New at Huffington Post: I Make the Case for Privatizing Crime Labs

Tuesday, June 14th, 2011

Very interested to see the reader reaction to this one.

Digg it |  reddit | |  Fark

64 Responses to “New at Huffington Post: I Make the Case for Privatizing Crime Labs”

  1. #1 |  Andrew S. | 

    Given the first comment (OMG! You’re talking about giving state money to private companies to do the job!), I’m not all that optimistic for the rest.

    (and I love how that comment calls you out for being biased based on your revelation that you used to work for Reason)

  2. #2 |  SJE | 

    I agree that this is a good idea, but I don’t think it goes far enough to solving the problems with crime lab data. Even if you make the crime lab one step removed from the cops and prosecutors, you still have a single large purchaser, the state, and so you still have an incentive to keep the customer happy.

    1. Combined fund that is run by defense attorneys, prosecutors and judges to do testing.
    2. Incentives and punishments for quality, with random testing
    3. You still need clear and thorough oversight to overhaul the way forensic science is done.

  3. #3 |  Difster | 

    And the idiots are already coming out over there.

  4. #4 |  Difster | 

    I’m having fun with the idiots. In case it wasn’t obvious (Difster = Diffey)

  5. #5 |  Whim | 

    Forensic crime laboratories should be independent in both function and reporting from the police authorities. The police want convictions. Crime labs associated with police departments want to please the police administration. They have been shown to repeatedly slant, distort, spike, and lie about forensic evidence, at crime labs all over the U.S.

    Another reason why we imprison more people than the rest of the world.

    It’s a big business.

  6. #6 |  roy | 

    I don’t see the incentive for private labs to do the job well. The state will have some gatekeeper for what labs are allowed to participate, and will game the system in favor of those who don’t rock the boat.

  7. #7 |  Radley Balko | 

    roy —

    The “evidence handler” would be outside the prosecutor’s office. If a DA wants to start excluding labs, he’d presumably need to explain why publicly.

    And if you’re a private lab who starts exposing problems at the state lab, you only gain prestige.

  8. #8 |  Andrew S. | 

    Cripes. The posters there are making my brain hurt. Seems that the only people making positive comments seem to be Agitatortots or Reasoners (given that the positive comments seem to be coming from people new to HuffPo)

    (I’m wildcatlh. Don’t ask, I’ve used that ‘nym online since 1997. I can’t even remember how it started.)

  9. #9 |  Mark Draughn | 

    Yeah, you’re challenging them now. The first two posts just lulled them into a false sense of security, but now you’re trying to lead them down the path…if the government is doing things that suck, maybe the government is/isn’t (pick one) the best entity to solve the problem…

  10. #10 |  DarkEFang | 

    I don’t know how feasible it would be, but ideally, there would be two crime labs testing evidence independently from one another. And there needs to be something that dissuades those crime labs from favoring the prosecution.

  11. #11 |  SJE | 

    “And if you’re a private lab who starts exposing problems at the state lab, you only gain prestige.”

    In theory, yes. However, won’t the prosecutor merely say that the lab is doing shoddy work that was not “confirmed” by all the other labs (i.e. shills)? What is to stop the prosecutor from finding all sorts of “violations”?
    How do you make this different from the individual in the state crime lab who gets railroaded for standing up?

  12. #12 |  Leon Wolfeson | 

    I don’t see a case at all for adding for-profit costs to the system. A country-wide standard of testing and proper system of auditing and checking is needed.

    Private companies add complexity into the system, no more and no less, and that increases – not decreases – the chance for screwups. If outside verification is needed, use another government lab in another area (as an anonymous sample, of course), which use the same procedures.

    (Unless the procedures are *identical*, it’s not convincing proof – a difference in assay chemicals, for example, between labs can lead to “variant” results)

  13. #13 |  Mark Draughn | 

    Side Note: We should probably start friending each other at HuffPo, so we can get those cool badges. Makes us look less like a bunch of outsiders. Or more like an invading army.

  14. #14 |  James | 


  15. #15 |  Radley Balko | 


    Why do you think a government lab is less likely to screw up than a private lab?

    And what about jurisdictions (most of them) where there’s only one government lab?

    And simply having uniform standards doesn’t preclude bias. Bias creeps in in more than just the choice of which procedure you’re going to use.

  16. #16 |  jb | 

    “a difference in assay chemicals, for example, between labs can lead to “variant” results”

    If we must use the same procedure so as to assure against any variant result it could happen that we are just duplicating a false result which the use of the other procedure would detect. The purpose of secondary testing is to verify, not duplicate. The secondary test is successful if it finds variation. The variation can be because the first test was performed improperly, the second test was performed improperly or because the evidence does not support a positive result under both tests; regardless of the reason variation will call into question the evidence, a good thing when it is being used by the state to convict a person. Futher testing becomes appropriate.

    I, for one, would not feel better knowing that I had been falsly convicted by evidence tested in dulicate because another test could have lead to a variant result.

  17. #17 |  CTD | 

    Shhhh. Radley = deep cover Kochtopus mole at the HuffPo. Don’t blow his cover.

  18. #18 |  J H | 

    I worked at the Michigan State Appellate Defender’s office as an unpaid volunteer on the crime lab project. We would review cases where forensic evidence was heavily relied on to see about retesting the evidence. What a dark cloud of shame this whole city sits under, having locked up people who *may* be innocent and leaving them no possibility of exoneration because of mishandling of evidence.

  19. #19 |  ktc2 | 

    This seems . . . wrong:

    ” including in the choice of which tests to run, how to records the results of those tests, how to interprets the results”

    Shouldn’t it be record the results and interpret the results?

  20. #20 |  ktc2 | 

    Sorry, I was an editor for a LONG time.

    Great article though! I think that would be a really good start.

  21. #21 |  Leah | 

    I’m just hoping someone helpfully sends you copies of your old crime lab for Reason asking if you’d seen the research.

  22. #22 |  ClubMedSux | 

    There are only three things a HuffPo commenter needs to make a sentence: a noun, a verb, and “the Koch Brothers.”

  23. #23 |  johnl | 

    Using private labs would have some marginal advantages. But procedural changes you’ve written about at reason, such as blinding, would be more important. The difference between “the boss wants you to say X” and “the customer wants you to say X” is pretty small.

  24. #24 |  André | 

    Haha, the same people who were lauding you for exposing the murderous cops who Diallo’d Jose Guereña are now crawling out and attacking Big Business Evil Corporations Oh My! You should probably follow up with some puppycide in your next article to return to your place as the abusive boyfriend of lefty comment threads. I’m still waiting for the day when people say “well, I’m not sure what I think about private crime labs, but I respect a lot of the writing Radley has done on criminal justice so I’ll give this a chance before rejecting it outright based on my existing biases.”

    Then again, when I read articles about how “society needs more Jesus/State-provided magnanimity!”, I also am pretty quick to reject it because it’s bullshit. So I dunno.

  25. #25 |  GreginOz | 

    Wow, HuffPo is SCARY! Such faith in the gummint…dumbing down really WORKS. ( Not, in any way, a chem/physics bloke but I still remember sumtink ’bout the Scientific Method; big wordz loik Empirical and Measurable EVIDENCE, which includes Archiving (i.e. preserving evidence), Peer Review (no cheating;-) and, um, Objectivity. See Wikipedia definition! How in hell is a lab, sucking off the teat of a police dept. objective?
    I recall a lewrockwell article, a psych study, that noted an inability of people to even consider, let alone examine logically, ideas or memes that contradicted that person’s world view, or threatened certain tenents of that person’s belief system. I think Radley is giving a whole new arsehole to his new audience, as shown by their comments!

  26. #26 |  Difster | 

    I just got back after being out for a few hours and checked in on the updates. Those people are freaking insane.

  27. #27 |  johnl | 

    What’s needed is a sanction for whores of the state (apologies to the other kind). Require prosecutors to use labs that do a maximum of 30% of their work for the state. If a lab is caught timting the scales for the state, the state might want to keep using them, but the plaintifs’ bar, insurance companies, and everyone else would be forced to stop. A mechanism that forced prosecutors to stop using a lab when everyone else does would leave little room for fudging.

  28. #28 |  Anthony | 

    It didn’t take very long in the HuffPo comments for someone to say that the labs need to be “fully funded”. Liberals always think that government fails because it doesn’t have enough money. That and the wrong people were/are in charge.

  29. #29 |  DarkEFang | 

    #13 – Mark Draughn

    I just friended everyone in that thread that I recognized.

  30. #30 |  Maggie McNeill | 

    The reaction of the Huffington Post commenters is like what one might expect after giving a biology lecture to a roomful of creationists. :-(

  31. #31 |  C.E. | 

    There are private companies that do forensic analysis. Their customers tend to be defense attorneys and lawyers involved in personal injury suits. Some of them are licensed to test controlled substances, and with a court order, the DEA will actually deliver substantial amounts of drugs to them to test independently. To me, the key would be having more than one lab, and having all labs available both to defendants and the Government. This would reduce somewhat the tendency to associate with one side–it would become just a job, after all.

  32. #32 |  Adam | 

    Ahhh, yes, the Huffington Post……I love this comment on your story, made by a HuffPo “community moderator”, no less:

    “Bull_shlt, this is just more “profitiza­tion” of our government­.”

    You’re a brave man, Mr Balko, for wanting to share your nut-punches with this crowd.

  33. #33 |  M | 

    It might be harder for a third party like the innocence project to get a private company to respond to records requests. And there’s a real risk of a private lab monopolizing local markets and developing the same codependent relationship that makes conflicts of interest so prevalent now. That is, I don’t really trust Quest or their competition all that much already so I’d rather see smaller new companies develop

  34. #34 |  Andrew S. | 

    My goodness. I think this comment just made my brain try to escape and run away.

  35. #35 |  Curt | 

    Very amused by reading the comments at HuffPo. Lots and lots of whining that private companies are motivated by their bottom-line instead of the “public good”. The current system is one that’s motivated by the “public good”… how’s that working out for us? Law-and-Order politicians take the word of cops and decide that locking up anyone accused of a crime is in support of the “public good”. The job of the crimelab is to prove that the prosecution is correct.

    The whole damn point of this is that private companies are motivated by their bottom-line. That’s the reason it has a chance to work. Currently, cops, prosecutors, and crime labs have no accountability. They can’t be held responsible in any way. Private companies who screw up would lose money and eventually their contract. If they withhold exculpatory evidence, they can be thrown in jail. That’s their incentive.

    Having said all of that, I still have concerns about the concept based on the fact that eventually some part of the system (Evidence Handler) reports to some portion of the government (if not the DA, then the mayor/city councilman/governor etc.). Someone in the system will report to an elected official who was potentially elected based on his “get tough on crime” stance. Making this system work depends on a solid method of insulating the crimelab system from the political whims.

  36. #36 |  Boyd Durkin | 

    I think this deserves a slow clap for trolling HuffPo readers and getting them all flustered…except it isn’t trolling when it is a legit good idea. This is how 99% of discussions go when you dare to present privatization ideas or anything that isn’t just a rehash of how to “fix” the existing state agency (with lots more money).

    Well done.

  37. #37 |  Mattocracy | 

    I like how all the commenters at HuffPo think that have another government agency to half-ass oversee another government agency is going to do the trick. Cause that’s been so effective in the past.

  38. #38 |  Mattocracy | 

    In theory, if a private crime lab fucks up and sends an innocent person to prison, does that mean the lab can be sued? Or is it going to get immunity?

  39. #39 |  Yizmo Gizmo | 

    There’s an unflappable reporter named Joe Neff in Raleigh/Durham NC who
    exposed the SBI here in NC, a crime lab that was just a puppet of
    law enforcement and the shite hit the fan when CNN’s
    Rogue Justice showed what happened to that guy
    who got convicted for murder on a mud splotch misidentified as blood.
    Then a story on another guy, a poor black drifter, in the area who got framed and conveniently SBI manufactured some evidence to support the case.
    SBI is getting relocated, I hear, and changing its name.
    Maybe no one will remember how many people they screwed.

  40. #40 |  Leon Wolfeson | 


    I’m arguing for uniform standards. That means, also, uniform training and standards, which are easier to audit. It doesn’t mean you’re less likely to have issues, it means issues are easier to catch. More, you can have a standard way of anonymizing samples – slashing bias.

    And yes, you may need to ship samples around to other areas for testing. Well, with uniform procedures, this is considerably easier. And samples from other labs /should/ be tested on a regular basis, to ensure standards.

    Having it as a national system also insulates the system from local politics, and lets the standards be set by a board of the top people in forensic science, nationally.

    As usual, if you want to argue that it should be a non-profit / company limited by guarantee*, then I’m quite willing to listen. I just don’t feel that it’s appropriate to have to pay for a profit margin for services which the government will always require.

    (*Like the UK examples of Network Rail, LINX or Nominet: holding important parts of infrastructure, with government oversight but not control)

  41. #41 |  Mattocracy | 


    You make it sound like that profit margin is going to greater than what the government doles out already in salary and operation costs for its own labs.

    I don’t think that is a fair assumption. And even still, if there is a profit margin that increases costs, this is about whether a person goes free or is imprisoned. The cost is worth it. Especially with multiple labs doing the work.

  42. #42 |  Irving Washington | 

    Great piece, Radley. Just FYI: you have some grammatical errors in the 8th paragraph. Looks like you or some editor changed “he” to “to.”

  43. #43 |  Cyto | 

    I got ‘fanned’! I got ‘fanned’! Yippie! I’m so excited!!

    Oh, wait a minute…. it’s just DarkEFang. When I made a HuffPo account just to post on Radley’s story I thought I might get to hang with the cool kids. Instead it’s just me and DarkEFang hanging in the corner with the other “why do you want to run other people’s lives?” misfits. We should all die our hair purple-black and wear black nail polish and black boots. Real Doc Martens too – not those lame Sketcher’s knockoffs. ‘Cause then we’d be sooo effin’ cool…

  44. #44 |  Nick T. | 

    These are good points Radley, and the overall point is really much more important than the specific solution. The bottom line is that their needs to be multiple labs, and reviews and a dismantling of the perverse incentive system. I have concerns that privitization would lead to cronyism, as labs are excluded for pretextual reasons when its obvious they destroyed a high profile case, or where people, driven by the same motivations that make others lie on the witness stand – to get the “bad guy” – leak information as to which cases they are peer reviewing at their private lab to the boys in blue so they know when they can take liberties.
    Part of being a libertarian -at least for me – is being skeptical of government to implement even the clearest and best-designed solutions.
    But all in all a very persuasive case that we need to get things back to real SCIENCE especially if we’re going to present it in criminal trials as such.

  45. #45 |  DarkEFang | 

    Who says I don’t already dress like that?

  46. #46 |  Chris Rhodes | 

    Agitator posters need to form our own comment bloc on HP. I’m under “Saro”, if anyone cares.

  47. #47 |  Jim Collins | 

    We still haven’t addressed the main problem. Who determines what evidence goes to the independant lab? We need to setup a system, where the Police aren’t agents of the Prosecutor.

  48. #48 |  Joshgeek | 

    Had to go on and try to set some foolish statists straight after reading some of the dribble the HuffPo commenters were spewing. It’s like they have an aversion to the term “private.” Thanks for the add, Dark. I’m psyched to be in teh kewl club in the corner with u and cyto. LOL. We’ll have to be vigilant in our lurking defense of Radley over there. Lots of work to do in that regard. I agree that the issue isn’t necessarily how to address this problem specifically as much as just exposing the issue to as many people as possible.

    Btw, if anyone wants to add me over there, it’s “jgeekrock”

  49. #49 |  Leon Wolfeson | 

    Mattocracy – The question not if the margin is “worth it” in justice terms (because you certainly won’t find me arguing against proper testing), but if it is reasonably avoidable by having government/not profit labs do the testing. I strongly believe that it is.

  50. #50 |  c andrew | 

    How about using tao as an identifier. For the Agitator Original. Everybody else would just thing we’re into eastern philosophy.

  51. #51 |  Cyto | 

    Win The Future!

    I responded to a HuffPoBot’s snark about rich criminals buying off labs and actually got cogitation:

    383 Fans
    18 hours ago (7:23 PM)
    makes sense, thanks for explaining­.

    Thanks HuffPo Super User mabinog! One of my favorite things in the entire world is to have my preconcieved notions proven wrong. You just proved me wrong about the HuffPoletariate.

    {Arlo Guthrie voice} And if one HuffPo superuser, just one HuffPo superuser was to listen to a libertarian argument….. Well, they’d think he was crazy and de-fan him. But TWO! If two HuffPo superusers were to listen to a libertarian argument…. well, they’d think they were gay and they’d make them go sit on the group W bench. But THREE.. That’s a movement!

  52. #52 |  Curt | 

    I love the comment from jeffrey678 at 11:35 am. It’s the perfect example of a mindset so completely different from the average agitatortot.

    Private labs make corruption worse… contractors are easier to intimidate. His solution, separate labs for quality control… one belonging to the state gov’t and one belonging to the federal gov’t.

    Government is inherently “good”. Private is inherently “bad”. On the extremely rare occasion that the government does something bad… the answer is to provide more government oversight.

  53. #53 |  plutosdad | 

    I just got done reading The Poisoner’s Handbook, and was amazed at the dedication to science and truth exhibited by Norris and Gettler, even to the point of testifying for the defense in cases.

    One of the important points was that before Norris, American coroners were political appointments who put truth far below their own pensions and political power.

    What happened to end the reforms and example those two set, where modern investigators don’t even try to blind their experiments to make them unbiased, and in the worst cases are little better than paid informants police use. Is it lack of money for doing proper experiments? is this because they have to work closely with prosecutors and police so they end up gravitating that way? Is it because their bosses are all politically appointed, and want to keep their jobs and power?

  54. #54 |  Mattocracy | 

    You’re never going to completely remove corruption from a government action. Even outsourcing still has someone in the government overseaing/giving out contracts/ etc.

    But it’s easier to fire a private entity for negligence than a government employee or agency. The government has a built in bias to do what is in its interest and find evidnce to prove it. A third party doesn’t carry that bias as readily.

    You might say that having a contract with the government might induce a private laboratory to skew results, but multiple labs, maybe private and government run, would start producing vastly different results if someone was screwing around.

    You can’t do that when your sample is one. You need a larger sample size of independent people doing the same work.

  55. #55 |  Leon Wolfeson | 

    A network of government labs, using identical testing methods, is much, much easier to monitor – and the same level of monitoring is required, of course – than a network of private labs, which have different standards and testing methods.

    “Even” outsourcing? No, especially outsourcing, because not only do you have to pay profit margins, plus people to watch the people auditing the labs, and you have to develop procedures based on a wide variety of labs…it’s extortionately expensive.

    Outsourcing is appropriate only when you can make major cost reductions (which is NOT going to be true for lab work, if it’s to be reliable), or when it’s not worth developing the capacity in-house (there’s a constant demand for forensic lab work).

  56. #56 |  yonemoto | 

    It’s distressing, too that at HuPo they automatically equate Privatization => for-profit => evil-profit-stakers.

    There’s no real reason why a crime testing lab would have to be a for-profit entity.

  57. #57 |  Mattocracy | 

    The Independent labs don’t have to have different standards. They just need to be able to double check each others work. You’re still making a lot of assumptions where they aren’t necessary.

  58. #58 |  Leon Wolfeson | 

    You’re making the assumptions that private labs, which don’t necessarily have the same hiring standards, equipment, procedures and so on WILL work in a uniform manner.

    This isn’t born out by current labs, for that matter. And by the time you’ve foisted a detailed manual on them, you’re simply paying a premium for something you could of done in-house with several layers less bureaucracy.

  59. #59 |  Cyto | 

    #55 | Leon Wolfeson |

    Yeah! Why would we outsource important testing!?!? Would you trust your medical lab tests to a private outsourced lab? Of course not! Those tests have life-and-death consequences for you. You’d never send your blood tests, or infectious disease culture, or MRI evaluation to a private company. What if their profit motive caused them to cut corners? How could you ever trust the results?

    In light of this, why on earth would anyone propose sending forensic evidence to private labs for testing??

  60. #60 |  plutosdad | 

    “You’re making the assumptions that private labs, which don’t necessarily have the same hiring standards, equipment, procedures and so on WILL work in a uniform manner.”
    No, actually rather the opposite will likely be the case. The exact procedures re: the science don’t matter as much as the results. If the results cannot be replicated by other labs, then the one that keeps getting things wrong will get booted. You don’t want politicians setting scientific procedures. They need to set procedures on how to handle evidence regarding privacy. It is in a lab’s own best interests to handle things to make sure to not contaminate samples, etc. That the labs can do themselves, most already do.

  61. #61 |  SJE | 

    Both liberatarians and liberals sometimes get stuck in absolutes of whether government or private industry is better.

    There is abundant evidence that the private sector CAN do all sorts of things better and cheaper than the government, and pay its people better. Things like better training, management, and use of technology, etc. Not all the time, mind you.

    When you have clear evidence of failure of government, it is time to consider alternatives. Cato showed that the DC Public Schools spent as much as tuition to elite private schools yet got terrible outcomes. Enter renewed push for charters and reform of the DCPS. Here, you have decades of corruption and inefficiency at crime labs, and people imprisoned on the basis of such labs. Time for a change.

  62. #62 |  Leon Wolfeson | 

    Sure, and I am arguing for a change, SJE: moving /away/ from locally run labs to a national network.

    The private sector can do better in some situations, a far from exhaustive list of factors being where full transparency and reproducibility is not required (not applicable), where need is variable (it generally isn’t), where only a few firms can handle the specialist needs of the field (not applicable), where there is strong price competition (not applicable), etc.

    You WILL end up simply paying a premium for a service which requires more cross-checking, and more overhead on the oversight.

    Plutosdad – I think you missed where I proposed the tests and standards were set by the top people in the profession, nationally. If there’s a standard procedure, it can be defended against tampering by local factors and politicians far more easily than isolated testing labs.

  63. #63 |  plutosdad | 

    @62 Leon Wolfeson: “I think you missed where I proposed the tests and standards were set by the top people in the profession, nationally”

    And how well has that worked in other cases? Anyone in a top position at the FDA that says marijuana is not that bad as alcohol and other drugs gets fired. The morning after pill was not approved for years, not because it is dangerous, but because of politics. You CANNOT remove politics from the equation.

    Also, a national standard is not necessary, it also interferes with the introduction of new techniques. If Gettler had to apply to change standards for every new test, criminal forensics would never have been overhauled and improved in this country. Labs already have standards, and with multiple labs, those with the best standards will produce the best results.

  64. #64 |  Leon Wolfeson | 

    plutosdad – That’s why I’m saying the body should be independent.

    And blow that – changes to procedures *done in state labs* which are used to test crime scene materials SHOULD have to go through standards. The defence can still do their own testing, with new techniques, privately.

    “those with the best standards will produce the best results”

    What rot. There are results which are correct for the test, and there are results which are wrong. Consistency is more important than a hypothetical “bestness”. You can be sure that “best” is ruled by the political filter you were complaining about as well…