¹ The EC is one of the pillars of the European Union. The European Union also includes a common foreign and security policy, and co-operation in the fields of justice and home affairs. This article focuses on legislative procedures in the EC. In the other areas, decisions are based on intergovernmental negotiations. For a detailed description of the EC institutions, see for example, Nugent, Neil, The Government and Politics of the European Community (London: Macmillan, 1994).CrossRefGoogle Scholar

² The 626 Members of the European Parliament (MEPs) are directly elected, the twenty Commissioners are appointed by common accord of the national governments, and in the Council each country is represented by a government minister.

³ Currently, the consultation procedure accounts for about 75 per cent of the legislation that passes through Parliament (199 opinions in 1993), the co-operation procedure accounts for about 20 per cent (fifty first readings in 1993), and the assent procedure accounts for very little legislation (eight cases in 1993).

⁴ Bieber, Roland, ‘Legislative Procedure for the Establishment of the Single Market’, Common Market Law Review, 25 (1988), 711–24.Google Scholar

⁵ Fitzmaurice, John, ‘An Analysis of the European Community's Co-operation Procedure’, Journal of Common Market Studies, 4 (1988), 389–400.CrossRefGoogle Scholar

⁶ Bogdanor, Vernon, ‘The June 1989 European Elections and the Institutions of the European Community’, Government and Opposition, 24 (1989), 199–214.CrossRefGoogle Scholar

⁷ Corbett, Richard, ‘Testing the New Procedures: The European Parliament's First Experiences with its New “Single Act” Powers’, Journal of Common Market Studies, 4 (1989), 359–72.CrossRefGoogle Scholar

⁸ Lenaerts, Koen, ‘Some Reflections on the Separation of Powers in the European Community’, Common Market Law Review, 28 (1991), 11–35.Google Scholar

⁹ Tsebelis, George, ‘The Power of the European Parliament as a Conditional Agenda Setter’, American Political Science Review, 88 (1994), 128–42.CrossRefGoogle Scholar

¹⁰ Crombez, Christophe, ‘Maastricht and the Co-Decision Procedure’ (working paper, Stanford University, California, 1994).Google Scholar

¹¹ The analysis can be extended to other types of single-peaked preferences with different countries being pivotal in the Council. If preferences are not single-peaked, the conclusions depend on the entire configuration of preferences.

¹² I assume for simplicity that no two countries have identical ideal policies.

¹³ In other words Black's median voter theorem applies. See Black, Duncan, The Theory of Committees and Elections (Cambridge: Cambridge University Press, 1958), pp. 14–25.Google Scholar In multidimensional extensions of the models, there will usually be no equilibrium unless the Commission and the Parliament are considered as unitary actors.

¹⁴ In particular, France, Germany, Italy and the United Kingdom have ten votes each; Spain eight; Belgium, Greece, Portugal and the Netherlands five each; Austria and Sweden four each; Denmark, Finland and Ireland three each; and Luxembourg two.

¹⁵ If p = 0 or p = 2_i country i is indifferent between p and the status quo. For simplicity I assume country i then votes in favour of p. For a country i with _i = 0, the acceptance set A_i is equal to the singleton {0}.

¹⁶ For the purposes of this theory it is sufficient to assume that a qualified majority is larger than a simple majority.

¹⁷ To obtain a qualified majority the proposal needs the support of at least eight countries. If two of the largest five countries vote against, the support of eleven other countries is necessary. If three of the largest countries vote against, no qualified majority can be obtained. Hence, _a ∈ {₃, ₄, ₅, ₆, ₇, ₈} and ∈ {₈, ₉, ₁₀, ₁₁, ₁₂, ₁₃} Since country 8 can be pivotal for a move to the right (left) only if the ideal policies of the largest five countries are to the right (left) of its ideal policy, it cannot be pivotal for moves to the right and to the left.

¹⁸ In the case of the assent procedure, the treaties do not use the term ‘common position’. I do it merely to preserve the analogy with the co-operation procedure.

¹⁹ The support S of the countries' ideal policies is equal to the set of policies p such that no policy is preferred to p by all countries.

²⁰ To locate the countries and institutions relative to one another I used data on party policy positions in a left–right policy space gathered by Laver and Hunt. See Laver, Michael and Hunt, W. Ben, Policy and Party Competition (New York: Routledge, Chapman and Hall, 1992).Google Scholar I assumed Commissioners and MEPs have the same policy positions as their national parties, and to obtain a country's policy position I averaged the policy positions of its governing parties.

²¹ See, for example, Garrett, Geoffrey, ‘International Cooperation and Institutional Choice: The European Community's Internal Market’, International Organization, 46 (1992), 533–60CrossRefGoogle Scholar; and Tsebelis, , ‘The Power of the European Parliament as a Conditional Agenda Setter’.Google Scholar

²² The Commission does not prefer the equilibrium policy to the status quo if CS∩A_c = Q∩S∩A_c = ø, where A_c is the Commission's acceptance set and is defined analogously to the countries' acceptance sets. Since Q∩S≠ø and Q∩A_c≠ø, CS∩A_c = ø ⇔ S∩A_c = ø.

²³ A coalition is a set of countries and institutions. It is winning if approval by all its members assures adoption of a proposal. A winning coalition that would no longer be winning if any member were to leave is a minimal winning coalition.

²⁴ The seminal paper on agenda-setting behaviour was written by Romer and Rosenthal (see Romer, Thomas and Rosenthal, Howard, ‘Political Resource Allocation, Controlled Agendas, and the Status Quo’, Public Choice, 33 (1978), 27–43).CrossRefGoogle Scholar

²⁵ Depending on the location of the Commission's ideal policy relative to the pivotal countries, four other situations are possible. (1) The Commission's ideal policy is located between country a's ideal policy and country b's ideal policy. The Commission then has agenda-setting power in interval IV, [_b, 2_b − _c), because both country b and the Commission want to move to the left. (2) The Commission prefers a policy to the right of all countries' policies. In that case it cannot obtain its ideal policy in intervals I and V (− ∞, 2_a − _n) and [_n, ∞), and _n then becomes EC policy. (3) The Commission's ideal policy is between country l's ideal policy and country a's ideal policy. Then the equilibrium policies are analogous to the equilibrium policies in Figure 4. (4) The Commission's ideal policy is to the left of country l's ideal policy. The equilibrium policies are then analogous to the equilibrium policies under (2).

²⁶ Since preserving the status quo also represents a decision, one might prefer the term ‘stasis’ to indecision.

²⁷ See Krehbiel, Keith, ‘Institutional and Partisan Sources of Gridlock: A Theory of Divided and Unified Government’, Journal of Theoretical Politics, forthcoming 1996.Google Scholar

²⁸ In a pure intergovernmental system, however, all decisions require unanimity in the Council. This increases the indecision because it eliminates one of the two possible minimal winning coalitions.

²⁹ Note, however, that the Commission proposes the ideal policy of the closest country if all countries have a higher (lower) ideal policy than the Commission.

³⁰ See, for example, Krehbiel, Keith, ‘Spatial Models of Legislative Choice’, Legislative Studies Quarterly, 13 (1988), 259–319.CrossRefGoogle Scholar

³¹ Country i in the case where _i < _a or _i > _b is considered a country with an extreme ideal policy. If a qualified majority wants to move to the right it is possible that countries that have ideal policies to the left of country a's ideal policy prefer the status quo to the equilibrium policy, depending on the Commission's ideal policy. Similar reasoning holds if a qualified majority wants to move to the left.

³² Depending on the location of the Parliament's ideal policy relative to the configuration of ideal policies of the pivotal countries and the Commission shown in Figure 6, four other situations are possible. (1) The Parliament's ideal policy is to the left of all countries' ideal policies. The Commission then prefers to attract country l's rather than the Parliament's support, and interval III is [₁, _a), interval II is [2₁ − _c, ₁) and interval I is (∞, 2₁ − _c). (2) The Parliament has an ideal policy to the right of country a's ideal policy but to the left of the Commission's ideal policy, i.e. _p ∈ [_a, _c). In that case the equilibrium policies are the same as under the consultation procedure. (3) The Parliament's ideal policy is to the right of the Commission's ideal policy but to the left of country n's ideal policy, i.e. _p ∈ [_c, _n). Then there are two differences with the consultation procedure: the status quo prevails in the interval [_c, _p) and the Commission has agenda-setting powers in the interval [_p, 2_p − _c). (4) The Parliament has an ideal policy to the right of all countries' ideal policies. Then there are also two differences with the consultation procedure: the status quo prevails in the interval [_c, _n) and the Commission has agenda-setting powers in the interval [_n, 2_n − _c). The equilibrium policies for different locations of the Commission's ideal policy are analogous.

³³ The last situation is in fact a combination of situations (2) and (3). It is not presented in Figure 6. It would occur in interval V, for example, if the Parliament's ideal policy were to the right of the Commission's ideal policy.

³⁴ In Figure 7 the parliament prefers at least one country's ideal policy to the status quo. The policies that are closest to the countries' ideal policies such that they are preferred to the status quo by the Parliament then belong to the support S of the countries' ideal policies. Thus, the set S is equal to the set S.

³⁵ The equilibrium policies for other configurations of ideal policies are analogous.

³⁶ More precisely, no policy other than the status quo is weakly preferred to the status quo by a qualified majority. Any proposal and any amendment other than the status quo are defeated. The set Q is equal to the singleton {0}.

³⁷ Note that if country a prefers a negative policy and country b prefers a positive policy the set CS is equal to the singleton {0}. As mentioned above, the status quo prevails, so the above statement holds in this case as well.

³⁸ It is easy to check that p _cs ∈ CS for all six cases: (1) p_cs = 0 ∈ Q, and ₁ < 0 < _a ⇒ p_cs ∈ S; (2) p_cs = ₁ > 0 ⇒ p_cs ∈ Q, and obviously p_cs ∈ S; (3) p_cs = PR_c ∈ CS; (4) p_cs = ₁ > PR_c ≥ 0 ⇒ ₁ ∈ Q, and p_cs ∈ S; (5) p_cs = ₁ ≥ 2_n − PR_c ≥ 2_n − 2_a > 0 ⇒ p_cs ∈ Q, and p_cs ∈ S; and (6) 0 < p_cs = 2_n − PR_c < _n < 2_a ⇒ p_cs ∈ Q, and ₁ < p_cs < _n.

³⁹ The optimal proposal is not amended by country 1 nor by any other country.

⁴⁰ In fact, the Parliament compares the common position to the policy that is implemented if it rejects the common position. This policy is not necessarily equal to the status quo. If a unanimous Council confirms the common position, it is the common position and the Parliament is indifferent. I assume the Parliament then votes sincerely. Therefore, it can be said that the Parliament compares the common position to the status quo. This assumption does not affect the equilibrium policy.

⁴¹ More precisely, the Council does not confirm any common position different from the status quo. The set U is then equal to the singleton {0}.

⁴² In fact, the Council compares to the status quo the policy that is implemented if it approves the proposal. This policy is not necessarily equal to the proposal. If neither the Parliament nor a unanimous Council prefer the proposal to the status quo, it is the status quo that prevails and the countries are indifferent. I assume countries then vote sincerely. Therefore, it can be said that the Council compares the proposal to the status quo. This assumption does not affect the equilibrium policy.

⁴³ In fact, the countries compare to the status quo the policy that is implemented if the amendment is approved. This policy is not necessarily equal to the amendment. If neither the Parliament nor a unanimous Council prefer the amendment to the status quo, it is the status quo that prevails and the countries are indifferent. I assume countries then vote sincerely. Therefore, it can be said that the countries compare the amendment to the status quo. This assumption does not affect the equilibrium policy.

⁴⁴ One only needs to substitute Q∩(A_p∪U) for Q to obtain the six cases under the co-operation procedure from the six cases under the consultation procedures.

⁴⁵ The implemented policy is preferred to the status quo by a qualified majority and belongs to the support of the countries' ideal policies for the same reasons as under the consultation procedure. Moreover, it is preferred to the status quo by the Parliament or by all countries in all six cases: (1) p_cp = 0 ∈ A_p; (2), (4) and (5) p_cp = ₁ ∈ U; (3) p_cp = PR_c ∈ CP; and (6) 0 < p_cp = 2_n − PR_c < _n ∈ A_p ∪ U ⇒ p_cp ∈ A_p ∪ U.

⁴⁶ If any policy the Parliament prefers to the status quo is left of all countries' ideal policies, then the set AS is the singleton of the most rightist policy the Parliament prefers to the status quo, i.e. AS = max {0, 2_p}. If any policy the Parliament prefers to the status quo is right of all countries' ideal policies, then the set AS is the singleton of the most leftist policy the Parliament prefers to the status quo, i.e. AS = min{0, 2_p}.