The statement ‘seven out of ten fastest growing economies are in Africa’ carries no real meaning. To utter it is merely stating that you subscribe to the hype August 26, 2014

‘Most of the time we simply do not know enough to assert accurate growth rates. There are also known biases and manipulations. Ethiopia, for example, is notable for having long-standing disagreements with the IMF regarding their growth rates. Whereas the official numbers have been quoted in double digits for the past decade, a thorough analysis suggested the actual growth rates were around 5 to 6 percent per annum. More generally, one study used satellite imaging of nighttime lights to calculate alternative growth rates, and found that authoritarian regimes overstate reported rates of growth by about 0.5 to 1.5 percentage points. Another recent study argues that inflation is systematically understated in African countries – which in turn means that growth and poverty reduction is overstated.’
http://africanarguments.org/2014/08/26/why-saying-seven-out-of-ten-fastest-growing-economies-are-in-africa-carries-no-real-meaning-by-morten-jerven/

Why saying ‘seven out of ten fastest growing economies are in Africa’ carries no real meaning

By Morten Jerven @ AfricanArguments
Before, during and after the US Africa summit one of the most frequently repeated factoids supporting the Africa Rising meme was that ‘seven out of ten fastest growing economies are in Africa.’ In reality this is both a far less accurate and much less impressive statistic than it sounds. More generally, narratives on African economic development tend to be loosely connected to facts, and instead are driven more by hype.

***

The ‘seven out of ten’ meme derives from a data exercise done in 2011 by The Economist. The exercise excluded countries with a population of less than 10 million and also the post-conflict booming Iraq and Afghanistan. This left 81 countries, 28 of them in Africa (more than 3 out of 10) and, if you take out the OECD countries from the sample, (which are unlikely to grow at more than 7 percent per annum), you find that every second economy in the sample is in Africa. It might not give the same rhetorical effect to say: ‘on average some African economies are expected to grow slightly faster than other non-OECD countries,’ but that would be more accurate.

And before we literally get ahead of ourselves (The Economist was reporting forecasts made for 2011 to 2015) there is a difference between forecasted and actually measured growth. According to John Kenneth Galbraith, the only function of economic forecasting is to make astrology look respectable. So how good is the IMF at forecasting growth in Low Income Countries?

According to their own evaluation, IMF forecasts “over-predicted GDP growth and under-predicted inflation.” Another study looked at the difference between the forecasts and the subsequent growth revisions in low income countries, and found that “output data revisions in low-income countries are, on average, larger than in other countries, and that they are much more optimistic.” Forecasts are systematically optimistic all over the world, but in Low Income Countries even more so.

***

Among those on the list of the fastest growers were countries like Nigeria, Ghana and Ethiopia. The news that both Nigerian and Ghanaian GDP doubled following the introduction of new benchmark years for estimating GDP in 2010 and 2014 should remind us that the pinpoint accuracy of these growth estimates is lacking. How confident should you be about a 7 percent growth rate when 50 percent of the economy is missing in the official baseline? Recent growth in countries with outdated base years is also overstated.

While Ghana has reportedly had the highest growth rates in the world over the past years, a peer review of the Ghana national accounts noted that “neither a national census of agriculture nor other surveys, such as a crop and live-stock survey, have been conducted…there is no survey to provide benchmark data for construction, domestic trade and services.” It was recently reported that an economic census is being planned for next year. What we do know is that Ghana (together with Zambia, another of the projected ‘top ten growers’) has returned to the IMF to seek assistance following their entry into international lending markets.

Most of the time we simply do not know enough to assert accurate growth rates. There are also known biases and manipulations. Ethiopia, for example, is notable for having long-standing disagreements with the IMF regarding their growth rates. Whereas the official numbers have been quoted in double digits for the past decade, a thorough analysis suggested the actual growth rates were around 5 to 6 percent per annum. More generally, one study used satellite imaging of nighttime lights to calculate alternative growth rates, and found that authoritarian regimes overstate reported rates of growth by about 0.5 to 1.5 percentage points. Another recent study argues that inflation is systematically understated in African countries – which in turn means that growth and poverty reduction is overstated.

***

Data bias is carried across from economic growth to other metrics. The pressure on scholars, journalists and other commentators to say something general about ‘Africa’ is relentless, and so the general rule is to oblige willingly. When talking about average trends in African politics and opinion, analysis is influence by the availability of survey data, such as Afrobarometer, and the data availability is biased. According to Kim Yi Donne, on The Washington Post’s ‘Monkey Cage’ blog, of the 15 African countries with the lowest Polity IV rankings, only seven have ever been included in the Afrobarometer, whereas all but one African country rated as a democracy by the same index is included.

Any quantitative study which says something about the relationship between growth and trends in inequality and poverty, relies on the availability of household survey data. One paper boldly stated that African Poverty is Falling…Much Faster than You Think! The data basis was very sparse and unevenly distributed. There were no data points for Angola, Congo, Comoros, Cape Verde, D.R. Congo, Eritrea, Equatorial Guinea, Seychelles, Togo, Sao Tome and Principe, Chad, Liberia, and Sudan. In addition, six countries only have one survey. The database included no observations since 2004 – so the trend in poverty was based entirely on conjecture. Famously you need at least two data points to draw a line. Yet the study included a graph of poverty lines in the Democratic Republic of Congo from 1970 to 2006 – based on zero data points.

A result of doubts about the accuracy of the official evidence, and a dearth of evidence on income distributions, scholars have turned to other measurements. Data on access to education and ownership of goods such as television sets from Demographic and Health Surveys were used to compile new asset indices. In turn, these data were used to proxy economic growth and in place of having a measure of the middle class. In both cases the data may paint a misleadingly positive picture. While claiming to describe all of Africa over the past two decades, these surveys are only available for some countries sometimes.

***

The statement ‘seven out of ten fastest growing economies are in Africa’ carries no real meaning. To utter it is merely stating that you subscribe to the hype. It is particularly frustrating, and it surely stands in way of objective evaluation, that the narratives in African Economic Development switches from one extreme to the other so swiftly. The truth lies somewhere between the ‘miracles’ and ‘tragedies’. It is nothing short of stunning that in a matter of 3-4 years the most famous phrase relating to African economies has turned from ‘Bottom Billion’ to ‘Africa Rising’.

Because of a lack of awareness on historical data on economic growth it was long claimed that Africa was suffering “a chronic failure of growth”, but growth is not new to the African economies, growth has been recurring. There is no doubt that there are more goods leaving and entering the African continent today than fifteen years ago. More roads and hotels are being built and more capital is flowing in and out of the African continent than before. But what is the real pace of economic growth? Does the increase in the volume of transaction result in a sustained increase in living standards? The evidence does not yet readily provide us with an answer. It is the job of scholars to give tempered assessments that navigate between what is make-believe and what passes as plausible evidence.

Morten Jerven is Associate Professor at the Simon Fraser University, School for International Studies. His book Poor Numbers: how we are misled by African development statistics and what to do about it is published by Cornell University Press. @MJerven

http://africanarguments.org/2014/08/26/why-saying-seven-out-of-ten-fastest-growing-economies-are-in-africa-carries-no-real-meaning-by-morten-jerven/

Related References:

Why Africa needs a data revolution

Since the term “data revolution” was introduced, there has been a flurry of activity to define, develop, and implement an agenda to transform the collection, use, and distribution of development statistics. That makes sense. Assessing the international community’s next development agenda, regardless of its details, will be impossible without accurate data.

Yet, in Sub-Saharan Africa – the region with the most potential for progress under the forthcoming Sustainable Development Goals – accurate data are severely lacking. From 1990 to 2009, only one Sub-Saharan country had data on all 12 indicators established in 2000 by the Millennium Development Goals. Indeed, of the 60 countries with complete vital statistics, not one is in Africa. While most African countries have likely experienced economic growth during the last decade, the accuracy of the data on which growth estimates are based – not to mention data on inflation, food production, education, and vaccination rates – remains far from adequate.

Inaccurate data can have serious consequences. Consider Nigeria’s experience earlier this year, when GDP rebasing showed that the economy was nearly 90% larger than previously thought. The distorted picture of Nigeria’s economy provided by the previous statistics likely led to misguided decisions regarding private investment, credit ratings, and taxation. Moreover, it meant that Nigeria was allocated more international aid than it merited – aid that could have gone to needier countries.

Contrary to popular belief, the constraints on the production and use of basic data stem not from a shortage of technical capacity and knowhow, but from underlying political and systemic challenges. For starters, national statistical offices often lack the institutional autonomy needed to protect the integrity of data, production of which thus tends to be influenced by political forces and special interest groups.

Poorly designed policies also undermine the accuracy of data. For example, governments and donors sometimes tie funding to self-reported measures, which creates incentives for recipients to over-report key data like vaccination or school-enrollment rates. Without effective oversight, these well-intentioned efforts to reward progress can go awry.

Despite these failings, national governments and international donors continue to devote far too few resources to ensuring the collection of adequate data. Only 2% of official development aid is earmarked for improving the quality of statistics – an amount wholly insufficient to assess accurately the impact of the other 98% of aid. And governments’ dependency on donors to fund and gather their core statistics is unsustainable.

In fact, stronger national statistical systems are the first step toward improving the accuracy, timeliness, and availability of the data that are essential to calculating almost any major economic or social-welfare indicator. These include statistics on births and deaths; growth and poverty; tax and trade; health, education, and safety; and land and the environment.

Developing such systems is an ambitious but achievable goal. All that is needed is a willingness to experiment with new approaches to collecting, using, and sharing data.

This is where the public comes in. If private firms, media, and civil-society organizations identify specific problems and call publicly for change, their governments will feel pressure to take the steps needed to produce accurate, unbiased data – for example, by enhancing the autonomy of national statistical offices or providing sufficient funds to hire more qualified personnel. While it may be tempting to bypass government and hope for an easy technology-based solution, sustainable, credible progress will be difficult without public-sector involvement.

The recognition by governments and external donors of the need for more – and more efficient – funding, particularly to national statistical systems, will be integral to such a shift. Establishing stronger incentives for agencies to produce good data – that is, data that are accurate, timely, relevant, and readily available – would also help, with clearly delineated metrics defining what qualifies as “good.” In fact, tying progress on those metrics to funding via pay-for-performance agreements could improve development outcomes considerably.

One concrete strategy to achieve these goals would be to create a country-donor compact for better data.

Read more @ http://forumblog.org/2014/08/africas-necessary-data-revolution/?utm_content=buffer4f4fd&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer