How Can You Tell When the Government Is Lying?

Always Check the Footnotes

Time for another very exciting post, this time about a research brief commissioned by the federal SMART office — the office responsible for federal policy and implementation with respect to sex offense registration. Who doesn’t love research briefs? I’ve tried to think of a way to make the lede pop a little more here, and the best I can come up with is that the federal government appears to be lying about the state of scientific research in order to justify an extremely costly program that doesn’t do anything except violate the constitutional rights of about a million men, women, children (we put children on the registry too, sometimes for their natural lives!) & their families.

It matters because a (negligently or not) sourcing error, in part, is what led us to have such sprawling, burdensome, and ineffective registries.

The research brief in that was commissioned has the eye-catching title Sex Offender Registration and Notification Act — Summary and Assessment of Research. Published by the Federal Research Division, it reads like police-state apologetics — doing its dead level best to dismiss the research casting doubt on various aspects sex offense registries. I had occasion to review it for a client, and if the title didn’t catch my eye, the abstract certainly did:

The top-level conclusion in the abstract is that “as of June 2019” (emphasis added) the research is “inconclusive” regarding the effectiveness of registration. This came as something of a surprise to me, because I’m aware of quite a few studies within the last decade that have found either no effect or a negative effect (i.e., an increase) in recidivism, but none that show it leads to less recidivism (and I feel like if there were one demonstrating effectiveness, it definitely would have been included in this brief).

So I was very eager indeed to see what evidence the Federal Research Division had amassed to support this incredible claim. On page 2, note 7: they cite two sources, one published in 2008, and one published in 2009:

Perhaps it goes without saying, but those sources themselves do not prognosticate about the state of research a decade into the future. Nor does the report grapple with several major research studies over the last decade that have concluded that SORN either do not have an impact on recidivism or make recidivism slightly worse. It simply pretends they do not exist.¹

I thought it might be a typo at first — never ascribe to malice what you can ascribe to incompetence etc etc — but the 2019 claim is repeated several times, and is a key finding in the abstract. Further, there are two sources that are offered to support the claim, but none of which are from the late 00’s. It’s hard to know what to make of it except that the government is simply lying because it can and no one cares if it’s to keep the public safe from ‘sex offenders.’

Here’s why the research brief matters: the scientific findings related to the effectiveness of registries are relevant to legal arguments about whether or not they constitute punishment, amongst other things. It’s not hard to imagine (if it hasn’t happened already) this getting dropped into an AG’s brief, and then getting regurgitated into an opinion of a court of not-insignificant note, and it becoming judicially accepted “fact” even though it has no support.

It’s happened before.

In the early 00’s, the Supreme Court considered a series of cases regarding extraordinary treatment of people convicted of sex offenses (e.g., whether registration constituted punishment). In those cases, Justice Kennedy claimed that recidivism rates of people convicted of sex offenses was “frightening and high.” The only support for this claim, as Ira and Tara Ellman discovered, was a 1986 Psychology Today article written by a clinician who was marketing his own services as a treatment provider (and who has since repudiated how his statement was used). The Ellmans discovered that the phrase “frightening and high” regarding purported recidivism rates was subsequently cited in 91 judicial opinions and briefs in 101 cases, but that was 8 years ago, and courts and government lawyers are still turning to that very phrase to ward off arguments that registries and registry-related paraphernalia (e.g., exclusion zones, regulation of travel, communication) are unconstitutional.

So a research brief that muddies the waters and concludes (falsely) that the state of research is inconclusive is going to undoubtedly be very attractive to judges looking for a way to punt to the legislature on constitutional claims over registration.

Likely very few people other than myself will care about this as it combines things that your average person would rather not think about: “sex offenders,” research briefs, footnotes, and the federal government. I also realize we are now in an era where SCOTUS openly ignores the factual record in cases to reach a desired conclusion (to say nothing of arguably abandoning stare decisis), so my concern about proper attribution in footnotes is kind of cute and anachronistic. But all the same, I still feel it’s worth spending my Saturday morning to spill some ink about it.²

1. For example, while the FRD brief considered four studies that it carefully selected, this 2021 meta-analysis considered 18 studies comprised of nearly half a million people and concluded that registries have no impact on recidivism rates. I realize the 2021 analysis is after the 2019 cut-off, which and it is plausible 2019 was picked to not have to reckon with this study (and others). Happily enough, the 2021 analysis includes a whopping 7 studies that were not included in the FRD brief but that otherwise fell within FRD’s self-selected window — but that they appear to have simply ignored.

2. I considered not writing about it for fear that it would make its way into government hands and then government briefs before realizing (1) no one will probably read this to begin with, get over yourself Hemingway and (2) it’s already in government hands given that it was published by the government.