Friday, June 27, 2014
Harmon on the fragility of knowledge in the Riley (cellphone and 4A) case
Prof. Rachel Harmon from UVA had an interesting post to the crimprof listserv that I thought warranted broader exposure, so with her permission I'm sharing it. (Rachel asked to also thank UVA law librarian Kent Olson for his help with the underlying research).
In light of the likely significance of the Court's opinion in Riley v. California, I may seem obsessed with the trivial, but I can’t help but note the Court’s odd support for one of its statements about policing, and the pathetic state of information about policing it reveals. On page 6, the Court states that “warrantless searches incident to arrest occur with far greater frequency than searches conducted pursuant to a warrant.” Though the proposition seems intuitively obvious, data on searches and seizures isn’t easy to find, so I was curious about the Court’s support.
Chief Justice Roberts cited LaFave’s Search and Seizure treatise, which struck me as an odd source for an empirical claim, so I looked it up. LaFave does indeed say, “While the myth persists that warrantless searches are the exception, the fact is that searches incident to arrest occur with the greatest frequency.” But that sentence has appeared unchanged since the first edition of the treatise in 1978. And LaFave’s support for the proposition is itself pathetic. It comes in a footnote which reads: “See T. Taylor, Two Studies in Constitutional Interpretation 48 (1969). ‘Comparison of the total number of search warrants issued with the arrests made is equally illuminating. In 1966 the New York police obtained 3,897 warrants and made 171,288 arrests. It is reliably reported that in San Francisco in 1966 there were 29,084 serious crimes reported to the police, who during the same year obtained only 19 search warrants.’ Model Code of Pre-Arraignment Procedure 493–94 (1975).”
Because I’m crazy, I pulled Taylor and the Model Code too.
Both sources suggest that they can’t really prove the original point. Taylor says, “[M]ost law enforcement agencies have been exceedingly lax with their record-keeping in this field. But there a few offices where the records are full enough to be meaningful, and from these it is abundantly apparent that searches of persons and premises incident to an arrest outnumber manifold searches covered by warrants.” He provides no further support for the claim.
The Model Code Commentary provides the numbers from 1966, but also makes it clear they are not to be taken too seriously. The New York data was apparently furnished directly to the Code’s Reporters from the NYPD, and the San Francisco numbers came from a New York Times’ reporter. (It was Fred Graham, the Supreme Court correspondent at the time and a lawyer.) According to a footnote to the Commentary, “Research efforts elsewhere foundered on the rocks of record-keeping failures. Law enforcement agencies do not commonly maintain statistical records pertaining to search warrants or searches and seizures generally.”
So the Supreme Court cited a source, unchanged since 1978, which cites two sources from the late 1960s, both of which suggest that there is very little evidence for the proposition because police record keeping is weak. I’m hardly one to criticize imperfect footnotes (since I’ve surely written many myself), but this one interests me. The Court is all too willing to make unsupported claims about policing, a problem I’ve noted before. See The Problem of Policing, 110 Mich. L. Rev. 761, 772-773 (2012). Moreover, for the Court, as well as scholars and policymakers there is a serious problem in finding credible information about what police do. See Why Do We (Still) Lack Data on Policing?, 96 Marq. L. Rev. 1119 (2013). The Riley/Wurie citation nicely illustrates both problems, and it won’t be the last to do so.