Click here for search results

Resources

Ghana: Census Methodologies

(extracted from case study)

Pre-1991

The 1986 Census

This exercise, involved preparing staff lists and identifying excess labor in the Civil Service for redeployment. In accordance with the target agreed with the World Bank, the redeployment program was to involve staff reductions at the rate of 5% per annum for 1987, 1988 and 1989 (approximately 15,000 per annum).

Questionnaires were prepared and sent to all the 10 regions to be completed by governmental organizations in respect of their employees. The exercise was to be executed over a period of four weeks. There was some indifference on the part of some heads, and a team therefore had to be sent round to collect the results. Two vehicles and four people were assigned to the task of collecting the missing data.

The 1987 and 1988 Censuses

These censuses were designed to establish an authoritative and useful database both for the functional and redeployment study and for other components of the Civil Service reform program. They were payroll-based.

The approach adopted was basically through the issue of circulars and forms to be filled by officers under the supervision of their heads of departments, ministers of state, regional ministers and district chief executives. The census forms were sent out to all pay points throughout the country with monthly pay vouchers. Officers were instructed by the Controller and Accountant General's Office to refuse to issue pay to staff members who had not completed the census forms. The pay officers were given a briefing when they came to Accra to collect the pay vouchers, but they had no formal training as such.

Reminders for the prompt return of the forms were also made through wireless messages to regions and districts.

The 1990 Census

The format for the data collection exercise in 1990 was different from the global censuses conducted in 1987/1988. It was recognized that the problem with the previous two exercises was in up-dating the results. However, the time required to conduct such a major census would be considerable, and officers might become unwilling to furnish the same details repeatedly. The Government of Ghana therefore sought a method to up-date the August 1988 exercise, without a further massive data collection process.

It was proposed to use PPMD resources to create first a senior and then a junior staff list by combining the 1988 database with information from the October 1989 payroll. The draft lists were then verified by requiring personnel officers in each ministry to check that all details are correct. This involved deleting staff no longer at their ministry, and filling in details (date of birth, date of first appointment) for staff who have joined their ministry or been transferred since August 1988. The senior staff lists and junior staff list were handled separately so that the processing of returns on senior staff can be completed quickly. Any lessons learnt during the senior staff list exercise could then also be used when completing the junior staff list, and briefing sessions were provided for personnel officers engaged in the exercise. It was anticipated that the senior list could be produced for distribution in January/February 1991. The junior staff list exercise was expected to take longer because of the larger numbers involved, with production scheduled for March/April 1991.

The creation of a staff list through use of the payroll with verification by personnel officers, rather than conducting a further data collection exercise on the 1987/8 model was considered preferable because:

  • It avoided a major field exercise.
  • It combined 1988 data with the October 1989 payroll to minimize the amount of data to be collected, hence considerably speeding up the exercise.
  • It could be regularly updated using the Controller and Accountant General's system and a microcomputer in the PPMD.
  • It would produce a physical senior and junior staff list which could be distributed and inspected by all civil servants, and systematically amended using quarterly staff gazettes.
  • Data obtained could be manipulated in PPMD using their new microcomputer to enable manpower planning, linkage with MSD establishment recommendations, and manpower budgeting procedures.

It was, however, observed that there were certain deficiencies stemming from excessive duplication of personnel information and the poor state of most of the personnel registries where the data was stored and maintained. There was also a problem with the flow of information from the regions to the Central Headquarters in Accra. There was no reconciliation between the staff lists and the payroll because the payroll, which might be expected to be useful source of data on all civil service members, was only able to yield a crude total head count as a result of the way in which it was structured.

Ghana: Post 1991 Census Methodologies

While annual census approach was seen as the most practical short-term solution to the lack of aggregate manpower data, but it was recognized that in the longer term, priority should be the installation of machinery and systems to generate and update data on a regular basis. It was in view of this that a study commissioned in 1989 recommended, among other things, the development of a computerized Integrated Personnel/Payroll Database (IPPD), finally approved in April 1990.

From the very beginning the consultants responsible for the IPPD Project recognized that source data for the system would be critical for success. Before IPPD went live in July 1995 work had been carried out in the OHCS service to gather the necessary data and enter them on an 'interim system'. The method used was to send out census forms to the pay point throughout the country with the monthly pay vouchers. The pay officers were instructed by the CAG to refuse to issue pay to staff members who had not completed the census form. The pay officers were given a briefing when they came to Accra to collect the pay vouchers, but they had no formal training. The system depended heavily upon the integrity of the pay officers and the individual civil servants competing of the forms. The returns were not verified against personnel files for accuracy.

It was recognized at the time that this was a 'rough and ready' approach justified by the limited numbers of staff that could be allocated to the task. Only a small number of foreign consultants and about 30 Ghanaians (drawn mainly from the manpower services division) could be spared to survey over 100,000 civil servants.

The project team produced a detailed strategy paper that analyzed the data available and recommended strategies for populating the database with personnel and payroll data. The report highlighted issues which were to prove crucial when the project was implemented. It divided the information the system required into three categories: payroll, pensions and personnel. The majority of payroll information was to be found in the existing computerized CAG Payroll System, which contained information on about 292,000 employees in the Civil Service, GES and other public agencies. Likewise, pensions information was largely to be found in the existing computerized CAG Pensions system. In both these cases it was recommended that the information be transferred electronically to the IPPD at the appropriate time.

The personnel information was a more formidable problem. Limited personnel information on 70,000 civil servants was held on the Dbase III system. In addition, GES had data on 153,000 employees on questionnaire forms completed by the staff themselves during a survey carried out in April 1991. The Ministry of Health had data on 37,000 employees, also on questionnaire forms, collected during a survey carried out in October and December 1991. Some of the information needed for IPPD had not been collected at all. The possibility of using personnel files as a data source was rejected at an early stage and not investigated further. The requisite information was gathered by means of survey forms.

IPPD went lie in July 1995 and replaced part of the old CAG database as the system for running the payroll.

Read the full case study…




Permanent URL for this page: http://go.worldbank.org/0J6TE9MRE0