One of the topics that came up at the Open Borders Bay Area Meetup (Sunday January 12) was the idea of slippery slopes towards open borders. My post is based on ideas raised in the discussions, but since the participants’ comments were largely off the record, I’ll take full responsibility (but not necessarily credit) for everything I write here.

Immediate, unconditional open borders, or even something fairly close to it, is unlikely to materialize in the near future. From a practical perspective, then, the best route to open borders is a long series of gradual steps. One key concern of people who have their eyes set on the end goal of open borders is that the first few steps should make future steps easy to achieve. A path that is ostensibly in the direction of open borders, but where the first few steps might lead to a shift backward (such as a nativist backlash that turns public opinion against more liberal migration) would be a bad path. A better choice would be a slippery slope: one where the initial steps may not necessarily look as “extreme” as open borders to the people who might object to them, but once those steps are taken, the next steps towards open borders become easier to accomplish. (UPDATE written after the post was drafted: Tyler Cowen’s recent post makes this type of backlash argument to explain a Swiss referendum to restrict immigration from the EU; see also responses from Bryan Caplan and Jason Brennan).

Some possible slippery slopes

• Gradual expansion and merging of free migration zones: For instance, the EU is a free migration zone for European countries, and it is gradually adding countries. Suppose the United States and Canada created their own free migration zone, Southeast Asian countries created their own free migration zone, and a few free migration zones emerged in Africa and Latin America (South America). The US-Canada free migration zone could, after some time, add Mexico and the Caribbean Islands, and then merge with the South American free migration zone. The European free migration zone could eventually merge with the new unified American migration zone. Over time, as the threat of terrorism, and the subjective sense of the threat, receded more and more into the past, the Gulf states could merge with the European free migration zone. And so on, until eventually the whole world would be a free migration zone.
• Gradual lowering of admissions standards for migrants: The idea here would be to begin with complete open borders for migrants who meet a particular skill level threshold or income threshold, and then to gradually reduce the threshold over time, either directly, or by failing to index for inflation, or by relying on global economic growth. A plausible argument for why this might work is that high-skilled and high-income migrants lead to fewer problems, both real and perceived, for natives, making the natives somewhat more pro-migrant (based on their experience interacting with migrants so far). It might therefore be easier to cross historical thresholds of migrant proportions in the population without any visible difficulties experienced by natives. Australia, Canada, and Singapore might be examples to look at, and it would be interesting to see the extent to which they avoid the nativist backlash phenomenon through an initial focus on high-skilled migration.
• Gradual increase in quotas: Hard quotas in various migrant categories may be gradually expanded, and new categories that permit more general forms of migration may be added, initially with low numerical caps but with the caps steadily rising. Although this could in principle be huge, it doesn’t seem to have had much slipperiness historically in and of itself.
• Family reunification, with a gradually expanding definition of “family”: Family reunification may begin small, with, for instance, the spouses or children, then it may expand to siblings, parents, or cousins, and finally it may expand to close friends or anybody whom a resident can provide some sort of sponsorship guarantee as a “personal acquaintance.” For a fixed definition of family, too, the amount of “chain migration” can be significant, and although insufficiently slippery to get to open borders all by itself, it can act as an amplifier to any other slippery tendencies. Quantitatively, this has been significant in both the US and Europe, and quite difficult to stop, even though governments have often attempted to crack down on various aspects of family reunification to prevent what they see as abuse.
• A tariff/tax scheme with a gradually declining tariff or tax rate: A DRITI-type scheme that begins with high surtaxes and high mandatory savings, to a level that meets the concerns both of restrictionists and people who want the government to earn revenue. If the revenue-earning part of it goes well, that would create a strong impetus to create and expand the program, and lower the surtaxes and mandatory savings in order to attract more migrants and generate more revenue. Even though staunch restrictionists may oppose that second step, governments and the majority of voters may have become too addicted to the revenue streams and expanded economy size afforded by the program to backtrack from it. Over time, the rates would come down to the point that it would look more like a bureaucratic inconvenience than a program specifically designed to regulate migration. At this stage, general egalitarian concerns, or general philosophical views about taxation, may take over and lead to an abolition of the surtax and of the forced savings program.

Reasons for continued slipping

• Growing tangible benefits (real or perceived) feeding into growing support for more migration at the margin: People see tangible benefits from the expansion so far, and want to keep going further in that direction. This may be because the constituency that has been created by partial expansion of migration is more eager for further expansion than the original set of natives (for instance, recent migrants might favor more liberal migration policies — something that seems empirically borne out). It may be because people get addicted to the revenues from a DRITI-like scheme and feel the need to expand it.
• Support for steady expansion as long as there are no tangible or major harms, based on a presumption that ceteris paribus, allowing more people to migrate is better: People don’t actually see much by way of tangible benefits, but they don’t see any tangible changes at all, so, given that they concede some sort of “presumption in favor of cautious expansion of migration as long as there are no visible drastic damages” they are happy to let it expand. Or perhaps the people aren’t even that aware of the exact state of things, but their representatives in government are willing to let migration gradually expand barring any unexpected developments.
• Slippery slope in people’s perception of the forms of migration that morality or justice require should not be forbidden: Each step in the slippery slope changes people’s perception of the moral necessity of open borders. For instance, people may go from “it’s unjust to forbid somebody from migrating if they have family in the country or are fleeing persecution” to “it’s unjust to forbid somebody from migrating if they are going to be a net fiscal plus for the nation” to “it’s unjust to forbid somebody from migrating if they’re honest and hardworking” to “it’s unjust to forbid somebody from migrating unless it can be demonstrated that they pose a significant threat to the life and liberty of people in the country.”
• Increasing share of the decision-making population that has a personal positive connection with migration: Migrants tend to be somewhat more pro-migration than natives (see the quote below), so increasing the proportion of migrants would increase the proportion of residents with pro-migration views. This may not have much effect alone, because migrants may be ineligible for direct participation in political decision-making, but they could influence the views of others they interact with. Natives who interact with migrants on a personal level may also become more pro-migration simply because of the “human face” of the issue, even if the migrants themselves don’t try to convince them. The children of migrants are also somewhat more pro-migration than natives. A caveat: the support for further migration may be limited to the migrant’s ethnic, racial, or religious group or socio-economic class.

The following quote from Bryan Caplan’s Cato Journal piece is relevant to the assertion made above about migrants having more pro-migration views:

Finally, there is at least one issue where immigrants are sharply more libertarian than natives: immigration itself. Materially, recent immigrants have the most to lose from additional immigration. Ottaviano and Peri (2008: 59) estimate that immigration from 1990–2006 depressed foreign-born workers’ wages by over 7 percent. But immigrants, like human beings generally, do not derive their political philosophies from material self-interest (Mansbridge 1990). The General Social Survey asks respondents to put their views on immigration on a 1–5 scale, with 5 being most hostile. People with two native-born parents have an average response of 3.9, with a
median of 4; people with at least one foreign-born parent have an average response of 3.1, with a median of 3. By way of comparison, people who call themselves “extremely liberal” have an average response of 3.3—versus 4.0 for the “extremely conservative.” People with foreign-born parents rarely favor open borders, but economists
and libertarians aside, no one is less opposed to immigration.

Reasons why we might expect slipping to not happen

Slipping is not an inevitability of nature. Without concerted effort, it may not happen. Some reasons:

• It is very difficult to keep making changes steadily. In general, we should expect change in fits and starts.
• As the proportion of migrants in the population reaches historically unprecedented levels, migration becomes a lot more visible as a phenomenon. Even with a moderate level of anti-foreign bias, people are likely to blame migration for any problems that arise. Also, appear to historical analogy fails to be a powerful argument for further expansion. The possibility of nativist backlash remains strong, and recent pushback to migration in the UK and Australia might be evidence of some natural limits in the amount of migration people will naturally tolerate given their current beliefs about how the world works.

Braking mechanisms

Braking mechanisms could matter for two reasons. First, we may be genuinely unsure of the extent to which open borders would be a blessing, even if we think it’s quite likely to be one. In this case, having a slope that slips by default, but that can be braked if things are turning out badly would be helpful. Just how much importance we place on a braking mechanism depends on what level of badness we deem sufficient to halt further progress. That level could vary between hardcore open borders advocates and people whose support for free migration is highly conditional to specific consequences.

Second, even if we (as open borders advocates) don’t believe that the braking mechanism would be necessary, others who are more skeptical of the change might be reassured by the presence of a braking mechanism. The braking mechanism would be like the chain passengers can pull for emergency-braking in some trains: it is rarely pulled, but its presence adds to the (perception of) safety (though the analogy also highlights that the braking mechanism may end up getting used for spurious reasons, just as miscreants misuse emergency brakes to get down wherever it’s convenient for them).

There could be different senses of braking and reversal:

• Outcome reversal: Additional migrants who have come in through the expansion so far are deported, bringing the situation as close as possible to how it would have been if the expansion had never occurred. Note: This isn’t intended to endorse such “outcome reversal” as either moral or even practically feasible; see the discussion below.
• Policy reversal: The migration policy for new migrants reverts to what it was before we started slipping the slope, but people who already migrated under the temporarily more liberal migration policy are not forced to leave. A slight variant of policy reversal, that I call punitive policy reversal, would be where policy is not just reverted, but made even more strict than it was in the past until such time as we effectively have an outcome reversal. For instance, if quotas are doubled for the next three years, and people then want to reverse the effects, quotas may be set at zero for the three years after that, and then returned to current levels. Note that punitive policy reversal can accomplish outcome reversal only in cases where the original quota was nonzero, and is best suited to numbers-based increases rather than the addition of totally new categories of migration.
• Policy freeze: The policy remains stuck at the new, expanded level, but further expansion becomes difficult.

(For the mathematically inclined, each sense references the derivative of the preceding one with respect to time).

(For an example of confusion that arose because of people in a conversation talking of reversibility in different senses, see the comments starting here).

Outcome reversal is nearly always difficult and often impossible to accomplish directly. Even if it were possible to force all the additional migrants to leave, the level of (visible) coercion and disruption necessary could give pause even to people highly critical of the expanded migration policy. Even those who wish to accomplish outcome reversal may end up suggesting the route of punitive policy reversal. Policy reversal is somewhat easier to accomplish if people feel the expansion has not worked well. Policy freeze, where further liberalization of migration is blocked, might be the easiest in so far as status quo bias argues in favor of it.

1. Gradual expansion of free migration zones can be stopped by the participating countries, even though it would be relatively hard to undo an existing free migration zone. Outcome reversal is near-impossible, policy reversal is very difficult, but policy freeze is relatively easy.
2. Gradual lowering of admissions standards can be stopped, and even reversed, if the results are deemed bad. Outcome reversal is near-impossible, policy reversal is of moderate difficulty (less so than with free migration zones), and policy freeze is relatively easy.
3. Gradual increases in quotas can be stopped, and even reversed. Outcome reversal is near-impossible, policy reversal is of moderate difficulty (perhaps a little easier than with admissions standards, because “how to get in” is anyway a sufficient mystery that people don’t feel they are being deprived of something they were definitively promised), and policy freeze is relatively easy. Note that punitive policy reversal can be accomplished in principle in cases where the original quota was substantially greater than zero anyway, and would be easier than outcome reversal through deportation, but would still be pretty difficult.
4. Family reunification would be difficult to undo. Outcome reversal would be near-impossible, policy reversal would be fairly difficult (though not impossible, and it might be easier to partially reverse at later stages), and policy freeze would be relatively easy.
5. For tariff/tax schemes, a continued reduction in the tariff or tax rate will happen only if there is sufficiently high revenue (so that it’s a quantitatively significant source of revenue, making revenue considerations more important relative to other considerations) plus strong evidence to suggest that lowering the tax or tariff rate would attract even more revenue by increasing the “customer” base considerably. Here, outcome reversal is near-impossible, while policy reversal and policy freeze are relatively easy.

The type of braking mechanism available for the policy might be a consideration in the choice of slippery slope. If you’re fairly confident that the braking mechanism is more likely to be abused than used out of necessity, then it makes sense to choose a slippery slope where braking is harder. If, on the other hand, you anticipate that the initial stages could reveal information that could lead to the need to brake, it makes sense to choose a slippery slope where braking is easier.

The utility of different slippery slopes in relation to estimates of the global impact of open borders

In a recent post, I argued based on Carl Shulman’s post that, since existing estimates of the global economic impact of open borders rely on more people moving than may be realistic, estimates may have to be adjusted downward to 20-60%. Whether this affects the overall utility of a particular partial move towards open borders depends on whether that move is targeted at increasing numerical caps on how many people may move, or at removing other policy barriers to movement (whose effect in terms of the number of people that move at the margin might change once we correct our estimates). In the Drake equation language, we are interested in the product $latex Wxy$, and the downward adjustment of $latex W$ might be accompanied by different compensating adjustments for $latex y$ depending on the sort of measure. The sensitivity of expected utility to uncertain estimates of the global economic impact of open borders might be a consideration in comparing different slippery slopes.

Slippery slopes versus keyhole solutions

On this site, we’ve discussed keyhole solutions a lot. Some people might use “keyhole solutions” and “slippery slopes” interchangeably, but there is a difference:

• A slippery slope involves a series of steps with each step liberalizing migration relative to the previous one, and with the end goal coming reasonably close to open borders. It represents a means or path.
• A keyhole solution is a (typically policy-based) modification of pure open borders that addresses a significant real or perceived problem with open borders while disrupting open borders as little as possible. It represents an end goal.

Slippery slopes and keyhole solutions could be related: the final stages of a slippery slope might resemble “open borders with a keyhole solution” and the initial stages might also be so described if “keyhole” is interpreted very generously. Also, some keyhole solutions with adjustable parameters (such as the surtax and mandatory savings parameters in DRITI) may lend themselves to a slippery slope: begin with highly restrictionist values of the parameters (so restrictionist that it would be a stretch to call it a “keyhole solution”), then gradually move the parameters in the direction of open borders.

PS: Also relevant is our moderate versus radical open borders page and Joel Newman’s blog post If Open Borders Are Instituted Gradually, What Should Be The Initial Number of Immigrants Admitted?
Thanks to Carl Shulman for helpful comments on a draft.

Vipul Naik

Vipul Naik founded the Open Borders website in March 2012. See also:

Vipul Naik’s personal statement
Vipul Naik’s description of open borders on his personal site
Page about Vipul Naik on Open Borders