Saturday, March 27, 2010

Sanctions, Sovereign Borrowing and Financial Integration

. Saturday, March 27, 2010

Over the last couple of months, Sarah, Will and I have been working on our MA theses, each hoping to make a sufficiently important contribution to the discipline (or at least show sufficient potential to some day contribute something) to warrant the department to continue funding us and let us start studying for comprehensive exams and work on our dissertation proposals. I'm writing my thesis on the intersection between economic sanctions and finance, specifically how financial integration and sovereign borrowing affects a target state's decision to acquiesce to sender demands in high politics cases.


Earlier this afternoon, I stumbled upon an article titled "Don't Sanction Dictators" by Jason McLure that was published in Foreign Policy last summer that has a bit in common with the argument I'm trying to advance. In the piece, McLure argues that sanctioning dictators is a futile policy choice for advanced countries and uses the threatened sanctions against Eritrea as he makes his case. McLure argues:
Sanctions are made to cut countries off from vital international exchange. The trouble is, Eritrea already trades less with the outside world than any country in Africa and places 210th out of all 226 countries and islands for global commerce.
He then discusses the specific case of Eritrea and how sanctions against dictators won't work because they won't respond to the coercive attempts of larger states and concludes:
These lessons apply to sanctions on dictators more broadly. How do you punish North Korea with sanctions when its trading partners are already limited to a handful of countries -- none of which are likely to pay heed to a harsher set of rules? How do you choke Zimbabwe's Robert Mugabe when his strongest rationale for staying in power is to save his country from the hands of countries who would (and do) impose sanctions? Perhaps it's no wonder that such countries' leaders not only survive sanctions, but use them to justify bad behavior.
I'm really happy to see McLure make this argument, and the beginning of his logic is similar to what I am arguing in my thesis. However, McLure stumbles in a couple of ways, specifically when he conflates "international exchange" with trade flows and makes a distinction between autocratic and democratic governments that I argue actually doesn't matter. Analyzing sanctions episodes using a political economy approach, specifically looking at the influences of trade and finance on sanctions success makes a lot of sense and should give our explanations greater traction (at least I hope it does). This is one way in which the extant literature on economic sanctions has been surprisingly weak.

I disagree with McLure and argue that it's not that dictators are better able to withstand the coercive pressure of larger powers simply because of the characteristics of dictatorships, it's the fact that most dictatorships are not intricately linked with the international financial system to the point where larger powers (i.e. the United States) can impose sufficient costs to induce them to cooperate. This characteristic can't be solely attributed to dictators. It's in fact possible for democratic states to have very little integration with global financial markets and thus hold the same financial characteristics that are common in certain autocracies, and you can also have an autocratic government (Singapore) that is heavily integrated with and dependent on financial markets. Furthermore, I object for an array of reasons with immediately relying on trade flows to make an argument for sanctions success without engaging the finance side (if you want to know why, I'll have my thesis up on my website sometime by late April).

Larger powers have very little bargaining leverage over states that are not integrated with and dependent on the international financial system. Why? Because the costs that matter (borrowing costs that are directly imposed on governments, rather than trade costs that are dispersed across the population) can't be unilaterally imposed by these states. They need the cooperation of financial markets, and more specifically institutional investors in order to impose costs on targets. These larger powers need to increase the risk associated with investing in a given country, which increases the borrowing costs of the target state as investors seek higher interest rates and/or decrease exposure to that market in order to compensate for the increased risk. A credible threat to act if a government does not change its policy coupled with the effects (or potential effects) of sanctions themselves are sufficient to induce an increase in a target government's risk profile.

Institutional investors can only impose costs on states that require foreign capital to continue to finance their international debt obligations, roll over their debt and fund their operations. Those that are not sufficiently integrated and not dependent on global markets for access to capital will have dramatically lower costs when engaging in sanctionable (risky) behavior because they don't have to worry about increases in borrowing costs when calculating the costs and benefits of a given policy bundle. It is not that states with one type of political regime or another are better able to withstand coercive pressure from advanced, industrial countries, it's the fact that those states that are dependent on external sources of financing have the most to lose from engaging in that activity and can't withstand the coercive pressure. Non-integrated and non-debt dependent states' individual cost-benefit analyses are not sufficiently altered by sanctions threats or impositions and thus the costs associated with defecting from the status quo are very low. Therefore, sanctions episode success should be based on two indicators: 1) a country's degree of financial integration and 2) it's external debt to GDP ratio. Or at least I hope it is so that my thesis committee will find it in their hearts to pass me!

9 comments:

Emmanuel said...

Also see Ian Bremmer's J-curve theory on the effect of sanctions on authoritarian regimes. It places a bit too much emphasis on American exceptionalism but it may be worth a read.

mdigiuseppe said...

Good Idea, I wish I'd thought of it. I work on the sovereign finance and armed conflict.

Alex Parets said...

Thanks Emmanuel. I'll definitely check out Bremmer's work this week.

ir - thanks! shoot me an email when you get a chance (aparets@unc.edu). I recently started some work on sovereign borrowing/exchange rates and armed conflict, specifically civil conflict. it's where my research agenda is going once the thesis is done. it'd be interesting to chat.

Thomas Oatley said...

Alex, this looks interesting. Please tell me you aren't testing an argument about political behavior using two measures (financial openness and government debt/GDP ratio) that are neither political nor behavioral.

Alex Parets said...

Thanks Thomas. I have an array of theoretically important political IVs in there like presence of an international organization and regime type along with other variables that the sanctions lit has identified as important. The results don't change when those variables are included. Regime type is not significant at .05 level (that level is meaningless anyway - the z-score is 1.73 so it is at .1). The effect of both financial openness and short term debt levels (as a % of GDP) are quite powerful and both highly significant (z of 2.35 and 4.12 respectively).

Both of the main political variables (IO and regime type) are significant at varying levels (my dummy for a multilateral international institution is highly significant as expected which would make Lisa Martin happy), but the explanatory power of the finance variables (short term debt and integration) are much greater than the political variables. So my argument that exposure to sovereign bond markets are critical in explaining sanctions success because sanctions costs are imposed on the government by these financial markets and increases the probability that targets will acquiesce seems to have some pretty good empirical support.

Thomas Oatley said...

What is your dependent variable?

Alex Parets said...

sanctions episode success. binary measure of success recoded from the TIES finaloutcome variable.

Thomas Oatley said...

So then, you are testing your argument using abstract aggregates that cannot possibly impose anything on anyone.

So, my concern is the following:

At best, you have specified a constraint. And as we all know, constraints don't cause behavior.

Thus, the inference from your empirics seems to be that targets are more likely to acquiesce when they are more exposed to financial markets, but your empirical model does not explain why they do so. And if your theoretical model does offer an explanation for this, then this empirical specification doesn't test it.

But, seeing as how I haven't read your MA, this may be off target.

My issue is really with the field rather than your MA. At some point we need to stop pretending that correlations among abstract economic aggregates tell us more about politics than a well-executed case study about the impact of financial markets on target government's decision to acquiesce. I'm not holding my breath.

Probably a good thing I'm not on your committee.

Alex Parets said...

First off, I agree that a well-executed case study would be a nice addition to my project (more on that later). However, before engaging in a single (or multiple) case study, there is the need to establish a firm correlation between theoretically important constraints on (and for this question, causes of) government behavior (and I engage how and why short term debt and integration constrain govt. behavior and cause a government to choose one policy over another in my theory). My research question and sanctions outcomes are precisely about cases at the margin, where constraints are a significant explanatory cause of behavior. Sanctions success is directly affected by constraints imposed by debt markets on government behavior and government decision making. These constrains at the margin matter dramatically in a state's cost-benefit analysis regarding a given policy bundle. They directly affect the costs associated with engaging in a given action and the more costly, specifically the more direct these costs are on the government itself (for which short term debt markets are a good measure), the more likely the government is to acquiesce to a sender's demands.

I definitely am testing my argument using aggregate measures and I'd argue that the impact of these measures, if correctly theorized, do have a major constraining effect on government behavior and explain why certain states acquiesce while others don't (above and beyond what can be explained simply by looking at general "cost" or regime type). What I believe I've identified is a significant constraint that the sanctions literature has heretofore ignored, because they've spent so much time arguing about trade/economic interdependence, the potential for future conflict, general costs and regime type as primary causal determinants of a state's decision as to whether or not to acquiesce.

Plus, what's wrong with abstract aggregates, both political and economic? I seem to remember a recent article on political institutions and foreign debt in the developing world just published in ISQ that used them. The degree of political participation as measured by a countries polity score or the size of its winning coalition and its growth rate place a constraint on a government's decision to borrow externally. Isn't that what you just argued? A constraint affecting behavior?

Now on to the case study. I would absolutely love to do a case study, and I will as this project goes forward and hopefully grows into my dissertation. But a single, or two or three case studies won't provide the generalizability that I am looking for or that most of us would like. The case study will add richness, depth, and further support to my theoretical argument. But if the correlation between these important variables on a large-n level aren't there, what will I get out of a case study?

If I had the time and resources, I'd love to be able to fly around and interview and/or survey government officials in sanctioned states and figure out exactly what influenced their decision-making. But I don't. The identification of a significant constraint that at the margin has a large impact on state behavior is I believe a good and important contribution to moving the literature forward.

Sanctions, Sovereign Borrowing and Financial Integration
 

PageRank

SiteMeter

Technorati

Add to Technorati Favorites