Does Morality Policy Exist? Testing a Basic Assumption
Mooney, Christopher Z., Schuldt, Richard G., Policy Studies Journal
After the 2004 presidential election, journalists, pundits, and pollsters claimed that moral values had become the main criterion on which voters assessed candidates. "Family and family values matter," opined the Atlanta Journal-Constitution, "they are not just relevant but dominant." (1) This hyperbole was quickly debunked by cooler heads in the social science community (Fiorina, 2005; Hillygus & Shields, 2005; Langer & Cohen, 2005; but see Mulligan, 2008), but in these reports, the popular press noticed a phenomenon that scholars had been exploring more systematically for over a decade, that beyond simple economic self-interest, basic moral values may have important and independent effects on political behavior (Carmines & Layman, 1997; Craig, Martinez, Kane, & Gainous, 2005; Layman, 2001; Weisberg, 2005).
In one branch of this literature, scholars have posed and tested a variety of hypotheses about how the politics of "morality policy" differ from the politics of run-of-the-mill, economically driven policies. Relative to other policies, morality policies are said to be adopted in greater congruence with public opinion (Camobreco & Barnello, 2008; Mooney & Lee, 2000) or not (Smith & Tatalovich, 2003), to have more ideological politics (Langer & Brace, 2005), to have noneconomic interest groups more powerful in their politics (Allen, 2005), to have their debates determined more by values than expertise (Lewis & Brooks, 2005), and to have their adoption influenced more by values than socioeconomic factors (Gibson, 2004; Mooney & Lee, 1995). In short, the politics of morality policy are said to differ from those of nonmorality policy in systematic, explainable ways.
One fundamental assumption upon which this line of scholarship is built is simply that this type of policy exists. That is, morality policy scholars assume that these policies have certain distinctive characteristics, namely that they generate conflicts of basic moral values, do not lend themselves to compromise, and are widely salient and technically simple (Mooney, 2001). (2) Indeed, any scholar evaluating Lowi's (1964) basic hypothesis that policy affects politics must assume that policies can be divided into distinct types based on theoretically relevant characteristics. While this assumption is often deceptively easy to make, it has been found to be quite problematic to support for some typologies (Anderson, 1997; Roberts & Dean, 1994; Tolbert, 2002).
We test this assumption of policy typology directly by asking: Does morality policy exist? Empirical support for hypotheses derived from the assumption of this typology provides indirect evidence of morality policy's existence. But heretofore the existence of morality policy has not been demonstrated directly. Indirect evidence of its existence may simply be spurious, attributable to some other factor that links those policies that have been deemed by scholars to be morality policies.
Using data from a telephone survey of just over seven hundred Illinoisans in 2005, we compare seven policies on four of morality policy's assumed characteristics, assessing whether they hang together in the pattern that scholars have suggested. We find that policies do indeed vary along most of these traits as morality policy scholars have assumed. In short, morality policy exists; there is a class of policies that have most of the bundle of characteristics claimed by morality policy scholars to distinguish them from other policies. Our analysis also suggests new avenues for research on this class of policies.
What Is Morality Policy?
Lowi (1964) popularized the notion that the nature of a policy under consideration could affect its politics. But ever since Lowi offered this tantalizing and fertile explanation for certain political behavior, scholars have debated how to categorize policies, both in theory and in research practice (Anderson, 1997; Roberts & Dean, 1994; Tolbert, 2002). To date, morality policy scholars have skirted this categorization problem. The typical study in the field begins with a theoretical discussion about what constitutes morality policy in general and then proposes hypotheses regarding its impact on some aspect of the policymaking process. Finally, the scholar simply selects one or more "obvious" exemplars of morality policy and tests these hypotheses. Policies relating to abortion and homosexuality have been used widely for this purpose (Camobreco & Barnello, 2008; Haider-Markel, 1999; Haider-Markel & Meier, 1996; Kane, 2007; Mooney & Lee, 1995; Oldmixon, 2005; Patton, 2007) since scholars seem to agree that in contemporary American politics, they are morality policies. In other such studies, policies relating to pornography (Brisbin, 2001; Smith, 2001), gambling (Pierce & Miller, 2004; Von Herrmann, 2002), and capital punishment (Langer & Brace, 2005; Mooney & Lee, 2000), among others (Cocca, 2004; Gibson, 2004; Glick & Hutchison, 2001; Lewis, 2006; Lewis & Brooks, 2005; Sharp, 2002) have also been assumed to be morality policies.
There is a fair degree of consensus in this literature about several common characteristics of morality policy. This distinctive bundle of qualities has been used to develop hypotheses about the distinctive politics of this policy type. Importantly, these distinguishing features of morality policy are assumed to exist in the perceptions of those involved in the debate rather than in anything related to policy substance (Eastvold, 2004; Mooney, 2001). That is, how an issue is framed, rather than its intrinsic content, leads to its classification as a morality policy. For example, Haider-Markel and Meier (1996, p. 333) argue that if "at least one advocacy coalition ... portray[s] the issue as one of morality or sin and use[s] moral arguments in its policy advocacy," then it is a morality policy. Thus, a policy is classified on this dimension through an assessment of people's beliefs and attitudes about it, rather than by, for example, an analysis of related legislation, the policy tools it employs, or the part of society it affects. So, for example, a policy that is in some way related to sex is not necessarily a morality policy; a policy is a morality policy only if people think about it in certain characteristic ways. In this article, we do not test this assumption about the existence and classification of morality policy. Rather, we rely on this assumption to assess the other assumed characteristics of morality policy, using surveys of morality policy scholars and the general public.
Among its scholars, the most frequently discussed feature of morality policy is that the conflict underlying at least some of its debate is based on fundamental first principles and basic moral values rather than, for example, economic interests (Button, Rienzo, & Wald, 1997, pp. 5-6; Gormley, 1986; Mooney, 2001; Tatalovich, Smith, & Bobic, 1994; Tatalovich & Daynes, 1988; Weisberg, 2005). As we have shown, political commentators and journalists, in addition to political scientists, find the conflict of such values especially unique and interesting. People do not chain themselves to cars or commit murder over a change in telephone regulation, but they have done these things--and more--in the debate over abortion regulation. Indeed, social scientists trying to explain such extreme behavior were the first to hypothesize that this unique type of policy existed (Hunter, 1994; Tatalovich & Daynes, 1988).
Morality policy has also been posited to have other characteristics in various studies: the difficulty of compromising on them, high public salience, and technical simplicity (Carmines & Stimson, 1980; Gormley, 1986; Mooney, 2001). These characteristics may be the result of the clash of basic values. For example, compromise may be especially difficult on morality policy because it is harder to split the difference between core moral values than, say, permissible levels of industrial benzene emissions. Furthermore, conflict over simple conceptions of right and wrong can generate debate that is less technically complex and engages a broader range of people. But while these assumed characteristics of morality policy might be better thought of as hypotheses about the impact of conflicts of basic values on politics, we follow morality policy scholars and consider them to be independent traits. This will yield a straightforward, multi-measure test of the morality policy assumption. A useful project of future research would be to sort out the causal relationships among these characteristics.
Assessing Morality Policy
To answer our research question, we conducted a measurement exercise in which the unit of analysis is a substantive policy category. The existence of morality policy as a broader policy class requires that the debate around certain policies generates more conflict of basic moral values than does that around other policies and that these same policies are more widely salient, technically simpler, and less amenable to compromise. We test the morality policy scholars' assumption about this classification by comparing how they assess certain substantive policies on these characteristics to how the general public assesses them. While the general public is usually not directly involved in the policymaking process, their views are relevant in assessing morality policy because, as studies have shown, morality policy politics are driven strongly by public opinion, or at least by elected officials' beliefs about public opinion (Camobreco & Barnello, 2008; Mooney & Lee, 2000).
To test whether morality policy scholars' policy classification is the same as the public's, we evaluated both scholars' and the public's assessments of substantive policies that take on a wide range of values on what might be thought of as a "morality policy scale." To do this, we first surveyed nine scholars who have published significant research on morality policy politics. (3) We asked these scholars to rank 14 policy areas in the order that they perceived their debate to have moral content in contemporary American politics. (4) We used these data from this survey to construct an extremely consistent Thurstone scale of morality policy for these scholars (Nunnally, 1978). The average pairwise Spearman's correlation between the rankings of these nine scholars was 0.90, with no correlation falling below 0.81; the Cronbach's alpha for the scale was 0.98, indicating high reliability. In short, these scholars' judgments on how these 14 policies fell on the morality policy dimension were very consistent. From these 14 policies, we selected the seven that scaled best and gave us the widest range of stimuli for our public opinion survey. From the highest to the lowest on our experts' ordinal morality policy scale, these policies were:
1. Gay marriage.
2. Abortion regulation.
3. Capital punishment.
4. Casino gambling.
5. Homeland security versus civil liberties.
6. National health insurance versus tax cuts.
7. Campaign contribution limits.
These policies represented the full range of these scholars' morality policy scale.
Next, in March-May 2005, we surveyed 707 Illinois residents on a variety of questions that related to four of the five assumed characteristics of morality policy. (5) Not every question was asked of every policy in order to keep the survey to a reasonable length. More problematic is the fact that since the survey was not conducted solely for this study, the questions do not parallel each of the assumed characteristics of morality policy perfectly. However, we have several good indicators of four of these assumptions, allowing us to draw valid conclusions about whether these policies and their politics are arranged as morality policy scholars suggests they should be.
Conflict of Basic Values
Existing scholarship suggests that the primary dimension that differentiates morality policy from nonmorality policy is the extent to which its public debate involves a conflict of basic moral values (Mooney, 2001). Morality policies are said to generate such conflict, while nonmorality policies do not. To assess our public opinion survey respondents' perceptions of our sample of policies on this dimension, we asked them four questions about the influence of values and objective information in the decision making of both themselves and others. We assume that our respondents saw these two sources of decision-making guidance as distinct and that they understood the difference between them. Our hypothesis is that morality policy decision making is based more on values, while nonmorality policy decision making is based more on objective information.
First, for each of our seven policy areas, our respondents were asked "Do you think this is an issue where more information helps people to have an opinion, or is this an issue where people just need to apply their basic moral values to have an opinion?" We posed this straightforward query to assess the respondents' beliefs about the relevance of values and facts to other people's decision making in each of these policy areas. In Table 1, we present the aggregated responses to this question for each policy. Since this was a forced-choice question, the percentages of respondents who chose "more information" were the inverse of those who chose "apply basic values" (except for those who chose not to respond).
The percentages in Table 1 demonstrate a clear pattern of variation along the morality policy dimension, with these policies clustering into three groups, as indicated in the table. By ratios of almost 1.6:1, more respondents believed that other people's decisions about gay marriage and abortion regulation policy are made by simply applying their basic values rather than by gathering information. On the other end of the dimension, by ratios in the range of 2.5:1 to nearly 3:1, more respondents believed that on civil liberties/homeland security, national health insurance/tax cuts, and campaign contribution limits, more information helps people make decisions. Capital punishment and casino gambling fall in the middle here, with either a large plurality or a majority reporting that information was more helpful than values but by margins far smaller than for the three policies on the right end of the table. For each group, the point estimates for policies are closer within the group than they are to estimates for policies outside the group.
For Groups I and III, the within-group point estimates are very close and not statistically distinct (< 0.05), and each is very different and statistically distinct from estimates in the other two groups. For Group II, the difference between the two-point estimates is statistically significant, but it is still less than the differences between these estimates and those in the other two groups. (6)
The respondents' ordering of these policy areas along this dimension and this clustering into these three groups follow closely the scale developed from our survey of morality policy scholars. Furthermore, a review of published morality policy scholarship shows that this is probably how researchers in the field would also rank these policy areas. That is, many morality policy studies examine the politics of abortion and gay rights (e.g., Camobreco & Barnello, 2008; Lewis, 2006; Lewis & Brooks, 2005; Mooney & Lee, 1995; Oldmixon, 2005; Patton, 2007), while somewhat fewer have looked at capital punishment and gambling policy (e.g., Langer & Brace, 2005; Mooney & Lee, 2000; Pierce & Miller, 2004; Von Herrmann, 2002); no published work has claimed that any of the other three policies in our survey was a morality policy.
Next, we examine the influences on a respondent's own decision making on these issues. First, for each of the seven policies, respondents were asked about the degree to which their "religious or moral beliefs and values" influenced their "thinking on this issue." We hypothesize that the reported influence here will be greater for those policies higher up the morality policy scale. Table 2 reports the aggregated responses to this question, showing the percentages of respondents who said their moral values influenced their opinion "a lot" or "some" versus the percentages who responded "not much" or "none." These percentages have a pattern quite similar to that in Table 1, although not an exact duplicate. Respondents' decisions on Group I policies are clearly influenced a great deal by their religious and moral beliefs, but so too are their decisions on capital punishment, a Group II policy. On the other hand, opinions on casino gambling (Group II) have a distribution of responses much more like Group III policies. In short, the policies on the extremes of our morality policy scale continue to satisfy expectations, but the policies in the middle group diverge somewhat here.
We also asked respondents about the degree to which other factors influenced their thinking on five of these policies, (7) including "the views expressed by leaders of your church or religion" (8) and "the views of policy experts." We hypothesized that morality policy decisions would be influenced more by religious leaders' opinions than those of policy experts for the same reasons that we hypothesized that religious beliefs and basic moral values would influence these decisions more than information. Table 3 shows that the survey responses support this hypothesis reasonably well. Religious leaders are more frequently cited as influential than are policy experts for both gay marriage and abortion regulation (Group I), while policy experts are far more frequently cited as influential for civil liberties/homeland security policy (Group III). For capital punishment and casino gambling (Group II), slightly more respondents cite policy experts than religious leaders. The differences between the percentages of influence reported for policy experts and for religious leaders are statistically significant for both Group I and Group III policies but in opposite directions, thus supporting our hypothesis. The percentages for Group II policies are not statistically distinct, again placing them between the two other groups on the morality policy scale. Furthermore, while the difference between these percentages is statistically indistinct within the groups in Table 3, they are each statistically distinct from each difference for each policy in each other group, (9) further supporting our hypothesis.
Next, we assess the relative influence of policy experts and religious leaders across four policy areas in a somewhat different way. After asking how likely it was that their views on the issue could change, respondents were asked, "How likely is it that you would re-think your opinion on this issue if the leaders of your church or religion (10) (policy experts) expressed an opinion different than yours?" Table 4 shows that religious leaders are more frequently cited as being influential than policy experts in this way only for a Group I policy (gay marriage). The other three policies in Group II and Group III show the opposite pattern, with increasing differences in the two percentages as we move across the morality policy scale. The percentages are not far apart for the first Group II policy (capital punishment), but for the Group III policy (civil liberties/homeland security), opinions are said to be influenced by policy experts more than 1.5 times as often as by the religious leaders. Results for the other Group II policy (casino gambling) fall between these two. Thus, the morality policy hypothesis is supported again. (11)
Lack of Compromise
Next, we use our public opinion survey to test the three other, perhaps less central, assumptions of the morality policy literature: that morality policy debate is less amenable to compromise and that the issues are more salient and technically simple. Due to restrictions on questionnaire length, we have less extensive data on these characteristics, but our results are nonetheless useful in understanding the existence and dimensionality of morality policy.
We assessed the amenability of our test policies to compromise in two ways. Our first approach was direct. For all seven policy areas, we asked the respondents, "Some issues are subject to the normal give-and-take of political compromise, while for other issues, compromise is difficult, if not impossible. For you personally, is this issue one where compromise is: [response]?" Our hypothesis was that fewer respondents would find it possible to compromise on morality policies than on nonmorality policies.
As seen in Table 5, substantially fewer respondents did indeed say that compromise was possible for gay marriage and abortion regulation policy (Group I) than for civil liberties/homeland security, national health insurance/tax cuts, and campaign contribution limits policy (Group III). In fact, while about the same number of, or somewhat more, respondents said that compromise was impossible than said it was possible for the Group I policies, over twice as many thought that compromise was possible than was impossible for Group III policies. The difference between the possible and impossible response percentages (last row of Table 5) is statistically distinct for each pair of policies across these two groups. Thus, all these Group I versus Group III comparisons support the morality policy hypothesis.
On the other hand, the Group II policies are quite different from one another on this question. Specifically, while capital punishment fits between Groups I and III, as the morality policy hypothesis suggests, casino gambling stands out, with many more people believing that compromise is possible on this than any other issue. Furthermore, the differences of percentages are statistically distinct between the Group II policies (as they are between the Group I policies). This anomaly may be due to the high-profile discussion of casino gambling in Illinois during and just before our survey was conducted. Specifically, this debate was about the extent of legalized gambling in the state, not its existence (a question that had been settled over a decade earlier), so compromise might well have seemed possible on it to our respondents. The results in Table 5 suggest the potential for a difference between morality policy debate over the extent of an activity and debate over its existence. This difference may have important theoretical and practical implications and should be investigated further in other contexts.
Our second approach to assessing a policy debate's amenability to compromise is to evaluate the strength of opinion on it. We assume that the stronger a person's opinion on an issue, the harder it will be for him or her to compromise with someone holding another view. We assessed the strength of our respondents' opinions on four of our policies by asking: "How likely is it that your views on this issue could change?" Table 6 shows the aggregated responses to this question, revealing a pattern of variation that is clearly correlated with the morality policy scale. The proportion saying they are not likely to change their opinions is largest for gay marriage (Group I), followed by capital punishment and casino gambling (Group II), and with the lowest being for civil liberties/homeland security (Group III). These percentages are all statistically distinct from one another. Thus, using either of these approaches, we find that the degree of a policy's amenability to compromise is correlated with the morality policy scale. Thus, the morality policy assumption is supported by these data.
The assumption that morality policies are technically simpler than nonmorality policies arises from the observation that their debates hinge on basic moral values, rather than on fact-based arguments about cause and effect and policy implementation (Meier, 1994). This dimension is difficult to measure with a public opinion survey, but three of the survey questions we have already discussed are relevant here.
First, consider the question of whether this is "an issue where more information helps people to have an opinion" or "an issue where people just need to apply their basic moral values to have an opinion." In addition to measuring perceived moral content, this question can also be interpreted as indicating the technical complexity of an issue. That is, it probably takes more information to make decisions about more complex issues. The top row of Table 1 shows that fewer respondents think that information is relevant to decision making (at least relative to basic moral values) on morality policies. Confirming this pattern, each of the between-group differences on this top row is statistically significant at the 0.05 level, and each of the within-group differences for Groups I and III is not. While the Group II top-row percentages are statistically distinct, they are in the predicted order and between those in Groups I and III.
Likewise, we can also reinterpret two of the questions comparing the impact of policy experts and religious leaders on a respondent's opinion and on his or her rethinking of that opinion. In Tables 3 and 4, we again see the pattern that, as we move from Group I to Group II to Group III, the influence of policy experts climbs. More specifically, in Table 3, we see that policy experts are an influence on far fewer respondents' opinions on gay marriage and abortion regulation than on civil liberties, with casino gambling and capital punishment falling between. And in Table 4, we see that when presented with a situation where their opinions differed from those of policy experts, far fewer respondents would rethink their opinion on gay marriage than would do so on civil liberties. Again, capital punishment and casino gambling fall between. Furthermore, all the between-group differences in these percentages in both tables are statistically significant.
Thus, the patterns in Tables 1, 3, and 4 show that our respondents believe that information is less influential in decision making for morality policy than for nonmorality policy. Interpreted as we have, this suggests that morality policy is thought to be technically simpler than other policies, as morality policy scholars have assumed.
Finally, we examine whether issue salience is related to our morality policy scale by asking our respondents how important each issue is to them. The word "salient" is too technical to include in a public opinion survey, and while "important" is more general and has some slightly different connotations than "salient," they are quite closely related. So for four of our policies, we asked our respondents: "How important would you say this issue is to you personally?"
Contrary to expectations, Table 7 shows that issue importance does not scale as assumed by morality policy scholars. Civil liberties/homeland security policy is far more important to our respondents than gay marriage policy, and capital punishment and casino gambling policy do not display importance percentages that are similar to one another or to either other policy. Indeed, 58 percent of the respondents said that gay marriage policy (Group I) is either not important to them or only somewhat important. This percentage is lower than that for casino gambling (Group II) but much higher than those for civil liberties/homeland security policy (Group III) and capital punishment (Group II). In short, these results do not support the assumption that morality policy is more salient to citizens than nonmorality policy. In fact, they suggest that morality policy may actually be far less important to the average citizen than some nonmorality policies. This is another relationship that merits further research.
Does morality policy exist? In a word, yes. Most important, we found that certain policies generate greater conflict over basic moral values than others and that these are the ones that morality policy scholars have often used as exemplars of the type, especially abortion regulation and gay rights policy. The variation on this dimension measured in different ways in our data scaled nearly perfectly with our Thurstone scale of morality policy scholars' judgments about these policies. Furthermore, these same policies appeared to be less amenable to compromise and less technically complex, just as morality policy scholars have assumed. Of course, all this is not too surprising to any close observer of the political process and the media's obsession with morals-based voting after the 2004 election. But we are the first to assess this basic assumption of morality policy scholarship systematically and confirm it; this is a contribution, if not a momentous one. But in addition to this simple confirmation, our study contributes to the understanding of morality policy and politics in a few more subtle ways.
First, we found no evidence that morality policy is more salient to the general public than nonmorality policy, as morality policy scholars had argued. In fact, our admittedly limited data indicate that gay marriage policy (which generates conflict of basic values and is thought not to be amenable to compromise and technically simple) was not at all important to most of our survey respondents, especially as compared to civil liberties/homeland security policy (which does not generate conflicts of basic values). Perhaps what scholars and journalists think is broad public salience for morality policy is really only salience among the interested and elite public. Or in a richer interpretation, perhaps the general public can be simply more easily motivated to attend to morality policy than nonmorality policy by these elites when public action is needed in the policymaking process. Further research is needed to assess this interesting hypothesis about elites and social movements on morality policy (Kane, 2007).
Another contribution of our study is that it highlights a very basic problem with much morality policy research: the lack of variation on the central independent variable. To draw an inference about the effect of a variable, a researcher must observe outcomes with cases that have different levels of that variable (King, Keohane, & Verba, 1994). Many morality policy studies have described the politics of only a single policy (12)--a morality policy--and then compared those politics to some assumed or implied politics of nonmorality policy (e.g., Gibson, 2004; Mooney & Lee, 1995). A better test would be to observe a given political process for policies that generate various levels of conflict on basic values and compare these politics directly, keeping all other conditions as equal as possible. Even if the level of such conflict cannot be measured quantitatively for a given set of policies, researchers should certainly be able to make relative judgments about this characteristic. Even comparing only two policies, a morality policy and a nonmorality policy, on the same aspect of politics would be far better than no explicit comparison at all.
Finally, our confirmation of the morality policy assumption also suggests a more general research direction, one that addresses questions in which policy type is the dependent variable rather than the independent variable. For example, morality policy scholars should examine why certain policy debates generate greater conflict over basic moral values than others. If policy type affects politics, as Lowi (1964) suggested, it behooves political scientists to explore the determinants of policy type. Perhaps the most interesting questions in this line of thought are how and why a policy can change its policy type (Haider-Markel & Meier, 1996; Pierce & Miller, 2004). Policy type change may not only be intriguing theoretically, it can also have important political implications. For example, Pierceson (2005) argues that when the debate about same-sex marriage (or civil unions) shifted from moral to legal arguments following court cases in Vermont, Hawaii, and Massachusetts in the past decade and a half, Americans began supporting such legal relationships to a degree that would have been unthinkable even 10 years before. Morality policy has done much to define the polarized, high-pitched politics of the United States today. Understanding how policies become framed in this way may help us learn how to disengage from moral political combat so as to address better our more tractable public problems.
Christopher Z. Mooney is a professor of political science in the Institute of Government and Public Affairs at the University of Illinois at Springfield. His research interests include morality politics and state legislative term limits.
Richard G. Schuldt is the director of the Survey Research Office at the University of Illinois at Springfield. His research interests include public opinion and voting behavior, including issue publics and civic engagement.
An earlier version of this article was presented at the 2006 meeting of the American Political Science Association, Philadelphia, PA. The authors wish to thank Sandra Vergari for her comments on this project and to acknowledge the financial support of the Center for State Policy and Leadership and its Survey Research Office, both located at the University of Illinois at Springfield, and the Institute of Government and Public Affairs at the University of Illinois. The survey of Illinois residents discussed in this article was conducted for the Public Policy Summit on Politics and Religion held by the Center for State Policy and Leadership, April 2005.
1. Anonymous, "ELECTION 2004: Religious Vote Fuels Victory for GOP," Atlanta Journal-Constitution, November 4, 2004, A19.
2. Morality policy is also said to generate greater citizen participation than average (Mooney, 2001). For two reasons, we do not include an assessment of this characteristic in our analysis. First, with so many policy stimuli, we needed to make some limiting choices to keep our public opinion questionnaire length manageable. Second, and more important, the rate of U.S. citizen participation on any given issue is so low that we felt that any true participation on these issues among our respondents would be overwhelmed by random error in our survey data.
3. We conducted this survey of morality policy scholars via email in January and February 2005.
4. The 14 policy areas we asked these scholars to rank were: homeland security versus civil liberties, abortion regulation, environmental protection, sex education, campaign finance regulation, medical use of marijuana, gay rights, capital punishment, road construction, cash assistance to the poor, casino gambling regulation, national health insurance versus tax cuts, teaching evolution as scientific truth, and right-to-die regulations.
5. Single-state studies must always justify their generalizability (Nicholson-Crotty & Meier, 2002). Is it reasonable to generalize from public opinion in Illinois to that in the entire country or even beyond? While no state is completely representative of the United States on all characteristics, Illinois is probably as representative as any and more representative than most. While Illinois is somewhat more liberal and Democratic than the national average, it is a very diverse state, paralleling the United States on a range of important social, economic, and political characteristics, such urbanization, per capita wealth, income distribution, higher education spending, women in the state legislature, and many others (Donovan, Mooney, & Smith, 2008, chap. 1). See also Stephen Ohlemacher, "Analysis Ranks Illinois Most Average State," Associated Press, May 17, 2007.
6. We assess the statistical significance of the differences between the percentages we report in two ways. The first approach is a simple z-test for the difference of proportions from different populations (Blalock, 1979, pp. 232-34). This requires the assumption that the policy populations are different, an assumption that probably holds. But to double check our statistical conclusions, we also calculate the 0.05 confidence interval around each percentage and compare the intervals of each pair of percentages between which we assess the difference. Where these intervals do not overlap, we consider the difference between the point estimates of that pair of percentages to be statistically significant For a sample of seven hundred from an infinite population, the following table reports the size of a p < 0.05 confidence interval around various percentages (the size of the interval varies with the size of the percentage):
Percentage p < 0.05 Confidence Interval 10 (or 90) +/-2.2% 20 (or 80) +/-3.0% 25 (or 75) +/-3.2% 30 (or 70) +/-3.4% 35 (or 65) +/-3.5% 40 (or 60) +/-3.6% 50 +/-3.7%
For each of the comparisons reported, these two approaches produced the same conclusions.
7. To limit the length of the survey, we asked our respondents about only a subset of these policies for some of our questions. See the Appendix for more details on the survey.
8. Since people who are not religious may also base some of their policy opinions on basic moral values, asking about the influence of "your church or religion" may underestimate the number of people basing their policy decisions on such values. However, we have no reason to believe that the proportion of nonreligious people basing their opinions on values would have a substantially different pattern of variation across the morality policy scale than is the case for religious people. So while the data in the second line of Table 3 likely underestimate the percentage of these decisions that are influenced by basic values, the relative level of basic-values influence across the morality policy scale that these figures indicate is probably not biased.
9. See Blalock (1979, pp. 234-36) for the statistical test of the difference of difference in proportions used in Tables 3, 4, and 5.
10. See Footnote 8.
11. This pattern is not driven disproportionately by those who are most or least likely to change their opinions (Mooney & Schuldt, 2006).
12. On the other hand, Sharp (2002) looks at the politics of more than one morality policy. But even this study does not explicitly compare these politics to those of parallel nonmorality policies.
Allen, Mahalley D. 2005. "Laying Down the Law? Interest Group Influence on State Adoption of Animal Cruelty Felony Laws." Policy Studies Journal 33: 443-58.
Anderson, James L. 1997. "Governmental Suasion: Refocusing the Lowi Policy Typology." Policy Studies Journal 25: 266-82.
Blalock, Hubert M. 1979. Social Statistics, 2nd ed. New York: McGraw-Hill.
Brisbin, Richard A., Jr. 2001. "From Censorship to Ratings: Substantive Rationality, Political Entrepreneurship, and Sex in the Movies." In The Public Clash of Private Values: The Politics of Morality Policy, ed. Christopher Z. Mooney. Chatham, NJ: Chatham House, 91-114.
Button, James W., Barbara A. Rienzo, and Kenneth D. Wald. 1997. Private Lives, Public Conflicts: Battles over Gay Rights in American Communities. Washington, DC: CQ Press.
Camobreco, John F., and Michelle A. Barnello. 2008. "Democratic Responsiveness and Policy Shock: The Case of State Abortion Policy." State Politics and Policy Quarterly 8: 48-65.
Carmines, Edward G., and Geoffrey C. Layman. 1997. "Issue Evolution in Postwar American Politics: Old Certainties and Fresh Tensions." In Present Discontents: American Politics in the Very Late 20th Century, ed. Byron E. Shafer. Chatham, NJ: Chatham House, 132-53.
Carmines, Edward G., and James A. Stimson. 1980. "The Two Faces of Issue Voting." American Political Science Review 74: 78-91.
Cocca, Carolyn E. 2004. Jailbait: The Politics of Statutory Rape Laws in the United States. Albany, NY: SUNY Press.
Craig, Stephen C., Michael D. Martinez, James G. Kane, and Jason Gainous. 2005. "Core Values, Value Conflict, and Citizens' Ambivalence about Gay Rights." Political Research Quarterly 58: 5-17.
Donovan, Todd, Christopher Z. Mooney, and Daniel A. Smith. 2008. State and Local Politics: Institutions and Reform. Belmont, CA: Wadsworth.
Eastvold, Jonathan C. 2004. "God, the State, and War: Toward a Containment Strategy for Religious Militancy." Presented at the Annual Meeting of the International Studies Association, Montreal, Quebec.
Fiorina, Morris P. 2005. Culture War? The Myth of a Polarized America. New York: Pearson Longman.
Gibson, M. Troy. 2004. "Culture Wars in State Education Policy: A Look at the Relative Treatment of Evolutionary Theory in State Science Standards." Social Science Quarterly 85: 1129-49.
Glick, Henry R., and Amy Hutchinson. 2001. "Physician-Assisted Suicide: Agenda Setting and the Elements of Morality Policy." In The Public Clash of Private Values: The Politics of Morality Policy, ed. Christopher Z. Mooney. Chatham, NJ: Chatham House, 55-72.
Gormley, William T. 1986. "Regulatory Issue Networks in a Federal System." Polity 18: 595-620.
Haider-Markel, Donald P. 1999. "Morality Policy and Individual-Level Political Behavior: The Case of Legislative Voting on Gay and Lesbian Issues." Policy Studies Journal 27: 735-49.
Haider-Markel, Donald P., and Kenneth J. Meier. 1996. "The Politics of Gay and Lesbian Rights: Expanding the Scope of the Conflict." Journal of Politics 58: 332-49.
Hillygus, D. Sunshine, and Todd G. Shields. 2005. "Moral Issues and Voter Decision Making in the 2004 Presidential Election." PS: Political Science and Politics 38: 201-9.
Hunter, James D. 1994. Before the Shooting Begins: Searching for Democracy in America's Culture War. New York: Free Press.
Jacoby, William G. 2000. "Issue Framing and Public Opinion on Government Spending." American Journal of Political Science 44: 750-67.
Kane, Melinda D. 2007. "Timing Matters: Shifts in the Causal Determinants of Sodomy Law Decriminalization, 1961-1998." Social Problems 54: 211-39.
King, Gary, Robert O. Keohane, and Sidney Verba. 1994. Designing Social Inquiry. Princeton, NJ: Princeton University Press.
Langer, Gary, and Jon Cohen. 2005. "Voters and Values in the 2004 Election." Public Opinion Quarterly 69: 744-61.
Langer, Laura, and Paul Brace. 2005. "The Preemptive Power of State Supreme Courts: Adoption of Abortion and Death Penalty Legislation." Policy Studies Journal 33: 317-40.
Layman, Geoffrey C. 2001. The Great Divide: Religious and Cultural Conflict in American Party Politics. New York: Columbia University Press.
Lewis, Gregory B. 2006. "Public Spending on the Arts as Morality Policy: The Structure of Public Attitudes." Policy Studies Journal 34: 131-8.
Lewis, Gregory B., and A. C. Brooks. 2005. "A Question of Morality: Artists' Values and Public Funding for the Arts." Public Administration Review 65: 8-17.
Lowi, Theodore J. 1964. "American Business, Public Policy, Case Studies, and Political Theory." World Politics 16: 677-715.
Meier, Kenneth J. 1994. The Politics of Sin. Armonk, NY: M.E. Sharpe.
Mooney, Christopher Z. 2001. "The Public Clash of Private Values." In The Public Clash of Private Values: The Politics of Morality Policy, ed. Christopher Z. Mooney. Chatham, NJ: Chatham House, 3-20.
Mooney, Christopher Z., and Mei-Hsien Lee. 1995. "Legislating Morality in the American States: The Case of Pre-Roe Abortion Regulation Reform." American Journal of Political Science 39: 599-627.
______. 2000. "The Influence of Values on Consensus and Contentious Morality Policy: U.S. Death Penalty Reform, 1956-1982." Journal of Politics 61: 223-39.
Mooney, Christopher Z., and Richard G. Schuldt. 2006. "Does Morality Policy Exist? Testing a Basic Assumption." Presented at the Annual Meeting of the American Political Science Association, Philadelphia, PA.
Mulligan, Kenneth. 2008. "The 'Myth' of Moral Values Voting in the 2004 Election." PS: Politics and Political Science 41: 1109-14.
Nicholson-Crotty, Sean, and Kenneth J. Meier. 2002. "Size Doesn't Matter: In Defense of Single-State Studies." State Politics and Policy Quarterly 2: 411-22.
Nunnally, Jum C. 1978. Psychometric Theory, 2nd ed. New York: McGraw-Hill.
Oldmixon, Elizabeth Anne. 2005. Uncompromising Positions: God, Sex, and the U.S. House of Representatives. Washington, DC: Georgetown University Press.
Patton, Dana. 2007. "The Supreme Court and Morality Policy Adoption in the American States." Political Research Quarterly 60: 468-88.
Pierce, Patrick A., and Donald E. Miller. 2004. Gambling Politics: State Government and the Business of Betting. Boulder, CO: Lynne Rienner.
Pierceson, Jason. 2005. Courts, Liberalism, and Rights: Gay Law and Politics in the United States and Canada. Philadelphia, PA: Temple University Press.
Roberts, R., and L. E. R. Dean. 1994. "An Inquiry into Lowi's Policy Typology: The Conservation Coalition and the 1985 and 1990 Farm Bills." Environment and Planning 12: 71-86.
Shaffer, Teri R., and Bradley S. O'Hara. 1995. Ethnicity and Framing Effects on the Distribution of Satisfaction Responses." Southeastern Louisiana University, Unpublished manuscript.
Sharp, Elaine B. 2002. "Culture, Institutions, and Urban Officials' Responses to Morality Issues." Political Research Quarterly 55: 861-84.
Smith, T. Alexander, and Raymond Tatalovich. 2003. Cultures at War: Moral Conflicts in Western Democracies. Peterborough, ON: Broadview.
Smith, Kevin B. 2001. "Clean Thoughts and Dirty Minds: The Politics of Porn." In The Public Clash of Private Values: The Politics of Morality Policy, ed. Christopher Z. Mooney. Chatham, NJ: Chatham House, 187-200.
Tatalovich, Raymond, and Byron W. Daynes, eds. 1988. Social Regulatory Policy. Boulder, CO: Westview.
Tatalovich, Raymond, T. Alexander Smith, and Michael P. Bobic. 1994. "Moral Conflict and the Policy Process." Policy Currents 4: 1-7.
Tolbert, Carolyn J. 2002. "Rethinking Lowi's Constituent Policy: Governance Policy and Direct Democracy." Environment and Planning 20: 75-93.
Von Herrmann, Denise K. 2002. The Big Gamble: The Politics of Lottery and Casino Expansion. Westport, CT: Greenwood.
Weisberg, Herbert F. 2005. "The Structure and Effects of Moral Predispositions in Contemporary American Politics." Journal of Politics 67: 646-68.
Appendix: Survey Details
The public opinion data presented in this article were collected through a telephone survey conducted by the Survey Research Office of the Center for State Policy and Leadership at the University of Illinois at Springfield in conjunction with the Public Policy Summit on Politics and Religion held by the Center for State Policy and Leadership in April 2005. Interviews were conducted from March 22 through May 25, 2005, with just over seven hundred respondents in randomly selected Illinois households (actual n = 707). The sample was stratified by region (city of Chicago, Chicago suburbs, north and central Illinois, and southern Illinois), and our reported statewide results have been weighted by region, gender, education level, age, and race (white versus nonwhite) to arrive at a more representative sample, correcting for response biases that are typical in telephone surveys. Randomly selected households were called up to six times at varying times of the week and day by trained research interviewers.
A pretested version of the interview schedule included the full morality policy follow-up question protocol for each of the seven issue areas. However, this procedure was found to be too time consuming, so we limited the full question follow-up protocol to four issue areas. Even with this editing, the average length of the interview was nearly one-half hour (mean = 28.5 minutes; median = 27 minutes).
The survey questions are given in the text and the tables. For the policy areas about which the respondents were asked, a respondent was first introduced to each policy area by (i) referring to it and then (ii) asking his or her position on it. In this way, the respondents had a consistent understanding of the policy area (Jacoby, 2000; Shaffer & O'Hara, 1995). We then asked the questions described in the text about each policy. The wording for the policy introduction and position questions in the survey was as follows:
People have differing opinions on the issue of gay marriage. Which of the following comes closest to what you think?
1. Gays should be allowed to marry.
2. Gays should not be allowed to marry but should be allowed to enter into legal relationships that give them the same rights as married couples have.
3. Gays should only have their most basic rights protected.
4. Government should not recognize gay relationships in any way.
People have differing opinions on the abortion issue. Which of the following comes closest to what you think?
1. Abortion should never be permitted.
2. Abortion should hardly ever be permitted.
3. Abortion should be prohibited by government only in rare circumstances.
4. Abortion should never be prohibited by government.
People have differing opinions on capital punishment--the issue of whether or not individuals can be put to death for committing certain crimes. Which of the following comes closest to what you think?
1. We should not have capital punishment in Illinois for any crime.
2. We should have capital punishment only in rare circumstances.
3. We should have capital punishment for a variety of serious and violent crimes.
People have differing opinions on the issue of casino gambling in Illinois. Which of the following comes closest to what you think?
1. We should increase the amount of casino gambling in Illinois.
2. We have about the right amount of casino gambling in Illinois.
3. We should eliminate casino gambling in Illinois.
Civil liberties versus homeland security:
People have differing opinions on the issue of homeland security and civil liberties. Which of the following comes closest to what you think?
1. We should do all we can to prevent future terrorist attacks even if this means restricting civil liberties.
2. Concerns about homeland security and civil liberties need to be balanced against each other.
3. In times like this, we need to do all we can to prevent the government from taking away our civil liberties.
National health insurance versus tax cuts:
Please tell me if you completely agree, mostly agree, mostly disagree, or completely disagree with this statement:
"The government should guarantee health insurance for all citizens, even if it means repealing most of the tax cuts passed under President Bush."
Campaign contribution limits:
People have differing opinions about aspects of campaign finance regulations in Illinois elections. When it comes to how much groups and individuals can contribute to political campaigns in Illinois, which of the following comes closest to what you think?
1. We should have very strict limits on how much groups and individuals can contribute.
2. We should have some limits on how much they can contribute.
3. We should have no limits on how much groups and individuals can contribute.
Table 1. Influences on Other People's Decisions--More Information or Basic Moral Values? (Percentage Choosing Each Forced-Choice Response) Group I Group II Gay Abortion Capital Casino Marriage Regulation Punishment Gambling "More information" 35.7% (a) 35.2% 48.4% 57.1% "Apply basic moral 55.2% 55.6% 43.9% 36.6% values" No response 9.1% 9.2% 7.7% 6.4% n 655 660 668 622 Ratio (b) of +1.55:1 +1.58:1 -1.10:1 -1.56:1 "values" to "information" responses Group III Civil Liberties National Health Campaign versus Homeland Insurance versus Contribution Security Tax Cuts Limits "More information" 69.0% 69.2% 66.2% "Apply basic moral 27.3% 23.7% 23.9% values" No response 3.8% 7.1% 9.9% n 654 654 653 Ratio (b) of -2.53:1 -2.92:1 -2.77:1 "values" to "information" responses Note: These data are the result of asking this survey question for each of these policy areas: "Do you think this is an issue where more information helps people to have an opinion, or is this an issue where people just need to apply their basic moral values to have an opinion?" (a) The differences between the percentages within Groups I and III are not statistically significant at the 0.05 level, but the differences between the percentages within Group II are statistically significant at this level. Each of the differences in one group is statistically distinct from each of those in each of the others, at the 0.05 level. (b) A positive ratio indicates more "values" responses than "information" responses; a negative ratio indicates the reverse. Table 2. Influence of Religious/Moral Beliefs on a Respondent's Own Opinion (Percentage Choosing Each Response) Group I Group II Abortion Capital Casino Gay Marriage Regulation Punishment Gambling "A lot" or "some" 61.9% (a) 64.4% 61.1% 40.3% "Not much" or "none" 35.4% 32.1% 37.2% 58.1% No response 2.7% 3.5% 1.8% 1.7% n 658 630 675 624 Group III Civil Liberties National Health Campaign versus Homeland Insurance versus Contribution Security Tax Cuts Limits "A lot" or "some" 48.7% 44.8% 28.5% "Not much" or "none" 49.9% 53.5% 67.9% No response 1.5% 1.6% 3.6% n 654 614 607 Note: These data are the result of asking this survey question for each of these policy areas: "I'm going to read you a list of factors. For each, I'd like you to tell me how much it has influenced your thinking on this issue." (a) The percentages of "a lot" or "some" for gay marriage, abortion regulation, and capital punishment are not statistically distinct at the 0.05 level nor are the percentages for national health insurance and civil liberties. The "a lot" or "some" percentage for campaign contribution limits is statistically distinct from all other policies, and the percentages for gay marriage, abortion regulation, and capital punishment are statistically distinct from the other three. The percentage for casino gambling is statistically distinct from campaign contribution limits and is actually in the opposite direction from that predicted when compared to civil liberties and national health insurance. Table 3. Influences on a Respondent's Own Opinion--Policy Experts Versus Religious Leaders (Percentage "A Lot" or "Some") Group I Group II Gay Abortion Capital Casino Marriage Regulation Punishment Gambling Policy experts 24.2% 26.3% 40.1% 30.7% Religious leaders 39.5% 42.6% 35.1% 26.7% Difference (religious +15.3% (a) +16.3% -5.0% -4.0% leaders--policy experts) n 656 630 675 623 Group III Civil Liberties versus Homeland Security Policy experts 56.0% Religious leaders 35.6% Difference (religious -20.4% leaders--policy experts) n 654 Note: These data are the result of asking this survey question for each of these policy areas: "I'm going to read you a list of factors. For each, I'd like you to tell me how much it has influenced your thinking on this issue." (a) The difference between the percentages of "policy experts" and "religious leaders" responses for each policy are statistically significant at the 0.05 level for Groups I and III but not for Group II. Also, the within-group differences of these percentages (third row) are statistically indistinct from one another, but the between-group differences are all statistically significant. Table 4. Likelihood of a Respondent's Rethinking His/Her Opinion if the Views of Policy Experts/Religious Leaders Are Different (Percentage "Very" or "Somewhat" Likely) Group III Group II Civil Liberties Group I Capital Casino versus Homeland Gay Marriage Punishment Gambling Security Respondent would rethink if the following differ: Policy experts 12.4% 25.5% 29.5% 41.1% Religious leaders 13.9% 20.2% 19.1% 24.8% Difference +1.5% (a) -5.3% -10.4% -16.3% (religious leaders--policy experts) n 656 672 623 654 Note: These data are the result of asking this survey question for each of these policy areas: "How likely is it that you would re-think your opinion on this issue if the leaders of your church or religion (policy experts) expressed an opinion different than yours?" Just prior to this question, respondents were asked: "How likely is it that your views on this issue could change?" (a) Each of the between-group difference of percentages is statistically distinct. Within Group II, the difference in the percentages in this row just misses being statistically significant. Table 5. Potential for Compromise on Respondent's Own Position on Each Issue (Percentage Choosing Each Response) Group I Group II Gay Abortion Capital Casino Marriage Regulation Punishment Gambling Possible 31.4% 23.8% 32.0% 59.8% Difficult 28.3% 29.9% 39.5% 17.1% Impossible 34.5% 39.6% 24.0% 17.6% No response 5.8% 6.7% 4.6% 5.5% n 656 661 670 623 Difference +3.1% (a) +15.8% -8.0% -42.2% (impossible-- possible) Group III National Health Campaign Civil Liberties versus Insurance versus Contribution Homeland Security Tax Cuts Limits Possible 45.9% 47.2% 50.0% Difficult 31.0% 27.9% 23.3% Impossible 19.8% 22.7% 21.5% No response 3.3% 2.2% 5.2% n 654 614 607 Difference -26.1% -24.5% -28.5% (impossible-- possible) Note: These data are the result of asking this survey question for each of these policy areas: "Some issues are subject to the normal give-and- take of political compromise, while for other issues, compromise is difficult if not impossible. For you personally, is this an issue where compromise is: [response]?" (a) The difference between the possible and impossible response percentages is statistically distinct at the 0.05 level for each pair of policies in different groups. The within-group differences of the difference of percentages are statistically distinct for Groups I and II but not for Group III. Table 6. Likelihood that a Respondent's Own Opinion on the Policy Could Change Group III Group II Civil Liberties Group I Capital Casino versus Homeland Gay Marriage Punishment Gambling Security Has opinion Not likely to 73.6% (a) 63.2% 53.8% 47.3% change Somewhat 14.1% 29.0% 32.9% 38.7% likely Very likely 3.5% 2.1% 3.1% 4.4% No response 0.9% 0.9% 0.9% 0.7% Has no opinion (so was not asked "opinion change" question) No opinion 8.0% 4.8% 9.4% 9.0% n 689 701 685 693 Note: For each policy area, these data are the result of asking respondents the substantive issue question (see Appendix) and the follow-up question: "How likely is it that your views on this issue could change?" (a) Using difference-of-proportions tests, we find that the differences in the "not likely" percentages between the adjacent policy areas are statistically significant at the 0.05 level. Table 7. The Importance of Each Issue to a Respondent Group III Group II Civil Liberties Group I Capital Casino versus Homeland Gay Marriage Punishment Gambling Security Very important 23.8% (a) 29.7% 14.7% 55.5% Important 14.5% 28.5% 8.5% 24.3% Somewhat 26.1% 30.7% 26.9% 14.8% important Not important 32.2% 8.3% 45.2% 2.9% No response 3.4% 2.8% 4.7% 2.5% n 693 706 690 694 Note: These data are the result of asking this survey question for each of these policy areas: "How important would you say this issue is to you personally?" (a) If these policies are ordered from most important (civil liberties versus homeland security) to least important (casino gambling), we find that the difference in "importance" response proportions between every adjacent policy area are statistically significant at the 0.05 level. However, this is not the hypothesized pattern, and, thus, our hypothesis is not supported.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Does Morality Policy Exist? Testing a Basic Assumption. Contributors: Mooney, Christopher Z. - Author, Schuldt, Richard G. - Author. Journal title: Policy Studies Journal. Volume: 36. Issue: 2 Publication date: May 2008. Page number: 199+. © 1999 Policy Studies Organization. COPYRIGHT 2008 Gale Group.