What a world assembly could look like

Abstract

I identify some basic institutional features that can make a world assembly viable in terms of size and degree of complexity. The most potentially satisfactory model is a bicameral assembly formed by a lower chamber with about 2000 seats (of which about one-fourth would be elected in single-member districts and about three-fourths in multimember districts of proportional representation with moderate magnitude) and an upper chamber of territorial representation based on about 700 territorial units all across the world (basically corresponding to the nearly 200 currently independent states plus about 500 regional governments).

Introduction

Numerous authors and global activists have proposed cosmopolitan democracy. Skeptics have retorted that democracy defined according to traditional attributes at the state level may not be feasible at the global level in a very large and complex world. Some scholars have postulated that, as long as democracy may require direct elections of representatives in a world assembly by broad suffrage, it may be “unviable” on a grand scale (for discussion, see Archibugi, 2008; Dahl, 1999; Falk and Strauss, 2001; Held, 1995; Keohane and Nye, 2000; Schwartzberg, 2012). ¹

This presumption is at least dubious on empirical grounds. For instance, the parliament of India is directly elected within a multiethnic federation comprising over 1.2 billion people, three dozen states and territories and two dozen official and scheduled languages, while the European Parliament is directly elected within a union comprising more than 500 million people, more than two dozen states and more than 20 official languages; levels of participation are similar to those in midterm elections for the United States Congress.

In order to assess the viability of a directly elected global assembly, it may be helpful to try to imagine what its basic features would be. Given that numerous institutional designs are conceivable, I will try to identify some institutional features that could be viable in terms of size and degree of complexity. In this piece, I specifically discuss the size and the basic electoral rules of an elected global assembly. I consider two models on empirical grounds by projecting from the relations among institutional formulas that have been observed as viable and durable at state and federation levels. The most viable and potentially satisfactory model seems to be a bicameral assembly formed by a lower chamber with about 2000 seats (of which about one-fourth would be elected in single-member districts and about three-fourths in multimember districts of proportional representation with moderate magnitude) and an upper chamber of territorial representation based on about 700 territorial units all across the world (basically corresponding to the nearly 200 currently independent states plus about 500 regional governments).

Size of the assembly

Firstly, an important decision is the number of seats that such a global assembly should gather. The best approximation to calculate the size of a directly elected lower chamber of the assembly in a democratic country is to take the cube root of the population. For most countries the number of inhabitants amounts to millions—that is, some figures with six zeroes—and so the cube root must be in the hundreds, or some figure with two zeroes. For example, as Spain has about 45,000,000 inhabitants, the cube root of this number closely approaches the 350 seats of its lower chamber of parliament (45,000,000^1/3 = 355). For most democratic countries, this is the best fit.

The relation is lower for very small countries, such as some islands in the Caribbean or the Pacific Ocean, as well as for regional or local legislatures. By analogy, it could be speculated that a very large global assembly would need, in contrast, a relatively larger proportion of seats. For the purposes of this piece, however, I work with the cube root relation.

For a world that will approach 8 billion people by 2025, a round number would be an assembly with 2000 seats (as 8,000,000,000^1/3 = 2000). This is indeed a large size. No democratic assembly approaches this number, as the largest ones, such as the European Parliament and the British House of Commons, are below 1000. The closest and highest number is the China National People’s Congress, which is formed of nearly 3000 seats. This, however, is a figure-head assembly in a non-democratic country, which meets for only two weeks a year and has almost no power regarding the ruling party. A global assembly of 2000 members would allocate one seat to every 4 million people, on average, which is not very much higher than the 2 million people and the three-fourths of a million in the current lower chambers of India and the United States, respectively. Given these assemblies, we may assume that such a global large assembly with a high number of specialized committees and a reduced standing committee could do the job. ²

Two-chamber electoral rules

How this could this assembly be elected? There is a relation between the number of territorial governments in large federations and the number of assembly seats elected in every district. An extreme case is the United States, which has the largest number of territorial governments in any country (50 states) and the smallest magnitude within electoral districts (just one seat). All the other countries have lower values for territorial governments (as low as one for unitary countries) and equal or higher values for electoral districts (as most democracies use rules of proportional representation). Just to mention a relatively extreme case on the other side: the Netherlands, which is a unitary country with no regional decentralization, elects all the 150 seats of its House of Representatives in a single, very large nationwide district.

Durable democratic regimes are based on a trade-off between these two variables—the number of territorial governments and the average magnitude of the electoral district, with values depending on the size of the country. For every country’s size, the lower the number of territorial governments, the larger the average magnitude of the electoral district and vice versa.

The logic of this relation is that a varied population can be aggregated into a single polity both on a territorial basis and through pluralistic politics permitted by large electoral districts. A large number of territorial political units in a federal structure can be the basis for a large, aggregative “union”, while an assembly based on proportional representation and multipartism can also be aggregative because it can lead to the formation of a broad government multiparty “coalition”. Both “union” and “coalition” can keep a large, diverse country together by using democratic means of governance. Yet in the quantitative trade-off between the values of the variables for the two institutional mechanisms, the number of territorial governments has more weight than the number of seats per district, especially for large countries.

For this relation to make sense at a global level, it must be assumed that there would be an upper chamber of territorial representation based on recognized territorial governments. We can thus construct a hypothetical institutional design for a global assembly with two chambers: a lower chamber with about 2000 seats based on worldwide distributed electoral districts and an upper chamber with delegates of recognized territorial governments. At least two possibilities would exist.

Firstly, if the recognized territorial units of the world were currently existing independent countries, as they are represented in the United Nations—approximately 200—then according to the empirical trade-off observable in federal and decentralized countries, we would estimate the average electoral district of the lower chamber should have about 25 seats. Yet 2000 seats in districts with about 25 seats each would mean that the world would be divided into 80 electoral districts, a number considerably lower than the number of independent countries. Most countries would have to share electoral districts with other countries. The lower chamber would thus have far fewer electoral districts than the upper chamber of territorial representation would have basic units. This arrangement may not favor adequate representation.

An alternative would be to consider as recognized territorial governments not only the nearly 200 countries that have representation in the General Assembly of the United Nations, but also the nearly 500 regional legislatures that exist in federal or decentralized countries (including, for example, the 50 states in the US, the 36 states and union territories in India, the 26 states in Brazil, the 16 laender in Germany, and so on). This results in the much higher number of about 700 territorial governments. According to the above-postulated empirical relationship, the average magnitude of the electoral district would therefore be much smaller: only three seats. For 2000 seats there would be 667 electoral districts, which is very close to the number of territorial units in the upper chamber. Essentially every state and region would work both as an electoral district for the lower chamber and as a unit for territorial representation in the upper chamber.

A possible apportionment for the lower chamber could imply, for example, approximately 500 districts based on all non-independent regions and smaller independent states could elect one seat each, while about 150 larger districts representing most independent states could elect close to10 seats each on average, with room for variance depending on the size of the country (>500 + 150 ≈ 667; 550 × 1 + 150 × 10 ≈ 2000). These values may also mirror the current variety of electoral systems at the country level, for while a few old, large democratic countries of mostly Anglo tradition use single-member districts, most current democracies use rules of proportional representation, the average district magnitude for all democracies being at about 11.

According to the well-established quantitative relationship between assembly size, electoral district magnitude and number of parties for state-based assemblies known as “the seat-product”, such an assembly would elect about nine parties. The effective number of parties, which weights the absolute number of parties by their relative size, would be a little higher than four. These values are not very distant from the averages in current democratic regimes, although it can be presumed that worldwide party federations would be looser than the typical state-based parties.

Conclusion

This note has explored some basic institutional features of a directly elected world assembly, in contrast to previous exercises that focused mostly on the design of an indirectly elected world parliamentary assembly formed by state parliamentarians. All in all, the basic values for a worldwide assembly could be:

• lower chamber size: approximately 2000 seats;

• lower chamber average district magnitude: approximately three seats (about 550 districts with one seat each and about 150 districts with multiple seats);

• number of electoral districts for the lower chamber and number of territorial units in the upper chamber: approximately 700;

• expected number of parties in the lower chamber: about nine;

• effective number of parties in the lower chamber: about four.

The above analysis suggests relatively moderate values for both the size of the assembly’s two chambers and for the size of the electoral districts, which are of a similar order of magnitude to assemblies and electoral districts in the largest democratic countries, federations and unions. These findings may support the viability of a directly elected global assembly in the current world. Future work should deal with the apportionment of the seats of the lower chamber across districts by applying existing formulas at state and federation level. Other issues, such as the powers of such an assembly, its internal decision rules, or the process for its establishment, rest beyond the scope of the present endeavor.