Centralization, Decentralization and Incentive Problems in Eurozone Financial Governance: A Contract Theory Analysis

Transcription

1 Internatonal Journal of Economcs an Fnance; Vol. 10, No. 3; 018 ISSN X E-ISSN Publshe by Canaan Center of Scence an Eucaton Centralzaton, Decentralzaton an Incentve Problems n Eurozone Fnancal Governance: A Contract Theory Analyss 1 Faculty of Economcs, ose Unversty, Toyo, Japan Yutaa Suzu 1 Corresponence: Yutaa Suzu, Faculty of Economcs, ose Unversty, 434 Ahara Macha Toyo Japan. Tel: E-mal: yutaa@hose.ac.jp Receve: December 4, 017 Accepte: January 30, 018 Onlne Publshe: February 5, 018 o: /jef.v10n3p93 URL: Abstract Ths paper uses a contract theory framewor to analyze the mechansms of eurozone fnancal governance, wth a focus on centralzaton vs. ecentralzaton an ncentve problems. By constructng a Stacelberg game moel wth n Mnstres of Fnance as the frst movers an the European Central Ban as the secon mover, we show that each government can create growth n ts own country (self-beneft) by ncreasng government spenng, but that ths wll ncrease nflaton, resultng n a ecrease n the value of the euro. As these effects are share equally by eurozone countres (cost sharng), an ncentve to free-re at the expense of other countres s present. We then analyze a penalty-base soluton to the free-rer problem an erve a secon-best soluton where a commtment not to renegotate penaltes ex-post s mpossble. The optmal soluton shows that lmte soveregnty, that s, substantally constrane fscal soveregnty, shoul be mpose as a hgh margnal cost for the ssuance of publc ebt. Fnally, we close the paper by scussng the possblty of Fscal Integraton (Fscal Unon). Keywors: monetary centralzaton, fscal ecentralzaton, free-rer problem, penalty schemes an renegotaton, Stablty an Growth Pact (SGP), fscal ntegraton 1. Introucton Ths paper analyzes the mechansms of eurozone fnancal governance through the Stablty an Growth Pact (SGP) (an the Fscal Compact (01) as a revse form) n theory an n realty, wth a focus on centralzaton, ecentralzaton, an ncentve structures. We begn by escrbng the nsttutonal framewor of the pero from 00 to 003, when the eurozone consste of 1 countres (Note 1). The authorty on nterest polcy s centralze wth the European Central Ban (ECB), an the authorty on fscal polcy s ecentralze, restng wth each country. Thus, the eurozone has a structure of monetary centralzaton an fscal ecentralzaton uner a common currency (the euro). Thus, a structure of strategc nterepenence emerges, wth tenson between the nterest rate polcy set by the ECB to mantan the value of the euro an natonal efcts ncurre by each country. The major factor mang fscal stablzaton ffcult s the Prsoner s Dlemma problem of fnancal governance. When other countres abe by fscal rules, one country has an ncentve to become a free rer. That country s neglect may cause other countres to follow, resultng n a loss of the force bnng the SGP together. Ths may lea to further ffcultes n mantanng the value of the euro, llustratng the theoretcal phenomenon of Tragey of the Commons. The SGP s the mechansm of fnancal governance uner an envronment of fscal ecentralzaton, whch mposes on member natons the conton of an annual natonal efct below 3% of GDP, an a government ebt of lower than 60% of GDP, at a regonal level. As an nsttutonal mechansm for the SGP, the European Commsson (EC) s allowe to act as an agent montorng the natonal efcts of eurozone countres, an the EC has the authorty to warn countres n volaton of the rules that penaltes wll be apple. The European Commsson (EC) checs for vergence from the fnancal stablzaton plans that member 93

2 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 natons have submtte. Ths serves as an early warnng of any rs of exceeng the 3% of GDP lmt. The European Commsson (EC) employs excessve natonal efct proceures (EDP) to guar aganst the contnuaton or worsenng of efcts. These proceures requre measures to be taen wthn four months an a resoluton to be reache wthn one year. A penalty warnng s gven for breaches. The penalty s prescrbe as a maxmum of 0.5% of GDP. When the amount of the penalty s aopte by a vote of the Eurozone Fnance Mnsters Councl, the penalty becomes bnng. Ths governance structure s ve, wth the European Commsson (EC) holng the warnng authorty regarng penalty applcaton an the Eurozone Fnance Mnsters Councl holng the ecson-mang authorty. Therefore, a country n volaton can approach the Mnstry of Fnance n each country to attempt to avo havng to pay a huge penalty,.e., there s room for renegotaton. Whether the penalty s nee mpose s etermne by qualfe majorty votng (QMV) n the Fnance Mnsters Councl. Aopton of the penalty requres a two-thrs or greater number of votes, wth the result that a mnorty of one-thr or more votes can bloc the penalty. Ths means that t may be relatvely easy for a major power n volaton to form a blocng mnorty. Two real cases from 00 an 003 expose these nsttutonal lmtatons. Case 1: Decson not to aopt an early warnng aganst Germany, January 30, 00. The European Commsson (EC) ssue an early warnng. owever, Germany was able to secure a blocng mnorty. To avo the loss of creblty of the SGP that woul result from a efeat n a vote, the Fnance Mnsters Councl not aopt the warnng ssue by the European Commsson n exchange for a commtment from Germany to reuce ts natonal efct. Case : Rejecton of the warnng aganst France an Germany by the Eurozone Fnance Mnsters Councl, November, 003. France an Germany ha been ssue secon warnngs of excessve natonal efct proceures (EDP), but the warnng to reuce ther natonal efcts was rejecte by the QMV. A proposal was passe to ssue a temporarly shelve penalty to Germany. These two cases expose certan nsttutonal lmtatons, emonstratng that even f the penalty nclue n the SGP s recommene by the European Commsson (EC), t s extremely ffcult for that penalty to be aopte by the Fnance Mnsters Councl, whch etermnes whether the penalty shoul actually be nvoe. The votng proceure that enable the above renegotaton n has recently been change. In the new Reverse Qualfe Majorty Votng (RQMV) proceure set out n the Fscal Compact n 01, Eurozone member natons must present a qualfe majorty (a two-thrs or greater number of votes) to prevent the mposton of sanctons, rather than to support the mposton of sanctons as n the former Stablty an Growth Pact (GSP). Ths change has mae t more ffcult for evatng countres to bloc the mposton of sanctons (penaltes) by collung wth other countres, as blocng now requres a two-thrs or greater number of votes n the new RQMV proceure. We present a theoretcal moel to analyze the effect of ths nsttutonal change from QMV to RQMV. We further vew ths process as a game structure n whch the European Central Ban (ECB) rases the polcy rate to mantan the value of the euro aganst any free-rer naton that accumulates natonal efcts when the penalty functon of the SGP to suppress a natonal efct uner fscal ecentralzaton oes not wor. Base on the above escrpton, the followng fgure epcts the nsttutonal framewor of the Eurozone. 94

3 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 Ths paper uses a game/contract theory framewor to analyze the mechansms of eurozone fnancal governance through the Stablty an Growth Pact (SGP) (an the Fscal Compact (01) as the revse form), wth a focus on centralzaton vs. ecentralzaton an ncentve structures n the European Unon (EU) (Note ). We frst construct a Stacelberg game moel playe by n Mnstres of Fnance (MOF) n n eurozone member countres as the frst movers, wth the European Central Ban (ECB) as the secon mover. We then post the followng basc ntuton. The Government n each country can create growth (economy) n ts own country, resultng n an ncrease n GDP (100% self-beneft), by ncreasng government spenng. Ths ncreases nflaton, resultng n a ecrease n the value of the euro. As these effects are share equally by euro countres (the cost s share equally), an ncentve exsts to cut corners on reucng the ssuance of publc ebt (a free re at the expense of other countres). An ncrease n the number of Mnstres of Fnance (MOF) or the number of natons wll ncrease the severty of the free-rer problem. Followng ths framewor, we analyze the penalty scheme n the SGP as a soluton to the free-rer problem. When an executor who commts hmself to an enforces the penalty schemes n the SGP exsts, the best soluton can be acheve as shown n the lterature (e.g., Bolton an Dewatrpont (005) an olmstrom (198)). owever, as we escrbe, re-scusson (votng) was conucte as to whether the penalty scheme shoul actually be mpose on the evatng country. In ths case, t s mpossble to eep the commtment to refran from conuctng ex-post renegotaton of the penalty. When major powers such as Germany an France brea the fscal rules n the SGP (through excessve government spenng), ex-post renegotaton may occur. At that tme, a gan from renegotaton wll exst. In exchange for se payments to other countres by Germany an France, no large penalty wll be ssue, or the amount of the penalty wll be sgnfcantly reuce. We characterze the secon best soluton as a stuaton where t s mpossble to meet the commtment to not renegotate penaltes ex-post. The optmal soluton shows that lmte soveregnty, that s, substantally constrane fscal soveregnty, shoul be mpose as a hgh margnal cost for the ssuance of publc ebt. 1.1 Relate Lterature Dewatrpont (001) explans the optmal vson of authorty between the MOF an the Central Ban (e.g., the Deutsche Bunesban - The Central Ban of Germany) n the settng of one MOF an a Central Ban. Motvate by hs wor, we have constructe an extene moel consstng of n Mnstres of Fnance (MOF) n n eurozone member natons an the European Central Ban (ECB), an explan the free-rer problem among the n member natons as a prsoner s lemma, whch s essental not only to our analyss but also for the unerstanng of urgent polcy ssues facng the EU toay. Wth ths moel, we evelop a soluton to the free-rer problem through the penalty scheme n the SGP (an then the Fscal Compact), an erve a secon best soluton where a commtment to not renegotate penaltes ex-post s mpossble; ths s also an urgent ssue face toay n the EU. For the free-rer problem, Char an Kehoe (008) frst note ebt over-ssue as a new type of free rer problem resultng from the lac of commtment on the part of the monetary authorty n ther Macro ynamc 95

4 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 moel of the monetary unon. In the non-cooperatve equlbrum, expectng the monetary authorty to nflate nomnal ebt, the government tens to over-ssue ebt, weghng only the cost of nflaton for ts resents an gnorng the cost of nflaton borne by the other members of the monetary unon. Our paper aopts a smpler contract/game theory framewor whch stll contans a smlar free-rer problem, although the monetary authorty mantans ts commtment to prce stablty. Further, one of the mportant ssues n the esgn of the monetary unon s the queston of the necessty of fscal constrants. Char an Kehoe (007) showe that ebt constrants are esrable n an envronment wthout commtment on the part of the monetary authorty. In contrast, our paper erves lmte soveregnty as the optmal soluton, ncatng that substantally constrane fscal soveregnty s an ex-ante soluton for ex-post renegotaton of the penalty scheme.. Basc Moel Usng Contract/Game Theory.1 Moel Settng The member players of the European Unon (EU) are the European Central Ban (ECB) an the Mnstry of Fnance (MOF) of each naton. The European Central Ban (ECB) s an ntegrate organzaton mae up of the central bans of member natons, wth the responsblty to manage monetary polcy n an ntegrate manner an to etermne the polcy rate (common nterest rate) (Note 3). owever, the Mnstry of Fnance (MOF) of each naton has the authorty to control ts own fnancal spenng. Thus, let the fnancal spenng level of the naton be, = 1,,.., n,. Ths s a structure of monetary centralzaton an fscal ecentralzaton. The GDP (Gross Domestc Prouct,.e., output) of naton s expresse as x =, = 1,,.., n, whch represents fnancal spenng (fnancng through the ssuance of natonal bons) mnus the polcy rate (Note 4). The nflaton rate π of the naton s expresse as π = β, = 1,,.., n (Note 5), an we assume that β > 1 (Note 6). Note that because nflaton represents prce ncreases, ths s equvalent to a ecrease n monetary value, that s, the value of the euro. x Next, let the objectve functon of the naton s Mnstry of Fnance (MOF) be naton s MOF places hgh value on an ncrease n GDP,.e., the so calle economc growth. Ths means that x, an that the ssuance of publc ebt an ncreases n outstanng publc ebt wll ncrease costs for the MOF. The nterpretaton of the ssuance cost of publc ebt s as follows. α s the nterest rate for publc ebt, an also a parameter that characterzes the margnal cost of the ssuance of publc ebt. s the cost of obtanng the approval or acceptance of the parlament to ncrease the ssuance of publc ebt (government fnancal spenng). Ths means that ncreasng publc ebt s costly n two ways. n. As the ECB s Meanwhle, let the objectve functon of the European Central Ban (ECB) be a guaran of commoty prces wth the msson to mantan the value of the euro, the frst term quaratc loss μπ means that nflaton/eflaton (prce ncreases/ecreases) an changes n the value of the euro are vewe as costs for the ECB. ere, μ s a parameter that captures the socal cost of nflaton. The secon term means that the ECB also recognzes the amount of government ebt of each naton as a cost. The most mportant pont s that the MOF acts base on an economc-expanson bas, n the sense that the MOF s objectve nclues x (GDP growth) but not μπ (cost of nflaton), whle the European Central Ban (ECB) has a prce-stablty (stablty of the euro s value) bas n the sense that the ECB s objectve nclues μπ but not x (Note 7). In summary, a conflct of nterest exsts between the MOF an the ECB. We have specfe the players of the game (Mnstres of Fnance 1,,.., n, European Central Ban), the strateges of each player (fnancal spenng through the ssuance of publc ebt, = 1,,.., n an a sngle polcy rate ), an the objectves that each player pursues. As for the tme structure (tmng), the MOF of each country smultaneously an nepenently chooses ts own fnancal spenng 1,,.., N, an, after observng 1 96

5 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 ths, the European Central Ban chooses a common polcy rate. The polcy rate can be ajuste much more qucly than fnancal spenng (the ssuance of publc ebt). Secton. theoretcally escrbes the followng basc mechansm. The government n each country can create economc growth n that country, meanng an ncrease n GDP (100% self-beneft), by ncreasng government spenng. Ths wll ncrease nflaton an ecrease the value of the euro. These effects are share equally by euro countres (costs are share equally). Thus, an ncentve for cuttng corners on reucng the ssuance of publc ebt (.e., a free re at the expense of other countres).. Moel Soluton by Bacwar Inucton In ths secton, we solve the basc moel usng bacwar nucton n one MOF case, two MOF cases, an n MOF cases, an note that the free-rer problem occurs n equlbrum uner the common currency (euro) system MOF an ECB Tmng (Note 8): T=1 The Mnstry of Fnance (MOF) chooses fnancal spenng. T= The European Central Ban (ECB) chooses the nterest rate (polcy rate) Ths s the Stacelberg Game playe by the Mnstry of Fnance (MOF) an the European Central Ban (ECB). T=: Gven the fnancal spenng n T=1, the ECB chooses the nterest rate (polcy rate). max Substtutng nto the payoff functon an optmzng wth respect to, we have the Frst-orer conton for optmalty: 0 Therefore, the best response functon by the ECB n T= s Ths shows that nflaton (the ecrease n the value of the Euro) ue to the ncrease n fnancal spenng n T=1 s suppresse by the polcy rate n T=. T=1: The Mnstry of Fnance (MOF) expects the best response functon by the European Central Ban (ECB) n T= an chooses the optmal fnancal spenng. max x s.t. That s, the Mnstry of Fnance (MOF) solves max. The Frst-orer conton for optmalty s ence, the optmal soluton n T=1 s an the equlbrum polcy rate set by the ECB n T= s The soluton concept s Sub-game Perfect Equlbrum, whch correspons to an. Note: The Nash Equlbrum of a Smultaneous Move Game between MOF an ECB s N 1 N 1,, hgher than n the case of the Stacelberg Game. In the Stacelberg game, the follower (ECB), after observng the leaer s (MOF) move (the fscal spenng at T=1), cancels out some of the effect of the ncrease n GDP by 97

6 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 ncreasng the polcy rate. Expectng ths, the MOF only aopts the optmal fscal expenture (the ssuance of publc ebt) to maxmze x. In other wors, the ECB has the authorty to ncrease the polcy rate, suppressng the ncentve for the MOF to ncrease fscal spenng n T=1 (Note 9, Note 10). Let us revew the argument thus far usng the best response functons. The best response functon for the ECB s epcte n Fgure 1. Expectng ths, the MOF chooses the optmal fscal spenng level at T=1, whch maxmzes the fference between 1 an. We can confrm that the margnal beneft 1 obtane by ncreasng s balance by the margnal ncrease n the optmal response 1 at The equlbrum nterest rate s then We epct ths Stacelberg equlbrum pont as S1 (for one MOF case). The Nash equlbrum pont n Fgure 1 s N. Fgure 1. Stacelberg equlbrum pont S1: 1 MOF an ECB.. MOFs an ECB The tme lne s as follows, an we solve ths game by bacwar nucton. T=1 T= Mnstres of Fnance, MOF1 an MOF European Central Ban (ECB) choose fnancal spenng 1, chooses the common polcy rate.. T=: Gven the fnancal spenng 1, of MF1 an MF, the ECB chooses the nterest rate (polcy rate). Gven the frst pero of fnancal expenture 1,, an n accorance wth the common polcy rate, whch the ECB chooses at T=, the expecte nflaton rate n countres = 1, at T= s π = β, = 1,. As the average nflaton rate s = β, the ECB solves the followng problem. 98

7 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; max max The Frst Orer Conton for optmalty s 1 0. Thus, the best response functon by the ECB at T= s, Ths means that the optmal acton for the ECB s to ncrease the polcy rate n proporton to the average ncrease n nflaton (meanng a ecrease n the value of the euro) ue to the ncrease n the frst pero of average fnancal spenng. T=1: The Mnstry of Fnance n each country = 1, expects the optmal response by the European Central Ban (ECB) at T= an smultaneously chooses the optmal level of fnancal spenng. The problem s Optmzng max x, 1, s.t x The soluton s 1 1 1, 1, (Note 11) wth respect to, 1,, we have the Frst Orer Conton: Thus, the Sub-game Perfect Equlbrum of ths game s 1 1 1, 1,, an 1, 1 1 We see that f there are two Mnstres of Fnance, equlbrum fnancal spenng at T=1 ncreases ), an the polcy rate etermne by the ECB at T= also ncreases ( , 1 1 ). ( 1 Intutve Explanaton Whle a country s GDP s ncrease by 100% of any ncrease n fnancal spenng by the MOF of that country, the ncrease n nterest rates (nflaton offset by an ncrease n nterest rates) by the ECB at T= wll be half of that when there s only one Mnstry of Fnance. Thus, the costs of the euro fallng ue to nflaton are equally share by the two countres. Then, the 100% self-beneft vs. the 50% cost buren wll lea to the free-rer problem, that s, an excessve fnancal ncentve to spen. An ncentve exsts to lmt reuctons n the ssuance of publc ebt (.e., there s an ncentve to free re at the expense of other countres). We wll now express ths n a smple mathematcal way. Each MOF = 1, smultaneously solves the followng problem at T=1: 1 max s.t. 1 The frst orer conton for optmalty of s Rearrangng the left han se of the equaton, we obtan MR MC 100% 1 of SMC Whle the GDP of ths country s ncrease by 100% of the ncrease n fnancal spenng, that s, by 1-α, the ncrease n nterest rates (nflaton offset by an ncrease n nterest rates) at T= by the ECB s one half of the ncrease when there s one Mnstry of Fnance 1 1. As the costs of the euro ecreasng n value ue to nflaton are equally share by two natons, the free rer problem (excessve fscal spenng) appears. We epct the equlbrum pont as S (for the two MOF case) n Fgure. 99

8 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 Fgure. Stacelberg equlbrum pont S: MOFs an ECB..3 n MOFs an ECB What happens f the number of Mnstres of Fnance (MOF) (or the number of natons) n ncreases? Then, each MOF smultaneously solves the followng problem at T=1 The frst orer conton for the optmalty of Rearrangng the left han se, we have 1 1 max s.t. n s n n MR MC 100% 1 n of SMC Whle the GDP of a country s ncrease by 100% of the ncrease n fnancal spenng, that s, by 1, the ncrease n nterest rates (nflaton offset by an ncrease n nterest rates) by the ECB n the secon pero s one n th the ncrease observe when there s one Mnstry of Fnance, that s, 1 n1 n. The costs of the euro fallng ue to nflaton are share equally by n natons. Thus, the free rer problem wll be more severe. Inee we reach the Subgame Perfect Equlbrum Soluton 1 1 1, 1,,.., n n ncreases as n ncreases (Note 1). Summarzng the argument presente n ths secton, we mae the followng proposton. Proposton1 an 1 1 1, whch n Uner the common currency unon wth Monetary Centralzaton an Fscal Decentralzaton, the free-rer problem for the ssuance of publc ebt (excessve fscal spenng) occurs n equlbrum. As a result, the equlbrum polcy nterest rate also ncreases. As the number of Eurozone member countres ncreases, the free-rer problem becomes more severe. Ths proposton mples that uner such a governance structure that no fscal rules (penalty schemes) are enforce among Eurozone countres, member countres always have the ncentve to over-spen, boostng omestc GDP at the cost of common currency s sustanablty. Ths s just the classcal tragey of the commons. That s, the exploson of publc ebt may brng own the common currency, followe by brea-up of the currency unon. 100

9 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; Penalty Scheme n the Stablty an Growth Pact (SGP): Commtment vs. Renegotaton 3.1 Commtment Soluton to the Free Rer Problem (Benchmar) When only one country evates from the agreement, that country pays penalty charges F to the European Commsson (EC). When another country evates, the evatng country pays the penalty charges F to the European Commsson (EC), whle complant countres receve the charges splt among n 1 countres. Thus, a country faces the followng penalty scheme: where s F f That country evates F n1 f One of the other countres evates 1,,.., s the vector of fscal spenng by each country n 1,,.., n When an executor exsts who commts hmself to an enforces such schemes where a penalty s mpose on countres that evate an the penaltes (payments) are transferre to those countres that abe by the rules, the frst best soluton can be acheve (Note 13). 3. What Is Suppose to appen When a Commtment not to Conuct Ex-Post Renegotaton of the Penalty Is Impossble? As scusse n the ntroucton, n real-lfe scenaros, re-scusson (votng) was conucte to etermne whether penaltes shoul be mpose on the evatng countres. When the major powers Germany an France broe the fscal rule of the SGP (by excessve government spenng), ex-post renegotaton occurre. At that tme, the countres nvolve were able to gan from the renegotaton. In exchange for se payments to other countres by Germany an France, no large penalty ( ) was assesse, wth a sgnfcant reucton n the amount of the penalty. Therefore, we propose the followng hypothess., an the barganng power of all remanng n-1 member natons be 1. When Germany (an France) follows an accepts the penalty, the penalty F s pa to the European Commsson (EC) an then allocate to the European Commsson (EC) an the other n-1 countres n the porton of :1. When the penalty s not followe an the se payment 1 F s pa to the remanng n-1 member natons after rect renegotaton, the offer from the evatng country s accepte (Note 14) by the n-1 natons, ecreasng the payment by the evatng country by F 1 F F 0. Ths s the same as the ea of b-rggng n tenerng or colluson n an aucton. Ths weaens the ex-post penalty scheme an maes t more ffcult to suppress ex-ante fscal ncentves at the proper level. ypothess: Let the barganng power of the European Commsson (EC) be 0,1 We now analyze to what egree fscal scplne can be mantane through ths penalty scheme wth ex-post renegotaton. Partal analyss: We focus on a case where equlbrum renegotaton s nuce, le by country α L (Note 15). Let 1 1 1, L, α L,α (Note 16) when there are two countres, n = (α L, α ). be the omnant strategy (Nash) equlbrum fnancal spenng level of types Lettng be the fnancal spenng level of country α, whch the SGP ntens to support (.e., fscal buget), (the free-rer level) shoul be satsfe. Then, the amount of the penalty F s set whle conserng that country α L evates from the SGP an renegotaton occurs between countres α L an α (Note 17). Now, we have the ncentve constrant for country α to support the fnancal spenng (fscal buget) level suggeste by the SGP F 1 Substantal Penalty 101

10 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; F Devaton Gan from to To support the fnancal spenng level of at least, the amount of the penalty F F 1 1 Devaton Gan from to Substantal Penalty s requre an s efne as F( ). Comparatve statstcs yel ΔF( ) Δ, ncatng that to reuce the fnancal spenng level or enhance the fscal scplne of country α, the amount of the penalty F shoul be ncrease. Now, when the ncentve constrant for country α s bnng (holng equal), country α L has an ncentve to evate to the equlbrum omnant strategy level 1 1 L 1 L aganst (1 )F( ) an L suggeste for country α L ( α ), resultng n the falure to satsfy the ncentve constrant on country α L. That s, L (Note 18), specfcally: L F L Devaton Gan from L to L L L Substantal Penalty At ths tme, a se-payment of (1 λ) ( ) emerges n the ex-post renegotaton between countres α L an α. For the se-payment from the country n volaton to other countres n ex-post renegotaton, t s assume that unnecessary spenng (excess buren) 0 occurs for each unt (Note 19). The total welfare n EU countres, whch correspons to fnancal spenng levels L an ( ), s expresse by the followng formula encompassng growth (economy, GDP), nflaton, the costs of the ssuance of publc ebt an the ea weght cost. L L L L x x F 1 L L L L F ( A) ( B) 1 ( C) As the polcy rate s chosen by the ECB so that = 1 ( ), the secon term (B) s zero at equlbrum. The Stablty an Growth Pact (SGP) sets the amount of the penalty F so that the fnancal spenng level s sustane at equlbrum, where the fference between the frst term, the total surplus (A) of the Mnstres of Fnance n both countres α L,α an the thr term (C), the ea weght cost, wll be maxmze. Now, we have the Frst Orer Conton for optmalty: F F margnal effect on the payoff margnal effect on the payoff of MOF of L country of MOF of country margnal ea weght cost 10

11 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 As the frst-orer conton for shows, the fscal scplne of = an the requre amount of the penalty F( ) are etermne so that the margnal ncrease n total surplus an the margnal ncrease n ea weght cost wll be balance. 1 F 1 nto the above conton, we have an explct soluton for Substtutng 1 1 optmal fscal scplne (fscal buget) FB level for Domnant Strategy Level for From (), we can etermne the followng. Frst, the targete level of fscal scplne s weghte by the frst best level an the omnant strategy level (the free-rer level) n the proporton of 1:. As ζ s larger, the vergence of from the frst best s more acceptable. Ths s nterprete to mean that as the ea weght cost ζ s larger when renegotaton occurs, the amount of the penalty cannot be ncrease. Secon, as α ncreases, mantanng the optmal fscal scplne for the hgh margnal cost country α becomes more ffcult,.e., t becomes more ffcult to mantan the fscal buget assgne to country α. Ths mples that when the hgh margnal cost country α loses the trust of the nternatonal bon maret an the nterest rate α for ts publc ebt ncreases (the prce ecreases), the fscal soveregnty of the hgh-cost country α shoul be lmte. In ths sense, the soluton () optmally nvolves the ea of lmte soveregnty. Thr, as ncreases, the optmal fscal scplne for the hgh margnal cost country α ncreases,.e., the fscal buget assgne to country α can be relaxe. Ths can be nterprete as follows. When β s hgher, the ECB can suppress nflaton suffcently an stablze the value of the Euro through the polcy rate (Note 0). Therefore, t s not necessary to mpose a har buget to suppress the fscal spenng of country α. Rather, the fscal buget shoul be relaxe to nuce economc growth. Fnally, as the European Commsson (EC) s porton λ ncreases, the effect of the penalty scheme weaens because a country n volaton can easly compensate countres wth small se payments of substantve reallocaton. ence, the amount of the penalty F becomes larger by the amount of the European Commsson s porton λ. The argument n ths secton s summarze n the followng proposton an corollares. Proposton Uner the Stablty an Growth Pact (SGP), t may be mpossble to honor a commtment not to renegotate the penalty scheme ex post. Then, equlbrum renegotaton occurs ex-post, whch s le by the low margnal cost country α L for the ssuance of publc ebt. The secon best soluton s an optmal fscal scplne (fscal buget) for the hgh margnal cost country α, whch s set at the weghte average of the frst best an omnant strategy (free-rer) levels. In ths sense, the optmal soluton nvolves the ea of lmte soveregnty (Note 1). Corollary.1 Effect of α on As the cost for the ssuance of publc ebt (the nterest rate of publc ebt) α becomes hgher, the optmal fscal scplne (fscal buget) for the hgh margnal cost country α becomes harer,.e., the fscal soveregnty of the hgh-cost country α shoul be more lmte. As the nterest rate of publc ebt α ecreases, the lmtaton on fscal soveregnty shoul be more relaxe. Corollary. Effect of on The better the ECB can suppress nflaton an stablze the value of the Euro through the polcy rate, the more the optmal fscal scplne (fscal buget) for the hgh margnal cost country α can be relaxe. 4. Dscusson 4.1 A Comment from Our Analyss on the Frst Eurozone Crss n 010 (Note ) Although the Eurozone crss emerge n early 010 from Greece, our paper emonstrates ts eep, funamental roots. The n MOF-ECB moel n secton ncates that a common currency wth fscal ecentralzaton 103

12 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 generates the economc cost of a severe free-rer problem, whch taes the concrete form of hgher fscal spenng than the best level through the ssuance of publc ebt by each member naton (MOF). As ths can result n a ecrease n the value of the euro (an an ncrease n the nterest rate on publc ebt), the SGP was establshe as a mechansm of fscal governance to prevent such a free-rer problem. In the Gree fnancal crss, Greece (an α country) shoul have abe by the fscal buget rule n the SGP, but there was a large evaton whch was hen wth forge evence. Eventually, the truth became clear to the maret. Ths cause a sharp fall n prces an an ncrease n nterest on Gree publc ebt. The fference n yel α α L greatly ncrease, where α L s the nterest rate on publc ebt n Germany. As a result, Greece ha to acheve 0, an ts GDP ecrease sgnfcantly. The collapse of the Gree economy has resulte n a ramatc loss of trust n the euro, whch coul result n the en of that currency. Thus, the run of the Gree economy was prevente (.e., Greece was bale out). In August 011, the slowown n the worl economy trggere concern about the solvency of much larger economes than Greece such as Span an Italy. Ths extenson of the crss was reflecte n rsng nterest rate spreas on publc ebt (above the nterest rate on German publc ebt). Durng these few years, some efensve barrers were constructe aganst the soveregn crss cause by Eurozone member natons (le Greece, Span, an Portugal) to prevent the crss from worsenng. One representatve barrer s the European Stablty Mechansm (ESM), whch has scretely reuce the possblty of the collapse of the Eurozone. Although there s stll no smple soluton to the Eurozone crss, our analyss suggests the followng practcal prescrpton. The ex-ante lmte soveregnty (ex-ante power restrcton) erve from the analyss of secton 3 as the secon best moe of fscal scplne an the commtment to ex-post sancton (penalty) (commtment not to renegotate t ex-post) wll both be mportant. Of course, montorng of the observance of the fscal buget rule shoul be strengthene as propose n the multlateral survellance proceure. 4. The Possblty of Fscal Integraton or Fscal Unon From Secton 3, we foun that uner such governance structure that wthout a fscal unon as a companon to the monetary unon, member countres are loosely subject to the penalty scheme n the Stablty an Growth Pact, evatng countres cannot be crebly punshe, especally f they are from the core countres. That s, wthout a fscal unon, a monetary unon tself can harly contan each member country s ncentve of overspenng through ebt fnancng, whch threatens the creblty of common currency Euro an may brng the entre monetary unon to the brn of collapse. So, fnally, we nee to scuss the possblty of Fscal Unon (Fscal Integraton). Fscal Integraton or Fscal Unon theoretcally correspons to the 1MOF + ECB moel n secton, where a Stacelberg Game s playe between the Mnstry of Fnance (MOF) an the European Central Ban (ECB). The net payoff of ths Regme s x Ths payoff s always larger than that of the Regme wth Fscal Decentralzaton an Monetary Centralzaton (MOFs + ECB moel) because the net payoff of Regme 3 s , an t follows that for 1 In Fgure 3, Pont FI represents Fscal Integraton (Fscal Unon) an the ECB equlbrum pont, whch has a greater net payoff than that of the Regme wth Fscal Decentralzaton an Monetary Centralzaton (Pont S). Economcally, ths means that the Fnancal Mnsters Councl (or unque fscal bugetary authorty) rectly controls fscal polces of all Eurozone member natons n orer to nternalze (free-rng) externaltes among them. In other wors, authorty regarng fnancal polcy tself shoul not be ecentralze, but rather centralze nto a unque fscal bugetary authorty. 104

13 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 Fgure 3. Fscal ntegraton (Fscal Unon) +ECB regme Why s ths regme unattanable? In the Eurozone fnancal governance, each member naton may erve some poltcal beneft from ts authorty on fscal polcy an hestate to elegate ths fscal authorty (.e., soveregnty) to some central authorty, wth the result that fscal ntegraton or fscal unon may not be achevable espte the fact that t woul be more economcally effcent. Ths pont seems to be the same as that referre to n the lterature of the Theory of the Frm regarng managers an worers who fear an resst losng ther prvate benefts n a shft from Non-Integraton to Integraton (Note 3). Thus, we have the followng Proposton 3. Proposton 3 If the Fscal Unon hols, effcency wll be enhance uner relatve soveregnty (by ECB an Fscal Unon) because the free-rer problem uner fscal ecentralzaton can be solve through the nternalzaton of externaltes. owever, the Fscal Unon woul not hol f the loss of poltcal benefts to each euro member naton ue to the loss of fscal soveregnty nuce by the fscal ntegraton (Fscal Unon) were large enough to lea member natons to brea away (Note 4). 5. Conclung Remars Ths paper use a game/contract theory framewor to analyze the mechansms of eurozone fnancal governance, wth a focus on centralzaton vs. ecentralzaton an ncentve structures n the European Unon (EU). By constructng a Stacelberg game moel playe by n Mnstres of Fnance (MOF) n the n Euro zone member countres as the frst movers, an the European Central Ban (ECB) as the secon mover, we frst expresse the basc ntuton that the government of each country can create growth (economy) n that country, that s, ncrease GDP (100% self-beneft), by ncreasng government spenng, but that ths wll ncrease nflaton an ecrease the value of the euro. As these effects are share equally by euro countres (the cost s share equally), an ncentve exsts to avo reucng the ssuance of publc ebt, ganng a free re at the expense of other countres. Followng ths framewor, we analyze a soluton to the free-rer problem through the penalty scheme n the SGP, an erve the secon best soluton when a commtment not to renegotate penaltes ex-post s mpossble. Ths analyss shows that lmte soveregnty shoul optmally be mpose on the ssuance of ebt by hgh margnal cost countres. Ths analyss was base on comparatve statstcs, an also explore the effect of the nsttutonal change from QMV (qualfe majorty votng) n the SGP to Reverse QMV n the Fscal Compact (01). Fnally, we close the paper by scussng the possblty of Fscal Integraton (Fscal Unon), where a unque fscal bugetary authorty rectly controls fscal polces of all Eurozone member natons n orer to nternalze (free-rng) externaltes among them, whch means fscal centralzaton as well as monetary centralzaton. Acnowlegements I woul le to than Phlppe Aghon, Olver art, esh Itoh, ea Murase, Taash P. Shmzu, Yoshhro Tsuranu, Tsutomu Watanabe, an partcpants n semnars at the arvar Contracts an Organzaton Lunch, the arvar Internatonal Lunch, totsubash Unversty (CTWE), ose Unversty, the Japanese Economc 105

14 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 Assocaton, the Japan Assocaton for Apple Economcs, an the 16th Worl Congress of the Internatonal Economcs Assocaton, for ther comments an scusson. I woul also le to than arvar Unversty for ts stmulatng acaemc envronment an for the hosptalty I experence urng my vstng scholarshp n Ths wor was supporte by a Grant-n-A for Scentfc Research from the Japan Socety for the Promoton of Scence (C) No ( ) an the Grant from the Zengn Founaton for Stues on Economcs an Fnance n Japan. References Bolton, P., & Dewatrpont, M. (005). Contract Theory. MIT Press. Char, V., & Kehoe, P. (007). On the Nee for Fscal Constrants n a Monetary Unon. Journal of Monetary Economcs, 54(8), Char, V., & Kehoe, P. (008). Tme Inconsstency an Free-Rng n a Monetary Unon. Journal of Money, Cret, an Banng, 40(7), Coase, R. (1960). The Problem of Socal Cost. Journal of Law an Economcs, 3, De Grauwe, P. (007). Economcs of Monetary Unon. Oxfor: Oxfor Unversty Press. Dewatrpont, M. (001). Authorty. Walras-Bowley Lecture at the Econometrc Socety. Grossman, S., & art, O. (1986). The Costs an Benefts of Ownershp: A Theory of Vertcal an Lateral Integraton. Journal of Poltcal Economy, 94, art, O., & olmstrom, B. (010). A Theory of Frm Scope. Quarterly Journal of Economcs, 15(), olmström, B. (198). Moral azar n Teams. Bell Journal of Economcs, 13(), Laffont, J., & Trole, J. (1991). The Poltcs of Government Decson-Mang: A Theory of Regulatory Capture. Quarterly Journal of Economcs, 106, Suzu, Y. (011). A Comparatve Theory of Non-Integraton, Integraton, an the Decentralze Frm. Journal of Internatonal Economc Stues, 5, 3-8. Retreve from Suzu, Y. (e.). (010). Cross-Sectoral Comparatve Analyss of Governance: A Contract & Game Theoretc Approach. ose Unversty Press (n Japanese). Suzu, Y., & Tsuranu, Y. (017). An Incomplete Contract Approach to Eurozone Fscal Governance Commtment vs. Flexblty. Internatonal Economc Journal, 31, Trole, J. (199). Colluson an the Theory of Organzatons. In J. Laffont (E.), Avances n Economc Theory: The Sxth Worl Congress. Cambrge: Cambrge Unversty Press. Trole, J. (013). What Future for the Eurozone Governance? Specal Lecture at Japan Socety of Monetary Economcs, totsubash Unversty, May 013. Tsuranu, Y. (005). Eurozone Fnance Governance through the Stablty an Growth Pact, January 00-November 003. ose Economc Revew, 73(1&). Notes Note 1. The nsttutonal escrpton below s manly base on Tsuranu (005), whch pursues the nsttutonal essence of eurozone fnancal governance from the vewpont of poltcal scence. De Grauwe (007) analyzes the costs an benefts assocate wth havng one currency, as well as the practcal ssues nvolve wth monetary unon. owever, ths paper oes not conuct contract/game theoretcal analyss. We argue that the early nsttutonal framewor of the pero from 00 to 003 was a funamental cause unermnng the creblty an effectveness of the penalty scheme n the Stablty an Growth Pact. Note. Trole (013) consers the Euro crss as an nsttutonal crss, an emphaszes the vewponts of the ncentves an nterests of Eurozone member natons. Suzu (010) proposes a cross-sector comparatve analyss of governance usng a Game/Contract theoretc approach, whch also consers supra-natonal 106

15 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 governance across the Eurozone. Note 3. The central ban of each naton can be regare as an agent that rectly mplements the polcy rate set by the ECB. In terms of ts busness organzaton, Suzu (011) referre to t as ntegraton, where top management gves orers an secton managers smply accept an execute ther orers. Note 4. Ths s bascally an analyss of a 45-egree lne, n whch fnancal spenng (government spenng) ncreases aggregate eman whle ncreasng nterest rates reuce aggregate eman through a eclne n prvate nvestment, a reflecton of the fact that they alter the equlbrum GDP. Note 5. Ths s the vew that an ncrease n fnancal spenng an a ecrease n the polcy rate wll lea to a growth n aggregate eman, resultng n upwar pressure on commoty prces. Note 6. The assumpton 1 means that the ECB s polcy rate has a greater effect on nflaton control (prce stablzaton of the euro) than on GDP (output) x, whch s an ncator of the real economy. Note 7. Ths s an extreme hypothess for smplfcaton. It s possble to generalze ths hypothess to let the Mnstry of Fnance conser economc growth an prce stablty (stablty of the value of the euro) as :1, where 1 1 an to let the European Central Ban conser them as 1 : n a weghte manner. Note 8. Ths tmng s ue to the fact that the ECB can ajust nterest rates much more qucly (every a few wees) than each soveregn naton can ajust ther fscal polces (annually, typcally). Note 9. Dewatrpont (001) etermnes the optmal arrangement of the ve authorty between the MOF an the Central Ban (e.g., the Bunesban n West Germany) for one MOF an Central Ban. Note 10. When 1, the MOF s scplne by the ECB to an excessve extent when choosng, as a hgher leas the ECB to ncrease to reuce nflaton, whch n turn leas the MOF to choose 0 an reuces GDP output to zero. Ths s the ECB-Integraton outcome (see Secton 4) replcate uner a ve authorty structure. We elmnate the case of 1 (a corner soluton case). Note 11. The optmal fscal spenng n T=1, 1, s the omnant strategy for each country, rrespectve of the opponent s strategy. Note 1. Ths result s consstent wth the economc growth n the eurozone, the ncrease n member natons an the accompanyng graual ncrease n the polcy rate untl the summer of 008, that s, before the Lehman shoc. The polcy rate was above 4% n the summer of 008. Note 13. Ths s a soluton to the moral hazar (free-rer problem) n teams through a penalty scheme. For varous forms of penalty schemes, see Bolton an Dewatrpont (005). Note 14. The se payment here can nclue the non-monetary (e.g. poltcal) support n the future from the evatng country. It proves an ncentve for the complant countres to accept the collusve (se payment) offer. Note 15. We assume that equlbrum renegotaton s nuce. The context s an nternatonal settng, mang t an ncomplete contract stuaton. Therefore, equlbrum renegotaton woul be more natural than renegotaton-proof. Note 16. As α s the margnal cost parameter for the ssuance of publc ebt, α ncates a hgh margnal cost ncates a low margnal cost country for the ssuance of publc ebt. country, whle L Note 17. The moelng s base on actual experences n 00 an 003 when the major low margnal cost powers (Germany an France) fale to abe by the SGP s fscal rule. Note 18. As α L α, the nequalty L> tens to hol. Ths mples that the major powers wth smaller margnal costs α L α, such as Germany an France, coul easly bloc the mposton of sanctons through colluson (the formaton of a blocng coalton) among eurozone countres before the vote at the Fnancal Mnsters Councl. Thus, an asymmetry n α L an α s essental to ths argument. Note 19. Ths ea s base on Trole (199) an Laffont an Trole (1991). If 0, we can attan the frst best effcency, just smlar to the Coase theorem (1960). In our nternatonal settng especally, a negotaton or persuason settng, the transacton cost 0 s a natural assumpton. Note 0. Remember that the nflaton rate s formulate as x. Note 1. Conser the effect of an nsttutonal change from QMV n the SGP to Reverse QMV n the Fscal 107

16 jef.ccsenet.org Internatonal Journal of Economcs an Fnance Vol. 10, No. 3; 018 Compact (01) referre to n the ntroucton. The Reverse QMV mples an ncrease n the penalty for evatng countres. owever, the Fscal Compact stll leaves room for ex-post renegotaton. The analyss n ths secton suggests that the Fscal Compact (01) coul mplement more rg fscal scplne than the SGP. Note. Suzu an Tsuranu (017) analyze the Secon Eurozone Crss n 015 by usng an ncomplete contract framewor, an reveals the conflct of nterests between the European Commsson (Germany) an Greece over the Gree ebt repayment plan, from the vewpont of austerty vs. pro-growth. Note 3. See, art an olmstrom (010) an Suzu (011). Note 4. Brext (ext of the Brtan from the EU ) coul be nterprete from ths pont of vew (the loss of poltcal benefts ue to the loss of soveregnty was large enough to lea the Brtan to ext from the EU whch was economcally benefcal). Copyrghts Copyrght for ths artcle s retane by the author(s), wth frst publcaton rghts grante to the journal. Ths s an open-access artcle strbute uner the terms an contons of the Creatve Commons Attrbuton lcense ( 108