The straight horizontal line across Nigeria at latitude 7˚10' divided it into two colonies, Northern and Southern, in 1899
In colonial Nigeria in the last years of the 19th century, a strange quirk of history led the British rulers to draw an arbitrary boundary line along the 7˚10′ N line of latitude, separating the population into two separate administrative districts.
Below the line, the colonial government raised money by levying taxes on imported alcohol and other goods that came through Southern Protectorate’s sea ports. Above the line, the administrators of the landlocked Northern Protectorate had no sea ports, and instead raised money through direct taxes. In the areas near the border, this took the form of a simple poll tax, where tax officials collected from each citizen the equivalent of between $4 and $20 in today’s dollars.
Could this seemingly minor difference—created over a century ago by a long-defunct colonial administration, and long ago erased by subsequent administrative divisions—possibly still matter today?
Yes, it could, according to Daniel Berger, a PhD student in politics at NYU. Berger’s paper, Taxes, Institutions and Local Governance: Evidence from a Natural Experiment in Colonial Nigeria, finds that the “simple act of having to collect taxes caused governments to be forced to build the capacity which can now provide basic government services.” As a result, governance today is “significantly better” in areas just above the line than in those just below it.
After looking at historical evidence and running statistical tests, Berger finds that there is no evidence of pre-existing differences among the people living very close to the arbitrary boundary on either side, and so is able to rule out the possibility that some third factor could account for the differences in governance that remain today.
The results are threefold. Berger uses Afrobarometer public opinion data to show that residents just above the line are happier with their local governments, and his use of demographic survey data shows that local governments just below the line spend 10 percent more of their budget on salaries (“an indicator of less competent government.”) Zeroing in on the propensity of mothers to vaccinate their child as a way to get at a precise measure of the quality of public service delivery, Berger finds that “living just below the line leads to a 10.7 percentage point reduction in the probability that her child will be vaccinated for polio.”
The conditions created by the administrative division led to two different equilibria, which help explain how the differences above and below the line were able to persist over time:
In the first, the local government does little except extract what few bribes it can….There is no incentive for hard work, as bureaucrats will neither be able to extract appreciably more rents (due to the limited amount of money available in the local economy) nor will they be able to improve government functioning on their own (since efficient functioning requires the entire bureaucracy working together). This also leads to a knock on effect on the human capital available to the local governments as the families which control the local government have no reason to steer their smartest children into local government service.
The second equilibrium is one in which significant services are actually delivered. Here, the local government is capable of delivering local basic public services with a reasonable level of efficiency and honesty. This grants sufficient legitimacy to the local government that they are able to collect local taxes, which never go to the center. They can then pay themselves regularly despite the fact that they are not regularly receiving the transfers they are due from the center. Here hard work does make a difference.
Berger’s conclusions also speak to the strength of norms and informal institutions. While the formal institutions—the idiosyncratic colonial structure of taxation—that created the original difference in bureaucratic capacity were long ago swept away, it is the informal norms, transmitted across generations, that persist and lead to the different outcomes we see today.
You may wonder what whim caused the British create this artificial boundary in the first place. The literature tells us that the British were worried that a colonial official senior enough to administer the whole undivided territory of Nigeria would be too old and too weak to survive the malarial climate. By cutting the province in two, the British could send two younger and heartier (but less-experienced) governors instead.
So, to simplify: measurable differences in the perception and quality of local government provision service persist between otherwise identical populations just north and south of Nigeria’s the 7˚10′ N latitude line…all for fear of a malarial mosquito.
13 Comments
This piece speaks in a uniquely illustrative – if perhaps unintentional way – to the fundamental and ultimately invariable attribute of aid: it is, by definition, non-local unearned income. As a paradigmatic contrast, taxation is local priced income, priced in that a population demands services in return for paying the government. Political scientists have long understood the very basic connection between local taxation and government accountability – something apparently lost on many economists.
I think that Jane Guyer (1992) in another essay on taxation and local governance in rural Nigeria summarizes this contrast most eloquently: “[l]arge aid flows result in a reduction in government accountability because governing elites no longer need to ensure the support of their publics and assent of their legislatures when they do not need to raise revenues from the local economy, as long as they keep the donors happy and willing to provide alternative sources of funding .”
Aid thus substitutes for the process of domestic resource mobilization; and when viewed from this perspective, the corrupting influence of aid in perverting basic links of accountability becomes much clearer.
the drama in the description of the two equilibria looks rather out of proportion to the 10pp difference in vaccination probabilities
@Ross, yes I see your point but your comment also makes me think I may have given some of the results short shrift in an effort to keep this post from being too long. In fact, the results are threefold.
1) Afrobarometer public opinion data shows that residents just above the line are happier with their local governments.
2) Demographic survey data shows that local governments just below the line spend 10 percent more of their budget on salaries (“an indicator of less competent government.”)
3) Vaccination data shows that living just below the line leads to a 10.7 percentage point reduction in the probability that a child will be vaccinated for polio.
I’ve added a sentence (para 6) to better reflect this.
What a superb finding! This Nigeria tax study illustrates that for development strategy to become sustainable and effective it should focus on growing the local tax base as a core objective.
Aid donor displacement of indigenous institutions creates aid dependency, including in the social sectors of health and education. Results that are bought through external help have high risk of failure when those resources are withdrawn. Local control, through taxation of a functioning formal monetary economy, provides the accountability for service delivery that is lost in donor dependence.
This study also illustrates a broader point regarding private sector development and poverty reduction. Aid could avoid creating dependency and become more sustainable and effective through a stronger focus on the binding constraints for the business enabling environment. The Nigerian data shows that taxation provides effective incentives for behaviour, with measurable results for poverty reduction. Rather than only working directly with the poor, aid would have better long term results by helping the local private sector to grow, so a country can fund social services through its own taxes. As well as paying tax, businesses provide jobs, customers, goods and services for the poor Work to improve the prospects for private firms has a strong multiplier effect for broad based development.
This governance finding regarding the centrality of taxation to development does not fit well in the MDGs, which seem to be more about delivering political targets than achieving sustainable results.
The quote below is an extract from a piece in the NYT. I am finding it hard to reconcile it with the rosy second equilibrium in the North and what is interpreted as better governance in the article.
” In the north, though northerners have ruled Nigeria for most of its post-colonial history, states like Zamfara and Katsina, with their grinding poverty and exhausted ecosystems, have few links to the global economy and offer their people little except an uncharacteristically harsh interpretation of Islam, manifested in the worst stereotypes of sharia, or Islamic law. Understandably, people who have been excluded from the future tend to grip tightly to what they are told are the certainties of the past. ”
http://www.nytimes.com/2003/12/30/opinion/30iht-edhenne_ed3_.html
A couple of points: first, taxation is important; I’ve blogged about this in the past:
http://aidthoughts.org/?p=591 .
The Christian Michelsen Institute has done a couple of interesting case studies about this too, linked in my blog post.
Secondly, it’s difficult to tell without seeing the actual d phil work but it sounds like it might be a piece of over-simplified history (not uncommon with economists and political scientists). Firstly, reducing these differences to one point, on taxation’s effect on governance is likely to be a gross over-simplification of what were very complex and involved historical processes going on in Nigeria in the colonial period; bear in mind that Nigeria was incredibly diverse in its history (state-wise) and taxation prior to that (slavery, tribute etc. – there were forms of taxation that were widely used). Hausaland and Yorubaland had state structures, Igboland was stateless. Within both Hausaland and Yorubaland there were periods of unity and periods of intense local differentiation. So its not surprising that when borders were drawn historical processes diverged. Don’t forget that many ‘tribal’ and regional identities only emerged in response to colonial impact and more specifically the borders it created. The border itself may have stimulated cultural differentiation beyond the effects of early policy on Governance.
Nigeria has, since colonialism, fractured from a nation of three states to one of more than fifty. Local variation is the norm and stems from various factors, few of which are ‘unusual’.
Secondly, how good is the data for early C20th Nigeria? Statistical tests will still rely on the quality of data from back then (or on assumptions made about historical development paths using data from now). Is the data robust enough to draw firm conclusions about C19th local differentiation?
I read this article and rolled my eyes. The reason is that it makes no sense to anyone who lives in Nigeria. I then read the paper and realised that his premise might be correct but only in a few states of the country that he looked. This colonial legacy argument makes no sense in the general context of Nigeria, but may make sense in the two middle belt states in which were on the border.
The problem with this post and the paper is that this very narrow conclusion about the border states (2 of 36 of the federation) with population less than 5% of the nation, can be generalised to the rest of Nigeria.
The problem with this generalization is that taking the north and the south of Nigeria and comparing you end up with a completely different view that invalidates these colonial impact results.
Local governments in the north are viewed as places where the the monthly allocations to the local government are shared between the workers. This has been observed by practically every youth corps (national service) member that has served in the north in recent times.
The second point is that polio pretty much only exists in the north and middle belt. This is because people in the north stopped taking polio vaccines. It was a political fight related to the implementation of sharia law a few years back.
However the most critical point is that most of the tax infrastructure that exists in Nigeria is in the south. Most of the tax collected in Nigeria is collected from the southern part of Nigeria. The VAT, PAYE and Company Tax receipts are evidence. It also turns out that the states with the highest tax receipts are those with the strongest institutions. However our author has decided that the tax realities 150 years ago are more important than the tax realities today.
Interesting, made me immediately think of Andrew Mwenda’s recent article, , which posits that taxation implies a negotiation between the govt and the taxed. With Aid, that negotiation doesn’t take place. I’m butchering the thesis, read the article.
Reductio ad absurdum. It takes more than a tax structure to generate proper services from the developing states. In my own country (Honduras) the issue with services, circles around the idea of a centralized government. Large amounts of Aid are channeled through the Central Government, which by the way generates contradicting incentives to contract local labour vs Central Government contractor.
We had a mayor of a city (Puerto Cortés) decentralize the aid and he got i.e. 3 miles of sidewalks for every mile built by a Central Government contractor. The idea of decentralizing governments should be one of the focuses of aid agencies as well as their own division of labour signed in Monterrey, Rome, Paris and Accra.
None of the initiatives is fully realized. It is more a pamphlet of good intentions.
“otherwise identical populations”
This is the most interesting phrase of the whole post, really. It’s just so interesting that in over a century, none of the people living a few miles south of the border realized how much happier they could make themselves and their children just by moving up the road a bit. Instead, they stayed put, generation after generation, and then complained about their plight to Afrobarometer and let their kids be deprived of vaccine. It’s a quaint and curious thing, how silly we humans can sometimes be.
Like Chinansa, I rolled my eyes @ this bit of “interesting information”.
As she said, the tax realities today certainly tell a different story.
Nevertheless, I’ll play devil’s advocate and offer a defense for the writer. The influence of resource allocations from the central government, the effect of which has been likened to aid flows by many competent development economists, might have skewed the observable results today.
Still, although I think using the arbitrary divide has problems (and ignores other contributory effects such as the British’s traditional preference for northerners; direct rule in the north and indirect rule in the south), the conclusion on taxes seems reasonable. The two states with the highest IGRs in Nigeria today (Lagos in the south and Sokoto in the North) are states with marginally better institutions.
However, I think this raises important questions about the causal relationship between taxes and good governance. Does good governance result in better and increased taxation or does taxation require good governance?
I hope the author clears this up in the final study
Anonymous @3.12:
“It’s just so interesting that in over a century, none of the people living a few miles south of the border realized how much happier they could make themselves and their children just by moving up the road a bit. Instead, they stayed put, generation after generation, and then complained about their plight to Afrobarometer and let their kids be deprived of vaccine. It’s a quaint and curious thing, how silly we humans can sometimes be.”
Unfortunate, yes, but silly? That would be a conclusion based on an assumption that people know and understand everything about the choices in front of them, and are able to make a choice.
As some others have suggested, I think the evidence is rather oversimplified, but the conclusion is certainly interesting and based on some sound theory.
I thought it was clear that Berger’s intention was not to generalize his findings to the whole of Nigeria, nor to make the overly reductive claim that tax structure is the only requirement for good government service delivery. If I inadvertently implied either of these things in my short summary of the research, this is my fault and not Daniel Berger’s. Please do not neglect to see the paper itself, linked to in the post above and again here: http://homepages.nyu.edu/~db1299/Nigeria.pdf
In criticizing the paper for failing to consider historical or current differences among populations in the north vs the south of Nigeria today, I think a couple of the commenters seem to really miss the point that the paper looked *only* at the populations just above and just below the temporary administrative boundary line.
People may be overly eager to generalize simplistic policy prescriptions from economics research (i.e. aha! If this paper shows a small effect of taxation on governance in a very specific area in Nigeria, we should redirect all development funds towards implementing poll taxes!). But this is not necessarily the fault of the research itself, and while any of you may agree or disagree with the paper’s methods or conclusions, I don’t think there’s anything wrong with the idea that this narrowly focused paper (which in fairness does rely on both statistical analysis and the historical record) could provide one small but valuable slice of evidence towards one of those fundamental–or “universal”– concepts that we can say tends to hold more often than not, i.e. that direct taxation can help build bureaucratic capacity, or that institutional norms can persist over time in surprising ways.
Note that Ranil Dissanayake over at Aid Thoughts has an interesting post on what he thinks are the limitations of this type of study. He uses my summary of Berger’s paper to illustrate his point that “economists tend too much towards isolation of specific issues (and this tends to lead to or benefit from reductionism).”
4 Trackbacks
[...] Instead we seek to apply the same reductionism to other disciplines. A great example is given in this Aid Watch post: It’s entitled History Matters, but the history it presents is remarkably reductionist, as [...]
[...] in Nigeria: 100 year-old decision creating a noticeable difference today. (Daniel Berger, via the AidWatch Blog) “Places where the British built local tax collection bureaucracies a century ago have [...]
[...] been made? Well, it appears that it is this sort of tax system which actually builds a functioning society: In colonial Nigeria in the last years of the 19th century, a strange quirk of history led the [...]
[...] happy tax story from Nigeria? Surely some mistake? Nope, but the tax dates from 1910 and has led to demonstrable improvements in [...]