A quick thought on hot spot policing and agglomeration economies

Over on Slate, Matt Yglesias, one of the nation’s leading voices on urban economic issues, notes that “hot spot” policing, or sending more police to specific high crime areas, actually seems to reduce crime and not just displace it, contrary to the expectations of some.  This is one of the findings in Frank Zimring’s fantastic book on crime in New York City, The City That Became Safe, which discusses how crime fell in New York faster and for longer than it did anywhere else in the country (you probably know about the murder rate, but here’s an amazing fact: auto theft fell by 94% from 1990 to today! AOL dial-up usage hasn’t fallen that far from its peak!).  One of the answers Zimring gives is that the police focused resources on specific problems, particularly outdoor drug markets.

Yglesias notes that, while it makes sense that “hot spot” policing would just displace open air drug markets to other places, it does not.  He doesn’t offer a theory why.   Here’s one – drug markets feature heavy gains from agglomeration.  For those of you who don’t know, agglomeration economics is the field that studies why economic behavior tends to clump together in cities, despite higher rents (Here’s the great Ed Glaeser, who with Paul Krugman, is the best known scholar in the field, explaining the basic types of gains from agglomeration).   Two agglomeration based stories can help explain why open-air drug markets exist and why disrupting them may reduce overall crime.  The agreement among dealers to locate in a specific place creates market depth, which allows consumers to pick among them, getting gains from specialization (different dealers may have access to different types of drugs, or drugs of different qualities and prices) and insurance (buyers know someone will sell them drugs in a big market, as opposed to counting on the reliability of a single drug dealer if they go to his corner).  And there is surely a great deal of learning among drug dealers, about how to avoid cops, how to deal with tweaking clients and the like.  And there are some economies of scale in providing common services, like security.   When a drug market gets displaced, the assumption that some other place will just pop up in equal size ignores that it is difficult to coordinate all of these people to show up in some other park (there isn’t a drug dealer trade group to house these negotiations, after all).  There are reasons to co-locate but coordinating everyone to move to one place is just really hard.

As Dan Rodriguez and I show in our piece The Location Market, the type of thinking that led people to believe shutting down open-air drug markets would have no effect pervades urban economic policy.  For instance, one of the central arguments for the congressional Height Act, which limits the height of buildings in Washington D.C., is that, if there are no skyscrapers in downtown DC, economic activity will move spread easily from K Street to Cleveland Park or Anacostia.  But it doesn’t – restricting supply just causes rents just go up downtown, which is now one of the most expensive office markets in the country.  The reason is that lobbyists want to be near other lobbyists.  Co-location provides easy access to “upstream and downstream linkages” (from restaurants set up for deal making to access to a deep market for temps) and most importantly information spillovers or learning from peers (As Rodriguez and I write, “A lobbyist talking to another lobbyist about congressional procedure is producing information spillovers that will improve the listener’s productivity at work. A lobbyist talking about congressional procedure to just about anyone else is a bore.”)

Effectively, the reason hot spot drug policing might work is the same reason the Height Act is bad.  Real estate is all about location, location, location.


You may also like...

5 Responses

  1. Orin Kerr says:

    Very interesting, David. Thanks for the links.

  2. Mike Madison says:

    I am always thrilled to see the phrase “agglomeration economics” in a blog post (really), but a distinction is in order: Not all cities display equal gains from agglomeration, and gains from agglomeration do not depend solely (or even primarily) on location-based solutions to coordination problems. Individual preferences related to agglomeration vary, firm sizes vary, industry structure varies, etc. Some cities (markets) display few gains from agglomeration. All of this is no news to Glaeser. For some useful supplemental reading, which Glaeser cites, take a look at Ben Chinitz’s 1961 paper, “Contrasts in Agglomeration: New York and Pittsburgh.“. One key point of that work is that using agglomeration economics to study drug policing or zoning law or any other aspect of urban policy requires an account of individual behavior.

  3. David Schleicher says:

    Thanks for the comment. Not quite sure how its responsive — I suggest that sellers in drug markets have substantial gains from co-location, which is an account of individual behavior. But if you’re interested in a more global account of how changes in agglomeration gains drove changes in local government law, you can check out my paper, The City as a Law and Economic Subject, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1471555

  4. Chris says:

    *”Real is estate is about co-location, co-location, co-location.”

  5. Mike Madison says:

    My point was that individual behavior varies, which is why gains from agglomeration are evident in some markets (including in some cities/regions) but less so in others. Not every individual behaves identically, let alone rationally. That, I hope, is obvious, but it’s important (I think) to flesh out that idea (what are the relevant dimensions of differential individual behavior?) and plug that fleshed out assumption into the agglomeration idea. Doing that makes it easier to understand why drug markets sometimes do re-appear elsewhere, and why clustered real estate development sometimes pops up in places other than downtown and sometimes does not. And why some cities thrive (New York), and how, and why some cities (Detroit) do not. Thanks for the link to your earlier paper.