The old saying tells us you have to walk before you run. In American politics, there is a strong bias in favor of incremental change -- taking small steps to learn what works and what doesn't rather than attempting to enact radical reforms in one fell swoop. In his latest paper, however, Dean Saul Levmore suggests that there might be some hidden dangers to the incrementalist model. Not always a cautious avenue towards prudent changes, incrementalism can act as a cover for powerful interest groups to achieve socially disoptimal outcomes through divide and conquer strategies.
Before Dean Levmore makes the bulk of his argument, though, he takes aim at the core mythos that gives incrementalism its positive aura: that it is a valuable teaching tool that allows us to feel our way gradually to the optimal social outcome. Certainly, sometimes incremental changes have this effect. But there is no reason to think that, as a general rule, incremental change is always a superior teacher than alternatives, such as a brief experiment with a dramatic change, or drastically different regulatory schemes across jurisdiction. A month where smoking was prohibited everywhere may well give society more useful information in planning future regulations than an "incremental" prohibition on smoking in restaurants.
Having dispatched the primary internal justification for incrementalism, the question of how incrementalism plays out in a world filled with various interest group rises in prominence. The problem of incrementalism is deeply tied to the problem of reversibility in regulatory events. A regulation is reversible if, upon its repeal, the affected parties can return to the status quo at little or no cost. So a smoking ban in bars, for example, is reversible, because if the ban is repealed, it is quite easy for the bar to revert to a smoker-friendly environment. By contrast, a regulation mandating wheelchair accessible ramps in buildings is not reversible -- even if the law was repealed, a building which had already constructed the ramp would not be able to recover its costs by getting rid of it.
When a regulation is irreversible, the problem of incrementalism manifests. Advocates for building wheelchair accessible ramps might start only by targeting new buildings, or buildings larger than a certain size. Doing this divides the opposition -- owners of older or smaller buildings unaffected by the legislation don't have an incentive to get into the fight -- making it easier to pass. Five years after this bill is passed, advocates press for a new bill covering all buildings. But now the buildings which already have installed ramps are on their side. They have been forced to incur costs that they can't get back -- from a competitive standpoint, they want their rivals to also have to spend money on the ramps. By proceeding incrementally, advocates for wheelchair ramps are not only able to divide the opposition, they are actually able to grow their coalition by pitting competitors against each other. It is easy to see how this dynamic can push regulation beyond its socially optimal level, as resistance to further expansions is continually weakened and all parties begin to have an incentive to engage in rent-seeking behavior.
This problem only occurs where regulation is irreversible. In the case of our smoking ban, by contrast, restaurants fighting a proposed extension to their businesses can entice bars to join the fight by promising in turn to press for a repeal of the original law as well, putting both businesses on equal footing again. Dean Levmore explores the idea of "retroactive compensation" as a way to simulate this dynamic even in the case of irreversible regulations. Retroactive compensation is compensation granted after the original regulation, triggered by another event. In the case of the wheelchair ramps, the newly affected buildings could demand that they only be forced to build the ramps if they were compensated for the cost. They could convince the already affected buildings to join their side by bundling that proposal with one for retroactive compensation for their past expenditures, once again placing both sides on equal footing. Unfortunately, beyond the easiest cases, the difficulties in assessing both what counts as a "reversal", and who has to pay for the compensation (taxpayers? advocates? prior beneficiaries?) make it a tough policy to implement in practice.
Alternatively, Levmore proposes a "disclose and delimit" rule which would force advocates to reveal at the start of their campaign the limits of their agenda, and then forbid them from proceeding beyond those bounds for a specified amount of time. Unfortunately, this solution runs into severe enforcement problems, as the legislatuures which could require the rule are also in the position to ignore it later. Moreover, there might not be perfect overlap (and indeed, possibly antagonism) between the advocates pressing for the original rule, and those who want the extension (though Levmore notes this might actually be an advantage insofar as it encourages more players to take the field early in the campaign process).