American Political Science Review (2018)112.4,1016-1035 doi:10.1017/S0003055418000400 American Political Science Association 2018 Reelection and Renegotiation:International Agreements in the Shadow of the Polls PETER BUISSERET University of Chicago DAN BERNHARDT University of Illinois and University of Warwick e study dynamic international agreements when one of the negotiating parties faces a threat of electoral replacement during negotiations,when agreements made before the election are the starting point for any subsequent renegotiation,and when governments cannot commit to future negotiation strategies.Conflicts of interest between governments may be softened or intensified by the governments'conflicts of interest with voters.We characterize when the threat of electoral turnover strengthens the prospect for successful negotiations,when it may cause negotiations to fail,and how it affects the division of the surplus from cooperation.We also show how changes in domestic politics- including uncertainty about the preferences of domestic political parties-affect a domestic government's ability to extract greater concessions in negotiations. "I decided rather than terminating NAFTA...we will renego- 1974 election manifesto promised to "seek a funda- tiate.Now,if I'm unable to make a fair deal,if I'm unable to mental re-negotiation."Upon entering government make a fair deal for the United States,meaning a fair deal for in 1974,Labour reopened negotiations,obtaining con- our workers and our companies,I will terminate NAFTA cessions in exchange for the UK's continued partici- But we're going to give renegotiation a good,strong shot." pation.After the 2015 general election,a Conserva- President Trump,April 27,2017 tive government initiated the renegotiation that culmi- nated in a vote for the UK to exit the European Union ("Brexit").In May 2017,the Trump administration no- INTRODUCTION tified Congress that it plans to renegotiate the North tates sign treaties,accede to international institu- American Free Trade Agreement(NAFTA).And in tions and organizations,and lend money to other June 2017,Donald Trump withdrew the United States states.The division of the surplus arising from from the Paris Climate accord with the prime intent to these activities is negotiated by the governments of the renegotiate better terms,asserting: day.However,these governments may,in turn,be re- placed by new governments over the life of the agree- "In order to fulfill my solemn duty to protect America and ment.This raises the possibility that arrangements its citizens,the United States will withdraw from the Paris signed by today's administrations may not be honored climate accord but begin negotiations to reenter either the by their successors. Paris accord or an entirely new transaction on terms that are fair to the United States." In fact,newly elected governments often try to rene- gotiate a predecessor's agreement.A Conservative government took the United Kingdom into the Eu- International negotiations may polarize and even ropean Economic Community (EEC)in 1973.That dominate domestic politics.In March 2010,the Euro- same year,the Labour Party declared that it "opposes pean Central Bank,EU,and IMF (the "Troika")es- British membership [in the EEC]on the terms ne- tablished emergency loan agreements to Greece.The gotiated by the Conservative Government,"and its first Greek bailout was negotiated between the Troika and the center-left PASOK government,which held a parliamentary majority of fewer than ten seats and Peter Buisseret is an Assistant Professor,Harris School of Public Pol- hence faced an ongoing threat of electoral replace- icy,University of Chicago(pbuisseret@uchicago.edu). ment over the life of the agreement.A domestic power Dan Bernhardt is IBE Distinguished Professor,Department of transition to an anti-bailout party could threaten the Economics,University of Illinois,and a Professor,Department of agreement's survival and the perceived harshness of Economics,University of Warwick(danber@illinois.edu) the initial terms could itself increase the chance of a We are indebted to Vincent Anesi,Mark Fey,and Amy Pond for their discussions of earlier drafts of this paper.In particular,we grate- more hostile future government via voters'dissatisfac- fully acknowledge Mark Fey's help in clarifying our intuition be- tion with the agreement.Both risks were realized:in hind Lemma 1.We also thank Emiel Awad,Steve Callander,Ker the next election.PASOK lost 119 seats.while Syriza- win Charles,John Duggan,Wiola Dziuda,Jon Eguia,Gilat Levy. the radical left-wing party that staunchly opposed the Kostas Matakos,Adam Meirowitz,Pablo Montagnes,Stephanie Rickard,Alastair Smith,and Richard Van Weelden for very help- bailout terms-became the second largest party.In Jan- ful discussions.We are grateful to seminar audiences at King's, uary 2015,Syriza came to power on the back of the Bath,Columbia.the Cornell Political Economy Conference.the Greek electorate's hostility to the austerity measures. Canadian Public Economic Group Meeting,and the UNSW-UQ Po- The new Greek government immediately reopened litical Economy Workshop.The Editor and three anonymous refer- ees offered many excellent suggestions that substantially improved this paper. 士 The first quote is from Labour's Programme for Britain (1973),and Received:August 10,2017:revised:March 2.2018:accepted:June 14. the second is from the Labour Party's February 1974 general election 2018.First published online:September 10,2018. manifesto. 1016
American Political Science Review (2018) 112, 4, 1016–1035 doi:10.1017/S0003055418000400 © American Political Science Association 2018 Reelection and Renegotiation: International Agreements in the Shadow of the Polls PETER BUISSERET University of Chicago DAN BERNHARDT University of Illinois and University of Warwick We study dynamic international agreements when one of the negotiating parties faces a threat of electoral replacement during negotiations, when agreements made before the election are the starting point for any subsequent renegotiation, and when governments cannot commit to future negotiation strategies. Conflicts of interest between governments may be softened or intensified by the governments’ conflicts of interest with voters. We characterize when the threat of electoral turnover strengthens the prospect for successful negotiations, when it may cause negotiations to fail, and how it affects the division of the surplus from cooperation. We also show how changes in domestic politics— including uncertainty about the preferences of domestic political parties—affect a domestic government’s ability to extract greater concessions in negotiations. “I decided rather than terminating NAFTA...we will renegotiate. Now, if I’m unable to make a fair deal, if I’m unable to make a fair deal for the United States,meaning a fair deal for our workers and our companies, I will terminate NAFTA. But we’re going to give renegotiation a good, strong shot.” President Trump, April 27, 2017 INTRODUCTION States sign treaties, accede to international institutions and organizations, and lend money to other states. The division of the surplus arising from these activities is negotiated by the governments of the day. However, these governments may, in turn, be replaced by new governments over the life of the agreement. This raises the possibility that arrangements signed by today’s administrations may not be honored by their successors. In fact, newly elected governments often try to renegotiate a predecessor’s agreement. A Conservative government took the United Kingdom into the European Economic Community (EEC) in 1973. That same year, the Labour Party declared that it “opposes British membership [in the EEC] on the terms negotiated by the Conservative Government,” and its Peter Buisseret is an Assistant Professor,Harris School of Public Policy, University of Chicago (pbuisseret@uchicago.edu). Dan Bernhardt is IBE Distinguished Professor, Department of Economics, University of Illinois, and a Professor, Department of Economics, University of Warwick (danber@illinois.edu). We are indebted to Vincent Anesi, Mark Fey, and Amy Pond for their discussions of earlier drafts of this paper. In particular, we gratefully acknowledge Mark Fey’s help in clarifying our intuition behind Lemma 1. We also thank Emiel Awad, Steve Callander, Kerwin Charles, John Duggan, Wiola Dziuda, Jon Eguia, Gilat Levy, Kostas Matakos, Adam Meirowitz, Pablo Montagnes, Stephanie Rickard, Alastair Smith, and Richard Van Weelden for very helpful discussions. We are grateful to seminar audiences at King’s, Bath, Columbia, the Cornell Political Economy Conference, the Canadian Public Economic Group Meeting, and the UNSW-UQ Political Economy Workshop. The Editor and three anonymous referees offered many excellent suggestions that substantially improved this paper. Received: August 10, 2017; revised: March 2, 2018; accepted: June 14, 2018. First published online: September 10, 2018. 1974 election manifesto promised to “seek a fundamental re-negotiation.”1 Upon entering government in 1974, Labour reopened negotiations, obtaining concessions in exchange for the UK’s continued participation. After the 2015 general election, a Conservative government initiated the renegotiation that culminated in a vote for the UK to exit the European Union (“Brexit’’). In May 2017, the Trump administration notified Congress that it plans to renegotiate the North American Free Trade Agreement (NAFTA). And in June 2017, Donald Trump withdrew the United States from the Paris Climate accord with the prime intent to renegotiate better terms, asserting: “In order to fulfill my solemn duty to protect America and its citizens, the United States will withdraw from the Paris climate accord but begin negotiations to reenter either the Paris accord or an entirely new transaction on terms that are fair to the United States.” International negotiations may polarize and even dominate domestic politics. In March 2010, the European Central Bank, EU, and IMF (the “Troika’’) established emergency loan agreements to Greece. The first Greek bailout was negotiated between the Troika and the center-left PASOK government, which held a parliamentary majority of fewer than ten seats and hence faced an ongoing threat of electoral replacement over the life of the agreement. A domestic power transition to an anti-bailout party could threaten the agreement’s survival and the perceived harshness of the initial terms could itself increase the chance of a more hostile future government via voters’ dissatisfaction with the agreement. Both risks were realized: in the next election, PASOK lost 119 seats, while Syriza— the radical left-wing party that staunchly opposed the bailout terms—became the second largest party. In January 2015, Syriza came to power on the back of the Greek electorate’s hostility to the austerity measures. The new Greek government immediately reopened 1 The first quote is from Labour’s Programme for Britain (1973), and the second is from the Labour Party’s February 1974 general election manifesto. 1016 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:04, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000400
Reelection and Renegotiation negotiations with EU member states.which nearly led for them given the agreement that was initially nego- Greece to exit the European Monetary Union. tiated.Following elections,domestic agents receive a In the context of a renegotiation,the effective bar- stochastic and publicly observed shock to their prefer- gaining power of a government typically derives from ences over the project.For example.civil unrest may its relative willingness to walk away from an existing raise the domestic political cost of the project regard- agreement,either in accordance with an exit process less of which political party holds power. stipulated in the agreement itself,or by simply abrogat- At date two,the transfer negotiated before the elec- ing the terms.2 This was manifest in the unilateral deci- tion serves as the transfer that would be made if sion by the Bush Administration to withdraw from the the new domestic government again implements the Kyoto Protocol in 2001,and in 2017 when the Trump project.However,either the foreign or domestic gov- Administration threatened to"terminate"NAFTA ab- ernment may renegotiate the existing terms by propos- sent a renegotiation that would deliver a"fair deal"for ing a new transfer.If accepted by the other govern- the United States.Indeed,when Margaret Thatcher ment,the proposed transfer replaces the standing offer: renegotiated a two-thirds rebate of Britain's contribu- but if rejected,the initial transfer remains in place.The tion to the budget of the EECin 1984,she is reported to foreign government then makes the prevailing transfer have succeeded only by threatening to withhold all of if and only if the date-two domestic government imple- Britain's contribution unless her demands were met.4 ments the project. The revised British contribution remained in place un- We explore how the prospect for initial agreements til2005.5 and the division of the surplus varies with(a)the pref- Our paper asks:How do pending national elections erences of the date-one domestic government,(b)un- determine (a)the prospects for initial cooperation be- certainty about the preferences of a future domestic tween states and (b)the division of the surplus from an government,(c)uncertainty about the preferences of vc士 agreement?And how do the terms of an initial agree- the domestic electorate,and(d)how agents discount ment affect the prospect of electoral replacement,the future outcomes. 4号元 bargaining attitude of a potential successor,or the risk Obviously,if agents care only for the short term,the that a successor will walk away from the agreement? foreign government wants to make the smallest date- Our Approach.At each of two dates,a domestic gov- one transfer that induces the domestic government to ernment negotiates an agreement with a foreign gov- pursue the project.But,suppose that agents care about ernment.The agreement specifies whether a binary the future consequences of an initial agreement.When policy project is undertaken and the extent of trans- a future domestic government takes power,it may want to negotiate a larger transfer than what it inherited. fers to be made between governments in return for implementing the project.The transfers could include But whether the foreign government would agree to a larger transfer depends on the credibility of the domes- budget contributions,rebates,or regulatory carve-outs. tic government's threat to abandon the project based Agents hold commonly known initial project valua- tions.The domestic incumbent is either intrinsically rel- on the existing terms-the more primitively hostile atively friendly or hostile to the project,but we assume is the date-two domestic government to the project, that neither domestic party initially wants to imple- (a)the smaller is date-two surplus,but (b)the greater ment the project without some concessions from the is the set of circumstances in which it would be willing to walk away from the existing agreement absent rene- foreign government. After the initial negotiations conclude,a national gotiation.This fundamental tension bears on all of our results. election determines whether the incumbent is retained or replaced by the other party.We first assume that When the election outcome is unaffected by ini- the uncertainty over who will hold power at date two tial negotiations,we prove that the two governments is unaffected by date-one outcomes.We then assume reach an agreement if and only if the immediate(date- that voters cast their ballots for whichever party is best one)total surplus from the project is positive.That is, static and dynamic conditions for an agreement coin- cide.Moreover,agreements always feature the small- est transfer that induces the date-one domestic govern- 2 The Treaty of Lisbon introduced an explicit procedure for a mem- ment to implement the project.Thus,beliefs about who ber country of the EU to exit 3 see https://goo.gl/UlBqBM will hold power in the future are irrelevant for whether 4 See Future Financing of the European Union.(6th Report session an initial deal is signed and for how the surplus from 2004-05,HL Paper62,page21,Q68). agreement is divided. 5 Power-sharing arrangements between central and peripheral gov- Matters are very different when domestic voters ernments within states are also subject to the threat of renegotia- select their date-two domestic representative taking tion,influenced by the threat or realization of electoral success by nationalist and secessionist regional parties,resulting in partial devo- into account initial negotiation outcomes.More hostile lution of policymaking(e.g,Catalonia or the Basque State in Spain: domestic governments can more credibly threaten to Quebec in Canada influenced by the Parti Quebecois;Scotland in walk away from an existing agreement.This raises the Great Britain,influenced by the Scottish Nationalist Party:or the prospect of appropriating more of the surplus and the Flemish Community and Walloon Region in Belgium).The terms attractiveness of electing a government that is more in- governing the division of policymaking responsibilities weigh heavily on elections and anticipation of possible renegotiations after futur trinsically hostile to the project.But,when represen- elections weigh on the current devolution of policymaking.We thank tatives are more hostile to the project than voters,the Laurent Bouton for proposing this application of our framework. misaligned interests also raise the prospect that the 1017
Reelection and Renegotiation negotiations with EU member states, which nearly led Greece to exit the European Monetary Union. In the context of a renegotiation, the effective bargaining power of a government typically derives from its relative willingness to walk away from an existing agreement, either in accordance with an exit process stipulated in the agreement itself, or by simply abrogating the terms.2 This was manifest in the unilateral decision by the Bush Administration to withdraw from the Kyoto Protocol in 2001, and in 2017 when the Trump Administration threatened to “terminate” NAFTA absent a renegotiation that would deliver a “fair deal” for the United States.3 Indeed, when Margaret Thatcher renegotiated a two-thirds rebate of Britain’s contribution to the budget of the EEC in 1984, she is reported to have succeeded only by threatening to withhold all of Britain’s contribution unless her demands were met.4 The revised British contribution remained in place until 2005.5 Our paper asks: How do pending national elections determine (a) the prospects for initial cooperation between states and (b) the division of the surplus from an agreement? And how do the terms of an initial agreement affect the prospect of electoral replacement, the bargaining attitude of a potential successor, or the risk that a successor will walk away from the agreement? Our Approach. At each of two dates, a domestic government negotiates an agreement with a foreign government. The agreement specifies whether a binary policy project is undertaken and the extent of transfers to be made between governments in return for implementing the project. The transfers could include budget contributions, rebates, or regulatory carve-outs. Agents hold commonly known initial project valuations.The domestic incumbent is either intrinsically relatively friendly or hostile to the project, but we assume that neither domestic party initially wants to implement the project without some concessions from the foreign government. After the initial negotiations conclude, a national election determines whether the incumbent is retained or replaced by the other party. We first assume that the uncertainty over who will hold power at date two is unaffected by date-one outcomes. We then assume that voters cast their ballots for whichever party is best 2 The Treaty of Lisbon introduced an explicit procedure for a member country of the EU to exit. 3 see https://goo.gl/UlBqBM. 4 See Future Financing of the European Union, (6th Report session 2004-05, HL Paper 62, page 21, Q68). 5 Power-sharing arrangements between central and peripheral governments within states are also subject to the threat of renegotiation, influenced by the threat or realization of electoral success by nationalist and secessionist regional parties, resulting in partial devolution of policymaking (e.g., Catalonia or the Basque State in Spain; Quebec in Canada influenced by the Parti Quebecois; Scotland in Great Britain, influenced by the Scottish Nationalist Party; or the Flemish Community and Walloon Region in Belgium). The terms governing the division of policymaking responsibilities weigh heavily on elections and anticipation of possible renegotiations after future elections weigh on the current devolution of policymaking.We thank Laurent Bouton for proposing this application of our framework. for them given the agreement that was initially negotiated. Following elections, domestic agents receive a stochastic and publicly observed shock to their preferences over the project. For example, civil unrest may raise the domestic political cost of the project regardless of which political party holds power. At date two, the transfer negotiated before the election serves as the transfer that would be made if the new domestic government again implements the project. However, either the foreign or domestic government may renegotiate the existing terms by proposing a new transfer. If accepted by the other government, the proposed transfer replaces the standing offer; but if rejected, the initial transfer remains in place. The foreign government then makes the prevailing transfer if and only if the date-two domestic government implements the project. We explore how the prospect for initial agreements and the division of the surplus varies with (a) the preferences of the date-one domestic government, (b) uncertainty about the preferences of a future domestic government, (c) uncertainty about the preferences of the domestic electorate, and (d) how agents discount future outcomes. Obviously, if agents care only for the short term, the foreign government wants to make the smallest dateone transfer that induces the domestic government to pursue the project. But, suppose that agents care about the future consequences of an initial agreement. When a future domestic government takes power,it may want to negotiate a larger transfer than what it inherited. But whether the foreign government would agree to a larger transfer depends on the credibility of the domestic government’s threat to abandon the project based on the existing terms—the more primitively hostile is the date-two domestic government to the project, (a) the smaller is date-two surplus, but (b) the greater is the set of circumstances in which it would be willing to walk away from the existing agreement absent renegotiation. This fundamental tension bears on all of our results. When the election outcome is unaffected by initial negotiations, we prove that the two governments reach an agreement if and only if the immediate (dateone) total surplus from the project is positive. That is, static and dynamic conditions for an agreement coincide. Moreover, agreements always feature the smallest transfer that induces the date-one domestic government to implement the project. Thus, beliefs about who will hold power in the future are irrelevant for whether an initial deal is signed and for how the surplus from agreement is divided. Matters are very different when domestic voters select their date-two domestic representative taking into account initial negotiation outcomes.More hostile domestic governments can more credibly threaten to walk away from an existing agreement. This raises the prospect of appropriating more of the surplus and the attractiveness of electing a government that is more intrinsically hostile to the project. But, when representatives are more hostile to the project than voters, the misaligned interests also raise the prospect that the 1017 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:04, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000400
Peter Buisseret and Dan Bernhardt date-two domestic government wants to terminate the of our results when voters can choose from a larger set project under conditions where voters want it to con- of political parties,or when voters cast ballots based on tinue.This raises the attractiveness of electing a more retrospective rather than prospective considerations, project-friendly government. or when parties can make limited commitments to their How voters resolve this tradeoff depends on date- negotiation strategies conditional on winning office one outcomes.Greater initial concessions by the for- Our model offers novel insights into how domestic eign government mitigate desires of domestic voters to politics affect international negotiations.First,demo- appoint a radical date-two government to extract even cratic governments should be most successful in ex- more.Instead,voters favor electing a government that tracting concessions from negotiating partners when is more likely to maintain the project.But if voters be- elections are imminent.This finding is consistent with lieve that the foreign government would be willing to evidence in Rickard and Caraway (2014)that labor offer far more concessions than are on the table,they market reforms demanded in exchange for IMF financ- prefer a more hostile government-regardless of their ing are less stringent for loans negotiated within six primitive preferences over the project,voters share a months of a pending election.Second,hawkish govern- common desire to extract as much surplus as possible ments that are the most ideologically opposed to in- from the foreign government.Thus,initial negotiations ternational agreements have electoral incentives to se- are both affected by,and partly determine,electoral cure less generous deals.A forward-looking electorate outcomes and subsequent negotiation outcomes. responds to a favorable status quo by appointing less risky governments that are more likely to preserve Main Results.Our main findings are as follows.If the it-that is,more project-friendly parties.So,a hawkish domestic government is initially relatively friendly,ini- incumbent that uses its leverage to secure better agree- tial agreements are signed whenever the static surplus ments hastens its departure from office!This may pro- between the foreign and friendly government is pos- vide insight into why,despite Syriza's failure to negoti- itive and,when the static surplus is instead negative, ate more favorable terms from the Troika,it retained agreements are signed if and only if elections are not its position as the largest parliamentary party in the too far off.The reason is that the governments'static subsequent election. conflicts of interest are attenuated by a dynamic con fluence of interest that is heightened by proximity to Contribution.The two crucial features of our frame- an election:both governments value more generous work are that (1)agreements are negotiated both be- standing agreements that encourage voters to return fore and after a domestic election and(2)the outcome the friendly party to office.This common interest may of pre-election negotiations determines the standing lead to even more generous offers by the foreign gov- offer in any subsequent negotiation. ernment than are needed to secure the friendly govern- Our focus on elections and renegotiation departs ment's participation.Thus,national elections not only from the "two-level games"framework developed by raise the prospect of agreements,but redirect surplus Putnam(1988)and explored in a vast body of work that away from the foreign government and toward the do- includes lida(1993).Mo(1995).Milner and Rosendorff mestic government. (1997)and Tarar(2001).The premise of this frame- If,instead,the domestic government is initially rel- work is that an agreement negotiated between a do atively hostile,agreements are never signed when the mestic and foreign government requires further do- static surplus is negative and,when the static surplus is mestic (i.e.,legislative)ratification;however,these positive,agreements are signed if and only if elections models do not allow for renegotiation once an initial are not too close.The reason is that the governments' bargain is struck.While important,this framework has static conflicts of interest are exacerbated by a dynamic its limitations.First.many countries do not have rat- conflict of interest that is heightened by proximity to an ification requirements.Second.executives often find election:more generous transfers harm the relatively ways to evade them-for example,via executive agree- hostile incumbent by reducing the prospect that it re- ments and nonbinding commitments in the United tains power,since voters then favor a friendly future States that are not subject to congressional approval.? government that will preserve the agreement.Finally, Third,regardless of ratification requirements,the possi- whenever an agreement is signed,the foreign govern- bility for future renegotiation remains relevant.Our re- ment appropriates all of the surplus from agreement. sults highlight how electoral concerns(both via policy More generally,dynamic considerations have a po and office motivations)drive negotiation outcomes- larizing effect on initial negotiations:static conflicts be- considerations missed by models that focus solely on tween the national and foreign government are mag- ratification.8 nified by other conflicts,including (1)policy and rent- seeking conflicts between the domestic political parties (2)policy conflicts between the parties and the elec For example,in the UK,the House of Commons may delay ratifi- cation of certain treaties,but there is no mechanism for Parliament torate,and (3)the policy conflict between the foreign to scrutinize,debate,or vote on treaties;see Lang (2017). government and the electorate.We show how changes For example,the Joint Comprehensive Plan of Action;see Mulligan in the project valuations of the domestic parties may (2017)for an introduction to the forms of international agreements drive more or less generous agreements,depending on and commitments in the US context The distinction between elections and ratification matters:a rati. the uncertainty about domestic voters'attitudes to- fier chooses between accepting an international agreement and pre- wards the project.Finally,we examine the robustness serving the status quo;voter choices reflect their induced preferences 1018
Peter Buisseret and Dan Bernhardt date-two domestic government wants to terminate the project under conditions where voters want it to continue. This raises the attractiveness of electing a more project-friendly government. How voters resolve this tradeoff depends on dateone outcomes. Greater initial concessions by the foreign government mitigate desires of domestic voters to appoint a radical date-two government to extract even more. Instead, voters favor electing a government that is more likely to maintain the project. But if voters believe that the foreign government would be willing to offer far more concessions than are on the table, they prefer a more hostile government—regardless of their primitive preferences over the project, voters share a common desire to extract as much surplus as possible from the foreign government. Thus, initial negotiations are both affected by, and partly determine, electoral outcomes and subsequent negotiation outcomes. Main Results. Our main findings are as follows. If the domestic government is initially relatively friendly, initial agreements are signed whenever the static surplus between the foreign and friendly government is positive and, when the static surplus is instead negative, agreements are signed if and only if elections are not too far off. The reason is that the governments’ static conflicts of interest are attenuated by a dynamic confluence of interest that is heightened by proximity to an election: both governments value more generous standing agreements that encourage voters to return the friendly party to office. This common interest may lead to even more generous offers by the foreign government than are needed to secure the friendly government’s participation. Thus, national elections not only raise the prospect of agreements, but redirect surplus away from the foreign government and toward the domestic government. If, instead, the domestic government is initially relatively hostile, agreements are never signed when the static surplus is negative and, when the static surplus is positive, agreements are signed if and only if elections are not too close. The reason is that the governments’ static conflicts of interest are exacerbated by a dynamic conflict of interest that is heightened by proximity to an election: more generous transfers harm the relatively hostile incumbent by reducing the prospect that it retains power, since voters then favor a friendly future government that will preserve the agreement. Finally, whenever an agreement is signed, the foreign government appropriates all of the surplus from agreement. More generally, dynamic considerations have a polarizing effect on initial negotiations: static conflicts between the national and foreign government are magnified by other conflicts, including (1) policy and rentseeking conflicts between the domestic political parties, (2) policy conflicts between the parties and the electorate, and (3) the policy conflict between the foreign government and the electorate. We show how changes in the project valuations of the domestic parties may drive more or less generous agreements, depending on the uncertainty about domestic voters’ attitudes towards the project. Finally, we examine the robustness of our results when voters can choose from a larger set of political parties, or when voters cast ballots based on retrospective rather than prospective considerations, or when parties can make limited commitments to their negotiation strategies conditional on winning office. Our model offers novel insights into how domestic politics affect international negotiations. First, democratic governments should be most successful in extracting concessions from negotiating partners when elections are imminent. This finding is consistent with evidence in Rickard and Caraway (2014) that labor market reforms demanded in exchange for IMF financing are less stringent for loans negotiated within six months of a pending election. Second, hawkish governments that are the most ideologically opposed to international agreements have electoral incentives to secure less generous deals. A forward-looking electorate responds to a favorable status quo by appointing less risky governments that are more likely to preserve it—that is, more project-friendly parties. So, a hawkish incumbent that uses its leverage to secure better agreements hastens its departure from office! This may provide insight into why, despite Syriza’s failure to negotiate more favorable terms from the Troika, it retained its position as the largest parliamentary party in the subsequent election. Contribution. The two crucial features of our framework are that (1) agreements are negotiated both before and after a domestic election and (2) the outcome of pre-election negotiations determines the standing offer in any subsequent negotiation. Our focus on elections and renegotiation departs from the “two-level games” framework developed by Putnam (1988) and explored in a vast body of work that includes Iida (1993),Mo (1995),Milner and Rosendorff (1997) and Tarar (2001). The premise of this framework is that an agreement negotiated between a domestic and foreign government requires further domestic (i.e., legislative) ratification; however, these models do not allow for renegotiation once an initial bargain is struck. While important, this framework has its limitations. First, many countries do not have ratification requirements.6 Second, executives often find ways to evade them—for example, via executive agreements and nonbinding commitments in the United States that are not subject to congressional approval.7 Third, regardless of ratification requirements, the possibility for future renegotiation remains relevant.Our results highlight how electoral concerns (both via policy and office motivations) drive negotiation outcomes— considerations missed by models that focus solely on ratification.8 6 For example, in the UK, the House of Commons may delay ratification of certain treaties, but there is no mechanism for Parliament to scrutinize, debate, or vote on treaties; see Lang (2017). 7 For example, the Joint Comprehensive Plan of Action; see Mulligan (2017) for an introduction to the forms of international agreements and commitments in the US context. 8 The distinction between elections and ratification matters: a ratifier chooses between accepting an international agreement and preserving the status quo; voter choices reflect their induced preferences 1018 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:04, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000400
Reelection and Renegotiation Electoral considerations may generate starkly differ- this reduces the foreign government's desire to insure ent implications for negotiation outcomes than those itself against future renegotiation.Both of these forces arising when the only relevant consideration is ratifica- encourage the foreign government to make less gener- tion.To illustrate this,consider a country that has a rat- ous concessions.Thus,the key prediction of the ratifica- ification requirement,and suppose that the relatively tion literature- -commonly referred to as the Schelling hostile domestic party holds ratification authority,but Conjecture (Milner 1997)-is entirely reversed in a the friendly party conducts initial negotiations with the context where electoral considerations are paramount. foreign government.This could arise with a divided Static models of intergovernmental negotiations government in which the friendly party holds executive generate predictions about the induced preferences authority,but the hostile party controls the legislative of voters over their representative (e.g.,government upper chamber.If the relatively more hostile party's or legislator)in subsequent negotiations,including preferences regarding international cooperation grow Persson and Tabellini (1992).Besley and Coate(2003) even more hostile-for example,its party leadership Buchholz,Haupt,and Peters (2005),and Harstad becomes more opposed to free trade or international (2008).In our framework,however,negotiations take environmental cooperation-does this benefit or im- place both before and after elections (i.e.,delegation pair the relatively friendly party's ability to extract con- decisions):and both governments'negotiating strate- cessions from the foreign government in its initial ne- gies account for the consequences of today's agree- gotiations? ment for subsequent election outcomes, generating The enduring prediction of the two-level games liter- strategic considerations that are absent in these static ature is that this change in domestic politics raises the models. friendly party's transfers from the foreign government There is a small literature on dynamic negotiations The reason is that the hostile party's increased intrinsic with interim elections.Wolford(2012)assumes that a hostility makes it more prone to refuse ratification.This domestic government is more likely to win reelection encourages the foreign government to make more gen when the share of the surplus it extracts from a for- 4号元 erous concessions to the friendly domestic government eign government in pre-election negotiations rises.We to sway the more hostile ratifier(Schelling 1980). show that when pre-election negotiations determine In our setting,by contrast,the friendly domestic gov the standing offer in subsequent negotiations,and thus ernment and foreign government anticipate the incum- voters'induced preferences,a relatively hostile incum- bent's threat of replacement by the relatively hostile bent suffers a lower prospect of reelection when it se- party in an election.If the hostile party is elected cures greater surplus.In Schultz(2005).an incumbent's the foreign government further anticipates a higher decision to“cooperate'”or“defect'”informs domestic risk that any agreement signed today may be renego- voters about his preferences,informing retention deci- tiated on less favorable terms to the foreign govern sions.However,there is no intergovernmental bargain- ment.More generous initial transfers (1)steer voters ing,and thus no mechanism to distribute any joint gains induced preferences in favor of the friendly govern- from cooperation (e.g.,transfers).In contrast,our fo- ment and(2)insure the foreign government against cus is precisely on how changes in preferences and un- renegotiation by either subsequent domestic govern- certainty affect the distribution of surplus from coop- ment,since more generous standing offers lower the eration across governments.10 Battaglini and Harstad prospect that a future domestic government will prefer (2016)show how an incumbent party might commit to to quit at the standing offer. inefficiently low sanctions (a"weak treaty")to differ- We derive the necessary and sufficient conditions un- entiate itself electorally from a challenger. 8 der which a more hostile opposition party lowers the Smith and Hayes(1997)also study a setting in which foreign government's transfers to the friendly party. countries may renegotiate an inherited pre-election To see why transfers may fall,notice that as the hos- agreement.In their model,governments bargain over tile party grows more extreme,it becomes less elec- spatial policies,rather than transfers as in our set- torally competitive,since more voters prefer the an- ting.They characterize renegotiation outcomes after ticipated negotiating position of the relatively friendly an election for a given inherited status quo and high- party.This has two implications.First,the foreign gov- light some properties that the status quo must satisfy ernment faces less urgency from using more generous if it is derived from pre-election negotiations.How- transfers to steer voters'induced preferences in favor ever,in contrast to our analysis,they do not charac- of reelecting the friendly party.Second,at any level of terize pre-election agreements,conditions on primi- 四 transfers-and thus at any future standing offer-the tives under which pre-election agreements are reached, friendly government is more likely to be reelected,in or how the surplus from agreement is divided across which case the post-election domestic government is governments. more likely to want to maintain any existing agreement; An empirical literature dating back to Thucydides studies the efficacy of internal democracy for foreign over the anticipated bargaining outcomes that their representatives policy commitments(see History of the Peloponnesian will achieve after the election.Once authority is delegated,voters no War. 8.70.1-2.)Recent researchers variously argue that longer influence negotiation outcomes and cannot trigger a reversion to an outside option. 士 9 The intuition is analogous to a Calvert-Wittman framework in 10 In Schneider and Slantchev(2018).governments are privately in which two parties offer differentiated platforms,and one party moves formed about the gains to cooperation,and the division of any sur- its platform further away from the other's. plus from an agreement is exogenous. 1019
Reelection and Renegotiation Electoral considerations may generate starkly different implications for negotiation outcomes than those arising when the only relevant consideration is ratification. To illustrate this, consider a country that has a ratification requirement, and suppose that the relatively hostile domestic party holds ratification authority, but the friendly party conducts initial negotiations with the foreign government. This could arise with a divided government in which the friendly party holds executive authority, but the hostile party controls the legislative upper chamber. If the relatively more hostile party’s preferences regarding international cooperation grow even more hostile—for example, its party leadership becomes more opposed to free trade or international environmental cooperation—does this benefit or impair the relatively friendly party’s ability to extract concessions from the foreign government in its initial negotiations? The enduring prediction of the two-level games literature is that this change in domestic politics raises the friendly party’s transfers from the foreign government. The reason is that the hostile party’s increased intrinsic hostility makes it more prone to refuse ratification.This encourages the foreign government to make more generous concessions to the friendly domestic government to sway the more hostile ratifier (Schelling 1980). In our setting, by contrast, the friendly domestic government and foreign government anticipate the incumbent’s threat of replacement by the relatively hostile party in an election. If the hostile party is elected, the foreign government further anticipates a higher risk that any agreement signed today may be renegotiated on less favorable terms to the foreign government. More generous initial transfers (1) steer voters’ induced preferences in favor of the friendly government and (2) insure the foreign government against renegotiation by either subsequent domestic government, since more generous standing offers lower the prospect that a future domestic government will prefer to quit at the standing offer. We derive the necessary and sufficient conditions under which a more hostile opposition party lowers the foreign government’s transfers to the friendly party. To see why transfers may fall, notice that as the hostile party grows more extreme, it becomes less electorally competitive, since more voters prefer the anticipated negotiating position of the relatively friendly party.9 This has two implications. First, the foreign government faces less urgency from using more generous transfers to steer voters’ induced preferences in favor of reelecting the friendly party. Second, at any level of transfers—and thus at any future standing offer—the friendly government is more likely to be reelected, in which case the post-election domestic government is more likely to want to maintain any existing agreement; over the anticipated bargaining outcomes that their representatives will achieve after the election. Once authority is delegated, voters no longer influence negotiation outcomes and cannot trigger a reversion to an outside option. 9 The intuition is analogous to a Calvert-Wittman framework in which two parties offer differentiated platforms, and one party moves its platform further away from the other’s. this reduces the foreign government’s desire to insure itself against future renegotiation. Both of these forces encourage the foreign government to make less generous concessions.Thus, the key prediction of the ratification literature—commonly referred to as the Schelling Conjecture (Milner 1997)—is entirely reversed in a context where electoral considerations are paramount. Static models of intergovernmental negotiations generate predictions about the induced preferences of voters over their representative (e.g., government or legislator) in subsequent negotiations, including Persson and Tabellini (1992), Besley and Coate (2003), Buchholz, Haupt, and Peters (2005), and Harstad (2008). In our framework, however, negotiations take place both before and after elections (i.e., delegation decisions); and both governments’ negotiating strategies account for the consequences of today’s agreement for subsequent election outcomes, generating strategic considerations that are absent in these static models. There is a small literature on dynamic negotiations with interim elections. Wolford (2012) assumes that a domestic government is more likely to win reelection when the share of the surplus it extracts from a foreign government in pre-election negotiations rises. We show that when pre-election negotiations determine the standing offer in subsequent negotiations, and thus voters’ induced preferences, a relatively hostile incumbent suffers a lower prospect of reelection when it secures greater surplus. In Schultz (2005), an incumbent’s decision to “cooperate” or “defect” informs domestic voters about his preferences, informing retention decisions. However, there is no intergovernmental bargaining, and thus no mechanism to distribute any joint gains from cooperation (e.g., transfers). In contrast, our focus is precisely on how changes in preferences and uncertainty affect the distribution of surplus from cooperation across governments.10 Battaglini and Harstad (2016) show how an incumbent party might commit to inefficiently low sanctions (a “weak treaty”) to differentiate itself electorally from a challenger. Smith and Hayes (1997) also study a setting in which countries may renegotiate an inherited pre-election agreement. In their model, governments bargain over spatial policies, rather than transfers as in our setting. They characterize renegotiation outcomes after an election for a given inherited status quo and highlight some properties that the status quo must satisfy if it is derived from pre-election negotiations. However, in contrast to our analysis, they do not characterize pre-election agreements, conditions on primitives under which pre-election agreements are reached, or how the surplus from agreement is divided across governments. An empirical literature dating back to Thucydides studies the efficacy of internal democracy for foreign policy commitments (see History of the Peloponnesian War, 8.70.1-2.) Recent researchers variously argue that 10 In Schneider and Slantchev (2018), governments are privately informed about the gains to cooperation, and the division of any surplus from an agreement is exogenous. 1019 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:04, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000400
Peter Buisseret and Dan Bernhardt the prospect of leader turnover raises (Gartzke and identity of the political party that holds power.We Gleditsch 2004)or reduces (Leeds,Mattes,and Vogel consider a two-party setting that features a relatively 2009;Gaubatz 1996;Leeds and Savun 2007)a govern- friendly party with date-one valuation 7,and a rela- ment's propensity to renegotiate or exit agreements. tively hostile party with date-one valuation v.These Others hold intermediate views.closer to ours.that the project valuations can be interpreted as flow payoffs degree of commitment is endogenous to the form of enjoyed at each date from the moment that the agree- the initial agreement(Lipson 1991;Abbott and Snidal ment is signed.If the project is not undertaken at date 1998:Rosendorff and Milner 2001).The empirical fre t,each agent receives a date-t payoff that we normalize quency and causes of treaty renegotiation have been to zero.All project valuations are common knowledge. subject to debate (e.g.,Downs,Rocke,and Barsoom Assumption 1 sets out the structure that the FG de- op//s 1996).We provide insights into how and when durable rives a higher value from the project than the relatively treaties are signed,and how this depends on the pref- friendly government,which,in turn,derives a higher erences of negotiating governments and the domestic value from the project than the relatively hostile party. political context. The idea that today's policies commit future Assumption1:vF>元>u. governments-and that such commitments can be All agents weight date-one payoffs by 1-8(0,1) used to manipulate electoral preferences-is well es- and date-two payoffs by 8.For example,1 -8 could tablished,for example in Alesina and Tabellini (1990), represent the time between the initial signing and the Milesi-Ferretti and Spolaore (1994),and Persson next election:when 8 is large,negotiations take place and Svensson (1989).In our setting,however,the relatively close to the election,after which there will degree of commitment itself is entirely endogenous.In be an opportunity to renegotiate the initial agreement. particular,initial negotiation outcomes change neither At the outset of negotiations,participation by DG the technology available to future governments, in the project with FG implies a transfer sER from nor their primitive valuation from post-election FG to DG1.In the EU accession example,s1>0 could 4号元 participation in the project.Finally,we contribute to represent a standard package of benefits,such as tariff a literature on dynamic political economy in which reductions or a share of regional development funds today's policy outcome serves as the reversion in that is awarded to a new member state upon joining.By subsequent negotiations (Dziuda and Loeper 2016: contrast,s10. We allow for negotiations between the countries in which FG encourages DGr to participate by offering MODEL more favorable terms.These negotiations unfold as fol- lows.At date one,FG is the proposer,and DGi is the Our two-date economy features two countries,a for- receiver.1 FG makes an initial offer b>s1,which is eign government (FG),and a date-t domestic govern- ment(DG,).FG can be interpreted either as an indi- a concession that it will give to DGi if and only if it participates in the agreement at that date.12 In the EU vidual government or a group of governments such as accession example,bi could represent additional con- the European Union,or an international organization cessions and carveouts on labor market or financial such as the International Monetary Fund.There is a sector regulations,budget contributions,or a more gen- project that the governments can undertake at each of erous share of regional development funds.After re- dates one and two;r,=1 indicates that the project is ceiving the offer b1,DGI chooses ri(b1)E(0,1),where undertaken at date t,and r=0 indicates that it is not. The project could represent the domestic country's ac- cession to an international organization such as the EU In the Supplemental Appendix,we show that our results extend the launch of a common currency,a climate agreement, or a region's participation in a federation or national prooinore generous union. terms than s1.This restriction is without loss of generality under many mild restrictions,for example that s+o +y is not too large At both dates,the project generates a value vF for and the likelihood that the median domestic voter places a very high FG.The value of the project to DG:depends on the value on the project is not too high. 1020
Peter Buisseret and Dan Bernhardt the prospect of leader turnover raises (Gartzke and Gleditsch 2004) or reduces (Leeds, Mattes, and Vogel 2009; Gaubatz 1996; Leeds and Savun 2007) a government’s propensity to renegotiate or exit agreements. Others hold intermediate views, closer to ours, that the degree of commitment is endogenous to the form of the initial agreement (Lipson 1991; Abbott and Snidal 1998; Rosendorff and Milner 2001). The empirical frequency and causes of treaty renegotiation have been subject to debate (e.g., Downs, Rocke, and Barsoom 1996). We provide insights into how and when durable treaties are signed, and how this depends on the preferences of negotiating governments and the domestic political context. The idea that today’s policies commit future governments—and that such commitments can be used to manipulate electoral preferences—is well established, for example in Alesina and Tabellini (1990), Milesi-Ferretti and Spolaore (1994), and Persson and Svensson (1989). In our setting, however, the degree of commitment itself is entirely endogenous. In particular, initial negotiation outcomes change neither the technology available to future governments, nor their primitive valuation from post-election participation in the project. Finally, we contribute to a literature on dynamic political economy in which today’s policy outcome serves as the reversion in subsequent negotiations (Dziuda and Loeper 2016; Acemoglu, Egorov, and Sonin 2014). The outline of this paper is as follows. We present our base model, analyzing a setting in which the uncertainty over who will hold future domestic political power does not hinge on the initial negotiation between the foreign and domestic government. We then consider endogenous elections, showing how the answers to our motivating questions change radically. We show how offers vary with primitives such as the intrinsic valuations that the domestic parties place on the project, as well as uncertainty about voters’ preferences. We then summarize extensions that are fully analyzed in the Supplemental Appendix. A conclusion follows. Proofs are in the Appendix. MODEL Our two-date economy features two countries, a foreign government (FG), and a date-t domestic government (DGt). FG can be interpreted either as an individual government or a group of governments such as the European Union, or an international organization such as the International Monetary Fund. There is a project that the governments can undertake at each of dates one and two; rt = 1 indicates that the project is undertaken at date t, and rt = 0 indicates that it is not. The project could represent the domestic country’s accession to an international organization such as the EU, the launch of a common currency, a climate agreement, or a region’s participation in a federation or national union. At both dates, the project generates a value vF for FG. The value of the project to DGt depends on the identity of the political party that holds power. We consider a two-party setting that features a relatively friendly party with date-one valuation v, and a relatively hostile party with date-one valuation v. These project valuations can be interpreted as flow payoffs enjoyed at each date from the moment that the agreement is signed. If the project is not undertaken at date t, each agent receives a date-t payoff that we normalize to zero. All project valuations are common knowledge. Assumption 1 sets out the structure that the FG derives a higher value from the project than the relatively friendly government, which, in turn, derives a higher value from the project than the relatively hostile party. Assumption 1: vF > v > v. All agents weight date-one payoffs by 1 − δ ∈ (0, 1) and date-two payoffs by δ. For example, 1 − δ could represent the time between the initial signing and the next election: when δ is large, negotiations take place relatively close to the election, after which there will be an opportunity to renegotiate the initial agreement. At the outset of negotiations, participation by DG1 in the project with FG implies a transfer s1 ∈ R from FG to DG1. In the EU accession example,s1 ≥ 0 could represent a standard package of benefits, such as tariff reductions or a share of regional development funds that is awarded to a new member state upon joining. By contrast, s1 0. We allow for negotiations between the countries in which FG encourages DGt to participate by offering more favorable terms. These negotiations unfold as follows. At date one, FG is the proposer, and DG1 is the receiver. 11 FG makes an initial offer b1 ≥ s1, which is a concession that it will give to DG1 if and only if it participates in the agreement at that date.12 In the EU accession example, b1 could represent additional concessions and carveouts on labor market or financial sector regulations, budget contributions, or a more generous share of regional development funds. After receiving the offer b1, DG1 chooses r1(b1) ∈ {0, 1}, where 11 In the Supplemental Appendix, we show that our results extend when DG1 is instead the proposer. 12 Throughout, we restrict FG to proposing weakly more generous terms than s1. This restriction is without loss of generality under many mild restrictions, for example that s1 + σ + v is not too large and the likelihood that the median domestic voter places a very high value on the project is not too high. 1020 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:04, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000400
Reelection and Renegotiation r(b)=1 indicates that the project is implemented at ing power or institutional features of the agreement date one and r(b)=0 indicates that it is not. that determine who can initiate renegotiations.We al- Between dates one and two,the date-one domes- low for arbitrary [0,1]to emphasize that results do tic government DG may be replaced by a new do- not depend sensitively on the distribution of future bar- mestic government DG2,according to a process that gaining power.3 The agent realized as proposer at date we describe below.After DG2 is realized.all domestic two can propose a new transfer,b2ER.If the date-two agents are hit by a common additive preference shock receiver accepts,this becomes the new date-two trans- A to the payoffs they derive from the project.We as- fer.Otherwise,the inherited terms from past negotia- sume that this publicly observed preference shock is tions remain in force,so that b2=s2.Next,DG2 decides drawn from a uniform distribution with support [ whether to quit the agreement and receive its outside o,o].This shock can capture an unanticipated wors- option of zero or to execute the agreement given the ening of the economy-unemployment may increase, date-two terms.FG then makes the agreed-upon trans- labor unions may organize industrial unrest,or there fer if and only if DG2 executes the agreement by im- may be civil unrest.Alternatively,new information may plementing the project come to light.For example,in 2004,an audit by the in- The expected lifetime payoff of a domestic agent coming Greek government found that,under a previ- with date-one project valuation v is ous PASOK administration,the government's statistics agency had misreported the country's debt and deficit (1-8)rn1(v+b1) figures to qualify for entry into the European single currency. Pr(v 2(v+b2+)f()d We first assume that date-one negotiations do not ∈{里， affect domestic election outcomes.Thus.DG2 is rela- tively hostile with exogenous probability Pr(v)E[0,1], and relatively friendly with probability Pr()=1- where f(A)is the density of the domestic preference 4号 Pr(v).This captures a benchmark in which the elec- shock,A.Here r(0,1]is the date-one domestic tion outcome is insensitive to the negotiation outcome government's initial decision to implement the project 'asn We later endogenize DG2's project valuation via an (r1=1)or not (r1 =0),r2 (0,1]denotes the project election.where electoral outcomes may depend on(1) outcome at date two,and b2 denotes the date-two whether the project was implemented at date one,and transfer from FG when the project is implemented the terms of the initial bargain;(2)how voters make at date two;that is,when r2=1.Note that domestic voting decisions (prospectively or retrospectively):and agents care about date-two policy outcomes regardless (3)the set of feasible replacements.We assume that of who holds office at that date.In addition to deriving there is sufficient variation in the domestic preference project-related payoffs like any other domestic agent, shockλ： we assume that each domestic political party derives an office-holding benefit of w>0 at any date that it holds office Assumption 3:UF+v-0 The analogous expected payoff of FG with project Assumption 3 says that there is enough uncertainty valuation vF is about the common domestic preference shock A that (a)it could exceed the expected surplus from the (1-8)rn(vF-b1) project between FG and the relatively project-friendly DG2 with valuation and (b)it could be even lower Pr(v r2(uF-b2)f(入)dλ than the expected value for the relatively hostile DG2 with valuation v from participating in the project at the initial standing offer,s1. One may observe that FG's project valuation does not After A is realized,the initial terms for the project evolve over time.This assumption eases presentation can be renegotiated,or if agreement was not reached at and analysis,allowing us to focus on the effects of un- date one,the governments can try again.The inherited certainty about DG2's valuation v.One can also in- date-one terms serve as the reversion point s2 for date- terpret the FG as the IMF or the World Bank,whose two bargaining.Thus,if the project was implemented at leadership is not expected to change over the course of date one with transfer b,the status-quo transfer is s2 negotiations. b;this transfer will be made at date two if the project is again implemented and new terms are not agreed upon For example,Thatcher's renegotiation of Britain's EU POLICY OUTCOMES AT DATE TWO budget rebate persisted from 1984 until 2005.If,in- We start by analyzing the long-term consequences of stead,the project was not implemented at date one, date-one outcomes.If the project was implemented at then the status quo transfer(i.e.,starting point for date- two negotiations in which the governments try again) iS S2=S1. 13 While does not play a key role,scholars have still considered With probability 0 [0,1],DG2 proposes the new how features of international institutions-for example,renegotia- tion protocols-might be chosen to maximize the prospect that an terms,and with probability 1-6 the FG makes the pro- agreement survives.See Koremenos,Lipson,and Snidal (2001).or posal.The parameter 6 could reflect intrinsic bargain- Koremenos (2001). 1021
Reelection and Renegotiation r1(b1) = 1 indicates that the project is implemented at date one and r1(b1) = 0 indicates that it is not. Between dates one and two, the date-one domestic government DG1 may be replaced by a new domestic government DG2, according to a process that we describe below. After DG2 is realized, all domestic agents are hit by a common additive preference shock λ to the payoffs they derive from the project. We assume that this publicly observed preference shock is drawn from a uniform distribution with support [ − σ, σ]. This shock can capture an unanticipated worsening of the economy—unemployment may increase, labor unions may organize industrial unrest, or there may be civil unrest.Alternatively, new information may come to light. For example, in 2004, an audit by the incoming Greek government found that, under a previous PASOK administration, the government’s statistics agency had misreported the country’s debt and deficit figures to qualify for entry into the European single currency. We first assume that date-one negotiations do not affect domestic election outcomes. Thus, DG2 is relatively hostile with exogenous probability Pr(v) ∈ [0, 1], and relatively friendly with probability Pr(v) = 1 − Pr(v). This captures a benchmark in which the election outcome is insensitive to the negotiation outcome. We later endogenize DG2’s project valuation via an election, where electoral outcomes may depend on (1) whether the project was implemented at date one, and the terms of the initial bargain; (2) how voters make voting decisions (prospectively or retrospectively); and (3) the set of feasible replacements. We assume that there is sufficient variation in the domestic preference shock λ: Assumption 3: vF + v −σ. Assumption 3 says that there is enough uncertainty about the common domestic preference shock λ that (a) it could exceed the expected surplus from the project between FG and the relatively project-friendly DG2 with valuation v; and (b) it could be even lower than the expected value for the relatively hostile DG2 with valuation v from participating in the project at the initial standing offer, s1. After λ is realized, the initial terms for the project can be renegotiated, or if agreement was not reached at date one, the governments can try again. The inherited date-one terms serve as the reversion point s2 for datetwo bargaining. Thus,if the project was implemented at date one with transfer b1, the status-quo transfer is s2 = b1; this transfer will be made at date two if the project is again implemented and new terms are not agreed upon. For example, Thatcher’s renegotiation of Britain’s EU budget rebate persisted from 1984 until 2005. If, instead, the project was not implemented at date one, then the status quo transfer (i.e., starting point for datetwo negotiations in which the governments try again) is s2 = s1. With probability θ ∈ [0, 1], DG2 proposes the new terms, and with probability 1 − θ the FG makes the proposal. The parameter θ could reflect intrinsic bargaining power or institutional features of the agreement that determine who can initiate renegotiations. We allow for arbitrary θ ∈ [0, 1] to emphasize that results do not depend sensitively on the distribution of future bargaining power.13 The agent realized as proposer at date two can propose a new transfer, b2 ∈ R. If the date-two receiver accepts, this becomes the new date-two transfer. Otherwise, the inherited terms from past negotiations remain in force, so that b2 = s2. Next,DG2 decides whether to quit the agreement and receive its outside option of zero or to execute the agreement given the date-two terms. FG then makes the agreed-upon transfer if and only if DG2 executes the agreement by implementing the project. The expected lifetime payoff of a domestic agent with date-one project valuation v is (1 − δ)r1(v + b1 ) + δ v ∈{v,v} Pr(v ) σ −σ r2(v + b2 + λ)f(λ) dλ, where f(λ) is the density of the domestic preference shock, λ. Here r1 ∈ {0, 1} is the date-one domestic government’s initial decision to implement the project (r1 = 1) or not (r1 = 0), r2 ∈ {0, 1} denotes the project outcome at date two, and b2 denotes the date-two transfer from FG when the project is implemented at date two; that is, when r2 = 1. Note that domestic agents care about date-two policy outcomes regardless of who holds office at that date. In addition to deriving project-related payoffs like any other domestic agent, we assume that each domestic political party derives an office-holding benefit of w > 0 at any date that it holds office. The analogous expected payoff of FG with project valuation vF is (1 − δ)r1(vF − b1 ) + δ v ∈{v,v} Pr(v ) σ −σ r2(vF − b2 )f(λ) dλ. One may observe that FG’s project valuation does not evolve over time. This assumption eases presentation and analysis, allowing us to focus on the effects of uncertainty about DG2’s valuation v2 D. One can also interpret the FG as the IMF or the World Bank, whose leadership is not expected to change over the course of negotiations. POLICY OUTCOMES AT DATE TWO We start by analyzing the long-term consequences of date-one outcomes. If the project was implemented at 13 While θ does not play a key role, scholars have still considered how features of international institutions—for example, renegotiation protocols—might be chosen to maximize the prospect that an agreement survives. See Koremenos, Lipson, and Snidal (2001), or Koremenos (2001). 1021 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:04, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000400
Peter Buisseret and Dan Bernhardt date one,that is,if r =1,then the status quo transfer s2 Finally,if the date-two surplus from agreement is is the transfer b that DG accepted.If the project was negative;that is,if Equation(1)does not hold,then no not implemented,that is,if r =0,then the status quo amendment will be agreed upon,as the joint surplus transfer that serves as the starting point for date-two from implementing the project is negative.The project negotiations is s2 =s1. will not be implemented and all agents receive date- Because there are no bargaining frictions,the project one payoffs of zero. will be implemented at the terminal date t =2 if and The expected date-two payoff of a domestic agent only if the associated surplus is positive;that is,if and with date-one project valuation v who anticipates that only if the DG2 will have project valuation vp and face status quo transfer s2 is thus v2+1+vF≥0←→入≥-(2+UF) (1) VD(v,vp,s2)= (v+s2+)f()d Even though the date-two implementation decision -(+) does not depend on date-one actions,the division of (听+s2 the surplus depends on (a)the status quo transfer and (v-v哈+(u哈+入+vF)f()d. (b)the shock realization A. Suppose,first,that DG2 has a high enough project (3) valuation p+A that it would receive a positive pay- off from implementing the project when it receives the The expected date-two project payoff of FG given s2 status-quo transfer s2: when it faces DG2 with valuation v is 令 2+1+52≥0←→入≥-(u哈+s2) (2) Vr(2,s2) (vF-s2)f()d -(听+s2) With probability 6,DG2 is recognized to propose a modification to the inherited terms,s2.Because DG2 (1-)(2+入+vF)f(2)d.(4) & prefers higher transfers,it never proposes a transfer (vD+UF) b2s2 will fail:if Equation (2)holds,FG recognizes A transfer of power from a friendly date-one domestic that DG2 will implement the project even if the initial government DG to a more hostile date-two domestic agreement is not amended.As a result,FG would reject government DG2(i.e.,from to v)carries two implica- the amendment,because a threat by DG,to renege on tions.First,it increases the prospect that DG2 can rene- the inherited agreement is not credible.With residual gotiate the initial terms to a more favorable arrange- probability 1-0,FG gets to propose a modification. ment.Second,it lowers the total surplus of the date-two Although FG would like to negotiate a reduced trans- negotiating parties.As a result,there will be situations fer,DG2 will refuse such amendments-it prefers to in which a hostile DG2 will fail to reach an agreement maintain the existing terms,which offer more favorable with FG in contexts where a more project-friendly DG2 concessions in return for implementing the project. would have successfully concluded the negotiation. Suppose,instead,that DG2 anticipates a negative value from implementing the project at the status-quo Discussion:The bargaining protocol is starker than transfer;that is,Equation(2)fails.This means that it necessary for our main results.What is crucial is that would prefer not to implement the project at date two the terms that the domestic government obtains at date unless the initial terms were amended to a higher trans- two improve as its valuation of the project falls,rela- fer.Suppose,first,that the surplus from agreement is tive to the status quo offer.This improvement in terms positive;that is,Equation(1)holds. holds regardless of the distribution of date-two bar- With probability 0,DG2 gets to propose a modifica- gaining power,6[0,1].When the domestic govern- tion to the inherited terms.If FG rejects the proposal, ment holds date-two proposal power,a more hostile the project will end when Equation(2)does not hold, representative can renegotiate the status quo transfer giving FG a payoff of zero.Thus,DG2 can renegoti- from s2 up to b2 Ur.When,instead,the FG holds pro- ate the date-two transfer from s2 to the larger transfer posal power,its offer holds the date-two domestic gov- b2=vF.That FG is held to its participation constraint ernment to its participation constraint,but its transfer is not essential-what matters is that there is a disconti- b2 =-(vp+)still increases as the domestic govern- nuity in the terms that DG2 can obtain when its threat ment becomes more hostile;that is,as vp decreases.A to break the existing agreement is credible;that is,at more hostile representative not only captures the up- the threshold on A defined in Equation(2).With prob- side of larger concessions-it also mitigates against the ability 1-0.FG is,instead,recognized.Since Equation downside of subsequent appropriation. (2)fails,FG must offer DG2 a larger transfer to secure its participation.It then raises the transfer from s2 to POLICY OUTCOMES AT DATE ONE b2 =-(vp+),leaving DG2 with value vp+A indif- ferent between implementing the project and quitting, Exogenous Power Transitions.In our benchmark set- allowing FG to claim the rest of the surplus for itself. ting,the date-two domestic government's (DG2's) 1022
Peter Buisseret and Dan Bernhardt date one, that is, if r1 = 1, then the status quo transfer s2 is the transfer b1 that DG1 accepted. If the project was not implemented, that is, if r1 = 0, then the status quo transfer that serves as the starting point for date-two negotiations is s2 = s1. Because there are no bargaining frictions, the project will be implemented at the terminal date t = 2 if and only if the associated surplus is positive; that is, if and only if v2 D + λ + vF ≥ 0 ⇐⇒ λ ≥ −(v2 D + vF ). (1) Even though the date-two implementation decision does not depend on date-one actions, the division of the surplus depends on (a) the status quo transfer and (b) the shock realization λ. Suppose, first, that DG2 has a high enough project valuation v2 D + λ that it would receive a positive payoff from implementing the project when it receives the status-quo transfer s2: v2 D + λ + s2 ≥ 0 ⇐⇒ λ ≥ −(v2 D + s2 ). (2) With probability θ, DG2 is recognized to propose a modification to the inherited terms, s2. Because DG2 prefers higher transfers, it never proposes a transfer b2 s2 will fail: if Equation (2) holds, FG recognizes that DG2 will implement the project even if the initial agreement is not amended.As a result, FG would reject the amendment, because a threat by DG2 to renege on the inherited agreement is not credible. With residual probability 1 − θ, FG gets to propose a modification. Although FG would like to negotiate a reduced transfer, DG2 will refuse such amendments—it prefers to maintain the existing terms, which offer more favorable concessions in return for implementing the project. Suppose, instead, that DG2 anticipates a negative value from implementing the project at the status-quo transfer; that is, Equation (2) fails. This means that it would prefer not to implement the project at date two unless the initial terms were amended to a higher transfer. Suppose, first, that the surplus from agreement is positive; that is, Equation (1) holds. With probability θ, DG2 gets to propose a modification to the inherited terms. If FG rejects the proposal, the project will end when Equation (2) does not hold, giving FG a payoff of zero. Thus, DG2 can renegotiate the date-two transfer from s2 to the larger transfer b2 = vF . That FG is held to its participation constraint is not essential—what matters is that there is a discontinuity in the terms that DG2 can obtain when its threat to break the existing agreement is credible; that is, at the threshold on λ defined in Equation (2). With probability 1 − θ, FG is, instead, recognized. Since Equation (2) fails, FG must offer DG2 a larger transfer to secure its participation. It then raises the transfer from s2 to b2 = −(v2 D + λ), leaving DG2 with value v2 D + λ indifferent between implementing the project and quitting, allowing FG to claim the rest of the surplus for itself. Finally, if the date-two surplus from agreement is negative; that is, if Equation (1) does not hold, then no amendment will be agreed upon, as the joint surplus from implementing the project is negative. The project will not be implemented and all agents receive dateone payoffs of zero. The expected date-two payoff of a domestic agent with date-one project valuation v who anticipates that the DG2 will have project valuation v2 D and face status quo transfer s2 is thus VD(v, v2 D,s2 ) = σ −(v2 D+s2 ) (v + s2 + λ)f(λ) dλ + −(v2 D+s2 ) −(v2 D+vF ) (v − v2 D + θ (v2 D + λ + vF ))f(λ) dλ. (3) The expected date-two project payoff of FG given s2 when it faces DG2 with valuation v2 D is VF (v2 D,s2 ) = σ −(v2 D+s2 ) (vF − s2 )f(λ) dλ + −(v2 D+s2 ) −(v2 D+vF ) (1 − θ )(v2 D + λ + vF )f(λ) dλ. (4) A transfer of power from a friendly date-one domestic government DG1 to a more hostile date-two domestic government DG2 (i.e., from v to v) carries two implications. First, it increases the prospect that DG2 can renegotiate the initial terms to a more favorable arrangement. Second,it lowers the total surplus of the date-two negotiating parties. As a result, there will be situations in which a hostile DG2 will fail to reach an agreement with FG in contexts where a more project-friendly DG2 would have successfully concluded the negotiation. Discussion: The bargaining protocol is starker than necessary for our main results. What is crucial is that the terms that the domestic government obtains at date two improve as its valuation of the project falls, relative to the status quo offer. This improvement in terms holds regardless of the distribution of date-two bargaining power, θ ∈ [0, 1]. When the domestic government holds date-two proposal power, a more hostile representative can renegotiate the status quo transfer from s2 up to b2 = vF .When, instead, the FG holds proposal power, its offer holds the date-two domestic government to its participation constraint, but its transfer b2 = −(v2 D + λ) still increases as the domestic government becomes more hostile; that is, as v2 D decreases. A more hostile representative not only captures the upside of larger concessions—it also mitigates against the downside of subsequent appropriation. POLICY OUTCOMES AT DATE ONE Exogenous Power Transitions. In our benchmark setting, the date-two domestic government’s (DG2’s) 1022 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:04, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000400
Reelection and Renegotiation valuation does not hinge on the date-one policy out- surplus arising from an agreement is the same as the come.At date one,the FG makes a proposal to the surplus in the event of disagreement.Thus,the total sur- domestic government DG1,which is either relatively plus from a date-one agreement versus no agreement friendly (vp=)or relatively hostile(vp=v).DG ac- is unrelated to its terms: cepts the offer,that is,r(b)=1,if and only if (1-8)(b+vF)+8Pr(u)△(D,y) (1-8)(2+b1) +8Pr(⑦)△(b,)-(1-8)(0+0) +8∑Pr( =]w+Vo(b,哈，b1) -8Pr(u)△(b,)-8Pr()△(2，) 吃也，可 (1-8)(v2+vr) (7) ≥(1-8)0 Because there is a constant surplus at each date,the sur- +8∑Pr()[呢=w+Vn(,s) plus across dates is also constant,and its division rep- resents a pure conflict of interest between the date-one negotiating parties.Starting from an offer that gives (5) DG its reservation payoff,suppose that FG can benefit from making larger initial offers that buttress its future where we recall that w>0 is the office rent that is en- negotiating position vis-a-vis an anticipated date-two joyed if and only if the incumbent is reelected;that is. domestic government.This could arise if both date-one p=vp.Thus,the FG's date-one proposal solves governments expect a significantly more hostile DG2 and the election is sufficiently imminent that FG's im- max (1-8)r(b1)(vF -b1) mediate losses from a larger transfer today are out- 4号元 D12s1 weighed by its expected future gains.Whenever a more generous offer raises FG's total expected payoff,how- +6 ∑Pr(哈)r(哈，1(b1)b1+(1-r1(b)s）, ever,the constant total expected surplus implies that 吃∈{但，可 this gain must come at the expense of DG,which therefore prefers to reject the offer. subject to the participation constraint that r(b)=1 if Thus,when agreement is reached,FG extracts all sur- Equation(5)holds,and r(b1)=0,otherwise. plus from agreement.Equation(7)reveals that the to- tal surplus is positive if and only if the total static sur- Proposition 1.When the identity of the date-two domes- plus is positive:uncertainty about the future has no tic government does not depend on the date-one agree- effect on whether an agreement is signed.Note,how- 是 ment,the project is implemented at date one if and only ever,that the transfer from FG to DG does not solve if the date-one surplus is positive;that is,vp+v0. the static participation constraint that vp+b>0,but Further,if the project is implemented at date one,the rather the dynamic participation constraint given by FG extracts all surplus,offering the transfer that satisfies Equation(5).14 Equation (5). For simplicity,we assume that the FG makes the of- Strikingly,uncertainty about who will hold future fer at date one.If,instead,the domestic government, domestic power has no effect on both (1)whether an DG,makes the initial offer,the conditions for agree- agreement is signed and(2)how the surplus from an ment in Proposition 1 still apply,but now the domestic agreement is divided between the governments.In par- government extracts all surplus. ticular,the static and dynamic conditions for a date- Exogenous power transitions create a constant to- one agreement coincide.Thus,Proposition 1 states that tal surplus between the FG and the date-one domes- a date-one agreement is signed if and only if such an tic government.So long as the static surplus from an agreement is efficient from the perspective of the date- agreement is positive,the FG can and will wish to one negotiating parties. induce the domestic government's participation.But, To understand the result,let A(vp,vp)be the ex ante there is no scope for both governments to benefit from expected date-two surplus from the perspective of the more generous offers-so if and only if the date-one date-one bargaining parties,when the date-one domes- surplus is positive,(1)an agreement is signed and(2) the discounted total expected surplus is fully extracted eys tic government DGI has project valuation vp and they by the FG. anticipate that DG2 has valuation vp: To facilitate a clear and tractable benchmark.we assume that the sole dynamic linkage across periods △(b,v2)=1[v哈=]w is that date-one agreements determine the date-two standing offer,s2.In practice,date-one agreements (b+入+vF)f()d.(6) -(+F) 14 It is routine to show that whenever the static surplus between FG /:sonu and hostile DG is positive,the hostile DG1 extracts a larger transfer Crucially,this surplus does not depend on the date than would a friendly DG.We establish this in the Supplemental two standing offer,s2.In particular,the total date-two Appendix. 1023
Reelection and Renegotiation valuation does not hinge on the date-one policy outcome. At date one, the FG makes a proposal to the domestic government DG1, which is either relatively friendly (v1 D = v) or relatively hostile (v1 D = v).DG1 accepts the offer, that is, r1(b1) = 1, if and only if (1 − δ)(v1 D + b1 ) + δ v2 D∈{v,v} Pr(v2 D) 1[v2 D = v1 D]w + VD(v1 D, v2 D, b1 ) ≥ (1 − δ)0 + δ v2 D∈{v,v} Pr(v2 D) 1[v2 D = v1 D]w + VD(v1 D, v2 D,s1 ) , (5) where we recall that w > 0 is the office rent that is enjoyed if and only if the incumbent is reelected; that is, v2 D = v1 D. Thus, the FG’s date-one proposal solves max b1≥s1 (1 − δ)r1(b1 )(vF − b1 ) + δ v2 D∈{v,v} Pr(v2 D)VF (v2 D,r1(b1 )b1 + (1 − r1(b1 ))s1 ), subject to the participation constraint that r1(b1) = 1 if Equation (5) holds, and r1(b1) = 0, otherwise. Proposition 1. When the identity of the date-two domestic government does not depend on the date-one agreement, the project is implemented at date one if and only if the date-one surplus is positive; that is, v1 D + vF ≥ 0. Further, if the project is implemented at date one, the FG extracts all surplus, offering the transfer that satisfies Equation (5). Strikingly, uncertainty about who will hold future domestic power has no effect on both (1) whether an agreement is signed and (2) how the surplus from an agreement is divided between the governments. In particular, the static and dynamic conditions for a dateone agreement coincide. Thus, Proposition 1 states that a date-one agreement is signed if and only if such an agreement is efficient from the perspective of the dateone negotiating parties. To understand the result,let (v1 D, v2 D) be the ex ante expected date-two surplus from the perspective of the date-one bargaining parties, when the date-one domestic government DG1 has project valuation v1 D and they anticipate that DG2 has valuation v2 D: (v1 D, v2 D) = 1[v2 D = v1 D]w + σ −(v2 D+vF ) (v1 D + λ + vF )f(λ)dλ. (6) Crucially, this surplus does not depend on the datetwo standing offer, s2. In particular, the total date-two surplus arising from an agreement is the same as the surplus in the event of disagreement.Thus, the total surplus from a date-one agreement versus no agreement is unrelated to its terms: (1 − δ)(v1 D + vF ) + δ Pr(v)(v1 D, v) + δ Pr(v)(v1 D, v) − (1 − δ)(0 + 0) − δ Pr(v)(v1 D, v) − δ Pr(v)(v1 D, v) = (1 − δ)(v1 D + vF ). (7) Because there is a constant surplus at each date, the surplus across dates is also constant, and its division represents a pure conflict of interest between the date-one negotiating parties. Starting from an offer that gives DG1 its reservation payoff, suppose that FG can benefit from making larger initial offers that buttress its future negotiating position vis-à-vis an anticipated date-two domestic government. This could arise if both date-one governments expect a significantly more hostile DG2 and the election is sufficiently imminent that FG’s immediate losses from a larger transfer today are outweighed by its expected future gains.Whenever a more generous offer raises FG’s total expected payoff, however, the constant total expected surplus implies that this gain must come at the expense of DG1, which therefore prefers to reject the offer. Thus, when agreement is reached, FG extracts all surplus from agreement. Equation (7) reveals that the total surplus is positive if and only if the total static surplus is positive: uncertainty about the future has no effect on whether an agreement is signed. Note, however, that the transfer from FG to DG1 does not solve the static participation constraint that v1 D + b1 ≥ 0, but rather the dynamic participation constraint given by Equation (5). 14 For simplicity, we assume that the FG makes the offer at date one. If, instead, the domestic government, DG1, makes the initial offer, the conditions for agreement in Proposition 1 still apply, but now the domestic government extracts all surplus. Exogenous power transitions create a constant total surplus between the FG and the date-one domestic government. So long as the static surplus from an agreement is positive, the FG can and will wish to induce the domestic government’s participation. But, there is no scope for both governments to benefit from more generous offers—so if and only if the date-one surplus is positive, (1) an agreement is signed and (2) the discounted total expected surplus is fully extracted by the FG. To facilitate a clear and tractable benchmark, we assume that the sole dynamic linkage across periods is that date-one agreements determine the date-two standing offer, s2. In practice, date-one agreements 14 It is routine to show that whenever the static surplus between FG and hostile DG1 is positive, the hostile DG1 extracts a larger transfer than would a friendly DG1. We establish this in the Supplemental Appendix. 1023 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:04, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000400
Peter Buisseret and Dan Bernhardt affect the total surplus through other channels.For ex- in exchange for implementing the project.The first as- ample,the possibility of participating in the project at piration depends on the voter's valuation,but the sec- date two might depend on whether an agreement was ond applies to all voters regardless of ideology,since formed at date one.In the Supplemental Appendix, all voters share a common value in extracting greater we consider this and related contexts,showing that surplus from the FG. while the conditions for an agreement to be signed may A domestic voter is more intrinsically aligned with depart from our benchmark condition,all surplus is the friendly party wheneverv:the friendly party nonetheless extracted by the FG whenever an agree- is relatively more likely to reach agreements with the ment is reached.15 FG in circumstances where the voter would prefer an op//s Endogenous Power Transitions.We now consider an agreement to no agreement.Yet,Lemma 1 reveals that electoral contest between dates one and two in which this domestic voter may nonetheless strictly prefer to domestic voters,who differ in their project valuations vote for the hostile party! vER,observe the date-one negotiation outcome and To understand why,suppose that e 1,so that then cast ballots in favor of their most-preferred date- DG2 always has proposal power in any renegotiation.16 two government A voter anticipates a positive prospect that either a Imminent elections have a polarizing effect on friendly or a hostile DG2 will renegotiate the standing date-one negotiations between FG and DG1.In the offer s2-indeed,with uniform preference shocks,each benchmark setting,initial negotiations are driven by party is equally likely to renegotiate the standing of- the conflict of interest between the date-one negotiat- fer in equilibrium.However,voters also anticipate that ing partners over the surplus division.When elections each party renegotiates with FG in different circum- respond to initial agreements,two other conflicts are stances,with the hostile party renegotiating when vot- critical:the policy and rent-seeking conflict between ers attach a higher valuation to the project than when the domestic incumbent and its possible successors,and the friendly party is renegotiating. First,ifλ∈[-（⑦+s2),-(u+s2)l,a friendly DG2 4 both domestic and foreign governments'conflict with the domestic electorate.As a result,initial agreements maintains the standing offer s2,but a hostile DG2 rene- no longer solely serve to divide the surplus:depending gotiates the offer from s2 to vr.Nonetheless,both do- on whether DG]is relatively friendly or hostile,initial mestic governments ultimately pursue the project.In agreements may themselves change both the division this context,a voter with valuation v +A derives a pos- and the size of the surplus from agreement. itive payoff difference from the hostile DG2 versus the Given status quo agreement s2,a voter with valua- friendly DG,of tion v prefers a date-two domestic government that, from her perspective,induces the best date-two nego- v+vF+入-(v+s2+入)=vF-S2>0. (9) tiation outcome.that is.that solves The voter's date-two project valuation (v +)does max VD(v,vD.S2). not affect her relative value because both governments would pursue the project:regardless of her valuation. she perceives that the hostile party is superior.because where s2=b if the project was implemented at date 1.501 it pursues the project and extracts a higher surplus in one with transfer b1,and s2=s if it was not imple- the same context that the friendly party would pursue mented at date one.The uniform distribution over the the project at the standing offer. preference shock yields Second,.ifλe[-(⑦+vr),-(u+vr)l,a friendly DG2 renegotiates the offer from s2 to vF,but a hostile Lemma 1.Given an inherited status quo agreement s2, DG2 instead walks away from the project.A voter with a domestic voter with project valuation v prefers to elect the hostile government if and only if project valuation v+A derives a payoff difference from the friendly DG2 versus the hostile DG2 of 卫+ 2 -+(UF-S2)=(S2) (8) UF+(v+入)-0. (10) A domestic voter's induced preference for the friendly Like the hostile DG2 in the first case,a friendly DG2 or hostile party depends on (1)her desire that a party in the second case appropriates FG's value from the reach an agreement with the FG in the same circum- project,vr.However,unlike the first case,this transfer stances where she would value the project and (2)her is partially offset by the fact that a friendly government desire to extract more generous transfers from the FG pursues the project in circumstances where a hostile government-and,possibly,the voter-wants to abandon it.When v+<0,the friendly government's 15 We consider variations on our benchmark setting in which(a)each continuation of the project generates a loss that is only L country faces start-up costs from pursuing the project in the first pe- partially offset by the transfer,vr.So,for example, riod an agreement is signed and (b)the distribution of the date-two preference shock in the domestic country depends on whether an voter's expectation of the payoff difference agreement is signed at date one.We thank an anonymous referee, who encouraged us to explore the robustness of our benchmark re- 16 We thank Mark Fey for conveying this intuition for our result in sult. an extremely insightful discussion. 1024
Peter Buisseret and Dan Bernhardt affect the total surplus through other channels. For example, the possibility of participating in the project at date two might depend on whether an agreement was formed at date one. In the Supplemental Appendix, we consider this and related contexts, showing that while the conditions for an agreement to be signed may depart from our benchmark condition, all surplus is nonetheless extracted by the FG whenever an agreement is reached.15 Endogenous Power Transitions. We now consider an electoral contest between dates one and two in which domestic voters, who differ in their project valuations v ∈ R, observe the date-one negotiation outcome and then cast ballots in favor of their most-preferred datetwo government. Imminent elections have a polarizing effect on date-one negotiations between FG and DG1. In the benchmark setting, initial negotiations are driven by the conflict of interest between the date-one negotiating partners over the surplus division. When elections respond to initial agreements, two other conflicts are critical: the policy and rent-seeking conflict between the domestic incumbent and its possible successors, and both domestic and foreign governments’ conflict with the domestic electorate. As a result, initial agreements no longer solely serve to divide the surplus: depending on whether DG1 is relatively friendly or hostile, initial agreements may themselves change both the division and the size of the surplus from agreement. Given status quo agreement s2, a voter with valuation v prefers a date-two domestic government that, from her perspective, induces the best date-two negotiation outcome, that is, that solves max v2 D VD(v, v2 D,s2 ), where s2 = b1 if the project was implemented at date one with transfer b1, and s2 = s1 if it was not implemented at date one. The uniform distribution over the preference shock λ yields Lemma 1. Given an inherited status quo agreement s2, a domestic voter with project valuation v prefers to elect the hostile government if and only if v ≤ v + v 2 + (vF − s2 ) ≡ vˆ(s2 ). (8) A domestic voter’s induced preference for the friendly or hostile party depends on (1) her desire that a party reach an agreement with the FG in the same circumstances where she would value the project and (2) her desire to extract more generous transfers from the FG 15 We consider variations on our benchmark setting in which (a) each country faces start-up costs from pursuing the project in the first period an agreement is signed and (b) the distribution of the date-two preference shock in the domestic country depends on whether an agreement is signed at date one. We thank an anonymous referee, who encouraged us to explore the robustness of our benchmark result. in exchange for implementing the project. The first aspiration depends on the voter’s valuation, but the second applies to all voters regardless of ideology, since all voters share a common value in extracting greater surplus from the FG. A domestic voter is more intrinsically aligned with the friendly party whenever v > v+v 2 : the friendly party is relatively more likely to reach agreements with the FG in circumstances where the voter would prefer an agreement to no agreement. Yet, Lemma 1 reveals that this domestic voter may nonetheless strictly prefer to vote for the hostile party! To understand why, suppose that θ = 1, so that DG2 always has proposal power in any renegotiation.16 A voter anticipates a positive prospect that either a friendly or a hostile DG2 will renegotiate the standing offer s2—indeed, with uniform preference shocks, each party is equally likely to renegotiate the standing offer in equilibrium. However, voters also anticipate that each party renegotiates with FG in different circumstances, with the hostile party renegotiating when voters attach a higher valuation to the project than when the friendly party is renegotiating. First, if λ ∈ [−(v + s2 ), −(v + s2 )], a friendly DG2 maintains the standing offer s2, but a hostile DG2 renegotiates the offer from s2 to vF. Nonetheless, both domestic governments ultimately pursue the project. In this context, a voter with valuation v + λ derives a positive payoff difference from the hostile DG2 versus the friendly DG2 of v + vF + λ − (v + s2 + λ) = vF − s2 > 0. (9) The voter’s date-two project valuation (v + λ) does not affect her relative value because both governments would pursue the project; regardless of her valuation, she perceives that the hostile party is superior, because it pursues the project and extracts a higher surplus in the same context that the friendly party would pursue the project at the standing offer. Second, if λ ∈ [−(v + vF ), −(v + vF )], a friendly DG2 renegotiates the offer from s2 to vF , but a hostile DG2 instead walks away from the project. A voter with project valuation v + λ derives a payoff difference from the friendly DG2 versus the hostile DG2 of vF + (v + λ) − 0. (10) Like the hostile DG2 in the first case, a friendly DG2 in the second case appropriates FG’s value from the project, vF . However, unlike the first case, this transfer is partially offset by the fact that a friendly government pursues the project in circumstances where a hostile government—and, possibly, the voter—wants to abandon it. When v + λ < 0, the friendly government’s continuation of the project generates a loss that is only partially offset by the transfer, vF . So, for example, voter v = v+v 2 ’s expectation of the payoff difference 16 We thank Mark Fey for conveying this intuition for our result in an extremely insightful discussion. 1024 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:04, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000400
Reelection and Renegotiation Equation (10)at the time of the domestic election is atively hostile domestic party is therefore zero.17 Since the payoff gain Equation (9)is strictly positive,a voter whose project valuation is equidistant from the two parties strictly prefers to support the Pr(vmed≤t(s2)）= (S2)-(ve-) 20 hostile party. Note that a voter's induced preferences over date- 学+p-s2-(-am） two governments differ from those of an agent who 11) 2a shares her project valuation.v.but chooses a date- two domestic government to maximize total expected The electoral consequences of a more favorable date- date-two surplus between that voter and FG.Such an two status quo s2 differ starkly for the friendly and agent would prefer the hostile government if and only hostile domestic parties.As s2 increases-for exam- if vs8 The reason for this divergence is that a ple,when the date-one domestic government extracts voter does not value total surplus,but rather her share a larger transfer in exchange for pursuing the project- of the surplus.This highlights a possible source of ineffi- the hostile party's electoral prospects fall,and the ciency in domestic election outcomes that are sensitive friendly party's electoral prospects rise.This is a key to a country's external negotiations. difference with Wolford(2012),who assumes that a do- Voters'induced preferences over date-two represen- mestic incumbent's re-election prospects rise with the tatives are manipulable by both date-one governments surplus it extracts from an FG in a pre-election ne- FG can manipulate a voter's tradeoffs via its initial of- gotiation.Our framework highlights that when initial fer,bi s1:more generous offers-if accepted-steer negotiation outcomes become the reversion point in voters toward the more project-friendly party.But DG future elections,voters'induced preferences may gen- can also manipulate voters'tradeoffs via its choice to erate the opposite relationship between an incum- accept or reject the offer,r(b)[0,1):rejecting an bent's negotiated share of the surplus from agreement 4号 offer bequeaths a worse status quo,steering voters to- and its reelection prospects. ward the more hostile party.How these concerns affect We earlier showed that when power transitions are the prospect of initial agreement,and the division of exogenous,total expected surplus is unaffected by the the surplus,will depend on the policy conflict between initial agreement.This is no longer true when date-one parties,between the parties and their electorate,and outcomes can alter electoral outcomes.To see why,rec- between all domestic agents and the FG.We now show ognize that from the perspective of the date-one bar- how these conflicts resolve. gaining parties,the expected date-two surplus derived Henceforth,we assume that the distribution of vot- from a status quo s2 is ers'project valuations has a unique median,vmed.The single-peaked structure of induced preferences then implies that the voter with this median valuation is de- Pr(umed≤i(s2)△(b,) cisive in an election:for any standing offer s2,the hos- +Pr(umed >i(s2))A(vp.) (12) tile party wins if and only ifd≤+(ur-s2)≡ (s2).We assume that,at date one,both the FG and where A(vp,vp)(defined in Equation (6))is the ex domestic parties are uncertain of the future median ante expected date-two surplus from the perspective of voter's project valuation: the date-one bargaining parties when DGi has project Assumption 4:The valuation vmed of the median voter valuation vp and DG2 has valuation v.Thus the is drawn from a uniform distribution on the interval relative total surplus from an agreement (versus no lve-a,ve +al,where (1)v -a and (2)v+ agreement)is a>+p-s1. (1-8)(vF +vp)+8(Pr(umed si(b1)) Uniform uncertainty is not essential for our results, but it facilitates tractable comparative statics (e.g.,on -Pr(umed≤i(s1))(△(vb,u)-△(vD,) v and a).There are many natural interpretations.For (13) example,the larger is a,the more uncertain are date- one negotiating parties about the preferences of the Lemma 2 highlights how the relative total surplus from domestic electorate.Conditions(1)and(2)imply that an agreement changes with the terms of the agreement, there is enough uncertainty about voter preferences depending on whether DG is relatively friendly or rel- that each party wins with positive probability given atively hostile any standing offer,s2E[s1,vF].The probability that the majority-preferred date-two government is the rel- Lemma 2.For any 8>0,the relative total surplus from an agreement with transfer b between the FG and the date-one domestic government is 17This follows from()(d=0. 18 Total expected date-two surplus between a voter with date-one (1)strictly increasing in b if DG is relatively /:sony project valuation and FG with valuation vF is maximized by setting friendly,with valuation v,and 6=业if and only if巴e+pu+F+)fa)d之户e+p)w+ (2)strictly decreasing in b if DG is relatively F+)f()da.which is equivalent tov hostile,with valuation v. 1025
Reelection and Renegotiation Equation (10) at the time of the domestic election is zero. 17 Since the payoff gain Equation (9) is strictly positive, a voter whose project valuation is equidistant from the two parties strictly prefers to support the hostile party. Note that a voter’s induced preferences over datetwo governments differ from those of an agent who shares her project valuation, v, but chooses a datetwo domestic government to maximize total expected date-two surplus between that voter and FG. Such an agent would prefer the hostile government if and only if v ≤ v+v 2 . 18 The reason for this divergence is that a voter does not value total surplus, but rather her share of the surplus.This highlights a possible source of inefficiency in domestic election outcomes that are sensitive to a country’s external negotiations. Voters’ induced preferences over date-two representatives are manipulable by both date-one governments. FG can manipulate a voter’s tradeoffs via its initial offer, b1 ≥ s1: more generous offers—if accepted—steer voters toward the more project-friendly party.But DG1 can also manipulate voters’ tradeoffs via its choice to accept or reject the offer, r1(b1) ∈ {0, 1}: rejecting an offer bequeaths a worse status quo, steering voters toward the more hostile party. How these concerns affect the prospect of initial agreement, and the division of the surplus, will depend on the policy conflict between parties, between the parties and their electorate, and between all domestic agents and the FG.We now show how these conflicts resolve. Henceforth, we assume that the distribution of voters’ project valuations has a unique median, vmed. The single-peaked structure of induced preferences then implies that the voter with this median valuation is decisive in an election: for any standing offer s2, the hostile party wins if and only if vmed ≤ v+v 2 + (vF − s2 ) ≡ vˆ(s2 ). We assume that, at date one, both the FG and domestic parties are uncertain of the future median voter’s project valuation: Assumption 4: The valuation vmed of the median voter is drawn from a uniform distribution on the interval [ve − α, ve + α], where (1) ve − α v+v 2 + vF − s1. Uniform uncertainty is not essential for our results, but it facilitates tractable comparative statics (e.g., on ve and α). There are many natural interpretations. For example, the larger is α, the more uncertain are dateone negotiating parties about the preferences of the domestic electorate. Conditions (1) and (2) imply that there is enough uncertainty about voter preferences that each party wins with positive probability given any standing offer, s2 ∈ [s1, vF]. The probability that the majority-preferred date-two government is the rel- 17 This follows from −(v+vF ) −(v+vF ) ( v+v 2 + vF + λ)f(λ)dλ = 0. 18 Total expected date-two surplus between a voter with date-one project valuation v and FG with valuation vF is maximized by setting v2 D = v if and only if σ −(v+vF )(v + vF + λ)f(λ)dλ ≥ σ −(v+vF )(v + vF + λ)f(λ)dλ, which is equivalent to v ≤ v+v 2 . atively hostile domestic party is therefore Pr(vmed ≤ vˆ(s2 )) = vˆ(s2 ) − (ve − α) 2α = v+v 2 + vF − s2 − (ve − α) 2α . (11) The electoral consequences of a more favorable datetwo status quo s2 differ starkly for the friendly and hostile domestic parties. As s2 increases—for example, when the date-one domestic government extracts a larger transfer in exchange for pursuing the project— the hostile party’s electoral prospects fall, and the friendly party’s electoral prospects rise. This is a key difference with Wolford (2012), who assumesthat a domestic incumbent’s re-election prospects rise with the surplus it extracts from an FG in a pre-election negotiation. Our framework highlights that when initial negotiation outcomes become the reversion point in future elections, voters’ induced preferences may generate the opposite relationship between an incumbent’s negotiated share of the surplus from agreement and its reelection prospects. We earlier showed that when power transitions are exogenous, total expected surplus is unaffected by the initial agreement. This is no longer true when date-one outcomes can alter electoral outcomes. To see why, recognize that from the perspective of the date-one bargaining parties, the expected date-two surplus derived from a status quo s2 is Pr(vmed ≤ vˆ(s2 ))(v1 D, v) + Pr(vmed > vˆ(s2 ))(v1 D, v), (12) where (v1 D, v2 D) (defined in Equation (6)) is the ex ante expected date-two surplus from the perspective of the date-one bargaining parties when DG1 has project valuation v1 D and DG2 has valuation v2 D. Thus the relative total surplus from an agreement (versus no agreement) is (1 − δ)(vF + v1 D) + δ(Pr(vmed ≤ vˆ(b1 )) − Pr(vmed ≤ vˆ(s1 )))((v1 D, v) − (v1 D, v)). (13) Lemma 2 highlights how the relative total surplus from an agreement changes with the terms of the agreement, depending on whether DG1 is relatively friendly or relatively hostile. Lemma 2. For any δ > 0, the relative total surplus from an agreement with transfer b1 between the FG and the date-one domestic government is (1) strictly increasing in b1 if DG1 is relatively friendly, with valuation v, and (2) strictly decreasing in b1 if DG1 is relatively hostile, with valuation v. 1025 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:04, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000400