Issues of data quality complicate the task of properly tagging capital expenditure on infrastructure, and constrain the range and quality of analytics. Some general considerations include the following:
(1) Only general government expenditures were included; that is, capital spending by state-owned enterprises (SOEs) was rarely captured. This has minimal impact on the overall amounts, given the limited role played by SOEs in the African context.
(2) Of-budget spending was also excluded from total amounts. In some countries, off-budget spending accounts for a significant share of total expenditure-although to a lesser degree for capital expenditures. Hence, total amounts are likely to be underestimated.
(3) Only information captured in treasury systems was included in the analysis. Foreign funded infrastructure spending kept off-budget is not included and might affect total spending-especially for high-aid countries.
(4) In some Francophone countries, budget classifications are not fully integrated with the chart of account. This makes it difficult to ensure seamless tracking of transactions covering a typical Engagement/Liquidation/Ordonnancement/Paiement (ELOP) expenditure chain.45
Additional country-specific caveats include the following:
• Actual data for 2015 were not available for a few countries at the time of writing this report; data for 2014 were used as proxy. This includes Burkina Faso, São Tomé and Príncipe, and Ethiopia.
• The fiscal year rarely coincides with the calendar year in several African countries. Therefore, 2015 refers to the fiscal year 2014/15.
• Several countries do not properly report sources of funding, especially distinguishing capital investments funded through foreign aid. Hence, they are excluded from the analysis of foreign versus domestic funding sources.
• Equatorial Guinea and Zimbabwe provided minimal information at the administrative and economic levels, which did not allow for proper identification of infrastructure spending. Hence, they were excluded from the analysis.
• Guinea-Bissau's treasury systems fail to capture execution amounts of capital expenditures; hence, only approved amounts were used in the analysis.
• Several Francophone African countries often accumulate spending by multiple ministries into one administrative unit ("depense communes"), which does not allow for proper functional identification. This was the case in Guinea and Mali. It potentially leads to underestimation of spending.
• Foreign funded capital expenditure was only available at the budget level for Togo, Benin, and Guinea. Other countries such as Uganda and Mauritania among others also presented some gaps. Hence, deviation analysis was only conducted for domestic funded levels.
_____________________________________________________________________________________________
45 For instance, Mali uses bridge tables to link accounting data with budgetary operations, effectively retrieving execution data from their systems that are fully consistent with budget nomenclature Others, like Togo, Mauritania, and Niger, only present data at the committed and payment order stages but not actual payments.