Public Spending on the Arts as Morality Policy: The Structure of Public Attitudes
Lewis, Gregory B., Policy Studies Journal
Successfully reframing a political issue as morality policy should strengthen the hand of those charging immorality. Changing issue frames can shift public opinion by changing what aspects of a problem take priority in public thinking (e.g., Jacoby 2000; Nelson, Clawson, and Oxley 1997). By emphasizing moral judgments, at which everyone is an expert, a morality policy frame tends to make issues easier and more salient (Carmines and Stimson 1980; Gormley 1986). Policy entrepreneurs have some freedom to recast issues as morality policy--the issues that attract moral reasoning vary over time and across societies (Mooney 2001)--especially if moral concerns already weakly align with divisions over the issue. Issue framers face a difficult challenge, however. They must be credible to their audience, the frame must be strong and speak to the audience's values, and the framing effort may need to be persistent (Chong and Druckman 2005). "Easy" issues have usually been on the political agenda for a long time (Carmines and Stimson 1980, 80).
One effort to redefine a minor spending issue as morality policy occurred in 1989, when Christian conservatives attempted to reframe the case against federal funding for the arts from waste to immorality, arguing that the National Endowment for Arts (NEA) subsidized antireligious, homoerotic, and pornographic art that, in the words of Senator Jesse Helms, "denigrated, offended, or mocked [moral, decent Americans] with their own tax dollars" (Hetherman 1999). For two years, the NEA attracted unprecedented media coverage, emphasizing battles between decency and immorality, free expression and censorship. To shift public opinion, however, the reframers needed to convince average Americans that arts spending raised sufficiently weighty moral issues to warrant attention it had never received before. Opposition leaders had limited credibility outside their base, however, and arts spending had never been a salient political issue (DiMaggio and Pettit 1999).
Using General Social Survey (GSS) data from 1985, 1990, and 1996--before, during, and after the NEA controversy--I examined whether the debate increased the importance of the decency immorality- and free expression-censorship dimensions in attitudes toward public spending on the arts. The analysis suggests that the reframing failed. Although those dimensions were related to support for arts spending in all three years, suggesting a preexisting cultural divide on the arts, NEA opponents appeared unable to exacerbate that division.
Reframing the Debate over Arts Funding
During its first quarter-century, the NEA attracted limited opposition, primarily from some Republicans and conservative Democrats who argued that the arts were an excessively costly luxury (Moen 1997). NEA spending rose three times faster than the rest of the federal budget during its first 15 years. In 1980, President Reagan achieved a 10% cut in the NEA budget, but based on objections to "the general principle of government support for the arts" rather than on the substance of its subsidies (Cummings 1991, 57).
In 1989, however, two grants--$15,000 to Andres Serrano for Piss Christ, a photograph of a crucifix immersed in urine, and $30,000 for The Perfect Moment, a traveling retrospective of photographer Robert Mapplethorpe's work, much of it explicitly homoerotic and sadomasochistic--turned the NEA into a battleground in the "culture wars." Christian conservatives led a campaign charging the NEA with giving federal imprimatur to antireligious bigotry, obscenity, and homosexuality. The Rev. Donald Wildmon of the American Family Association initiated the campaign against Piss Christ and the NEA, warning that "[t]he bias and bigotry against Christians, which has dominated television and movies for the last decade or more, has now moved over to the art museums" (Bolton 1992, 27). The Christian Coalition placed a full-page ad in the Washington Post warning Congress that taxpayers would punish them for "wasting their hard-earned money to promote sodomy, child pornography, and attacks on Jesus Christ [and] ... to teach their sons how to sodomize one another" (Bolton 1992, 316). Patrick Buchanan warned that "America's art and culture are, more and more, openly anti-Christian, anti-American, nihilistic," and called for conservatives to lead "a cultural revolution" (Bolton 1992, 32-3); in a 1992 campaign commercial, he blamed the Bush administration for funding "so-called art [that] has glorified homosexuality, exploited children and perverted the image of Jesus Christ" (Loth 1992). Jesse Helms, leader of the fight in the Senate, tied the NEA's activities to "[t]he homosexual 'community,' the feminists, the civil libertarians,... [and] their dangerous anti-family and anti-American agendas" (Bolton 1992, 306). In a letter to the NEA chair signed by over one hundred members of Congress, Rep. Dick Armey demanded an end to subsidies to "morally reprehensible trash" like the Mapplethorpe exhibit (Bolton 1992, 107).
NEA supporters typically defended artistic freedom rather than the work attacked and repeatedly emphasized that only a handful of the NEA's 80 thousand grants had generated controversy. Supporters framed the issue as freedom versus censorship, especially after the Corcoran Gallery canceled its exhibition of The Perfect Moment, and the director of a Cincinnati art museum was indicted on charges of pandering obscenity for showing the same exhibit. The arts community defended near-absolute freedom of expression, even for artists who received public funds (Bolton 1992, 63).
Data and Design
To successfully reframe arts spending as morality policy, opponents needed to get voters to think first of religious and moral values in assessing the NEA, rather than attitudes toward government taxing and spending, the traditional grounds for opposing arts spending. NEA supporters hoped to mobilize opponents of censorship and restrictive sexual morality. Fortuitously, the GSS asked whether respondents "would like to see more or less government spending [on] ... culture and the arts" in 1985, 1990, and 1996--before, during, and after the NEA controversy. (Details on variable coding are in the appendix.) Large minorities (41 to 46%) preferred to see less spending on culture and the arts, while much smaller minorities (14 to 16%) wanted more. However, support for arts spending did not decline significantly between 1985 and 1996.
DiMaggio and Pettit (1999) have already analyzed these data, and Brooks (2001) looked at a similar question on the 1998 GSS. Both looked at how religion and attitudes toward government affected support for arts spending, but neither explicitly examined the salience of sexual morality and free expression, the axes on which the new debate turned. Like these authors, I examined the impact of political ideology, religion, and taste for the arts. I expected conservatives, Republicans, and opponents of an active role for government in the economy to be most likely to favor cutting spending on the arts (Brooks 2001; DiMaggio and Pettit 1999). Fundamentalist Protestants and perhaps Catholics, because the crucifix in Piss Christ offended Catholics most, should be more likely to favor cuts in arts spending, though DiMaggio and Pettit found that Catholics differed from nonfundamentalist Protestants only at the height of the debate. Unlike these authors, I focused on whether the salience of religion, sexual morality, and free expression rose.
I included three measures of political ideology to capture the fiscal frame. Support for domestic spending was measured as the mean response to six questions on spending for the environment, health, the police and law enforcement, education, and retirement and unemployment benefits. Respondents who supported more spending on all six items scored 2 and those who favored less spending on all six items scored 0. Two dummy variables identified Democrats and Republicans, and a 7-point ideology scale ranged from "extremely conservative" to "extremely liberal."
To capture the salience of the morality politics frame, I included measures of religion, sexual morality, and commitment to free expression. Five dummy variables distinguished fundamentalist Protestants, Catholics, Jews, members of other religions, and the nonreligious from moderate and liberal Protestants (the reference group). A dummy variable coded 1 for those who attend religious services weekly added a measure of religious intensity. Two four-point scales measured whether "a married person having sexual relations with someone other than the marriage partner" and "sexual relations between two adults of the same sex" were "always wrong, almost always wrong, wrong only sometimes, or not wrong at all." For both, those who answered "always wrong" were coded 1 and those who answered "not wrong at all" were coded 4. The free expression measure combined 15 questions on whether "people whose ideas are considered bad or dangerous by other people"--atheists, Communists, racists, militarists, and homosexuals--should be allowed to teach college, give advocacy speeches in one's community, and have advocacy books in one's public library. The measure is the proportion of these civil liberties questions that each respondent answered positively.
As proxies for how highly respondents valued the arts, I added education, gender, residence, and marital status. Arts attendance rises with education and is higher for women, single and divorced people, and city-dwellers than for men, married people, and those in more rural areas (Heilbrun and Gray 2001; Throsby 1994).
I ran ordered-logit models on whether respondents thought government should spend less, the same as now, or more on culture and the arts for each year separately (see the first three columns of Table 1). I then combined data for all years and added dummy variables for survey years to determine whether support declined among comparable individuals (column 4). Finally, I added interactions between these year dummies and each of the independent variables (not shown). This model replicates all relationships found in the three years separately. The coefficient on each 1996 interaction term, for instance, showed the difference between the coefficients on that variable in 1996 and 1985, allowing tests of whether coefficients changed significantly.
First, the fiscal frame is the strongest determinant of support for arts spending in every year. Support for government domestic spending has the highest standardized odds-ratio (Long and Freese 2001) in all models, suggesting that a one-standard-deviation increase in support for domestic spending increases the odds of favoring increased arts spending by about 60 percent. Though the coefficients on party identification and ideology are statistically significant only in the combined sample, they are reasonably stable across years. In the combined sample, holding all other variables at their means, extreme conservatives are estimated to be 14 percentage points more likely than extreme liberals to favor cuts in arts spending, and Democrats were about 7 percentage points less likely than independents and Republicans to favor cuts.
Second, "culture war" attitudes toward sexual morality and free expression also affect support for arts spending in all years. In 1985, before the NEA controversy, the coefficients on support for civil liberties and disapproval of homosexuality were already statistically significant and that on disapproval of extramarital sex was significant in a one-tailed test. Holding the other variables at their means, giving all pro-free expression answers rather than none and calling homosexual or extramarital sex "not wrong at all" rather than "always wrong" are each expected to increase the probability of favoring more spending on the arts by 6 to 9 percentage points.
Third, once these attitudes are controlled, religion has little impact, though predictors of arts attendance remain predictors of support for arts spending. Christian conservatives led the opposition to the NEA, but fundamentalist Protestants were indistinguishable from moderate and liberal Protestants with similar attitudes and demographics. Catholics were more likely than comparable nonfundamentalist Protestants to favor cuts in 1990. Jewish respondents were the most likely to favor arts spending, with the effect largest in 1990. Those who attended religious services weekly and would thus seem most willing to incorporate moral judgments into their thinking about the arts were more likely than similar others to favor arts spending. Women, the better educated, the unmarried, and city dwellers were more likely than comparable others both to attend the arts and to favor increased spending.
Fourth, and most importantly, the structure of attitudes does not change over time; in particular, the salience of religion, sexual morality and civil libertarianism does not rise. The model did the worst job of predicting support in 1990, at the height of the debate; the adjusted count [R.sup.2] and coefficients on domestic spending, homosexuality, and free expression were all at their lowest. However, we are unable to reject the null hypothesis that the effect of each variable remained constant across all three years. Although comparable respondents were about 6 percentage points more likely to favor cuts in 1990 and 1996 than in 1985 (column 4), only one of 34 coefficients on interaction terms was individually significant at the 0.05 level, less than one would expect by chance. Likelihood-ratio and Wald tests of joint significance could not reject the null that all variables had the same impact in each year. The tests were run three ways: for all interaction terms, for interaction terms for 1990 and 1996 separately, and for each pair of interaction terms on the same variable. The test for support for civil liberties came closest to finding change over the three years (p < 0.06), but the coefficients suggest that attitudes toward censorship mattered less during and after the controversy than before.
The attempt to reframe public arts spending as morality politics appears to have tapped into a preexisting cultural divide rather than to have opened a new one. Though the media debate was vitriolic, the conflict does not appear to have exacerbated the divide. Attitudes toward domestic spending remained the strongest predictor of arts attitudes in all years--the fiscal frame did not diminish in importance. Among those with similar attitudes toward domestic spending, support for arts spending was already divided along the morality-censorship dimension, and those who disapproved of homosexuality and were willing to limit free expression wanted to cut public arts support before Christian conservatives highlighted NEA grants for Piss Christ and Robert Mapplethorpe's homoerotic photography. As DiMaggio and Pettit (1999) argue, the debate had a negligible impact on public opinion, perhaps because the issue never achieved the salience elites perceived. Congressional Republicans did cut the NEA budget in half almost immediately after the 1994 "Republican Revolution," suggesting that the debate turned the Republican elite against public funding of the arts, though it had little impact on the public's perceptions of a spending item that had attracted little prior attention. Though preexisting divisions on arts spending that aligned with a morality politics frame gave the re-framing some chance of success, the formidable obstacles to shifting public thinking stymied the effort.
Gregory Lewis is professor of public administration and urban studies at Georgia State University and director of the joint Georgia State-Georgia Tech doctoral program in public policy. He has published widely on the career patterns and attitudes of public employees, on public support for lesbian and gay rights, and on morality policy more broadly.
A previous version of this paper was presented at the annual meeting of the American Political Science Association, Philadelphia, August 30, 2003. The author is grateful to Charles W. Gossett and to several anonymous reviewers for helpful suggestions and to the Interuniversity Consortium for Political and Social Research for providing the General Social Survey.
Bolton, Richard. 1992. Culture Wars: Documents from the Recent Controversies in the Arts. New York: New Press.
Brooks, Arthur C. 2001. "Who Opposes Government Arts Funding?" Public Choice 108: 355-67.
Carmines, Edward G., and James A. Stimson. 1980. "The Two Faces of Issue Voting." American Political Science Review 74: 78-91.
Chong, Dennis, and James N. Druckman. 2005. "Competitive Framing." Paper presented at annual meeting of American Political Science Association, Washington, DC, September 3, 2005.
Cummings, Milton C., Jr. 1991. "Government and the Arts: An Overview." In Public Money and the Muse: Essays on Government Funding for the Arts, ed. Stephen Benedict. New York: W.W. Norton.
Davis, James Allan, Tom W. Smith, and Peter V. Marsden. 2003. General Social Surveys, 1972-2002: Cumulative Codebook. Chicago: National Opinion Research Center.
DiMaggio, Paul, and Becky Pettit. 1999. Public Opinion and Political Vulnerability: Why Has the National Endowment for the Arts Been Such an Attractive Target? Working Paper 7, Center for Arts and Cultural Policy Studies, Princeton University, NJ.
Gormley, William T. 1986. "Regulatory Issue Networks in a Federal System." Polity 18: 595-620.
Heilbrun, James, and Charles M. Gray. 2001. The Economics of Art and Culture, 2nd ed. Cambridge: Cambridge University Press.
Hetherman, Luke. 1999. The National Endowment for the Arts: Essential for Cultural Growth, Democracy and Creative Expression in America [Online]. http://www.burbank.com/arts/nea.htm. Accessed September 22, 2005.
Jacoby, William G. 2000. "Issue Framing and Public Opinion on Government Spending." American Journal of Political Science 44: 750-67.
Long, J. Scott, and Jeremy Freese, 2001. Regression Models for Categorical Dependent Variables Using Stata. College Station TX: Stata Press.
Loth, Renee. 1992. "Buchanan says Bush Funded Obscene Art; Ad Uses Clip from Film on Gay Blacks." Boston Globe (February 27): 14.
Moen, Matthew C. 1997. "Congress and the National Endowment for the Arts: Institutional Patterns and Arts Funding, 1965-1994." Social Science Journal 34: 185-200.
Mooney, Christopher Z., ed. 2001. The Public Clash of Private Values: The Politics of Morality Policy. New York: Chatham House Publishers.
Nelson, Thomas E., Rosalee A. Clawson, and Zoe M. Oxley. 1997. "Media Framing of a Civil Liberties Conflict and its Effect on Tolerance." American Political Science Review 91: 567-84.
Throsby, David. 1994. "The Production and Consumption of the Arts: A View of Cultural Economics." Journal of Economic Literature 32 (1): 1-29.
Wooldridge, Jeffrey M. 2003. Introductory Econometrics: A Modern Approach, 2nd ed. Mason, OH: Thomson South-Western.
Exact question wording for all questions is on the Interuniversity Consortium for Political and Social Research's (ICPSR's) General Social Survey web page (http://webapp.icpsr.umich.edu/GSS/) and in Davis, Smith, and Marsden (2003). The dependent variable, SPARTS, was recoded from five values ("spend much more" to "spend much less") to three (combining "much more" with "more" and "much less" with "less"), because the simpler variable meets the proportionality of odds assumption for ordered logit but the five-value variable does not.
The domestic spending variable is the mean of SPENVIRO, SPHLTH, SPPOLICE, SPSCHOOL, SPRETIRE, and SPUNEMP. All six items load at least 0.4 on a single factor, and Cronbach's alpha is 0.73, indicating acceptable reliability. Dropping any item lowers alpha. Support for spending on the military and defense (SPARMS) was essentially uncorrelated with this variable. POLVIEWS is reversecoded to become the level of liberalism. The Democrat and Republican dummy variables are created from PARTYID and do not include independents who lean toward those parties. The religion dummies are recodes of RELIG, FUND, and ATTEND: 34% are fundamentalists, 24% are Catholics, and 30% attend religious services weekly. The sexual morality variables are XMARSEX and HOMOSEX: 78% said extramarital relations were always wrong and only 2% said they were not wrong at all, with no clear trend over time; 68% also said homosexual relations were always wrong, but the percentage saying "not wrong at all" was ten times as high (20%) and doubled from 15% in 1985 to 29% in 1996. The support for free expression variable is created from the variables with prefixes SPK, COL, and LIB and suffixes ATH, RAC, COM, MIL, and HOMO. All 15 items load at least 0.57 on a single factor, and Cronbach's alpha is 0.91. The mean proportion rose from 0.58 in 1985 to 0.66 in 1996.
Ordered logit analysis typically assumes an unobserved, interval-level dependent variable (in this case, support for public arts spending), represented imperfectly by an ordinal-level variable (Wooldridge 2003). Coefficients represent the change in the log-odds of being above any threshold or cut-point (favoring at least keeping spending at current levels as opposed to cutting it, and favoring spending more as opposed to less or about the same) from a one-unit increase in the independent variable. All models presented pass likelihood-ratio and Wald tests of the proportional odds assumption, meaning that it is reasonable that each variable has the same impact at each threshold. Exponentiating the coefficients converts them to odds ratios. The Stata prchange procedure allows converting logit coefficients to percentage changes in the dependent variable, setting the independent variables to set values (their means in this article). To assess the relative strength of relationships, I used standardized odds-ratios, which show the proportional change in the odds of taking a more positive stance on arts spending from a one standard deviation increase in the independent variable (Long and Freese 2001, 132-36).
Table 1. Ordered Logit Model for Support for Arts Spending 1985 1990 1996 Combined Support for domestic 1.428** 1.038** 1.538** 1.384** spending (5.20) (3.54) (5.85) (8.87) Level of liberalism 0.129 0.120 0.067 0.095* (1.81) (1.84) (1.08) (2.56) Democrat 0.393 0.051 0.390* 0.271* (1.89) (0.26) (2.12) (2.42) Republican 0.232 -0.160 -0.230 -0.044 (1.06) (0.80) (1.13) (0.37) Acceptability of 0.228* 0.008 0.149* 0.124** homosexual sex (2.54) (0.10) (2.25) (2.80) Acceptability of 0.203 0.313* 0.314** 0.280** extramarital sex (1.71) (2.22) (2.63) (3.93) Support for free 1.135** 0.474 0.122 0.567** expression (3.75) (1.65) (0.41) (3.37) Fundamentalist Protestant 0.024 -0.150 -0.088 -0.082 (0.11) (0.76) (0.42) (0.69) Catholic 0.149 -0.423* -0.062 -0.099 (0.68) (2.01) (0.28) (0.81) Jewish 0.772 1.609** 0.358 0.830* (1.17) (2.93) (0.67) (2.52) Other religion 0.420 -0.769 0.404 0.141 (0.65) (1.61) (1.11) (0.54) No religion 0.014 0.202 -0.053 0.047 (0.04) (0.62) (0.20) (0.27) Attends religious services 0.379* 0.319 -0.005 0.214* weekly (1.98) (1.81) (0.03) (2.04) Years of education 0.002 0.080** 0.082** 0.055** (0.05) (2.59) (2.63) (3.09) Female 0.200 0.210 0.132 0.185* (1.13) (1.28) (0.84) (1.97) Single or divorced 0.103 0.160 0.405* 0.250* (0.54) (0.92) (2.53) (2.53) City dweller 0.537** 0.206 0.282 0.348** (2.96) (1.19) (1.75) (3.57) Surveyed in 1990 -0.255* (2.21) Surveyed in 1996 -0.273* (2.36) Adj. Count [R.sup.2] 0.195 0.137 0.163 0.189 Observations 575 645 699 1,919 Absolute value of z statistics are in parentheses. *significant at 5 percent; **significant at 1 percent.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Public Spending on the Arts as Morality Policy: The Structure of Public Attitudes. Contributors: Lewis, Gregory B. - Author. Journal title: Policy Studies Journal. Volume: 34. Issue: 1 Publication date: February 2006. Page number: 131+. © 1999 Policy Studies Organization. COPYRIGHT 2006 Gale Group.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.