Friday, October 20, 2006

Economic Origins of Dictatorship and Democracy

I had a depressing conversation while seated next to a man from Cote d’Ivoire while traveling about a month ago. What particularly struck me about our conversation was when he was telling me about the second coup in his country since 2000, again with factions breaking mainly along tribal ethnic lines, I realized how similar that situation was to the situation in Europe before the formation of nation-states. At the time I had been reading Oscar Jaszi’s The Dissolution of the Hapsburg Monarchy, which is a detailed illustration of (failed) state formation. Feudalism is noting more then tribalism plus agriculture (and indeed, it may be more accurate to describe contemporary tribalism in Cote d’Ivoire and elsewhere as feudalism). It occurred to me that I could not think of any successful nation-state (besides the neo-Europes, US, Canada, New Zealand, and Australia¹) that did not go through a period of rule by an oppressive autocratic regime, similar to the Hapsburgs: authoritarian regimes which imposed a national identity over the existing clan and tribal identities of feudalism. Henry the VIII in England demanded loyalty to the state church, as did the Hapsburgs in Spain and the Austrian Empire, the result of the consolidation of the French and Russian nation-states famously triggered revolutions. China had Qin and Japan had Tokugawa. India had the British and Turkey had Attaturk.

But most of Africa didn’t have a similar experience. Most of Africa can be characterized as having ethnic conflicts that discourage investment, causing a catch-22 of development: people fight because they are poor and they are poor because they fight.

Recently Delong posted a review of Daron Acemoglu’s new book Economic Origins of Dictatorship and Democracy, and after reading the book I believe my hypothesis concerning absolutism as a necessary stage in state formation is supported by his theories.

To summarize the book: Democracy happens when there is a credible threat of revolution while there is also a large part of the wealth tied up in easily destroyed items (like factories, railways, and equipment, which can be smashed, as opposed to a mine or fertile land, which are practically indestructible). The effect is that democracy will be granted by elites and anti-democratic coups will not be attempted after the establishment of democracy, both because elites are afraid their easily destroyed physical capital will in fact be destroyed in the possible turmoil.

That is all well and good, and indeed I can’t think of any successful democratizations (with the possible exception of Taiwan²) that do not fit into his model. However, Acemoglu and Robinson don’t really talk about transition: how do these countries caught in the development cathch-22 of belligerence due to poverty and poverty due to belligerence break that cycle and become stable successful democracies like most of the Western World? According to Acemoglu’s model, violent destructive governmental transitions happen when there is little cost to that violence; if there are no factories that could burn, then this “punishment” for the violence doesn’t exist, decreasing the disincentives for violence. This seems to mirror the development catch-22 within the Acemoglu and Robinson theoretical framework; countries that are prone to violence cannot develop physical capital, countries with little physical capital are prone to violence. How then can the fragile physical capital that will eventually be held hostage in order to stabilize democracy come into existence in this turbulent atmosphere?

It can’t. It seems that a necessary corollary from Acemoglu’s argument is that a strong and probably authoritarian leader must control and pacify all the various competing factions of the society, possibly through repression, long enough for physical capital to build up to a point where it can be ransomed for democracy.

The direct implication is that pluralistic idealism in the foreign policies of the developed world may impede the development of poor unstable countries if executed improperly. However, there are possible non-passive roles for benevolent foreign actors in this development process, chiefly, ensuring that the economy develops to a sufficient level under any extant authoritarian regime, as well as possibly assisting the destabilization of autocrats that have worn out their welcome (i.e. dictators who stay on past some point of physical capital accumulation), although this can be done non-militarily in some cases (e.g. Taiwan and South Korea, whose authoritarian governments both buckled in the eighties, partially due US policies developed by Carter that put diplomatic and public pressure on the regimes).

1: The exclusion of the neo-Europes can be explained by the fact that immigration tends to destroy existing tribal identity. Arguably, the only exceptions are immigrant criminal organizations, which still are significantly weakened by immigration and tend to disappear over time.
2: Taiwan is an interesting case in that it was in many respects similar to Singapore in anti-democratic nature. Singapore was held up as an example of a country which accumulated fragile physical capital but remained anti-democratic due to the supposed benign and egalitarian nature of this particular authoritarian regime. This egalitarianism, manifested in the wealth distribution most importantly for Acemoglu, caused a lack of demand for revolution. I would argue that a similar situation existed in Taiwan, except that political pressure from the US nudged the dominant elites towards democracy.

Friday, October 06, 2006

Partisan Markets?

Ezra wrote a thing on Tapped about how markets are flawed because some Stanford economists correlated suprise democrat election with stock market reaction. I don't read Tapped so I didn't see this a couple days ago. Some other guy critqued his attack on markets in general based on this one study, so Ezra responded on his site, which I do read.


Possible explanations abound, from rich investors rejoicing over the redistribution they're about to enjoy to businesses salivating over expected corporate welfare or regulatory favoritism. Whatever the reason…


There is a good chance that these findings are essentially inconclusive. There’s a reasonable chance that the Bloomberg story misinterprets what the authors found. And there’s a decent chance that there is a rational explanation…

You know what the great thing is about that interweb thingy all the nerds are talking about? You can look stuff like this (pdf) up and decide for yourself. If you look at the article referenced by the Bloomsburg blurb (same pdf) you’ll see the chart on page 23 which is the important reference for the article (that’s the chart that looks at stock market reactions since 1880, the rest of the paper looks much more closely at simply Dubya I and II).

If you clicked through and look at the graph, you’ll see that the correlation is heavily reliant on three outliers that occurred during the FDR and Truman. If we confine our study to the contemporary Post-War era, looking at data since 1952, the correlation is destroyed. The coefficient on Republican election goes from 0.02 to 0.0073 and probably more importantly, the Pearson’s coefficient, (which measures how well a correlation fits) going from an already weak 0.2 to a laughable 0.048.

This seems to say that the markets really liked Republicans, until they figured out that politics won’t influence equity prices that much. So the market really does figure it out eventually. That, or possibly the market really was much more heavily influenced by politics in the pre-war era, with the policies of the Republicans much more explicitly pro-big business and/or flat out corrupt.

Wednesday, October 04, 2006

They Ripped Me Off!

Well, probably not really.

I've been meaning to post about this for a couple days now. The NYT published a comparison between ancient pirates and terrorism (full text for non-subscribers) strikingly similar to my own. A lot of bloggers really liked the idea, including one of Leiter’s guest bloggers, so I guess I should feel flattered?

I think mine is better because I bring up the fact that Rome had also recently defeated another world superpower that helped keep these non-state bad guys in check.

I also have a comparison of Bush and King Richard I that is probably worth looking at if you like this sort of thing.

Sunday, October 01, 2006

Deficit reduction

Thanks to the tables Max references here I can make an estimate on what it would cost to balance the budget.

Max’s post is about the ridiculous comments made in the right blogosphere in response to these numbers. I’d like to point out that by my calculations, about 30% of income of all Americans not in the top bracket is taken by taxes if the 15% FICA is included (FICA is paid half by the employee directly and half by the employer directly making the cost of the employee all that much higher and therefore impacting salary). If the disproportionate effect of FICA on high incomes is taken into account (no income over $94,000 is taxed by FICA), people in the top bracket would also have an effective tax rate of about 30%.

Assuming away disincentive effects, removing the $94,000 limit would increase receipts by 120 billion and increase the effective tax rate (total adjusted income over tax paid) of the top bracket to about 42%. If we also increase the top two brackets’ marginal rates to 55% the marginal rate of taxation for the top bracket would be about 70% after including the expanded FICA and the effective rate would be about 55%, but we would get about another 130 billion on top of the 120 billion from increasing FICA. And if we eliminate the 10% and 15% capital gains brackets, moving both to 20%, we would get about 50 billion.

That would be about 300 billion without touching people who made less then $150,000 (except a few miscellaneous capital gains earners). The effective tax rate for people outside the top two brackets (people who make under $150,000) is about 16% not including FICA. If we increase that to 20% it would mean taking about a quarter more income taxes from the rest of the people; ten percent bracket would go to 12.5%, the 15% would go to 18.75%, and so on. This would increase the effective rate on people earning less then $150,000 to 34% and reduce the deficit by about 130 billion.

This all would leave about 40 billion in deficit, which seems like a reasonable amount to be able to cut from discretionary spending (most especially the bloated Bush defense budget). Note that all of this assumes away disincentive effects from increasing the top marginal rates. If indeed the increase in income inequality is due to increases in returns to education, then the disincentive effects will probably be small.