پوشش رسانه ای و دید موضوع: پاسخ قانونی دولت به زورگویی مدرسه ای

Abstract Dealing with the schoolyard bully is an age-old problem; however, legislators have only recently tackled it at the state level. This study examines the adoption of anti-bullying policies from the policy diffusion and innovation perspective with an emphasis on the role of print media coverage. The study contributes to the policy diffusion literature by examining both national and local media coverage as conduits for diffusion and adds to the expanding views of the diffusion process beyond the traditional geographic proximity argument. Further, it provides the first examination of an emerging policy area important to education policy scholars. The findings show issue saliency via national media coverage drives policy adoption beyond any geographic proximity.

1. Introduction In March 2011, President Obama joked about his “own big ears and funny name” at a White House summit to address growing concerns about bullying and to launch a six-department federal initiative to deal with the issue (Calmes, 2011). This federal response follows 49 states that adopted anti-bullying policies during the previous decade. School officials and lawmakers have a moral and concurrent legal obligation to protect children in their schools and go to great lengths to protect children from potential dangers that exist in schools. A school culture must provide for the security and safety of its students and employees or run the risk of shortchanging their students by failing to provide a fair and equitable opportunity to learn and experience academic growth. In the aftermath of incidents such as Columbine and other acts of school violence, school officials and lawmakers focused more on methods to protect children through various policy and legislative initiatives (Birkland & Lawrence, 2009). As such, one important step has been to pass state statutes designed to reduce the prevalent problem of bullying in schools, to punish perpetrators, and offer protections to victims of bullying. While education scholars have investigated the role of bullying interventions in schools, there is much less knowledge the policy and political aspects of this issue. This paper examines the states’ responses to school safety through the enactment of anti-bullying policies. The purpose of this paper is to address an on-going debate in the policy diffusion literature, specifically challenging the traditional view of a geographic relationship for policy diffusion in the case where media attention raises the saliency of the policy. This study demonstrates the pivotal role media can play in the spread of policy across the country. Specifically, the national media can have a significant role in the diffusion process in emerging policy areas that gain a lot of nonpartisan attention in a relatively short time span. Bullying makes an ideal test of this theory, especially at the state level. It is also worth noting what this paper does not argue or test. First, the success or failure of these policies is not examined; it is the determinants of a state passing legislation in this area. In fact, passing bullying legislation by the states may be nothing more than an exercise in symbolic politics. However, this does not mean it is a costless exercise or an unimportant one, regardless of the impact of the law. Passing a policy establishes it on the legislature's agenda and provides a starting point for attention to the issue. Second, while the paper examines anti-bullying policy adoption, the real purpose is the interesting test case this makes for understanding the role of media coverage in a policy's diffusion across the country; it is not about the effects of these policies. Specifically, this study looks at the public policy aspects of this process by analyzing the influences on the adoption and diffusion of the policies across the country. First, the background on the bullying issue across the country is provided followed by a theoretical framework with a focus on the role of issue saliency via print media coverage and issue complexity within the innovation and diffusion literature. Next is the analysis with an overview of the data and the event history analysis (EHA) models used before turning to the results and discussion. The findings indicate national print media coverage raises the saliency of the issue and is an important determinant of policy adoption while traditional geographic variables do not explain the diffusion of the policy adoption across the country. The study concludes with a discussion of the contribution of these findings for studies of policy adoption and diffusion and for future research into bullying policy. The main contribution of this study is a more complete understanding of the role of media coverage in policy adoption and diffusion. Building from previous studies, the findings show non-geographic factors, such as issue visibility vis-à-vis media coverage, can trump the traditional arguments of geographic proximity when a policy is newly emerging on the political scene without a clear partisan or ideological division. This forces states to consider the growing scope of the problem, specifically when national coverage becomes a dominant factor in the issue's salience.

5. Results Table 1 presents two models with the second including the proxies for bullying related outcomes. The results show strong support for the importance of issue salience, but no apparent role for geographic diffusion. In both models, national media coverage has a positive and significant influence on the likelihood of adopting a bullying policy in that year. There is not a significant influence for state level coverage and given Fig. 2, this is not surprising as national coverage routinely outpaced state coverage. As the issue became more salient nationally, via print media coverage, the more likely a state was to adopt the policy. It appears that as national media coverage of events linked the effects of bullying to ongoing negative events, the issue moved from a latent policy to one that quickly emerged on the political agenda. Additionally, neither of the bullying proxies in the full model are statistically significant, suggesting the media coverage itself increased the visibility of the issue and was not a result of increases in behaviors or outcomes associated with bullying. Table 1. Determinants of state adoption of anti-bullying policies. Base model Full model Media coverage State media stories (t − 1) −0.0018 (0.0068) −0.0007 (0.0075) National media stories (t − 1) 0.0308* (0.0127) 0.0303* (0.0135) Geographic diffusion % Neighbor adoption (t − 1) 0.4859 (0.8245) 0.9990 (0.9131) % Regional adoption (t − 1) 0.3310 (1.0143) −0.2020 (1.0324) Political variables Democratic controla 1.5290** (0.5130) 1.2177* (0.5587) Divided controla 0.4745 (0.4337) 0.2921 (0.5117) Legislative professionalism −8.8530** (3.0476) −10.5082*** (2.8439) % liberal population 1.2551 (3.2070) 0.0060 (3.2822) Education variables Student enrollment (log) 0.5886** (0.2108) 0.5448* (0.2668) Per-pupil spending ($1,000s) 0.3279** (0.1135) 0.3761** (0.1306) Graduation rate 0.0038 (0.0258) 0.0227 (0.0361) Student/teacher ratio 0.1983* (0.0905) 0.1565 (0.1040) Bullying proxies % of students in fight at school 0.0979 (0.1208) State youth suicide rate (t − 1) −0.0732 (0.1101) Time trend −0.0539 (0.2771) 0.0185 (0.2953) Constant −18.0257*** (4.4629) −19.0272** (6.1436) Observations 390 362 Wald chi2 82.90*** 87.72*** Log pseudolikelihood −118.10 −105.91 Notes: Logit estimates with robust standard errors in parentheses, clustered by state. Dependent variable is state-year coded as 0 for no bullying policy adoption and 1 for adoption of bullying policy during the year. Each state drops out of the data set once they enacted an initial policy. Years range from 1998 to 2012. a Republican control of state is the base category * p < .05. ** p < .01. *** p < .001. Table options There is no support for the diffusion hypotheses as neither the neighbor nor regional adoption variables reaches statistical significance in Table 1. This suggests that traditional mechanisms of communicating policy innovation are not at work with the adoption of general anti-bullying policies. Most likely this is due to the salience of the issue as captured by the media coverage. Lawmakers did not have to rely on learning about the adoption of neighboring states’ anti-bullying policies as the issue began gaining traction in both national and local media outlets across the country during the previous decade. States with more students, larger student-teacher ratios, and states that spend more per-pupil are more likely to adopt the policy. These results suggest the educational environment is also an important part of the process as states with bigger needs, more students, fewer teachers per student, and greater resources are more likely to adopt the policy. From a political perspective, more liberal states are not any more likely to adopt the policy. However, Democratically controlled states are significantly more likely to adopt the policy, and states with more professionalized legislatures are less likely to adopt the policy. While this result does not change the underlying findings and a full exploration is beyond the scope of the paper, it is worth noting Bernick and Myers’ (2012) argument that politicians revert to their parties’ traditional policy stances when issues become highly salient. The negative finding for professionalization fits with other research that finds more professionalized state legislatures may be less responsive to new policy areas, although this is an area without a clear consensus. Specifically, Herrick (2008) finds similar results in terms of partisan support and professionalization when examining emerging policies designed to protect gay, lesbian, bisexual, and transgender rights. To get a sense of the substantive influence of the findings, Fig. 3 shows the predicted probabilities from model 1 for national news coverage while holding all other variables at their appropriate means or modal categories. The dashed lines represent the 95% confidence interval. When media coverage is low, there is an extremely low and statistically insignificant likelihood of policy adoption. The likelihood begins to increase and becomes statistically significant as the number of national stories in a year reach 50. At that point, the influence for media coverage becomes statistically significant while the predicted probability remains low at just over 5%. As the number of stories increase to 100 stories in a year, the likelihood of adoption increases to 15% and continues to increase by about 6% for every additional 10 stories. Predicted probabilities for anti-bullying adoption by national print media ... Fig. 3. Predicted probabilities for anti-bullying adoption by national print media coverage. Sources: Results calculated from results in model 1. Figure options Overall, Fig. 3 highlights two important trends in the likelihood of adopting a bullying policy. First, print media coverage matters once it reaches a critical mass. As the coverage increased in the latter part of the decade, states that had yet to adopt had to pay attention. The important role of national coverage over state coverage is not surprising given the nature of the issue. Bullying is not constrained to one part of the country as kids harass one another regardless of where they live, and several high profile incidents pushed the everyday concern to the forefront. As the issue became more prominent in the national dialog, states responded. Lawmakers could not afford to ignore the issue by assuming it was not an issue in their own state.