Is Africa much richer than we think? No one knows

(CNN) – What do we know about poverty, income and growth in sub-Saharan Africa? The answer is: much less than we like to think. The data is unreliable and potentially seriously misleading. The question is of great importance. Economic growth rates or per-capita income estimates are commonly used in statements about development in Africa.
If income and growth statistics in Africa do not mean anything, a great part of current development analysis and policy targets are similarly meaningless. So far, debates on whether Africa is rising or not have failed to ask the question whether the growth data is telling us what we would like to think.
The research I conducted for my book “Poor Numbers: How We Are Misled by African Development Statistics and What to Do about It” clearly shows that any evaluation of whether Africa’s rise is real or not must begin and end with a careful evaluation of the growth and income evidence. Without such analysis, one runs the risk of reporting statistical fiction.
In November 2010, Ghana Statistical Services, the official provider of statistics in Ghana announced that it was revising its GDP estimates upwards by over 60%, suggesting that economic activities worth about $13 billion had been missed in previous estimates. After the revision a range of new activities were accounted for, and as a result Ghana was suddenly upgraded from a low-income country to a lower-middle-income country.
In the fall of 2011 Nigeria also announced an upward revision of its GDP. This revision is not yet complete, but it has been suggested that revision will cause a similarly large jump. If GDP doubles in Nigeria following the revision it will mean that the GDP for sub-Saharan Africa will increase by more than 15%. The value of the increase amounts to as much as 40 economies roughly the size of Malawi’s.
These revisions, which were caused by changes in the base year, methods and basic data of the national accounts, have resulted in confusion and disbelief in the development community.
If we know so little about growth and income in Ghana, one of the best studied economies in Africa, how shall we interpret the data from other African economies? In response to these emerging uncertainties, Shanta Devarajan, the World Bank’s Chief Economist for Africa, has declared the current state of affairs “Africa’s statistical tragedy.”
In “Poor Numbers” I show that these well publicized statistical events are only the tip of the iceberg.
These big jumps in the income statistics for Nigeria and Ghana mean that a large amount of economic activity has previously gone missing in the years since the 1990s, and it thus becomes guesswork to write or rewrite the economic history of Nigeria and Ghana based on the official statistics. It also means that any ranking of African economies today according to GDP or similar national income derivatives are meaningless.
In the research I conducted I surveyed methods and data in use in national statistical offices in sub-Saharan Africa. I could only obtain such information for 34 countries and of these, 21 reported having a base year that is within the last decade (i.e., 2001 or more recent), while 13 countries have base years from the 1980s and 1990s. That means a lot of such revisions may be forthcoming, and that many African economies are currently much richer than we think.
In “Poor Numbers” I document how these data problems mislead us. First of all the data misinform analysts and academics. This is first and foremost a knowledge problem. Academics get led astray when they argue about the causes for slow growth in some places and rapid place in others, or when they suggest fundamental reasons for why some countries are richer than others.
If the numbers they base such analysis on are unreliable, the conclusions they draw are flawed. One such example is that for many countries, such as Tanzania, and Zambia, because of upward revisions in the data series, the growth evidence artificially overstates the positive growth effect from liberalization policies in the 1980s and 1990s.
In turn, such erroneous scholarly findings may inform policy, which may result in very costly policy mistakes. Sometimes the data feed directly into policy decisions.
Funding bodies, such as the World Bank, and donor countries may make decisions to allocate countries based on their income status. This means that when Ghana was classified as a low-income country, it was eligible for concessional lending; now it is not. More fundamentally, these macroeconomic data inform policy makers whether projects and policies are working or not, and decisions to discontinue, reform or intensify efforts will be made on such basis.
The numbers also enter the policy domain, such as in decisions whether to fund fertilizer subsidies or not. The debate over the merits use of fertilizer subsidies has centered on the case of Malawi and the government’s decision to break with the IMF and the World Bank by re-introducing fertilizer subsidies.
According to Jeffrey Sachs, writing in the New York Times, President Bingu wa Mutharika of Malawi “broke old donor-led shibboleths by establishing new government programs to get fertilizer and high-yield seeds to impoverished peasant farmers who could not afford these inputs. Farm yields soared once nitrogen got back into the depleted soils”.
According to statistics published by the Ministry of Agriculture the maize was 3.4 million tons, in 2006/7. However, a long-repressed agricultural census (2006/2007) indicated a maize output of 2.1 million tons that same year. If the previously circulated numbers would have been correct, it would have meant that either huge stockpiles of maize accumulated around the country or a significant portion of the population was getting fat, neither of which was happening.
What did happen was that donor pressure and political pressure to show numbers that prove the success of the fertilizer voucher program pushed the numbers upwards and upwards.
The demand for numbers is overwhelming. The data will be created, invented or otherwise imputed to fill the gap. When the data basis is meager and the aggregates rely on a lot of guessing and assumptions the statisticians have little defense and are vulnerable to pressure. This is not only a matter of costly waste of donor money. Discrepancies such as those found here can be vital differences in countries where large parts of the populations are living in, or are close to living in, absolute poverty.
The main lesson is that numbers do inform scholarly conclusions, impact donor decisions and inform policy choice. In most cases data users do not know or choose to ignore that they are being misled by development statistics.
A new agenda for better data for development is sorely needed, because poor numbers are too important to be dismissed as just that.