Successful government bureaucracies

Jason asks:

What are some examples of successful government bureaucracies?

Wars aside, here is a short and very incomplete list: the NIH, the Manhattan Project, U.C. Berkeley, the University of Michigan, Fairfax County, the World Trade Organization, the urban planners of postwar Germany, some of the Victorian public works and public health commissions, most of what goes on in Singapore, anywhere that J.S. Bach worked.

The European Union has been very good for eastern Europe.  I'll leave aside the health care issue because we've debated that plenty already.  The real question is what all these examples have in common.


I want to throw out the Tennessee Valley Authority.

No success is forever, but US public schools at all levels were a tremendous success for a long time. In some respects, they still are.

Let me give you a hint on what each successful bureaucracy has: a clear mission. Failure becomes almost inevitable when an organization's mission goes away or get diluted by extraneous concerns. The public schools are again a classic example.

US public schools are questionable at best, especially when results vs dollar spent are compared. If a clear mission is all that it takes, then how come the DMV is such a failure no matter what state you reside in? Their mission seems simple enough: register cars, oversee driver's licenses, and in some states test emissions. Much more simple than the task of elementary education, imo.

Hong Kong under te British flag -1945/97 - was a pure apolitical bureacracy; and a very successful one.

The problem with government bureaucracies is not that they cannot succeed. Quite often they do their job very well indeed, and you can pick up the hum that comes from the organisation. It is that they need to be really accountable. They must have somebody with authority and the will to judge their objectives and performance, and impose change if needed. (The Governor of Hong Kong was always accountable to London.) They will also slip into decline if that somebody does not have a grasp of what realistic objectives should be.

Japanese post-war growth came unstuck when the civil service ceased to be accountable to its LDP 'masters'. The sprawling US Federal structure has two masters; and its salavation has been that both Congress and the Presidency are occasionally willing and able to judge objectives and performance. (Its curse is that the sprawl means that a lot of the bureaucracy is neglected by both masters much of the time.) The formidable British civil service is faltering because its political masters for the last 20 years seem to have lost the grasp of what are realistic objectives.

A good bureaucracy is like a good horse. It can get you where you want to be surprisingly fast over bad terrain. On occasion,it will even get its masters home when they are drunk or delirious. But if you persist in not controlling it, you will end up falling hard, stumbling into the ditch, or both.

SSA did an amazing job until the 1970s.

I don't see many complaints about the CDC either.

It is always amazing that for less than .50 cents I can mail a letter anywhere the US and it will be delivered in a two-three days.

This thing called the internet seems to have worked out ok.

Most of these organizations (I'm not too knowledgeable about Singapore) are staffed by a relatively small number of highly-educated, highly-motivated people. The same is not true of, say, the Department of Health and Human Services.

Public schools are statist indoctrination camps that happen to teach subjects like the 3Rs on the side.

Success hasn't been defined here, but in the private sector (i.e. the real world) it's defined by profit and loss, satisfied customers, productive employees, and wealthier capitalists.

Does the Post Office count?

I'm very happy to see you put the NIH up here.

In 2002, a mysterious respiratory disease that was killing 9.6% of those it infected was discovered in rural China.

In 22 months, scientists at the NIH had not only discovered the virus that caused SARS, but had advanced a vaccine to the point where they had begun human trials. 22 months.

It's easy to focus on bureaucratic inefficiencies and waste that can be produced with public institutions (though people don't like to focus on private waste and private bureaucracies as much). Still, we have to remember we have some incredible resources in this country, particularly in the field of basic research.

I would take issue with the claim that the DMV is a failure in every state. In California, the DMV seems to work as well as any business in terms of ease of access and wait times. A lot of renewals can be done through the mail or online and when you have to go in person, you can make an appointment online. They seem to be pretty good about ensuring there are enough staff to handle everyone at a given time. My wait time never seemed any worse than, say, waiting for a table at a crowded restaurant or waiting to see a doctor or dentist.

I'm sure this didn't happen by accident, though. As others have noted above, having a clear mission with incentives geared toward achieving that mission probably make a big difference. If DMV offices were judged by average wait times or amount of forms processed per day and managers were expected to account for any poor performance, that could be one way to make more DMV offices more efficient. When you just pay people by the hour and don't supervise them or care about how well they do their jobs, on the other hand, that's asking for trouble.

Here in Switzerland, we have one example after another of well-run and successful bureaucracies. Even the local equivalent of the DMV.

I think a big part of it is that people are RESPECTED for being members of these organizations, and many of the best and brightest choose to work in them. I also think there is a cultural appreciation for competence and disdain for incompetence that creates great cooperative pressure in response to poor performance.

If there was a government bureaucracy that provided education, training, and a structured meritocracy to 1.45 million Americans (the majority of which with only a high school degree), surely that would be laudable, wouldn't it?

But it's the military, so let's not talk about it and resume the regularly scheduled programming of intellectual self-regard and extol the wonders that are Cal-Berkeley and the U of M.

Brian Timoney

1) Here's one thing that many of them have in common: they are designed to innovate technologically to solve relatively narrowly defined problems. That's true for the Manhattan project, NIH, the internet, DARPA, and to a lesser extent research universities

2) @ Bob Murphy: if the Manhattan project were private and "could reap the profits from selling A-bombs", Iran would have a nuclear bomb, along with many others. So maybe we would have saved some money in creating it, but that's clearly not the whole picture.

3) @ Andrew: "People forget that governments are just people. My guess is the parts that are successful aren't really governments. Successful schools are parents and teachers. Should the government get credit for this?"

It's very hard to credit an organization for something if you disconnect its members from the organization. By analogy, would you say that Apple deserves no credit for the iPhone because it was its employees, not the company itself, that actually invented it? That would be silly, since an organization is defined by its members. And that is true for companies (like Apple) and for governments.

I think NOAA deserves to be on this list. I'll claim ignorance of their budget, but they seem pretty damned good at what they do.

"But it's the military, so let's not talk about it and resume the regularly scheduled programming of intellectual self-regard and extol the wonders that are Cal-Berkeley and the U of M."

Ah, there's no snobbery as warm and fulfilling as reverse snobbery. That kind of obsession with quantity (1.45 million!) as the ultimate measure of things, along with the conviction that only giant state institutions represent the will of the people rather than useless coddled elites, is what made the bureaucracies of the Soviet Union and China so great.

Doesn't the answer depend on what your objectives are? If your objective is to reduce poverty and improve health for the elderly, I'd say the Social Security Administration is highly "successful".

What are some examples of successful government bureaucracies?

Stazi, KGB, Gestapo, IRS, etc.

One of the basic necessities of a highly efficient and successful government bureaucracy is the ability to destroy the lives of anyone who gets in their way.

I don't believe there's a big difference between a successful mid size corporation and a successful government program: In the end, they are large organizations of people that tend to fail, but sometimes succeed, and can continue succeeding for a few decades.

Both need an achievable mission, a leadership group that is competent and believes in the mission, a business plan of sorts, quality hiring, and some luck.

The reason most people think of the evils of government while they ignore the failure of corporations is that, for the most part, corporations start small, and truly dismal enterprises go away way before they are large enough to reach the public consciousness. Government programs start with near their target optimal funding, so when they fail, they fail bigger. The exception might be dot com startups, which received too much funding for their actual level of promise, and therefore failed more spectacularly.

If we started government programs as pilots, and only expand them after we've seen that the administrators knew what they were doing and had a plan that is proven to be viable, we'd get as many failures, but the ratio of money spent on success and failure would improve.

Now, that'd not be viable in projects with huge startup costs, but that's why those projects are seldom attempted in the private sector: The risk is just very high.

Bob Murphy: Another thing, how is the Manhattan Project an example? Yes, they ended up creating an atomic weapon. But did they do it for less money than it would have taken a private company to do, if that had been legal and the company could reap the profits from selling A-bombs?


Okay, now that my hair is back in a flat position, is there any way this could work?

a) Contract to the low bidder. At best, we have an airline result with our nuclear deterrent built with the cheapest materials possible. I'm rather worried about safety. In the mediocre case we get $10k hammers. Worse case (not worst, but worse) - Haliburton with H-bombs. We could name it the HALiburton-9000.

b) X-prize style. Get the best and brightest at dozens of private and public institutions working on it. It worked for space flight. Okay.... Security leaks are going to be killer... Undergrads at MIT working on nuclear warheads... Be afraid.

Just as an aside, I was watching Spiderman 2 for the first time last night while cleaning, so add to my list of fears Doc Oc with Einstein's hair. Jurassic Park comes to mind as another instructive example. Compared to the list of ways the entire project could have gone wrong, no matter who did it, I think it comes down as a success.

It might be instructive, though, to compare the Manhattan Project to other countries' efforts that also produced bombs: security, cost, time taken, etc.

A parking lot of around 100 spaces at a nearby train station, run by the New York-based Metropolitan Transportation Authority, recently reopened after a five-month rehabilitation, the sort of work that would take no more than a couple of weeks for a privately owned lot. An adjacent lot is now being rehabilitated, when all is said and done 300 spaces will be redone at a cost of $4 million, which is about $13.6K/space. Building new surface parking lots should cost about $5/space.

The DMV in Sauk County Wisconsin is perfectly customer-focused as well with short wait times. And the IRS really is good at collecting withheld taxes from regular paycheck earners like me, and also mailing back refunds more promptly than a rebate company would. I know their computers and enforcement suck but hey, Verizon's phone support sucks too.

I know mentioning Bach was a half joke, but I thought I'd mention that Bach succeeded despite being under the thumb of a generally intrusive, meddling, and unappreciative church council in Leipzig. They even had him busy training the council member's children as a simple music teacher. They were disappointed he staged Matthaus-Passion a total of twice (though nine years apart). It was old news... everybody had heard it before.

He was, not surprisingly, always looking for a better post with more freedom.

I am surprised that no one has mentioned the Indian Civil Service under the British. They did a great job of following their political masters and yet being able to bring some benefit to the natives. The model was replicated in successful countries like HK and Singapore. The IAS is still one of the few institutions in India that works given the scale of problems.

They are all very limited in scope and mostly a-political. And yes the NOAA needs to be on that list.

The Nazi concentration camps were very successful at what they set out to do.

What exactly is the definition of success in the context of this thread? Creating a list of things without their counterfactuals? Did an economist really endorse such an idea?

Case in point, the counterfactual of my first paragraph is that the world would almost certainly be a much better place without such government success, regardless of what the alternative would have been.

However, most of the things listed in the comments here can't make anywhere close to such a claim, and probably have the sign wrong. Anyone who thinks government run K-12 schools hasn't been paying attention to the voucher studies.

It might be more fruitful to look at agencies that were considered competent at one point, but declined. Knowing how elite organizations like NASA of the 60s or Hoover's FBI declined into the mess they are in today would be instructive.

I think hibikir is closest to the right answer, but I'd simplify it thus: government bureaucracies have no useful failure modes.

If a private enterprise stops working, it weakens, and eventually dies out. Even if it has spent years being a great success (GM comes to mind, though its failure took decades). Moreover, there's a fairly hard limit on the quantity of resources they can consume (the enterprise's built-up reserves and what they can convince others to lend them).

A bureaucracy that has failed needs to be actively killed, no matter how ridiculous its results are. The market will not provide enough information to define its viability.

A test of this theory is that you'd expect to see "successful" bureaucracies appear about as often as "successful" companies, but the failed bureaucracies would stay around instead of disappearing. (scare-quotes because I don't think a tight, comparable definition of success for bureaucracies and companies is a trivial matter).

Pearl Harbor, Cold War, Gulf of Tonkin, 9/11....all did their job...see government can get it's people to support mass destruction for the benefit of bankers and the military industrial complex. yay government!

Organizations do well when they have clear goals, measured progress, room to innovate, and accountability. That's true for bureaucracy as well as the private sector. And the smaller the OODA loop the better.

Growing up in New Jersey the DMV was deservedly the butt of every joke, despite clear goals and the ability to measure progress. Without accountability and mired in 1,001 rules & regs, they were stuck. Later New Jersey stopped paying by the hour and started paying by the person served ($10 for registration, $15 for licenses, etc.) and gave local managers freedom to innovate how the local branch was run. It's not a marvel of efficiency, and I would add it to your list of bureaucracies that do well.

Accountability is really important, but that doesn't always mean "pay for performance." Many professionals understand just how important their work is, and they hold themselves accountable. Many doctors and teachers are like this, trying to save lives or educate minds. These types will be more productive with a flat salary and freedom to innovate than with any sort of externally imposed incentive structure. But that means HR has to have the incentive & freedom to hire and fire the right sort of people.

Many government bureaucracies suffer from a lack of clear goals, mainly because of mandates written by ten different Congressmen with different lobbyists to answer to. NASA is an example of an organization that cannot decide whether it's primary purpose is to explore the farthest reaches of space, build a low-cost LEO infrastructure, put men on Mars/Moon for King & Country, or just employ constituencies in Florida, Alabama and Texas by any means necessary. It also suffers from multi-year OODA loops.

The problem that the vast majority of bureaucracies suffer from though is a poisonous combination of Monopoly + Special Interest Capture. Consider a public works Dept. that answers to politicians that answer to union bosses (for both votes and campaign funding). The clear goal of that Dept., whatever it started out as, will quickly become "Transfer as much wealth to the union, for as little work, as we can get away with." Service goes to crap, costs skyrocket, and the public cannot go to a competitor (because of the monopoly).

"Bureaucracy" Vol 6; Selected works of Gordon Tulloock (Liberty Fund 2006)

What measurements or standards comprise "successful?" Politically; or, in productive structure?

Since the term "governmental" is the modifier, the implied area is public administration. Politically.

In an exchange-oriented society, how are we to measure or judge the effectiveness of the exchange of services (service for service here being in exchange for taxation) by segments of the the governmental mechanisms identified as bureaucracies.

I'm Singaporean myself and I'd like to know too haha. Our civil service prides itself on taking in the "best and brightest" (at least when referring to the upper echelons of its hierarchy), either through scholarships or attractive remuneration packages. Indeed, one might go on to argue that the civil service has a virtual monopoly on talent, and perhaps this successful bureaucracy comes at the expense of something else (namely, entrepreneurship).

Still, it's something everything Singaporean can reasonably be proud of, myself included.

how good is berkeley?! but what about the rest of the UC system? esp UCLA and UCSD.

@Tim “A couple of the government successes in the lists are government successes at limiting themselves in some way.† I’d agree, but want to unpick the reasons†¦

The modernization tautology is that public bureaucracies are successful in the successful countries (roughly speaking, two dozen OECD countries). They work badly in the other 175 or so countries of the world. Public bureaucracies work better in the advanced countries because they are accountable in historical order of importance to (1) economic markets (2) the legal system and (3) parliament – all of which demand that bureaucracy delimit itself through standardized goal-oriented procedural impersonality, i.e. demand it does not discriminate among individuals and groups to achieve economic or political outcomes. ‘Success’ is all relative because advanced society bureaucracies are in evolution. If the OECD trend to social-democratic big-government continues to be powered by political interest-group pressure we can expect declines in the quality of bureaucracy. If the countervailing trend towards new and experimental technological forms of delimited impartial governance accelerates we might expect the quality of bureaucracy to improve. All this can be noticed in differential trends within countries, where most comments focus.

I am done with this website that provides no substance, but only gives pithy remarks and generality. Where is your proof of such pronouncements to back up your claims?

WTO? Wars aside, but a war-time weapon project of mass destruction?

None at all, empty rhetoric like most academics from state schools.

No wonder you are all dinosaurs and you know it.

Can the concept of success exist outside of a system of private property based ethics? Now that's the REAL question!

Given a general mandate of universal service to both the rich and the poor, powerful and weak, I'd say that one can determine the nature of success rather objectively.

For example, how does private property come into play in determine the success of the courts, other than the degree to which your property rights are protected by a court without regard to whether you are rich or poor, politically powerful or weak? It seems to me that if the judgement made when you are poor and weak is the same as when you are rich and powerful, then the court is successful.

If your mail is accepted and delivered for the same rate between the same addresses whether you are rich or poor, weak or powerful, then the Post Office is a success.

And the one thing that you have with the courts and Post Office is a say as both a "customer" and "shareholder".

It is a very small minority that complains the Post Office or now USPS is poorly run, and their claims are absurd. The clain is that the cost of Postal Service would be cheaper if privately run. And the evidence for this is what? The cost of shipping by USPS Express Mail compared to FedEx or UPS for the same parcel? By all comparisons I see, the USPS is cheaper, and the only thing it doesn't do is offer the enhanced delivery like within 12 hours, 16 hours, and 24 hours.

Universal transport is provided only by government - before 1900, it was the government that provided the land, subsidies, and incentives, and rates to provide rail service pretty much everywhere, until 1980, it was the government that provided the infrastructure, the subsidies (airmail was a big one), and rates to provide service pretty much everywhere. Today you can't get service by rail or air to lots of places well served, but the government provides service almost everywhere with its roads.

Of course, as both customer and shareholder, we both benefit and play a role in deciding what kind of service is delivered.

I suppose the minority thinks that they don't have enough control, as if you have control in a market served only by private firms. Let's say you have a pre existing medical condition and no government regulation, and as a result of the medical condition, you have difficulty making a lot of money. Do you think the free market is going to serve you? If you are forced to sell off all your private property in order to get medical service, and now you have no private property left, how exactly do property rights help you get the medical care you need?

Is a right to life, liberty, and property something you have if your only option is dying from a medical condition that you can't afford because you sold off all your property?

How do you define success under those terms?

Erie Canal earned a return above market rates.

I don't think these agencies have anything in common. If governments try thousands upon thousands of programs it is just plain statistics that a few will return more than they cost.

50 posts and no one defines a successful bureaucracy?

You mean successful by its own standards at its own claimed goals? Or at maintaining its power? You mean it is successful at delivering goods and services to a person in need of their products or services? Does its job with the least intrusion? Pick a metric!

It is a good question and I don't have a good answer, and doubt there even is one.

It's worth noting that the " magic of the market" isn't in creating better organizations but on weeding out the ineffective ones. Government bureaucracies aren't necessarily typically ineffective because government breeds more ineffective organizations, but that it let's the ineffective ones that it does breed stick around a lot longer.

Comments for this post are closed