Kenya
Fiscal Transparency Evaluation
Author: International Monetary Fund. Fiscal Affairs Dept.
Publication Date:
12 Jul 2016
eISBN:
9781475538069
Language:
English
Keywords:
ISCR; CR; central government; balance sheet; annual budget; budget year; real GDP
Download PDF (2 MB)

This paper discusses key findings of the Fiscal Transparency Evaluation of Kenya. Since 2010, the Treasury has made important changes in Kenya’s public financial management framework, the impact of which can clearly be seen in its performance against the Fiscal Transparency Code. The prospects for quick improvements in the fiscal reporting area are clearly within the grasp of the National Treasury. Fiscal forecasting and budgeting practices are generally in line with good practice under the Fiscal Transparency Code, reflecting more than a decade of experience with medium-term budgeting and preparing and presenting macro frameworks.

Abstract

This paper discusses key findings of the Fiscal Transparency Evaluation of Kenya. Since 2010, the Treasury has made important changes in Kenya’s public financial management framework, the impact of which can clearly be seen in its performance against the Fiscal Transparency Code. The prospects for quick improvements in the fiscal reporting area are clearly within the grasp of the National Treasury. Fiscal forecasting and budgeting practices are generally in line with good practice under the Fiscal Transparency Code, reflecting more than a decade of experience with medium-term budgeting and preparing and presenting macro frameworks.

I. Fiscal Reporting

A. Introduction

1. Fiscal reports should provide a comprehensive, timely, reliable, comparable, and accessible summary of the government’s financial performance and position. To do so, all fiscal reports, notably budget execution reports, fiscal statistics, and government accounts, should:

  • Cover all institutional units engaged in fiscal activity;

  • Capture all assets, liabilities, revenue, expenditure, financing, and other flows;

  • Be published in a frequent and timely manner;

  • Be classified according to an internationally recognized classification system;

  • Reconcile any unexplained discrepancies within or between fiscal reports; and

  • Be prepared by an independent agency (in the case of statistics) or scrutinized by an independent national audit institution (in the case of accounts).

2. There are a range of deficiencies in Kenya’s reporting practices that limit the value of current fiscal reports, including:

  • The coverage of fiscal reporting is almost exclusively budgetary central government on a cash basis, excluding significant fiscal operations by off-budget units, and no balance sheet reporting.

  • The timeliness of reporting varies, from frequent and timely reports on the basic exchequer flows, to considerable delays in the more comprehensive budget outturns and audited annual statements leaving a vacuum of reliable information for policymakers.

  • The quality of reporting is diminished by the existence of different estimates of actual fiscal outcomes that are not internally inconsistent with verifiable information, such as the stock of debt. While the classification of reporting is good, there are large fiscal operations that remain outside of international standards.

  • The integrity of reporting is weak, with no consolidated financial statement and almost half of expenditure being subject to adverse or disclaimer audit opinions.

3. A range of institutional changes have been introduced to address some of these deficiencies, although their implementation will take time to complete. Notable amongst these planned changes, along with an indicative timeline, are:

  • Accounting standards – implementation of the decisions of the newly established Public Sector Accounting Standards Board - IPSAS for general government and IFRS for commercial state corporations. (3-5 years).

  • Financial statements – implementation of improved annual financial statements based on new reporting formats through training, functionality included in the re-engineered IFMIS. (2-3 years).

  • Consolidated statements – preparation of consolidated central government annual financial statements, including a financial position statement by Accountant General’s Department and their submission to the Auditor General in a timely manner. (2–4 years).

  • Classification structure – full implementation of the revised classification structure built into the re-engineered IFMIS for general government. (1–3 years).

  • IFMIS – completing the rollout of the re-engineered IFMIS to all budget entities. (2–4 years).

Table 1.1.

Kenya: List of Fiscal Reports

article image

1.1. Coverage of Fiscal Reports

1.1.1. Coverage of public sector institutions (Not Met)

4. Kenya’s public sector comprises 419 separate institutional entities. Table 1.2 provides an overall picture of the expenditure, revenues and flows of each subsector of the public sector and their institutions:

Table 1.2.

Kenya Public Sector Institutional Composition and Finances (2012–13)

(Percent of GDP)

article image
Source: Staff estimates derived from BPS, 2013-14 Budget Annex 2, IMF STA technical assistance report and company financial statements, and Water Service Sector annual report.

Shows the transactions with non-public sector, excluding transfer received and paid to other public sector units.

  • Budgetary central government funded through “the Exchequer” comprises the 19 government ministries and 16 other institutions who have their total revenue and expenditure authorized by Parliament through the annual Appropriation Act.

  • 48 Extra-budgetary funds, who receive transfers from the budget estimates and/or donors, as well as some own source revenues, and distribute these funds direct to their intended purpose, such as the Road Maintenance Fund, the Constituency Development Fund and the Water Services Trust Fund. In most cases, these funds carry balances across years, creating a difference between the transfer recorded in the budget as expenditure, and the actual expenditure undertaken by the funds.

  • 129 Non-Commercial Parastatals, or Semi-Autonomous Government Agencies (SAGAs) that perform direct policy actions for the government, and are primarily funded through budget transfers but also collect own source revenues. They include a wide range of entities, including authorities such as Kenya National Highways Authority, autonomous agencies such as Kenya Revenue Authority and Universities such as Kenyatta.

  • Sub-National Governments made up of 175 local authorities until 2012–13, replaced by 47 counties starting in 2013–14. These used to receive a roughly equal share of intergovernmental transfers and own-source revenue. However, under the new Constitution, their responsibilities were increased, accompanied by a substantial increase in intergovernmental transfers. They own 120 water service providing corporations that are primarily funded through own revenues.

  • 39 Commercial Parastatals in which the government owns a controlling stake or exercises management control. These include large utility companies such as Kenya Power and Lighting Co and Kenya Airports Authority; agricultural companies such as Nzoi Sugar Company; as well as a number of banks, such as Consolidated Bank of Kenya. While not large recipients of government grants or transfers, they do receive support through loan guarantees and on-lending and capital transfers.

  • The Central Bank of Kenya.

5. There is no comprehensive register of public sector institutional units maintained in Kenya. An annex to the budget provides the most comprehensive list available, however it is incomplete, in that it is missing a large number of county water service corporations and a number of commercial parastatals.

6. Kenya’s public sector expenditure accounted for around 41 percent of GDP in 2012 13, significantly greater than the 25 percent of GDP in budget expenditure. This is due to large expenditures from the non-commercial and commercial parastatals sectors, combined with a small amount of spending from the local authorities. The net figures net out the 5.8 percent of GDP of inter-public sector transfers comprising of transfers from the exchequer to parastatals, local governments and extra-budgetary funds, as well as some dividend payments—primarily from the Central Bank—and surplus funds from the parastatals to the budget.

7. The devolution in 2013–14 changes the mix of expenditures between the central and sub-national governments. Sub-national expenditure increases from 1 percent of GDP by local authorities in 2012–13 to around 4½ percent of GDP by counties in 2013–14, with an equivalent increase in transfers from the budget (primarily the Equitable Share). While this would be expected to result in no increase in overall expenditure, problems around the transitional arrangements may result in direct central government expenditures not decreasing by a commensurate amount.

8. Kenya’s key fiscal reports focus primarily on the budgetary central government, with very little information on the other sectors, even within the central government. The Financial Statements report solely on the receipts into and issues from the exchequer, as well as the use of those releases and own source revenue by BCG entities, on a cash basis. While these include transfers to funds and parastatals, they do not provide details on their expenditure or own-source revenue. The coverage of government expenditure estimated in the Quarterly Economic and Budgetary Review is similarly limited. The KNBS’ Annual Economic Survey does provide a slightly higher degree of coverage, by including information on the county/local authority expenditures. The Controller of the Budget also provides quarterly reporting on the expenditure and revenue outcomes of the counties as well as budgetary central government.

9. Kenya’s substantial parastatal and extra-budgetary sectors remain outside the key fiscal reports. This leaves around 28 percent of both central and general government expenditure unreported, and 44 percent of public sector expenditure unreported in fiscal reports (Figure 1.1). This is a significant share, and results in Kenya not meeting the basic requirement to report on all central government entities. The current reporting framework needs to be strengthened in order to differentiate correctly between market and non-market activities, essential for reporting on the different subsectors of government so as to enable general government reporting. Incomplete information on parastatals and extra-budgetary funds are provided as an annex to the budget documents, though these are produced irregularly, not accounted for in the aggregates and not included in ex post fiscal reporting.

Figure 1.1.
Figure 1.1.

Kenya: Coverage of Public Sector Institutions in Fiscal Reports

(Percent of expenditure)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: Staff estimates.

10. Kenya is required to move to full general government reporting under its East African Monetary Union (EAMU) commitments. This will be needed to report on the EAMU general government fiscal deficit and public debt convergence criteria, which must be achieved for three years from 2021 in order to enter the monetary union. It will be important to prepare this information well in advance of these timelines in order to take policy action to meet them.

1.1.2. Coverage of Stocks (Basic)

11. There is no reporting of the government’s balance sheet in any of the fiscal reports, although basic information on public debt and cash deposits are provided. The Treasury’s debt report and QEBR provide information on both domestic and external debt, while cash and deposit information for the budgetary central government is provided in the debt report. The CBK also provides information on public debt as well as government and other public depositors’ cash balances at the Central Bank, though similar information on deposit balances at commercial banks is not available. These reports put government debt liabilities at June 2013 at 42.3 percent of GDP, with pending bills of 0.7 percent of GDP, and cash assets of 1.3 percent of GDP. There is also information provided for on-lent loans of 3.7 percent of GDP.

12. Kenya’s balance sheet is likely to be far more extensive than reported, with public sector liabilities estimated for this report at 93 percent of GDP—more than double the reported level of public debt (Table 1.3 and Figure 1.2). This difference comes from a number of reasons:

Figure 1.2.
Figure 1.2.

Kenya: Public Sector Balance Sheet and Coverage in Fiscal Reports

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: Staff estimates.
Table 1.3.

Estimate of Kenya’s Public Sector Balance Sheet, 2012–13

(Percent of GDP)

article image
Source: Staff estimates derived from CBK Statistical Bulletin, EBF and Parastatal Annual Financial Reports (30 largest entities, latest available, not always 2012–13); 2011–12 Local Authority Financial Statements, Report on Actuarial Valuation and Projected Benefit Expenditure, internal Treasury estimates.

  • The inclusion of unreported public service pension liabilities of 22 percent of GDP. These liabilities represent the amount of pension obligations the government already owes to existing and past public servants. This can be roughly thought of as the lump sum amount the government would have to pay out should all civil servants have retired on June 30, 2013.

  • The extension of coverage from budgetary central government to the public sector—including liabilities of sub-national governments of 1.6 percent of GDP; non-commercial parastatals of 4.2 percent of GDP; and commercial public corporations of 21.6 percent of GDP—largely in the form of loans and bond issues, as well as liabilities of the publicly owned banking sector of 10.9 percent of GDP, of which the Central Bank’s currency on issue represents a large component.

  • These liabilities are reduced by consolidating the major cross holdings of assets and liabilities within the public sector. For example, many of the loans owed by public corporations are actually on-lent from central government, which need to be consolidated to avoid double counting. Similarly, much of the government’s debt on issue is held as an asset by public sector units - such as the NSSF, the Central Bank and commercial banks.

13. The size of Kenya’s public sector liabilities is relatively small compared to available figures for other countries (Figure 1.3). However, as these data are not readily available for comparable countries to Kenya, it is difficult to draw any firm conclusions, particularly given Kenya’s relatively limited access to international debt markets—especially for public corporations—compared to advanced economies. The other African FTE to date in Mozambique found a similar level of liabilities, though public corporations’ share was significantly larger.

Figure 1.3.
Figure 1.3.

Public Sector Gross Liabilities

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: Staff estimates for Kenya, Portugal, Ireland, Russia, and Romania; National Financial Statements.

14. Public sector assets estimated at 92 percent of GDP offset most of the liabilities -and are far greater than the reported assets of 5 percent of GDP. These additional unreported assets include:

  • Nonfinancial general government assets of 53 percent of GDP. These include buildings, roads, property and equipment that the government has accumulated in investments over the past 40 years. This estimate is very unreliable, as there is no register or valuation of public assets, and therefore has been calculated using the perpetual inventory method based on historical investment and depreciation rates. This estimate does not include the value of public lands or sub-soil assets (such as oil and gas reserves), as no data are available, representing a significant gap in the balance sheet.

  • Central Bank assets of 12.5 percent of GDP including foreign exchange reserves, balances due from local banks and bonds held for trading.

  • Loans issued by publicly controlled banks of 1.3 percent of GDP. These are largely made up of mortgage and other lending by the banks to the private sector.

  • A number of consolidated asset cross holdings within the public sector, particularly the value of publicly owned parastatals (both commercial and non-commercial), whose net worth are recorded as an asset by the government, but are then consolidated resulting in net worth across the different layers of government remaining unchanged.

15. Kenya’s net worth is estimated to be roughly balanced at minus 0.9 percent of GDP, falling in the mid-range of available international comparators, though caution must be expressed around this figure. The very large uncertainties around the value of non-financial assets—and high likelihood of overvaluation of the estimate due to poor quality investments in the past and insufficient maintenance of existing infrastructure—mean that the true net liability figure could be significantly higher. Nevertheless, this is relatively low compared to existing estimates, mainly from advanced economies, and in line with other developing economies (Figure 1.4).

Figure 1.4.
Figure 1.4.

Public Sector Net Worth in Selected Countries, 2012-13

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: Staff estimates and national financial statements.
1.1.3. Coverage of flows (Basic)

16. Fiscal reports are predominantly prepared on a cash basis and do not recognize accrual flows. The accounting and budget reports are consistent in applying a modified cash basis, under which invoices in respect of goods and services received prior to the year-end may be paid during the first two weeks of July but expensed in the previous year. The annual financial statements disclose amounts for pending bills outstanding at the year-end, but these are expensed against the following year’s budget. The in-year QEBR includes committed expenditures. Public finance statistics reports are based on the transactional cash flows and, excepting changes in the value of external borrowing stocks, take little account of other economic flows.

17. Fiscal reports recognize the cash revenues, expenditures, and financing flows and are supplemented with some information on accrual balances. Budget documentation and in-year reports distinguish between revenues, expenditures, and financing. They further differentiate between recurrent and capital investment flows. Information on borrowing stocks, on-lending, and guarantees are routinely reported.

18. Significant accrual-based flows that would impact on net worth remain unreported. This is because the cash basis concentrates on cash transactions arising during the reporting period and their impact on cash balances. Other economic flows are not recognized. For instance, increases in pension obligations for public servants and military employees are neither expensed nor recognized as a liability in the accounts. The planned introduction of IPSAS accrual compliant standards and formats will progressively capture and report on the other economic flows and balances.

19. Cash to accrual adjustments would increase the 2012-13 deficit from 5.2 to 8.0 percent of GDP. Table 1.4 illustrates the impact of the significant adjustments. The computation excludes value and volume changes to liabilities (e.g., impact of exchange rate variation on external debt) and assets (e.g., stocks, inventories, investments and depreciation). The significant changes comprise:

Table 1.4.

Kenya: 2012/13 Cash to Accrual Adjustment

article image
Source: QEBR, Debt Management Report, 2012-13 Auditor Report, Report on Actuarial Valuation and Projected Benefit Expenditure.Note: other than the net accrued pension liabilities, the other adjustments are reported in various public documents.

  • Pensions – public service employees are for the most part entitled to non-contributory, benefit-defined pension schemes. The May 2014 actuarial review estimated the value of pension benefits accrued to date and indicated that pension liabilities would continue to accrue at an estimated 22.1 percent of the current payroll costs per annum. For 2012/13 this equated to 1.9 percent of GDP.

  • Pending Bills – these are disclosed in the annual financial statements. They represent invoices for goods and services delivered or consumed during the year, but which remained unpaid at 30th June. Settlement subsequently takes place in the new financial year. The change between the 2012-13 opening and closing balances of pending bills was 0.9 percent of GDP.

  • On lending arrangements – at June 30, 2013, state corporations were in arrears in servicing their debt to central government—KSh12.3 billion was overdue in respect of principal repayments and KSh24.2 billion interest (including KSh507 million which had accrued during 2012/13). In managing the portfolio, a number of adjustments were agreed upon during that year, including: the conversion to equity of KSh2,058 million principal and KSh12,989 million interest; the write-off of KSh347 million principal and KSh2,912 million interest; and the conversion of KSh1,862 million principal to a grant. The accrued interest, write-off and conversion to grant are reflected as cash to accrual adjustments.

  • Annual investments in PPPs – assets constructed under PPP arrangements typically have service lives longer than the contract period. The government controls the assets and bears the risks for a major portion of the asset lives. Imputed costs should be recorded during construction. No details on the actual costs of PPP construction in 2012/13 are available but, where they exist, would further increase the GFS fiscal deficit.

1.1.4 Tax expenditures (Not Met)

20. There is no reporting of tax expenditures. Despite a constitutional requirement for a public record of tax waivers to be published, and the PFM Act’s requirement for an annual report, no such reports are available. An internal record is kept of tax exemptions approved by the Cabinet Secretary Finance, which would be a good starting point for reporting.

21. However, the methodology for identifying and measuring tax expenditures is yet to be formally established. The various tax laws provide for exemptions and the manner in which these are to be approved. These exemptions include among others: investment deduction allowances (which can reach 150 percent if investment is done in remote areas); export promotion incentives, through Export Processing Zones; zero-rated VAT products; and tax reliefs on interest earnings, on capital gains, and on mortgage interest payments.

22. Studies done by the Kenya Revenue Authority and the World Bank estimate tax expenditures of around 2½ percent of GDP, broadly similar to countries in the region (Figure 1.5). These studies identified the major tax expenditures as accelerated depreciation and investment deduction allowances and exemptions to Export Processing Zones.

Figure 1.5.
Figure 1.5.

Tax Expenditures

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: World Bank and Kenya Revenue Authority.

23. A number of tax expenditures have been removed over recent years. The VAT Act of 2013 reduced significantly the incentives given to investors by repealing VAT remissions that existed on some activities, and by reducing the number of items that fall under preferential rates. The proposed Excise Duty Bill, currently under review, also streamlines exemptions.

1.2. Frequency and Timeliness

1.2.1. Frequency of in-year reports (Basic)

24. Kenya produces regular in-year fiscal reports which are submitted to Parliament and are publicly accessible. These include the Quarterly Economic and Budgetary Reviews (QEBR) produced by the Treasury, and quarterly budget implementation review reports covering the national government (BIRR) and county governments (CBIRR) prepared by the Controller of Budget (CoB). Additionally, the Treasury publishes a monthly Statement of Actual Receipts into and Issues from the Exchequer Account.

25. The reports concentrate on flows and vary in their coverage. The QEBR provides the most comprehensive synopsis of the national budget implementation and is the key in-year fiscal report. It includes discussion on economic and fiscal developments, provides an update on the stock of public debt and guarantees, and contains summary expenditure/commitment tables on the implementation of the national budget by implementing agency and by economic category. The BIRR also provides information on implementation of the national budget, reporting on the flows to and from the Exchequer account and payments made against the Exchequer issues (with expenditure analysis by sector, implementing agency and major economic category). The CBIRR does the same for county budget implementation. The Statement of Actual Receipts into and Issues from the Exchequer Account is the most limited, reporting only on the flows into and out of the Exchequer Account, with no analysis of how the funds have been applied.

26. The QEBR is published quarterly within 45 days of the end of the quarter. The calendar for its submission to the Parliament and publication is prescribed in the PFM Act and has been adopted as standard practice. The Controller of Budget, in her role as independent overseer of budget implementation, routinely submits the BIRR and CBIRR within two months of the quarter end.

1.2.2. Timeliness of annual financial statements (Basic)

27. Audits of annual financial statements for individual ministries are completed within twelve months of the year-end. Accounting Officers prepare and submit annual financial statements for audit in respect of the Votes and funds for which they are accountable. These are submitted by October 31 (four months after the year-end) and audits completed by the following May/June (11–12 months after year-end). The Auditor-General submits to the Parliament a report on the annual audits detailing his findings. The report is published on the Kenya National Audit Office (KENAO) website.

28. The Treasury produces and submits to the Auditor-General an annual Statement of Receipts into and Issues from the Exchequer Account and Summaries of Unaudited Financial Statements. The Statement is audited for its account of the receipts credited to the Exchequer, of the distribution of those funds and the Exchequer balance. This audit is completed by 31st December (i.e., within six months of the year-end), and the audited statement with the Auditor-General’s opinion are included in his annual report to the Parliament. The Unaudited Financial Statements extract and aggregate specific information from the annual financial statements of the individual ministries. The summaries, however, are not audited and, despite significant changes made to the individual ministries financial statements as a result of the audit, are neither updated to reflect the corrections nor published.

1.3. Quality

1.3.1. Classification (Basic)

29. Kenya produces fiscal reports analyzing revenues and expenditures by administrative, economic and functional classifications. The QEBR and annual financial statements provide breakdowns by administrative and economic classifications. Annual reports published by the Kenya National Bureau of Statistics (KNBS) provide GFS and COFOG compliant reports. Currently the KNBS does not produce in-year public sector reports.

30. Kenya’s reporting is supported by a standard chart of accounts introduced in July 2013. The structure provides multiple reporting segments, is harmonized across budget and accounting functions and is shared by the national and county levels of government. It was introduced as part of the re-engineered IFMIS and its structure is consistent with international standards, as illustrated in Table 1.5. However, its benefits have yet to be fully realized as the improvements are still to be exploited in the annual financial statements. In this regard, a Public Sector Accounting Standards Board (PSASB) has been established and has started to prescribe international public sector accounting standards (IPSAS) compliant standards and formats to be progressively introduced from 2013/14.

Table 1.5.

Kenya’s Classification vs. International Standards

article image

31. Some current practices diminish the credibility and usefulness of GFS-compliant reports. Notable amongst these is the high proportion of the national budget designated for general government purposes but classified under Grants with no detailed breakdown of budget or actual (e.g., transfer to Defense). For instance, the Economic Survey’s Table 6.9 shows the values of Grants to General Government Units as a proportion of Total Recurrent Expenditure varying between 23 percent and 33 percent over the four years from 2009/10 to 2012/13. Incomplete adoption of the standard chart of accounts introduces further complexity, requiring manual intervention to complete the mapping from program codes in budgetary central government and administrative codes in counties to appropriate COFOG codes.

1.3.2. Internal consistency (Basic)

32. Kenya reliably publishes only one of the three internal consistency checks on fiscal data required under the code. Information on debt issued and debt holdings are provided in various documents, primarily the CBK’s statistical bulletin, providing details of the nationality of holders of external debt, and breakdowns of domestic debt holdings by instrument and sector. There are various comparisons of net lending to net financing, however these vary between reporting institutions. The fiscal data of statistical record from the KNBS do not reconcile, with a large average absolute discrepancy of 1.7 percent of GDP over the past four years between the above the line budget balance and below the line net financing (Figure 1.6).

Figure 1.6.
Figure 1.6.

Kenya: Discrepancy between Budget Balance and Net Financing Estimates

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: KNBS Economic Survey.

33. There are no reconciliations provided between the net financing and change in net debt estimates, though available data suggest that large discrepancies exist for most years. While there is no reconciliation available, the required data are provided in various documents, resulting in a series of large stock flow adjustments. As the public debt data are solid and can be compared to external data, this calls into question the reliability of both the above-the line budget balance, and the net financing figures. There are no data provided to explain these large stock-flow adjustments, and no explanation provided in any documents for the four percent of GDP increase in net debt in 2011 above financing needs for the budget, or the similarly large discrepancies in 2002 and 2003.

Figure 1.7.
Figure 1.7.

Kenya: Stock-Flow Adjustments

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: Central Bank of Kenya Statistical Bulletin and WEO database.* Change in gross public debt up to 2008-09, net public debt from 2009-10 onwards.

34. While some data that could potentially explain the source of the stock-flow adjustments are published, they do not resolve the discrepancy. Potential causes for the stock flow adjustments include changes in government cash deposits, on-lending and equity holdings, or exchange rate variations. These are published—to a greater or lesser extent—however the available information does not explain the discrepancy. A potential underlying cause may be revisions to budget balance data (see the next section) that are not recorded in either the budget outturn or central bank reports, and while published in the statistics, are not reflected in revised net financing estimates.

1.3.3. Historical consistency (Not met)

35. There is no systematic reporting or explanation of revisions to fiscal aggregates in any of the fiscal reports, despite a pattern of large upward revisions to budget deficits. The fiscal statistics are updated as the estimates of fiscal aggregates move from provisional to final following the outcome of the auditing process. However, these appear as new figures in the following year’s report, without any comparison, explanation or bridging tables reconciling the different vintages.

36. Two of the key fiscal reports do not revise their fiscal data following external audits, providing an incorrect picture of the public finances. The Treasury’s QEBR and the Central Bank’s Statistical Bulletin continue to publish provisional fiscal outcome data, even after the results of the external audit have been provided. This is an area of concern, as over the past five years, revisions between provisional and audited figures have resulted in an average increase of 1.3 percent of GDP to the budget deficit (Figure 1.8).

Figure 1.8.
Figure 1.8.

Kenya: Revisions to Fiscal Balance

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: KNBS Economic Survey.

37. There have been no major revisions to historical fiscal statistics over the past five years other than the move from provisional to audited figures, though large revisions are likely in the near future. This is an indication that historical data are not revisited once they have been audited. As the government moves to IPSAS and GFS reporting over coming years, there will be changes to coverage and basis of accounting that should be applied retrospectively over historical fiscal data to provide consistent time series. These should be accompanied with clear explanations for revisions. In addition, the move to adopt System of National Accounts 2010 is anticipated to result in a larger estimate of GDP that will reduce the key fiscal ratios. As this standard is applied, all figures should be provided in nominal terms, with the old and new ratios to GDP provided for clarity and transparency around the new figures.

1.4. Integrity of Fiscal Reports

1.4.1. Statistical Integrity (Advanced)

38. Annual fiscal statistics are produced and disseminated by the KNBS in line with GDDS practices. The independence of the KNBS is provided by the Statistics Act of 2006, which lays out the mandate, responsibilities and roles of various actors within the KNBS. The head of the KNBS is appointed for a term of three years, renewable once, and is appointed by the Cabinet Secretary of Planning based on a shortlist of three candidates provided by a board of directors. The board is made up of representatives from universities, research institutions and the private sector, as well as from the Treasury and Ministry of Planning and Devolution. The director can only be removed from office by the Cabinet Secretary Planning on a recommendation from the board. While there are some issues with the fiscal statistics, overall the agency does appear to be both independent and professional.

1.4.2. External Audit (Basic)

39. The Kenya National Audit Office (KENAO) provides independent oversight of public finances.1 The Office of the Auditor General is an independent body established under the Constitution to audit public entities and report on their management of public funds. It has a complement of 974 staff, of which 714 are auditors or examiners, and can outsource audits for public corporations, projects and specialist investigations. The Auditor General is nominated and, subject to the approval of the Parliament, appointed by the President for a non-extendable period of eight years. The Auditor General may only be removed from office in cases of misconduct or incapacity in accordance with the provisions of the Constitution.

40. KENAO applies international standards and a risk-based approach to its audits. Amongst others, the KENAO is required to carry out audits of: the national and county governments; public funds; statutory bodies, commissions and independent offices; public corporations; and projects. It plans and performs audits to obtain reasonable assurance about whether the statements are free from material misstatement and issues opinions as to whether the statements are unqualified (“representing a true and fair view)”, or are deficient—with opinions of unqualified, adverse or disclaimer reflecting increasing degrees of deficiency

41. The 2012/13 audit of national government entities resulted in 42 percent of expenditures being subject to adverse and disclaimer audit opinions, though many of these are cleared before the report is adopted by Parliament. Table 1.6 illustrates the analysis for expenditure statements and the chart shows the proportion of expenditures within each opinion category. Issues that lead to adverse or disclaimer opinions include: discrepancies between reported figures and ledger balances; inconsistencies between the closing balances for one year and opening balances for the next; exclusion of transactions from the statements; unexplained, unsupported or unreconciled balances; failure to comply with treasury instructions; and bookkeeping errors. The nature and scale of discrepancies follows a similar pattern to the previous year’s audit. Audits of the 2012/13 revenue statements also resulted in 13 of the 23 statements attracting an adverse or disclaimer opinion.

Table 1.6.

Kenya: 2012/13 Audit Opinions – National Government Expenditures

article image
Source: Executive Summary of the Auditor-General’s Report 2012/13.

42. Article 229 of the constitution requires that the audit report should be submitted to Parliament within six months after the end of each financial year. In order to meet this deadline, the Auditor General must work with the information that is available at that time and it is inevitable that the first draft of the report has significant audit queries. However, ministries continue to provide the required information and by the time the report is adopted by parliament, most of the queries have been addressed although KENAO does not clear them. The National Treasury is working with KENAO to ensure that queries are addressed before the report is published, beginning this year. There will also be a tracking of all queries to expedite any unresolved issues arising from the audit.

43. The absence of consolidated annual financial statements prevents the Auditor-General from expressing an opinion on the government’s overall financial performance and position. Irrespective, the magnitude of errors detected in the individual financial statements suggests that their aggregation could not produce reliable consolidated annual statements worthy of a “true and fair view” audit opinion.

1.4.3. Comparability of Fiscal Data (Basic)

44. Fiscal reporting has only a basic level of comparability between reports, with the budget outturn reports presented on the same basis as the Budget. The budget documentation reports on the key fiscal aggregates, whose revised figures and actual outcomes are subsequently reported in the Treasury’s QEBR document. The budget also provides reporting on expenditures and revenues flows on the Exchequer account, while actuals are reported by the Controller of Budget’s quarterly report. However, the exchequer basis of reporting is not easily converted into fiscal aggregates, due to inclusion of principal debt repayments as expenditure.

45. The published audited statements are limited to reporting the flows on the Exchequer account, and do not include information on the fiscal aggregates, making it difficult to compare to the budget. It is possible to perform a basic reconciliation from the Exchequer flows, as reported in the unaudited financial statements, to the fiscal aggregate outturns reported by the QEBR. However, this requires a number of adjustments, whose basis would not be clear to anyone without strong understanding of underlying workings of the budget, and still leaves a residual difference of between 1 to 5 percent of expenditure over the last five years.

46. The fiscal statistics should in principle be on the same basis as both the budget and the QEBR outturn documents, however differences between reported actuals from the two sources leave this in question. There should be no difference between reported outturns for the most recent year between the Economic Survey and QEBR, as they both report on provisional fiscal outturns, have the same institutional coverage, and are on a cash basis. However, over the last 5 years, the two reports have varied in their estimates of provisional expenditure by an average of 2.5 percent. The variances between actual figures are significantly larger, owing to the reporting of revised audited figures by the KNBS (though these do not always match the audited outcomes in the audit report), and unrevised provisional figures in the QEBR, resulting in markedly different actual outturn estimates for key fiscal aggregates (see Figure 1.9).

Figure 1.9.
Figure 1.9.

Kenya: Reconciliation of 2012–13 Total Expenditure Between Key Fiscal Reports

(KSh billion)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: QEBR, Unaudited financial statements, Audit Report, Economic Survey 2014.

1.5. Recommendations

47. The above assessment highlights the following priorities for strengthening fiscal reporting in Kenya:

1.1 Expand the institutional coverage of the key fiscal reports, to include at a basic level the direct expenditure and revenues of extra-budgetary and parastatals sectors. Include an informational annex in the key reports, combining data from the Controller of Budget on county expenditures and revenues, and present aggregate fiscal data on a general government basis.

1.2 Begin preparing a balance sheet for the central government in the key fiscal documents, using data that is already available within the existing systems and reports. This should be on a provisional (i.e., non-audited basis) initially. As government reporting systems move to an IPSAS basis, improve the quality of the estimates, and expand the coverage to first, general government, and eventually to the entire public sector.

1.3 Improve the quality and timeliness of the audited financial statements, with Accountant General’s Department producing consolidated annual financial statements in accordance with PSASB standards, and building capacity to lessen reporting deficiencies. The Auditor-General should complete the audit within 6 months of the year-end. The National Treasury should publish the audited statements.

1.4 Move the basis of reporting for all fiscal documents to internally consistent final audited figures and begin reporting the major fiscal aggregates in the financial statements. Investigate and report on any remaining discrepancies between the fiscal balance and net-financing and stock-flow adjustment. Undertake an annual reconciliation between the net financing and change in debt stock outturns during the preparation of final accounts, preparing a reconciliation table to be audited by the Auditor General. Develop and implement a methodology for preparing and explaining historical revisions.

Table 1.7.

Summary Assessment of Fiscal Reporting Practices

article image

II. Fiscal Forecasting and Budgeting

A. Introduction

48. Budgets and their underlying fiscal forecasts should provide a clear statement of the government’s budgetary objectives and policy intentions, and comprehensive, timely, and credible projections of the evolution of the public finances. To do so, fiscal forecasts and budgets should:

  • Provide a comprehensive overview of fiscal prospects, and be presented on a gross basis (revenue, expenditure and financing);

  • Be presented, debated, and approved in a timely and orderly manner, guided by clearly defined powers and responsibilities of the executive and legislative;

  • Be presented in a way that facilitate policy analysis and accountability; and

  • Be credible and subject to independent evaluation.

49. Kenya’s fiscal forecasting and budgeting has evolved over the past decade, but several challenges remain:

  • Budgetary documentation does not provide a comprehensive coverage of revenue, expenditure and financing of central government;

  • Selection and budgeting of public investment projects is fragmented resulting in limited disclosure of their financial obligations, and a lack of systematic cost-benefit analysis;

  • Current fiscal principles do not translate into clear, measurable and binding fiscal policy objectives; and

  • The multi-year planning and budgeting process is yet to produce reliable expenditure priorities and credible forward estimates that underpin predictability in funding.

50. Reforms have been introduced to address some of these challenges, but are yet to deliver the desired objectives. Planned actions which remain outstanding include:

  • Information on government owned entities (GoE): Drafting legislation for reporting and inclusion of off-budget resources in budget document;

  • Information on public investments: Extending the coverage of E-ProMIS2 to include GoK-funded projects;

  • Fiscal Objectives: Preparing the regulatory framework to implement the fiscal responsibility principles in the PFM Act 2012; and

  • Credible forward expenditure estimates: Strengthening and rolling-out program-based budgeting to improve expenditure prioritization and costing of policies and programs.

2.1. Comprehensiveness

2.1.1 Budget Unity (Good)

51. The annual budget documentation provides a generally satisfactory picture of the fiscal operations of the central government, including the main social security funds. The documentation and its purpose are summarized in Table 2.1. The main estimates books provide the fiscal operations for the budgetary central government, including transfers to parastatals, EBFs and sub-national governments. From 2012, an annex has been included in the budget documentation summarizing revenues and expenditures of most parastatals though this was not included in the 2014–15 Budget.

Table 2.1.

Kenya: Fiscal Forecasting and Budget Documents

article image

52. While the coverage of the budget is not complete, the entities remaining outside of the Summary Annex are presumed to be relatively small. The annex reports on 128 of the 132 parastatals, with the 4 unreported entities assumed to be relatively small; and the 15 major EBFs3 out of a total of 75. Data on the unreported funds are not available, however after extensive discussion with the authorities, these are believed to be either small or dormant. Nevertheless, all efforts should be made to include these entities within the next annex.

53. However, there are a number of issues with the presentation of these estimates that significantly obscure transparency, holding the assessment back from an advanced rating:

  • The size of central government is significantly understated in the fiscal aggregate figures presented in the BPS and budget estimates as they do not contain the parastatals or EBF’s revenue, expenditure or financing. While the separate numbers are provided in the annex, these are not easily integrated, due to consolidation issues.

  • There are large inconsistencies between transfers to parastatals and EBFs from the budget estimates, and receipts of grants in the summary annex, of 1 percent of GDP for recurrent grants, and 9 percent of GDP for development grants. This suggests not all recurrent transfers are being picked up in the annex, and that some development grants are being channeled directly to parastatals without being included in the budget estimates.

  • The annex fails to distinguish the commercial parastatals that are outside of the central government such as KENGEN, who are engaged in market activities, from the non-market core central government entities - such as the Kenya Revenue Authority, that are inside.

54. The budget documentation presents all expenditures and revenues of the budgetary central government departments on a gross basis. The Budget Estimates provide a clear distinction between expenditures funded through: (i) domestic revenue (ordinary revenue); (ii) external loans and grants channeled through the exchequer; (iii) external loans and grants disbursed outside the exchequer; and (iv) own source revenues. The last two categories of funding are authorized as appropriation-in-aid (AIA), shown on both the expenditure and revenue side of the budget.

2.1.2 Macroeconomic Forecasts (Good)

55. Budget documentation provides detailed medium-term macroeconomic forecasts and discusses the main underlying assumptions. Comprehensive forecasts are presented in the BROP, which is published in September, and then later updated in the BPS, which is published in February. Forecasts over the budget and two outer years cover real GDP growth, inflation (Consumer Price Index and GDP deflator), the terms of trade, investment and savings, reserves and the current account balance. The BROP and the BPS also discuss some of underlying assumptions behind growth and inflation forecasts, such as global and regional trends, domestic conditions, international oil prices, and fiscal and monetary policies. The Budget Statement provides a final update in April, though this is fairly minimal.

56. The budget documentation does not provide explanations of key macroeconomic variables and their components. A more complete analysis would focus on the key components behind the forecasts, such as sectors of the economy (agriculture, industry and services), and provide an explanation on their contribution to GDP growth. A similar exercise could be done using the expenditure or income components of GDP. That way, policy measures, such as measures presented in the last budget to improve irrigation and infrastructure, could be discussed through their impact on the components of GDP. The set of projected variables could also be extended to cover asset prices and indicators of the long-run underlying trend of the economy (productivity growth, and the output gap).

57. While Kenya’s year-ahead forecasts errors for real GDP are relatively high, they do not present a strong bias. Among comparable countries, Kenya has one of the highest absolute average forecast errors, indicating that the accuracy of the forecasts is relatively low (Figure 2.1). Nevertheless, its forecasts, for both budget and outer years, do not seem to suffer from any major bias, compared to a set of sub-Saharan and European countries (Figure 2.2). However, there are two distinct periods within that decade: in the first half, forecasts were conservative, with growth outturns stronger than forecasts, in the latter half, the reverse was true, with outcomes coming in well below forecasts.

Figure 2.1.
Figure 2.1.

Absolute Forecast Errors of Real GDP Growth, 2003–12

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: Staff calculations.
Figure 2.2.
Figure 2.2.

Forecast Errors of Real GDP Growth, 2000–13

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: Staff calculations.
2.1.3 Medium-term Budget Framework (Advanced)

58. The government introduced a detailed medium-term expenditure framework for central government in 2001. The integration of top-down and bottom-up expenditure forecasting has resulted in a more consistent budget, and greater focus on the medium-term impact of implementation of the national development priorities.

59. The medium-term budget framework, published in the BROP and BPS, presents updates of macroeconomic fiscal performance, fiscal policy objectives and expenditure projections over the medium term. The documentation explains the performance against the targets on the macro-economic fiscal objectives of the two preceding years as well updates on the current year, and states the medium-term macro-fiscal outlook. Specifically, the BPS summarizes revenue, expenditure and financing of aggregate central government for the two preceding years, upcoming budget and two outer years broken down by sector, ministry and broad economic category (see Table 2.2) and the Estimates present the medium-term figures on the same basis as the annual budget. Following the introduction of program-based budgeting in 2012, the government has published Estimates of Expenditure for the National Government for 2013/14 and 2014/15 stating the gross expenditure estimates on a rolling three-year horizon for each program under a department or “Vote” with clear program objectives and performance indicators.

Table 2.2.

Example of an MTBF 2014/15 Budget (KSh million)

article image
Source: National Treasury.

60. However, the real benefits of the MTBF approach in Kenya are yet to be fully realized. While the annual budget discipline has improved, medium-term expenditure discipline remains elusive. Kenya’s fiscal forecasting record has been relatively cautious in the near-term but mixed over the medium term. Generally, expenditure comes in under budgeted amounts in the budget year, due primarily to underspends in the development budget. However, the medium-term expenditure projections have not yet proven to be a solid anchor on future years’ expenditure, with actual expenditure on average 9 percent higher than projected over the medium term (Figures 2.3 and 2.4). While those upward revisions to expenditures over the medium term are at the lower range for the EAC region, they are considerably higher than even the worst performing countries in Europe (Figure 2.3).

Figure 2.3.
Figure 2.3.

Kenya: Government Expenditure: Outturns versus Forecasts (2001/02–2013/14)

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: BPS and QEBR.
Figure 2.4.
Figure 2.4.

Medium-Term Forecast Errors of Expenditure (2000–12)

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: BPS, QEBR, and European Commission.

61. The failure of MTBF to improve the predictability of medium-term expenditure can be attributed to a number of weaknesses in the underlying PFM system. These include: poor prioritization; unrealistic budgeting, and poor policy and program implementation. The Treasury should design and implement a capacity building program as part of the roll out of the program-based budgeting to address these weaknesses and ultimately improve the credibility of the framework.

2.1.4. Investment Projects (Basic)

62. While public investment management regulations and procedures are in place, public investment projects are not systematically subject to cost-benefit analysis. All new capital project proposals are appraised by the technical departments in the relevant line ministry in accordance with rules established by the Public Procurement and Disposal of Assets Authority. However, while externally funded projects are subject to rigorous cost benefit analysis, the same cannot be said of domestically funded public investment projects and furthermore, where this is done, the results are not always made public. Furthermore, within the National Treasury, the approval and budgeting of public investment projects remains fragmented with the External Resources Department coordinating externally-funded projects and the Budget Supply Department coordinating domestically-funded projects.

63. The annual Estimates for Development Expenditure do not include the total financial obligation of projects, just the expenditures and source of funding for the budget and two outer years. In addition, the budget documentation provides insufficient information on the scope and implementation progress of each project, essential to promote accountability and ensure effective oversight. Furthermore, the documentation includes on an equal footing both capital projects and “softer” development activities such as capacity building, making it difficult for policymakers and the public to have a clear picture on how the budget addresses Kenya’s infrastructure needs. The recent MTEF guidelines for 2013/14 budget year requested the central government departments and agencies to submit detailed information on each project, but the response from sectors was mixed.

64. The provision in the PFM Act of 2012 requiring both central and county governments to allocate a minimum of 30 percent of budgetary resources to development projects will impose additional strain on already unsatisfactory absorption capacities. As shown in Figure 2.5, there has been persistent under-spending of the development budget in recent years, indicating a significant absorption capacity problem which will need to be overcome if the new target of 30 percent is to be achieved. As noted above, increasing absorption will require streamlining institutional roles and coordination arrangements, particularly within the National Treasury.

Figure 2.5.
Figure 2.5.

Kenya: Annual Execution Rates of the Development Budget 2007/08–2012/13

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: National Treasury

2.2. Orderliness

2.2.1 Fiscal Legislation (Advanced)

65. The Constitution and the Public Financial Management Act of 2012 provide a comprehensive legal and regulatory framework that governs the collection and use of public resources. The framework states the principles of public finance, guides the sharing of the national revenues and states the processes as well as responsibilities of the executive and legislature in public finance management. In addition, the law includes a limit on the legislature’s power to amend the executive’s budget proposal. Specifically, the PFM Act includes a number of elements that are part of modern financial management legislation, namely: introduction of stronger fiscal rules and medium-term objectives; strengthening the macro-fiscal focus of the law; and extending the coverage of the budget to all entities included within the general government.

66. Sections 35-40 of the PFM Act state the timetable for budget preparation and approval. They provide the timing of the major interactions within and between the executive and the parliament, detail the process for submission of the main budget and accompanying documents, including the process for preparing the parliamentary budget, and specify the approval and publication of the Appropriation Act. Specifically, Section 38 lays out the content requirements for the budget documentation focusing on key requirements that should facilitate an informed debate on fiscal operations of the entire general government.

2.2.2. Timeliness of Budget Documents (Good)

67. The budget is presented to Parliament at the end of April, two months prior to the budget year, and the budget estimates are approved and published before the beginning of the fiscal year. This conforms to the legal framework requirements mentioned above and is in line with good practice under the Fiscal Transparency Code. The medium-term Budget Policy Statement (BPS) is submitted, together with the medium-term Debt Management Strategy, to Parliament in February two months before presentation of the annual budget. The BPS sets the allocation of resources across government, both within central government and between levels of government, and is adopted by Parliament along with any changes it deems appropriate. The budget documentation is published on National Treasury Website to allow the public sufficient time to scrutinize and engage with the executive on the annual budget and related forecasts.

2.3. Policy Orientation

2.3.1. Fiscal Policy Objectives (Basic)

68. Section 15 of the PFM Act requires the National Treasury to enforce a number of fiscal responsibility principles, though the implementing regulations have yet to be finalized. By nature, many of these are fiscal objectives rather than principles, setting floors or ceilings for a number of key fiscal aggregates over the medium term, notably:

  • a minimum of 30 percent of the national budget to be allocated to development spending;

  • government borrowing to only finance development spending;

  • the wage bill to revenue ratio not to exceed a percentage to be defined by regulation; and

  • public debt and obligations to be capped at levels to be defined by regulation.

69. Taken together, these principles do not provide operational guidance for policy makers’ decisions on the main fiscal aggregates such as the overall balance, total expenditure or public debt. There are still no limits defined for the wage bill or public debt, as the key regulations are still pending. The requirement that government borrowing be used only to finance development expenditure does not in practice provide a reasonable limit on the deficit, given that development expenditure is targeted at some 30 percent of the budget (10 percent of GDP). The lack of any binding constraint on the fiscal deficit can be seen in the gradual upward drift of the deficit to the current level of 5.7 percent of GDP (Figure 2.6).

Figure 2.6.
Figure 2.6.

Kenya: Overall Fiscal Balance, Including Grants

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: QEBR.

70. These fiscal principles are clearly stated and some are reported against. The BPS reiterates the main fiscal principles set in the PFM Act 2012. The Memorandum to the Budget Statement, first submitted to Parliament in April 2014, reports against some of these principles, notably on the extent to which development spending is meeting its target and borrowing is being undertaken to finance development projects.

71. Progress on operationalizing these principles is expected to be made in the near future. The Public Finance Regulations, currently under preparation, will provide numerical targets to the wage bill and the debt-to-GDP ratio. Capping the wage bill may help contain spending pressures. A cap on debt can provide an anchor to the deficit (as long as it is sufficiently binding in the medium term). Going forward, Kenya will be required to adhere to the fiscal convergence criteria set out in the East African Monetary Union (EAMU) Protocol, if it wishes to enter the monetary union planned for 2024. This would require the annual overall deficit not to exceed three percent of GDP and the debt stock to stay below 50 percent of GDP in net present value. However, these criteria are not yet embodied in legislation or set out as government objectives.

2.3.2. Performance Information (Good)

72. The Budget documentation presents information on targets for, and performance against, the outputs delivered under each major government policy area. The Estimates of Expenditure for the National Government for fiscal year 2014/15 provide: (i) high level policy objectives and outcomes related to the second Medium-term Plan Kenya Vision 2030—which are further cascaded into programs or sub-programs; (ii) a summary of the performance against output targets for the previous years; and (ii) output targets to be delivered under each program within the medium term. This has helped to focus the policy debate on department’s policy objectives and track their achievement against set targets.

73. However, the production of the program budget estimates—initiated in 2012—is still in its infancy and faces several challenges. Programs are still constructed around departments and agencies of the newly restructured state departments making it difficult to understand their link with the second medium-term plan priorities. Furthermore, many of the outputs are not clearly defined and not measurable, undermining the generation of reliable costs of program implementation. As a consequence, the performance information has not yet had significant influence on budgetary resource allocation.

74. Detailed performance information on the financial impact of government policies on citizens is produced but not submitted as part of the budget documentation. The Fourth Annual Progress Report 2011–12 reviewed the implementation of major flagship policies and projects and performance indicators articulated under the First Medium Term Plan 2008–12 of the Kenya Vision 2030. The Annual Public Expenditure Review explains the trends in fiscal policy management as well as budget allocation and execution, and provides an empirical investigation of the efficiency and effectiveness in the various sectors of the Medium-Term Plan. However, publication of both these reports has consistently been late and not synchronized with the budget process, thereby missing the opportunity to influence budgetary decisions.

2.3.3 Public Participation (Advanced)

75. Formal public participation has been part of the budget process since the mid-2000s with a significant influence on budget decisions in select sectors. The budget calendar provides several platforms for consultation with the public on various fiscal or sector policy issues. The line ministries invite the public through the print media or the website to send their opinions on sector spending proposal or tax measures. The non-government organizations also participate in sector budget discussions led by sectors or line ministries and public hearings on macro-economic and fiscal outlook, including sector spending conducted by National Treasury. The public hearings have been able to influence spending decisions that would address issues of equality in human development and poverty reduction. Further, the legislature also seeks views of the public on the annual budget proposals at county level. The existing practices are enshrined with the legal framework which also provides for a publication of budget summary that can be accessed by the citizens.

76. A Citizen’s Budget Guide has been produced for the last three years. This provides an accessible description of (i) recent fiscal performance and economic prospects; (ii) summary of the implications of the budget for a typical citizen; (iii) the implications of the budget for different demographic groups, namely programs aimed at cushioning the poor and vulnerable-social safety nets, pro-poor spending, food insecurity; and (iv) equity and regional balance and Youth and Women Programs.

2.4. Credibility

2.4.1 Independent Evaluation (Good)

77. The non-partisan Parliamentary Budget Office (PBO) provides independent medium-term macroeconomic and fiscal forecasts. The PBO publishes its separate forecasts in an annual Budget Options document. This usually appears a week after the Treasury’s Budget Policy Statement. The Budget Options paper critically assesses past fiscal performance against targets, and proposes policy options for the annual budget based on its own projections of resource constraints. While it does not explicitly evaluate the credibility of the Treasury’s forecasts against its own, it does provide a benchmark against which the Treasury’s forecasts can be evaluated by parliamentarians and the public. It also provides an explanation of the differences due to differing underlying assumptions. It is worth noting that the PBO has generally been more cautious in its projections of government revenue in the past, which has often proven to be more accurate (see Figure 2.7).

Figure 2.7.
Figure 2.7.

Comparison of One-Year Ahead Revenue Projections Excluding Grants

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Sources: BPS, Budget Options, QEBR 2014 (Q4).
2.4.2. Supplementary Budget (Good)

78. Any material variation to the approved budget is subject to authorization by the National Assembly. The interpretation of material variation is relatively narrow. It includes increases to aggregate budgeted expenditures, as well as transfers between administrative votes, programs, development and recurrent activities, and economic groupings (e.g., personal emoluments and goods & services). The approval takes the form of a supplementary appropriation bill. In Kenya the number of supplementary bills is limited—generally to one per year following a mid-year budget review, but a second if exceptional demands arise.

79. The practice is more stringent than legally required. The legal framework (Constitution Article 223, PFM Act Article 44) permits the retrospective authorization of budget increases of up to 10 percent of the aggregate approved budget and requires the parliamentary submission to be accompanied by a statement demonstrating how any additional expenditure relates to the fiscal principles. In practice, parliamentary approvals are sought prior to funds being made available for commitment or payment purposes. The revised budget is released in the IFMIS, used by all central government units for expenditure control and accounting.

80. However, the expenditure outturns are always significantly lower than both supplementary and original budgets, undermining the budget as an effective guide to expenditure. Whilst Figure 2.8 illustrates that with the exception of 2011/12 the new Constitution with major reduction of ministries and transfer of functions to counties—supplementary changes were relatively small in aggregate terms, the actual outturns are on average 12 percent lower than the original estimates. Figure 2.9 shows that there is no regular pattern whereby savings on development are reallocated to recurrent purposes or vice versa.

Figure 2.8.
Figure 2.8.

Kenya: Total Central Government Budgets and Outturns

(2004/05–2012/13)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: National Treasury.
Figure 2.9.
Figure 2.9.

Percentage Change between Original and Revised Budgets

(2004/05–2012/13)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: National Treasury.
2.4.3. Forecast Reconciliation (Basic)

81. Budget documentation provides limited information on the size and causes of forecast revisions in the medium term. The BROP and the BPS provide a reconciliation table which shows the current and the last published forecast for the medium term (budget year and two outer years) for revenue, expenditure and financing over the medium term. An update on the changes in fiscal performance during the current year is provided, as well as some explanations on how these changes, including policy changes, can explain the forecast revision. There is no discussion on the changes to the economic outlook and how it has led to revisions in the forecast.

82. Forecasts variations have tended to be large and should be reconciled. In the past three years, outer-year expenditure forecasts have been revised upwards by nine percent on average. These revisions suggest that the previous outer year forecasts were set at unrealistically low levels. An example of forecast reconciliation is presented in Table 2.3.

Table 2.3.

Kenya: Indicative Reconciliation Table for Expenditure Ceilings

article image
Source: National Treasury and IMF staff estimates.

2.5. Recommendations

83. The above assessment highlights the following priorities for strengthening fiscal forecasting and budgeting in Kenya:

2.1 Address discrepancies of information on grants across the various budget documents. Ensure consistency between the transfers budgeted in the estimates document and the grants received shown in the annex on public corporations, and provide explanations of any residual differences in the annex document.

2.2 Clarify institutional roles for public investment management, strengthen the coordination, appraisal and monitoring of public investment projects to raise absorption capacity, require cost-benefit analysis for all major infrastructure projects, and ensure that budget documents reflect the full lifetime cost of projects, as well as the stock of related multi-year contract commitments being carried over into the new budget year.

2.3 Clarify the fiscal policy principles to give practical guidance to policy makers as to what the major fiscal aggregates should be, and provide clear measurable reporting against them in the budget documents.

2.4 Improve the current forecast reconciliation table by introducing a clear decomposition of the three sources of forecast variations: (i) new policy decisions taken during the period; (ii) changes to the economic outlook; and (iii) accounting and technical changes.

Table 2.4.

Summary Assessment of Fiscal Forecasting and Budgeting Practices

article image

III. Fiscal Risk Analysis and Management

A. Introduction

84. A government’s ability to cope with fiscal risk depends on the quality of its information about risks, its powers to mitigate such risks, and its capacity to absorb the fiscal implications of those risks should they materialize. As a minimum, this implies the need for institutional and procedural arrangements, and associated capacities, to collect, analyze and report those fiscal risks with a potential to disrupt fiscal objectives and medium-term fiscal sustainability. These would usually include risks arising from macroeconomic volatility, guarantees, state-owned enterprise losses, court judgments, and disasters. Mitigation against such risks is usually handled through budgetary reserves or contingencies funds. A more comprehensive approach to risk management would cover all potential risks, including those arising from PPPs, natural resource exploitation, financial sector difficulties, management of assets and liabilities, and long-term fiscal sustainability issues such as pension liabilities.

85. While macro-economic volatility has been low in recent years Kenya is still exposed to a wide range of fiscal risks. The volatilities of GDP and government revenues have been relatively low compared to other countries, making macroeconomic risks less challenging. Additionally, the government’s exposure to the financial sector is limited. Debt is currently judged to be sustainable and below the ceiling set out under the EAMU convergence criteria. Additionally, demographic trends are not expected to exert much pressure on age-related expenditures in the long term, under current policy settings. However, other fiscal risks are emerging notably from devolution—where contingent liabilities could increase as newly established counties exercise their powers under the Constitution, from oil and gas exploitation, and from the deteriorating security situation.

86. The present disclosure and management of fiscal risks is not very comprehensive. A fiscal risk statement has been consistently included in the BPS in recent years and is a step in the right direction. Kenya displays advanced practice against the Fiscal Transparency Code in a few areas such as the disclosure and management of contingencies funds and environmental risks. However, the statement’s documentation of fiscal risks is not comprehensive, omitting significant contingent liabilities of about 11 percent of GDP, including sizeable PPP activities and government guarantee schemes such as the one related to the NSSF. Additionally, it does not fully quantify those risks that are discussed. Also, despite the potential realization of fiscal risks from devolution, there is still no systematic monitoring or reporting of activities outside the perimeter of the central government, including the inherited assets and liabilities of the newly formed counties and their fiscal framework.

3.1. Risk Disclosure and Analysis

3.1.1. Macroeconomic Risks (Basic)

87. The Statement of Specific Fiscal Risk in the annual Budget Policy Statement analyzes the potential impact of macroeconomic shocks. This is a requirement of the PFM Act 2012 and has been produced for the past three years. It considers the impact of a one percent shock to GDP, inflation and import prices, and a ten percent shock to the exchange rate.

88. The sensitivity analysis assesses the magnitude of the shocks on the fiscal balance but does not provide a discussion of the mechanisms at play. This makes it difficult to follow the impact on particular variables and some counter-intuitive results make the numbers seem implausible. For instance, in the last statement, a one percent positive shock to GDP leads to an increase of government expenditure by eight percent of GDP, which would necessitate unrealistic adjustments to the other components of GDP. In addition, a review of the last three statements shows that the same shocks have markedly different impacts. For example, a ten percent depreciation of the domestic currency increases government revenue by one percent in 2013 but reduces it by 5.5 percent in 2014. The analysis does not consider alternative scenarios that reflect the potentially wider distribution of shocks that would occur in practice. The underlying modeling for this analysis is still work in progress.

89. The volatility of macroeconomic and fiscal variables is not as high as in comparable countries, suggesting that macroeconomic risks may be relatively contained in Kenya (Figure 3.1). However, this assumption may change in the future. For instance, as the share of commercial borrowing increases, debt is likely to become more sensitive to changes in interest rates. As shown in Figure 3.2, the risk of exceeding the 50 percent EAMU debt target is low at around 5 percent probability.

Figure 3.1.
Figure 3.1.

Volatility of Nominal GDP and Nominal Government Revenues (2003–13)

(Standard Deviations of Percent Growth)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Sources: IMF WEO database.
Figure 3.2.
Figure 3.2.

Kenya: Macro-Sensitivity Analysis of Public Debt

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: Probabilistic forecasts based on DSA assumptions in the 2014 Article IV report; EAMU Debt Limit.
3.1.2. Specific Fiscal Risks (Basic)

90. The government produces a Statement on Specific Fiscal Risks which discusses risks to the public finances qualitatively. The statement is included as a separate section of the government’s Budget Policy Statement, which is submitted to Parliament and published annually in accordance with the provisions of the PFM Act of 2012. It discusses the potential exposure to various shocks including environmental shocks, contingent liabilities, and financial sector exposure. It also makes policy recommendations to help mitigate these risks.

91. There is some quantification of fiscal risks but it is not exhaustive and omits significant items with potentially large fiscal implications. The statement on specific fiscal risks omits a discussion on existing commitments including: (i) active PPP arrangements such as the Kenya-Uganda Railways and power purchasing arrangements that are associated with Independent Power Producers (IPPs); (ii) indemnities and letters of support associated with these IPPs; (iii) pending court proceedings related to pension awards to teachers; (iv) exposure under the Deposit Protection Fund; (v) exposure under the new government guarantee scheme of the National Social Security Fund; and (vi) existing callable capital to international financial institutions such as the African Development Bank (Table 3.1).

Table 3.1.

Contingent Liabilities in Kenya, 2013

article image
Sources: IMF Staff calculations and data from authorities.

92. Some fiscal risks that are not included in the Statement on Fiscal Risks are quantified in various other reports. These risks include: (i) outstanding on-lent loans to state corporations that have roughly doubled in 2012/13 (4 percent of GDP), which are analyzed in the Annual Debt Report; (ii) debt financing risks which are analyzed under different scenarios (including one with country borrowing) in the Medium Term Debt Strategy document; and (iii) contingent liabilities related to the deposit protection fund, which are outlined in the Financial Stability Report of the Central Bank of Kenya, and the annual report of the Deposit Protection Fund Board. However, the value of this information is somewhat reduced by its fragmentation.

93. In total, unreported contingent liabilities are assessed in this report to be KSh869.4 billion or 17.2 percent of GDP (Table 3.1). The identified items, and their quantification, are explained in the various specialized sections later in this chapter.

3.1.3. Long-term Fiscal Sustainability Analysis (Not Met)

94. The government does not produce or publish projections of the main fiscal aggregates or the social security funds over the long term. Projections for the main fiscal aggregates (revenues, expenditures, deficit, and debt variables) are made over three years only. Overall, the government relies on the Debt Sustainability Analysis report produced jointly by the IMF and the World Bank for its assessment of fiscal sustainability. Projections for pension liabilities are available over more than 30 years but are not published. Furthermore, projections for the potentially increasing liabilities of the National Social Security Fund are not produced.

95. The lack of comprehensive long-term sustainability analysis by the authorities is not an immediate concern given Kenya’s relatively low debt and low demographic pressures. The government is committed to a fiscal consolidation plan that aims to reduce debt levels in order to meet the EAMU ceiling in 2021. Unlike many advanced countries, Kenya’s old-age dependency ratio is not expected to rise rapidly in the coming decades, limiting the demographic pressure on age-related expenditures (Figure 3.3).

Figure 3.3.
Figure 3.3.

Old Age Dependency Ratio in 2050

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: Staff calculations

3.2. Risk Management

3.2.1. Budgetary Contingency (Advanced)

96. Kenya’s budget includes an allocation for budgetary contingencies with transparent access criteria and reporting on its utilization. The PFM Act of 2012 provides the regulatory framework for managing the Contingencies Fund established in the 2010 Constitution. The Contingencies Fund is to be used for urgent and unforeseen expenditure during the fiscal year, including natural and man-made disasters. The Constitution caps the size of the Contingencies Fund at KSh10 billion, but in practice the government has allocated only half that amount in the last two budgets. The Treasury is required to report to Parliament on the utilization of the Contingencies Fund within two months.

97. Utilization of the Contingencies Fund requires a case to be made and is subject to approval by the Cabinet Secretary Finance. Records are kept for supporting evidence when reporting to Parliament and for audit. Figure 3.4 shows the utilization of the Contingencies Fund over the last few years, as well as the various purposes for which the fund was used during 2013/14. Based on these practices, Kenya achieves an advanced score on this indicator.

Figure 3.4.

Contingency Fund Spending in Kenya

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: National Treasury data.
3.2.2. Assets and Liabilities (Basic)

98. There is no report that shows the central government’s consolidated assets and liabilities. Major liabilities include public debt, pension liabilities and payables. The major assets can be grouped as cash, investments, on-lending and debtors, and fixed assets.4 The extent to which these liabilities and assets are disclosed and managed is discussed below:

  • Public Debt – government borrowing is controlled by the National Treasury, and subject to approval by Parliament. The National Treasury submits a Medium-Term Debt Strategy (MTDS) annually to Parliament and publishes it. The MTDS analyzes the current stock of debt and provides guidance on the amounts and types of borrowing and guarantees to be undertaken over the medium term. It evaluates the costs, risks and sustainability of various scenarios and proposes implementation strategies.

  • Pension Liabilities - most central government public servants are entitled to a defined-benefit pension on retirement. Currently pensions are funded from the national budget. Actuarial surveys are periodically carried out to determine the present value of current pension liabilities and to project this liability forward. The actuarial study completed for the Salaries and Remuneration Commission in May 2014 showed a current liability of KSh991.9 billion5 for 2013 (22 percent of GDP). Strategies to manage the liability are under discussion and include the implementation of the Public Service Superannuation Act 2012.

  • Investments – the government has majority holdings in 39 state owned enterprises and minority share holdings in 29 companies. Details are contained in an annual Consolidated National Government Investment Report. State owned enterprises negotiate annual performance contracts with the National Government and report against the agreed measures. Poor performing enterprises are identified and strategies developed to restructure, privatize or dissolve. Registers of shares are maintained for minority shareholdings but there is little active management of this portfolio.

  • On-lending6 - As at June 30, 2013, the government had on-lent KSh167 billion (3.7 percent of GDP) to state corporations. However, debt service on these loans, as at June 30, 2013, were in arrears by KSh12.3 billion on principal repayments and KSh24.2 billion on interest payments (together equivalent to 0.8 percent of GDP). During the course of the year management actions had been taken in respect of a number of these non-performing loans, including rescheduling of payment dates, write-off of unrecoverable sums, and conversion to equity or grants.

  • Cash balances – Exchequer account operations result in a substantial overdraft from the Central Bank of Kenya, despite the fact that the government holds large unremunerated cash balances with the Central Bank and commercial banks. The Treasury is currently working with the Central Bank to establish a Treasury Single Account, as required by the PFM Act of 2012. This will treat all cash holdings as one, and improve their management.

  • Fixed assets – There is no comprehensive asset management framework to guide the strategic planning, acquisition and disposal of assets, operations, security, maintenance, accounting and performance monitoring.

99. Schedules of liabilities and assets handed over to the county governments are not available, presenting a significant fiscal risk. The 47 counties inherited the liabilities of the 175 local authorities that they absorbed. Available information from counties such as Nairobi that have taken on the liabilities, as well as the 2011/12 audit statements, suggests that they are significant and include outstanding loans, court compensations, and arrears in salaries, pension contributions, and payments to suppliers. The Transition Authority is mandated to collate and verify this information but has not yet completed the task. In the meantime, some counties have taken over the liabilities and others only partially. A recent court ruling confirmed county responsibility for these liabilities. Meanwhile, some counties have chosen to transfer assets in settlement of their debts.

3.2.3. Guarantees (Basic)

100. Parliament imposes a ceiling on the stock of guarantees in line with the PFM Act of 2012. The current ceiling is set at KSh200 billion (4 percent of GDP). The PFM Act 2012 limits the power to issue new guarantees to the National Treasury.

101. The government reports on loan guarantees, although the coverage is not exhaustive, especially in relation to implicit guarantees. The stock of publicly guaranteed debt and their beneficiaries are disclosed in the Annual Debt Report. The latest debt report of May 2014 show that government loan guarantees amounting to around KSh45 billion (0.9 percent of GDP), consisting of loans contracted by seven public corporations and one sub-national government (Nairobi) with four lenders (Japan, Canada, U.S.A. and IDA).

102. However, there is no disclosure of a number of other commitments, which are financially similar to guarantees, even if they have a different legal form. These collectively represent around KSh287 billion (6.4 percent of GDP). These are:

  • National Social Security Fund: Article 70 (b) of the National Social Security Fund Act of 2013 requires government to protect members against default. As such, should the Fund default with zero net worth and defaults, the government will be liable to cover all members’ contribution plus accrued interest. Based on the National Social Security Fund’s 2013 balance sheet, this liability is currently estimated at KSh113 billion or 2.2 percent of GDP. The liability is expected to rise due to the increased contributions and coverage provided in the new Act, including the conversion of public servants to a defined contribution scheme under the National Social Security Fund.

  • Deposit Protection Fund: The Deposit Protection Fund guarantees deposits up a maximum of KSh100,000 (USD1,176) in the event of bank failure. The Deposit Protection Fund’s assets currently cover around 20 percent of the value of insured deposits in Kenya, valued at KSh198.9 billion by the Bank of Kenya in 2013. The uncovered portion of the deposits amounts to KSh153.4 billion or 3 percent of GDP.

  • Guarantees issued by public corporations: an analysis of the financial statements of the ten largest public corporations shows that public corporations have issued loan guarantees to the amount of KSh3.4 billion as of June 2013.

3.2.4. Public Private Partnerships (Not Met)

103. Information on active Public Private Partnership (PPP) contracts has not been made publically available by the government and their fiscal implications are yet to be fully assessed. Recent estimates produced by the World Bank suggest a total contract value of active PPP projects is around USD3.4 billion (5.7 percent of GDP), a significant total which should warrant full disclosure and risk management by the National Treasury (Figure 3.5). There are two PPP arrangements that pose the highest risks at this time:

Figure 3.5.
Figure 3.5.

Contract Values of Public Private Partnerships (PPPs)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: World Bank, and data from the authorities. *** Based on data from the authorities

  • The 25-year Kenya-Uganda Railways PPP concession contract signed in 2006 includes exclusivity provisions which may be affected once the Mombasa to Nairobi Standard Gauge Railway construction project becomes operational.

  • Twelve power purchasing agreements, totaling USD3.1 billion or 5.2 percent of GDP, have been signed between the Kenya Power and Lighting Company, a state corporation, and Independent Power Producers between 1996 and 2013, before the PPP Act of 2013 was enacted and enforced. These contracts involve an obligation for a minimum demand/revenue guarantee (“take” or “pay”). They are supported by letters of support by the government providing coverage for political risks and underwritten by Partial Risk Guarantee instruments provided by the World Bank (Table 3.2).7

Table 3.2.

Recent Indicators of Banking Financial Sector Stability

article image
Source: GFSR database.

104. New PPPs planned over the period 2014-17 will pose additional risks if their fiscal implications are not properly managed and disclosed. All new PPPs will be negotiated and signed within the new PPP Act of 2013. The PPP unit in the National Treasury is mandated to collect and disclose information on all PPP contracts (Section 25 of the PPP Act 2013). A list of 47 national priority-PPP projects has already been identified in several sectors notably transportation, tourism, housing, and energy (Figure 3.4). This will add to contingent liabilities going forward, and urgently requires a PPP assessment framework to be put in place for which technical support has already been agreed with the World Bank.

3.2.5. Financial Sector Exposure (Good)

105. The fiscal risks linked to the financial sector are relatively low, as the government has limited exposure in the sector. The Kenyan banking sector comprises 43 commercial banks, 1 mortgage finance company, and 9 deposit taking microfinance institutions. Total bank assets amounted to around 54 percent of GDP in 2013, with the three remaining public banks accounting for only 4.6 percent of GDP. The ratio of nonperforming loans to total loans has increased recently, but data indicate that banks remain sound and profitable (Table 3.2).

106. There has been no government support to the banking sector since the last bailout of a major public bank in the mid-2000s. In 2008, the government injected liquidity (0.7 percent of GDP) into the National Bank of Kenya by paying off debt owed to the bank by state corporations. This decision was made in order to enable the public bank to meet key statutory and prudential requirements at the time. Since then there have been no other instances of government support to failing banks.

107. A deposit insurance scheme exists, providing coverage for depositors and serving as a liquidator for potentially failed financial institutions. The maximum coverage of Ksh100,000 has been in place since the scheme’s inception in 1985, but is currently under review. The Deposit Protection Fund Board reports annually on the fund. Currently, the fund’s balance covers around 20 percent of insured deposits which are estimated to be around 4 percent of GDP. The coverage value is somewhat below the international benchmark of 40 percent, but is in line with the minimum coverage threshold set by the International Association of Deposit Insurers. The uncovered portion of deposits (3 percent of GDP) constitutes contingent liabilities for the government, and thus potentially poses additional fiscal risks to public finances.

108. Reporting on explicit financial sector exposure of the government is adequate. The government regularly discloses all support to the financial sector in various reports. The Budget Policy Statement and analyzes the potential exposure of the government to financial sector risks and would disclose any explicit support or intervention in the system. The Central Bank of Kenya publishes a bi-annual Financial Stability Report and an annual Bank Supervision Report where it identifies risks and vulnerabilities in the financial system, and suggests policy measures to mitigate them. These reports encompass a fairly thorough assessment of financial sector stability, and conducts micro stress tests that are based on a fairly plausible range of shocks, including credit risk shocks, liquidity risk shocks, and market wide shocks. The Central Bank is working on incorporating macro stress testing in its reports, which will be based on a wide range of macroeconomic assumptions and a plausible range of financial market scenarios.

3.2.6. Natural Resources (Not Met)

109. The exploitation of natural resources at present is not very significant in Kenya. According to the Wealth of Nations Database by the World Bank, the net present value of known deposits of minerals, subsoil assets and forests (non-timber) is only around 4 percent of GDP. The government discloses the volume and sales of these assets in the Economic Survey published by the Kenya National Bureau of Statistics (KNBS).

110. There have been several recent discoveries of oil and gas in Kenya which will need to be closely monitored once the size of exploitable reserves is known. As part of the fiscal risk framework, disclosure of the government’s assessment of reserves and the analysis of their potential fiscal implications will be required. Currently there are 23 investors undertaking explorations in 44 separate license blocks. Crude oil has been found in 7 wells (Tullow Oil with Africa Oil as Joint Venture Partner), and natural gas and oil were found in both onshore and offshore blocks (Apache and British Gas Group). It is early days in the exploration, and commercially viable quantities are yet to be proven. However, Tullow, one of the main investors in oil fields in Kenya, estimates reserves to be well over 600 millions of barrels of oil equivalent, comparable to the reserves of Equatorial Guinea and the Republic of Congo. If this is confirmed, it could generate substantial revenues for the government, and potentially bring Kenya’s external current account to surplus soon after exploitation starts.

111. The government discloses information on some ongoing exploration activity in the oil and gas sector. Basic information on the status of the explorations is provided in a report published by the Ministry of Energy. The current medium-term budget envelope does not include potential revenues from these sources.

3.2.7. Environmental Risks (Advanced)

112. Annual costs from natural disasters in Kenya are not particularly high compared to other countries, but they have been increasing over recent years. Kenya is prone to cyclical drought shocks due to its location in the Horn of Africa. The average annual cost of damages due to natural disasters over the last decade was around 0.1 percent of GDP (Figure 3.6). However, some past drought episodes have cost substantially more, the most recent one being in 2012 which necessitated a larger policy response financed mainly through access to funds set aside for drought relief and the contingency fund (0.3 percent of GDP), as well as support under the IMF’s Rapid Financing Instrument. The government quantifies the fiscal implications of potential drought episodes based on past experiences in its annual Budget Policy Statement. In addition to drought risks, Kenya has recently faced significant challenges related to terrorism events which could have important implications on the economy going forward.

Figure 3.6.
Figure 3.6.

Environmental Risks in Kenya and Comparator Countries

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

IMF Staff calculations and data from World Bank.

113. To coordinate and effectively respond to drought shocks, the government has outlined a well-defined strategy to manage drought risks. This strategy is published in the Vision 2030 document. The government has also established the National Drought Management Authority which has adopted an effective approach to monitoring, managing and coordinating these risks through its monthly Drought Monitoring and Early Warning Bulletin reports. These reports identify likely zones of drought and issue specific warning signals. These reports are produced for each county in Kenya and are regularly published on the authority’s website. The authority is preparing an investment plan on Ending Drought Emergencies, which has drawn potential support from several development partners. Additionally, the government is currently developing and adopting policies that help mitigate drought shocks in the long-term, including a scaling up of investment projects related to geothermal energy production (which reduced the reliance on hydro), and the expansion of irrigation.

3.3. Fiscal Coordination

3.3.1. Sub-National Governments (Not Met)

114. There has been a four-fold increase in subnational spending with the establishment of counties in 2013 which sharply increases fiscal risks in this sector. Budgets of the old local authorities (LAs) amounted to 1 percent of GDP while those of counties increase subnational spending to 4.3 percent of GDP. Counties are now responsible for significant expenditure programs, including health and agriculture. They have far greater autonomy than the old LAs and are allowed to borrow though only with a central government guarantee. The reporting framework for counties to central government is yet to be fully developed, with the exception of constitutionally mandated reporting to the Controller of Budget and the Auditor General. Without adequate oversight and reporting to provide early warning of fiscal problems, the counties represent significant risks to fiscal sustainability over the medium term.

115. Financial reporting in the old Local Authorities was weak and the new counties have yet to submit their first annual accounts. The 2011–12 audit of Local Authorities found significant weaknesses in their financial reporting practices, with 169 out of the 175 local authorities having disclaimer or adverse audit opinions. Furthermore, the closing balances of local authorities at the time of their merger with counties are yet to be fully verified and audited, undermining the credibility of the balance sheets of the new counties. Among the liabilities were outstanding loans and unpaid obligations the value of which cannot always be ascertained due to lack of evidence—Nairobi County has current liabilities of over KSh 40 billion. Many counties are disputing responsibility for these past debts, some of which are now accruing penalty interest.

116. Although the Constitution allows counties to borrow with central government guarantee, so far there has been no significant county borrowing. While counties have not yet approached the central government for guarantees for long-term borrowing, as required by Constitution and PFM Act 2012, some counties are already resorting to in-year liquidity borrowing from local banks. The framework for long-term borrowing is currently under discussion, including the sharing of the aggregate annually agreed national borrowing limit.

117. Overestimation of own revenues are contributing to fiscal imbalances in some counties. Own revenue estimates in 2013/14 county budgets was over KSh 60 billion while collected revenues are estimated to reach only KSh 25 billion. The worst affected counties are also those that fared less well in the redistribution of resources under the 2012 revenue sharing formula. Figure 3.7 suggests that those counties with lower ratio of equitable share to cost of devolved functions (i.e. more fiscally imbalanced) systematically overestimated their own revenues compared to those with higher ratios (i.e. less fiscally imbalanced).

Figure 3.7
Figure 3.7

Overestimation of Own Revenues and Fiscal Imbalances in Counties

(FY2013/14)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: IMF staff calculations based on data from the authorities and the World Bank
3.3.2. Public Corporations (Basic)

118. All transfers to public corporations are included within the budget estimates, and are published in a parastatal annex to the budget. This annex provided revenue, expenditure and operating balance information on all parastatals, both commercial and non-commercial, with a relatively clear summary of the entire sector. Although it is sometimes produced after the budget is adopted (e.g. in the case of the 2014-15 budget, which has not yet been published), this annex does provide sufficient basic information on the sector. Every public corporation is subject to an annual performance agreement with its parent ministry that set performance targets.

119. There are no other summary reports detailing the financial performance of the public corporations. Financial information is available for the majority of the 39 corporations in the annual financial statements included in their annual reports that are produced according to IFRS, although many are not available on the internet, and where they are, are often outdated.

Figure 3.8:
Figure 3.8:

Public Corporations Balance Sheet, 2012–13

(Percent of G DP)

Citation: IMF Staff Country Reports 2016, 221; 10.5089/9781475538069.002.A001

Source: Annual Financial Statements, 2012–13 where available.

120. There are no summaries of the size or health of the sector’s balance sheets, which while not particularly large by international standards, include 9 percent of GDP of liabilities (see Table 1.3). The majority of the balance sheets however seem relatively healthy, with an unusually high equity share, equivalent to 8 percent of GDP – almost equal to liabilities, and almost half of assets. This is a legacy of large state involvement through equity injections and grants, as well as relatively limited access to financial markets from which to borrow.

121. There are also a number of public corporations in difficulties and the sector is set to expand, warranting closer monitoring of financial performance. Examples include the sugar companies and ferry services, some of which are in negative net equity positions. The public corporation sector is anticipated to expand in coming years—with revenues more than doubling in the 2014–15 budget, off the back of large donor loans, particularly to the railway and power generation sectors. In addition, there are pressures to create new corporations, particularly at the county level, who inherited the 120 water service providing corporations (whose activities are well reported and monitored by the Ministry of Water), but have not yet begun setting up any other corporations. These issues warrant close monitoring, with early steps taken should problems emerge. This is currently being done at line ministry level, but the lack of summary reporting—either internally or publicly means the overall risks that the sector poses are not being considered.

122. Public corporations have been the beneficiary of indirect support in the form of on-lent and guaranteed loans from the government. The government has borrowed 4.6 percent of GDP from international donors, and on-lent those loans to public corporations. A number of these loans have not been repaid, or are in arrears. So far the government has written off 0.2 percent of GDP of these loans—the largest of which were to Telkom and Kenya Railways, with further write-offs likely given there are another 0.8 percent of GDP of payment arrears. In addition, the sector has been the recipient of government guaranteed loans of 0.9 percent of GDP, with the guarantees called for 15 percent of these loans.

123. The sector is required to perform a variety of quasi-fiscal activities, though information on the size or value of these activities is not provided. These are centered around the power, transport and agriculture sectors, and include price ceilings on power, loss making sales of fertilizer and free ferry transportation services. These are usually covered through transfers from the budget, and in some cases have set up incentives for poor performance (with loss-making activities automatically covered) resulting in continued governance problems amongst some firms.

3.4. Recommendations

124. The above assessment highlights the following priorities for strengthening fiscal risk analysis and management in Kenya:

3.1 Report and quantify all fiscal risks in the fiscal risk statement and discuss their implications for fiscal forecasts. Coverage should be exhaustive and comprehensive, including currently unreported contingent liabilities arising from public-private partnerships and all existing guarantees (NSSF, Deposit Protection, pending court cases, and callable capital to international financial institutions).

3.2 Disclose all the rights and obligations and other exposures under all existing and planned PPPs and PPP-type contracts at least annually (including at the sub-national government level). It would also be important to report the expected annual receipts and payments over the life of these contracts. Assessing the fiscal implications of these contracts should also be undertaken within a framework maintained by the PPP unit at the National Treasury.

3.3 Ensure a regular reporting (at least annually) on potential oil and gas resources going forward. This includes regular reporting on the volume and value of these assets. Additionally, it would be important to initiate the process of setting up a macro-fiscal framework for oil and gas resource revenues.

3.4 Develop an enhanced county fiscal reporting system. For each county, the system should generate: (a) key debt indicators—e.g., debt service/debt stock, debt service/current revenue, debt stock/current revenue; (b) key fiscal performance indicators—e.g., operating expenditure/operating revenue, own revenue collection/revenue potential, personnel expenditure/operating expenditure, actual development expenditure/planned; and (c) key liquidity indicators—e.g., accounts payable/cash balance, cash balance/operating revenue.

3.5 Prepare an annual financial report on the public corporation sector. The report should provide information both aggregate sectoral fiscal information, as well as summary financial performance indicators for each of the corporations.

Table 3.3.

Kenya: Summary Evaluation of Fiscal Risk Practices

article image

Annex 1. Kenya: Mapping of Fiscal Transparency Indicators to 2013–18 PFM Reform Interventions

article image
Indicator evaluation acronyms: NM – Not Met; B – Basic; G – Good; A – Advanced