Journal of European Public Policy 6:1 March 1999: 155Ð 79

Review section symposium: The

choice for Europe: Social purpose
and state power from Messina to
Maastricht
Helen Wallace
James A. Caporaso
Fritz W. Schampf
Andrew Moravcsik

Andrew Moravcsik (1998) The Choice for Europe: Social Purpose and State Power
from Messina to Maastricht (Ithaca: Cornell University Press; published in the
UK by UCL Press, London, December 1998. ISBN 1 85728 191 8 (hb) £40.00.
ISBN 1 85728 192 6 (pb) £14.95.

PIECING THE INTEGRATION JIGSAW TOGETHER

Helen Wallace
Andrew Moravcsik’s long-awaited volume assembles a huge jigsaw puzzle. The
sharply defined picture of west European integration reveals calculating statecraft
by highly strategic key governments to promote the core economic interests of
Britain, France and Germany. The picture, drawn on a large canvas, spreads from
the early years of the European Economic Community to the Treaty on European
Union in Maastricht. No concessions are made to pointillisme or impressionism in
the execution of this picture, with its careful observation and enriching detail,
combined with a grand sweep of unified design.
The result is an unusually wide-ranging analysis of five history-making episodes
in the evolution of what is now the European Union (EU). The breadth, depth and
variety of primary sources underpin a compelling interpretation of the underlying
dynamics of west European integration. This jigsaw is composed of many
thousands of pieces, painstakingly sorted – and re-sorted – into a coherent whole.
The training in history shows in the emphasis on substantive evidence, while the
commitment to robust theorizing about the international political economy
permeates the flow of the analysis. The central objective is to contest the sweeping,
soft and often sloppy assertions about European integration as either a special form
of supranationalism or driven by ‘geopolitical ideology’. On the contrary, the
process and the outcomes are much more ‘normal’ (p. 5), in that the representatives
of the leading states have for fifty years behaved logically in using the EU to

Copyright © Routledge 1999 1350–1763

156 Journal of European Public Policy

promote their economic interests. The volume constitutes a formidable statement

of liberal intergovernmentalism, which Moravcsik has, himself, done so much to
develop. It builds on historical argument – Moravcsik has a great deal in common
with Milward as a debunker of carelessly received wisdom. It contributes to
international relations theory, by removing the study of west European integration
from its idiosyncratic corner and locating it in the mainstream of theorizing about
the relationship between the state and the international system.
Much hangs on the attempt to disprove two alternative interpretations of
European integration. One is the view that geopolitical ideology explains the process.
The other is that some form of supranational entrepreneurship or engrenage has been
at work, whether as defined by neofunctionalism or as reinterpreted in versions of
historical institutionalism. Moravcsik produces vigorous arguments against, and
detailed rebuttals of, both camps. He is, for this reviewer at least, more focused and
persuasive in his attacks on historical institutionalism than in his onslaught on
geopolitics. His targets are clearer in the former case, a lively and productive arena
of academic contention. Somehow the rebuttal of geopolitics is more contrived, in
aiming at something of a straw man; few authors rely on geopolitical explanation
alone. To recognize the primacy of political economy factors need not require us to
throw all geopolitical considerations out of the window. Indeed, as Moravcsik
himself honestly acknowledges (p. 478), in Germany, a non-trivial example,
geopolitical factors seem periodically to have been quite influential. An analysis that
weighted, rather than ‘exceptionalized’, the geopolitical component in relation to
political economy considerations might have served better.
Thus, in spite of the vigour and richness of the argument, the volume presents a
picture, not a photograph. The eye of the artist is selective, capturing some parts of
the process more thoroughly than others; and it slides over some of the elements
that displease the eye. Or, to put it another way, this is one of those irritating jigsaws
(like those hideous ones of only baked beans) where the pieces can be put together
in more than one combination; or, even more irritating, a jigsaw with similar, but not
identical, pictures on each side. And not quite all the pieces of the jigsaw are on the
table.

SOME MISSING PIECES OF THE JIGSAW

One obvious criticism of the volume arises from its tight focus on British, French
and German policies. Their relative importance is undeniable in the history of the
EU. None the less, other member countries, albeit much less powerful, have played
some part in the intriguing form of complex multilateralism that comprises the EU.
The pattern of coalition-building that underpinned strategic bargains and which
linked the three to the rest might have merited more explicit attention.
A second underdeveloped element lies in the description of how state
preferences are formed, a process at the heart of Moravcsik’s analysis. The argument
and evidence concentrate on the behaviour of incumbent governmental élites in the
three countries and their symbiosis with the interests articulated by leading
economic agents. This begs some questions about the nature of politics within the
 H. Wallace: Piecing the integration jigsaw together 157

three countries, about the capacity of incumbent governments to call the shots and
to act as gatekeepers, and about the primacy of particular economic agents. I would
have preferred a more finely grained assessment of domestic politics. The British
story might, for example, have mentioned the weakening influence of peak
associations and the privileged relations, during the 1980s especially, of a partisan
government with particular firms in the business and financial sectors. More
attention could have been paid to the deep political controversies in Britain about
the goals and implications of European integration, these being only partly
contested over political economy preferences and often over other elements of raw
politics. The ousting of Mrs Thatcher from office barely rates a mention in the
account of the period leading up to Maastricht. British preferences as regards
integration have not been quite so consistent and stable as Moravcsik would have us
believe. In France and Germany too those engaged in the wider political arena have
not always been quite so convinced that their incumbent governments were making
wise policy choices about when and whether to delegate to EU institutions.
A third dog that does not really bark is the voice of comparative politics. The
study somewhat too easily identifies similarities in the approaches of the three
countries, arguing that differences on particular issues (agriculture, liberalized
internal and external trade, European currency, and so forth) are in each case related
to consistent differences of economic interest. Surely there is more to it than that. I
am myself quite nervous about the assumption so widely found in the available
comparative assessments of individual EU member states that there is a more or less
standard model for evaluating national responses to integration. Some glimmerings
of this concern can be founded in Moravcsik’s admission that in some respects the
German picture is at variance with the French and British, in embracing more
diverse elements of state policy.
A fourth lacuna has to do with absent pieces of evidence. Moravcsik has
assembled so much data that it becomes hard to believe that in his 501 pages he
might have left the odd stone or two unturned. However, the long-running
argument about the British position vis-à-vis the Community budget comes out in
rather a muddle. It was incumbent British politicians who chose (rightly or
wrongly) not to press the budget issue in the accession negotiations, and their
successors who in 1974 accepted a Dublin mechanism that manifestly did not solve
the problem of budgetary inequity. Rational and well-judged statecraft could surely
have produced a more nuanced and possibly more effective negotiating strategy.
Indeed the long shadow cast by the subsequent debilitating arguments surely
merited more weight in the analysis. One footnote here – the Commission had in
1970 recognized this and of its own initiative calculated alternative figures and
sketched a formula that might have avoided the subsequent rows. Why these figures
were never fed into the accession negotiation, and how it was that the Commission
a few years later categorically blocked as not allowable under EC law a mechanism
that was close to that eventually agreed in Fontainebleau in 1984 are episodes that
deserve scrutiny to identify Commission influence on these negotiations between
governments. The close relationship between the Commission and the British in
devising the European Regional Development Fund (ERDF) in 1973/5, not least
158 Journal of European Public Policy

the famous ‘Thomson map’ (devised with the lubrication of a good whisky), had
more of a role than Moravcsik admits, so concerned is he with linking the ERDF
story to the discussion of economic and moneary union and the European
monetary system. There is a confusion here of time, events and causality. One
confusion is the reference (pp. 350–2) to the eventual British budget solution as a
form of ‘generalized juste retour’. If only it had been that, the EU would not now be
facing such a potentially bruising argument over net national contributions.

MONOCHROME OR POLYCHROME?
Moravcsik’s picture is drawn parsimoniously in black and white only. I agree with
the thrust of his argument that domestic politics and domestically sourced prefer-
ences explain a great deal about both the delegation of policy powers to the EU and
refusals to delegate. Yet to rely quite so heavily on incumbent political élites’ own
evidence on their goals and behaviour is to build in something of a distortion. Too
much is drawn from the memoirs and witness of these protagonists for my taste.
Too little account is taken of the self-justification and self-rationalization of such
accounts. Few national politicians or officials are likely to credit the Commission
with critical influence or to admit to uncertainty or inconsistency.
Moreover, the asserted hierarchy of decision in three stages – preference
formation, negotiation and institutional choice – is too neat and tidy. Of course,
actor-centred rationality is more important and more amenable to systematic
analysis than many students of European integration have acknowledged. But the
implied control of process and outcomes needs more justification than Moravcsik
adduces. Space needs to be made for irrationality, for confusion and for mistaken
judgements (vide here the management of Britain’s policy over the Ioannina
compromise in 1994, or the infamous ‘non-cooperation’ policy prompted by BSE
in 1996, two events outside the scope of this volume). In brief, a little more nuance,
more varied colours, and a little less emphasis on successful rationality would make
the core of the analysis more persuasive.

THE PICTURE ON THE OTHER SIDE OF THE JIGSAW

Quibbles apart, does Moravcsik’s overall argument stand up and has he proved his
case beyond reasonable doubt? This volume is an extraordinarily powerful and
well-substantiated statement of the case for bringing and keeping the state at the
centre of the analysis of European integration. But methinks he doth protest too
much about the limited and mostly uninfluential role of supranational policy entre-
preneurs and transnational coalitions, whether of political or economic agents. My
own view is not that supranational entrepreneurs can control or direct the process,
but that in some circumstances they can be – and sometimes have been – important
catalysts of plausible ideas and brokered agreements. It is simply not the case that in
the Council, the Committee of Permanent Representatives (COREPER) or the
Special Committee on Agriculture the Commission is uninvolved (pp. 224–5, 306).
Indeed, all of us who at one stage or another in our careers have played committee
 H. Wallace: Piecing the integration jigsaw together 159

politics know that often getting the conclusion you want is more important than
winning credit for its adoption.
The Thomson map for the distribution of ERDF funds is one example. The
never-publicized meeting, convened in the Ardennes by the Commission in
December 1983, of British, French and German officials to stitch a deal on mutual
recognition is another example. The ability now and again of Commission officials
to feed into the Council presidency usefully defining texts for others to claim
parentage, even on critical and contentious issues, is not to be discounted so easily.
Or, a different kind of illustration, the skill of a behind-the-scenes coalition of big
employers to get their text on pensions and the Barber judgment adopted in the
Maastricht IGC suggests that forces other than statecraft are sometimes at work.
Sure, there are principal–agent relationships in play, but the agency function is
sometimes more than responsive.
Two other points might emerge from looking at the picture on the other side of
the puzzle. One is that the institutional setting of the EU has interesting impacts on
the behaviour of state actors, both constraining their behaviour and providing
opportunity structures. In May 1982, when the then British minister sought to
invoke the Luxembourg Compromise on an issue of agricultural prices, there was a
pause before the result was clear. The Danish and Greek ministers followed the
presumed code in supporting the British in principle, as their routine instructions
required. The French, omitted on this point from Moravcsik’s account (p. 350),
sought instead clarifying instructions from Paris; to the surprise of the French on
the spot the instruction was, for the first time, to block the application of the
formula. The consequential procedural uncertainty was to deter British negotiators
from invoking the Luxembourg Compromise even on more serious issues.
The second point has to do with the role of ideology or doctrine. Over the period
of EU evolution more of a tussle has occurred between competing economic doc-
trines than Moravcsik acknowledges. In their turn the common agricultural policy
(CAP), the single market and the single currency have been achieved as a particular
political-economic doctrine gained a kind of transnational ascendancy at a partic-
ular moment. Whether or not each winning idea necessarily benefited the economies
of the countries whose governments accepted them is a matter of judgement or
interpretation, not a matter of clear determination. There were after all always
critics in France of the CAP outcomes sought – and won – by French ministers.
These criticisms notwithstanding, Moravcsik’s careful analysis provides formid-
able arguments against those who portray European integration as necessarily a
cumulative and irreversible process. The deliberative and calculated strategies of
leading negotiators can just as well be directed at denying or unravelling supra-
national agreements as at consolidating collective arrangements. The pendulum can
indeed swing back towards the individual countries when the costs of co-operation
seem too high or the rewards for commitment too uncertain. Credence for this
interpretation can be found in recent and current attempts, not least from German
negotiators in Amsterdam, to ‘repatriate’ policy powers (notably in competition
policy) or to entrench ‘subsidiarity’. Among both practitioners and commentators
the arguments will continue, much fortified by this stimulating study.
160 Journal of European Public Policy

TOWARD A NORMAL SCIENCE OF REGIONAL INTEGRATION

James A. Caporaso
In The Choice for Europe: Social Purpose and State Power from Messina to
Maastricht, Andrew Moravcsik provides a new synthesis for the field of regional
integration studies generally and European Union (EU) studies in particular. This
major new study combines three theoretical strands with five important,
history-making case studies (the Rome Treaty, the common agricultural policy, the
European monetary system, the Single European Act (SEA), and European
monetary union). For all five cases, Moravcsik provides an account of how the
preferences of national politicians are formed, drawing on and extending liberal
theory. Second, he utilizes negotiation theory to describe and analyse efficient as
well as distributive bargaining. And, third, he creatively applies institutional theory
to help us understand how bargains are enforced and maintained. Despite great
complexity in the realm of the evidence, Moravcsik leaves us with a central message.
There is a master variable at each of the three stages. He states the conclusion boldly
and efficiently: ‘I argue that a tripartite explanation of integration – economic
interest, relative power, credible commitments – accounts for the form, substance,
and timing of major steps toward European integration’ (p. 4).
This is not a harmless book. Many idols are smashed. Neofunctionalism suffers
regarding the autonomy, influence, and agenda-setting power of international
entrepreneurs. Functionalists will have to concede that transnational society
(Monnet’s unite des faits) does not have the influence that many once thought.
Realists will have to give ground regarding the systemic determination of
preferences (security externalities explain little). Indeed, Moravcsik argues and
adduces strong evidence to show that national interests are produced through a
national political process, albeit a process that operates on domestic and
transnational economic interests. Regulatory theorists will have to recognize that
the EU is decidedly not ‘government by information’, by experts, by scientists, or
by task-oriented specialists, except to the extent that experts are carefully chartered
by governments to carry out specific tasks. There are quite simply too many hard
distributive bargains in the EU’s history to come to the conclusion that the EU is
solely a collective gains project. For functionalists, the good news is that the EU is
for real, i.e. it counts in a broad variety of ways. The EU is not limited to the
management of market imperfections at the regional level. The bad news is that the
process is controlled by state politicians.
The strengths of this book are easy to discuss and should not be dismissed too
quickly in the interest of getting to the critical comments. Since the book has been a
long time coming, the fear was that it would be a patchwork, reflecting incremental
adaptation to critics along the way, but diluting itself in the process. What we have
instead is a major synthesis, an integration of three major lines of theoretical
development in comparative and international politics. To write a book such as The
Choice for Europe requires the analyst to embrace three daunting questions. First,
why do state leaders have the preferences they have? A corollary but still important
question has to do with whether central decision-maker preferences are the most
 J.A. Caporaso: Toward a normal science of regional integration 161

important, in contrast, for example, to the preferences of transnational societal

actors. Second, once preferences have been accounted for, how do they (along with
other factors such as power) translate into negotiated outcomes? Three, how do
leaders get the results of negotiation to stick?
The first question, in some ways the most difficult, is tackled through a
sophisticated application of liberal theory, which Moravcsik himself has done a lot
to develop in this book and elsewhere. The second question is the province of
negotiation theory, a well-developed literature in its own right. Moravcsik skilfully
uses this theory and applies it to his cases of non-coercive bargaining. Finally, the
third question leads us to institutional theory, particularly the theory of credible
commitments. The author takes us to the frontier in each of these bodies of theory,
isolates what is essential for his story, and skilfully applies all three.
The second strength of the book is that the scholarship, i.e. the empirical
research, is extraordinarily good. Moravcsik has carefully investigated each of his
five cases and has judiciously weighed and sifted the evidence. He spends a
considerable amount of time informing the reader about his research procedures,
lending an air of transparency and reproducibility to how he arrives at his
conclusions. He relies on statistical evidence when appropriate, archival material
when available, in-depth interviews, the documentary record, biographies, diaries,
newspaper accounts, and secondary literature. The case studies cover a wide range,
from the formation of the European Economic Community (EEC) to the Treaty on
European Union. The important eleven-year period after the EEC came into
existence is subjected to a close scrutiny. I am impressed at the depth, thoroughness,
and creativity of the scholarship, all the more so because it is not done in the spirit of
‘all the facts we can get’. Quite the contrary, the evidence provided is implied by the
theories, sometimes distinctively so. I am trying to recall the last time I read a book
in the social sciences where the author told me, in great detail, how he would treat
the secondary literature in terms of its primary evidence, conclusions, and
theoretical claims.
A third strength is that The Choice for Europe mainstreams EU studies and
regional integration theory. Many scholars consider the EU to be quite distinct and
therefore that it calls forth its own genre of research. It is this assumption, perhaps
more than any other, that has retarded the growth of cumulative knowledge in this
area. Indeed, in descriptive terms, the EU is quite different from other regional
associations (it is larger, more institutionally developed, has a greater share of
intra-regional trade). However, in Moravcsik’s hands, these descriptive differences
(across sectors, countries, time periods) are simply the raw material for a general
theory. This book helps to locate integration studies within the overall body of
knowledge of international relations, comparative politics, and political science,
and takes EU studies out of its self-constructed theoretical ghetto. The result may
well be that students of integration will have to take more seriously the professional
literatures of international relations, comparative politics, and political economy.
Standards that apply in other sub-fields, for example, with regard to research design,
data collection, and analysis, are more likely to extend to regional integration
studies also.
162 Journal of European Public Policy

The Choice for Europe also prompts a number of critical comments and
questions. Let me raise the first point as a question. Do domestic institutions
matter? Do differences in institutions among the member states affect the co-
operative outcomes and choices analysed? In the abstract, one would think the
answer is ‘yes’. In a way, the field of comparative politics depends on a positive
answer. The most important thing about domestic politics is how a society
constructs its institutions and the way these institutions are built is thought to affect
political outcomes. Yet Moravcsik’s underlying model is one of societal pluralism,
particularly economic pluralism. Economic factors, particularly the distribution of
commercial and productive interests, work their way up toward the national
leaders, who in turn carry these preferences into international negotiations. The
results of the negotiations are then locked in by institutions. Presumably the story
of regional co-operation can be told with a universal Olsonian pluralism (economic
interests plus a logic of collective action), a negotiation analysis driven by the
relative cost of non-agreement to the parties, and institutional lock-in. This is the
essential story, despite the episodic causality of geopolitical factors, ideology, and
the autonomous force of supranational technocrats.
If Moravcsik’s basic model is correct, should we draw the implication that
differences in domestic institutions do not matter for either the preference
formation or negotiation phase of his study? Admittedly, this would be a tough pill
to swallow for many comparativists who are inclined to think that differences in the
organization of interest groups (pluralist vs. corporatist), political parties (two party
vs. multiparty), and executive–legislative relations (parliamentary vs. presidential)
make a difference. I would like to have seen this issue addressed in systematic
fashion, if not explicitly modelled. I, of course, do not have evidence that the results
would be changed, but there are good theoretical reasons for thinking that variables
such as the number of veto points, the distribution of domestic political power, and
the degree of divided government count.
A second point concerns the logic of case selection and how this bears on the test
of rival theories. Moravcsik has chosen to focus on five very important,
history-making cases. While this seems reasonable (why not look at the most
dramatic cases?), how does it affect the functionalist focus on the repeated, normal,
and mundane? We can ask pointedly ‘what is a case?’ Moravcsik has a meta-theory
of cases and case selection. For him, a case is a relatively discrete, bounded event or
set of events. While his cases are protracted and have considerable temporal range,
they are distinctively not what we would call processes. I do not think that the
functional alternative would imply cases at all, in the sense used here, of instances
where a problem exists, there is serious conflict among the relevant parties, and
negotiations to solve problems result in some policy or explicit non-agreement. For
evidence, a functionalist would look at the slow accretions in domestic and
transnational society resulting from trade, capital flows, movements of workers,
capitalists, and tourists, cross-border activities of professional organizations, and so
on. Moravcsik is perfectly aware of these methodological differences (indeed, many
of the factors I mention are incorporated into the liberal side of his theory) but the
point still remains: the selection of cases makes it easier to confirm the
 J.A. Caporaso: Toward a normal science of regional integration 163

intergovernmentalist story and implies less ability to probe the merits of

process-based theories. The difference between intergovernmentalism and
functionalism is somewhat redolent of the differences between pluralists and
structuralists in the much older community power structure debate, with
Moravcsik standing in as the modern Robert Dahl but no one to represent Floyd
Hunter and the sociologists.
Two criteria of case selection are internal and external validity. External validity
has to do with generalization, i.e. with the extension of research results to data
realms outside the ones directly studied. Large-N studies are best equipped to do
this. Internal validity has to do with the nature of causal inference. While not
completely separate from external validity, it has its own logic. It asks ‘how do we
know it is X and not Y that is causing Z?’ Since Moravcsik is doing a small-N study,
what is the best way to sample cases, given his particular meta-conception of cases?
All five cases represent major choices for members of the EU. With the partial
exception of the single market, I believe all the cases represent a mix of gains and
losses. It is not clear that these cases can be ordered on any underlying continuum so
it is also not clear that we should (again with the exception of the SEA) expect
differences in outcome. While the results of the case study analyses are therefore
descriptively robust (economic factors and government preferences are important),
variation in cases does not lead us to expect differences. If the cases had been ordered
along several dimensions (say, more or less controversial, or from collective gain to
hard distributive gains and losses, or different types of policy sector), comparison of
the results across cases would have given us more leverage.
Third, the theoretical formulation employed in this book obscures one of the
most important theoretical divides in integration theory, that between neo-
functionalism and intergovernmentalism. While intergovernmentalism obviously
owes a debt to realism, Moravcsik divests himself of realism’s albatross of systemic
determination of preferences and paves the way for a more modest ‘domestic
realism’ in which government leaders take on the preferences of key domestic
groups. The predictions of intergovernmentalism are less bold but more accurate.
The move with regard to neofunctionalism is more controversial. For interpretive
purposes, Moravcsik retains only one of two legs of the neofunctionalist approach,
the importance of international technocracy and international institutions. He
definitionally cuts away the importance of transnational society, severing the link
between society and supranational institutions, and reattaches it to liberalism,
admittedly in a much more sophisticated form. Neofunctionalism is left standing on
one leg (international technocrats without transnational society) while
intergovernmentalism acquires a new ally, domestic and transnational society. We
have a much more refined analysis of the social, more in tune with modern theories
of preferences, but neofunctionalism is impoverished.
Perhaps I should not be disappointed at the failure of an ultimate showdown
between intergovernmentalism and neofunctionalism. I must admit the wish is
partly for nostalgic reasons in any case. Moravcsik’s new synthesis does a better job
of theorizing the importance of the economic and the social and he has provided us
with a richer and more theoretically productive formulation. His synthesis is not a
164 Journal of European Public Policy

casual eclecticism but rather a reconceptualization of the older terms of debate in

which he upgrades previous insights in light of modern theory, and places regional
integration studies, of which the EU is a part, within the fields of comparative and
international political economy. It will be hard to please both EU specialists and
theoretically minded students of comparative politics and international relations.
Indeed, Moravcsik intends to change our minds about the way many of us look at
the relations among these fields. We should read the book carefully and grapple with
the issues raised by it. It may be the most significant book on the subject since Haas’s
The Uniting of Europe.

SELECTING CASES AND TESTING HYPOTHESES

Fritz W. Scharpf
This book has been long in coming, after so many previews, colloquia, seminar
presentations, and papers based on the underlying research. Now that it is finally
here, 500 pages of text, there is a great sense of satisfaction in seeing so much of this
material integrated into a well-organized and well-written opus that backs up
earlier arguments by impressively researched historical case studies of the
substantive preferences, bargaining strategies and institutional arrangements
pursued by France, Germany and the United Kingdom in five ‘constitutional’
negotiations that have shaped the process of European integration. Somehow,
however, the excitement that one had been prepared for isn’t quite there – perhaps
because some of the most controversial and most interesting propositions of these
earlier papers were simply left out here.1 But this in no way detracts from the value
of this important book.
Moravcsik presents the most complete, theoretically disciplined and
methodologically self-conscious historical account yet available of the antecedent
conditions, bargaining processes and outcomes of five intergovernmental
negotiations that have shaped European integration – the original Treaties of Rome,
the settlement of fundamental disputes preceding the establishment of the common
agricultural policy in the 1960s, negotiations about the co-ordination of monetary
policy in the ‘Snake’ and the European monetary system in the 1970s, the Single
European Act in the 1980s, and the Maastricht Treaty and European monetary
union in the 1990s. Even more remarkably, these case studies, each of them
amounting to about seventy pages, are used to ‘test’ a small set of competing
hypotheses, all with respectable pedigrees in the theoretical literature on European
integration. Moreover, the historical descriptions are extremely well documented,
with an emphasis on primary sources and a critical awareness of the limitations of
secondary sources.
The hypotheses derive their salience from highly generalized, theory-oriented
controversies within the subdiscipline of international relations: Who are the
relevant actors, supranational organizations or national governments? What shapes
their preferences, geopolitical or economic interests? What explains the outcomes of
distributional conflicts, international institutions or the relative bargaining power
 F.W. Scharpf: Selecting cases and testing hypotheses 165

of the states involved? What explains the delegation of decision-making authority

to supranational institutions, federalist ideology, the superior information
efficiency of international organizations, or the interest of governments in assuring
the credibility of mutual commitments?
As the author repeatedly reminds the reader, it is one of the strengths of this book
that it explicitly specifies and tests competing hypotheses. It must also be said that,
although there is never any question of the author’s preference for one answer to
each of his questions (always the last-mentioned one in the above listing), all the
hypotheses are fully elaborated and presented with great care to suggest their
prima-facie plausibility. The same is true of Moravcsik’s treatment of the evidence in
the case studies. Their length is mainly owing to the fact that he bends over
backwards in searching for, and then disproving, suggestions that might favour one
of the hypotheses that he is determined to reject. There is never any doubt,
nevertheless, that in the end they will be rejected; in fact, the book reads very much
like an ex-parte brief against most of the existing literature.
But that is not meant to suggest that Moravcsik’s conclusions are
unconvincing. If anything, one might say that – given his selection of cases –
most of his preferred hypotheses have such a high degree of a priori plausibility
that it seems hard to take their competitors quite as seriously as he does. Since
only intergovernmental negotiations are being considered, why shouldn’t the
preferences of national governments have shaped the outcomes? Since all case
studies have issues of economic integration as their focus, why shouldn’t
economic concerns have shaped the negotiating positions of governments? And
since only decisions requiring unanimous agreement are being analysed, why
shouldn’t the outcomes be affected by the relative bargaining powers of the
governments involved?
The only one of Moravcsik’s hypotheses with which I am tempted to take issue is
the last one regarding the choice of an implementation mechanism for agreements
reached through intergovernmental bargaining – the choice being between
self-implementation by the negotiating governments (requiring the resolution of
disputes through further negotiations and unanimous decisions in the Council), or
a ‘pooling of sovereignties’ (allowing future decisions to be reached by qualified
majority vote in the Council), or finally the ‘delegation’ of decision powers to the
Commission and the European Court of Justice. With regard to the last two
options, Moravcsik claims that the choice is explained by a concern for the
credibility of commitments in the face of ubiquitous temptations to defect. That
explanation presupposes that the problems governments face in the
implementation stage are in the nature of a (perhaps asymmetrical) Prisoner’s
Dilemma constellation. That is plausible enough for some issues – mainly in the
fields of ‘negative integration’, competition policy and government subsidies,
where the common interest in having open and undistorted markets needs to be
safeguarded at the implementation stage against the ubiquitous protectionist
temptations of individual member states.
But delegation seems also a plausible device in policy areas in which the
temptation to defect plays no role at all. This is true of a wide range of non-tariff
166 Journal of European Public Policy

barriers that were to be removed in consequence of the Single European Act. In this
context, the harmonization of product standards was delegated, on the basis of
rudimentary specifications by the Council, either to the Commission and its
elaborate ‘Comitologie’ (that is not discussed at all in the book) or to
non-governmental standard-setting bodies. Here the constellations of national
interest resemble the battle of the sexes, rather than the Prisoner’s Dilemma –
implying that it is in everybody’s interest to apply a common standard once it is
agreed upon. Thus the justification for delegation is not the fear of defection, but the
high transaction costs associated with intergovernmental negotiations in a situation
where countries differ in their preferences for specific definitions of a common
standard. That justification, however, seems relatively close to the ‘information
efficiency’ hypothesis that Moravcsik rejects.
The reason, I suggest, is not an insufficient understanding of game-theoretic
distinctions. While Moravcsik does not aspire to membership in the ‘analytical
narratives’ club, he has a good intuitive grasp of analytical approaches, and he is
clearly much too sophisticated simply to assume that all the world is a Prisoner’s
Dilemma. Rather, the reason seems to be an overly strong commitment to the idea
that the social sciences should produce general theories about empirical regularities.
In this quest, theories that apply without exception are to be preferred to
‘sometimes true theories’ whose conclusions depend on contingent antecedent
conditions; and studies that are able to demonstrate a universal regularity are
considered more successful than studies producing explanations of significant
differences among superficially similar cases. In general, I see little point in
criticizing such methodological predilections (which may also be reinforced by
membership of a particularly theory-happy sub-discipline of political science) –
what matters is what we can learn from a study. But since Moravcsik pushes his
methodological views with so much indignation against students of European
integration who have followed different drummers, it is perhaps appropriate to
discuss the price he is in fact paying for his choices.
First, the preoccupation with general theory tends to impede theoretically
interesting discoveries. Deviant cases are in danger of being ignored or downplayed
as ‘exceptions’, rather than being exploited for the development of more
reality-congruent theory through the introduction of analytically pertinent
distinctions. The example discussed above points to such opportunity costs. Others
include the under-theorized discussions of the ‘exceptionally’ strong role of the
Commission in defining the agenda for the Single European Act, or of the
institutionalized capacity of the Bundesbank to influence the negotiating positions
of the German government in the 1970s and the 1990s in ways that were at odds
with the interests of industrial producers (which according to Moravcsik should
‘generally’ control government preferences).
More subtle but equally costly is the influence which the universalistic
aspirations seem to have on case selection. In the ongoing debate between
‘intergovernmentalist’ and ‘supranationalist’ interpretations of European policy
choices, most authors (including Moravcsik in some of his earlier contributions)
have focused on a wide range of decisions, some of ‘constitutional’ significance,
 F.W. Scharpf: Selecting cases and testing hypotheses 167

some of a more ‘everyday’ variety. The attempt to settle these controversies with a
definitive study that focuses only on decisions of constitutional significance may
appear a bit tricky unless care is taken to theorize the distinction between the two
classes of decisions, and the reasons why and how that distinction should make a
difference. Moreover, the focus of the present book is restricted to those
constitutional decisions that have been reached by way of intergovernmental
negotiations, while constitutional changes reached through Commission strategies
and Court decisions (which have constitutional force when they interpret the
Treaty) are completely excluded from consideration. Thus, we do learn that
demands for the liberalization of transport policy ran into ‘complete deadlock’ in
the negotiations of the 1960s (p. 217), and that France and Germany remained
opposed to it in later rounds of Treaty negotiations as well – but we learn nothing
about the ‘supranational’ processes and mechanisms through which air, road and
rail transport (or telecommunications, postal services and electric power, for that
matter) were nevertheless brought under the rule of European competition law in
the late 1980s and the 1990s. In other words, the generality of confirmed hypotheses
is in part bought by the exclusion of ‘deviant’ cases – which would, of course, have
to be accounted for in a fully specified general theory.
I would not have dwelled on these points if Moravcsik did not insist so much
on the superiority and uniqueness of his own approach.2 In any case, they
should not overshadow my conviction that this is a very good book. Its
empirical chapters are so rich in well-documented detail and insightful
interpretation that their intrinsic value far exceeds the function they have in
supporting or disconfirming the few and somewhat thin theoretical hypotheses
they are supposed to test. Moreover, both the introduction and the first
theoretical chapter contain a wealth of theoretical and methodological insights
that have, in fact, not yet been grasped by many students of European
integration. To mention just two: it is indeed true that the causal influence of an
actor, say, the Commission, on negotiating outcomes cannot be inferred from
the observable activities of that actor, but must be assessed in the context of an
explicit bargaining theory that takes account of the structure of asymmetrical
interdependencies among the actors involved (pp. 9, 53–4). Even more
important is the clarity with which Moravcsik has defined the concept of
preferences as ‘an ordered and weighted set of values placed on future
substantive outcomes’ (p. 24). The concept, in other words, is meant to describe
relatively stable criteria of evaluation, rather than the contingent choices among
alternative courses of action that may be more or less useful for reaching
preferred outcomes. If this were generally understood and accepted, many
confusing controversies over the need to ‘endogenize preferences’ in rational
choice approaches and to allow for the possibility of ‘policy learning’ could be
avoided.
But these are merely appetizers – there is so much more food for thought in this
book, as well as an immense wealth of historical information, that no serious
student of European integration can afford to ignore it. Highly recommended.
168 Journal of European Public Policy

NOTES
1 The earlier article arguing that ‘the European Community strengthens the state’
(meaning the national executive) is mentioned in passing in a footnote (p. 76), but
its thesis does not fit well with the present conceptualization according to which
‘“State” actors are treated as proxies for the underlying social forces’ (p. 36,
footnote 29).
2 For instance, it is surely true that, for identifying the preferences of actors, their strategic
calculations, and the information at their disposal, primary sources are to be preferred
over secondary sources, and hard primary sources over soft ones (pp. 80–2). But if
reliance on hard primary sources is illustrated by the fact that the author himself
conducted over a hundred interviews, it is not clear that his procedures differ so much
from the practices of most colleagues who are doing serious empirical work in
comparative politics.

THE CHOICE FOR EUROPE: CURRENT COMMENTARY AND

FUTURE RESEARCH: A RESPONSE TO JAMES CAPORASO,
FRITZ SCHARPF, AND HELEN WALLACE
Andrew Moravcsik
It is a pleasure and a privilege to read and respond to three such distinguished
and varied critics as James Caporaso, Fritz Scharpf, and Helen Wallace.1 Trying
to do justice to their disparate critiques recapitulates the tensions we all face in
conducting basic research on European integration. Relevant scholarship spans
the disciplines of history, economics, legal theory, sociology, and political
science – and, within the latter, the sub-disciplines of international relations,
comparative politics, and policy analysis. No surprise, then, that each
commentator, representing a distinct area of political science, situates the major
theoretical conclusions of The Choice for Europe differently. Caporaso treats
them as audacious attempts to smash long-standing sub-disciplinary idols.
Scharpf views them as straightforward, occasionally even obvious,
generalizations about the political economy of the type of global issues the
European Community (EC) handles. Wallace treats them as inherently
incomplete, perhaps one-sided generalizations squeezed out of a far more
complex historical reality. From his or her particular perspective, each seems
justified.
Despite differences of taste and training, however, the commentators
converge on at least four concrete criticisms. These are: (1) the historical data
and cases are incomplete and perhaps biased; (2) the analysis omits potentially
powerful explanations; (3) the theories employed explain ideal-typical
generalities but insufficient variation across cases; (4) the cases are biased in
favour of intergovernmental politics, because they do not explicitly consider
incremental change over time via judicial or administrative politics. In
responding to these four concerns, I shall pay primary attention to future
opportunities to advance the research programme underlying The Choice for
Europe.
 A. Moravcsik: The Choice for Europe 169

DOES THE CHOICE FOR EUROPE REST ON A BIASED SAMPLE

OF HISTORICAL DATA?
I begin with Wallace’s accurate observation that I have not related the entire story of
European integration. The Choice for Europe sharpens and extends the most
prominent explanations for preferences, bargaining, and delegation in the EC
literature, then collects and analyses limited categories of data pertaining to specific
hypotheses.2 Despite its length, it privileges generalization about broad outcomes
over subtle details, thereby quite deliberately omitting many secondary aspects and
influences of major EC decisions.
Omission alone, however, is no ground for complaint – as Wallace is well aware.
Any effort to illuminate general patterns in complex events must sacrifice some
richness of detail. Even if I could convince readers to read – or, in these thrifty times,
an academic publisher to publish – more than 520 pages on European integration,
the mass of journalistic reportage, secondary sources, oral histories, and
governmental documents on major European Union (EU) decisions is so vast that
one can quite literally cite dozens of plausible conjectures to explain any aspect of
these decisions, followed by dozens of sources to support each conjecture. Chasing
each one down in complete detail would expand the project to encyclopedic scope.3
The real question is thus not whether the history in The Choice for Europe is
incomplete – it obviously is – but whether it is biased. At the heart of the matter lies
Wallace’s scepticism that an unbiased sample of the available evidence really resolves
theoretical debates as decisively as I claim it does. If I had told more of the story or
consulted more sources, she hints, my conclusions – notably those about
commercial motivations and the weakness of supranational entrepreneurs – would
have been more ambivalent. Wallace openly questions whether the evidence so
clearly disconfirms geopolitical and ideological explanations and whether the
existing literature so strongly asserts them. Even Scharpf, who generously
acknowledges my attention to methodology, cannot resist adding that ‘there is
never any question of the author’s preference for one answer to each of his questions
. . . always the last-mentioned one’ (p. 165).4 In this, I suspect, Wallace and Scharpf
capture the instinctive reaction of many readers. Why, readers will ask, should we
accept a series of empirical claims that concede so little to the bulk of existing
scholarship? Isn’t this because, as Wallace believes, ‘the eye of the artist is selective’
and ‘slides over some of the elements that displease’ (p. 156) it?
Underlying the question of bias is a fundamental issue: By what methodological
criteria should qualitative analysts select and analyse data and hypotheses? While I
can claim neither to have consulted every available document nor to have
considered every conceivable conjecture, I do claim to have employed a
standardized procedure for selecting, analysing and presenting empirical evidence
that strives explicitly to be both transparent and unbiased.5 I aim to test explicit
hypotheses that express, extend and refine the most prominent arguments about
European integration. I repeatedly bend the structure of the book to consider
plausible conjectures that did not quite fit the mould. As Scharpf and Caporaso
observe, moreover, hypotheses are formulated to offer not just an even-handed test
between theories, but in some important ways to bias the results away from my
170 Journal of European Public Policy

final conclusions. In addition, I discard secondary-source speculation entirely,

including the citations to the Financial Times and Economist on which many
analyses rest, and rely instead on observable patterns of behaviour and a selection of
more reliable primary documents. I weight those sources carefully and openly. The
underlying goal of such techniques is to generate findings that are replicable – a
quality, as Caporaso and Scharpf concede, largely absent from existing studies of
the EC.6
To see the difference between what Wallace suggests I have done and what I
actually did, consider the case of General de Gaulle. Three points here. First,
Wallace believes I overstate the existing consensus in favour of geopolitical and
ideological explanations of European integration. Yet among the secondary articles
and books on the General, which number many thousands, I was not able to find a
single one that attributed his major EC decisions (e.g. opting for membership, the
Fouchet Plan, vetoing the British, provoking the ‘empty chair’ crisis) primarily to
commercial interest. Some mention economic interest as a secondary consideration;
some dismiss it altogether. Second, Wallace’s scepticism and the secondary literature
notwithstanding, the available primary evidence runs about 5:1 in favour of
economic motivations – with the most reliable evidence also being the most
favourable.7 I am open to suggestions as to why this finding might be spurious. I
have thought through most of the propagandistic possibilities and come to the
conclusion that it is an accurate representation of the General’s thinking. Third, this
evidence is not based, as Wallace implies, on selective citation of the opinions and
speculations of those memoir writers and interview subjects. I weight the evidence
carefully, with memoirs and interviews employed primarily for factual information
and greater emphasis accorded to those witnesses backed by corroborating
evidence and with little apparent incentive to mislead. Much of it relies on verbatim
transcripts of Cabinet meetings and confidential discussions. Sheer speculation
after the fact (unless openly presented as such in the text) is discounted to zero – no
matter who the author is. Where such interpretive judgements are problematic, they
are explained in detail to the reader.8
By explicating the precise methodological, theoretical and empirical bases on
which I reach conclusions, I have given potential critics a leg up. In contrast to
non-replicable studies based on inductive theory, I thereby render it far easier for
historians and political scientists to challenge the objectivity and accuracy of my
analysis. Such challenges are inevitable; some are sure to be telling, particularly
where they rest on newly available documents and data. That is the sort of debate we
should be having.
Yet, for the moment, Wallace in fact neither questions the methodological
criteria I employ nor offers an explicit demonstration of my failure to meet them.9
While conceding that my conclusions are counterintuitive – as is my intent – she
fails to show that they are in any way biased. I do not for a moment mean to
denigrate the more detailed and open-ended policy analysis of the kind practised by
Wallace and, in a somewhat different way, by leading diplomatic historians. These
have been unique and indispensable sources of insight and information for me as
well as many others – as exemplified by Wallace’s elegantly detailed critique. Yet in
 A. Moravcsik: The Choice for Europe 171

the end I side with Caporaso and Scharpf (and also in the end, I suspect, Wallace) in
my conviction that our understanding of European integration would also benefit
greatly from more disciplined and focused social scientific debates among
theoretically and methodologically replicable claims.

ARE IMPORTANT THEORIES NEGLECTED?

A more telling and not entirely unrelated criticism, advanced by all three
commentators, is that The Choice for Europe neglects important factors to which
more attention might fruitfully be given. I do not dispute the value of considering
new variables and welcome future research drawing on theories of policy ideas,
domestic institutions, and two-level games.10 To the contrary, testing such theories
against a rigorous baseline should be a major priority for future research in this area.
Wallace rightly observes that I fail explicitly to consider, except in passing,
explanations based on economic policy ideas. While the tight correlation between
structural economic interests, interest group mobilization, and national positions
calls such explanations into question, they surely merit more rigorous testing.11 The
analysis in The Choice for Europe predicts in general that the weaker and more
diffuse the domestic constituency behind a policy and the more uncertainty there is
about cause–effect relations, the greater the role of economic (like geopolitical)
ideas is likely to be.12 Thus I predict – as is the case in domestic policy – that we shall
observe a relatively modest autonomous impact for ideas in agricultural or perhaps
industrial tariff policy, a greater role in regulatory policy, and the greatest role in
monetary policy, where fundamental uncertainty about the consequences of policy
is greater and costs or benefits more diffusely distributed.13
Caporaso and Wallace observe that I downplay the role of domestic political
institutions, including political parties. Again, the correlation between structural
economic interests and national policy calls this explanation into question. One is
repeatedly struck, for example, by the longer-term continuity of national policy
during periods of partisan change. Still, one might point to areas in which partisan
differences matter or institutions occasionally play a covert role. We would expect
political parties to be more important in those broad redistributive issues – such as
social and monetary policy – as they are in domestic politics, as well as in areas of
overtly ideological motivation, such as the delegation of power to the European
Parliament. Examples of the background importance of institutions might be de
Gaulle’s role in reforming the French economy, surely a function in part of the
centralized constitution of the Fifth Republic, and the role of central banks in
monetary policy, which I treat as an exogenous constraint but (as Scharpf notes) do
not fully integrate with the generally pluralist thrust of the argument. Endogenous
tariff theory, which I appropriate, has moved recently toward more explicit
discussion of institutions – a trend of which EC scholars might usefully take note.14
Scharpf wonders, in a related criticism, why two-level game theory – in par-
ticular my claim elsewhere that European integration ‘strengthens the state’ – was
not included.15 The use of international institutions and foreign policy prerogatives
to make domestic policy does not simply involve exchanges among governments
172 Journal of European Public Policy

representing social interests, but also redistributes domestic influence among state
and societal actors. Theories of this kind surely deserve closer scrutiny, particularly
when explaining aspects of the EC’s ‘democratic deficit’. Yet the best way to test
whether such factors actually have an impact on policy is to compare them to a
baseline unitary-actor theory; anything less only invites confusion.16 In addition, I
believe such forces are empirically secondary. My purpose in The Choice for Europe
was firmly to establish such a baseline; I invite debate with any equally carefully
controlled empirical challenge.

DO THE THEORIES EXPLAIN SUFFICIENT ‘SECOND-ORDER’

VARIATION?
The Choice for Europe is explicitly multi-causal. It divides major EC decisions into
three analytical stages; within each stage, it seeks to assess the relative explanatory
power of two or three theories. It seeks to explain major intergovernmental
bargains by accounting for variation in national preferences, bargaining outcomes,
and institutional choices. All three commentators, led by Scharpf, suggest that more
could have been done to explain what might be termed ‘second-order’ variation –
that is, variation in the relative power of the theories that compete to explain each
stage. Rather than simply assessing the relative power of each theory in each
category – economic vs. geopolitical influences on preferences, intergovernmental
vs. supranational influences on bargaining, ideological vs. technocratic vs.
commitment considerations on institutional choice – could not more be done to
endogenize variation in the relative importance of each across cases? This is an
important and subtle observation, the implications of which merit serious
consideration.
All three commentators are correct to assert that greater attention to antecedent
conditions – and, though this is not made explicit, case selection more attuned to
variation on the independent variables – might have permitted us to learn more
about the conditions under which each theory holds. The major cost of not doing
this, Scharpf insightfully observes, is that ‘deviant cases are in danger of being
ignored or downplayed as “exceptions”, rather than being exploited for the
development of more reality-congruent theory through the introduction of
analytically pertinent distinctions’ (p. 166). On the question of preference
formation, Wallace notes that ‘an analysis that weighted, rather than
“exceptionalized”, the geopolitical component in relation to political economy
considerations might have served better’ (p. 156). Caporaso adds that ‘if the cases
had been ordered along several dimensions (say, more or less controversial, or from
collective gain to hard distributive gains and losses, or different types of policy
sector), comparison of the results across cases would have given us more leverage’
(p. 163).
There is much truth here. Yet Scharpf and Caporaso rightly present this not as
criticism, but as recognition of the trade-offs inherent in research design.17 While
openly aspiring to both, The Choice for Europe is in the end more problem-driven
than theory-driven, in that my primary goal is to isolate the most important
 A. Moravcsik: The Choice for Europe 173

determinants of a series of major decisions and mould them into a synthesis

that can establish certain baseline expectations about European integration. (It is a
bit surprising, given her comments above, that Wallace chides me for this choice.)
When forced to choose between the integrity of this goal and optimizing case
selection to advance particular lines of theory, I lean somewhat toward the former.
The result, as Caporaso hints, is that we gain somewhat more empirical knowledge
about the relative importance of each theory in the case of the EC (internal validity)
and somewhat less about the extent of its generalizability to other cases (external
validity). Refining useful theories by specifying antecedent conditions more
precisely and thereby explaining residual variation is surely one of the most
important tasks in the research programme that follows from The Choice for
Europe. That being said, however, it would be quite misleading to conclude, as one
might be tempted to do on the basis of Scharpf’s comments alone, that The Choice
for Europe lacks theory sufficiently well grounded in rigorous deductive
assumptions to support clear predictions about empirical scope or generalizability
beyond the EU. The book in fact extends existing bodies of theory to answer
precisely the questions that Scharpf, Wallace, and Caporaso pose. Two examples
must suffice.
First, Scharpf and Wallace suggest the need to specify conditions under which
supranational entrepreneurs are more or less likely to exercise influence over
international negotiations, thereby explaining the ‘exceptional’ case of the Single
European Act (SEA), in which Commission and Parliament officials played a
significant (if still secondary) role. My argument in The Choice for Europe, drawing
on social choice models of entrepreneurship, is that supranational influence is
possible only where two conditions are met: national governments face high ex-ante
transaction costs and significant informational (or ideational) asymmetries favour
supranational entrepreneurs. This, for reasons I elaborate in more detail, appears to
require that there should be domestic (and transnational) co-ordination problems,
which vary in predictable ways. In sum, if we know some basic things about
domestic politics, we can predict ‘windows of opportunity’ for supranational
entrepreneurship.18 In an article forthcoming in International Organization I
extend this line of argument, testing five fine-grained theories of supranational
entrepreneurship, each with subtly different informational assumptions and
antecedent conditions, which I then employ – precisely as Scharpf and Wallace
recommend – to endogenize the SEA ‘exception’. The resulting hypotheses can also
be applied across a wide range of international institutions.19
Second, Wallace and Caporaso suggest the need to specify more precisely the
conditions under which economic interests or geopolitical interests or ideology
matter more – thus explaining the ‘exceptional’ impact of geopolitics that I found in
approximately 20 per cent of the episodes of national preference formation. Here,
too, The Choice for Europe advances a distinctive argument. For the Olsonian
pluralist reasons Caporaso correctly attributes to me, when governments balance
these two imperatives, the presumption tends to be in favour of the issue-specific
(hence generally, in the case of the EC, economic) concerns.20 Geopolitical ideology
is more important where issue-specific consequences are essentially incalculable
174 Journal of European Public Policy

(e.g. recurrent debates over the powers of the European Parliament) or where core
national economic interests are already satisfied (e.g. German acceptance of a
customs union over a free trade area in the 1950s). While this hypothesis seems to
test out well, there is surely much room for greater specificity.21
In sum, we all agree that more precise, deductively grounded theory with clear
antecedent conditions should be a central objective of future research on European
integration. I believe that The Choice for Europe sets forth some significant
empirical propositions as steps toward such a goal – including most of those
recommended by Scharpf, Wallace and Caporaso – and that their empirical
confirmation further supports the underlying research agenda found there.

IS THE CASE SELECTION BIASED?

Caporaso and Scharpf suggest that the case selection may be biased. Rather than
selecting cases of major policy change per se – which might have included such
incremental processes as the establishment of a distinctive Brussels-based
bureaucratic style and culture, the development of formal and informal norms
around the Committee of Permanent Representatives (COREPER) or the Council
of Ministers, and, most important, the gradual assertion and acceptance of the
supremacy of European law – I focus on intergovernmental bargains per se. Scharpf
goes so far as to suggest that each element of my central conclusion that the relative
power of governments pursuing commercial interests determines the course of
integration has such ‘a high degree of a priori plausibility that it seems hard to take
[its] competitors quite as seriously as he does’ (p. 165).
This criticism need not detain us long, for two reasons. First, while Scharpf, like
myself, was trained to believe that basic political economy and bargaining theory
has ‘a high degree of a priori plausibility’ (p. 165), this is not, even now, accepted by
most scholars and public commentators on European integration. ‘Today,’ I write
in the book, ‘no claim seems more radical than the claim that the behavior of EC
member governments is normal.’ Other critical responses, including those of
Caporaso and Wallace, clearly illustrate this.
Second and more fundamental, it is simply not my intention to offer a
comprehensive theory of European integration. What I offer instead is far narrower,
namely a proposed solution to what is arguably ‘the most fundamental puzzle
confronting those who seek to understand European integration’, namely the
determination of which factors most strongly influence major intergovernmental
bargains.22 This constrained focus should trouble only those committed to the
venerable notion, of which Caporaso is rightly critical, that a single theory can
explain all of EC politics at one go. Following Haas’s admirably honest self-critique
in the early 1970s, I believe the search for such a theory to be futile, even
counterproductive.23 There cannot be a ‘theory of regional integration’ or ‘theory
of the EC’ any more than there can be a ‘theory of comparative politics’ or ‘theory
of American politics’. The EC is a complex, institutionally diverse, multi-faceted
political system, and I know of no convincing reason why a theory of incremental
legal or administrative change under delegated institutional constraints should have
 A. Moravcsik: The Choice for Europe 175

the same basis as a theory of intergovernmental bargains in a classical diplomatic

setting.24 Many events in the EC are not properly within the specified domain of my
theories – something that Scharpf, with his admirable attention to antecedent
conditions, should be the first to concede. The task of synthesizing all the disparate
elements of European integration into one model, if possible at all, still lies far
before us.
To be sure, I confess to just a bit of nostalgia – vicarious, in my case – for what
Caporaso terms an ‘ultimate showdown’ (p. 163) with neofunctionalism. Like him,
however, I feel it inappropriate to indulge it. Instead, the research programme
underlying The Choice for Europe seeks (in Caporaso’s apt colloquialism) to
‘mainstream’ integration studies (p. 161). Rather than refute neofunctionalism, we
should dismember it, either appropriating or challenging selective hypotheses. I
appropriate its interest group theory of politics and economic functionalism, while
challenging its supranationalist theory of bargaining and technocratic theory of
delegation.
Yet Scharpf and even Caporaso fail to acknowledge that, while the research
design of The Choice for Europe is not optimized for this purpose, the case studies
are none the less deployed to conduct a preliminary test of more dynamic ‘historical
institutionalist’ (HI) claims about endogenous processes of integration over time –
thereby addressing many concerns of those who believe incremental change is more
fundamental than grand bargains. For this task I employ what is surely the most
rigorous formulation of HI to date, one proposed by Paul Pierson. Pierson
persuasively argues that if HI claims are correct, we should observe over time
unintended or unforeseen consequences, unpredictable and unexpected shifts in
national preferences, and, perhaps, a powerful role for supranational entrepreneurs.25
This is consistent with, though does not necessarily imply, movement toward
Wallace’s view that we need to consider more fully the role of ‘irrationality . . .
confusion and . . . mistaken judgements’ (p. 158).
We observe, I argue in the conclusion to The Choice for Europe, none of these
things.26 To be sure, structural circumstances and state preferences evolve over time
and some (though decidedly not most) of these changes appear endogenous to prior
decisions to integrate. To the extent, moreover, that these adjustments involve
‘asset-specific’ investments – that is, investments dependent on maintenance and
continuation of integration – underlying support for integration is likely to deepen
over time as a result. Yet most of the significant endogenous changes in structural
circumstances and preferences between major decisions were foreseen, indeed
intentional. Though governments have often issued spurious denials, developments
such as large common agricultural policy (CAP) surpluses driven by high price
supports, the increasing size and influence of international exporters and investors,
and the tendency of qualified majority voting to impede opposition by govern-
ments with extreme preferences were far from unforeseen or undesired. The
primary purpose of European integration was to bring about just such results.
Unintended and undesired consequences have been secondary; changes in national
preferences have been incremental and linear over long periods.
176 Journal of European Public Policy

CONCLUSION: FEDERALISM AND THE FUTURE OF

INTEGRATION
I have focused throughout on the future of EC scholarship. My central point is that
The Choice for Europe in fact addresses a good number of the criticisms raised by
Caporaso, Scharpf and Wallace. Despite sub-disciplinary and methodological
differences, there is in fact considerable consensus among all four of us concerning
the proper path of future fruitful scholarship within the research programme set
forth in The Choice for Europe.
I would like to close by turning very briefly to the implications for the future of
European integration itself.27 Wallace suggests in closing that my analysis ‘provides
formidable arguments against those who portray European integration as
necessarily a cumulative and irreversible process’ (p. 159). This assessment rests on
the belief – a legacy not just of scholars like Karl Deutsch, but of the European
federalist movement – that ultimately only fundamental transfers of sovereignty
and shifts in values can lock in integration.
I believe The Choice for Europe supports a more optimistic prognosis. In an
era where democratization has pacified Western Europe, talk of federalism
triggers deep public suspicion, and technocratic planning (central banking
excepted) has fallen out of fashion, Europe is none the less proceeding toward
enlargement, monetary integration, and an ever deepening single market. There
is an underlying functional reason for this, namely the consistent increase in
social support, above all from producer interests, for the economic integration
of Europe. Over time, underlying socio-economic developments and the prior
success of the EU in achieving its objectives have created invested economic
interests that are the major guarantors of its future stability. From this
perspective, are not the true ‘Eurosceptics’ those who believe that the EU is
fragile because it rests on fears of refighting the Second World War, hopes of
realizing federalist dreams, the intermittent ‘political will’ of national leaders,
and the unintended consequences of prior actions? And are not the true
‘Europeans’ those who view the EU as a stable form of pragmatic co-operation
deliberately tailored to the enduring, increasingly convergent national interests
of European firms, governments, and citizens?

Address for correspondence: Andrew Moravcsik, Associate Professor of

Government, Center for European Studies, Harvard University, 27 Kirkland
Street, Cambridge, MA 02138, USA. Tel: 617 495 4303, ext. 205. Fax: 617 495 8509.
email: moravcs@fas.harvard.edu

NOTES
1 I am grateful to James Caporaso, Robert Keohane, Paul Pierson, George Ross and
Helen Wallace for comments on this article.
2 Throughout I employ the name European Community, not European Union, as the
book in question covers the period from 1955 to 1991.
3 Such constraints should not be underestimated. A distinguished reviewer for a leading
academic press – an outlier, fortunately – recommended that I cut The Choice for
 A. Moravcsik: The Choice for Europe 177

Europe in half by truncating the theoretical analysis, citing no more than five sources
per page, and eliminating the United Kingdom!
4 For the record, this sequence was imposed only in the last manuscript revision, after
consultation with my editor, for presentational reasons.
5 Andrew Moravcsik, The Choice for Europe: Social Purpose and State Power from
Messina to Maastricht (Ithaca: Cornell University Press, 1998), pp. 77–85, also 10–13.
(Hereafter ‘CFE’).
6 Gary King, Robert O. Keohane and Sidney Verba, Designing Social Inquiry: Scientific
Inference in Qualitative Research (Princeton: Princeton University Press, 1994), pp.
26–7. Without a shared commitment to replicability, for example, it is hard to know
what to make of Wallace’s assertion that the outcome of the British accession
negotiations in the early 1970s was decisively shaped by policy errors on the part of the
British government – a result she believes my approach is too crude to capture. While
my book does not investigate this case in depth – the section was cut in revisions – I do
consider British accession negotiations a decade earlier, which raised very similar issues.
Consistent with my practice of entertaining explanations that are widespread in the
literature, even when they are not among the standardized theories I test, I consider and
reject precisely this explanation – that is, the near universal view that British leaders
were incompetent or benighted. I conclude that the bulk of the primary evidence reveals
Macmillan’s diplomacy as far more foresighted and more economically motivated than
most diplomatic histories suggest. For this reason among others, I suspect that the
bargain that the British government struck in the early 1970s was about as good as could
be expected, given its internal and external weaknesses. The unfavourable outcome to
Britain fits, moreover, the pattern of subsequent Greek, Iberian and perhaps also
Scandinavian/Austrian accessions, as well as current negotiations with Central and South
European candidates, wherein applicants do badly in initial accession negotiations, in
which they are demandeurs, but subsequently exploit the prerogatives of membership to
extort side-payments. The British themselves did just this, if somewhat inadvertently, in
1975. My basic point is not that my interpretation is correct and Wallace’s is not. It is
instead that there are always many a priori plausible conjectures. To determine which are
more accurate, we must commit ourselves to rigorous and replicable interrogation of
unbiased data – and then debate the result with equally constrained critics. From this
perspective, Wallace’s empirical assertion is intriguing, but hardly conclusive.
7 CFE, pp. 176–97, 225–30. For a more systematic presentation of the evidence, see
Moravcsik, ‘De Gaulle and Europe: historical revision and social science theory’,
Center for European Studies Working Paper Series 8.5 (Cambridge, MA: Harvard
University, 1998).
8 For explication of sources on de Gaulle, see in particular CFE, p. 178n.
9 But she occasionally seems tempted: ‘Whether or not each winning idea necessarily
benefited the economies of the countries whose governments accepted them is a matter
of judgement or interpretation, not a matter of clear determination. There were after all
always critics.’ (p. 159).
10 Though there is not space here, a number of other areas might be added, including
political socialization, social construction of ideas, and public opinion.
11 Wallace cites the example of French debates over CAP reform as a matter where
government officials disagreed fundamentally about the welfare effects of integration.
Yet it is fair to say that no major EC policy was, at the time of its creation, recommended
by most objective observers – notably academic economists – on aggregate welfare
grounds. Certainly the preponderance of the evidence suggests that the decision of the
well-intentioned Giscard government, to which Wallace refers, turned on interest
group opposition, not aggregate welfare gains. And about the power of French farmers
there could be little uncertainty! For details of the Giscard government’s calculations,
see the sources cited in CFE, pp. 273–4, especially Michael Tracy (ed.), Farmers and
Politics in France (Enstone: Arkelton Trust, 1991).
178 Journal of European Public Policy

12 Also Judith Goldstein and Robert Keohane (eds), Ideas and Foreign Policy: Beliefs,
Institutions, and Political Change (Ithaca: Cornell University Press, 1993).
13 For sophisticated treatments, see Kathleen R. McNamara, The Currency of Ideas:
Monetary Politics in the European Union (Ithaca: Cornell University Press, 1997); Paul
Krugman, Peddling Prosperity: Economic Sense and Nonsense in the Age of Diminished
Expectations (New York: Norton, 1994); Alberto Giovannini, ‘Economic and
monetary union: what happened? Exploring the political dimension of optimum
currency areas’, in Centre for Economic Policy Research, The Monetary Future of
Europe (London: CEPR, 1993); Peter A. Hall (ed.), The Political Power of Economic
Ideas: Keynesianism across Nations (Princeton: Princeton University Press, 1989).
14 James Alt, Jeffry Frieden, Michael Gilligan, Dani Rodrik and Ronald Rogowski, ‘The
political economy of international trade: enduring puzzles and an agenda for inquiry,’
Comparative Political Studies 29(6) (December 1996): 689–717.
15 Andrew Moravcsik, ‘Why the European Community strengthens the state:
international co-operation and domestic politics’, Center for European Studies
Working Paper Series No. 52 (Cambridge, MA: Harvard University, 1994).
16 On the perils that befall those who conflate the two lines of argument, rather than
treating them as separate, see e.g. Karl-Orfeo Fioretos, ‘The anatomy of autonomy:
interdependence, domestic balances of power, and European integration’, Review of
International Studies 23(3) (July 1997): 293–320.
17 Indeed, this is not unlike the trade-off made by those Scharpf refers to as the ‘analytical
narratives club’ (p. 166) in focusing on the internal validity of a limited number of cases.
Like members of this club, I believe that the strongest support for my empirical results
lies in the number of facts, often new and unexpected, about a small number of cases that
some theories are able to explain. Accordingly, much of the detailed evidence on the
cases – e.g. on the positions of social groups and the distribution of information among
international actors – had never before been collected or assembled, and was absent
from early drafts of CFE. One difference, as Scharpf notes, is that the ‘analytical
narratives’ school places more emphasis on formal modelling and less on unbiased
comparative theory testing, whereas my priorities are in this book the opposite. Which
imposes tighter constraint on empirical explanations is open to debate; the answer no
doubt varies. Cf. Robert Bates et al., Analytic Narratives (Princeton: Princeton
University Press, 1998).
18 CFE, pp. 483–5.
19 Andrew Moravcsik, ‘A new statecraft? Supranational entrepreneurs and interstate co-
operation’, International Organization (forthcoming). For a preliminary version, see
the paper of the same name, Weatherhead Center for International Affairs Working
Paper Series No. 98-10 (Cambridge, MA: Harvard University, 1998).
20 CFE, pp. 50, 477–9. It is important to note that my theoretical formulation is in fact
slightly more complex, but thereby more generalizable. The issue is whether
issue-specific interdependence or linkage to general foreign policy concerns dominates
national preference formation. It is only because the issues with which the EC is
concerned are generally (but not always) economic, while general foreign policy
concerns are generally geopolitical or ideological, that this formulation generates the
expectation that (more focused) commercial interests will trump (more diffuse)
political-military motivations. CFE, pp. 26–7, 49–50.
21 This criticism was previously debated in Daniel Wincott, ‘Institutional interaction and
European integration: towards an everyday critique of liberal intergovernmentalism’,
Journal of Common Market Studies 33(4) (December 1995): 597–609; Andrew
Moravcsik, ‘Liberal intergovernmentalism and integration: a rejoinder’, Journal of
Common Market Studies 33(4) (December 1995): 611–28. One might extend this
analysis by theorizing more rigorously the sources of variation in the strength of
geopolitical interest – as suggested by the tendency of Germany to be more influenced
by geopolitical concerns than Britain or France. Even more comprehensive, at least in
 A. Moravcsik: The Choice for Europe 179

principle, would be an integrated theory of trade-offs between commercial and

geopolitical concerns, rather than a simple recognition of a trade-off between the two –
a ‘theory of grand strategy’. Wallace’s and Caporaso’s encouragement notwithstanding,
I suspect fine-grained predictions from such models are currently beyond the reach of
IR theory – perhaps for very fundamental reasons. For state of the art, see David A.
Lake, Entangling Relations: American Foreign Relations in its Century (University of
California, San Diego, unpublished manuscript, 1997).
22 CFE, p. 1.
23 CFE, pp. 13–17.
24 CFE, pp. 13–17. Of course, it remains an open question to what extent the pattern of
national preferences (while mitigated by institutional constraints) remains a, perhaps
the, decisive factor in daily decision-making or judicial interpretation. Some doubt this,
but for evidence (explicit and implicit) of decisive intergovernmental influence on daily
decision-making, see Adrienne Héritier, Susanne Mingers, Christoph Knill and
Martina Becka, Die Veränderung von Staatlichkeit in Europa. Ein regulativer
Wettbewerb: Deutschland, Grosbrittanien, Frankreich (Opladen: Leske & Budrich,
1994); Jonathan Golob, ‘The path to EU environmental policy: domestic politics,
supranational institutions, global competition’ (paper presented at the Fifth Biennial
International Conference of the European Community Studies Association, 29 May–1
June 1997, Seattle, Washington). For clear arguments that even where supranational
institutions and actors are important in daily decision-making, it is appropriate to begin
with an intergovernmental explanation, see Mark Pollack, ‘Delegation, agency and
agenda setting in the European Community’, International Organization 51(1)
(Winter 1997): 99–134; Moravcsik, ‘Liberal intergovernmentalism.’
25 Paul Pierson, ‘The path to European Union: an historical institutionalist account’,
Comparative Political Studies 29(2) (April 1996): 123–64.
26 CFE, pp. 489–94. A truly rigorous test, it is worth noting, would require sophisticated
input–output models of the domestic and international economy.
27 For a more detailed argument, see Andrew Moravcsik, ‘Europe’s integration at
century’s end’, in Moravcsik (ed.), Centralization or Fragmentation? Europe Facing the
Challenges of Deepening Diversity, and Democracy (New York: Council on Foreign
Relations Press, 1998), pp. 1–58.