This paper presents evidence suggesting that the relationship between income and economic structure is shifting over time, with countries across the income distribution uniformly increasing the share of labor in service sectors and an increasingly less stark relationship between manufacturing intensity and gross value added per capita.

The paper then assesses historical patterns of productivity convergence at a more detailed sector disaggregation than has been previously available. The analysis finds suggestive evidence that, at least in recent decades, convergent pressures in services industries are stronger than in manufacturing. Focusing on African economies, the paper presents a country-by-country historical analysis of structural change over the past four decades.

Given the varied patterns and trends in structural change across African countries, it is difficult to characterize structural change from a single, continent-wide perspective. Some countries saw an early transition of labor out of agriculture, with manufacturing absorbing this labor in the decades prior to the 1990s, while another group of countries saw a later transition out of agriculture, where the services sector played a large role in labor reallocations in the 1990s and 2000s.

Finally, the paper provides a country-by-country structural transformation scorecard to assess patterns of structural change in jobs and growth.

---

Introduction

A new and a more optimistic narrative on Africa’s growth and structural transformation has emerged from a series of high level reports – the 2014 African Transformation Report, the African Union’s Agenda 2063, the African Development Bank’s Long Term Strategy, the UN Economic Commission for Africa’s Economic Report on Africa 2013, UNCTAD’s 2012 report on structural transformation, multiple recent IMF papers, among others, which point to the fact that Africa’s recent progress is not based solely on natural resources. A recent report by the World Bank on economic transformation and poverty reduction in Africa found that the region’s economies are developing in unexpected ways. The region is largely “bypassing industrialization as a major driver of growth and jobs”. The demographic dividend will make the region’s labor force much larger and better than that of any nation, including China or India, if the young children in Africa could be better educated than in the past.

Africa’s recent economic performance has vastly improved: its annual growth rate approached 3% in per capita terms after 2000. Rodrik (2014) claims this growth is not just driven by investment, but also by total factor productivity growth – seen for the first time since the 1970s. While there was growth during the 2000s, the economic decline prior to this decade was so profound that many countries have yet to catch up to post-independence income levels.

The historical empirical reality shows that with the exception of the European periphery and East Asia, convergence has been the exception rather than the norm. Growth theory has adapted to empirical realities by differentiating between conditional and unconditional convergence: growth in developing countries is dependent on overcoming several obstacles or country characteristics, such as weak institutions, poor infrastructure, disadvantageous geography and inadequate economic policies. Consequently, Rodrik (2014) concludes that developing countries converge to rich country income levels conditional on these problems being overcome.

An ongoing debate is whether improvements in growth fundamentals can be expected to improve economic stability. Multiple studies have shown the relationship between good policy and economic growth is not particularly strong, though Acemoglu, Gallego, and Robinson (2014) suggest that differences in institution quality account for 75% of the variation in income levels around the world. This latter work points to a long run effect in levels, rather than an insight into short- or medium-term growth rates.

It is now well-established that the modern organized manufacturing sector exhibits unconditional convergence, contrary to other economic sectors; that is, manufacturing industries converge to the global productivity frontier regardless of the condition of growth fundamentals. The question is whether Africa can generate a “growth miracle” based on the manufacturing industry. In a model that incorporates this unconditional convergence, structural change – the shift of labor from low productivity industries to high productivity industries – plays an essential role.

The level of employment shares across sectors in developing countries conditional on income level shows that African countries fit into a global pattern: shares of employment in the different sectors are what one would expect them to be. So just like elsewhere, structural change in sub-Saharan Africa has been through a decline in the share of labor employment in agriculture, the least productive sector in sub-Saharan economies. Unlike other developing countries, African countries (generally) have not seen a significant increase in the share of labor force employment in manufacturing – instead, the shift in employment share has been towards the services sector.

Thus African structural change is unlike the pattern exhibited by Asian and European industrializers: labor is currently largely shifting towards the services industry, which historically has been less productive than manufacturing in other countries. Additionally, the manufacturing sector is dominated by informal firms, which are less productive than the formal organized firms required for unconditional convergence. Rodrik (2014) notes that African countries are de-industrializing much faster than Asian and European countries did in the past. African countries thus need to assess what type of growth model is both suitable and feasible.

A Historical Lens

Using a recently available data set from the Groeningen Growth and Development Centre’s 10-sector database (hereafter “GGDC data”) that provides employment and value added for ten disaggregated product sectors for a large set of countries historically, we begin the paper with high-level analyses investigating cross-country trends in sectoral productivity convergence and how the relationship between levels of development and economic structure has changed over time.

The GGDC data “provides a long-run, internationally comparable dataset on sectoral productivity performance in Africa, Asia, and Latin America. Variables covered in the data set are annual series of value added, output deflators, and persons employed for 10 broad sectors”. These ten sectors comprise agriculture, mining, utilities, construction, manufacturing, wholesale and retail trade, FIRE industries, transport, government services, and other services (restaurants, hospitality, etc.). Among Sub-Saharan African countries, it contains data from Ethiopia, Botswana, Ghana, Kenya, Malawi, Mauritius, Nigeria, Senegal, South Africa, Tanzania and Zambia.

These data allow a more nuanced look at the role of specific sectors in structural change, particularly separating construction and mining from manufacturing aggregates, and FIRE industries, government services, transport, and utilities from services aggregates, allowing greater specificity in understanding which sectors are drivers of structural change. Many of the countries in the recent release have reliable data going back to the 1970s or earlier, so we are able to build a long historical quantitative view of structural change. This, in turn, allows us to place economies’ current and recent trends in comparative historical context.