50 International Organization counterfactual leverage:What was the range of possibilities without these ideas?12 Cross-case comparisons help in this respect,suggesting alternatives in similar situations.But cross-national comparisons are rarely similar enough for analysts to ascribe all their variation to ideas.Critics may suspect that unnoticed differences in structural or institutional situations account for part of the divergence.3 Overall,laments one sympathetic reviewer,ideas in today's literature are"simply another rather than the causal factor."14 For skeptics,these"how much"questions make all the difference.If variation in ideas alone cannot be tied to specific variation in major outcomes,concludes a less sympathetic reviewer,ideas can still be downplayed as"a valuable supplement to [objective-]interest-based,rational actor models."15 One Way to Tie Ideas to Outcomes We need not conclude that ideas and clear causal arguments are incompatible,as a third reviewer implies.6 Settings where ideas cross-cut prevailing lines of organi- zation can clearly display their causal impact.French decision making in early European integration is one very significant example.While such cases cannot prove that ideas matter everywhere,they should convince skeptics that ideas can be major causes in politics.This section first considers how to isolate ideas'effects on individuals'views.Then it suggests how individual-level ideas can affect major outcomes. Cross-Cutting Ideas and Individual Variation Cross-cutting ideas have particularly clear effects because they offer the sharpest possible contrast to the expectations of objective-interest theories.All such theories define actors'interests in terms of objective conditions at some level of organiza- tion.For realists,individuals in a state share interests defined by the distribution of power;political party theorists trace members'interests to electoral or coalitional constraints;bureaucratic theorists highlight units'interests within organizational rivalries;and liberals or Marxists derive group interests from economic constraints (in different ways).Ideational approaches posit,in contrast,that actors interpret their interests through ideas that can vary independently from their objective positions. Their greatest divergence with objective-interest theories thus occurs where ideas vary as independently as possible from organizational lines in an arena,strongly 12.On counterfactuals and ideas,Blyth 1997,235;in general,Fearon 1991;and Tetlock and Belkin 1996. 13.Berman1998.11. 14.Blyth 1997,236.A reviewer of constructivism echoes that"norms are invoked as one of several causal variables with little or no insight given on how much of the outcome they explain."Checkel 1998, 339. 15.Jacobsen1995.,285 16.Yee1996.102.counterfactual leverage: What was the range of possibilities without these ideas?12 Cross-case comparisons help in this respect, suggesting alternatives in similar situations. But cross-national comparisons are rarely similar enough for analysts to ascribe all their variation to ideas. Critics may suspect that unnoticed differences in structural or institutional situations account for part of the divergence.13 Overall, laments one sympathetic reviewer, ideas in today’s literature are “simply another rather than the causal factor.”14 For skeptics, these “how much” questions make all the difference. If variation in ideas alone cannot be tied to specific variation in major outcomes, concludes a less sympathetic reviewer, ideas can still be downplayed as “a valuable supplement to [objective-] interest-based, rational actor models.”15 One Way to Tie Ideas to Outcomes We need not conclude that ideas and clear causal arguments are incompatible, as a third reviewer implies.16 Settings where ideas cross-cut prevailing lines of organi￾zation can clearly display their causal impact. French decision making in early European integration is one very significant example. While such cases cannot prove that ideas matter everywhere, they should convince skeptics that ideas can be major causes in politics. This section first considers how to isolate ideas’ effects on individuals’ views. Then it suggests how individual-level ideas can affect major outcomes. Cross-Cutting Ideas and Individual Variation Cross-cutting ideas have particularly clear effects because they offer the sharpest possible contrast to the expectations of objective-interest theories. All such theories define actors’ interests in terms of objective conditions at some level of organiza￾tion. For realists, individuals in a state share interests defined by the distribution of power; political party theorists trace members’ interests to electoral or coalitional constraints; bureaucratic theorists highlight units’ interests within organizational rivalries; and liberals or Marxists derive group interests from economic constraints (in different ways). Ideational approaches posit, in contrast, that actors interpret their interests through ideas that can vary independently from their objective positions. Their greatest divergence with objective-interest theories thus occurs where ideas vary as independently as possible from organizational lines in an arena, strongly 12. On counterfactuals and ideas, Blyth 1997, 235; in general, Fearon 1991; and Tetlock and Belkin 1996. 13. Berman 1998, 11. 14. Blyth 1997, 236. A reviewer of constructivism echoes that “norms are invoked as one of several causal variables with little or no insight given on how much of the outcome they explain.” Checkel 1998, 339. 15. Jacobsen 1995, 285. 16. Yee 1996, 102. 50 International Organization