Full opinion text
FINDINGS OF FACT AND CONCLUSIONS OF LAW KENNELLY, District Judge. Plaintiffs Entertainment Software Association, Video Software Dealers Association, and Illinois Retail Merchants Association sued several state and local officials seeking to enjoin the enforcement of Illinois’s Violent Video Games Law and Sexually Explicit Video Games Law. Plaintiffs moved for a preliminary injunction, and the Court ordered a hearing. All parties consented to combine the preliminary injunction hearing with trial on the merits. This constitutes the Court’s findings of fact and conclusions of law under the Federal Rules of Civil Procedure 52(a). For the reasons stated below, the Court concludes that both statutes violate the First Amendment and therefore issues a permanent injunction against their implementation. Findings of Fact The plaintiffs are associations of entities that create, publish, distribute, sell, and rent video games. The Entertainment Software Association (ESA) represents video game creators and publishers; the Video Software Dealers Association (VSDA) represents home video retailers; and the Illinois Retail Merchants Association (IRMA) represents Illinois merchants. The ESA is comprised of publishers of interactive entertainment, including video games, computer games, and mobile games. Lowenstein Dep. at 18. Its mission is “to serve the business and public affairs interests of the entertainment software industry.” Id. at 25. To achieve this mission, ESA implements a number of programs, such as government relations, anti-piracy enforcement, intellectual property policy, and media relations programs. Id. at 26. Among its many activities, ESA also evaluates proposed legislation regulating the entertainment software industry and responds to those proposals based on the mandate given by the membership. Members are updated on the ESA’s positions on particular legislation, including state statutes regulating video games based on content, contemporaneously. Id. at 38. The VSDA is a trade association for the home video industry that includes retailers and distributors of home video products, including video games. Andersen Dep. at 7. The association’s many activities include tracking proposed regulations of video games, responding to those regulations, and pursuing litigation if necessary. Decisions to engage in litigation are made by the VSDA board of directors. Although members do not vote on the VSDA’s stance on all relevant legislation or on the association’s decision to initiate litigation, they can weigh in on VSDA’s actions in a numbers of ways, including telephone calls, emails, and letters. Id. at 30-43. IRMA has approximately one thousand retail members, though it is unclear how many sell or rent video games. Vite Dep. at 26. IRMA’s activities include examining pending legislation in Illinois, developing position papers, and communicating the association’s position to the General Assembly. IRMA communicates with its members about these activities through its website and a weekly newsletter published while the General Assembly is in session. Vite Dep. at 8-10. With regard to the legislation at issue in this case, IRMA discussed the proposed legislation with between fifty and one hundred members before it was passed. Id. at 61. IRMA staff typically determines whether the organization should participate in litigation after discussing the issue with affected members. Id. at 40-41. In deciding to join this case, IRMA president David Vite consulted with several leaders of the association and also consulted members directly affected by the legislation. Id. at 43, 55. The defendants are Illinois public officials: Rod Blagojevich is the Governor of Illinois; Lisa Madigan is the Attorney General of Illinois; and Richard Devine is the State’s Attorney for Cook County, which includes Chicago. Video games are one of the newest and most popular forms of artistic expression. They most resemble films and television shows by telling stories through pictures, text, and sound, but they also parallel popular books, such as the Choose Your Own Adventure series, which enable readers to make decisions about how the plot and characters will develop. Video games are generally designed to entertain players and viewers, but they can also inform and advocate viewpoints. They are therefore considered protected expression under the First Amendment. See Am. Amusement Machine Ass’n v. Kendrick, 244 F.3d 572, 579 (7th Cir.2001). The video game industry has adopted a voluntary rating system to advise consumers about the content of video games. The Entertainment Software Rating Board (ESRB), a division of the ESA, assigns ratings to games and provides content descriptors. The rating system is voluntary. Most video game publishers, however, submit video games to the ESRB for review, and most video game retailers refuse to sell or rent games that have not been rated by the ESRB. Lowenstein Dep. at 72-73,103-04. Illinois Public Act 94-0315, which was adopted in July 2002, expands the state’s regulation of the sale and rental of video games. The Act amends 720 ILCS 5/11-21, which criminalizes the distribution to an individual under the age of eighteen of material that is “harmful to minors.” It also creates two new criminal statutes: the Violent Video Games Law (WGL) and the Sexually Explicit Video Games Law (SEVGL). Plaintiffs challenge the WGL and the SEVGL, but they do not challenge the amendments to 720 ILCS 5/11-21. The WGL establishes criminal penalties for selling or renting violent video games to minors, allowing such games to be purchased using a self-checkout electronic scanner, and failing to label such games with a two inch by two inch label stating “18”. 720 ILCS 5/12A-25. The statute defines “violent video games” as those that include: depictions of or simulations of human-on-human violence in which the player kills or otherwise causes serious physical harm to another human. “Serious physical harm” includes depictions of death, dismemberment, amputation, decapitation, maiming, disfigurement, mutilation of body parts, or rape. Id. 12A-10(e). Violations of the statute are punishable by fines ranging from five hundred to one thousand dollars. Id. 12A-15(a)-(c) & 12A-25(b). The SEVGL creates criminal penalties for selling or renting sexually explicit video games to minors, allowing such games to be purchased using a self-checkout electronic scanner, and failing to label such games with a two inch by two inch label stating “18”. 720 ILCS 5/12B-15; 5/12B-25. The SEVGL has two additional requirements. Video game retailers must post eighteen by twenty-four inch signs within five feet of every video game display or point of sale or rental informing customers about the ESRB rating system, and they must make brochures about the ESRB rating system available to customers. Id. 12B-30 & 12B-35. The statute defines “sexually explicit video games” as including:. those that the average person, applying contemporary community standards would find, with respect to minors, is designed to appeal or pander to the prurient interest and depict or represent in a manner patently offensive with respect to minors, an actual or simulated sexual act or sexual contact, an actual or simulated normal or perverted sexual act or a lewd exhibition of the genitals or post-pubescent female breast. Id. 12B-10(e). In adopting the WGL and the SEVGL, the Illinois General Assembly made findings about the accessibility of violent and sexually explicit video games to minors. Specifically, the legislative record includes reports by the Federal Trade Commission and the Illinois State Crime Commission about the ability of minors to purchase M-rated video games. In 2004, the FTC found that sixty-nine percent of unaccompanied teenagers were able to purchase M-rated video games. Def.’s Exh. D to 56.1(a) Statement at BL 169. In 2005, the Illinois State Crime Commission found that a fifteen-year old boy was able to buy M-rated games at eleven of fifteen, or seventy-three percent, of retailers visited. Id. at BL 266-69. The 2004 FTC study also examined whether unaccompanied teenagers could purchase analogous media products in other formats. The FTC concluded that eighty-one percent of unaccompanied minors could purchase R-rated movies, and eighty-three percent could purchase music with explicit lyrics — far more than were able to purchase M-rated video games. Id. at 152,160. With regard to the WGL, the legislative record includes scholarly articles, news articles, written testimony, and transcripts of House debates and House Civil Judiciary Committee Hearings discussing the impact of violent video games on minors. See generally Ex. D. to Def.’s Rule 56.1(a) Statement. Among other things, the record includes seventeen scholarly articles contending that minors become more aggressive when exposed to media violence, including video game violence. See generally, id. at BL 00075-00090, 000096-00131, 00328-00633. The same person — Dr. Craig Anderson — authored or co-authored fourteen of these articles; Dr. Douglas Gentile, a colleague of Dr. Anderson’s, authored one article; and Dr. William Kro-nenberger, who has relied on Dr. Anderson’s research in developing his own studies, authored the other two articles. See id. With regard to the SEVGL, the Illinois General Assembly found “sexually explicit video games inappropriate for minors.” 720 ILCS 5/12B-5. The legislative record does not include scholarly articles or expert testimony on this issue, but there are comments from legislators contending that the sexually explicit content in many video games is inappropriate for children. See e.g., Ex. D. to Def.’s Rule 56.1(a) Statement at BL 00006, 00066 (Rep. LaVia, discussing video games with “vivid pictures of nudity,” images of characters “defecating on people,” and rewards for “sleeping with prostitutes”). On July 25, 2005, the Act was signed into law. On the same day, plaintiffs filed this lawsuit. Specifically, plaintiffs sought declaratory and injunctive relief against enforcement of the WGL and SEVGL on the ground that the statutes violate their constitutional right to free expression. The plaintiffs also moved for a preliminary injunction to prevent the statutes from going into effect on January 1, 2006. Defendants responded by filing motions to dismiss for lack of standing, and, in the case of defendants Madigan and Devine, based on their purported immunity from suit under the Eleventh Amendment. Defendant Blagojevieh also filed a partial motion for summary judgment on the claims regarding the SEVGL, in which defendant Madigan subsequently joined. On November 14 and 15, this Court held an evidentiary hearing on the effect of violent video games on youth. With the agreement of the parties, this Court consolidated the hearing on the preliminary injunction with trial on the merits pursuant to Rule 65(a)(2) of the Federal Rules of Civil Procedure. The evidence presented via live witnesses concerned two main issues: first, whether minors who play violent video games experience an increase in aggressive thoughts, aggressive affect, and aggressive behavior, and second, whether minors who play such games experience a decline in brain activity in the region of the brain that controls behavior. The parties also agreed that the Court could consider as evidence the affidavits that had been submitted as part of the preliminary injunction briefing, as well as deposition testimony of certain witnesses. 1. Effect of violent video games on aggressive thoughts and behavior Dr. Craig Anderson, a psychologist and professor at Iowa State University, testified on behalf of the defendants. Dr. Anderson summarized research, including his own, regarding the relationship between minors’ exposure to violent video games and aggressive thoughts and behavior. Based on this research, Dr. Anderson testified that “it seems clear that exposure to violent video games increases aggressive behavior, aggressive thinking, physiological arousal, aggressive feelings, and is also associated with a decrease in prosocial behavior.” Tr. 212-13. Dr. Anderson’s studies on the connection between media violence and aggressive cognition and behavior are rooted in his broader research into aggression. Specifically, Dr. Anderson developed his “general aggression model” to explain how an individual’s personal characteristics and experiences trigger aggressive thoughts or responses to particular situations and how this cycle or cycles can make aggressive thoughts and behaviors more accessible. According to Dr. Anderson, an individual’s personal characteristics and experiences— particularly with violence — prime or activate aggressive thoughts and teach aggressive “scripts,” making it more likely that he or she will have aggressive reactions to particular situations. Tr. 212-21. Dr. Anderson testified that playing violent video games is one activity that primes aggressive thoughts and teaches aggressive scripts. He stated: In violent video games, you rehearse really the whole sequence [of aggression], You rehearse, you practice being vigilant, that is, looking for the source of the threat. You practice identifying sources of threat. You practice making decisions about how to respond to that threat. And eventually, you actually carry out some form of action, typically a violent action to deal with that threat, clicking a mouse or something on the keyboard or a pretend sort of gun of some kind. Tr. 227-29. As a result of regularly playing violent video games, Dr. Anderson testified, these scripts or knowledge structures become “chronically accessible” and ultimately become “automatized.” Tr. 229-30. The research underlying Dr. Anderson’s testimony, however, does not support such a stark and sweeping conclusion. We begin by providing an overview of the studies cited by Dr. Anderson to support his conclusions. One of the articles Dr. Anderson published involved two studies of the effects of violent video games on measures of aggression on college students. This study provided the basis for the Indianapolis ordinance regulating violent video games that was struck down in Kendrick. Kendrick, 244 F.3d at 578. In the first study, Dr. Anderson and Dr. Karen Dill of Lenoir-Rhyne College conducted a survey of college students regarding exposure to violent video games and aggressive behavior, using components of the National Youth Survey as a measure. Dr. Anderson found “a strong positive correlation between video game exposure and aggressive behavior.” Tr. 251. He conceded, however, that once the results were adjusted to exclude non-serious behavior, such as throwing snowballs, less than ten percent of the participants reported engaging in aggressive behavior. Tr. 298-302. Dr. Anderson also indicated that exposure to violent video games only incrementally affected the amount of aggressive behavior they engaged in. Tr. 278-83. In a second study, Dr. Anderson conducted an experiment in which participants played a violent video game, Wolfenstein 3D, or a non-violent video game, Myst. The participants were then asked to participate in a task in which they were supposed to “compete” with someone outside the room and administer a noise blast to the “loser” whenever they won. Based on this experiment, Dr. Anderson concluded that violent video games caused an increase in aggressive behavior, because participants who played Myst administered a longer noise blast than the participants who played Wolfenstein 3D. In his testimony, however, Dr. Anderson stated that the difference between the two groups was a matter of milliseconds. Tr. 329. Dr. Anderson also discussed three additional studies he conducted using college students. In the first experiment, half the students played violent video games and half played non-violent video games. They were then given a list of partially completed words — many of which could have been completed to form an aggressive word— and asked the participants to finish the words. The students who played the violent video game were more likely to fill in the blank to form an aggressive word, leading Dr. Anderson and his colleagues to conclude that exposure to the violent video game had increased their aggressive thoughts. Tr. 235-38. In the second experiment, half the students played a violent video game and half played a non-violent video game. They then went through time trials in which they were led to believe they were competing against someone in the next room. In the first series of time trials, the participants went through a series of time trials in which they were “punished” with a noise blast from a “competitor” when they lost. There were, in fact, no competitors: the noise blasts administered to the participants were controlled by a computer. Half received random blasts, and the other half received blasts that increased in intensity. In the second series of time trials, the participants administered a noise blast to their “competitor” if they won. Dr. Anderson found that the students who had played violent video game and received random noise blasts administered more intense noise blasts than all of the students who played the non-violent video games. He therefore concluded that exposure to violent video games increased an individual’s aggressive behavior. Notably, the students who played violent video games and received increasingly intense noise blasts administered the lowest intensity noise blasts. Though that seems to contradict his finding, Dr. Anderson stated that previous research indicates that results from participants who played violent video games and received random noise blasts are more relevant. Tr. 238-40. Though the Court is skeptical about the explanation of these contradictory results, it is willing to assume for purposes of discussion that Dr. Anderson’s conclusion is correct. The Court, however, questions the overall import of Dr. Anderson’s findings, given that on a one to ten scale of intensity, the most “aggressive” violent video game players administered an average blast of 5.93, and the least “aggressive” non-violent video game players administered an average blast of 3.98. There was only a two point difference, and both averages were in the middle of the intensity scale. See Craig A. Anderson et. al., Violent Video Games: Specific Effects of Violent Content on Aggressive Thoughts and Behavior, 36 Advances Experimental Soc. Psychol. 199, 215-24 (2004) (hereinafter, “Violent Video Games: Specific Effects”). In the final experiment, half the students played two violent video games and half of them played two non-violent video games. The violent video games were subdivided into those with human targets and those with non-human targets. They then engaged in the same noise blast task as the participants in the second experiment. The students who played violent video games gave “more or higher punishment levels to their opponents than those who played one of the nonviolent games.” Tr. 241-42. Dr. Anderson’s study, however, indicates that those participants who played a violent video game administered an average noise blast that was less than one point higher than the average noise blast of non-violent video game players. See Anderson, ‘Violent Video Games: Specific Effects” at 229. Dr. Anderson also found that the participants’ reactions to playing a violent video game with human targets was the same as their reactions to playing one with non-human targets. Tr. 327-28. Dr. Anderson stated that there has only been one reliable longitudinal study into the impact of violent video games on aggression in minors. Such studies, which examine the effects of a particular variable at two or more points in time, are important because they can measure longer term effects, rule out alternative explanations for particular behavior, and help identify the cause of a particular behavior. Tr. 312. In this study, which is still undergoing peer review, Dr. Douglas Gentile, a colleague of Dr. Anderson’s at Iowa State University, interviewed third-, fourth-, and fifth-graders about the level of violence in the video games they played during the six month period of the study. Then, using reports from parents and teachers, he determined the child’s level of aggressive behavior at the beginning and end of the study to assess whether such behavior had increased. According to Dr. Anderson, Dr. Gentile’s longitudinal study showed that those children with a higher exposure to violent video games were more likely to have been in a fight by the end of the study, even after controlling for whether the child had been in a fight — -and likely had a proclivity for aggression — before the study began. Tr. 254-261. From this longitudinal study, Dr. Anderson concluded that “[w]hat is clear is that regardless of the initial cause, playing violent video games still makes children more aggressive.” Def. Exh. 8, Douglas A. Gentile and Craig A. Anderson, Violent Video Games: The Effects on Youth, and Public Policy Implications, in Children, Culture, and Violence, at 232 (N. Dowd et. al. eds.) (forthcoming) (emphasis in original). The total increase in aggressive behavior between the beginning and end of the study, however, was not very large; there was a high (0.4 to 0.5) correlation between aggression at the beginning and end of the study; and at most, only four percent of the increase in aggression was associated with exposure to video game violence. Tr. 318-21. Finally, Dr. Anderson discussed the results of three meta-analyses he conducted in 2001, 2003, and 2004. In a meta-analysis, a researcher compiles all of the studies that have been conducted in a particular area, combines the results from those studies, and makes conclusions based on the body of research as a whole. In each of these meta-analyses, Dr. Anderson found that exposure to violent video games was associated with aggressive thinking and behavior. The studies differed, however, because he added new data and adjusted his methodology with each meta-analysis. For example, in the 2003 meta-analysis, Dr. Anderson conducted a separate breakdown of studies involving participants who were eighteen or younger, and in the 2004 meta-analysis, he did a separate analysis of studies that used “best practices.” Nonetheless, his conclusion about the link between exposure to violent video games and aggression remained the same. Tr. 263-64. Dr. Jeffrey Goldstein, a social psychologist at the University of Utrecht in the Netherlands, and Dr. Dmitri Williams, an assistant professor of communications at the University of Illinois at Urbana-Champaign, testified for the plaintiffs, and they responded to Dr. Anderson’s testimony. Dr. Goldstein has conducted research into whether video games help improve cognitive skills, Goldstein Dep. at 53-54; Dr. Williams conducted an intensive one-month study for his doctoral dissertation on individuals who played a violent, multiplayer computer-based video game. Tr. 152-58. They agreed that there is a correlation between exposure to video game violence and increases in aggressive cognition and behavior, but disagreed with Dr. Anderson’s conclusion that the research establishes that exposure to violent video games causes increases in aggressive thinking and behavior. Goldstein Dep. at 84; Tr. 130. Dr. Goldstein and Dr. Williams shared a number of the same concerns with the methodology and conclusions of research regarding violent video games cited by Dr. Anderson. With regard to methodology, they were concerned that these studies defined aggressive thoughts and behavior vaguely (e.g., equating aggressive play with aggressive behavior), administered problematic tests for measuring aggression (e.g., Stroop tests and noise blasts), used violent and non-violent video games that were too dissimilar (e.g., Wolfenstein 3D and Myst), and failed to address the context of game playing (e.g., asking subjects to play for too short a time and without others around them). Goldstein Dep. at 36, 70-71, 79-82, 89-90, 97-98, 116-18, 121; Goldstein Aff. ¶¶ 7-8, 11-12; Tr. 138-51, 157-71. With regard to their conclusions, Dr. Goldstein and Dr. Williams noted that Dr. Anderson not only had failed to cite any peer-reviewed studies that had shown a definitive causal link between violent video game play and aggression, but had also ignored research that reached conflicting conclusions. Dr. Goldstein and Dr. Williams noted that several studies concluded that there was no relationship between these two variables. They also cited studies concluding that in certain instances, there was a negative relationship between violent video game play and aggressive thoughts and behavior (e.g., initial increases in aggression wore off if the individual was allowed to play violent video game for longer period). Goldstein Aff. ¶¶ 10, 22 41-42; Goldstein Dep. at 182-89. Though the Court believes that many of the measures of aggression used in violent video game research are likely valid, we agree with Dr. Goldstein and Dr. Williams that neither Dr. Anderson’s testimony nor his research establish a solid causal link between violent video game exposure and aggressive thinking and behavior. As Dr. Goldstein and Dr. Williams noted, researchers in this field have not eliminated the most obvious alternative explanation: aggressive individuals may themselves be attracted to violent video games. Gold-stein Aff. ¶ 33; Tr. 133. Even if one were to accept the proposition that playing violent video games increases aggressive thoughts or behavior, there is no evidence that this effect is at all significant. Dr. Anderson provided no evidence supporting the view that playing violent video games has a lasting effect on aggressive thoughts and behavior — in other words, an effect that lingers more than a short time after the player stops playing the game. Based on general psychological theories and long-term studies of television and movie violence, Dr. Anderson hypothesizes that frequently and intensely playing violent video games will have a lasting effect on young players. Tr. 291-92. He does not, however, cite any data or studies to back up his hypothesis. In most studies, test subjects play a game for between ten and seventy-five minutes, and only one study — Dr. Gentile’s — included an assessment of the effect of regular violent video game play over a longer period of time. This research is insufficient to draw conclusions about the long-term impact of video games on minors. Dr. Anderson also has not provided evidence to show that the purported relationship between violent video game exposure and aggressive thoughts or behavior is any greater than with other types of media violence, such as television or movies, or other factors that contribute to aggression, such as poverty. In fact, several of the studies he uses to support his conclusions examine media violence generally and do not disaggregate the effect of video game violence or compare the effects of video game violence to these or other forms of media violence. Finally, the Court is concerned that the legislative record does not indicate that the Illinois General Assembly considered any of the evidence that showed no relationship or a negative relationship between violent video game play and increases in aggressive thoughts and behavior. The legislative record included none of the articles cited by Dr. Goldstein or Dr. Williams. It included no data whatsoever that was critical of research finding a causal link between violent video game play and aggression. These omissions further undermine defendants’ claim that the legislature made “reasonable inferences” from the scientific literature based on “substantial evidence.” See Turner Broadcasting System v. FCC, 512 U.S. 622, 665, 114 S.Ct. 2445, 129 L.Ed.2d 497 (1994). 2. Effect of violent video games on brain activity Dr. William Kronenberger, who testified for the defendants, is a clinical psychologist at the Indiana University School of Medicine who focuses on working with and studying children and adolescents with behavior disorders. Tr. 12. He has conducted research into the relationship between adolescent exposure to media violence and aggressive thinking and behavior. Tr. 23. In his testimony, Dr. Kronenberger cited studies that he said indicate that increased exposure to media violence has negative effects on adolescent brain activity. To measure how exposure to media violence affects brain activity, Dr. Kronen-berger uses functional magnetic resource imaging, a neuroimaging technique that measures blood flow to certain regions' of the brain, allowing researchers to infer the level of activity in those areas. Tr. 31. Researchers use fMRI imaging to determine how a particular task affects brain activity by having the test subject perform the task inside an fMRI scanner, having a control subject perform a different, but similar task, and comparing the level of blood flow to the relevant areas of the test subject’s and control subject’s brains. Tr. 32. Dr. Kronenberger has published an article in a peer-reviewed journal discussing a study that used fMRI imaging to examine how exposure to media violence affects aggressive thinking and aggressive behavior in adolescents. Tr. 24. He and his colleagues specifically examined two parts of the brain — the anterior cingulate cortex (ACC) and the dorsolateral prefrontal cortex (DLPFC), which is comprised of the middle frontal gyrus (MFG) and the inferi- or frontal gyrus (IFG). Dr. Kronenber-ger’s research team decided to focus on these areas because prior research has indicated that lower activation, deficiencies, or injuries in these areas are associated with aggressive or violent behavior. Tr. 36. For the experiment discussed in his article, Dr. Kronenberger’s research team recruited two groups of adolescents — one consisted of adolescents diagnosed with a behavior disorder, and the other was a control group. Based on interviews, the researchers subdivided the groups into those with high and low exposure to media violence. Tr. 37-43. All of the subjects underwent fMRI imaging, and while they were in the fMRI scanner, they performed something called a counting Stroop test. In this test, a multiple-digit number is flashed onto a screen. The test subject is asked to state the number of digits, but the actual numerals are different from the number of digits listed. Stroop tests can be administered in a number of ways, but they are used to measure an individual’s ability to inhibit his initial impulses in favor of a more difficult response. See Tr. 19-20. Dr. Kronenberger and his colleagues found that the control subjects as a whole experienced activation in the ACC and the left DLPFC, while the adolescents with behavior disorders experienced activation in the MFG portion of the DLPFC but no activation in the ACC. Furthermore, when the researchers compared the control subjects with high media exposure to the control subjects with low media violence exposure, they found that control subjects with high media violence exposure had activation in the left MFG, the same region in which subjects with behavior disorders experienced activation, but control subjects with low media violence experienced activation in the ACC and the left IFG. Tr. 52-53. This experiment was part of a larger study. During the first phase of that study, Dr. Kronenberger conducted a second experiment, the results of which had not yet undergone the process of peer review. In this experiment, the methodology was the same except that while the participants were in the scanners, half watched simulated play of a violent video game (James Bond), and the rest watched simulated play of a non-violent game (ear racing). The participants did not actually play the games, but they were told to press a button whenever they would have performed an action had they actually been playing. Tr. 46-47. The researchers found that along the axis of high versus low media violence exposure, the control subjects with high media violence exposure experienced activation in the IFG, and the control subjects with low media violence exposure showed activation in the MFG. Along the axis of watching the violent versus the non-violent video game, the control subjects who watched the violent video game showed activation in the left and right MFG while the subjects with behavior disorders who played violent video games showed activity in the IFG. Tr. 56-58. During the second phase of the study, Dr. Kronenberger and his research team expanded their research to examine how exposure to media violence affects the am-ygdala, a portion of the brain that he testified has been associated with emotional functioning, particularly “threat arousal stimuli.” The results of this portion of the study have not yet been published in a peer-reviewed journal. Dr. Kronberger had the test subjects complete what' he termed an “emotional Stroop color-word test.” Both the adolescents with behavior disorders and the control subjects had to state the color of the word, but half of each group was asked to read words of aggression or harm (i.e., “kill, murder, rape”) while half of each group was asked to read words without violent connotations. Dr. Kronenberger found that the control group experienced activation of the ACC and the DFC while the adolescents with behavioral disorders showed increased activity in the amygdala and the parahippocampal gyrus, which are also associated with emotion. Additionally, test subjects who had high media violence exposure also showed activation in the amygdala and the parahippo-campal gyrus while those with low media violence exposure showed activation in the DLPFC. Tr. 68-73. Dr. Kronenberger conceded that his studies only demonstrate a correlative, not a causal, relationship between high media violence exposure and children who experience behavioral disorders, decreased brain activity in the ACC and the DLFPC, and increased activity in the amygdala and the parahippocampal gyrus. See Tr. 77-78. Dr. Kronenberger has also published two articles in peer-reviewed journals which report studies on the use of neuro-cognitive testing to examine how media violence affects cognitive activity. Tr. 28-29. In the first article, Dr. Kronenberger reported the results of a study in which he evaluated the impact of exposure to media violence on a child’s executive functioning — the ability to plan, organize, direct, and carry out behavior that is oriented toward a goal. Specifically, Dr. Kronen-berger used four measures to evaluate executive functioning — two neurocognitive tests performed in a lab and two self-reporting tests. He then determined the children’s level of media violence exposure. After controlling for a number of variables, Dr. Kronenberger found that “there was a significant correlation between media violence exposure ... and poorer scores on each of those four measures of executive functioning.” He ultimately concluded that the higher the media violence exposure, the worse the test subjects performed on measures of executive functioning. Tr. 61-64. In his second article, Dr. Kronenberger compared adolescents with behavior disorders and control adolescents, and evaluated whether those with behavior disorders and their parents reported greater exposure to media violence in the last week and in the last year. Dr. Kronenberger found that adolescents with behavior disorders and their parents reported more media violence exposure over the last year. After the study concluded, he also used a post hoc analysis to examine the unique effects of television violence and video game violence, and he found that each was related to whether an adolescent had a behavior disorder or was in the control group. Tr. 65-67. Dr. Howard Nusbaum, a cognitive psychologist at the University of Chicago, testified for the plaintiffs and responded to Dr. Kronenberger’s testimony. Dr. Nusb-aum uses fMRI imaging in his research, and he has written several articles on methodological issues on how to conduct fMRI research and interpret the results. Tr. 351-53. He thoroughly reviewed Dr. Kronenberger’s research and identified fundamental problems with “the background assumptions of the work, where it starts from, the methodology that is represented in the work and the conclusions that are drawn.” Tr. 358-59. Initially, Dr. Nusbaum testified, Dr. Kronenberger made two incorrect assumptions. First, he assumed a one-to-one relationship between various parts of the brain and particular behaviors. Dr. Nusb-aum testified that particular brain activity can affect multiple behaviors, and specific behaviors can be influenced by activity in multiple areas of the brain. Specifically, he stated that activity in the ACC and DLPFC — the areas Dr. Kronenberger stated were associated with impulse control, self-regulation, choice, attention, and concentration — affect many behaviors, and the behaviors Dr. Kronenberger traces to the frontal lobes also implicate other areas of the brain. In fact, according to Dr. Nusbaum, “[t]hese areas certainly are not highly associated with people who are aggressive, or activity in these areas is certainly not highly associated with people who are aggressive.” Tr. 374-75. Second, Dr. Kronenberger assumed that decreased activity in one part of the brain equaled impaired or deficient brain activity. Dr. Nusbaum disagreed, stating that decreased activity can signal expertise or use of an alternate method to complete the assigned task. Tr. 359-68. Next, Dr. Nusbaum discussed problems with Dr. Kronenberger’s methodology. In evaluating Dr. Kronenberger’s fMRI studies, Dr. Nusbaum noted several problems. First, the images published in Dr. Kronen-berger’s study are composites of the images of all the individuals in the study. Such images, Dr. Nusbaum explained, can appear to show activity in areas where no individual subject actually showed activity. Second, with regard to both types of testing, Dr. Nusbaum stated that Dr. Kronen-berger’s methods for measuring aggressive thoughts, such as the Stroop tests, the Conners performance test, and the behavior checklist, were not appropriate for measuring the type of behavioral control required to prevent oneself from reacting aggressively to a situation. Tr. 368-70, 372-73, 376-77, 387-88. Third, in discussing Dr. Kronenberger’s neurocognitive testing study alone, Dr. Nusbaum testified that such testing used particular patterns of behavior to infer the part of the brain that was activated, but because of the many-to-many relationship between brain regions and behavior, it is not possible to make “those clear kinds of inferences.” Tr. 368. Finally, Dr. Nusbaum discussed his concerns about Dr. Kronenberger’s conclusions. In discussing Dr. Kronenberger’s single published study, Dr. Nusbaum identified several problems with the researchers’ interpretation of the data. He noted that the researchers had drawn conclusions about similarities between the brain activity of adolescents with behavior disorders and control adolescents with high media violence exposure and about the alleged impairments in brain activity of these groups. Dr. Nusbaum stated that the scans showed that the composite brain activity of these groups differed and that it was unclear whether they truly had decreased brain activity as compared to adolescents without behavior disorders and with low media violence exposure. Even if the images were read to show decreased brain activity for these groups in certain areas of the brain, Dr. Nusbaum stated, there were several alternative reasons, such as the development of expertise or the use of another part of the brain to perform the same function, that Dr. Kro-nenberger’s research team had not considered. Tr. 373-81. Even if there were some deficit in brain activity, Dr. Nusbaum noted, Dr. Kronen-berger’s two experiments show adolescents with behavior disorders, and those without such disorders but high media violence exposure, with decreased activity in different areas in the two studies. According to Dr. Nusbaum, “I know Dr. Kronenberger has argued that it’s a different task, and so there’s different demands on it, and I agree with that, but in respect of the model of deficit, if the brain is deficit, it should be in both cases.” Dr. Nusbaum stated that the decreased activity in both experiments likely reflected the participants’ use of another method for completing the task, not mental impairments. Tr. 386. In discussing the second experiment in Dr. Kronenberger’s fMRI study — the experiment in which the participants watched a simulated video game and then took an “emotional Stroop color-word test” — Dr. Nusbaum again found that the researchers lacked the information to make definitive conclusions because they had not controlled for certain characteristics of their participants. Even if they had, he said, there were several alternative explanations for why the amygdala was activated in adolescents with behavior disorders and control subjects with high media violence exposure. Tr. 387-93. Dr. Nusbaum also disputed the conclusions of the neurocognitive study in which Dr. Kronenberger used neurocognitive testing to evaluate whether exposure to media violence impairs executive functioning. The problem with this study, according to Dr. Nusbaum, was that Dr. Kronen-berger and his colleagues compared the response times of the participants in the study “raw,” without accounting for the overall quickness or slowness of the participants in any activity they engaged in. As a result, Dr. Nusbaum testified that it was impossible to draw any conclusions from the data as reported. Tr. 381-83. The Court found Dr. Nusbaum’s testimony credible and persuasive, and Dr. Kronenberger’s unpersuásive. Consistent with Dr. Nusbaum’s testimony, the Court finds that Dr. Kronenberger’s studies cannot support the weight he attempts to put on them via his conclusions. The defendants have offered no basis to permit a reasonable conclusion that, as the legislature found, minors who play violent video games are more likely to “[ejxperience a reduction of activity in the frontal lobes of the brain which is responsible for controlling behavior.” Tr. 356. 3. Chilling effect of the WGL and the SEVGL The parties submitted affidavits and deposition testimony from the heads of the ESA, VSDA, and IRMA, and from a video game creator, discussing their concerns about the WGL and the SEVGL. According to Douglas Lowenstein, president of the ESA, the WGL and the SEVGL will have a chilling effect on the speech of video game creators, publishers and retailers: by restricting the sale of games and creating a regulatory structure that requires retailers to make game by game decisions about what they can and cannot sell, ... the ripple effect of that is potential retailers deciding not to sell a wide variety of games, therefore having a chilling effect on the speech rights of ESA members and other people who publish video games. Lowenstein Dep. at 101-02. Lowenstein stated that retailer decisions “will prevent the expression of ESA members from reaching a wide range of willing recipients.” Lowenstein Aff. ¶ 15. Because the definitions in the WGL and the SEVGL are vague, Lowenstein states, retailers will make subjective individual decisions about what is regulated by the “fairly vague and ill defined terms in the statute.” Lowenstein Dep. at 105. For example, he states that to avoid criminal liability, certain retailers might, out of an abundance of caution, label certain T-rated games— which are, under the ESRB ratings, acceptable for anyone thirteen or older— with an “18” sticker. According to Lowen-stein, labeling games that are not actually covered by the Act could in turn confuse consumers, because the “18” sticker conflicts with the ESRB ratings, and they may refuse to purchase the games due to the “scarlet letter quality” of the label rather than because of any real concerns about the content of the game. Id. at 126-28. Lowenstein contends that this requirement will not educate consumers, but will “primarily serve to stigmatize ESA members’ games.” Lowenstein Aff. ¶ 18. Additionally, Lowenstein stated, because the statutes are extremely burdensome-requiring retailers to put in long hours to ensure compliance — they will have a direct impact on what retailers sell and how they sell it. He stated that this, in turn, will affect members of the ESA, because they cannot publish products that retailers are unable or unwilling to sell. Lowenstein Dep. at 130-31; Lowenstein Aff. ¶¶ 18-19. Lowenstein conceded, however, that the ESA has no specific indication from Illinois retailers that they would refuse to sell or would be unable to sell video games that would require the “18” sticker. Lowen-stein Dep. at 106-07,130-31. Lowenstein also explained that conducting more outreach and education about the ESRB rating systems would offer án alternative means for achieving the General Assembly’s goals while preserving plaintiffs’ First Amendment rights. According to Lowenstein, Gail Markels, a representative of the ESA, shared this proposal with the Illinois House Civil Judiciary Committee and Governor Blagojevich, contending that this alternative would be effective because “as the FTC has found, parents are involved in 83% of video game purchases for minors.” Both the General Assembly and Governor Blagojevich rejected this proposal. Lowenstein Aff. ¶¶ 21-23. Crossan Andersen, president of the VSDA, discussed the effect of the statutes on home video retailers. He stated that the statutes placed an immense burden on retailers and the “affirmative defenses provided in the Act do not reduce the burden.” Andersen Aff. ¶¶ 16-17. Andersen discussed the issue with VSDA members in Illinois who expressed “[a] conviction that they could not determine what could be covered.” Andersen Dep. at 51-52. He stated that “it is impossible for VSDA’s members to determine with any degree of certainty whether any number of the titles in their inventories is covered by the definition of ‘violent video games’ ” because of the pace of certain games (e.g., there is no way of knowing what actually happened to a character in the last screen) and the special powers possessed by many of the characters who appear “human” in video games. Andersen Aff. ¶ 10-11. Andersen said his members also expressed “grave concern” about the “message that would be conveyed” by the “18” sticker required on violent and sexually explicit video games. Andersen Dep. at 91. In discussing the burden posed by these requirements, Anderson noted three problems: the labels would be particularly large, they would force retailers to eliminate product information they consider important, and they would conflict with the message retailers seek to advance (e.g., displaying games that are T-rated by the ESRB with an “18” label). Andersen Dep. at 51-52; Andersen Aff. ¶ 12; 18-19. Andersen expressed concern that the “18” symbol will create a stigma that: M-rated, T-rated, or even lower rated games carrying the “18” label will no longer be perceived as products which simply contain material some parents might find unsuitable for their children (like R-rated or PG-13-rated movies) and instead be treated as though they were sexually explicit adult products— even pornography. Andersen Aff. ¶ 22. As a result, he states that some adult customers will be chilled from acquiring such games. Id. Andersen stated that his members were concerned about the “particularly obtrusive” signage requirements. Andersen Dep. at 51-52; 89-91. He expressed concern that this requirement would force retailers to communicate regarding the ESRB rating system in a way that they have not chosen for themselves and that the requirement that signs be placed at multiple locations will be extremely burdensome. Andersen Aff. ¶ 13. Andersen conceded that he did not know of any members who conducted research into the burdens or cost of meeting the requirements in the WGL and the SEVGL. Andersen Dep. at 51-52; 87-89. Andersen stated that the ban on use of self-scanning machines to purchase or rent the regulated games would be extremely burdensome, but he stated that he had not heard any concerns specifically about the legislation’s self-scanning ban. Andersen Aff. ¶ 15; Andersen Dep. at 51-52; 89-91. Finally, Andersen noted that many retailers rent video games under contracts with their customers. These agreements allow customers to designate family members, including children, to rent video games on an account and enable customers to set limitations on the types of games their children can rent. According to Andersen, “[t]his Act burdens retailers and their customers (who are consenting parents under these agreements) by obligating the retailer to disregard the parent’s wishes and instead restrict their children to the 18’ rating imposed by the state.” Andersen Aff. ¶ 23. David Vite, the president and chief executive officer of IRMA, discussed the impact df the WGL and the SEVGL on Illinois retailers. He stated that IRMA members had expressed concerns about many of the statutes’ provisions. First, he stated that the labeling requirement would be burdensome and confusing. Multi-state retailers would have to require their Illinois stores to individually label the regulated materials, and they would be forced to make a “subjective judgment on what is violent or sexually explicit.” To make that decision, the retailer would have to play the entire game and determine, “based on 102 state’s attorneys’ judgments, what is violent or sexually explicit.” Id. at 63-65. Vite conceded that no studies had been done to calculate the costs imposed by this requirement. Id. at 72-76. Vite stated that parents would be confused to find potentially conflicting ESRB ratings and “18” stickers on the game, particularly with regard to M-rated games that the ESRB considers appropriate for seventeen-year olds but that might be subject to the Illinois labeling requirement. Id. at 66-67. Second, Vite discussed the burdens of the requirement to post signs regarding the ESRB rating system. He stated that members were worried about the costs but agreed that he did not have a definite figure about the actual costs of. printing and hanging the signs. He also stated that members were concerned about the number and location of the places where a retailer would be required to place these signs, both because they would displace other marketing information and because they would be difficult to maintain. As an example, he referenced stores with sixty cash registers or points of sale, which would have to maintain a sign at every cash register. If a sign were to fall down, “a state’s attorney or some other person [could] raise the issue that [the signs] were not at the point of sale,” leaving the store open to criminal liability and damage to its reputation. Id. at 79-83. Finally, Vite stated that the statute’s limitations on the use of self-scanning machines to purchase regulated games would be unduly burdensome. He did, however, concede that there are members of IRMA who have self scanning machines but have developed alternate ways of selling products, such as alcohol and cigarettes, that are prohibited to individuals under a certain age. Id. at 86-88. Ted Price, the president and chief executive officer of Insomniac Games, discussed the chilling effect of the WGL and SEVGL on video game creators. He stated that “game creators not only strive to entertain their audiences but to communicate messages and ideas including a variety of political, social, and moral commentary.” Price Aff. ¶ 4. Price stated that the Illinois statutes would chill those messages, because the Act’s definition of violence — particularly the phrase “human-on-human violence” — is so vague that creators will be unable to determine what' is regulated, forcing them to eliminate anything in their games that resembles violence. As way of example, Price stated that he “would feel very uncomfortable about including even cartoon violence,” such as the type included in Insomniac Games’ T-rated Ratchet and Clank series. Id. ¶¶ 7-10 (noting that Ratchet and Clank’s characters are humanoid, not human, but could be misinterpreted by some retailers as human). Discussion In deciding whether to grant permanent injunctive relief, a court must consider whether the plaintiff has succeeded on the merits; whether it has an adequate remedy at law or will be irreparably harmed if the injunction does not issue; whether the threatened injury to the plaintiff outweighs the threatened harm the injunction may inflict on the defendant; and whether the granting of the injunction will harm the public interest. See, e.g., Plummer v. Am. Institute of Certified Public Accountants, 97 F.3d 220, 229 (7th Cir.1996). With this standard in mind, we examine the Violent Video Game Law and the Sexually Explicit Video Game Law in turn. Initially, however, we address defendants’ motions to dismiss based on Eleventh Amendment immunity and plaintiffs’ purported lack of standing. 1. Eleventh Amendment Immunity Defendants Madigan and Devine have moved to dismiss this case on the grounds that they are immune from suit under the Eleventh Amendment. Generally, the Eleventh Amendment bars suits against a state or state officials. There is an exception, however, for suits to enjoin the enforcement of unconstitutional statutes. Ex Parte Young, 209 U.S. 123, 28 S.Ct. 441, 52 L.Ed. 714 (1908). Under Ex Parte Young, a state official may be sued so long as he or she has “some connection with the enforcement of the Act.” Id. at 157, 28 S.Ct. 441. Defendant Madigan argues that she is not involved with local enforcement of criminal statutes, and defendant Devine states that he is not involved with the enforcement of misdemeanors like the one created by the Act. Plaintiffs note that if these arguments succeeded, there likely would be no one for them to sue to challenge the Act. The Court disposed of the defendants’ arguments in an oral ruling at the outset of the hearing on November 15, but we nonetheless repeat our reasoning here. The Court finds that because Attorney General Madigan is charged with instituting proceedings on behalf of the state; representing the state in appellate criminal proceedings; providing advice and assistance to state’s attorneys regarding state law; and conducting criminal investigations and prosecutions at a county prosecutor’s request, she has a sufficient connection to enforcement of the statute to permit a suit against her under Ex Parte Young. With regard tó defendant Devine, the Court finds that the fact that he is charged with instituting prosecutions under the state criminal laws in Cook County provides a sufficient connection to enforcement of the Act to enable suit against him to enjoin its enforcement. For these reasons, neither defendant Madigan nor defendant Devine is immune from suit in this case. 2. Standing To establish its standing to sue, an associational organization, like each of the plaintiffs in this case, must show “(a) its members would otherwise have standing to sue in their own right; (b) the interests it seeks to protect are germane to the organization’s purpose; and (c) neither the claim asserted nor the relief requested requires the participation of individual members in the lawsuit.” Hunt v. Wash. State Apple Adver. Comm’n., 432 U.S. 333, 343, 97 S.Ct. 2434, 53 L.Ed.2d 383 (1977). Defendants argue that the ESA, VSDA, and IRMA have not provided a factual basis to establish standing, and that they are precluded from establishing standing because of conflicts of interest between the associations and their members. The Court rejects defendants’ arguments and finds that plaintiffs have standing to sue. Each of the plaintiff associations was created to serve the business and public affairs interests of its members, which include the creators and publishers of video games (ESA), home video retailers (VSDA), and Illinois retailers (IRMA). The members of these associations rely on the ESA, VSDA, and IRMA to file suits like this to protect their rights, and they will experience a chilling effect on their free speech rights if the WGL and SEVGL go into effect. See generally Lowenstein Dep. (ESA); Price Dep. (ESA Member); Andersen Dep (VSDA); Vite Dep. (IRMA). Moreover, despite defendants’ intimations, there is no indication of a conflict of interest among the various associations’ members. The Court therefore finds that the plaintiffs have standing to bring this action. 3. Constitutionality of WGL Sale, Rental, and Check-Out Restrictions The First Amendment embodies a principle that is at the core of our political system and our national ethos: “each person should decide for himself or herself the ideas and beliefs deserving of expression, consideration, and adherence.” Turner Broadcasting System, 512 U.S. at 641, 114 S.Ct. 2445. A law that restricts speech because of its message “contravenes this essential right.” Id. For this reason, content-based regulations are presumptively invalid. R.A.V. v. St. Paul, 505 U.S. 377, 382, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992). All parties agree that the WGL is a content-based regulation subject to the strictest scrutiny under the First Amendment. R.A.V. v. City of St. Paul, 505 U.S. 377, 382, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992). Under this standard, the state may impose a content-based restriction on speech only if it has a compelling interest and has chosen the least restrictive means to further this interest. Sable Communications of California, Inc. v. FCC, 492 U.S. 115, 126, 109 S.Ct. 2829, 106 L.Ed.2d 93 (1989). Defendants identify five interests in regulating violent video games that they claim are compelling, including preventing violent, aggressive, and asocial behavior; preventing psychological harm to minors who play such games; eliminating societal factors that may inhibit the physiological and neurological development of, its youth; facilitating the maturation of Illinois’ children into law-abiding, productive adults; and assisting parents in protecting their children from such games. 720 ILCS 5/12A-5(d)-(h). The Court agrees that the legislature has a compelling interest in preventing violent behavior by children, protecting children from violence, and assisting parents in achieving the same goals. When the state defends a regulation of speech as a means to “prevent anticipated harms,” however, “it must do more than simply ‘posit the existence of the disease sought to be cured.’ ” Rather, “[i]t must demonstrate that the recited harms are real, not merely conjectural, and that the regulation will in fact alleviate these harms in a direct and material way.” Turner Broadcasting System, 512 U.S. at 664, 114 S.Ct. 2445. Legislative determinations about the scope of a particular harm and the benefits of a potential solution must be accorded substantial deference, but they are not insulated from meaningful judicial review. A court must consider the facts bearing on an issue of constitutional law to ensure that the legislature drew “reasonable inferences based on substantial evidence.” Id. at 666, 114 S.Ct. 2445. The plaintiffs’s challenge to the WGL is, contrary to defendants’ claims, governed by the Seventh Circuit’s decision in Am. Amusement Machine Assoc. v. Kendrick, 244 F.3d 572 (7th Cir.2001). In Kendrick, the Seventh Circuit examined the constitutionality of an Indianapolis ordinance which limited the access of minors to video game machines deemed “harmful to minors” that were located in public places, such as movie theaters and arcades. The ordinance defined a video game machine that was “harmful to minors” as one that depicted “graphic violence” or “strong sexual content” that “predominantly appeals to minors’ morbid interest i