Loretta Lynch’s Hot and Cold Running Data-Sharing

[See update below: Lynch says she didn’t mean how these statements came out.]

It’s bad enough that Attorney General Loretta Lynch refuses to force police to keep records on how many people they kill.

In a conversation with NBC journalist Chuck Todd on a range of criminal justice issues, Lynch said on Thursday that she does not support a federal mandate to report people killed by police.

“One of the things we are focusing on at the Department of Justice is not trying to reach down from Washington and dictate to every local department how they should handle the minutia of record keeping, but we are stressing to them that these records must be kept,” she said at the Washington Ideas Forum, hosted by AtlanticLIVE and the Aspen Institute.

It’s her reasoning I find really troubling.

Lynch said the Justice Department does “encourage” local departments to maintain records on police shootings but that improving police-community relations is more important. She noted that the small size of the average police department could make record-keeping difficult.

“The statistics are important, but the real issues are: ‘what steps are we all taking to connect communities … with police and back with government?’” she said.

It’s all well and good to say communities and their cops just need to get along.

But cops are claiming a Ferguson Effect that statistically doesn’t exist and the NYT is reprinting the claim only because the cops say so.

Here’s what the crime story said: “Among some experts and rank-and-file officers, the notion that less aggressive policing has emboldened criminals — known as the “Ferguson effect” in some circles — is a popular theory for the uptick in violence.” A paragraph later, the story continues: “Others doubt the theory or say data has not emerged to prove it.” Two experts are quoted, and the story moves on from there.

Bill Michtom of Portland, Ore., wrote to me about it, calling it a “classic example of false equivalence.” Ta-Nehisi Coates called the suggestion of a Ferguson effect “utterly baseless” in a piece for The Atlantic, noting that one of the experts quoted said that the rise in violent crime in St. Louis had begun before the large protests last year over a white police officer’s fatal shooting of an unarmed black teenager.

One of the story’s reporters, Monica Davey, and the national editor, Alison Mitchell, strongly disagree that this is false equivalence or that it was misleading to readers. In fact, they told me, it would be wrong of The Times not to report something that some police officers are identifying as part of their mind-set.

Ms. Davey, who agrees that false balance is infuriating and must be avoided, said in an email that this example simply doesn’t fit the description. For one thing, she said, there is no established truth here: “The question about the validity of this theory simply has not been definitively answered in the way that the earth’s shape has.” And, she said, “police officers must be given some credence in assessing whether they themselves feel that they are behaving differently now — the essence of what some of them have called the ‘Ferguson effect.’ ”

Or, as Ms. Mitchell puts it: “We have the police suggesting that police are pulling back — should we not report that?”

My view is that the introduction of this explosive idea didn’t serve readers well because, in this context, it was mentioned briefly, sourced vaguely, and then countered by disagreement. If police officers are indeed pulling back from their duties, and are willing to be identified and quoted, and if there’s evidence to back it up, that would be worth a full exploration in a separate article. But this glancing treatment could easily have left readers baffled, at the very least.

Things aren’t going to improve so long as cops can just make shit up, in spite of data to the contrary.

Just as importantly, since 9/11, the mandate throughout the Federal government — and especially for FBI — has been to share information promiscuously, including down to local police departments. Some of that information includes untested leads; some of it includes cyber and terrorist threat assessments.

If Lynch is telling us these local police departments don’t have the ability to handle reporting back and forth from the federal government, than the rest of the info sharing should stop too, because it could violate Americans’ privacy and/or expose intelligence streams.

But we all know that’s not going to happen.

Which means Lynch is supporting an asymmetrical reporting system that can’t be used for oversight of the larger system.

Update: Lynch says her statements last week weren’t what she was trying to say.

The point I was trying to make at that conference related to our overall view of how we deal with police departments as part of our practice of enforcing consent decrees, or working with them and I was trying to make the point that we also have to focus on building community trust which is a very individual – very local – practice.  Unfortunately, my comments gave the misperception that we were changing our view in some way about the importance of this data – nothing could be further from the truth.  This data is not only vital – we are working closely with law enforcement to develop national consistent standards for collecting this kind of information.

More from her statement:

“The department’s position and the administration’s position has consistently been that we need to have national, consistent data,” said Attorney General Lynch.  “This information is useful because it helps us see trends, it helps us promote accountability and transparency,” said Attorney General Lynch.  “We’re also going further in developing standards for publishing information about deaths in custody as well, because transparency and accountability are helped by this kind of national data.”

24 replies
  1. Don Bacon says:

    Why doesn’t Lynch just tell the truth, that such reporting would aggravate cops and it’s not smart to aggravate cops, no matter who you are. Cops having their way even to kill citizens is like bearing arms (in fact they are connected), to be allowed without any lefty reporting requirements (they believe).

  2. TarheelDem says:

    What’s the figurative or literal gun pointed at the head of the US Attorney General when it comes to the accountability of law enforcement? There is a reason that there are black appointees at Attorney General and Department of Homeland Security. What appears to be happening is some sort of internal revolts in both agencies with likely roots in groups promoting institutionalized racism in law enforcement. As I often say, the ghost of J. Edgar Hoover still wallks the halls both at DOJ and in his eponymous building.

    There is in a number of places a movement afoot to use those 331 million guns to trigger a race war. The impunity of the Oath Keepers and the tightness of their paramilitary operations with local law enforcement in St. Louis means that we are moving toward the sort of situation that existed in 1875-1877 when not even the military to could protect black people from terror because of the collaboration of white people with the terrorist organizations. And in that period, local law enforcement and state militia were definitely inside the terror movements like the Ku Klux Klan, White League, and Red Shirts.

  3. ne plus ultra says:

    A fair-minded person can not look at the St. Louis murder rate by month and say that it’s ridiculous to see an inflection point in August 2014.

    The City Lab piece used an odd graph to try to make it look that way – ratio of murders from previous year, which highlights a seeming huge spike in April 2014.

    This is a result, not of a spike in murders in April, 2014, but the statistical outlier of April 2013, which had the fewest murders of any month in the 2-year period.

    June 2014 was lower than June 2013. June 2015 was a little higher. After that, every month was higher.

    I can’t say that Ferguson is the cause. The trend may have started in July, or it may be attributable to other causes despite really taking hold in August. But given the spiky nature of the graph and the small increase in July, it’s also unreasonable to insist that the cause could only have been changes in drug markets in July.

    To claim it ‘statistically doesn’t exist’ is going well beyond the data. Further beyond the data than what the NY Times is reporting.

  4. P J Evans says:

    If Lynch doesn’t want those records kept, it might be because she believes the police versions of what happened – even when there’s evidence they’re making shit up. Which makes me wonder WTF she thinks her job actually is. (Since she’s not prosecuting banksters for their crimes, and not prosecuting cops for theirs, and apparently is all too willing to go after the rest of us even when it requires making shit up….)

  5. orionATL says:

    as for the “ferguson effect” it can be said with confidence that both national and local police will abuse and misuse any social science, statistical reporting, or psychological theory they can get their hands on.

    i find it useful, in part to achiev some distance from emotional entanglement, to model police and prosecutors as athletes who are obsessed with winning and who are determined to win whatever cheating may be required. obviouly not all athletes nor all police act this way, but those who publicly defend policing tactics and police misconduct, e.g. police union, district attorneys, routinely do so.

    another example of this kind of spurious justification based on psuedo-social science, is the doj/fbi reliance on radicalization theory to con juries and obtain conviction of muslims accused of tertorism. one can almost guaranted that the less solid evidence is available against a “terrorism” suspect, the more the prosecution will depend on radicalization theory.

    i don’t know of any instances, but i would expect this convenient abuse of psychology to soon makevits appearance in vourt rooms where domestic leaders, aka troublemakers, are targeted with radicalization grenades.

  6. orionATL says:

    as for heather macdonald and her institution’s defense of the extraordinary fact of race-based executions by american police,

    from misswiki:

    “…[Macdonald] is a John M. Olin Fellow of the Manhattan Institute.[12] In addition, she is a contributing editor to New York’s City Journal,…”

    and what is city journal?

    “City Journal is a quarterly magazine published by the Manhattan Institute for Policy Research,[1] a conservative think tank based in New York City. Its current editor is Brian C. Anderson. The magazine was started in 1994.[1] Myron Magnet, its editor from 1994 to 2006, is now editor-at-large. Many City Journal articles focus on urban policy. Contributors include John M. Olin and Senior Fellow Heather Mac Donald…”

    and who was john olin and his olin foundation?


    [… John M. Olin Foundation was an American grant-making foundation established in 1953 by John M. Olin, president of the Olin Industries chemical and munitions manufacturing businesses. … The Foundation is most notable for its early support and funding of the law and economics movement and the Federalist Society. “All in all, the Federalist Society has been one of the best investments the foundation ever made”, wrote the Foundation to its trustees in 2003.[1]…]

    one can debate facts and ideas in debate-club fashion ’til the cows come home, but when dealing with american right-wing propagandists it is unwise not to know the web of associations. there is very little that is spontaneous with an individual or intellectually honest.

    • orionATL says:

      further, re john m. olin:

      from misswiki

      “… Olin started his career in 1913 as a chemical engineer for his father’s Western Cartridge Company, a predecessor of Olin Industries, Inc. In 1935, following Western Cartridge’s acquisition of the Winchester Repeating Arms Company, he was named First Vice-President of the merged Winchester-Western and head of the Winchester division. …”

  7. Don Bacon says:

    I suppose that Lynch’s decision had a lot to do with the general problem of crime reports.
    In New York City, and I imagine elsewhere, cops have manipulated crime reports to show less serious crime. They have also made it difficult for citizens to report crime. “Come down to the station between 2 and 3 and fill out a crime report.” That sort of thing. Instead, the cops have concentrated on petty infractions, failure to signal, inoperative lights, stop-and-frisk, etc, just the general harassing of citizens. –Broken window crimes. The result has been that reports indicate a huge drop in crime. Success! So let’s not rock the boat now.
    The US “justice” system on major crime is entirely corrupt — why pretend otherwise. From the Constitution on down, it’s all set up to protect the favored few against the rabble.

    • ne plus ultra says:

      There’s only so much tinkering a police department can get away with in reference to murder stats. Most criminologists consider the numbers quite accurate.

      • orionATL says:

        it’s never the stats, the numbers, that are important, but the facts, concepts, and theories the numbers are said to represent.

        if the stats mix spitting gum on the sidewalk and jaywalking with robbery and arson, or less rhetorically, if they mix different levels of severity of a specific type of crime, say, assault, then the stat may convey misleading info.

        for any crime stats to be comparable across counties, cities, metropolitans, states, the categories have to be very specifically and rigidly defined and the techniques for aggregating have to be simiiarly equal across those entities.

        for stats to avoid being misleading for analyzing, e.g., one city, over time, the categories, counting, and techniques for aggregating have to be very stable from one governing entity to all others.

        without knowing this area of sociology, but knowing how any collection of numbers can be toyed with for political or propaganda gains, i’d guess there is quite a bit of tinkering police and law enforcement can get away. doing so only requires making sure the numbers agree with whatever is the desired objective.

        but with the central stat powering this discussion, how many american citizens are killed by police each year, there can’t be much doubt.

        discussing a spurious effect like the “ferguson effect” does serve the very useful purposeof taking eyes off of that cetntral issue.

      • orionATL says:

        murder ststistics can and are manipulated by officials

        – most prominently, now, police kilings declared justified

        – instances where a medical officer (coroner) declares no homicide
        – instances where a medical officer is pressured to change a homicide report
        -instances where officials, say a prison official, fail to report a hoicide

        – instances where levels of severity of homicide are mixed, e.g., negligent vs premeditated.

      • orionATL says:

        from #21 “…But it needs to be pointed out that you’re repeated insistence on the radical unreliability of murder stats isn’t a widely held opinion…”

        what needs to be pointed out is this:

        -all my responses re murder stats are responses to your #12 & #14 which are unambiguously:

        – naieve at best

        -an appeal to authority

        – quite possibly a lawyer’s willful ignorance.

        my responses #s 13, 15, 16,17, 18, & 20 are a series of connected solid responses.

        my point about the potential to manipulate even murder statistics (statistics are different from rates- rates are ex post facto the actual sausage making) is that they, and any statistic, can be manipulated. one analogy is with the manipulation (not always intentional) of publicly published opinion polls.

        as for what you label my “insistence on the radical unreliability of murder statistics”, i’ve simply pointed out potential problems with the integrity of the lynchpin (actually dependent variable) of a new urban myth, one i am beginning to suspect you clearly hold dearly, “the ferguson effect”.

        any experienced statistician or analyst worth their salt (i didn’t say their 6 figure salary) would find nothing whatsoever “radical” about my criticism of your “naievete” about police statistics.

        one thing is certain, IF murder statistics are manipulable and IF some were to have been manipulated recently, then the “ferguson effect” would be knocked into a cocked hat.

        but, funny thing, murder statistics (or is it rates?) are not what the “ferguson effect” hangs on, as i pointed out in #22.

        the “ferguson effect” is the dependent variable. the independent variable(s) is what caused the ferguson effect, if it exists.

        so there are two hurdles:

        1- has there been an increase in murders or some other crime measurse in specified localities?


        2- what caused that increase(s) if it exists? the latter are the independent variables. see #22.

        so far, the only “data” on independent varuables (causes) are police union and police self-reports transcribed by media stenographers and the chitchat of police-affairs hangerson.

        there is nothing whatsoever “radical” in my questioning of any statistics, even the police-generated murder statistics you seem worried to protect.

        what may be labelled radical by some is my insistance that this sudden-to-appear “ferguson effect”, disguised as trustworthy instant criminal sociology, is instead another urban myth, one with a clever propaganda purpose.

  8. earlofhuntingdon says:

    ““One of the things we are focusing on at the Department of Justice is not trying to reach down from Washington and dictate to every local department how they should handle the minutia of record keeping, but we are stressing to them that these records must be kept,” she said at the Washington Ideas Forum, hosted by AtlanticLIVE and the Aspen Institute.”

    Lynch’s claimed restraint flies in the face of the plethora of ways that the USG has federalized local law enforcement over the past three decades, from weapons, training and budget largesse to communications protocols and data sharing. At least since Reagan, the feds don’t collect or stop collecting data on any issue that might embarrass them.

    • earlofhuntingdon says:

      “Minutia of record keeping” is how Ms. Lynch defines sudden death at the hands of local law enforcement? If we had a press corps, it would take exception to such dismissive arrogance and claims of procedure so at odds with reality.

  9. orionATL says:

    “no one has show the ferguson effect spurious”.
    no one has shown that a concept concocted by individuals with a strong motive to concict it for pokitical reasons is worthy of serious discussion.

    with science one first has the idea, then puts the theory to test, then comes to a tentative conclusion. right now the so-called “ferguson effect” is a political concept, all propaganda and no supportingt theory or data. the cart has been put before the horse.

  10. orionATL says:

    this part of your #12 comment

    “… Most criminologists consider the numbers quite accurate… ”

    is a red flag, in this case a rather blatant appeal to authority (see #’s 13 and 15 above).

    for the record i didn’t see that bacon’s comment focused on murder.

  11. orionATL says:

    from e.w.’s update:

    “… we need to have national, consistent data,” said Attorney General Lynch. “This information is useful because it helps us see trends .. We’re also going further in developing standards for publishing information about deaths in custody as well …”

    read: oh, this knucklehead of an attorney general finally got the memo, did she?

  12. Don Bacon says:

    Regarding my comments in #9 above regarding manipulation of crime statistics by NYPD, I’ve found the article which describes it.

    The NYPD Tapes: Inside Bed-Stuy’s 81st Precinct

    Two years ago, a police officer in a Brooklyn precinct became gravely concerned about how the public was being served. To document his concerns, he began carrying around a digital sound recorder, secretly recording his colleagues and superiors.

    He recorded precinct roll calls. He recorded his precinct commander and other supervisors. He recorded street encounters. He recorded small talk and stationhouse banter. In all, he surreptitiously collected hundreds of hours of cops talking about their jobs.

    Made without the knowledge or approval of the NYPD, the tapes—made between June 1, 2008, and October 31, 2009, in the 81st Precinct in Bedford-Stuyvesant and obtained exclusively by the Voice—provide an unprecedented portrait of what it’s like to work as a cop in this city.

    They reveal that precinct bosses threaten street cops if they don’t make their quotas of arrests and stop-and-frisks, but also tell them not to take certain robbery reports in order to manipulate crime statistics. The tapes also refer to command officers calling crime victims directly to intimidate them about their complaints. . . .

  13. orionATL says:

    let’s suppose some years in the past that a police dep’t was under pressure to report a decline in crime or just wanted to make its performance look better. further, suppose that the resulting apparent drop in crime was accomplished by jiggering reports of crimes, say, police intimidating victims of robbery about their report of being robbed so the report is erased from the record or reclassified as a less serious matter.

    then suppose some years later, that that police department is publicly attacked for its abusive treatment of citizens, especially lower class blacks and hispanics.

    how then might that police department, especially its union, its command structure, and hangerson like the manhattan institute, create a public counterattack, i.e., a loud public p.r. response to allegations of brutality and summary execution?

    why, the dep’t can BEGIN ACCURATELY REPORTING THE CRIMES IN ITS PURVIEW, punctilliously recording each crime at its most severe level and then LOUDLY TRUMPET THE “SUDDEN” RISE IN CRIME.

    sweet, no? clever, no?

    well, here’s what started this fantastical thought experiment:


    and some background:



  14. ne plus ultra says:

    What you’ve written applies well to most crime stats, including most violent crime stats.

    What you’ve written doesn’t apply well to murder stats.

    That’s why when departments want to spin the facts, they start citing changes in crime stats, rather than murder stats. It’s too easy for people to zero in on exactly what they’re mischaracterizing.

    You don’t have to believe me, or every criminologist in the country, of whatever political or other background. But it needs to be pointed out that you’re repeated insistence on the radical unreliability of murder stats isn’t a widely held opinion.

    Allowing that a “Ferguson effect” might be occurring doesn’t automatically lend itself to the spin police departments put on it. (And I continue to be agnostic. It’s rather early to determine whether there is any national change. The numbers aren’t in.)

    But it would be suprising if murder rates didn’t have some relationship to broad changes in the relationships between cops and the communities they serve. That’s a key tenet of most left and liberal attitudes to crime. I think Tracey Meares in particular has made some fascinating points, focusing on the concept of legitimacy.

    It would be strange indeed if the public’s sense, and particularly the African American public’s sense of police legitimacy had not worsened in the wake of Ferguson.

    Some departments seem to draw the conclusion “See, you have to trust us or you’re screwed.” But other people would draw a different conclusion – that it will take changes in police approaches to create a calmer environment in which all communities can trust the police.

    And since I mention her and her theories, I should note that 3 months ago, Tracey Meares did not believe there was much evidence of a Ferguson effect. I’m not saying I see strong evidence. But the St. Louis data cited above isn’t sufficient to deny a Ferguson effect either. We need to be open to learn what is happening before drawing conclusions on what is to be done.

    • orionATL says:

      – what “most criminologists” think true about murder stats may correspond to reality or may not.

      – murder stats are clearly manipulable. the question is are they? and how much so?

      – my concern is not with the numbers (or any crime stats), but with the validation thru media repetion of an idea, “the feguson effect” which has no scientific validation whatsoever, an idea generated by police and for police for the specific purpose of countering unfavorable publicity over large numbers of police shootings of citizens.

      – such an origin in whst is likely sophistry does not bode well for finding out what really is happening in cities.

      – looking for the “ferguson effect” may well stir research. that might turn out to be its most positive legacy.

      – the “ferguson effect” is being expanded rapidly from merely “more murders” to “more crime”. the talk is still of murders, but the implications are “crime wave”.

      – there is no reason to think that police reluctance to assault citizens inappropriately, in particular shoot them, is causing a change in crime statstics. why in the word should it? the police are still patroling ad arresting as necessary aren’t they? or are they sulking and shirking due to pique and fear the law might be applied to a tiny proportion of their behavior?

      – apparently, the “ferguson effect” effect is a consequence of police not doing their job. or not doing it thoroughly. or not doing it violently. or not doing it menacingly enough though legally. exactly which?

      – the measurement needed is not just of murder rate – or as being expanded, serious crime, or all crime. the measurement needed involves how police are doing their job activities. and it involves not just current measurement of police job activity by city, but, to allow for comparison, pasrt measurement of police policing behavior by city.
      – have any measuremets of police policing behavior been made.? i think not. what is our source of info at the moment? why, police self-reporting, self-reporting at a time the police are under severe pressure to stop shooting citizens, especially those from the lower classes.
      – clearly the assertion of a “ferguson effect” is as i described it above, a propaganda term designed to counter criticism of and, more seriously, application of legal consequnces to pllice for police shootings.

      – the proposition:

      if there are more murders now than in the past after the public furor over the shooting of a young black man in ferguson, missouri, then that is caused by that public furor and the subsequent refusal of police to do their jobs

      is a testable proposition, but a complex one requiring measurements some of which are needed but probaby unavaiable.

      at present we are testing nothing while asserting, or at least skillfully implying, a conclusion.

      at this point, “the ferguson effect” is a propaganda/p.r. phrase, not a scientific concept. it is merely a disguised appeal to “broken windows”. any change in murder rates, or crime rates, doesn’t change that.

  15. orionATL says:

    consider “the ferguson effect” hypothesis in the light of an if-then statement in logic.

    the “ferguson effect” is clearly dependent on something causing it to occur. therefore, it is a dependent variable, a consequence of something else.

    abstractly, this can be written,

    if p, then q.

    q, therefore p.

    in our case:

    if the police change their activities, x,y,z,

    then “the ferguson effect” willl occur, i.e., murder rates will increase.

    the first thing useful to notice is that “the ferguson effect” is a substitute for “the murder rate (or some crime rate) will increase”. this suggests its propaganda origin, otherwise, why not just say “if police .., then murder rate…”, and leave it at that? why introduce ferguson at all?

    one error in reasoning that can occur with “if p, then q” reasoning is called affirming the consequent. in this situation it would be written as:

    if “ferguson effect”, i.e, murder rate up,

    therefore police changed their activities, x,y,z.

    the fallacy here is that other things could have caused the murder rate to go up (for example, the current heroin rampage) not just a police “work to rule” action.

    formal fallacies aren’t as deadly as they sound; they are a consequence of the way words are arranged. nonetheless, they offer insight into what a line of reasoning is trying to do to convince someone of, e.g., the reasoning, appearing in the media these days, especially the intellectually slovenly wsjournal.

    in summary, the usefulness of this formal exercise in logic in analyzing the usefulness of what we are hearing and reading is to indicate just how literally backwards is the reasoning supporting media palaver about “the ferguson effect”. the cart (proposition q, the consequence of someone(s) doing something) has been put before the horse (proposition p, the activity that someone did to “cause” q).

Comments are closed.