Common use of MAI Clause in Contracts

MAI. Corollary 2: If B = B∗, then the only equilibrium is one where all countries join Proof: See appendix 7. The intuition of Corollary 2 is also straight forward. The benchmark B∗ is the value of the rent extraction rate that is chosen by a social planner in a world without information asymmetries. Since the lobbying distortion pushes the desired rent extraction rate below B∗, the time inconsistency distortion alone causes a rent extraction rate above B∗. However, MAI rule B∗ provides to all countries a commitment device to solve the time-inconsistency problem by joining MAI at no cost. Hence, opting out of MAI makes no longer sense. Overall, a weak MAI (large B) is not sharply binding and every country joins MAI to avoid losses from signalling high rent-extraction rates in case of staying out. Next, I compare the incentives of governments to protest against MAI negotiations. Every government objects negotiation if it expects a loss in a world with MAI compared to one without. Proposition 2 Governments of all countries that do not join MAI lose compared to a world without MAI; governments of all countries with χi > χ gain. At least some govern- ˜ ments of countries that join MAI with χi such that χ∗ > χi > χ lose. Proof: See appendix 8. This proposition can explain why some countries object other countries to negoti- ate a MAI even though they are neither forced into nor excluded from membership. A partial MAI, i.e. a MAI where some countries join and others opt out, exerts a negative information externality on non-members; outsiders signal that they are inclined to extract large rents from MNEs. The resulting investment diversion harms governments that do not decide to join. It need to be kept in mind, however, that a loss for a government does not necessarily imply a welfare loss of the country, since government objectives are distorted by lobbying groups.24 Proposition 2 can explain the protest storm of some LDCs against the negotiation of MAI by the club of the OECD countries, although they were both free to opt in or out. According to my explanation, they were fearing the information externality that may arise from the decision to opt out. The protest comes from governments that are ex post but not ex ante contra free-market spirited. Again, a quote by the former Commerce Secretary to Government of India supports this model feature: "Selective and judicious government intervention is therefore widely considered necessary to support or protect domestic industry and technology creation ... Adequate freedom and flexibility to pursue their own policies towards FDI and foreign technology is therefore regarded by developing countries as a matter of fundamental importance ..." (Ganesan, 1998, p. 5) 3 Endogenous MAI Formation I extend now the game by two additional stages to endogenize which countries start ne- gotiating agreements among themselves. I superimpose on top of the previous stages the choice of countries with which other countries to start negotiation and the choice of each negotiation group of how strict the MAI rule is going to be. The new timing is given in Figure 3. 24 The result in proposition 2 mirrors the one in ▇▇▇▇▇▇▇ and ▇▇▇▇▇ (2001). However, I derive the result analytically in a world of partial MAI while the previous study derived this result for a complete MAI. Government type ⎢ ⎥ 0, χ χ ∈ ⎡ _ ⎤ randomly drawn from uniform dis- tribution (private knowledge) Choosing negotia- tion group C Determi- ning strict- ness of MAI B Government decides to join MAI or stay out Z ∈{I , O} MNEs decide to locate in t=1 Governments decide on rent extraction rate ß MNEs decide to re-locate in t=2 3.1 Negotiating MAI In this section, I turn to the stage when MAI is negotiated. I assume that there exists a club C of a countable number of countries that starts exclusively negotiating an agreement.25 In particular, this group chooses the threshold B. The club is assumed to have more favorable political-risk characteristics than the world as a whole. All countries are then free to opt in or out after the agreement is written. I assume also that MAI takes the form of a rule β ≤ B. Then the strictness of MAI, i.e. the threshold value B, can be found from a simple Nash bargaining solution where the Nash product is defined as c∈C |C| Q [W (χ ,β , B) − W (χ ,β )] 1 , (19) |C| is the number of group members, and the government objective function of successful negotiation WcI (χc, βiM , B) obtains an additional argument B, since the strictness of MAI is now allowed to vary. In addition applies the participation constraint due to the assumption of unanimity among negotiators [WcI (χc, βiM , B) − WcN (χc, βiM )] ≥ 0 (20) for all countries c ∈ C. I denote with χs the country with the smallest weight on capital 25 By the chosen set-up, Northern home country interests are consistently excluded from shaping the agreement, because MNE profits are zero in any case. Any Northern home country will thus have interests similar to host countries. This feature clearly falls short of reality. Section 5 argues, however, why this particular model feature does not upset the model mechanics. { } income in the negotiation group C, i.e. χs = min χc . Likewise, I denote the country c∈C { } with the largest weight ▇▇, ▇.▇. ▇▇ = max χc . Then the constraint (20) is not binding for c∈C any country unless it is binding for country χs, since χs is the country that is first hit by a welfare loss according to Proposition 2 when the agreement gets too strict. When maximizing the Nash product (19) with respect to MAI strictness B under the participation constraint (20), one obtains the following first order condition WcI (χc, βiM , B) − WcN (χc, βiM ) P ∂WcI (χc, βiM , B) 1 c∈C