I. Introduction

II. Gallarotti

A. Basic argument: IOs have effects, but sometimes they are negative. They can exacerbate existing problems and create new ones.

B. Systematic failures of IO (i.e., when IO destabilizes international relations)

1. Technical difficulty: When attempting to manage complex, tightly coupled systems

a) Economic management is difficult and action can have unintended effects that make things worse rather than better, example of Louvre Accords leading to weakening of dollar because market actors saw the need to strengthen the dollar as evidence that, in fact, it was weak.

b) Important question here is also to ask whether non-action would have been better

c) In complex contexts, there are side effects that are complicated and hard to predict, and many of them may be contrary to your goals

d) Is this a question of avoiding responsibility for problems? Is non-action blamable?

2. Adverse substitution: When discourage states from seeking substantive or long-term resolutions and replaces responsible foreign policy (moral hazard problem – e.g., League of Nations as cause of WWII)

a) IO as substitute for world government

b) IO as substitute for responsible state policies that could deal with problems better than IOs do

3. Dispute intensification: Can intensify disputes when used as weapon of foreign policy

a) IOs become a venue for political battle to be done

b) Linkage can be used to fuel rather than de-fuel conflicts

c) IOs as locus for collusion among members against nonmembers – e.g., most alliances, OPEC and other cartels

4. Moral hazard: mitigates consequences of irresponsible behavior makes actors more prone to engage in it. Often leveled at IMF and World Bank.

a) IEA insures against energy shortages and thereby makes them more likely

b) IMF and WB provide loans in response to bad economic policies which makes those policies more likely in the future because those who committed the policies don’t have to feel the results of those policies

C. Once again, back to situation structure as important factor in understanding IR

1. 394ff: most games are not PD or stag hunt and responding as if they are causes more problems than it solves

2. Is IO limiting "public bads" or trying to create "public goods"? Gallarotti argues that IO is better in former case than latter, but IOs even more than creating public goods create private goods, which would be created anyway so IO is redundant

3. IOs would be more effective in creating long term answers (states take care of short term ones reasonably well, G argues) but they have difficulty doing so

D. Gallarotti ends with a call for more limited forms of international organization – is this convincing? What would you need to show to accept this?

E. Basic argument of Gallarotti is that, in many cases, the counterfactual non-regime world would be better than the observed regime world. This shows the importance of careful analysis of counterfactuals

III. Smith

A. A puzzle worth explaining in first para of article: "more than thirty regional accords were officially notified to the GATT between 1990 and 1994 alone, a total exceeding that of any previous five-year period" (201). Why? Note that it is NOT the formation of the EU – since that had occurred before.

B. Basically an article about treaty formation – why do these treaties have dispute settlement provisions of the type they have? "At issue in this study is the nature of ex ante institutional design, not the record of ex post state behavior" (202). "My question is not why or when nations coopertate in reciprocal trade accords, but how they do so" (214).

C. DV = level of legalism (203)

1. How do we tell different dispute settlement provisions apart in terms of their level of legalism?

2. Review of third party complaints

a) Right to third party review of complaints at one end

b) Consultations only at other end

c) Note, here, need to identify type of variation in DV before can explain it

3. Status of rulings from process

a) Formally binding in legal terms

b) Or not

4. Type of decision-makers and their selection

a) Standing tribunal that rule on any and all disputes

b) Ad hoc arbitrators for each dispute

5. Who has standing to file complaint and obtain rulings? States, individuals, groups, parts of IOs?

6. Remedies available

a) Direct effect – national law makes decisions binding automatically

b) Retaliatory measures allowed

7. Nice chart on 207 of elements of DV – NOT the IV here but simply defining and identifying and operationalizing the DV

D. DSPs as forums for reducing costs of agreement while still receiving benefits thereof

1. Tradeoff between policy discretion and treaty compliance by others. That is the exchange that is involved

2. Can’t identify who will be harmed in advance so DSP provides way of dealing with harmed actors in own state

3. Also benefits of inducing other states into compliance through procedures

4. DSPs "constrain the opportunistic behavior of foreign governments" and "promotes the confidence of the private sector" (211).

E. The IVs: Preferences for different forms of DSPs

1. Trade dependence (state specific) – more trade dependence leads to preference for more legalistic

2. Relative economic power (state specific) – more power leads to preference for less legalistic

3. Depth of liberalization (treaty specific) – more liberalizing agreement leads to more legalistic

4. Note direct relation of IVs to DV as defined earlier

5. Who dictates terms? Largest economy since has lowest value from agreement (needs it least) and so dictates terms to other countries. Therefore more equal pacts have more legalistic arrangements – implication, NAFTA should have less legalistic terms than EU (notice EU may be unequal but its less unequal than NAFTA).

F. Analysis

1. Selection of cases to "ensure comparability", i.e., to use control variables in case selection process

2. Notice "coding" of DV from legal text which is complicated to summary of legal text to five possible codings of none, low, medium, high, and very high. Describes and defends his coding system by a few examples of how he coded as he did.

3. Measuring of IVs

a) Trade dependence : notice disconnect here – describes as important variable on 211 but doesn’t measure.

b) Relative economic power : Devises a thoughtful metric that tries to capture economic inequality – complicated math but conceptually interesting and convincing

c) Depth of liberalization : uses IMF categories of free trade areas, customs unions, common markets, and economic unions (four values of IV)

d) Interaction of b and c is third variable, not trade dependence.

G. Note that Smith goes through rival explanations of his results – acknowledges shortcomings of analysis and attempts to address them. Trying to get at truth rather than support his own argument.

H. Conclusions

1. "Legalistic mechanisms are unlikely where asymmetry is high or integration is shallow" (237).

2. Suggests that case studies of individual cases would support large number analysis done here. But note that does fairly non-mathematical analysis of large number of cases.