Politics-US voting rights and rules
THE PERVERSE CONSEQUENCES
OF E{-HCT{}KAL REFOR&,{ TN THfl L]ru{Ttr* S"TCTHS
ADANI J. BERthtSHY l
fuIassachusetts Inslitute of Tech*a1cgy .
A nunrher olelcctoral refirrms have been cnact*d in dre iinitcil States in the past three d*cade-o thal are dtsignccl to incrcrss tnrnout tr)),casing restricticns on the casting ul billlots. Iloth prcp*- nijnts iiird opponents of ,-I,-ctora! retbrnis irgree ahat these refnrms sirculd increaso the denrc- gr-;rphii: represerltrtivei,ess 01'Lhe electoratc by reduoing the direct cLrsts r*'i,oting, thereb), incre*sirg iumont among less-privilegeii grtmps rvho. presumrb11," ue rnosi !cnsitive to the costs oi curning to the polls. fu lact. these relarrnr l.lave been greatll,'contesled because hrrth n.r:rjur" political prrties b.lievr: that incrcasing lrirnout amons lcss-pliviicgcd groups r.vi1l bcnclit lfcmo- cratic l-roliticians. I revieu, evidence tiom numerons studies olelectoral reform to demDnstraic that retbrms d*signed io rnake it ersirr l+r i*gistered vot.rs to cast their b:illots actually increase, rather than rerhice^ socioecr.rnon:ic lri:rses in ihe ct mposition ol the vr:Ling pubiic. I oonclude with ii retlrmmen{iation that s,e shift lhe tircus ol'electoral relorm lrcm an emph:rsis on iristituiional ciringes lL) ir concentration ou pclilic:rl rngagenrent.
&c-t$orrJs.' elcL:lion re.fb*it: t'rli!n_q; po!ititul ltar'Lit:il;tt[io;t: tttrtu 'Lti: q !)ft'\t tltoti 'rl
ln the past three decades, a series of electcral let"orms have been eilacted that iire designed to increase tllrnout b-v: easing restrictions on tire casting of ballots" i:rcluciing voting-by-iiuail iVBL,{), early voting, the relaxing of stringent allsentee balloting procedures, and {nternet vating. Both proponents and oppoflents of electoral refbrms agree that these reforms should increase the demographic r:epresentativene ss of the eiectorate by reducifig the direct costs of voting, thereLry, increas- ing tumout aittong gr*L:ps *f, lcrwer socioeconomic status who, pr"e-
,4utiror's Nott: This article draws hr:irviil, up+n.ioint ivork rvith Nanc;, Burns anJ X.,fike Iraiigott. Thel. tre eiiu;rlly responsible 1'or the iievelopntent of the lrvr:-path ct*ception cl'turrloul ilis- enssed here but ;rre innocent ol an1, em'urs in this arlicie. For conrlnents rrn e:rrlier ilratis of this ;rrLi<"'le. I u,uultl liire to lllal)li Jastn B;rrahas. Janrie flrucknran. Jtn:r.r Girnpel. Ben Highton. .i*nnitir.lcrit. Girbricl Lcnz, Andrerv lLetlve s. ancl Sarah Slcd. I ivtuld also iikc to thank N'Iichae I Hann:er ii,.r pror,iiling lccr-ss to his rccoded CPS d:rt;r and Jarncs McGhec ilnd Alice Savage 1*r reselr*il assistance.
AMEltIald\ POLITICS RESEARCII, \bl. -r-3 Nc. -1. .luhi 1005 j7 l-.1!l I tl0i: L0.i llji 153267-iX0j26a4l9 O 2t)0-5 Sage Puhlications
+,.]
17) AI4I:RICAFI POLITICS RESII.,\RCH r'.IULY l0{.}.5
silmabiy, are m{}st s*nsitive lo {he e{}sts *f clrnting tr> the pllnls. A r11{.'}r"e derllogr"aphie ally rliverse and rrpresefit:rtive etr*r:torate r-trrl_y har,f, impi:rtant ir-r:pli*atior"rs far the .Ameriean p*litieatr s.v-steir.r Eiectoral refornu have. in iact, been gre;rtl--v contested because both nr;rior poiit- ical i:arti*s beli*ve that incl'easins turn{}ut alnong less-privilegeil groups rn,ill he**t'it Denrocratic politicians. Hor.l,evel', contrary- t{) expectatirns. th* rvave of electoral reforms may fi".'" increased tum- out sligirtly hut has iiot had the hypothesizecl partisa* effects.
Ei*et*i:al reloi:::s. ner.efiheless, ]tavr: had a significant. it unin- te::cied. efTect. i\{ost politiciaiis anrl scho}ars have focused rr:fi:rm efiorts on the tangible barriers to voting. The se retrlrrns are designed to lclr,cr or erase the barriers to vritiris, making il easier for s/l citizens to vote, regarclless of their personal circumstances. llcrrvever, thsl riircct costs of registrati<;n and getting t+ {he hall<}t box arc onlv ;:art of fi:e picture. The firclre significant costs are the c*gnitive eosts of hecoming engaged u,i{h and infor:ned about the p*litical i.r,orld. Pr:iit- ir:al inierest ancl engagernent, afier ali, cleter"mil:e to a large extent s,ho l,*ies anll u,ho cioes n<tt. Becanse ievels of political eng;lgemfnt cui:- rent!1' fr:ll*u,, rather than cross, dernographic dil,isions in the *lectar- ate. retbrms designed to nlake voting "*asier" magnily th* existing socio*:coiiomic lriases in the coirposilion af the elecaorate. 'ni:es* retbrms ensllre that those r":itizens rvho ar* most engaged rvith the politic al world-tho-ce with poli tic all5, relel'ant re ! c) u rc Lr (-c on tin ue to parlicipate, rvhereas those individuals lvithout such resources fall bi; rhe rvayside.2
Xn this r'sS&Y, X re vifu, errictrence from numerous studie s of ele ctoral prae{ices to den:onstr:rte that refbrm,c desig*erl l* ma}.,e it easier ibr regist*reri voiors io east their baitrots inr:re;rse" rather than redLlce, socioeer:t+mic biases in the co*:p*sition r;1'the vclting puhlie .r Tliis artir:tre dis,--usses oirl--v those ref'olms that :l"lake it easier to vote once registered. Adrt"rittedl-v" the registrati** +f vr:tels is a criticai ste;: in the votileg pr{}cess. But, ;rs I disr--uss in great*:: deti:ii i* wh:i{ tr-rllolvs, man.r,' sf the argunler:ts made lrere c*:rcerning tlre ac{ cf' voting calr also be appiied to th* stud.r,- of'v*ter registration (Traugott, ?CI04; see also }{ightc:r,r f2**41 for a rl*tailer.J disr":ussian of'thr effects of easing re,{istration require ir:*nts i.
I begin b3,, advancing a tu/o-process dl,ualnic eonerption of the electorate. \\rhen discussing v*ting reforrns, politicians and acar-1efir-
Bcrinskli / LIONSEQUEN{IES {lF F,LECTURAL RLii{)RI{ 173
ics of,ten eoncenh'itte ou etlcouraging-*r "'stipullting" new votels tit join th* elector".ate. Horvever. stimulatian cn1,v tetrls half the stori/. W€ uust aiso consider tire rctention of voiers fron: eiectii:n to elec- tir:n. {Jsing the stirnulation/retention fi"ameu,ark" I argue ti:ili ref'cr*rs desigr"l*d to make the ;ict of voting easier increase the socio*co**mic Lrias of the electorate. Rather than stin:ulati::g the unengageri, *,ho are relatively defir":ien{ in political and ec*noinic resources, these retbrms reiain engaged vr:iers-those rvhc arc rich in poliliqnlly relevant rt:sourlr*s. Thus, althougtr e lectora! leti:r*:s *ov'irrr.*urrl turnout, il:e.v do so hy ensuring lllat politically en-{aged vo{crs ccntinric to elrme to polls elee tion after election. I enrl r.vith a ealtr to refocus the stud3,. of voting-aiid political participati** illilre se!1erall3.'-ar,r,a1,' frorn an *,r*mination of the clirect costs r:{the act of voting ta a sir"rdy oi' the Less perceptible, hut ec1ualtr.v- irilportailt, costs Lrl'plrliti*a1 eilgage- ment. Iiesti{utionai cirange al*nc rvill not hring ah*ert a democratic electorate. Nii matter hor,v lolv the clireci c$sis to casting a ballcit are set. the onl.v wav to ar-r-**lpiish b*th goals of increasing tunlout and eliminating socioecon*r.nic biases in tlie voting population is tir i*crease the e:rgagement of tl-ie l:roade r mass publir- x,ith the political
\-. rvoilcl. Political infon:ratii:* aad interest, not the high tangible costs of the act of vctii:g, are the real barriers tt: a tnily democralic vating puhlic"
ETECTOILAL REP]*ETrI,CS
This levier.v ilre uses on four partii:lltrar retbrms rtresigneri {o ease the act of voiilg l'or citizens regist*re cl to vc]te in a given-jurisdiciion: per- tnissive *bsentee l-ralllrting, early v*ting, VBM. and lnternet voting.] I describe {?rese reibrms in turn here.
PF],IIMISSTV*] A}iSE]!TEE I}AI,T,{}TEF]{;
&'{aay states resirici the use *f ahsentee bal}ots to those citirens s,ho cat?n*t make it to the ponls on Electio* DaSi bssr,ur* ot r':ccupation, bnsiness, stndies" travel, or other reasons. Hclrvever, other states have liberalized {}:e use of ahsentee ba}l*rs. Altogether 22 states rtorr, make an abser:tee ballot availahrle ta any registered vot*r rvli+ requests one.
414 Ai.{IIRICAN P{)LIT{CS RESIIAIICH r'.ll.rl-Y :(}(.j.5
withclut *ee.1 to shol,l,' r;luse (Hnnser:, :001).: Ai-rsentee ball*ting has becr:me extren'Iel)i p{}pular in these staies. Feir example. in th* past ttv{i elections in Cetrifoinia {the 2iX}2 general *l*ctii:n and the 2ti03 guber- natorial electicn), ffi{}r'e ihan ?5% {}t voters cast their vctes thr*ug}: atrsentee hallots"
E.{RI,\'lOTINI;
Siiice 1988, voters i* Texas anci I -l ilther slates h*v* beeu permittecl to citst ballots, in pers*n, *g: to 3 ..r,eeks irelbre the electi** *t de sig- nated poling sites {F{a:is*ia,20{}}: Stein. 1998)." As is th* cas* rvit}: permissive aLrsentee ball*tir;g. vo{ers do llot har,e t* *-ttai:lisk prior cause to cast earl,v balli:ts. tsailoting sites are oiten piaced in conve- nient plar:e s, sueh as supermarkets and post office s" ln addition, in cer- tain states-sueh as T'ex;rs-cilizens do rrot have {o rrote in {heir orvn precinct. Insteart. tht'v cair vote ;lt any early voting site rhel, find c*rr- venien{. Like absentee balloting, eae'tr;- r,oting is used rvirlel1,. rvhen avaihble. t* 2iX]0, tor instance, 39?o *f the presid*irtial vr:te in Texas ,,vas cast eartr,v iF{anse*, 20CIn).
VOTINC.BY-N,IATI,
lJ*del"VE&'f s_r-'sterns, everv reqiste:ecl r.oter in a gir.en jurisdiction r'***ives a halj*t thilt the_v r:a* dr*p r:ff or mail biick io the proper eiee - tion r:fficial. B,v mailing ballcis to the honles cf every resisterecl voter tp tt 3 weeks in arlvance of Election Ilay, elecfion off,icials give poter"ltial voters a:::i:le ti*:e to cast th*ir votes at their ccnvcnience. VBlr'I elections begarlr at tlie local lev*l ir: the late l!]70s a*d have grorvn incl*asinglv popular. By o::cr estimate" lg states have used Vts&'l far at least one *lectir:n and, si*ce i998, *rcgcn has conducted al1 its elections b-v rxail itserinsk3,. 8urns. et I'r*ugott, 20i)1).
I}'J?flRNET VOTING
i*temet votirig is a vctiirg systerx &at rvouid e:iable l,ifters t$ cast a seclir!' and secret balk:t over the Inten:e t. These hatrlots could be e ast ilt computers i:laced at trarliticn;rl polling plaees, ri*signateel Xnterner kiosks uneler the cclnirc,l ol electi*n *ificials, r:r 1i'orn ritizens' rirvn
Berin-skr, / COIiSE{}lllil{CES OF ELECTORAL REFCRhJ 4,-5
h+mes-u,hat Ah,arez and Hall {2$04) cail "remote internet vi}ting." lnteruet voting has nq:t been ael*pt*d in any votirrg jurisrlie lio:-i in the Uniteel States, bul proponenis of this meth*d point to its promise in easing the costs of- r,oting, especiali-v f'rx' milital'y personnei and the disabled lAlvar'*z & Hall. 2004).
TT{E T}YI\AI\,,IICS OF ?E{E YOTII{G PROCESS
To begin an examination of the effects of elecloral }aws *n regis- tered voters, we fiIust start r.vi{h a m*del of the electorate" Cne e*m- 1111]fi e oncepticur of thr: elect*rale . adviinced most explicitl5r b5; Camp- h*lt i1966) is that of a polentiall,v expanding series of coni-:entric cir*les. Campbell rnakes a distinction tretw'een a har:t-l "'core" of voters i.vhi: always turn out. regardless of the panicnlars ot- an etrection, and sets of "periilheral" voters who p;uticipate only in high stakes elec- tions. As Campbell iirgues, "The turnout in an,v specilic election is largel,v a questior"l of how manv of the Iess interested, less respo:;sive peo;:l*" are sufllcientl,v stimulated hv the political circnmsiances of the m<:rn*nt ti: malie the effort to votc" ( 1966, p. .13.).'Fron'r this pr:int of vielv, the lva.v to expaud the electcrate is i* scmehc*, compel groups rf peripheral voters to tllrrl out.
Such a vierv of ii:e e lectorate r*sts on tlre assun-iption that "per:iph- eral voters" are the mosl liliely to ehange their voti*g behavior fr*m eiection to election. thereb_v" driving tire cliarge in aggr"egat* lur:rr:ul ler,,els. UnrJer this vierv, onl-v the "cllre" el*cl*rale votes in lor,r,-stimu- lus elections. s*eh as lccal elcrciions. A-c {he elc*toral rnr.,ironmen[ trecrrmes mor"e chlrg*d, as it does lbr close races and eleetions fru' nationai of{ice, the voting population erpanrls outwarrl ir+m {he r:*re. bri*ging i* uroups of increasingl3' apathetic \ioters frcrn ttrre peripher-v (Sigelman & Jervell, 1986).
Citizens are &ot, h*rvever, so easil.v so:tecl intc "core" ancl "peliph- eral" votet's.3 Certaini3, some indiviti!.tals l'ote itt *l,erv electiorl, ancl soine individuals nev*r \rote, brit nranrl,'citizells sonretinl*s c&-rt a 1'ote ancl sonre{im*s do not. Tun:*ut e Lrhs and fl*rvs as {he characteristics of electians change; for instance, ma*_v mole citizens vote in statelvide elections than school board elec€ians. M*r*q>v*r'. ti:e volatility of incli,
4,-b AI,IIIRICAN POLITICS RISL,AITCH 1 .lLrLY :(l{}5
vidual turnout from election to eiectirrn-changes lvhich in the aggre- gate result in a chumir:g clt the electorate-is relativeiy high even tr.etrveen higir intensit_v- elections, such as Fresidential elecricins.e F*cusing on aggregate tllfixrut ti'om election to election ohscures the volatility in individr"ral-level turnilut behavior. r\n:r11,sis of the voting t:eharrior of indivicluals or.er tim* clemonstrates that the n:ilvernent *f citizet:s in a*d out of the eler:torat* cloes not contorm to a vi*lv *f eitr:- t*:rate as ctlrnposed oft "peripheral" vcters rvi,o jr":in ihe "core" l,otels clly in l"ligli-stirnulus elections. Sige imali and Jewell (1986) oxarn- i;:ei1 the tllrnolrt pattenls of registered voters ol:er l0 e ]*cti*r:s in l{en- tueL,v ironr 1978 ta 1982 and fcurrel tremc*rir:us volctilit,v in i*divid- uai vr>ting tle haviar over thai periacl. N{ore reeenll3,. in a series oi live elections in Oleg*u in the earl.v l99ils. Berinsky, Burns. alld lraugott i200i) l*und {hat the vast niajority *l r*gistered voters. 65c/o, were "transier:t r,*ters" r,r,h{r raoved in anri out cf tire eleclcirate over the *Iection series. By, e*n-rparison. 257c rvere "constant vlrlers" r,r,h{r turnertr cul i* a]i five *iections, aurl onl1, l0% rvere chr*nic "'non-r,'{}t- ers" "uvho ne!'er cast a ballot, tl"rough they r,vere etigible to do so.r{)
Taking a drvnamic perspect;ve of the e]*ctol'ate, rve can see tllat stilrtulatine nerv \rotl:rs to join the ranks of the voring public is *r:ly t.rn* path to increasing tnre size of the elect*rate. Electot:al ref*rl:is can also i:.lr'rt-ase the size of the elcctorate ov*:: t!:* lorig haul tl:rough vgvslisi$l-fueeping those people wh* have voted iil thr past virting iii the tr.ltui'e. in loi.v-interrsity local *lections as r.l,ell as irigh-intensit,v presidentiai elc:cti**s {B*rirrsky. Bnlns, & Traugott, 2001; 'Ii'atl,qotr. 20CI4).it Under this sceuario. turnout may increase in the flgglegate relatir.,e to the sitr"ration rvhet"e retorms r/,,ere not imirlemented because i'otrrs r.l,ho previor-lsly tulirertr out are transfbrm*d frcm'"transient vrlt- els" to '"corlstant voters,"' th*r*brv reriucing the movement o1'eitize*s **t *f the ranks of-the voting public. Thus, not onl,v rvill the c*mposi- tian *l the electorate change bi:t there r.l,ill h* ;r ::et illcl'ease in tur* uut.1l
This trvo-path d-y;:arnie u-i:nception of'tt;rl:*ut has consequetrces for the stuclv *f e le*oral refor*:. Scr]:alars a*d p*liticians argue-,--either irnplicitl,v or cxplicitlv-tilat *lectolal refor:ls that rnake voting eas- ier ivill incrrase the size $l {he *lectorate. and tha[ this increase rvill res*lt fu-om th* stirculalion *l registered nonvoters. '{'hat is. rhey itssltilie thiit electoral refoi:ms i,l,il1 expand the "cortr" electorate out-
llcrinskv / CIONSEQUIII{CES alF FILIICTOR.AL REF()Rh{ 477
wald ii) hring in new gloups of "per"ipheral" rrclters. By lor,vering the elirect c*sts of voting. these scholars assurne that groups of indiviciuals rvho ;ire close to the threshoXd of par-ticipation rvill join the ranks of the constii*t rroters" Such logic is incomplete. Electoral reforms ma-v increase turnout through stimulation. F{olerrer, electoral refrrrms also serve to retain transient votels fiorn election to election, through hoth higl-r intensi ty and I orv-intensity contests.
Disringuishing retention effects fiom stimulati*n effects is irnpi:r- tant becarise. if the stimulated votefs'and retained voters have difterent socioccononlic characteristics, these processes rnay afTect the coiltpo- sitiot ol the voting public ill different rvay,'s. SpecificallS,, the relii[ive size and characteristics of these populations rvil1 deterurine the corn- positiori of the electorat*. lf thcse trvo qroulls cliffer in ilreir personal cher*cteristics. and the se cirarai:teristics aft'ect the ability t:f electoral reforms to change voting behavior', electoral refonns may lxrve signii- ieant ccmpositional efleets" Thus, if we are ci:ncerned abii*t the clemographic representativeness of the electorate" lve must s*e how the processes of stirnutration end retention proceecl, and pa5,' attention tr: th* characte ristics and sizes of the graups of resisterecl vcters who are r:.rost likel.v to change their i:ei:avior in respollse to eiectoral reforms"
Here is lvhere lhe perl.erse conseqlleflee s i:f *lectoral refot'rn take rool. Both prclponents arrd oi:ponents of ele.rtoral refoms agre* lhal these rcforms wauld increase the demographir": represeiltativeness of the electorate b3; making it easier l'or *itizens r:f lou, socio*csxroilic statr.rs to cast their hallot. That is. they believc that electrrral relbrms rvil1 sti*-rulate the less w,el1-alT-w]:o historicall3, u-omprise "ilon-\,ot- s1c"-1s -ioin the etreet*rate.rr These results do no{, in liict, occrrr. Beciiuse ttrre population of "transient" r,*{*rs is *nck:rved r,vith greater ievels r:f politicalll,' l'elevant resoLlr'.rcs^ voting ret'orms hl'i*g lvell-off citizeus tr: the p*}ls repeatedly. The reas** ferthis efi-ect is straightfbr- rvard. Sometimes, erren ttrle nlost engageil and inlerested citizens will iail to cast tl"leir vote on election day because they are sick, the rveather is bad" they have missed a registratiori deadJine, tXrey have moved, cr they are called out of to$/n Llnexpectedl,v on business. Electoral relorms therefore inr:reas* the prope*sity of likely voters to consis- tently turnout bv smoothing over the idiosyncrasies that cause etrgaged ci{izens to somelimes miss castin*s their votes in pafiicuiar
J-78 AI,IERICAN P0LITICS RF,.{iIARCH 1.il]LY JO{}.5
elections. Fut simpn3,. *iectoral refcirms h*ve a greater effect on retention iltau on stimul*tion.
The net ef'fect *l this process is that, over tiine, electorai reitrrms r.r,'iil likel,v increas* the rate of turn*ut re]*tive t+ the situation u,here the letorms ',vere nct in place. Eut ii *,ill achieve this inerease b.v ret;li:ri::g ihose citizens r.vhc hav* a high proper:si{' to r,ntr, }lot t}v stimulating r,"rl'-presumahl.v less engaged-citizer:s t* .!*fur tire ranks *l the voting piiblic. r\s Southr.vell and Burchett (20CI0i prit it. voting retbrms bring th*',ei:ealr: of the u:r:p" ta the p*11s. Th*s, elec- toral reforms clt"l not charige the nature of the electorate. Rather, lhev give us "rntfre of the same"-tliey'. exacerhate. ra*rer than e*se, erisi- ing st>ciaeconomie biases in {lie corcposition of the voting public.
THE Ul/IililiYcil
Such a concl*sii:n might be somer.r4xrt surprising ti: those whr: fol- lor.i.,the p;lfiisa* dehate concerning electolal refixm" but the er,,idenee is clear. A large nnnrber of empirical studies of reforins rlesigr"red t<t make the actual castiieg of the vote easier have convergeii on a coin- *:cn cenclusiorr. Although these rctbrms do not crea{e partisan hiase s i* thc electorate" thel'reintbrce the demographic compcsitional hias r:f ahe electorate a*d r]13-v oven heighten that bias.
{deall"v, to obsen,e the eftects of voting reforrn on the composition cf the electorate, lve should examine over tirne individual-level data to se* il'these refcrrms have a gre$ter elfect on stimulation or on reten- ti*n, u,,hich coukl not he seen in the;rggregate results. Such data allow us tr': assess hon,particular elee tola] reforms change the ordin;u'y pat- te*rs of the churning of the eleelorate.
Berinsk5,. Bums. ancl Tiaug*tt i200t) conrluctecl one such investi- gation. str"rd-ving the eff-ects of VBful in Oregon. The autl:rors fau*d tl"rat VBl,'i .iid increase the tumout in that state, hut it did so primaril;1 through reterltion. Specifica}l3,', they I'ouncl that VBtrI was not very,' effective at prilling resistant registered nonvoters into the el*cia::aie, hut it did a vel-y good job of retaining existirug lrcters. folore in:por- ta*t f*r present pury)oses, they found that \rBM irad clissimilar effects fcrr different groups in the eleeiorate. 'Ihe.v fou*ii that V$}lt stimu- lated those individLrals belonging to sloups rvho were ali'ead,i liliei.v io
Llerinskr.,,1 COI\ SIlt)UEliCi :S {-}F ELECTi}RAL RLITORT\.'I 17 t)
participate irr elections-oidel" voters, those who are weil educaterl. ;ind those lvitir substantial a*rounts of campaign inierest. ,{s they conclude:
\'ilfuf surbilizes the electol'ate i* a ura-v that increases turrlolrt ove r the long run" It does this not so much hy mobilizins neu, votl:rs into thrr clectorate, but rather b3, retaining political participaut$ over a series i:f elections. . . . L{arreo,;,cr. tlre irtobilizaiioc pro*rrss r.i,orks more eftbc- tivelv li:r the rcsource ricl: than the resoi-rrce lloor. " . . In the end, the voters who rnove in and stay in the electorate look iluch like the exist- ing voters in tire s.vstem. By reducing man.,- of the direct costs of elec- toral participatiou, VB&{ alloivs the habitLral beh*r,ior of vorers a*d noii!'r-]ters to take center sl&ge. Those rve lvoukl expect to votf-the resource rich-are rnore likel-v to -custain their vote Gver {he long h;tutr. And those rve r,."'ould expect to abstain-the resource poor-are iltorr likel.v to sl*y'ont ol"the electr:rate over lhe icne haul. (;r. t93)
Though panel data ffiay he t}:e best way to Lrxamine tl:* riy*amic comp*sition of the electorate, such stridies are rar* dr:e to the cost cf acquiring validated vr:te clala. I-{ou.,ever" sturlies u.cing cross-sectional clata have co*1f, rt) .Dnclusiorls similar t* thitse al Berinsky et al. (200 1). lt is p*ssible to look at crass-seetiorr*l rtata to see the residuals of the effects ol voting refbrm. Specificall,v, hy crimparing [he charac- teristics of those whc: {ake advantage of novel torms of v-oting and those r.vh* do not, we r*n *saimat* the eftects of eleetr:ra1 ref<rrms in :i single electioll. Thus. alih*ugh rve cannot distinguish the efl'ects Erf retention tiom those of stimutration, we c;in see if, as Berinskv *t a]. argue, it is the resource rich p,h* are likel3; ta take advantage of il"le easj*r tbrms of votitlg. The eviclence from such studie s is consistent ;rnd unanrbiguous: Yoting refonr:s do *r:t amelicrate the resi:urce- trased biases of the electorate. Irl all cases, these ref'arms r*ii:frlrce ol increase such biases. I revier,v this eviiience here rvith respeet to lb*r spe cific refonls de sisned to make the act of voting e asier: Y8&,5. earl-y voting, liberal absentee ili-'rting. a*d -[nlernet lri]tin_g.
\iBh,T SYSTE,E{S
Like Beriruk-v et al. {2CIi}1), So*thr.r,ell anel Burchett (1997, 200i}) eramined the eitects of VEM in *regoll. Using an RSD sur\rey of vot- ers at the tin:e of the 11)1)S special election, the5; rlivided their respon-
48(i AI,IERICAN P{}LITICS RllSllA}iCH r'.I{ILY :i)f }.5
dents into three hroad groups; (a) traditional v'.oters, those vl,hr; r,oteil in the 1992 p*lling-ptrace election, had no pr{}blercs getting to th* p*11s. iind voted frequentlli in lhe past: {h} "'VBh,t voters," voter"s \vtr}o voted in tire 1996 vote-h-y-il1ail eiection but ciid not viltr in I !)92, lTa<i trouhle geitiilg to the polls. ol voted infiequentlir in the past; anri {c) n{xivoters, those ciiizens u,'ho dicl not vote in the 1996 r,oie:l-:}::::ail clection. Southrvell and Burchett ( 1997, 2000) fcund that "\rBfuf vot- ers" were stightiy older, nicre urbaii. ancl ]ess partisan than rver* th* traditional voters. Nevertheless- these trvo groups rvcrc morr sinrilal to eacii cther {han ihev rvere to ihc $onr/oters. Both the tladitirxal voters ancl the VEL,{ voters .vvere belter ertuc;ited, urore politieall-1, klolvledgi:able. ani,1 cared :uore ahrrut tl':.e outcornl: of the eleciicn lharr nonr:oters. r'r The authors therefor* e rme luded ttrat V.tlil.{ does $ct impriive the deinr:graphic representtti:reness r:f the electorate. Ktrp and Banduce i {2*00) cal11e to simiiar cc*r:lusicns rvitlr very rliilerent rieta. Using agsregate precinct*level tun:out rlata lirim three $regon counties matched rvitl: ri;l{;;. ircm the 1990 census. t}re_v tound that VB&{ ir:creesed turnt:*t hy Whit*, educated. older, lTigir-income vot- ers. \tite-b,v-mail procedures clid not increase turnftut bv minority or rlrral voters. Finally, h,'tagleb,v (l9B7i examincd VBIv'I elections i:r Califonrja, Or*gon, and Washington in ihe earlv 1980s iising aggre- gat* data anrl concluiled that in most mail-brallat i:lections, the mail vele{-{ closel-1,' resenlbled municipal polling place elecrols.'s
EAR'Y VOTING
Stein has str"lrtied the impact of ear13, .,,oting proceclures irt ?'exas (Stcil, 1998; Stein & Garcia-Monet, 19!]7i. Anatr,vsis of tr.r,o exit palls rvith early vixers and election-day voters suggests that tliere are no demographic rtifrerences in the trvr: populations. T'irus, earI,'- vcting prriduces iln *lectorale that is no *l${e ref}resentalil,e *t'the filll popu- iation than traditional election-du.r,v voting methods" Neelrt,and Rich- ardsou lztX]1) came t{} sir::iiar cc::clusions regarding the eff'ects of ear:ly voting using a 1996 telephone survey of registered voieis in Knox Count,v, Tennessee. The.v lbund that e;irly voters te*ded to ire rnor* politica]ly engaE:ed than the rest of tl:e *lectorate. That is, eariv r,'oters lverc p:u'tisans and those citizens :,vl"lo follow elections closel;1. Ffolvever, alrhough earlv voting n:*.v havc increaseel th*
llelin-rky i fiol\SEQLlllliCgS OIj LLECTOI1AL REF{)Rh{ 4ii1
asgregate levei of political engagefir*nt ilf thc v{}ting pCIpulatir}lr. it clicl not incre&se ttrre demographic representative*ess of th* etre.torilte" h'lultivariate airrll,sis found llo significant den:*graphic rtril fere*ces betr,veen early ancl electian-ciay voters" though bivariate analSrsis indi- cated thrt r.veaith,v voter"s used early voting at a higher r*te than otlter inccme gl'ilups. Neel-v ancl Richardson concluded that early voting sinrpl-v allor.vs those rvha rvould have voted on Election l)a-v to vote at il l))ore eonvcnicnt tirne.
.4Tisiii'{TliE V$?ING
Karp and Eandueei (2001) investigated rvh*th*r liberai abs*ntee voting larvs expa::i1 ih* electorate using a national sanrple (42 states) of National Eierti*ra Stud3i data pooled from flr,e successive eiections. The authr:rs tound that oicler. rnore er{ueated. p*litii:a}1.v actirre r..oters are fiiarc likelS, ti: r,,ote absenlee than other voters. 'Ihe,v also {*rend lhat m*r"e liberal absentee laws incre;rse participati{,}n among trv* l*r,v vot- i*g groups. students and the disahleri, but rlo ne:t increase v*ting r;ites amoilg minorities or independents. States rvith lii:era] abseni** lau:s hiirl higher rates i:f absentee voting than states that did not aciopt su*h proc*dures, br.rt th*y also had higher rates of in-person r..oting. Thus, the authnrs theorized that higher turrloui in staies with liheral absentee voting laws might l":e an efTect af a general paltiCIipatory pc,liticai cul- ture rati"rer than thc existrnce of particular absentee votiug pl"ocL'dure s" Ka;:p and Banducci's results therefore indicate that liberal absentee las,s do little, if an,vthing, tr ameliorate tlre inherent clen:ographic biases of the electorate. {n an iridividual-level anal,vsis of CPS daLa. *liver (1996) LrotrIles to the sa*te conclusion as Karp anrl Banducci (2001): votrrs u,ith leigher incomes aird more edur:ation are fi1ore likely than r>ther voters tr: take advantage ol'absente* brtrlot la*'s (for a rer.'ie'uv that comes ta sin:ilar lindings, see Hansen, 2001).
I}iTER]!F]T V*TTN{I
fuiuch of tlie p*tentiatr ilr:pact of the lnternel on v*ii::g is specula- tive. in future elections, son:e :nembers of the [-1.S. lllitritar_y ancl drler- icans living abroad may he abie to cflst thei{ haliot cnline.'r'Horvever. there is one prc*.li::ent example of wiilesg:;:ead lnter**i voti*g in a
4$l 41,{ERIC.I\N P{ILITICS RESEAIi.CH i'.llJLY :0{}.5
trajor U.S. election-ihe 2000 Arizona Demccratic Frimary'. Though Democrats in Arizcuil had only useri a primary systen"I since 1984" in 2f)00, a recclril nL:mbrr of votes lvere cast, inclicatir:g that Internet vot- ing proc*dures ma.v har.,e increased aumout. Alvarez ;rild Nagler (20*1 ) used ecolo-eical inf'erence *f'tulnout in tl'r* primary and fbund that *on-lVhites. the unemploved, ancl lural residents were less likely than cther vott:rs to Llse Internet tralloting {as rvere ra/G:re1l anel the elderll,i. They roncludecl '"futcrnct r,oting is likely to exacerbate the eu;Te ilt problem ci class-l:ias in American e lectii:ns if ir is introducerl aiiy t;me in the iieai: fiiture" ip. 28). Gibso:'r (2*01; also exa:xined the effc'ct of lntemei voting iir that salne Ariz*na *leetion. Like Anvarez a:ri"1 l{agler (200f ). she used courtl,'-lcvel dat* and conciuded that despite the irrcreasecl tumouf in the 20C* electiolr. the Arizona electr:r- ate r,r,,as ili) uxlre representatir.,e c,f the tuli population tiran previous elections. SpecifiealiS,', she found that lntemet voting i:ccurred ;rt a signiticantlv higher ra{e in eounties u,ith higher media* i*r-:on:e and highel percentage rf lVhite resirients.
SUI{&IARY 0F RESE:{RC}I
Alt told. using a variet_v of methods and data sollr"ces to assess the etlects of dift"erent r.oting refor"nls, a series of scholal's hal,e coi'l-r* tr: a single coi:clusion. Acr*s-s broth ag_rregate and inclividual-level clata sets. using both pane land cr*ss-sectiona] designs. ancl enrploS,ing exit polls, validated r.oae lecords, ancl telephane sulvrys" the resuit-q are e*nsistent. Individuals r.vlro utilize eas.v r;oting proce<lures t*::d to be inore politicall;, en-taged and interested than th*re rvho do not talie adviinta_ge of tirc opp*rtu*it,v (irr fhe case r:{ VB&1, earlv voting, and abse*tee voting). &{oreover, inciiviituals re,ho make use af -l*s*c r*stri.c- tive voting prmedures are better educated iin the case of Vtsl.,{ and abse*tee voting) and havr'e higher inc*:'r'ie s ifor al] r'elbrm"r,). ?hus" r,ot- ing relbrms rio not cLrnect the biases inherent in the electorate, anrl in some cases, retbrms may even worsetr tleese hiases.
1VHAT CAN *g r}{}}iE?
The elect*ral ret'orms *f the past three dec:ldes have not, rls s{}l:le p*liticians f'eared, led to a chaltge in the partisan composirir:n af ahe
Berinskr. i C()NStrQlIEli{lES O}; ELI1CTCRAL RIIF{}RI{ .is.l
eleetorat*. How'ever, fcr scholars and practitioners iilterfisted in increasing the rie*ecgraphic reprfsentation of the electorat*, this revielv sounris n disrnll note. Eleciclral reliirms. on their own, cilrlnf,)t an:eliorate the present socioeconomic hiases in iir* composition of the viiting puhlic because the increase in turno*t these ref'omrs bling is th* result of the retentir;n *f transi*nt v{ltris, not the stimulaticn of nonvoters.
h{ore*ver, the problenrs irlenti.fied lrere car,lrrct be solveel h_v :-:'r*v- ing thc foeus of ref'orm to electoral reeisiraii*n. Sonte auth*rs har.e argued ihal t]:e scr:ioeconoinic bias in th* comp*sition of the elector- ate is the resuit *f restrictive regisirarian laws (iliven & Cloward, f 989i. ELrt the e rnpiricai evidence enot's nr:t sL4lp** this r.,iell,. Over tnre past 3* vears. it lias hecome easier lor citize*s in the Urtiter,l .States to legister to \,ote {.see Highion l2i}0,11 for ;l e*mprehensiv:e rel,ietv ol these ciriurgesi. \tr/ith ililss;lge of the $l*tionel Voter Regisrration Act ii{VRA) i* 1993. most slate barri*rs to vcrting }vere relrl{}v*i1. This lib- eralizatior: has nct ied t+ g:"*ater socioeconontic equality in the cnm- p*sition ot th* electorate, as Fiven and Clos,arri t1989) predicted. Curent Population Surverv {CPS) ciata indicates that the propoffi;on *f registercd votels remained steady ftom I 992 to 200CI.'7 &{orcover" as tire ;tppenclix ciernonstrates. voti:rg refol'nrs dkl little if an.r,thing to ease thc socioeconorlic gap betr.veen rrotfrs a:rd ncnvoters. reeardless of r*gistration status. l['llese ilndings are consistent r.vith the balanc",: of academic ivcrk ecncerning the effccls oI r*gistration reform. Pri*r to the passage o{ the l\tr/RA, several studie s fcrund {hat iibe ralizing l,*ier registration lar.l,s might incrcase turllout brrt t*-ouLl have al*rcsl iro impact lrn lhe socioeel:n*mic con'ip*sition lrl'the electr:rate (&4itchell & Wlezien, 19951 \\illllngcr" rt R.csensto*e. lg8{i;.r' Stuciies con- ducteri since the passage lrl'the NVR.A ,vi*ki sin:il;u' ilnclings. For instance, Knack ancl Whiae {21}00) t'i:rd that the implenrenrario* of eiection-day' registration pl'oceclnres does uot enhance the participa- ti*n of lorv inc*me and low eduration citizens.'' Revier,r,i*g the litera- aure on the effecis of'reeis{ration lax,s olt tunlcut. F{ightrilr (20S.1} eor, clucies "Ther* is norl, little roam f'or cr*ancing tllrilorit turtlier b-v making registration easier" {p. -iI3,}.:r'{n sulr'r, the prahlem is not e.xpanclirig the size ol{ire pote*riatr elee l*r*te: ihe probienl ir in trming those potential volers tc aetual votrrs.-ri Focusiitg on the direer c*sts ol
48.1 .AI,{L.RICAN P0LITIL]S RBSEARCH i JI]LY ]00.5
voting-both in registration anil the act of castilrg a hall*t-in,i11 not get us \,-ei} {ar tor,vaiei a goal of truly representatil'e elee lr:l'ate.
Horvever" a pcsitive agenda can emerge from these resuits. The {ail- ure of eler-t*ral refcr*rs to produce a moi'e repr*sentatir:e electorate casts d*ubt +n t1:* possihility of ;reirievinq fuli *r1ualit1,- i:f political voice through instituri*nal refornt" Periraps, tl:en, institutional refbrm is ui;t th* proper response to the p;:ohlem" There is another sclutio*- one that inr,t:"lves reframing the puzzic *f participation. Reseai'*h r:n tufil$ua has consistentl.v shou,n that it is political intcrest and engage- ir:ent that cirives citizens to tlre polls and ke*ps thern voting fro:m elec- tion to electiau (Rosen-rtone & F{ansen, 1993; \,erba, Schlozman, & BrarJv, 1995). Tire studies rer,irrved tlere untlerscore tliis poi*t. PoiiLi- call,v engageel eitizens ire the ones r,viro lake advantage of li}:eral vot- ing rules. Th*se lacking p*liticatr interest remain nonl,tters. The prob- 1em, from the point of achieving ;r rlemilgrapirically reirresentative ejectoraie. is that pr:lrtical eng*gement cLrmentl,l' folklv,,s. rather than crosses" de*e*graphic <jivisions in the eiectcl"ate. trt is rvealthier, mcre educatecl ciiizens rvl"l* i**d to he more engaged ancl i*i*rmed about politics iDelli Carpini & Keeter, 19q5)" Thes* citizens pc-<sess the politicaily relevant resources tl"rat e::al'rle them ta rnake mra*ingtul iii:ks between their personal i*te r*sis anr-1 values or] oi'ie hai.:d and con- troversies in the politi*alw'odcl on the other. The costs oLri:aking these links ir* lhe renl c*sts of palitical partie ipati*n {s*e fierinsliSr" :0CI4). Thrrs. to expand the size atul the diversit,v of tlie eleetorate. $,e shorrld not i*st rarake it easier to vifle thr*ugh institutionan relorm-r.r,hich. through the process of, re tention. produces a larger elee tarate, hllf one that is riemograptrricall;- similar to th*{ l'*und i* eieetir;*s eoniiucteci using trarlitional methortrs *{'casti*g ballots. $,re must also expancl the scape of the politicali3r engagertr citieenry. Pul a**therlr'a,v, irrslearl of ma&;ing it incrementall.v easier f*r citiz*r"rs t* pa$icipate ir"l p*litics, rve slrouid rneke people w'ttftt to participate. This is a more dillicult and ]ess -rtraightfonr:*rd task ttrrln the in:plementiiiion of electoral reibrn:s. Ferhaps p*litical activists coulcl present political grievances
Bcrinskl i a{}}iSIQIlIl}iCES OF L.LECTORAL Ri:rI(}Rirl 4S5
ir: rvavs that eng;rge new groups of cilizens, or elites could reframe issues in rv..l,vs that speak m{}re rlirectl,v to ttrle coneerirs ol rhe ilisad- \iantii*led {Sehattschneirt*r. 1960)" Sin:i}arly, politicai leader-' anrl political partr','' lead*rs c*ulci expand efli:rts at n:cbilizatio:: i* lva;-'s that more clirectlv appe al t* itrre politic*l interests of orclinary citizens {Rosenstone & F{ans**. i9q3}. *efining a specific course of relbrm is h*,voncl the scope of this article, Lrut r*locusing politiciil refol'nrs to addr*ss thes* eoncerns c*ulii be an importalrt stcp in refornring the elecl*ral procrss.
tr\.rl:at is most irnpoltailt far ftrtlire researchr is that u,e adopt a mote catholic e **e*p{ion of the e *sls of political parlicipati*n" A focus orr thc' tarrgible casts af th* *e I *i voti*g has drirninatecl thc acallemic and popular r"liscnssirrn ol vcting reforrn lor m*re than 3{},vears. Speaking of the potential ol an1ine v*ting, Jesse Cordi* sums up the con\ren- tional lvisdoul i!'ell ll,her: hc ar::r"lrs. "People ahv;11,'s r,l,onder r,vhy tlo so ter,r,, pelrple vilie? Well, it's prett,r- obvions. Because it's harri" {A}thcfl, 2*fi4, p. 16). T}.re act of r.oting is indeed harci, but it is hard ir-l u,ays rliff'erent fi"orn thos* ider"rtifieri trrv political conlmentators. Rather iha* ibcus iln lhe eoncrete ccsts <if rtrting, rae must ir.,cns orl tlie less percrptiblr costs of h*:coming engaged rvith the p*litieai rvorld. Ti:e tr*e costs of partieipation Iie r:at.just in the expression of rrpinion l:ut alst'r inthe.{t;rw*ticrr of political opi*rons" Existi:-rg e l*,.:- toral refor:iis do nothing to minimize the cogtitivrJ cosls ass*ciated s,ifh voting. Thus, reclucing the r-lireet cos{s r:f opinion e ,<pressi** hrt'' *raking l,oting easier r,vill not, in and ol itself, neccssxily gir,* eq*i:l vclic* to all groups i* society. Certainly, ntaking the act cl'p*rtieii:a- tir:n as simple as possible is a r,vorthy goal, but ilrslitutional relofilis harre aaken us as far as they can to-uvard a rlemslcratic electorate. To l"en'leriv the dell*iencies of the American potritie a1 system and ini:rcase tur*out arrorlg hahitual *onl,oters" we mltst iceus retorm efil;fis o* increasing the engagenxent of'the electi:rate u,ith il:* p*litical lr'orld, thereb,v ]r:s,ering {he cognitive c{}sts ass<xiated rvith poiitical pa::ticipation.
36.--r
jlg.l 21.1
11.9
30.9 .+.1.4
24.4 1 ..1
36.-5
26.4 :Jt.5 12._1
28.2
41.5 :8.4
1.7
38.0
"1"1.4
!5.i 1-1.0
41.4
3r).4
l. l
"17.8
lo.0 2*.5
-!.1, 38.5
,1 1 .--1
1a..1
21.9
2-i.2
ti.'l
20.1
--17.7
38.s
-7.1
tB.4 24.{}
:7.7 29.t
r2.3
15.1
9.rl
19.8
ir.5 29.0 :9.7
10.8
-tf).5 41t.7
1{1. t
456 A},TERICAN POLITICS RESEARCH I JULY:{)0,5
APPENDIX Socioeeonomie Characteristics of the Electsret* {in perr*i}taccs }
Qmrtile Not Rcpi.gtctztl Registe rccl, Dkl Not Vote Voteri
199] Income level
First quartile
\concl quartile.^.\, r rr{d qu3ru le I.ourth quartile
Educltion Less than high scho.rl High -cch60l graduate Sorne college College graclirate
i996 ine+rce level
Fimt quartilc: Sccond quartile
Third quarrtile Fourth quarlil*
Erluct'rtion
I-.ess thrn high scht ol High scirool grii.luilLe Sonrt- colk:ge Colleee gmduete
200{J
Incr:me level First quartile
Second quartile Third quartile Foufth quartile
Educrtion Less ihan high school High school graduato Som,: cr-.llcge
College graduate
"_16.1 2{}.4 31.1 29.: ti .1 :1 .8 15.6 28_ ,1
l0.t) q.l 38.11 30.0 --1$.8 "50.5 3.: 10.7
SOIJRCE: CrrfiEnt P.rpulalion Survel: \rirter Slpplen.rent File" Novenber 1991. Novr'nrher I 996, Novembi:r f{J{X}. NOTE:'fhe co!unrn enries are the peruant;Ige sia pafticulargroup (1lot regisiere{llregistered. drd ndt \(ll(,i \iite(ll ;r1 ,, gi\Cn tilti,)Ctrrt)Onii|. CitlEt{1f-\.
Bcrinskv / C()NSIlQUlli.iCIlS {-)F alLljCTi)RAL RLI{)11\{ ,ifil
H*TES
i . Thc cnriliri*:rl rccord cencer*ing lh* reirresenteti i eness oltilc rttitutles *f 1xr1 itical partic- ipants is sonicrr.hi* mirrd. lYolfinger';r*d Rosf nstone ( l r80) use dlia frrxi j !)71ia lind thrlt \1]t- er-\ iurd r.r{rnvoters cio not r.liflir signilicantll irr their poiicl preleiel:ces. lising ilata collccted iriu.rcst tr.vo decades later. \'elba. Schlrrzn'ian. flrarly. iird N ie I i 9!t3 j e ome to * sinrilar e irrtclusiirlt. Othe r schuiar-r. l'tlruier er. Iravu ilund lhtl volers auil norrvclers do n{)l src c}c-.r)-d)re un lt-}aia€r:i rr i puhlic pclic;r. llcrinsk,v i2{X}2i 1rnds that non!ot.r\ al! rnore lihr:ra! tx social we}tarc policl' clucsti*ns tlia* r.re voaers- Siniilariy" Bennett and Rrsnick ( 1 990) fjnd that rloil\'oters are "sligi':tly
moie in i:iv*r of al incr*l-oeil gol'eniincnt role in the domcslii: xrena" (p. 75'.) j. l. {}i' cou lse" I i1m rl{rt the Jir-st io mtke l liis ir(}ini {see A lvarez & Iria5ller, 'Jl.it) 1 : I-.eightei, &
liaglor'" I 992; Highicn " 1004 ). trUhat is nor.,el ;rbuul tho argunrcrit prcldrric.l hLrrc is thc trvo-path
ri,rt.eflii):r ol' i:i1'n*nric t(liruul disrusseri l:rok.nr,. 3. Fr-r insl:i*ce" thc passil:i. i,.l tliu Nation:il Votcr Rcgisirati()r1 Act (NVRA) dicl n..rl
increascs iuiiiciit icvels as prr,,pixenis of ilie hill expeetecl.'I'he more intefesting que,ctir:n" iioni il.ic pcrspective olthis rer,:ielv. is u,irethcr this r*lrr:l r:hangi:cl thc cr:l:rpcsiti+n oi-ihc eleciorlte"
4. Fc:-mi anai,v-'sis, I e h{rst' llt exirll]ine iir+se reJtrrrr.ls design*ii speciiira!l1.,io eas.'t-he direcl dosls {-}!riasli*s ir balior. Aithlrugh I licus nn particiilar relirrrns here^ th+ criliilue arhtnccd hcre r:an be applie d t0 ixhr:r lelrrrrs designed hr tast ilti i'r:sls ill lh* acl ui [:1ir:g, ir:cl-rriirrg polie ie ' "1esig*ed
ic fni.rke votr! regisiratirn rasicr ise* bei*iv ). I Li+ nri sp*aL ic cther helistirg rL'li)rms sircli irs lhe vi:tiiig cquip*ient r*iilrms arldressed in the l1elp Arnericr Vote Act ol2iliil illA\,4) or prr,.:lsi.rnal {railots (those i-.aliilts cast in a polling place hv ltlter.s wh{}se namc-s !i. n+i appear r.lrl thr precinct rosier).
5. ftr srrrre j uri siiicLions. \,{rrrers c ar} reqilesi l{-\ lldf (}fiiri "pef l'nl11o!'}l itble n la-ri \'oltrs" si} .1 hal- 1ot iviil aiira.vs bc nrailed ti) thcir lionles.
6. J* 2ljt)-tr, lirr in-suncc, n:sidr:nts olTi:ras can cast carly ,,,ctcs heginning 17 da;"s brrfbrc Lhc eicctii:n and ending "i dals bcfbir Llection Da1, 1'Iexas Secreta{_\'of State !!'ehsite. littp:ll Er'r.,,.sLls.stilt..l\.us/elcctitrnsr'.,,oierlearlytr:ling.sh1ml. acc,-ssed ll,l;ry,,25, 10041.
7. Sei: alsu llenurdo ti980; rrnd h'agel and l\,lcNultr'(i!)9ir]. 8- Camphe l1 i i 9(r(r). it shilul.1 be noiuri. ilccr-.unted lirr lhij tvpes o1'churrring ol'lhu *iectinilte
c0ncid.rcd hcra. 9. l,rsing drtl firrrB the M7l- i 976 lriatioral lllecrions Studl Panc!, \,.,* c3r]
'lb.e r\'r ihij Vot-
ing llehlr'itrr cf 1,115 responalents ircrcss tlie tr.v* prcsirlcnti;rl rL-ciicns. Arn*ng ihese respon- ilents. in bot[i e iecticns. repodeil tlrn]4rit rl'ars l i:'ti:;iili, *nchzt:ti:ed, hr:iilirrg si"-;rri-v at lrhrost B0% in both eleetions {78% in 1!}71 ;tr.di<}':io in 197{r). }'{o!v+v!-r. .rm{xg {ht' sarnple. 1(r% olthe
'L'sporldtlrlls voted iir one clcelion. bul n0r tht filhgr" rr,herets 7{-}li ',,rtcd in hi:th *lcctiorts. rntl
I4% nr',,ei rilteil. Pane i dlrlir lir.inr ihc I9?2- I 9.15 e lectir.n cycl*s dcnrrrnsirxtcs a sirliilirr patiern. ,\mr;ng thc -545 rr:sponrlcnt-c rvho rcprirtc.l their r'*ii*g he liav!i;r in l:i-.th ! 9r]l ;rrcl 1996. tu nrul rvas fil?,in i9!)2 irnd 78%, ir 199(>.ln tiris sample. 71!,i ofrespondenis voi*.i in both eleciions" 171/l r'oteri in ou* elrcticn irut nut the dher. and 12i.f ue.;er rri*d. Alrhough this ie:,el olchurn ing ;-! n.)l as ltror)()uncerl ils tire lufi]ovirr toirnd in clrclir:il-r l(:l iorver r;iilr:c. liicr resrlts deiltunstrate fhaleveri in thu rrrr-.sr salicni arrd lrighestprclileelectirrns, registeied,.,iri*r: *-,,jr'e ir: ar:il irutrri thc alecti)r*tc lionr clccti0n iU clcctirx. h{*rrr in}porliint. in l:olh r,.f thesr citsLrs. liid irrl-rr:lrrlic* rrl' transir-ni vr.tri-c is largcr tl.ran thc populeiion of nonvoters. It si:ould also he nct*d thar these rcsnlis ar* not dependcnt on the use olsclf-i'cport,:d clata. ivhich research has shot,* ta bc sril:jtcr to thc cver-rep(iliir:g r.ll le.,els i;f tun'lr:ul (Presser & fra:.:gctt, l!l!il: Silver. Anderscn. & Atrtimsr--n. 1i186). The I 972- lATb Prncl crnlains laliii;rted ru{;ng dati} in irdrltiot ti; the solf: r'ti){rll rl;.rla. Amung lhc 1 .-i02 pane I resprxden{s rvho-sc loting bchirliur could irr: valiclatcrd. 85'*
488 Ah{ERIa,{}'l Pi)LITXCS RESLARCH 1'JULY :{ii.}5
voted in I !]?l rrnd 5B!li, r,tti:e! in I 976, again denlonstr;rtirg a high degreo of s{ahility ir aggreg;rtc lu ruoul. H r}!.",e\;ra. ci lhese rcsF,-rr(ieuts^ 77'% r'oitd in br-.th electioirs, 5%, neve r \'alted. aild 1 5% vr-.ted in *ne eleclio!.r. hxt ll;t the otlicr. Thr.is. thc patlem firr the vlrlidated dilla i! {he sanre as 1i!:' the reporied daia. thereb-v underscoilrgl lhe r-otJustrirss oi'th*sa resulls.
l {}. Sce Brod,r, ;ind S riid*rman { I 977 ) and Gcrbe r" {irr-cn. r*d Shachcr t l(.}{)ll tirr a disc rission *l l'rabinral paiterns or'\oting behavior
I 1. ln siim, rh* indirrdual 1*rel shitis in votine L,ch:rviortiret nftlct irgglegate trn-l1out ar. ah. result r:l'i.a'i)(]i-urin*t processes. I{egisterd nor'}\'ot()rs r:an tre brought inll} the e!eclorate^ as Can-rp'nr:ll { l!,}{i6) envisioned (stimul+tiou), hut transienl vot.rs car} illso he oncc-umged rr n:air:- tain thcil voting perticigr:itirx and -stu-r'iii ih* clccklra{c irctcnlion}-
I l. lJiicierstanilirlt this secilild path to increirsine lh. elccterltc is inrportrrnt" so I v"iil spend somc timr'crplicatilrg the pro':i:ss o1'stimuiati*u--tiro ectir,;lrioir cin,rt.,'oi*is.-. .xnd rcientiLll--- *nsnring lhal- t.ir'lsierr1 u--lcrs keeil coming ttl ihe polts *sing an irnalugv i-cnr popuiatirm dr.,nan:ics. liir::siC,-r l.ru gloups of ii-le san.lc ir.ririai size. If ih"- bi:'lh r-i:ies anLi iieath rales ilre thc same itr h(-:th 1-:irpulatiour. ilrt' t\4,{) groups r.;ill trc thc s;rnte si:c lte '.,citr latrr. Hltrier, if'Lltc birth rate of lhc -cecond triiup is highcr thirn that lf tlie iilsi. the sizc of the s*crrlcl gr*r:p rr'i1l inere;rse lclntivc ti) th* llrsi grorry aiicr rxc 1,car: This ccndiiion is irkin to iucreasing thc popula- tion of \ ote $ thi'origh an incr*rsc in the rate of st}r#lrli,iS rloitrciers ao ioin ihi e lectorate. Slmi- larl1.'. ilthe rleath rate tl'ihe s*ccnd group is snraller ihlrn that ol'lhe lirst, ihe size ofthe seconrl grr'lup u,ill irlso ini:rease relaiive ro thai. ol'the first grr:up. I'his condili,:rn is akin to increasing ili* -rizc r;i lh* l,'iiiing p{}Fillai;{)rr b1,' }ieeping trar'}sieni v(}lcrt in tht rrleclrrratc {iom electiun tr clcc- tion. Thai is" this -rxlcr-ts is akin tu ir,'irrilicn. Thr sirnlt pcili car be mriale .lh{-.ut ah,roiille iur:}out leyels lor;r siilele {:r(lup. li biith and delth rates are equal icr a give* gi'oup. then the size cfthe popiil*ticn rvill uot changc aiter o*c year'. llthciigir the com';rsiIi*n ofthe li.:iiig vrill.'l'his is ltin to a cliunring elecl{-rrai{e *'hcre iotal aggregate tumoul is unchirngeri elen tho,"rgh Lhe pcpuialiun
\v Ul r'oters does trhanlre lronr oloction to eleelian. Horvel*r. il'ilxl df:ith ralc (l*errr:isrl. ihcn Lhs-ro uiil be popuietion gi!)rvtll as hirlhs outrtumber delith-q. This is al;in tr-. increurius lugrcgri* itrlttoLti hr inirc.l.u)g r(lirlti,lt).
1 3. l-'ollor.r'ing ihe lead of lVbitinger ancl Ros*nstone 1 I 980,t. most rcarlernic.r argile that sui:h chanl:es.rill nol.hafige ihe piutisan balance i:fthc *lrlctoraii:. Reccnt r.vork. horvever. hls icund ihai increasing- turnout iu the Li.S. rniglrt change the outcome rll'eleclions {see Ciirin, Schiekler-, and Sid*r f2t)t)-il r*gar-ding uatir;n+i elections attd Hirjn:rl an{i Tr-iiuirslinr fin ;.u'essl regarilinr Lhc p1)tcr)tially scilcus ir--ns*iluunces ti)r !oael elecLions r11'cl:rss an(i ril.e-based irr,thal*ncer il Lrit:t-r- ctrl). Politiciirns" an llic (iihcr hand. 3.r lcss conliclcnl about thr: naturc r..l'!:utisan ctl,:tls. f.r ernnplc, Rcpriiirlicrtn Phil Cnim sai.i p:rssagr of IlR2. tlrr- grrecurs+r to the NVIiA. u,ould "tilt tht: polii ical sl,stcnr^^ tri the b*nclii *lthr l)emocreiir: panf iiyuotcd in {Jalr,crt :iix! Cilchr ist I I !j!}.1 I).
I.1. IheseconcirisitusaredrarvnlitrnLrir,iriilte[niilvsisintheIq!i7pieco.-lheseditierences hlld up in thr m*ltil*riate c{}*ar;{t :rr; r',,ell. dlthough rhc ;ruthors did not rliri:ctll corllprre thr thrcc groups iii such ir i:irntr:ii. logil alrll,sis in thcir 10{i0 piecc indl{3irll that although liting- b1,,-nitil IYBNl t vi)ter-s and trilditirlnal .,,c:ie rs rId not riil'ie r on e duc:r'Lion :r*ri iroliticai kno.wlcdge" VLIN'i i,oters aucl registcred norr\,(llers dillered on thcse twi) quantitirs.
1 -5. In diau,ing inie rencos ol'ii.rriii,iii*al-ievel behavior frixr aggregate d;riu^ these siudids sul: ter l'r.or.r.r pol"-rtial ecclogii:al inleience prr:blen:s (l{lbinson, 1i)50: for iilodrm ircatnlcnis" lcd Achen a*d Shi..,ei"v [19!51 and King f iri97]). Ho\t,ever. ti]e cur]r'er-gence o{ the lirrrlings o1'thr' VBI'I studies usiii,i. rlggregate d.1ta :ind th*se using indii'idual-lcvcl data -suggests thai ihis 11au, is not ia1al.
I 6. tlnline i ciins ilr cl erseal i::iiit:rr;7 pcrsonnel r,,,as Flaiureil lbr 2{i*"1 tlirrirgh thc Pcnia- tiru-run Secui:.- L.l*etlonir: Ilegislr;:iirrn anri Voiing Exp"-r'iinent. Ihis proglarn rvas canceled in Fqbrurrrr. I'l{}-+. Jlts irr ccluri!-r rufl( i*t..
Erri n-r k1 l a{)NStsQlJE}iCtrS {)}' ELFICT{)RAL REF{}R.ni
17.'f he rcgistraticn ral* in thc Llnitecl States r.vas 78.1','i in 1992, 77.1% in 1996. aril 78,.11;t ni 2{!{}i} ar:ctlding tl mrv *nai. sis itith.- Cuuent Pcpr:l*iiix Sru:.,ev (CPSt di:iii.
I 8. lt :r sliglrtli di fl rrent r,,eir. but one {iriit srjfl{)(-\ the fi ntlin5:s * i' Lhis article^ 1\ ci iing*r;rnJ Higirton (19!)8) iind ih{il th* adopiion.)f a r})otor-voterprogran} in Coiorario incrcased fiu-n+ul c.,,crirll" but tircsc gains c:inrc prinrerill,iinrng thosc irith high schooi dcgrccs;ind sonrc ct;llegc eclui:ntiix. 1'h.- glin;.ivere lorvest anrong iir+sc r.vith icss than a higli sehcol *duciltion.
19" Sp.cciticaill,. *sing CPS rlata lioni i !)!)0 and I g!)4. tlie v iinri thirt strrtr:s thrt ;t"lti1xr:d LDR i*rproied eilu:riiili ol reprcstlrlt:rti$r l-1y age and rllbiliil slltils. t{r}!vef *. ihey fcund no dilTcr- dnire itn urlrieati{)n ;rnil incoirle {r.rnee tlti ptevi(}us lcvel ol'state turnotli tvas in(rlrduced as:r cr:n- troi r,;rriirblel.'Ihel ccncluds thsi "ahllsr,ruplorting lrcgisiraiion] relbrins in th. hopts ol' iniproving piuliciiisii(-.n ain*ng the pocr irnrl l*rs *dLrcetcd (groups. rcasoi:al:!;,,. allreil ovct-sint- plisticalll:. rrssociaiecl r.vrth fii:erai anel L)emocratic L1a$: carjs.s) mal,Lrr clisal:groini.-d b:; cur finilings. . . . t)ul resr.rlts tlo not suiist;rntiaie ihe ci;rir-rr that pr*-FiVRA administmtive bairie;:s lo r-egis1,ri*g aited as 'de lir:to inr:cnre ;rnd lii*rar:5' tcst "' t100i), p. 41 ). S ini ilarly. in a stud;, ,-rf rci- ing; i;rh*r icr i n sLirles !vith niinirnirl .esldcLi(rni on rcgislration, Highrr:n ( t 997 ) linis liitb :ult- lioretiix of thc strcirr*cononric -sl,rrv ai'the e !ectlxiiia corlip,irir lil sralas r.,,iih mcrc rcstrictirc lau,s. H* ioniludes th,r1 registration iaivs. "do not eili]ait. to llr tha main reason ibr ihe sociiiecrr- nomic sLcs, oiA!.no!-i.sn \ oters" { p. 573.t. Ilanmer's {2ii(}-1) stur!y oi Llll{ rnd motor-uc,te r rcuir- triltio* .i)nlt'J to a sii:iilar conclusion. i-ie finds th:rt while ih* l,-:rsl eihicat*d citi,:eris :r1-.pear-ln cain the nrr.lst 1'rom easier regislration prci:edul*s. thc overrll dit'flrei:ces licross education level: *re c;trite srrrall. Hlnr*er conclutlrs that f hese regislr-irliorr reiirrrls dc. lit{ Ie kr redur:* in<xlnalit-r irr tiirnout- lt shoulci l,.e notr-d th:ri ihc *rldencc is noi uniinirnilus an this poi*i: rrthcr studics havir acidr,-ssecl the ctl*cts of rcgistriiaioii re form ancl llnd r:r'iclcirce thai is scinewirat mired. Ah,;rrez. AusolrtL.':herc. r*ii! lViisan i2llilii rrna!.r,ze CPS rlata ticm l{i0(i to arLir*ss th. couutorf:ictiral .luestion .rf hcli, ihe cc,ilrposilir.rir ol trhe e lec{ornt.- s,rni ld r:hani:e i I EI)I{ rvas udqrted in a! I slutes. Tlr*.r' conch,rile tirat tilrnouL \','cillil irlcreasL' across Llie bo;rrtl hut F-f)R worild dinrirrirh thr *ducr- liorr inrhi'rlancr i* lhr ci:rnpasition af tlie eleciorair (lhou!!h ihuir analv-rrs irl:r inilicate:hal lhc inclnt str:rtiijcaiiitn o1' thc elcction rreulcl pr:rsist).
10. -Iraugr,-tr-
{2i}{}-l} i*r'i*rvs ihs lit.ratlrre cl1 th. riTcLis *ili}::r':ilizcrl ri:gistr:itio* lar-,,s airil f,nils thi:t thesc retorms have no apprrciable r:ifcct on the sociocconi.rmic conrposition of thr eiectorate.
2I. The slr;ic is iruc lf ihe :erios oi'electoral re hr:ur rrriildaiod brv HA1A. buch as vilti!r!l technolr:gy ir))ffi-\\,eirier)ts ailti Llie r:rcatiorr ol're:rlraiiuod lrter regi-rlr-rlicn 1isl.s. Thesc relir|nrs erc in4rc:-i;uii in thilt tha,! ian hclp rnsurr thiit rolrs rast iil pirriicular rlectir;ns r*eorcl lhe intencled vot*: ari citizens. Blt liaeratu!-e revierverl ir*r'c suggcsis ihat r:vcn a c+mpeiril;nin of impr.rvi:d eiec{oral institutiorls rvill not significlntly ehange the ei:rnposition rrfthe elLrctoratg.
RE!-FIRENCI.IS
Aclrr*. {1. H.. & Shivt!y,W. P. (lqq--i). {l;ttss-it:t'ti iiili:rtticr:. Chic:r:lo: l,lnivtlsit:' ol{lhicega Pre s -o.
.41r,3162" R. L4. L.. Ansolah*herc, S.. & \\ilson. C. t{. (2002). flrciicn riat i,oter re gis[ratioii ii: tlt t Li niteti Str:tcs: H ow ot!t - \r?r1 \:.tit g tLt.ti clnilSe lhc ,:ottpositicn ol t!1r An!(!rbiiti t'!tt tor- air,'. CaltechA{IT Voting Technr-.logi f'i-trject \\brking Pap*r.
A lvare z. I{. l'{. I -." & Hall. 'I- F.. (2{.}il,+ }. fi)nrl, i:iirk, itirj .,,rtr<:: Tht.furtu'r: ttl ittt<:rrrer rutjrrg. Wash- ilrrl',n- I )( : l{i .,1:ltr' l})\ttt{tli.tt.
49{i A\,II1RIC,AN POLITICS RESEARCH I JUI.Y 100.5
Alvr,-2. li.. h{." & Nagler. J. ilti01). 'i'lr iikell constq*t:rc.s ctlilt.rnet ,.,{ting f cr p+iiticsl rr:p resenlaiion. {.,11i:7i!" l-4'' Ert,ictt', -t-i. I i i5-1 1 5-1.
AhhoiI. S. r1004. llfareh 7,r. P*inl. elich. elecL. Eri.stoti {iiitbr: ivJ,rgl,.iti{:"p.26. Benne tl. 5. F,.. & E{esnick^ D. r 1990}. The itrplicatiun,s oi'nonvoting l*r domoeracy in thi: l-Inited
Stai*s. ,i;r:arlla;t Jourtni i| PoLititoi St'ierire^ -l1.71 1-St)2" Berinsky. A. J. (2001). Silent vrticcs: Social ra,eliirre policl opinions and political eqri:rlitr in
Ani*ricer. ,\rtta icari Jou rntt! oj' Poiitica! Scicite t. 16, 2i 6-28 t- . Fterinstl., A. J. t1004,t. Silcrrt lclcls. {};tiiiion po!ls ond Ttitiiticri particiitati*:i *t Amr:ri..tt.
Frinculorr, N.l: Princeti.rn {.lrir*rsit,v Prers. Bcrin-rk1,.,'\..I".N.Burns.&Tr-i'rugoti.I,i.\\r.(20i]1).\\'h.)\,otrsb!. miriI?;\clynnnricrn*il*l ol thc
incliviclrial-!ci,el consequellc.s of l,ilie -b1,-*r.til sy-1ie*rs. Frbik ili;i*it;n Quttrter{.t'.5.5, 17s- 197.
Btotl1,. I1. A.. & Sniilerinail, P. i\,1. ( I.177). From liie spaco to lxrliing plar:e: ')'he relelance r.ri'pe r- sornl concerns lor ictinr: behavicr. Briiislt Journul ci f'oiitirr;l S,-icnte, 7.331-3$0.
Caiirrt. .T. 1\'. " 't Gilchrist. .I. ( ! (i9.1).
S upf ose lhe,v held ;in eleclion and almr:st cr'*ry'but1,rr canle I PS: Po!itk:ai Sti?tii:t itilL! &tlirit;t.26. aY-{-i00.
Canrpirr:ll. ;\. i 1966). Surgr: and dcrline : A studl'olclcctolal cirirngc. In A. Crutptttll" P. lr'i. CLU- yerse , W. l,,lillet. & D. Stokes qEils .';. Eie,:t itus <ltt! iitc t'oliti-ct ardt,.(pp. -tril-61). i',terv lbrk: Wiley.
{litrin^ .tr." SchickJer" B., & Sides, i. ( j'103i. \Vhat if ever,vone voLecl? Siilulating the impact of increa-r** tuflrorit in -raniitu elecliolls. ,\tnerii:ttit.loii*rci l_i Folitital Stisti<'e. 47, 7.<-90.
Ilr'l1i Cer-piri. fui. X., & K*ttlr. S. {.199{:). \lhat.,1*it:ri<:.tti., kfio11 {i!}i.;iif 1.tt!ilit.t l;u! r1,1ry il fi?.{tt- te;:. Nera,Have n. C'l': laic Uliversitl Press.
lleNarrlr;..1. { lqii{,i}. Tr:inrrut lnd thc \'ot.:'l'he.1oke's on the L}*mocrsts. -4i;;,:r"iriiit !'oiiticcl .\'-'i- i'ai, r' J{ai rr'}1. ,'4 +tI{l-+lr}-
Gerher. A. S., Green, D. P. & Shlchar, R. (2i]{}3). \i}iing rral,bc habit-ibrn:i*g: Evirlenct l-nrrn a ranilontized liuld expcrinterrt.;\*icrit:ln Jour;ii::i r1f fu;iitit:ol St ittii:;t. d7. 5iltl-.5--50.
Gibson. R. (20t)1t. Iilectioli-r r-.nline: Asse ssing intcrn*t i,oting in light cllhe Arizrne dcmuciltic prirrlrrry. Ptt ! i t ic tt I S c i ? t tc { Q.uti t i {: rl,i; J,r {>. -5(r I -5 Ii.
Ilzlinal. 2.. & Tn;iinstine . .1. t in pr*-.s;. Wh':re aurnout mattels: Thr i:oiiscqricfice,s +iun,'ven turn- rxit in cit1" politics. Joruriiril rifFriiiii<'s^
Har:nrer. X,t. .L (:*t)4). Ftitn s<:ii:tiiait to t!tt'ticn ard btrond: Un<!e rsl*r.ilitt,1 iiis <'rsit:;t:.s tttti t;.fist'eueti{ t}s .f el*:ktra! rs-lisfitt ifi .4meri:*. {lnpuhlisl.u.:C ili:e i{)r:tl iiissollrliUtt. l-T*il'trsiLr o1'N'Iicliigat.
Ilanseii. J. h'1. il{}{}li. iili{.', rrtii:g, ilt!lTslt-i.!{{i Libs?it!!? ,.,oiir1.i, *riti i,ciititr li.r'rr;*il. Llnputi- lisheil rpart ofthc'lhsk }-orce or1 thc l;rdrrai Llr.tiari Syst.ni.
il ighton^ Ii. ( 1 997)" h.asy legistrnti;-.n ailti voter turnou L. Jourliirl e f l:'oiirics " 59. 505--s 7-<. Hig!:trrn. R. (200.1). Regisir-:rtion anii urting in thc Iinited States. Persyitctit'ts an Polilli.'s, l. .5tl?-
Ki..rp..I. A.. & llanducci. S. A. (l{X.}{)). Goi*g postal: Horv iril-n:ai! clcctions intlu*nce tirmout. Pi; I i i ic a i B e i rctt' i {) it 2:, :2.\-2 3.}.
Karp. J. A.. .t tla*ducr:i, S. A. i2{lill ). Al.seittec voting^ niohiiizuricn" end pariir:ipation.;r?( /'l' ciut {>olitii:.s Rcieaich" 24, 1,33-19-5.
King. G. (1997). ,.1 sollltion tc rlu: ci:tllgirol iitfrrr:iri.e rrr;itk:in: Rerotisttiittir? inli,itiunl bcirtttior t'tr;;tt .tiittt!.1!{!r:ri*fa. Princ*lix" I{.1: Princeton {-hivcrsit,v Pre,l-r.
Irnack. S.. & 11hite ,.i. {2{-}i}i}). Eirciir:ii-da1'regi-stlaticn and t*rntut ineiluitlit-:-. Pitlitic'tti Beitttr,- iar.22- 2*-44.
1..*ighlt:},. J. F.., & ll'agler'. -1. ( 1'.)i;l). S+*ioectni:ntir: +iass L:i;rl in tut::+ut. 11]{:4- 198&: 'fh* vtxers remain the sami:. A;relir:oii Poiititg! Stie nu &r'',irr,,, E6^ 7:,s-7--14.
Eerinskri i {X}NSE(}UIINCES {)F ELF,CT{)RAL REFORh{ 491
N'l*gl*br,, D. ( 1987). Partir:ipation in mail ballot eler:tions. \'Ve stcrri Poiiticcl QLiartril 4/-;.7i)- ql.
N{iLcltrll, G. E.. & 1\.rlir::i*rr. C. i I !i95 t. Tltc: intpacl r>1'1c5Ja1 e i;n rLrri* i\ 1)n \ { )lrr risislr,rti, rn. t unt- r-:rrl. and the cr-.mposiiion ul'the A!.l.rcriciril ElecLr:raie. Poiitii:*! Eaiirii'ior 17. L-9-:{}2.
iriagrl" J. H-. & \.IcNult1,. J- E. ( 1996). Parti-can e ll.ecLs oi..'otel turnout in scniltorial iind gLrbcr*a- torial elections. Anerican Po!itical,9cience Rev!ew, 94. 78(.)-79--!.
Neel1, G. W.. & ltichardson, L. E.. Jr. (2001 ). l\jho is ealiv votin*'.'An individual ler'""I cxainina- ti t>n. T k t S <; r: i u.i.ir iu u' c J itu fi .ri, -1E. 3 8 1 -3r:.
Oiiver, J. E. (1996). Thc *flccts of *ligibiliLy restriclioDs and parly lclivily on al-.senltje'.,clirly irrtrl irversll turnout. Arnar-lran Jaurnai i:;f Politii;al Scienct.4fl. 4{)S-511.
Piven. Ir'. F-.. & Clorva|d. R. A. ( 1 989). Goverilrncnt statistics and contlicting expcclations +l'fion- votinc. P,5.' f'oiittcitl Sti,ttice anri Pitlitic.s,l:. 578-589.
Presser. S., &'[iaugi]it" L{. (li}1}2). I-ittle white lies an'l :ocilrl scienc"- models: {i-.rrelaied reslroos{r en-ot-s it.} a panel srud;. oi'\,otiitg. Fttbiic Olsit'tiLtrt Quartar!1," -15, 77-86.
Rohinson, \.\,i S. {195f1). Ecukrglical eoneln{ions and tho bchavior ol'indivirluals. i\tnerii:nn 9oc io log ictt ! Ret'leu: 1.1, l5 1 -.157.
Rrtsenstonc. S. J.. & H:nse n. J. H. a I 991). M obiligi ir-tt, ystrri it iS:taiirtn, it:iri iie nirx; rai:v ir i .\xtt r- iirr. Nerv \brk: l,lircl,lillirn.
Schattschneic!er. i'-.. E. ( 196{il.'l lia sc;tiisr:,-eteigil Tttoplc:,4 rea!i.st'.s vie]..,: o.f t',ctttocrtti:t'ittAtner- iiii. Fofi \\'rrirh,'tr'X: LIariouil Brace Jovanovich {bliege Publishers.
Sigelman, L., .t Jeivell. }{. F.. i ll)${it. Frun: rorq l{r peripher3'; A rr--Le on lhe imagerl l-rlconcu*- tric clec'.ilratrs- .iai;rnal ri' Rtlitic.r. 45" 440-:tr-tr9.
S ilver. B. l]., A*de rson, ts. A.. & Abrar:rsal, P R. ( 1 a86 ). \Vho orerrepoi'ts voting'J Tiie A//,./ l- c it t Pa ! i i it ;i i 5 t i t n t e ll et i c vt,. 8t]. 6 1 l-{:1"1.
Southweli. l'." & Burchett. J. { l q97). Survev ol vote-bv-n-raii senate election in ihe stafe (t'L}re- son. P,t.' Pttiititci S{:ir."ilce cfii! Pt-i1i;ir'.i',:i0. 53-.57.
Souihrvell. P^ & Burchutt, J- i:i-lt.}i;}. I-)oes chi,Lnging the rules ch*ngu the players? The elIecL o1' all-mail elections on thg colnpositio* i:f the r:lcctorilte. Suti.ol Stittu:e Quarterlr.8,/, 8.17- $.15.
Sisrn. R. N'I. (1998i. Earl1,r,r-,ting. Pttblic {}pinion Qu*rrerll ti2. -\7-6r. St*in, Ii. N'{.. & G;rrcia-lvlonet^ l-.}. A. { 1i)97 }. loting early. hut not often. Sot io I Sc icnt e Quarle r!'t ,
7.^" 657-{;71 .
Traugolt, l,{. \\i (200,:l). \\.hv electorill relornl has l;rilct1: I1'you build ir, rvill the1, coil.re? ftr A. }i. fldgler" l.'I. R. .fuist. & 1r. &IcCaflir"v {Eil:i. ). Rsr./,li?iin{! t!1e ro!e : T!rc paiitit's anii yr.)siti(t.\ t ii .liiierit{lit eiectictti i"tfbrnt. N.-rv \-ork: Oxford Llniversit}'Pres-q.
ll'olfinger. t{., & ttighton. 13. t 1 9!}8)" Estim*ting thc eflects ol thc ilatiLrnal volr-r rrgistratiorl iurt i,l l !it]--r. Po i i r it it! E c h o v ! ti r. 10, 79- 1 04.
\i'oilinger^ R.. & Rosenstoue. S. .1. i 198*). t4rl111 ynlEs J lrerv Haven, llT: l ale l.iniversit), Prcss. Verbir. S.. Schloznran, K. L., & llrarly. H. E. (1995). \itit:t cnti uiuuiit.v. {-ivit t,olmttari:ni itL
,\t i rc ric ttit,:.,ulilils.Cirnrhridgt:, L,fA: Harvald ltnii'crsit)' Frcss- Verba. S., Scirlozman. K. L., llrady, ll. tr.. & Nie. Ii. lt. (1991). f,litizrn rctivit],: Who pafii.i-
p:iles? trVh:it do the1, say'?,1r;et itart Po!itical Si:i<,ttc( R?vittr,8. 3t)3-3 I E.
Atiant ,1. Be rirtsk| is an ossit: iLtte ltiitit'sso r tl' Pi;iit itttl SL it nt:e iti M lT- tle htts 1;Ltlsi i sirci! lrtiL:ie s in i/r r\n:cricirn .ir-.umal of Prlitical Science, .Iournal ol Politic,r, Politic;il fsy- cliologrt cirrlPublic Opinion Quartcrll,. Hr is tha uuthor ofSilcnL Viricr:s: Fuhlic Oplniol artd Ptrlitical l{epresentatiou in Americl (Pirti:tlcn Utit.r:r,ti?, Pres,s, 2{}{i1).