Appendix II. Methodological Notes to Chapter IV

International Monetary Fund. African Dept.
Published Date:
April 2007
  • ShareShare
Show Summary Details

The analysis uses UN Comtrade data on SSA merchandise (physical goods) exports, captured as imports to the reporting country. The reporting countries are the United States; the pre-2004 European Union (referred to as EU15); China; a group called Industrial Asia (Hong Kong SAR, Japan, Korea, and Singapore); and a group called Developing Asia (India, Indonesia, Malaysia, Thailand, and Taiwan, Province of China). Occasionally India is a sufficiently large component of the last group to warrant its being broken out separately. While these groups do not include all of SSA’s trade partners, they capture the main reporting partners for SSA in Comtrade. Two years are chosen for detailed comparison: 1985 and 2004, or 2005 when data are available for the latter. The data cover merchandise (including manufacturing) exports from SSA to each of the countries or groupings and are expressed in nominal US dollars.

On the SSA side, the analytical subgroups are resource-intensive (oil and non-oil), coastal, and landlocked countries. The listing is provided in the Statistical Appendix. Collier (2006) explains why the division is analytically useful. Collier and O’Connell (2006) explain the classification of countries by group. For reasons explained below, the non-oil resource-rich classification is of limited usefulness with Comtrade data. The income classifications, drawn from the World Bank, are based on gross national income per capita.

In Sections B and C, the definition of SSA corresponds to that of the World Bank and therefore includes Sudan, Mauritania, and Djibouti. These countries, which are not covered by the IMF’s Africa Department, are not included in the subgroups used in Section D. To maintain consistency of aggregation between the subgroups and the total for all countries, this section therefore excludes the three countries from the SSA aggregate calculation.

While comprehensive in many respects, the Comtrade data have some important limitations. One of the most important for the analysis of SSA subgroups is that SACU countries other than South Africa have only been reporting separate data to Comtrade since 2000. These countries are included in several SSA subgroups; for instance, all are middle-income countries, and two (Botswana and Namibia) are important non-oil resource-rich countries. This impedes the ability to analyze trends by comparing years before and after 2000; such comparisons are therefore not reported for the non-oil resource-rich group.

Comtrade data also have some reporting gaps. China does not report data for many product categories for 1985, and no Indian data are yet available for 2005. Furthermore, India does not report fuel imports to Comtrade. While the values involved in each gap are relatively small, they should be borne in mind in interpreting the tables and charts of the chapter.

The four broad sectoral classifications are formed from single-digit SITC categories: food and beverage (SITC 0, 1, and 4), crude materials (SITC 2), fuels (SITC 3), and manufactures and chemicals (SITC 5–8).

    Other Resources Citing This Publication