Open borders and the viability of democracy

Let me raise a very large question: Is democracy viable under open borders?

Of course the question has no single answer: terms need to be defined, for one thing.

Democracy nowadays is usually taken to mean “one person, one vote,” which in turn seems like a political expression of the claim, in the Declaration of Independence, that “all men are created equal.” Yet in the US Senate, voters in small states have 10 times as much per capita representation as voters from large states, or more. Nonetheless, I think most Americans would regard with horror the idea that anyone could be given 2 votes, or 7, or 0.3. The clause in the Constitution which reads…

Representatives and direct Taxes shall be apportioned among the several States which may be included within this Union, according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons, including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons.

… is regarded with horror, rightly, but I think for somewhat the wrong reason. The right reason for horror is that the Constitution seems to be tacitly condoning slavery. Worse, the slaves’ weight in this count increased masters’ representation. The wrong reason to be offended by the clause is to assume that it implies that black slaves were only three-fifths of a human. Logically, this is a nonsequitur, particularly since the slaves didn’t get to exercise three-fifths of a vote. If the slaves had been able to vote, but their votes had counted only three-fifths as much as those of free whites, that would have been much better than the system that actually prevailed in the Old South.

In fact, I would raise the question: is there really anything so wrong with giving different people different numbers of votes? If a municipality gave long-term residents two votes, recent arrivals one, would there be anything especially wicked about that? What if an electoral system gave an extra vote to anyone who had voted for a losing candidate in the most recent election, as a means of preventing tyranny of the majority? If there is some reason to regard such practices as unwise or even immoral, should “one person, one vote” be considered part of the definition of democracy, or is it just a desirable feature that some democracies might have?

This is relevant to open borders, because one way to safeguard a country’s institutions while opening the borders to immigrants might be to let immigrants vote, but give them fewer votes than natives. That way, if a very large number of immigrants came, say 200 million in the US case, they could have a political voice while leaving natives secure in their electoral dominance, with possible benefits for institutional continuity. But whether this represents a way that democracy could be viable under open borders depends on whether “one person, one vote” is part of the definition of democracy.

Loosely speaking, democracy seems to mean everyone can vote, as opposed to, say, gentry or aristocrats or oligarchs. Nowadays Britain is usually considered to have become democratic only in the later 19th century as the franchise was extended to all or almost all adult males, but of course Britain had voting for centuries before that. So, for that matter, did ancient Sparta, after its own fashion. Indeed, while to say that ancient Athens was a democracy and ancient Sparta was not captures an important truth, it is not a truth easily defined, let alone reconciled with any modern definitions of democracy, for the Athenians had lots of slaves and there were also many metics (foreigners) in the city: by this estimate, less than one-third of the population were citizens, and of those only the males could vote. For citizens, though, Athenian politics was highly participatory and inclusive; much more so than any democratic polity that exists today. Sparta was also probably more participatory and inclusive than any modern democracy for the members of its fierce citizen-elite lording it over masses of helots with whom the Spartan state was nominally in a state of war for centuries, so as to absolve its Krypteia or secret service of manslaughter as they made routine killings of uppity helots. If you’re going to deny Sparta the status of democracy on the “one man, one vote” criterion, Athens fails that test, too. Bottom line: democracy means everyone can vote, but who is “everyone?” All “citizens”– however defined? All residents? All natives? All whom the state regards as subject to its authority? Probably no answer could be offered which would correspond with all our traditional intuitions and historical judgments about what polities qualify as democratic.

As important as these questions are, I’ll put them to one side, because I want to highlight a mind-blowing conundrum which is rarely noticed, namely, how odd it is that democracy ever works at all. Public choice has exposed just how strange, how paradoxical almost, it is, that democracy works at all.

First exhibit: the paradox of voting. People aren’t paid to vote, nor, in most countries, are they required to vote. Voting isn’t very costly, of course, but it does involve some trouble. One has to go to the polling station. One vote has almost no chance of tipping an election. So why bother? If people were “rational agents,” in the economist’s sense, they’d stay home.

Many reasons have been proposed: civic duty, expressive voting, altruism. Altruism is my favorite because it lends itself to mathematical expression, and because of the nifty fact that a tiny degree of altruism suffices to make you vote. Suppose you believe that if Good Guy wins, all your 10 million fellow citizens will, on average, reap benefits worth $10,000 relative to the situation if Bad Guy wins. You’ll also benefit by $10,000. And you estimate your chances of tipping the election at 1 in 1 million. If you’re selfish, you value your vote at $10,000 / 1 million = 1 cent. Your time, plus the gas money to get to the voting booth, let alone watching the news to assess candidates, is worth much more than that. You won’t vote. Now suppose you’re slightly altruistic, valuing the welfare of other people 1% as much as your own. Now you value your vote at 1% * $100 billion / 1 million = $1,000. That’s probably enough to get you to drive to the polling station.

Second exhibit, related: rational ignorance. Again, odds of tipping the election are next to nil, why bother to follow politics and get informed? Here I have three answers: (a) people learn about the laws to which they are subject, in order to comply with them, and become informed voters as a side-effect of this; (b) people follow politics because it’s interesting; and (c) altruism again. I think there’s much truth in Bryan Caplan’s Myth of the Rational Voter argument that people are “rationally irrational” in politics because they don’t suffer the consequences of their choices. Your vote won’t tip the election, so you might as well vote to make yourself feel good, supporting pious, bone-headed policies that sound good but don’t work. But a lot of people do take an interest in politics and put some effort into making their own choices and trying to persuade others. (My own family seems to talk more about politics these days than anything else.)

Third exhibit: cycling. Suppose three voters, A, B, and C are choosing between three policies, 1, 2, and 3, and their preferences are as shown below:

A     B     C
Best        1      2     3
Middle   2     3     1
Worst     3     1     2

If A, B, and C try to decide by majority vote, policy 1 beats policy 2 by a 2-1 vote, policy 2 beats policy 3 by the same margin, and policy 3 likely beats policy 1. If “>” signifies preference, 1 > 2 > 3 > 1 > 2 > 3… etc., in an infinite loop. If there is unlimited freedom to keep proposing new choices, the voters will cycle through the policy options forever. If the voting rules are set up to cut off this process, then the choice will be a function of the voting rules only. This is the simplest of what turns out to be a huge set of cases in which choice by majority rule is indeterminate. Once one has understood this, a phrase like “the will of the people” comes to seem meaningless.

A more optimistic take on democracy is that implied by the median voter theorem. In this model, voters are arranged on a spectrum, and party competition drives them to the center of the spectrum. But this optimistic result turns out to depend on the very special condition that voters are arranged along a spectrum, i.e., either there is only one issue, or there are several issues but people’s views on various issues are highly correlated, so that the distribution of opinion is essentially linear, albeit in a multi-dimensional issue space. If people’s opinions on the various issues are imperfectly correlated, the median voter theorem fails, and cycling comes into its own.

Fourth exhibit, related: Arrow’s impossibility theorem, summarized by Wikipedia thus:

In short, the theorem states that no rank-order voting system can be designed that satisfies these three “fairness” criteria:

  • If every voter prefers alternative X over alternative Y, then the group prefers X over Y.

  • If every voter’s preference between X and Y remains unchanged, then the group’s preference between X and Y will also remain unchanged (even if voters’ preferences between other pairs like X and Z, Y and Z, or Z and W change).

  • There is no “dictator”: no single voter possesses the power to always determine the group’s preference.

These criteria seem like rather minimal criteria to meet, yet it can’t be done. Any nondictatorial voting system is vulnerable to manipulation and arbitrary churn. This, incidentally, is why the arcane work of party “whips” and “majority leaders” is so important.

Fifth exhibit: rent-seeking and principal-agent problems. Government policies tend to benefit private actors. Private actors know this, and lobby for them. Even if elected officials are supposed to serve the interests of those who elected them– although that statement is rather question-begging and perhaps meaningless, as the analysis of cycling shows– it may not be in their interest to do so. Rational ignorance exacerbates the problem, because if a politician behaves differently in office than he promised, or than his constituency wants, it’s really not worth their time to find this out. Campaign cash may be worth more to him than the slight risk of alienating voters. There are principal-agent problems between voters and elected officials, and also between elected officials and bureaucrats.

Given that all these problems are inherent in democracy, we should not so much be surprised that democracy broke down in places like Weimar Germany or post-Soviet Russia, but rather, puzzled as to how it has survived in the United States and post-WWII western Europe and some other places. To Americans, democracy seems obvious, reasonable, mandatory, common sense, and I think deep down Americans are a bit puzzled that so many peoples have so long failed to do their obvious duty, and serve their obvious interests, by establishing durable democracy. From politeness, we in varying degrees hide even from ourselves the condescension we feel, deep down, towards such foreigners, but I think a large part of the reason why Americans are reluctant to let too many foreigners come here is that we think they are likely to ruin our democracy, since so many of them have ruined theirs at one time or another. We don’t want to be the Weimar Republic. Perhaps we would not be running that risk by opening the borders, but we might be; certainly, as long as we don’t understand why democracy fails elsewhere, we can’t be very sure what would cause it to fail here. Why take the risk? Well, no, scratch that– doubling world GDP is a good reason to take it. But still, it’s a big risk to take, and even a huge possible upside might not justify it, if the downside is the possible destruction of American democracy.

To answer such doubts, I think we need a theory of why, after all, American democracy does work. I’ve given part of an answer already. I think something like altruism goes a long way towards solving the paradox of voting and the problem of rational ignorance. But Americans just take an interest in politics as a cultural trait– not all of them, and of course we’re not the only ones to do so, but enough people to make public opinion a force to be reckoned with and to some extent even a principled custodian of political affairs. On cycling, this recent post by Garett Jones is better than what I’ll be able to write. But my hunch is that (a) altruism mitigates the cycling problem by binding/arranging the disparate interests into coalitions with a certain rough ideological coherence, (b) well-established rules, procedures, protocols, and traditions put some limits on the opportunism of cycling, as does press supervision and public opinion, and (c) despite the surface quarreling there is a rather large basis of fundamental consensus in the American republic which makes the scope for arbitrariness and cycling less dangerous.

Now, under open borders, even under a DRITI scheme with substantial migration taxes, I would expect upwards of 100 million immigrants to the United States in a generation or so. In terms of their abstract commitment to freedom and democracy, I suspect they wouldn’t be that different from Americans, first because freedom and democracy are now much admired, and to a lesser extent practiced, around the world, second because immigrants to America, of all places, would know coming in that that’s what they had to expect and would self-select, third because on arrival they would learn the rules of the game and be influenced by American public opinion. If immigrants were given easy access to the vote, however, I think that would substantially change US policy. It might lead to more redistribution, but what I’d be even more worried about is that it would exacerbate cycling problems, because immigrants’ altruism wouldn’t be directed in the same ways. They’d be less likely than Americans to think hard about what’s in the interests of the country as a whole– a meaningless phrase, perhaps, if one tries to cross-examine it philosophically, but nonetheless with a meaning in the minds of voters and a real power to coordinate public opinion. The point at present is not that people ought to think do so. Morally, I’m rather inclined to think voters ought to put the interest of the world as a whole first, which is quite a different matter. An immigrant who cared mainly about his ethnic group and his country of origin and voted in the interests of these is not obviously in a less defensible position than an American who votes for what he thinks is in the interests of the US (but not the world) as a whole. The point, rather, is that I suspect it would lead to more electoral chaos. The theoretical prediction of cycling, which the current configuration of politics and public opinion somehow mitigates, would become more of a problem. It would also be hard to induce foreigners to share the perhaps narrow-minded but civically useful veneration of Americans for the Constitution.

It would be better, in my view, to admit unlimited immigration (again, subject to DRITI taxes) but not to give them easy access to the vote. Yet this would involve a certain fundamental rethinking of democracy. Of course, there are already many millions of non-voting foreigners living on US soil, but they are still a small share of the population. Under DRITI, non-voting immigrants might become an absolute majority of the resident population of the US, and they would surely become a majority in many areas. Probably in some areas non-voting immigrants from particular countries would become the predominant population. And DRITI immigrants would be paying taxes to subsidize large classes of natives better off than themselves. How much resentment this might arouse is impossible to guess, but I think the fact of consent, the fact that DRITI immigrants would have specifically agreed to those taxes before coming, would mitigate the resentment considerably. Still, I think you’d lose something by giving DRITI immigrants no representation, because the experience of voting would induce them to take more of an interest in the national community; and also, democratic accountability probably does tend to improve the quality of laws and the competence of officials and bureaucrats, even if far less effectively than market accountability. DRITI immigrants’ votes would be a useful informational resource. And so I suppose my ideal would be to give DRITI immigrants limited voting rights, e.g., in municipal but not national elections, or even– if I’m correct in thinking there’s not actually anything morally objectionable about this– giving them 1/8 of a vote, or something.

What’s problematic is that the current notion of democracy seems to be that “everyone” gets to vote. People aren’t accustomed to there being huge swaths of the population who are subject to the laws but can’t vote. Certain standard arguments for obeying the law, e.g., “you voted for it, so you should obey it,” or “if you don’t like it, vote to change it, but meanwhile obey,” would suddenly fall flat. It would be inconsistent with what many people mean by democracy, though certainly not with all and perhaps not with the best definitions of democracy. It would be essential to preserve free speech and freedoms of association and religion, and it would probably be infeasible, certainly undesirable, and I think unjust, to deny these to immigrants; but without the vote they might often take to the streets with demands, and the government would often find itself defying or ignoring, perhaps sometimes pandering or yielding to, political street mobs.

The meaning of American democracy has changed several times. The Jacksonian revolution, the Civil War, the New Deal, and the civil rights movement all involved fundamental changes in what it meant. Open borders would require yet another fundamental change, one which might come to regard deportation as fundamentally criminal, intolerable, and inconsistent with democratic norms, while at the same time reconciling itself to huge populations with no or limited voting rights due to foreign nationality, even as they enjoyed the right of residence and the protection of the law. The American electorate would enjoy both less power– having relinquished the right to exclude foreigners– and more– it would govern huge populations of resident foreigners– while (assuming open borders became general) American citizens would enjoy the right to live and work abroad as well as at home.

Nathan Smith is an assistant professor of economics at Fresno Pacific University. He did his Ph.D. in economics from George Mason University and has also worked for the World Bank. Smith proposed Don’t Restrict Immigration, Tax It, one of the more comprehensive keyhole solution proposals to address concerns surrounding open borders.

See also:

Page about Nathan Smith on Open Borders
All blog posts by Nathan Smith

6 thoughts on “Open borders and the viability of democracy”

  1. I came up with a proposal (written in the Indian context in that particular answer, but there isn’t anything India-specific about the proposal per se) that can allow for different people to have differently weighted votes, while potentially mitigating most of the “second-class citizen” downsides:

    The key is to have the extent to which a vote counts determined by a simple three-question quiz administered at the time of the vote, rather than permanently attached to the voter, and this weightage is not even made public — so there is no concern about second-class citizen treatment through differential weightage to votes. In the immigration context, then, the idea would be that migrants who have been in the country for a certain period of time can vote just as citizens do, but if, in fact, their political knowledge is less than that of natives, their votes will count for less. If, on the other hand, they have more political knowledge than natives, their votes will count for more.

    The main counter-argument I expect from restrictionists is that some migrants may have high levels of political knowledge, but may not be assimilated into the political preferences of the host country. (In the US context, they may be mavens of history, civics, and geography, but think that the First Amendment should be scrapped, for instance). Objectively, I think that high political knowledge correlates with “good” political preferences — and those who argue otherwise should bear the burden of proof. That said, there is also the danger of diversity itself being problematic for the “cycling” reasons you mention, which is not a problem that tests of political knowledge could easily counter.

    1. I’ve actually heard this suggested before (in the American context – “how many seats in Congress? How many judges are on the Supreme Court? Who is the Speaker of the House?” etc.), but I heard it suggested as an actual prerequisite TO vote, not just as a vote-weight system. The problem with either of those is that political knowledge might not correct for all the systematic biases that harm the democratic process. I’d like to see a test that actually corrects for the four biases Bryan Caplan posits in Myth of the Rational Voter. If you want to add political knowledge as well, great! Personally, I think more questions would be better – if you make the test longer without making it harder, it will discourage people from voting. Which is a good thing, since people that would choose not to vote because they’d have to take a test are probably not people we’d miss the votes of. Everyone should be ALLOWED to vote, but that doesn’t mean everyone should vote.

  2. Mark Twain wrote an essay once detailing the concept of “increased suffrage” – in his idea, everyone always gets at least 1 vote, but you could qualify for more if you met certain criteria. Own a business? Extra vote. Graduate college? Extra vote. Etc. The idea was that the better-educated you were, and the more “skin in the game” you had, the more votes you’d possess. The system involved no disenfranchising, since everyone always got at least 1 vote, and everyone (at least in theory) could accomplish the other tasks to get more.

    There may be something to this system. Reading “Myth of the Rational Voter,” it’s obvious that the biggest problem with democracy is the systematic biases of the voters. If a weighted system could be created that weighted against those biases (say, by giving more votes to those least likely to have them), much of that problem could be solved while still maintaining a democratic government. Which is, of course, the worst form of government “except all the other ones.” (Apologies to Mr. Churchill.)

    The question is – do immigrants possess these biases in greater or lesser frequency than native-born citizens? For example, if it was discovered that immigrants consistently vote pro-market, pro-immigration (I’d imagine that one to be true, at least), pro-free-trade, anti-war, etc., then I would definitely be in favor of giving them more votes! We shouldn’t assume that immigrants will, by nature, vote poorly. We should check – and I’d be more than willing to incorporate them (and everyone else) into a system that weights votes based on biases.

      1. Fascinating! I find especially interesting the notion that parents should get an additional vote for each dependent child. Not saying I think it’s a good idea (haven’t had enough time to think about it), but it’s certainly intriguing. I’d really, really love to see a proposal for plural voting in the US, even if just as a thought experiment.

        1. The main reason why I prefer my secret ballot political knowledge multiplier proposal to open multiple voting is concerns about the corrosive effects of openly assigning more votes to one individual relative to another can do for social status. In earlier times, however, the technology to use this kind of secret multiplier voting didn’t exist, so plural voting by external criteria was probably the best one could do, despite its drawbacks in terms of social inegalitarianism.

Leave a Reply