Arab Spring and the stability of monarchy

Victor Menaldo has a new paper:

This paper helps explain the variation in political turmoil observed in the MENA during the Arab Spring. The region’s monarchies have been largely spared of violence while the “republics” have not. A theory about how a monarchy’s political culture solves a ruler’s credible commitment problem explains why this has been the case. Using a panel dataset of the MENA countries (1950-2006), I show that monarchs are less likely than non-monarchs to experience political instability, a result that holds across several measures. They are also more likely to respect the rule of law and property rights, and grow their economies. Through the use of an instrumental variable that proxies for a legacy of tribalism, the time that has elapsed since the Neolithic Revolution weighted by Land Quality, I show that this result runs from monarchy to political stability. The results are also robust to alternative political explanations and country fixed effects.

Here are his other papers.


The crazy exception to this rule is Lebanon. Lebanon is the most democratic Arab country and normally the most unstable, but so far it's also been spared.

I suspect you mean "spared the recent wave of violence".

It has the highest Christian population, and a Muslim population of just under 60%. And the lebanese disaspora has provided a vent for elites, leaving the country to sectarianization of a permanent proletariat. Average IQ still looks really low, but that's probably partlly the result of the diaspora. Hard for a monarchy to form in that country without a dominant 'tribe'. Prediction is that

As I've matured I've begun to feel that the fall of the soviets was the disaster that Gorbachev predicted. It put an upper limit on the expansion of the new-communism that is islam. And while the Arabs were unable to modernize, and so were the Ottomans, it was the weakness of the ottoman empire, that made the west's expansion possible. Friedman suggests the only solution is for Islamic civilization to develop a 'core state'. Iran is the only candidate and it's just another North Korea. The other alternative is to favor the strategy of the Germanic peoples that Hoppe recommends, and that's to create a multitude of small kingdoms.

States are vehicles for war.

States are vehicles for war.

It's amazing that anyone could believe otherwise.

States are vehicles for self defense.

Hans Herman-Hoppe. That is all.

Because timing of Neolithic Revolution totally didn't affect economic growth through any other channel.

But it says instrumental!

Aren't most of the monarchies oil-rich so they can buy off opponents? I guess Libya had oil too but maybe Quaddafi was just too pigheaded to go this route.

Yemen also had both oil and political turmoil. And Jordan produces no oil at all, but is very stable.

The rationale for his instrumental variable:

"The logic behind this instrument is
that in MENA countries where settled agriculture has historically been difficult, if not impossible,
tribal social structures have been more likely to survive. By extension, so have monarchies."

The justification for an instrument is not that it is correlated with the intermediate variable, but that the only plausible mechanism for the instrument to affect the end variable is THROUGH the intervening variable AND that the instrument is not subject to endogeneity.

So many instruments these days get the second part and the authors think they've brilliantly isolated a problem... forgetting that the first part is probably even more important.

So, we are to believe that a history of tribalism itself can ONLY impact political divisiveness THROUGH the probability of developing a monarchy? Because... tribalism itself doesn't cause conflict directly?

Instrumental variables should not be an excuse for sloppy thinking. Nor do the controls justify this thinking. Ethnic fractionalization is a poor control for tribalism (and yet still yields high significance).

Indeed, TRIBALISM itself might be a mere proxy for the original variable - agrarian land quality, since agrarian societies are known to engender/require more developed property rights structures and specialization of labor, which creates greater incentive for the population to invest in security of these rights.

There's no doubt about the simple correlation, but the argument for monarchies as causal is awful. SCIV - swiss cheese instrumental variable.

[Also, using an OLS based metric when the dependent variable is so lumpy overstates the t statistics... So, we are to really believe that the chance of a variable randomly being causal yields a t value of 10, which is a p value of 7.6 * 10^-24? REALLY? Because, as we all know, the dependent variable follows a conditionally normal distribution... As in, conflict measures never get bunched up or anything like that.]

You better be pretty old to be this cynical.

Repeat after me: "Publish or Perish". "Publish or Perish". "Publish or Perish".

Maybe you could do a paper where you systematically destroy lots of other papers for their lack of rigor--remember "Publish or Perish".


I'm that cynical and I'm not very old

Damn, StatsGuy - this is a great post.

We should ask Tyler sometime about the right way to read papers with instrument variables.

Didn't he write something a while back about how the more often an IV is used, the less likely it is that the IV variable is a good one?

But clearly we do like papers with IVs, and they do influence our thinking (and get published in damn good journals). And yet they clearly are not very good "evidence" for a position.

So - what do we do with IV papers? Also what is the single best IV paper? (My favorite has always been the one by Lakshmi Iyer that used an obscure British Imperial policy (the Doctrine of Lapse) to estimate the impact of colonial rule).

I think you're applying the wrong standard.

He's making an historical and a praxeological argument and looking for evidence in scarce and lumpy data that shows some sort of correlation. He is not looking at sparse and lumpy data in order to deduce causality. So he's basically saying that the data doesn't *conflict* with his argument. He's not saying that the data demonstrates the argument. Essentially, it's really hard to put data together on something like this, but what he can do is see if the data conflicts. Aside from that I agree with your criticism in principle but not in application: at face value, it seems that monarchies are better able to persist, no matter what era we're in.

The fact that we apply the mathematics of the physical world to the social sciences, given that we have so little data about ourselves that is other than vague aggregates, is bad enough. (We lack data and computing power for a more fractal mathematics than the Dynamic Stochastic Equilibrium Model.) But we must not also make the mistake of positivism on top of it. At least within the social sciences, if one can have a bias in the fist place it simply proves that biases or observations, or opinions are indeed facts. :) But that's a whole 'nother discussion.

If he were only claiming consistency, not causality, I would be less harsh. However:

"These results hold after controlling for oil rents, other possible
confounders, and country fixed effects. Finally, they appear to be causal: they hold after isolating
the exogenous variation in the MENA’s regime types via the use of an instrumental variable that
captures countries’ legacy of tribalism, measured as how much time has elapsed since the
Neolithic Revolution weighted by a country’s land quality."

Your argument is invalid. Stop being a bitter little man and wash up for dinner.

Monarchies tend to fall in waves. It didn't look like a very stable form of government from 1775-1875, the time period when many, if not most, of the world's European monarchies fell or were relegated to symbolic status.

The population ruled by the British Crown and the Tsar of Russia in 1875 was probably larger, both absolutely, and as a proportion of the world population, than the total population ruled by monarchies in 1775.

By 1875, the British Commonwealth was a monarchy in name only. The Tsar of Russia, of course, fell in the Russian revolution not all that much later.

I should have checked dates more carefully - it wasn't until 1876 that Victoria became "Empress of India", and though she ruled through a Governor-General, India was definitely under monarchical rule through that period.

This appears to ignore some important variables.

Namely, the 800 lb. gorilla in the room - American "empire" (for lack of a better word) and geopolitics.

Many of these monarchs have been American clients and proxies.

A lot of the countries that had big uprising were client states too - Egypt, Yemen

I wonder if they're comparable to something like Saudi Arabia. Saudi Arabia is more important isn't it?

Saudi Arabia has a lot of oil money to bribe the population with. Bahrain had a big uprising, it's one of the key US client states in the region, has oil wealth and is a monarchy. There's exceptions to everything.

"Many of these monarchs have been American clients and proxies."

In the Arab Spring years, yes. Otherwise, no.

MR, Mencius Moldbug edition?

Yes. Thus endeth the lesson given by Hans Herman Hoppe. :)

"The best reason why Monarchy is a strong government is, that it is an intelligible government. The mass of mankind understand it, and they hardly anywhere in the world understand any other." - W. Bagehot. Bagehot then goes on to discuss what Bryan Caplan has more recently called "myth of the rational voter". But the fact that government is complex and citizens are easily confused, misled or frustrated is not sufficient to answer the question of monarchy.

But the better answer, provided By hoppe, is that elected officials always create a Tragedy of the Commons out of the society, its institutions and its economy - elected officials have every incentive to spend, to sell off, to destroy traditions, and to create factions so that they can profit from dispute resolution among them. A monarch has the opposite interest: he has every incentive to create an enduring nation for his people, so that he can persist his family heritage. And even better, since the monarchy -- by its social status alone -- creates at the very least, mating advantages, and at the very best, wealth and a place in history for family members, the family will happily commit regicide if the monarch acts against their interests. (And history is full of examples.) You can kill a monarch, but removing politicians is an exercise in futility. As soon as one is gone, another pops up in his stead.

The answer that I advocate, is that in a society where political power is unattainable, the only venue for status seeking is the market. And success in the market is good for consumers and entrepreneurs alike. Conversely, all political action is merely a distraction - a waste of time and effort in lost productivity an liesure because political power must both be pursued and defended against. All commercial action is a benefit to someone, somewhere.

The best and most stable form of government we have yet discovered, consists of a rigid constitution, under hereditary monarchy, where the monarch has limited power of veto, perhaps limited to dismissing the government, with an upper house having rigid criteria for membership, and whose responsibly is limited to ascent or veto, and a lower house from citizens who meet rigid criteria for membership, and who alone can initiate bills, where both houses are appointed by lottocracy, and where there is no compensation for service, and where all administration performed in the private sector, by organizations and individuals capable of being fired. Balanced an independent judiciary that administers the common law.

This system is a defense against the usurpation of the government, a defense against the natural corruption of bureaucracy, a defense against the fashion and passion of the public, a defense against the politicization and factionalization of society, a direction of competitive energies to the market, and relegates reward for public service to that of social status. But best of all, society is socially bimodal, and having houses of government that represent their interests, and force a compromise provides a vent for stress, and a means of cooperating through compromise and exchange.

Hmm. "Appointed by lottocracy" and "no compensation for service" seem contradictory - most people cannot afford to spend years working without pay. Or is one of the "rigid criteria for membership" a family fortune? Otherwise, I see a large fraction of your citizen-legislators either declining the dubious honor, or trying real hard to figure out how they can make an immediate financial profit from their legislative office.

I'm not sure why it's a contradiction. (See Is Democracy Possible? by John Burnheim (Australia)) It's certainly been suggested in almost every era. It's worked quite well in our past.

There is very little discourse in government that is not open to comprehension by any literate individual who is interested in working reasonably hard. There is very little in government that is not open to undue influence (direct or indirect) and political privatization or bureaucratic privatization.

Schumpeter's observation that the struggle for power was not between labor and capitalists, but between public intellectuals and entrepreneurs was illuminating. Politics of personas and ideology is detrimental and expensive. The money made, conflict created, and productivity misallocated wasting time on the politics of personalities, and the ideologies by which those personalities make decisions, instead of elucidation of the issues is probably quantifiable. The question is whether it is a form of corruption or not? :)

It is simply foolish to think that you can somehow eliminate politics and the political sphere.

"All commercial action is a benefit to someone, somewhere. "

Couldn't you say the same thing about political action?

CBBB: All political action is by definition coercion, isn't it? :)

So is a lot of market actions. We weren't talking about coercion, however; you said that “All commercial action is a benefit to someone, somewhere. ” This doesn't rule out coercion - the political action doesn't have to benefit everyone (nor do all market actions benefit everyone) they just need to benefit SOMEONE.

This will get to be too long of a discussion. But no, there is a very big difference between alliances of producers, and alliances or rent seekers.

And high prices are not coercion. Barrier to competition is coercion. If you subscribe to the position (supported by evidence) that market coercion is only possible in concert with the aid of government, and that all monopolies are caused by government, then, indeed, all politica activity is coercive and all market activity is voluntary. If your matter of account is subjective, and your means voluntary, all political activity can only be coercive, and therefore not subjective or voluntary.


Why can't high prices be used to coerce?

Because one always has the option of not paying.

Only in the abstract, in reality....not so much

If you look at the UN Human Development Index for 2011, four of the top five countries (i.e., all except for the USA) is a monarchy, as are nine of the top fifteen ( By contrast, 44 of the bottom 46 (by my count) HDI countries are (nominally, at least) republics. Pretty clear that, ceteris parabus, you're much better off living in a monarchy than a non-monarchy. Of course, it's not at all clear that there is a causal relationship (surviving monarchies are, by definition, countries that have had more stable histories and peaceful constitutional development). But, on the other hand, it's not entirely implausible that a monarchy is a good thing.

The USA has an elective monarchy. Hasn't it?

Not at all.

Each of the last three presidents have failed to pass some desired legislation despite having control of both houses of Congress. Each of them have passed some laws with only a bare majority - their own party members voting against it for fear of losing their seats.

Being a US president is far from being a king. The office isn't even as powerful as being a Prime Minister.

Who said Kings were more powerful than Prime Ministers?

"I show that monarchs are less likely than non-monarchs to experience political instability,"

Perhpas the reason is because, when a monarchy experiences instability, sooner cease to be a monarchy?

Good point.

I'm pretty sure the data is in, that monarchies are more stable. And it would be pretty hard to refute it. Most of human history consists of monarchies of one form or another. Certainly the most stable form of government in the past 300 years has been the English model. The American model might have done well were it not for the Louisiana purchase, which set the southern export economy and the northern manufacturing economy at odds -- competing for the opportunity to expand into the new territory. Democracies and republics have pretty fragile heritages. They are, it seems, a 'Luxury Good' we eulogize without logical reason. It's a convenient means of claiming legitimacy during a transition of power from one caste to another, and little more. That's what Pericles used it for: to get support for building an empire - the Delian League - so that he could spend the next two years at war. That's what the Jacobins did with France. It's certainly what the Founding Fathers did in the states. "Expand enfranchisement to justify a transition in power". Let's not forget that the reason for the American revolution against England was to escape paying for the war that England had fought almost entirely on our behalf - and nearly bankrupted her.

Monarchies have a long (low) time preference, and they deny people access to political power, focusing their energies on useful activities more so than ideologies when in pursuit of social status. All that's left is to rely upon is ideology under democracy. Voting MUST be based on ideology. It is not possible for voters to know enough. It could be argued that ideology is actually enough to go by. But I'm not sure how that's tested. :)

"Most of human history consists of monarchies of one form or another."

Actually, most of human history consists of relatively egalitarian, non-hierarchical hunter-gatherer bands; behaviorally modern humans have existed for at least ~50,000 years, while monarchy only became possible after the Neolithic Revolution 10,000 to 7,000 years ago.

All moot points. By their very nature monarchs, like religions, cannot be created de novo. New religions, like communism or antropogenetic warming etc, tend not to stick long term because they lack historical legitimacy. Similarly, someone trying to become monarch and create a dynasty would not stick - he would be thrown out by a competing group, because that is what politics is, because he would lack transcendental legitimacy.

All those yearning for monarchies a) have clearly never lived under one b) have never proposed how to create one de novo.

Conservatism is useless if it has no plausible recommendations for the future.

anthropogenic: the global warming is anthropogenic

oops. no preview

those yearning for monarchies a) have clearly never lived under one
yes, just imagine those poor wretches living under the monarchical yolk in Canada, Australia, Denmark and the Netherlands.

that should be "yoke". oops.

I have lived in one of the monarchies above, another one and two republics, so I think I know what I am talking about.

By their very nature monarchs, like religions, cannot be created de novo.

Say what? Both ends of the comparison are wrong. Consider 19th century Sweden; consider Mormons and Baha'is.

But USA may have been able to bring back the king in Afghanistan.

Living under a good monarchy sounds pleasant. The problem is living under a bad one. How long are we supposed to put up with a succession of Loiuses?

I am still for abolishing monarchies, including the one in the United Kingdom:

Then you're not knowledgeable enough about economics and politics.

And of course, on your site, all you say is "I disagree" and you don't put forth any arguments. The one daft attempt that you do is a humor sketch that entirely misrepresents the economic realities of the manorial system. Furthermore, Knights were literate, they ran businesses, they were poor and weak. Peasants were illiterate semi-human animals, quite unlike the rhetorician in your example. Education was incredibly expensive, mostly because books were. Democracies are justifications for the state, and the purpose of the state is to make borrowing money for war - either offensive or defensive, possible. A united germany is a defense against France. A soviet union is a defense against germany.

About monarchies being more stable:

european countires with the same political regime in 1900 and today (ecludinf micro-nations): United Kingdom, Denmark, Sweden, Norway, Switzerland, Netherlands, Belgium and Luxemburg (all other countries had revolutions, cival wars, coups, etc.).

At the first look, seems a good point for monarchy : 7 in 8 "stable" countries are monarchies. However, almost all european countries in 1900 were monarchies (only France and Switzerland were republics, I think). The rate of political survival of monarchies are really higher of republics? I doubt.

If we count by countries, we have a 50% survival rate for republics (1 in 2) ; for monarchies it is perhaps 39% (7 in perhaps 18 monarchies in 1900);

By population it is more complex - Switzerland is much less populated than France, meaning that the, "survival rate" or republics is less than 50%. But, by the same reasoning, the overtrowed monarchies were the giants of Europe (Russia, Germany, Austria-Hungria, even the Ottoman Empire), while (excepting UK) the surviving monarchies are tiny populated countries (meaning that the survival rate of monarchies is also smaller than 39%).

I would guess monarchies are an older form of rule. Nobody makes themselves King anymore. If you take over now, you become Beloved President For Life. As a result, monarchies that are in existence today are those that developed stable institutions. But as I said, that's basically just a guess.

Comments for this post are closed