Table I

Preparation, presidencies, and the efficiency of six IGCs

Case		Preparation	Presidencies		Efficiency
1.	IGC of 1985	Effective	Luxembourg:	Effective	High
2.	IGC on EMU (1991)	Effective	Luxembourg: Netherlands:	Effective Effective	High
3.	IGC on Political Union (1991)	Ineffective	Luxembourg: Netherlands:	Effective Medium	Medium
4.	IGC of 1996-97	Effective	Italy: Ireland: Netherlands:	Ineffective Medium Ineffective	Medium
5.	IGC of 2000	Ineffective	Portugal: French:	Medium Ineffective	Low
6.	IGC of 2003-04	Medium	Italy: Ireland:	Medium Effective	High (but only in a second attempt)

Explanation: We used three categories for our classification of the preparation and the presidencies: effective, medium, and ineffective. For the empirical indicators used, see table 2. Our theoretical argument made us not consider the presidencies during the preparatory phases of the IGCs.

Table II

Operationalisation of the variables

Variable	Cause of Variation	Empirical Indicator
Effectiveness of the preparation	Time-span available for preparation Preparatory group's willingness to abstain from distributional bargaining	Does the final document put a large number of issues on the negotiating table, and does it outline different options for how to achieve a compromise?
Effectiveness of the presidency	Preference of state holding presidency as compared to the preferences of other participants (outlier or not?) Exogenous events (elections, crises) that reduce the capacity of the presidency to guide the negotiations	Does the presidency build on the work done during the preparatory phase or by previous presidencies? Does the presidency try to exclude specific issues from the negotiations? How do participants in the negotiations evaluate the work of the presidency?