Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
B-240963. November 2,199O : The Honorable Al Swift. Chairman, Subcommittee on Elections. Committee on House Administration. House of Representatives.
Typology: Schemes and Mind Maps
1 / 84
This page cannot be seen from the preview
Don't miss anything!
I --___-- IJni1.d^ States^ General^ Accounting^ Office
142582
GAC)/I’EMiI-Yl- 1
.-.. .----- ..^^ - .-....I... _.- .-.- -.- lll-.. ll--..“. -.“.^.---v”-. (^) I!
Executive Summary
Compared to an averageturnout of 80 percent in other democracies, voter turnout in the United States is low and has been declining continu- ously since 1960. In the 1988 presidential election, only 60 percent of those old enough to vote participated, down from 63 percent in 1960. Turnout is even lower in congressionalelections-only 36 percent voted in 1986, for example.
In view of this situation, the Subcommittee on Elections of the Com- mittee on HouseAdministration asked GAOto provide information on how participation in federal elections could be increased.Specifically, the Subcommittee requested that GAOidentify the practices of other democraciesthat encouragevoter participation. Additionally, the Sub- committee requested that GAOprovide information on state and local efforts to increasevoter turnout and, if feasible, to identify particularly successfulpolicies and practices that could be the basis for federal action. At the request of the Subcommittee, we did not fully evaluate registration systems.
Someobservers have attributed the low and declining turnout to increased levels of political alienation among U.S. citizens. Others have noted that electoral policies and practices make voting difficult in this country compared to other democracies.Still others believe that, rather than a consequenceof political alienation or procedural barriers, low turnout is a reflection of the low level of organized political conflict in the United States compared to other countries.
Prom a review of the relevant research, GAOconcluded that the compar- atively low American voter turnout is not the consequenceof political alienation. Rather, the evidencepoints to international differences in the characteristics of political parties and election procedures.Political par- ties in the United States are less closely aligned with demographic groups than are parties in many other democracies,which may reduce the significance to voters of election outcomes.Moreover, election proce dures in the United States make voting cumbersome.
Investigating the interstate differences in voter turnout, GAOfound that three election procedures are associatedwith higher levels of voter turnout: (1) registration deadlines that fall on or closeto election day, (2) toll-free phone numbers that allow a voter to request an absentee ballot, and (3) the practice of voting by mail rather than in polling places.
Page 2 0A0/PJMw1-1 vothlg
Executive Summary
Contrary to popular belief, voter information campaigns generally are not associatedwith higher turnout. One notable exception is that states that mail information about propositions and referendums to individual householdshave higher turnout than states that do not provide this service.
Finally, someevidencesuggeststhat two actions could slow the contin- uing decline in voter turnout: (1) providing toll-free phone numbers that allow the voter to find out if he or she is about to be purged from the registration rolls, and (2) staging mock elections in high schoolsusing actual voting equipment and materials.
With respect to electoral procedures,the United States differs from other democraciesin several ways. In somecountries, voting is manda- tory. This seemsto increaseturnout even though specific penalties are rarely administered. More importantly, almost two thirds of all democ- racies employ a system of automatic voter registration, whereas in the United States the individual voter must take the initiative to register prior to a deadline that could be as long as 60 days in advanceof the election day.
GAOfound that registration deadlines are strongly associatedwith voter turnout. Basedon a statistical analysis, GAO estimated that adopting election-day registration could increaseturnout in somestates from 1 percent to more than 12 percent in the state with the earliest registra- tion deadline. As many as 37 states might increasetheir turnout 6 or more percent by adopting election-day registration. Further, in states having registration deadlines well in advanceof the election, informa- tion campaigns designedto educate voters about registration deadlines, registration drives, or places and hours of registration were not associ- ated with higher turnout. More extensive information campaigns,there- fore, are not a substitute for registration reform. GAO found that the availability of toll-free telephone numbers was associatedwith higher turnout, or with a smaller turnout decline since 1980. Specifically, the availability of toll-free telephone numbers that a voter could use to request an absenteeballot was positively associatedwith turnout, and the availability of toll-free phone numbers through which the voter could learn about the intent to purge him or her from the registration rolls was related to a smaller decline in turnout from 1980 to 1988.
However, the procedural changethat could have the most dramatic effect on turnout is the all-mail election in which ballots are mailed to
Page8 GAO/PEMD-Sl-1Votlng
Executive Summary
While GAO identified the opinions of responding public officials and out- reach organization directors on what they believed to be effective local practices to increasevoter turnout, a major finding is that little empir- ical support for these beliefs exists. The outreach organizations GAOsur- veyed did not conduct evaluations that would enable them to substantiate their views. Moreover, the beliefs reported by officials about the effects of voter information activities seemto be contradicted by GAOfindings that voter information campaigns generally are not associatedwith higher levels of voter turnout.
Basedon the findings included in this report, the Congressmay wish to consider making voter registration more convenient, such as by adopting a system of automatic registration. In addition, the Congressmay wish to explore the feasibility of encouraging efforts to increaseparticipation through the following:
. assessingthe appropriateness of all-mail balloting for federal elections, paying particular attention to the differences between all-mail-ballot elections and conventional polling-place elections in turnout, cost, fraud, abuse,and public satisfaction; l using toll free phone numbers in each state and the District of Columbia by which the voter could (1) request that an official absenteeor mail ballot be sent to his or her legal residence,and (2) obtain registration information, including the intent to purge the voter from the registration rolls; l mailing pamphlets explaining propositions and referendums to house- holds of registered voters; and finally, l placing polling booths and other materials in high school civics or other appropriate classesso that students could conduct mock elections using authentic equipment and materials.
Page 5 GAO/PEMD-91-l Voting
Contents
Objectives,Scope,and Methodology Study Strengths and Limitations Outline of This Report
International Comparisons Why Is US. Turnout So Low? What Do the Higher Voter Turnouts in Other Democracies Mean for the United States? Electoral Participation in the United States Summary of Research Matter for CongressionalConsideration
Election Procedures The Experience of All-Mail Balloting Conclusion Matters for CongressionalConsideration
Voter Info~ation ad Registration-Related Information
Voting-Related Information Issues-RelatedInformation Voter Information CampaignsTargeting Youths, Minorities, and Women
Voter Information and the Decline in Turnout Since (^1980 ) Are Voter Information Campaignsa Consequenceof Low (^66) Turnout Rather Than a Causeof High Turnout? Conclusion Matters for CongressionalConsideration
Specific Practices of Local Outreach Organizations (^61) Conclusion (^71)
Page 6 GAO/PFND-Bl-1 Voting
C!att.enta ,
Table 1.2:Multiple RegressionAnalysis of Election Proceduresand Voter Information Activities on 1980- 88 Turnout Decline for 60 States and the District of Columbia
Figure 2.2: U. S. Voter Turnout in Non-Presidential Elections, 1960-86.
Figure 4.1: Frequency of Voter Information Activities for High-, Medium-, and Low-Turnout States. Figure 6.1: Registration-Related Activities Assumed by Local Outreach Organization Directors to Be Effective in Increasing Voter Turnout Figure 6.2: Voter Information Media Assumed by Local Outreach Organization Directors to Be Effective in Increasing Voter Turnout
Figure 6.3: Informational ThemesAssumed by Local Outreach Organization Directors to Be Effective in Increasing Voter Turnout.
Figure 6.4: Get-Out-The-VoteActivities Assumed by Local Outreach Organization Directors to Be Effective in Increasing Voter Turnout
Page8 GAO/PEMDBl-1 Voting
d
Page 9 GAO/PEMD-91-l Voting
Chapter 1 Introduction
In this environment of concern for the continuing decline in turnout and limited knowledge of effective methods for improving voter participa- tion, observers have noted the need for research addressingthe opportu- nity to vote and the role played by voter information and motivation. In recent reviews of voter participation, for example, both the Committee for the Study of the American Electorate (an independent research and advocacy group) and the National Association of Secretariesof State have called for consideration of federal action to support voter outreach efforts and research on promising election practices at the state and local levels.
Efforts to enact legislation to enhanceaccessto voting have been common in the US. Congressin recent years. However, with the excep- tion ofI-I.R. 2190, the National Voter Registration Act of 1989#these proposals have not succeededin advancing beyond the committee stage. Resistanceto proposals for reform in this area appears to be broad and basedto some extent on concernsfor maintaining the security of elec- tions from fraud or abuse,and possibly becausethere has been little public pressure for registration reform. Resistancehas also been noted in the caseof elected officials who may seek to perpetuate the adminis- trative ground rules by which they were elected.
The National Voter Registration Act of 1989 was passedby the Houseof Representatives,but legislation was not passedby the Senate.Under the bill’s provisions, states that require voter registration would be under federal obligation to include a voter registration section in the standard application form for a motor vehicle driver’s license, as well as to pro- vide opportunities for mail and in-person registration. No recent or cur- rently proposed federal legislation addressesthe need for improvement of voter information or outreach practices.
The House Subcommittee on Elections requested that we gather and
integrate information on voter turnout that could assist in possible con- gressional actions to enhanceparticipation in federal elections. Our approach involved first assessinginternational, state, and local (district, county, multicounty) differences in voter turnout and then, to the extent possible, determining the electoral procedures,voter information cam- paigns, and get-out-the vo- te activities that are associatedwith these turnout differences.
Page 11 GAO/PEMD-Bl-1^ Voting
Chapter 1 Introduction
Our study revolves around eight questions with focusesranging from international, to national, to state, to local.
International:
National:
State:
Local:
The existing research pertinent to these questions is uneven. There are a number of studies that addressthe reasonsfor the international differ- encesin turnout. State variations in US. turnout, along with the causes of the decline in U.S. turnout since 1960, also have been the focus of research. However, there are virtually no studies that addressthe effects of voter information campaigns,or the efforts of local outreach organizations, on voter turnout.
Page 12 GAO/PEMD-fU-1Voting
chaptm 1 lntroductlon
research of this type. Moreover, the number of caseswas sufficiently large to allow multivariate statistical procedures that led to stronger conclusions.
Second,the state and local surveys we conducted answered important questions concerning election procedures,voter information campaigns, and get-out-the-vote activities, someof which had not been previously addressed.This included an assessmentof both the main effects and possible interaction effects of these procedures and activities on voter turnout. For example, we addressedthe possibility that voter informa- tion campaigns both increase voter turnout and ameliorate the turnout effects of other election procedures.
Finally, our study avoided certain significant methodological problems of the past in comparing the turnout and election costs of all-mail elec- tions with those of polling-place elections. Earlier efforts had suffered from comparability problems as a result of comparing all-mail with pol- ling-place elections in different jurisdictions or different elections.
research that attempts to understand the causesof high and low voter turnout. While our design allowed us to identify whether policies and practices were associatedwith high voter turnout at the state level, we could not be absolutely sure that such policies or practices causedthe higher turnout. There might have been other causal factors that were not identified or not included in the study.
A secondlimitation involved the problem of measuring the effects of local outreach activities. To determine whether the efforts of an out- reach organization actually causedpotential voters to vote would require a comparison of the voting behavior of potential voters con- tacted by the organization with the voting behavior of those not con- tacted by the organization. However, outreach organizations typically do not keep records of the potential voters they contact. In addition, it is difficult to measurethe extent to which organizations actually engagein outreach activities.
In order to get someidea of the promising practices of these outreach organizations, we solicited the opinions of the directors or heads of the local outreach organizations about what they believe are effective activ- ities, and why they think these activities are effective. The limitation of
Page 14 GAO/PJSMIWl-1Voting
this method, of course,is that directors of outreach organizations may not make accurate estimates of effectiveness.
Questions 1,2, and 3-the “international” and “national” questions- are addressedin chapter 2. Questions4 and 6, which concern the effects of election procedures, are discussedin chapter 3. Question 6 regarding the effect of voter information campaignsis dealt with in chapter 4. Finally, questions 7 and 8, which examine the possible effects of local outreach organizations, are discussedin chapter 6.
Page 16 GAO/PEMDW1 Voting
chapter 2 Amertcan Voter Turnout: Low and DecUdng
Table 2.1: Rank Order of 21 bemocracier by Voter Turnour (^) Country Rank Turnout Italy 1 94.0% Austria 2 09. Belgium 3 86 Sweden 4 86. Portugal 5 85. Greece 6 04. Netherlands 7 84. Australia 8 83. Denmark 9 82. Norway 10 81. West Germany (^) - 11 81.1- New Zealand 12 78. France 13 78. United Kingdom 14 76. Japan 15 74. Spain 16 73. Canada 17 67. Finland 18 63. ireland 19 62. United States 20 52. Switzerland 21 39. aBased on the most recently published compilation, covering elections up to 1981. Source: D. Glass, P. Squire, and R. Wolfinger, “Voter Turnout: An International Comparison,” Public Opinion, 6 (Dec.-Jan. 1984) p. 50.
Not only is voter turnout in the United States lower than in many other democracies,it has been declining. Figure 2.1 shows that the turnout in presidential electionsjumped from 61 percent in 1948 to 63 percent in 1962, before dipping to 60 percent in 1966. In 1960, the turnout returned to 63 percent, but in 1964 American voter turnout began a steady downward trend. The decline was only one or two percent per election until 1968, and then from 1968 to 1972 turnout declined from 61 to 66 percent.
Y
Page 17 GAO/PJCMD-91-lVoting
Flgun 2.1: U. 8. Voter Turnout in Prealdentlal ElectIona, 1948-88. 06 Pueml d voting qa popuhtbn vatlng 64
0
1949 1962 1954 1999 1964 1969 1972 1979 1929 1984 1988 Ebctlon yoan
It is important to note that someof the decline from 1968 to 1972 reflects the enfranchisement of 18 to 20-year-olds. 1972 was the first presidential election in which this agegroup could vote. Researchcon- sistently indicates that young people are substantially less likely to vote than are older citizens. (US. Bureau of the Census,1989) It should be noted, however, that since virtually all industrialized democracies enfranchise 18 to 20-year-olds,allowing young people to vote cannot be cited as one of the reasonswhy the United States has comparatively lower voter turnout.
From 1972 to 1988, the decline continued, although at a much slower pace. In 1972,66 percent of the voting agepublic voted; by 1988, the turnout had dropped to 60 percent.
While turnout for congressionalelections in non-presidential (midterm) election years has always been lower than turnout in presidential years, it has followed the samedownward trend. Figure 2.2 shows that in 1960 the turnout was 43 percent, that it rose to 48 percent in 1966, dropped
Page 18 GAO/PEMD-@l-lVoting
