After robust sensitivity analyses and reviews by experts, the team that developed ARII composed the index of 16 indicators that were chosen to reflect the state of regional integration in Africa and efforts to integrate the continent further. The indicators are grouped into 5 dimensions

Sources of data

ARII’s 16 indicators are drawn from national and regional data, United Nations agencies, the African Development Bank, the African Union, the African Airlines Association, the World Bank, and other sources. To avoid double-counting, ARII uses indicators that are not linked to each other. See the Methodological Note for details about sources of data and years and countries of coverage.

Data limitations

Indicators were included where good data was available for a majority of countries. In a few cases where data was missing, ARII used the most similar datasets or an average of scores from a comparable country. See the Methodological Note for details about the treatment of missing data.

The weighting system

Although all of the indicators used in ARII are relevant to measuring regional integration, their influence may vary. Using an equal weighting system – a system where all indicators weigh the same – can overweigh or underweigh some indicators, producing biased scores. But using expert judgements to assign weights among indicators can be costly. It can also add subjectivity to the process. 

To counter this problem, ARII 2019 uses principal components analysis (PCA), a statistical methodology that allows robust computation of weights while maintaining objectivity. PCA derives weights based on the structure of the data. It also preserves variations in the data. ARII uses PCA to determine the weights of both indicators and dimensions. More specifically, ARII uses a two-stage weighting procedure. First, it assigns a weight to each of the indicators within each dimension. Second, it assigns weights to the dimensions. See the 2019 Methodological Note for details about ARII’s use of PCA.

The scaling system

From the rate of inflation differential to the ratification of the African Continental Free Trade Area (AfCFTA) agreement, ARII is made up of different indicators measured in different units. A common scale is required to aggregate the indicators into a composite index and apply PCA. For that reason, ARII used simple min-max rescaling procedure to normalise the indicators to range between 0 and 1, where 0 denotes the lowest level of integration and 1 denotes the highest level of integration. See the 2019 Methodological Note for details about ARII’s normalisation of the data.

Why countries sometimes have different scores

Some countries in Africa are members of more than one regional economic community. Because of this, a country may have different scores and rankings on the same dimension. For example, Libya scored 0.462 on trade integration within COMESA, where it ranked ninth out of 21 countries, but it only scored 0.390 within AMU, where it ranked fourth out of 5 countries. 

Some differences in scores and rankings can be explained by historical links, comparative advantages, and topography. Regional policies also play a role. If, for example, a country imposes visa restrictions on the countries that are members of the first regional economic community to which it belongs, but not on countries that are members of the second regional economic community to which it belongs, its score on the free movement of people dimension may be lower within the first community than within the second.

Control centre

ARII 2019 was built in six steps:

1. Selection of indicators
2. Standardisation and weighting of the data
3. Calculation of scores for each dimension
4. Calculation of scores for each country
5. Calculation of scores for each regional economic community
6. Calculation of scores for the continent

The Index aims to be an accessible, comprehensive, practical and results-focused regional integration tool that focuses on the policy level and on-the-ground realities.

You are currently offline. Some pages or content may fail to load.