On paper, federations generally seem like a good idea. A federal state can pool the resources of what would otherwise be separate economies to provide public goods more efficiently, which is what states are there for, at least from an economist’s perspective. Public goods, to expand Samuelson’s 1954 definition a little bit,1 are those which can be enjoyed in common, either because they can be consumed by each individual without detracting from anyone else’s consumption (e.g. national defence or clean air), or because one individual’s consumption of them makes everyone else better off through positive externalities, even if not everyone can consume that good or use that service at the same time (e.g. highways or public, free-at-the-point-of-access healthcare).2
A federal state is efficient at providing those public goods because it benefits from economies of scale, as fixed and initial costs to set up the bureaucracy and infrastructure needed can be very high and so the marginal cost of expanding the provision tends to decrease with scale (think about the cost of creating a judicial system, and how adding every additional judge afterwards is cheaper once you have the system up and running). Large federal states (NB: not big federal governments necessarily!) also appeal to economists because they provide the legal and political framework for market integration: without internal tariffs for commodities or restrictions to the movement of factors, the allocation of capital and labour should be more efficient, encouraging specialization in the comparative advantage of each province (because, as Adam Smith said and Brexit will perhaps prove once again, the extent of the division of labour is limited by market size).
Federations are also machines of spatial redistribution. They redistribute fiscal resources generated in one province3 either individually to another province (via federal subsidies or transfers) or to all of them (when using those resources to support federal-level public goods, as in defence or monetary policy, from which the rich provinces get less utility relative to the share they’ve paid for). Federations redistribute policy responsibility spatially as well, when they decide which level of government should organize the provision of the public goods in question.
The problem with federations, then, is that sharing is difficult, as most kids learn early on (and a particularly difficult lesson that is when, like me, they don’t have siblings). By sharing I mean really sharing—economic solidarity as opposed to charity: giving a non-negligible part of your income, enough to make you less well-off significantly, to someone poorer. When I was a kid for some reason the inner joy brought about by the warm glow of giving wasn’t enticing enough: if, through no fault of my own, I happened to get better gifts than my cousin, then why should I compromise and lend him my shiny new Batmobile for a whole weekend? To put in terms of today’s World Cup match: if you’re 1900s Buenos Aires and happen to have the best seaport in the country and a very productive export agriculture, why should you not be allowed to keep most of Argentina’s tariff revenue? And if oil reserves are concentrated in the Niger delta, why should the majority of oil revenue be spent elsewhere in Nigeria? Also, many economists would say, making very productive regions pay disproportionally (via increased taxes) for the costs of a federation could distort incentives to invest in them, thereby affecting the regional engines of national economic growth.
Belle époque Argentina (c. 1870-1913) and post-civil war Nigeria (since 1970) faced these problems, compounded by challenging population geographies. By 1869, half of the Argentine population lived far from the littoral areas, where most of the capacity for raising public revenue was concentrated. Buenos Aires and Entre Ríos had been home for decades to an extremely productive livestock export economy, and Santa Fe became the centre of a rapidly growing grain agriculture, which would eventually make Argentina the world’s largest exporter of corn and linseed, and one of the three largest wheat producers in the world. The inner provinces, in the west (the region known as Cuyo) and the north, were largely disconnected from the agricultural export-led boom, but they were home to comparatively large populations, which reflected the preceding history of the River Plate in the colonial period, when the gravitational pull of the economy came from the Potosí mines to the north-west (already by 1809 more than half of the ‘Argentine’ population was in the north). Looking at this pattern, and comparing it to Australia’s population geography concentrated in the most productive states (New South Wales and Victoria), Gerchunoff and Torre have argued that the Argentine population was in the wrong place.4 Until 1890, each provincial government was responsible for raising its own taxes, whilst the federal government kept all the tariff revenue. The classic problem of instability of fiscal revenue based on taxing foreign trade led the federal government to create some new internal taxes, superimposed on provincial ones, but tariffs still accounted for three quarters of revenues to the federal coffers, and fiscal federalism remained very limited. Inflows of capital and labour (in the form of hundreds of thousands of mostly southern Europeans) were therefore concentrated in Buenos Aires (both the city and the surrounding province), which became home to almost half of Argentina’s population by 1914, whereas the west and north’s combined share dropped to under 30%.
In 1970s Nigeria, the demographically largest regions were geographically, and to a large extent, ethnically separated from the comparatively more sparsely populated delta states which held the oil reserves ultimately responsible for more than 80% of national fiscal revenue. Until 1970, fiscal federalism in Nigeria had been limited by the derivation principle established during British rule as early as 1946, under which allocation of fiscal resources benefited almost solely the region where revenue was generated, which greatly benefitted the south-west in the 1950s during its cocoa export boom. Derivation was progressively abandoned in favour of the Distributable Pool Account (DPA) as the allocation principle over the late 1960s, and once the (perhaps related?) civil war was over in 1970, the military regime reduced the share of allocation based on derivation to 50%, which meant that half of the oil-fuelled DPA was allocated to states on the bases of their population.5 In 1971, the federal government itself started collecting oil receipts, of which only 20% were devolved to the oil-producing states. During the oil boom the mid-west Niger delta was still the region with more fiscal expenditure per capita, but its share was almost eight times smaller than it would have been under the derivation principle. Furthermore, as Olowononi argued, the fact that states depended so significantly on oil revenues which reached them only through the federally administered DPA weakened federalism in a deeper sense, creating a fiscal landscape more resembling a unitary system.6
Let’s add just a few numbers to the comparison. If we divide each country intro four large regions, and define a very simple per capita fiscal share indicator, dividing each region’s share of total national fiscal expenditure over that region’s share of total national population, Nigeria’s 1970s fiscal federalism seems to have been much stronger than 1910s Argentina’s, as all major regions had at least 90% of the fiscal revenue that would correspond to them on the basis of their population sizes (which is the single best indicator of the scale of the public goods they had to provide). Argentina’s north, on the other hand, got only about two-thirds of what it would have received on a per capita basis.
|Region||Pop share||Public spending share||Spending share/Pop share|
|Buenos Aires, Argentina||48.2||58.9||1.22|
|Rest of litoral, Argentina||22.1||18.2||0.82|
|Cuyo (West), Argentina||6.8||7.1||1.04|
NB: for Nigeria population shares for 1963 and fiscal shares for 1977; for Argentina population shares for 1914 and fiscal shares for 1916. Sources: Nigeria: Population Census of Nigeria, 1963, and Mbanefoh and Egwaikhide (1998). Argentina: Censo nacional, 1914, and Porto (1990).
Economies are also polities, and sharing tends to be good politics: Argentina and Nigeria both endured civil war after their independence, and in both cases many contemporary observers doubted that these federal countries would remain united as such. An imperfect but still consequential fiscal federalism possibly played a part in that. But sharing can also be good economics. First, it can be an effective antidote to the changing fortunes of economic cycles: provinces which once were net recipients of federal monies can later on become net contributors (such as Western Australia after the mining boom). Second, and more importantly, since economic development is ultimately about improving people’s material welfare, the strongest economic argument for intra-federal fiscal transfers is that they have the potential to create greatest happiness for the greatest number, even if they are not Pareto-efficient. The decision to share or not to share fiscal resources and economic prosperity also translates into population geography: if people can’t get richer where they are, they will try to go to a place where they can.7 When fiscal federalism suffers, either by policy choice or because the fiscal muscle for intra-federation transfers is not as strong, large-scale internal migration often occurs. In 1930s Argentina, when the agricultural export economy stagnated in the 1930s and imports (and therefore the tariffs providing most of the fiscal revenue) contracted further crippling the already very limited fiscal federalism, hundreds of thousands of internal migrants moved towards the capital, creating the greater Buenos Aires as we now know it. In so doing the also gave Peronism a social base, but that is a topic for another matchday.
His definitions could generally do with expanding, and this one isn’t an exception. P. A. Samuelson, “The Pure Theory of Public Expenditure,” The review of Economics and Statistics (1954).[return]
It’s a shame the US didn’t make it to the World Cup, if nothing else because it would have provided us with many opportunities to discuss the economics of healthcare, but I guess given the current state of the NHS we might get to it as England progresses to the next round.[return]
By “province” I mean the second-level government structure below the national state in a federal country (known as states in most cases, e.g. Nigeria, Mexico, the US, Australia, Brazil, etc.). In Argentina they’re called provinces, which I find more convenient as it allows me to reserve the word “state” for the national state.[return]
P. Gerchunoff and I. Torre, “¿Estaba la población argentina en el lugar equivocado? Un enfoque de economía política sobre las migraciones (1880-1914),” Desarrollo económico (2014).[return]
The distributional logic of the DPA under Gowon and beyond also encouraged pressures for the creation of more states, as each state was guaranteed a minimum of oil-derived fiscal transfers regardless of its population. T. Forrest, Politics and Economic Development in Nigeria (Boulder: Westview Press, 1995), 134[return]
G. D. Olowononi, “Revenue allocation and economics of federalism,” in Federalism and Political Restructuring in Nigeria, ed. Kunle Amuwo, et al (Ibadan Nigeria: IFRA, 1998).[return]
This is also true for the international economy, as Milanovic argued in one of our (I’m sure Tom agrees) favourite papers ever: B. Milanovic, “Global Income Inequality in Numbers: in History and Now,” Global policy 4, no. 2 (2013).[return]