One good first step would be to look at high state capacity, effective governance polities and see 1- what are they doing right, and 2- what is culturally/institutionally "copyable"? And this includes not only other countries or states or cities in certain areas (China infrastructure, South Korea nuclear power, Spanish high speed rail, Milan subways, Singapore healthcare) but also in the past (US highways, French nuclear power in 70s, US Manhattan Project then ICBM project then Apollo, etc.)

One white pill is that because the current dysfunction is SO bad and dire in the US, that there's massive room for improvement via low hanging fruit, 20% reform that yields 80% of the benefit, etc.

Expand full comment

Government bureaucracies are also permanently left-leaning and undermining of the political policy agenda of any conservative-leaning elected government. "Unsurprisingly, neither governmental bureaucracies and quangos nor other civil institutions keep statistics on the political leanings of their employees. But there are clues. Unherd columnist Peter Franklin reflecting on his own experience of working in two UK government departments comments: “How many of the civil servants that most closely serve this Conservative government are actually Leftwing? Well....I would say approximately all of them”. And it’s not just the UK. Research in the US context finds that “the political beliefs of the median federal government employee lie to the left not only of the median Republican, but also the median Democrat”". https://grahamcunningham.substack.com/p/carry-on-governing

Expand full comment
Jul 30, 2023·edited Jul 30, 2023

You should be ashamed of yourself for repeating false partisan rhetoric without looking at the data, and so should the people liking your post.

The claim that "the political beliefs of the median federal government employee lie to the left ... [of] the median Democrat" would be quite sensational if it were true! So, you shouldn't just repeat such a sensational claim without checking up and finding a reputable primary source. That's how fake news spreads. When I follow your links they lead me to an Amazon page for a book and no indication of where the claim originated. Where's the study? Is there a study?

After checking up, it turns out it isn't true. Federal employees have practically the same political party membership as the general public. Here's some real data. (ignore the headline, look at the numbers) https://news.gallup.com/poll/146786/democrats-lead-ranks-union-state-workers.aspx

Unionized federal employees: 27.2% Republican, 39.5% Democrat, 31.4% independent.

Unionized non-government employees: 23.8% Republican, 41.9% Democrat, 32.9% independent

Non-unionized federal employees: 32.5% Republican, 29.3% Democrat, 36.3% independent.

Non-unionized non-government employees: 31.6% Republican, 30.1% Democrat, 36.8% independent

For comparison: 33% of federal employees are unionized, and 10% of private sector workers are unionized (based on https://www.washingtonpost.com/business/2023/03/17/unions-federal-workers-harris-biden-white-house-task-force/ ). Combining those numbers with the Gallup poll, I calculate the overall percentages are:

Federal employees total (union + nonunion): 30.7% Republican, 32.7% Democrat, 34.7% Independent

Non-government employees total (union + nonunion): 30.8% Republican, 31.3% Democrat, 36.4% independent

Expand full comment

Do you have a more recent source for the partisanship of government employees? The Gallup link is good but it's from 2011.

Expand full comment
Aug 1, 2023·edited Aug 1, 2023

I found https://www.govexec.com/workforce/2022/11/poll-federal-employees-slightly-prefer-democrats-upcoming-midterms/378843/ and see the full survey results at https://admin.govexec.com/media/gbc/docs/pdfs_edit/102422poll.pptx .

The headline is that federal employees who were subscribers to GovExec somewhat favored Democrats in midterm House elections for 2022 (46% to 35%) and 2022 Senate elections (37% to 33%).

The figure that is directly comparable to the Gallup poll is the party membership, on page 12 of the powerpoint. 32% of Federal employees in this 2022 survey are registered Democrats, 33% are registered Republicans, and 35% are independents. That's pretty much the same as the 2011 Gallup poll results.

Caveat: this survey was given to 298 subscribes to GovExec. That's a fairly small sample size and as it is restricted to GovExec subscribers it might also not be a representative sample of all federal employees. Although, the agreement with the larger Gallup poll on the topic of party membership suggests it's not too biased.

Expand full comment

I said "there are clues". I then cited two 'clues' that I came across in two highly respectable publications - Unherd and Law & Liberty. l do not feel remotely 'ashamed' and very much doubt if the people who 'liked' my comment do either. I do however feel slightly irritated by your intemperate attempted ticking off.

Expand full comment

You ought to feel ashamed. The "unherd" quote is an anecdote and a personal subjective perspective, not evidence whatsoever. The "law and liberty" quote lacks a primary source (I told you how it dead-ends at a link to an Amazon book) and also completely wrong, based on the Gallup poll. You should be ashamed of placing your trust in such sources and ashamed of not fact checking them.

Expand full comment

I said "there are clues". just CLUES - nothing more assertive than that. You need to get over yourself but I somehow doubt if you are ever going to manage it. This is the end of this exchange as far as I'm concerned.

Expand full comment

Extraordinary claims demand extraordinary proof, not just "clues." If you read something and you think, "wow that's amazing and I'd like to share it with people!" then you should be *extra suspicious* of it and extra careful to check the facts.

Fake news spreads because people share things they think are amazing without bothering to check the facts. For good statistical reasons, "boring" information is more likely to be true than sensational information. Double-check the sensational information.

Expand full comment

"outcome accountability" sounds a lot like Futarchy.

Expand full comment

I believe this was the overall point, prediction markets could be a tool for accountability

Expand full comment

I think you're right. This shows that:

(a) Robin is smarter than I am - to him this seemed too obvious to mention; I thought he was criticizing the book, and

(b) since most people are not as smart as Robin, he needs to make more effort to explain things in a way that people will understand.

"Outcome accountability" sounds way more reasonable and explanatory than "Futarchy".

Anything with "future" in it seems inherently flaky, as cranks of all kinds think their ideas will succeed "in the future" (someday). It sounds like it's out there with nuclear powered flying cars and jetpaks. (Not that those things can't ever happen, but few people take them seriously.)

I know the intent is that Futarchy evaluates things afterwards (therefore in the future), but this isn't obvious until you spend time understanding the idea. The name "Futarchy" discourages most people from taking it seriously enough to even read about it.

Expand full comment

Naming/branding is a reason prediction markets haven’t taken off, perhaps “outcome accountability” is too broad sounding (laws, contracts, norms could also fall under outcome accountability). The larger reason prediction markets aren’t used more is because they take power away from individual decision makers who don’t want to adopt such a system. There’s some kind of bias that makes us value the word of charismatic individuals over aggregate data.

Expand full comment
Jul 30, 2023·edited Jul 30, 2023

If you give total freedom to contractors to do whatever they want, you're just kicking the problem of accountability down a level. What's to stop an unscrupulous and unsupervised contractor from taking the money and siphoning it off to their pals and allies, just as you're concerned about politicians doing?

> “We’re not following the letter of the law,” [Yadira] admits. “But we’re producing the results that we know Congress intended.” (p. 215)

Who would believe that a typical contractor will accurately read what Congress intended? They're going to interpret it in whatever way balloons the project budget. The problem is not competence, it's corruption. A competent but corrupt actor is more of a problem than an incompetent but sincere actor.

If the public can enforce outcome-based accountability among government officials first, then the government officials will in turn be motivated to enforce outcome-based accountability among their contractors. If the public can't enforce outcome-based accountability among government officials, things will proceed as they have been. Letting officials disclaim responsibility for supervising the contractors is a step backwards.

But, how can it be fixed? I think we ought to try some radical ideas, see if they work. For example, one of the problems is that politicians are attracted to politics because of the opportunity for power and prestige. That's not the right kind of personality to serve the public interest. What if we anonymized politicians, made them shave their heads and go around in rags with their faces covered, to make the profession unappealing to ego-driven personalities? What if we prevented politicians from trading stocks while in office, to reduce perverse incentives? What if we surveilled politicians 24/7 and made the surveillance public, to eliminate backroom deals? (Rare exceptions might be made for when the politician must view sensitive military information.) And what if we tried outcome-based prediction markets? Of course, none of this will actually happen, not least of which because the politicians would never vote to reduce their own power.

Expand full comment

Yadira is a civil servant, not a contractor. But yes, similar issues still apply.

Expand full comment

Why not just do like the private sector does? Have a pool of $10B/year that gets paid out to politicians (current and former) on the basis of performance since they entered office. You don't have a convenient proxy like share price, but you could come up with a reasonable weighted performance score easily enough (GDP growth, crime rate, consumer confidence, etc.). As you say, it's silly that we pay these people so little money. It's no wonder corporations have so much influence, and that so few good people enter politics.

Expand full comment

A good plan, though of course there's https://en.wikipedia.org/wiki/Goodhart%27s_law : "When a measure becomes a target, it ceases to be a good measure." For instance, politicians pursuing short term gains to maximize their metrics, that screw over the future. Or politicians deliberately screwing things up in their last days in office, so that the consequences fall on their successor if their successor is a political adversary. (Plenty of that already.)

Rather than performance since the politician entered office, it would be better to run markets of what people *expect* the long-term future performance to be, conditional on different politicians getting elected. And reward the politician based on how much he outperforms the baseline of the market (e.g. let him buy shares, or give him some shares). That way the politician would be neither rewarded nor punished for changes that everyone knows are coming regardless of his actions. The market could also account for long-term consequences beyond the politician's term, unlike using direct performance metrics.

Expand full comment

I know this is a late response, but this is very reminiscent of the ideas laid out in Jane Jacobs underrated dialogue "Systems of Survival." I don't think I could adequately summarize her ideas, but the idea of overhauling state accountability and not mixing commercial and government responsibilities is very similar to Jacobs ideas on guardian vs commercial symptoms, and she would probably describe the current practice of mixing government responsibilities and commercial development software as a "monstrous hybrid." In the context of Jane Jacobs ideas, converting the role of government software development to a guardian role, like a judge or military role, makes sense.

Expand full comment

I actually do work in such a structure: French "state start-ups", small teams who build great agile solutions for the french government.

Funnily enough, I do agree wholeheartedly with Pahlka, while realizing that "hope" is not a great model, it might be the best we have. Government oversight is so ridiculously hard to achieve well, especially in tech when even private-company-oversight is a huge challenge.

Me and my team have been doing great work for the French state. We could work ten times less and bill just as much. We don't, because we are dedicated professionals, and we've been trusted 100%. We are in a great position to do great things, hence we aren't incentivized to breach said trust.

Expand full comment

That was way too many typos to be typos; Google Voice or the like? rapid libertarians

the coffered

has spend

bilious of dollars

to this same things

lack of deskilled technologists

regulte less


Expand full comment

<-"She just says to hire good people (she can suggest who) and free them to do what they see as best. And then hope."

Seeing how well China and Singapore have done with this strategy, I can understand her enthusiasm.

Expand full comment

Just as we need separation of Church and State, we need complete separation of Economy and State. Government “regulation” has been and will always be captured by Big Corporates to build moats against their smaller competitors. Consumers must have the right to sue for damages under tort law to keep Business honest, and have the freedom to vote via their wallets, absent coercive government mandates.

Expand full comment