Small assemblies in non-independent territories

I am going to do a little crowd-sourcing here. What do people think is a reasonable way to define “autonomous enough” to include a territory in a set of small assemblies worthy of comparative analysis alongside independent nations with small assemblies?

That is, there are various countries with very small assemblies that are recognized as independent states, such as St. Kitts and Nevis (assembly of 11 seats) or Antigua & Barbuda (17). Like the two just mentioned, most of the small-assembly independent states that are also democracies with small assemblies are in one world region (Caribbean) and use one type of electoral system (FPTP).

Now suppose one wanted to branch out and include small territories that were either not in the Caribbean or used PR. Suppose further that one did not want to include obviously fully dependent territories that just happen to hold elections for an internal legislative council. Where would one draw the line?

For instance, are the Faroe Islands and Greenland “autonomous enough” to include? What about Aruba and Curaçao? These use PR systems, and the first two are not Caribbean. Or the Cook Islands, with is FPTP but non-Caribbean?

One would need a reasonable standard for autonomy. I sort of feel the places I just named might qualify, but I do not know why I feel that way. And I do not want the can of worms opened whereby I’d be asked–legitimately–why did you exclude Turks and Caicos (for example)? (Other than, well, I already had enough FPTP Caribbean cases.)

The smallest currently included independent country in my related datasets seems to be St. Kitts & Nevis (pop 54k). One of the territories I mentioned is much smaller than that (Cook Islands only 15k), but others are of the same order as St. Kitts (like Faroe Islands and Greenland, 53-55k). I probably have a floor somewhere on population–which might well exclude Cook Islands–but my current query is for a reasonable standard on what is sufficiently self-governing to be comparable to small independent states for purposes of analyzing their assemblies and electoral systems.

What do readers of this site think?

31 thoughts on “Small assemblies in non-independent territories

  1. I’m not certain how wide a net you want to cast, but “worthy of comparative analysis” certainly describes Scotland, Bavaria, and Baden-Württemberg.
    Curaçao and Aruba, sure. Until the most recent election, Hong Kong’s system was certainly interesting, and likely Macau’s. French Polynesia’s looks quite interesting, and includes gender parity. Jersey has just changed its electoral system, so why not? As for the Isle of Man, in 1881 its Parliament, Tynwald, became the first national legislative body in the world to give women the right to vote in a general election (although perhaps this does not count since this excluded married women).


  2. Thank you, Wilf. That is precisely the issue–I do not know how wide a net to cast! I’d probably err on the side of wide unless I discover a good reason to trim further.

    However, I should have made clear that this will not include state/provincial electoral systems. It is cross-national, so the question is really where do we draw the line between “dependency that is close enough to being an independent nation” and “dependency that is more like a state or province.”

    It also will be only simple electoral systems, so while I agree that HK and others mentioned are interesting, they will be outside the purview. That is, just single-tier, single-round, with party/list vote pooling within districts if there’s more than one candidate per party or list per district. (I remain conflicted about whether free-list systems belong in the set, but open list clearly do, by the criteria as defined. I just wrote a separate post about that.)

    Also, for the moment, not including presidential systems, but at least some semi-presidential systems likely will stay in.


  3. Pingback: Is free-list PR a “simple” electoral system? | Fruits and Votes

  4. Given some actual countries today are highly limited in their real independence (think of EU member states and many Pacific Island nations) without their party system being part of a larger whole (unlike most federal units). and given that you are not studying something like federalism or autonomy where the degree of autonomy might be important to the dependent or independent variable, I think you have good rationale for casting a pretty wide net here. The one thing I would alert you to thinking about is whether your sample might be biased by excluding (as I expect you will) political systems dominated by independents rather than political parties, which is relatively common among this particular set of jurisdictions (Channel Islands, Falkland Islands come to mind). Something to think about.


    • I agree with all that regarding how to think through what “autonomy” means, thanks.

      On the independents, yes, I also have given this some thought. We need to exclude such cases because the effective number of parties is thereby rendered quite meaningless. But that rule applies regardless of country size, and results in the exclusion of Papua New Guinea and Ukraine*, too. I don’t think this introduces bias with respect to the specific proposal of addition of a few small dependent territories with “national” party systems, although if anyone wants to make the argument that it does, I am willing to rethink.

      (*When Ukraine used nationwide closed-list PR, the issue of independents magically went away! And those two elections are in the existing dataset, but I am considering taking them out because it was just 2006 and 2007, a short interlude between more independent-friendly rules (MMM). Besides, it is a system with a strong presidency, and we will be focused mostly on parliamentary systems, plus the odd assembly-independent system.)


  5. I guess I would be inclined to exclude places which have representation in the national legislature of the sovereign country, which would allow you to include the Cook Islands, Aruba and Curacao but exclude the Faroes and Greenland (and afaik all French overseas territories). I fear this might create an Anglophone bias in your database, and I don’t know how well theoretically grounded it is, but it seems like the only rule I can think of that lets you draw a clean line.


    • Thanks. Yes, I thought of this as a possible exclusion rule. But I could not convince myself that such representation externally would undermine the claims that the territory’s own party system was a product of its assembly electoral system. More important, I think, is that the “metropolitan” parties not compete in the territory or not dominate the competition.

      And I am indeed attempting to avoid Anglophone bias, which is already very strong in the dataset for one particular class of electoral system!

      I was planning to exclude the French overseas departments because the classification as “department” implies lack of autonomy. However, this may be a poor criterion, as maybe that does not mean much in practice. Of course, if their representation in the sovereign metropolitan legislature were a criterion, they would be out (as would the two Danish territories as you note).


      • AFAIK New Caledonia and French Polynesia have far more autonomy than metropolitan ‘départements’ but most other French overseas territories do not.


      • That sounds correct, and I have considered including at least New Caledonia (if its electoral system meets the other criteria for inclusion). But I still do not know how to decide what territories are “autonomous enough.”

        For whatever it may be worth, I have found that data for several of the territories mentioned in the original planting are available in CLEA, but no French territory is included. So I might just go with what they have collected, which means I don’t have to defend any additional ones I selected on my own. (If I do that, no Cook Islands, I am afraid.)


      • Actually, I did not have to look very far (a few paragraphs into a Wikipedia article) to be warned off New Caledonia, as its electoral system, and voting eligibility seem pretty non-simple.

        At least after 2007, French Polynesia uses a two-round system, which definitely puts it outside the set I am interested in for current purposes. Data on earlier elections appear sparse. I won’t be expending effort to locate more.


      • As regards to what Henry said, I was under the impression that the Cook Islands and Niue are both more freely associated states while Tokelau is a dependent territory.

        I don’t have a clear reason for saying so, but my gut would be to say that Aruba, Sint Maarten, and Curacao would be “autonomous enough” while Bonaire, Sint Eustatius, and Saba aren’t.

        I would ask whether indigenous nations in North America should also be considered. They have a considerable degree of autonomy which has only increased since the McGirt decision.


    • Besides the British “Crown Dependencies” and “Overseas Territories”, US “non-incorporated territories” and the very complex case of West Bank/Gaza Strip, there is non-autonomous territory in the world without at least the pretense of representation in the national parliament of the sovereign country (at least, for countries with parliaments)?


      • I would add the Cook Islands, Niue and Tokelau to your list but, yes, it is not a very long one. I’d also add the US territories which have non-voting representatives

        I guess I’m inclined to use this list as a starting point, to which other regions can be added, in order to limit the degree to which legal and practical autonomy needs to be considered. Autonomy is a sticky enough concept at the best of times but in this case I think the concept Professor Shugart is looking for is detachment from the metropolitan party system, which is downstream from autonomy and needs further theoretical justification.


      • I had not even thought of the Palestinian Legislative Council. Given how much blog space I spent on that 2006 election, and that I have a (currently dormant) paper specifically on it electoral system, perhaps it should have been on my mind. It certainly meets one of my core case-selection criteria–the party system can plausibly be said to be “domestic” and not a function of the other state to which the territory is somehow affiliated. But it has a directly elected non-ceremonial presidency and a complex electoral system. Therefore it would be out anyway. I also am not generally inclined to include cases with a complete breakdown of democracy, especially if it took place after only one election. (So please don’t ask why one election from Somalia in the 1960s is already in the dataset! (I may drop it now.))

        The US territories are also excluded as having presidential systems (and in some cases non-simple electoral systems).


    • Interesting!

      In any case, I had ruled Turks & Caicos out of the sample because it uses multiseat plurality, a non-simple electoral system.

      In other case-selection news, I was going to include Anguilla till I discovered that my data source has only 7 constituencies for the one election it includes in a way that appeared useable, but there are actually 11. So that is a pretty substantial percentage of districts missing!

      I also have less than a crying need for more British or formerly British Caribbean (or Caribbean-adjacent) island states/territories.


  6. Pingback: The political system of Guyana | Fruits and Votes

  7. On further thought, I think it may be pretty defensible to include cases where the unit in question has a completely different party system from the main part of the country, especially when that unit has really tiny representation in the national legislature. I’m thinking of Greenland, but there are other examples.


  8. Pingback: St Kitts and Nevis 2000–Crazy result | Fruits and Votes

  9. The non-partisan legislatures in Nunavut and the North West Territories may be worth a look? Nunavut in particular may be interesting given the bilingualism with Inuktitut (their first Speaker was a unilingual

    I was wondering if you’re following the Yukon electoral reform process at all? The minority government agreed to examine the electoral system as part of their confidence and supply agreement with the NDP. I don’t have high hopes for it but we’ll see. Seems relevant to this discussion too.

    The Yukon seems to be a major outlier from your seats-product rule too (19 seats and they’ve had a pretty robust 3 party system for a while). 19 seats seems like a big deviation from the cube root rule too (for a pop of 42,000 it should be 35).


    • The Yukon deviation is not out of line with many other sub-national assemblies. In a discussion of Canadian elections and the SPM in October 2018 I noted:

      “Although Canadian sub-national assemblies are too small according to the cube root rule, they are too small in a very systematic way. Using current values for assembly sizes, I find assemblies of the 13 provinces and territories coming in at 0.55 of the cube root value, with a standard deviation 0.05. Other former British colonies also have undersized assemblies in states/territories: Australia 0.48 (SD 0.08); India 0.45 (0.12); United States excluding NH 0.65 (0.27).”


      • For Ontario, the cube root is 244. Using your 0.55 rule, we would have 134. We actually have 124, which is the 121 federal constituencies plus 2 exceptional northern ones and one grandfathered northern one. This is a potential issue in Ontario where the official opposition NDP has a policy of adopting MMP which might usefully include additional MPPs. One recalls the New Zealand slogan “I’d rather live in a democracy with 120 MPs than a dictatorship with 99.”


    • I am vaguely aware of the electoral reform discussions in Yukon, but know very little about what is going on.

      I would not consider territorial assemblies for the project needs that prompted this thread, as they are clearly not autonomous, plus in the specific case of Yukon, the party system includes the parties of the wider federal system. It is still of interest to me, but not as a case for this specific research. I definitely can’t include nonpartisan assemblies, as it is obviously impossible to calculate the effective number of parties in such cases!

      It is not unusual for subnational assemblies to be both smaller than the cube root law (which is explicitly modeled for national assemblies) or to have party systems that deviate from the seat product model. It would be very interesting to explore such a dataset for testing the limits of the predictions, and perhaps fine-tuning them; on this general theme, there is the earlier post that David refers to on Canadian assemblies. I do have a paper currently underway on local councils. It is a wide open research agenda (and one I do not expect to get very far in to myself). (Come to think of it, I probably should make a planting all about that paper!)

      Also, please: “seat product model” not “seats-product rule.” It is a set of quantitatively predictive logical models, not just an empirical rule. And “cube root law” not “rule” because it has a logical basis and likewise not only an empirical tendency. In Votes from Seats we refer to the individual predictive formulas for key outcomes (effective number of parties, largest seat share, etc.) as “laws” in much the same way as the cube-root law of assembly size.


  10. Ryan,
    I haven’t looked at other cases at the level of state/province/territory.

    At a sub-sub-national level, I did look at 18 “cities” in Alberta of populations between 10,000 and 1,200,000. The scale factor for a cube root law came out as 0.23 (SD 0.06). There may be a framing effect: the Municipal Government Act sets the minimum size as 5-7 councillors and all but Edmonton and Calgary have 7 or 9 (including the mayor) on council.


    • That is an important thing to keep in mind: Municipal council sizes are often set, or limited, by law at a higher level of government. The cube root law can’t override national, state, or provincial law!


      • This point was brought home forcefully when I looked yesterday at state/province assembly sizes in Germany, Brazil and South Africa. The slopes of log-log plots were 0.35 for the German states, 0.39 for Brazil, but 0.48 for South African provinces.

        It turns out that the South African constitution sets the size of provincial legislatures at 1 member peer 100,000 inhabitants, but also sets a floor at 30 members and a ceiling at 80 members. Consequently, the log-log plot has a sigmoidal shape: South African provinces cannot be used in any study of the cube root law.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.