Category Archives: Political Science

A Hayekian and Buchananite Interpretation of Legal Prohibitions of Disassociation

I have criticized some libertarians for communicating the worst aspects of libertarian philosophy. I, in contrast, prefer to emphasize other traits, such as cosmopolitanism. I think that a free society tends towards an erosion of undesirable prejudices, where their undesirability is decided collectively (e.g. if a majority of people morally oppose racism). But, my position does not imply that those “Auburn libertarians” — a category I use for convenience — are wrong. In fact, I am implicitly saying that they are more right than wrong, in most cases. My argument boils down to, “Even in a world where people are allowed to disassociate based on socially unacceptable prejudices, social outcomes will be similar to those in a world where these outcomes are forced through law.”

The problem is that us “bleeding hearts” — well, actually, I shouldn’t speak for those who know more than I do — are still in a pickle when it comes to issues like civil rights legislation. I’m going to use civil rights legislation as the main example throughout this post. I see three general approaches,

  1. We justify these laws;
  2. We continue to skirt around the issue by emphasizing other forces;
  3. We explicitly oppose these laws.

(1) is seemingly at odds with basic libertarian philosophy. (2) has its limits, and there is that ultimate dissatisfaction with leaving a controversial problem essentially unanswered. (3) would essentially put us in the same camp as the libertarians we disagree with.

I suppose there is some complicated way of justifying civil rights legislation. We can talk about past injustices. We can talk about minimizing injustice (utilitarianism). I don’t know much about these approaches to the problem, so I won’t embarrass myself by talking about them. Rather, I want to propose an alternative, influenced by Hayek and Buchanan.

In the first volume to Law, Legislation, and Liberty, Hayek discusses institutions (rules) that arise spontaneously. These kind of rules are abstract, often difficult to pin down and communicate. They are followed, oftentimes without even knowing of their existence, because they allow for a more accurate interaction between different persons’ expectations. The system of rules evolves over time, because better rules allow for more well-off societies, meaning bad rules will go extinct and good rules will be adopted by other societies.

As these rules come to be better understood by people, they can be encoded in a written rule of law, legislated by government. Legislating versions of these rules is not the best approach to institutional evolution, but it is a natural outcome of the belief that we can rationally plan society — the best rules aren’t implicit and spontaneous, but explicit and planned. Hayek argues that most “planned” rules are just versions of spontaneous institutions that are finally understood by legislators.

Racial prejudices were always strong in the United States, motivated by whatever reason. This fact is manifested in things like pre-1965 immigration laws, the Jim Crow laws, and the reaction to the civil rights movement. But, I think it’s fair to assume that racial prejudices were slowly eroding, at least in the context of specific races. It’s likely that in 1960 a greater share of Americans supported extending civil rights to African-Americans than in 1860. Reasons for this include a network effect to crumbling prejudices (the more people who accept some previously “alien” trait, the more socially acceptable it is), and the fact that racial prejudice is a cost — it’s a cost to do business, and it’s a personal cost (hate damages the hater more than the hated). Further, new generations will grow up surrounded by different cultural forces.

My point is that it could be that the U.S. government was legislating law that already implicitly existed, even it constrained to a lesser degree. Maybe the extension of civil rights, within the polity, was already taking place, because the share of people actively seeking to restrict them was falling. Old rules, based on racial prejudice, were costly, and newer rules allowed for better societies. If this is true, the U.S. government was merely strengthening a trend towards the extension of civil rights.

Still, how would we justify forcing people to accept these rules? Is it right to enforce a prohibition against disassociation on the basis of race? Or, is it better to just let these forces work themselves out voluntarily. This is where Buchanan’s and Tullock’s The Calculus of Consent comes in.

Ideally, democratic institutions allow for mutually beneficial exchanges that would otherwise be impossible, or much more difficult, to make through market institutions (think public goods). This is even more true of representative, and complex, democracies, which make the political process more inclusive and allow for a superior economization of knowledge. What do these exchanges look like? Suppose we are talking about an exchange between two people, where one wants the government to invest in building a road and the other in space exploration. Depending on the intensity of their preferences, there may be an opportunity for gains from trade. Person A can offer Person B a vote in favor of space exploration, as long as Person B votes in favor of road construction.

We don’t have ideal democratic institutions. But, that doesn’t mean our democracy doesn’t work at all, even if it’s too complex for us to fully understand (this reminds me of people who oppose markets, largely because their complexity make them difficult to understand — people don’t like what they don’t understand). Even if we are mostly unaware of how the process works, and we don’t feel directly involved in its planning, American democratic institutions are probably still better than the alternatives in allowing for mutually beneficial exchange outside of markets.

If we think of it in terms of exchange, civil rights legislation seems to have less to do with coercion than we might think. The intensity of prejudice was falling, and the intensity against prejudice was rising, so it makes sense that some kind of compromise was struck. Of course, given our non-perfect democracy, there are probably a large number of people who were forced to accept these new rules, without receiving of greater value in return. But, still, in principle, it’s not clear that things like civil rights legislation is “unlibertarian,” given that we can conceive of a situation where such legislation would arise voluntarily.

Borrowing a point from Walter Block, the most egregious aspects of public legislation would probably be avoided if people had the right to disassociate from government. This right would allow people to cease a relationship with a government, and, of course, the latter would cease its relationship with the former — no access to public benefits (unless they’re externalized). I have made this point before. And, it’s not radical, because it’s essentially what motivates the Tiebout model of tax revenue competition. I don’t think allowing the right to disassociate would lead to anarchism, only to greater competition between providers of public goods. The benefits of being able to exploit gains from trade through political institutions are just too strong.

Free Societies are Cosmopolitan Societies

Suppose, for the sake of argument, that “we” — society in general, the United States, western civilization, take your pick — are tending towards greater efficiency in social coordination. I don’t want to use the term “free society,” because I think the term “freedom” can be misleading. But, I do envision a world where the limits to government are stricter than they are today. Imagine if people could disassociate and associate with different governments at will (see “Cosmopolitanism and the State“). Assume that all social relationships are voluntary, and that we can explicitly perceive them as such. Would such a society be cosmopolitan and largely border-less or would it be composed of various small communities, each populated by people of similar beliefs and prejudices? My prediction is the former.

There is the fact that what we expect is for the division of labor to grow towards its limit, the global labor supply. To exploit potential gains from trade, we expect people to associate with each other. Prejudices can prevent these associations from taking place, but prejudice doesn’t pay in the long-run. When discriminative barriers are high, profits — gains from trade foregone — are high. “Savvy entrepreneurs” are going to exploit those profits, and the loser is the bigot. Over time, discriminative barriers make less and less sense.

Alternatively, frame prejudice within the context of spontaneous order. Institutions — rules — change through a competitive process, where (we hope) the best of alternative institutions are chosen on the basis of what it adds to the process of social coordination. Societies guided by rules that aren’t burdened by prejudice are more likely to be successful than societies where bigotry handicaps the division of labor. Spontaneous order, in this case, promotes inclusive rules. These rules influence members of society; people tend to conform to them. In fact, those who don’t conform to broadly accepted rules are typically shunned, serving as an incentive to constrain your range of choice.

And the evidence, so far, generally confirms my hypothesis. Immigrants tend to assimilate into our society over a period of generations, and the period of assimilation is shortening. As society becomes more cosmopolitan, the cultural shocks that immigration waves tend to bring with them attenuate. Consider the case of Irish and German immigrants in the United States. Initially, they were met by hostility; they had upset the existing order (especially in the labor market). The same is true of southern Europeans, who would come to the U.S. in large numbers decades later. But, these new peoples — and, later, their successors —  gradually assimilated, and society conformed to them, as well. The U.S. is known as the “melting pot” for the reason that our culture is multicultural. That’s not to say that prejudice has ended, but I am saying that prejudices have tended to erode over time.

The other option is for society to fracture into various communities. It’s implied that the territorial borders to these communities will be, in some way, enforced. They are populated by relatively homogenous agents, who live together because of their compatible belief sets. Maybe an all-white group of families wants to live in an area where only whites live. Or, maybe heterosexual couples want to live in an area where there are no homosexuals, and they enforce that by ostracizing potential members who don’t fit their belief sets. They can still associate with people outside of their communities on a limited basis, but otherwise society doesn’t exhibit the cosmopolitanism I predict.

We do see patterns of fracturing in the real world. The clearest example that is still relevant today are residential communities — especially walled residential communities — that are populated by, on some level, homogenous agents. But, the barriers to entry to these communities have tended to fall. Today, they’re typically based on income, not necessarily race. And, even in the case of income, the barriers are falling. Non-white households are earning growing incomes, and they’re using their new incomes to buy in expensive areas. My interpretation of the evidence is that homogenous communities tend to break down over time, and they tend to become smaller and smaller. This last part is because these types of barriers are hard to enforce, and the only other option to accepting the facts is to recede into smaller-and-smaller communities.

The liberal world I foresee is not one where people tend to exercise their right to disassociate on the basis of some prejudice, but rather one where these tendencies gradually die out. Cosmopolitan societies are usually more successful than homogenous societies: the forgone gains of trade fall, there is a more dynamic intellectual mix — making cosmopolitan societies much more innovative —, and societies that impose less barriers against trade are typically those that are the most peaceful. Communities that impose barriers to entry will generally shrink in size, or fade, as these barriers are toppled. What liberalism promotes, then, is a reduction of the reasons people use to justify disassociation. Instead, liberalism (and capitalism) promotes association — mutual dependence to improve each others’ standard of living. Prejudice is inconsistent with this world.

What is Liberalism?

Many libertarians argue that modern liberals hijacked the term “liberal” and re-defined it. These libertarians like to say that classical liberals would have eschewed modern liberalism, and that their beliefs were closer to those of modern libertarianism. One way a modern liberal could defend from this attack is to argue that, no, classical liberalism is nothing like libertarianism, and that modern liberalism follows from its classical counterpart. She might also argue that libertarianism may also follow from classical liberalism, but that there are many directions the original political philosophy could go. I like this line of reasoning, but I want to take it one step further. To me, liberalism is not about a particular set of beliefs — e.g. equality however defined, natural rights, et cetera —; rather, it’s about what institutions should like, to maximize the communication of beliefs and to make governance responsive to changes in preferences exhibited by the state’s citizens.

One thing I should get out of the way, right away: I am not arguing that my interpretation is what classical liberals had in mind. I do not consider myself well read in this area; I have read very few classical liberal tracts, and have confined myself mostly to excerpts. I am not even in a position to make any argument of that sort. The reason I referenced the debate between modern liberals and libertarians is because thinking about it is was what reminded me about my own position on liberalism, because of some similarities between my theory and the modern liberal defense of their liberalism.

Most liberals and libertarians talk about liberalism as if it were a philosophy about specific beliefs. Paul Krugman, for example, likes to say that the facts have a well-known liberal bias. He means the facts corroborate liberal beliefs — he implicitly assumes that liberalism is about a particular set of social preferences. Libertarians, and many classical liberals, like to talk about natural rights and limited government, making the same implicit assumption as Krugman: (classical) liberalism is about enforcing a particular set of beliefs. I can’t accept these positions, because to me they seem much closer to authoritarianism than they are to liberalism.

While I associate with Hayek’s brand of liberalism, I think it’s flawed, too. In “Individualism: True and False,” he distinguishes between continental liberals and Scottish–English liberalism. According to Hayek, continental liberalism was attracted to extreme rationalism, which he associates with a planning mentality. By contrast, Scottish–English liberalism emphasized the concept of spontaneous order, where society progresses through a competitive, decentralized process, rather than it being planned and rationalized. Modern liberalism stems from continental liberalism, adopting its rationalism and attraction to planning. Hayek’s liberalism, and libertarianism to one degree or another, follows classical liberalism in its advocating for unplanned order. But, this assumes — without justification or explanation — that democratic outcomes are planned outcomes, rather than an alternative institutional means of promoting “spontaneous” change. The word “spontaneous” in spontaneous order also seems difficult to pin down: human action is inevitably planned at some level. Finally, ultimately, Hayek’s position also boils down to a debate over which concrete set of beliefs liberalism should defend: planning or spontaneous order.

In my view, liberalism takes for granted a plurality of concrete belief sets. You can think that business owners should be allowed to discriminate against race — as taboo as that belief is — and still be a liberal. Liberalism is the exact opposite of belief enforcement. It is a political philosophy that aims to erode the ability of those in power to enforce their own set of beliefs. Liberals support government institutions that help communicate social preferences between citizens; it supports a process of belief reconciliation. Apart from allowing the satisfaction of the political preferences of a broader class of citizens, the sort of institutions liberalism supports make it so that these institutions are responsive to changes in the preferences of society’s citizens.

Conservatism, then, is not about a particular belief set, either. It’s about disrupting the liberal process of change. Conservatives do this by enforcing their belief set over alternatives. I don’t know if this is what modern Republicans support, but the fact is that modern Republicans are working in the context of liberal institutions. Modern Republicans do have a particular set of preferences, but they communicate these through the democratic process, which is a liberal one. Whether their beliefs are enforced or not depends on the outcome of this process, and how their preferences are reconciled with others’.

My point might be made clearer by means of (a controversial) example. Let’s talk about the issue of gay marriage. Social preferences over gay marriage are changing and the tide is turning in its favor. Liberalism is not about enforcing gay marriage because it’s a belief that defines the philosophy. Rather, liberalism is about enforcing gay marriage if and only if social preferences are changing in that direction. We can conceive of a world where gay marriage advocates are actually conservatives. Suppose that gay marriage has been legal for the past 500 years and that social preferences, for whatever (unfathomable) reason, are shifting against it. The liberal thing to do would be to allow the institutions of governance to reflect this change, which in this world is to make gay marriage illegal.

I don’t think my interpretation of liberalism is unique. I just don’t know enough about others’ views to make any well-documented connections. It seems to me that John Rawls’ brand of liberalism comes close to mine, from what I’ve been told and the excerpts I have read — as A Theory of Justice sits on my shelf, waiting to be read cover-to-cover. It’s also what I’d like to believe that Hayek was grasping at. This might be a good example of confirmation bias, but I also see streaks of what I’m talking about in David Hume, James Buchanan, and the bleeding heart libertarians amongst others.

Murray Rothbard, while in my opinion definitely not a liberal, had an understanding of the point I’m making. From what I know, his anarchist vision necessitates, to one degree or another, homogeneity of beliefs. Those with different belief sets can disassociate from others and form their own societies. While I reject this vision in favor of cosmopolitanism, reconciliation, and political exchange, I think Rothbard’s philosophy, in a way, confirms (or, is consistent with) mine.

Quickly coming back to Hayek, I think the debate between planning and spontaneous order is not about what the right brand of liberalism is, but is a scientific debate about the method by which liberal institutions arise. My argument might make more sense if I compare it to a similar situation in economics, one that most people are already familiar with. People like to brand themselves as Keynesians, Austrians, Monetarists, et cetera, and, although these different schools of thought often throw jabs at each other, they are all advocates of capitalism. They all favor a process where individuals can communicate their preferences, and these preferences will (ideally) be satisfied by rank of value. The disagreement is over how the institutions of capitalism should come about: by government decision or through the competitive capitalist process. (Of course, this distinction suffers from the same problems as Hayek’s, but assume these away, just for the sake of understanding the point I’m getting at.) These differences in scientific beliefs also must be resolved through communication and reconciliation, which is what (ideally) modern institutions of science attempt to achieve. The situation is transferable pretty much as-is to the debate between planning and spontaneous order between political philosophers.

My belief is that liberalism is about promoting governance that adapts to changes in social preferences, which requires pluralism. This is opposite to authoritarianism, which enforces a particular set of beliefs over all others, despite social preferences. Conservatism, if my definition is right, is authoritarian. Those who think that liberalism is about a particular set of beliefs, and that these should be enforced over all others, are not liberals at all — they are conservatives, they are authoritarian.

Glasner the Hoppean

One disincentive to investing in the monopoly over money in peacetime is that government decision makers have a limited tenure in office and have no transferable property rights in the assets they create while in office…

Without ownership of, or marketable shares in, the state, the incentives to undertake investments that redound to the future benefit of the state are much weaker than those to invest in privately owned enterprises.

The lack of transferable property rights in the state suggests that government decision makers have an inflationary bias. Under an absolute and hereditary monarchy, the bias would be reduced, though not eliminated.

Followed by a caveat,

On the other hand, hereditary monarchies have an inordinately high probability of producing incompetent rulers who are unable or unwilling to follow policies in their own, and the state’s long-term interest. Thus a small, homogenous ruling elite, such as the aristocracy of a city-state like the Venetian Republic, would seem to be the form of government most likely to invest in the monopoly over money in peacetime.

— David Glasner, Free Banking and Monetary Reform (Cambridge: Cambridge University Press, 1989), pp. 39–40.

Something to consider, however, is that politicians rarely run for a single office, meaning democratic elections are oftentimes repeated games. This creates a reputation constraint on the politician, as the costs to bad policies will be internalized in subsequent campaigns. Another way to partially internalize the costs is through party dynamics. Being an important member of a political party probably requires you to have a positive reputation amongst those who vote for that party. I actually don’t know much about how the principal–agent problem is resolved in democracy, but blaming finite terms on bad policy has always struck me as weak. It seems to require a much more simple form of democracy, lacking some of the more intricate institutions that define modern democracies in developed nations.

I wouldn’t say monarchies are more likely to produce bad leaders, either. The probability might be the same, or it might be even larger in democracy (we’re selecting from a larger sample pool — one that isn’t trained to rule during their childhood). But, limited terms and democracy create competition, and it’s easier to get rid of bad leaders. Democracy creates a selection process that monarchy doesn’t have.

Modern democracy isn’t perfect, of course; it has its costs, just like every other alternative political framework. But, I tend to gravitate towards Jeffrey Friedman’s theory of radical ignorance, and my guess is that radical ignorance problems were just as bad, or even worse, under monarchies than they are under democracies.

N.B. These are only a few paragraphs of a (very interesting and informative) book that deals with a different topic entirely (free banking), so I don’t mean to give more importance to them than is really due. I just thought they were interesting, especially given that I’ve criticized Hans-Hermann Hoppe’s argument on democracy v. monarchy before.

Why I Was an Anarchist; Why I’m No Longer an Anarchist

Here is Gene Callahan’s version. Like him, the anarcho-capitalist message captured me at first, but then I realized that a free world could look very different. My change in opinion came with different reasoning, however.

Why I was an anarchist: I sympathize with the idea of a society where individuals can associate and disassociate with whomever they please. I don’t think a society where individuals can truly opt in/out of different organizations dedicated to governance is utopian (see “Cosmopolitanism and the State“).  The anarchist “vision” captured me, because I believed that that was genuinely what such a world would look like. I thought it would look very different to that of today. All goods, including justice and security, would be provided for by private firms on the market. The assumption was that the only relevant institutions of resource allocation are those of the market.

Why I’m no longer an anarchist: I no longer assume that the only worthwhile institutions of resource allocation are those of the market. I now think that it’s conceivable — not necessarily always a priori true; it’s an empirical question — that alternative institutions, such as those of governance, can have their own processes of resource allocation which could have a comparative advantage in certain situations. Specifically, they would have a comparative advantage where market institutions are comparatively weak in providing certain goods and/or services.

I have come to believe that the criteria anarcho-capitalists have for a free society are fulfilled in one that looks very much like the one we’re in today. Influenced by Buchanan and Tullock, I think that voluntary (at the constitutional level) governments are possible and that they may look a lot like modern governments (although, most likely smaller in terms of both range of goods/services provided and territorial size) — for an analogous distinction between the constitutional and decision-making levels, see “Are Property Rights Coercive.” In fact, I think that many anarcho-capitalists who criticize modern institutions suffer from a fatal conceit, and that these institutions are closer to their ideal than they think (see also the much more tame “The Case Against Anarchy“).

Are Property Rights Coercive?

At a Liberty Fund colloquium on Keynes and Hayek, I was almost persuaded towards the position that property rights are coercive. The basic idea, if I understand it correctly, is that a claim on property necessarily restricts the choices (we can call them “liberties”) of others. Therefore, by claiming some property, you are coercively restricting others’ ability to use that property. See, for example, Matt Zwolinski’s “Libertarianism, Freedom, and Coercion,”

Just as libertarians recognize that a regime of property rights restricts liberty, so too do they recognize that it does so through the use of coercion.

But, I’m no longer so sure. If we look at property rights both on a constitutional and decision-making level, the degree of coercion is much more tame. What am I missing?

To get an idea of what the constitutional and decision-making levels are, let me provide two examples. The first example is from The Calculus of Consent. Buchanan and Tullock use these two concepts to devise an abstract voluntary, democratic government. At the decision-making level — when, for instance, we’re actually voting on different policies — we can expect disagreement, and some will not directly benefit from the policy. But, we can conceive of unanimous consent on the constitutional level, which defines the set of decision-making rules. Why would someone agree on the constitutional level when they expect to disagree at the decision-making level? Because they gain when the decision-making goes their way, and the rules guide a bargaining process allowing for gains from (political) trade. In a sense, then, those who live in this ideal world are voluntarily accepting policies that may not benefit them (and, in fact, may hurt them), because overall they gain from the set of institutions.

The second example comes from a libertarian conception of law, and I recall this theme being present in David Friedman’s Machinery of Freedom. At the decision-making level, we all expect to suffer costs from any legal system. There is some probability that we will receive a traffic ticket, or that we’ll be fined for whatever reason, or maybe even that we’ll be arrested for murder. Given this, why would a society agree to any legal system? After all, everyone expects to lose at some point, right? But, there is an overall gain. Even thieves enjoy having their own property protected by the law. Likewise, it’s possible that we all gain from traffic regulation. Thus, even though at the decision-making level we may suffer costs, at the constitutional level we all tend to support a set of laws. To make my point clearer, in this ideal legal framework, it would be hypocritical of me to argue that a police officer coerced me by issuing me a speeding ticket, if I had originally agreed to the set of policies.

The concept applied to property rights is the same. At the decision-making level we can expect disputes, and we suffer costs when others deny us use of their property. But, we also benefit from property rights, including the choices we ourselves have with our own property (and the restrictions we place on others). Thus, most people accept property rights at the constitutional (or institutional) level — we all benefit from having rules that define property rights. In that sense, just like with (ideal) government and law, we are voluntarily accepting the reality of costs, because the benefits outweigh them. Therefore, to me it seems a stretch to argue that property rights are coercive. They only appear coercive if you don’t look at the bigger picture.

Or, we can still call imposed costs coercion at the decision-making level, but then we need to find an easy way of distinguishing between that and coercion at the constitutional level — when a set of rules are imposed on a society (e.g. all economic goods ultimately belong to the king, who can take them and use them at his discretion).

Democratic Complexity

A lesson in the difficulty of [centrally] planning democratic institutions,

The Afghan government collects an extremely low level of revenue (less than ten percent of GDP), and a large share of this comes from customs rather than taxation. In effect, Afghans are not really charged by their government for the services they are provided. Moreover, for the most part, the Afghan government neither funds nor delivers the key public services offered in the country. According to estimates by the U.S. Government Accountability Office, in recent years, the United States and other donors paid for about 90 percent of Afghanistan’s total public expenditures, including funding for the Afghan National Security Forces. In addition, the provision of may key services remains highly dependent on foreign advisers and experts.

In their 1967 book, The United States in Vietnam, George Kahin and John Lewis wrote that “U.S. aid thus provided [South Vietnamese President Ngo Dinh] Diem with a degree of financial independence that isolated him from basic economic and political realities and reduced his need to appreciate or respond to his people’s wants and expectations.” Like Diem, Karzai has little reason to improve his state’s effectiveness or accountability.

— Karl W. Eikenberry, “The Limits of Counterinsurgency Doctrine in Afghanistan,” Foreign Affairs 92, 5 (2013), p. 67.

When foreign aid is used to maintain extractive political regimes, the root problem is institutional weakness. When a country like the U.S. invades another one with the intent of reorganizing government along democratic lines, they set themselves the task of planning out the set of institutions that constrain governance. This goes beyond the organizations of government — i.e. congress, parliament, the executive, judicial, et cetera —, and includes the various rules and customs that develop over time, from local branches of authority to national governments. These type of issues make humanitarian wars, with the intention of introducing democracy, less predictable, and therefore more costly — not just to the invader, but to the country being invaded, as well.

I selectively quote Eikenberry. In the preceding paragraph, he argues that what keeps democratic executives in check are taxes. If the level of public goods and taxes are not set to the taxpayers’ preferences, citizens will vote for another party. Since most of Afghanistan’s income is made up of tariff revenue and foreign aid, the implication is that Afghans are not paying the majority of the costs of public mal-expenditure. As a result, the ruling party doesn’t really face a political constraint in that direction; since the people aren’t paying for it, the people don’t care what the government is buying. But, tariff income probable made up the bulk of U.S. revenue during the late 18th and early to mid- 19th centuries, yet the U.S. didn’t suffer from the same degree of corruption.

Does democracy invite nationalism?

Before democratization begins, nationalism is usually weak or absent among the broad masses of the population. Popular nationalism typically arises during the earliest stages of democratization, when elites use nationalist appeals to compete for popular support. Democratization produces nationalism when powerful groups within the nation not only need to harness popular energies to the tasks of war and economic development, but they also want to avoid surrendering real political authority to the average citizen. For those elites, nationalism is a convenient doctrine that justified a partial form of democracy, in which an elite rules in the name of the nation yet may not be fully accountable to its people. Under conditions of partial democratization, elites can often use their control over the levers of government, the economy, and the mass media to promote nationalist ideas, and thus set the agenda for debate. Nationalist conflicts arise as a by-product of elite efforts to persuade the people to accept divisive nationalist ideas.

— Jack Snyder, From Voting to Violence (New York: W.W. Norton & Company, Inc.), p. 32.

I don’t remember exactly where I saw this book referenced, but I’m pretty sure the referrer was Bryan Caplan. I haven’t read it, but I was skimming through it and I read the above. Since I haven’t read the book, I’ll hold off from commenting on it.

Expectational Voodoo

A few months ago, Daron Acemoglu and James Robinson commented on current events in Egypt by using Turkey as a foil. The first half of the post introduces us with a brief political history of Turkey, starting with the Democratic Party’s (DP) 1960 election victory. Sometime after coming to power, however, the DP turned to authoritarianism. The military responded by toppling the DP in a coup, and they would launch three more in the years to come. Acemoglu and Robinson then give an interesting answer to a good question,

What would have happened without the military coup? Nobody knows. Perhaps. Menderes and other DP elites would irreparably damage the economy or somehow cow into a total submission before the next election to effectively set up their own dictatorship. Possible. But unlikely. Rather, they would have probably been kicked out of power in the next election, cementing Turkish democracy’s credentials (ed. emphasis mine).

This and other parts of the post reminded me of Bryan Caplan’s argument that “every social system requires favorable expectations to survive.” Speaking of democracy, he makes the point, for example, that peaceful transfer of power will only occur if people expect them to. The argument sounds a lot like a belief in self-fulfilling expectations. Caplan’s discussion of communism seems to confirm my interpretation. His essentially asks whether self-fulfilling expectations can save communism, but concludes that they can’t because the incentive to shirk is too strong.

The self-fulfilling expectations argument has parallels in economics. One example is NGDP targeting. One of the main transmission mechanisms is expectations. If a central bank buys assets on a large-scale and people come to expect NGDP to rise the demand for money will fall, as people exchange cash for assets, goods, and services. The increase in purchases, in turn, causes an actual increase in the NGDP, and it continues in a cycle (with diminishing returns, I assume).

But, I’ve always thought that the “expectations channel” is a weak theory. Yes, expectations in economics are very important, and yes expectations of the future inform present decisions, but sometimes the thing people are supposedly expecting is so broad that the theory starts to lose persuasive power. Firm owners do consider expectations of future demand and revenue to plan their production processes, individuals do think about their expectations of future income when making spending decisions, but how many non-economists even know what NGDP is? How many people can you think of that make daily decisions based on where they think NGDP will be x-time into the future?

For people to care about NGDP, it has to be an indirect channel. Monetary policy has to affect revenue streams, which means that there has to be some transmission mechanism that influences expectations. In other words, there is something that prompts the person to revise their expectations. I don’t think central bank monetary policy is enough, because the fact is that most people consider many other variables, if they consider monetary policy at all. There are people who do include expectations of monetary policy in their decision-making, but to rely on these people to actually cause a significant rise in NGDP is a bit desperate, in my opinion.

The expectations channel explanation makes even less sense once you consider that in the real world expectations are heterogeneous, and oftentimes people outright disagree with each other. The economics profession itself can’t agree as to whether there will be inflation, deflation, stagnation, a double-dip, or a straight-forward recovery. The safest assumption is that the range of opinion is even broader outside of the economics profession, since the average non-economist is not very likely to have the educational background to even know much about what they supposedly expect. Finally, most people find that they have to correct their beliefs over time (i.e. their expectations are falsified). This reinforces my belief that changes in expectations need to be prompted, and the method of communication can be indirect — e.g. a child learns to not touch fire not because she expects some intricate chemical reaction that will damage her hand’s tissue, but because she expects that touching the fire will hurt (and she knows this out of experience).

My sentiment is the same when it comes to the use of the expectations channel theory in political science (it’s probably not coincidental that the people cited here invoking it are all economists). Is his discussion of communism, Caplan implicitly says as much. He argues that the incentive to shirk is too strong for expectations to self-fulfill in communism. But, the incentive to shirk is a product of that social system’s institutions: there is no constraint that disincentivizes shirking. (Not too mention that we know that the strongest argument against communism doesn’t have anything to do with expectations, but about the lack of institutions that allow for an efficient allocation of inputs.)

Similarly, a precondition of successful democracy isn’t a democratic expectational equilibrium, but of the right institutions (e.g. rules of the game). I agree that people come to expect peaceful transitions of power, and that the expectation of such outcomes is important (otherwise there is an incentive to resist), but I’d argue that these expectations form over time, as people see peaceful transitions occur. This is how I interpret what I excerpt above out of Acemoglu and Robinson’s piece. The event has to occur so that people have a reason to expect it to occur again in the future. Sometimes this reason is a completely random event; other times, expectations form slowly, over time, and reflect gradual changes in institutions. It would be difficult to substantiate an argument that the expectation of peaceful democracy came suddenly, rather than develop over time as institutions of governance changed and incentivized these expectations to form. For example, people trace American democracy too the English magna carta.

Calling “expectations!” strikes me almost as hand-waving. Theories of change need transmission mechanisms to explain the relationship between variables, especially when they’re forward-looking (remember, expectations are heterogeneous). In the 1930s, Mises and Hayek argued that economists’ emphasis on changes in the price level hid them from the consequences in changes in relative prices, including the business cycle. This may not be so drastic, but it really does seem that the more broad and aggregated the expectations channel is the higher the probability that the theory is wrong.

Cosmopolitanism and the State

I think it’s fair to assume that, on some level, we can distinguish between liberal Libertarians and conservative Libertarians, where the major difference is in the prediction of what a close-to-ideal libertarian society would look like. The liberals predict a plural world where borders matter less and where differences can be resolved through trade, whether through institutions of the market or institutions of governance. Conservatives, especially those least interested in trade-offs and compromise, foresee a world where (some or all) people migrate to form relatively ethnically homogenous communities — this prediction may not be explicit in their writing, but this is my interpretation.1 Both doctrines are consistent with the broad Libertarian value of limited government, they just differ in the prediction on human behavior and values (i.e. the debate is over an empirical question). This division doesn’t always work, but I think it does to a sufficient extent.

While discussing Libertarian populism, Matt Zwolinski bemoans that Libertarian populism is losing its cosmopolitanism. He argues that “thoroughgoing cosmopolitanism”  is”one of libertarianism’s most attractive and distinctive elements.” While I think it depends on the interpreter, I agree with Zwolinski. I suspect it’s because both Zwolinski and I take on a pluralist view of society, which embraces disagreement and where progress is measured, at least in part, by changes to the rules which guide the decision-making process in a plural world. I also suspect that this interpretation reflects certain idiosyncrasies; that is, we are cosmopolitan people (or, at least we [I] think we [I] are [am]). On the other hand, my interpretation of a small sample of evidence also tells me that greater cosmopolitanism is in some way causally related to the institutional change necessary to bring about greater constraints on government.

About a year ago, I was involved in writing a literature review on Catalán nationalism. My family is from Castilla-La Mancha and is deeply nationalistic (a strong Spanish identity), and so, I must admit, I was at first interested in seeing how much of the periphery nationalism was just political rhetoric. It turns out my prior is about 50 percent right, and if I were to adjust this figure I’d adjust it down.2 But, the data shows results that conflict both with the nationalist Spanish interpretation and the nationalist regional interpretation; Spanish nationalism has been declining very quickly and regional nationalism has been declining gradually or remained more-or-less stagnant. The graphs below are missing lines for “Spanish only” and “autonomous community [AC] only” identities, but they show the trends for three different broad identity groups: more Spanish than “regional,” equally Spanish as “regional,” and more “regional” than Spanish (“regional” should be replaced by the region’s actual national identity; e.g. Catalán, Basque, Galician, et cetera),

Spain regional identification— Enric Martínez-Herrera, “From Nation-Building to Building Identification with Political Communities,” European Journal of Political Research 41, 4 (2002), pp. 435–437.

One discernible long-run trend is a decrease in the amount of people who identify as “more Spanish than regional.” Two others, with Galicia failing to fit both, is a stagnant, but also somewhat cyclical, pattern amongst those who consider themselves “more regional than Spanish” and a slight rise in those who identify as “equally Spanish and regional.” (Ironically, Galicia’s nationalist movement, in my experience, is talked about very little. It doesn’t make a lot of noise.) But, equally, or more, important is the fact that a super-majority of people in these regions consider themselves both “regional” and Spanish, even if the latter identity is diminishing in importance.

There are some important and relevant facts that may help put this data into context. First, Cataluña has been subject to at least two major waves of immigration following the Spanish Civil War (1936–39). One of these occurred during the Franco regime, where Spaniards of other regions emigrated to Cataluña, seeking work. The other began in the 1990s and is related to the immigration boom that affected Spain right up until the most recent financial crisis and depression. Many of the political programs to revive the Catalán language and identity were in response to the “threat” of immigration. Second, historically, Cataluña has been one of the, if not the, wealthiest regions of Spain. Barcelona is a major city and port, overseeing a significant chunk of Spanish trade. If trade helps introduce people of different cultures and preferences, it makes sense to claim that Barcelona was, and continues to be, a milieu for the exchange of ideas. All of this suggests that greater cosmopolitanism is to be expected, and there’s evidence that this cosmopolitanism informs the decision of forming a national identity.3 The fact is that the majority of Catalanes have a multicultural identity, including not only a tinge of Spanish and a hint of Catalán, but also a speck of European (see Medrano and Gutiérrez [2001]).

With this cosmopolitanism in mind, let’s consider cyclical movements in identity. Look at the “S < AC” line for Cataluña (top left): you see periodic increases and decreases in the proportion of people who identify as “more regional than Spanish.” These changes are multicausal, but identity is sensitive to exogenous conditions. For example, a financial crisis that points to Federal corruption is likely to bias people towards their regional identities, and vice versa. It’s not coincidental, I think, that some of the largest pro-Catalán independence demonstrations are occurring during one of the bleakest moments of contemporary Spain’s economic history. (Game theory fans see Aragonès [2007].)

In the framework of Spain’s political institutions, these fluctuations of identity are important. Remember that modern Spanish democracy came about only relatively recently, following the death of Franco in 1975, and that the debate over the constitution included discussion of periphery national identities. The result is a central government that surrenders quite a bit of power to the autonomous regions (equivalent to American states), and a competitive dialectic has emerged between local and central governments. This implies a dynamic that differs from what we see in the United States, where intra-political conflict is not as visible. In Spain, by contrast, there’s a primitive battle over jurisdiction that is driven, in large part, by the cultural preferences of that region’s society. I don’t pretend to know much about this process,but it’s clear to me that one of the most important factors is the conflict over national identity. In other words, the political dynamic in Spain, or even Cataluña specifically, owes much of its existence to the fact that members of that society are willing to switch allegiance.

My interpretation of what I’ve read on Catalán nationalism has influenced my views on what future governance may look like. Consider the concept of Tiebout competition (Tiebout [1956]). Suppose there are such things as public goods, which are goods that cannot be optimally provided for through the market process. An alternative way of providing them is through government. The main issue with this approach is that we don’t know the process by which the government can optimally allocate inputs towards the provision of public goods. Tiebout’s model of political competition is an early contribution to this debate. According to the model, if there is freedom of mobility, governments that provide an excess supply of public goods will suffer from emigration and falling tax receipts. It is a political profit and loss model. “My” theory doesn’t require the assumption of freedom of mobility. Rather, it assumes that the transaction costs are low enough for territorial changes of allegiance, such that governments have to peacefully compete for land (well, the clients who own that land). It is Tiebout competition, but on a different dimension.

My approach is broadly consistent with light postnationalism. The kind of postnationalism I have in mind would be something like Yasemin Soysal’s The Limits of Citizenship (admittedly, some ideas in this book have seen better days — e.g. the resurgence of anti-immigrant nationalism in Europe conflicts with some of her predictions). One theme in that book is the provision of welfare in a world where borders are becoming increasingly meaningless. For example, what if Turkey is providing for Turks living in Germany? This kind of action would fall outside the scope of a typical welfare system,  focused on those living within a specific territorial jurisdiction. If Soysal talks about overlapping jurisdictions, what I have in mind are fluctuating jurisdictions. Ultimately, “fluctuate” might imply the same thing that it does when we talk about people choosing between internet providers.

I tend to think that cosmopolitanism, pluralism, and limited government are all related, in the sense that freer societies tend to be more plural and more cosmopolitan than their less free counterparts. Part of this comes from a clear belief in the Whig interpretation of history, as far as liberalism goes. “My” theory, however, doesn’t predict that society will necessarily become more free, more plural, and more cosmopolitan, but that if it does become more free, it will also be more plural and more cosmopolitan.


1. Take for instance Hans-Hermann Hoppe’s writing on immigration (e.g. “Natural Order, the State, and the Immigration Problem“). I don’t want to ridicule the article. I think it’s better than some would suspect without reading it, especially compared to Hoppe’s other work. But, Hoppe focuses on the right to exclusion, which to me reveals a preference for exclusive, or relatively homogenous, social groups. Another example is Rothbard’s prediction that a free society requires a ethically homogenous social group (I may be wrong; the only Rothbard book on ethics I own is The Ethics of Liberty, and I haven’t read it in full).

2. For some evidence that may confirm my priors see Thomas Jeffrey Miley, “The Discourse of Language and Nation in Catalonia,” Berkeley Journal of Sociology 46 (2002), pp. 46–78. Miley confirms that social/political role is related to identity.

3. Miley (2002) is an interesting interpretation of the political movement to revive Catalan identity. He distinguishes between movements that focus on differences between identities and those which focus on similarities. Miley’s narrative is that Catalán nationalism is driven more by similarities, attracting a more diverse population of constituents.