Regional economic analysis of power, elections, and secession

And we obtain that integration will occur when the policy preference in two regions issimilar and that secession will occur when the cost of policy adjustment is large.Lastly, we show th

New Frontiers in Regional Science: Asian Perspectives 21

Makoto Tawada (General Managing Editor), Aichi Gakuin University

Kiyoko Hagihara, Bukkyo University

Lily Kiminami, Niigata University

Editorial Board

Yasuhiro Sakai (Advisor Chief Japan), Shiga University

Yasuhide Okuyama, University of Kitakyushu

Zheng Wang, Chinese Academy of Sciences

Yuzuru Miyata, Toyohashi University of Technology

Hiroyuki Shibusawa, Toyohashi University of Technology

Saburo Saito, Fukuoka University

Makoto Okamura, Hiroshima University

Moriki Hosoe, Kumamoto Gakuen University

Budy Prasetyo Resosudarmo, Crawford School of Public Policy, ANU

Shin-Kun Peng, Academia Sinica

Geoffrey John Dennis Hewings, University of Illinois

Euijune Kim, Seoul National University

Srijit Mishra, Indira Gandhi Institute of Development Research

Amitrajeet A Batabyal, Rochester Institute of Technology

Yizhi Wang, Shanghai Academy of Social Sciences

Daniel Shefer, Technion - Israel Institute of Technology

Akira Kiminami, The University of Tokyo

Jorge Serrano, National University of Mexico

Advisory Board

Peter Nijkamp (Chair, Ex Officio Member of Editorial Board), Tinbergen InstituteRachel S Franklin, Brown University

Mark D Partridge, Ohio State University

Jacques Poot, University of Waikato

Aura Reggiani, University of Bologna

This series is a constellation of works by scholars in the field of regional science and

in related disciplines specifically focusing on dynamism in Asia

Asia is the most dynamic part of the world Japan, Korea, Taiwan, and Singaporeexperienced rapid and miracle economic growth in the 1970s Malaysia, Indonesia,and Thailand followed in the 1980s China, India, and Vietnam are now risingcountries in Asia and are even leading the world economy Due to their rapideconomic development and growth, Asian countries continue to face a variety ofurgent issues including regional and institutional unbalanced growth, environmentalproblems, poverty amidst prosperity, an ageing society, the collapse of the bubbleeconomy, and deflation, among others

Asian countries are diversified as they have their own cultural, historical, andgeographical as well as political conditions Due to this fact, scholars specializing

in regional science as an inter- and multi-discipline have taken leading roles inproviding mitigating policy proposals based on robust interdisciplinary analysis ofmultifaceted regional issues and subjects in Asia This series not only will presentunique research results from Asia that are unfamiliar in other parts of the worldbecause of language barriers, but also will publish advanced research results fromthose regions that have focused on regional and urban issues in Asia from differentperspectives

The series aims to expand the frontiers of regional science through diffusion ofintrinsically developed and advanced modern regional science methodologies inAsia and other areas of the world Readers will be inspired to realize that regionaland urban issues in the world are so vast that their established methodologies stillhave space for development and refinement, and to understand the importance ofthe interdisciplinary and multidisciplinary approach that is inherent in regionalscience for analyzing and resolving urgent regional and urban issues in Asia.Topics under consideration in this series include the theory of social cost andbenefit analysis and criteria of public investments, socio-economic vulnerabilityagainst disasters, food security and policy, agro-food systems in China, industrialclustering in Asia, comprehensive management of water environment and resources

in a river basin, the international trade bloc and food security, migration and labormarket in Asia, land policy and local property tax, Information and CommunicationTechnology planning, consumer “shop-around” movements, and regeneration ofdowntowns, among others

Researchers who are interested in publishing their books in this Seriesshould obtain a proposal form from Yoshiro Higano (Editor in Chief,

higano@jsrsai.envr.tsukuba.ac.jp) and return the completed form to him

More information about this series athttp://www.springer.com/series/13039

Moriki Hosoe

Faculty of Economics

Kumamoto Gakuen University

Kumamoto, Japan

ISSN 2199-5974 ISSN 2199-5982 (electronic)

New Frontiers in Regional Science: Asian Perspectives

ISBN 978-4-431-55895-8 ISBN 978-4-431-55897-2 (eBook)

The use of general descriptive names, registered names, trademarks, service marks, etc in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.

The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer Japan KK

The registered company address is: Shiroyama Trust Tower, 4-3-1 Toranomon, Minato-ku, Tokyo

105-6005, Japan

The main object of this book is to explain some of the mechanisms of integrationand secession among regions from the point of view of game theory To attain thisgoal, the incentives of regions in a country to unite or to secede and the conditions

in the way for the member countries of a federation to leave peacefully or by way

of violence are examined

Since 1990, the Soviet Union drastically split into 15 independent countries;Yugoslavia was divided into 6 countries after the bloody civil war In Asia, tensionsover secession have become more and more serious, including separatist tensions inChina, India, Iraq, Myanmar, and Sri Lanka Why do such regional conflicts occur?Are there any devices to solve them?

In this book, we try to answer such questions, and study the movement of regionalsecession and integration in each region, like culture, ethnicity, or religion There areseveral important resources of power in regional conflicts One of the most importantfactors in the bargaining power of conflicting parties is ethnicity Ethnicity is anelement that dominates the common feelings of people in many cases Therefore,committed power in regional conflicts is very important Another factor is religiousfervor A religious mind and religious enthusiasm are different things However, incountries or regions in ethnic crisis situations, the two are often united

We also investigate the influence of power of a third region on secession We canraise the recent consolidation by Russia of the Crimean Peninsula as an example ofthis third party effect Election is also an important power to solve regional conflicts.Therefore, how the formats of regional election or referendum are designed iscritical as some commitment devices of secession The decentralization of power

in tax-subsidy policy and in the political governance form in two regions is shown

to be important as a way to avoid wasteful conflict for the secession incentive of

a minority region We point out that the natural resources in secessionist are oftencritical as bargaining power over secession conflict How those powers depend onheterogeneity costs associated with different preferences over the type of region andthe relative size of the two regions is shown

v

Lastly, we would like to thank the editorial board at the Japan Section of theRegional Science Association for compiling this volume as one of the series of NewFrontiers in Regional Science: Asian Perspectives, in particular, Professor YoshiroHigano, who kindly reviewed the contents of this book as editor in chief.

August 2017

1 The Issue of Regional Conflict and Secession 1

Determines National Size? 161

Takashi Kihara

Index 193

vii

The Issue of Regional Conflict and SecessionMoriki Hosoe

1.1 The Issue of Regional Conflict and Secession

Regional conflicts, which have persisted from ancient times to modern times, haveincluded ethnic wars and wars of conquest Secession has been the cause and result

of regional conflicts During the last few decades, many national breakouts haveoccurred Aleksandar and Radan [2] provide a comprehensive guide to the secessionissues The most dramatic examples were the big bang in Eastern Europe afterthe disintegration of the Soviet Union Even today, from Catalonia and Scotland

to South Sudan and Kashmir, secessionist movements demanding independenceare occurring throughout the world under all kinds of political and economiccircumstances A short look at two histories in Asia is illustrative of the causes

of the breakout and formation of states One is related to Bangladesh The other isrelated to Timor

In the sixteenth century, the Bengal region developed into the center of commerceand industry under the Mughal Empire The UK controlled most of the Bengalregion as England’s East India Company was formed at the end of the eighteenthcentury After Britain withdrew from colonial India in 1947, districts in whichMuslims constituted a majority separated and became independent as Pakistan Thedistrict in which a majority of inhabitants believed in Hinduism was separated as

M Hosoe (  )

Faculty of Economics, Kumamoto Gakuen University, Kumamoto, Japan

e-mail: hosoe@kumagaku.ac.jp

© Springer Japan KK 2018

M Hosoe (eds.), Regional Economic Analysis of Power, Elections, and Secession,

New Frontiers in Regional Science: Asian Perspectives 21,

https://doi.org/10.1007/978-4-431-55897-2_1

1

India What is now Bangladesh became East Pakistan Partition was followed by warwith India over Kashmir, mass migration of Muslims, and resettlement of Hindusand Sikhs within the new borders This led to violence, financial loss, and death

on a large scale Pakistan therefore comprised East Pakistan and West Pakistan,geographically separated by 1600 km Muslims lived in both places, but economicand political disparities between the two regions spread As a result, two regions hadbeen under serious conflict Since the Urdu language used in West Pakistan Provincebecame the national language, the civil war began The Awami League, based in EastPakistan with a high population, was eventually victorious In December 1974, theindependence of Bangladesh was acknowledged Several million people were killed

or displaced by the India-Pakistan war and the Bangladesh war for independence.Given this history, the main causes of independence and secession of Bangladeshare the differences in language, geographical remoteness, economic disparity, andWestern colonialism

Let us raise the case of Timor-Leste as another example of regional conflict inAsia Since the sixteenth century, Timor Island was occupied by the Netherlands

in the western half and Portugal in the eastern half After World War II, EastTimor continued as a Portuguese colony after the former Dutch territory becameindependent as Indonesia The independence movement also strengthened in EastTimor, as a result of the democratization of Portugal East Timor declared itsindependence in 1975 However, the independence faction and a conspiracy withIndonesia conflicted, erupting into civil war Reportedly, about two million peoplewere killed The Indonesian Soeharto administration came to consolidate thisregion However, in 1999, following the United Nations-sponsored act of self-determination, the referendum had a majority in favor of independence, and EastTimor established the independence through intervention of a multinational force ofthe United Nations In this case of Timor-Leste, ethnic conflict or religious conflictwas not necessarily a main cause of regional conflict Development of nationalismwas promoted since Portuguese colonization Severe poverty under Indonesianrepression provided the main momentum driving the push toward independence

In modern democratic societies, regional conflicts that arise with secession areresolved in most cases by peaceful means, contrasting against bloody conflict inregions in Asia These controversies are subjected to ardent parliamentary delibera-tions and become important issues in parliamentary elections or referendums Theseare often brought to judicial decisions

For example, regarding the independence issue in Quebec, Canada, the centralgovernment had asked the court to rule on the legality of a unilateral Que-bec declaration of independence after Quebec implemented the referendum TheSupreme Court of Canada ruled in 1998 that neither the Canadian constitution norinternational law allows Quebec to secede from Canada unilaterally It also showed

that if a clear majority of Quebecers unambiguously opts for secession, the federalgovernment and the other provinces would have a constitutional duty to negotiate.The problem of Scotland secession so far seems to be in the peaceful process.However, regional conflicts do occur in gray zone in modern times The recent con-solidation by Russia of the Crimean Peninsula is regarded as a solution by peacefulcamouflaging of power In this case, fierce conflict arose because the independentcountry of Ukraine, created after the disintegration of the Soviet Union, consists

of two regions, Western Ukraine and Eastern Ukraine, with ethnical, cultural, andeconomic differences Western Ukraine is anti-Russian, whereas Eastern Ukraine

is pro-Russian Crimea is home to many Russians and is characterized by beingclose to Russia in geopolitical terms In Ukraine, the constitution established duringdisintegration of the Soviet Union stipulated that separation of the country should

be decided by referendum of the entire country Following disintegration of theSoviet Union, people in Eastern Ukraine were dissatisfied because democratic,pro-European, and pro-American governments were formed Under the backing

of Russia, the Crimean parliament conducted a referendum for the separationand independence of the Crimea from Ukraine, obtaining overwhelming approval.Through it, the Crimean parliament declared separation and independence Westerncountries have since imposed economic sanctions against Russia (see A Bebler[3]) Most countries of the international community have not approved this inde-pendence, labeling it as de facto intervention by Russia However though there areproblems in the political procedure, but from an ethnic point of view, this secessioncan be understood Anyway this case shows that intervention of third countriesinto secession conflict is an important issue in understanding situations from theperspective of international politics, which we discuss in Chaps.4 and 5 of thisbook

Why do regional conflicts and secession movements occur? To understand this, it isbetter to consider the reasons for national existence From an economic perspective,

a fundamentally important role for the state is to offer public goods to its citizens.Alesina and Spolaore [1] and Bolton and Roland [5] argue that the size of sovereignnations is a trade-off between benefits and cost A fundamentally important rolefor the state is the supply of public goods for consumption, which are generallynoncompetitive Therefore, public goods are provided on the scale of the nationaleconomy This is a main merit of the scale of nation However, larger national scaleentails larger management costs and congestion costs Particularly, independence

is likely to bring diversification of preferences for public policy Such diversitiesinclude different cultures, languages, ethnic groups, and religions These mightmake it difficult for society to harmonize and coordinate Costs rise concomitantlywith larger size Such a trade-off between economies of scale and heterogeneity

of preferences might occur Therefore, the number of countries is determined from

prevailing conditions of economy and geography, culture, and power of violence.These movements of breakup of countries correspond to development of secessionmovements around the world.

In Chap.9 of this book, we present an empirical examination of the effects

of an increase in “country size” in terms of total population or real GDP Thoseeffects are assessed particularly on government expenditures per GDP as a proxyfor public goods provision inefficiency, on the trade volume per GDP as a proxyfor trade openness, and on the per-capita GDP growth rate as a proxy for economicperformance Data are averaged over 11 nonoverlapping 5-year periods Particularly,costs of heterogeneity, exemplified by religious and linguistic fractionalizationindices, are found empirically to have significant effects on the inefficiency of publicgoods provision Scale economies in supplying public goods are found empiricallyeven for such populous countries with greater than 500 million people

1.2 Power of Conflicting Parties in a Secession Movement

The title of this book is The Regional Economic Analysis of Power, Elections, and

Secession What is power in the context of secession? It is related to the respective

bargaining capabilities of conflicting parties There are several important resources

of power in regional conflicts As the above examples illustrate, an important factor

in the bargaining power of conflicting parties is ethnicity Ethnicity is an elementthat dominates common feelings of people in many cases Therefore, it is a veryimportant committed power in regional conflicts Another factor is religious fervor.Religious mind and religious enthusiasm are different things However, in countriesand ethnic crisis situations of countries or regions, the two are often united Ingeneral, regional conflicts are regional joint actions to seek common interests orjoint values Collier and Hoeffler [7] reports the following

Secessions depend upon the invention of an imagined political community and natural resources will often be instrumental in transforming this invention from the pipe-dream

of a handful of romantics to the reality of a large political or military organization.

In this meaning, the symbols or common values in secessionist movement areimaginary, as is nationalism However, as the example given above illustrates, mostregional conflicts are based on some substantial values at least among people in thesecession

Another factor that can be cited as reflecting bargaining power in regionalconflicts is cultural conflict, such as language Because of economic disparity,opposition also becomes a springboard in regional conflicts, in some cases Ofcourse, differences related to ethnicity, region, language, and culture do not transferdirectly to conflict over secession, but confronted by a combination of these factors,

a secession movement might occur

This book does not analyze separately the problem of power derived from thecharacteristics of people, such as ethnicity, religion, language, and culture This

line of research is left to several empirical research books We summarize severalcharacteristics embedded in people as an important one of bargaining power in asecession movement.

Actually, in many chapters of this book, we introduce the typical utility of

citizens in a region as presented below U A is a representative utility of citizen in

region A.

U A D a N X A  X A/2 s.X A  X B/2C C (1.1)

Here, X A stands for the contents of a single issue of public policy NX A represents

an ideal point of the issue for the citizen in region A This is the value of the characteristics embedded in citizen in the region A, like ethnicity , religion, language, and culture Then, the regional policy is to ascertain the contents X A of

the issue X B is determined by the authority of another region B If parameter s is positive, then the harmony of this public policy is desirable in the region A If s

is negative, the disparity of public policies is desirable C is a private good Then

our fundamental problem is who determines the public policy and how two regionscoordinate public policies This is our basic model of the secession problem Thisideal point of public policy is a power derived from the characteristics embedded incitizen in the region

In this book, we introduce elections, third parties, and natural resources as otherpowers of bargaining We investigate the influence of various power in a bargainingprocess of secession on the behaviors of conflicting parties and the results ofsecession movements

1.2.1 Election as a Commitment Device of Interest

Coordination Among Diverse Actors

Though citizens have rigid characteristics that include irreversible ones, the teristics of them are considerably diverse A useful device to coordinate such diverseopinions and feelings is an election In modern democratic societies, politicaldecision-making is based on elections People in democratic society more or lessbelieve that political decisions under election are acceptable, although we havesome exceptional cases Therefore, even in the shadow of recession, the decisions ofpublic policy are taken under elections In many chapters of this book, we investigatesome election models to cope with secession difficulties Elections are the onlyinstrument of democratic and peaceful transfer of power with the consent and choice

charac-of the majority Therefore, results charac-of election reveal the bargaining position oversecession We adopt citizen candidate models as election models in many chapters

of this book The citizen candidate model has three stages: (1) citizens decide onwhether to stand for office; (2) citizens vote for candidates who stand; then thewinner is elected by plurality rule; and (3) the candidate who is elected implements

the candidate’s own most-preferred policy (see Osborne and Slivinski [10]) Usingthis citizen candidate model, we can form public policy in each region in the shadow

of secession Though the citizen candidate model reflects the ideal image of election

in a democratic society, more or less, it describes closely the reality of representativeelections based on a mature democratic society

However, it is important to distinguish between direct democracy, in which zens participate actively in governing their own political affairs, and representativedemocracy, in which the people elect representatives to carry out the business ofgovernment Direct democracy is connected directly to the idea itself of democracy.Therefore, societal dissatisfaction often requires direct elections

Elections are often adopted to resolve regional conflicts One type is the referendum

In 2016, Britain voted to leave the European Union in a historic referendum.Sometimes referendums are subject to populism Therefore, designing the votingsystem is an important problem A mere simple majority does not provide sufficientlegitimacy An example of a supermajority requirement was used in 2006 inMontenegro, where the law stipulated that independence would be approved ifsupported by 55% of those eligible to vote (see Qvortrup [12]) According toPavkovic and Radan [11], the total turnout of the referendum over breaking thestate union with Serbia was 86%; 55.5% voted in favor and 44.5 were against

In addition, ex ante approval is necessary in many cases to hold a referendum

on independence For example, the Scottish independence issue requires a simplemajority of the local referendums in the region claiming separation However,prior to that, approval of the referendum implementation by the UK Parliament isnecessary How the central government intervenes in the procedure of secessionbargaining is difficult to solve As an extreme case, China does not allow anyregional separation by its anti-separation law This is possible only in politicaldictatorship Ukraine is premised on nationwide referendum implementation beforethe referendum in the conflicted region This mechanism discourages a movementfor a referendum vote However, in rare cases, if the area in question is bad for thecentral government, then consent for independence can be obtained by referendum

In the case of Ukraine, although not conducting a nationwide referendum, a localreferendum in Crimea was held as an excuse for the international community, withillegal referendum voting of Ukraine against the background of Russian ArmedForces in Crimea The inability to prevent Parliament’s declaration of independencemade virtual independence virtually feasible Radan and Pavkovic pointed out thefollowing: Between 1900 and 2010, 44 of 60 secessions or secession attempts werepreceded by referendums Among those 44 referendums, war broke out in 6 cases

In other words, secession was achieved peacefully in 38 (86%) of the cases

In Chap.7 of this book, we analyze the referendum system mechanism as ademocratic procedure for coordinating public opinion by a simple voting model

based on the Hotelling model This study introduces an additional vote as thedemocratic procedure The referendum must be approved by most of the wholenation A main outcome is that the vote for approving referendums as a democraticprocedure engenders suppression of secession when secession is socially desirable.

When secession is a looming risk, central governments can use political ization as a substitute for costly violence to reduce the risk For example, Myanmarhas suffered one of the world’s longest running civil wars, one rooted in ethnicconflict The creation of a federal state is an issue that has been at the heart of thecountry’s politics long before it gained independence in 1948 This is a key issuefacing the country’s first democratically elected government in decades Politicaldecentralization refers to the granting of local autonomy over policies important

decentral-to local identity such as education and culture, as well as overt policies important

to local welfare such as environment and health, combined with the buttressing ofself-government by allowing local voters to elect their own representatives.Federalism can be seen as a way of stabilizing, or making credible, decentralizedgovernmental structures Through fiscal federalism, considerable policy and fiscalpowers are deleted to subcentral governments According to Oates [9], Weingast[15], and Sorens [13], the typical form of federalism has the characteristics ofthe following Firstly, subcentral governments exercise considerable control overeconomic policy and expenditure competencies but face a hard budget constraintbecause they largely fund their own expenditures from their own taxes The hardbudget constraint is a critical point of successful federalism because the budget isoften soft under the threat of secession Secondly, a common market permits freemovement of labor, capital, and goods across jurisdictional borders The system

is institutionalized and self-enforcing because the central government has madecredible commitments not to interfere in the legally authorized decision-making

of subcentral authorities Therefore, governments that are concerned about the risk

of secessionist insurgency and their ability to combat one might provide politicalautonomy without much economic autonomy The central government has to designcarefully how the balance of political autonomy and economic autonomy should

be taken Since Myanmar has so many ethnic groups with different economicsituations, it might be desirable to follow a soft federal system because some ethnicgroups want to promote the rights of their people and the others are federal statesthat ensured equal rights and included power-sharing agreements

However, decentralizing economic power to regional governments might alsoraise the risk of actual secession Federalism was once regarded as an effectivegovernance system as a way of integration in a pluralistic society However, incountries where the federation system was adopted and promoted decentralization,various problems such as separation and independence problems have burst out inrecent years, and doubt is given to the effect Is the federal system a governance

system that brings conflict and separation? Or resolve the conflict and bring aboutintegration?

In the second half of Chap.2, we investigate the relationship between cal governance and secession We consider two integration schemes, centralizedintegration and decentralized integration Under the centralized integration, publicpolicy is unified in two regions However, under decentralized integration, deferentcontents of public policy may be applied to each region Decentralized integration

politi-is a kind of federation Then we show that centralized integration will occur whenthe relative population in minority region is small, and secession will occur whenthe relative population in the minority region is larger than a certain level And

we obtain that integration will occur when the policy preference in two regions issimilar and that secession will occur when the cost of policy adjustment is large.Lastly, we show that, depending on the cost level of public goods, the two dynamicprocesses of going from centralized integration to decentralized integration, viasecession, and of going directly from centralized integration to decentralizedintegration will occur with the increase of the relative population in the minorityregion

Geographical bias of natural resources is often the cause of regional conflict,and it intensifies the conflicts When the presence of a local ethnic group has ahigh ability of access to a large natural resource, this local ethnic group might

be better off if it were independent In such a case, they have incentives to start

a secessionist rebellion Secessionist movements have been linked to importantlocal natural resources Angola, Myanmar, Democratic Republic of Congo, andMorocco are good examples As a recent example, we have a secession movement

in Scotland According to House of Lords (2013), if Scotland were able to controloil and gas from the North Sea perfectly under a secession economy, it might besaid that the per-capita income in Scotland would increase under an integratedeconomy

To date, several studies have analyzed the influences of natural resources onthe regional conflict (see Hodler [8]) Actually, natural resources engender moreconflicts in fractionalized countries Besley and Persson [4] emphasize that weakinstitutions, low income, and large natural resources engender a greater risk of civilwar Caselli and Coleman [6] specifically examine the decision of the dominantethnic group to exploit or not to exploit other groups in terms of the proceedsfrom natural resource extraction, but do not address how the geographic distributionand the economic features of natural resources affect the risk of ethnic conflict ofdifferent kinds In Chap.6 of this book, we also consider the problem of a profitdistribution of a natural resource under secession Results show that if tax revenuefrom natural resources increases, then the optimal level of transfer to deter secession

increases Furthermore, if the degree of public policy externality increases, then theoptimal level of transfer to deter secession decreases.

Natural resources engender more conflicts in fractionalized countries Attempts

to secede often entail the threat or actual use of force, the mobilization of vastresources, and high human and material costs The threat of civil war is theworst power of the bargaining process in secession Spolaore [14] shows that theprobability of secession and the amount of resources diverted to separatist conflictdepend on factors such as heterogeneity costs, economies of scale, relative size, andexternal threats In Chap.8of our book, we investigate the possibility of civil war interm of management of resource and regional redistributive policy Using a contestmodel, we show that the optimal initial sharing rate of vital resources in the regionincreases with the rate of sharing in the case of secession and also increases with thepenalty sharing rate The winning probability of a region in the civil war decreases

in all budget scales and also increases in the scale of resources

Wimmer, Cederman, and Min [16] show that the likelihood of armed frontation increases as the center of power becomes more ethnically segmentedand as greater proportions of a state population are excluded from power because

con-of their ethnic background In the last section Chap.7 of this book, we showthat lower income, larger national size in terms of population and total product,greater linguistic heterogeneity, larger military force, and deterioration of policyand institutional environments are associated with more frequent armed conflicts,including territorial as well as international conflicts These factors commonlyincrease the probability of the occurrence of armed conflict

1.2.5 Influence of Third Parties on Secession Movements

rival region benefits by integration But there is a possibility that the rival regionbecomes a second-mover Chapter5investigates the secession game of rival region

as a second-mover Then it is critical how integration works as a commitmentdevice The commitment effect makes the integrated region more aggressive, sothat the region selects the more extreme policy The other region does not favor theintegration when this effect is large

Finally, we raise the difficulty of openness of information related to politicalinterests and economic interests Asymmetric information influences the bargain-ing power among conflicting parties in the case of secession In nondemocraticcountries, and even in democratic countries, bribes and corruption often occur.They constitute a source of power In Chap.3, using a citizen candidate model, weexamine the effect of spillover and rent-seeking on the integration and secession

in two regions The residents vote for a representative considering not only withbargaining of integration but also with rent-seeking activity by a representative.Consequently, in case there are spillover effects and rent-seeking effects, we showthe possibility of secession even if the spillover effect is large and show that theexistence of rent-seeking raises the possibility of secession

8 Hodler, R., (2006), “The curse of natural resources in fractionalized countries”, European Economic Review, 50(6): 1367–1386.

9 Oates, W., (1985), “Searching for Leviathan: An Empirical Study”, American Economic Review, 75: 748–57.

10 Osborne, M J and A Slivinski, (1996), “A Model of Political Competition with

Citizen-Candidates”, The Quarterly Journal of Economics, 111(1), 65–96.

11 Pavkovic, A and P Radan, (2007), Creating New States: Theory and Practice of Secession,

14 Spolaore, E., (2008), Economics of Governance, 9: 45–63.

15 Weingast, B R., (1995), “The economic role of political institutions: Market-Preserving

federalism and economic development”, Journal of Law, Economics, and Organization, 11(1):

1–31.

16 Wimmer, A., L Cederman, and B Min, (2009), “Ethnic Politics and Armed Conflict: A

Configurational Analysis of a New Global Data Set”, American Sociological Review, 74(2):,

316–337.

Secession, Representative Elections, and Political Governance

Moriki Hosoe

2.1 Introduction

In this chapter, we analyze the endogenous determination of integration andsecession by introducing policy preference and policy conflict cost, stressing onrepresentative election and political governance A seminal paper in this field isBuchanan and Faith [6]

On history, the confrontation of policy has often produced the conflict amongregions For example, the conflict on the slave policy among states triggered the civilwar of the United States of America Most of ethnic conflicts have the difference ofpolicy preference among ethnic groups We have often seen that a third region had agreat influence to the decision of integration and secession among regions Westerncountries had significant influence of the secession of Eastern European countriesfrom the block of Soviet Union Also the Japan-United States Security Treaty hasbeen related to the possible threat of Far Eastern regions We have seen numerouspolitical episodes where the existence of third region influenced the result of theintegration or secession among several regions

In modern democratic society, regional conflicts with secession are in most casesresolved by peaceful instruments, as in Quebec or Scotland Regional conflictsoften become the important issues in parliamentary elections or referendum andare discussed under ardent parliamentary deliberations These are often brought tothe judicial decisions As for the literatures of new political economics, we havePersson and Tabellini [16], Persson et al [17,18] We firstly clarify the working ofelection for the integration and secession using the citizen candidate model To do

so, we introduce a simple policy preference model with two regions and examine

M Hosoe (  )

Faculty of Economics, Kumamoto Gakuen University, Kumamoto, Japan

e-mail: hosoe@kumagaku.ac.jp

© Springer Japan KK 2018

M Hosoe (eds.), Regional Economic Analysis of Power, Elections, and Secession,

New Frontiers in Regional Science: Asian Perspectives 21,

https://doi.org/10.1007/978-4-431-55897-2_2

13

how the distance of policy preference in two regions influences the integration andsecession via the bargaining of their representatives who are elected by citizens ineach region (see Grossman [10] and Grossman [11]) The idea of policy preferencecomes from the stream of political science (i.e., Epstein and O’Halloran [7]) When

a majority region commits a policy, we assume there is a policy externality whichmeans that the utility loss may occur to citizen in a minority region Therefore thispolicy externality will influence the possibility of integration between two regions.Secondly we examine how decentralizing political power to the regional levelmay address citizens’ demands and reduce future secessionism Decentralizing eco-nomic power to regional governments may raise the risk of actual secession Accord-ing to Sorens [19], governments that worry about the risk of secessionist insurgencyand their ability to combat one may therefore provide political autonomy withoutmuch economic autonomy Federalism can be an effective means for preventingseparatism However, it may also promote separatism as we mentioned in Chap.1

In Sect.2.2, we introduce a basic model in which the situation of regional conflictbetween two regions is formalized and the problem of representative electionand policy decision are raised In Sect.2.3, under the possibility of integration,

we analyze several properties of representative elections in two regions In anequilibrium, we clarify the mechanism of integration and secession In Sect.2.4,

we show that a centralized integration will occur when the relative population inminority region is small and secession will occur when the relative population inthe minority region is larger than a certain level In Sect.2.5, we investigate aboutthe two types of integration, centralized integration and decentralized integration

2.2 The Model

Suppose that there are two regions A and B, with the population of N A and N B The

region B is a minority region, that is, N A > N B Let us normalize N A C N B D 1

Each region faces the policy choice of a single and common policy issue X, and

the bliss point of policy of each representative citizen is assumed to be NX R; NX R,respectively Then as the actual policy decision is far from the bliss point, the utility

of representative citizen decreases If the policies that each region implements for acommon policy issue are different, a friction for the regional economy may occur.This is called policy externality We confine our discussion to the policy externality

of majority region to minority region

Meanwhile this economy is under the return-to-scale economy, and the per capitaincome is 0i D h i N i for i D A; B Here, h iis marginal productivity with respect to

the scale for region i Let assume that h A > h O> 0

When each region implements a policy X A ; X B representatively, then the utility

U A ; U Bof representative citizen is shown as follows:

U B D  NX B R  X B/2 s.X B  X A/2C 0B (2.1)

U AD  NX A R  X A/2C 0A (2.2)

The utility of the bliss point is normalized to be zero s is marginal cost of policy

externality due to the policy difference and is assumed to be a positive constant lessthan 1

In the beginning, let us seek the equilibrium policy that each region independentlymakes Each representative citizen in two regions chooses the optimal policy for his

region Then the optimal policy X B0in region B, given the policy X O

Therefore, as the policy externality increases, the utility of minority region

decreases Obviously, the utility of majority region A is not influenced by the

increases and personal income increases So when region A and region B integrate and a policy X A is implemented, the utility of the representative citizen for region

A is

U I AD  NX A R  X A/2C I A;where

Then the utility of representative citizen in the minority region B is

U I BD  NX B R  X A/2C I B;where

I B D h A N A C N B/  B h A  h B /N B

Generally, it is held that B> A

The process of the bargaining for integration is assumed that region A offers a unified policy X A to the minority region O Furthermore region A may offer the lump-sum money F for the region B not to refuse the offer This lump-sum money

is offered to guarantee the utility of region B under the state of the secession That

Therefore, the optimal policy is represented by



N AC s

1 C s

 NX R A NX B R/2C F AB 0 ; (2.10)where

ABD N A

N B

.I A 0A/ C I B 0B/;

which is the total scale effect of integration between A and B.

From (2.7), the integration incentive does not occur without scale effect That is,

Lemma 2.1 In two regions, the integration does not occur without scale effect And

the efficient policy is realized by integration.

Also, from (2.7), as the distance of policy preference in two regions is large, thepossibility of secession is large, and as the marginal policy external cost is large, thepossibility of integration is large in Fig.2.1

Lemma 2.2 As policy conflict cost s is sufficiently large, the possibility of

integra-tion becomes large.

By rearranging (2.10), we can show how the population ratio in two regions andthe distance of policy preference between two regions influence the possibility ofintegration and secession:

Let us denote the left-hand side of (2.11) by S.N A ; s/ Then S.N A ; s/ increases

in N A when policy conflict cost is small From this, we show that as in Fig.2.1,

given a constant distance of policy type, secession occurs when the population N A

in majority region is above a certain level On the other hand, as in Fig.2.2, when

policy conflict cost is large, S.N A ; s/ is of U-shape in N A Then if the type distance

is small, (1) integration occurs when the majority population is not so large (2)Secession occurs when the majority population is in intermediate interval, and (3)integration occurs again when to majority population is sufficiently large On theother hand, in Fig.2.2, if the type distance is large, (1) secession occurs when themajority population is smaller than a certain level, and (2) integration occurs whenthe majority population is larger than the level

These properties are explained as follows In general, policy coordination effect

of integration increases when the majority population decreases Furthermore theeffect of integration increases as the policy conflict cost increases Therefore sincepolicy coordination effect of integration is small when the policy conflict cost issmall, the increase of majority population strengthens the merit of integration Onthe other hand, when the policy conflict cost is large, the decrease of majoritypopulation strengthens the policy coordination effect of integration and widens themerit of integration in addition to the case of the large population in majority region.Therefore we obtain the following lemma:

N A S(N A , s)

integration secession integration

( ¯X A − ¯ X B)

1/2

1

Fig 2.2 Large policy-conflict cost case

Lemma 2.3 When the policy conflict cost s/ is small, the possibility of integration

increases as the population of majority increases On the other hand when the policy conflict cost is large, the possibility of integration increases not only in the large population of majority but also in the sufficiently small population.

2.3 Election of Representative and Integration Conditions

So far we have discussed the integration problem as the policy preference ofrepresentatives in two regions given Here we will endogenously decide the type

of policy preference of each representative via regional elections

To do so, we assume that the distributions of policy preference for each citizen

in two regions are given and the average of policy preference, NX Band NX A, are equal

to median value, respectively Without loss of generality, the following is assumed:

Assumption 2.1 NX B< NX A

2.3.1 Representative Election Under Secession

In the bargaining, let us examine the case of no integration Considering (2.3) and(2.4) of equilibrium policy when the type of representative in each region is given,citizens in each region will elect a representative

The optimal type of representative for each citizen in majority region is obviouslyhimself since there is no policy externality from minority region Therefore, the type

of representative in majority region is elected by median voter theorem

On the other hand, when each representative NX R, NX R is elected, the utility of a

citizen i in region B with N X Biis:



N

Then the optimal type NX R

Bi of representative for citizen i is obtained by maximizing

the above utility From this maximization problem, NX R

Biof representative for citizen

i is

X Bi R D NX Bi

That is, the optimal type of representative for citizen i is also himself though there

is a policy externality This is explained by the fact that the type selection of the

representative in majority region is not influenced by the type of the representative i

minority region

Thus the elected types of the representative in two regions are the median citizen

In case of no secession, the policy externality does not influence the behavior ofcitizen in two regions

Lemma 2.4 In the case of no integration, the elected representative in each region

is the median citizen, which implies that the policy externality does not influence the behavior of citizen in two regions.

2.3.2 Representative Election Under the Bargaining of

Integration

Now we consider the representative election problem under the bargaining ofintegration In Sect.2.2, we examined the optimal policy under the bargaining ofintegration between the representative NX R and NX R Therefore, consider that theoptimal policy is (2.7)

So when each representative NX R and NX R is elected, using (2.7), the utility of a

citizen i in region A with bliss point N X Aiis shown as follows:

Therefore the optimal type NX R

Ai of the representative for a citizen i in region A is

obtained by maximizing the above utility From the first-order condition of thismaximization problem, we get

N AN B

N A

s 1Cs

From this condition, the optimal type of the representative is a linear function of thebliss point of the citizen Therefore, using median voter theorem, we show that as anoptimal response for the type NX B R of the representative in region B, the elected type

N AN B

N A

s 1Cs

(2.12)

The elected type of the representative in region B is obtained in the same way.

When each representative NX R

B and NX R

A is elected, using Eq (2.8), the utility of a

citizen i in region B with bliss point N X Biis shown as follows:

Therefore the optimal type NX R

Bi of the representative for a citizen i in region B is

obtained by maximizing the above utility From the first-order condition of thismaximization problem, we get

the representative in region A, the elected type N X R

B in region B is obtained as follows:

.N A  N B/

N AC N1

A

s 1Cs

But the elected type of the representative in region A is smaller than that in region B

if the adverse inequality of (2.17) is held Therefore, when the population in majority

region is not so large, the elected type of the representative in region A is smaller than that in region B if the policy conflict cost s is sufficiently large.

Here, let us check the sign of the denominators of Eqs (2.14) and (2.15) Thedenominator is positive if (2.17) is held, and the denominator is negative if theadverse inequality is held Furthermore, when (2.17) is held, the coefficient of NX Ain(2.14) is shown to be positive from the following inequality:

A is not elected between the median types in two regions, but elected outside of the

median type in minority region

On the other hand, if (2.17) is not held, the denominator of the right-hand side in(2.14) is negative So we will rearrange (2.14) as follows:

.N A  N B/

N AN1

A

s 1Cs

then the type of the representative in majority region is elected between the mediancitizens in two regions On the other hand, if (2.20) is not held, the type of therepresentative in majority region is elected outside of the median citizen in majorityregion Therefore, the following lemma is held from (2.14).

Lemma 2.5

1 If (2.17) is held, in an equilibrium, the type distance of the elected representatives

decreases in policy conflict cost s.

2 If (2.17) is not held and (2.20) is held, the type of the representative in majority

region moves to the same direction with the median citizen in majority region.

3 If either (2.17) and (2.20) are not held, the type of the representative in majority

region moves to the opposite direction with the median citizen in majority region.

Now let us examine the type of the representative in minority region When thedenominator of (2.15) is positive, the type of the representative in minority region iselected outside of the median citizen in minority region since the coefficient of NX B

is positive and the coefficient of NX Ais negative Furthermore, when the denominator

of (2.15) is positive, (2.15) can be rewritten as follows:

N

Since the coefficient of NX Bis negative and the coefficient of NX Ais positive, the type

of the representative in minority region is elected remote from the median citizen inmajority region and outside of the median citizen in majority region

Then as for the coefficient of NX Bin (2.15), the following inequality is held:

Lemma 2.6 In equilibrium, the type of the representative in minority region moves

to the same direction with the median citizen in minority region and to the opposite direction with median citizen in majority region.

The decrease in the distance of median type between two regions has two effects

on the utility in minority region One is the effect of the increase of the utility inminority region through the policy externality, and the other is the indirect effect ofthe decrease of the utility in minority region through policy making The latter effect

is intensified by the scale effect in majority region Therefore, as in Lemma2.4,

when N A  N B C s=.1 C s/, the type of the representative in majority region moves

to the opposite direction with the median citizen in majority region, and when N A

N B C s=.1 C s/, the type of the representative in majority region moves to the same

direction with the median citizen in majority region

Fig 2.3 The elected type of representative

For the above discussion, in equilibrium, the elected type of the representative intwo regions is allocated in Fig.2.3

Therefore, we obtain Proposition2.1

Proposition 2.1

1 if 1Cs s  N2

A , the type of the representative in two regions is more extreme type (Fig.2.3(1)) This may occur when the population in minority region is not so large.

2 if 1Cs1  N2

A and N A  N B/  s

1Cs , the type of representative in majority region

is between each median citizen in two regions, but the type of representative in minority region is outside of the median citizen in majority region (Fig.2.3(2)) This may occur when the population in majority region is sufficiently large and the policy conflict cost is not high.

3 if 1Cs1  N2

1Cs , the type of representative in minority (majority) region is outside of the median citizen in majority (minority) region (Fig.2.3(3)).

As the bargaining of integration by the elected type of representative in tworegions, from (2.7), (2.14), and (2.15), the equilibrium policy X Iis shown by

.N A  N B/

N AC 1

N A

s 1Cs

N



N

Proposition 2.2 In equilibrium, the policy externality does not influence the policy

level in integration.

Now let us compare this equilibrium policy in integration with the first-best

policy Since the first-best policy X FB is N A XNA C N B XNB, the distance between thesetwo policies is

Note that N A 1=2 Then the following proposition is held See Fig.2.4

Proposition 2.3 The equilibrium policy in integration is closer to the median

type of citizen in majority region than the first-best policy Furthermore, as the population of majority increases, these two policies move to the median type of citizen in majority region.

is, the problem of incentive of integration This incentive for integration is shown

in (2.10) From (2.10), the distance between the types of two representatives has tosatisfy a certain level in order to have an incentive for integration By substituting(2.16) into (2.10), we can obtain the condition where both types of representatives

in two regions want to integrate This is shown as the following proposition

Proposition 2.4 The condition where integration is realized through representative

elections in two regions is

Then we can examine the effect of policy conflict cost on the possibility of

integration Let us differentiate the right-hand side of the above equation in s:

3

From this equation, we have

 0, the possibility of integration decreases as the policy

conflict cost increases.

 0, the possibility of integration increases as the policy

conflict cost increases.

Now let us investigate the effect of the type distance between two representativesand the change of the population on the possibility of integration Then (2.11) isshown as follows:

2



N A1Cs s 

.N A N B/2  NX A NX B/2 (2.25)Let us denote the left-hand side of (2.25) as W.N A ; s/ Then W.N A ; s/ is an increasing function in N A, when the policy conflict cost is small Then, if thepopulation on the majority region is smaller than a certain level, the secessionremains, and if the population on the majority region is larger than the certain level,the integration occurs

On the other hand, when the policy conflict cost is large, W.N A ; s/ is an U-shape curve in N A Then if the distance between two median types is small, the integrationoccurs when the population of majority region is large, the secession remains whenthe population of majority region is large in a degree, and the integration againoccurs when the population of majority region is sufficiently large Furthermore,

if the distance between two median types is large, the secession remains when thepopulation of majority region is smaller than a certain level, and the integrationoccurs when the population of majority region is larger than the certain level.Generally, the policy coordination effect of integration increases as the popula-tion of majority region decreases and the policy conflict cost increases Therefore,since the effect of integration on policy coordination becomes small if the policyconflict cost is small, the merit of integration increases as the population on majorityregion increases On the other hand when policy conflict cost is large, the policycoordination effect of integration becomes large as the population of majority regiondecreases Therefore, when policy conflict cost is large, the integration occurs notonly when the population of majority region is large but also when the population

of majority region is sufficiently small

From the above discussion, we have the following proposition

Proposition 2.6 In equilibrium election for each representative in two regions:

1 when policy conflict cost is small, the integration occurs if the population of

majority region is sufficiently large.

2 when policy conflict cost is large, the integration occurs not only when the

population of majority region is large but also when the population of majority region is sufficiently small.

2.4 Centralized Governance and Secession

So far we have examined a simple public policy model as an integrated politicalform But in reality, there are several governance forms of integration From thissection, we introduce a centralized governance form and a decentralized form andinvestigate the influence of these governance forms on the possibility of secession

In order to clarify this discussion, we assume that a constant level g of a public good

is needed to manage an independent country and the level g is same, irrespective of

the size of population, as in Alesina and Spolaore [2], Alesina et al [3], and Alesinaand Spolaore [4] Of course the public good is assumed to be of non-rivalry In each

region i, typical citizen has an income y i

Let us consider the situation of secession between two regions In region i, per

capita income tax i is levied to produce the level g of a public good, namely, a governmental service Then g D N ii y i i D A; B/ is held y i.1  i/ goes to private

consumption Let us denote the utilities U A and U Bof the representative citizen for

each region as follows, when each region implements a policy X A ; X B:

U A D a N X A  X A/2C y A.1  A / C g (2.26)

U B D b N X B  X B/2 s.X B  X A/2C y B.1  B / C g (2.27)Here the utility at the bliss point is normalized to be zero We assume thatwhen there is a policy conflict between the majority region and minority region,the minority region incurs a policy external cost s, while the majority region doesnot Furthermore, we assume that the policy difference between two regions doesnot influence the utility of the majority region, but not that of the minority region,which reflects the difference of the population

Now let us seek the equilibrium policy that each region makes under secession The

optimal policy for region B, given the policy X A for region A, is shown by

X B OD b N X B C sX A

b C s

On the other hand, the optimal policy for region A is obviously the bliss point N X A,

whatever the policy for region B may be Therefore, the optimal public policy in region B is shown as follows:

2.4.3 Policy Decision Under Centralized Integration

Let us consider the case that two regions integrate Here there are two forms

of integration One is a centralized integration (CI) The other is a decentralizedintegration (DI) Under the centralized integration, the same level of public policy

is implemented Under the decentralized integration, there is a room to implementdifferent levels of the public policy for two regions In this subsection, we considerthe case of centralized integration In this case, the first effect of regional integration

on the majority region is in general a policy effect By integration, the majorityregion can dominate the policy in the minority region But in this two-region model,the policy effect for the majority region is neglected The second effect is a scaleeffect By integration, the population increases and the tax burden decreases So

when region A and region B integrate and a policy X Ais implemented, the utilities

of the representative citizen for regions A and B are

U A CI D a N X A  X A/2C y A.1  A / C g

U B CI D b N X B  X A/2C y B.1  B / C g

where A y A N AC B y B N B D g is held We have the three effects of integration for

the minority region One is the deletion of the policy conflicts, which is positive forthe minority region Second is the delegation of policy decision, which is negative.And the third is of course a scale effect

The region A proposes a pair of public policy X A and tax policy A; B/ to the

region B Region B can refuse the proposal if it makes region B worse than in the secession Therefore, region A has to propose the policies so as to satisfy the

following integration-acceptance condition:

Proposition 2.7 Under centralized integration, the optimal public policy becomes

efficient.

By substituting the value of the optimal public policy into (2.32), we have the

maximum utility of representative citizen in region A under integration as follows:

in region A under integration with the utility in region A under secession If the

following condition is held, the integration will occur:

Fig 2.5 Integration condition 1

This condition is shown as follows:

in region B.

As the relative population in the minority region increases, the optimal publicpolicy becomes close to the ideal point for the minority region This implies theincrease of policy adjustment cost for the majority region Figure2.5shows thissituation

Proposition 2.8 While centralized integration will occur in the case of small

relative population in the minority region, secession will occur when the relative population becomes larger than the certain level.

On the other hand, the per capita tax in the majority region is smaller than inthe case of secession due to scale merits of integration Since the effect of the scalemerits on the per capita tax in the majority region is larger than the effect of theincrease of the threat point in the bargaining with the minority region on the percapita tax in the majority region when the relative population in the minority region

U I∗

A

U A O∗

N B /N A Integration Secession

Fig 2.6 Integration condition 2

is small, the per capita tax decreases as the relative population in the minority regionbecomes large Therefore, Proposition2.8is held, which is shown in Fig.2.6.Using (2.36), let us compare the utility on region A under integration with

that under secession From the comparison, we show the integration condition

in Fig.2.6 And since the decreasing effect of per capita tax in the majorityregion increases as the level of public goods increases, we obtain the followingpropositions:

Proposition 2.9 As the level of public goods becomes high, the possibility of

integration becomes high In the case of low level of public goods, when the relative population of the majority region is low, the possibility of integration is high, and when the relative population of the majority region becomes high, the possibility of secession becomes high.

Proposition 2.10 Let us consider the case of middle size of public goods and high

policy conflict cost Then the possibility of integration is large, when the relative population of the majority region is small However, as the relative population of the majority region increases, the possibility of secession increases, and as the relative population of the majority region increases more, the possibility of integration will occur again.

We also obtain the following proposition regarding the ideal point gap betweentwo regions:

Proposition 2.11 As the ideal point gap of public policy between two regions

becomes large, the possibility of secession becomes large.

( ¯X A − ¯ X B)

Integration

Secession

Fig 2.7 Integration and secession

In this chapter, the level of public goods is assumed to be given This assumptionimplies that the level of public goods (civil engineering, armaments, etc.) in acountry may depend on the natural environment or geopolitical situation Forexample, it may be important that the country is vulnerable or not to natural calamity

or invasion from a third strong country

The integration-acceptance condition is rearranged as follows:

From this inequality, the possibility of secession increases as the ideal point gap

of public policy between two regions becomes large and the cost of public goodsbecomes large, as shown in Fig.2.7

2.4.4 Tax Policy Under Centralized Integration

Let us examine the tax policy for each region under integration The total amount of

tax in A is shown as follows: