“Iraq has never seen this kind of fighting in its battles with ISIS”

That is the headline of a WaPo story.

Here is one bit.

“It is a bitter fight: street to street, house to house, with the presence of civilians slowing the advancing forces. Car bombs — the militants’ main weapon — speed out of garages and straight into advancing military convoys.”

The reports of ISIL tactics are alarming. Not to pander but I’ve been stunned and, frankly, sickened reading how horrific ISIL is as an organization. From rape to using chemical weapons on children, ISIL is one disgustingly barbaric group.  Most worrying is that, as many have cautioned, the bigger challenge will come once ISIL is pushed out of their strongholds of Mosul and Aleppo and they start to look more like a  traditional islamic terrorist organization.

For those advocating a bigger role the US in Syria and Iraq, it’s important to note that ISIS has it’s roots in foreign interventions. First the Russian invasion of Afghanistan and then the American lead conflict in Iraq. This isn’t to blame Russia or the USA but to warn against proposals that don’t take into consideration unintended consequences.

Vox has a very nice video explaining the rise of ISIL here.





The Selectorate theory and Syrian regime survival.


Demand is growing for US involvement in Syria but not all rebellions are created equal. There is no indication that the Assad leadership will abide to any power sharing transition like what was observed in Tunisia or Egypt. Not only is Assad’s party stubbornly clinging to power, Syria stands apart from other movements of the Arab spring for its level of conflict. The violence taking place in Syria is both qualitatively and quantitatively different. It is estimated that approximate 11 percent of Syria has died, some of which have occurred through chemical attacks.

There are a host of historical and cultural reasons why some regimes do not gracefully transition to power. Outside powers, such as Iran and Russia, are aggravating the situation and there is a precedent in Syria of taking a heavy hand towards rebellions. But little discussion has been made of how the existing Syrian political economy exacerbates the length and amount of conflict being observed. I suggest that the reason why Syria is such as violent mess is because of the incentives of those in power to remain loyal to the leadership. I draw on insights produced by the selectorate theory which argues that regime loyalty and the lengths they will go to retain power are shaped by two institutions, the winning coalition and the selectorate.

For the uninitiated, the selectorate model argues two institutions shape regime stability. The selectorate is the subset of residents that have a genuine opportunity to select state leadership. Selectorate membership is derived from state specific institutional characteristics such as lineage, wealth, or legal qualifications. In democratic regimes, especially those commonly referred to as western, the selectorate is generally all citizens over the age of 18. The size of the selectorate in authoritarian or military regimes is usually more narrow and generally consists of a small coalition of elites. The winning coalition is that subset of the selectorate that abides by state specific institutional conditions to confer political authority over the residents of the state. Political candidates of any style of regime must by definition form their winning coalition of supporters from the selectorate. This is done through public policy which can be allocated to either private or public goods. What shapes the allocation of state resources is the pressure of a challenger. If an incumbent fails to offer members of the winning coalition the best public policy bundle, a challenger will be able to persuade members of the winning coalition to defect. Holding utility constant, it is assumed that members of the selectorate would prefer to receive private goods to public goods. In general, the larger the wining coalition, the more incentive leadership has to offer public policy in the form of public goods. This takes the form of efficient institutions, the impartial rule of law, and transparent governing. When the winning coalition is small, there is an incentive to offer public policy to the winning coalition in the form of private goods. This includes state licenses, income from rent seeking, and other forms of corruption.

The relationship is made clear through an example. Let R be the total amount of income produced in a state where R= $1,000. If W is the size of the winning coalition and all of the state resources are allocated towards private goods, then $1,000/W decreases as W increases. That is, $1,000/2 > $1,000/4, 1,000/4 > $1,000/6, and so forth. Public goods, however, are not discriminatory and their consumption by one individual does not preclude consumption by others. It is assumed that the preference for public goods are uniformly distributed in the winning coalition so that all members of the winning coalition benefit evenly from their production. Unlike private goods, the size of the wining coalition does not effect the consumption of public goods. If R=$1,000 and the public good gives each member of the winning coalition the equivalent of $30 worth of utility, then for a state with a winning coalition of 40 members, public goods will be produced because $1,000/40=$25 per member, which is less than $30. If the winning coalition consists of 20 members, then $1,000/20 = $50 and private goods will be produced because $50>$30. The more members in a winning coalition, the less state resources are available to be distributed among those in that coalition. For those interested I provide this more formally as an addendeum.

The theory has a host of interesting applications including the democratic peace theory (winning a war is treated as a public good so democracies will divert a large amount of resources towards winning them) to how leadership responds to natural disasters.

In addition to shaping what sort of public policy state leadership will offer, the selectorate and winning coalition explain regime loyalty. The amount of loyalty is proxied by the ratio of the winning collation to the selectorate, or W/S. If W/S is large, the loyalty of those in W will be low because if W is large or S is small then the probability of being included in the next coalition is high. If W is small and S large than the probability of being in the next coalition is low. Let W = 300 and S = 2,000,000 then the probability of being in the next coalition is 300/2,000,000 or 0.0015. This is a very rough approximation of North Korea. If W = 65,000,000 and S = 150,000,000 then the probability of being in the next coalition is 65,000,000/150,000,000 or 0.43. This is a very rough approximation of the United States.

A small winning coalition with a large selectorate induces loyalty as public policy is largely defined by private consumption and the fact that a member of the winning coalition is easily replaced. A member of W is less valuable to the challenger when W/S is small. The probability of being in the next winning coalition is too small and the costs are too great to expect any orderly regime change. Unfortunately for American options in Syria, a large selectorate and a small winning coalition is what Syria has. The ratio W/S explains a good amount of why Syria is approaching its 5th year and why the violence only seems to being increasing. Those in the winning coalition have demonstrated that they will go to great lengths to retain power including the use of chemical weapons against civilians. This means that the Assad regime will not go quietly as other regimes such as Egypt which has about half the size of a selectorate as Syria does. As I’ve wrote before on this blog, a sober assessment of American priorities should be taken prior to taking initiatives towards involving itself in Syria. Those advocating that “Assad most go” needs to mindful of the existing political economy in Syria and what that means for a power transition. After 5 years of civil war, it’s obvious that a regime change will not occur with a light footprint strategy. If the United States wants a power transition, it needs to fully commit because the Baathist party of Syria has far too much at stake to entertain any minor compromise of its political power.



Below is a simplified version of the selectorate theory. I’m still new to wordpress so if someone has a better way of writing equations let me know, but I was forced to cut and paste using my MAC’s screen shot feature.

I reduced the model to only demonstrating the relevant parts for the above discussion. The model is more involved and elaborate than presented here, almost to the point of sacrificing parsimony for granularity. For the interested reader, I highly recommend the following book. It is the most comprehensive treatment of the Selectorate Theory to date and has deeply shaped my thinking on foreign policy.

State leadership will attempt to maximize discretionary income, which is the difference between total state revenue and public policy.



Total state revenue explicitly modeled takes a common form. It is assumed that the tax rate is applied evenly across all residents. I treat as exogenous.


Public policy offered by the incumbent is a bundle of public and private goods so that


Public policy of the incumbent can be modeled explicitly as


It is assumed that private goods are distributed evenly to the winning coalition which allows W to serve as a proxy price for private goods. If M­g is the amount of an incumbents public policy allocated towards private goods, then each member of W receives Mg/W share. If the winning coalition increases, so does the overall cost of providing private goods. If the winning coalition decreases, so does the overall cost of providing private goods.

The objective function of the incumbent fully assembled and explicitly stated is below.



Leadership will choose to maximize If a Markov Perfect Equilibria is assumed, then holds due to the challenger constraint. That is, to prevent defection from his winning coalition, the incumbent was required to offer a bundle of private and pubic goods in the previous round that provided at a minimum the same amount of utility to the least affiliated member of his coalition that a challenger did. The model assumes that each member of the selectorate has a level of affinity for political leadership that is based on cultural and ideological connections. The full set of affinities for those in W can be linearly ordered based on the strength of affinity for leadership. If is affinity then the set of all affinities of those in the selectorate can be arranged where 0< < 1. Affinity for leadership is treated as a residual of support after utility from public policy is consumed. This will allow for any stalemates between incumbent and challenger to be resolved if utility derived from public policy is the same between incumbent and challenger. That is, a member of the selectorate will arrive at their choice of leadership based firstly on utility derived from the combination of private and public goods offered by leadership. Affinity for leadership will only decide support of leadership if public policies offered by candidates are the same. A political rival will attempt to persuade the member in the incumbents winning coalition with the lowest affinity (or lowest affinity member (LAM)) to defect by offering a public policy that offers them more utility than the incumbent. An incumbent only need to match this utility as he benefits from what is described as the ‘incumbency advantage’. This is the advantage enjoyed by incumbents when the utility offered from the challenger is discounted by the fact that a challenger must first attain power and the utility offered by someone not in leadership is uncertain. To defect from the winning coalition would result in a forfeiture of all utility offered by the leadership’s public policy.

The constraint is that the incumbent must match the best public policy bundle of any challenger.






Syria and mission creep.

During the third presidential debate, the probable 45th president has confirmed her support for a no fly zone (NFZ) in Syria. Calls for a NFZ are mostly driven by humanitarian reasons. The vivid imagery of Syrian suffering pressures the west to respond and a NFZ is a way for the United States to address this suffering while minimizing risk. The problem is that the NFZ is a policy proposal that doesn’t match the facts on the ground and will most probably result in mission creep.

Mission creep is the tendency of small scale interventions to incrementally expand into more ambitious projects. The term was originally coined by Jim Hoagland when describing US involvement in Somalia. What started off as a humanitarian mission in a failed state was modified into a police action against a local warlord. The consequence was the death of 18 American soldiers and approximately 200 Somalis.

How does this apply to a NFZ in Syria? A NFZ would fail to properly address the civilian casualties in Syria and would lead to a demand for either a safe zone or result in a dent in American credibility. A NFZ is just what it says. It’s airspace that is patrolled by the west to prevent belligerent parties from entering with their aircraft. Note that a NFZ is conceptually different from a safe zone as a NFZ does not address conflict on the ground. Because the demand for a no fly zone is driven by concerns over civilian casualties, it’s important that a NFZ properly address the cause of these causalities. Unfortunately a NFZ in Syria does not and this is confirmed by the data.

Finding reliable and current casualty rates of Syrian civilians is difficult to attain. Finding how they were killed is even more difficult. Violations Documentation Center (VDC) is the only source I know of that produces data on the number and method of civilian casualties but they do not publish regularly. In September of 2015 the NYT published VDC data which noted that of the 85,404 civilians casualties they were able to record, approximately a fifth were from Syrian aerial bombings. Their most current report is part of their monthly series (which is also not published regularly) which reports that for August 2016, 39 percent of the 1,737 civilians deaths were from aerial or barrel bomb attacks. An increase but most probable from the involvement of Russia and I would assume that most of these deaths are in Aleppo which would not be included in any proposed NFZ.

What does this have to do with mission creep? Everything as the policy prescription does not match the reality on the ground. As noted, most civilian casualties of the Syrian conflict are not the result of aerial bombardments but from sources other than Russian and Syrian aircraft. Additionally, most aerial attacks are in areas that would never be plausibly included in a NFZ. But even if a NFZ were to reduce civilian deaths from Syrian and Russian aircraft, overall civilian casualties may not significantly drop as the Syrian government could substitute tactics and rely less on aerial attacks and more on mortar and artillery. What is to stop renewed demands to address the civilian casualties that continue after the NFZ is erected? What do we expect the US response to be to ground conflict between the multiple belligerents occurring inside the NFZ? What would happen if a Russian aircraft entered into the NFZ? These are all questions that have not been discussed at this point and all are plausibly answered by some form of mission creep. A NFZ only serves to deepen the United States into a conflict by way of an inappropriate policy. It is deeply unfortunate but the truth is Syria has no strategic or economic (oil) value to justify American use of force in the area. If the aim of the NFZ is to save Syrian civilians, the most obvious solution would be to have some sort policy which allows for the resettlement of refugees. Such a policy would be politically difficult, especially in the age of heightened demographic anxiety, but would directly address the tragedy of civilians being caught in the middle of a civil war.