Trust and Political Information: Attitudinal Change in Participants in the Youth Parliament in Brazil

Abstract

This article analyses the impact of socializing experiences on the political attitudes of youngsters. More specifically, our goal is to evaluate the impact of the Youth Parliament program on youngsters’ confidence levels in the Minas Gerais State Assembly (MGSA). The analysis focuses on the cognitive foundations of attitudes and results show a substantial increase in confidence levels in MGSA, an increase associated with the acquisition of information on the institution. It is asserted that the increase in confidence in MGSA represents and attitudinal “gain”. The study design involves quasi-experimental research on a non-random sample. We conducted two rounds of interviews in 2008, prior and subsequent to the program, with 335 participants (167 in the treatment group; and 168 in the control group).

Keywords:
Political attitudes; Political knowledge; Political socialization; Civic education

Introduction

This article assesses the impact of the Youth Parliament on political attitudes of its participants, focusing on their confidence in the Minas Gerais State Assembly in Brazil. We examine whether changes in attitudes have a cognitive basis, that is, whether they are related to changes in political knowledge.

The Youth Parliament project is sponsored by Minas Gerais's State Assembly (MGSA) and the Catholic University of Minas Gerais for public and private school students from Belo Horizonte. Students from these schools participate in a series of activities (lectures, workshops, roundtables, legislative activities etc) during one semester. At the end of the program, students convene on the house floor to vote on a series of proposals, which are subsequently sent as a bill to MGSA. The project aims to enhance high school students’ civic education “through activities that lead them to better understand the organization of powers, especially the Legislative, and prepare them for political participation as citizens”.¹

The Youth Parliament (YP) is part of a broader initiative in institutional development promoted by MGSA. Since the early 1990s, the state assembly has been conducting an ambitious project of institutional innovation, which in addition to modernizing their structure and increasing their technical staff's level of professionalism, aims to foster local communities’ participation in the legislative process and improve access to the institution. For this purpose, the project includes debates, seminars and forums that eventually lead to the introduction of a bill in the MGSA.

As we shall see, young people's perceptions in regards to MGSA have changed, and their confidence in same has increased, both as a result of participation in the YP. In the absence of direct evidence, we suggest that the institutional environment of the innovative MGSA has contributed to this change, particularly in a positive direction.

We argue that this increased confidence in the MGSA has a cognitive basis because it is associated with information gained through frequent contact with the state legislature, its employees, and representatives, as well as through participation in workshops on civic education and legislative activities. The existence of a cognitive basis for attitudinal change depends on whether the YP provides any cognitive gains to participants and whether those gains can be associated with an increase in confidence in the MGSA.

Our research design follows the logic of quasi-experimental research (Campbell and Stanley 1979). In 2008, a non-random sample of 670 participants² completed two rounds of interviews³, 335 before and 335 after the program. We interviewed 167 YP participants and 168 non-participants. The 2008 session of the YP, with the theme “youth and violence,” included seven participating schools. In the selection of the control group, we sought to ensure symmetry between the school types and between different family background factors. Thus, seven schools with the same socioeconomic and cultural profiles as the schools of participating students (public, private and military) were selected. Likewise, students from our control group were selected on the basis of similar characteristics, such as gender, age, grade, social class and parental education.

This article consists of three parts. First, we present studies in civic education and political sociology and examine their contribution to the understanding of the gains acquired through political socialization. Second, we analyze the initial attitudes and changes of attitude among young people in an attempt to assess the impact of the Youth Parliament on young people's confidence in the MGSA. Finally, we verify whether the program provides information gains and whether they are associated with increased confidence in the MGSA.

Political Socialization and Civic Education

This paper discusses two streams in the literature on political socialization. One stream is in the field of political sociology and involves the core idea that attitudes, knowledge and political skills are acquired through socialization (Andolina, Jenkins, Zukin and Keeter 2003; Galston 2001; Jennings and Markus 1984; Sears and Funk 1999; Sears and Valentino 1997). The classical view is that the formation of core political attitudes occurs in childhood, under the predominant influence of the family, and remains stable over the years. Recent studies (Jennings and Markus 1984; Jennings and Niemi 1974; Sears and Valentino 1997) have questioned this view, suggesting that 1) important socialization processes occur after childhood, 2) other agents, depending on the context, are as significant as the family and 3) many attitudes and beliefs formed in childhood are malleable and therefore subject to change throughout life (Sapiro 2004).

A second tradition investigates the role of civic education in the development of competent citizens (Campbell 2008; Finkel 2003; Finkel and Ernst 2005; Finkel et al. 2000; Finkel and Smith 2011; Galston 2001; Niemi and Junn1998; Slomczynski and Shabad1998). These studies are generally related to projects and programs for civic education in fledgling democracies and countries undergoing democratic transitions. Researchers in this area are interested in determining whether these programs, when successful, can make citizens more informed about politics, more likely to engage in political activities, and more democratic in their attitudes.

In a manner similar to the literature on political sociology, our research emphasizes the importance of socializing experiences in an individual's political development after childhood. In addition, and similarly to prior research on civic education, our research investigates programs designed to better inform citizens, and foster political activism, attitudes, and values consistent with democratic regimes.

The central concept of our study involves socialization “gains.” This concept has been used in studies that attribute importance to “events” in an individual's political trajectory, especially among youngsters. For example, Jennings and Markus’ (1984) panel study (1965, 1973 and 1982) of high school students and their parents concluded that flexibility and openness to change in attitudes of young people are related to their limited experience with the world of politics. This inexperience makes young people more vulnerable than adults to the influence of social (such as war or economic crises) or personal events (such as migration and marriage). In another study, Sears and Valentino's (1997) findings on the significant effect voting for the first time has on young people's attitudes and knowledge shows the importance of their first contact with politics. Sears and Valentino suggested that young voters have more to gain in terms of knowledge acquisition and crystallization of political attitudes than do adults who have voted several times. The authors argued that voting is a socialization experience that reduces the gap between those with a greater stock of information and more stable attitudes towards politics and those with virtually no information.

The notion of “gains” is also present in the area of civic education. The quality of democracy is assumed to depend on citizens’ ability to act according to democratic principles (Galston 2001). This requires knowledge of those principles and their basic rules of operation, as well as participation, and political tolerance (Finkel 2003; Finkel and Ernst 2005; Finkel et al. 2000). Gains are found when civic programs are able to develop political behavior that is appropriate in democratic regimes.

More importantly, gains do not occur evenly across every dimension of political behavior. While most studies conclude that individuals who participate in civic education programs become more active and better informed citizens, those studies are not as optimistic in regards to the effects this participation may have on political attitudes.

If we consider the distinction between different political objects (Almond and Verba 1989; Dalton 2004; Easton 1965; Norris 1999), we find that specific objects, such as trust in political institutions, are more malleable than abstract principles, such as political tolerance and adherence to democracy. Thus, changes in institutional trust may occur at the same pace or even faster than cognitive change and political participation.

To better understand the socialization gains in political attitudes, we must establish criteria for discussing two complex issues: the direction of the expected change (either positive or negative) and its basis.

It is not easy to determine the direction of change in institutional confidence as a result of civic programs, unlike identifying knowledge acquisition and political participation. In the debate on political trust, some authors have called attention to the importance of trust for the stability of democracy or specific policies (Almond and Verba 1989; Offe 1999; Warren 1999), whereas others have proposed the need for some mistrust to ensure responsiveness to societal demands (Inglehart 1999; Inglehart and Welzel 2005; Norris 1999). How can we tell if a particular event or program produces attitudinal gains for the participants? Finkel et al. (2000, 1855) argue that it is only possible to speak of “gains” when change in political trust happens as a result of its association with democratic attitudes or the acquisition of relevant information.

To summarize: if civic education has positive effects on institutional trust, we expect those effects to operate through the individuals’ rejection of previous anti-democratic regimes and through a positive spill-over effect from increased efficacy, interpersonal trust, and other supportive democratic values. If civic education has negative effects on trust, we expect the effects to operate through increased awareness of poor system performance, through an increased sense of the unresponsiveness of local political elites, and through a heightened perception of the gaps between democratic ideals and current political practices.

In this paper, we adopt a similar approach in dealing with attitudinal gains. We consider gains have occurred whenever a change in attitudes toward political actors and institutions is founded on acquired information. Therefore, attitudinal changes represent socialization gains when they are accompanied by informational gains. Conversely, changes in attitudes are not considered gains if they are based on purely emotional ties. The normative reference standard is the citizen who, through relevant information, is able to judge political institutions and, as a consequence thereof, to increase or decrease the degree of trust they have placed in them (Norris 1999).⁴

One can consider that the Youth Parliament program has generated gains if participants’ attitudinal changes have a basis in information on the MGSA acquired during the program. The data produced by our research do not allow us to examine this question directly because YP participants were not asked to associate their level of confidence in the state assembly with any political information they may have acquired about it. In anticipation of one of our conclusions, we found young people who participated in the YP to have achieved attitudinal gains because, during the program, they acquired both knowledge in politics and confidence in the MGSA. While we cannot conclude that there is a causal relationship between these changes, they stem from the same process of political socialization. Teenagers who are more informed of MGSA processes, actors, and activities come to rely more on the institution.

Our expectation is that the Youth Parliament's dynamics, including its political content and activities related to legislative practices that are coordinated by the MGSA's staff, positively affect young people's confidence in the institution. The YP enables them to acknowledge the MGSA and get involved in itsinstitutional development project (Anastasia 2001). This is especially true of project initiatives in increasing the MGSA's outreach to society.

Initial Conditions and Changes in Time

An examination of the level of young people's political confidence in the MGSA prior to the YP reveals that it was not distributed evenly among high school students in Belo Horizonte who participated in our study. In confirmation of a widely accepted thesis (Finkel and Ernst 2005; Ichilov 2007; Slomczynski and Shabad 1998), school and family environments are primarily responsible for these attitudinal differences. Individual characteristics, such as political efficacy and the degree of knowledge of politics, tell us nothing about differences in confidence in the MGSA among young people.⁵

Table 1
Dependent variable: Trust in MGSA(binary)

	B	Std Error	Exp(B)
Treatment group	0.127	0.254	1.135
Parents’ education	-0.653	0.232	0.520***
Parents’ participation	0.092	0.132	1.097
Military school (0 = Public school)	0.232	0.446	1.261
Middle-class school (0 = Public school)	0.747	0.522	2.111
Elite school (0 = Public school)	1.386	0.604	3.997**
Political efficacy	0.415	0.260	1.515
Exposure to information in the media	0.057	0.047	1.059
Sex (0 = woman)	0.430	0.259	1.537*
Constant	-1.805	0.538	0.164***
N			284
Nagelkerke R2			0..082
Overall percentage			59,7%
Chi-square			18.001

• ^***
  Statistically significant 0.01 /
• ^**
  statistically significant 0.05 /
• ^*
  statistically significant 0.10

However, this consideration of initial attitudes does not inform us in regards to the subject of our study: change eventually caused by the YP. A first way to explore this question is a graphical analysis of the curve of change over time for the groups that were studied.

Figure 1 provides a clear picture of change. There is a decrease in trust among young people who did not attend the YP and an increase in trust among the participants. If we consider that students who did not participate in the YP showed a decline in confidence in the MGSA, the YP's effects become more apparent: not only does the program help to prevent this decline; it also increases participants’ confidence.⁶

There is clearly a growing confidence among YP participants in the MGSA. We ask if this increase is attributable solely to participation in the Youth Parliament. To answer the question, we used a regression model with a variable indicating time (before and after the YP), and another, differentiating the treatment group and the control group. The interaction between these two variables shows the effects of the Youth Parliament. Because other factors may interfere with attitude changes during the program, we included in the model a set of controls associated with socioeconomic conditions and individual attributes.

We organized the data to analyze changes both on an individual level and in the context of program implementation in school.⁷ On an individual level, we classified the students into a treatment group and a control group. We created two categories: 0 = no participation and 1 = YP participant.⁸ For contextual analysis, the classification criterion was the success the program showed in each school. Two types of information were used as a proxy for success: the number of times students in each school engaged in debate and the average percentage of aggregate attendance in Youth Parliament workshops at each school. The volume of talk indicates whether the YP was a dynamic experience. In the second case, the average aggregate attendance indicates whether the experience, as a whole, was motivating. The scale of success included the following categories: 0 = no participation, 1 = participant in a school in which the YP was not successful (neither dynamic nor motivating), 2 = participant in a school in which the YP was successful (dynamic or motivating), 3 = participant in a school in which the YP was very successful (dynamic and motivating).

Graph 1
Trust in MGSA

Our two-level model of analysis aims to demonstrate the advantage of using explanatory models of change triggered by political socialization experiences. These models include distinctions in terms of participation in the program and its context.⁹

Table 2 shows the effect of the YP on the individual level of analysis. The data are unequivocal: participation in the program increased confidence in the MGSA. Even when we control for family and school environments and for individual characteristics, such as political efficacy and exposure to political information in the media, participants in the Youth Parliament were more than twice as likely to trust the MGSA than were students who did not participate in the program.

Table 2
Dependent variable: Trust in MGSA(binary)

	B	Std Error	Exp(B)
Time	-0.432	0.235	0.649*
Treatment group	0.152	0.246	1.165
Treatment group * Time	0.928	0.350	2.530***
Parents’ education	-0.134	0.159	0.874
Parents’ participation	-0.030	0.093	0.970
Military school (0 = Public school)	-0.150	0.315	0.860
Middle-class school (0 = Public school)	0.119	0.363	1.127
Elite school (0 = Public school)	0.404	0.417	1.498
Political efficacy	0.134	0.181	1.144
Exposure to information in the media	0.053	0.034	1.055
Sex (0 = woman)	0.269	0.182	1.309
Constant	-0.991	0.385	0.371***
N			570
Nagelkerke R2			0.070
Overall percentage			60,9%
Chi-square			30.680

• ^***
  Statistically significant 0.01 /
• ** statistically significant 0.05 /
• ^*
  statistically significant 0.10

Table 3
Dependent variable: Trust in MGSA(binary)

	B	Std Error	Exp(B)
Time	-0.432	0.235	0.649*
Treatment group in school not successful	0.002	0.383	1.002
Treatment group in school not successful * Time	0.705	0.518	2.024
Treatment group in school successful	0.352	0.311	1.421
Participant in school successful * Time	0.860	0.433	2.363***
Treatment group in school very successful	-0.258	0.516	0.773
Treatment group in school very successful * Time	1.573	0.676	4.821***
Parents’ education	-0.103	0.162	0.902
Parents’ participation	-0.026	0.093	0.974
Military school (0 = Public school)	-0.039	0.343	0.962
Middle-class school (0 = Public school)	0.068	0.368	1.070
Elite school (0 = Public school)	0.443	0.421	1.557
Political efficacy	0.133	0.182	1.143
Exposure to information in the media	0.054	0.034	1.056
Sex (0 = woman)	0.263	0.183	1.301
Constant	-1.020	0.388	0.361***
N			570
Nagelkerke R2			0.077
Overall Percentage			60,7%
Chi-square			33.628

• ^***
  sStatistically significant 0.01 /
• ** statistically significant 0.05 /
• ^*
  statistically significant 0.10

Program effects were found to also depend on the program's success in schools. The program showed significant effects on participants in schools in which, according to our criteria, it was considered successful. These effects were much larger when the program was entirely successful. In this case, the students’ trust in the assembly was almost five times higher than trust displayed by students in the control group.

In summary, the impact of the YP on students’ attitudes depends on an individual factor – participation in the program – and on a contextual factor associated with the quality of the program in the school.

The Cognitive Basis of Attitudinal Change

How should the increase in confidence in MGSA found in young people who participated in the Youth Parliament be interpreted? We consider two possible interpretations. First, the change in attitude could have occurred spontaneously as a result of the bond formed with the institution, its employees, and representatives. In this case, the sudden increase in confidence and positive attitudes toward the state assembly would be tied to a purely emotional basis.

The other possible reason – which we argue to be the case – is that confidence in the MGSA occurs in an environment of constant information flow regarding the legislative process, the institution, its function and activities, and its interaction with the community. This does not mean that high school students acquired, in only a few months, deep and detailed knowledge of issues that even most educated citizens lack. Even so, their attitudinal change is expected to be accompanied by cognitive changes, albeit modest ones.

Our research provides data that allow us to address the cognitive foundations of attitudinal changes. At first we tested a political information index. This model leads to the inevitable conclusion that the Youth Parliament did not provide cognitive gains.¹⁰ Young people who participated in the program “learned” next to nothing of prominent political leaders, party ideologies, or even about acronyms that stand for political and social organizations. This is not surprising because the YP does not prepare young people for a test of their general knowledge of politics. A significant portion of the program's content involves broader issues, such as concepts of citizenship, democracy, and political participation.

However, it is not the case that the YP taught participants only abstract concepts. Upon completion of the program, participants were able to name a greater number of state representatives. The YP also proved efficient in transmitting information about events created by the MGSA to stimulate dialogue and the participation of organized sectors of civil society in the legislative process.

Table 4
Dependent variables: Mentions of deputies and events in the MGSA

	Deputies			Legislative events
	B	Std Error	Exp(B)	B	Std Error	Exp(B)
1 = Medium level of knowledge	-3.447	0.651	1.134	-0.712	0.395	1.290
Intercept	0.126	0.418	2.969***	0.255	0.246	1.570*
Time	1.088	0.387	1.361	0.451	0.269	0.953
Treatment group	0.308	0.529	1.035	-0.048	0.391	1.032
Treatment group * Time	0.035	0.244	1.088	0.031	0.172	1.114
Parents’ education	0.084	.0131	1.024	0.108	0.108	1.091
Parents’ participation	0.024	0.513	0.629	0.087	0.327	0.845
Military school (0 = Public school)	-0.464	0.606	1.577	-0.168	0.384	0.571
Middle-class school (0 = Public school)	0.455	0.640	0.639*	-0.560	0.447	0.809
Elite school (0 = Public school)	-0.448	0.258	1.174***	-0.212	0.194	1.079**
Political efficacy	0.160	0.054		0.076	0.036
Exposure to information in the media
2 = High level of knowledge	-5,431	1.055	0.314	-3.148	0.619	1.171
Intercept	-1.160	0.842	2.593*	0.158	0.331	1.088
Time	0.953	0.555	7.268**	0.084	0.356	3.008**
Treatment group	1.983	0.950	1.262	1.101	0.486	1.525*
Treatment group * Time	0.233	0.365	1.411*	0.422	0.234	1.534***
Parents’ education	0.344	0.181	0.322	0.428	0.124	1.892
Parents’ participation	-1.134	0.814	0.528	0.638	0.506	1.065
Military school (0 = Public school)	-0.638	0.867	0.345	0.063	0.570	0.594
Middle class school (0 = Public school)	-1.063	0.975	0.438***	-0.520	0.640	1.244
Elite school (0 = Public school)	-0.826	0.376	1.537***	0.219	0.250	1.270***
Political efficacy	0.430	0.097		0.239	0.050
Exposure to information in the media
N			570			570
Nagelkerke R2			0.186			0.186
Chi-square			104.77			104.77
Reference category: no knowledge

• ^***
  Statistically significant 0.01 /
• ^**
  statistically significant 0.05 /
• ^*
  statistically significant 0.10

The YP provided an excellent education on legislative procedures, especially the ones students were in contact with during the program. As an example, students who participated in the YP were fifteen times more likely to be familiar with markup procedure (by means of which bills are amended during debate), even after controlling for variables of family and school environments.¹¹ The Youth Parliament also contributed to informing its participants of issues in local politics. One indication of this knowledge is that participants knew the names of more mayoral candidates in Belo Horizonte city in the 2008 elections than the control group.

These data suggest that the increase in confidence in and knowledge of the MGSA and related issues occurred during the same period and resulted from the same political socialization experience. The simultaneity and common origin of cognitive and attitudinal changes does not necessarily mean that these two phenomena are linked. However, they suggest that the program had double effects. Even if young participants failed to use knowledge acquired during the program, or the knowledge contributed to supporting (or even reducing) their previous trust in the state legislature, YP participants were prepared to form their attitudes on more solid cognitive bases.

A set of questions about the participants’ perceptions of the MGSA contributed to the analysis of the relationship between attitudinal and cognitive changes. Weasked the participants if they considered the MGSA transparent, dynamic, independent and open to the participation of civil society. As opposed to expressions of pure confidence in the MGSA, these questions have a clear evaluative component that depends on respondents’ existing information regarding the institution. For this reason, each question offers the alternative “I don't know”.

Table 5
Image of the MGSA (data in %)

	Openness						Dynamism
	Wave 1		Wave 2			Wave 1		Wave 2
	T C		T	C		T	C	T	C
It is open	56.3	39.3	79.6	47.0	It is dynamic	26.5	22.6	41.0	21.4
It is not open	12.0	21.4	6.0	17.9	It is not	35.5	31.0	26.5	32.7
Don't know	31.7	39.3	24.8	35.1	dynamic	38.0	46.4	32.5	45.8
					Don't know
Transparency					Independence
	Wave 1		Wave 2			Wave 1		Wave 2
	T	C	T	C		T	C	T	C
It is transparent	32.3	24.4	48.5	25.7	It is independent	38.9	31.0	51.8	38.9
It is not transparent	41.9	42.9	34.1	41.9	It is not independent	15.6	22.6	15.1	24.0
Don't know	25.7	32.7	17.4	32.3	Don't know	45.5	46.4	33.1	37.1

• T = treatment group / C = control group

We can see a significant shift of responses to positive attitudes among the YP participants.¹² If this were simply an indication of a migration of negative to positive attitudes, there would be no novelty in relation to what we already know about the increase in confidence in the MGSA. The additional information supplied by the data is the migration of “I don't know” responses to favorable opinions of the MGSA. After participating in the program, high school students became more favorable toward the MGSA and more confident in their opinion thereof.

Therefore, we know that young people who participated in the program changed their attitudes in the same direction, learned about politics, and began to feel more secure in regards to their evaluation of the MGSA. However, can we conclude that the increase in confidence in MGSA represents attitudinal “gains”? There are strong indications that these changes were, in fact, gains, but any measure of certaintythat increased confidence in the MGSA has a cognitive basis would require the identification of a direct association between attitudes and cognition.

To answer this question, we used the method of “difference in difference.” The variables used were derived by subtracting the values at time 1 (before the YP) from values at time 2 (after the YP took place) for each of the two groups (treatment and control). The result of this transaction represents a change in the variable values over time for each group.

The answer, although modest, is positive. We find a positive correlation between knowledge of MGSA-related events and trust in the institution. The same is not true for knowledge regarding Minas Gerais’ state representatives. However, this does not mean that the knowledge gains and attitudinal gains are independent processes. This point is emphatically supported by the association between perceptions of the assembly and knowledge items. A comparison between the group that participated in the program and the control group shows that the attitudes of the former have stronger cognitive bases. First, there is a clear association between the increase of knowledge about the representatives and the increase in the MGSA's positive image. The correlation between knowledge acquisition regarding legislative events and an increase in positive evaluations of the state assembly is even more significant. The control group gave no indication of an association between cognitive and attitudinal change.

Table 6
Correlations between attitudes and knowledge

	Non-participants		Participants
	Name of deputies	Knowledge of events	Name of deputies	Knowledge of events
Trust in MGSA	-0.049	0.005	0.058	0.159**
Evaluation of MGSA	0.124	0.088	0.168*	0.245***

• ^***
  Statistically significant 0.01 /
• ^**
  statistically significant 0.05

Thus, it can be said that the political socialization that took place during the program not only resulted in an increase in knowledge and confidence in MGSA but also that many participants who changed their attitudes acquired information that established consistent attitudes regarding the MGSA. This does not mean that a “gain” in confidence is devoid of any affective component; it means that this gain cannot be accounted for solely by that component.¹³

One question remains: are there grounds, in political reality, for these attitudinalgains? Why does learning about theMGSAincrease respondents’ level of confidence? The data fromour studydo not providea direct answer to this question. We know, however, that what makes the Minas Gerais State Assembly unique among Brazilian subnational legislatures is the new institutional setting that emerged in the early 1990s. In an atmosphere of institutional innovation (Anastasia2001), the MGSA modernized its structure and administration, increased staff professionalism, and became more democratic in its organization, by outreaching to the community and establishing an ongoing dialogue and participationin the legislative process. The Youth Parliament is an offshoot of this new context, and participants have witnessed and experienced this reality.

Concluding Remarks

The research results presented in this article are straight forward. In as little as a few months, the students who attended the Youth Parliament activities showed increased confidence in the Minas Gerais Legislative Assembly. That confidence increased even further when the program in the school was found to be dynamic and motivating. Thus, attitudinal changes can occur at the same pace as (or an even faster than) changes in political knowledge and participation.

Furthermore, after attending the program, youngsters became more aware of legislative procedures and the state representatives, more knowledgeable in regards to the MGSA and more willing to obtain information from the media about the state assembly. They did not necessarily gain a great deal of knowledge of politics. Far from leaving the program with an encyclopedic knowledge (Lupia 1994) of the topic, students’ cognitive gains were found to be limited and fragmented; however, these gains were more than enough to broaden their understanding of politics and, particularly, of the MGSA. Knowing the names of Minas Gerais’ state representatives and mayoral candidates, events organized by the MGSA, and legislative procedures seems to say little on the formation of politically competent citizens. However, one may be demanding too much in expecting high school students who took part in a dozen meetings in a three-month program to show even more significant gains .

Besides, we assume that it is possible to relate the limited knowledge items of our questionnaire to a wider repertoire of knowledge. It is reasonable to expect that students who can name representatives also have more knowledge of their activities than those who cannot. Likewise, students who are able to identify the activities promoted by the MGSA are more likely to have more information about the institution than those who do not.

In short, we found that participation in the Youth Parliament did lead to attitudinal gains. Let us keep in mind, however, that public confidence in political institutions is not intrinsically valuable to democracy. In this article, “gains” mean that the increased confidence in the MGSA has cognitive bases, which is a desirable outcome.

It is important to note that gains from the YP presented in this article are specifically related to the state legislature and not to other institutions or political actors. This distinction works as an antidote to the common and undesirable effect of “contamination” (Parker 1977; Rennó and Gramacho 2010) in perceptions of political institutions and actors. We would not be surprised if new evidence revealed that the shield that prevents positive attitude changes towards the MGSA from extending to state deputies and other political institutions is built by cognitive gains provided by the YP program.

• 1.
  Available at http://www2.almg.gov.br/hotsites/parlamentoJovem/oqueE.html. Accessed 12 April, 2011.
• 2.
  The project was designed to include 351 interviews in the pretest and 351 in the post-test, for a total of 702 interviews. However, we were only able to interview 335 students in the post-test. Thus, we decided to exclude from the analysis cases missing in the post-test.
• 3.
  The first round was conducted between February and April 2008, and the second round was conducted between June and November 2008.
• 4.
  Since the 1990s, there has been a fruitful debate on trust (or mistrust) in political institutions and its effect on democracy (Dalton 2004; Inglehart 1999; Pharr and Putnam 2000; Norris 1999). This debate calls into question the assumption of the classical theory of political culture, which suggests that confidence in political institutions fosters the stability of the democratic regime.
• 5.
  Regarding trust in the MGSA, there was no difference between the treatment group and the control group prior to the experience of the YP, which confirms the reliability of pairing both groups. On a scale of 0 to 3, the average confidence of the control group was 1.41, whereas for the participants it was 1.50. This difference is not statistically significant (0.289).
• 6.
  A test of means indicates that the differences shown in the chart are statistically significant. The difference between time 1 and time 2 was −0.12 for non-participants (significance 0.06) and 0.19 for participants (significance 0.008).
• 7.
  Because there were only 14 schools, we decided not to use multilevel regression models. The dependence of observations within each school was corrected by using the robust cluster in Stata.
• 8.
  The total number of workshops (including those that occurred at school and at the MGSA) ranged from ten to twelve, depending on the school. The total number of meetings in school ranged from seven to ten.
• 9.
  Ideally, this model would integrate the individual factor (including the degree of exposure of individuals to the program) and the contextual factor (the success of the YP at school). However, because of the limited number of observations in our database, it was not possible to create an integrated model with variables that included both levels of analysis.
• 10.
  The construction of the information index that we tested appears in the appendix of this article. We used a linear regression model in which the variable of political information was the dependent variable and the variables of time, treatment and their interaction were independent. The coefficients were not statistically significant.
• 11.
  We used a logistic regression model with the following controls: parental education, parental political participation, school type, subjective efficacy, media exposure and sex. The significance level was 0.001.
• 12.
  The questions were as follows. 1) In your opinion, is the Minas Gerais State Assembly closed to the participation of society or open to the participation of society?; 2) In your opinion, is the Minas Gerais State Assembly a transparent institution that informs people about its acts or an institution that is not transparent and does not show the public what happens there?; 3) In your opinion, is the Minas Gerais State Assembly an institution without initiative that simply fulfills its duties or a dynamic and innovative institution?; 4) In your opinion, does the Minas Gerais State Assembly lack freedom of choice and yield to the will of the governor, or is it an independent institution that acts according to its own will? All questions offered the possible answer “don't know.”
• 13.
  However, our research has no data with which to test this hypothesis.
• Revised by Marcus Lessa

Bibliographical References

• Almond, G., and S. Verba.1989. The civic culture: Political attitudes and democracy in five nations Princeton: Princeton University Press.
• Anastasia, F. 2001. “Transformando o Legislativo: a experiência da Assembleia Legislativa de Minas Gerais”. In O poder Legislativo nos estados: diversidade e convergência, ed. F. Santos, 23-83. Rio de Janeiro: Fundação Getulio Vargas.
• Andolina, M. W., K. Jenkins,C. Zukin, and S. Keeter. 2003. Habits from home, lessons from school: Influences on youth civic engagement. Political Science and Politics 36(2): 275-80.
• Campbell, D., and J. Stanley. 1979. Delineamentos experimentais e quase-experimentais de pesquisa São Paulo: EDUSP.
• Campbell, D. E. 2008. Voice in the classroom: How an open classroom climate fosters political engagement among adolescents. Political Behavior, 30: 437-54.
• Dalton, R. J. 2004. Democratic challenges, democratic choices Oxford: Oxford University Press.
• Easton, D. 1965. A systems analysis of political life New York: Wiley.
• Finkel, S. E. 2003. Can democracy be taught? Journal of Democracy, 14: 137-51.
• Finkel, S. E. and H. R. Ernst. 2005. Civic education in post-apartheid South Africa: Alternative paths to the development of political knowledge and democratic values. Political Psychology 26 (3): 333-64.
• Finkel, S. E., C. A. Sabatini, and G. G. Bevis.2000. Civic education, civil society, and political mistrust in a developing democracy: The case of the Dominican Republic. World Development, 28: 1851-74.
• Finkel, E. S., and E. A. Smith. 2011. Civic education, political discussion and the social transmission of democratic knowledge and values in a new democracy: The 2002 Kenya. American Journal of Political Science 55 (2): 417-35.
• Galston, W. A. 2001. Political knowledge, political engagement, and civic education. Annual Review of Political Science 4: 217-34.
• Ichilov, O. 2007. Civic knowledge of High School students in Israel: Personal and contextual determinants. Political Psychology 28 (4): 417-40.
• Inglehart, R. 1999. Trust, well-being and democracy. In: M. E. Warren, ed. Democracy and trust Cambridge: Cambridge University Press.
• Inglehart, R., and C. Welzel. 2005. Modernization, cultural change, and democracy: The human development sequence New York: Cambridge University Press.
• Jennings, M. K.,and R. G. Niemi. 1974. The political character of adolescence: The influence of families and schools Princeton: Princeton University Press.
• Jennings, M. K.,andG. B. Markus. 1984. Partisan orientations over long haul: Results from the three-wave political socialization panel study. The American Political Science Review 78 (4): 1000-18.
• Lupia, A. 1994. Shortcuts versus encyclopedias: Information and voting behavior in California insurance reform elections. American Political Science Review 88 (1): 63-76.
• Niemi, R., and J. Junn. 1998. Civic education: What makes students learn New Haven and London: Yale University Press.
• Norris, P. 1999. Critical citizens: Global support for democratic government Cambridge: Oxford University Press.
• Offe, C. 1999. How can we trust our fellow citizens? In M. E. Warren, ed. Democracy and trust Cambridge: Cambridge University Press.
• Parker, G. R. 1977. Some themes in congressional unpopularity. American Journal of Political Science 21(1): 93-109.
• Pharr, S. J., and R. D. Putnam. 2000. Disaffected democracies: What's troubling the trilateral countries? Princeton: Princeton University.
• Renno, L., and W. Gramacho. 2010. Let's blame everyone: Executive and Legislative evaluations of economic performance in Brazil and Chile. Journal of Politics in Latin America 2 (1): 53-78.
• Sapiro, V. 2004. Not your parents’ political socialization: introduction for a new generation. Annual Review Political Science 7: 1-23.
• Sears, D. O., and C. L. Funk. 1999. Evidence of the long-term persistence of adults’ political predispositions. The Journal of Politics 61 (1): 1-28.
• Sears, D. O., and N. A. Valentino. 1997. Politics matters: Political events as catalysts for preadult socialization. The American Political Science Review 91 (1): 45-65.
• Slomczynski, K. M., and G. Shabad. 1998. Can support for democracy and the market be learned in school? A natural experiment in post-communist Poland. Political Psychology 19 (4): 749-79
• Warren, M. 1999. Democracy and trust. Cambridge: Cambridge University Press.

Appendix

Parents’ education. A construct was created by factor analysis (principal axis). It generated a variable of parents’ education, which combined the years of schooling of the respondent's father and mother. The two original variables were measured in terms of a complete or incomplete educational degree. To turn these variables into continuous variables, following the strategy of Bills and Haller (1984), we assigned the following values to the years of schooling: 0 (never attended school), 2 (1st to 4th grade of basic education, incomplete), 4 (1st to 4th grade of basic education, complete), 6 (5th to 8th grade of basic education, incomplete), 8 (5th to 8th grade of basic education, complete), 9.5 (1 to 3 years of secondary education, incomplete), 11 (1 to 3 years of secondary education, complete), 13 (incomplete higher education), 15 (college degree), 16 (graduate education, incomplete) and 17 (graduate degree).

Parents’ political participation. A construct was created using factor analysis (principal axis). It generated a variable of parents’ political participation, which combined the political participation of the respondent's father and mother. The original individual values for the political participation of the father and mother were the sum of the number of associations, parties, and trade unions with which they were engaged.

Types of schools. One of the variables in our database sought to capture the school environment in a single indicator. The categories were defined as follows: public school, public military school, middle-class private school and elite private school. We created dummy variables from the four categories. The criteria used for the classification of schools as elite and middle class were tuition and geographic location. No middle-class school tuition was higher than R$ 520, whereas the fees of elite schools ranged from R$ 642 to R$ 738. Elite schools, in addition to being publicly recognized as the “best schools,” were located in the most valued neighborhoods of Belo Horizonte city. In the case of public schools and the military (also public), the classification represents their formal identity.

Subjective political efficacy. We provided the statement, “Sometimes politics and government seem so complicated that a person like you cannot really understand what is happening.” The answer was given on a Likert scale of five points, with the following options: strongly agree, partly agree, neither agree nor disagree, partly disagree and strongly disagree. We created a binary variable, excluding the middle option, so that each of the new categories included all of the attitudes expressed in one of two directions (agreeing or disagreeing).

Exposure to information in the media. We asked how often the interviewee was exposed to political information through mass media, specifically through three types of media: a) TV and radio, b) newspapers and magazines and c) the Internet. Each of the three types of media was given a rating for exposure (0 = never, 1 = rarely, 2 = a few times a month, 3 = a few times a week, 4 = daily). The final variable was the sum of the answers to three questions.

Factual knowledge. We constructed an index of respondents’ political information by factor analysis (principal axis) from four different sets of questions about political affairs. The four sets were classified as follows. 1) Knowledge of leading politicians’ offices and political parties: the questions were open and asked respondents to state the party and office held by four leading politicians in Brazil: AecioNeves, Fernando Pimentel, José Serra, and Martha Suplicy. There was also a question about the office and country of Hugo Chavez. The questions were coded according to four levels of accuracy (3 = full accuracy, 2 = incomplete accuracy, missing some information either about the name of the office or its state level, 1 = stated the politician's previous office, 0 = error or no response). The final index for the set of variables was the result of the factor analysis (principal axis). Cronbach's alpha was 0.867, and all variable loads were satisfactory. 2) Number of parties that the respondent was able to name: this set of variables was a list of the parties the interviewee knew. For every Brazilian party named correctly, the interviewee earned a point, while a non-response or error was worth zero. The index was constructed by the sum of the scores and ranged from 0 to 11. 3) Knowledge of the meaning of abbreviations for political and social organizations: respondents were asked to state the meaning of TRE (Regional Electoral Tribunal), ONU (United Nations), UNE (National Union of Students) and MST (Landless Workers’ Movement). The questions were open and were post-coded into three levels (2 = full meaning; 1 = partial meaning, restricted its general subject, 0 = error or no response). The index was constructed by factor analysis (principal axis). The Cronbach's alpha was 0.597, and all variable loads were satisfactory. 4) Knowledge of the ideological position of parties: the respondent was asked to state whether each of four Brazilian parties (DEM, PCdoB, PSTU and PP) were left or right. Each correct answer was worth 1 point. The index was the sum of correct answers.

Trust in Minas Gerais State Assembly. The question asked was, “Regarding Minas Gerais State Assembly, you (1) never trust it, (2) trust it sometimes, (3) trust it most of the time or (4) always trust it.” A binary variable was created, with “never trust” and “trust sometimes” = 0 and “trust most of the time” and “always trust” = 1.

Knowledge of the names of deputies. The following question was asked: “Could you tell me the name of any deputy of the Minas Gerais State Assembly?” Two points were awarded for the representative's full name, and one point was awarded for an incomplete name. This scale was then categorized for the multinomial model, with 0 = 0, 1 = 2, and 3 = all others.

Knowledge of Minas Gerais State Assembly forums for civil society. The question asked was, “The Minas Gerais Legislative offers many possibilities for participation to the population. Could you name any of them?” In case of no spontaneous response, the interviewee was asked if he knew of the following: a) legislative seminars, b) technical forums, c) public hearings or d) commission of popular participation. The variable was then constructed from the spontaneous and forced responses. Each spontaneous response was assigned two points. Each forced response was assigned one point. The responses were added to create a scale from zero to four. This scale was then categorized for the multinomial model, with 0 = 0, 1 = 1, and 2 = all others.

“Markup.”. The question asked was, “Markup is a legislative procedure. A bill is subject to markup a) when the representative wants to express full support for its content, b) when the subject was superficially addressed, c) to promote changesto the bill, d) to block a vote or e) you don't know.” The correct answer (c) was assigned the value 1, and all other answers were assigned the value 0.

Knowledge of the candidates for mayor in the 2008 elections. The following question was asked: “Who are the mayoral candidates in Belo Horizonte's 2008 election whose names you remember?”.

Publication Dates

• Publication in this collection
  27 Oct 2023
• Date of issue
  2012

History

• Received
  Feb 2012
• Accepted
  May 2012