Campbell - Self-Interest, Social Security, And the Participation Patterns of Senior Citizens

7/29/2019 Campbell - Self-Interest, Social Security, And the Participation Patterns of Senior Citizens

1/11

Self-Interest, Social Security, and the Distinctive Participation Patterns of Senior Citizens

Author(s): Andrea Louise CampbellReviewed work(s):Source: The American Political Science Review, Vol. 96, No. 3 (Sep., 2002), pp. 565-574Published by: American Political Science AssociationStable URL: http://www.jstor.org/stable/3117930 .

Accessed: 26/08/2012 23:24

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 ofcontent 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].

.

American Political Science Association is collaborating with JSTOR to digitize, preserve and extend access to

The American Political Science Review.
http://www.jstor.org/action/showPublisher?publisherCode=apsahttp://www.jstor.org/stable/3117930?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/3117930?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=apsa


2/11

American Political Science Review Vol. 96, No. 3 September2002

Self-Interest, S o c i a l S e c u r i t y , a n d t h e D i s t i n c t i v e ParticipationP a t t e r n s o f S e n i o r C i t i z e n sANDREA LOUISE CAMPBELL Harvard University

ecadesof participation esearch howthatpoliticalactivity ncreaseswith ncome,but thepar-ticipationof senior citizensspecificallywith regard o Social Securityposes an exception othispattern.Social Security-orientedarticipationdecreasesas incomerises, n part becauselower-incomeeniorsare moredependent n theprogram.Thenegativencome-participationradientsespeciallypronounced or letterwritingabouttheprogram,but evenSocialSecurity-relatedotingandcontributing re lesscommonamong higher-incomeeniors.This s an instance n whichself-interestshighly influential:Those who aremoredependent re more active.It is also an exampleof lower-classmobilizationwithregard o an economic ssue,somethingquiteunusual n the UnitedStates.

t is well established that political participation in-creases with income, even for activitiesin whichincome seems irrelevant, uch as voting,or coun-terintuitive, uch as protesting (e.g., Rosenstone andHansen1993;Verbaand Nie 1972;Verba,Schlozman,andBrady1995).Verba,Schlozman, ndBrady(1995)suggest several reasonsfor this pattern.Income is apoliticallyrelevantandfungibleresourcethatdirectlyor indirectly acilitatesactivity; he affluentare morehighlyengagedwithpublicaffairs, sthey typicallyhavemore at stakein a varietyof policyareas;andhigher-income ndividuals remore requently ecruitedopo-liticalactivity. n the United States,which acksstronglaborunionsandsocialistparties,ower-income itizensareless mobilized,especiallyon economic ssues.Political activity concerningSocial Security may,however,be different.Social Security s not only aneconomic ssuebutalsoagovernmentprogram.Ratherthanpersonalizeor "morselize"heir fate as in manyotherdomains, itizenscanattribute heirwell-being ogovernment ction Lane1962,353; ee SearsandFunk1990).Moreover,politicalactivity s seen as an appro-priatevehicleforexpressingdissatisfaction r concernabout the policy and its operation.Because the pro-gramis self-contributory, o welfarestigmaor moralbardiscouragespoliticalactivismon the issue,as is thecase withothergovernmentprogramsbenefiting ow-income individuals.Moreimportantly, ocialSecurityis unlikemany ssuesin that the stakesaregreater orlower-incomerecipients,who derive a larger portionof theirincome fromthe program.The usual income-interest gradientis reversed.Accordingly,participa-tion based on Social Securitymay be greateramonglower-income hanamonghigher-incomeeniors,a re-versal of the typically positive income-participationrelationship.

Poorerseniorsmayseemincapableof participatingat highrates,giventheir ow resources.However,sev-eral characteristicsf seniorsand theirprogramsmaketheir unusualactivityconcerningSocialSecuritypos-sible. Seniors'resourcesareoffsetin a varietyof waysandlessdetrimental o theirparticipationhanlow re-sources are to the participationof nonseniors.Also,the political parties mobilize low-income seniors atthe samerate ashigher-incomeeniors,andmoderate-incomeeldersat evenhigherrates,againdifferent romthe usual pattern,where mobilization ncreases withincome.A negativerelationshipbetween ncomeandpartic-ipatory activityconcerningSocial Securitywould beimportant n several regards.It would representanexceptionto a long-standing egularity n behavioralresearch-that politicalparticipations morecommonamongthe affluent. t would be a rareexampleof self-interestexertinga significant nfluence on individualbehavior,since the most dependentare hypothesizedto be the most active.Finally, his case would add an-otherexample o thesmallbutgrowingiterature ntheeffects of policyon thepoliticalbehaviorandattitudesof masspublics(e.g.,Mettler1998;Mettlerand Welch2001;Soss 1999).SocialSecuritymaydemocratize e-niorparticipation yraising heactivity evelsof lower-incomeseniorsrelative to the affluent or the portionof seniors'participationhatconcernsSocialSecurity.INCOMEANDPARTICIPATIONNAMERICADecades of researchamply llustrate he higherpoliti-cal participation ates of the affluentand offerseveralexplanations or the traditionally ositiverelationshipbetween ncomeandpoliticalactivity. erhaps hemostobvious s that ncome s anecessary esource orsomeactivities, uchasmaking ampaign ontributions. venbeyond that, political participation s a luxury goodmore readilyavailable o higher-income itizens.Theaffluentare not preoccupiedwith securing he neces-sities of daily life and can purchase he labor of oth-ers such as gardenersand babysittersto free theirtime (e.g.,Conway1991, 25;Milbrathand Goel 1977,98;Rosenstone 1982).They can "simplyaffordto domore-of everything-than citizenswith ittlemoney,"including political participation (Rosenstone andHansen1993,12).

Andrea Louise Campbell is Assistant Professor of Government,Harvard University, 321 Littauer Center, Cambridge, MA 02138([email protected]).Thanks go to Steve Ansolabehere, Henry Brady, Barry Burden,and seminar participants at MIT and Yale University for their com-ments on an earlier draft. The data analysis was carried out usingSPSS 9.0. The data used in this study were obtained from the Inter-University Consortium for Political and Social Research, the Centerfor Political Studies at the University of Michigan, and the RoperCenter forPublic Opinion Research at the University of Connecticut,Storrs.565


3/11

Self-Interest,Social Security,and the Distinctive ParticipationPatterns of Seniors September2002Efforts to recruit citizens to political activity alsotend to focus on the affluent. In Europe, socialistparties and labor unions mobilize the lower class, insome countries to levels of political activity rivalingthose of the middle class (Powell 1986; Verba, Nie, andKim 1978). In the United States, in contrast, no labor-based political party emerged in the early twentiethcentury (Skocpol 1995, 26), and labor unions have

been relatively weak. Contemporary mobilizationtechniques such as direct mail and door-to-doorcanvassing are directed predominantly toward high-status individuals and neighborhoods (Rosenstoneand Hansen 1993). Low-income Americans are rarelymobilized to political activity, except by religiousinstitutions, which typically focus on moral rather thaneconomic issues (Verba, Schlozman, and Brady 1995).Finally, ncome affects participation by enhancing in-dividuals' engagement with politics and sense of stakein the system. Higher-income people have a greaterstake in many policy areas.Perhaps the best example istax policy, where those with high incomes have more towin or lose. Affluent homeowners, for example, weremore likely than other Californians to support Propo-sition 13, the 1978 property tax rollback initiative, be-cause of the large tax savings the measure promisedthem (Sears and Citrin 1982). High-income individualsare also more likely to perceive their great stake; theirpolitical interest is stimulated by their high-level jobs,in which they learn how government affects their busi-ness and personal lives (Conway 1991). Additionally,they belong to voluntary organizations at high rates.These institutional affiliations provide civic skills, re-cruitment opportunities, and proximity to other activeindividuals who both share information about issuesand enforce participatory norms (Verba, Schlozman,and Brady 1995).

Thus, resource, engagement, and mobilization fac-tors all contribute to the higher participation rates ofthe affluent. The positive relationship between incomeand political participation is so pervasive that it appearseven in explanations of political activity among disad-vantaged groups. Black and Latino political activists,for example, have higher incomes and are less likelyto receive government assistance than the groups fromwhich they are drawn (Verba, Schlozman, and Brady1995, chap. 8). The resource mobilization theory ofsocial movements argues that groups with higher lev-els of resources are more likely to overcome collec-tive action problems and engage in political activity(McCarthy and Zald 1977). Next to education, incomeis the political resource most important in explainingboth conventional and unconventional activity.SOCIALSECURITYAND SENIORPARTICIPATIONFor most participatory activities, income operates inthe same way for senior citizens: Affluent seniors aremore politically active (Campbell 2000). However, inthe context of Social Security, income plays a differentrole-it is an indicator not of "resources" but of"dependence on government transfers." As shown inFigure 1, poorer seniors derive a higher portion oftheir income from Social Security.Seniors in the lowesttwo quintiles get over 80% of their income from theprogram, compared to just 18% for the highest quintile.Because the stakes in this issue area are greater forlower-income seniors,they should exhibit more interestin Social Securitythan their more affluent counterparts,and they should be more likely to engage in SocialSecurity-based participation. In this issue domainincome measures dependence rather thanparticipatory

FIGURE. Percentageof SeniorIncome romSocial Securityby Total ncomeQuintile,1992100%

S90%-.

82% 81%80% -

670%- 64%

6 0 %

w) 50% - 45%E00 40%30%0-0 18%20%10%.0%

5th (lowest) 4th 3rd 2nd 1st (highest)SeniorIncomeQuintile

Source: OlderAmericans 000: KeyIndicators f Well-BeingFederalInteragencyForum n AgingRelatedStatistics2000, 66).

566


4/11

AmericanPoliticalScienceReview Vol.96,No.3capacity, ndtheusuallypositiverelationships etweenincome and political interest and participationarereversed.Not allpoliticalactivitieswouldbe equallyaffected.A negativeincome-participation elationship s mostlikely for letterwritingabout SocialSecurity.Citizen-initiated,issue-specific,and independentof the elec-toralcycle,letter writing s the chief meansby whichseniors can voice their concernsaboutSocialSecurity.Thisactivitymight ncludeparticularizedaseworkre-queststo solve individualproblems,moregeneralizedstatementsof policyconcern,orobjections o deleteri-ous reformproposals. n any of these instances, etterwritingis the participatory ehicle by which specificissueconcernscan be raised,andlower-income eniorsshouldbe morelikelyto write suchletters.There may be spillover effects on voting andcontributingas well. For these acts, however, theincome-participationrelationshipis unlikely to bemonotonically negative, as is the case with SocialSecurity-basedontacting.Contactings a "purer" ct:Letters are usually about a specifictopic, so surveyrespondents'assertionsthat they wrote about SocialSecurityareprobablyaccurate. ncontrast,ndividualsvote and contributefor a varietyof reasons,and soSocialSecurity-based otingand contributingmaybe"contaminated"by other issues. Since the income-participationrelationshiparound these other issuesis positive, the relationship between income andreportedSocial Security-basedparticipationwill notbe monotonicallynegative.In addition,contributingrequiresdisposableincome, suggestinga curvilinearrelationshipfor that activity,with moderate-incomeseniorswho areconcernedaboutSocialSecuritybeingmore likely to contributewith the program n mindthaneither ow-income eniorswho cannotafford o orhigh-income eniorswho are notinterested ndoingso.DATAANDMETHODSThreedatasets arenecessary o test these hypothesesaboutseniors' nterest n SocialSecurityandtheirpar-ticipation oncerningheprogram ndto describe heirpatternofpoliticalpartymobilization, s nosingledataset containsall of these dependentvariables.RoperSurvey8108(RoperOrganization 981)askedrespon-dentshow closelythey follownews of varioustopics.1Iusethisdatasetto assessseniors' nterest nSocialSe-curitycompared o otherdomesticand nternationals-sues.The1990CitizenParticipation tudy Verbaet al.1995)provides he SocialSecurity-based articipationitems.Program ecipientswere askedwhether heyhadevercontactedanelected official ocomplainaboutSo-cialSecuritypayments,aken ntoaccountacandidate'spositionon SocialSecurity n decidinghow to vote,ormade a campaign ontribution ased atleast inpartonconcernaboutSocialSecuritypayments.Finally,partymobilizationsmeasured nthe 1996NationalElectionStudy,where respondentswere askedwhethersome-

one from a politicalpartycontactedthem duringtheelectionseason.TheCitizenParticipation tudy temsmeasuringSo-cialSecurity-based articipation equirea closerlook.Not only do the activities themselvesdiffer in their"purity,"s notedabove,but alsothesurvey temsmea-suringSocialSecurity-basedparticipation aryin thelocus of self-interest.With SocialSecurity-based on-tacting, he self-interest s in the motivation o writealetter;a seniorcitizendependson the program, xpe-riencesa problemwith benefitsor feels concernabouta proposedpolicy change,and choosesto writea letterto an electedofficial o requestassistanceor to expressthe policy concern.The hypothesizedrelationship sbetween ncome anda particular indof participation,and this participation s higher among lower-incomeseniors.SocialSecurity-based otingandcontributingare a bit different.The question wordingfor SocialSecurity-based otingplacestheself-interestnot inthemotivation o vote, but in vote choice. The item mea-sures(1) whethera respondentvoted and(2) whetherthe respondentfelt that Social Securityfiguredintothe vote choice. Social Security-based ontributingsalso a dependentvariable with two parts:the act ofcontributing ndtheattribution f themotiveto SocialSecurity.Thus orvotingandcontributingn thecontextof SocialSecurity, he relationship s not between in-comeandacertainkindof participation, s withSocialSecurity-based ontacting,but betweenincome and apropensityoattribute heprimarymotivation oSocialSecurity.However,while thesedistinctionsareimpor-tant onote,affluent eniors houldnonethelessbeleastlikely to engagein these acts,whether the dependentvariable s contactingwith Social Security n mind orattributing vote decisionor politicalcontribution oconcernwith the program.The relationshipbetweenincome andeachtype of SocialSecurity-based artic-ipationshouldbe different romthatbetweenincomeand non-SocialSecurity-based articipation.In each analysis, he dependentvariable s modeledas a functionof income,education, ender,race,mar-italstatus,workstatus,and age (whereavailable).In-come is representedcurvilinearily y the inclusionofa squared ermexcept where tests for multicollinear-ity indicated hatincome-squaredwasredundantwithincome.Themodelsareestimatedwithordinary eastsquares OLS)or,where the outcome s dichotomous,logisticregression."Seniors" rerespondentsaged65andover,and"nonseniors" reages 18 to 64.SENIORS' PARTICIPATORYCAPACITYBefore testing the hypotheses about the greaterpropensityof lower-incomeseniors to be interestedin Social Securityand to engage in Social Security-basedparticipation,t is necessary o discusswhetherseniors,particularly he less affluent,have sufficientmeansto participate.Resourcessuch as educationandincomeareamong hestrongestpredictors fparticipa-toryactivity Verba,Schlozman,andBrady 1995),yetseniorcitizenshave the lowest formaleducationandhousehold income of any age group. One-third of

1 This is the most recent Roper survey available assessing So-cial Security-specific interest. See Appendix A for the news itemsincluded.

567


5/11

Self-Interest,Social Security,and the DistinctiveParticipationPatterns of Seniors September2002seniorsin the CitizenParticipationStudylack a highschooldiploma,compared o only 9%of respondentsaged35 to 49. Theaverage amily ncome for seniors sjustover$24,000, omparedo almost$46,000 or thoseaged35 to 49.Despite their low resourcelevels, however,seniorcitizensin the United States participateat very highrates. nthe 1996NationalElectionStudy, eniorswere7% more likely to vote than 35 to 64 year olds, and27%morelikelyto vote thanpersonsunder35.Twelvepercentof seniorsmadecampaign ontributions,om-paredto 9% of the middleage groupandjust 3% ofthe youngestgroup.Seniors'activeparticipation uggests hat theirmod-est resource evels are counterbalancednseveralways.Withretirement, ne tradesone politically elevantre-source, ncome,for another, ree time. Seniorsreportover 12 hours of free time per day,compared o fivehours for respondentsaged 35 to 49.2Also, seniors'incomes"gofurther"hanthose of nonseniors.Seniorsdo spend a greatershare of their income on healthcare-12% of their before-taxincome, comparedto5%amongall households n 1997-but spend ess thannonseniors,both in absolute terms and as a percent-ageof income, n categories uchasentertainment, p-parel, ransportation,nd,mostnotably,personal axesand mortgagecosts (U.S.Bureauof the Census1999,471). Seniors also hold less nonmortgageconsumerdebt (Duensing1988). Theybenefit from a varietyoffinancialbreaksthatareconferredon the basis of agerather hana means test:Theyareexemptfrom SocialSecurityand Medicare axes on earnings; hey get anextradeductionontheir ederal ncometaxes; heypaytaxeson a smallershare of theirincomes;andtheyre-ceive discountsatmanyretailstores, estaurants,otels,and so on.3The mean seniorfamily ncome sjust53%that of 35 to 49 year olds in the CitizenParticipationStudy,but adjusting or householdsize bringsseniors'percapita ncometo 88% thatof the youngergroup.Thatseniorshavelow education evelsmaybe evenless relevantfor participation han their low incomessince educationrelative to one's cohort is what mat-ters for manyparticipatory ctivities(Nie, Junn,andStehlik-Barry1996). But even seniors' low absolutelevels of formaleducation are offset in severalways.Tosome extent,life experiencemayserveas a substi-tute for formal education(Wolfinger ndRosenstone1980).Over a lifetime one may absorbparticipatorynorms and values, learn to process political informa-tion, and become accustomed to voting through sheerhabit (Plutzer 2002; Weisberg and Grofman 1981); along life in the "school of hard knocks" may impartthe same experience with bureaucratic relationships

that helps educatedpeople overcome the proceduralhurdlesnecessary o registerandvote (Wolfinger ndRosenstone1980,35-36,60).Theremayalso be cohortdifferences n education content and quality.Earliercohortsmay have been more steeped in democratic,"American" alues n school(Rosenstoneand Hansen1993).Perhapsgreatereducationqualityexplainswhyseniorsscore ashighas the moreeducated35 to 49yearolds on the Citizen ParticipationStudy vocabularytest.4Lack of formaleducationmayalso be offset byvoluntarygroupand churchmemberships. eniorsbe-longto thesegroupsat the sameratesas,orhigherratesthan,youngerpeopleandpracticemanypolitically el-evant skillssuch as letterwritingandspeech making.5Thus ow resourcesdo notrepresent he samebartoseniorpoliticalactivityas to nonsenioractivity.Manyseniors, ven the lessaffluent,have sufficient esourcesto actontheirgreat nterest nSocialSecurity,owhichI now turn.RESULTSPoliticalInterestInRoperSurvey8108,respondentswereaskedwhethertheyfollow news of variousdomesticand nternationalissues closely,follow it casually,or pay no attention.Eighty-onepercentofseniors aid hat hey ollownewsof SocialSecurityclosely,compared o 54% of nonse-niors. Seniors followed news of Social Securitymorecloselythananyothertopic.Indeed,the highest evelof interestshownby any age groupin any issue wasseniors' nterest n SocialSecurity.6To examine how the dependence gradient influ-ences interest n SocialSecurity, modeledinterest n

2 In the Citizen Participation Study,free time is defined as time notspent working, doing housework, studying, or sleeping. The largestdifference between seniors and nonseniors is time spent working.Retired senior men have even more free time, 14 hours per day.3 The median senior couple with an income of $25,000 would owe$294 in federal taxes,or 1.15%,whereas the median nonsenior couplewith an income of $45,000 would owe $5,164, or 11.5%. On an in-come of $40,000, seniors would pay $2,664, or 6.6%, in federal taxes,compared to $4,414 or 11.0% for nonseniors (Moon and Mulvey1996, 25).

4 On a 10-item test, seniors got 6.3 definitions correct on average,compared to 6.6 for the younger group.5 In the Citizen ParticipationStudy, respondents were asked whetherthey had engaged in the following activities in the last six monthsas part of their job, organizational involvements, or church activ-ities: written a letter, come to a meeting where they took part inmaking decisions, planned or chaired a meeting, or given a presen-tation or speech. The job skill questions were asked of those whowere employed. The church skill questions were asked of those whowere active members of their churches or who had served on theirchurch board in the past five years. The organization skill questionswere asked of respondents' "main"organization, and only skills ex-ercised in nonpolitical organizations are analyzed here (see Verba,Schlozman, and Brady 1995, Appendix B.9, for furtherexplanationsof the skills variables). Seniors are just as likely to belong to nonpo-litical organizations as 35 to 49 year olds (74 to 72%) and practicethe same number of skills there, 1.3 of a possible four. Seniors areeven more likely to be members of a religious institution or to attendreligious services two or more times per month-79%, vs. 67% of 35to 49 year olds-and, again, practice as many skills, 1.6 on average.Only 10% of seniors work, and the survey did not ask about skillsacquired on the job before retirement.6 Senior interest in Social Securityremains strong contemporarily.Inthe 1998 midterm election exit poll administered by the Voter NewsService, seniors were most likely to cite Social Security as the issuethat mattered most in deciding their vote for the House of Repre-sentatives (Toner 1999, A10). Thirty-one percent of seniors choseSocial Security as the issue that mattered most, followed by "moraland ethical standards" (20%), education (14%), taxes (11%), theeconomy/jobs (10%), health care (8%), and the Clinton/Lewinskymatter (6%). The "most important problem" mentioned most fre-quently among seniors in the 2000 NES was Social Security;amongnonseniors, it was education.

568


6/11

AmericanPolitical cienceReview Vol.96,No.3

FIGURE. Predicted nterestnSocial Security,Domestic,and Internationalssues amongSeniorsbyIncome1

0.9M0.8

0.70.6 - - t

S0.5."00.4.o 0.3o - Social SecurityS0.2 .. Domestic

0 - - -- IntemrnationalS0-0 10 20 30 40 50

Income 000s)Source:RoperSurvey8108 (RoperOrganization 981).Note: InterestnSocialSecurityestimatedwith ogisticregression; nterest ndomesticand internationalssues, withOLS.

domestic topics, international opics, and Social Se-curityas a function of income and other covariates.7I scored interestin SocialSecurityas a dichotomousvariable-one forthosefollowingnews of theprogramclosely,zero for those followingit casuallyor not atall-and estimatedthe modelwithlogistic regression.I constructed calesfor interest n domesticand inter-national issues by addingthe number of issues eachrespondent ollowedclosely (zero to five),whichI re-codedto zero to one andestimatedwithOLS.8Figure2 shows the effectof income on issueinterestamongseniorswith the other covariatesheld to theirmeans.Interest ndomesticand nternationalssues n-creaseswithincome,the traditionalpattern.Incontrast,nterest nSocialSecurity allsathigh n-come levels.Strictly, predictedamonotonicallynega-tiveincome-interest elationship,whiletheresultshereare curvilinear. t could be that the incomevariable nthe Roper datapicksup some of the (positive)effectsof education even when educationis controlled for.

Alternatively, t maybe that self-interest s a weakerinfluenceon attitudes hanparticipation,or actingonan issue s morecostly hanvoicinganopinionon asur-vey,andso typicallyonlytheself-interestedact(Citrinand Green1990).In anycase,even with issueinterest,SocialSecurityhas a differentpatternvis-a-vis ncomethan other domestic and international ssues. Theseeffects are evenmorepronounced n participation.Political ParticipationTable1 showsthe threetypesof SocialSecurity-basedparticipationmodeledas afunctionof income, ncome-squared,and other covariates.As hypothesized, n-come has a negative nfluenceon contactingbased onSocialSecurity,heoppositeofits usualsign(column2).The relativelysmall numberof casespreventsthe co-efficientfor income from achievingstatisticalsignifi-canceitself,but a log likelihood est showsthat ncomeandincome-squared rejointlysignificant p < .001).9Figure3a illustrates he probabilityby income that asenior citizenwillwritea lettercomplaining boutSo-cialSecurity, nd it fallsprecipitously,n greatcontrastto thetypicallypositive ncome-participationradient.As predicted,the resultsfor Social Security-basedvoting and contributingare somewhatdifferent.Thelikelihood of voting on the basis of Social Securityfirst rises with income and then decreasesat higherincomelevels (Figure3b). If the SocialSecurityvotingitemtrulymeasuredonlySocialSecurity-orientedot-ing andwasnot contaminatedby otherissues (peoplevote formanyreasons even whenthey say theyvotedwith regardto Social Security),the line would drop

7 A factor analysis on the news items produces two scales, with thedomestic issues (prices, income taxes, interest rates, air traffic con-trollers' strike, and the Conoco buyout) producing one scale with aCronbach's a of .58 and international issues (royal wedding, Arab-Israeli relations, Polish situation, English riots, Irish hunger strikes)roducing another with a Cronbach'sa of .75.Thenews interest items are ordinal variables. I could have estimatedeach with ordered logit, but I wanted to combine the domestic andinternational items, and doing so other than by this method wouldimpose a metric on them. While scoring close interest as 1 and casualor no interest as 0 loses some information, the basic result-thatinterest in domestic and international issues rises with income whileinterest in Social Security falls at higher income levels-remains thesame whether (1) the items are estimated as shown; (2) interest inthe domestic and international items is scored 1, 2, 3 and combinedinto additive scales that violate the ordinal nature of the variables,or (3) the percentage of respondents who closely follow each item isaveraged across domestic and international issues and Social Securityby income level (a bivariate model).

9 While the Citizen Participation Study had, 2,517 respondents, theSocial Security-specific participation items were asked of SocialSecurity households, including 235 senior respondents.569


7/11

Self-Interest,Social Security,and the Distinctive ParticipationPatternsof Seniors September2002

TABLE . Explaining ocialSecurity-BasedParticipationySeniorsSocial Security-BasedParticipationIndependentVariable Contacting Voting ContributingIncome -0.014a 0.017a 0.095a(0.024) (0.014) (0.068)Income-squared 0.00002a -0.00009a -0.001 a(0.0001) (0.00008) 0.001Education 0.391** 0.090 0.126(0.144) (0.097) (0.166)Male 1.310* 1.215** -0.354(0.610) (0.379) (0.614)Married -0.829 -0.020 -0.073(0.566) (0.323) (0.608)Black -0.071 -0.367 0.988(1.050) (0.604) (0.899)Working 1.243 -1.830** 0.252(1.408) (0.678) (1.223)Retired 1.233 - 1.330*** 0.933(1.122) (0.399) (0.755)Age 0.023 -0.043# 0.005(0.042) (0.027) (0.047)Constant -6.838* 1.934 -5.403(3.341) (1.999) (3.745)

N 235 232 234%predicted 94.1 70.4 94.3Cox & Snell R2 0.07 0.09 0.02Source: 990Citizen articipationtudyVerbat al.1995).Note:Figuresncellsare ogistic egressionoefficients.tandardrrors re nparentheses. esults re orrespondentsged65andover. nOLS stimates f thesamemodels,hetolerancesoreachpredictorregreaterhan 10,indicatinghatmulticollinearitysnota problem.p < .10;*p < .05;**p< .01;*** < .001(two-sided).Log-likelihoodestsshow hat ncome nd ncome-squaredrejointlyignificantneach ofthe modelsp < .001).

monotonicallywith income.Evenso,SocialSecurity sless common at highincomelevels,different romtheusualparticipation attern.SocialSecurity-based ontributing lso rises at firstwithincome and then drops.Makingcontributions fany kind,even those based on a concern with SocialSecurity,s unlikelyat low incomelevels,where inter-est in the issue is greatbut activity imitedby incomeconstraints. uchactivitybecomesmorelikelyat mod-erate income levels, where there is more disposableincome but still substantialdependenceon Social Se-curity.As predicted,participationallsat highincomelevels(Figure3c).Althoughvotingandcontributing onot fallmonotonicallywithincome as does contacting,both are lesscommonamonghigh-incomehanamonglow-andmoderate-income eniors.MobilizationSocial Security may be of such importanceto low-income seniors that they recognize their interest inthe programwithoutbeing promptedby other polit-icalactors.Perhaps he impactof governmentaction sas "transparentlybvious" to seniors as it is to gov-ernmentemployees and farmers,who recognizetheeffectsof governmentpolicieson theirwell-beingandwhoparticipateat rateshigher hanpredictedby theirsocioeconomicstatus(Lipset[1959]1981, 196-97;seeLewis-Beck1977andWolfinger ndRosenstone1980).

As it turnsout,the unusualpatternof seniors' nterestin SocialSecurity s reinforcedby a distinctivemobi-lizationpattern:Low-incomeseniors are as likely astheir higher-income ounterparts o be mobilizedbythe politicalparties,andmoderate-income eniors areeven morelikely.Figure4 shows the likelihoods hat seniorsandnon-seniors at various ncome levels were contactedby apoliticalpartyduring heelectionseason,aftercontrolsfor otherdemographic haracteristics. ornonseniors,the likelihood of being mobilized rises with income.This s thepattern ound npastresearch:High-incomeindividualsare more likely to be asked to participatethan their ow-incomecounterparts.Once again,senior citizens are different.The likeli-hood of partymobilizationduring he campaign ea-son rises with income to a point but then falls athighincome levels."1 artiesmayappearto be actingrationally, onservingimitedresourcesby declining o10 This pattern is most pronounced for mobilization by the Demo-cratic party. The likelihood that a senior respondent is mobilizedby the Democratic party rises from .05 at low income levels to .13 atmoderate incomes, fallingto nearlyzero among the most affluent. Thepattern of mobilization by the Republican party is much flatter,andthe curve,althoughmuch more shallow,opens upward: he likelihoodof mobilization by the Republican party is .08 at low income levels,decreasing slightly,to .06, at moderate incomes, and increasingto .08among the affluent. In neither case are affluent seniors mobilized athigher rates than poorer seniors.

570


8/11

AmericanPoliticalScience Review Vol. 96, No. 3

FIGURE. SocialSecurity-BasedParticipatory ctivityamongSeniorsbyIncomea. Social Security-basedcontacting

0.06cnt 0.050.04m 0.03

9 0.020.01

0 .o 0 50 100 150 200 250Income (000s)

b. Social Security-based voting

3 0.40Cno 5 0.300.25

0.20e 0.15-0.10 -0 0.050.00 0 50 100 150 200 250

Income (000s)

c. Social Security-based contributing.0 0.10

A 0.08 o.o4e 0.040.02

So.oo0000.00 0 50 100 150 200 250Income (000s)

Source: 1990 Citizen Participation Study (Verbaet al. 1995).

mobilizehigh-income eniors,whoalreadyvote athighrates.Presumablymobilizinghighly participatory f-fluent nonseniorswouldbe a waste as well, yet that ispreciselywhatthepartiesdo,making he differentpat-tern for seniorsall the more notable.Thus lower- andmoderate-income eniorsare more ikelytoparticipatewithregard o SocialSecuritynotonlybecause of theirgreat nterest n theprogram,but also becausetheyaremobilized o politicsat highrates.

DISCUSSIONThe role of Social Security helps explain the highrates of political participation by seniors. Amongnonseniors,the extremely modest activity levels ofthe less affluentpull down the overall participationrate. In contrast, low-income seniors participateathigherrates thanwouldbe predictedbytheirresourcelevelsbecause of theirdependenceon SocialSecurity.On average,seniors receive almosthalf their incomefrom Social Security (Moon and Mulvey 1996, 29).Like some otherspecialsocietalgroups-farmers andgovernmentemployees-seniors have a large stakein governmentaction. Social Security is especiallyimportant o poorerseniors-the 40% of seniorswhoreceive four-fifthsof their income from the program.Their dependence on Social Security boosts theirparticipationand works againstthe usually positiveincome-participation radient.SocialSecurity's ole is bothimportantandunusual.The SocialSecurity-based articipation f low-incomeseniors s a rare nstance n whichself-interests highlyinfluential.Researchers often have difficultydetect-ing self-interestedbehavior(Citrinand Green 1990).Peoplefail to act in their self-interestbecauseof infor-mationalor cognitive imitations Simon1983)or be-causetheychoose as socialcreatureso actaltruistically(Sen 1977;Stoker1992;Wildavsky1987).With SocialSecurity, hough,the stakes are "visible,""tangible,""large," nd "certain"CitrinandGreen1990,18),andthere is no moralprohibitionon the exercise of theirself-interest.11ocialSecuritys a non-means-testedn-titlementprogram o whichrecipientsor theirspousescontributed during their years in the workforce.Althoughacademicsandpolicyexpertsmayassert hatthe systemis reallya pay-as-you-go ntergenerationaltransferprogramand that the relationshipbetweenone's contributionsand one's benefits is indirect, heprograms popularlyunderstoodas an insurancepro-gram. As FranklinD. Roosevelt said, Social Secu-rity'sself-contributory esign gives recipientsa "legal,moral,andpoliticalright" o their benefits(quotedinSchlesinger1958,309).SocialSecurity lso sdistinctiven Americanpoliticsas a mobilizer of a low-incomegroup.To the extentthat economic self-interest influences behavior, itoften simply augments he high participation ates ofthe most active segments of the citizenry.Those af-fluent homeowners who opposed Proposition 13 were"the most politicized and vocal" portion of the popu-lation to begin with (Sears and Citrin 1982, 220). It islow-income seniors, however, who are most active onthe basis of Social Security, able to be so because oftheir sufficiently high participatory capacity, interestedin doing so because of their great stake in the program,11In addition, the self-interest that is influential here is the rarest andmost rigorous type: "narrow"self-interest, the pursuit of "relativelyshort-run material benefits for the actor or his immediate family,"to use David Sears's definition (found in Citrin and Green 1990, 4).Usually scholars are unable to detect the influence of self-interesteven when it is more broadly defined.

571


9/11

Self-Interest,Social Security,and the Distinctive ParticipationPatternsof Seniors September2002

FIGURE. PoliticalPartyMobilization f Seniors andNonseniorsbyIncome0.50

* 0.45S0.40

o 0.35ES0 0.300.25-

' .0 5 - --. - Nonseniors

CU 0.20o0 0.150.10 Seniors2040 60 80 100 120 140Source: 1996 Nationallectiontudy (Rosenstone et al. 1998). Nonseniors? 0.050.00

0 20 40 60 80 100 120 140Income (O00s)

Source: 1996 NationalElectionStudy(Rosenstone et al. 1998).

and recruited o do so in partbecause SocialSecuritycreates a basis for mobilization.Indeed, this bottom-up self-intereststory is aug-mented by a top-downmobilizationprocess.Seniorcitizens are one of the few groupsin Americansoci-ety in which both the high-and the low-incomemem-bers are mobilized with regardto economic issues,and not only by the political partiesas shown here,but also in other venues. One of every two seniorsbelongs to the AARP, and as with voluntarygroupmembershipngeneral,AARP membersaremoreedu-cated,affluent, ndwhitethannonmembersCampbell2000).However,poorer eniors,unlikemanyother ow-incomegroups, lso havesourcesofmobilization, olit-icalinformation, ndpoliticaldiscussion.Thousands fseniorcentersand nutritionprograms cross he coun-try,manyrunby the Administration n Aging,attractlower-income eniors.Inevitably eniorshearpoliticaldiscussionandpolicynews.One studyfrom the 1960sfound hatalthough eniorsactive nold-agesocialwel-fare clubs were demographicallyimilarto nonmem-bers, club members were more likely to talk abouthealth and medical care (25%versus2%) and otherproblemsof old age (16%to 5%),were morelikelytobelievethat"olderpeople ought o organize o demandtheir rights" 68%to 39%),and were more likely tobelieve that older people ought to be more active inpolitics(74% o48%)(Rose 1965).Hence, ow-incomeseniorsboth takea keen interest n governmentpolicyand learn in these group settingshow to make theirpreferencesheard n the politicalarena.CONCLUSIONAccording o Lipset([1959]1981,190),one factor af-fecting urnoutsthe "relevance fgovernment oliciesto the individual."Groupsthat see a visible effect ofgovernmentpolicyon theirwell-being-including gov-ernmentemployees,farmers,veterans,and,as shown

here,SocialSecurityrecipients-have a greatstake ingovernment ctivityandparticipate thigher ates hanwouldotherwisebe expected(Lipset[1959]1981,190-94;Wolfinger ndRosenstone1980).The SocialSecuritycase is also unusualandimpor-tant because the program'sparticipatory ffects aremost pronounced or low-income seniors.Ordinarilyit is difficult o mobilizepoorer ndividuals, utlower-income seniors'dependenceon Social Security s sogreatand so obvious that seniors are able to act in aself-interestedmanner,protecting heirSocialSecuritybenefits.SocialSecurity husprovidesanimportantn-stance of self-interest ignificantlynfluencingpartici-patorybehavior.It would not be an exaggeration oconclude that the program s largely responsibleforthe creationof the seniorconstituency nd ts electoralsignificance.APPENDIXA: ROPERSURVEY8108QUESTIONWORDINGSIssue-specificnterest.Of course,everyone s more nterestedinsomethingsbeingcarriednthenews han n others.Totakesome differentkindsof examples-Is news about[each temin turn] somethingyou have recentlybeen followingfairlyclosely,orjustfollowingcasually, r not payingmuchatten-tion to? Whichpricesaregoing upor down and how much?:the legislation o reduce ncometaxes; helegislation o dealwithchanges n SocialSecurity;nterestratesbeing chargedbybanks; ttemptsbySeagrams,Mobil,Texaco,DuPont,andothersto buy Conoco;the marriageof Prince CharlesandLadyDiana; hehunger trikesbyIrishprisoners;heriots nEngland; elationsbetweenIsraeland heArabcountries;hesituation n Poland; he air traffic ontrollers' trike. Coded1 if followedclosely,0 if followedcasuallyor not much).Income. Now here is a list of incomecategories.Wouldyou call off the letterof the category hat best describes hecombinedannual ncome of all membersof thishousehold,includingwagesorsalary, ensions,nterestordividends, ndall other sources?Nine-category ariable.

572


10/11

American Political Science Review Vol. 96, No. 3Race. Recorded by interviewer. Coded 1 if black,Ootherwise.Gender. Coded 1 if male, 0 if female.Marital status. Are you married, single, widowed, sepa-rated, or divorced? Coded 1 if married,0 otherwise.Workstatus.Are you at present employed, either full-timeor part-time? [If no]: Are you a housewife, unemployed, astudent, retired, or what? Coded 1 if working, 0 otherwise.Education. What was the last grade of regular school thatyou completed-not counting specialized schools like secre-tarial, art, or trade schools? Seven-category variable.

APPENDIXB: CITIZEN ARTICIPATIONSTUDYQUESTIONWORDINGSSocial Security voting. In the past five years, have you takeninto account the position of a candidate in relation to SocialSecurity Payments in deciding how to vote? Coded 1 if yes,O f no.Social Securitycontacting. In the past five years, have youcontacted a government official to complain about SocialSecurity Payments? Coded 1 if yes, 0 if no.Social Security contributing. In the past five years, haveyou given a campaign contribution based, at least in part, onyour concern about Social SecurityPayments? Coded 1 if yes,O f no.Income. Which of the income groups listed on this cardincludes the total 1989 income before taxes of all membersof your family living in your home? Please include salaries,wages, pensions, dividends, interest, and all other income. [Ifuncertain: What would be your best guess?] Sixteen-categoryvariable.Race. What is your race? [Which category describes yourracial background?] Coded 1 if black, 0 otherwise.Gender. Coded 1 if male, 0 if female.Marital status. Are you currently married, living in amarriage-like relationship, widowed, divorced, separated,or have you never been married? Coded 1 if married,0 otherwise.Workstatus.Last week, were you working full-time forpay,working part-time for pay,going to school, keeping house, orsomething else? Coded 1 if working, 0 otherwise.Education. What is the highest grade of regularschool thatyou have completed and gotten credit for? If necessary say:By regular school we mean a school that can be countedtoward an elementary or high school diploma or a collegeor university degree. Did you get a high school diploma orpass a high school equivalency test? Do you have any collegedegrees-that is, not including degrees from a business col-lege, technical college, or vocational school? What is the high-est degree that you have earned? Eight-category variable.Age. Calculated from year born.

APPENDIX C: 1996 NATIONALELECTIONSTUDY QUESTION WORDINGSParty mobilization. As you know, the political parties try totalk to as many people as they can to get them to vote for theircandidate. Did anyone from one of the political parties callyou up or come around and talk to you about the campaignthis year? (v961162; coded 1 if yes, O f no.)Income. Please look at page 21 of the booklet and tellme the letter of the income group that includes the incomeof all members of your family living here in 1995 beforetaxes. This figureshould include salaries,wages, pensions, div-idends, interest, and all other income. Twenty-four-categoryvariable.

Race. (v960067; coded 1 if black, 0 otherwise).Gender. (v960066; coded 1 if male, 0 if female).Marital status. Are you married now and living with your(husband/wife)-or are you widowed, divorced, separated,or have you never married? (v960606; coded 1 if married,0 otherwise).Work status. We'd like to know if you are working now,temporarily laid off, or are you unemployed, retired, perma-nently disabled, a homemaker, a student, or what? (v960616;coded 1 if working, 0 otherwise.)Education. What is the highest grade of school or yearof college you have completed? Did you get a high schooldiploma or pass a high school equivalency test? What is thehighest degree that you have earned? I used the summaryof the above three questions (v960610). Seven-categoryvariable.Age. Calculated from year born (v960605).

APPENDIXD: PREDICTINGSSUEINTERESTAMONGSENIORSANDPARTYMOBILIZATIONF SENIORSANDNONSENIORSTABLE l. Predicting ssue InterestamongSeniors(Figure2)Independent Social Domestic InternationalVariable Security Issues IssuesIncome 0.135# 0.006* 0.008**(0.075) (0.002) (0.003)Income-squared -0.003*(0.001)Education 0.176 0.034* 0.045*(0.151) (0.016) (0.019)Male 0.078 0.064 0.010

(0.440) (0.051) (0.061)Married -0.261 0.031 0.034(0.407) (0.046) (0.055)Black 0.314 -0.030 -0.023(0.599) (0.069) (0.082)Working 0.158 0.046 0.074(0.779) (0.086) (0.106)Retired -0.001 -0.008 0.048(0.502) (0.059) (0.071)Constant -0.089 0.254*** 0.078(0.670) (0.073) (0.087)N 202 196 196%predicted 78.2R2 0.04 0.14 0.15F-test p


11/11

Self-Interest,Social Security,and the Distinctive ParticipationPatterns of Seniors September2002

TABLE 2. PredictingPartyMobilization fSeniors andNonseniors(Figure4)IndependentVariable Nonseniors(18-64) Seniors(65+)Income 0.002 0.025a(0.003) (0.018)Income-squared -0.0002a(0.0002)Education 0.190*** 0.063(0.049) (0.093)Male -0.166 0.030(0.146) (0.293)Married 0.454** 0.096(0.162) (0.305)Black -0.370 -0.149(0.271) (0.501)Working -0.002 1.206*(0.203) (0.611)Retired 0.261 0.703(0.376) (0.486)Age 0.020** 0.004(0.007) (0.019)Constant -2.967*** -2.072(0.380) (1.562)N 1,113 281%predicted 73.9 63.3Cox&Snell R2 0.05 0.04Source: 996National lectiontudy Rosenstonetal.1998).Note:Figuresncellsare ogistic egressionoefficients.tan-dard rrors re nparentheses.ncome-squaredaseliminatedfromhenonseniorquation ecausea checkormulticollinear-ityrevealed high orrelationith ncome.nOLSestimatesof the modelsshown, he tolerances orthe predictorsregreaterhan.1 ,indicatinghatmulticollinearitysnotaproblem.* < .05; **p< .01; ***p .001 (two-sided).Log-likelihoodtest shows hat ncome nd ncome-squaredre ointlyignifi-cant p < .001).

REFERENCESCampbell, Andrea Louise. 2000. "The Third Rail of American Pol-itics: Senior Citizen Activism and the American Welfare State."Ph.D. diss. University of California at Berkeley.Citrin, Jack,and Donald Philip Green. 1990. "The Self-Interest Mo-tive in American Public Opinion." Research in Micropolitics 3:1-28.Conway, M. Margaret. 1991. Political Participation in the UnitedStates. 2nd ed. Washington, DC: CO Press.Duensing, Edward E. 1988. America's Elderly:A Sourcebook. NewBrunswick, NJ: Center for Urban Policy Research.Federal Interagency Forumon Aging Related Statistics. 2000. OlderAmericans 2000: Key Indicators of Well-Being. Washington, DC:

U.S. Government PrintingOffice.Lane,Robert E. 1962. Political Ideology: WhytheAmerican CommonMan Believes What He Does. Glencoe, IL:Free Press.Lewis-Beck, Michael S. 1977. "Agrarian Political Behavior in theUnited States."American Journal of Political Science 21: 543-65.Lipset, Seymour Martin. [1959] 1981. Political Man: The Social Basesof Politics. Baltimore: Johns Hopkins University Press.McCarthy,John D., andMayerN.Zald. 1977. "Resource Mobilizationand Social Movements: A Partial Theory."American Journal ofSociology 82: 1212-41.Mettler, Suzanne. 1998. Dividing Citizens:Gender and Federalism inNew Deal Public Policy. Ithaca, NY: Cornell University Press.Mettler, Suzanne, and EricWelch. 2001. "PolicyFeedback and Politi-cal Participation:Effects of the G.I. Bill for World WarII VeteransOver the Life Course."Paper presented at the Annual Meeting ofthe American Political Science Association, San Francisco.

Milbrath, Lester W., and M. L. Goel. 1977. Political Participation.2nd ed. Chicago: Rand McNally.Moon, Marilyn, and Janemarie Mulvey. 1996. Entitlementsand theElderly: Protecting Promises, Recognizing Reality. Washington,DC: Urban Institute Press.Nie, Norman H., JaneJunn,and Kenneth Stehlik-Barry.1996.Educa-tion and Democratic Citizenship in America. Chicago: Universityof Chicago Press.Plutzer, Eric. 2002. "Becoming a Habitual Voter: Inertia, Resources,and Growth in Young Adulthood." American Political ScienceReview 96: 41-56.Powell, G. Bingham. 1986. "AmericanVoting Behavior in Compara-tive Perspective."American Political Science Review 80: 17-43.Roper Organization. 1981. Roper Reports8108 (August 1981) [com-puter file] (Study No. USRPRR1981-08). Storrs:Roper Center forPublic Opinion Research, University of Connecticut [distributor].Rose, Arnold M. 1965. "GroupConsciousness among the Aging." InOlderPeople and Their Social World:TheSubcultureof theAging,eds. Arnold M. Rose and Warren A. Peterson. Philadelphia:E A.Davis.Rosenstone, Steve J. 1982."Economic Adversity and VoterTurnout."American Journal of Political Science 26: 25-46.Rosenstone, Steven J., and John Mark Hansen. 1993. Mobilization,Participation,and Democracy in America. New York:Macmillan.Rosenstone, Steven J.,Donald R. Kinder,and Warren E. Miller,andthe National Election Studies. 1998. American National Election

Studies,1996: Pre- and Post-Election Study [data set]. 3rd release.Ann Arbor: University of Michigan, Center for Political Studies[producer and distributor].Schlesinger, Arthur M., Jr. 1958. The Coming of the New Deal.Boston: Houghton Mifflin.Sears, David O., and Jack Citrin. 1982. Tax Revolt: Something forNothing in California.Cambridge,MA: HarvardUniversity Press.Sears, David 0., and Carolyn L. Funk. 1990. "Self-Interest inAmericans' Political Opinions." In Beyond Self-Interest,ed. JaneJ. Mansbridge. Chicago: University of Chicago Press.Sen, Amartya K. 1977. "Rational Fools: A Critiqueof the BehavioralFoundations of Economic Theory." Philosophy and Public Affairs6: 314-44.Simon, Herbert A. 1983. Reason in Human Affairs. Stanford, CA:Stanford University Press.Skocpol, Theda. 1995. Social Policy in the United States:Future Possi-bilities in Historical Perspective.Princeton, NJ: Princeton Univer-sity Press.Soss, Joe. 1999. "Lessons of Welfare: Policy Design, Political Learn-ing, and Political Action." American Political Science Review 93:363-80.Stoker, Laura. 1992. "Interests and Ethics in Politics." American Po-litical Science Review 86:369-80.Toner, Robin. 1999. "Older Voters' Shift to G.O.P. Is Democrats'2000 Challenge."New York Times(May 31) Al.U.S. Bureau of the Census. 1999. Statistical Abstract of the UnitedStates 1999. Washington,DC: U.S. Government PrintingOffice.Verba, Sidney, and Norman H. Nie. 1972. Participationin America:Political Democracy and Social Equality. New York: Harper andRow.Verba, Sidney,Norman H. Nie, and Jae-on Kim. 1978. Participationand Political Equality:A Seven-Nation Comparison. Cambridge:Cambridge University Press.Verba, Sidney, Kay Lehman Schlozman, and Henry E. Brady.1995. Voice and Equality: Civic Voluntarismin American Politics.Cambridge,MA: Harvard University Press.Verba, Sidney, Kay Lehman Schlozman, Henry E. Brady, andNorman Nie. 1995. American Citizen Participation Study, 1990[computer file]. ICPSR version. Chicago: University of Chicago,National Opinion Research Center (NORC) [producer]. AnnArbor, MI: Inter-University Consortium for Political and SocialResearch [distributor].Weisberg,Herbert F.,and Bernard Grofman. 1981. "CandidateEval-uations and Turnout."American Politics Quarterly9: 197-219.Wildavsky, Aaron. 1987. "Choosing Preferences by ConstructingInstitutions:A CulturalTheory of Preference Formation." Amer-ican Political Science Review 81: 3-21.Wolfinger, Raymond E., and Steven J.Rosenstone. 1980. Who Votes?New Haven, CT: Yale University Press.

Campbell - Self-Interest, Social Security, And the Participation Patterns of Senior Citizens

Documents

Transcript of Campbell - Self-Interest, Social Security, And the Participation Patterns of Senior Citizens