Government Payroll Management: Removing the Ghosts in the Databases
Posted by Franck Bessette
The wage bill is usually one of the biggest items of government expenditure and susceptible to weak control, misappropriation, and even corruption. From an expenditure policy perspective, the political economy of wage bill management often creates inflexibility because of pay increases linked to the electoral cycle, political preferences for wage scale compression, political or electoral based hiring, or the role of trade unions which favor seniority based promotions and management through the number of positions. This policy issue nevertheless requires a proper understanding of public financial management perspectives and the use of appropriate tools for budgeting and controlling personnel expenditures.
This post, based on a small sample of countries, looks at some PFM instruments and their linkage to personnel management issues.
a) Personnel database, establishment, individual personnel records and payroll
The payroll is underpinned by a personnel database (sometimes called “nominal roll”), which provides a list of all staff who should be paid every month, and which can be verified against the approved establishment list, the list of budgeted positions and the individual personnel records (or staff lists). In many countries, linkages between these databases are often problematic. If these databases do not provide reliable information, this might lessen the quality of budgetary estimates and thus the budgetary process as a whole since the wage bill represents very often the bulk of current expenditures.
In Yemen, personnel databases remain highly fragmented. The Civil Service Ministry operates Oracle based databases at the headquarters and in 18 of the 22 Governorates. These databases are not directly linked and are not updated regularly and systematically to reflect new hires, terminations, retirements, and transfers. One consequence of such fragmentation of the personnel database has been the inclusion of multiple entries of a single civil servant, allowing them to receive multiple wages. The consequence is that budgetary estimates are based on the establishment, determined by means of a 1998 assessment of the personnel rolls which included a census. Since then the establishment figure employed has been calculated by assigning annual increases between 9% and 11% and further allocating 75% of the assigned increase to the social sectors. This approach to budget preparation has assured that the process is principally top-down and not based upon bottom-up personnel roll estimates and therefore does not reflect actual needs of ministries. An effort is currently underway in Yemen to develop a new comprehensive personnel database including biometric data control, based on physical characteristics such as fingerprints. The new system will, however, still exclude staff from the ministries of Defense, Interior, the Judiciary, and partly those of Education.
A comprehensive or partial census is often necessary to increase reliability of personnel databases. It is good practice in low-income countries to start with a comprehensive census complemented later on with more targeted reviews. In Côte d'Ivoire, a staff census was conducted in 2002 with financing from the World Bank. Only civil servants who physically presented themselves were paid. This helped remove a number of ghost workers from the lists. Since then, only assessments at the sector level were conducted.
A comprehensive census is often not sufficient. Systematic update and integration with older lists is also necessary. In Madagascar, the ministry of public service conducted a comprehensive census in 1999, which dragged on until 2001. The problem was that reconciliation with the personnel database of the payroll directorate was never completed and that the census-based list was never updated. The situation is even more complicated because of the coexistence, usually for historical reasons, of two competing personnel databases in the payroll directorate.
b) Linkages and coordination
Linkages between personnel database and payroll data are key as Premchand has noted (1983): “An organizational dualism has developed where the creation of posts is done by personnel management offices, while funding is provided by the finance ministry. This division of duties, in some countries, has led to several practical problems. The separation implies that the creation of posts will take place without resource constraint and that, once posts are created, either more resources are provided or the posts reduced. As budget reviews take place at different times, confusion between the creation of posts and their funding is to be expected. The main issue in personnel budgeting and controlling the government wage bill is the role to be assigned to the ministry of finance. The integration of selected sections of personnel offices with finance ministries would bring a coherent policy for the fast reorganization of the line ministries and revising pay structures.”
To avoid “confusion between creation of posts and their funding”, the ministry of finance should be involved in the management of the wage bill, in determining manpower levels in line ministries (establishment), long- and short-term financial implications of staff reductions and retrenchment policies, as well as financial components of the pay structure for the civil service as a whole.
In a solid budgeting system, the MOF should impose both spending limits and staff ceilings. In Haiti, civil service comprises open-ended staff and contractual staff for seasonal or occasional needs (decree of May 17, 2005). In practice, criteria for recruitment of contractual staff are never applied by line ministries and are systematically protracted. There is virtually no control over recruitment of contractuals by departments. This has resulted in arrears. The solution would be to introduce staff ceilings for open-ended and contractual staff to control the wage bill and to introduce internal control systems to check availability of funds and legal validity of staffing decisions. These ceilings should be reflected in the budget circular sent to line ministries.
Allen and Tommasi (2001) consider nevertheless that personnel expenditure ceilings are in general insufficient to control personnel expenditures in transition countries because they are often regarded, in practice, as a floor aimed at protecting personnel expenditures rather than a genuine spending limit. There is a need for more clearly identifying constraints on personnel expenditures at the budget preparation stage, and reinforcing mechanisms to ensure that the legal commitments related to them (recruitment and other decisions) comply with the spending limit. Besides, to avoid encouraging hiring of consultants and other contractuals instead of regular staff, the recommendation should be, like in Haiti, to include contractuals within the ceiling.
The challenge is then often the coordination between MOF and line ministries. In Morocco, the MOF manages the wage bill for approximately 650,000 civil servants who are also managed by line ministries in terms of staffing, training, or regulatory aspects of their situation. This situation creates coordination difficulties which could only be solved through automated linkage between personnel databases of ministries maintained by HR directorates (DRH) and the centralized payroll database maintained by the treasury (TGR). This implies some extent of harmonization in staffing regulations and practices as well as statutory measures across ministries. This implies also that different software systems are not used in various ministries. In Morocco, the process is still ongoing in spite of sustained efforts and of the efficiency of the “@jour” software and involves reorganization of administrative circuits and responsibilities. Currently, and until a complete and immediate computerized linkage is achieved, the system in place involves monthly reconciliations between DRH and TGR.
This delay in reconciling personnel records and payroll is a very convenient benchmark to measure the performance of payroll management. There is a PEFA indicator, in which the best score is given to countries where “required changes to the personnel records and payroll are updated monthly, generally in time for the following month’s payments. Retroactive adjustments are rare (if reliable data exists, it shows corrections in max. 3% of salary payments”.
c) Internal controls and audits
Internal controls and audits are the keystone of any payroll management system. Authority to change records and payroll trail should be restricted and should result in an audit trail. Payroll audits should be undertaken regularly to identify ghost workers, fill data gaps, and identify control weaknesses. Very often, payroll management lies within the scope of internal or external audit bodies (Inspection générale des finances and Cour des comptes in a Francophone context). In addition, internal control mechanisms should be put in place in line ministries. In Morocco, there is coordination between TGR and line ministries for these controls. Up to now, they resulted in no more than 225 wage suspension and about 363 radiations between 2005 and 2007, which is still a modest result but with exemplarity value. Since 2007, it was decided that at least five joint payroll audits would be conducted every year.