Skip to content

American bureaucracy is pretty good

I happened to read my millionth tweet this evening about how wildly inept the federal government is, and that suddenly got me curious: how do we compare with our peer countries? I'm talking about non-tiny, non-poor countries.

There's no concrete measure of this, but the World Bank has a measure of government efficiency that they keep up to date. They're a reasonably reliable bunch, so I used their scores. I had no idea how this would turn out before I did it, but here it is:

For some reason the World Bank doesn't rate Taiwan, so they aren't included even though they qualify.

In the end, 16 countries made the cut and the US ranked 6th. There's no reason that we should do especially well since we're comparing ourselves to other first world countries. But we did, placing at the very high end of average.

I have no idea how the World Bank calculates this. As Rod Serling said, I'm just presenting it for your consideration. There are plenty of things we could improve, but apparently our federal bureaucracy isn't quite the hellhole Elon Musk thinks it is.

30 thoughts on “American bureaucracy is pretty good

    1. Jasper_in_Boston

      Same reason? lol. I’m pretty sure they don’t refrain from rating California because of pressure from a Communist dictatorship.

    2. Batchman

      For the same reason they don’t rate California.

      Thank you for saying that. I am so sick of reading "California is the world's fifth largest economy" nonsense.

  1. kenalovell

    Kevin, Kevin, Kevin. The data you link to is for government effectiveness, a very different thing to efficiency. The World Bank explains how the index is compiled:

    Government effectiveness captures perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies. This table lists the individual variables from each data source used to construct this measure in the Worldwide Governance Indicators. Please note that the table refers to the questions in the most recentlyused edition of each source in the WGI. Questions in some sources have changed over time and some sources have been discontinued from the WGI. For more details, please refer to the data files for each source available at http://www.govindicators.org

    The Bank also publishes data on the efficiency of government spending, which is an entirely different thing. I'm inclined to be sceptical about this data, given the efficiency of government spending in the US is supposed to have shot from 3.06 in 2016 (slightly below world median) to 5.89 in 2017. https://prosperitydata360.worldbank.org/en/indicator/WEF+GCIHH+EOSQ043

    1. Batchman

      A fellow coder of my acquaintance had a sign hanging in his place of work that read:

      Efficiency is doing the job right.
      Effectiveness is doing the right job.

    2. Narsham

      That World Bank data on this point is compiled purely from an "Executive Opinion Survey," meaning CEOs were polled. There doesn't appear to be any hard data behind it at all. Also, in the original study document it's labeled "Wastefulness of Government Spending" and not efficiency. I can't explain why the World Bank report is different, but have no reason to think they didn't convert the numbers.

  2. Jimm

    Most of these folks have no idea what they want from government, except security for the in group, which doesn't presuppose freedom, liberty, justice, dignity or any of that, especially for anyone not in the in group ("other"), a wide range of authoritarian to Nordic model states can provide this "security", some not free at all (the ultimate and ironic insecurity).

  3. martinmc

    Elon Musk's, and Trump's I'm sure, definition of Government Efficiency is "Let me do what I want." Anything that stops them from doing what they want is inefficient. We really don't need charts and models to figure out how they plan to make the government more efficient.

    1. Yikes

      No kidding. Republican positioning on this has evolved from a sane (relatively) desire for arguing that its better for government to have a limited scope of operations to the following bullshit, which, unfortunately works for millions upon tens of millions of voters.

      1. Point out (you don't even need to actually point this out), that the taxes on a particular person add up to a "big" number - this is easy because of two things, first taxes on even a lower middle class income add up to thousands of dollars, and second, the total amount of taxation always adds up to an astronomical number.

      2. Argue that because the numbers in point one are "big" government "must" be a waste.

      FFS, its hard to even type it but that's how dumb millions of people are. Government could be wasteful in a particular area, so could private industry, but its equally, if not more probable, that government spending is no worse or better than private spending, and, for that matter, as this chart shows, that we get at least as much out of our Federal government as other first world peers get out of theirs. As opposed to, I guess, government spending on the Klingon home world or government spending in some theoretical libertarian paradise.

      1. aldoushickman

        I imagine that Qo'noS government spending is actually pretty limited, as the Klingons seem to have some sort of House-based feudalism society. I guess they do manage to build lots of interstellar starships, whereas we have yet to build a single one, so maybe that's something.

    2. Narsham

      Step 1: fire everyone at NASA.
      Step 2: Contract all of NASA's functions to a company with X in its name (the "X" means "Xtra efficient."
      Step 3: Claim to be able to do everything NASA did for a fraction of the cost.
      Step 4: Take twice as long and have cost overruns so that everything ends up costing 50-100% more.
      Step 5: Profit.

      Hey, that's why Musk is so rich: his plans that lead to making a profit don't have "????" in the middle of them. They may not be ethically sound, but they work for him.

  4. eannie

    My decades of living in the Netherlands made me well aware of the machinations of one of the most thorough….and ruthlessly efficient bureaucracies on the planet. Dutch “ambtenaren”. Know all the tricks that citizens might pull and are way ahead of you. They never saw a stamp..or document..they didn’t love…and it doesn’t pay not to have your ducks in a row.. however…they run a country that was a joy to live in…for all the red tape..these guys know how to run a railroad.

  5. reino2

    It's all a bait and switch. They justify lowering taxes by promising to make the government more efficient. Then they make it more efficient by cutting $20 somewhere and don't care about the deficit.

  6. coynedj

    I know someone who works for the US Geological Survey, and has worked in private industry previously. He says that the government workers are far more dedicated and effective at providing quality service than the private sector. Of course, firing half the workers and not changing the workload will make them even more efficient, right?

  7. middleoftheroaddem

    US infrastructure project costs are materially higher than similar OECD countries. For example, the Second Avenue subway extension in NYC was six times more expansive than a similar addition in Paris. There are many similar examples.

    It is hard to see how bureaucracy, versus say materials, are not the driver of the US sky high costs....

    1. emjayay

      That's NYC though, complicated by the fact that the MTA is actually a state agency.

      London has put in two new subway lines, the slightly more trainlike Elizabeth line, a new automated elevated light rail, and extended many existing subway lines including getting the subway (along with several other rail options) to Heathrow in the past half century while NYC has added exactly four stations, all in Manhattan. And a lot of the design in London is really good while those new stations in NYC are all completely mediocre other than a few cool things like the mosaics at one station.

      Same kind of thing in various EU cities and elsewhere.

    2. KenSchulz

      Multiple citations needed. One ought also to address the underlying geology. Paris is built over sedimentary deposits; Manhattan bedrock is metamorphic. (cf. Wikipedia)

      1. middleoftheroaddem

        KenSchulz - IF there was only one example, THEN I would agree with your questions.

        Perhaps read the Vox article above, or use Google to see multiple examples of US infrastructure costs being much higher than the OECD.

        1. KenSchulz

          I didn’t dispute that US costs are higher; but immediately concluding that bureaucracy accounts for the difference is unwarranted. Other factors that affect costs must be considered: geology, land values, density of above-ground development, environmental impact, and so on.

          1. middleoftheroaddem

            KenSchulz - while certain US projects might have troublesome characteristics, in total I find it hard to think the US is that unique. Frankly, the US generally has lower density than much of the OECD: that would, I imagine, conceptionally reduce construction costs...

            Is there creditable evidence to support your supposition ?

            1. aldoushickman

              "Frankly, the US generally has lower density than much of the OECD: that would, I imagine, conceptionally reduce construction costs..."

              That just sounds like you'd need to build *more* infrastructure to serve the same number of people.

            2. KenSchulz

              You are the one who made a strong claim, "It is hard to see how bureaucracy, versus say materials, are [sic] not the driver of the US sky high costs....", so it is you that needs to provide evidence.

  8. name99

    Apart from the technical issues with the graph, the issue in the US is not so much "do we get the maximum number of governations per dollar", it's "do we even WANT the governations we get"?

    Back when the soviet union was a going concern, there was a genre of jokes about this sort of thing - yes the party gets what it ordered, but that wasn't what it wanted.

    The same is true in the US. The issue is less "are we getting what the law [at least as interpreted by someone] demands" and less "does the law represent what we actually want". Getting to this point via cutting/limiting appropriations (and even better, authorizations) may be suboptimal, but it's the way things are done in the US system.
    You'd think someone who's constantly complaining about AI's giving something that's superficially what was asked for, but not what's wanted would GET this point...

    I've said repeatedly that to understand politics you need to take things seriously, but sometimes not literally. This is one of those times.

  9. Narsham

    Efficiency isn't. It's all about what it is you want to accomplish and why.

    For example, consider the IRS. Nobody loves paying taxes. And there's some people who make mistakes on their returns, or deliberately lie or distort and don't pay what they lawfully owe. So consider two basic systems of taxation, and the efficiency decisions they lead to:

    System 1: Everyone prepares their own tax documents, or pays someone to do them. The wealthy and big corporations may have teams of employees who can look for every loophole; some may even have full-time lobbying agents in Washington proposing new tax rules to create additional savings. Those without that kind of wealth and influence may do their own taxes or rely on tax preparers.

    It's inefficient for the IRS itself to independently compute every tax-payer's documents and compare with what's been filed. So it needs to spot check or look for patterns which might call for an audit. Audits can be time-consuming and expensive. Auditing someone who never paid any income tax might be important if they cheated to the tune of millions of dollars; if they never had enough income to owe a tax, auditing them is only worthwhile if it turns out they were lying about their income. You have to make decisions somehow.

    But it may also be inefficient to audit massive corporations or the megarich, because you then have to go look through years of records, potentially deal with dozens of employees, and even if you find problems you may need to resort to legal action to recover any owed tax. Further, each audit which might bring in millions or even billions of back tax will cost a lot. If these big audits cost $1 million each, Congress can starve you out of conducting them.

    In system 1 you could improve efficiency by simply allowing the IRS to "charge" all the costs of an audit to the money recovered by it so long as the recovered money meets a minimum threshold: spend $1 million to recover $55 million, and the whole cost of that audit is paid for out of the recovered money. Spend $1000 to recover $100 and you're out $1000. (DOGE will not advise doing that.) Arguably, you could make the IRS more "efficient" by conducting a minimum number of audits per year: the IRS budget goes down, even if revenues suffer. Is the most "efficient" approach to drop audits to the minimum required to keep anyone honest? A min/max point between auditing nobody and auditing enough people that many returns are accurate? Do you determine that point based on number of incorrect returns, or money owed per return? Is it inefficient to catch mistakes where people paid too much tax, or is that a desirable state of affairs?

    System 2: The IRS computes everyone's taxes itself. The wealthy, well-educated, or accounting nerds can double-check the IRS computation and dispute it. This system would be inefficient if computing income taxes for the majority of Americans weren't fairly simple: much of the tax code covers situations most people simply don't fall into. You need a robust system to catch obvious and bad errors (computed tax higher than someone's total income, say). You'll still have accountants working for private citizens doing a once-over on the IRS computations; you'd have many more tax preparers employed directly by the IRS.

    Is System 2 less efficient than System 1 because you'd have a huge IRS with offices around the world (US citizens living abroad pay taxes, too)? Or is it more efficient because tax cheating would be more difficult (though certainly not impossible)?

    I will observe System 2 may lend itself to more government corruption, in the sense that the incentive to cheat the system is going to put pressure on whoever at the IRS handles complex cases like multinational corporations or billionaires. OTOH, System 2 might also mean less systemic corruption. If you're only examining government systems, do you care about the effect of changing approaches on non-government systems?

    I can more clearly identify the places in System 1 which reward cheating the system (or playing the system for your advantage) than I can in System 2. That does mean a lot more interest groups would support using System 1 (where they can profit) or System 2. But we don't have System 2, so you'd have to guess at what would go wrong were we to switch; more to the point, there's powerful political pressure not to switch because System 1 is so much more obviously advantaging the wealthy and powerful who have political influence. Plus the powerful potentially have more to lose with a change.

Comments are closed.