Murphy Argues for Private Police Services...

because the state's police screwed up.

But when policing was private, how did it do?

It turns out the English homicide rate in 1200 was roughly 20 times today's rate. (London had a record-low number of homicides in 2012, by the way.)

All that decline came during a time of pretty continuous state growth.

Correlation is not causation, and so on, sure, but does anyone really think this is just a coincidence, when we have such an excellent theory as to what happened? (The war of all against all, leviathan, etc.)


  1. In the limit each police force has size one, and each citizen is both client and enforcer. Predictions?

  2. Anonymous4:48 PM

    Well, wasn't all violence greater then than it is today? It would make sense to control for that, no?

    Further, what significance does police organization have with regard to actual homicide rates? If you're going to kill somebody I don't think that most people will weigh the pros and cons of whether the police force is public or private. Are you trying to say that if a police force is private that all of the sudden everybody says, "ah, now I can finally kill that bastard?".

    The problems with your point of view here are far more varied than merely correlation doesn't imply causation.

    1. "Well, wasn't all violence greater then than it is today? It would make sense to control for that, no? "

      This high rate of violence is EXACTLY what Hobbes says we get without Leviathan, and is just what causes the high homicide rate! That you think we should "control" for this is stunning.

    2. Anonymous12:43 PM

      You were talking about private police and violence, so yes, one should control for the more general trend. How is it "stunning" to say that? If instead you were talking about statelessness and violence, then you should have been more clear. These are separate arguments. However, I still think that attributing the general decrease of violence throughout human history to the state ignores innumerable other factors, as well as ignoring that the state is the product of society (not the other way around).

    3. Why not 'control' for the growth of the central state?

    4. Anonymous8:58 PM

      Control for whatever you want. I'm just saying that if you wish to reveal a trend on a specific case, then you must also control for the correlative general trend in order to differentiate the data.

      It sort of reminds me of how pro-gun people are now pointing to the data showing a general decrease in violence since 1993(?) and showing data of how gun sales are increasing at record numbers, then concluding that guns are reducing violence. I think that people should be able to have a gun if they want one, but I still don't buy this argument.

      While I still think there is merit to what I said, my original intention in those first two sentences above was to just mess with Gene a little. I know that he already buys into the idea that violence has decreased historically due to the state, but here he is talking about a single service that the state provides. These are two separate arguments, so how does one differentiate the data?

      We all accept that violence has decreased. Some believe it is due to the state, others do not. However, for those who believe that the state is the reason for the decline, then they can pretty much point to any increase in state-provided services to support their position (which is what I see Gene doing, as well as blurring the lines between the two separate arguments). Shoot, in some twisted way you could say that war has decreased violence.

      I admit that the state has had influence along with innumerable other things, but I deny that it is the primary. You can quibble about what discrete changes to the trend a state vs stateless society will have, or even amongst different systems of states, but I still think that the general trend would have still been an overall reduction in violence. The state is just a part or product of society, it isn't society, and most people don't even pay it any mind (whether this is good or bad is up for debate).

  3. I think this says less than you think it does. When you make these time series analyses, you have to control for other factors which can also account for a decrease in the homicide rate. This includes the possibility that policing, in general, was very ineffective ~1600, and policing/security, in general, gradually became more effective with new technologies. So, it may be the case that in modern day private security wouldn't fare as poorly as that yr. 1600 figure suggests.

    1. Well, Jonathan, I don't think it says all that much! Of course a lot more was going on in this period than just two changes.

      What I do think is that it says a a heck of a lot more than Murphy's single anecdote does!!

    2. Also, the long term trend in falls in homicide rates was most pronounced from about 1300 to 1700, where rates fell from about 50–100 per 100,000 to about 10 per 100,000 (Pinker, Steven. 2011. The Better Angels of our Nature: Why Violence has Declined. Viking, New York, NY., p. 63). Furthermore, for the period before 1800 the current data from historical records are likely underestimates, so the downward plunge was probably much greater.

      Also, you can control for other factors: e.g, it was not, say, modern medicine that caused it, for effective, science-based medicine is an invention of the 20th century, and most of the decline happened before 1700 when medical science was very crude, and doctors possibly “killed as many patients as they saved” (Pinker 2011: 62).

      The big factor is the rise of the Leviathan and national state: it brought an end to feudal system, and "nationalised" justice, and made murder a criminal offence punished by the state, rather than just a private tort, or private law offence.

      Gene Callahan, if you haven't read the relevant chapter in Pinker's book, I summarise it here:

      It's a great book.

    3. "This includes the possibility that policing, in general, was very ineffective ~1600, and policing/security, in general, gradually became more effective with new technologies."

      As a matter of factor big declines in homicide rates had already happened by 1600. What new technologies revolutionised policing in the 1400s or 1500s? I can't think of any.

      Medical science did not do it either.

  4. I also oppose monopolies in electric utilities. There were no government-backed utilities in the year 1200, and yet indoor lighting sucked! Sorry, competition fails again.

    1. Very interesting point Bob. Let's see how it works out for you:

      In 1200, we lacked today's technology of lighting, so of course since then the quality of indoor lighting has gone way up.

      In 1200, we lacked today's technology of violence, so of course since then the murder rate has gone way...

      Oops, it is down 95%. This is not exactly working out in your favor.

    2. Most people want more light (there are some complaints about light pollution, and technology could help with that). Most people don't want more violence. The analogy doesn't completely hold.

    3. TGGP, no analogy "completely holds," or it would be the same case, not an analogy!

      But clearly, some people in 1200 liked violence. By 2000, those people had a LOT more technology with which to "provide" it. And yet homicide dropped by a factor of 20.

  5. Anonymous4:24 AM

    decide if one thinks this is bad or good, etc.


Post a Comment

Popular posts from this blog

Central Planning Works!