This Fiscal Transparency Evaluation (FTE) assesses the quality of fiscal reporting in Kenya against the principles set out in the Fiscal Transparency Code. Kenya has experienced a lot of structural and economic changes since 2014. One of the key objectives of this FTE is to estimate Kenya’s balance sheet, and to cover as many as possible of the entities in the public sector. The coverage of Kenya’s reporting of fiscal statistics has improved considerably. The report discusses that Kenya continues to perform well in the overall transparency of its fiscal forecasting and budgeting practices (Pillar II of the Code), which is based on a strong legal framework. It does so against a backdrop of significant ongoing reforms, including far-reaching fiscal devolution to counties, and the introduction of performance-based budgeting. A recent important change in the law is expected to synchronize the submission and approval of the government’s spending proposals and the tax measures in the Finance Bill. The recommendations set out under each of the pillars of this report aim to address several challenges. The report also encourages the authorities to continue with the implementation of the recommendations set out in the 2014 report, on which good or satisfactory progress has been made in about half the cases.

Abstract

This Fiscal Transparency Evaluation (FTE) assesses the quality of fiscal reporting in Kenya against the principles set out in the Fiscal Transparency Code. Kenya has experienced a lot of structural and economic changes since 2014. One of the key objectives of this FTE is to estimate Kenya’s balance sheet, and to cover as many as possible of the entities in the public sector. The coverage of Kenya’s reporting of fiscal statistics has improved considerably. The report discusses that Kenya continues to perform well in the overall transparency of its fiscal forecasting and budgeting practices (Pillar II of the Code), which is based on a strong legal framework. It does so against a backdrop of significant ongoing reforms, including far-reaching fiscal devolution to counties, and the introduction of performance-based budgeting. A recent important change in the law is expected to synchronize the submission and approval of the government’s spending proposals and the tax measures in the Finance Bill. The recommendations set out under each of the pillars of this report aim to address several challenges. The report also encourages the authorities to continue with the implementation of the recommendations set out in the 2014 report, on which good or satisfactory progress has been made in about half the cases.

I. Fiscal Reporting

1. Fiscal reports should provide a comprehensive, relevant, timely and reliable overview of the government’s financial positions and performance. This chapter assesses the quality of fiscal reporting in Kenya against the principles set out in the Fiscal Transparency Code (FTC). In doing so, it assesses the following: the coverage of institutions, stocks, and flows; the frequency and timeliness; the quality of fiscal reporting; and the integrity of fiscal reports. Fiscal reports, which include in-year budget execution reports, fiscal statistics, and annual financial statements, should:

  • Cover all institutional units in the public sector classified according to international standards;

  • Record all assets, liabilities, revenue, expenditures, financing, and other economic flows;

  • Be published in a frequent and timely manner;

  • Reconcile the different balances calculated and have comparable data across reports; and

  • Be prepared by an independent agency in the case of statistics and audited by an independent external audit authority in the case of financial statements.

2. Kenya publishes many fiscal reports and has been progressively improving its reporting practices. Improvements in coverage can be attributed to the completion of the Public Sector Institutions Table which has been developed using the Government Finance Statistics Manual (GFSM) 2014 standards. Kenya now publishes unaudited financial statements within nine months of the year-end, significantly improving the timeliness of information available to the public. The Kenya National Bureau of Statistics (KNBS) publishes data on spending by the United Nations’ Classification of Functions of Government (COFOG). The National Treasury (NT) has published consolidated financial statements for the different institutional sectors of government, which are expected to be audited soon, a development that provides a good platform for the accrual accounting reforms which are currently being implemented. Table 1.1 sets out the list of fiscal reports published.

Table 1.1.

Kenya: List of Fiscal Reports

article image

3. Since 2014, the rating on three principles of Pillar I of the Code has been increased, and nine have remained unchanged (Table 1.2). The improvements contributing to the increased ratings are discussed below. Since 2014, several important improvements have been made to fiscal reporting practices, summarized in Table 1.3. In this report we make several recommendations to enhance the presentation of fiscal reports, extend the use of the Integrated Financial Management Information System (IFMIS), ensure the consistency and reconciliation of fiscal reports from different sources, initiate the reporting of tax expenditures, and progressively expand the balance sheet coverage of fiscal reports.

Table 1.2.

Kenya: Fiscal Reporting: Summary of Changes since 2014

article image
Source: IMF staff.
Table 1.3.

Kenya: Fiscal Reporting: Progress on Recommendations since 2014

article image

4. Table 1.4 sets out several detailed measures which, once implemented, would further improve the quality, regularity, and timeliness of information disclosed in the government’s fiscal reports.

Table 1.4.

Kenya: Measures to Strengthen Fiscal Reporting

article image
Source: IMF staff.

1.1. Coverage of Fiscal Reports

1.1.1. Coverage of Institutions (Good, Increased from 2014)

5. Kenya’s public sector comprised 519 institutional units in 2017–18 compared to 419 units in 2012–13. Most of the additional institutional units were extrabudgetary units. They can be broken down into the following subsectors:

  • Central government, which includes 120 budgetary central government units (BCG) and 213 extrabudgetary units (EBUs). The BCG sector includes 21 ministries, the Presidency, the Judiciary, the National Assembly, non-ministerial departments and agencies, and cost centers under these entities. Ten EBUs hold almost two-thirds of the total assets.1

  • Local governments comprise 47 counties including Nairobi City;

  • Social security funds include the National Social Security Fund (NSSF), the National Hospital Insurance Fund (NHIF) and the National Social Security Assistance Authority (NSSAA);

  • Nonfinancial public corporations include 121 commercial entities most of which are controlled by the central government. The ten largest entities hold 92 percent of the total assets of this sector;2

  • Financial public corporations comprise the Central Bank of Kenya (CBK) and 14 other financial intermediaries most of which are controlled by the central government. The five largest entities hold 86 percent of the total assets of this sector excluding the CBK.3

6. Kenya’s public sector expenditure is estimated to be around 39 percent of GDP in 2017–18. Kenya’s GDP has almost doubled since FY 2012/13 partly due to the rebasing that took place in 2015 and which raised GDP by about one-quarter. Table 1.5 summarizes the distribution of public sector revenue and expenditure across the different subsectors and shows that:

Table 1.5.

Kenya: Public Sector Institutions and Finances, 2017–18

(Percent of GDP)

article image
Source: IMF staff estimates.

Relates to transfers between entities within the same layer of the public sector.

Gross revenue less inter public sector (PS) transfers.

Gross expenditure less inter PS transfers.

Net revenue less net expenditure.

Using GDP estimates rebased in 2015.

  • General government expenditure accounts for 31.4 percent of GDP on a consolidated basis, of which over three-quarters flow through the central government and the rest through subnational governments and social security funds. This high level of expenditure might be attributed to large infrastructure investments which also contributes to the increase in physical assets for the public sector; and

  • Public corporations’ expenditure accounts for approximately 8 percent of GDP, of which about 90 percent is spent by nonfinancial public corporations.

7. The coverage of Kenya’s reporting of fiscal statistics has improved considerably.

In 2014, the coverage was mainly limited to the BCG sub-sector but has now expanded to cover the consolidation of county governments as well. Factors contributing to this improvement are summarized in Box 1.1. Reporting follows international standards. Through its annual Economic Survey, the Kenya National Bureau of Statistics (KNBS) compiles and publishes data on the general government and its subsectors, classified in accordance with the GFSM 2014. The consolidated financial statements of the government cover the BCG, EBUs and county governments. These serve as building blocks to improving the quality of data for general government and components of the public sector. However, there is need to reconcile data disseminated by the KNBS and reports of the NT which are not always consistent.

Improvements in Coverage

The coverage of public sector expenditure has seen a significant improvement since the 2014 FTE (Figure 1.1). This is largely attributable to the financial reporting reforms undertaken by the NT’s Directorate of Accounting Services. These reforms include:

  • Establishment of the Public Sector Accounting Standards Board (PSASB) and specification of the accounting standards for the general government and PC sectors.

  • Establishment of a Government Finance Statistics Technical Working Group to support fiscal data compilation and periodically review fiscal data to address gaps.

  • The preparation and publication of the Public Sector Institutions Table based on sectors aligned to GFSM2014—this has been instrumental in improving the coverage of public sector entities.

  • Strengthening accounting skills in the NT’s Public Accounts Unit—this has been facilitated by East AFRITAC through technical advice and regional and national training events.

  • The production and publication of consolidated financial statements, which now include separate consolidated statements of Statutory Corporations, SAGAs and Funds. The statements also now present summary data for each individual entity, such as revenue, expenses, operating balances, assets and liabilities. This development is part of the phased implementation of accrual-based IPSAS, under which the government will ultimately consolidate the whole of government by sector in accordance with GFSM2014.

Figure 1.1.
Figure 1.1.

Public Sector Expenditure and Coverage in Fiscal Reports, 2017–18

(Percent of expenditure of each level)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: IMF staff estimates derived from the unaudited consolidated financial reports of the NT.

Reforms are ongoing with East AFRITAC support, and disclosures under full accrual-based IPSAS will further enhance reporting and transparency.

Source: IMF staff.

8. The NT reports quarterly fiscal information through the QEBR that also uses the GFSM 2014 classification. The NT started producing consolidated annual financial reports which include data on the sub-sectors of BCG and counties although the information on (noncommercial) EBUs (Semi-autonomous Government Agencies, (SAGAs)) is consolidated with data on (commercial) public corporations (PCs). This is a major improvement and publishing sperate information on EBUs and PCs can easily be achieved. The NT has also initiated quarterly reporting by the market and non-market establishments through a standardized financial reporting form.

9. The institutional coverage has been improved through reports from the Integrated Financial Management Information System (IFMIS) and use of standard reporting templates. All BCG entities and counties are now using the IFMIS. The standardized templates for EBUs and PCs developed by the government and the PSASB facilitate consolidated reporting of the sector. Compilation of general government fiscal statistics is facilitated by reporting through the IFMIS. The commitments made on the EAC integration agenda have contributed to the rapid expansion of fiscal coverage in Kenya, which has been driven by a significant capacity development effort by the authorities.

10. Annual consolidated financial reports cover over 95 percent of all the EBUs and PCs, a major improvement since the 2014 FTE. Figure 1.1 shows the coverage of fiscal reports in 2017–18 compared to 2012–13. Efforts are underway by the NT and the KNBS to publish fiscal reports that cover the entire public sector, and a preliminary table for 2017–18 has already been prepared for internal use. The unreported expenditures in GFS reports include pending bills and some on-lending of loans.

1.1.2. Coverage of Stocks (Basic, Unchanged from 2014)

11. Fiscal reports largely cover financial assets since the consolidated financial statements are prepared on a cash basis of reporting. The stock of public debt is reported in the annual Economic Survey Report of the KNBS and quarterly public debt reports covering only the BCG are published in the QEBR. Quarterly debt reports cover both the stock of domestic and external debt. The Statistical Bulletin of the CBK also provides information on public debt as well as government and other public depositors’ balances at the central bank. The GFS reports cover only flows. The NT’s Department for Accounting Services prepares and submits for audit each year a separate Public Debt report that shows both the flows and the stocks of debt for the budgetary central government.

12. The balance sheet of Kenya has grown significantly since 2012–13. Public sector assets and liabilities are estimated at 116 percent and 121 percent of GDP, respectively (Table 1.6 and Figure 1.2). Assets and liabilities have increased by an estimated 24 and 29 percentage points respectively since 2012–13. The increase is largely due to the government’s significant investment in infrastructure. Unreported non-financial assets have increased by 24.3 percent of GDP over the same period. The unreported liabilities include:

Table 1.6.

Kenya: Estimate of Public Sector Balance Sheet, 2017–18

(Percent of GDP)

article image
Source: IMF staff estimates derived from CBK statistics bulletin, the KNBS Economic Survey, unaudited consolidated financial reports of the NT, and an internal report of the World Bank on civil service pensions.

EBUs include social security funds.

Refers to cross holdings between public corporations and general government.

Refers to entire public sector after adjusting for cross holdings between the subsectors.

Methodological Note: The estimation of the balance sheet relied on multiple sources of government fiscal reports and documents that provide information on the different variables and instruments that form part of the balance sheet. The estimates of non-financial assets of the general government were based on the perpetual inventory method of estimation, using data on investment and assumed depreciation rates. Financial reports of about 350 public entities were used to generate information on social security funds, EBUs and public corporations. Information available at the KNBS on the operations of the BCG sector and county governments was also used. IMF databases and technical assistance reports complemented these official sources.
Figure 1.2.
Figure 1.2.

Kenya: Coverage of Public Sector Balance Sheet in Fiscal Reports, 2017

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: IMF staff estimates.
  • Public service pensions estimated at about 30 percent of GDP which include accrued-to-date pension obligations for existing and retired civil servants.

  • Public-private partnership (PPP) liabilities to government of about 3.5 percent of GDP.

  • Actuarial obligations to the social security sub-sector equivalent to 3.3 percent of GDP.

13. The main composition of the public sector balance sheet on a consolidated basis is as follows:

  • Nonfinancial assets account for about 90 percent of GDP, of which 48 percent of GDP is attributed to the BCG sector, 20 percent of GDP to EBUs, and the remainder to PCs;

  • Financial assets account for 26 percent of GDP on a consolidated basis, after eliminating cross-holdings of assets and liabilities within the public sector. The general government owns about half of these assets while PCs own the other half before consolidation. The assets of the public sector are composed mainly of currency and deposits, shares and equity.

  • Liabilities account for 121 percent of GDP on a consolidated basis. The major reported instruments include debt securities, loans and payables or other liabilities which represent 22 percent, 31 percent, and 26 percent of GDP respectively.

14. The stock of Kenya’s public sector liabilities is somewhat modest compared to other emerging markets and low-income developing economies (Figure 1.3). Pension liabilities (about 30 percent of GDP) and the growing stock of public debt are important contributors. The public corporations in Kenya have a more moderate stock of liabilities compared to other countries in the sample.

Figure 1.3.
Figure 1.3.

Public Sector Gross Liabilities in Selected Countries, 2017

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: IMF Fiscal Monitor, October 2018.

15. Kenya’s public sector net worth and net financial worth are estimated to be minus five percent and minus 94 percent of GDP in 2017–18, respectively. Comparisons with other countries are shown in Figures 1.4 and 1.5. It should be noted, however, that the estimates of non-financial assets in Kenya are highly uncertain. The process of developing an asset register for government has been initiated but, in the meantime, Kenya has many fixed assets that have not been valued, while the government has been ramping up its investments in infrastructure. Net worth therefore might thus be significantly higher than the estimates reported.

Figure 1.4.
Figure 1.4.

Public Sector Net Worth in Selected Countries

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source. IMF staff. Latest available data: 2017–18 (Kenya, Rwanda, Lithuania), 2016 (Mexico, Malta, Armenia), 2015 (Uganda, Georgia, Columbia), 2014 (Brazil), 2013 (Macedonia), and 2012 (Mozambique).
Figure 1.5.
Figure 1.5.

Public Sector Net Financial Worth in Selected Countries

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source. IMF staff. Latest available data: 2017–18 (Kenya, Rwanda, Lithuania), 2016 (Mexico, Malta, Armenia), 2015 (Uganda, Georgia), 2014 (Brazil).

1.1.3. Coverage of Flows (Basic, Unchanged from 2014)

16. Fiscal reports cover cash revenue, expenditure, and financing, mostly on a cash basis. The annual financial reports disclose accounts payable and receivable as a stepping stone for the transition to an accrual basis of reporting. There are no fiscal reports on other economic flows. Annual and quarterly GFS reports show revenue and expenses on a cash basis, the net acquisition of nonfinancial and financial assets, and the net incurrence of liabilities, classified in accordance with GFSM 2014. The consolidated financial statements of the government include a statement of cash flows from operating, investing, and financing activities as well as a cash-basis statement of revenue and expenditure. Expenditure arrears are disclosed in the notes to the accounts but are not recorded on the face of the financial statements. Accrued pension liabilities are also not recorded.

17. Other economic flows arising from changes in the value of public sector assets may be sizable in Kenya. GFSM 2014 requires a statement of other economic flows, which is composed of holdings, gains and losses, and other changes in the volume of assets. Given the size of public corporations that hold approximately 42 percent and 31 percent of GDP in assets and liabilities respectively, there is a likely sizeable amount of flows due to price and volume changes. These other economic flows should be progressively reported during the phased implementation of accrual accounting standards in the coming few years.

1.1.4. Coverage of Tax Expenditure (Not Met, Unchanged from 2014)

18. Kenya does not publish any regular report that comprehensively discloses the estimated revenue losses from tax expenditures. The Kenya Revenue Authority (KRA) produces a report on annual tax exemptions which is submitted to the Auditor General, but it is not published due to concerns over the reliability of the data. The Parliamentary Budget Office (PBO)’s Budget Options Report includes a table estimating the costs of tax exemptions given to specific sectors or companies but does not estimate the total cost of tax expenditures. The World Bank undertook a study in 2017 which provided estimates of some tax expenditures: for example: 1.8 percent of GDP on corporate income tax and 3.1 percent on VAT. Current reporting practices on tax arrears do not comply with the constitutional requirement to publish all tax waivers or the PFM Act’s requirement for publishing an annual report on these exemptions and concessions.

19. The lack of a formal methodology for estimating tax expenditures is one of the key challenges. Capacity would need to be developed in the KRA and the NT, which still has not created a unit responsible for tax policy issues. East AFRITAC had previously provided training on calculating tax expenditures but most of the government officials involved have left and capacity in this area is now low. Box 1.2 below provides guidance on the methodology for calculating tax expenditures. The process of developing tax expenditure reports is being initiated.

Tax Expenditures

Tax expenditures may be defined as provisions of tax law, regulations, or practices that reduce or postpone revenue for certain taxpayers relative to a benchmark tax. Governments are sometimes attracted to tax expenditures because they reduce revenue rather than the alternative of increasing spending (and so appear not to add to the size of government). On the other hand, tax expenditures can reduce the transparency, efficiency, and equity of the fiscal system.

Tax expenditures usually are given legal status by amendment to revenue laws. They can take several forms:

  • Allowances: amounts deducted from the benchmark to arrive at the tax base;

  • Exemptions: amounts excluded from the tax base;

  • Rate relief a reduced rate of tax applied to a defined class of taxpayer or taxable transactions;

  • Tax deferral: a delay in the requirement to pay tax; and

  • Credits: amounts deducted from the tax liability.

While the underlying concept of the cost of tax expenditures is ‘revenue foregone,’ the methodologies for the measurement of tax expenditures are a matter of debate and discussion, especially regarding the definition of the tax base and measuring the expected behavioral responses of taxpayers in the absence of the tax expenditure.

Tax expenditures are typically less transparent and subject to less scrutiny than conventional spending, and overall reduce the integrity and quality of the budget allocation process. For that reason, consideration should be given to converting tax expenditures into conventional spending where possible. Where tax expenditures continue to exist, they should be subject to rigorous scrutiny in budget decision-making processes alongside other spending proposals, and the costs and extent of the set of existing tax expenditures should be made transparent through publication of these data as part of the annual budget documentation.

Several countries—for example, Australia, France and the UK—publish detailed information on tax expenditures.

Sources: OECD, Tax Expenditures in OECD Countries. 2010; Healy, C. and M. Mansour, “Tax Expenditure Reporting and its Use in Fiscal Management: A Guide for Developing Countries”. IMF, Fiscal Affairs Department, How to Notes, March 2019.

20. Significant reductions in the number of exemptions and waivers were noted in the 2014 FTE but since then there has been a significant increase. The number of exempted and zero-rated items has more than doubled since 2013 (Table 1.7). This growth emphasizes the importance of transparent disclosures of tax expenditures, especially considering the underperformance of revenues in recent years and the growing pressures on controlling the budget deficit. The KRA cites tax expenditures as a major impediment to its ability to meet revenue targets, with tax expenditures estimated at Ksh 478 billion in 2017 against a shortfall in budgeted collections of about Ksh 300 billion. Tax expenditures are relatively high compared to other sampled countries (Figure 1.6).

Table 1.7.

Kenya: Number of Tax Exempted and Zero-Rated Items

article image
Source: Kenya Revenue Authority.
Figure 1.6.
Figure 1.6.

Annual Revenue Loss from Tax Expenditure in Selected Countries

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: IMF Staff. Latest data available: 2018 (Italy), 2017 (Lithuania, France, Netherland, Portugal, Romania, Spain, UK), 2016 (Estonia, Ireland), 2015 (Colombia), 2014 (Poland, Uganda), 2012 (Mozambique).

1.2. Frequency and Timeliness of Fiscal Reporting

1.2.1. Frequency of In-Year Fiscal Reporting (Basic, Unchanged from 2014)

21. Kenya produces regular in-year fiscal reports which are submitted to the National Assembly and published. The NT produces a QEBR which is published on the NT’s website within 45 days of the end of each quarter. The Controller of Budget (CoB) publishes within two months of the end of each quarter a Budget Implementation Review Report (BIRR) for the national government and a CBIRR for the 47 county governments. The Financial Regulations require the Accounting Officer of each public sector entity to submit its monthly financial reports by day 10 of the following month to Treasury with copies to the CoB and the Auditor General.4 These reports are consolidated in the production of the quarterly reports. Monthly statements are produced on the actual receipts into and disbursements from the Exchequer Account, but these are not published.5

22. There are some technical constraints that prevent the publication of in-year reports within one month. The process is based on the manual submission of reports from the public entities, which then need to be consolidated physically. For example, the CoB stated that some counties undertake transactions outside of the IFMIS, and manual processing sometimes result in delays in receiving their statements. This points to a need for further capacity development—no transactions should be undertaken outside of IFMIS. If the NT and the CoB had confidence that the data in IFMIS were fully comprehensive, they would be able to consolidate the entities’ data through the system. However, analysis of the data would still be required, which takes some time.

1.2.2 Timeliness of Annual Financial Statements (Good, Increased From 2014)

23. The law requires that audits of the annual financial statements of individual ministries, departments and agencies (MDAs) be completed within six months of the financial year-end.6 The MDAs are required to prepare their annual financial statements in accordance with the standards prescribed by the PSASB and submit them to the Office of the Auditor General (OAG) within three months of the financial year-end. The OAG publishes the audit report and accompanying financial statements on its website. The Treasury also produces a “Statement of Receipts into and Issues from the Exchequer Account” which is audited but not published.

24. Since the previous FTE, the NT now produces consolidated financial statements the most recent of which were published within eight months of the end of the financial year. 7

These statements are presented and published as “unaudited” documents but the OAG is currently in the process of auditing the accounts for the past three financial years. Once this audit is completed it will represent the first publication of Kenya’s audited consolidated financial statements. The final audit report of the Auditor General is submitted to the National Assembly or the relevant County Assembly, who must within three months debate the report. The NT prepares a Treasury Memorandum, providing management’s responses to the audit findings.

Consideration could be given to publishing both the audit opinion and management’s responses with the consolidated financial statements.

1.3. Quality of Fiscal Reports

1.3.1. Classification (Good, Increased from 2014)

25. Kenya’s fiscal reports are based on a standard chart of accounts, and transactions are analyzed by administrative, economic and functional classifications. The Program Based Budget Book also presents the budget in a program format along with key performance indicators, but no budget execution reports by program are published, a deficiency noted by the PBO.8 No information is published on changes in the performance of delivering public services by program and the associated changes in budget allocations for these programs. The consolidated financial statements and QEBRs present data with a breakdown by administrative and economic classifications. The KNBS publishes an annual Economic Survey, which since 2014 now presents tables showing the analysis of government expenditures in accordance with the UN’s COFOG standards.

26. A standard chart of accounts is used for general government budgeting and reporting across national government and the counties. The chart is multi-dimensional and includes separate segments for administrative units, economic classification, programs, functional classification (mapped to programs), and sources of funds. The chart of accounts is currently being redeveloped to support the move to accrual accounting, whilst also supporting cash-based budgeting. The wider rollout of the IFMIS which incorporates the standard chart of accounts has improved consistency in its use.

27. Identifying the ultimate nature of spending from intra-governmental transfers remains a challenge. The different segments of government are currently consolidated separately. There is no consolidation of the whole of government. Under the accrual accounting reforms, financial reports will cover the whole of government broken down by segment in accordance with GFSM standards, i.e., the BCG; central government; county governments; general government; and public corporations.9

1.3.2. Internal Consistency (Basic, Unchanged from 2014)

28. The FTE Code requires that fiscal reports be reconciled on three summary fiscal aggregates. Kenya largely reconciles on one consistency check and good progress is being made on reconciling the other two. The CBK’s statistical bulletin provides information on debt and debt holdings by residence of the holder and a breakdown of debt by instrument. The 2014 FTE highlighted large internal inconsistencies. Figure 1.7 indicates that since 2014, efforts to narrow the gaps are being made especially on reconciling net lending/borrowing with financing although the discrepancy still exceeds 0.3 percent of GDP.

Figure 1.7.
Figure 1.7.

Discrepancy Between Budget Balance and Net Financing

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: Economic Survey of the KNBS, IMF staff estimates.

29. Kenya’s stock-flow adjustments need further reconciliation. The previous FTE pointed out the large inconsistencies on these adjustments. Figure 1.8 shows that whereas efforts have been made to narrow the gap on stock-flow adjustments since 2014—which have reduced from approximately 4.2 percent of GDP in 2015–16 to 0.9 percent of GDP in 2017–18—there remain discrepancies between changes in government debt and the fiscal deficit/surplus which need to be explained and disclosed. Some of these discrepancies are due to below-the-line transactions, identified from various reports. On-lending raises cash requirements without affecting the deficit and expanded significantly in 2015–16 and 2016–17. Exchange rate changes affect the value of debt denominated in foreign currencies and played an important role in in 2010–11 and 2011–12. Changes in cash balances have also led to significant discrepancies, notably since 2013–14, which highlight issues relating to the transfer of proceeds from the issuance of Eurobonds to offshore bank accounts in 2013–14, transactions that were the subject of a special audit.

Figure 1.8.
Figure 1.8.

Stock-Flow Adjustments of General Government

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: NT, Economic Survey of the KNBS, Statistics Bulletin of CBK, IMF staff estimates.

30. There is also a need to reconcile the disaggregated debt information reported by the NT with the aggregate public debt data shared through fiscal reports. Data in the KNBS’s annual fiscal report is not consistent with fiscal information reports published by the NT. These discrepancies may be attributed to delays in information sharing between the two government agencies.

1.3.3. Historical Revisions (Not Met, Unchanged from 2014)

31. There is need to improve on the reporting of revisions to data presented in fiscal and financial reports, with a public explanation of why these revisions have been made.

The KNBS’s annual Economic Survey includes fiscal tables aligned with the GFSM 2014, but no revisions are made to the historical data. The GFS Technical Working Group has developed a draft fiscal and debt data revision policy which will provide guidelines on data revisions.10 This policy will provide users of fiscal reports with greater certainty on the revisions that might take place. There is currently no fiscal data revision calendar on the website of the KNBS. When major historical revisions are made, it is important that they are reported with an explanation, in accordance with international standards.11 Kenya rebased its GDP in 2015 resulting in an approximate 25 percent increase. Although the historical data series for GDP were re-published, there was not a comprehensive revision of historical series for fiscal data to facilitate the consistency of fiscal reports.

32. Revisions to historical fiscal statistics may be significant in Kenya. Kenya has produced consolidated financial reports from FY 2014–15 to FY 2017–18 and have sizeable amounts indicated as prior-year adjustments to the data for payables, receivables and deposits. For BCG such adjustments have been as large as 1.2 percent of GDP (in 2015–16), though smaller for county governments (Figure 1.9). Once audited, these adjustments should be incorporated in fiscal reports as revisions. It is important that the KNBS produce clear guidelines on the treatment of historical revisions in Kenya.

Figure 1.9.
Figure 1.9.

Prior-Year Adjustments to Consolidated Reports

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: Consolidated financial reports of Government of Kenya, IMF staff estimates.

1.4. Integrity of Fiscal Reports

1.4.1. Statistical Integrity (Advanced, Unchanged from 2014)

33. Fiscal statistics are compiled by an independent professional body, the KNBS, and disseminated in accordance with international standards, namely the IMF’s Enhanced General Data Dissemination System (e-GDDS). The KNBS national data summary page includes a list of general government fiscal reports and central government gross debt.

34. The QEBR produced by the NT is validated through the GFS Technical Working Group. The KNBS has an independent status with responsibilities as the “only competent institution” for approving the methods for the compilation and dissemination of statistical information.12 The GFS Technical Working Group reviews the institutional classifications and scrutinizes the quality of fiscal and other statistics. Its establishment has had a large impact on improving the quality and integrity of fiscal reports since the last FTE carried out in 2014.

1.4.2. External Audit (Basic, Unchanged from 2014)

35. The OAG undertakes audits and provides an audit opinion on the financial statements of all public entities. The Auditor General draws his mandate from the Constitution of Kenya, which provides for the independence of the Office. The Public Audit Act 2015 provides for the functions and powers of the Office. The OAG undertakes audits of national government, including extrabudgetary units, county governments, public funds, statutory bodies, commissions and independent offices, public corporations, and projects. It publishes the individual audit reports on its website and also publishes an annual report which provides an overview of all audits undertaken along with selected findings. The OAG undertakes its audits in compliance with the International Standards of Supreme Audit Institutions (ISSAIs).

36. The OAG’s audit opinions on the financial statements of the national government have improved over the years. An unqualified opinion signifies that the statements are free from misstatement and present a true and fair view. Varying degrees of material errors may attract a qualified or adverse opinion, or a disclaimer, reflecting increasing materiality of deficiency. Table 1.8 shows that of 129 national government entities audited over the past four years, 82 percent attracted either an unqualified or qualified opinion and 18 percent attracted either an adverse opinion or a disclaimer. Some slight improvement in the opinions given may be noted over this period.

Table 1.8.

Audit Opinions on Government Financial Statements

article image
Source: Office of the Auditor General.

37. The OAG is currently working on the audit of the consolidated sets of financial statements for the fiscal years 2014–15 to 2017–18. Receiving an audit opinion on the consolidated financial statements and appending this to the statements, along with a management response, would significantly enhance the transparency of reporting. Users of the reports could then be assured of the reliability of the data contained therein.

1.4.3. Comparability of Fiscal Data (Basic, Unchanged from 2014)

38. The QEBR, the BIRR produced by the NT and the Economic Survey compiled by KNBS are prepared on a comparable basis to the budget, but no reconciliation of data is made between successive reports. There is need to reconcile data in the QEBR, the consolidated financial reports and the Economic Survey since they are based on outturns. The primary statements in the annual financial reports and the GFS reports are prepared on a GFS economic item basis, which is a different basis to the budget. Separate consolidated financial statements are prepared for the Consolidated Fund and the operations of the MDAs. The consolidated statements, which were not produced when the 2014 FTE was undertaken, now present basic fiscal aggregates which can be compared to the data disclosed in other fiscal reports. Once the accrual-based IPSAS is implemented, one of the key documents would be the “Statement of Actual Amounts on a Comparable Basis to the Budget”. This would enable a clearer comparison of data from the financial statements and the budget. There would also be merit in publishing a table presenting the budget on a GFS economic item basis.

39. There are large discrepancies in the aggregate fiscal data published in the consolidated financial statements and fiscal statistics. As shown in Figure 1.10, the difference between the general government net borrowing reported in the annual GFS table (8.6 percent of GDP) and the balance reported in the consolidated financial statements (6.5 percent of GDP) was equivalent to 2.1 percent of GDP in 2017–18. This is partly because the consolidated financial statements do not include the net lending of social security funds in addition to differences in coverage. The consolidated financial reports reflected an adjustment related to financing of 2.3 percent of GDP as a balancing item in the financial reports. This discrepancy is expected to be resolved in the coming years as the authorities strengthen financial reporting. Since consolidated financial statements and fiscal statistics differ in their coverage and classification, it is important to publish explanations and reconciliations of these differences.

Figure 1.10.
Figure 1.10.

Kenya: Reconciliation of General Government Net Borrowing and Consolidated Financial Statements Balances, 2017–18

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: IMF staff calculations. Gray bars show explained differences in net borrowing figures between reports.

1.5. Recommendations

40. Based on the above assessment, the evaluation highlights the following priorities for improving the transparency of fiscal reporting:

  • Recommendation 1.1. Enhance presentation and transparency of fiscal reports:

    • Separate the reporting of EBUs and PCs in the consolidated financial statements;

    • Consolidate financial reports according to the GFS classification of the sectors of government, with a consolidation for the whole of the public sector;

    • Publish GFS tables covering the entire public sector, including stocks;

    • Publish the OAG’s audit opinion and management responses in the consolidated financial statements.

  • Recommendation 1.2. Harmonize fiscal reporting to ensure consistency between and within fiscal reports:

    • Use the IFMIS as the primary source of data for general government transactions;

    • Produce numeric statements directly from the IFMIS for consolidated statements, QEBRs, and BIRRs;

    • Publish reconciliation statements of the differences between the fiscal aggregates published in various fiscal reports.

  • Recommendation 1.3. Initiate tax expenditure reporting:

    • Develop methodology and guidelines for compiling tax expenditures, based on international good practice;

    • Build capacity on the compilation of tax expenditures in the NT and the KRA;

    • Publish an annual report on the impact of tax expenditures and their budgetary implications, highlighting the costs of new exemptions, waivers, reliefs, and allowances.

  • Recommendation 1.4. Progressively expand the balance sheet coverage of fiscal reports:

    • Disclose all national and subnational government nonfinancial assets and liabilities in the notes to the consolidated financial statements, and progressively update the statement of the financial position onto an accrual basis;

    • Produce an aggregate financial balance sheet of the sectors of government as an annex to the annual GFS table;

    • Finalize the policy guidelines on the valuation of assets and liabilities and commence the valuation of all stocks;

    • Produce an estimate of the public sector balance sheet as an annex to the annual GFS table.

Table 1.9.

Summary Evaluation: Fiscal Reporting

article image

II. FISCAL FORECASTING AND BUDGETING

41. 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. It is important that fiscal forecasts and budgets:

  • Are based on credible projections of macroeconomic developments;

  • Provide comprehensive information on the government’s fiscal objectives and budgetary plans, facilitate policy analysis and accountability;

  • The budget calendar gives the legislature enough time to scrutinize and approve the plans before the budget year begins.

42. Kenya continues to perform well in the overall transparency of its fiscal forecasting and budgeting practices, which is based on a strong legal framework. It does so against a backdrop of significant ongoing reforms, including far-reaching fiscal devolution to counties, and the introduction of performance-based budgeting. However, ongoing issues include the poor credibility of Kenya’s fiscal forecasts, which have been affected in recent years by significant revenue optimism (annual overestimates of about 8 percent), leading to repeated fiscal slippages and in-year spending curtailments. Substantial in-year spending reallocations through supplementary budgets undermine the credibility of the budget preparation process. Poor project appraisal and weaknesses in budget allocation has contributed to a large number of stalled and delayed public investment projects and cost overruns.

43. Since 2014, the rating of one principle has been increased, two reduced, while nine remain unchanged (Table 2.1). The increase in the use and reporting of performance targets, particularly through the government’s Performance Based Budgeting and Sector Reports, leads to an upgrade from ‘Good’ to ‘Advanced’ in the rating of the Performance Information principle. The lack of reporting of gross revenues, expenditures and financing of EBUs has led to a downgrade in the rating of the Budget Unity principle to ‘Basic’. Weaknesses in the government’s reporting against its Fiscal Responsibility Principles undermine the transparency and credibility of fiscal policy. This has led to a reduction in the rating of this principle from ‘Basic’ to ‘Not met’.

Table 2.1.

Kenya: Fiscal Forecasting and Budgeting: Summary of Changes since 2014

article image

44. There has been mixed progress in implementing the recommendations of the 2014 FTE (Table 2.2). Positive reform measures have been implemented in PIM reforms, although more can be done in this area. Issues of budget unity have improved with the integration of EBUs in the budget. The specification and implementation of the Fiscal Responsibility Principles has not prevented fiscal slippage, and a lack of a decomposition of macroeconomic and fiscal forecast variations hinders a proper assessment of the sources of forecast error.

Table 2.2.

Fiscal Forecasting and Budgeting: Progress on Recommendations since 2014

article image
Source: IMF staff.

45. Building on this progress, the authorities should prioritize investment reforms, and strengthen its capacity to assess sources of forecast variation, particularly in revenue projections. In addition, the latest assessment highlights the need to clarify the limits, measurement and compliance with the government’s debt rule; increase public scrutiny and improve the presentation of supplementary budgets; fully operationalize the public investment management (PIM) unit and relocate it within the NT; base the budget on more realistic revenue projections; and improve the current forecast reconciliation table by including the decomposition by source of forecast variations.

2.1. Comprehensiveness

2.1.1. Budget Unity (Basic, Reduced from 2014)

46. The annual budget provides a good coverage of the main fiscal operations of the central government. The expenditure, tax and financing of central government MDAs are covered in the annual Budget Estimates, as are approximately 90 percent of EBUs, the number of which has expanded significantly in recent years.

47. Expenditures and revenues of budgetary central government are produced on a gross basis. Financing of expenditures distinguishes between domestic revenues, loans and grants, and own source revenues. Grants to EBUs from the central government are netted out during the reporting of consolidated financial statements.

48. The size of central government is understated in the fiscal aggregate figures. As noted in the 2014 FTE, the Budget Policy Statement (BPS) does not contain the revenues, expenditure or financing of EBUs. The Annex to the BPS and budget estimates present separate numbers, but these are not easily integrated due to consolidation issues. It is also difficult to assess whether there are inconsistencies between transfers to EBUs and the receipts of those transfers, and therefore whether grants are being channeled directly to EBUs without being included in budget estimates. Fig 2.1 shows that EBUs are significant in Kenya relative to other countries that have recently undergone FTEs, illustrating the importance of these issues. The lack of incorporation of gross EBU figures into budget documentation results in Kenya meeting only basic practice.

Figure 2.1.
Figure 2.1.

EBU Expenditure in Selected Countries (percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: IMF staff calculations

2.1.2. Macroeconomic Forecasts (Good, Unchanged from 2014)

49. The budget documentation includes medium-term forecasts of key macroeconomic variables. The Budget Review and Outlook Paper (BROP) and BPS present forecasts of key macroeconomic variables, such as real GDP growth, the GDP deflator, and the external balance, for the budget year and two outer years.13 These data are supported by an in-depth quantitative analysis of recent economic and fiscal performance.

50. Real GDP forecast errors have demonstrated an optimistic bias since 2010. The underperformance of real GDP relative to forecasts is larger for the outer-year projections (Figure 2.2). Forecasts for nominal GDP show no systematic bias for the year of the forecast, although some pessimism bias for the outer years (for further discussion, see Principle 2.1.3). The biases in opposite directions appear to reflect under-forecasting of the GDP deflator, particularly in outer years.

Figure 2.2.
Figure 2.2.

Average Forecast Errors of Real and Nominal GDP Growth 2010–18

(percentage point)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: NT, IMF staff calculations.Note: T = the fiscal year in which the projections are made; T + 1 = the first year following the fiscal year; T+2 = the second year following the fiscal year.

51. A more in-depth explanation of the underlying assumptions of the macroeconomic forecasts and their components would increase the transparency and plausibility of the forecasts. The underlying assumptions of the outlook are relatively high level—for example, an increase in real GDP growth is justified by “a pickup in manufacturing and agricultural activities.”14 The economic forecasts would benefit from an explanation of the assumed growth in specific sectors of the economy (e.g., agriculture, industry and services), as well as the assumed level of employment and wages. Such information would better highlight risks as well as supporting discussions on relevant policies that may affect them and assist the PBO in its assessment of the realism of the forecasts.

2.1.3. Medium-Term Budget Framework (Advanced, Unchanged from 2014)

52. The BPS presents fiscal, revenue and expenditure projections over the medium term. This documentation shows the outturns of the previous year as well as projections for the current budget year and two outer years.15 A detailed breakdown of expenditure ceilings is provided by sector, program, and shows the split between recurrent and capital spending. The BPS also details the main priorities for spending within each sector. Estimates present the medium-term figures on the same basis as the annual budget. A breakdown of revenue by major tax head is provided, though there is comparatively little discussion.

53. The fiscal forecasts would particularly benefit from an explanation of the assumptions underpinning revenue growth. Revenue forecasts have been consistently and significantly optimistic, with a striking increase in over-optimism since 2012–13 (Figure 2.3). This creates numerous challenges in the PFM system, reducing fiscal transparency and the credibility of the budget (Box 2.1). There is little discussion of the assumed link between growth and revenue, factors that may affect revenue growth of individual revenue heads, or whether new policies may be required to meet the projections.

Figure 2.3.
Figure 2.3.

Aggregate Revenue Outturn vs. Forecasts

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: NT, IMF staff calculations.

Revenue Forecasts and the PFM System

Optimistic revenue projections leading to in-year revenue shortfalls have become commonplace. Revenue has underperformed against the original budget by an average of 8 per cent between 2010-2018. This has contributed to the need for in-year spending reductions relative to the budget over the course of the year. Over the same period, expenditure was on average 4 percent lower than budgeted for at the start of the year.

These errors create numerous problems throughout the PFM system by:

  • Reducing budget credibility—as some areas of spending will be curtailed throughout the year, the budget is no longer an accurate guide to the spending priorities of the government.

  • Disrupting services and project delivery—implementing in-year budget cuts undermines careful budget preparation, contributing to inefficiencies, delayed and stalled projects, and disruptions to services.

  • Mounting pending bills—cash rationing against commitments leads to expenditure arrears, and can affect government suppliers, the integrity of fiscal reports, and future budgets.

  • Creating fiscal slippage—where spending reductions do not take all the strain of the revenue shortfall, then borrowing must increase, leading to higher projected/targeted in the budget.

Institutional dynamics can contribute to revenue optimism. Annual revenue targets set by the NT are typically higher than the KRA’s internal forecasts. This is seen to strengthen KRA’s revenue collection efforts, reducing inefficiencies in revenue administration. Adopting these stretching targets as revenue projections creates the illusion of a larger spending envelope than what is likely to be available. This allows more projects and activities to be financed in the budget than what might be realistic, deferring hard spending prioritization decisions to later in the budget cycle, with disruptive consequences.

There is always inherent uncertainty in revenue projections, but setting expenditures based on the full realization of a stretching target—which has consistently fallen short in the past—will amplify risks to the implementation of the budget. Risk mitigation would require making a clear distinction between stretching targets and realistic projections and could involve focusing on other incentives structures for increasing revenue collection or budgeting a proportion of spending contingent on the realization of revenues at a defined point in the year. It would also benefit from a clear assessment of how much of historic revenue shortfall has been due to macroeconomic factors, policy changes, or other technical/one off factors, set out in a forecast reconciliation table (as discussed in Principle 2.4.2).

Source: IMF Staff

54. Revenue shortfalls have emerged despite overperformance of nominal GDP relative to forecasts. The revenue forecast errors are not well correlated with the nominal GDP forecast errors, with many falling outside the expected range with an elasticity of revenue to growth of between 1–1.5 (Figure 2.4). This has particularly been the case in recent years, where revenue growth has been more likely to under-perform nominal GDP growth, indicating possible structural issues with revenue forecasting models and assumptions.

Figure 2.4.
Figure 2.4.

Budget Forecast Errors on Nominal GDP and Revenue Growth

(percent forecast error on T+1)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: NT, IMF staff calculations.

55. The Medium-Term Budget Framework (MTBF) provides a guide to changing sector priorities, although aggregate spending ceilings have not established a strong anchor for future years’ expenditure. While the reporting of expenditure ceilings for the outer two years merits an advanced rating, the spending projections themselves do not yet provide a credible binding framework for setting medium-term spending priorities, as shown in Figure 2.5. Between 2010 and 2018, actual expenditure was on average 5 percent higher than projected over the medium term.

Figure 2.5.
Figure 2.5.

Aggregate Expenditure Outturn vs. Forecasts

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: National Treasury

2.1.4. Investment Projects (Basic, Unchanged from 2014)

56. Progress has been made on establishing a PIM unit in the NT, but it is not yet fully operational nor having a significant impact on the budget.16 The human capacity, IT systems and analytical tools used by the unit are still being developed, and the staffing structure, and the location of the unit within the NT is under consideration by senior management.

57. The budget documents (Annex 7) disclose multi-annual contractual commitments for investment projects over a three-year period. The Budget call circular—issued at the start of each budget process—requests MDAs to update this information to reflect any changes to the cost estimates and include outstanding arrears and counterpart funding requirements.

58. Not all major projects are subject to a cost-benefit analysis, nor competitive tendering. The 2018 Public Investment Management Assessment (PIMA) carried out by the IMF found that many project appraisals do not include either a cost-benefit analysis nor a standard methodology for project appraisal and selection.17 In addition, several large government-to-government (G2G) contractual arrangements, such as the Nairobi-Mombasa SGR project, operate outside the standard procurement process. Since 2018, the authorities have been taking active steps to address these weaknesses through the establishment of the PIM unit and the production of draft PIM regulations, which include standardized templates for preparing a project’s conceptual design, and pre-feasibility and feasibility studies. These reform measures and any changes to procurement arrangements, however, have yet to be fully implemented.18

59. The lack of an effective gatekeeping function has allowed many new projects to enter the budget, which has created challenges to finance ongoing projects. The rapid increase in public investment since 2010 (Figure 2.6) occurred without enough screening for project viability and readiness before they entered the budget and the stock of projects under implementation is now estimated at approximately 1,000. There has been a subsequent squeeze on ongoing projects in the absence of fiscal space (Figure 2.7), which is now accruing large costs to the government.19 First, pending bills have been accumulating, particularly for the State Department of Infrastructure, with an amount estimated amount of Ksh. 78 billion in June 2019 (0.1 percent of GDP). Second, the number of stalled projects is increasing, and is currently estimated at approximately 500 (half of all ongoing projects), because of non-payment to contractors, insufficient allocation of funds to projects, and litigation cases in court.20 Expenditure estimated at Ksh.1 trillion (12 percent of GDP) is required to finalize these projects.

Figure 2.6.
Figure 2.6.

Public Investment, Kenya vs. Comparator Countries

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: IMF PIMA Database, NT and IMF World Economic Outlook (WEO) database. Comparator countries are Cameroon, Ghana, Nigeria, Senegal, Tanzania, Uganda and Zambia.
Figure 2.7.
Figure 2.7.

Expenditures, Revenues and the Deficit

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

2.2. Orderliness

2.2.1. Fiscal Legislation (Advanced, Unchanged from 2014)

60. A comprehensive legal and regulatory framework for the collection and use of public resources, and for budget preparation is provided in the PFM Act 2012 and the Constitution. The PFM Act sets out the timetable for budget preparation and approval, the content requirements for budget documentation, and the respective role and responsibilities of the authorities and the National Assembly. Regulations set out in the 2015 Legal Notice No.34 made amendments to the PFM Act, including clarification of fiscal responsibility principles, and procedures if the Appropriations Bill is not passed.

61. An important recent amendment to the budget calendar synchronizes the submission and approval of the annual Finance Bill to coincide with the Appropriations Bill. Currently, spending estimates are submitted by April 30 each year and the Appropriations Bill must be approved by June 30. By contrast, revenue estimates are submitted two months later, by June 30, and the Finance Bill must be approved by September 30. This means that spending plans are approved prior to revenue plans being scrutinized, amended and approved by the legislature. In practice, this misalignment allows revenue measures to vary from what was assumed for the Appropriations Bill, leading to potential revenue gaps in the budget.21 A recently approved law will align the timing of the two Bills for the 2020/21 budget cycle, helping to ensure that the budget is based on consistent spending and revenue plans.22

2.2.2. Timeliness of Budget Documents (Good, Unchanged from 2014)

62. The budget is presented to the National Assembly two months ahead of the start of the fiscal year, and estimates are approved and published before the beginning of the fiscal year. These deadlines are as required by the PFM Act where estimates of recurrent and development expenditure are published two months before the start of the financial year (Table 2.3).

Table 2.3.

Kenya: Key Dates in the Budget Calendar (FY: July 1–June 30)

article image
Source: PFM Act Section 35–45.

63. Discussions are underway to provide an extended period for the National Assembly to review the BPS. The PBO has recommended that more time be made available to the National Assembly for debate and amendments through a change to the PFM Act.

2.3. Policy Orientation

2.3.1. Fiscal Policy Objectives (Not Met, Reduced from 2014)

64. The Fiscal Responsibility Principles for the national government were set in the PFM Act 2012, with further elaboration in a Legal Notice in 2015 (LN 35 25). This Legal Notice also sets out similar fiscal responsibility principles for county governments and includes a numerical ceiling on public debt. The PFM Act provides for an escape clause in the event of a major natural disaster or some other significant economic shock or unforeseen event. The principles can be summarized as follows:

  • A minimum of 30 percent of national and county government’s budgets shall be allocated to development expenditure.

  • The national government’s expenditure on wages shall not exceed 35 percent of annual revenue.23

  • Over the medium term, the national government’s borrowings shall be used only for financing development expenditure and not for recurrent expenditure.

  • Public debt and obligations must be maintained at a sustainable level, and national public debt shall not exceed 50 percent of GDP in net present value (NPV) terms.

65. The transparency and credibility of the debt sustainability principle is undermined by conflicting reports of the level of debt and the acceptable limit. While the 2015 Legal Notice sets a maximum debt level of 50 percent of GDP in NPV terms, budget documents instead assess debt to be sustainable—and therefore acceptable—if it is less than 70 percent of GDP in NPV terms. There is also no authoritative figure for the level of debt in a given year, making compliance with either debt limit difficult to assess. Figure 2.8 shows the different levels of reported debt in NPV terms for FY 2017–18. A lack of transparent methodology for the construction of the debt measure, which depends heavily on the discount rate chosen by the NT, also hampers the operational credibility of the debt principle.

Figure 2.8.
Figure 2.8.

Reported Debt Levels 2017–18 and Thresholds

(Percent of GDP in NPV terms)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: NT, Parliamentary Budget Office.

66. The above-mentioned principles are insufficient to anchor fiscal policy in the medium term. The principles that limit the wage bill and promote development expenditure affect the composition of spending but do not bind the deficit or debt. As development expenditure is approximately 10 percent of GDP, the borrowing principle only prevents the government from running a very loose fiscal policy. While the government is uncomfortably close to the legally-imposed 50 percent limit on debt as a share of GDP in NPV terms, it is still some distance from the 70 percent GDP debt threshold which it holds itself accountable to, thus not providing a meaningful constraint on fiscal policy in the medium-term. Simple projections suggest that it could be possible to run a deficit of 12 percent of GDP over the period of the Medium-Term Fiscal Framework (2019–20 to 2022–23) without necessarily breaching the 70 percent limit. The government has also adopted the fiscal rules in the East African Monetary Union (EAMU) convergence criteria, and provides annual submissions on compliance to the EAC secretariat.24

2.3.2. Performance Information (Advanced, Increased from 2014)

67. Kenya prepares an ambitious array of development plans that guide expenditure decisions, and which are monitored and updated periodically, and make substantial use of performance information. The government has launched its third medium-term development plan (2018–2022) which is designed to deliver the broad strategic objectives—focusing on the country’s ‘Big Four’ policy agenda—set out in the Kenya Vision 2030.25 Sector Plans for all MDAs as well as County Integrated Development Plans (CIPDs) are used to support alignment with the priorities of the 2018–22 plan. A Comprehensive Public Expenditure Review was conducted in 2017, with the support of the World Bank, assessing how public expenditure addresses county and national level priorities, setting out how sector spending was allocated between 2013–14 and 2015–16, what was achieved, and in some sectors, the incidence of spending on different demographics.

68. Kenya uses program-based budgeting, and key performance indicators (KPIs) for each program are produced for each of the three years of the MTBF. These KPIs include a mix of output and outcome-based indicators. Program achievements are updated annually, and are reviewed on a multi-year basis, most recently through sector reports which assess achievements against targets for their respective programs (Table 2.4 shows an example from the education sector). Annual progress reports have started to be conducted by counties to review progress against performance targets set in their CIPDs. The government should ensure that these targets are used effectively, particularly in discussions between the NT and MDAs during the budget preparation process, to ensure harmonization of strategic priorities between national, county and sectoral plans, and budget allocations. The effectiveness of performance information is much reduced where targets and achievement indicators are divorced from budget allocation decisions, particularly for underperforming programs.

Table 2.4.

Examples of KPIs from the Education Sector

article image
Source: Government of Kenya, Education Sector Report, 2018.

2.3.3. Public Participation (Advanced, Unchanged from 2014)

69. Providing the public with a formal voice in budget deliberations is enshrined in the legal framework, mandated for both the national government and counties. Public hearings are held ahead of budget debates in the National Assembly, and cover the full range of sectors, such as health and agriculture, as well as wider topics such as the economic and fiscal outlook. Sector reports on the summaries of these public hearings are published. The BPS 2019 includes a detailed section (Annex, Table 7) on comments received from the public on the budget proposals and what action the government proposes to take in response. Similar arrangements are made to elicit responses from business groups and citizens to tax proposals included in the Finance Bill. There is more variability in the coverage and reporting of public participation at county level. Public engagement in the budget process is aided by the Budget Summary Highlights paper which is published alongside the Budget Statement.

2.4. Credibility

2.4.1. Independent Evaluation (Good, Unchanged from 2014)

70. The Parliamentary Budget Office (PBO) provides advice to the National Assembly, its committees, and the public on budgetary and economic issues. Established in 2007, the PBO produces independent assessments of the budget and its underpinning economic and fiscal forecasts, as well as spending and revenue estimates, through its ‘Unpacking’ and ‘Budget Options Paper’ reports.

71. The PBO assesses compliance with the government’s fiscal responsibility principles (see Section 2.3.1), and reviews the realism of the government’s macro-fiscal forecasts. While it has done so in the past, it is not currently producing its own forecasts to reflect the most likely central scenario. However, it does publish its own alternative economic policy scenarios, with their implications for the medium-term fiscal framework. The PBO’s Budget Options Paper assesses the government’s compliance with its fiscal rules, although, as discussed in Section 2.3.1, there is currently a lack of clarity in the government’s interpretation of the debt limit and how it should be measured. The PBO also produces topical analyses of special issues, offering analytical deep dives into areas such as stalled projects and the economic and fiscal impact of cartels.

72. The PBO should continue to grow its role in supporting budgetary scrutiny. The impressive range of PBO reports would further benefit from a stronger comparison of the government’s forecasts and those published by the PBO and other independent forecasters. as well as a clearer presentation of the assumptions underpinning the PBO’s forecasts. Such analysis would be supported by strengthening the PBO’s technical forecasting capabilities, thus providing a greater challenge to the NT’s forecasts. Given the growth in the budgets of county governments, the PBO could provide welcome support to county assemblies in improving their budgetary documents. Finally, the PBO could consider publishing its ‘unpacking’ reports which analyze the government’s supplementary budgets.

2.4.2. Supplementary Budget (Good, Unchanged from 2014)

73. The rules governing the need for supplementary budgets following variations to the approved budget are clearly set out in the legal framework. Any material variation to the approved budget requires authorization by the National Assembly. The Cabinet Secretary of the NT is authorized to increase program spending by up to 10 percent of the aggregate approved budget before requiring retrospective legislative approval within two months.26 This puts the legislature in a difficult position if the funds have already been disbursed. In practice, two supplementary budgets have become the norm over the last 5 years. These budgets incorporate many changes to the original budget.

74. Since 2014–15, the final revised estimates resulting from supplementary budgets have increased recurrent spending by an average of six percent of the original budget and have made no net change to development spending. Nonetheless, development spending has tended to be substantially under-executed relative to the budget estimates. Supplementary budgets for recurrent spending tend to push up appropriations relative to the original budget, but the final execution tends to fall back in line with the original budgeted levels approved by the National Assembly (Figure 2.9). This may raise issues of ex-post compliance with the Fiscal Responsibility Principles in some years, which require development spending to remain above 30 percent of total expenditure.

Figure 2.9.
Figure 2.9.

Supplementary, Budget Estimates and Spending (percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: NT and Controller of Budget. Figures exclude Appropriation in Aid (AiA).

75. However, these net movements mask larger gross budgetary reallocations, particularly in the development budget, which undermine the budget preparation process, project planning and implementation. While supplementary budgets should primarily be used to respond to unforeseen emerging circumstances, these reallocations may also reflect ongoing adjustments to the budget that was prepared at the start of the year. Changes to the development budget are most significant, with net changes masking larger gross reductions in some projects offset by gross increases in others (Figure 2.10). Gross increases and reductions to the development budget averaged 13 percent over the period. Such churn means that development spending ends up not bearing a good resemblance to what was intended in the original budget, reducing its transparency and credibility. Large changes to in-year budgetary allocations can also disrupt project planning and implementation. The recurrent budget suffered less churn, with gross increases averaging 9 percent, and gross reductions averaging 3 percent.

Figure 2.10.
Figure 2.10.

In-Year Reallocations in Development and Recurrent Budgets

(Change in Spending as a Share of the Original Budget)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: NT. Figures exclude Appropriation-in-Aid (AIA).

2.4.3. Forecast Reconciliation (Basic, Unchanged from 2014)

76. Budget documentation provides deviations of previous forecast vintages from actuals but there is limited discussion of the sources of variation. The statement of specific fiscal risks sets out deviations in the main macroeconomic and fiscal aggregates, including revenue, expenditure and financing. There is some discursive assessment of the major deviations, though this is at a very high level, and there is no formal or quantitative breakdown of how much of the variation is due to macroeconomic determinants as compared to new policies.

2.5. Recommendations

77. Based on the above assessment, the evaluation highlights the following priorities for improving the transparency of fiscal forecasts and budgets:

  • Recommendation 2.1. Reduce systematic forecast variations in revenues and seek to mitigate their impact on budget implementation. This could be supported by:

    • A quantified analysis of the sources of revenue variation from forecast, through a forecast reconciliation table in the BPS, with a clear decomposition into: a) new policy measures; b) macroeconomic factors; and c) other technical/accounting factors;

    • Clearer plans in the budget on how spending and borrowing plans will respond if revenue collection materializes significantly below targets at set points in the year.

    • Strengthen the role of the Parliamentary Budget Office in scrutinizing official revenue forecasts.

  • Recommendation 2.2. Fully operationalize the PIM unit, and consider its location in the NT’s organizational structure:

    • Ensure that the unit’s location maximizes its impact on budget decision making;

    • Develop relevant tools and templates for project prescreening, the estimation of budget baselines (for government-financed and externally-financed projects), accurate costing of multiyear commitments, and prioritizing stalled projects for inclusion in the budget.

  • Recommendation 2.3. Clarify compliance with the debt limit specified in the statutory Fiscal Responsibility Principles:

    • Produce a single authoritative measure of the level of debt as a percentage of GDP in NPV terms in the budget documentation;

    • Publish the methodology and assumptions underpinning the construction of the debt measure;

    • Publish compliance with the debt limit set out in the Fiscal Responsibility Principles in budget documentation and clarify its relationship with other assessments of fiscal sustainability.

  • Recommendation 2.4. Increase the transparency of supplementary budgets and focus on responding to unforeseen events:

    • Publish timely summaries of supplementary budgets, including major changes and implications for programs in the BPS;

    • Seek to reduce the size of reallocations within supplementary budgets to focus on genuinely emerging and unforeseen developments.

Table 2.5.

Kenya: Summary Evaluation: Fiscal Forecasting and Budgeting

article image

III. FISCAL RISKS

78. Governments should disclose, analyze, and manage risks to public finances and ensure effective coordination of fiscal decision-making across the public sector. This chapter assesses the quality of Kenya’s fiscal risk analysis, management and reporting practices against the standards set by three dimensions of the IMF’s Fiscal Transparency Code:

  • General arrangements for the disclosure and analysis of fiscal risks;

  • The management of risks arising from specific sources, such as government contingencies and guarantees, PPPs, and the financial sector; and

  • Coordination of fiscal relations and performances between central government, local governments, and PCs.

79. Kenya discloses and analyzes fiscal risks, but their quantification and comprehensiveness could be improved with emphasis on the most pressing specific risks. The government discloses and assesses many of the fiscal risks it faces in the Statement of Specific Fiscal Risks. Deeper analysis of the most serious risks—including PPPs, subnational governments and PCs—should be given priority given their importance for asset and liability management. In addition, several risks and mitigation measures, which are reported in standalone reports (outside of the NT), are not consolidated. These include financial sector exposure, natural resources, and environmental risks. In addition to strengthening its capacity to analyze risk and develop mitigation strategies, the NT should improve its collaboration with other MDAs that have lead responsibility for specific policy areas to gather data and better understand the risks involved.

80. Since 2014, the rating of two principles has been reduced, one has improved and nine have remained unchanged (Table 3.1). The lack of in-year reporting of how budget contingencies have been utilized and non-reporting of the fiscal cost of past natural disasters have led to downgrades for budget contingencies and environmental risks. Individual and consolidated reporting at the county level has improved since the last assessment, which now allows for cross-county assessments of the main fiscal indicators.

Table 3.1.

Fiscal Risk Analysis and Management: Summary of Changes since 2014

article image

81. Progress has been made on all but one of the recommendations made in the 2014 report (Table 3.2). Economic and fiscal developments since 2014 have seen the PPP pipeline grow to approximately 13 percent of GDP (and the increased use of non-financial PCs as a vehicle for infrastructure investment, which has led to a growing stock of their assets and liabilities (25 percent and 17 percent of GDP respectively). Strengthening risk analysis and reporting in these areas should be prioritized over the medium term. Several counties, including Nairobi and Mombasa had negative net assets in 2018, so require close monitoring. Continuing reforms in all these areas therefore should be a priority for the government.

Table 3.2.

Fiscal Risk Analysis and Management: Progress on Recommendations since 2014

article image

82. Building on this progress, the recommendations in the present report focus on strengthening fiscal risk reporting and institutional arrangements within the NT. Improvements in reporting should be aimed at: (i) better quantification of major specific fiscal risks; and (ii) the development of consolidated reports on PCs and environmental risks. Greater transparency in the presentation of projections of oil and non-oil natural resources and their fiscal implications is also encouraged. Supporting this, the NT could undertake a capacity assessment and functional review to identify where risk management practices could be strengthened across all the areas considered under this pillar of the Fiscal Transparency Code.

3.1. Fiscal Risk Disclosure and Analysis

3.1.1. Macroeconomic risks (Basic, Unchanged from 2014).

83. The Statement of Specific Fiscal Risks (SSFR) provides an assessment of the impact of deviations in selected economic variables on the fiscal position. Sensitivity analysis illustrates the impact of a one percentage point change in real growth, inflation, the exchange rate and the value of imported goods on expenditure, revenues and the fiscal balance, but does not show the impact of these changes on public debt. The SSFR presents a combined shock scenario but this analysis provides little explanation of the plausibility of the assumptions and the fiscal policy implications of the analysis and is of limited use for policy makers and the public.27

84. Macroeconomic volatility has been lower than most other countries in the region over the past decade (Figure 3.1). The relatively low revenue volatility suggests that there is significant room for improvement in addressing the revenue forecast bias discussed under Principle 2.1.3 (the Medium-Term Budget Framework).

Figure 3.1.
Figure 3.1.

Volatility of GDP and Revenue

(Percent, 2007–18)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: IMF WEO database, April 2019.Note: Volatility measured as the standard deviation of annual nominal growth rate from 2007–2018.

3.1.2. Specific Fiscal Risks (Basic, Unchanged from 2014)

85. The government’s SSFR mainly analyzes risks to public finances qualitatively and includes only a limited analysis of risks that are managed by MDAs outside the NT. The SSFR is included as a separate section of the BPS, which is submitted to Parliament and published annually in accordance with the provisions of the PFM Act. It discusses the potential exposure to various risks including macroeconomic risks, the sustainability of public debt, explicit and implicit contingent liabilities, and fiscal risks related to devolution. It also makes policy recommendations to help mitigate some of these risks. Risk analysis is undertaken in several other reports produced by the CBK and individual SOEs, but this work is not reflected in the SSFR.28

86. Kenya’s maximum gross exposure to specific fiscal risks is estimated at around 40 percent of GDP. The main exposures are the liabilities of PCs (17 percent of GDP), long-term fiscal pressures from the ageing of the population (9 percent of GDP), PPPs (8 percent of GDP), the bank deposit insurance fund (3 percent of GDP), legal claims against the government (0.3 percent of GDP) and environmental risks (0.1 percent of GDP). These risks are summarized in Table 3.3. Potentially significant risks associated with sustainability of the pension scheme (Section 3.1.3), PPP pipeline projects (Section 3.2.4), PCs (Section 3.3.2) and legal claims are not sufficiently reported on, whereas the bank deposit insurance fund (Section 3.2.5) and environmental risks (Section 3.2.7) are generally well covered in the SSFR.

Table 3.3.

Kenya: Selected Specific Fiscal Risks, Gross Exposure

article image
Source: IMF staff calculations.
1/

There is no publicly available information on the government’s exposure in these contracts so total project values are stated.

2/

Estimated as outstanding pending bills for land compensation reported by the State Department for Infrastructure. Many sources of legal claims (e.g., litigation on tax issues or procurement disputes) are not included.

87. The government should consider expanding the coverage of specific risks and improving their quantification. The SFFR covers seven of the eleven major risk categories identified as important internationally but does not provide estimates of their magnitude.29 The lack of quantification does not enable an assessment of the potential fiscal impact, nor help develop a prioritized risk mitigation strategy. The establishment of the Fiscal Commitments and Contingent Liabilities (FCCL) framework should allow the NT to play an important role in identifying and analyzing risks in public corporations (Section 3.3.2) and PPPs (Section 3.2.4) and their associated lending programs and guaranteed support, as well as the growing level of risk in subnational governments (Section 3.3.1). But the NT’s capacity may need to be expanded to analyze risks according to international standards and develop appropriate risk mitigation strategies.

3.1.3. Long-term Sustainability of Public Finances (Not Met, Unchanged from 2014)

88. The government does not produce or publish projections of the main fiscal aggregates or of the social security funds over the long term. Projections of the main fiscal aggregates (revenues, expenditures, deficit, and debt variables) are made over three years only. The NSSF undertakes actuarial projections of its liabilities for internal use, but these are not published. A World Bank study from 2016 estimated pension liabilities as 30 percent of GDP on an NPV basis.

89. Risks around long-term fiscal aggregates and sustainability are moderated by favorable demographic conditions. The lack of comprehensive long-term sustainability analysis by the authorities is not an immediate priority given Kenya’s relatively low demographic pressures. 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 3.2). Other pressures on fiscal sustainability could also be explored in this context, such as increases in health and education spending as Kenya migrates to upper middle-income status and drives towards the completion of the Sustainable Development Goals (SDGs).

Figure 3.2.
Figure 3.2.

Old Age Dependency Ratios (Age 65+ compared to 20–64 age population)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: United Nations Population Database, 2017 and IMF staff estimates.

3.2. Fiscal Risk Management

3.2.1. Budgetary Contingencies (Good, reduced from 2014)

90. Kenya’s budget includes an allocation for budgetary contingencies to deal with unexpected spending. The utilization of the Contingency Fund has fallen in recent years (Figure 3.3) and has been restricted to addressing emergency relief for drought- and flood-affected victims.30 However, data were not made available for 2014 and 2015.

Figure 3.3.
Figure 3.3.

Contingency Fund Spending

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: National Treasury.

91. The Contingency Fund has transparent access criteria governed by the 2010 Constitution and the 2012 PFM Act.31 These criteria include a detailed definition of cases in which the Fund can be used, how and when it should be replenished and reporting arrangements for its use (Table 3.4).

Table 3.4.

Access Criteria, Replenishment and Reporting for the Contingency Fund

article image
Sources: 2010 Constitution and 2012 PFM Act.

92. In-year reporting on the utilization of the Contingency Fund is not included in the NT’s QEBR or other in-year fiscal reports. This has led to a downgrade in the scoring of this principle.

3.2.2. Management of Assets and Liabilities (Basic, Unchanged for 2014)

93. Public debt is authorized by the PFM Act 2012, and the risks surrounding the public debt portfolio are regularly reported, albeit with some delay. The published Annual Public Debt Management Report (APDMR) and the Medium-Term Debt Management Strategy (MTDS) include a detailed analysis of the debt portfolio. The MTDS, for example, sets out an analysis of alternative debt strategies for lowering borrowing costs and mitigating currency, interest and refinancing risks. The APDMR highlights non-performing elements of the debt portfolio for guaranteed debt and on-lending arrangements, which are generally low.

Figure 3.4.
Figure 3.4.

Public Sector Financial Assets

(percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: IMF staff. Latest data: 2016 (Armenia, Georgia, Lithuania, Malta, North Macedonia), 2015 (Colombia, Uganda), 2014 (Brazil).

94. Whereas disclosure of public debt data is comprehensive, the reporting of other assets and liabilities is more limited. Since 2014 the balance sheet has grown significantly. Assets and liabilities are now recorded at 116 and 121 percent of GDP respectively (see Section 1.1.2). This is largely a result of an expansionary capital development program, but is also related to sustained expenditure pressures, such as growing pension liabilities and the rising cost of devolution. As a result, general government debt has risen to 57 percent of GDP, which is the highest in the EAC region, except for Burundi.32 This has led the debt servicing to revenue ratio to more than double since 2011 and the ratio is projected to reach 40 percent in 2019, a level typically only associated with countries at high risk of debt distress. Financial assets are valued at 26 percent of GDP, which is broadly comparable to other countries. Whilst improvements have been made to the reporting of various categories of assets and liabilities several areas could be strengthened further from a risk management perspective (Table 3.5).33 For example, the oversight and management of pending bills (expenditure arrears) is currently very weak. Box 3.1 describes a possible approach to managing these risks, which has been used in other countries of the region, including Uganda and Tanzania.

Table 3.5.

Kenya: Disclosure Gaps in Asset and Liability Management

article image
Source: IMF staff.

Expenditure Arrears

“Pending bills” is the term used by government to define payables which have not been paid at the end of the financial year. Under the current cash basis of accounting, neither the expenditure nor the liability are disclosed in the financial statements for these items. Pending bills include both current payables as well as expenditure arrears and it is important to clearly separate the two.

The current strategy for clearing expenditure arrears of BCG and counties rests on their being the first call on the budget. There is no in-year reporting or monitoring of arrears. The consolidated financial statements of public corporations and SAGAs do not show an ageing analysis of the aggregate payables, which is important in developing a strategy to manage them.

A strategy to improve the oversight, management and clearance of arrears could include the following elements:

  • A clear definition of arrears;

  • Prevention measures (e.g., improved commitment control, management of multi-annual commitments, and improved costing of the budget baseline);

  • Responsibilities for oversight and management;

  • Plausible measures for clearing the accumulated stock of arrears;

  • Reporting requirements on the consolidated stock of arrears, with analysis by: the entity incurring the arrears; the expenditure category of the arrears; and their vintage. Reporting should be undertaken on a quarterly basis and used as a basis for discussion and decision making.

3.2.3. Guarantees (Basic, Unchanged from 2014)

95. The stock of publicly guaranteed debt is disclosed in the government’s APDMR and the MTDS. Appendices to the Public Debt Report disclose a breakdown of guaranteed loans with analysis of: the borrowing entity; the year in which the loan was issued; the purpose of the loan; the creditor; and the balances over the past five years. Non-performing loans are disclosed along with the annual costs to the government for calls on the guarantees. As at June 2018, the total outstanding guaranteed debt stood at Ksh.141.4 billion (1.7 percent of GDP), of which Ksh.1.9 billion (1.3 percent of the total) was non-performing. During 2017/18, the cost to government on the defaults amounted to Ksh.1.4 billion.

96. The government has other explicit and implicit guarantees which are not quantified or disclosed. Examples of such guarantees include the National Social Security Fund, where the government is legally obligated to protect the members rights under the scheme in the event of failure of the Fund.35 Liabilities of the Fund stand at Ksh. 223 billion, approximately 3 percent of GDP. Similarly, the Deposit Protection Fund is underwritten by the government. However, there is no report analyzing the full stock of guarantees issued by the government including the inherent risks. Furthermore, there is no assessment (even qualitative) of exposure to implicit risks, e.g., where the government would be forced to step in and bail out an organization such as a county government or financial institution in the event of a severe economic or financial shock.

3.2.4. Public-Private Partnerships (Not met, Unchanged from 2014)

97. Kenya has a well-defined legal and institutional framework for the management of PPPs, but this is yet to become fully operational. The PPP Act 2013 provides the legal framework for PPPs and the new FCCL framework was adopted in 2018 to ensure that the potential liabilities to the government are identified and managed at all stages of a PPP project from inception to operation.36 The PPP unit is responsible for collecting, analyzing and disseminating information on contingent liabilities to ensure government commitments are sustainable. Whilst significant advances have been made on the institutional and legal framework for PPPs, the operational tools and capacity to implement them are still work in progress, and the PPP unit currently has limited capacity to undertake effective oversight.37

98. There is limited reporting on the potential risk exposure to government from existing PPPs. A list of 30 PPP contracts was presented in the 2019 BPS with a project value of 8 percent of GDP.38 Elements of termination risk, government guarantees, and annual fixed payments are stated, but several types of risk are missing.39 Previous versions of the BPS (for example, February 2016) included more explicit risk analysis, which would have scored a basic rating under the Code.

99. No risk analysis is undertaken for pipeline projects, which are sizable and growing in number. A list of 78 pipeline projects was published in the PPP Pipeline Status Report (Table 3.6). The total value of these projects is estimated at $USD 11.4 billion, almost half of which are concentrated in six projects, all of which are at the procurement stage (Figure 3.5).40 There is no reporting on government’s rights, obligations and other risk exposures, some of which are on a fixed term basis, such as the road annuity projects.41

Table 3.6.

Number and Status of PPP Pipeline Projects (July 2019)

article image
Source: PPP Pipeline Status Report, July 2019. www.pppunit.go.ke.
Figure 3.5.
Figure 3.5.

Major PPP Pipeline Projects

(Total project value, as a percentage of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: Annual Debt Report, FY2017/18.

3.2.5. Financial Sector (Good, Unchanged from 2014)

100. The CBK produces regular and comprehensive assessments of the financial sector. Their annual Financial Stability Report undertakes a risk assessment of the entire sector, which comprises both deposit and non-deposit taking institutions.42 Most financial assets (60 percent) are concentrated in the banking sector and an annual bank supervision report provides a deeper risk analysis of each banking sub-sector and entity.43 A deposit insurance scheme exists, which guarantees deposits of up to Ksh100,000. In 2017, this scheme covered almost nine percent of deposits (3.1 percent of GDP). It is managed by the Kenya Deposit Insurance Corporation which produces an annual report and a financial statement, including information on how the scheme is capitalized.44

101. Recent events in the economy and banking sector have contributed to a deterioration in Kenya’s financial soundness indicators. First, intervention was required to cover deposits in three commercial banks which defaulted in 2015 and 2016 (Dubai, Imperial and Chase). Second, the deteriorating financial condition of the majority state-owned National Bank of Kenya has strained the aggregate position of the sector. It is the second highest contributor to non-performing loans (NPLs) over the past two years accounting for 12 percent of total NPLs.45 On an aggregate basis, total NPLs increased by 23 percent from 2016 to 2017. As a result, Kenya performed comparatively worse than other EAC member states (excluding Burundi) in most indicators of banking sector stability in 2018 (Table 3.7). The higher NPL ratio relative to other countries, however, may be partly attributable to the interest rate cap that was imposed on the banking sector in September 2016.

Table 3.7.

Indicators of Banking Sector Stability (2018)

article image
Source: IMF Financial Soundness Indicators.

102. These indicators should be considered in the context of a series of turbulent shocks to the economy, including recent parliamentary elections and drought, which underscores the resilience of the banking sector. Recent restructuring of the banking sector has also made it stronger and more stable. In particular, the state-owned National Bank of Kenya (NBK) was acquired by the CBK (the largest bank in Kenya) in September 2019 and KCB’s final receivership offer for Imperial Bank was accepted by CBK on Aug 2019. The partial acquisition of Chase Bank’s assets and liabilities was finalized in August 2018. Moreover, in a major consolidation, one large bank, CBA, and one medium-sized bank, NIC, merged to make up the third largest bank in Kenya.

103. Discussions have begun with the authorities about the possibility that the IMF and World Bank will conduct a comprehensive Financial Sector Assessment Program (FSAP) of Kenya in 2020. The FSAP would review potential vulnerabilities in both commercial banks and other financial institutions, e.g., savings and credit cooperative societies, insurance and reinsurance companies, and the stock exchange.

3.2.6. Natural Resources (Not Met, Unchanged from 2014)

104. Kenya has relatively small reserves of natural resources and estimates of the volume, value and fiscal management of these are yet to be fully established. The current breakdown of the gross value-added production from natural resources is estimated at approximately 3.2 percent of GDP.46 Based on World Bank estimates of natural resource rents, Kenya ranks amongst the lower end in Sub Saharan Africa (Figure 3.6). Non-oil mining is, however projected to grow to 10 percent of GDP by 2030 and oil production could peak at an estimated 1.5 percent of GDP.47

Figure 3.6.
Figure 3.6.

Natural Resource Rents, 2017

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: World Bank World Development Indicators.Total natural resource rents are the sum of oil, natural gas, coal (hard and soft), mineral and forest rents.

105. The fiscal management framework has not yet been adjusted to accommodate the discovery and initial exportation of oil.48 A neighboring country, Uganda, which has somewhat greater amounts of proven oil and gas reserves to Kenya,49 provides a good model by taking early active steps to ensure greater transparency to ensure resources are put to effective use, establishing a sovereign wealth fund, and initiating the process to join the Extractive Industries Transparency Initiative, EITI (Box 3.2).

Planning for Oil Revenues in Uganda

Uganda has made important strides since it announced the discovery of 1.7 billion barrels of recoverable oil in the Lake Albert area in 2006. The joint venture partners have received their production licenses, a national oil company has been created to hold interest in these licenses, and a Petroleum Authority has been established to regulate and monitor the sector. The Uganda Oil Project currently consists of two major upstream developments, an export pipeline via Tanzania, and a domestic refinery. First oil is expected in early 2023, provided the oil companies make their final investment decision by the end of 2019. After first oil, production is expected to reach a peak of 230,000 barrels per day within a few years, and to gradually diminish until the mid-2040s. At current market prices, this would represent a gross revenue of the sector of about 7–8 percent of GDP at peak, of which about half would represent the government’s take.

The authorities have taken important steps to set up a coherent oil revenue management framework. Cognizant of the inherent volatility of oil resources, and its short-lived nature in Uganda, the PFM Act 2015 lays the foundations for fiscal policy after the advent of oil. It establishes a Petroleum Fund to which all oil revenues will flow. Part of these revenues will be transferred to the Consolidated Fund, and the remainder will be invested in a sovereign wealth fund - the Petroleum Revenue Investment Reserve (PRIR). The Act also requires the government to set targets on the fiscal balance excluding oil revenue—the non-oil balance, with the underlying principle that government spending should be delinked from oil revenue flows.

The government intends to further improve this framework in coming years. Progress has been made to develop the Petroleum Regulations which would operationalize the principles set in the PFM Act 2015. This is expected to include a clear and comprehensive definition of petroleum revenue – an issue which has been subject of some recent debates in Uganda; to operationalize the management of the Petroleum Fund and the PRIR; and to set the transparency standards in the oil sector. In addition, an Investment Committee has been recently set up to advice the government on the investment policy of the PRIR. Work is also under way to revise the set of fiscal rules that would underlie the next Charter for Fiscal Responsibility. Finally, Uganda has also recently committed to join the Extractive Industry Transparency Initiative (EITI).

3.2.7. Environmental Risks (Basic, Reduced from 2014)

106. Annual costs from natural disasters in Kenya are not particularly high compared to other countries, but recent droughts and floods have imposed a significant fiscal cost. Kenya is prone to periodic droughts due to its location in the Horn of Africa. The average annual damages due to natural disasters over the last decade was around 0.1 percent of GDP (Figure 3.7). The 2012 drought was estimated to cost 0.3 percent of GDP and the BPS cites recent droughts as a primary cause of reduced agricultural sector growth.

Figure 3.7.
Figure 3.7.

Average Annual Damages from Natural Disasters, 2003–12

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 002; 10.5089/9781513526089.002.A001

Source: World Bank Development Report, 2014.

107. Environmental risks and their economic impact are discussed comprehensively in several government publications, but these risks are not quantified. The Climate Change Action Plan (2018–21), the Green Economy Strategic Investment Plan (2016–30) and the Climate Expenditure and Budget Review (2016) are predominantly forward-looking strategic documents and do not include any estimates of the fiscal costs of natural disasters or environmental shocks, which has led to the downgrade of this indicator. Regarding risk mitigation, the government has outlined a well-defined strategy to manage drought risks, including the use of early warning drought bulletins.50

108. Reporting of environmental risks could be made more comprehensive with information consolidated from various MDAs into a single report. Rwanda has adopted this approach by developing a national risk atlas, which provides a useful platform for assessing environmental fiscal risks (Box 3.3).

The Rwanda Natural Risk Atlas

The Natural Risk Atlas analyzes the risks from five main categories of natural disaster—droughts, floods, earthquakes, landslides and windstorms. Analysis is made of demographic, infrastructure and other socioeconomic data, the estimated lapses of time between environmental shocks, and the probability of future occurrence by geographical area based on an analysis of historic data. The report presents detailed methodologies for assessing the exposure to various categories of environmental hazard, and potential losses of human life, crops, cultivated areas, and infrastructure from each type of hazard. The assessment of losses is supported by detailed costing tables.

https://www.gfdrr.org/en/publication/rwanda-national-risk-atlas

3.3. Fiscal Coordination

3.3.1. Subnational Governments (Basic, Increased from 2014)

109. The NT prepares and publishes annual consolidated financial statements of the county governments and the COB produces quarterly reports. The consolidated financial statements disclose aggregate payables and the net financial assets for all 47 counties, but excludes the individual payables figures, which can be found in the counties’ own financial statements that are audited and published. The County Budget Implementation Review Reports produced by the COB provide data on the utilization of funds against exchequer releases, own-source revenue and pending bills.

110. Since 2014, support to subnational governments has been ramped up significantly. At the time of the previous FTE, the county government structure had only recently been implemented, systems support was limited, and capacity was low. Since then, the IFMIS has been rolled out to the counties and extensive capacity building for county officials has been undertaken. The Inter-Government Fiscal Relations Department in the NT has been strengthened to deliver improved support to the counties. Devolved government brings services closer to the population and worldwide experience has demonstrated wide ranging benefits, including improved education, health delivery and poverty reduction. Further strengthening of subnational fiscal performance will be fundamental to realizing benefits in Kenya.

111. Despite these improvements in reporting, the lack of reporting on pending bills and inherited debt masks the true financial condition and performance of counties.51 In 2018–19, the National Treasury commissioned a special audit of County Governments’ pending bills. After deducting pending bills, 13 out of 47 counties showed negative net assets (Figure 3.8). Whilst the PFM Act sets limits on borrowing for counties, the challenge of addressing inherited debt from former local authorities potentially undermines the limits set for SNG borrowing. Examples include loans taken out by water companies transferred from the former local authorities to the counties.52,53

Figure 3.8.