Rethinking Sex-Offender Registries
As they bicycled and scootered back to their homes from a trip to the local convenience store in the 9 p.m. darkness of Sunday, October 22, 1989, Jacob Wetterling, his brother Trevor, and their friend Aaron Larson were accosted by a masked gunman with a raspy voice. After ordering them to lie face down in a ditch, the man told all three boys to turn over, asked their ages, and examined their faces. Brandishing his gun, the kidnapper ordered Aaron and Trevor to run toward a nearby forest, threatening to shoot if they turned back. He took Jacob, then 11 years old.
Jacob's mother, Patty Wetterling, spearheaded an all-out effort to find her son. FBI agents, National Guard troops, and volunteers descended on St. Joseph, Minnesota. Posters were hung. Jacob's face appeared on the back of milk cartons. Tips flooded in, but no firm leads materialized.
Jacob remains missing. Mrs. Wetterling, for her part, wondered if anything could have been done differently. The answer, she believed, came in part from what the police told her: If only they had a list of suspects — a registry — they would at least have a place to start.
Mrs. Wetterling proved herself an effective lobbyist: In 1991, thanks largely to her efforts, the state of Minnesota established the nation's first public sex-offender registry. Three years later, President Bill Clinton signed the Jacob Wetterling Crimes Against Children and Sexually Violent Offender Registration Act that required all states to establish their own registries. Votes to establish and fund state registries and maintain national standards passed with almost no dissent.
The registries grew over time. Megan's Law, a 1996 amendment to the Wetterling Act, required community notification for certain sex offenders and placed many records on the then relatively new World Wide Web. In 2006, another new law, the Adam Walsh Act, established new national standards for the registries, assessed penalties on states that didn't follow them, built a national internet database of offenders, established an office to track them, and expanded the registries. Today, all 50 states, the District of Columbia, and Puerto Rico maintain registries. The practice has spread internationally, and the United Kingdom, Canada, and Australia have all established registry systems of their own.
Life on a registry imposes many burdens on those required to take part. Individuals included on registries must inform police or other public-safety officials of their places of residence and work. Failure to register in a timely fashion can result in additional felony charges. They must obtain permission to move and, often, to travel. Most have their names posted in publicly accessible internet databases. A number of states — including Florida, Oklahoma, Tennessee, and Nevada — require some classes of sex offender to have special state ID cards or driver's licenses identifying them as such.
Many states and localities have laws forbidding sex offenders from living anywhere near schools or daycare centers, which often requires them to live far outside any city or reasonably dense suburb. Many are even barred from homeless shelters. Positions that bring sex offenders into regular contact with children — nearly all jobs at schools — are also off-limits. In many places, people on registries cannot patronize sexually oriented businesses, own firearms, and even hand out candy on Halloween. Laws to increase penalties on registered sex offenders even further — restricting them from visiting playgrounds or barring them from living with their own children — also have widespread public support. Indeed, it appears that no proposed sex-offender registration law has ever failed a free-standing, regular-order floor vote in any state legislature. No state that has passed a sex-offender registration law has ever repealed it, and no law has ever been weakened in a substantial way — even when stories emerge of serious consequences for former offenders. In California, for instance, which keeps some of the most detailed public statistics on sex offenders, 20% have no place to live as a result of residency restrictions. Such stories evoke little public sympathy and inspire few calls for reform.
In short, few new public policies have become so widespread so quickly or attracted such unanimous support from across the political spectrum. The reason for this is obvious: All parents are horrified by the thought of their children being snatched from them and sexually abused. Sexually oriented crimes committed against children are, for deep-seated cultural and perhaps innately human reasons, considered particularly grave violations of human dignity.
The registries have, in an important sense, worked: Patty Wetterling's successful crusade correlated with improvements in public safety. Rape rates, tracked by the Federal Bureau of Investigation's Uniform Crime Reporting Program, declined from roughly 37 per 100,000 in 1995, after the first national registry law passed Congress, to about 26 per 100,000 today, a 30% drop. Even as the population has grown by roughly 13%, the number of child sexual-abuse cases fell from about 88,000 in 1999 (the first year for which the Department of Health and Human Services collected data on a national level) to fewer than 61,000 in 2013. While these numbers (and any others associated with sex crimes) are probably best considered as relative measures since so many sexual offenses go unreported, they reflect a significant drop in the offenses that registries are intended to prevent.
Despite all this good news, however, a closer look at sex-offender registration reveals a more nuanced and disturbing story. Although effective in some respects at reducing crime, today's sex-offender registries do not work as well as they could. Current registries are too inclusive, are overly restrictive, and end up hurting some of those they are intended to help. With some common-sense reforms, sex-offender registries could become far more effective in improving public safety.
Lawmakers and public-safety advocates should consider reforms to limit the number of people in the registries. Though it may seem counterintuitive, they must roll back some of the restrictions placed on those who register if we are to have any hope of re-integrating them into society. We must do more to keep the most dangerous offenders out of schools, and we must monitor the most potentially dangerous criminals more closely and even increase the use of the most severe sanctions (like lifetime civil commitment) that are currently available. Registration of sex offenders can be an effective law-enforcement tool, but over-registration and overly restrictive rules on all those who are registered may do more harm than good.
A COMPOSITE SKETCH
Any examination of the registries must start with a look at the demographics of sex offenders who target children; they are far different than many people imagine. Sex offenders come from all walks of life. People convicted of sex offenses are slightly more likely to be white than non-white, relative to other felons. They have slightly higher levels of income and educational attainment (most are high-school graduates) than those incarcerated for other serious crimes. Insofar as they pursue adult sexual relationships at all, the overwhelming majority are men sexually interested in women. But few broad demographic characteristics give evidence as to who is likely to become a sex offender.
According to the Bureau of Justice Statistics' overview of sex offenders, most sex offenders targeting children have some sort of prior criminal record. Only about 15%, however, have been convicted of another sex offense, and only about a third of these prior offenses involve children. Among violent sex offenders, a category that includes all those who have sex with someone under the age of consent, the BJS data show that over 40% are arrested again within three years for some crime, but only about 5% actually commit another sex offense within three years. Indeed, just over 3% of released child molesters are arrested within three years for committing another sex crime against a child. When non-violent sex offenders — a category that includes those who deal in child pornography and expose themselves in public — are included in the dataset, recidivism rates drop substantially. And by all accounts, the recidivism of sex offenders is well below that of felons in general.
This does not mean, as some left-of-center academics seem to contend, that convicted sex offenders pose no danger to society and should not be monitored. They are at least 50 times more likely to commit sex offenses than are randomly selected men from the population as a whole. According to scholar Emily Horowitz, roughly 90% of sex offenders know their victims. Random kidnappers, like the man who took Jacob Wetterling, are quite rare. By most estimates, about a third of victims are family members of their abusers and most of the rest are victimized by someone else their parents know. Pedophiles seldom "kidnap" victims, as seen in movies and portrayed in popular novels. The Polly Klass Foundation estimates that fewer than 100 children are kidnapped by strangers each year in the manner that Jacob Wetterling was. Many of these "stranger kidnappings" involve children who were sitting in the back seats of stolen vehicles or interrupted another crime in progress. Parents wanting to protect their own children should worry much more about their own friends and relatives than random strangers.
Although no pedophile, by definition, has a healthy adult sex life, about 98% of male pedophiles, who account for over 90% of all pedophiles, classify their adult relationships as heterosexual or predominantly heterosexual.
It's not clear, however, if it's correct to think of any pedophile as "gay" or, for that matter, "straight." An influential, although hardly uncontested, body of research led by Fred Berlin of the Johns Hopkins School of Medicine suggests that pedophilia itself is a sexual orientation that, like heterosexuality or homosexuality, results from a complex interplay of difficult-to-measure social, environmental, and perhaps genetic factors. If this is the case, then, like other sexual orientations, it may well be essentially impossible to modify in adults. Even if true, however, this finding would not mean that people who are attracted to children are uncontrollable and untreatable: People with all sorts of sexual orientations can abstain from sex altogether. And this is precisely what we would expect pedophiles to do if they cannot overcome their attraction. Furthermore, it's not entirely clear (and may be impossible to know) whether every person convicted of molesting a child is a pedophile by "orientation." A large number of people may engage in sexually pathological behavior involving children for reasons such as a sex addiction, a desire to transgress social rules, or a non-age-related sexual fetish rather than an attraction to children per se.
Whatever the case, pedophiles exist, molest thousands of children each year, and pose a clear and present danger to society.
EFFECTIVENESS AND CLUTTER
The correlation between widespread sex-offender registration and falling rates of sex offenses does not establish that the offenses have declined because of registration. The falling rates of rape closely track a decline in all forms of violent crime. One could name any number of theories explaining the causes of the overall drop in violent crime. They include but aren't limited to better policing, higher rates of incarceration, demographic trends, bans on lead-based paint and gasoline, changes in the architectural design of cities, the wider legalization of elective abortion, and cultural shifts that more harshly sanction violent behavior. Reductions in child sexual abuse also closely track a more-or-less equal reduction in non-sexual abuse of children.
The best research on the efficacy of sex-offender registration does show positive effects, in terms of reduced sexual offenses. What appears to be the single most comprehensive, robust, and recent review of the national data, a 2011 study published in the Journal of Law and Economics, finds registration reduces the number of sex offenses by about 13%, after controlling for a number of relevant variables. But while the literature finding a causal reduction from registration is reasonably robust, this result is by no means universally confirmed. A more limited study published in the same journal that confined its work to Washington, D.C., found no effect at all. It's also not clear in which direction the causation flows. Since 95% of sex offenses are committed by people who have not committed a prior sex offense, a large part of the value of registration may come from deterring some number of sex offenses by people who might otherwise commit them.
Most important, virtually no well-controlled study shows any quantifiable benefit from the practice of notifying communities of sex offenders living in their midst. No study of the practice has shown notification, as opposed to registration, to have deterrence value in preventing sex offenses. The literature does show overwhelming evidence of large costs to neighbors in the form of reduced real-estate prices. A major study published in the American Economic Review measured an average 4% loss in the value of homes within a tenth of a mile of a sex offender. While there are some anecdotal cases of community notification helping to catch individual sexual predators, it's not clear that any sex offender who re-offended has ever been caught by neighbors solely because of public notification of his presence. In other words, the biggest quantifiable cost of sex-offender notification appears to be borne by the neighbors it is intended to help, with no measurable improvement in public safety.
When it comes to the most important presumed function of the registries — keeping pedophiles out of schools — they seem to be failing dramatically. Although 46 states and the District of Columbia maintain procedures to keep pedophiles out of schools (and nearly all sizable school districts in the remaining four states have procedures of their own), a Government Accountability Office report found the system simply doesn't work and has allowed hundreds of sex offenders into direct contact with children.
Possibly due to bureaucratic confusion stemming from a patchwork of government agencies that lacks a single point of contact, a surprisingly large number of pedophiles find work in school settings with the very types of children they victimized. Some state laws and union contracts may even limit schools' ability to fire pedophiles or parents' ability to sue. Fear of lawsuits can instead lead some districts to counsel sex offenders out of jobs and subsequently send them on to other districts with letters of reference. In a typical year, school personnel commit roughly 400 sexual offenses against students. While the monitoring of sex offenders in society may be too harsh in some respects, efforts to monitor them in schools do not seem extensive enough.
While some people on the registries certainly are public threats, many are not. Journalist and lawyer Chanakya Sethi found that 12 states require registration for urination in public and six states do for prostitution-related offenses. Teenagers who have consensual sex with other teenagers can be forced to register (sometimes for life) in 29 states. Numerous states permit and some even require registration for kidnapping, even where it has no sexual element. Consensual incestuous sex between adults (while deeply abnormal) can require registration, even though it presents no public danger.
Most disturbingly, about 40 states put juveniles on sex-offender registries, and Nicole Pittman of Impact Justice has found that six states can require juveniles to register for life. Indeed, the federal Adam Walsh Act created some incentives for doing exactly that. At least 5% and perhaps as many as a quarter of all people on the registries around the country are there for offenses for which they were tried as juveniles. Many of the offenses these juveniles have committed are as trivial as indecent exposure. In Pittman's fieldwork, she has uncovered numerous children younger than 10 years old who have ended up on the registry for "assaults" that involved games of "doctor" and other sexually oriented play they may not have even understood. In 2015, one Michigan judge handed down a sentence of 25 years on the sex-offender registry to a young man who, at 19, had consensual sex with a 14-year-old girl who had claimed to be 17. (After a public outcry, the judge reluctantly agreed to reconsider his sentence.) Prosecutors in Archbold, Ohio, brought charges that could have resulted in mandatory registration for high-school students caught exchanging nude "selfies."
Certainly, some juveniles may commit heinous and violent sex crimes for which registration is appropriate. Where that is the case, all states but New Mexico allow people under the age of juvenile jurisdiction to be tried as adults for at least some sex offenses. But the presence of non-violent and non-threatening juveniles on sex-offender registries contributes to registry "clutter" that makes it difficult for police and social workers to monitor the truly dangerous sex offenders. Phillip Garrido, who kidnapped and held Jaycee Dugard in his backyard for 18 years and abused her repeatedly, is a good example of someone who slipped through the cracks. He was on a sex-offender registry for prior incidents of molestation and kidnapping. His home was visited by parole officers and social workers numerous times. But, overtaxed by the need to monitor California's more than 83,000 registered sex offenders, officials never performed the thorough search of his house that would have located Dugard. Instead, it took sharp-eyed officials at the University of California, Berkeley, to bring about her eventual rescue. In a time of stretched budgets, effectively monitoring truly dangerous sex offenders is going to require pruning the registries.
People looking at the system of registration are thus left with a paradox: It seems to do some good, but many of its features also do a great deal of harm. Ending the registries would be both unwise and hugely unpopular, but responsible policymakers should focus on some sensible ways they could be improved.
REDUCING THE REGISTRIES
Making the registries more effective should start with reducing the number of offenders listed. Removing those who do not pose any particular public danger would both remedy the injustices done to them and improve public officials' ability to monitor those who remain. Two groups in particular deserve speedy release from the registries: those convicted of minor, sometimes non-sexual offenses and those whose convictions were handed down by juvenile courts.
Adults convicted of offenses like indecent exposure, public urination, prostitution or soliciting prostitution, kidnapping their own children as part of a custody dispute, and consensual incest with other adults all deserve various forms of social censor or punishment or both. But there's no evidence they pose public dangers beyond those associated with these relatively minor criminal offenses. None of these behaviors have been linked to child molestation or violent sexual assaults anywhere in the academic literature. Requiring such offenders to remain on registries wastes public resources, ruins lives, and does nothing to improve public safety.
For many of the same reasons, people convicted in juvenile court should, as a class, be removed from registries; their continued presence is perverse and undermines the purpose of the juvenile justice system. Juveniles who act out sexually get branded as "pedophiles" under laws that consider victims' ages but not those of offenders. A 17-year-old boy who has consensual sex with a 15-year-old girl might need counseling or punishment from his parents, but he certainly isn't a pedophile. Two teenagers who swap naked "selfies" may deserve to lose their smartphones, but they certainly aren't "child pornographers." Laws that fail to take these obvious realities into account impose huge consequences on juveniles convicted of sex offenses: the threat of being banned from living with their own siblings, being forced into foster care, and expulsion from their high schools (the same schools doing such a poor job of ensuring that pedophiles don't get hired). None of these collateral consequences does any good for society, for the offenders, or for their victims.
Moreover, the long-lasting, sometimes lifelong, nature of sex-offender registration runs counter to the purpose of the juvenile justice system. Juvenile courts are intended primarily as therapeutic and rehabilitative mechanisms. They have looser rules of evidence than adult courts; they maintain far fewer public records; and, at least in theory, they hand out sanctions based on the "best interest" of the accused, rather than a desire to punish. Only a few states allow jury trials in juvenile court, and even then they are quite rare. Most states allow juvenile records to be sealed; the process is sometimes even automatic. Even people with unsealed records typically retain the rights to vote, receive government benefits, and live where they choose.
If prosecutors or police believe that a juvenile is so dangerous that he merits long-term registration, they ought to avail themselves of procedures to try him in an adult court. Any other standard undermines the very idea of maintaining a distinct system for younger offenders.
Estimating precisely how many offenders would be removed from registries as a result of this change in policy is difficult. Registries rarely report the age at which their registrants were convicted. What data do exist suggest that those convicted as juveniles make up as much as a third of registered offenders in the 40 states that have some form of juvenile registration. It's estimated an additional 10% of non-juvenile registrants are guilty of offenses that pose no obvious public harm, although this may differ a good deal from state to state. Whatever the ultimate figure, it would be easy to reduce the size and scope of sex-offender registries — and the hardships imposed on those who have committed only minor offenses — while actually increasing public safety.
By any count, however, the majority of people on the sex-offender registries are adults who committed reasonably serious crimes. They are more likely than members of the population as a whole to commit such acts again, even though most of them will not. Of course, the same can be said of almost anybody with any sort of criminal record. As with other people who commit crimes, it's unfair and unjust to brand all sex offenders as social pariahs for the rest of their lives, particularly since they have lower recidivism rates than other types of felons.
Making it impossible for sex offenders to live in most places contributes directly to their becoming homeless, which in turn makes them harder to track — and harder to keep away from potential victims. Far-reaching residency bans, although politically popular, simply do not pass the most basic cost-benefit test. Every dataset makes clear that children are far more likely to be sexually abused by family members than by strangers who happen to live near their school or daycare center. Judges, police, and probation officers can and should still be able to require many classes of sex offenders to stay off of school grounds during school hours and avoid other areas where children congregate (something modern GPS-monitoring can assure cheaply and easily), but blanket residency restrictions simply do not serve any valid public-safety purpose.
Forcing convicted sex offenders to the margins of society also tends to remove them from the orbit of family, friends, and houses of worship, making it more likely that they will turn to crime again. For instance, it's difficult to see why sex offenders should be automatically denied commercial driver's licenses or barred from working as insurance agents. Aside from obvious restrictions on working with children and perhaps carrying out certain medical tasks, most restrictions on sex offenders should be tailored to fit individual circumstances and levels of dangerousness. Restrictions on professional licensing should be set to fit the specific sex offense, rather than applied to every person convicted of any sexually oriented crime.
Moreover, the lack of any evidence that public notification reduces crime, coupled with its negative effects on property values, counsels in favor of restricting the practice. Notification helps attach an unnecessary stigma even to those convicted of only minor sex offenses. A person who sexually gropes a stranger once has done something wrong and perhaps traumatizing, but he does not pose the same public danger as a murderer, who is not required to notify his neighbors of his prior conviction. Yet, because of registries, he faces a greater public stigma than a murderer. Eliminating public notification completely would face huge political hurdles and, given the ease with which information already on the internet can be preserved, is probably impossible anyway. The most practical change might be limiting mandatory community notification and internet recording to actual predators over the age of 21 who have sexually assaulted young children. Even in these cases, the value of notification likely comes more from the fact that the public wants it than from any demonstrable benefit it actually provides.
On the other hand, efforts to keep sex offenders out of schools ought to be enhanced and improved. Finding the resources to do this would be reasonably easy if much of the excess currently cluttering sex-offender registries were removed. In this context, a new, bipartisan proposal by Senators Joe Manchin and Pat Toomey deserves serious consideration. The bill would set federal standards to prevent child predators from working in schools and would penalize states where districts try to "pass the trash," or counsel sex offenders to resign quietly before they are sent along to other schools with positive letters of reference.
THE WORST OF THE WORST
For serious offenders, who constitute the majority of those currently on sex-offender registries, the practice of registration offers a deterrent value that appears effective at reducing sexual assault and child sex-abuse rates. Three careful and deliberate policy changes could help law enforcement deal more effectively with these truly bad actors: increased mandatory outpatient treatment; increased use of indefinite civil commitment for the worst offenders; and more targeted focus of federal resources on serious, mostly internet-based child predators and other serious sex offenders, rather than the child pornographers who currently make up the lions' share of the federal case load.
Insofar as sexual attraction to children is an essentially fixed sexual orientation, it may be impossible to truly "cure" it. Comprehensive literature reviews led by a team from the University of Illinois at Chicago have mixed findings: While the best-run treatments do reduce actual recidivism among sex offenders, the reduction is only by about one-third, and even then it's far from clear that pedophiles are made to let go of their sexual attraction to children altogether.
Interestingly, after adjustment for a variety of variables, outpatient treatment outside of secure facilities appears to work even better than forcing treatment behind bars. In fact, a number of studies show that treatment for sex offenders behind prison walls is counterproductive. This suggests it may be better to focus prison sentences for child molesters almost entirely on deterrence and punishment, while augmenting treatment efforts outside the jailhouse walls. For those who fail to participate in treatment programs, a version of the rapidly spreading "swift and certain sanctions" regimes — which provide short, often immediate jail stays every time an offender slips up — may provide an incentive to stick with the program and receive treatment. They have worked to encourage many drug addicts to break their habits, and they may help pedophiles in the same way. Many offenders who are removed from registries or kept on law-enforcement-only registries might continue to be subject to long-term GPS monitoring to keep them away from schools and other areas where they might pose a threat.
Some sex offenders may be resistant to all treatment and unable to control their urges to molest children. In these cases — which comprise a small but non-trivial percentage of sex offenses — moves toward increased civil commitment may make sense. All states allow for civil commitment of the dangerous mentally ill in hospital-like settings when the individual is deemed to pose a risk to himself or others. Currently, 20 states and the District of Columbia have statutes that provide for an additional level of review following the release of certain sex offenders. A small number of offenders at very high risk of offending again can, under these regimes, be detained indefinitely in hospital-like settings.
Such treatment, of course, is advisable only as an absolute last resort. But just as it's possible to detain a mental patient who experiences a drive to kill or maim others, it should also be possible to detain someone in situations where expert testimony convinces a court that they will commit sexual violence if released. A mandatory review process for certain grave sex offenses may be desirable. In exceptional cases, civil commitment of a tiny number of particularly dangerous juvenile sex offenders (who might otherwise be released with no public record) might be justified as well. Indefinite civil commitment is a very powerful tool to put in the hands of the state, and, certainly, it carries a risk of being overused. But it should not be ruled out in all cases for sex offenders, and its use likely deserves expansion.
The most difficult cases to deal with involve individuals found guilty of possessing child pornography. It goes without saying that any use of sexual materials involving children deeply offends social norms, and its mere possession ought to be subject to significant criminal sanction. Despite efforts of many left-leaning researchers to minimize the problem, furthermore, it is a truly serious one that has grown with the internet. Indeed, a recent study of the "Dark Web" conducted by scholar Gareth Owen found that roughly 80% of users visiting the secret websites that use untraceable Tor network technology were seeking child pornography.
But current laws involving child pornography — often prosecuted under federal law — may need to be updated. Child-pornography laws were written largely with the idea of prosecuting those who distributed magazines, print photographs, videotapes, and celluloid film strips depicting minors in sexual situations. Today, nearly all child pornography gets shared on peer-to-peer networks that make all consumers "distributors" simply by virtue of participation.
The average sentence for child pornography is now nearly eight years, longer than the average sentence for rape, which is just over five years. Whatever harm looking at a picture of a child in a sexual situation causes (and it's significant), it is probably not greater than the harm resulting from actual sexual assault. Nonetheless, the BJS finds that child-porn offenses make up 70% of the federal sex-offender registry caseload.
Rather than try to effect a change in federal law or prescribe punishments federally, it would be better to focus federal resources on the greatest dangers. These include human-trafficking rings and actual predators who lure children across state lines. Meanwhile, states should be encouraged to take on a greater share of the child-porn caseload and decide punishments based on local attitudes and beliefs. In any case, mere possession of child pornography should remain a reasonably serious crime, albeit one that is dealt with, for the most part, on the local level.
ADDRESSING THE REAL PROBLEM
The practice of requiring sex offenders to register with law-enforcement officials is effective and has contributed to a sizable drop in sex offenses committed against children in the United States. Notifying the public of sex offenders, on the other hand, is ineffective and should be limited if not eliminated. The registries that exist, furthermore, do tremendous harm to some people who, although clearly guilty of various wrongs, do not pose a significant threat to children or anyone else in society.
The nation needs to reconsider its headlong rush into ever-expanding sex-offender registration and target the registries more carefully at the most genuinely dangerous individuals. Certain petty restrictions should be dropped and many individuals should be deleted from the registries in order to minimize unnecessary damage to individuals and communities and to allow law enforcement to focus on the most dangerous offenders. In certain cases, serious punishments, including indefinite civil commitment for certain offenders, also ought to be expanded. Efforts to keep sex offenders out of schools also deserve expansion.
More than two decades after her initial success in establishing Minnesota's registry, Patty Wetterling — now a political activist who has run twice for Congress — expresses second thoughts about the registries she fought to establish. While she still supports the idea of the registries, Wetterling thinks they have gone too far and should drop juveniles and many other categories of offenders. "We can't just keep locking [sex offenders] up," she told Minneapolis's City Pages in 2013. "That doesn't change the problem."