The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent

Slippery Slope Arguments

Equality Slippery Slopes


[This month, I'm serializing my 2003 Harvard Law Review article, The Mechanisms of the Slippery Slope.]

Multi-peaked slippery slopes can happen when a significant group of people prefers both extremes to the compromise position. One such situation is when A without B seems unfairly discriminatory. Consider the following example:

  • Position 0 is no school choice: the state funds only public schools.
  • Position A is secular school choice: the state funds public schools but also gives parents vouchers that they can take to private secular schools but not to religious schools. (Note that I published this in 2003; this very month, the Supreme Court appears poised to rule that this position is unconstitutionally discriminatory against religion, but let's set that aside for now, and focus on the pure legislative-legislative slippery slope, in which future voters or legislators consider whether to shift from to even without a court decision so ordering.)
  • Position B is total school choice: the state funds public schools but also gives parents vouchers that they can take to any private school, secular or religious. (As it happens, that's my personal preference, but I'm not talking here about what's best normatively—just about whether one's endorsing a move from 0 to may indeed increase the likelihood of others pushing things further from to B.)

And let's say that voter preferences break down just as in the previous example:

Group Most prefers Next preference Most dislikes 0→A A→B 0→B Attitude Voting strength
1 0 A B "As little school choice as possible" 10%
2 0 B A + "No school choice is best, but better total school choice than discriminatory exclusion of religious schools 20%
3 A 0 B + "Secular school choice is better than none, but definitely no inclusion of religious schools" 20%
4 A B 0 + + "Secular school choice is best, but we can live with including religious schools" 10%
5 B 0 A + + "Total school choice is best, but better no school choice than discriminatory exclusion of religious schools" 10%
6 B A 0 + + + "As much school choice as possible" 30%


Because 30% of the voters (groups 2 and 5) have multi-peaked preferences driven by their hostility to discrimination against religious schools, there is an equality slippery slope. Total school choice would have gotten only 50% of the vote (groups 4, 5, and 6) if it had been proposed without the intermediate step of secular school choice. But proceeding one step at a time, we have a 60% vote for secular school choice (groups 3, 4, and 6), and then a 60% vote for total school choice (groups 2, 5, and 6), driven largely by group 2's strong preference for equality.

Once the system has gone all the way to total school choice, group 3 will likely regret its original support for A (secular school choice). Total school choice is the worst option from group 3's perspective, and yet it was group 3's support for the halfway step of secular school choice that made total school choice possible.

This example illustrates that an equality slippery slope can happen even when A and B are distinguishable. Here, a majority of voters concludes that A and B needn't be treated equally; but the slippage happens because a minority (here, 30%) exhibits a multi-peaked preference by preferring either form of equal treatment (0 or B) to unequal treatment (A). {Even the minority that accepts the analogy between A and B could recognize that the two are logically distinguishable, but still conclude that the similarities are substantial enough that the distinction shouldn't lead to a difference in treatment.} Thus, even those who support A on its own, and who believe that A and B can be logically distinguished, might be wise to oppose A if there's enough risk that implementing A will lead others to also end up supporting B.

Consider also the assisted suicide debates, where allowing "those in the final stages of terminal illness who are on life support systems … to hasten their deaths by directing the removal of such systems" (A) has led to arguments that it's wrong for "those who are similarly situated, except for the previous attachment of life sustaining equipment, [to be] not allowed to hasten death by self-administering prescribed drugs" (B). {The Supreme Court has rejected an argument that this distinction is unconstitutional (though two judges on the Second Circuit had accepted it); but if two Second Circuit judges found the equality argument persuasive enough to constitutionally command such equal treatment, at least some listeners may find it persuasive enough to justify such equal treatment as a policy matter, within the context of legislative debate.} Even some of the people who are hesitant about B at first (though probably not those who bitterly oppose B) might also be reluctant, once A is allowed, to deny to some of the dying a release that is offered to others. The acceptance of A may thus increase the chances that B will be enacted, even if A's supporters had sincerely insisted that they were only seeking A and not B.

Likewise, one might reasonably worry that once B (assisted suicide for the terminally ill) was implemented, equality concerns would push some decisionmakers to allow assisted suicide for still more people (C), such as the "chronically ill, who have longer to suffer than the terminally ill, or … individuals who have psychological pain not associated with physical disease"—"[t]o refuse assisted suicide or euthanasia to these individuals would be a form of discrimination." And even if courts can roughly distinguish categories A, B, and C in a way that's generally sensible, though arbitrary in close cases, judges may be reluctant to apply this distinction to a real person whose particular close case they are deciding.

This sort of equality-based slippage seems to have happened in the Netherlands. Dutch courts began by declining to punish doctors who assisted the suicides of the terminally ill. They then extended this principle to cover patients who were victims of "unbearable suffering," without any requirement that the patients be terminally ill. They then extended the principle to cover a patient who was in seemingly irremediable mental pain, caused by chronic depression, alcohol abuse, and prescription drug abuse, on the theory that the suffering of the mentally ill is "subjectively experienced as unbearable" by them, comparably to how the physically ill experience physical suffering.

Dutch courts then extended this principle to cover a fifty-year-old woman who was in mental pain partly caused by the death of her two sons, again on the theory that her suffering was unbearable. "Intolerable psychological suffering is no different from intolerable physical suffering," the doctor in that case reasoned, and the court agreed, concluding that the relevant question was "the irreversibility of the intolerably experienced suffering, not the source of it."

In these examples, the bottom of the equality slippery slope is more government funding or more freedom from restraint, but the slope could also lead toward greater government power and greater restrictions. For instance, when a free speech exception is created for one constituency, others may resent even more the absence of an exception for their own favored cause. Consider one argument in favor of campus speech codes:

Powerful actors like government agencies, the writers' lobby, industries, and so on have always been successful at coining free speech 'exceptions' to suit their interest—copyright, false advertising, words of threat, defamation, libel, plagiarism, words of monopoly, and many others. But the strength of the interest behind these exceptions seems no less than that of a black undergraduate subjected to vicious abuse while walking late at night on campus.

Or consider the similar argument that the existence of the obscenity exception justifies bans on Nazi advocacy because "[t]here is no principled reason to permit the banning of material that appeals to a depraved interest in sex but not the banning of material that appeals to a depraved interest in violence and mass murder."

Some people who make such arguments might have supported proposal B (the creation of a new free speech exception) even had proposal A (the creation of the old free speech exceptions) never been implemented. But their use of the equality argument suggests that they think some listeners might be moved by the analogy between A and B. This attitude may be characterized as a worthy love of consistency, or as unworthy "censorship envy"—but in either case, it is a real phenomenon. So far, U.S. courts have resisted these arguments, but American political leaders, future U.S. courts, and politicians and courts in countries that have a narrower view of free speech may well find them logically and emotionally appealing.