Childhood Vaccination Policy

bs_bs_banner

The Policy Studies Journal, Vol. 42, No. 4, 2014

Cultural Worldview and Preference for Childhood
Vaccination Policy
Geoboo Song, Carol L. Silva, and Hank C. Jenkins-Smith
In the face of the reemerging threat of preventable diseases and the simultaneous vaccine risk controversy, what explains variations in Americans’ policy preferences regarding childhood vaccinations? Using original data from a recent nationwide Internet survey of 1,213 American adults, this research seeks to explain differing public opinions on childhood vaccination policies and related issues of governance. As Mary Douglas and Aaron Wildavsky’s grid-group cultural theory of policy preference formation suggests, cultural biases have a significant impact on the
formation of preferences toward various vaccination policies. Hierarchs are in support of mandatory vaccination, oppose religious and philosophical exemption, and believe the government should preside over vaccinationrelated decisions. Fatalists strike a bold contrast in their opposition to mandatory vaccination policy and support for religious and philosophical exemptions and the role of parents in deciding on vaccinations. Falling between hierarchs and fatalists, egalitarian support for vaccinations is stronger than individualists‘.
KEY WORDS: childhood vaccination policy, health policy, cultural theory, public opinion

Introduction
The recent resurgence of deadly, but preventable, communicable diseases in the
United States has grown into a serious public health concern. In 2000, for instance, the United States announced successful elimination of measles (Centers for
Disease Control and Prevention [CDC], 2012), but recent outbreaks continue to pose a threat to public health. As the disease resurfaces in areas around the United States, the general public faces the risk of sustained transmission; 402 confirmed measles cases in 18 states, including Ohio, California, and New York, were reported to the CDC so far in 2014, which will make 2014 a record year for measles outbreaks in the past 18 years, and if this trend continues, 2014 may become the worst year for measles in decades. Improvement of public health by immunization through vaccinations has been a consistent policy in the United States for decades. In the late 1950s and

528
0190-292X © 2014 Policy Studies Organization
Published by Wiley Periodicals, Inc., 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington Road, Oxford, OX4 2DQ.

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

529

early 1960s, the federal government established a nationwide vaccination policy based upon the Vaccination Assistance Act (Calandrillo, 2004; Rein,
Honeycutt, Rojas-Smith, & Hersey, 2006). Accordingly, by the late 1960s, several states established mandatory vaccination policies for children upon school entry against an array of infectious diseases and, by the late 1970s, all 50 states had adopted this vaccine mandate. Currently, all states require vaccinations against measles, polio, rubella, and diphtheria while vaccinations against other major contagious diseases are either required or recommended.1 Meanwhile, the government has also instituted policies to address various concerns associated with vaccination. In order to address the possibility of rare and scientifically unverified, but potentially grave, side effects from vaccinations, for example, the federal government established the National Childhood Vaccine
Injury Act in 1986. The National Childhood Vaccine Injury Program, based upon this legislation, provides compensation for injury or death resulting from an adverse reaction to a vaccination without requiring a confirmatory investigation of the responsible party, through a federal “no-fault” system (Barringer, Studdert,
Kachalia, & Mello, 2008; Elliott, Narayan, & Nasmith, 2008). Beyond these compensation allowances, public health policy provides avenues for vaccine avoidance.
Currently, all 50 states allow medical exemption from vaccinations for those children who can be expected to develop serious allergic reactions.2 Another 48 states, excluding Mississippi and West Virginia, allow for religious exemption from the vaccine mandate.3 In 19 states,4 a philosophical (nonreligious belief) exemption5 is also permitted.
In sum, vaccine policy generally seeks near universal vaccinations to maintain herd immunity levels for the population at large, while at the same time providing multiple avenues for “opting out.” This array of seemingly divergent policy directions makes evident the ongoing struggle between efforts to enforce vaccination requirements for the benefit of public health and defense of individual rights based upon religious or philosophical convictions. Proponents of mandatory vaccinations argue that the government should limit the scope of religious and philosophical exemptions (which, if widely exercised, will result in a declining vaccination rate) because in their view, the benefits of freedom from infectious diseases, both at the individual and societal level, far outweigh the costs of restricted parental choice or the physical risks posed by vaccinations. Therefore, proponents urge that exemptions be limited to those based upon verified health threats to the individual (Salmon et al., 2005). By contrast, vaccination opponents argue that the focus should now be shifted to the risks of vaccinations6 because the threat of infectious diseases has been diminished in modern societies, and because individuals (and parents) should have the right to make decisions about vaccinations based upon their personal beliefs (Mariner, Annas, & Glantz,
2005; Frontline, 2010; Wallace, 2009; Woo et al., 2004). In the face of the reemerging threat of preventable deadly diseases and in the midst of the vaccine risk controversy, how can we account for these differences in policy preference regarding vaccinations? 530

Policy Studies Journal, 42:4

Theoretical Conjectures
Grid-Group Cultural Theory of Policy Preference Formation
Much previous research implies that personal values and beliefs exert a critical influence on policy preference in various domains (e.g., Colcord, 1977; Doan &
Kirkpatrick, 2013; Gerber & Neeley, 2005; McAvoy, 1998; Pereira & Van Ryzin, 1998;
Sabatier & Jenkins-Smith, 1993; Wildavsky, 1987).
Because public policy can be considered an institutional element designed to resolve a particular social problem, and because public policy based upon due process and social consensus can be understood as a norm and rule that defines social relationships, one’s preference for a particular public policy is derived not from a simple benefit-cost calculation, but rather from individual evaluation of the nature of influence a given policy, rule, or norm has upon a preferred “way of life” (e.g., Jenkins-Smith & Herron, 2009;
Jenkins-Smith & Smith, 1994; Kahan, Braman, Cohen, Gastil, & Slovic, 2010; Lodge,
Wegrich, & McElroy, 2010; Silva & Jenkins-Smith, 2007; Swedlow, 1994, 2011;
Thompson, Ellis, & Wildavsky, 1990).
Though holding much broader theoretical and empirical applications, gridgroup cultural theory, in this case, can also be understood as providing an avenue
for a direct application of such theoretical postulations in understanding diverse public attitudes toward the current state level childhood vaccination policies. Generally, cultural theory claims that individuals’ cultural bias or distinctive cultural worldview, as more intrinsic personal values and beliefs constituting “thick” or culturally constrained rationalities, shape their policy preferences (Wildavsky, 1987), which translate into their preferred institutional arrangements (Ferejohn, 1991;
Lockhart, 2001). Furthermore, cultural theory asserts that, cognitively and institutionally, diverse cultural biases or predispositions coexist within various levels of the system, and, in fact, rely on one another, not only for comparative reinforcement of distinctive cultural identities, but also for compensation for each others’ deficiencies for various socio-cultural contingencies, which leads to a long-term systemic sustainability, or the principle of “socio-cultural viability” (Douglas, 1996). We argue that divergent public attitudes regarding key issues pertaining to current childhood vaccination policies, such as mandatory vaccinations, various exemptions, and the issue of governance over vaccine decision making, are molded by individuals’ diverse cultural outlooks and are, in turn, interactively reflected in the related policies, an important component of institutional arrangement concerning childhood vaccinations, as cultural theory suggests.
More specifically, cultural theory seeks to characterize the scope and nature of preferred ways of life—different philosophical orientations for social relationships— based upon two theoretical dimensions: group and grid (Douglas, 1970; Ripberger,
Song, Nowlin, Jones, & Jenkins-Smith, 2012; Thompson et al., 1990; Wildavsky, 1987;
Wildavsky & Dake, 1990).
As depicted in Figure 1, group refers to what degree individuals’ social relations are governed by group membership or “bounded units” within a society (Thompson et al., 1990, p. 5). Grid indicates to what degree individuals’ social relationships are

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

531

Grid
Strong

Fatalists
(Life is luck!)

Hierarchs
(Institutionalized authority)

Group
Weak

Strong
Individualists
(Liberty)

Egalitarians
(Equality)

Weak

Figure 1. Grid-Group Cultural Theory Framework.

determined by “externally imposed prescriptions” such as rules or social norms
(Thompson et al., 1990, p. 5). So, the strength or weakness of “group boundaries”
(group) and the number, nature, and diversity of the various “prescriptions” (grid) enacted upon individuals formulate their cultural outlook, i.e., “shared values legitimating social practices” (Wildavsky, 1987, p. 6). Based upon these two theoretical dimensions of sociality are four prototypical—and theoretically orthogonal— cultural orientations: hierarchism, egalitarianism, individualism, and fatalism.
The hierarch orientation is rooted in a preference for strong group attachment and numerous social prescriptions that clearly define roles in society. This cultural bias emphasizes authority in the belief that social division of labor based upon specialization and expertise (rather than upon equality among the members of society) contributes to the well-being of society as a whole (Douglas, 1970;
Ripberger et al., 2012; Thompson et al., 1990; Wildavsky, 1987; Wildavsky & Dake,
1990). Considering that most experts assert that vaccinations are the most effective way to improve public health through prevention of diseases and that the health risks when weighed against the benefits are negligible, it can be conjectured that those with strong hierarchical tendencies will favor mandatory vaccination policy.
This policy preference is also attributed to the fact that mandatory vaccination policy is characterized as a government prescription emphasizing collective benefits over individual risk. With the same reasoning, those with a strong hierarch bias will tend to reject the various exemption policies. That is, we expect hierarchs to oppose such policies because they are seen to resist expert opinion, focus on individual concerns rather than societal benefits, and are based upon “exceptional” cases which encourage defection from the existing institutional order (of mandatory vaccination policy). From the perspective of risk governance, which is related to the broader framework of policy decision making, group-oriented hierarchs will tend to believe that eradication and elimination of infectious diseases is the responsibility of the community and not of individuals. Therefore, hierarchs will believe

532

Policy Studies Journal, 42:4

that the government, not children’s parents, should be the chief decision maker regarding childhood immunizations.
The egalitarian orientation, like that of hierarchs, is based upon strong group affinities. However, its grid orientation is weak; egalitarians do not desire social relationships depending upon stratified institutions or the rules imposed by expert
“outsiders” (Douglas, 1970; Ripberger et al., 2012; Thompson et al., 1990; Wildavsky,
1987; Wildavsky & Dake, 1990). That is, egalitarians prefer equal social relationships and dislike institutional infringement and authority imposed from outside the group. A robust sense of group orientation makes egalitarians more concerned about societal risk posed by jeopardizing public health through decreased herd immunity than about infringement upon individuals’ choice not to be vaccinated. Because they are most concerned with societal-level well-being, they will tend to support government-mandated vaccination policy. However, egalitarians are likely to support such mandates to a lesser degree than hierarchs due to their aversion to coercion by concentrated authority based on expert (rather than community) consensus. This cultural characteristic will also have an impact on various vaccine exemption policies; considering that exemption policies focus more on individual rather than collective benefits, like hierarchs, egalitarians will oppose them, but to a lesser degree, due to their skepticism concerning policies based upon centralized institutional controls rather than community consensus. Group-oriented egalitarians will be prone to think that the elimination and eradication of infectious diseases is a societal, not an individual, responsibility. Accordingly, they will tend to believe that the government, not parents, should be the ultimate decision makers with regard to immunization of children. However, egalitarians know that emphasizing the role of government in the immunization of children also means the institutionalization of mandatory vaccination, which lessens their level of agreement with this reinforcement of governmental authority in comparison with hierarchs.
The individualist orientation can be characterized as having both weak group and grid orientations. Individualists prefer laissez faire, contract-based social interactions based upon self-regulation and competitiveness rooted in equal opportunity (Douglas, 1970; Ripberger et al., 2012; Thompson et al., 1990; Wildavsky, 1987;
Wildavsky & Dake, 1990). They dislike authority, external prescription, and the idea of equity based on equal outcomes rather than individual merit and effort.
With respect to mandatory vaccine policy, individualists (like egalitarians) are likely to be conflicted. Individualists may tend to oppose mandatory vaccination policy because they rely on institutional prescription and coercion. At the same time, individualists will be sympathetic with these policies because they do not want non-vaccinated individuals to impose disease on others against their will; contracting a contagious disease due to the negligence or choice of others violates individualists’ preferences for individual autonomy. Nevertheless, individualists are likely to support vaccination exemptions because they oppose imposition of choice on private individuals by governments. They will tend to believe that the elimination of infectious diseases is, in the end, an individual problem. Therefore, in their view, children’s parents should make the key decisions regarding their immunization, not the government.

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

533

Table 1. Cultural Type and Hypothesized Preferences for Childhood Vaccination Policy

Hierarch
Egalitarian
Individualist
Fatalist

Mandatory vaccination

Religious exemption

Philosophical exemption

Parent should decide?

Strongly support
Conflictingly support
Conflictingly oppose
Strongly oppose

Strongly oppose
Conflictingly oppose
Conflictingly support
Strongly support

Strongly oppose
Conflictingly oppose
Conflictingly support
Strongly support

Strongly disagree
Conflictingly disagree
Conflictingly agree
Strongly agree

Based upon a weak group orientation coupled with a perception of capriciously imposed constraints, fatalists lean toward nonparticipation in social relationships and (where possible) seek to avoid the requirements imposed by society (Douglas,
1970; Ripberger et al., 2012; Thompson et al., 1990; Wildavsky, 1987; Wildavsky &
Dake, 1990). Fatalists perceive ubiquitous societal rules and distinctions, exhibiting strong grid orientation, while simultaneously perceiving life’s events as chiefly random and uncontrollable. By implication, fatalists would tend to believe that becoming infected with communicable diseases is part of one’s destiny or luck, and therefore will be skeptical of the mandatory vaccine policy designed to prevent such
diseases.
Given that life is largely governed by random events, fatalists are also likely to view the potential risks of vaccines as being as great as the benefits. Because they are likely to be skeptical of vaccine benefits and concerned about the risks, they will tend to support various vaccination exemption policies that would free them from responsibility for vaccinating themselves and their children. This type of cultural orientation will urge people to think that efforts to eliminate infectious diseases are
(at best) left to the individual, rather than depending on a (probably ineffectual) societal mandate. For fatalists, parents, not the government, should be the chief decision makers regarding children’s vaccinations.
This discussion regarding the hypothesized relationships between each distinctive cultural type and vaccine policy preferences are succinctly summarized and presented in Table 1, which is our primary concern in the following empirical analysis. It is noteworthy that although we proceeded with our theoretical discussion as if there existed a fixed prototypical cultural type of individuals, such
as hierarch, egalitarian, individualist, and fatalist, in this study, these are theoretical constructs utilized to make our discussion clearer and more understandable, and, in practice, individuals may hold multiple cultural orientations at the same time, one of which may become relatively more dominant than others, contingent upon the circumstances they encounter, as Thompson, Grendstad, and Selle (1999) suggest. To clarify our conceptual argument, when we refer to a hierarch, for example, we imply an individual who tends to possess relatively strong hierarchism, while remaining cultural orientations are consistently subdued throughout various contingencies.
Organic Culture, Ideology, Religion, and Demographics
Grid-group cultural orientations are not expected to operate in isolation, and there are a number of competing (or complementary) conjectures regarding the

534

Policy Studies Journal, 42:4

sources of vaccine policy preferences. We include a subset of those that are most promising, both as rival explanations to cultural theory and as controls in our analytical models. In doing so, we first introduce other important components of personal belief systems, such as post-materialistic “organic” subculture, political ideology, and religious beliefs, before addressing demographic characteristics, including vaccine related knowledge level and household characteristics. The post-materialistic “organic” subculture promotes personal well-being, favoring naturally based remedies and holistic, homeopathic treatments (in the belief that the human body can heal itself) over mechanized, mass-produced, and synthetically derived modern medicine based upon methodological individualism. Therefore, strong adherents to organic culture are expected generally to dislike vaccinations
(Ernst, 2001; Gellin, Maibach, & Marcuse, 2000; Lehrke, Nuebling, Hofmann, &
Stoessel, 2001), oppose mandatory vaccinations, and support exemption policies, reducing the problem of dealing with infectious diseases to the individual (or local community) level and delegating parents as the chief decision makers for their children’s vaccinations. Political ideologies are also expected to influence policy preferences over a wide range of policies (e.g., Colcord, 1977; Fiorino, 1989; Plutzer,
Maney, & O’Connor, 1998; Ripberger, Rabovsky, & Herron, 2011; Rothman &
Lichter, 1987; Wiener & Koontz, 2010). Those who are politically conservative generally tend to dislike expansive government that infringes on individual liberties.
Therefore, with regard to vaccination policies, conservatives are more likely than liberals to oppose mandatory vaccine policy, which is based upon government enforcement, and are also more prone to support various exemption policies and to believe that parents should decide whether their children should be given immunizations. Beyond political ideologies, we also expect that those for whom religion is very important in their personal lives will tend to oppose mandatory vaccination policy, support exemptions (especially those which are religiously or philosophically based) and believe that parents should be the sole decision makers for childhood vaccinations. Demographic characteristics, including vaccine related knowledge level and household characteristics, can also influence policy preferences regarding vaccinations. Considering the fact that the majority of scientists champion the effectiveness of vaccinations and have verified very few cases of severe adverse reactions, those who are more knowledgeable about the scientific consensus regarding vaccines are likely to support mandatory vaccination policy, oppose exemptions, and favor the government’s role in managing childhood vaccinations. Household characteristics also help form policy preference. For instance, the benefit/risk perceptions (and therefore, policy preference) of parents of infants and young children who are the targets of mandatory vaccine programs may differ from those of people with no children or with adult children. Additionally, females, who tend to be more riskadverse than men across a wide array of hazards (Finucane, Slovic, Mertz, Flynn, &
Satterfield, 2000; Flynn, Slovic, & Mertz, 1994; but see Kahan, Braman, Gastil, Slovic,
& Mertz, 2007; Palmer, 2003),7 may generally perceive higher levels of vaccine risk than do males. Therefore, parents with infants or young children and females may be

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

535

more likely to oppose mandatory vaccination policy, support exemptions, and prefer to make their own decisions regarding immunization of any children in their care. Other demographic characteristics such as age, race, education, and income level may also have an impact on vaccination policy preferences and interact with the other aforementioned characteristics (Timmermans, Henneman,
Hirasing, & Van der Wal, 2005). Prior research has found that on average, highly educated, older, and wealthier non-Hispanic whites8 are more likely to be more knowledgeable, which may in turn lead them to support mandatory vaccination policy and to dislike exemptions (e.g., Baker, Dang, Ly, & Diaz, 2010; Luman, Barker,
McCauley, & Drews-Botsch, 2005; Smith & Stevenson, 2008; Wooten, Luman, &
Barker, 2007).
Utilizing these rival explanations and controls, this research systematically investigates conjectures drawn from a variety of theoretical explanations for policy preferences toward childhood vaccination policy and the factors that impinge upon the formation of such preferences. However, we primarily focus upon how the general public’s cultural biases shape their vaccination policy preferences, as previously emphasized.
Data, Variables, and Measures
Survey Data
A February 2010 nationwide Internet survey involving respondents recruited by
Survey Sampling International (SSI) was conducted to measure public perceptions of vaccination risks and policy preferences. The Institutional Review Board approved the survey and overall research design for Human Research Participant Protection.
The sample for this survey study was drawn from SSI’s regular panel of approximately 400,000 Internet survey recruits. More specifically, when our Internet survey was open to SSI’s online survey panel, SSI filtered through survey entrants in a manner that the demographic characteristics of survey participants reflect national census characteristics.9
A total of 1,213 volunteers aged 18 and older who accepted an e-mail invitation describing the study completed the survey, received five dollars in compensation, and were entered into a larger cash drawing. Each respondent provided a range of background information including age, gender, education level, and household income, revealing that the average age of survey participants was slightly over 45. Nearly 52% of respondents were female, 77% were non-Hispanic whites,10 and 45% had college or more advanced degrees while 97.3% had high school or higher degrees. Participants’ median annual household income was between $40,000 and $50,000. Of the 64% of participants who were parents, roughly half had children living at home. The survey encompassed over 100 questions focused on issues regarding vaccination practices, perceived benefits and risks of vaccinations, preferences for government vaccination policies, and acquisition of health information from the Internet, with an average response time of 22 minutes. 536

Policy Studies Journal, 42:4
Table 2. Variables and Associated Models

Dependent variable Primary independent variable

Control variable Preference toward mandatory vaccination policy

Model 1

Preference toward religious exemption policy

Model 2

Preference toward philosophical exemption policy

Model 3

Parents, not government, as chief immunization decision makers

Model 4

Cultural worldview

Hierarchism
Egalitarianism
Fatalism

Other personal values/beliefs Individualism
Organic culture
Political ideology

Demographic characteristics Personal importance of religious faith
Vaccine-related knowledge
Family characteristics
Age
Gender
Race
Education
Income

Model 1,
2, 3, and 4

Variables and Measures
In order to more systematically investigate the theoretical conjectures previously discussed in this article, we apply a two-step analysis of the survey data. First, we employ ordinary least squares (OLS) regression with robust standard errors to examine the relationships between individuals’ cultural orientations and vaccination policy preferences. Though informative and intuitive, because such analytical results do not necessarily offer direct answers for how the formations of vaccination policy preference vary by different cultural types, not by cultural biases, we further conduct
Bayesian posterior simulations based on findings from the first step of the analysis.
The variables and measures used in the first step of regression analysis are the focus of the discussion in this section, and are summarized in Table 2.
The dependent variables are the preferences for the existing government vaccine policies: mandatory vaccine policy and religious and philosophical exemption policies. For a given vaccination policy, each respondent’s preference is measured on a
7-point Likert scale ranging from 1 (strongly oppose) to 7 (strongly support). In addition, dependent variables also include the general public’s preference for the intrinsic governance framework bearing on vaccine policies, such as whether parents or the government should make decisions about immunization of children. The respondents were asked to rate their level of agreement (or disagreement) with the relevant statement on a 7-point Likert scale ranging from 1 (strongly disagree) to 7
(strongly agree).11
The primary independent variables include the four cultural dispositions based upon cultural theory: hierarchism, egalitarianism, fatalism, and individualism. Three cultural bias-related survey questions were asked for each bias (provided in random

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

537

order), for a total of 12 questions. Specifically, the three hierarchism items reflect one’s propensity for obedience to social norms, order, authority, and rule, while subtly affirming the importance of the group or society over individual members.
The three egalitarianism items are related to the individual’s proclivity toward social, economic, and political fairness and equality while resonating ideals of less stratified social structure with strong sense of group welfare. The three items related with individualism are associated with a personal conviction that the moral foundation of society lies in compartmentalized individual competition with no societal constraints, not either in allowing society to help the disadvantaged or in emphasizing the welfare of the group as a whole rather than that of its individual members. The final three fatalism items imply a passive personal outlook on life, and a belief that success in life is a matter of random chance and that forces beyond our control largely determine the course of our lives. Respondents rated the degree of agreement with each related statement on a 7-point Likert scale ranging from 1 to 7. The index for each cultural bias is calculated by taking the mean of each set of three related survey items.12 These cultural theory measures were based on prior research (e.g.,
Jenkins-Smith & Smith, 1994; Jones & Song, 2014; Silva & Jenkins-Smith, 2007; Silva,
Jenkins-Smith, & Barke, 2007; Song, 2014; Wildavsky & Dake, 1990), which tested and verified the validity and reliability of the measures. In order to check empirical validity of the measures in use in this study, we conducted factor analysis. The results demonstrated that four unique latent dimensions are extracted from the 12 cultural theory measures, and each of the four conceptual dimensions, one for each cultural disposition, are comprised of three related cultural theory measures (see Appendix
4). Alpha scalability scores for the survey measures used for cultural theory indices were all in the acceptable range with scores from 0.63 to 0.80, which indicates the reliability of the measures in use.
Control and rival explanatory variables include other values and beliefs that have an impact on policy preference as described above.13 To create an index of organic culture, we measure respondents’ degree of agreement with three relevant statements, grading their responses to each on a 7-point scale, with higher scores representing greater agreement. The mean value of all three responses is then used as our organic culture index.14 Political ideology is measured on a 7-point scale
(from 1 to 7) with lower scores exhibiting stronger liberalism and higher scores demonstrating stronger conservatism. Personal importance of religious faith was measured on an 11-point Likert scale ranging from 0 to 10, with higher scores indicating that respondents consider religion more important in their lives. In order to create an index that measures individuals’ knowledge levels regarding vaccinations, we posed six basic vaccine related yes-no questions that are derived from the Centers for Disease Control and Prevention’s website that provides information regarding vaccines and immunizations.15 Individuals are given one point for each correct answer and a final score from 0 to 6, with higher scores indicating greater knowledge (as a higher number of the six questions was answered correctly). Finally, we control demographic variables that could impact personal opinion on vaccination policies. Parental status identifies parents of children under age 18, of children 18 and older, and non-parents. In the categories of gender and

538

Policy Studies Journal, 42:4
Table 3. Descriptive Statistics

Variable
Preference toward mandatory vaccination policy
Preference toward religious exemption policy
Preference toward philosophical exemption policy
Parents, not government, as chief immunization decision makers
Hierarchism index
Egalitarianism index
Individualism index
Fatalism index
Organic culture index
Political ideology
Personal importance of religious faith
Knowledge index
Age
Education
Income

n

Mean

S.D.

Median

Min

Max

1,203
1,201
1,205
1,208

5.5
3.5
3.3
4.4

1.6
2
1.9
1.9

6
3
3
4

1
1
1
1

7
7
7
7

1,196
1,199
1,195
1,198
1,196
1,203
1,209
1,194
1,212
1,194
1,198

4.5
4.2
4.4
3.6
4.5
4.1
6.5
4.4
45.2
3.5
6.1

1.3
1.6
1.3
1.5
1.3
1.6
3.4
1.2
15.8
1.3
4.3

4.3
4.3
4.3
3.7
4.3
4
7
5
45
3
5

1
1
1
1
1
1
0
0
18
1
1

7
7
7
7
7
7
10
6
88
7
21

Table 4. Frequency Table
Variable
Parent with children over 18
Parent with children under 18
Gender
Race

n

Category 1

Category 2

1,201
1,201
1,201
1,207

No (68.3%)
No (67.7%)
Female (51.9%)
Non-White (23.0%)

Yes (31.7%)
Yes (32.3%)
Male (48.1%)
White, Not Hispanic (77.0%)

race, respondents are coded 1 for male and for non-Hispanic white. Levels of education and household income are measured on 7-point and 21-point rising scales, respectively. As presented in Table 3, central tendency measures (e.g., arithmetical mean or median), dispersion measure (e.g., standard deviation) and the range of values of the interval scale variables used for the analysis reveal no extreme distributions. In general, respondents prefer mandatory vaccination policy to various exemption policies. However, when they were asked who should decide about immunizing children, many respondents felt that parents rather than the government should have the final say. In terms of cultural biases, respondents are more inclined toward hierarchism, egalitarianism, and individualism than they are toward fatalism. Respondents exhibit a modest level of affinity with “organic culture” and a normal distribution over the range of political ideologies. Overall, respondents indicate that religious faith is relatively important in their lives and possess moderately high levels of vaccine-related knowledge as measured on our index. The distribution of the income variable displays the typical characteristic of being skewed to the right.
The frequencies of the categorical variables are shown in Table 4. While 381 participants (31.7% of total valid responses) are parents who do not have any child under 18, 388 participants (32.3% of total valid responses) are parents who have at least one child under 18.

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

539

Table 5. OLS Regression Results
Independent
variable

Hierarchism
Egalitarianism
Individualism
Fatalism
Organic culture
Political ideology
Personal importance of religious faith
Knowledge
Parent with children over 18 (1 = Yes)
Parent with children under 18 (1 = Yes)
Age
Gender (1 = Male)
Race (1 = White, not
Hispanic)
Education
Income
Intercept
F
Adjusted R2
Degrees of freedom

Dependent variable
Preference toward Preference toward Preference toward
Parents, not mandatory religious exemption philosophical government, as vaccination policy policy exemption policy chief immunization decision makers
Model 1

Model 2

Model 3

Model 4

0.272***
(0.044)
0.137***
(0.039)
0.034
(0.041)
−0.101***
(0.038)
−0.166***
(0.040)
−0.070**
(0.034)
0.005
(0.014)
0.173***
(0.040)
0.088
(0.133)
0.055
(0.115)
0.009**
(0.004)
−0.150*
(0.090)
−0.071
(0.117)
0.058*
(0.035)
0.018
(0.011)
3.584***
(0.430)
10.01***
0.112
1,054

−0.238***
(0.055)
−0.110**
(0.048)
0.162***
(0.051)
0.195***
(0.048)
0.234***
(0.050)
0.048
(0.043)
0.135***
(0.018)
−0.190***
(0.048)
−0.058
(0.171)
−0.025
(0.140)
−0.026***
(0.005)
−0.049
(0.113)
0.402***
(0.141)
0.018
(0.048)
−0.018
(0.014)
3.199***
(0.533)
15.48***
0.169
1,053

−0.258***
(0.060)
−0.062
(0.047)
0.205***
(0.055)
0.194***
(0.049)
0.236***
(0.048)
0.039
(0.041)
0.090***
(0.018)
−0.219***
(0.047)
−0.005
(0.160)
0.128
(0.139)
−0.023***
(0.005)
−0.005
(0.111)
0.337**
(0.139)
−0.050
(0.048)
−0.011
(0.014)
3.297***
(0.535)
15.94***
0.173
1,056

−0.209***
(0.057)
−0.024
(0.047)
0.253***
(0.050)
0.175***
(0.048)
0.185***
(0.048)
0.100**
(0.041)
0.091***
(0.018)
−0.152***
(0.046)
0.076
(0.168)
0.256*
(0.134)
−0.018***
(0.005)
0.040
(0.109)
0.461***
(0.134)
−0.132***
(0.046)
−0.003
(0.015)
3.376***
(0.545)
14.84***
0.162
1,058

*p < 0.10. **p < 0.05. ***p < 0.01. Robust standard errors in parentheses.

Empirical Findings
Table 5 displays the OLS regression results with robust standard errors.16 In order to tackle the problem of the heteroskedastic error distribution17 in statistical inference based on the results acquired from the fitted regression models, we applied the procedures for heteroskedasticity consistent covariance estimation suggested by
White (1980) and used the results to make adjustments to the standard errors of regression coefficients derived from the OLS estimation in order to improve our statistical inference.

540

Policy Studies Journal, 42:4

The results show that cultural biases systematically influence vaccine policy preferences even when controlling for the effects of other variables on the dependent variable.18 First, as we conjectured earlier, those with strong hierarchical culture tend to support mandatory vaccine policy (+0.272, p < 0.001 in Model 1); oppose various exemption policies (−0.238, p < 0.001 in Model 2; −0.258, p < 0.001 in Model 3); and feel that the government, not parents, should determine childhood vaccinations
(−0.209, p < 0.001 in Model 4). Second, egalitarian bias has a comparatively weaker and less consistent impact on vaccine related policy preferences. Those who have a strong egalitarian bias tend to support mandatory vaccine policy (+0.137, p < 0.001 in
Model 1) and oppose religious exemption policy (−0.110, p < 0.05 in Model 2).
However, egalitarianism has no statistically significant impact on any other policy preferences. Third, the individualist cultural bias also has an inconsistent impact on vaccination policy preferences. Strong individualists are more likely to support various vaccine exemption policies (+0.162, p < 0.001 in Model 2; +0.205, p < 0.001 in
Model 3) and agree that parents, not the government, should decide on children’s vaccinations (+0.253, p < 0.001 in Model 4). However, individualism has no statistically significant impact on other policy preference. Finally, the fatalist bias exhibits a consistent influence on various vaccine-related policy preferences. As expected, strong fatalists tend to oppose mandatory vaccination policy (−0.101, p < 0.001 in
Model 1); support various exemption policies (+0.195, p < 0.001 in Model 2; +0.194, p < 0.001 in Model 3); and believe that parents, not government, should decide if their children are immunized (+0.175, p < 0.001 in Model 4).
The effects of several control variables on the policy preferences are noteworthy.
In general, those who are characterized by a strong organic culture, who are politically conservative,19 who hold strong religious beliefs,20 or who are non-Hispanic white, older, or who are less knowledgeable about immunizations are more reluctant to support mandatory vaccination policy (Model 1); tend to favor vaccine exemption policies (Model 2 and 3); and believe that parents, not government, should chiefly decide about immunization of children (Model 4).21
The next analysis involves a prediction of the distributions of vaccine policy preferences according to prototypical cultural type utilizing the fitted regression model and technique of statistical simulations suggested by King, Tomz, and
Wittenberg (2000). Based upon this approach,22 we took the following analytic steps to acquire predicted distribution of various vaccine policy preferences for each of the four prototypical cultural types in this paper. First, we fit OLS regression models to the sample used for previous regression analysis for our hypothesis test. We used the same dependent variables employed in the previous models (i.e., preferences for various vaccination policies), but used only the four cultural measures (i.e., hierarchism, egalitarianism, individualism, and fatalism) as explanatory variables.23
Second, based upon the estimated parameters and variance-covariance matrix24 of these parameters acquired from the first step of the analysis, we ran iterative simulations (1,000 times) as suggested by Gelman and Hill (2007). From this iterative simulation, we obtained 1,000 different vectors of estimated regression coefficients
(including coefficients for the intercept term) for each model. Third, we defined prototypes for each of the four cultural types by assigning one standard deviation

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

Strongly oppose

Strongly oppose
Hierarch

541

Strongly support

Strongly oppose

Strongly support

Strongly disagree

Egalitarian

Individualist

Strongly support

Strongly agree
Fatalist

Figure 2. Vaccination Policy Preferences by Cultural Type.

above the mean of a particular cultural orientation index and one standard deviation below the mean of the remaining three cultural orientation indices.25 This step reflects the suggestion that each cultural type derives its identity from both its own particular biases and also from the negation of other cultural types’ biases
(Ripberger, Jenkins-Smith, & Herron, 2011). Finally, in each of the 1,000 different simulated regression equations, we entered in the cultural measure values for each prototypical cultural type determined in the previous step in order to obtain a distribution of predicted preferences for each of the four cultural types for the respective childhood vaccination policies.
Figure 2 displays the results of this analysis highlighting the four most contested childhood vaccination policy issues. Panel (a) shows distribution of predicted policy preferences for mandatory vaccinations, while panels (b), (c), and (d) show predicted policy preferences for religious exemption, philosophical exemption, and opinion on parental decision-making power regarding vaccinations, respectively. As shown in the legend, the solid black histogram represents hierarchs, white outlined in dark red represents egalitarians, white outlined in light orange represents individualists, and solid gray represents fatalists. The vertical axis of the histograms shows the density function of the distribution, while the horizontal axis represents either the degree of support for, or level of agreement with, the given policy issue. Overall, the panels in
Figure 2 reveal that hierarchs and fatalists are the two cultural types exhibiting the sharpest contrast in policy preference (histograms show no overlap in the distribution of predicted policy preferences for these two prototypes). As expected, hierarchs are typically in support of mandatory vaccination, oppose religious and philosophical exemption, and feel that government should preside over vaccination-related decisions. Fatalists strike a bold contrast in their opposition to mandatory vaccination

542

Policy Studies Journal, 42:4

policy and support for religious and philosophical exemptions. In addition, they are much more likely to support the role of parents in deciding on vaccinations. In general, the divergence in policy positions between the two groups is driven by hierarchs, who are, for instance, clearly and strongly in support of mandatory vaccine policy, and oppose religious and philosophical exemptions, whereas fatalists neither strongly oppose (or support) mandatory vaccinations nor are in clear support of (or opposition to) exemptions based on religious and philosophical beliefs. Instead, fatalists’ opinions on both issues tend to fall closer to the scale midpoint, reflecting both less support for mandatory vaccination positions and a less decisive position overall. Falling between hierarchs and fatalists, egalitarian support for vaccinations is essentially stronger than individualists’, while the two exhibit notable overlap in their distribution of predicted preferences toward mandatory vaccination.
Conclusion and Discussion
This research seeks to explain varying public opinions in the vaccine policy subsystem of the United States, where conflicting principles coexist within the same policy. Consistent with the argument of grid-group cultural theory (Douglas &
Wildavsky, 1983; Wildavsky, 1987), we find that cultural biases have significant impacts on the formation of preferences toward vaccination policies. Hierarchs and egalitarians are more likely to be pro-vaccination, while individualists and (especially) fatalists tend to oppose this view. Hierarchs advocate mandatory vaccination; disapprove of religious and philosophical exemptions; and believe that government, not parents, should preside over childhood immunizations. By contrast, fatalists are inclined to negate mandatory vaccination policy and uphold religious and philosophical exemptions and the role of parents in determining vaccination of children.
Egalitarians’ pro-vaccination inclination is relatively weaker and less consistent than hierarchs’, while individualists’ anti-vaccination leanings are overall less robust than those of fatalists.
The most important element of these findings is that the vaccine policy debate is not solely based upon efficacy in reduction in disease or the resulting societal benefits and costs. Rather, it gains considerable momentum from the clash of worldviews. An intrinsic value dimension, notably in the form of cultural orientation, is reflected in the way this debate has become a surrogate for an overarching contest among competing sets of societal norms. Many government health authorities and experts believe that people oppose vaccinations because of their own inability to access quality vaccinerelated knowledge or due to dissemination of false information. In response, health advocates have tried to enlighten the general public and thereby increase compliance with mandatory vaccination policy, thereby improving public health. Of course, proliferation of quality knowledge and sound information provided by the scientific community is essential. However, the results of this study show that preferences for vaccine-related policies are significantly influenced by individual values and beliefs regarding desirable social relationships, notably in the form of cultural predispositions. Furthermore, from a cultural cognition perspective, individuals’ cultural biases work as a set of heuristics in the processing of the information and in the course of

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

543

reasoning (Jones & Song, 2014; Kahan & Braman, 2006; Kahan, Jenkins-Smith, &
Braman, 2011; Silva & Jenkins-Smith, 2007; Silva et al., 2007). That is, when individuals with a particular cultural bias encounter new information, they will reinterpret that information through the filters of their own cultural bias and use the results in their reasoning and policy evaluation. Even the seemingly unrelated matter of crediting expertise, which is integral to deciding whether a policy is beneficial or risky, is subject to cultural cognition (Kahan et al., 2011). With this in mind, three related concepts from risk communication perspectives may prove useful to public health authorities in the effective dissemination of vaccine-related knowledge: identity affirmation, pluralistic advocacy, and narrative framing (Kahan et al., 2011, pp. 23–24). First, identity affirmation has shown that individuals are more likely to embrace information that appears to reinforce their own worldviews and reject conclusions that undermine their values (Cohen, Aronson, & Steele, 2000). Individualists, for example, are more likely to respond to vaccine messages that make the point that the decision not to vaccinate children imposes involuntary risks on others than they are to appeals to authority and expertise. Additionally, the concept of pluralistic advocacy highlights how individuals reject messages from experts they believe do not share their cultural values and reflexively trust information from experts whose (presumed) values align with their own (Earle & Cvetkovich, 1995). Moreover, if experts representing an array of different values appear to fall on both sides of the fence on a given issue, then the individual may be less prone to engage in identity-protective cognition with respect to the information these experts provide. Third, our results suggest that information campaigns regarding vaccine programs would benefit from narrative framing26
(Anderson, 1997; Jones & McBeth, 2010; Jones & Song, 2014; Shanahan, Jones, &
McBeth, 2011; Shanahan, Jones, McBeth, & Lane, 2013; Shanahan, McBeth, &
Hathaway, 2010), in which custom-fit templates or culturally nuanced narratives are employed that bolster feelings of validation for a particular cultural group. Such customized messages are designed to appeal to their target group by assigning positive value to their worldview, thereby garnering more attention for crucial public messages carrying vaccine-related information.27
Meanwhile, this research holds some limitations that are to be considered in interpreting the results and extending related research in the future. First, as the primary concern of this research is to evaluate public attitudes toward current vaccination policies, the analysis is not inclusive of any other possible policy options that various cultural types might prefer. For instance, one could argue that though hierarchs and egalitarians seem to share similar preferences toward current vaccination policies in the analysis discussed, if a more comprehensive list of policy alternatives were considered in the analysis, their preferences might suddenly begin to diverge. For that end, more comprehensive vaccination policy options should be included in the analysis in future research.28 Second, though we provide a brief theoretical discussion on how cultural theory relates to alternative theories
(or related factors) in explaining vaccination policy preferences, this discussion and related empirical analysis are certainly not thorough enough to describe the complex interactions between cultural orientations and other factors in explaining vaccination policy preferences. One could argue, for instance, that certain demo-

544

Policy Studies Journal, 42:4

graphic characteristics of individuals may nurture their unique cultural worldviews, which, in turn, could form the political ideology and “organic” culture that translate into vaccination policy preferences. However, such theoretical and empirical comprehension is not the focus of this research, and we reserve such aspects of more complex analysis for future research while focusing on the cultural theory explanation of vaccination policy preferences in this paper. Third, there are some methodological concerns to be addressed with regard to survey measures used for the analysis. For instance, one could argue that the hierarchism survey questions, presented in Appendix 3, are disproportionally focused on measuring the “grid” dimension of cultural orientation and not the “group” dimension, which can be thought of as providing legitimacy for the structural stratification and sources of authority of the “grid.” A better way of incorporating both dimensions of hierarchism should be considered in crafting hierarchism measures in future research. Fourth, though the focus of this research is to explain public attitude’s effect on formation of childhood vaccination policies, our analysis does not explain how such policy preferences translate into actual vaccination behavior, which holds more direct practical ramifications in achieving health policy goals from the perspective of government health authorities. Such linkage between attitude and behavior should be further studied in the future. On a related note, one could also question whether the findings of this research would hold external validity and be applicable to other health policy domains, such as mandatory health insurance policy, which can be discussed in future research.
Geoboo Song is an assistant professor in the Department of Political Science at the J. William Fulbright College of Arts and Sciences of the University of
Arkansas.
Carol L. Silva is an associate professor in the Department of Political Science and the Director of the Center for Risk, Crisis, and Management at the University of
Oklahoma.
Hank C. Jenkins-Smith is a George Lynn Cross Research Professor in the Department of Political Science at the University of Oklahoma. He is a co-Director of the
Center for Energy, Security and Society, which is a jointly run center of the University of Oklahoma and Sandia National Laboratories.
Notes
The authors would like to thank Jon Palfreman, David L. Weimer, Brendon Swedlow, Christopher M.
Weible, Joseph T. Ripberger, Matthew C. Nowlin, Kuhika Gupta, Sarah R. Trousset, A. Kate Miller, and anonymous referees for their helpful comments on this article.
1. For more detail, see the following CDC webpage: http://www2a.cdc.gov/nip/schoolsurv/ schImmRqmt.asp. 2. These medical exemptions are determined by a physician.
3. Those who opt out from vaccinating their children for religious reasons often believe that according to biblical or other religious teachings, they must avoid vaccinations’ “tainted” or “toxic” contents.
Some religious groups also view vaccinations (and other aspects of science and technology) as human interference with the will of God.

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

545

4. These states are Arizona, Arkansas, California, Colorado, Idaho, Louisiana, Maine, Michigan, Minnesota, Missouri, North Dakota, Ohio, Oklahoma, Pennsylvania, Texas, Utah, Vermont, Washington, and
Wisconsin.
5. Those who avoid vaccinating their children for philosophical reasons often believe that citizens in a free society, regardless of their views on the efficacy of vaccinations, ought to have the right to make such decisions based upon their own moral and ethical imperative, which is seen as superior to the government’s legal authority and mandates.
6. A good example is found in the recent controversy regarding a vaccine-autism link. Though most experts within the scientific community disagree (e.g., Dales, Hammer, & Smith, 2001; Kaye, del Mar
Melero-Montes, & Jick, 2001; Taylor et al., 1999), a few scientists have argued that thimerosal, a preservative found in vaccines, can cause damage to neurological development and result in autism in infants and children (e.g., Stratton, Gable, Shetty, & McCormick, 2001; Wakefield & Montgomery,
2000; Wakefield et al., 1998). At the same time, a recent federal vaccine court decision gave momentum to the vaccine-autism link controversy when it was ruled that vaccinations did not “cause” Hannah
Polling’s autism but “resulted” in it by aggravating an unknown mitochondrial disorder (Attkisson,
2010). The government was ordered to compensate her family with $1.5 million for the first year and
$500,000 every following year for Polling’s lifetime. One of the primary and most controversial reasons vaccine opponents try to avoid vaccination is their belief that vaccines can cause neurologic disorders such as autism or Asperger’s syndrome.
7. All of these cited works claim that females generally tend to perceive higher levels of risk than males.
However, what Palmer (2003) and Kahan et al. (2007) argue differs from Finucane et al. (2000) and
Flynn et al. (1994) in that the former scholars essentially emphasize differences in shared values between different gender groups, not gender per se, in understanding differences in perceptions of risk.
8. This racial group may have more sociopolitical and economic resources and a greater sense of control, which may foster more confidence in various aspects of science and technology, including vaccines, and more support for mandatory vaccination policy than most minority groups. Alternatively, one could argue that this racial group tends to be more individualistic compared to its counterparts and could be less likely to support vaccine mandates, not because of the belief that vaccines are dangerous, but because they do not like the idea of compulsory action required by the government.
9. For more detail on this Internet sampling method, see Best, Krueger, Hubbard, and Smith (2001) and
Berrens, Bohara, Jenkins-Smith, Silva, and Weimer (2003). For a demographic comparison of the survey sample and U.S. national population, see Appendix 1.
10. The racial composition of remaining survey participants is as follows: American Indian (1%), Asian
(6%), African American (8%), Hispanic (6%), and Other (2%).
11. The question wording used for measuring these policy preferences is shown in Appendix 2.
12. The question wording of these variables is presented in Appendix 3.
13. The question wordings for the control items are shown in Appendix 5.
14. The alpha score for the organic culture scale is 0.64.
15. For more detail, see http://www.cdc.gov/vaccines/vac-gen/common-faqs.htm.
16. One could argue that ordinal logistic regression is a more appropriate approach in this type of analysis, because dependent variables are measured on a non-interval scale, which violates OLS regression assumptions. We conducted the ordinal logistic regression analysis using the exact same model specifications we used for the OLS regression. As presented in Appendix 6, the results from the ordinal logistic regression analysis are consistent with those from OLS regression (presented in
Table 5). For the purpose of simplicity of interpretation, we will focus on the results from OLS regression in the following discussion.
17. Besides this heteroskedasticity issue, we did not find any serious violations of the OLS regression assumptions from any estimated models presented in Table 5. For instance, the results of Ramsey’s
RESET test and Shapiro-Wilk test of normality revealed that the functional relationships specified in the fitted models are linear and estimated errors are normally distributed, respectively. The VIF
(i.e., Variance Inflation Factor) index for each variable specified on the right side of the regression equations ranged from 1.05 to 1.88, which indicates that there is no alarming level of multicollinearity concern. 546

Policy Studies Journal, 42:4

18. With regard to the interpretation of regression results, we understand that, technically, regression does not offer any causal relationship between variables, but rather provides co-variation between them.
What we are emphasizing in reporting the regression results here, however, is the conceptual relationship along with the consistency of our theoretical discussion on the cultural theory explanation of vaccine policy preferences, which is based upon causality claim by nature.
19. The effect of support for vaccine exemption policies based upon religion or philosophical belief is not statistically significant.
20. Their propensity to support mandatory vaccination policy is not statistically significant.
21. Though the major concern of this research is to examine how variations of individuals’ vaccine policy preferences are explained by their cultural predispositions, it is worthwhile to make a more comprehensive note of how other variables derived from alternative explanations play into such a relationship. For instance, one could argue that individuals’ demographic characteristics (that reflect the environment wherein the nurturing and shaping of their values and beliefs, notably cultural predispositions, occur) in turn provide the basis of their attitude formation towards vaccine policy preferences. We, accordingly, conducted additional multivariate analysis and found that hierarchism is stronger among the less educated and non-Whites; egalitarianism is stronger among those who are parents, younger, non-White, less educated, and less wealthy; individualism is stronger among the wealthy and Whites; and fatalism is stronger among younger, non-Whites, males, and those who are less educated.
22. King et al. (2000) argue that parameter estimates from a regression model (e.g., regression coefficients) fit to the sample can never be absolutely certain due to the finite nature of the sample. To address this uncertainty more directly in our analysis, we can draw many probable sets of parameters from their posterior or sampling distribution. These simulated results can then either be displayed graphically for a more effective description of the original regression results or used for further analysis.
23. We constructed a simplified model mainly because the focus of this analysis does not lie in the hypothesis test (which was the focus of previous regression analysis) but in the predictions based on the estimated effects of primary independent variables (i.e., cultural measures) on dependent variables (i.e., vaccine policy preferences) which were statistically verified in the previous regression analysis that also contained control variables derived from other major competing theoretical claims. In addition, hypothesis tests involved in previous regressions showed that many of the estimated regression coefficients for the control variables are not statistically significantly different from zero, which was another reason we discarded these variables in this regression analysis for the prediction. The results of this regression analysis are shown in
Appendix 7.
24. We are using variance-covariance matrix of estimated parameters derived from OLS regression rather than the one adjusted to heteroskedastic distribution of errors, which we used for statistical inference in the original multivariate model, because we did not find any heteroskedasticity problem in this simplified regression model.
25. For instance, Table 3 in the previous section of this paper reveals that the mean of hierarchism is 4.5, with a standard deviation of 1.3. Therefore, the prototypical hierarch is one who scores 5.8 on hierarchism measure (the combined value of hierarchism mean and standard deviation) and 2.6
(egalitarianism mean [=4.2] minus its standard deviation [=1.6]), 3.1 (individualism mean [=4.4] minus its standard deviation [=1.3]) and 2.1 (fatalism mean [=3.6] minus its standard deviation [=1.5]) on measures of egalitarianism, individualism, and fatalism, respectively.
26. For the similar argument made by more interpretive approaches, which helps understand how related public messages are constructed within a given “context” and “meanings” associated with the policy positions the public holds, see, for instance, Fischer and Gottweis (2012).
27. One anonymous reviewer suggested that it is equally important to acknowledge that this approach runs some risks that the “target group” would become more radical (as a result of positive feedback) by only highlighting certain elements in the policy debate on vaccination. He or she asserted, accordingly, that such risks could be mitigated by making clear that the target group’s view is only one part of the policy debate—though an essential part—and ought to be combined with other various views
(i.e., negative feedback or “surprise”), which results from socio-cultural viability argument provided in the theoretical discussion section of this paper.

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

547

28. With regard to mandatory vaccination policy, for instance, as strong egalitarians would not trust the government, they would dislike such government mandates. However, local or community-based mandates would be much more acceptable, if set up and implemented by support groups with which they affiliate and trust. This would differentiate their policy preference from strong hierarchs’. References
Anderson, John L. 1997. “Governmental Suasion: Refocusing the Lowi Policy Typology.” Policy Studies
Journal 25 (2): 266–82.
Attkisson, Sharyl. 2010. “Family to Receive $1.5 M + in First-Ever Vaccine-Autism Court Award.” CBS
News Investigates [Online]. http://www.cbsnews.com/news/family-to-receive-15m-plus-in-firstever-vaccine-autism-court-award. Accessed September 5, 2014.
Baker, Dian L., Michelle T. Dang, May Ying Ly, and Rafael Diaz. 2010. “Perception of Barriers to Immunization among Parents of Hmong Origin in California.” American Journal of Public Health 100 (5):
839–45.
Barringer, Paul J., David M. Studdert, Allen B. Kachalia, and Michelle M. Mello. 2008. “Administrative
Compensation of Medical Injuries: A Hardy Perennial Blooms Again.” Journal of Health Politics, Policy and Law 33 (4): 725–60.
Berrens, Robert P., Alok K. Bohara, Hank C. Jenkins-Smith, Carol L. Silva, and David L. Weimer. 2003. “The
Advent of Internet Surveys for Political Research: A Comparison of Telephone and Internet
Samples.” Political Analysis 11 (1): 1–22.
Best, Samuel J., Brian Krueger, Clark Hubbard, and Andrew Smith. 2001. “An Assessment of the
Generalizability of Internet Surveys.” Social Science Computer Review 19 (2): 131–45.
Calandrillo, Steve P. 2004. “Vanishing Vaccinations: Why Are So Many Americans Opting Out of Vaccinating their Children?” University of Michigan Journal of Law Reform 37 (2): 353–440.
Centers for Disease Control and Prevention. 2012. “Measles-United States, 2011.” Morbidity and Mortality
Weekly Report (MMWR) 61 (15): 253–57.
Cohen, Geoffrey L., Joshua Aronson, and Claude M. Steele. 2000. “When Beliefs Yield to Evidence:
Reducing Biased Evaluation by Affirming the Self.” Personality and Social Psychology Bulletin 26 (9):
1151–64.
Colcord, Frank C. 1977. “Urban Transportation and Political Ideology: Sweden and the United States.”
Policy Studies Journal 6 (1): 9–19.
Dales, Loring, Sandra Jo Hammer, and Natalie J. Smith. 2001. “Time Trends in Autism and in MMR
Immunization Coverage in California.” JAMA: The Journal of the American Medical Association 285 (9):
1183–95.
Doan, Alesha E., and Kellee Kirkpatrick. 2013. “Giving Girls a Shot: An Examination of Mandatory
Vaccination Legislation.” Policy Studies Journal 41 (2): 295–318.
Douglas, Mary. 1970. Natural Symbols. London: Barrie & Rockliff.
———. 1996. Thought Styles: Critical Essays on Good Taste. London: Sage.
Douglas, Mary, and Aaron Wildavsky. 1983. Risk and Culture. Berkeley: University of California Press.
Earle, Timothy C., and George Cvetkovich. 1995. Social Trust: Toward A Cosmopolitan Society. Westport, CT:
Praeger Publishers.
Elliott, E. Donald, Sanjay A. Narayan, and Moneen S. Nasmith. 2008. “Administrative ‘Health Courts’ for
Medical Injury Claims: The Federal Constitutional Issues.” Journal of Health Politics, Policy and Law 33
(4): 761–98.
Ernst, Edzard. 2001. “Rise in Popularity of Complementary and Alternative Medicine: Reasons and
Consequences for Vaccination.” Vaccine 20: S90–93.
Ferejohn, John. 1991. “Rationality and Interpretation: Parliamentary Elections in Early Stuart England.” In
The Economic Approach to Politics: A Critical Reassessment of the Theory of Rational Action, ed. Kristen
Renwick Monroe. New York: Harper-Collins, 279–305.

548

Policy Studies Journal, 42:4

Finucane, Melissa L., Paul Slovic, C. K. Mertz, James Flynn, and Theresa A. Satterfield. 2000. “Gender,
Race, and Perceived Risk: the ‘White Male’ Effect.” Health, Risk & Society 2 (2): 159–72.
Fiorino, Daniel J. 1989. “Technical and Democratic Values in Risk Analysis.” Risk Analysis 9 (3):
293–99.
Fischer, Frank, and Herbert Gottweis. 2012. The Argumentative Turn Revisited: Public Policy as Communicative Practice. Durham, NC: Duke University Press.
Flynn, James, Paul Slovic, and C. K. Mertz. 1994. “Gender, Race, and Perception of Environmental Health
Risks.” Risk Analysis 14 (6): 1101–8.
Frontline. 2010. “The Vaccine War: Interviews-Jenny McCarthy.” Public Broadcasting Service (April 27)
[Online].
http://www.pbs.org/wgbh/pages/frontline/vaccines/interviews/mccarthy.html/1.
Accessed September 5, 2014.
Gellin, Bruce G., Edward W. Maibach, and Edgar K. Marcuse. 2000. “Do Parents Understand Immunizations? A National Telephone Survey.” Pediatrics 106 (5): 1097–102.
Gelman, Andrew, and Jennifer Hill. 2007. Data Analysis Using Regression and Multilevel/Hierarchical Models.
Cambridge: Cambridge University Press.
Gerber, Brian J., and Grant W. Neeley. 2005. “Perceived Risk and Citizen Preferences for Governmental
Management of Routine Hazards.” Policy Studies Journal 33 (3): 395–418.
Herron, Kerry G., and Hank C. Jenkins-Smith. 2006. Critical Masses and Critical Choices: Evolving Public
Opinion on Nuclear Weapons, Terrorism, and Security, Pittsburgh, PA: University of Pittsburgh Press.
Jenkins-Smith, Hank C., and Kerry G. Herron. 2009. “Rock and A Hard Place: Public Willingness to Trade
Civil Rights and Liberties for Greater Security.” Politics & Policy 37 (5): 1095–129.
Jenkins-Smith, Hank C., and Walter K. Smith. 1994. “Ideology, Culture, and Risk Perception,” In Politics,
Policy and Culture, eds. D. J. Coyle, and R. Ellis. Boulder, CO: Westview Press, 17–32.
Jones, Michael D., and Mark K. McBeth. 2010. “A Narrative Policy Framework: Clear Enough to Be
Wrong?” Policy Studies Journal 38 (2): 329–53.
Jones, Michael D., and Geoboo Song. 2014. “Making Sense of Climate Change: How Story Frames Shape
Cognition.” Political Psychology 35 (4): 447–76.
Kahan, Dan M., and Donald Braman. 2006. “Cultural Cognition and Public Policy.” Yale Law & Policy
Review 24 (1): 149–72.
Kahan, Dan M., Donald Braman, Geoffrey L. Cohen, John Gastil, and Paul Slovic. 2010. “Who Fears the
HPV Vaccine, Who Doesn’t, and Why? An Experimental Study of the Mechanisms of Cultural
Cognition.” Law and Human Behavior 34 (6): 501–16.
Kahan, Dan M., Donald Braman, John Gastil, Paul Slovic, and C. K. Mertz. 2007. “Culture and IdentityProtective Cognition: Explaining the White-Male Effect in Risk Perception.” Journal of Empirical Legal
Studies 4 (3): 465–505.
Kahan, Dan M., Hank C. Jenkins-Smith, and Donald Braman. 2011. “Cultural Cognition of Scientific
Consensus.” Journal of Risk Research 14 (2): 147–74.
Kaye, James A., Maria del Mar Melero-Montes, and Hershel Jick. 2001. “Mumps, Measles, and Rubella
Vaccine and the Incidence of Autism Recorded by General Practitioners: A Time Trend Analysis.”
British Medical Journal 322 (7284): 460–63.
King, Gary, Michael Tomz, and Jason Wittenberg. 2000. “Making The Most of Statistical Analyses: Improving Interpretation and Presentation.” American Journal of Political Science 44 (2): 347–61.
Lehrke, Philipp, Matthias Nuebling, Friedrich Hofmann, and Ulrich Stoessel. 2001. “Attitudes of
Homoeopathic Physicians towards Vaccination.” Vaccine 19 (32): 4859–64.
Lockhart, Charles. 2001. Protecting the Elderly: How Culture Shapes Social Policy. University Park: Penn State
University Press.
Lodge, Marin, Kai Wegrich, and Gail McElroy. 2010. “Dodgy Kebabs Everywhere? Variety of Worldviews and Regulatory Change.” Public Administration 88 (1): 247–66.
Luman, Elizabeth T., Lawrence E. Barker, Mary Mason McCauley, and Carolyn Drews-Botsch. 2005.
“Timeliness of Childhood Immunizations: A State-Specific Analysis.” American Journal of Public
Health 95 (8): 1367–74.

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

549

Mariner, Wendy K., George J. Annas, and Leonard Glantz. 2005. “Jacobson v Massachusetts: It’s Not Your
Great-Great-Grandfather’s Public Health Law.” American Journal of Public Health 95 (4): 581–90.
McAvoy, Gregory E. 1998. “Partisan Probing and Democratic Decisionmaking Rethinking the Nimby
Syndrome.” Policy Studies Journal 26 (2): 274–92.
Palmer, Christina G. S. 2003. “Risk Perception: Another Look at the White Male Effect.” Health, Risk &
Society 5 (1): 71–83.
Pereira, Joseph A., and Gregg G. Van Ryzin. 1998. “Understanding Public Support for Time Limits and
Other Welfare Reforms.” Policy Studies Journal 26 (3): 398–418.
Plutzer, Eric, Ardith Maney, and Robert E. O’Connor. 1998. “Ideology and Elites’ Perceptions of the Safety of New Technologies.” American Journal of Political Science 42 (1): 190–209.
Rein, David B., Amanda A. Honeycutt, Lucia Rojas-Smith, and James C. Hersey. 2006. “Impact of the
CDC’s Section 317 Immunization Grants Program Funding on Childhood Vaccination Coverage.”
American Journal of Public Health 96 (9): 1548–53.
Ripberger, Joseph T., Hank C. Jenkins-Smith, and Kerry Herron. 2011. “Cultural Theory and National
Security: How Cultural Orientations Shape Public Beliefs About Nuclear Weapons and Homeland
Security.” PS, Political Science & Politics 44 (4): 715–9.
Ripberger, Joseph T., Thomas M. Rabovsky, and Kerry G. Herron. 2011. “Public Opinion and Nuclear
Zero: A Domestic Constraint on Ditching the Bomb.” Politics & Policy 39 (6): 891–923.
Ripberger, Joseph T., Geoboo Song, Matthew C. Nowlin, Michael D. Jones, and Hank C. Jenkins-Smith.
2012. “Reconsidering the Relationship Between Cultural Theory, Political Ideology, and Political
Knowledge.” Social Science Quarterly 93 (3): 713–31.
Rothman, Stanley, and S. Robert Lichter. 1987. “Elite Ideology and Risk Perception in Nuclear Energy
Policy.” The American Political Science Review 81 (2): 383–404.
Sabatier, Paul A., and Hank C. Jenkins-Smith. 1993. Policy Change and Learning: An Advocacy Coalition
Approach. Boulder, CO: Westview Press.
Salmon, Daniel A., Jason W. Sapsin, Stephen Teret, Richard F. Jacobs, Joseph W. Thompson, Kevin Ryan, and Neal A. Halsey. 2005. “Public Health and the Politics of School Immunization Requirements.”
American Journal of Public Health 95 (5): 778–83.
Shanahan, Elizabeth A., Michael D. Jones, and Mark K. McBeth. 2011. “Policy Narratives and Policy
Processes.” Policy Studies Journal 39 (3): 535–61.
Shanahan, Elizabeth A., Michael D. Jones, Mark K. McBeth, and R. R. Lane. 2013. “An Angel on the Wind: How Heroic Policy Narratives Shape Policy Realities.” Policy Studies Journal 41 (3):
453–83.
Shanahan, Elizabeth A., Mark K. McBeth, and Paul Hathaway. 2010. “Narrative Policy Framework: The
Influence of Media Policy Narratives on Public Opinion.” Politics & Policy 39 (3): 373–400.
Silva, Carol L., and Hank C. Jenkins-Smith. 2007. “The Precautionary Principle in Context: US and EU Scientists’ Prescriptions for Policy in the Face of Uncertainty.” Social Science Quarterly 88 (3):
640–64.
Silva, Carol L., Hank C. Jenkins-Smith, and Richard P. Barke. 2007. “Reconciling Scientists’ Beliefs about
Radiation Risks and Social Norms: Explaining Preferred Radiation Protection Standards.” Risk
Analysis 27 (3): 755–73.
Smith, Philip J., and John, Stevenson. 2008. “Racial/Ethnic Disparities in Vaccination Coverage by 19
Months of Age: An Evaluation of the Impact of Missing Data Resulting From Record Scattering.”
Statistics in Medicine 27 (20): 4107–18.
Song, Geoboo. 2014. “Understanding Public Perceptions of Benefits and Risks of Childhood Vaccinations in the United States.” Risk Analysis 34 (3): 541–55.
Stratton, Kathleen R., Alisha Gable, Padma Shetty, and Marie McCormick. 2001. Immunization Safety
Review: Measles-Mumps-Rubella Vaccine and Autism. Washington, DC: National Academies Press.
Swedlow, Brendon. 1994. “Cultural Influences on Policies Concerning Mental Illness,” In Politics, Policy, and Culture, eds. Dennis J. Coyle, and Richard J. Ellis. Boulder, CO: Westview Press, 71–89.
———. 2011. “Editor’s Introduction: Cultural Theory’s Contributions to Political Science.” PS, Political
Science & Politics 44 (4): 703–10.

550

Policy Studies Journal, 42:4

Taylor, Brent, Elizabeth Miller, C. Paddy Farrington, Maria-Christina Petropoulos, Isabelle Favot-Mayaud,
Jun Li, and Pauline A. Waight. 1999. “Autism and Measles, Mumps, and Rubella Vaccine: No
Epidemiological Evidence for a Causal Association.” Lancet 353 (9169): 2026–29.
Thompson, Michael, Richard J. Ellis, and Aaron B. Wildavsky. 1990. Cultural Theory. Boulder, CO:
Westview Press.
Thompson, Michael, Gunnar Grendstad, and Per Selle. 1999. Cultural Theory as Political Science. London:
Routledge.
Timmermans, Danielle R. M., Lidewij Henneman, Remy A. Hirasing, and Gerrit Van der Wal. 2005.
“Attitudes and Risk Perception of Parents of Different Ethnic Backgrounds Regarding Meningococcal C Vaccination.” Vaccine 23 (25): 3329–35.
Wakefield, Andrew J, and Scott M. Montgomery. 2000. “Measles, Mumps, Rubella Vaccine: Through a
Glass, Darkly.” Adverse Drug Reactions and Toxicological Reviews 19 (4): 265–83.
Wakefield, Andrew J., Simon H. Murch, Andrew Anthony, John Linnell, David H. Casson, Mohsin Malik,
Mark Berelowitz, et al. 1998. “RETRACTED: Ileal-Lymphoid-Nodular Hyperplasia, Non-Specific
Colitis, and Pervasive Developmental Disorder in Children.” Lancet 351 (9103): 637–41.
Wallace, Amy. 2009. “An Epidemic of Fear: How Panicked Parents Skipping Shots Endangers Us All.”
Wired Magazine (October 19) [Online]. http://www.wired.com/magazine/2009/10/ff
_waronscience/all/1. Accessed September 5, 2014.
White, Halbert. 1980. “A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for
Heteroskedasticity.” Econometrica: Journal of the Econometric Society 48 (4): 817–38.
Wiener, Joshua G., and Tomas M. Koontz. 2010. “Shifting Winds: Explaining Variation in State Policies to
Promote Small-Scale Wind Energy.” Policy Studies Journal 38 (4): 629–51.
Wildavsky, Aaron. 1987. “Choosing Preferences by Constructing Institutions: A Cultural Theory of Preference Formation.” The American Political Science Review 81 (1): 4–21.
Wildavsky, Aaron, and Karl Dake. 1990. “Theories of Risk Perception: Who Fears What and Why?”
Daedalus 119 (4): 41–60.
Woo, Emily Jane, Robert Ball, Ann Bostrom, Sean V. Shadomy, Leslie K. Ball, Geoffrey Evans, and Miles
Braun. 2004. “Vaccine Risk Perception among Reporters of Autism after Vaccination: Vaccine Adverse
Event Reporting System 1990–2001.” American Journal of Public Health 94 (6): 990–95.
Wooten, Karen G., Elizabeth T. Luman, and Lawrence E. Barker. 2007. “Socioeconomic Factors and
Persistent Racial Disparities in Childhood Vaccination.” American Journal of Health Behavior 31 (4):
434–45.

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

551

Appendix 1. Demographic Comparison of Survey Sample and
U.S. National Population
Survey Sample

U.S. Population*

48.1%
51.9%

48.1%
51.9%

12.2%
56.1%
31.7%

13.2%
57.0%
28.8%

97.3%
45.0%

83.1%
24.3%

76.6%
8.0%
6.0%
9.4%

72.7%
11.5%
11.0%
4.7%

50.3%
34.6%
15.1%

57.3%
29.3%
13.4%

Gender
Male
Female
Age
18–24
25–54
>54
Education
High School or Higher
College Grad or Higher
Race
White, non-Hispanic
Black
Hispanic
Other
Household Income
$0–$49,999
$50,000–$99,999
$100,000 and above
*Source: Herron and Jenkins-Smith (2006, p. 180).

Appendix 2. Dependent Variables and Measures
Variable
Preference toward mandatory vaccination policy
Preference toward religious exemption policy
Preference toward philosophical exemption policy
Parents, not government, as chief immunization decision makers

Measure
How do you feel about vaccine requirements for school entry? (1 = Strongly oppose to 7 = Strongly support)
How do you feel about religious exemptions from vaccine requirements? (1 = Strongly oppose to 7 = Strongly support) How do you feel about exemptions from vaccine requirements based on the parents’ philosophy or beliefs?
(1 = Strongly oppose to 7 = Strongly support)
Parents, not the government, should make decisions about immunizing their children. (1 = Strongly disagree to
7 = Strongly agree)

552

Policy Studies Journal, 42:4

Appendix 3. Independent Variables and Measures
Variable
Hierarchism

Hierarchism index
Egalitarianism

Egalitarianism index
Individualism

Individualism index
Fatalism

Fatalism index

Measure
The best way to get ahead in life is to work hard and do what you are told to do. (1 = Strongly disagree to 7 = Strongly agree)
Our society is in trouble because we don’t obey those in authority. (1 = Strongly disagree to 7 = Strongly agree)
Society would be much better off if we imposed strict and swift punishment on those who break the rules. (1 = Strongly disagree to 7 = Strongly agree)
Index of above three items (α = 0.63)
What our society needs is a fairness revolution to make the distribution of goods more equal. (1 = Strongly disagree to 7 = Strongly agree)
Society works best if power is shared equally. (1 = Strongly disagree to
7 = Strongly agree)
It is our responsibility to reduce the differences in income between the rich and the poor. (1 = Strongly disagree to 7 = Strongly agree)
Index of above three items (α = 0.80)
Even if some people are at a disadvantage, it is best for society to let people succeed or fail on their own. (1 = Strongly disagree to 7 = Strongly agree)
Even the disadvantaged should have to make their own way in the world.
(1 = Strongly disagree to 7 = Strongly agree)
We are all better off when we compete as individuals. (1 = Strongly disagree to
7 = Strongly agree)
Index of above three items (α = 0.70)
Most of the important things that take place in life happen by random chance.
(1 = Strongly disagree to 7 = Strongly agree)
No matter how hard we try, the course of our lives is largely determined by forces beyond our control. (1 = Strongly disagree to 7 = Strongly agree)
For the most part, succeeding in life is a matter of chance. (1 = Strongly disagree to 7 = Strongly agree)
Index of above three items (α = 0.77)

Appendix 4. Results of Factor Analysis of Cultural Orientation Measures
Factor 1
Fatalism 1
Fatalism 2
Fatalism 3
Egalitarianism 1
Egalitarianism 2
Egalitarianism 3
Individualism 1
Individualism 2
Individualism 3
Hierarchism 1
Hierarchism 2
Hierarchism 3

SS loadings
Proportion Var.
Cumulative Var.

Factor 2

Factor 3

Factor 4

0.82
0.75
0.80
0.80
0.82
0.80
0.81
0.78
0.70
0.64
0.78
0.74
Factor 1

Factor 2

Factor 3

Factor 4

2.14
0.18
0.18

2.10
0.18
0.36

1.98
0.16
0.52

1.71
0.14
0.66

Note: Factor loadings less than 0.60 are not shown. We utilized the varimax rotation method in the analysis.

Song/Silva/Jenkins-Smith: Childhood Vaccination Policy

553

Appendix 5. Control Variables and Measures
Variable
Organic culture

Organic culture index
Political ideology

Personal importance of religious faith
Vaccine-related knowledge

Knowledge index
Parent with children over 18
Parent with children under 18
Age
Gender
Race
Education
Income

Measure
Man-made toxins are much more dangerous than those toxins found in nature. (1 = Strongly disagree to 7 = Strongly agree)
It is almost always better to try natural or homeopathic remedies first. (1 = Strongly disagree to 7 = Strongly agree)
In general, organic fruits and vegetables are healthier for you than non-organic ones. (1 = Strongly disagree to 7 = Strongly agree)
Index of above three items (α = 0.64)
Which of the following categories best describes your views?
(1 = Strongly liberal to 4 = Middle of the road to 7 = Strongly conservative) How important is religious faith in your life? (0 = Not at all important to 10 = Extremely important)
Even with mandatory vaccine programs, infectious diseases including measles, whooping cough and chickenpox, still occur in small numbers in the United States. (0 = No; 1 = Yes)
Vaccines typically cause many harmful side effects, illnesses, and even death. (0 = No; 1 = Yes)
Infants have natural immunity for most infectious diseases.
(0 = No; 1 = Yes)
Getting vaccinated will substantially reduce the likelihood of getting the disease, but it will not eliminate the chance of getting it completely. (0 = No; 1 = Yes)
Most health officials recommend that infants and children receive multiple vaccinations for different diseases at the same time.
(0 = No; 1 = Yes)
Diseases had already begun to disappear before vaccines were introduced, because of better hygiene and sanitation. (0 = No;
1 = Yes)
Index of above six items (A total number of correct answers)
1 = Parent who does not have any child under 18
1 = Parent who has at least one child under 18
Age on last birthday
1 = Male
1 = White, Not Hispanic
The highest level of education completed (1 = Elementary or some high school to 7 = Doctorate [of any type])
Total estimated annual household income (1 = 0–$10,000 to
21 = $200,000 or more)

554

Policy Studies Journal, 42:4

Appendix 6. Ordinal Logistic Regression Results
Independent
variable

Hierarchism
Egalitarianism
Individualism
Fatalism
Residual Dev.
AIC
Degrees of freedom

Dependent variable
Preference toward Preference toward Preference toward
Parents, not mandatory religious exemption philosophical government, as vaccination policy policy exemption policy chief immunization decision makers
0.370***
(0.056)
0.189***
(0.047)
0.094
(0.050)
−0.175***
(0.048)
3,262.03
3,304.03
1,054

−0.273***
(0.054)
−0.099**
(0.046)
0.173***
(0.049)
0.205***
(0.047)
3,816.07
3,816.07
1,053

−0.300***
(0.055)
−0.052
(0.046)
0.226***
(0.049)
0.209***
(0.046)
3,791.90
3,833.90
1,056

−0.235**
(0.054)
−0.018
(0.046)
0.282***
(0.049)
0.185***
(0.047)
3,669.15
3,911.15
1,058

**p < 0.05.
***p < 0.01. Standard errors in parentheses.
Note: Estimated coefficients for control variables and intercepts are not shown.

Appendix 7. OLS Regression Results Used for Posterior Simulation
Independent
variable

Hierarchism
Egalitarianism
Individualism
Fatalism
Intercept
F
Adjusted R2
Degrees of freedom

Dependent variable
Preference toward Preference toward Preference toward
Parents, not mandatory religious exemption philosophical government, as vaccination policy policy exemption policy chief immunization decision makers
Model 5

Model 6

Model 7

Model 8

0.264***
(0.041)
0.097***
(0.033)
0.015
(0.037)
−0.171***
(0.035)
4.504***
(0.219)
16.85***
0.052
1,149

−0.174***
(0.053)
−0.029
(0.043)
0.166***
(0.048)
0.233***
(0.046)
2.827***
(0.284)
11.38***
0.035
1,148

−0.210***
(0.050)
0.011
(0.041)
0.204***
(0.046)
0.266***
(0.044)
2.395***
(0.271)
17.94***
0.055
1,151

−0.141**
(0.050)
0.010
(0.040)
0.250***
(0.046)
0.229***
(0.043)
3.098***
(0.268)
18.120***
0.056
1,154

**p < 0.05.
***p < 0.01. Standard errors in parentheses.

This document is a scanned copy of a printed document. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material.