Provisional Legislative Council
Panel on Public Service
Meeting on 22 September 1997

Pay Level Survey Mechanism


Members have asked to be briefed on the mechanism used in the last pay level survey carried out in 1986; the results of that survey and the Administration's views on the need for another survey on civil service pay levels. This paper describes the methodology and approach adopted in the 1986 survey and subsequent developments and sets out our current thinking on pay level surveys.


2. The 1986 Pay Level Survey was conducted against a background of Staff Side requests for increase in the salaries of the non-directorate civil service following increase in the salaries of the directorate officers in 1985 and allegations of a shortfall in the annual pay adjustment for 1983. The Government decided that a pay level survey should be conducted to establish whether civil servants�remuneration was in line with that of their counterparts in the private sector. The survey was carried out by the Standing Commission on Civil Service Salaries and Conditions of Service which in turn appointed Hay Management Consultants to undertake the survey.

Methodology and approach

3. Since many civil service jobs do not have analogues in the private sector, it was decided not to directly match jobs in the civil service with those in the private sector. Instead a method of job evaluation was employed under which a representative sample of civil service jobs were compared with a similarly representative sample of jobs in the private sector based on three elements :

  1. know-how, i.e. the qualifications, knowledge, skills, and experience required to enable the job to be carried out properly;
  2. problem-solving, i.e. the original thinking required by the job for analysing, evaluating, creating, reasoning, arriving at and making conclusions; and
  3. accountability, i.e. the answerability for decision/actions and for the consequences thereof.

The total number of points scored for each job were then calculated and matched with the salary and the total remuneration of the job. As regards fringe benefits, the Consultants valued them mainly on the basis of notional maximum value to employees.

The results

4. The results of the 1986 survey showed that generally, with the exception of Model Scale 1, and D3 and above pay packages, civil service remuneration package compared favourably with the private sector. The pay packages for Model Scale 1, D3 and D4 were found to be below private sector levels. There was not sufficient information to make a valid comparison with the private sector on pay packages for ranks above D4. The general results of the survey were accepted by the Government.

5. Having regard to the results of the 1986 survey, the hours of work of Model Scale 1 staff were reduced from 48 hours net to 45 hours net, Model Scale 1 pay scale was adjusted upwards by 3% in value and extended by adding a new pay point so that all serving officers would receive a salary benefit of one pay point on conversion to the next higher point on the revised pay scale. No adjustment was proposed for the D3 and above pay as a result of this survey.

Staff reaction and subsequent development

6. The findings outlined in para 4 above were rejected by the Staff Sides of both the Senior Civil Service Council and the Police Force Council. The main complaint about the 1986 Pay Level Survey was centred on the defects which were perceived in the pay level survey methodology, in particular, the method of job evaluation and the valuation of fringe benefits. More specifically, the Staff Sides felt that :

  1. the methodology of the survey was too broadbrush and had a limited statistical basis. The Consultants�methodology took into account only three factors i.e. know-how, problem solving and accountability. This method was highly subjective, prone to error and open to manipulation. This ignored other important factors e.g. physical effort, working conditions, etc. It was therefore unsuitable for evaluating complex civil service jobs;
  2. the use of maximum notional value as a means of calculating the value or benefits was biased against civil servants and the valuation of civil service benefits in particular quarters, private tenancy allowances, and pensions were inflated;
  3. the time-table for the Staff Sides to comment on the survey was too tight and they were not provided with adequate information; and
  4. the methodology did not take into account the special features of disciplined services work, e.g. danger, stress, on call, restriction of personal freedom.

7. The Staff Sides reacted strongly to the Government's decision to accept in principle the general results of the survey and indicated that implementation would seriously affect staff morale and would be strongly resisted. While discussion on this continued, there was further disagreement between the Administration and the Staff Sides over the size of the 1988 pay adjustment. In the event a Committee of Inquiry was appointed in August 1988 to examine, inter alia, the methodology and findings of the 1986 Pay Level Survey and comment on their validity as a basis for making adjustments to civil service pay. The Committee concluded that :

  1. the methodology used by the Consultants for the pay comparisons was sound and reputable but job-for-job comparisons would have been preferable and would have created greater confidence in the results;
  2. the methodology used for the evaluation of fringe benefits tended to overvalue civil service benefits especially in relation to housing; and
  3. there were nevertheless no convincing grounds for disputing the general tenor of the results though the degrees of discrepancy between the private sector and the civil service must be in doubt.

8. At the same time, the Committee of Inquiry noted that while the Consultants were only asked to make broadbrush comparisons, the results of the survey came to be widely regarded as identifying precise indicators of the extent to which civil servants were over or underpaid. The Committee of Inquiry also noted the Standing Commission's view that broad comparability with the private sector, though important, should not necessarily be the first principle or major consideration for determining civil service pay. Having regard to these and the fact that the employment and remuneration scene in Hong Kong had altered radically between 1986 and 1989, the Committee of Inquiry concluded that the 1986 Pay Level Survey did not provide a sufficient basis for making specific adjustments to civil service pay either then or in the future. This was accepted by Government subject to the improvements already awarded to Model Scale 1 staff (outlined in para 5 above) remaining intact.

Future plans

9. When commenting on the system of pay level surveys in respect of the Committee of Inquiry's findings in 1989, the Standing Commission on Civil Service Salaries and Conditions of Service made the following points :

  1. there is practical difficulty in finding enough private sector job analogues for making job-for-job comparisons;
  2. frequent adjustments to external relativities at the cost of inevitable disruption to internal relativities would not be conducive to the stability of the civil service; and
  3. frequent pay level surveys would bring about considerable changes to the existing patterns of civil service pay structure, staff consultation, and pay determination.

These are very valid points. We therefore need to consider very carefully before proceeding with another pay level survey. At the moment, we do not have such plans. We will however continue to monitor the situation and review our position when circumstances warrant. We will also continue to undertake reviews of individual grades as and when justified. Such reviews were for example recently conducted in respect of the Social Work Assistant, Supervisor of Typing Services, and Land Inspector grades.

Civil Service Bureau
13 September 1997

Last Updated on 24 October 1997