Secessionist Demands, Parties, and Outcomes
As
noted in the previous chapter, this dissertation focuses on political parties
as vehicles for secessionist demands in a population. Secessionist demands are assumed to derive from assessments of
the costs and benefits of secession, but demands do not translate automatically
into policy outcomes. In democratic
countries, political demands are expressed through political parties,
intraparty factions, pressure groups, and extralegal direct action, and these
methods of political activity are in turn directed at policy outcomes. Since secessionist demands threaten the
existing political establishment, it is likely that they will be expressed most
often through the first and last methods.
Traditional political parties will tend not to be receptive to these
demands, making intraparty secessionist factions nonviable and pressure group
activity unfruitful without an independent electoral vehicle. For the reasons discussed in Chapter 1, this
dissertation ignores extralegal direct action and focuses on political
parties. Therefore, it is worthwhile
looking more closely at how secessionist demands arise, how these demands
translate into secessionist political parties, and how these political parties
affect policy outcomes.
A
region has a good chance of developing a secessionist movement whenever the
“base conditions” for secessionist demands are present. Not all regions possess the base conditions,
and they do not develop secessionist parties.
Neither a separate cultural identity nor economic grievances are usually
enough on their own to generate secessionist demands. Together, however, they often constitute a potent
combination. Even where secessionist
demands exist in a segment of a region’s population, they may not be expressed
electorally, through the vehicle of a political party supporting a platform of
secession or wide-ranging autonomy. The
political system may discourage the formation of new political parties.
Most obviously,
the electoral rule matters. Under
single-member district (SMD) plurality rule, at most two political parties can
be competitive in a given electoral district.[1] In most SMD countries, this rule would mean
that only a left-wing and a right-wing party, defined by their positions on
socioeconomic issues, could survive and garner votes. Parties campaigning along a second dimension, such as
centralization versus secession, would not win votes because voters would not
consider them to have a chance of winning.
Of course, it does not work exactly this way in practice; even in SMD
countries there are usually small third and fourth parties which do not win
seats and yet achieve positive vote percentages, usually from voters either who
do not regard the socioeconomic axis as not very important relative to another
dimension of policy or who regard the two “major” parties as being virtually
identical and thus not worth choosing between.
To predict which
regions will develop stronger secessionist parties, then, it is necessary to
include variables that influence secessionist demands, as well as variables
measuring institutions that have a persistent effect on the translation of
these demands into political parties.
Such a model would be a “cross-sectional” model, that is, it explains
variations across sections of data, abstracting from temporal variation within
the sections. This chapter develops a
set of hypotheses about the cross-sectional determinants of secessionism.
Once we have
developed a cross-sectional model of secessionism, we might well want to know
about the factors influencing year-to-year secessionist electoral success. The base conditions for secessionist demands
mostly do not change much over time.
Certainly, aspects of cultural identity take a very long time to
change. On the other hand, factors
affecting the economic benefits of independence might change, and the success
or failure of policies aimed at reducing the secessionist sentiment should
cause upticks and downturns in secessionist vote. Finally, there may be external conditions affecting secessionist
vote indirectly; for example, a variable that tracks the salience of the
secession issue should be an important predictor of secessionist vote. Even when fundamental support for the idea
of independence remains constant, a decline in the salience of the issue should
result in a decline in secessionist vote.
We might think of these “salience factors” as transitory institutions:
they do not affect secessionist demands so much as how these demands translate
into electoral support. This chapter
also develops hypotheses about the over-time determinants of secessionist vote.
Institutions
also mediate how parties affect policy outcomes. There is a large literature on how parliamentary rules, types of
coalitions, separation of powers, and other mechanisms influence the
policy-making process. The major
relevant institutional context for secessionist parties is how autonomous the
region of operation is from the central government. A regional legislature is probably necessary for a secessionist
party to achieve power and hold a referendum on independence. In most countries, the central government
would never do this for them.[2]
However,
actual secession is a rarity in democratic countries. Instead, what more frequently occurs is that the central
government offers concessions to the voters of the region, thereby dampening
support for secession. Concessions may
take the form of fiscal incentives (subsidies) or greater legislative autonomy
or representation.[3] The trigger for these concessions is usually
an unexpected electoral success for a secessionist political party. Furthermore, this causal relationship is
apparent to the voters, and the voters may have an incentive to vote for a
secessionist political party even if they don’t favor independence, in order to
prod the central government into offering concessions.[4] Sophisticated secessionist parties often
take advantage of this logic implicitly with slogans noting how their existence
is good for the region.[5]
Thus,
the incentives to vote for a secessionist party and the likely effects of
secessionist electoral victories are partly endogenous to each other as a
result of a strategic game between a region’s voters and the central
government. This chapter will present
Bayesian and spatial models of the game between a potentially secessionist
region and a central government, demonstrating that regions with strong
secessionist parties should receive more autonomy than regions without such
parties.
This
chapter thus contains three theoretical sections: explanations of
cross-sectional variation in secessionist vote, explanations of over-time
variation in secessionist vote, and explanation of the consequences of
secessionist vote.
Hypotheses about Inter-Regional Variations in
Secessionist Electoral Support
Perhaps the most
obvious model for predicting support for independence is to assume that it
varies according to the degree of basic cultural difference between the
peripheral region and the rest of the country.
Cultural differences can be linguistic, ethnic, religious, or
ideological. Of these “dimensions of
difference,” ethnicity is the most difficult to measure, as ethnic identity is
malleable and often defined in terms of the other kinds of difference or even
in terms of state citizenship (which would make the concept circular for any
attempt at using it to predict attempts at changing state boundaries). Ethnic identities are notoriously malleable
to political manipulation; therefore, it is problematic to assert that
secessionism is straightforwardly an effect of ethnic difference – it is
as likely to be the other way around.
Linguistic (and
in some countries, religious) differences thus seem to be, if anything is, the “primordial”
influences on a people’s conception of itself as different.[6] It is clear that secession has become an
issue in Quebec, for example, chiefly because Francophones in Quebec see
themselves as a “distinct society” separate from English Canada.[7] Notwithstanding, focusing on only linguistic
and religious differences overlooks certain, possibly empirically relevant
factors in what makes a group of individuals consider themselves a
“people.” A major factor seems to be a
historical memory of independent statehood.
The Scots, for example, largely share the language and religion of the
English,[8]
but they have a strong national identity based upon a long historical
experience of independence, while the Welsh, who are more different from the
English in terms of language than are the Scots (many more Welsh speak Welsh
than Scots speak Gaelic), do not seem to have quite as strong a sense of
national identity, possibly because their independence is shrouded in distant
memory.[9]
Hypothesis
1: Linguistic difference promotes secessionism.
Hypothesis
2: History of independence promotes secessionism.
Hypothesis
1 requires modification, however, for there are linguistically distinctive
regions that are nevertheless candidates not for secessionism, but irredentism. These regions are typically small enclaves
that due to accident of history or spoils of war have been cut off from their
“native homeland” and incorporated into a neighboring state. Such enclaves include Ahvenanmaa (Swedish: Åland)
in Finland, South Tyrol and Valle d’Aoste in Italy, both Flanders and Wallonia
in Belgium, and of course Northern Ireland in the U.K.[10]
In most countries secessionism and irredentism are quite distinct phenomena,
and a region can neatly be categorized as either potentially secessionist or
potentially irredentist. The peculiar
case of Belgium is perhaps the only counter-example, where support for union
with France is rather strong among Walloon activists, but pan-Netherlandic
sentiment is nearly nonexistent among their Flemish counterparts. Most potentially irredentist regions receive
substantial autonomy as part of a negotiated settlement between the countries
involved. Cross-border cultural
exchange with the “homeland” reduces feelings of insecurity among the
population of the linguistic enclave.
Minority language groups without such ties, such as the Welsh and
Basques, realize that if the language fails in their region, it dies out
completely.[11] For all these reasons, secessionism should be low
in potentially irredentist regions.
Hypothesis
1b: Secessionism will be lower in “unredeemed regions” adjacent to their
linguistic homelands.
Ideological
differences between voters in a region and voters in the country as a whole are
a more variable component of identity and can indeed serve as an impetus for
secessionist movements. The SNP draws
some of its appeal from this kind of difference: since the 1960s Scottish
voters have moved increasingly to the left of English voters. Movements drawing mostly on this kind of
difference are aiming explicitly at securing economic policy autonomy for the
region. There is reason to suspect,
however, that globalization makes autonomy of this sort less viable, to the
extent that it reduces the variance in economic policies across countries. Political-economic theory suggests that we
should expect such a reduction. As capital mobility increases,
capital-holders seek out jurisdictions that are likely to produce higher
returns to capital. A large element of
a return to capital is the level of tax on earnings from capital. Therefore, we would expect that to prevent
economically damaging capital flight and erosion of the tax base governments
worldwide would reduce taxes on capital.
We therefore would expect that regions that wish to have higher taxes on
capital than existing independent states have will find their aspirations
frustrated by higher levels of capital mobility. Thus, capital mobility dampens the positive effect of ideological
difference on secessionism. Though the
evidence of convergence remains limited,[12]
it seems clear that voters in peripheral regions that seek a more favorable
economic policy regime should begin to favor other methods than autonomy to
achieve their goals, including the transfer of economic policy from the
existing central government to a supra-state entity. Nevertheless, if ideological difference has any effect on a
region’s tendency toward secessionism, it should be positive.
Hypothesis 3: Ideological
differences between a region and the rest of the country cause greater
secessionism.
While
cultural differences between periphery and center can serve as a baseline model
of potential secessionist sentiment, surveys consistently show that voters in
peripheral regions are far more interested in economic issues than in
protection of heritage and culture. Secessionist
parties succeed not because they appeal to a primordial past, but because they
are able to present independence or wide-ranging autonomy as beneficial in
political and economic terms. Cultural
identity provides a sense of separateness, but voters consider this
separateness relevant only when it can be mobilized to achieve political and
economic goals that are important to them.
The appeal of secession stems from the desire for government that is
closer to home, more attuned to popular demands, more flexible, and in many
cases smaller and less powerful.
Accordingly, the demand for independence should correlate with variables
that affect popular perceptions of how sustainable and desirable an independent
government would be.
Territorial grievances are frequently based on
perceptions that the region is a net “loser” from the existing political
union. In particular, when a region
pays more in taxes than it receives in expenditures it is likely to present
demands for fiscal autonomy and possibly even secession if those demands are
not met. Regions that receive more in
expenditures than they pay in taxes are poor ground for secessionism, because
independence would mean the loss of subsidies.
Revenue
drains from the center are frequently issues even for movements that are
left-of-center. The Scottish National
Party famously exploited the issue of North Sea oil discoveries in the 1974
elections with the slogan, “Rich Scots or Poor Britons.” Likewise, there is great controversy in
Canada over whether Quebec is a net recipient or net contributor of federal
government funds. The left-wing
secessionist Parti Québecois naturally claims that Quebec is not subsidized,
while the anti-secessionist Liberal Party asserts that it is. In Spain, the two regions most receptive to
secessionist appeals, Euskadi (the Basque country) and Catalonia, are also the
wealthiest regions and thus presumably contribute more to the Spanish treasury
than they receive in return. In modern
democracies the state redistributes from the high-income to the low-income;
thus, in modern democracies higher-income regions are likely to be more
inclined to support secession or robust fiscal federalism than lower-income
regions.[13]
Hypothesis
4: Relative regional affluence promotes secessionism.
In some
cases cultural difference might not matter much at all for secessionist
sentiment. For example, Northern
Italian (“Padanian”) secessionists are motivated mostly by the fact that Northern
Italy pays much more in taxes to the Italian government than it receives back
in services, and by corruption in Rome.
There are few linguistic or ethnic differences between Padania and the
rest of Italy, and the ideological differences that exist are mostly explicable
in terms of the revenue-expenditure disparity and stereotypes about regional
corruption.[14] Both rich North and poor South vote for
Forza Italia and allies, each for its own reasons, while the Center votes for
the left.
Another
way to interpret the “wealth effect” is that citizens of a wealthier region are
more likely to view their region as potentially a viable independent state than
are citizens of a poorer region. The
problem with this interpretation is that absolute wealth should be a better
indicator of viability than relative income.
Even a relatively poor region in the West (Andalusia, for example) is
higher-income and accordingly more “viable” than many independent countries in
Africa and Asia. Concerns about
viability do have a high profile in debates over independence, however, with
opponents of secession typically stressing that the region concerned is not
viable. Larger territories are thought
to be more viable as independent states than small ones, since most public
goods enjoy declining average costs over population because of a large
fixed-cost component in their supply.
The success of small countries such as Iceland seems to render suspect
this argument and the economic logic underlying it; however, there is no
denying that arguments to viability have a prominent place in anti-secessionist
rhetoric. Thus, variables perceived to
affect viability (whether they actually do or not) should make a difference to
some voters weighing the arguments.
Hypothesis
5: More populous regions will have stronger secessionist movements.
Besides
cultural and economic variables, political institutions also matter for
secessionism. Horowitz argues that
secessionism is more likely where territorial units within a country are few
and large.[15] In these cases there are only a few vectors
of conflict, and compromise becomes more difficult. In addition, when an ethnic group is unified in a single
territory, collective action against the state is easier. It is no accident that the most centralized
states in the world (e.g., Turkey, the Baltic states, Greece) tend to have many
very small provinces rather than a few large ones.
Hypothesis 6:
Regions with greater representation in the country as a whole are more likely
to support secessionist movements.
Finally,
we have already examined the importance of electoral institutions. For many secessionist parties, a two-party
system will discourage voters from “wasting their votes” on a secessionist
candidate. To be sure, secessionist movements are by
definition geographically concentrated and therefore do not necessarily face
the same problems as small liberal parties or ideologically extreme parties of
the left and right that presumably have relatively even country-wide support. Secessionist parties with a proven record of
winning seats, like the SNP and the Parti Québecois (PQ), probably do not
suffer very much from strategic voting, but in general we should
expect that multi-party systems will be more congenial to secessionist parties.
Hypothesis 7:
Multi-party systems are more favorable to secessionist parties than systems
with few parties.
Finally, regions
with their own legislatures should see higher secessionist vote in regional
elections only. Secessionist parties
should do better in regional elections because they have a greater chance of
winning an election and forming a government at the regional level than at the
country level. In most cases, it is
mathematically impossible for secessionist parties to win a countrywide
majority because they run candidates only in the region for which they seek
independence or autonomy. Voters might
then see a vote for a secessionist party in a countrywide election as a wasted
vote. Furthermore, secessionist parties
have an incentive to cultivate knowledge and expertise in regional policy, as
they base their appeal on defense of regional interests. Voters thus may see secessionist parties as
particularly well qualified to run regional government.
Hypothesis 8:
Secessionist parties do better in regional elections.
Hypotheses about Inter-Temporal Variations in
Secessionist Electoral Support
Many of
the hypotheses just identified cannot explain short-run variations in
secessionist vote. In particular, they cannot
explain why, for example, the SNP and Plaid Cymru obtained record vote totals
in the early 1970s, declined from the late 1970s into the early 1990s, and are
now enjoying success paralleling or even surpassing that of the early 1970s, or
why the Lega Nord rose from humble beginnings in the early 1980s to become one
of the most important political forces in northern Italy in the mid-1990s.
The growth of secessionism in the 1990s attracted the
attention of economists. Their argument
is that globalization has increased the credibility of regions’ threats of
exiting from existing countries. This
contention appears at first counterintuitive.
The fact that political secessionism has increased around the world at
the same time as the advance of global economic integration does suggest that
the two trends may be mutually reinforcing rather than antagonistic. Much of the recent economics literature on
the equilibrium size and number of countries has added to the intuition that
free trade makes political separation less costly in terms of efficiency.
The logic is that in a world of free trade, borders mean
less. When a small region becomes an
independent country, it does not need to become more self-sufficient: the
adjustment from the comparative-advantage allocation of resources for a
separating country is less in a world of free trade than in a world of autarky.[16] Secessionist movements may also think more
strategically; under global free trade a putative country can forge economic,
cultural, and political ties with countries other than the metropole. Thus, Quebec sovereigntists may favor
increasing economic ties with the United States as a counterweight to the
province’s dependence on Canada.
Jacques Parizeau, sometime leader of the sovereigntist Parti Québecois,
argued that increased openness in the international economy would allow Quebec
to be politically independent and economically viable.[17]
Note, however, that the globalization argument does not
necessarily hold that regions more integrated into the world economy are more
likely to secede. Rather, the level of world
economic integration makes secession more viable for regions all over the
world, even if they are currently closed.
Indeed, closure of regional markets to global producers can be a positive
inducement to secession: the region may desire free trade but have its
aspirations frustrated in the existing country. The “Tariff of Abominations” was the stimulus of the South
Carolina Nullification Controversy in the early 1830s, and tariffs were a
secondary but important issue in the eventual secession of the Confederate
states.[18] Moreover, regional integration into the
global economy can buttress the arguments of parties that support regional
autonomy but explicitly oppose independence, as Pierre Martin argues has been
the case for the Liberal Party of Quebec.[19] These parties can argue that high existing
levels of globalization mean that economic autonomy from the existing state and
integration into the global economy do not require formal political
independence. Thus, a regression of
secessionist party vote totals on regional integration into the global economy
would not be an appropriate test of the globalization argument.[20]
In addition to economic integration, it may be that
certain kinds of political integration actually encourage certain kinds of
political separation, insofar as that integration serves to reduce
transnational transactions costs. Thus,
customs unions like the European Union may actually encourage
secessionism. Thus, Alex Salmond,
leader of the Scottish National Party, is able to say:
People say,
“What’s a wee country like this going to do for an army? Who’s going to
do food and
drug testing? Who will issue the
patents?” And people always worry
about the
money, you know: “We’ll have a Scottish currency that nobody wants and
a central bank
that nobody listens to”. . . The whole debate on independence has
been changed
by a single idea. And that’s the
European Union.[21]
The European Union might be seen as a challenge to the
economists’ argument. If free trade is
associated with political disintegration, why do we see apparent political
integration in the European Union proceeding at the same time as trade barriers
are falling? Alesina, Spolaore, and
Wacziarg (2000, 1293) suggest classifying the European Union as a case of “deep
economic integration” rather than as an area of political integration. Put another way, globalization increases the
number of public goods that can be effectively provided through international
cooperation rather than outright political integration. An alternative argument would be to note
that while globalization reduces the efficiency costs of separation (or
non-integration), it might also provide political incentives for
integration. Particularly, capital
mobility theoretically should erode governments’ autonomy in certain areas of
fiscal, monetary, and regulatory policy.
If political leaders are concerned with maintaining policy autonomy
vis-à-vis private economic actors as an end in itself, they may find that their
only recourse is political integration.[22] Doubtlessly, it is impossible to categorize
the EU strictly as either a functional-economic institution or a
proto-superstate; it has elements of both.
The point is that globalization has reduced both the economic costs of
secession from existing states and the political costs of the non-integration
of existing states into a European superstate.
Hypothesis 9: Secessionist vote should be
increasing over time in close correlation with the progress of world economic
integration.
It is important to note that the economic argument about
the effects of globalization on secessionism assumes that secessionist
movements care more about the trappings of formal sovereignty than they do
about effective political autonomy.
Free trade, capital mobility, and especially customs-union arrangements
constrain political autonomy in important ways. Under free trade, governments are less able to engage in
comprehensive industrial policy, central planning, and other kinds of
micro-management of the economy, as well as unsustainable macroeconomic
policies. The reasons are multiple:
distorting relative prices in the domestic economy can lead to poor export
performance; comparative advantage tends to move quickly in world markets,
rendering central planning ineffective at best; large budget deficits can
stimulate imports and discourage exports, resulting in abnormally large balance
of payments deficits. Factor mobility
tends to limit governments’ options further, theoretically forcing governments
to lower taxes on more mobile factors and to eliminate non-productive
spending. Customs unions can explicitly
regulate monetary and fiscal policies, as well as microeconomic policies that
affect factor costs.
As argued in the previous section with respect to the
effects of ideological difference on secessionism, if the primary goal of
secessionist movements were the establishment of greater economic policy
autonomy, we would expect the kinds of economic and political integration just
described to make secessionism a less viable strategy over time.[23] But secessionist movements frequently care
about something else: cultural autonomy.
For these secessionist movements, linguistic and other cultural policies
take on primary significance, even though when measured as a percentage of the
budget or for their effects on economic aggregates these policies seem
insignificant. Indeed, it is
indubitable that for secessionist movements of this sort there is a pure
psychological benefit from the prospect of having one’s own ethnic or
linguistic group recognized as a sovereign nation-state, apart from what that
status may entail in practice. To put
it another way, the benefits of secession in terms of maintaining cultural
difference cannot explain the recent increase in secessionist sentiment because
these benefits are constant, but the decline in the costs of secession in terms
of economic efficiency might be able to explain part of the secessionist
revival.
The globalization argument, however, ignores the
interactive effects of ideology. There
is a new kind of secessionist movement on the rise: the sort that seeks to
secede precisely in order to reduce the government’s economic policy autonomy. Significant issues for movements of this sort
include net revenue drains from the home region to the other parts of the
country and what they see as intrusive policies from the center. The Lega Nord in Italy and the Alaska
Independence Party in the U.S. derive their support from issues of this kind. In general, if global economic integration
entails smaller government (still a large “if”), then regions that favor
smaller government will be more likely to support secessionist movements when
economic integration is high.[24] This tendency is reinforced to the extent
that secession itself has an independent effect on the constraints on
government size, by means of increasing factors’ mobility across borders and
stimulating fiscal competition.
(Holding all patterns of goods, services, and productive-factors
movements constant, an increase in the number of countries in the world
increases measured cross-border mobility and presumably incentives for
governments to engage in fiscal competition.)
Nevertheless, the mediating effects of ideology on globalization are
still at this point speculative, as it has not yet become clear to what extent
globalization does force governments to become smaller in the first place.
Though the globalization argument can perhaps explain the
trend in secessionist vote over time, it does less well explaining the
year-by-year ups and downs in secessionist support. However, there are some economic indicators that do vary in the
short run and may help to explain voters’ calculations at election time. For example, if a region is suffering from
high unemployment, the region’s voters might become more dissatisfied with the
current political arrangement and more likely to support a radical solution
like separation. On the other hand,
high unemployment might cause a region’s voters to be fearful of the region’s
economic prospects if it were to become an independent country. This
Catch-22 situation has been dubbed the “fear-confidence antithetical effects”.[25] In order to sell secession, a secessionist
party must inspire confidence in the voters that the region would do well after
independence, and fear that the region would do poorly without it. These two goals are sometimes in
conflict. However, in the case of
unemployment there seems to be an additional factor weighing against secessionist
parties. When times are bad, voters are
more likely to seek immediate policy solutions for their problems, and major
constitutional change is likely to become a non-issue during such times. Thus, unemployment can be viewed as an
“institutional” barrier to the mobilization of secessionist preferences into
votes. On the other hand, if
unemployment is not very high in an absolute sense but remains noticeably
higher than unemployment in the rest of the country, the situation may be
interpreted as central government bias against the peripheral region, and
independence may be seen (whether rationally or not) as a means to long-run
economic success. If the economy has
gone into steep decline, however, the relative affluence effect described in
the previous section may take hold: a region’s voters may reject secessionism
when relative regional affluence is falling because they foresee the prospect
of more subsidies from the central government.
Note that voters are likely to undertake this sort of calculation only
when a strong secessionist movement already exists in their region. Otherwise, frustration at high unemployment
is likely to be expressed through other avenues.
Hypothesis 10: Poor economic conditions in absolute
terms decrease the perceived relevance of constitutional changes and therefore
reduce secessionist vote.
Hypothesis
11: Higher unemployment in the region compared to the country as a whole,
contrariwise, feeds accusations of central government unfairness and therefore
increases secessionist vote.
Hypothesis
12: However, if relative regional affluence falls, indicating a move toward a
net-subsidized status for the region, secessionist vote should decline, as the
costs of independence would rise.
Policy
changes also have a major impact on secessionist support. One such policy change is an offer of
autonomy. If a central government
offers a regional legislature to a region that did not previously have one, or
concedes further powers to an already existing regional legislature, many
voters in the region will feel satisfied, temporarily at least, that the new
arrangements are sufficient. These
voters, best thought of as “conditional secessionists,” will in the near term
no longer vote for the secessionist party.
By contrast, if a promised offer of autonomy falls through, dashed hopes
may lead to greater resentment than would have existed if the offer had never
been considered, and the secessionist alternative may benefit.
Hypothesis
13: Recent increases in regional autonomy reduce regional support for
secessionist parties.
Hypothesis
14: Unexpected setbacks for the progress of regional autonomy originating in
external sources cause temporary surges in secessionist support.
The clause about
“external sources” is a way of taking into account failures for the autonomist
cause that are due not to central government failures but to rejection from a
region’s own voters. If a region’s
voters vote down a proposal for more autonomy, they have no cause for
resentment against the central government.
This result only highlights divisions within the regional community and
eliminates the “us versus them” dialectic that often results in heightened
pressure for autonomy or independence.
Thus, if a central government fully expects to win a referendum on
independence, there is no reason, from its perspective, why it should not hold
one.
These
considerations suggest that a central government (with a relatively short time
horizon) has an incentive to offer autonomy to troublesome regions in order to
prevent secessionist electoral success.
Recent significant decentralization in Belgium, Spain, and the U.K. –
along with more tentative efforts in France and Italy – seem to support the
hypothesis that globalization induces a general increase in regional autonomy
arrangements. Whether this phenomenon
is related to secessionism is an open question in the literature: perhaps
decentralization is a purely functionalist response to globalization, a means
of ensuring institutional flexibility and promoting economic growth.[26] If this alternative hypothesis were true,
then increases in autonomy should be proceeding at a roughly even rate in
regions that have secessionist parties and regions that do not.
A General Model of Secessionist Threats and
Government Concessions
This
section formalizes the preceding intuitions and describes more rigorously when
governments will offer concessions, and when regions will decide to
secede. As already noted, if governments anticipate secessionism and
head it off with fiscal transfers or offers of autonomy, then observed
secessionist sentiment will be lower than it would have been had the government
not made such offers; if voters anticipate that the government will make such
concessions, they may vote strategically for a secessionist party even if they
do not support independence above all other options. However, voters will vote for a secessionist party only if they
view independence (or the secessionist party’s platform) as a more desirable
outcome than the status quo. Thus, a
large vote for a secessionist party constitutes a credible threat of secession
from the central government’s point of view.
A model can be specified as follows. In the first stage an election takes place
that indicates to the central government the probability that a region is type
1 or type 2. A type 1 region is one
that will secede if it does not receive concessions. Type 1 regions are represented by the upper branch of the game;
they receive a payoff of a for seceding, where a>0. A type 2 region is one that will not
secede even if it does not receive concessions. Type 2 regions are represented by the lower branch; they receive
a payoff of b for seceding, where b<0.[27]
In either case, secession results
in a payoff d for the government, where d<0.[28]
Figure 2.1
-g(c),h(c)
G M
f(v) d,a
~M
S
R
N ~S 0,0
-g(c),h(c)
1-f(v) M
S’ d,b
G ~M
R
~S’ 0,0
The election is modeled with a move by
Nature (N). The probability that a
region is type 1 is presumably a positive function of the secessionist vote
total in that region: f(v), where f(v) is monotonically
increasing in v. The central
government (G) may then choose to offer concessions c. The payoffs to regions for not seceding are
a function of such concessions: h(c), where h(c) is monotonically
increasing in c and where h(0)=0. The payoffs to a government for concessions made are –g(c),
where g(c) is monotonically increasing in c and g(0)=0. Theoretically, c is continuous: the
government may offer any level of concessions, whether fiscal or
constitutional. However, in practice the
government has just two choices: it can set c so that a type 1 region
would be indifferent between seceding and staying in the union and receiving
concessions (a=h(c)), or it can set c=0. It would be futile to set c below the
point where a=h(c) and unnecessary to set c above that
point. Thus, the game models G’s choice
as one between M (making sufficient concessions and no more) and ~M (making no
concessions). If the government makes
concessions, the game ends and the players receive payoffs. If the government chooses not to make
concessions, the region then has a choice between seceding, S and not seceding,
~S.[29]
The expected utility for G from playing M
is –g(c), and the expected utility from playing ~M is f(v)d. G will thus play M when –g(c)>f(v)d
and ~M when that condition does not hold.
This result means that a central government will appease when it
estimates that the costs of making concessions are not as great as the costs of
the region’s secession multiplied by the probability that the region is type 1:
the sort that will secede if concessions are not made.
It is straightforward to determine how the
region plays if the government decides not to make concessions. A type 1 region always secedes (since a>0);
a type 2 region never secedes (since b>0). Formally, the perfect Bayesian equilibria are characterized as
follows: (M; S,~S: -g(c)/d>f(v)); (~M; S,~S: -g(c)/d<f(v)). In any given situation, however, there is
only one equilibrium. That equilibrium
depends on the central government’s assessment of the probability of the
region’s being type 1 or type 2.
That some regions will secede if they do not receive
concessions, while others will not secede under any circumstances, is hardly a
revolutionary result. However, the
model does show that sometimes non-secessionist regions will receive
concessions, and sometimes conditionally secessionist regions will not – and
will therefore secede (or commence louder threats of secession). The central government may make mistakes,
because it does not possess certain knowledge of the region’s “type.” On the other hand, elections provide the
government with clues about the region’s type and therefore whether to make
concessions; over time, it is difficult for a central government to make big
mistakes, as few regions in the developed world have actually seceded. Furthermore, the model illustrates the fact
that an exogenous increase in a, the benefits of secession, will
increase the concessions c required to maintain union. Therefore, if the benefits of secession are
increasing around the world due to globalization, some of the evidence for this
phenomenon should be found not just in outright calls for secession but also in
the level of concessions that central governments are offering to their
peripheral regions to forestall such demands.
From the government’s perspective, an increase (decrease) in the costs
of secession for itself or a decrease (increase) in the costs of concession
should result in more (less) appeasement and less (more) observed secessionism.[30]
A Spatial Model of Autonomy and Secessionism
Now that we have examined the logic of concessions
in general, let us look at the issue of autonomy more closely. Political scientists can subject autonomy to
a more informative kind of game theory, spatial analysis, since autonomy is a
continuous vector ranging from complete centralization to complete
independence. Figure 2 presents a
continuum ranging from “0” (complete centralization) to “6” (complete independence). Between these two poles, moving from left to
right, are administrative decentralization (“1”), devolution with an elected
regional chamber having authority over a central government grant and perhaps
cultural matters (“2”), full-fledged federalism with regional tax-setting
authority (“3”), confederation or “associated state” status with the regions or
associated states being considered the fundamental political units of the
polity (“4”), and customs union arrangements (“5”). The spatial distances between each number roughly represent the
substantive differences between them: formal independence under a customs union
is generally closer to full independence than to confederation (the difference
between France under the EU and France not under the EU is less than the
difference between France and the Isle of Man); administrative decentralization
is close to full centralization, since no law-making powers are devolved to the
regions.[31] These points are only a few of the possible
points that could be laid out on this continuum: there exists a variety of
institutional forms mixing the structures described here.
Figure
2.2
0------1----------2-----------3---------4----------5------6
Centralism – Admin. Decent. – Devolution –
Federalism – Confederation – Customs Union – Independence
Assume the
ideal point for the median legislator in the governing coalition in the
central-state legislature lies at 1.
There is a secessionist party in Region A, which takes position 5. The median voter in Region A lies at point
4. Preferences are estimated by a
negative loss function:
(1)
U(xi)=-a(|Pj-Ii|)2
where
U(xi) represents the utility of player xi, a
is a constant, Pj is the point where a current or proposed
policy j resides, and Ii is the ideal point of player
xi. This situation
approximates a typical situation in a European country within the EU dealing
with a surge of secessionist sentiment: the situation with regard to the United
Kingdom and Scotland, for example.
Assume further that a majority vote
in Region A for independence of any kind is sufficient for secession to occur.[32] The secessionist party may propose a
referendum of this kind at any time.
Assume that the status quo is at point 1, the ideal point of the median
central-state legislator. Suppose then that
the secessionist party proposes a referendum on its ideal point: point 5, which
I have interpreted as independence under a customs union, though the
substantive interpretation is not material to the model.[33] The median voter in Region A will vote for the
referendum because it is closer to her ideal point than the status quo:
(2)
|P5-P4|<|P1-P4|
Now
suppose, more realistically, that the central-state legislature is the first
mover. The central-state legislature
will set a new policy PN that approaches the following
equality:
(3)
|PN-P4|=|P5-P4|
The
left-hand term will be infinitesimally smaller than the right-hand term, just
enough to make the median voter in Region A prefer the new policy to the
secessionist referendum.[34]
If we assume that the central-state
legislature can control the wording of the referendum, the prospects are even
dimmer for the secessionists. Assume
the median voter is actually at point 5, favoring independence under a customs
union. Assume the status quo is at
point PR, where:
(4) PR-P5<0
(5) |PR-P5|<|P5-P6|
Equation (4) means that PR is to the left of
point 5, if we imagine that the points to the right are assigned higher numbers
along a dimension of “how much independence” exists.[35] Equation (5) means that PR
is closer to the median voter in Region A than is point 6. If the central-government legislature can
control the wording of the referendum, it can propose a referendum advocating
type 6 independence, which will be rejected.
Even though a majority of voters in Region A support independence of
some kind, no referendum on independence will be accepted if the
central-government legislature can control the wording of the referendum in
addition to adjusting the status quo as in the first example.
In fact,
central governments may not need to control the wording of the referendum to
cause a referendum on sovereignty to fail.
Since arrangements short of full independence (confederation, customs
union) would require successful negotiations between the central government and
the seceding region, the federal government can insist that if the region
declares sovereignty, it will refuse to negotiate. In this situation, the federal government would be forcing the
referendum choice to be one between full independence and the status quo,
whatever the text of the referendum might read. Prior to the ruling of the Canadian Supreme Court that the
federal government would be required to enter good-faith negotiations with
Quebec should that province decide to secede, the federal government argued
that it would not be required to enter negotiations, while Quebec secessionists
insisted that Canadian officials were bluffing. Secessionist leader Lucien Bouchard claimed that had a guarantee
of negotiations existed before the 1995 referendum, it would have passed.[36]
Secessionist
political parties face similar considerations in elections, apart from
referenda. The central government
cannot control the secessionist party’s manifesto in the same way that in some
circumstances it can control the wording of a referendum. However, the central government does control
policy and can make offers of increased autonomy just prior to elections in
order to mollify voters (and in some cases retract or hedge the offers
afterwards).
The
foregoing model has several implications.
First, secession in democratic countries occurs only if a majority of
citizens in the affected region support secession, but majority support is not
a sufficient condition for secession to occur if the central-state legislature
has agenda-setting power. Second, if we
think of a more nuanced real-world situation in which policy is sticky and the
central government does not know where the regional median voter lies (and the
region’s voters know this), voters can vote for a secessionist party even if
they do not support separation (formally: independence is not the voters’ ideal
point) in order to convince the government that it needs to move policy toward
greater autonomy. This outcome is a
corollary of the result that a central government can give more autonomy to a
region to dampen secessionist support.[37] Third, if the median voter in a region does
not support full independence, then in less proportional electoral systems
secessionist parties may have incentives to moderate in order to win seats. Thus, we see parties ranging from the merely
regionalist who explicitly eschew independence or even autonomy, to autonomist
parties with independentist factions, some of whom advocate intermediate
measures such as seats in the UN without full political independence (as Plaid
Cymru advocates for Wales), to full-fledged independentist parties. This dissertation focuses on the latter two
sorts of parties.
For
the purposes of this dissertation, the hypothesis of particular interest is
that electoral secessionism induces decentralization short of independence.
Hypothesis 15: Regions with
secessionist parties are more likely to receive more autonomy, and at earlier
dates, than regions without secessionist parties.
Conclusion
Cultural, economic,
and institutional variables explain the cross-regional variation in
secessionist vote. Cultural variables
include a regional language, history of independence, lack of irredentist
potential, and ideological differences.
Economic variables include relative regional affluence and
population. The cultural and economic
variables together explain how secessionist demands arise in a region. Demands lead to political parties, but
institutions mediate this causal relationship and in some cases can squelch the
electoral expression of secessionism altogether. Thus, institutional variables like representation of the region
in countrywide institutions, multi-party electoral system, and elections to a
regional legislature may be important.
Once a region possesses
a secessionist party, economic and policy factors explain the over-time
variation in its electoral fortunes.
Secessionist parties should do poorly when absolute economic conditions
are bad, when unemployment relative to the rest of the country is low, and when
relative regional affluence is falling.
The latter two factors are undoubtedly tightly correlated with each
other, despite their predicted opposing effects: usually when regional income
falls relative to the rest of the country, regional unemployment relative to
the rest of the country rises. It may
be difficult to parse the different effects empirically. Secessionist parties do better over time as
world economic integration advances.
This effect measures an actual decrease in the perceived costs of
independence, whereas the three other economic variables just mentioned do not
measure changing assessments of the costs and benefits of independence, but
rather changing assessments of the costs and benefits of voting for a
secessionist party rather than a statewide party. Policy changes can affect both calculations. Fiscal appeasement and autonomy offers may
reduce the perceived benefits of independence relative to the new status
quo. On the other hand, failures or
unexpected reversals of such policies may stimulate a short-run reaction
against the parties of central government and an increase in support for
secessionist parties.
While
secessionist parties have not yet achieved their ultimate goals in the
countries studied in this dissertation, they have provoked central governments
to respond. The main effect of
secessionist electoral success has been increasing political decentralization,
especially in regions where secessionist support is strongest.
[1] Gary W. Cox,
Making Votes Count: Strategic Coordination in the World’s Electoral Systems
(Cambridge, UK: Cambridge University Press, 1997).
[2] Before the
recent establishment of a Scottish Parliament, the SNP claimed that if they won
a majority of the Scottish seats in Westminster, they would have a mandate for
independence and would withdraw from Westminster to draw up a constitution for
an independent Scotland. There were
questions, however, about the legality of this course of action, and whether the
rest of Britain (or indeed, of Scotland) would accept this procedure as
valid. The existence of a Scottish
Parliament with legislative authority considerably simplifies the issue for the
SNP. See Jack Brand, The National
Movement in Scotland (London: Routledge, 1978), p. 187.
[3] Treisman finds
that fiscal appeasement worked to limit secessionism in Russia in the
mid-1990s, while fiscal appeasement was not practiced in Eastern European
states that broke up: the Soviet Union, Czechoslovakia, and Yugoslavia. Daniel Treisman, After the Deluge: Regional
Crises and Political Consolidation in Russia (Ann Arbor: University of
Michigan Press, 1999).
[4] However,
such voters must at least be “conditional secessionists,” in that they do not
fear the consequences of electing a secessionist majority. With the concessions that they expect, they
would oppose independence; without them, they view the prospect of independence
with equanimity.
[5] The example
of Scotland is again instructive here.
In 1974, over 76% of Scottish respondents to a BES survey thought SNP
victories were “good for Scotland,” including 70% of Labour and 68% of
Conservative voters (William L. Miller, Bo Sarlvik, Ivor Crewe, and Jim Alt,
“The Connection between SNP Voting and Demand for Scottish Self-Government,” European
Journal of Political Research, 5 (1977), 83-102; see also Keith Webb, The
Growth of Nationalism in Scotland (Glasgow: Molendinar Press, 1977), p.
97). Meanwhile, the majority of SNP
voters in 1974 supported devolution above independence, and independence above
the status quo (Roger Levy, Scottish Nationalism at the Crossroads
(Edinburgh: Scottish Academic Press, 1990), p. 25.
[6]
Nevertheless, even the definition of what counts as a “language” can be
manipulated politically, as regionalist parties in Spain have discovered. Valencian regionalists opposed to Catalan
secessionism have tried, for example, to define “Valencian” as a language
separate from Catalan. All the same,
there are fairly objective, scientific criteria for establishing what is a
language and what is a dialect, whereas there are no such criteria for
establishing what is an “ethnic group.”
[7] Quebec
sought to have its status as a “distinct society” enshrined in the Canadian
Constitution, but opposition from other provinces blocked the plan.
[8] The
Presbyterian churches in Scotland (Church of Scotland, Free Church, Free
Presbyterians) are stronger than the Episcopal Church in Scotland, while in
England the Episcopalian/Anglican Church of England is stronger than English
Presbyterian churches, but both Scots and English are overwhelmingly
Protestant.
[9] Scotland
united with Great Britain in 1707; Wales was conquered by England in 1282.
[10] Ethnic
identity in Northern Ireland tends to track religious affiliation more than
language loyalty, but the Irish language does tend to be associated with the
republican cause there.
[11] A “between”
case might be that of Québecois, who speak a language spoken in many other
countries, but none of them adjacent to Quebec. French would not die if the Québecois were assimilated, but it
would almost certainly die in North America.
There is no potential irredentist solution for Quebec; therefore,
secessionism is the likely outcome in these “between” cases.
[12] See
Geoffrey Garrett, Partisan Politics in the Global Economy (Cambridge, UK:
Cambridge University Press, 1998); Daniel W. Drezner, “Globalization and Policy
Convergence,” International Studies Review 3, 1 (2001): 53-78; Steven K.
Vogel, Freer Markets, More Rules: Regulatory Reform in Advanced Industrial
Countries (Ithaca, NY: Cornell University Press, 1996).
[13] By
contrast, the theory of “internal colonialism” holds that regions less well off
relative to the rest of the country are more likely to harbor resentment toward
the metropole for historical ill treatment (Michael Hechter, Internal
Colonialism: The Celtic Fringe in British National Development, 1536-1966
(Berkeley: University of California Press, 1975)). Horowitz (1985, pp. 229-88) argues
that backward groups in backward regions are those most likely to endorse
secessionist movements, as their elites are fearful of competition for public
positions from outsiders. This argument
does not explain why ordinary members of the group would go along with the
elites, given that the elites’ rent-seeking behavior would damage them more
than anyone else. Horowitz does note
that his hypothesized relationship does not seem to hold in wealthier
societies, a fact that suggests that the posited correlation may be conditional
on some attribute common to wealthy societies, such as democracy. In democratic countries wealthy regions are
likely to be losers from interregional redistribution, while in authoritarian
countries, poorer regions might be more likely to seek separation, as they
would expect domination from stronger groups from wealthy regions. Since this dissertation focuses on
democratic countries with well-developed party systems, the wealthy
secessionist region is expected to be the more common observation in the
sample.
[14] However,
Patrizia Messina (“Opposition in Italy in the 1990s: Local Political Cultures
and the Northern League,” Government and Opposition 33, 4 (1998)) argues
that the strongholds of the Lega Nord in northeast Italy differ from the rest
of Italy in other ways, being the center of small-scale export-led development
and traditional Catholic values.
[15] Horowitz
(1985), p. 621.
[16] I believe
that Mario Polèse (“Economic Integration, National Policies, and the
Rationality of Regional Separatism,” New Nationalisms of the Developed West:
Toward Explanation, ed. Edward A. Tiryakian and Ronald Rogowski (Boston:
Allen and Unwin, 1985)) first proposed this argument; see also Donald Wittman,
“Nations and States: Mergers and Acquisitions; Dissolutions and Divorce,” American
Economic Association Papers and Proceedings 81, 2 (1991), 126-29. Patrick Bolton, Gerard Roland, and Enrico
Spolaore, “Economic Theories of the Break-up and Integration of Nations,” European
Economic Review 40, 3-5 (1996): 265-87; Alberto Alesina and Enrico
Spolaore, “On the Number and Size of Nations,” Quarterly Journal of
Economics 112, 4 (1997): 1027-56; Patrick Bolton and Gerard Roland, “The
Breakup of Nations: A Political Economy Analysis,” Quarterly Journal of
Economics 112, 4 (1997): 1057-90; Alberto Alesina and Romain Wacziarg,
“Openness, Country Size, and the Government,” Journal of Public Economics 69,
3 (1998): 305-21; and Alberto Alesina, Enrico Spolaore, and Romain Wacziarg,
“Economic Integration and Political Disintegration,” The American Economic
Review 90, 5 (2000): 1276-96 have contributed to its formalization.
[17] Jacques
Parizeau, “Interview,” in Austin R. Riggs and Tom Velk, eds., Canadian-American
Free Trade: Historical, Political, and Economic Dimensions (Montreal:
Institute for Research on Public Policy, 1987). See also Hudson Meadwell, “The Politics of Nationalism in
Quebec,” World Politics 45 (January 1993): 203-41.
[18] See Walter
Edgar, South Carolina: A History (Columbia, S.C.: University of South
Carolina Press, 1998).
[19] Martin
1997. Consistent with the argument
presented here, Martin finds that the Parti Québecois, while similarly
supportive of globalization, stresses the institutional framework of
globalization, as represented by NAFTA and GATT, since these institutions would
guarantee a secure trading environment for an independent Quebec.
[20]
Accordingly, Pieter van Houten finds that “globalization and European
integration have changed and reinforced autonomy demands in already assertive
regions” in six West European countries, but regional-level globalization
variables fail to account for “variation in autonomy demands across regions”
(1). Pieter van Houten, “Globalization
and Regional Autonomy Demands in Western Europe,” Working Paper, Faculty of
Social and Political Sciences and Churchill College, University of Cambridge,
July 2001.
[21] Quoted by
T.R. Reid, “EU’s Potential Lifts Scots’ Hope of Independence,” Washington
Post Foreign Service, December 12, 2000; Page A01.
[22] This
reasoning seems to be behind the pronouncement of Joschka Fischer, Green
Foreign Minister of Germany, that “Europe is an objectively left project”
(translation from interview given to Profil, June 29, 1997).
[23] Keating
(Michael Keating, Nations against the State: The New Politics of Nationalism
in Quebec, Catalonia, and Scotland (Basingstoke, UK: Macmillan, 1996))
argues that minority nationalist groups are wary of decentralizing economic
policy, and that for this reason minority nationalisms are fundamentally
“dangerous” to business (220). Surely,
however, independence would involve a decentralization of economic policy. Minority nationalist groups that serve
merely as territorial lobbies rather than as secessionist parties might act in
a manner more consistent with Keating’s analysis.
[24] The exception
may be labor mobility, which can stimulate a backlash effect benefiting parties
of both left and right that support immigration controls. It is interesting to speculate why labor
mobility has to date provoked more of a backlash in most countries than has
capital mobility. Perhaps the reason
has to do with the visibility of labor movements relative to capital movements.
[25] Stéphane
Dion, "Why is Secession Difficult in Well-Established Democracies? Lessons
from Quebec," British Journal of Political Science 26 (1996):
269-283.
[26] Hueglin
argues for example: “[H]ierarchically centralized systems pass from the complex
to highly complex stage of organization, and there the efficiency benefits of
centralized systems’ control and regulation may be superseded by the costs of
system maintenance” (455). Thomas O.
Hueglin, “Regionalism in Western Europe: Conceptual Problems of a New Political
Perspective,” Comparative Politics (July 1986): 439-57.
[27] Whether a region
is Type 1 or Type 2 is considered here to be exogenous, but the section on
“Inter-Regional Variations in Secessionist Vote” presents several hypotheses
about the factors influencing whether a region is Type 1 or Type 2.
[28]
Hypothetically, there might also be regions that would secede no matter what
concessions the government offered them.
This logical possibility is well outside the limits of empirical
probability, however, as in the limit we could imagine a central government
offering to move the country’s capital to the region, spend 100% of the
country’s budget on the region’s citizens, and grant the region’s legislators
100% representation in the central parliament in perpetuity. Governments are of course unlikely to offer
such concessions, but that decision process is captured by the model.
[29] In the real
world, of course, the process is more complicated. A region doesn’t usually secede right away if it doesn’t achieve
what it wants. Instead, the vote share
of the secessionist party might go up at the next election. The government could then take this trend as
evidence that concessions are needed.
[30] An
alternative possibility is that the costs of secession for central states are
falling while the costs of appeasement are rising; this phenomenon might
explain part of the trend toward increasing secessionism.
[31] The
distances between the points are merely suggestive and are not meant to be
taken too seriously. The spatial model
below assigns different ideal points to the actors based on this graph, but the
reader is free to differ with the substantive interpretations given of these
points: they do not affect the model’s results.
[32] This
assumption is not unreasonable for advanced Western democracies, where a
majority victory in a secession referendum is usually assumed to signal the
beginning of negotiations for independence (e.g., Quebec). On the other hand, some democratic
constitutions specify supermajority requirements for secession. Nevis’ secession from St. Kitts was foiled
by just such a provision: a referendum on secession in 1998 gathered 61.7% of
the vote, shy of the two-thirds requirement (Ralph Premdas, “Self-Determination
and Decentralisation in the Caribbean: Tobago and Nevis,” St. Kitts and Nevis
Country Conference paper, latest revision September 2000).
[33] Surveys
show that in almost all peripheral regions around the world there are few
people whose ideal point is 6. Most
people seem to be willing to take a “good deal” in terms of autonomy and fiscal
incentives over the uncertainty of full independence.
[34] In an
iterated scenario, of course, secessionist leaders would respond with another
referendum, the government would compromise again, and the referendum would
fail again. Ultimately, policy would
approximate the median voter’s ideal position.
However, if regional legislatures promote confidence in independence in
the long run, voters’ ideal points would continue moving to the right on the
centralism-independence scale, all other factors being held constant.
[35] Again, we
don’t need to translate the points representing different kinds of arrangements
into any specific numbers along this dimension for the model to work.
[36] “The other
side, they understood very well that was the weak spot,” said Bouchard. “I must say that it was quite damaging. Obviously we lost a lot of support because
of that.” Source: Canadian Press
Newswire, August 21, 1998.
[37] These
predictions appear to contradict in some respects those of Rudolph and Thompson
(Joseph R. Rudolph, Jr. and Robert J. Thompson, “Ethnoterritorial Movements and
the Policy Process,” Comparative Politics April 1985: 291-311), who
argue that organizations challenging the boundaries of the state are “extreme
political separatists” with demands not susceptible to accommodation
(294). They argue that moderation of
separatist objectives has been necessary to open the door to electoral success
and to stimulate the center to negotiate, which negotiation eventually leads to
a decline in the electoral success of ethnoterritorial movements (296;
300). By contrast, the spatial model
suggests that a radical secessionist party is more likely than a moderate one
of equal electoral success (a crucial constant) to elicit concessions from the
central government. Hudson Meadwell (“A
Rational Choice Approach to Political Regionalism,” Comparative Politics
July 1991: 401-21) argues accordingly that “regionalists . . . may present
themselves publicly as supporters of independence . . . as a bargaining
strategy designed to win concessions from the state” (402).
