone's focusnaturallyfallson descrip-
tiveindicatorsof social conditionsand trendsand on interpretations and
explanationsthereof.
In eithercase, certainpresuppositionsmaybe hidden.For instance,the
definitionof normative welfareindicatorspresupposessubstantialconsen-
suson "socialobjectivesorgoals." Similarly,thedefinition
ofglobalquality
of lifeindicatorspresupposesthe abilityto aggregatemeaningfully across
individualsand socialconditions.Finally,theidentification
of keyvariables
thatdetermine socialconditionsand changespresupposescausal knowledge
about the processesgenerating the conditions.Whetheror not thesepre-
sumptionsare factuallyvalid is an empiricalquestion.

SOCIOLOGICAL CONTRIBUTIONS TO
SOCIAL-INDICATORS RESEARCH
In spiteoftheirdifferences
in orientation,
however,it wouldbe misleading
to portraythesocial-engineering
and social-reporting on social
perspectives
 SOCIAL INDICATORS I1

indicatorsas opposites.Manysociologists working in a social-reporting

modehavea long-run goalsimilarto thatofpolicy-analysts/microecono-
mistsworking mode-namely,
in a social-engineering to deviseexplana-
tionsofsocialtrendsthatlead to thedevelopment ofpoliciesforcoping
and/orcontrolling
with,altering, socialchange.But,perhaps becausethey
comefroma discipline richinempiricalstudiesofdiversesocialphenom-
ena,sociologiststendto be skepticalaboutthepossibility of developing
socialindicatormodelsbased on microeconomic conceptsthatlead to
effectivepolicies.Insteadtheyemphasizetheneedforbetterand more
complete measurements and analyses.Again,thisattitude founditsex-
pressionin Ogburn'sworkoverfifty yearsago:

timeseriesshowa veryimportant
Ourvariousstudiesof statistical thing,namely,thatthe
measured trendofeventsandphenomena is thebestguidethatwe yethavefortheprediction
ofthefuture. ofmystery
Knowledgeis theantithesis Andtheknowledge
anduncertainty. of
whathasoccurred andofwhatis happening is thesafestguidewe have.Withmorecomplete
andwithbetter
statistics measurement we shallattainfullerknowledge ofwhatis happening
tous andwherewe aregoing.Onlywiththeseshallwe be ina positioneventobegintospeak
of control.(Ogburn1929 [1964]: 101).

Working outofthistraditionofsocialmeasurement, bysoci-

contributions
toresearch
ologists on socialindicators on thedevelop-
haveconcentrated
mentofmeasures ofsocialconditionsandtheiruseinlongitudinal
designs,
ofsocialchangeandsocialreporting,
studies
analytical andformalmodels
ofsocialchange.

Replicationand LongitudinalData Series

Fromthelimited experience ofthepastfewyears,itmight be assumedby
thattheideaofrepeating
a novicesociologist majorcommunity studies and
nationalsamplesurveys themeasurement
to facilitate of long-term trends
andothersocialchanges hasalwaysbeenacceptedamongsocialscientists
Thatthiswasnotthecaseeventwodecadesagois borne
andstatisticians.
outbyrecallingthereaction totheBauer(1966)volume
ofthecontributors
to thetaskwithwhichtheywerecharged. The dearthofrepeated studies
andofotherdatausefulinthemeasurement of change led O. D. Duncan
(1969:10) to assert:"Obviously, a minimum requirement formeasuring
changeis thatthesamevariable be measuredat two distinct pointsin
time.... In addition..., the two (or measurements
more) must be suffi-
cientlycomparableand reliable thatthe of
inference change from the differ-
of
ence theirmagnitudes is trustworthyto an acceptable degree of
confidence."Duncanwentontoarguethatan improved abilitytomeasure
socialchangeis fundamental toprogressinthefieldofsocialreporting. He
12 LAND

proposed thata strategyofreplicating baselinestudieswouldmakeoptimal

useofexisting datasourcesandcitedseveralexamples ofexisting baseline
studiesthatmight meritreplication. Duncan
Finally, formulated guidelines
forselectingand replicating baselinestudies.
In theyearssincepublication of Duncan'smonograph, muchof his
researchprogram hasbeenrealized.First,severalmajorstudies doneprior
to 1969havebeenreplicated. Examples aretheDuncan,Schuman & Dun-
can (1973) replicationofitemsfromprevious DetroitAreasurveys; the
Featherman & Hauser(1978)replication ofBlau& Duncan's(1967)study
ofoccupational mobility patterns ofAmerican males;Davis's(1975)repli-
cationofStouffer's(1955)study ofAmericans' attitudestowards civilliber-
ties;and thecurrent replication byT. Caplowand associates(Caplow&
Bahr1982)oftheLynd& Lynd(1937)studiesofMiddletown (Muncie,
Indiana).Second,severalmajornationalsamplesurveys havebeensetup
on a repeatedbasisbothwithingovernment agenciesand in academic
survey organizations(fora comprehensive survey ofexisting socialindica-
tordataseries,seeTaeuber& Rockwell, 1982).Examples arethe(currently
NationalSurvey
triennial) ofFamilyGrowth takenbytheNationalCenter
forHealthStatisticssince1973andlinkedtoearlier fertility
surveys (P.A.A.
Affairs,Summer, 1981:3); theGeneralSocialSurvey(currently biennial)
takenbytheNationalOpinion Research Center since1972(underNational
ScienceFoundation sponsorship) andlinkedtopriornational opinionsur-
veys(Smith1980);andtheAmerican NationalElectionStudy(currently
biennial)takenbytheUniversity ofMichigan Survey Research Center since
1972(underNationalScienceFoundation support)and linkedto earlier
electionstudies(Inter-University Consortium forPoliticaland SocialRe-
search,1980-81:185-92).
Sociologistshavealso beenactivein theapplication oftwootherlon-
gitudinalresearchstrategies to producedatasetsthatcanbe analyzedfor
socialreportingpurposes: panelandevent-history designs.Examples ofthe
former aretheSewell-Hauser PanelonWisconsin Youth(Sewell& Hauser
1975);theNationalLongitudinal StudyofLaborForceExperience (Parnes
1972);andthePanelStudyinIncomeDynamics (Morgan& Smith1969).
Anillustrationofthelatter istheColemanetal (1972a,b) study oftheearly
occupationalcareersofwhiteandblackmales.
AnalyticalStudiesofSocial Changeand Social Reporting
Theincreasing ofover-time
availability datasetsintheformofreplications
of baselinestudies,panels,event-histories,
and timeserieshas allowed
andothersocialscientists
sociologists toteaseoutplausible,
informed expla-
nationsofcontemporary socialtrends.
Someofthesearecloselyassociated
withthesocialindicators rubric,as whentheyappearin collections of
 SOCIAL INDICATORS 13

essaysthatcomment on governmentcompendia of statistical

series(see
Taeuber1978,1981)orwhentheyaremembers ofthenew"SocialTrends
in theUnitedStates"monograph seriessponsored
by theSocialScience
ResearchCouncil(see Cherlin1981). Most,however,are published in
journalsor in otherresearchreportsand are viewedby their
refereed
authorsas contributions
to basic knowlegeratherthanto social reporting.
Because thelatterbodyof literatureis bothlargeand diverse,reviewof a
fewexamples-chosento illustrate a broaderrangeofissues,data sets,and
modes of analysis-will have to suffice.I introducetheseas answersto
rhetoricalquestions.
After a decadeofcivilrights activismandlegislativegainsinthe1960s,
didtherelative incomegapbetween blacksandwhiteswidenornarrow in
the1970s?UsingPublicUse Sampledatafrom the1970CensusofPopula-
tionandtheMarch1976AnnualDemographic FileoftheCurrent Popula-
tionSurvey, Bianchi(1980)foundthattheratioofblack-to-white average
household income(in 1975dollars)was.63in 1970and.64in 1976.From
thisone mightinferthattherewas littleor no progress towardsincome
equalitybetween theracesintheperiod1970-75.However, bydecompos-
ingthedatabytypeofhousehold Bianchifoundthatthisratiogrewfrom
.70 to .75 forhusband-wife households and from.72 to .79 forfemale-
headedhouseholds. Thustherewas substantial movement towardsracial
equality in incomeforthesehousehold types.
Howcanthesefindings bereconciled? Usingdemographic decomposition
techniques, Bianchi(1980)combined twofactsintoan explanation. First,
bothin 1970and 1976,theaveragehousehold incomeforfemale-headed
households was substantiallylessthanthatofhusband-wife households.
the
Second, proportion of black households headedby womenincreased
fasterfrom1970to1976thandidthatproportion ofwhite households. Thus
eventhoughthetrendwastowardsgreater racialincomeequalitywithin
eachhousehold type,thegreater rateofflowofblackhouseholds intothe
moreincome-disadvantaged type(duringthisperiod)yieldslittleoverall
progress. Whilethesefindings andtheirexplanation maynotleadtospecific
operational for
suggestions government policiesdesigned to promote in-
comeequality between theraces,theyindicate changes that have occurred
andsuggest areasinwhichtodevelopgeneral policiesthatmight alterthese
trends.
Conceiving ofcontemporary liberalism"
"political as a bodyofattitudes
thatarereformist, democratic (participatory),
supportive ofcivilliberties,
regulatory and interventionist, centralist,
humanitarian, egalitarian,and
permissive, has theUS adultpopulation beenmovinglargely in a liberal
direction overthelastfourdecades?By splicingtogether and analyzing
trends pertaining to relevantattitudinalitemsspanning theyears1938-
14 LAND

1978 fromthe General Social Survey,the AmericanNational Election

Studies,andearlier GallupandHarrispolls(seeSmith1980),Smith(1982)
concludesthattherehas indeedbeena generalshifttowardliberalism
during thepost-World WarII period.A plurality ofattitudetrends moved
towardliberalism. Smithfoundevidencethatthisliberalshifthas weak-
ened,butnotreversed, duringthelastdecade.Furthermore, theliberal
movement hasnotbeenuniform acrossall subjects-attitudes toward abor-
tion,civilliberties,racerelations, and religionhavemovedmostconsis-
tentlyintheliberaldirection, whilecrime/violence andspending/taxation
itemshaveshownmoreconservative trends.
Howcanthesetrends beexplained? Smith(1982)concludes thatthemain
causesofthegeneral trendtowardliberalism were(a) modernization (the
twentieth-century shifttoa post-industrial
economy basedontechnological
innovation, scientific
progress, professionalization,and theefficient orga-
nizationofpeopleandknowledge) and (b) America'sliberalidealismas-
sistedbytheNewDeal Realignment andinstitutionalleadership fromthe
SupremeCourtand otherorganizations. By contrast, he findsthatthe
conservative ofpublicspending
shift inthe1970scanbeexplained
priorities
by(a) therising burden oftaxation,(b) theweakening ofeconomic growth,
(c) thedisillusionment withGreatSocietypromises, and(d) thefactthat
spending doesnotpresent a clearconflictbetween liberalidealsandliberal
attitudes.
Finally, he attributestheconservative trendin attitudes toward
crimeandviolence totheupsurge inviolent andproperty crime(during the
1960sand 1970s)and theassociateddecreasein judicialpunishments.
Again,thesetrends andtheirexplanations arenotmeanttoleadtospecific
operational programs. Buttheyareinformative andenlightening, andthey
leadtoexpectations abouthowfuture economic andsocial-structural trends
mayaffect shiftsin politicalattitudes.
Has theUS welfare system createda newcasteofAmericans-perhaps
as muchas onetenth ofthepopulation-who arealmosttotally dependent
on thestate,withlittlehopeor prospect ofbreaking free(as claimedby
Anderson1978:56)?Administrative recordsshowhowmuchthepublic
sectorspendsonwelfare programs eachyearandhowmanyfamilies receive
incomefrom welfare sources.Whenputintotime-series form, thestatistics
showsubstantial growth inthewelfare rollsoverthepastseveraldecades,
withcloseto one-tenthofthepopulationnowreceiving in one form
benefits
or another.But analysesofdata fromthePanel Studyin IncomeDynamics
on thepatterns
ofwelfare useduringthelate 1960sand 1970sshowthat
whilemanyfamilies received
incomefromwelfare sourcesat leastoncein
thedecade,fewer thanonefifth
oftheserecipients
(about4.4% oftheUS
couldbe characterized
population) as dependent uponthisincomefor
extendedperiods(G. Duncanetal 1982).Manyfamiliesreceiving welfare
 SOCIAL INDICATORS 15

wereintheearlyphasesofrecovery
benefits from an economiccrisiscaused
bydeath,departure, ordisabilityofa husband, a processthatoften culmi-
natedin findingfull-timeemployment, remarriage, orboth.Furthermore,
mostofthechildren offamilies aidedby thewelfare systemduringthis
perioddidnotthemselves receivewelfarebenefits aftertheylefthomeand
formed theirownhouseholds. Theseandmanyrelatedfindings aboutthe
dynamics ofpoverty, welfare,workhours,andearnings (G. Duncanetal
1982)provide a rationalbasisfromwhichtoevaluateandsuggest changes
in thenation'swelfare system.
Whatis therelationship between fluctuations in economicactivity (as
indexedbytheunemployment rate)andhealth(as indexed bycause-specific
mortalityrates)?In a reportto theJointEconomicCommittee oftheUS
Congress,Brenner (1976)claimedthatan increasein theunemployment
ratewasassociated withsubsequent increasesinage-,sex-,andrace-specific
ratesofhomicide, deathfrom
suicide, cirrhosis oftheliverandcardiovascu-
lardisease,andtotalmortality. Brenner reachedtheseconclusions byre-
gressingannualUS timeseries(datingfromthe1930sthrough theearly
1970s)foreachmortality indexonannualratesofunemployment, inflation,
andrealpercapitaincome.Cohen& Felson(1979)pointed outthatBren-
ner'sequationsweremisspecified, anassertion corroborated byautocorrela-
tionand otherdiagnostic statistics
reported byBrenner (1976).Thatthis
mayindeedhaveproduced spurious findings is suggestedbya subsequent
studyofLand& McMillen(1980).Usinga moreelaborate dynamic struc-
turalequationmodeling strategy forpost-World War II US timeseries,
theseauthorsconfirmed Brenner's findings onlyforthecardiovascular
diseasemortalityrate.Furthermore, theyfoundthatthepositive effectof
theunemployment rateonthismortality ratebecamestatistically insignifi-
cantwhencigarette consumption-which tendsto riseduringa recession
-is controlled.In a similartimeseriesstudyofdataon Canadaand its
provinces,Adams(1980)corroborated theLand-McMillen findings.The
implicationsofthelatterstudies forgeneral policiesbywhichtosoften the
impactsofeconomic downturns onincreases inmortality arequitedifferent
fromthoseofBrenner's (1976)study.
FormalModels ofSocial Change
Inresponsetotheanalytical
problemsandopportunitiesassociatedwiththe
developmentofnewandexpanded dataseries, andstatisticians
sociologists
havedevelopednewanalyticalmodelsbywhichtodistillinformation
about
socialchangefromthedata.For instance, the
byconjoining development
modelsfortheanalysisofdiscrete
oflog-linear surveydata in theearly
1970s(seee.g.Goodman1972a,b) withthereplications
ofbaselinesurveys
launchedduringthatperiod,Davis (1975) and 0. D. Duncan(1975b)
16 LAND

developed discrete multivariate modelsfortheanalysisofsurvey replica-

tions.Regression andpath-analytic to
approaches replication analysis were
describedby Treiman & Terrell (1975), Winsborough (1975),and Feather-
man& Hauser(1975).
Whenseveralreplications ofcross-sectional surveys areavailablefrom
successiveperiods andspanning a broad range of ages in eachsurvey, a new
analyticproblemoccurs in both the regression and the log-linear ap-
proachesto theanalysis of change. This is the problem of identifying and
allocatingcomponents ofthetotalobserved changes ina dependent variable
uniquelyto the effectsof the various age groups, periods, and cohorts
(definedby period ofbirth or some other common event)-the so-called
age-period-cohort (apc)problem. Becausetheageofan individual at some
is
period(date) exactly equal to that date minus her/his birthcohort date,
solutionsto thisproblem generally requiretheintroduction of a priori
constraints and/orexternal information of someotherkind.Definitive
statements ofthesemodeling issues were givenbyMasonet al (1973)for
theregression case andby Fienberg & Mason (1978)forthelog-linear case.
Substantive applications of apc and related models to US data that are
relevanttosocialreporting include Winsborough (1975),Fienberg & Mason
(1978),B. Duncan(1979),Pullum(1980),Mason& Smith(1981),and0.
D. Duncan(1982).
The emergence during the1970sofpaneldatasuitablefortheanalysis
ofnationalsocialtrends similarly has stimulated workon formal models.
Forinstance, Goodman(1973)developed log-linear modelsforthecausal
analysisofcategorical paneldata,whereasHannan& Young(1977)sur-
veyedand assessedtheperformance ofseveralestimation algorithms for
regression-type causalmodelsina panelcontext. Bothoftheseapproaches
modelchangesin thestate-distributions of panelmembers at successive
wavesofa panel.Bycontrast, Singer & Spilerman (1974;1976a,b),building
on a tradition ofsociological analysisthatdatesbackto Coleman(1964,
1968),explored a number of statistical issuesintheestimation ofcontinu-
ous-time Markovmodelsofmobility between successive panelwaves;see
alsoCohen& Singer (1979);Singer & Cohen(1980),andSinger (1981).This
researchhasrevealedtwodifficulties-called theembedability andidenti-
problems-infitting
fication stochastic processmodelstopaneldata.To
circumvent theseproblems, Tumaetal (1979)advocated theuseofevent-
historydataanddevelopedestimation methods forthecausalanalysisof
suchdatavia Markovmobility models.
The foregoing stochastic modelsmakeefficient use of individual-level
longitudinal datatoanalyzemobility trajectories. However, manydescrip-
tivesocial-indicatorsdataareavailableonlyin theformoftimeseriesof
populationstocks(e.g.theproportion ofindividuals between theagesof20
 SOCIAL INDICATORS 17

and24 whoareinthelaborforce)andflows(e.g.thenumber ofdeathsper

calendaryearofmalesaged55-59).A majorresearch question tobe asked
of suchdata is howthestocksand flowsare influenced by othersocial,
demographic, andeconomic Untilrecently,
changes. however, theanalysis
ofsuchdatatypically employed ad hoc combinations of econometric and
time-seriestechniquesthatignoretheunderlying population-dynamics/so-
cial mobilitystructure timeseries.In response
ofsocial-indicator to this,
Land (1979)developed an ergodicmodelofaggregate socialchangethat
generalizestheclassicallifetableandstablepopulation modelsofpopula-
tiondynamics. Whenthereareno environmental changes inducing period-
to-periodshiftsinmobility patterns,whenmortality is ignored,andwhen
thereis no agedifferentiationintransition
rates,thismodelreducestothe
standard Markovsocialmobility model.Thisergodic modelhasbeenused
to guidetheanalysis ofsocialtrends in suchdiverseaspectsofAmerican
societyas marriage,
family, andpopulation (Land& Felson1977),school-
ing(Felson& land 1978),mortality, morbidity and disability (Land &
McMillen1980,1981b),and crime(Cohenet al 1980).

SOME CURRENT RESEARCH PROBLEMS

Social Accounting
Systems
Contemporary American societyproduces vastquantitiesofstatistical
in-
formation aboutitscitizens, andactivities.
institutions, Accounting systems
attempt toorganize suchinformation so as toanalyzesocialconstancy and
change.NationalIncomeand ProductAccounts(NIPA), forexample,
organize monetary informationonmarket transactionswithina framework
suitableforanalyzingtheoperation ofthenational economic system.From
theoutsetofthesocial-indicators movement, socialscientists
andstatisti-
cianshaverecognized theinadequacies oftheNIPA forassessing broad
societaloutcomes, foridentifyingsocialproblems, formeasuring social
change,andforassessing policiesanddeveloping plansandprograms for
thefuture. Beginning withconceptual essaysbyGross(1966,1969),this
recognitionhasstimulated workon thedevelopment ofgeneralizedsocial
accounting systems thatincorporateinformation on socialconditionsnot
indexedintheNIPA. In theyearssincethepublication ofGross'sessays,
threetypesofaccounting unitshaveprovided thefoundation forimproved
ornewapproaches tosocialaccounting thatseemamenable toapplication
in social-indicators
research.
In net national welfareaccounting,based on a monetary
accountingunit,
variousimputations
andotherdevicesareusedto taketheexternalities
of
markettransactions
outofthecalculation
ofnational
incomeintheNIPA.
Thisprocedure
attemptstoobtaina better oftheimplications
estimate for
18 LAND

humanwelfare oftheactionsofthemarket economy. (Externalities arethe

costsandbenefits of individual and corporate economic activitiesthat are
experienced bythe larger society,such as the healthcosts of airpollution,
thatdo notenterintodecisions bytheindividuals or organizations under-
taking the activities.)For instance, Nordhaus & Tobin (1973)have incorpo-
ratedintotheGNP suchthingsas thebenefit of leisureand thecosts
ofpollution. While thisis a useful contribution towards making theNIPA
moreadequateaggregate measures ofeconomic welfare,therearelimits to
what the national accounts can include by means of such corrections.
A secondapproachis time-based accounting. Mostproposedsystems
seeka comprehensive accountoftheuseoftimebyall individuals inboth
market andnonmarket production andconsumption (Land& Juster 1981:
5-7). These resource-accounting systems, as indicatedabove,reston
householdproduction and consumption theory.By developing imputed
dollar-equivalents ofnonmarket household production and consumption
activities,thesesystems propose toexpandtheNIPA toincorporate mone-
taryindexesofbothmarket andnonmarket production ofeconomic wel-
fare.
Thisapproach canprovide valuableinformation aboutthelimitations of
inferences basedon theconventional NIPA. For instance, timespentat
workis apparently measured badlybyconventional dataonhoursworked.
Comparisons ofconventional measures ofweeklyhourswithtime-budget
diaryestimates indicate thattheconventional measure moreseverely over-
estimated timeactuallyspentat theworkplacein themid-1970s thanin
themid-1960s (Land & Juster1981:11). Thesedifferences mayexplain
about1% peryearofthewidelynotedproductivity slowdown during this
period.
Nonetheless, therearebothpractical andconceptual problems withthis
approach tosocialaccounting. Thepractical problems pertain toitsextrav-
agantdata demands:Somevariants(e.g. Fox & Ghosh 1981)seemto
requiresamplesizeson theorderof 100,000personsperyearin orderto
achievethestatistical reliabilityanddetailnecessary foroperationalization.
Conceptually, thehousehold production andconsumption theory embodied
in thesesystems recognizes causationonlyat thelevelofindividuals or
households. But manysocialphenomena are subjectto "compositional
effects" thatemergeas populations of individuals and organizations in-
teract.Thismeansthatadequatecausalexplanations ofaggregate social
conditions cannotbe extrapolated fromcausalexplanations developed at
thelevelofindividuals or households and viceversa(Firebaugh 1978).
Demographic accounting, inwhichtheaccounting unitis thenumber of
personsor organizations occupying socialstatesor categories at a given
time,is a thirdapproach tosocialaccounting. Developedinitially byStone
 SOCIAL INDICATORS 19

(1971) and subsequently relatedto thepopulation-dynamics and social-

mobility modelsofdemography andsociology byLand(1979),theempha-
sis in thistypeof accounting systemis on thespecification of various
sequences of"state-occupancy distributions"through whicha population
ofindividuals ororganizations flowsintime,andon theestimation ofthe
corresponding "transition regimes"thatgovern transitionfromonedistri-
butionto another. Whensuchaccountsare formulated froma "cohort"
ratherthana "period"perspective, theycan be linkedwithtime-based
socialaccountsthrough notions.
life-cycle Furthermore, Land (1979)has
shownhowthestatistical measures bywhichdescriptive socialindicators
usuallyare constructed (described above)can be embedded in a suitably
formulated system ofdemographic accounts.
Demographic accounting restsupondouble-entry accounting notions
requiring thattheflowsofpopulation intoa givengeographical areaand
period(births, immigrants, andsurvivors fromthepreceding period)must
exactly equaltheflowsfromthatareaandperiod(deaths,emigrants, and
survivors intothenextperiod).Thisrequirement can be usedto evaluate
theinternal and temporal consistency of data fromdiversecensusvital
and sample-survey
statistics sources.Usingsuchconcepts, forinstance,
Land & McMillen(1981a) concludedthatpublished annualpopulation
estimates fortheUnitedStatesweretemporally inconsistent.
Theybegan
to reformulate theseintoa consistent setofsingle-year-of-age-,sex-,and
race-specificestimates thatcanbe usedas denominators forratesandother
socialindicators.
Demographic accountsallowtheuse ofa variety oftheoretical frame-
works.For instance, in hiseffortsto linktheNIPA to demographic ac-
counts,Stone (1981) uses conventional input-output notions.But
sociologicalnotionsof institutional interdependence and population
ecologytheory mayalso be appliedin a demographic-accounting frame-
work(Land& McMillen198la). Furthermore, demographic accounts can
be used to operationalize themodelsof socialmobility and population
dynamics describedinthepreceding section (Land& Rogers1982).Inbrief,
demographic of
accountsseemcapable accommodating theheterogeneity
ofthe"social"insocialindicators andarethusmostdeserving ofadditional
research efforts.
and MeasurementIssues
Conceptualization
oftraditional
Fivecriticisms economic indicators
thatgrewoutoftheearly
programmatic of socialindicators
discussions werecitedearlierin this
Researchduring
chapter. thepasttwodecadeshasbeendifferentiallysuc-
inredressing
cessful theinadequaciesindicated.
Forinstance,methodologi-
cal workon satisfaction
indicatorshas obviatedthe equationof GNP
20 LAND

measures withpsychological satisfaction, happiness, orlifefulfillment; sur-

veyresearch instruments nowexistforthedirectmeasurement ofthelatter
concepts. Similarly,netnational welfare accounting research hasproduced
adjustments to theNIPA formarket externalities so thattheybetter indi-
catethecontributions ofmarket goodsand services to socialwell-being.
Finally,research ondescriptive socialindicators hasproduced newwaysof
conceptualizing, measuring, analyzing, and reporting nonmarket social
conditions and changestherein.
In spiteoftheseadvances, littleprogress hasbeenmadeontheconceptu-
alizationandmeasurement of(a) thesocialrules,values,norms, andstruc-
turesthatsupporttheeconomicsystem and (b) theinterdependence of
changesin theformer and changesin thelatter.One approachto this
problem attempts thedirect measurement ofvaluesandvaluesystems; some
theoretical and methodological workhas beendone(e.g. Rokeach1973,
1979),andsomeempirical evidence onshifts invaluesfrom themid-1950s
to 1971isreported intheDuncan,Schuman & Duncan(1973)Detroit Area
Surveyreplication. But efforts to collectdata on thevaluesand value
systems oftheAmerican population aredwarfed bythosedevotedto the
studyoflifesatisfaction andattitudes moregenerally. Thusthedatado not
nowexisttosupport studies ofshifting valuesystems thatwouldbecompa-
rableto,say,Smith's(1982)research on generalliberalism. In fact,social
scientistsnowknowlittlemoreabouttheinterdependencies ofeconomic,
social,anddemographic changes withchanges invaluesystems incontem-
porary American society thanwasknownat theoutsetofthesocial-indica-
torsmovement.
An alternative approach is to focuson theorganizations andstructures
thatmediatebetweenindividuals and thenation-state. Social processes
involve manylayersofsuchintermediate structures: tiesofmutualsupport
amongfriends andwithin families; flows ofinformation through kin,friend-
ship,andcollegial networks; linkages ofschooling organizations tohomes,
andfactories;
offices, thecontrol ofhumanservices bylocalgovernments,
etc.Sociologists haveproduced manystudiesofintermediate socialstruc-
turesandprocesses, butlittle progress hasbeenmadetowards conceptualiz-
ing and measuringtheir significant changes. Several theoriesin
organizational sociology merit exploration forpotential applicabilitytothis
task.Two illustrations mustsuffice.
Blau's (1977) theoryconcernshow variations in the inequality and
heterogeneity ofsocialstructures (conceived ofas multi-dimensional spaces
ofsocialpositions amongwhicha population is distributed)affect processes
ofsocialassociation (communications amongindividuals). Rogers(1971)
hasdocumented howsocialheterogeneity slowsthediffusion ofnewideas.
Notingthis,Lewis(1982)has arguedthattheweakening oftechnology-
 SOCIAL INDICATORS 21

basedeconomic growthin theUnitedStatesduringthepasttwodecades

maybe relatedto increasing socialheterogeneity.
Takentogether, these
studiessuggesta needforsystematic researchon thedimensions along
inAmerican
whichsocialheterogeneity societyhasincreasedordecreased,
on howthesechanges haveaffectedprocessesofsocialassociation,
andon
howthisrelatesto changes in theproductivity
oftheAmerican economy.
White's(1981a,b,c)theory concernshoworganizations produceinter-
faces:contracts,markets,employee-employer relationships,
negotiations,
understandings, etc.Lewis(1982)has arguedthatincreases
lawsuits, in a
certainclassof"interfaces"-namelylitigations-may havecontributedto
theUS productivityslowdown. Suchrelationships
and,moregenerally, the
causesand consequences ofchangesin attributesofinterfacesshouldbe
studiedsystematically.
Social Forecasting
Thesystematic useofsocialindicatorstoforecasttrendsinsocialconditions
and/orturning pointsintrends has so farnotrewarded earlierhopes.To
be sure,descriptivesocialindicators,
especially thosein traditional
demo-
graphicareas,continue to be used in conventional projection exercises
(Ascher,1978:Ch. 3); anda largefuturology literature
makesvarioususes
of socialindicatorsin thepreparation of projectionsand scenarios(e.g.
Snyder1973).Buttherehasbeenvirtually no systematic useoftheresults
ofanalyticalstudiesofsocialchangeandformal modelstogenerate out-of-
sample-period ofsocialconditions
(ex ante) forecasts (an exception is Co-
henet al 1980).
As a consequence, thistopicgreatly needsresearch attention,bothto
produce improved forecastsandtocontribute tothetheory andmethodol-
ogyofforecasting. Surelytheapplication of resultsfromanalytical and
forecastingstudiescan lead to moreabundant and higherqualityshort-
range(0-2 yr)andmedium-range (2-10 yr)socialforecasts, eventhough
longer-rangeforecastswillprobably remain speculativeatbest.In thelatter
case,systematic studymightlead to thedevelopment of "impossibility
theorems" thatdefine theaccuracylimitsofsocialforecasts.

CONCLUSION
Contemporary insocialindicators
interest hasbeensustainedamongsocial
fornearlytwodecades.Muchofthefirst
scientists decadewasdevotedto
theproductionofprogrammatic statementsandthesorting outofpromis-
ingresearchdirections.
Theseconddecadesawtheproduction of(a) repli-
newlongitudinal
cationstudies, datasets,andtimeseries,whichstimulated
(b) numerous studiesofsocialchange,socialreports,
analytical andwork
on formalmodelsofsocialchange.
22 LAND

Theresults ofthisactivityshouldnotbe underestimated. Substantively,

muchmoreis knownaboutsocialconditions and changesin American
societytodaythanwhenthestudythatledtotheBauer(1966)volumewas
commissioned. Perhapsevenmoreimportantly, twoprinciples forthecon-
ductofresearch on socialindicators havebeenestablished. First,without
carefulreplicationsofbaselinestudiesand/orregularly repeated observa-
tions,measurement andanalysisofsocialchangecan be neither scientific
norrelevant topolicy.Second,economies ofscaleandscientific
comparabil-
ityare facilitatedbythecreation ofcommondatabases.
Becausesocial-indicators researchdepends morethanmostsocialscience
researchoncooperation between thefederal government andtheacademic
researchcommunity, thereare,andprobably alwayswillbe,challenges to
thecontinued scientific
development ofthefield.
Foronething, theproduc-
tionofthemajornational datasetsbythefederal statistical
system-from
whichsocialindicators are derived-isonlypartlyunderthecontrolof
socialscientists.Furthermore, theallocationof publicfundsto produce
theseandprivate sectordatasets(suchas theNationalElection Studies and
theGeneralSocialSurvey) is subjecttolargerpolitical
forces.Thusfederal
budgetcutsmayremove muchofthestatistical infrastructureuponwhich
social-indicatorsresearchrests.
Thispossibility mightevokethecynical judgment thatthestudyofsocial
bornintheAmerican
indicators, society ofthe1960s,is notneededinthe
1980sandbeyond. An alternativeview,basedon a sociological assessment
ofthecriticalimportance ofinformation forbothpublicandprivate sector
is thatneedsforstatistical
decisions, dataon socialconditions andtrends
willnotdisappear. Ofcourse, therearenewissuesofpublicconcern inthe
1980s,and someissuesof criticalimportance in earlierperiodsare less
salienttoday,as Sheldon& Freeman(1970)anticipated. Buttheneedfor
information aboutcurrent conditions and trendsremains.
The discipline ofsociology shouldnotethepowerful influence sociolo-
gistshavehad on thedevelopment ofthefieldof socialindicators. The
descriptivesocialindicators perspective-emphasizing theproduction of
data,analyses, andmodelsbywhichtoimprove socialreporting-has been
oneofthemostsuccessful linesofdevelopment pursued overthepasttwo
decades.Thisapproach hasstimulated accumulation ofmuchnew,socially
and scientificallyusefulknowledge, onlya smallsampling ofwhichwas
reviewed here.Yet, as a profession, sociologyhas beenneither notably
cognizant of,norcommitted to,thedevelopment ofsocialindicators and
socialreporting. Neworganizational arrangements may,therefore, be re-
quiredinthefuture. Onlythecontinued commitment ofsociologiststothe
studyofsocialindicators willensurethattheachievements ofthepastare
consolidated andbuiltupon.
 SOCIAL INDICATORS 23

ACKNOWLEDGMENTS
Preparationof thispaperwas made possible,in part,by NationalScience
FoundationGrantSES-8104767.The writing ofthischapteralso was facili-
tatedby interaction, over severalyearsof service,withmembersof the
Social Science ResearchCouncil Advisoryand PlanningCommitteeon
Social Indicatorsand withmembersofthestaffoftheCouncil'sCenterfor
Coordinationof Researchon Social Indicators.In particular,I thankOtis
DudleyDuncan,Bill Mason,JohnModell,RobertParke,RobertPearson,
AlbertJ. Reiss, StephenH. Schneider,and Nancy Brandon Tuma for
commentsthatimprovedthemanuscript.

LiteratureCited
Adams, 0. B. 1980. Health and Economic
Activity.Ottawa: Statistics Canada.
71 pp.
Anderson,M. 1978. Welfare.Stanford, CA:
Hoover Inst. Press.251 pp.
Andrews,F. M., Inglehart,R. F. 1979. The
structureof subjectivewell-beingin
nine Westernsocieties.Soc. Ind. Res.
6:73-90
Andrews,F. M., Withey,S. B. 1976. Social
Indicators of Well-Being:Americans'
Perceptionsof Life Quality.NY: Ple-
num.455 pp.
Ascher,W. 1978. Forecasting: An Appraisal
for Policy-Makers and Planners.Balti-
more:JohnsHopkins.239 pp.
Bauer, R. A., ed. 1966. Social Indicators.
Cambridge,MA: MIT. 357 pp.
Bianchi,S. M. 1980.Racial differences
in per
capitaincome,1960-76:theimportance
of householdsize, headship,and labor
force participation.Demography17:
129-44
Biderman,A. D. 1966.Social indicatorsand
goals. In Social Indicators,ed. R. A.
Bauer, pp. 68-153. Cambridge,MA:
MIT. 357 pp.
Biderman,A. D. 1970. Information, intelli-
gence,enlightened publicpolicy:func-
tionsand organizationof societalfeed-
back. PolicySci. 1:217-30
Blau, P. M. 1977. Inequalityand Hetero-
geneity:A PrimitiveTheoryof Social
Structure. NY: The Free Press.307 pp.
Blau,P. M., Duncan,0. D. 1967. TheAmeri-
can OccupationalStructure.NY: Wi-
ley. 520 pp.
Brenner,H. 1976.Estimating thesocial costs
of nationaleconomicpolicy: implica-
tionsformentaland physicalhealthand
criminalaggression.Joint Economic
Committee,Congress of the United
States,Paper No. 5. WashingtonDC:
USGPO. 229 pp.
Campbell,A., Converse,P. E., eds. 1972. The
Human Meaning of Social Change.
NY: Russell Sage Found. 547 pp.
Campbell,A., Converse,P. E., Rodgers,W.
L. 1976. The QualityofAmericanLife:
Perceptions, Evaluations,and Satisfac-
tions.NY: RussellSage Found. 583 pp.
Caplow, T., Bahr, H. M. 1982. Middletown
Families: Fifty Years of Change and
Continuity. Minneapolis:Univ. Minne-
sota. 436 pp.
Carley,M. 1981.Social Measurement and So-
cial Indicators:IssuesofPolicyand The-
ory. London: George Allen & Unwin.
195 pp.
Carlisle,E. 1972.The conceptualstructure of
social indicators.In Social Indicators
and Social Policy,ed. A. Shonfield,S.
Shaw, pp. 23-32. London: Heinemann
EducationalBooks. 151 pp.
Cherlin,A. J. 1981. Marriage,Divorce,Re-
marriage.Cambridge,MA: Harvard.
142 pp.
Cohen,J.E., Singer,B. 1979.Malaria in Ni-
geria: constrained continuous-time
Markov models for discrete-time lon-
gitudinaldata on humanmixed-species
infections.Lect. Math. LifeSci. 12:69-
133
Cohen,L. E., Felson,M. 1979.On estimating
the social costs of national economic
policy: a critical examinationof the
Brennerstudy.Soc. Ind. Res. 6:251-60
Cohen,L. E., Felson,M., Land, K. C. 1980.
Propertycrimeratesin theU.S.: a mac-
rodynamicanalysis,1947-77, with ex
ante forecastsforthemid-1980s.Am.J.
Sociol. 86:90-118
Coleman,J. S. 1964. Introduction to Mathe-
maticalSociology.NY: The Free Press.
554 pp.
24 LAND

Coleman,J.S. 1968.The mathematical

study ships with educational trendsin the
ofchange.
InMethodology
inSocialRe- UnitedStates,1947-1974.In Research
search, ed. H. M. Blalock, Jr.,A. B. in PopulationEconomics,ed. J. L. Si-
Blalock, pp. 428-78. NY: McGraw- mon,pp. 93-126. Greenwich,CT: JAI.
Hill. 493 pp. 347 pp.
Coleman,J.S., Blum,Z. D., Sorenson,A. B., Ferriss,A. L. 1969. Indicatorsof Trendsin
Rossi, P. H. 1972a. White and black AmericanEducation.NY: RussellSage
careersduringthefirstdecade of labor Found. 454 pp.
forceexperience.Part I: occupational Ferriss,A. L. 1970. Indicatorsof Changein
status.Soc. Sci, Res. 1:243-70 theAmericanFamily.NY: RussellSage
Coleman,J. S., Berry,C. C., Blum, Z. D. Found. 145 pp.
1972b.Whiteand black careersduring Ferriss,A. L. 1971. Indicatorsof Trendsin
the firstdecade of labor forceexperi- theStatusof Women.NY: RussellSage
ence. Part II: occupationalstatusand Found. 451 pp.
incometogether.Soc. Sci, Res. 1:392- Fienberg,W. E., Mason,W. M. 1978.Identi-
404 ficationand estimationof age-period-
Davis, J. A. 1975.The log linearanalysisof cohortmodelsin theanalysisofdiscrete
surveyreplications. In Social Indicators archivaldata. Sociol. Methodol. 1979:
Models, ed. K. C. Land, S. Spilerman, 1-67
pp. 75-104. NY: Russell Sage Found. Firebaugh,G. 1978. Individual-levelrela-
411 pp. tionships from aggregate data. Am.
de Neufville,J.I. 1975.Social Indicatorsand Sociol. Rev. 43:457-72
Public Policy. Amsterdam:Elsevier Fox, K. A. 1974.Social Indicators
and Social
PublishingCo. 311 pp. Theory:Elementsofan Operational
Duncan, B. 1979. Change in worker/non- Sys-
tem. NY: Wiley-Interscience. 328 pp.
workerratiosforwomen.Demography
Fox, K. A., Ghosh,S. K. 1981. A behavior
16:535-48
Duncan, G. J.,Coe, R. D., Corcoran,M. E., setting approach to social accounts
Hill, M. S., Hoffman, S., Morgan,J.N. combiningconceptsand data fromeco-
1982.A Summary ofResearch logical psychology,economics, and
Findings studiesoftimeuse. In SocialAccounting
fromthePanelStudy ofIncome Dynam-
ics. Ann Arbor,MI: Inst. Soc. Res. Systems:Essayson theState of theArt,
Duncan, 0. D. 1969. Toward Social Re- ed. F. T. Juster,K. C. Land, pp. 132-
porting:Next Steps. NY: Russell Sage 217. NY: Academic.479 pp.
Found. 41 pp. Glatzer,W. 1981. International actorsin so-
Duncan,0. D. 1975a.Does moneybuysatis- cial indicatorsresearch.Soc. Ind. News-
faction?Soc. Ind.Res. 2:267-74 lett. 16:1, 8-12
Duncan, 0. D. 1975b. Measuring social Goodman,L. A. 1972a. A modifiedmultiple
changevia replication ofsurveys.In So- regressionapproachto the analysisof
cialIndicator Models,ed. K. C. Land, dichotomousvariables.Am.Sociol.Rev.
S. Spilerman, pp. 105-128.NY: Russell 37:28-46
Sage Found. 411 pp. Goodman,L. A. 1972b.A generalmodelfor
Duncan,0. D. 1982.Recqntcohortslead re- the analysisof surveys.Am. J. Sociol.
jection of sex typing.Sex Roles 8: 77:1035-86
127-33 Goodman, L. A. 1973. Causal analysis of
Duncan, 0. D., Schuman,H., Duncan, B. data frompanelstudiesand otherkinds
1973.SocialChangein a Metropolitan of surveys.Am. J. Sociol. 78:1135-91
Community. NY: RussellSage Found. Gross,B. M. 1966. The stateof the nation:
126 pp. social systemsaccounting.In Social In-
Easterlin,R. A. 1973.Does moneybuyhap- dicators,ed. R. A. Bauer,pp. 154-271.
piness?Publ. Int. 30:3-10 Cambridge,MA: MIT. 357 pp.
Featherman, D. L., Hauser,R. M. 1975.De- Gross,B. M., ed. 1969.SocialIntelligencefor
signfora replicatestudyof social mo- America'sFuture. Boston,MA: Allyn
bilityin theUnitedStates.In Social In- and Bacon. 541 pp.
dicatorModels,ed. K. C. Land, S. Spil- Hannan,M. T., Young,A. A. 1977.Estima-
erman,pp. 219-52. NY: Russell Sage tionin panelmodels:resultson pooling
Found. 411 pp. cross-sectionsand time series. Sociol.
Featherman, D. L., Hauser,R. M. 1978. Op- Methodol.1977:52-83
portunity and Change.NY: Academic. Headey,B. 1981. The qualityof lifein Aus-
572 pp. tralia.Soc. Ind. Res. 9:155-82
Felson,M., Land, K. C. 1978.Social,demo- Inter-University Consortiumfor Political
graphic and economic interrelation- and Social Research.1980-1981.Guide

to Resourcesand Services.Ann Arbor,
MI: Inst. Soc. Res. 389 pp.
Juster,F. T., Courant,P. N., Dow, G. K.
1981. The theoryand measurement of
well-being:a suggestedframework for
accountingand analysis.In Social Ac-
counting Systems:Essayson theStateof
theArt,ed. F. T. Juster,K. C. Land,pp.
23-94. NY: Academic.479 pp.
Land, K. C. 1971.On thedefinition ofsocial
indicators.Am. Sociol. 6:322-25
Land, K. C. 1975a. Social indicatormodels:
an overview.In SocialIndicatorModels,
ed. K. C. Land, S. Spilerman, pp. 5-36.
NY: Russell Sage Found. 411 pp.
Land, K. C. 1975b.Theories,modelsand in-
dicatorsof social change.Int. Soc. Sci.
J. 27:7-37
Land, K. C. 1979. Modelingmacro social
change.Sociol.Methodol.1980:219-77
Land, K. C., Felson, M. 1977. A dynamic
macrosocialindicatormodelofchanges
in marriage,family,and populationin
the U.S.: 1947-74. Soc. Sci. Res. 6:
328-62
Land, K. C., Juster,F. T. 1981. Social ac-
countingsystems:an overview.In So-
cial Accounting Systems:Essayson the
StateoftheArt,ed. F. T. Juster, K. C.
Land,pp. 1-22.NY: Academic.479 pp.
Land, K. C., McMillen,M. M. 1980.A mac-
rodynamicanalysisof changesin mor-
tality indexes in the U.S., 1946-75:
somepreliminary results.Soc. Ind. Res.
7:1-46
Land, K. C., McMillen,M. M. 198la. Demo-
graphicaccountsand thestudyofsocial
change,withapplicationsto the post-
WorldWar II UnitedStates.In Social
Accounting Systems:Essayson theState
oftheArt,ed. F. T. Juster, K. C. Land,
pp. 242-306. NY: Academic.479 pp.
Land, K. C., McMillen,M. M. 198lb. Deter-
minants of morbidityand disability
trendsin the U.S., 1958-77. Soc. Ind.
Res. 9:313-45
Land, K. C., Rogers,A., eds. 1982.Multidi-
mensionalMathematicalDemography.
NY: Academic
Lewis, J. D. 1982. Technology,enterprise,
and Americaneconomicgrowth.Sci-
ence 215:1204-11
Liu, B. C. 1976. QualityofLifeIndicatorsin
US. MetropolitanAreas: Statistical
Analysis.NY: Praeger.315 pp.
Lynd,R. S., Lynd,H. M. 1937.Middletown
in Transition.NY: Harcourt,Brace &
World.604 pp.
Mason,K. L., Mason,W. M., Winsborough,
H. H., Poole, W. K. 1973.Some meth-
odologicalproblemsin cohortanalysis
SOCIAL INDICATORS 25
of archival data. Am. Sociol. Rev.
38:242-57
Mason, W. M., Smith,H. L. 1981.Age-peri-
od-cohortanalysis and the study of
deaths from pulmonarytuberculosis.
Revisionofpaperpresented at Conf.on
AnalyzingLongitudinalData forAge,
Periodand CohortEffects, June,1979,
Snowmass,CO.
Michalos,A. C. 1980.Satisfaction and happi-
ness. Soc. Ind. Res. 8:385-422
Moore,W. E., Sheldon,E. B. 1965.Monitor-
ingsocialchange:a conceptualand pro-
grammaticstatement.Proc. Soc. Stat.
Sec. Am. Stat. Assoc.,pp. 144 49
Morgan,J. N., Smith,J. D. 1969. A Panel
StudyofIncomeDynamics.AnnArbor,
MI.: Inst. Soc. Res. 344 pp.
NationalCommissionon Technology,Auto-
mation,and EconomicProgress.1966.
Technology and theAmerican Economy,
Vol. 1. WashingtonDC: USGPO. 115
PP.
Nordhaus,W. D., Tobin,J. 1973. Is growth
obsolete?In TheMeasurement ofSocial
and Economic Performance,ed. M.
Moss, pp. 509-31. NY: Nati.Bur.Econ.
Res. 605 pp.
Ogburn,W. F. 1929.The responsibility ofthe
social sciences. In Recent -Social
Changesin the UnitedStatessince the
Warand Particularly in 1927,ed. W. F.
Ogburn.Reprinted(1964) in WilliamF
Ogburnon Cultureand Social Change,
ed. O. D. Duncan,pp. 99-101. Chicago:
Univ. Chicago. 360 pp.
Parke,R., Seidman,D. 1978. Social indica-
tors and social reporting.Ann. Am.
Acad. Pol. Soc. Sci. 435:1-22
Parnes, H. S. 1972. Longitudinalsurveys:
prospectsand problems.Monthly Labor
Rev. (Feb.):11-15
President'sResearch Committeeon Social
Trends. 1933. RecentSocial Trendsin
the UnitedStates. NY: McGraw-Hill.
1568 pp.
Pullum,W. T. 1980. Separatingage, period,
and cohorteffects in whiteU.S. fertility,
1920-1970.Soc. Sci. Res. 9:225-44
Rogers,E. M., Shoemaker,F. F. 1971. Com-
municationof Innovations.NY: Free
Press.476 pp.
Rokeach, M. 1973. The Nature of Human
Values. NY: Free Press.438 pp.
Rokeach, M. 1979. Understanding Human
Values: Individualand Societal. NY:
Free Press.322 pp.
Sewell,W. H., Hauser, R. M. 1975. Educa-
tion, Occupation, and Earnings:
Achievement in theEarly Career. NY:
Academic.237 pp.
26 LAND

Sheldon,E. B. 1971.Social reporting

forthe Am. Acad. Po. Soc. Sci. 453 (Jan.)308
1970s.In FederalStatistics:
Reportof PP.
Commission,
thePresident's pp. 403- Taeuber,R. C., Rockwell,R. C. 1982. Na-
35. WashingtonDC: USGPO. 555 pp. tional Social Data Series. A Compen-
Sheldon,E. B., Freeman,H. E. 1970. Notes diumofBriefDescriptions. Washington
on social indicators:promisesand po- DC: Soc. Sci. Res. Counc. 148 pp.
tential.PolicySci. 1:977-111 Terleckyj,N. E. 1975. Improvements in the
Sheldon,E. B., Land, K. C. 1972.Social re- QualityofLife:EstimatesofPossibilities
portingforthe 1970s:a reviewand pro- intheUnited States,1974-1983.Washing-
grammaticstatement.Policy Sci. 3: tonDC: Nati.Plan. Assoc. 285 pp.
137-51 Terleckyj,N. E. 1981. A Social production
Sheldon,E. B., Moore,W. E., eds. 1968.In- framework forresourceaccounting.In
dicators ofSocialChange: Concepts and Social AccountingSystems.Essays on
Measurements. NY: Russell Sage theStateoftheArt,ed. F. T. Juster, K.
C. Land, pp. 95-131. NY: Academic.
Found. 822 pp. 479
Sheldon,E. B., Parke,R. 1975.Social indica- Treiman,pp.D. J., Terrell,K. 1975. Women,
tors.Science 188:693-99 work,and wages-trendsin thefemale
Singer,B. 1981.Estimationofnonstationary occupationalstructuresince 1940. In
Markovchainsfrompaneldata. Sociol. Social Indicator Models, ed. K. C.
Methodol.1981:319-37 Land, S. Spilerman,pp. 157-200. NY:
Singer,B., Cohen,J.E. 1980.Estimating ma- Russell Sage Found. 411 pp.
laria incidenceand recoveryratesfrom Tuma,N. B., Hannan,M. T., Groeneveld,L.
panel surveys. Math. Biosci. 49:273- P. 1979.Dynamicanalysisofeventhis-
305 tories.Am. J. Sociol. 84:820-54
Singer,B., Spilerman,S. 1974.Social mobil- United Nations StatisticalOffice.1975. To-
ity models for heterogeneouspopula- wardsa Systemof Social and Demo-
tions. Sociol. Methodol. 1973-74:356- graphicStatistics.NY: UnitedNations.
401 187 pp.
Singer,B., Spilerman,S. 1976a. Some meth- U.S. Departmentof Commerce.1974.Social
odologicalissuesin theanalysisof lon- Indicators, 1973. Washington DC:
gitudinal surveys. Ann. Econ. Soc. USGPO. 258 pp.
Meas. 5:447-74 U.S. Departmentof Commerce.1978.Social
Singer,B., Spilerman,S. 1976b.The repre- Indicators, 1977. Washington DC:
sentationofsocialprocessesbyMarkov USGPO. 564 pp.
models.Am. J. Sociol. 82:1-54 U.S. Department of Commerce.1980. Social
Smith,T. W. 1980.A Compendium ofTrends Indicators, III. Washington DC:
on GeneralSocial SurveyQuestions. USGPO. 585 pp.
Chicago:Natl. Opin.Res. Cent.260 pp. U.S. Departmentof Health,Education,and
Smith,T. W. 1982. Generalliberalismand Welfare.1969. Towarda Social Report.
social change in post-WorldWar II WashingtonDC: USGPO. 101 pp.
America: a summaryof trends.Soc. White,H. C. 1981a. Interfaces. Presentedat
Ind.Res. 10:1-28 Conf. on Indicatorsof Organizational
Snyder,G. S. 1973. 1994: The Worldof To- Change,December,1981,Washington,
morrow.Washington DC: U.S. News & D.C.
World Report,Inc. 191 pp. White,H. C. 1981b.Productionmarketsas
induced role structures.Sociol. Me-
Stone, R. 1971.Demographic Accounting and thodol.1981:1-57
Model Building.Paris:Org. Econ. Co- White,H. C. 1981c. Wheredo marketscome
op. Devel. 125 pp. from?Am. J. Sociol. 87:517-47
Stone, R. 1981. The relationshipof demo- Wilcox,L. D., Brooks,R. M., Beal, G. M.,
graphic accounts to national income Klonglan,G. E. 1972.Social Indicators
and product accounts. In Social Ac- and Societal Monitoring.San Fran-
counting Systems:EssaysontheStateof cisco: Jossey-Bass.464 pp.
theArt,ed. F. T. Juster,
K. C. Land,pp. Winsborough, H. H. 1975. Age, period,co-
307-376. NY: Academic.479 pp. hort,and educationeffects on earnings
Stouffer,S. A. 1955. Communism,Confor- byrace-an experiment witha sequence
mity, and CivilLiberties.
Garden City, surveys.In Social In-
of cross-sectional
NY: Doubleday. 178 pp. dicatorModels,ed. K. C. Land, S. Spil-
Taeuber,C., ed. 1978.Americain theseven- erman,pp. 201-218. NY: Russell Sage
ties: some social indicators.Ann. Am. Found. 411 pp.
Acad. Pol. Soc. Sci 435 (Jan.) 355 pp. Zapf,W. 1976.International, public,and pri-
Taeuber, C., ed. 1981. America entersthe Soc. Ind
vateactorsin social reporting.
eighties:some social indicators.Ann. Newslett.10:4-5