Morten Jerven's Poor Numbers is an entertaining and topical read, predating the ‘Data Revolution’ – that is, the current drive to improve data quality and transparency – in development statistics. The book provides an ethnography of national income accounting in Africa, assessing the history, underlying theory and process of production of gross domestic product (GDP) indicators. The book is organized into four chapters, following three main themes. Poor Numbers (1) gives an impression of the extent of data inaccuracy for poor African countries, and shows (2) why this matters and (3) what can be done about it.
The limited reliability of statistics for most poor African economies is generally accepted. However, Poor Numbers demonstrates how severe these limitations are for even such a commonly used indicator as GDP, exposing the dangers of glossing over data problems – an all too frequent occurrence in quantitative economic analysis. Scrutinizing three main sources of GDP data – the World Development Indicators, the Penn World Tables and the Angus Maddison dataset – Jerven shows in Chapter 1 that levels of per capita income for African economies vary strongly across data sources. As a consequence, relative country rankings – when ordered from richest to poorest – become meaningless for one-fifth of African economies.
The book's persuasive power lies in Jerven's ability to place abstract numbers into both historical perspective and local context, as demonstrated in Chapter 2. The author discusses those aspects of African statistics typically disregarded by their users: statistical offices in poor countries are often sparsely equipped and under pressure to generate comprehensive and regular economic indicators. This has been especially problematic during economic crises when budgets are cut. The presence of a large informal sector – which is notoriously difficult to monitor – makes this task even more daunting.
An important argument of the book is that statistics are not simply standardized numbers gathered in a rule-based process, but rather convey power. Data can be used to push for certain economic policies and to prevent others. Therefore, economic figures and their gathering process can be highly politicized, as discussed in Chapter 3.
It is typically not fraudulent figures that are the problem but the inadequacy or absence of data, paired with limited data literacy in developing societies. Therefore, in the final chapter, Jerven proposes changes to data collection and usage in poor economies. Given the bold message of Chapter 3, the policy recommendations seem timid, mostly focusing on technical aspects of data gathering: poor countries require greater resources for the collection of statistics. Help from donors is needed, but must not distort data collection with ad hoc projects and bias towards data coverage. Instead, reliable, frequently conducted (small-scale) surveys based on local knowledge are necessary. Crucially, the book advocates a qualitative analysis of statistics in which economists should be guided by the knowledge of historians and other social scientists.
Economic Growth and Measurement Reconsidered follows up on Poor Numbers, providing detailed histories of growth and their measures for Botswana, Kenya, Tanzania and Zambia. The book goes into the nitty-gritty of statistics and growth histories. As a result, it might be less entertaining to a general readership but more informative to development and Africa scholars.
Jerven challenges the popular stereotype that African development has been characterized by stagnation. He questions the conventional narrative that Botswana and Kenya were free-market success stories – with ‘good’ and ‘open’ policies – while Tanzania and Zambia represented failed socialist experiments – implementing ‘bad’ and ‘closed’ policies. The categorizations ‘open’ and ‘closed’ refer to trade regimes, broadly distinguishing between liberal and protectionist policies. The book discusses the mainstream interpretation of growth in sub-Saharan Africa and these ‘good’ and ‘bad’ policies in the first chapter. It then moves on to assess GDP data and their reliability in Africa in general and in the four countries in particular, before discussing the GDP statistics for each individual country in detail. Finally, Chapter 8 offers a comparative perspective on growth in the four economies. Jerven finds that Botswana's early growth is most probably overestimated, that there is not that much difference between Tanzania's and Kenya's performance, and that Zambia's poor growth was mainly driven by declining copper prices.
While, in both books, Jerven provides convincing evidence for why GDP data in poor African countries is often deficient, he does not explore the crucial issue of to what extent GDP is an adequate measure of development. It was Joseph Schumpeter, as well as the structuralists around Raúl Prebisch, who distinguished between growth and development. While the former is simply doing more of the same, the latter is the actual challenge for societies – how to combine resources in a new way, and how to innovate. The concluding remarks of Poor Numbers point towards these pressing questions when Jerven speaks of ‘conventional GDP measures’ (p. 113) and briefly touches upon the Millennium Development Goals. All in all, however, Jerven does not demonstrate to his readers how they are misled by African development statistics but by African growth statistics.
The presumption that growth and development are interchangeable phenomena relies on the trickle-down effect. In development economics, the belief in trickle down implicitly underlies a whole body of quantitative research: growth regressions. These attempt to identify drivers of growth in search of viable development policies. However, the assumption that growth automatically benefits the poor has been severely challenged. As Thomas Piketty has shown in Capital in the Twenty-First Century, inequality has been rising alongside (modest) GDP growth in high-income countries, eroding the welfare of a large share of households. It is becoming increasingly clear that even a well-constructed and precisely measured GDP figure is a poor indicator of people's living standards. While Poor Numbers might be forgiven for excluding these discussions, so as to powerfully depict poor data quality, this narrow focus is less understandable for the follow-up publication. Jerven's work would have benefited from a (short) discussion of GDP growth as a measure of development, especially since the author himself claims to have discussed the ‘progress and widespread economic growth following independence’ (p. 191) in the four countries. Note that here ‘progress’ is put first.
One of the chosen case studies could have been the starting point for such a reflection: Botswana. Anyone familiar with Botswana – regularly praised as an impressive growth success – understands that growth and development can be very different phenomena. Botswana experienced the longest and strongest growth episode in recent world history. According to the Spence Commission's Growth Report, Botswana grew by at least 7 per cent annually on average between 1960 and 2005, increasing its per capita GDP by a factor of eighteen over the period. Despite this growth, Botswana has one of the highest HIV/AIDS prevalence rates in the world, suffers from high income inequality, and struggles to diversify its economy in the face of depleting diamond deposits, on which the country is dependent for tax revenue and foreign exchange. Hence, even though challenging GDP data is a useful exercise, questioning growth regressions needs to go beyond the data, calling into doubt the theoretical foundations upon which these regressions are conducted.
Nevertheless, both books are insightful contributions to current debates in development policy. Poor Numbers has been a huge success well beyond specialist circles and scholars will surely greet Economic Growth and Measurement Reconsidered with similar enthusiasm. In fact, both books should be required reading for anyone dealing with development statistics.
