Comparative lessons on state structure, identity, compromise and ‘fluidity’ from a veteran of global federalism
Guest contributor
Igor Blažević
In the midst of Myanmar’s ongoing struggle for democracy and federalism, I had an in-depth conversation with George Anderson, an expert on federalism and post-conflict state-building.
The Q&A, moderated by Igor Blažević, brought together Anderson and a group of Myanmar anti-coup activists to explore the challenges and options for creating or modifying the state structure in the nation’s future federal democratic union.
Anderson brings a rare combination of academic insight and practical experience. A former President and CEO of the Forum of Federations, he also served for over three decades in Canada’s federal public service.
He has advised on federal transitions in over 20 countries and authored foundational texts such as Federalism: An Introduction and Fiscal Federalism: A Comparative Introduction.
Drawing from global comparative experiences—in India, Nigeria, Canada, Ethiopia, Nepal, and beyond—Anderson examined the dilemmas Myanmar faces: ethnic representation, state boundaries, institutional design, and the difficult path from resistance and liberation toward constitutional government, a sense of shared citizenship, a spirit of compromise, ‘fluid’ politics, and the establishment of a functioning and peaceful state.
His insights offer valuable guidance for those shaping the future of Myanmar’s governance.
Balancing ethnic identity and shared citizenship
IB: Let’s start with one of the key questions about the future of Myanmar. Should Myanmar be organized as an ethno-centric federation, a territorial federation, or some kind of hybrid of both?
On one hand, we have 70 years of clear aspirations from Myanmar’s ethnic nationalities for self-determination and self-rule, including a high degree of autonomy within their own states. They have now liberated significant territories and are already building their ethnic nation-states. It’s hard to imagine how this can be reversed in any future scenario.
On the other hand, we face a very complex demographic situation. Many existing states and regions do not have a clear ethnic majority. What is your take on this? Is it possible to reconcile the demand for ethno-centric states with Myanmar’s extreme diversity and demographic complexity?
GA: Obviously, the challenges that Myanmar faces are considerable. It’s very difficult to make the transition from civil war to democracy, and it’s doubly difficult to make the transition to a democratic federation, because federalism adds a whole other dimension in terms of the types of issues that need to be resolved.
To make a successful transition to federalism, I think you need three fundamental commitments. First, you need a commitment to constitutional government, because federalism defines the roles of the federal and state governments. It sets up the central institutions, and respect for the constitution is critical.
Second, you need to have a sense of belonging not just to your local community—your ethnic group or region—but to the national community as a whole. If people only identify with one part of the country, that creates significant tension. So it’s essential to work on building a sense of shared citizenship across Myanmar, alongside the more localized forms of identity.
Third, you need a spirit of compromise and mutual tolerance. In a federation with diverse populations and experiences, nobody is going to get everything they want. It’s about understanding priorities. Coming out of a civil war, the number one priority is peace and security—establishing a functioning state. Compromises will be necessary to build a system that works across the whole country, and that spirit of compromise is absolutely essential for success.
IB: We could begin by saying: ‘Let’s try to create as homogenous as possible ethnic states in a future federal Myanmar, and try to achieve that by redrawing borders. But as someone coming from Bosnia, I am always concerned when people propose redrawing internal borders in countries with a mixed population.’ In Bosnia, attempts to do so triggered horrific ethnic cleansing. On the other hand, India showed that under certain mechanisms, it has been possible to carve new states from existing ones to form more homogeneous linguistic units. What do you think is the right approach for Myanmar?
GA: You only need to look at the map of Myanmar to see the extent of population overlap and how ethnic groups are divided regionally. They’re not neatly organized into single, separate territories. Trying to draw a map where each ethnic group has its own clear-majority territory would be extremely difficult—virtually impossible.
Nepal had a similar experience. They initially committed to ethnic federalism, but when they confronted the demographic reality, it proved unworkable. In the end, they moved away from ethno-centric model. When they created their states, they included a provision that no state could be named after an ethnic group, recognizing that every state would contain a mix of people. We’ll see how things work out in Nepal, but at the moment the system seems to be functioning fairly well.
This goes to the point I was making about compromise—it’s absolutely essential for people to recognize the impossibility of achieving all they want.
You also mentioned India, which is a useful example. When India gained independence, it knew it would be a federal country, but it started with 27 states inherited from the British colonial administration. The Indian leadership understood this arrangement wouldn’t work in the long term.
A few years later, they established a commission to examine how to reorganize the states. This led to the creation of 18 new states, and over time, more were added in response to social and political pressures. India adopted a progressive and evolutionary approach to federal restructuring rather than attempting to resolve all state boundaries at the moment of independence.
IB: If we imagine a scenario where the military junta has been removed, and we begin a political process to build the country’s future institutions. We start with 14 administrative units inherited from the previous system. But on the ground, we now have de facto new states—Wa, Kokang, Ta’ang, as well as Arakan, Chin, Kachin, Karenni, and Karen states. What is your advice, based on comparative experience? Should we begin with these 14 units and gradually adjust them, or should we quickly try to agree on a new state structure formed during the liberation struggle?
GA: Experiences vary widely across countries. In Nepal, they created a new parliament tasked with drafting a federal constitution. But during the first phase, they did not yet have states—it remained a unitary system in transition. The constitution needed a two-thirds majority to pass, and it took two elections to get there.
In Ethiopia, after the 1990s transition, the ruling party was strongly committed to ethnic federalism. They drew the map accordingly. Still, in many regions, there was no clear ethnic majority in the new states. Over time, they created a few more states. So, like India, it was an evolutionary process—they started with a fair measure of consensus and gradually adapted.
In Nigeria, when the country was first created, it had three states—one in the north, which had the majority of the population, and two in the south. After the military took over, they gradually redrew the map, eventually creating 36 states. The aim was to break up large ethnic blocs and foster more fluid politics. Some say they created too many states, but it did facilitate more fluid and more consensual politics than the earlier structure with big blocks one against the other.
A key point if you delay redrawing the federal map is: ‘What’s the decision rule?’ In India, the central government had the authority to create new states. In every case but one, they did it with the consent of the states affected, but they could act unilaterally if needed. In Nigeria, the military imposed the changes. In Spain, the constitution allowed and required amalgamation of units to create larger states but did not allow redrawing borders. So if you’re going to defer state restructuring, you need to define how that process would work and who decides.
IB: You’ve mentioned ‘fluid politics’ a few times. Can you explain what you mean by fluid vs. rigid politics, and why fluidity matters?
GA: When politics becomes too strongly based on identity—ethnic, linguistic, religious—you risk people always voting for the same party or one aligned with their group. Over time, you get stuck. The same forces keep disagreeing, and they do not move to make change.
Let me give an example from my country Canada. We have one French-speaking province—Quebec—about 25 percent of the population, and nine predominantly English-speaking provinces.
Over time, the alliances between Quebec and parts of English-speaking Canada have shifted significantly, depending on the issue. We’ve never been locked into a pattern where Quebec votes one way and the rest of Canada another. As issues change, alliances change.
Political parties get elected in all parts of the country, not just one part of the country. That creates fluidity of politics and reduces the temperature of political conflict. It’s important to remember that there are more issues in politics than those linked to ethnic identity.
For example, in federations, some states are rich in extractive resources, some are heavily agricultural, there are big cities and rural areas, so economic interests differ amongst states and these can cut across ethnic lines: States with different ethnic identities may have strongly shared interests on non-ethnic issues.
Over time the importance of different issues shifts and this can lead to shifting political alliances of territories and states—in other words to fluid politics. It also helps create an environment in which the losers in one election can hope to win in a future election and where political parties can remain sensitive to the interests of regions or groups that did not support them.
Thus in Canada, India, and Switzerland—all federations with several states having the same language—the main political alliances are not strictly along language or ethnic lines, but reflect a variety of issues.
Balancing majorities and minorities in federal structures
IB: What’s your take on the idea of having one or two ethnic “Bamar states” in a future federation? There is an increasing number of actors in Myanmar who are calling for that—not only on the ethnic side but also increasingly among Bamar actors as well.
GA: If you have one Bamar state that includes 65 percent of the population, it would result in a very unbalanced federation. That one state would be so large and powerful that it would fundamentally alter the nature of the country.
There are federations where one or two states are very large—but not to that extent. I mentioned Nigeria, where they deliberately broke up the large northern and southern states to encourage more fluid politics.
You’re more likely to find shared interests between the Bamar population and the various ethnic communities in a true federal system if states are somewhat comparable in size. Some Bamar majority states might find their economic interests, for example, align more with some of the ethnic majority states.
Currently, in Myanmar, the biggest states like Yangon and Mandalay have around 6–7 million people, while the smallest—like Karenni and Chin—have about 300,000 to 500,000 people. There’s already a significant variation, but creating one state that encompasses 65 percent of the population would overwhelm all the smaller states.
IB: If there is no single “Bamar State” in a federal Myanmar, then where would Bamar people live? How would they be represented?
GA: The existing structure of seven states and seven regions could be retained. Instead of creating a single Bamar State, you could keep this current structure, especially for the Bamar population.
There are examples from other countries. India has several Hindi-speaking states. Canada has multiple English-speaking provinces. Switzerland has several German-speaking and several French-speaking cantons.
A single linguistic group doesn’t have to be concentrated in one state. In fact, it may be more beneficial for that group to have localized governments. For example, Bamar people in Mandalay may have different interests than those in Yangon, and a localized government could better respond to those specific needs.
IB: But the concern I hear from colleagues in Myanmar is that if we have more states where the Bamar are the majority, then the Bamar population could dominate decision-making in a future federal system.
Ethnic nationalities worry they will always be overruled—both in the Lower House and Upper House—because there would be more Bamar majority states. So they feel they will always be the losers.
How can we balance politics in a country with such a disproportion in population size—with one group being 65 percent and others just 5–7 percent, or even less?
GA: There are a lot of ways to address that concern.
Normally, a Lower House is based on representation by population, so yes, the Bamar population would likely have a majority in the Lower House. But it may not be a full 65 percent majority if the electoral design gives smaller states slightly more representation than their population would normally justify. That’s one adjustment—giving extra weight to minority population in the Lower House.
Another approach is to introduce special voting procedures on sensitive issues. For instance, in Belgium, certain issues that are especially sensitive to either the French-speaking or Dutch-speaking communities require a majority within each group. Myanmar could adopt similar mechanisms—sparingly—to protect minorities in key areas of legislation.
You can also design the electoral system to encourage political parties to seek votes across the entire country. That’s difficult but possible. If done right, it can incentivize parties to represent a broad national interest, rather than being confined to a single ethnic or geographic base.
Still speaking about the Lower House—you can create rules for government formation that require representation from multiple regions or groups, not just a simple majority. If these rules are too rigid, it might become difficult to form a government. But if done thoughtfully, such rules can ensure inclusive coalitions and a broader sense of representation.
Finally, there needs to be a sense of common citizenship, regardless of ethnicity, which means all have the same fundamental rights. There can also be measures to promote a representative mix of backgrounds in various government institutions—the bureaucracy, the military, the courts.
IB: And what could be the role of the Upper House in a federal Myanmar in protecting ethnic nationalities from being dominated by the Bamar majority?
GA: Almost all federations have an Upper House, though their powers and structures vary significantly.
In Canada, for example, the Upper House is unelected and relatively weak. In Belgium, they are even considering abolishing it altogether. So while an Upper House is common in federations, it’s not mandatory. But assuming Myanmar decides to have one, there are different models to explore.
In most federations, Upper Houses are based on representation by state or province. In some, each state has equal representation—the same number of members or senators. In others, the number of representatives is weighted by population, though still less proportionally than in the Lower House. For instance, large states may have six or seven senators, while smaller ones might have only one.
You can play with this. This structure allows adjustments and balance. This flexibility means there are many ways to adjust representation to reflect the country’s diversity and demographics.
I know that there is a concern in the ethnic communities about equality and the sense that if there is one Bamar State that would make the Upper House a place where all ethnic states would have a lot more weight compared to the Bamar population.
However, a scenario where the Bamar majority is reduced to a small minority in the Upper House would likely not be acceptable to them. Such an arrangement would not be realistic or sustainable.
It’s also important to consider the actual powers of the Upper House. In the United States, the Senate is powerful—it can approve treaties and key government appointments that the House of Representatives cannot.
In Germany, the Upper House (Bundesrat) plays a more limited role. It’s technically not even part of Parliament. Its powers are mostly restricted to state-level issues. Some matters are handled only by the Lower House; others require approval from both chambers.
So the key takeaway is: There’s room for balancing mechanisms between the two houses. A well-designed Upper House can ensure meaningful participation from all states—including ethnic states—without marginalizing the majority. And all of this can still work within Myanmar’s existing framework of 14 regions and states.
Protection, decentralization, and a roadmap for transition
IB: Apart from the federal system and the existence of Upper and Lower Houses, what other mechanisms can help ensure equality among different ethnic groups?
GA: Looking beyond the two houses, one important consideration is the protection of minority rights. Minorities are often most concerned about language—specifically, access to education and government services in their own languages. These rights can be constitutionally guaranteed, both at the national and regional levels.
Another approach, rather than creating entirely new states, is to use special administrative areas or municipalities with significant responsibilities. Myanmar already has some experience with this—you have several special administrative regions.
But you have to be careful. For example, in Mizoram in northeast India, they created a special administrative region that has almost become a state within a state. You want to avoid creating a special administrative region within a state that basically takes all the responsibilities of the state, and just puts them down to the next level.
It’s important to carefully determine what responsibilities should remain at the state level and what can be devolved to special regions or municipalities.
Finally, regardless of which ethnic group is dominant in a given state, the principle of equality for all citizens or residents should be constitutionally enshrined to prevent discrimination and ensure fairness.
IB: What do you think about deep decentralization—not just federalism at the state level, but pushing governance and budgeting down to the township level, as Indonesia has done? Could that be a solution for future Myanmar?
GA: Absolutely, I think deep decentralization is worth serious consideration. But there are a few key issues to think through.
First is capacity—will local governments have the necessary human and financial resources to manage the responsibilities given to them?
Second is constitutional design—should this three-tier structure be explicitly defined in the national constitution, or should individual states be allowed to manage their own internal governance structures?
IB: Now that many ethnic armed groups have grown stronger through the Spring Revolution and control large territories, won’t they have a stronger voice in the future? They have de facto power and control—won’t that influence the shape of the future?
GA: It’s inevitable that, during the transition, the National Unity Government (NUG) will need to engage with ethnic armed organizations—and vice versa. Neither side can impose its will. There needs to be a process of dialogue and consensus-building.
We’ve spent a lot of time talking about the design of a federal system, but you need to spend just as much time thinking about how you design the transition process. How will you organize discussions that lead to a new constitution?
Are you going to try to solve everything at once, or proceed in stages? In Nepal, for example, they had a two-step process: First, they elected a Parliament, and then that Parliament created the constitution. You might also decide to settle some issues now and leave others for a later stage.
IB: Given Myanmar’s ethnic diversity, which system is better: Presidential or Parliamentary?
GA: My sense is that there’s a broad consensus leaning toward a parliamentary system. Parliamentary systems are generally better at power-sharing and supporting political fluidity. The risk with a presidential system is the emergence of strongman politics—even in well-established democracies.
IB: What are the pros and cons of states having their own armies versus establishing a single national military?
GA: In the long run, the goal must be to have one national army. You simply can’t have multiple military forces loyal to different ethnic or regional authorities. How you get to that point is extremely complex and needs the input of security-sector specialists. But the end goal must be the disarmament of independent armed groups and their integration into a unified military.
Before the 2021 military coup, I visited Myanmar several times. It was a country with so much promise and potential. The coup has been a tragedy, but I support your struggle and am willing to help however I can. You face tremendous challenges, but you’re on the right side of history. Your country has enormous potential if it can transition to a better place.
Igor Blazevic is a lecturer at the Educational Initiatives Myanmar and Senior Advisor at the Prague Civil Society Centre.
George Anderson became president of the Forum of Federations in June 2005. He was previously Deputy Minister of Natural Resources Canada from 2002-2005.
DVB publishes a diversity of opinions that does not reflect DVB editorial policy. We’d like to hear what you think about this or any of our stories: [email protected]
