Response to Franchino

Fabio Franchino (2013) provides an insightful and critical examination of the findings of my book Veto Power, as well as the volume by Finke et al., Reforming the European Union. Naturally, I welcome Franchino’s praise, but I also gladly accept his criticism. I would like to respond, though, to some of the issues Franchino raises. I briefly discuss Franchino’s replication study of my main empirical results, and then move on to discuss my treatment of intergovernmentalism, and specifically liberal intergovernmentalism (LI) (Moravcsik, 1998).

I welcome Franchino’s replication of my main empirical findings using J-tests to distinguish between the competing models. Throughout the course of working on the project, I explored many approaches to testing these models, but it did not occur to me to take Franchino’s approach. As Franchino notes, the two independent variables representing the institutional and intergovernmental models are very similar to one another – they are, after all, based on the same member state preferences, just weighted differently. Franchino’s method provides a very clean way to assess the explanatory power of these different theories. I am quite pleased, but not surprised, that the results of this different method confirm my original findings that issue inclusion is best explained by the salience weighted model, while the saliency-capability weighted model performs poorly. The results of the J-tests provide further support for my theoretical argument that bargaining capabilities at intergovernmental conferences (IGCs) are not related to a member state’s size, but rather to preferences and the equal rights afforded to all states to block change.

With respect to my treatment of intergovernmentalism, Franchino states that I have misunderstood the LI argument regarding bargaining power. I do not believe I have misunderstood LI, or its treatment of bargaining power; rather I have chosen to treat it as only one of a variety of forms of intergovernmentalism. My decision to refer to ‘intergovernmentalism’ – rather than ‘liberal intergovernmentalism’ – when discussing bargaining power was a very conscious one. Throughout the book I was careful not to cite Moravcsik as the sole proponent of an intergovernmental approach, citing in addition realist intergovernmental work by Grieco (1995) and others (see Slapin, 2011: 12, 14). Indeed, I use Bueno de Mesquita’s (1994) model to capture intergovernmentalism, not Moravcsik’s (see Slapin, 2011: 69). When I do refer specifically to LI, it is with respect to government preference formation and not bargaining (see Slapin, 2011: 15, 34–35).

Of course, my response may raise the question why I chose not to focus more attention on LI, given its prominence in the literature. There were several reasons. First, unlike LI, my goal was decidedly not to offer a new ‘grand theory’ of European integration. Rather, my goal was to treat IGCs as a forum in which member states make important decisions about institutional design. Given the institutional critiques of intergovernmentalism (e.g. Garrett and Tsebelis, 1996; Tsebelis and Garrett, 2001), I do not believe it is possible to offer a complete theory of EU integration that focuses solely (or even primarily) on IGCs and interstate bargaining (see Slapin, 2011: 1).

Second, as Franchino notes, Moravcsik relies on a similar conceptualization of bargaining power to that found in institutional studies. ¹ Both approaches suggest that power is related to member state preferences and, in the words of Moravcsik, ‘the relative value that it [each government] places on an agreement compared to the outcome of its best alternative policy’ (Moravcsik, 1998: 62). In fact, one could read my findings as a limited confirmation of some aspects of LI. After all, Moravcsik views credible threats to cast a veto as a potential source of power (Moravcsik,1998: 63). I, however, test an institutional model of bargaining power that is narrower than the LI model. The model I find best explains the data is one in which each member state has recourse to a veto, and each government’s willingness to cast its veto is a function of its preferences with regard to the treaty and the status quo. Moravcsik, on the other hand, argues that states have a variety of sources of power, including the ability to both exit and veto. How states use these sources of power depends upon the policy alternatives available to them, something I explore in a formal model in chapters 7 and 8. In Moravcsik’s model, some states may be made worse off than the status quo – they need only to be better off than their best alternative to the negotiated outcome (BANTA). My empirical findings imply that state size does not provide governments with more favourable outside options, or better BANTAs.

Third, my approach has been to examine all issues discussed at the IGC, rather than focus solely on substantive ones, such as monetary union or asylum policy. This is in contrast to Moravcsik, who, when examining the Treaty on European Union, focuses exclusively on monetary union, and not the major institutional innovations that the treaty produced, such as the codecision procedure (Moravcsik, 1998: chapter 6). Because of the outsized attention given to IGCs in intergovernmental ‘grand theories’, too little attention is given to the institutional issues that take up so much of governments’ time at these negotiations, such as the reweighting of Council votes, altering the number of Commissioners per member state, and increasing the powers of the European Parliament. By focusing on a non-random subset of issues, intergovernmental studies that ignore disputes over institutions may provide biased estimates of state bargaining power.

Lastly, in addition to examining all issues, I also examine the positions of all member states. Franchino states that a misunderstanding regarding Moravcsik’s conceptualization of bargaining power may arise from Moravcsik’s decision to focus his empirical analysis on the preferences of France, Germany, and the United Kingdom. Franchino writes, ‘short of an unlikely elementary mistake in research design, the focus on France, Germany, and the United Kingdom means that, for Moravcsik, the value of disagreement is predominantly determined by factors other than capability’ (Franchino, 2013: 324–337). I agree with Franchino to a point – the lack of variance in size across Moravcsik’s cases leaves us unable to say anything about the importance of size in bargaining. I disagree though, that Moravcsik’s choice of cases (focusing solely on large states) is innocuous. Rather, it affects our ability to make inferences about member state power with respect to size and all other variables that could affect bargaining power. By focusing on the preferences of large states, the empirical analysis implicitly assumes that small states exert no influence over the outcome. Imagine for the moment a hypothetical organization with three members, j ∈ ( 1 , 2 , 3 ) , who take decisions on i ∈ ( 1 , 2 , … , N ) independent issues. Each member has a position, x ij , on each issue. The outcome of negotiations on each issue, y i , is a function of the positions of the members. Thus, we can estimate the average ‘power’ or influence of the actors using a regression model (with the constant suppressed):

y i = x i 1 β 1 + x i 2 β 2 + x i 3 β 3 + e i

In this model, the β ’s represent each actor’s average influence over all issues. It is precisely the model I estimate in chapter 5. Now imagine that we fail to account for the positions of actor 3 in this model, and instead estimate the model omitting x 3 . To the extent that actor 3 exerts influence over the outcome, and has preferences correlated with actors 1 and 2, we now have omitted variable bias. We will attribute to actors 1 and 2 the influence of actor 3. Moreover, if actor 3 has influence for reasons that are different from actors 1 and 2, we would potentially misstate the reasons for the (now biased) estimates of influence of actors 1 and 2, as well. Without accounting for the preferences of all actors involved in a multilateral negotiation, it is difficult to say which actors have influence and why.

In sum, I hope my approach has advanced the study of intergovernmental negotiations by focusing on the preferences of all actors across all issues. Without doing so, it is difficult to assess the true nature of state bargaining power. I appreciate Franchino raising these points and I hope that I have been able to clarify any misunderstandings.