Counting matters! Statistics are the backbone of proper planning

Africa, over the past 20 years, has moved from a position of Afro pessimism through Afro optimism to a state of Afro enthusiasm. In Africa, we all like it but we should also ask why the narrative has changed so drastically? Is the new narrative coming from Africa? Partly, yes. However, most of what we hear is still an outside view and for good reason. The one who controls the information controls the narrative.

Carlos Lopes, executive secretary of the UN Economic Commission for Africa.

Carlos Lopes, executive secretary of the UN Economic Commission for Africa.

As much as we like to hear about Africa’s growth figures we should be concerned about the quality of the data. Maybe the picture will be even better if we had good data but most likely less glamorous than it looks. The question is, can we verify the numbers behind telecoms, retail, banking, corruption, poverty and the like? What is the statistical basis for putting out these figures? And can they be corroborated by the facts on the ground?

In this 21st century and this age of technology and innovation there is no excuse for Africa to do things the same way it did it 50 years ago. The use of mobile phones, internet and new technology is growing at an incredible pace, part of moving the world into a data revolution.

In a recent article, Rick Rowden questioned the basis of Africa’s supposed growth figures. He argued that due to the unavailability and/or unreliability of data or what he refers to as “unhelpful indicators” it is difficult for economists and analysts to believe the sheer hype. Derek Blades, as far back as 1980, also suggested that African GDP level estimates had large errors and should be considered with a plus-minus margin. In 2011, when lead statisticians in 23 countries where asked the direct question: ‘Do you think that GDP rates are fine?’ only Namibia and Lesotho answered in the affirmative. The other statisticians were convinced that their country’s GDP were underestimated.

The dangers of using incorrect statistics

So what do we lose by our continuous affiliation to forecasting and making projections based on nothing more than outdated data and hunches? Is the proverbial garbage in, garbage out syndrome happening? There is the danger of wrong analysis leading to wrong or insufficient development of policy solutions that, instead of alleviating a dire situation, aggravates it. Let me elaborate the point.

Let us recall how GDP is calculated. To work out real GDP, government statisticians use the prices of goods and services from a “base year” as a reference. It is recommended that base years should be changed every fifth year but only seven out of 54 African countries regularly do this. In a study of 47 African countries by Morten Jerven, an economic historian, he found that only 10 African countries used base years less than a decade old. Up to seven countries still had base years in the 1980s and several others were unsure of their base years.

In 2010, Ghana revised its base year from 1996 to 2006 thereby shooting up its GDP by 60%. This resulted in the World Bank reclassifying it overnight from a poor to a middle income country. This meant in concrete terms that about US$13bn of economic activities had been systematically overlooked. In response, African development expert Todd Moss (2010) exclaimed: “Boy, we really don’t know anything.” Given this level of error margin in the GDP estimate on Ghana, arguably one of the most studied countries on the continent, what should we think about economic statistics deriving from many other African countries?

Nigeria is now planning to rebase its economy from its 1990 base year rate to 2008. “Rebasing” could see its economy swell from $273bn to $382bn, just behind South Africa at $420bn. Jerven’s analysis on the impact of this rebasing within the economy of sub-Saharan Africa leads to the conclusion that there are about 40 Malawis unaccounted for in the Nigerian economy to date.

But even with these drastic changes, can we bet on these figures?

Morten further compares the World Bank’s calculations of country-specific GDP with GDP figures calculated by national statistics offices and finds that only three country figures totally agree. In a few other cases there are discrepancies such as 476% for Ethiopia, or 10,432% for Sudan. The list of discrepancies and inaccuracies in the calculation of such important economic variables goes on.

In a second example, researchers from the London School of Economics argue that over the last 20 years, household consumption in Africa has actually been growing from 3.4% to 3.7% per annum, based not on the GDP calculations but on factors such as the number of households with TV sets, electricity access or percentage of people with mobile phones, compared to the 0.9%-11% suggested by income statistics.

Another example is the estimation that there are 316m new subscribers of mobile phones in Africa since 2000. How were there numbers calculated? Most likely with industry giving the figures of sale of handsets.

According to the Food and Agriculture Organisation (FAO), the quantity and quality of agricultural statistics coming from national statistical offices have been on a steady decline since the early 1980s, particularly in Africa, and official submissions from countries in Africa are at their lowest level since before 1961, with only one in four African countries reporting basic crop production data.

Take the Millennium Development Goals for example: only 17 African countries have collected data to measure changes in poverty in the past decade and 47% of African countries have not carried out a household income or expenditure survey in more than five years.

Finally, we have to acknowledge that the census and demographic statistics are left wanting. Without good knowledge of how many people live in certain areas and what their living conditions are like, it is difficult to give credibility to any statistics that finish with “per number of people”.

What to do?

Governments in Africa are increasingly strengthening their statistical departments in line with the trend of developing longer term development plans. In 2012, PARIS21 supported several African countries to design or implement a National Strategies for the Development of Statistics. Niger, Benin, Burundi, and Guinea benefited, with other countries to follow, but sadly, most national statistical machinery is still quite deficient. Much more needs to be done. The United Nations Economic Commission for Africa is ready to take the challenge now.

Joseph Schumpeter used to say: “We need statistics not only for explaining things, but also in order to know precisely what there is to explain.” How can Africa catch up with the current momentum and enhance its statistics capacity without being able to plan? And how can planning be meaningful without good and reliable statistics?

Africa needs to make use of technology in this new era of the ‘data revolution’. The increase in the use and access to new forms of technology is not limited to the developed world. The spread of mobile phones and internet in Africa can and should be harnessed for improving and reducing the cost of collecting statistics. It is estimated that while internet traffic will grow by more than 50% in Africa, Latin America and the Middle East, it will only grow 25%-30% in North America.

Also, the collection of data through the use of mobile devices can accelerate the rate of collection, analysis, interpretation and use of real time data to solve the continent’s problems. At the macro level, the use of technology in health and agriculture, and surveys to collect and analyse proper data needs to be encouraged.

Another important development is big data. The use of big data gives access to real time information and therefore makes it easy and quicker to analyse and respond to crisis and project trends in diverse areas such as weather, agriculture, health, population and so on.

There is, of course, debate on the use of big data due to privacy and the possibility of revealing personal confidential information while doing analysis. Moreover, most of the data generated through these means are perception as opposed to facts. Notwithstanding, there are some very good applications of big data that have turned out to be very useful in providing users with timely and quality policy and development information.

In conclusion, African leaders and policy makers must wake up to the incredible opportunity ahead and make use of its institutions, structures and systems that can support the churning out of credible data as a means to ably measure our developmental growth in a credible manner.

Carlos Lopes is executive secretary of the United Nations Economic Commission for Africa. The article was first published on his blog.