Disperse the Federal Government

I have a random suggestion for the new Administration: disperse the Federal government. The idea is to move agencies that do not really need to be in Washington to depressed areas of the country. This would improve the labor market in those areas.

We could move HUD to Detroit. We could move the Department of Energy to West Virginia. We could move the Department of Education to rural Mississippi.

I know that some of you do not think we need these agencies at all. But dispersing them might accomplish some of what you want. Many of the employees would be unwilling to relocate.

17 thoughts on “Disperse the Federal Government

  1. I like it. And potentially, it could also save the government quite a bit of money. The median wages and cost of living is much lower in a lot of those areas than in DC, so pay scales could be lower also.

  2. Here is a way to think about it. Think of a declining population, everyone born reaches reproductive age, everyone has offspring, but every couple has 1 child. What happens to the information? Half of all genetic information is lost every generation (taking every gene on its own). Now take a stable population, everyone reaches reproductive age and every couple has 2 children. What happens? 25% of all genetic information is lost every generation, because there is overlap in which genes the children get.

    High fidelity at the genetic level of the genes that are passed leads no low fidelity at the population level when looking at all of the genes from one generation to the next, and it gets less when you look at how genes interact with each other, epigenetics etc.

  3. This is what the military does, though not originally or principally for reasons of geographically distributing the benefits of concentrating tax expenditures and government employment in particular locations to produce local economic positive externalities. Though, when it comes time to try to remove or consolidate bases through something like a BRAC commission then watch people fight.

    As a historical note, the 2005 BRAC told the Army to get rid of a lot of expensive square feet and locality housing allowances used for its Human Resources Command (many in the National Capital Region) and move everyone to Fort Knox. Kentucky is cheaper and the state, local area, and even nearby city of Louisville has greatly benefited from the action.

    However, excepting the pentagon and the DC-area defense agencies, the military provides an almost idea institutional setting to accomplish this goal, where the costs of moving people to almost anywhere are fairly low since the forces and support commands are already and purposefully geographically distributed, and units are insular, modular and compartmentalized by function.

    That is to say that these units have something like constant returns to scale and low economies of agglomeration. But management functions aren’t like that, and have high returns to agglomeration.

    So, ‘distributing the Pentagon’ would be a bad idea, because the management and control functions that occur there require a lot of in-person social interactions, proximate teamwork, face-to-face meetings, and coordination. Most people have functions which effectively, physically ‘tether’ them to at least a few other key individuals, and those people have tethers to other people, and in the spiderweb you end up with people who have tethers to other important DoD agencies and government entities and especially the White House.

    And you want certain public employees in roles common across the government to be able to move around from department to department without having to move house. That happens a lot and produces big win-win benefits for both the government and the civil servants.

    The overall trend is definitely towards centralization and consolidation, and there are important economic and technological reasons for it. It’s not that the government couldn’t technically scatter a million or so employees to the four winds and force them to use the inadequate substitute of IT and teleconferencing, but there would be big costs to those measures to achieve the intended local benefits.

    I would expect you to usually support something more along the lines of “Just give those locations money instead of producing those costs,” though of course importing a group of people with higher incomes and human capital than the local average can produce other kinds of benefits. But actually the federal government already kind of does this through all sorts of grants and transfers than subsidize local government, education, and health care.

    Now, think about your reaction if someone suggested that the government should insist on distributing any white-collar corporate campus greater than 500 employees. Can’t have another 500 employees in one place closer to an existing place than, say, 50 miles. Well, there’s probably a good reason certain corporations want a big campus in one place where all the employees and managers are close to each other.

    Or what if the government told cities that they must quota their populations to no more than 500,000 within 10 miles of the center. That would certainly produce some benefits for the residents in terms of infrastructure congestion and land per person, but obviously lots of costs too. These apply to certain cities especially, like Silicon Valley. I mean, why should all those top brain and wealth creation be concentrated in one place, bidding real estate prices to the sky and jamming the roads?

    If it was all really a zero-sum rat race without big benefits for individuals and businesses, then it would present a kind of ‘market failure’ and it would be a great, enormously surplus-creating idea for the government to break up Silicon Valley and spread it across the country, perhaps even to such an extent that it would be worth the obnoxious, liberty-reducing intervention.

    The same logic that applies to campuses and cities applies to the federal government.

    • 1. If it is less profitable for a company to be in Silicon Valley than to be in Utah or Texas, it will make less money and its stock will suffer. There is no such “profit test” for the federal government. A very different logic applies.

      2. For 20 years, I have been hearing about how information technology–all the networked computers and smart phones and printers and whatever comes next–make centralization less and less necessary. Has that all been a lie?

      • Not necessarily, but there was definitely a lot of hype based on wishful thinking.

        Consider: if I can use IT to push some function at a distance of ten miles away, then why not ten thousand, outsourced to the cheapest place on earth? Well, we do that, and there are entire cities in India and the Philippines which specialize in those functions.

        That means that those tasks won’t be performed much in developed counties with more expensive labor, at least not without government imposed distortions.

        So it was naive to think that the main worker beneficiaries of such technological decentralization would be domestic ones in an era of globalization.

        And that means all the tasks that remain are the ones that can’t be outsourced or decentralized via technology for some reason related to the fundamental nature of the task itself.

        And if those are all the tasks left, then through this mechanism IT has increased, not decreased, the force of centralization in developed / first world countries, and exacerbated the issue, not alleviated it at all.

  4. I don’t think it works very efficiently having the headquarters physically distant from the White House and Congress. A lot of the government already has field offices distributed around the country and world, from the IRS, DHS, and FBI to the DoD and State. Typically the operational tempo differs quite a bit- the field is where stuff gets done. Perhaps a better approach would be to strictly limit the number of employees at the headquarters location, thus forcing greater distribution. For example, the size limit of 100 staff for the NSC seems to be effective, if still high.

    More reliable and high quality video conferencing is making this easier, maybe VR will help too.

    • You can have McDonald’s restaurants or Walmarts in thousands of little places. They obviously have to be close to where the customers are. Yes, there are store managers and regional managers, but they are performing only a small part of organizational control. The headquarters management functions are always all in one place, and for good reason.

      You obviously need your border patrolmen distributed in field offices along the border, and your customs agents distributed in ports of entry (and even forward deployed in foreign countries), nearly all of which are far from Washington. But CBP headquarters is only half a mile from the Oval Office. Even agencies that are ‘only’ 10, or God forbid 20 or more, miles away usually start to suffer from ‘out of the loop’ and ‘out of sight, out of mind’ problems.

      There is not much sense in setting some arbitrary staff-size limit for any particular government function. This is for a variety of reasons, but the most obvious one is that it ends up being treated as a merely ‘formal’ requirement that is simply too easy to circumvent in practical terms. Ok, the NSC is ‘limited at 100’, but then if the workload requires more than 100 people, other parts of the government are commandeered (and sometimes literally ‘detailed’) to perform those same functions. You’ve gained nothing but perhaps a politicians benefit by some symbolic signal displays, but you’ve lost another little shred of normative institutional respect for both elected politicians in general and for the principle of adhering to both the letter and spirit of the law. Not worth it.

      Bottom Line: the only good way to have fewer government workers in DC is simply to do less governing.

      • “Bottom Line: the only good way to have fewer government workers in DC is simply to do less governing.”

        Agreed. I would rather have a smaller government that worked well rather than a large government that worked poorly.

  5. This is a great idea. It would also help counter the criticism that government is too insular, sheltered in the DC bubble. As you mention, many current employees might object due to unwillingness to relocate. Is there a natural constituency that would push for dispersion? Maybe, public choice explains the current concentration of agencies within the DC beltway. Public employees are actually a powerful special interest group.

  6. I disagree with this idea. When it is dispersed, the government has a bargaining chip. If you want to keep your government job factory running in Flint Michigan, then you better approve my boondoggle project over here in Florida.

    Instead of trying to attract private sector investment, state governments chase after government bureaucracies in the same way that many pathetic cities chase after museums, sports teams, halls of fame, memorials, river walks, mass transit projects, and other unproductive ventures.

    If an obscure yet bloated bureaucracy is concentrated in DC, the rest of the country has little interest in keeping it going when the budget hits the fan.

  7. Why not take an axe to the budgets of the agencies? 10% of last year’s “continuing resolution”, let the agency “stop doing its job”.
    Chop 90% of the headcount.
    No more new regulations, only doing what is really absolutely needed.

    “Automatic approval” on requests for approval if not rejected within 6 months would be good, too.

    Most folk won’t notice much bad, certainly not at first.

  8. I’m not sure what you mean by disperse. There are tens of thousands of federal (civilian) employees spread throughout the country. The EPA, just to take an example with which I am familiar, has 10 regional offices outside of Washington. HUD, to use your example, also has 10 regional offices and 117 local offices. Yes, the head of each agency is located in DC, but they are all part of the Executive Branch, the head of which, the President, is unavoidably located in DC.

  9. When Mike Harris was Premier of Ontario, he moved many provincial government agencies and jobs out of Toronto to rural areas and smaller towns. Whether it was due to cost savings or to pay back smaller communities that voted for the Progressive Conservatives, the only real effect 20 years on has been to distribute Liberal, statist voters across the province.

Comments are closed.