Log in

No account? Create an account

Previous Entry | Next Entry

The Revision History of Democracy

Apart from being one of our nation's founding documents and an eloquent call to resistance against tyranny, the Declaration of Independence happens to be a great template for how to structure a logical argument as a springboard for further action. Its form is that of a syllogism, one of the oldest methods of deductive reasoning. First it establishes the major premise, a proposition which sets up the second half of the conclusion. Paraphrased, it looks like this:

All governments which fail to protect the natural rights of their citizens, or derive their powers from their citizens' consent, must be altered or replaced.
Then the minor premise, a proposition which sets up the first half of the conclusion. I'll just quote directly this time:

The history of the present King of Great Britain is a history of repeated injuries and usurpations, all having in direct object the establishment of an absolute Tyranny over these States.
Now the rules of inference apply: the King's actions harm his American citizens; citizens who are being harmed by their government have a duty to alter or replace that government; therefore, the inevitable conclusion is

That these United Colonies are, and of Right ought to be Free and Independent States; that they are Absolved from all Allegiance to the British Crown, and that all political connection between them and the State of Great Britain, is and ought to be totally dissolved; and that as Free and Independent States, they have full Power to levy War, conclude Peace, contract Alliances, establish Commerce, and to do all other Acts and Things which Independent States may of right do.
But right now I'm not as interested in the overall logical structure as I am in the snapshot of history that occupies the twenty-seven points that Jefferson included as support for the minor premise. Why? Because they're the first entries in the United States' bug tracking system. Each one is a complaint from users, and more importantly, the Constitution provides patches for each one. A few examples:
  • "He has refused his Assent to Laws" — Article 1, Section 7. Congress can override a presidential veto.
  • "He has made Judges dependent on his Will alone" — Article 3. The judiciary and its independence are established by Constitutional authority, not that of the executive.
  • "He has affected to render the Military independent of and superior to the Civil power" — Article 2, Section 2. The president is Commander-in-Chief of the military and it is subject to the executive branch.
  • "For Quartering large bodies of armed troops among us" — Third Amendment. Bans this practice entirely.
The Declaration calls out the abuses of tyrannical individuals, and the Constitution aims to establish a system of government — a protocol, if you'll forgive the nerdy indulgence — that is resistant to bad actors. But over the last 200+ years, we've discovered a terrible, terrible thing: the Constitution is not sufficiently Byzantine-resistant.

No, no, put the guns down, we're not being invaded by ancient Greeks. If a protocol is Byzantine-resistant, then it still does what it's supposed to even when some pieces fail on their own or are compromised by an outside party. The checks and balances established in the Constitution are an early example of this: if any one branch is compromised, the other two can halt or reverse actions taken by that branch. The judiciary can even revert actions taken by a compromised legislature and executive acting in concert, as long as a member of the citizenry is willing to bring action — the extra check on the incredible power of the judiciary is that the people have to invoke it.

But all these systems have a lot of moving parts, and what we've discovered, to our horror and disbelief, is that it's not necessary for an attacker to compromise an entire branch of government in order to bend the government to its interests without regard to the interests of the citizenry. In fact, the amount of effort required is comparatively quite small. We can break this down with a small abstract example. Suppose that there's a network made up of several thousand nodes, of types A, B, and C. Controlling the entire network so that it always spits out whatever output you want, in any situation, requires you to take over at least half of the A nodes, at least half of the B nodes, and some number of C nodes. However, if you only want to control some outputs in certain situations, then you only need to subvert a particular 10%1 of the A nodes and a particular 10% of the B nodes. If you control any C nodes, that's gravy.

Oh, and each node advertises its function.

Since it's easy to identify which nodes to target, a smart attacker will compromise only the exact nodes it needs to control in order to get the results it wants. Anything else is a waste of effort. This beautiful system, designed to resist tyranny by making all actions contingent upon the will of an accountable majority, has a giant, gaping flaw that renders it vulnerable to having its actions controlled by a small but clever minority.

The name of this vulnerability is regulatory capture.

Regulations are established by offices within the executive branch. They are not laws. Congress can amend regulations, but there is no public accountability for the regulations themselves. Nor is there any accountability for other actions regulatory agencies take; the populace cannot vote in a different chairman of the FCC during the next election cycle if the previous one refused to investigate telcos that conspired with the NSA to wiretap U.S. citizens illegally, nor can they call for a vote of no confidence in P.J. Crowley or Janet Napolitano for dropping millions of dollars on possibly dangerous and widely hated full-body scanners peddled by former DHS head Michael Chertoff when it turns out that the scanners don't even do what Chertoff claims they do.

It isn't in the telcos' interests to let their customers know "oh, by the way, we were snooping on your phone calls and email." It's bad for business. It is in Rapiscan's interest, as well as Michael Chertoff's personally, to grab as big a piece of the TSA's $300 million advanced-imaging-technology pie as they possibly can. The safety, privacy, and other fundamental human rights of the citizens — whose interests are supposed to be protected by the regulatory agencies! — have been left by the wayside, because the agencies have fallen victim to regulatory capture.

Now, there are perfectly good practical reasons to have regulatory agencies. If you're going to have publicly funded infrastructure of any kind, it follows that there will be policies on how money will be spent on building and maintaining that infrastructure, how the infrastructure will be operated, &c. The postal service (whose regulatory authority over the mails is implied in the Postal Clause of the Constitution) is one of the less broken examples of this; it's run afoul of the First Amendment in the past, but the Supreme Court has been vigilant about this in the last century. However, the postal service is almost unique among United States regulatory agencies in one respect: it regulates itself and individuals, but not corporations. USPS holds nearly no regulatory authority2 over Federal Express, UPS or DHL. (Technically, these aren't even mail carriers. USPS holds a legal monopoly over non-urgent letters.) This is why you never hear anyone railing about "postal special interests" on the news, the way you do about the pharmaceutical industry or the insurance industry. There is little that a package carrier can do to stack the deck in its own favor. (The flip side, of course, is that USPS can't make guarantees about contract carriers. You have a guarantee of privacy in USPS-delivered mail that you don't from FedEx. Except from postal inspectors, of course; who watches the watchmen? Which is why I said it's less broken.)

Cue defendants of regulatory agencies citing all the reasons why the FDA, FCC and so on were established in the first place. You'll talk about it in the comments anyway and far be it from me to stop you, but keep in mind, I'm already well aware. I've read my Upton Sinclair, and when it comes to the FCC you might say I'm a little obsessive. We attempted to establish regulatory agencies to protect the rights and interests of citizens. This intention is noble, but the implementation has failed. The onus now is on us, as citizens, to track down the bug — or bugs — that allow regulatory capture attacks to succeed, file the report, and if the current maintainers are unable or unwilling to implement a patch, replace the maintainers with ones who will.

If we can't pull that off, I suppose we could just fork the codebase.

ETA: bramcohen tweets a fascinating article about spontaneous order, superlinear growth, and economies of scale with respect to cities. Tangentially related, but related all the same.

110% is probably a huge overestimate in terms of raw numbers. However, the cost to compromise a node varies, tending to increase as you go up in the bureaucratic hierarchy. It takes a lot of social capital, and probably a lot of money, to influence the Secretary of the Interior. It takes a lot less to influence the director of a national park, so if you want that contract to build a new lodge, you're better off schmoozing the director. And if all you want to do is dump hazardous waste, you can probably buy off a couple of park rangers much more cheaply than disposing of it properly would cost.
2USPS gets to decide what constitutes a hazardous good, e.g., ammunition.


( 53 comments — Leave a comment )
Dec. 18th, 2010 09:08 pm (UTC)
Genius. You. Are.
Dec. 18th, 2010 09:39 pm (UTC)
Aw, I'm just assembling puzzle-pieces. Regulatory capture's been a known problem since the 1950s, and Pease, Shostak and Lamport described the Byzantine generals problem when I was three. I have to think about Byzantine attacks in a lot of the threat modeling I do in my research, so it just seemed like the natural thing to do. But thank you. :)
(no subject) - jodi_davis - Dec. 18th, 2010 09:44 pm (UTC) - Expand
(no subject) - morbid_curious - Dec. 18th, 2010 11:16 pm (UTC) - Expand
(no subject) - maradydd - Dec. 18th, 2010 11:26 pm (UTC) - Expand
(no subject) - ndgmtlcd - Dec. 19th, 2010 02:17 am (UTC) - Expand
(no subject) - maradydd - Dec. 19th, 2010 02:31 am (UTC) - Expand
(no subject) - smallmercies - Dec. 19th, 2010 11:48 pm (UTC) - Expand
Dec. 19th, 2010 12:38 am (UTC)
There is fairly strong evidence that much of the movement for regulatory agencies, back in the 19th century, was primarily driven by the desire of large business firms for effective cartelization arrangements. "Wicked big businessmen are despoiling the public; government has to protect us!" was just the sales pitch. See for example Gabriel Kolko's Railroads and Regulation, or more recently Spulber and Yoo's discussion of AT&T in their book on networks and telecommunications. That is, it's not just regulatory capture; a lot of regulatory programs are born already captive.
Dec. 19th, 2010 12:44 am (UTC)
I've read a bunch about the Federal Reserve lately, and it followed the same path. "Protect the banks' business interests" was the real goal, but not a great sales pitch.
(no subject) - maradydd - Dec. 19th, 2010 02:32 am (UTC) - Expand
(no subject) - darthzeth - Dec. 19th, 2010 02:48 am (UTC) - Expand
(no subject) - maradydd - Dec. 19th, 2010 03:13 am (UTC) - Expand
(no subject) - darthzeth - Dec. 19th, 2010 03:43 am (UTC) - Expand
(no subject) - maradydd - Dec. 19th, 2010 04:58 am (UTC) - Expand
(no subject) - jrtom - Dec. 20th, 2010 12:30 am (UTC) - Expand
(no subject) - whswhs - Dec. 20th, 2010 08:55 pm (UTC) - Expand
(no subject) - vatine - Dec. 19th, 2010 02:23 pm (UTC) - Expand
Dec. 19th, 2010 02:24 am (UTC)
It's a hard sales pitch to counter, too, because anyone who argues against expanding the power of regulatory agencies quickly ends up on the wrong end of a "so when did you stop beating your wife?" flavour of accusation.

I've been struggling to think of an appropriate check on this part of the executive branch, but all I can come up with is some form of referendum. Even when the legislature tries to check a regulatory agency, the agency often just says "no," as happened when Congress asked the FCC to investigate the telcos' complicity with the NSA.
(no subject) - whswhs - Dec. 19th, 2010 02:40 am (UTC) - Expand
(no subject) - maradydd - Dec. 19th, 2010 03:11 am (UTC) - Expand
(no subject) - whswhs - Dec. 19th, 2010 04:36 pm (UTC) - Expand
Dec. 19th, 2010 11:04 am (UTC)
The Chairman of FCC is appointed by (and serves at the pleasure of) the president, and has to be confirmed by the senate. So you can't say that there's no accountability, but clearly there isn't a sufficient amount.

I also wonder how much of the effects you speak of stem from regulatory capture, rather than from the general problem of lobbying / special interests—it's not uncommon for congress to enact laws that protect corporations against the public interest.
Dec. 19th, 2010 01:54 pm (UTC)
[links] Link salad remembers eating lamb yesterday
User jaylake referenced to your post from [links] Link salad remembers eating lamb yesterday saying: [...] — A very strange photo from x planes. The Revision History of Democracy [...]
Dec. 19th, 2010 05:04 pm (UTC)
Incidentally, your argument about the specific details of the Constitution and the Bill of Rights being fixes for the problems that the Declaration of Independence identifies strikes me as a really fascinating one that I have not seen before. Would you feel able to develop it systematically at some point? I would enjoy seeing a full length essay. In fact, I have in mind a site that could suitably carry it and might well be interested; e-mail me if you'd like details.
Dec. 19th, 2010 06:12 pm (UTC)
The Constitution made a lot more sense to me when viewed in it's historical context as trying to fix specific abuses during it's time. It also tries to fix problems with the Articles of Confederation, as well.
(no subject) - maradydd - Dec. 21st, 2010 01:49 am (UTC) - Expand
Dec. 20th, 2010 12:26 am (UTC)
Another interesting aspect of the Declaration of Independence: almost all of the claims made in it against the crown are stark, raving, paranoid nonsense which bore little to no connection to reality and were the 18th century equivalent of Troofer conspiracy theories. That may have something to do with the later failure of the Constitutional structure as well.
Dec. 20th, 2010 08:57 pm (UTC)
I think that claim really needs documentation.
Documentation (Part 1 of 2) - kenshi - Dec. 20th, 2010 10:14 pm (UTC) - Expand
Re: Documentation (Part 1 of 2) - whswhs - Dec. 20th, 2010 11:33 pm (UTC) - Expand
Re: Documentation (Part 1 of 2) - kenshi - Dec. 20th, 2010 11:47 pm (UTC) - Expand
Documentation (Part 2 of 2) - kenshi - Dec. 20th, 2010 10:14 pm (UTC) - Expand
Re: Documentation (Part 2 of 2) - whswhs - Dec. 21st, 2010 12:52 am (UTC) - Expand
Re: Documentation (Part 2 of 2) - kenshi - Dec. 21st, 2010 01:23 am (UTC) - Expand
Re: Documentation (Part 2 of 2) - maradydd - Dec. 21st, 2010 02:06 am (UTC) - Expand
Re: Documentation (Part 2 of 2) - kenshi - Dec. 21st, 2010 02:55 am (UTC) - Expand
Re: Documentation (Part 2 of 2) - whswhs - Dec. 21st, 2010 04:58 am (UTC) - Expand
Dec. 20th, 2010 12:28 am (UTC)
This is a very interesting perspective on (Constitutional) law.

A couple of thoughts that occur to me, using the same lens:

(1) It seems like a lot of our system of law is basically a conglomeration of special cases. It seems like there ought to be a way to design the system so that while we include as many tests as we like, and add new ones as we develop new criteria, the law itself is not built in this way.
(Reality check: this would presumably work better if we weren't using natural language, and if the facts were always available and never under dispute. Ah, well.)

(2) Extending the above analogy, perhaps one thing that might help would be to be able to have a limited constitutional convention: mark certain parts of the codebase as 'frozen' and others as subject to change. (Mmm. Nomic.)

(3) I'd like to be able to mark certain pieces of the law as 'deprecated'. :)

On a related note, I'm not sure I would actually want to fork this codebase. It's kind of a mess.
Oct. 24th, 2015 10:40 pm (UTC)
I agree that we are not being invaded by the Greeks, but I heartily disagree with your request to put the guns down.
Oct. 25th, 2015 05:24 pm (UTC)
This is a fascinating argument. I've generally thought of regulatory capture as being a consequence of other sorts of corruption (campaign finance, etc.). And I think that's how reformers generally think about it too.

After all if the executive makes the appointments, s/he could appoint academics, idealists, prosecutors, and every other kind of informed critic whose career does not depend on industry approval. But instead we end up with industry-friendly regulators, mostly as favors to funders.

But I think you're arguing that even in a situation where the executive, judicial, and legislative nodes were basically okay, regulatory capture would still occur. So we have to think of a separate patch for this issue.

And I think that's right. There are a limited number of positions any powerful interest has to capture, and there are a number of techniques that don't even involve legislators or executive, such as

- covertly influencing research and information gathering (e.g. funding a university);
- coordinating among industry to present misleading information to the regulator;
- obtaining a cushy job for the regulator after their appointment is over

The libertarian pull request would be to remove such agencies. Maybe worth considering, in many cases. It's not clear to me if education needs to be regulated the way it is now.

In cases where we still need some sort of operating rules (climate change, securities exchange), perhaps regulation could be tied to legislation somehow. For instance, there's a concept of "failure standards" for laws; you must declare in advance a metric of what you're trying to achieve; if legislation + regulation is not having the effect, it's declared to have failed, and it comes up for immediate review. This might get a responsible legislature (responsible in the legal sense) very interested in the details of the regulation, and its execution.

That's more of a tweak than a real fix though. One thing that's changed in the last 50 years, I think to the detriment of good government, is that government jobs aren't prestigious any more. I'm not sure where the death spiral starts, but it's not a place where smart and ambitious people go to get things done any more. And that creates some kind of self-fulfilling prophecy. So being a regulator is just a job that either people of inferior qualifications do (apparently this happens a lot at the SEC) or something that people do as a career move to be rewarded by industry later. How do we recreate regulation as a real career track?

Edited at 2015-10-25 05:26 pm (UTC)
Oct. 25th, 2015 05:26 pm (UTC)
One even sees capture occurring in non-governmental organisations. Boards of directors can be captured, for instance, whether of nonprofit or for-profit corporations. It seems to me that "structural capture" captures (har!) the essential commonality: some measure of influence over the decision-making process of a decision-making body, whether through that's over the people who constitute the body or the rules they and the rest of the body have to follow.
( 53 comments — Leave a comment )

Latest Month

July 2015


Powered by LiveJournal.com
Designed by Tiffany Chow