I’ve heard it said that at graduation, each newly-minted architect should receive two things: a diploma, and a lifetime supply of dynamite — the diploma to create anew; the dynamite to rub out mistakes.
Does this sentiment apply to legislators as well? In a recent essay, David Hitt suggests that every law have a built-in expiration date (sunset clause). Hitt proposes that under automatic expiry, a debate would be forced on each and every law on the books at the time of its demise. Would this lead to the retention only of that legislation which has passed the test of time?
I’m afraid not, but Hitt does get at something else entirely.
Given standard legislative practice, you’d be more likely to see the following exchange:
—-
Speaker: I hereby put to a vote the Omnibus Extension Act of 2005. As a reminder, the OEA-2005 renews the term of all laws which would otherwise expire this calendar year, extending each for the statutory maximum of 20 more years. Any opposition?
Floor: (sound of crickets chirping)
Speaker: And now, new legislation…
—–
Furthermore, legal code exhibits the same brittleness as compiled software code. Once a component (or law) is written and built into the system, other parts of the system come to depend on this newly created code. Therefore, it is difficult to rip out laws piecemeal at a future date. Hitt’s proposal fails to take into account this interdependency of legal code.
In addition, every change to the legal code entails costs above and beyond the direct cost of the legislation, exclusive, of course, of the need to print new law books! There is the charge for relevant lawyers to digest the legal code, the expense of agencies adapting themselves to whatever changes are necessary, and cost for private individuals and corporations to both change in the present and modify their future planning in light of the new legislation. Automatic expiry of laws is only an instance of this problem — if laws started disappearing willy-nilly, the change costs alone might mark Hitt’s proposal as infeasible.
Despite these problems, Hitt is right that for any given legislator, that there is every incentive to consider additional legislation, and little incentive to reconsider existing legislation. The question has to be: what would give legislators incentives not only to pass new laws only as absolutely necessary, but also to intelligently kill and embalm laws on the books? One way would be if the ongoing cost of running a government could be clearly apportioned to each law. To do this requires a clear causal tie between a law on the books and the cost of that law, in both financial and social terms. As noted above, elucidating this chain of causation can be as difficult as decompiling source from object code. There are always multiple explanations consistent with a given set of evidence, so the question is how to trace back to the source code (the law) the situation in which we find ourselves as a result of the code of law, taken holistically. Don’t lose heart, however — possible solutions will be the topic of a future essay.
What’s really lacking here is a lack of clarity about the cost of ideas. We love to believe in a free market for ideas, and I agree that such a market is an unqualified good. And as long as we remain in the realm of pure thought, there is no downside to their fecundity and immortality. True, they can clog the mind, but the brain is decently good at blurring and eventually rubbing out stale knowledge and ideas. The problem comes when these ideas are reified. Once reified, ideas do entail costs. Once a law is enacted, it determines the actions which people take. These actions do have real consequences, and bills of Congress lead without fail to pushing around bills of Washington, Lincoln, and Jefferson.
This points to a facet of the legislative process which standard treatments underplay, namely the psychological consequences of making lawmaking anybody’s primary, if not sole, creative outlet. As humans, we have a need to create, a need as basic as loving or eating. Often, it is our creativity that gives each of us landmarks in our search for meaning. A corollary: make legislating a person’s dominant creative outlet, and you will see a lot of legislation.
Now, some may see nothing special about this. After all, in today’s society of professionals, for each expert, the exercise of their intellect in their foveal domain is their primary creative outlet. The difference with legislators is that, a priori, there is only one legislature in any given jurisdiction. If a software engineer goes crazy and develops bloated products, go buy another one. If an architect suffers tectonic tics, move to a different house. In these other professional outlets, the cost of excess creativity is inherently limited by the possibility of choice. It is, therefore, much more critical in the domain of government to govern the innovations of the governors. While standard “balance of power” structures keep the ideological bent of the legislative mix within bounds, a balance holds level whether loaded with 20 pounds or 200, as long as the weight is evenly divided. Such relative measures fail to account for the fact that the most perfectly balanced scale will collapse when overloaded. Controls on relative authority fail to constrain the total quantity of laws when all branches of government have a vested interest in passing additional laws.
Hitt’s proposal is one attempt at this, but several others exist. One of the more widely-adopted is the part-time legislature. Arguably, it’s not the time constraint which limits the lawmaking in such beasties, because even if they meet only on occassion, there’s still plenty of time to do mischief. Rather, the limit exists because out of both boredom and financial pressure, no part-time legislator can be unemployed when not in the State House or on Capitol Hill. By virtue of having some other creative outlet, and some other basis for identity, it becomes less necessary for any individual to feel that to create, they must pass laws.
On occasion, a software system becomes so bloated with additional code that the best path forward is to discard it and start anew. It is, in many ways, the least costly solution, but because it has a highly concentrated, short-duration cost, such a total restart can be extremely difficult choice to make. As the law exists, it is impossible for laws to just drop off the books. Further, it would be extremely costly to replace an entire system of code, legal code in this case, with an entirely new design, so it often piles cruft on cruft until the system collapses catastrophically (this is also known as a violent revolution.) What are the means by which this can be avoided? Can lessons from the construction of software systems can be taken to the construction of legal systems? Larry Lessig has talked about how geeks create law, but only in a limited domain. How different would our legislatures look today if they were composed not of legislators with a legal background, but rather of architects, engineers, or artists? The danger is that these new legislators must have empathy, and a broad capacity for empathy may be lacking in such professionals. Without empathy, without a strong ability to imagine themselves in the place of another, it is dangerous to let artists or engineers run the show, because what is acceptable for the objects of their art or design is forbidden when applied to other humans. Without empathy, we threaten this.