Political socialization in the classroom revisited: the Kids Voting program.

American schools have long been viewed as agents of political socialization, helping to build support for the prevailing societal norms. Along with family, friends, the surrounding environment and the mass media, schools are seen as a major influence on young people as they develop a political awareness.

The general decline in voter turnout in the United States since 1960 has raised doubts about the effectiveness of schools and other socialization agents in spurring political participation. Yet a new civic education program, based in Arizona, has offered a fresh approach to using the schools as a direct and indirect means of boosting voter turnout. Six years after its first trial run, Kids Voting USA created and administered a curriculum in 1994 that was used by 2.3 million students in 20 states and the District of Columbia.

This study evaluates the success of Kids Voting in meeting two of its three objectives: to increase student awareness and intellectual involvement in the 1994 election, and to use these students to get additional adults to vote in larger numbers that they would have in the absence of the program. The third objective of the program--to increase the rate at which these students eventually vote as adults--would require a massive longitudinal study and is beyond the scope of this effort.

The significance of this study rests in its evaluation of a unique civics education program that is expected to be used by 5 million students in 40 states in 1996. The study, grounded in an area that has seen reduced social science research in the past decade,(FN1) also challenges the field's pessimism about the ability of a civics curriculum to influence students in a predictable manner, at least in the short-term.

POLITICAL SOCIALIZATION AND THE SCHOOLSFor more than 30 years, researchers have explored the role of the school as one of the locations where children develop the knowledge, attitudes and behaviors that shape their roles as future participants in a democracy (Banks and Roker, 1994; Greenstein, 1965; Hess and Torney, 1967; Jennings, 1993; Langton, 1967). While family and home background are often viewed as the primary agents of socialization, schools are seen as a significant secondary agent, along mass media exposure (Atkin, 1981; Bronstein, 1993), the political context of the times (Niemi, 1974) and the role of the individual as an independent factor in the process (Chaffee, Pan and McLeod, 1995; Haste and Torney-Purta, 1992; Jennings and Niemi, 1974; Knutson, 1974; Niemi, 1974).

In their seminal 1967 work, Hess and Torney conclude that "the public school is the most important and effective instrument of political socialization in the United States" (p. 221). While some researchers have minimized the role of schools (Easton and Dennis, 1969), many others note the role of schools in teaching both formal and informal civic values (Dawson and Prewitt, 1969; Palonsky, 1987).

The inculcation of political attitudes begins well before high school (Hess and Torney, 1967; Moore, Lare, and Wagner, 1985). Greenstein notes that "during the last five years of elementary school, children move from near--but not complete--ignorance of adult politics to awareness of most of the conspicuous features of the adult political arena" (1965, p. 1). At the very earliest grade levels, students begin forming supportive, uncritical opinions toward the abstract political community (Moore, Lare, and Wagner, 1985). Children in grades 2 and 3 can relate to government in terms of the president, while those in grades 4 to 8 are increasingly able to use voting and congress as a way to explain government (Easton and Dennis, 1973). At about age 7 many children experience a "cognitive revolution" that continues to about age 13 when they enter the final stages of cognitive development (Easton and Dennis, 1973).

Mass media exposure and the political context of the times also are seen as having an impact on political socialization. Exposure to campaign information from television and newspapers can have an important impact on cognitive processes (Atkin, 1981; Bronstein, 1993; Chaffee and Yang, 1990; Hawkins, Pingree, and Roberts, 1975). While researchers in the 1960s found that children's images of the president were usually positive and non-cynical (Easton and Hess, 1962; Greenstein, 1965), studies in the Watergate era and beyond show a decline in the idealization of the president that may be due to the political mood of the times (Bronstein, 1993; Hawkins, Pingree, and Roberts, 1975).

CIVICS PROGRAMS IN SCHOOLSWhile the role of schools in political socialization appears strong, there is little evidence to show it is due to the kind of civics courses that the general public might commonly associate with learning about the political system. Civics classes and other formal political education efforts in schools have not been found to have a major impact on political attitudes or behavior, due to the nature of political socialization being more of a long-term and gradual process rather than one that can be affected by a single high school course (Beck, 1977; Jennings, Langton and Niemi, 1974; Langton, 1969; Litt, 1963). Traditional high school-level civics programs have been found to have an impact on some subgroups such as minority students (Langton, 1969) and those generally from lower social class backgrounds (Beck, 1977). But many students peak in their early political interest during the junior high school years; their attitudes become so entrenched that little change occurs in high school, despite the earnest efforts of civics teachers (Easton and Hess, 1962). One recent study of young children found they had no lack of interest in political campaigns and no lack of faith in the American political system, but the study attributed such attitudes to the information the children had received from the broadcast media, not information from the schools (Bronstein, 1993). Given the lack of success, some educators have suggested a shift in the emphasis of such courses away from citizen-skills and mechanics-of-government approaches and toward the teaching of moral education and the values needed to ensure a healthy community and common life (Leppard, 1993; Levitt and Longstreet, 1993; Mabe, 1993).

A federal study of "get out the vote" efforts in various states concludes there is no significant relationship between school-centered civic education courses and turnout (General Accounting Office, 1990). The study found 78 percent of all states allowed youths to register early for their first election, 76 percent supplied instructional materials to schools on voting and 36 percent used youth-oriented media in voter outreach and information campaigns. Low turnout states were more likely to engage in such activities, yet the GAO report found no evidence that such efforts were associated with increased turnout.

CAUSES OF LOW VOTER TURNOUTThe apparent lack of success in socializing students so they become adult voters parallels a general decline in turnout levels in U.S. elections since 1960 and a traditionally low turnout rate for young people (Teixeira, 1992). In presidential election years, turnout of the voting age population has dropped from 62.8 percent in 1960 to 50.1 percent in 1988 before picking up to 55.2 percent in 1992's three-way race. In off-year elections, turnout nationally has dropped from 47.0 percent in 1962 to 36.5 percent in 1990 before picking up slightly to 39.6 percent in 1994 (Crocker, 1994, 1995).

Reasons for the long-term decline may include a drop in voter partisanship and political efficacy (Abramson and Aldrich, 1982), changes in voter age, and reduced reliance on newspapers for campaign news (Shaffer, 1981; Teixeira, 1987, 1992). Kleppner (1982) attributes virtually all of the decline to changes in partisanship, political efficacy and lowering the voting age to 18. Cassel and Luskin (1988) argue such conclusions are too simple and that none of the models explain the bulk of the drop in turnout. The voting decline has prompted renewed interest in the role of schools in addressing the low turnout levels of younger voters (Teixeira, 1992). Cohorts now entering the electorate vote in such very low numbers that their pre-adult impressions of government and the political system may not have been positive ones (Teixeira, 1992).

THE KIDS VOTING PROGRAMIn this environment of lagging voter turnout and academic pessimism about the effectiveness of civics programs, the Kids Voting program was launched. Based in Tempe, Arizona, it began on a trial basis in six Arizona communities in 1988 and was extended statewide in 1990. In 1992, the program was rechristened Kids Voting USA and received a limited national run in 11 states. In 1994 it expanded to 20 states plus the District of Columbia, reaching 2.3 million students in kindergarten through grade 12 at a budgeted cost of $5 million in cash and in-kind contributions, all from private sources.

The KV curriculum is individualized for grades K-12 and stresses cooperative learning, group problem-solving, and active, hands-on experiences. Typical lessons for primary grade students include role playing, craft activities and classroom elections, while secondary students consider policy options, research the positions of candidates and hold formal debates. The curriculum stresses information gathering, especially from the news media. Teachers generally use the program for six to 12 hours of classroom instruction in the fall prior to an election. The low-cost program--budgeted at about $2 to $3 per student--is designed to elicit higher-order thinking while instilling in students the value of voting and the enpowerment that the voting act can give to a participant. By increasing awareness of the electoral process, the program hopes that graduating high school seniors will feel knowledgeable and comfortable about voting when they turn 18 and will vote in greater numbers than students not exposed to Kids Voting.

A second KV component calls for students to discuss political issues with their parents or adults within their household environment. The discussions are designed to reinforce the classroom instruction and to increase the chances the parents or affected adults will become aware and involved in the election and cast a ballot. On Election Day, students are encouraged to accompany these adults to the polls and cast an unofficial ballot. Students in K-8 could not cast a ballot without an adult being present; the goal was to increase parental turnout due to pressure from the children to take them to the polls. The impact of the program on adults also may have been increased by such community-wide activities such as a "KidsConvention" for students to discuss political issues, the recruitment of hundreds of adult volunteers to handle the student voting booths on Election Day, and heavy news media reporting on the project.

The first formal analysis of Kids Voting (Merrill, Simon, and Adrian, 1994) was based on the 1990 Arizona statewide election. A linear regression model was used to project expected turnout, based on past off-year elections, and it showed that turnout was 3.9 percent higher than expected. The study also used a statewide survey of registered voters that attempted to isolate the portion of voters who said they cast a ballot solely due to KV. Respondents who said they voted were asked if KV were "in any way a factor in their decision to vote Tuesday." If the response was yes, they were asked: "Was the Kids Voting program the determining factor in your decision to vote on Election Day? That is, would you have not voted if it were not for the Kids Voting program?" Some 2.6 percent of voters identified KV as the determining factor in their decision to vote. Turnout in precincts that used KV for two consecutive elections also was higher than comparable precincts, suggesting a cumulative effect of repeat usage.

A similar study was conducted in 11 states where individual communities used the program in 1992: Alaska, Arizona, California, Colorado, Florida, Georgia, Kansas, North Carolina, Ohio, Tennessee and South Dakota (Merrill, 1993). First, a direct comparison of areas using KV and matched, adjacent areas without KV indicated that turnout was 2.8 percent higher in areas with the KV program. Second, a linear regression model was used to project turnout, based on historical patterns, in areas with KV and the matched, adjacent areas without it. The projected turnout was then compared to the actual turnout in the KV and non-KV areas. The analysis found that turnout was 1.7 percent higher in areas with KV than in the matched, adjacent area. Finally, postelection surveys in five states indicated that 2.9 percent of registered voters said they cast a ballot solely due to the existence of the KV program.

In summary, the various measures to analyze the program in 1990 and 1992 found KV has been uniformly associated with increases in turnout ranging from 1.7 percent to 3.9 percent.

For the 1994 election, an analysis of Kids Voting in the San Jose, California area also found strong effects in terms of student communication behaviors, such as media usage and discussions about politics (Chaffee, Pan, and McLeod, 1995). There was no measurable effect of Kids Voting on turnout, in part because 86 percent of the parents interviewed in the small sample of 477 families said they had voted, thereby limiting any comparison due to the small number of non-voters.

The present study is based on the expanded, national Kids Voting program that was used in 20 states and the District of Columbia in 1994. The analysis is aided by the fact that it was an off-year election and the higher turnout associated with a presidential election is absent.

This study focuses on three research questions: (1) Was adult turnout higher in the areas where the KV program was used, compared to matched, adjacent areas that did not use the program?; (2) Did participating students actively discuss the election in class, follow the campaign in the news media and talk about the campaign with their family? Would they want to take part in such a process again?; and (3) What value, if any, did participating teachers assign to the KV program?

METHODTo determine whether Kids Voting increased voting in the areas where the program was used, turnout in each area was compared to a matched, comparable area in the same vicinity that did not use the program. Census data, registration statistics and voting statistics were used to ensure that matched areas had similar racial, socioeconomic and partisan characteristics. Interested social scientists or election officials in each area were consulted to ensure the matches were comparable and reasonable.

It was not always possible to find matching areas for comparison because boundaries for geopolitical units like precincts or counties frequently crossed school district boundaries. However, matches were found for communities in 15 of the 21 states studied. An alternative method was used in Arizona, since KV was used in 99 percent of Arizona's school districts and no matching areas could be found. In this case, a statewide post-election survey was conducted to evaluate the program and to isolate the percentage of voters who said Kids Voting was the specific reason they went to the polls on Election Day. The resulting percentage (1.5 percent) was multiplied by the total number of voters who cast a ballot (1.15 million voters) and than multiplied by the percentage of school districts using the program (99 percent) to approximate the program's impact on statewide turnout.

To gauge the program's impact in the classroom, a sample of 24,976 students was selected in 20 of the 21 states using the program in 1994. Participating schools were stratified by racial, political and demographic characteristics, then selected at random in a manner designed to yield approximately the same number of respondents in grades K-3, 4-6, and 7-12. Individual classes in each school were then selected at random, and all students in a chosen class completed a questionnaire. Twelve questions were given to students in all three grade groups. They differed only in phrasing for the youngest children (e.g., the gender question asked whether students in K-3 were "a boy or a girl" instead of "male or a female"). Six of the questions were designed to gauge student involvement in Kids Voting-related activities (e.g., "When you were at home, did you ask questions about voting") and exposure to mediated campaign information. Two questions dealt with political efficacy and trust (e.g., "Do you think the President is the kind of person who would help you if you were in trouble?") and are not reported here. Three questions asked about the student's grade level, gender and race. A final question asked, "Would you like to have Kids Voting as part of what you learn in school the next time there is an election?".

The surveys for grades 4-6 and 7-12 included seven additional questions, centering on family decision-making patterns (e.g., "When family decisions are made, do your parents let you have, 1. A lot of input; 2. some input; or 3. None?"), plus additional media use and efficacy questions.

The student survey yielded 7,928 responses from students in grades kindergarten through three, 8,323 responses from grades four through six, and 8,725 responses from grades seven through 12. Slightly more female students completed a survey (52 percent). There was representation from a variety of ethic groups (16 percent of students identified themselves as African American, 6 percent as Hispanic or Latino, 5 percent as Native American and 4 percent as Asian and 69 percent as White or Anglo).

Finally, in an effort to determine what value, if any, did participating teachers assign to the program, a random sample of more than 1,000 teachers (n=1,084) from 20 of the 21 participating states was used. The 19-question survey asked teachers to rate KV on a variety of dimensions, including its ability to increase student knowledge about elections, increase student enthusiasm for voting, and increase parental involvement in the schools.

There were several limitations to the study. Ideally, an evaluation study of a program like Kids Voting should focus on the political socialization of students at school in conjunction with the influence of socializing factors at home and elsewhere. However, the federal "Buckley Amendment" privacy law limited the ability of KV to identify students by name in collecting data on any political, attitudinal or behavioral questions that might be seen as invading the privacy of students or parents.

KV also did not collect pre-test information in the aggregate that could have been used to evaluate the program in a quasi-experimental framework. Aggregate student attitudes and behaviors could have been measured prior to exposure to Kids Voting (Time A), and then immediately after the treatment (Time B) and contrasted with a control group. Instead, the program just collected data after the election, making it impossible to isolate changes in individual or collective attitudes and behavior that may be due to the stimulus of Kids Voting.

KV also did not question students (and teachers) in the matched, adjacent non-KV areas and use their responses as a control group. A comparison of the students using KV and those in a control group would have been useful in gauging the program's impact on political efficacy and trust, family decision-making processes and media usage.

Due to these limitations, the present study is unable to directly show any causation between KV and a specific student's attitudes before and after the election. Instead, it focuses on participant attitudes toward the program and external evidence of the program's possible impact, such as changes in turnout in areas where it was used compared to areas where it was not used.

FINDINGSThe geographic areas using Kids Voting in 1994 experienced a small but consistent increase in turnout, compared to similar matched areas, as shown in Table 1. In 14 of the 15 geographic areas examined, turnout increased from +0.1 percent (California) to as much as +9 percent (Washington state).

Within this study design, it was impossible to demonstrate in a causal sense that KV was the only factor responsible either for the increase in turnout experienced by the 14 areas (or the single decrease, in Colorado). However, the median increase across all 15 areas in 1994 (2.0 percent) was consistent with KV research conducted in 1988, 1990 and 1992 that found usage of KV was associated with an increase in turnout of 1.7 percent to 3.9 percent.

Table 2 presents the results of student attitudes toward the program. The student survey indicated 95 percent of the student participants in KV said their teachers had talked to them frequently (63 percent) or occasionally (32 percent) about voting, suggesting the KV program was used by virtually all teachers in the areas where school administrators had contracted for it (Table 2). There was remarkably little variance across grade level to this question or any of the other questions dealing with campaign activity. Most students also reported taking part, occasionally or frequently, in such activities as debates on political or social issues at school (76 percent) and asking questions at home about elections (71 percent). Most students said they were able to go the polls with their parents and vote on Election Day (55 percent) and said their families frequently or occasionally had political discussions at home (59 percent).

Kids Voting participants also were associated with high levels of news media use for campaign information. Seventy-three percent of students said they frequently or occasionally watched news about the campaign on television; 73 percent also said they followed the campaign on radio. Usage of the print media was less frequent: 49 percent of students said they read about the campaign in newspapers and 29 percent said they read magazine campaign coverage.

Developing an appreciation for voting and for the importance of casting a ballot is a central thrust of the Kids Voting program. Seventy-six percent of the participating students said they felt it was very important for people to vote on Election Day; 18 percent said it was somewhat important and 6 percent said it was not very important. There was virtually no difference across grade levels on this key question. Students also indicated a strong desire to use KV in future classroom discussions. Asked whether they would "like to have Kids Voting as part of what you learn in school the next time there is an election," more than 85 percent of students (with a specific response) said yes. With the inclusion of a "not sure" category, there was about a 5-to-1 ratio of students favoring retention of the program for future election campaigns.

Finally, Table 3 indicates that teacher evaluations of KV also were positive regarding the program's value. Asked for an "overall impression of the Kids Voting program," 39 percent said it was very favorable, 50 percent said favorable, 9 percent said somewhat unfavorable and 2 percent said very unfavorable. Teachers wer likely to rate it favorably for grades 7-9 (91 percent) and 10-12 (98 percent) as they were for grades K-3 (89 percent) and 4-6 (89 percent). Asked to rate KV in terms of its impact on students and parents, 99 percent of teachers said the program increased student knowledge. Ninety-nine percent of teachers said it increased student enthusiasm, while 81 percent said it increased parental involvement in the schools.

In terms of the research questions, the study found that adult turnout was higher in the areas where the KV program was used, compared to comparable areas that did not use the program. Participating students actively discussed the election in class, followed the campaign in the news media, talked about the campaign with their family, and said they would want to take part in such a process again. Finally, most participating teachers said the program had a positive impact in increasing short-term student knowledge and enthusiasm about the election.

DISCUSSIONThis study suggests Kids Voting can provide a positive introduction of students to the voting process. Seventy-six percent of participants said they think it's important for people to vote on Election Day, an encouraging sign in a country where many citizens cannot be bothered to even register. The uniformity of results across grade levels was much higher than in earlier studies that had suggested such civic-oriented programs were less likely to succeed at the high school level (Beck, 1977; Easton and Hess, 1962; Jennings, Langton and Niemi, 1974; Litt, 1963).

The matched comparisons of areas with KV and those without the program suggest that turnout was, on average, 2.2 percent higher in the 15 states where it was used, consistent with past studies. No causal argument can be made due to the many potential intervening variables. For example, despite the best efforts at pairing districts, those with KV may have had more aggressive social studies administrators whose zeal for such a program energizes civics teaching at all levels and produces students who vote at higher levels later in life. The increase can be viewed by supporters as a positive step in reversing the general decline in turnout (and perhaps by critics as a negligible change that was not worth the time of students, teachers, volunteers, and $5 million in funding).

Kids Voting also may have boosted turnout in an indirect manner. To implement the program, KV enlisted 45,000 volunteers on Election Day to staff polling areas and allow students to cast their mock ballots. Many of these volunteers would have voted any way; however, some may cast a ballot because they were already at the polling location. Pre-election publicity about the program also may have spurred additional turnout. The program is considered "good copy" by the news media. KV generates dozens of newspaper, television, radio and magazine stories in every area where it is used. A study of the 1990 KV program in Arizona found it generated more than 300 newspaper articles in the final six weeks of the campaign; a statewide survey also found nine in 10 registered voters said they had heard of the program by Election Day (Merrill, Simon, and Adrian, 1994).

Beyond the immediate focus on attitudes toward elections, Kids Voting also may have several secondary benefits, such as increased student awareness and use of the news media and increased discussions of public affairs with family and friends. Such activities are associated with long-term political socialization (Atkin, 1981; Bronstein, 1993). It remains to be seen whether these apparent effects are short-term or lasting when the students assume their adult roles as voting citizens in society.

The limitations imposed by federal privacy law on student data collection, as discussed earlier, must be acknowledged. However, organizers can measure changes in aggregate attitudes simply by using the elements of a classic experiment. Such an approach would be hard to do in a state like Arizona, where 99 percent of school districts use the program and a large control group does not exist. But Chaffee, Pan, and McLeod (1995) showed the value of conducting such a study in a community like San Jose, California, where only some schools use KV. Pre-testing a sample of all students in a single city, allowing the KV curriculum to be used for half the students, and then post-testing all the students after the election would provide a more valid approach to isolating the impact of KV, especially if such a design were used in a variety of cities and states. Even better, of course, would be a long-term study that would track tens of thousands of KV students, along with a control group, over decades to see if turnout levels were different for KV participants. The logistics of such a study, given the high transience of Americans at the close of the century, is staggering.

Yet the potential for the program to have an impact on the political socialization of students, where many civics programs have failed before, suggests it is worthy of broader future study.

Added material.

JAMES SIMONDirect all correspondence to: James Simon, Department of Communication, University of the Pacific, Stockton, California 95211. Telephone: (209) 946-3049. E-mail: jsimon@vmsl.cc.uop.edu.

BRUCE D. MERRILLArizona State University.

Table 1. Comparison of Turnout, Areas Within States That Used KV vs. Matched Areas that Did Not Use KV.

(TABLE) % Turnout Area % With KV % Without KV Difference Additional VotersWashington State 62.0 53.0 +9.0 1,046Georgia 54.7 47.5 +7.2 3,287Tennessee 62.2 56.6 +5.6 9,108South Carolina 61.2 55.6 +5.6 2,314Kansas 68.0 62.7 +5.3 4,826Michigan 58.6 55.0 +3.6 1,232Ohio 58.0 55.0 +3.0 9,959Washington, DC 54.0 51.9 +2.1 336Maryland 66.0 63.9 +2.1 1,347Alaska 61.8 57.7 +4.1 2,258North Carolina 35.0 34.0 +1.0 1,063Florida 65.5 64.5 +1.0 11,777Arizona (FN*) (FN*) +1.5 17,133California 56.1 56.0 +0.1 322Colorado 56.5 59.9 -3.4 0Total Additional Voters 66,008Median Increase Across All 15 Areas: 2.0%.

FOOTNOTE* Arizona percentage based on projection from post-election statewide survey of registered voters.

Table 2. Political Attention and Campaign Activity Of KV Participants, By Grade.

(TABLE) % K-3 % 4-6 % 7-12 % All StudentsHow often did ...... your teacher talk to you in classabout voting Frequently 60 64 65 63 Occasionally 33 32 30 32 Seldom/Never 7 4 5 5 Total 100 100 100 100 (n) (7,581) (8,246) (8,671) (24,498)... you ask questions at home aboutelections Frequently 20 20 15 18 Occasionally 44 57 59 53 Seldom/Never 36 22 26 29 Total 100 100 100 100 (n) (7,570) (8,285) (8,702) (24,557)... you participate when politicaland social issues debated at school Frequently (FN*) 31 22 26 Occasionally 46 53 50 Seldom/Never 23 24 24 Total 100 100 100 (n) (8,198) (8,654) (16,852)... your family has political discus-sions at home Frequently 21 13 11 15 Occasionally 36 48 47 44 Seldom/Never 43 38 42 41 Total 100 100 100 100 (n) (7,484) (8,234) (8,670) (24,398)Did you go to the polls and vote onElection Day? Yes 56 56 .53 55 No 44 44 .47 45 Total 100 100 100 100 (n) (7,658) (8,174) (8,473) (24,305)How often did you watch/ readthings about the election ... ... On television? Frequently 7 30 26 .28 Occasionally 37 50 43 44 Seldom/Never 36 20 20 26 Total 100 100 100 100 (n) (7,611) (8,275) (8,708) (24,594)... In the newspaper? Frequently 17 15 16 16 Occasionally 24 35 40 33 Seldom/Never 59 50 44 51 Total 100 100 100 100 (n) (7,594) (8,279) (8,699) (24,572)... On the radio? Frequently (FN*) 33 31 32 Occasionally 43 39 41 Seldom/Never 24 30 27 Total 100 100 100 (n) (8,275) (8,700) (16,975)... In magazines? Frequently (FN*) 7 7 7 Occasionally 20 24 22 Seldom/Never 73 69 71 Total 100 100 100 (n) (8,234) (8,688) (24,922)How important is it for people tovote on Election Day? Very Important 78 75 75 76 Somewhat 14 20 20 18 Not Very 7 5 5 6 Total 100 100 100 100 (n) (7,713) (8,232) (8,442) (24,387)Would you like to have Kids Votingas part of what you learn in schoolthe next time there is an election Yes 86 65 59 No 14 8 12 Not sure (FN*) 26 29 Total 100 100 100 (n) (7,768) (8,232) (8,474).

FOOTNOTE* not offered as a response at that grade level. Some percentages do not add up to 100due to rounding.

Table 3. Evaluation of Kids Voting By Participating Teachers, By Grade Level.

(TABLE) Grade Level K-3 4-6 7-9 10-12 TotalA. Overall Impression of the Kids Voting ProgramVery Favorable 36% 42% 45% 38% 39%Favorable 53% 47% 46% 60% 50%Somewhat Favorable 9% 10% 9% 3% 9%Not Very Favorable 3% 1% 0% 0% 2%Total 100% 100% 100% 100% 100%(n) (397) (339) (114) (37) (1,084)B. Impact of Kids Voting on Increasing ... ... Student ... Student ... Parental Knowledge Enthusiasm InvolvementA Great Deal 61% 64% 23%Somewhat 38% 35% 58%Not At All 1% 1% 19%Total 100% 100% 100%(n) (1,075) (1,066) (1,084).

NOTESome percentages do not add up to 100% due to rounding.

FOOTNOTE1. The concept of political socialization can be traced to Plato's search in "The Republic" for a way of developing an individual's character so that he or she would be supportive of the ideal polity. In the current literature, the term is often used to describe a "developmental process through which the citizen matures politically" (Dawson and Prewitt, 1969, p. 17). It also can be viewed as a conservative process "facilitating the maintenance of the status quo by making people love the system under which they are born" (Sigel, 1969, p. 1). Niemi has criticized the latter definition for assuming that the politicization of the citizenry is "inevitably purposeful, goal-oriented, rational and organized, and it precludes an understanding of individual differences in ability and willingness to learn" (Niemi, 1974, p. 8). He argues that the politicization of the individual is a process in which a citizen both accommodates and assimilates the themes of the political culture. While the topic remains a fertile area for research by education, sociology, psychology and communication scholars, political science research in the area seemed to peak between 1965 and 1983, prompting one recent critic to dismiss the field as a "political fad" (Peng, 1994, p. 100). In a similar overview, Sears is less critical, but concludes that "the great burst of activity of the 1960s and early 1970s appears to have subsided and in some respects has nearly disappeared altogether" due in part to the belief that a child's political attitudes are not very strong and have little impact on their political attitudes in adulthood (Sears, 1990, p. 69).

REFERENCESAbramson, P. R. and J. H. Aldrich. (1982). The Decline of Electoral Participation in America. American Political Science Review, 76: 502-521.

Atkin, C. K. (1981). Communication and Political Socialization. In D. Nimmo and K. Sanders (Eds.), Handbook of Political Communication, pp. 299-328. Beverly Hills, CA: Sage.

Banks, M. H. and D. Roker. (1994). The Political Socialization of Youth: Exploring the Influence of School Experience. Journal of Adolescence, 17: 3-15.

Beck, P. A. (1977). The Role of Agents in Political Socialization. In S. A. Renshon (Ed.), Handbook of Political Socialization, pp. 115-141. New York: Free Press.

Bronstein, C. (1993, August). Tellin' it like it is: Children's attitudes toward the election process and the '92 campaign. Paper presented at the meeting of the Association for Education in Journalism and Mass Communication, Kansas City, Mo.

Cassel, C. A. and R. C. Luskin. (1988). Simple Explanations of Turnout Decrease. American Political Science Review, 82: 1321-1327.

Chaffee, S., Z. Pan, and J. McLeod. (1995, June). Effects of Kids Voting San Jose: A quasi-experimental evaluation. Report to the Policy Study Center, Annenberg School for Communication, University of Pennsylvania.

Chaffee, S. and S. Yang. (1990). Communication and Political Socialization. In O. Ichilov (Ed.), Political Socialization: Citizenship, Education and Democracy, pp. 137-157. New York: Teachers College Press.

Crocker, R. (1994). Voter Registration and Turnout: 1948-1992 (Congressional Research Service Rep. 94-22). Washington, DC: Congressional Research Service.

Crocker, R. (1995). Voter registration and turnout, 1994: Preliminary results (Congressional Research Service Rep. 95-237). Washington, DC: Congressional Research Service.

Dawson, R. E. and K. Prewitt. (1969). Political Socialization. Boston: Little, Brown.

Easton, D. and J. Dennis. (1969). Children in the Political System. New York: McGraw-Hill.

Easton, D. and J. Dennis. (1973). Governing Authorities: The Child's Image of Government. In J. Dennis (Ed.), Socialization to Politics: A Reader. New York: John Wiley.

Easton, D. and R. D. Hess. (1962). The Child's Political World. Midwest Journal of Political Science, 6: 229-246.

General Accounting Office (1990, November). Voting: Someprocedural changes and information activities could increase turnout (GAO Rep. PEMD-91-1). Washington, DC: General Accounting Office.

Greenstein, F. I. (1965). Children and Politics. New Haven, CT: Yale University Press.

Hawkins, R. P., S. Pingree, and D. F. Roberts. (1975). Political Socialization of Children: Two Studies of Responses to Watergate. American Politics Quarterly, 3: 406-422.

Haste, H. and J. Torney-Purta. (1992). The Development of Political Understanding: A New Perspective. San Francisco, CA: Jossey-Bass.

Hess, R. D. and J. Torney. (1967). The Development of Political Attitudes In Children. Chicago: Aldine.

Jennings, M. K. (1993). Education and Political Development Among Young Adults. Politics and the Individual, 3: 1-24.

Jennings, M. K., K. P. Langton, and R. G. Niemi. (1974). Effects of the High School Civics Curriculum. In M. K. Jennings and R. G. Niemi (Eds.), The Political Character of Adolescence. Princeton, NJ: Princeton University Press.

Jennings, M. K. and R. G. Niemi. (1974). The Political Character of Adolescence. Princeton, NJ: Princeton University Press.

Kleppner, P. (1982). Who Voted? The Dynamics of Electoral Turnout. New York: Praeger.

Knutson, J. N. (1974). Prepolitical Ideologies: The Basis of Political Learning. In R. G. Niemi (Ed.), The Politics of Future Citizens, pp. 7-40. San Francisco: Jossey-Bass.

Langton, K. P. (1967). Peer Group and School and Political Socialization. American Political Science Review, 61, 751-758.

Langton, K. P. (1969). Political Socialization. New York: Oxford University Press.

Leppard, L. J. (1993). Discovering a Democratic Tradition and Educating for Public Politics. Social Education, 57: 23-26.

Levitt, G. A. and W. S. Longstreet. (1993). The Social Studies, 83 (July/August): 142-148.

Litt, E. (1963). CiviC Education, Community Norms and Political Indoctrination. American Sociological Review, 28: 69-75.

Mabe, A. R. (1993). Moral and Practical Foundations for Civic Education. The Social Studies, 83 (July/August): 153-157.

Merrill, B. D. (1993). Evaluating the effectiveness of Kids Voting USA in the 1992 presidential election. Unpublished.

Merrill, B. D., J. Simon, and E. Adrian. (1994). Boosting Voter Turnout: The Kids Voting Program. Journal of Social Studies Research, 18: 2-7.

Moore, S., J. Lare, and K. A. Wagner. (1985). The Child's Political World. New York: Praeger.

Niemi, R. G. (Ed.). (1974). The Politics of Future Citizens. San Francisco: Jossey-Bass.

Palonsky, S. B. (1987). Political Socialization in Elementary Schools. The Elementary School Journal, 87: 493-505.

Peng, Y. (1994). Intellectual Fads in Political Science: The Cases of Political Socialization and Community Power Studies. PS: Political Science and Politics, 27 (March): 100-108.

Sears, D. O. (1990). Whither Political Socialization Research? The Question of Persistence. In O. Ichilov (Ed.), Political Socialization, Citizenship Education and Democracy, pp. 69-97. New York: Teachers College Press.

Shaffer, S. (1981). A Multivariate Explanation of Decreasing Turnout in Presidential Elections, 1960-1976. American Journal of Political Science, 25: 69-95.

Sigel, R. (1965). Assumptions about the Learning of Political Values. Annals of the American Academy of Political and Social Science, 36: 1-9.

Teixeira, Ruy. (1987). Why Americans Don't Vote: Turnout Decline in the United States, 1960-1984. New York: Greenwood Press.

Teixeira, Ruy. (1992). The Disappearing American Voter. Washington, DC: Brookings Institution.