Patterns of Budgetary Execution: Programmed, Revised, and Executed
In Cape Verde, as in many developing countries, the execution rate of recurrent expenditures is much higher than that of capital expenditures. Recurrent expenditure execution rates typically are more than 90 percent of the originally approved budget. However, the best implementation rate for capital expenditure during 2002–06 was 81 percent in 2006. This difference stems from the already discussed characteristics of the budgetary execution process: (i) nondiscretionary expenditures determine the implementation of recurrent expenditures; and (ii) the fluctuating flow of foreign aid determines the implementation of the investment program.
Table 2.14: Actual Expenditure as % of Budgeted Expenditure,1 2002–06
In the category of recurrent expenditures, the implementation rate of nondiscretionary expenditures (salaries and benefits, transfers and subsidies, and interest payments) is typically high. For example, in the subcategory salaries and benefitsexpenditure, the outlay is usually over 90 percent of that originally programmed. In the subcategory interest payments, this rate exceeded 100 percent in 2004, 2005, and 2006. The subcategory transfers and subsidies had an execution rate above 90 percent in 2002–06, and in 2005 and 2006. It dropped to approximately 80 percent (as noted above, the payment of some subsidies is not included in the budget). This high implementation rate of nondiscretionary expenditure often is obtained at the expense of discretionary expenditure such as goods and services. In the category of goods and services, the expenditure execution rate versus planned budget varied from 69 percent in 2002 to a peak of 92 percent in 2005, and then down to 70 percent in 2006.
Programming and implementation of the Public Investment Program (PIP)
PIP programming has been over optimistic (table 2.10). The highest execution rate during the period of analysis was 81 percent in 2006, and it had fallen as low as 59 percent (2003). The average execution rate during the period was 69 percent. This variable performance is due primarily to the fact that PIP implementation depends to a large extent on foreign aid materialization. Table 2.11 shows that, during the 5-year analysis, more than 80 percent of the investment was funded by external resources. In 2004 this rate climbed to 95.2 percent. Whenever the disbursement of external funds is delayed or does not materialize, implementation may be suspended or even cancelled.
Some projects, mostly the ones with important social dimensions, have been implemented, even when foreign aid does not materialize or is insufficient. A wide range of secondary investments and small maintenance tasks have been implemented through domestic funding. Similarly, when a project to be funded with external resources has priority but lacks significant endowment, the Treasury mobilizes funds to bridge the financing gap over a short period, that is, until foreign aid arrives. In this manner, it is possible to continue project implementation. However, in the case of very large investments such as infrastructure projects, implementation is delayed or remains incomplete.
Table 2.15: Public Investment Program (PIP) - Execution and funding sources, 2002–07
Source: Ministry of Finance and Public Administration.
Notes: 1 Preliminary accounts.
2 Approved budget.
Variability in foreign aid funds is rooted in two main factors: (i) the nature of aid (project versus budgetary aid); and (ii) delays related to donor or recipient political or policy implementation processes. In the first place, as is typical in many developing countries, most foreign aid to Cape Verde is channeled toward project financing. This type of aid is prone to delays whenever the bureaucracies of the donor and the recipient differ with respect to public procurement, financial management, or other requirements related to domestic regulation. Furthermore, this type of aid generates project coordination units that often operate outside national budgetary processes, thus making the budgetary unification process difficult. Second, foreign donors frequently must abide by their own national objectives or policies, which are not always aligned with the objectives of the recipient country.
It is expected, however, that aid predictability will improve, as budget support has gained prominence during the last years, thanks to the creation of the Budget Support Group. Created in 2005, the BSG included the World Bank, the European Union, and the Dutch cooperation.15 The coordinated approach used by the BSG to align and harmonize their support for poverty reduction around the GPRSP-1 has played a catalytic role in bringing in new partners. As a result, BSG was expanded in 2007 to a total of 6 partners when the African Development Bank (AfDB) and Austria, and Spain cooperations joined the group. The BSG expanded further when the Portuguese cooperation announced at the GAT meeting (Grupo de Apoio à Transição) in December 2007 that it will join too. During the last 3 years, approximately 25 percent of foreign aid was in the form of budget support.
Deficiencies in monitoring project implementation for projects implemented through direct aid makes difficult in-depth and efficient analysis of investment expenditure execution. The deficiencies in monitoring project implementation financed with external sources are due, to some extent, to lack of capacity. Capacity building in monitoring is a priority. Furthermore, it also is important that development partners collaborate by making data available at the appropriate times.
Recommendations Given that recourse to large amounts of credit is not sustainable, it is recommended that the Cape Verde authorities implement a set of actions to guarantee a more reliable inflow of foreign aid:
Prepare a pluri-annual framework with the donors as part of the MTEF process.
Encourage partners to make directly available updated financial data of funded projects to establish transparency in the presentation of quarterly and annual state accounts.