Posts Tagged ‘secession’

On March 15, I had the opportunity to testify at the U.S. House Foreign Affairs Subcommittee on Europe, Eurasia, and Emerging Threats, chaired by California Rep. Dana Rohrabacher, on the topic of whether the U.S. government should change its policy toward national self-determination movements. I’m posting here my written testimony (my oral testimony had to be briefer). The conversations with the congressmen were interesting, and I think they revealed something about the foundations of U.S. policy toward such movements. Rep. Weber grilled me a bit about whether countries such as Spain, France, and Italy that prohibit secession in their constitutions should not enforce their constitutions. I responded that these countries should change their constitutions. However, another point I would make now is that the purpose of constitutions is to constrain governments, not citizens. Statutes and administrative regulations constrain citizens, but the constitution in turn restricts the ways in which government can create and implement these rules. So a constitutional prohibition against secession could really be enforced only against government agencies, and a government would be well within its legal rights to allow citizens to pursue secession, even in the face of a constitutional prohibition.


National self-determination movements seek greater self-government for a national minority, typically including the right to vote on forming a new independent state. Recent examples of successful self-determination movements include South Sudan, Kosovo, Montenegro, and East Timor. Ongoing self-determination claims are found in Scotland, Catalonia, the Faroe Islands, Kashmir, Tamil Eelam, Somaliland, Western Sahara, West Papua, Tibet, Mindanao, and many other places. Like other states, the U.S. government faces decisions about whether to recognize declarations of independence, to enter into diplomatic relations with new states, and to engage in diplomacy with other states about self-determination movements within their borders.

In my testimony, I will first describe the current state of self-determination movements around the world, then summarize what scholars have learned about the relationship between self-determination conflicts and violence. I will conclude by assessing the validity of claims advocating the creation of new states or changes to national borders.

The Current State of Self-Determination Movements

Self-determination movements generally take one of two forms: political parties and armed groups. Currently, secessionist political parties that seek at least a vote on independence are found in Belgium (Flanders), Canada (Quebec), Denmark (Faroe Islands and Greenland), Finland (Åland), France (Brittany and Corsica), Germany (Bavaria), Italy (Veneto and Sardinia), Spain (Catalonia, the Baleares, the Basque Country, Navarre, Canary Islands, and Galicia), the UK (Scotland and Wales), and the United States (Alaska and Puerto Rico). In addition, irredentist parties, which seek to move territory from one country to another, are present in the UK (Northern Ireland) and Italy (South Tyrol).

Armed self-determination movements are typically found in the developing world. Figure 1 shows where intrastate armed conflicts on territorial issues (generally, self-determination) occurred during the 2011-2014 period. These conflicts require at least 25 battle deaths in a single year to be counted.

territorial conflicts map

Developing countries usually forbid self-determination movements from organizing as political parties. For instance, Turkey, India, Pakistan, Sri Lanka, and Russia make advocacy for the self-determination of a particular region a criminal offense — an act that would be protected by the First Amendment in the United States.

Western, liberal democracies typically allow secessionist parties to organize and contest elections, but they do not all allow for secession. The Canadian Supreme Court has ruled that Quebec secession is negotiable if the province votes by “a clear majority on a clear question” for independence. The United Kingdom negotiated the terms of an independence referendum for Scotland and agreed to be bound by the result. The Danish government concedes a right to independence for Greenland and the Faroe Islands, and breakup is routinely discussed as a legal option for Belgium. St. Kitts and Nevis and Liechtenstein have constitutional clauses protecting the right of secession. On the other hand, France, Spain, and Italy all have constitutions explicitly defining their countries as indivisible, thus proscribing secession.

Majority support for independence in a population is rare. As of this writing, in all the high-income democracies of Europe, North America, and the Pacific Rim, there is only one region in which parties clearly favoring short-run independence have won an absolute majority of votes in any recent election: Scotland. Furthermore, in Scotland, many voters voted for the Scottish National Party (SNP) without favoring independence, and support for independence has been below 50 percent in polls since that election, including
the September 18, 2014 referendum itself. Using data from the Minorities at Risk project, I found that as of 2003, 107 ethnonational minorities, 38 percent of the total number in the data set, had a secessionist organization of any size (Sorens 2012, 56). In a recent article, I estimated the percentage of the population supporting independence in every state of India, finding figures no higher than 20 percent anywhere (Sorens 2014, 264).

The Causes of Self-Determination Conflicts

Popular demand for independence comes from a combination of a distinctive cultural identity, territorial coherence, and either political or economic benefits of independence (Sorens 2005; Hale 2008; Sorens 2012). Having just one of these elements is not enough, which is why the vast majority of minority nations around the world do not have any secessionist movement at all.

One worry about allowing secessionist movements is so-called “contagion” across regions or countries, but secessionism does not in fact seem to be contagious across countries, although it does have a tendency to spread within a country (Ayres and Saideman 2000; Sorens 2012), which is why governments often crack down on them (Walter 2006).

At the individual level, there is some evidence that voter support for independence is rational, that is, related in the expected way to the expected benefits of independence (Howe 1998). However, there is a difficult-to-resolve debate about the extent to which independence support is caused by voters’ assessments of the benefits of independence, or if instead independence support causes those estimates of benefits through a process of rationalization (Mendelsohn 2003).

Secessionism is strongly associated with violent conflict (Toft 2003). In general, separatist civil wars last longer than other kinds of wars, implying that the warring parties cannot find negotiated settlements even when the conflicts are stalemated (Fearon 2004; Sorens 2012).

I find that providing a legal path to independence is associated with less ethnonationalist rebellion (Sorens 2012). The United Kingdom, Canada, Denmark, and Belgium have had much less secessionist violence than France, Spain, and Italy — and secessionist violence has gone away in Puerto Rico since the U.S. government informally recognized their right to independence. Clauses permitting secession were also crucial to peace agreements ending the conflicts in Northern Ireland, South Sudan, and Bougainville (part of Papua New Guinea). The European Union’s Treaty of Lisbon explicitly recognizes member states’ right to secede from the Union, because no country would want to join a union they could never leave.

Implications for U.S. Policy

A legal path to independence can promote peace by constraining secessionists and central governments to pursue their aims through electoral and legislative means. On the one hand, secessionists have no excuse for resorting to violent tactics; to do so would be to admit failure to persuade a majority of the people they claim to represent, while imposing costs of violence on the very people they purport to represent and from whom they would have to recruit. On the other hand, central governments often cannot commit to respecting
a negotiated regional autonomy compromise without also conceding a right to secede. The South Sudanese and Bougainvillean secessionists would probably not have agreed to a peace deal without a referendum guarantee. These conflicts lasted 22 and nine years, respectively. Authoritarian and especially nationalistic central governments will face both desire and opportunity to renege
on previously negotiated autonomy arrangements; only a right to secede may be sufficient to deter them and thereby induce secessionist rebels to lay down arms in the first place. I also find that central governments permitting a legal path to independence are more likely to decentralize to ethnic minority regions and have never recentralized power in the post-World War II era (Sorens 2012).

If every country recognized its minority nations’ right to secede, only a few would apparently exercise such a right. Moreover, the overall level of global violence would likely decline by replacing intrastate conflicts with interstate conflicts. Intrastate conflicts are far more common than interstate conflicts (see Figure 2). Since World War 2, civil conflicts have killed seven times more people than interstate conflicts (Collier and Sambanis 2005; PRIO n.d.). Civil wars last much longer than interstate wars (Fearon 2004). Civil wars are also more likely to happen in more populous countries (Fearon and Laitin 2003). These findings suggest that a global increase in the number of independent states and a decrease in their average size would decrease the total number of conflict deaths.

conflict types

There are good reasons for the U.S. government to avoid assertively internationalizing other countries’ self-determination conflicts, which can look like meddling in other countries’ internal affairs. The U.S. arguably erred in refusing to negotiate a democratically authorized partition of Kosovo; as a result, an independent Kosovo lacks broad recognition from other states and is having trouble entering international institutions. Nevertheless, once a declaration of independence is issued, the U.S. government has no choice but to respond. In such an event, the U.S. government might wish to consider not only the interests of the host state, but also the interests of the seceding state and the effect of secession on regional stability. On average, replacing a state-to-nation relationship with a state-to-state relationship reduces violence.


Ayres, R. William & Stephen Saideman. 2000. “Is Separatism as Contagious as the Common Cold or as Cancer? Testing the International and DomesticDeterminants of Secessionism.” Nationalism and Ethnic Politics 6(3):92–114.
Collier, Paul & Nicholas Sambanis. 2005. Preface. In Understanding Civil War: Evidence and Analysis: Volume 1 (Africa), ed. Paul Collier & Nicholas Sambanis. Washington, D.C.: The World Bank.
Data on Armed Conflict. 2013. Peace Research Institute of Oslo. http://www.prio.no/Data/Armed-Conflict/, accessed September 20, 2013.
Fearon, James D. 2004. “Why Do Some Civil Wars Last So Much Longer than Others?” Journal of Peace Research 41(3):275–301.
Fearon, James D. & David D. Laitin. 2003. “Ethnicity, Insurgency, and Civil War.” American Political Science Review 97(1):75–90.
Hale, Henry E. 2008. The Foundations of Ethnic Politics: Separatism of States and Nations in Eurasia and the World. Cambridge, U.K.: Cambridge University Press.
Howe, Paul. 1998. “Rationality and Sovereignty Support in Quebec.” Canadian Journal of Political Science 31(1):31–59.
Mendelsohn, Matthew. 2003. “Rational Choice and Socio-Psychological Explanations for Opinion on Quebec Sovereignty.” Canadian Journal of Political Science 36(3):511–537.
Sorens, Jason. 2005. “The Cross-Sectional Determinants of Secessionism in Advanced Democracies.” Comparative Political Studies 38(3):304–326.
Sorens, Jason. 2012. Secessionism: Identity, Interest, and Strategy. Montreal, Que.: McGill-Queen’s University Press.
Sorens, Jason. 2014. “Legal Regimes for Secession: Applying Moral Theory and Empirical Findings.” Public Affairs Quarterly 28(3):259–288.
Toft, Monica Duffy. 2003. The Geography of Ethnic Violence: Identity, Interests, and the Indivisibility of Territory. Princeton, N.J.: Princeton University Press.
Walter, Barbara F. 2006. “Building Reputation: Why Governments Fight Some Separatists but Not Others.” American Journal of Political Science 50(2):313–330.

Read Full Post »

Catalonia will hold a de facto independence referendum through regional elections on September 27, 2015. This one will have “real” effect, unlike the 9N, because the Catalan independence parties would form a unity government and set up the institutions of an independent state, ultimately declaring independence at a date yet to be announced.

Read Full Post »

My paper on the political philosophy of secession is now out in Public Affairs Quarterly, an open-access journal. Read it here. Teaser:

The United Kingdom currently sets the gold standard for management of secessionist politics. The British and Scottish governments negotiated in good faith over the terms of the independence referendum that Scotland held on September 18, 2014. If Scotland had voted to secede, the British government would have recognized its independence, thus affirming that the United Kingdom is a free partnership among its peoples.

Spain presents a different scenario altogether. Catalonia intends to hold its own “consultation” on independence, but the Spanish government has denied its right to do so, thus denying that Spain is a free partnership. The Catalan government has repeatedly sought to hold negotiations on the self-determination process, but has been rebuffed. What ought the Catalan government to do? By the criteria set forth in this paper, Catalonia has tried to conform to a just institutional regime for regulating secessionist politics, while Spain has not. Catalonia would be justified in using all proportionate means to secure a just outcome.

Read Full Post »

Constitute.org is a useful website designed by political scientists to let researchers search for and compare constitutional texts on particular topics. Here for instance is a search on secession clauses. Although one of the site’s creators, Zachary Elkins, says that 22 states contemplate some process for state divorce, only three constitutions expressly authorize some part of the country to secede: Ethiopia, Liechtenstein, and St. Kitts and Nevis. Ethiopia lets each people or nationality secede by a supermajority vote of its legislature, Liechtenstein lets each commune secede (I believe this was an addition of the 2003 constitution), and St. Kitts and Nevis lets Nevis secede by a supermajority referendum vote.

In addition to these, Britain’s Northern Ireland Act of 1998 lets the majority of Northern Irelanders decide to join the Republic of Ireland, and the constitution of Uzbekistan lets Karakalpakstan secede with the consent of the Uzbekistan government.

It would be interesting to see how many states define themselves as “indivisible,” thus tying a government’s hands and preventing it from authorizing secession. A search on the term brings up some irrelevant cases, but 72 constitutions contain the term.

By the way, the Prince of Liechtenstein is a moderate libertarian, and their constitution is fairly consistent with the philosophy. Check out his book on the topic.

Read Full Post »

Last week I was in Barcelona for two days, giving a talk at an event on “the right to decide,” sponsored by the Centre Maurits Coppieters (nonprofit arm of the European Free Alliance, the European Parliament group for ideologically mainstream minority nationalist parties) and by the Fundació Josep Irla (nonprofit arm of the Catalan Republican Left [ERC], largest pro-independence party in Catalonia). I also did some media interviews. You can see some excerpts of my interview with Catalonia’s TV3 evening news here (in Catalan).

I was interested in going to find out more about Catalonia’s independence movement and its prospects. For background on the Catalan movement, see my post here on Pileus from September 24, 2012, two weeks after the massive Catalan National Day demonstrations that kicked off the current process. (That post, including its forecasts, has held up pretty well, I’d say.)

Now that the Spanish Constitutional Court has invalidated the consulta (consultative plebiscite) that the Catalan Government had authorized with the support of over two-thirds of Catalan MP’s and three-quarters of the Catalan electorate, the way forward is murky. An official consultation will not now happen. Instead, tens of thousands of volunteer poll workers are signing up to help with an unofficial poll that will involve ballots and ballot boxes and occur on November 9.

It remains to be seen how successful the November 9 consultation will be. The pro-independence parties and civil society organizations are trying hard to mobilize voters and volunteers for the event. The anti-independence parties are boycotting the vote, as indeed are some far-left types who hate Artur Mas, such as the leader of the ex-communist, ecosocialist party ICV-EUiA, which otherwise supports the “right to decide” and remains agnostic on independence.

If the November 9 consultation is successful, then the pro-independence parties will try to negotiate a “unitary party list” for early elections to the Catalan Parliament. They will treat this election as a plebiscite-by-proxy, and if an absolute majority of seats and votes go to the pro-independence list, Artur Mas will take it as a mandate for independence.

However, several difficulties remain. The more radically independentist party, ERC, wants to declare independence right away after a successful “plebiscitary election.” Artur Mas’ party, Convergence and Union (CiU), is divided between independentists and those favoring a solution like confederation. (Technically, the party is a long-standing alliance between two separate parties, the now-independentist Democratic Convergence of Catalonia and the autonomist Democratic Union of Catalonia.) Generally, the last few days have seen more division and acrimony among secessionist leaders than the previous two years, and if it continues, that division will alienate voters. Civil society groups continue to call for unity among the pro-independence leaders.

Another difficulty is that while a majority of Catalans with an opinion on the matter favor independence (a recent El Mundo poll had the anti-independence side ahead within the margin of error, but their polls have always been biased in an anti-independence direction), polls suggest the pro-independence parties would not together gain a majority in early elections. The reason for this is that many independentists are not in the secessionist parties. For a successful result, the “unitary list” will need to contain important leaders from civil society and non-secessionist parties.

If the Catalan process stumbles now, it will be a shame, because it will show the Spanish government that they can face down demands for more autonomy simply by standing pat and threatening to arrest politicians. Spanish autonomous communities like Catalonia enjoy far less autonomy than American states (they are not allowed, for instance, to vary the overall tax burden from a central standard).

The most likely outcome of the process now seems to be (more…)

Read Full Post »

Unless the polls are systematically biased or there is a late-breaking surge in support for “Yes,” the “No” campaign looks set to squeak by with a narrow victory in the Scottish independence referendum. On the betting markets, a “Yes” vote has plunged below an implied probability of 20%. What has this decline in the prospects for independence done to capital markets? In my last post on the subject, I found that British firms and the pound were nearly untouched by what was at the time significant momentum for “Yes,” but that a nine-firm Scottish equity index was hit hard. If those losses reflected unease about independence, then the latest news should have caused growth in my Scottish equity index.

The biggest decline in the Yes team’s chances actually came overnight September 11-12, when the chances of Scottish independence abruptly fell about 10 percentage points on the release of new polls in the evening of September 11. (For a full list of recent polls, see Wikipedia.) The Yougov poll showing “No” in the lead (a dramatic reversal from its previous poll) seems to have been leaked just before the closing bell on September 11.

betfair independence odds

Accordingly, I examine the performance of the Scottish equity index on the London Stock Exchange between 4:30 PM and 5:00 PM local time on September 11, when the odds of Scottish independence declined so rapidly. These are the nine stocks I include in the index: SL, SSE, FGP, WEIR, SGC, AGGK, WG, ADN, and MNZS. Of these, eight of nine rose on the poll news. Again, I weight by each stock by its market cap to create the index. The index rose 0.5% on the news, a rather small increase compared to the 1.7% decline after the shock Yougov poll showing “Yes” ahead. The overall patterns were pretty similar, though. The two transportation companies, Firstgroup and Stagecoach Group, were basically unchanged between the two. Energy-linked firms and Standard Life led gainers. Aggreko (temperature control systems) registered a small gain, and Aberdeen Asset Management a somewhat larger one.

Roughly a ten-percentage-point drop in independence likelihood led to a 0.5% gain in the value of Scottish equities, less than a third of the loss in Scottish equities after an eight-percentage-point gain in independence likelihood just a few days prior. On balance, these results suggest we should revise downward the costs of secession suggested by the prior post.

One objection to this interpretation might be that the leak of the Yougov poll just before the closing bell gave traders little time to respond. But this does not appear to be the case. The Scottish equity index was flat at the opening bell on September 12, suggesting that there was no new information for traders to consider.

Here are two more interpretations. First, betting markets are less liquid and well-capitalized than financial markets. The actual gain in the probability of Scottish independence after the first Yougov poll may have been greater than the immediate response on betting markets. Second, the shock of a poll actually showing “Yes” ahead may have led traders to overestimate the likelihood of Scottish independence, and perhaps even the costs of secession (in a moment of panic). Having been inured to the initial shock and its aftermath, traders then took later news with more equanimity.

Overall, though, the results are still suggesting net economic costs to Scottish independence. How much of the emphasis should be put on the “Scottish” part of that phrase and how much on “independence” remains a matter of debate, but clearly energy and financial firms are more affected than transportation and service ones.

Read Full Post »

What can we learn from capital markets about the likely consequences of Scottish independence? A trio of recent polls has shown the “Yes” side to have pulled roughly even with “No.” With momentum on their side, it’s not unthinkable at all that “Yes” will pull it out, resulting in the first secession from a Western democracy since Iceland withdrew from Danish union in 1944. Most American commentators, from Paul Krugman to Tyler Cowen, oppose Scottish independence and forecast economic disaster for the new country. Are they right?

Let’s look at the behavior of capital markets in Britain since these polls’ release to find out. First, let’s set the stage by looking at how betting markets price the probability of Scottish independence. Unfortunately, there are no nice InTrade-style charts for implicit probabilities anymore, at least not that I can access from the United States. From oddschecker.com, I am able to pull odds from different exchanges from the beginning and end of each day. Looking at the markets with most liquidity, it looks as if the odds for independence moved from about 19.5% Friday night to about 25% Saturday night, after the release of the YouGov and Panelbase polls (the Panelbase poll suggested “No” might still have a small lead). On Monday morning the odds stood at about 23.3%. After the release of the TNS poll Monday evening (confirming the dead heat), the odds moved in to 26.0%.

Next, let’s look at the behavior of capital markets over this period. Here is how the pound has fared against the euro:

euros per pound

Not much of a correlation. To be sure, the pound fell against the euro when trading opened Monday morning, following the shock weekend poll from Yougov and the somewhat-reassuring poll from Panelbase, but the TNS poll released late Monday night appears to have had zero effect on the pound, even though it did have a small effect on the betting markets.

Now, the pound has fared a little worse against the dollar, because the euro has also dropped against the dollar. This may reflect that traders believe Scottish independence raises the probability of British exit from the EU. But this would not be a direct cost of Scottish independence, and it would ultimately be up to English, Welsh, and Northern Irish voters whether they want to withdraw from the EU.

What about the stock market?


I’ve got the Dow in there for comparison (in green). So the FTSE fell about 0.3% on opening Monday, then drifted downward throughout the day, finally recovering all that ground except the initial 0.3% drop. On opening Tuesday after the TNS poll, it actually rose. As of this writing is down just about 0.4% from Friday’s close. This looks like a muted response to me.

But what about Scottish-exposed stocks in particular? I took the list of top 25 Scottish companies here and winnowed the list down to those listed as having Scottish ownership and being publicly traded. Nine companies fit that test. I then constructed a weighted average of their share prices at Friday close, Monday open, Monday close, and Tuesday open, the weights being each stock’s market cap according to investing.com. Recall that there were two surprise polls, one over the weekend and one released Monday evening, the former having the greater effect on betting markets.

The Scottish index I created lost 1.7% of its value on opening Monday morning, a noteworthy drop because it happened right away. It’s plausible to attribute this drop to the increased risk of independence. However, today it lost nothing on opening – in fact, it was up 0.1%. Still, the total loss to these nine firms’ market value amounts to about $800 million. The fact that there was no further response of capital markets to the TNS poll, even though betting markets did respond, weakens our confidence somewhat that investors are responding negatively to the prospect of independence, but let us work with the assumption that they are.

What would happen to these firms’ value if independence were dead certain? Expected utility analysis helps us here. They lost $800 million in value on an increase in the probability of independence of 5.5+2.7=8.2%. We can infer that an increase from 20% to 100% would wipe out $800 million*8/.6=$7.8 billion. That’s a fair proportion of their existing value: about 16%. Of course, investors are risk averse, and the very uncertainty of the outcome might be driving a fair proportion of the losses.

A closer look reveals that different stocks responded differently to the poll news. Two transportation companies, FirstGroup and Stagecoach Group, lost virtually nothing, and Aggreko, which rents temperature control systems, lost absolutely nothing. Financial and energy/power companies were pounded. An engineering company closely linked to the oil industry, the Weir Group, took a more modest 1.0% loss.

How to sum up? So far (more…)

Read Full Post »

Older Posts »

%d bloggers like this: