Personnel Expenditure Review and Wage Policy Principle: briefing

Meeting report

PUBLIC SERVICE AND ADMINISTRATION PORTFOLIO COMMITTEE
9 February 2000
PERSONNEL EXPENDITURE REVIEW AND WAGE POLICY PRINCIPLE

Documents handed out:
Draft Remuneration Policy

Do not quote this document. For discussion only.

EXECUTIVE SUMMARY
This paper sets out the aims of the public service remuneration policy within the historical context of public service employment, and proposes basic changes to the salary system and benefits as implied by this policy

1.Remuneration in the public service must contribute to transformation in four key respects:
· Supporting strategies to reconstruct and develop society, including
improved service delivery. Benefits and other types of remuneration should support changes in employment that will expand services in poor and historically underserved areas. To support changes, remuneration measures must provide appropriate incentives and must align with other national policies, for instance on labour relations, healthcare, pensions and housing.

· Continuing to overcome the legacy of apartheid in employment in the
public service. Remuneration policies must ensure a decent wage for all employees, end unfair discrimination and privileges, and open up opportunities for career paths and training for lower level workers.

· Contributing to efficient, fair and democratic administration. Salary and benefit systems should be transparent and easy to manage. They should define appropriate limits to the discretion of managers to set conditions for individual employees. The basic salary should provide the bulk of remuneration, with other benefits or allowances only where no other solution seems viable.

· Ensuring cost-efficiency and synergy with fiscal policies. Synergy is a two-way process. Remuneration policies must take resources into account At the same time, as these policies make personnel costs more understandable and predictable, they should assist fiscal planners to reflect the realities of public-service employment. In particular, as far as possible we need to stabilise real salaries for the foreseeable future.

2. The public service can be separated into six main occupational groups:
· the major occupations of educators, nurses, police, correctional officers and defence personnel. This group constitutes two thirds of all public servants;

· elementary employees - about 20 per cent of the public service;

· administrative employees - 10 per cent;

· smaller professions, departmental managers and skilled production workers, who each comprise 5 per cent or less of the total workforce.

Each of these groups pose different challenges to the public-service remuneration policy.

3. Public-service remuneration must take into account the varied reasons people join the public service, which means that it may not compare with private-sector pay. If it proves impossible to attract or retain people with appropriate skills, managers must consider all possible causes, including the level of remuneration.

4. The system of pay progression and bonuses should:
· provide strong, easy-to-administer incentives for improved (relevant)
competency as well as acceptance of demanding responsibilities, good work and innovation.
· The system should give managers enough flexibility to take the particular circumstances of occupations and individual employees into account; but must also establish controls to ensure managers use their discretion only in ways that are fair and equitable.
· The system should ensure equal pay for work of equal value both within and across occupations.

5. To make an effective link possible between the salary system and performance incentives, the DPSA will develop measures to support improved performance management throughout the public service.

6. A new salary system will replace the current rank and leg promotions, notches and bonuses with
· post promotions in the context of defined career paths for all occupations, and
an annual performance incentive in addition to the normal annual general increase. · The incentive will be paid according to norms defined within the various sectors, but should average 1 per cent for the public service as a whole.

7. By the end of 2001, sectors must define realistic career paths for major
occupations, including elementary employees. While employees will not be guaranteed automatic promotion, in every occupation they should know the requirements and general time frames required for a realistic chance to move up. Career paths should be linked to the National Qualifications Framework.

8. In terms of general increases,
· the public service will no longer provide disproportionately large increases on the minimum levels, and
· unless severe fiscal difficulties intervene, general increases should reflect inflation or realistic inflation targets.

9. Where high levels of turnover or vacancies indicate a problem in attracting
and retaining necessary skills, departments must
· review the need for skills and the possibility of substituting more easily available or lesser competencies
· ensure that people with the skills are deployed so as to maximise the benefits,
· consider more flexible forms of employment, such as part-time or contract work,
· improve the working environment,
· avoid upgrades in salaries until all alternatives have been explored, and
· avoid costly competition between departments or institutions for key groups of professionals.

10. Special measures for senior managers include:
· all managers should be employed on performance contracts, although the term need not be fixed,
· the DPSA must take greater responsibility for ensuring the appropriate deployment and redeployment of managers, in order to ensure an increasingly competent and skilled cadre, and
· The DPSA will sponsor an investigation by independent researchers into conditions of service for senior managers.

11. Sectors will replace the PAS with more modern and efficient agreements
on pay and grading.

12. Benefits should aim to meet specific needs as cost-effectively as possible, rather than providing fairly artificial incentives for long service. Retention in the main professions will be better served by appropriate pay progression. Still, it may prove impractical to reform some existing benefits quickly or at all. In any case, all benefits will be reviewed to ensure that they contribute adequately to redeployment where restructuring requires it.

13. As soon as possible, collective procurement systems will be developed for the homeowner allowance and medical aids. This should generate substantial savings for the employer and employees without reducing the benefit. Savings will be used, in part, to develop more appropriate Systems for low-income and rural employees and, in the case of medical assistance, for HIV-positive public servants.

14. To enhance mobility,
· Appropriate transfer mechanisms be designed for pensions.
· Relocation costs will include only basic actual costs, which sectors will define within a national framework.

15. In terms of overtime,
· Departments that spend over 5 per cent of their salaries on overtime must take steps to reduce overtime to normal proportions by reorganising work, giving upgrades if necessary, and defining appropriate numbers of positions.
· The DPSA will help sectors define cost-effective ways to comply with the BCEA, and establish models for policies and controls on overtime.
· Sectors will review smaller allowances and, where practicable, incorporate them into base pay.

16. Vacation leave should be a right and meet its fundamental aims providing for necessary relaxation - and not act as a source of additional retirement income or an incentive for long service. Similarly, sick leave should be used for illness, not extra time off. Generally, the BCEA will establish the minimum norms for leave.

Section 1
Introduction: Aims of a remuneration policy for the public service

A policy on remuneration for the public service should respond to a variety of challenges.

· Real salaries in the public service fluctuated heavily in the past decade. In the late 1980s and the mid- 1990s, falling real salaries caused protest action and loss of skills. The large pay increases that followed made budgeting difficult.

· By and large, the existing benefit Systems use inefficient systems and provide inappropriate incentives. They leave the government open to spiralling costs, but do not help shape the public service to support reconstruction and development.

· In many occupations, remuneration practices have not helped to recruit and retain skilled employees as cost-effectively as possible. Some relatively small groups with very important skills have high vacancies and turnover; and other groups have obtained pay increments in the form of upgrades, overtime or merit awards, without a systematic assessment of needs.

· The failure to establish principles for public-service remuneration through open discussion with stakeholders means that government and labour have little to guide them during collective bargaining. This situation leads to unnecessary conflict and may end up with inappropriate agreements.

· The public service must anticipate changes in the environment, including HIVIAIDS, which may lead to significant modifications in remuneration and costs.

A remuneration policy will not pre-empt collective bargaining over salaries. It cannot and should not give a formula for annual increases. Rather, it should assist both unions and the state as employer to define long-term linkages between remuneration and the transformation of the public service. It should set a framework for more productive and stable negotiations.

This remuneration policy rests on significant developments in the past four years, culminating in the introduction of the new Public Service Regulations. We do not here review these measures in detail, but focus instead on ways to take them further.

This document sets out
· In Section 2, the principles and context of the remuneration policy.
· In Section 3, a problem analysis of the current salary and benefits system, in light of those principles.
· In Section 4, proposals for reform based on the problem analysis.

Section 2
Principles and context of remuneration in the public service
A public-service remuneration policy must take into account basic principles that guide transformation of the public service. At the same time, it must satisfy the diverse needs of major occupational groupings, and take into account the specific dynamics of the public service. This section therefore first reviews the basic principles that must guide remuneration policy. It then explores the structure of public-service employment and some unique factors in determining remuneration for public servants.

1. Principles
Remuneration in the public service must contribute to transformation in four key manners:

1. Supporting strategies to reconstruct and develop society, including improved service delivery. Benefits and other types of remuneration should support changes in employment that will expand services in poor and historically underserved areas. To support changes, remuneration measures must provide appropriate incentives and must align with other national policies, for instance on labour relations, healthcare, pensions and housing.

2. Continuing to overcome the legacy of apartheid in employment in the public service. Remuneration policies must ensure a decent wage for all employees, end unfair discrimination and privileges, and open opportunities for career paths and training for lower level workers.

3. Contributing to efficient, fair and democratic administration. Salary and benefit Systems should be transparent and easy to manage. They should define appropriate limits to the discretion of managers in setting conditions for individual employees. The basic salary should provide the bulk of remuneration, with other benefits or allowances only where no clean wage solution seems viable.

4. Ensuring cost-efficiency and synergy with fiscal policies. Synergy is a two-way process. Remuneration policies must take resources into account. At the same time, as these policies make personnel costs more understandable and predictable, this should assist fiscal planners to reflect the realities of public-service employment. In particular, as far as possible we need to stabilise real salaries for the foreseeable future.

2. The context: A diverse public service
A particular challenge to remuneration in the public service lies in the need to support highly diverse occupations. To understand this problem, we first review the structure of the public service.

The public service can be separated into six main occupational groups. Table 1 indicates, these groups differ markedly in size and diversity. [Table not included] Each group poses different challenges to the public-service remuneration policy.

Major Professions
The major professions are found in education, health, policing, correctional employees and defence. The major professions constitute almost two thirds of all public servants, or over 600 000 employees. On the whole, these occupations have high levels of education relative to the national workforce. Close to two thirds of employees in education and health, for example, have post-secondary degrees. There are 370 000 teachers alone - close to 40 per cent of the public service, and this group account for over one in ten South Africans with a tertiary degree.

In these large professions the public service ethos was associated with virtually permanent employment, and stability remains important. Salaries and benefits appear to be more or less appropriate, given the relatively high level of skills in these jobs. The security services - police, defence and corrections - have historically received much better medical aid and pensions than other public servants.

Managing remuneration in the major professions requires care to avoid conflict over disparities within and between the sectors. Furthermore, in the absence of co-ordinate measures, regions could end up competing for skills, leading to a brain drain and spiraling costs. At the same time, the public service has found it difficult to staff remote rural areas and historically deprived communities, pointing to the need for more targeted incentives.

Elementary Workers
Elementary workers, most of whom have matric or less, make up 20 per cent of the public service. About 60 per cent work on domestic jobs, such as cleaning and gardening, in hospitals, schools and other institutions. The rest are mostly construction, forestry and agricultural workers. In many of these jobs, essentially colonial work organisation persists. In construction and rural occupations, for example, elementary workers still fall under foremen.

For these workers, remuneration policy faces particular contradictions. Government has a commitment to ensuring a living wage for all employees. Over the past five years, the average salary at this level has increased over 20 per cent in real terms. But managers have not helped these workers to improve competencies or raise productivity. Elementary employees are almost all still in dead-end jobs, with no defined route for promotion into semiskilled positions. In the absence of improved outputs, rising salaries have led to pressure to eliminate lower-level jobs.

In addition, the existing housing medical and pension schemes do not give elementary employees equal access to benefits, despite formal equality.

Administration
Administrative employees - the conventional civil service - are about one in ten public servants. They are generally grouped under a few major job titles, such as administrative officers; but these job titles span a diversity of functions, ranging from institutional managers to policy development to clerical functions. As with the major professions, existing incomes appear more or less adequate to maintain long-term employment. But considerable pressure exists to differentiate existing occupational groups in order to reflect the differences in the nature and value of jobs.

Skilled Production
Skilled production workers include a host of small occupations, including artisans, farm foremen, court stenographers and the like. They constitute a relatively small share of the public service - about 5 per cent - with unusually high seniority in most occupations.

Small Professions
Small professions include experts such as doctors, veterinarians, legal personnel and others. These employees form just under 5 per cent of all public servants. Some of these professions show high rates of turnover or vacancies, indicating problems with the current system of financial and non-financial incentives.

Given the national shortage of skills facing South Africa, a few occupations can command extremely high pay in the private sector. For these professions, government generally cannot compete financially with the private sector. Even more than with the large professions, an uncoordinated approach can also lead to competition between components of the public service. In some cases, it has also led to the misuse of benefits and allowances, such as bonuses and overtime, in order to raise incomes without increasing salaries. These strategies may lead to unfair discrepancies and also make budgeting difficult. Instead of focusing on simply raising pay, the remuneration policy must help define more innovative approaches in obtaining the necessary skills for the public service

Departmental Management
Departmental management includes employees from directors and above, as well as most assistant and deputy directors. Although they comprise only about 1 per cent of the public service, they have a disproportionate impact on its effectiveness. A particular risk can arise in these jobs if managers can authorise higher salaries and other remuneration, such as overtime or bonuses, for themselves.

In contrast to the major professions, for senior management a greater degree of openness and career mobility seems desirable. Opening the higher echelons of the public service to applicants from outside has led to improved skills_, greater understanding of new policies, and better representation. But remuneration policies must help manage possible negative effects, especially job-hopping and induction issues.

In order to close the apartheid wage gap, the past five years have seen a substantial drop in real salaries at these levels, which probably cannot persist. For senior managers, however, the decline in salaries has been offset by a car allowance. This allowance makes up over a third of remuneration - a rather odd incentive structure - and rises in line with the price index for luxury cars.

Representivity
The major occupations vary greatly in terms of representivity.
· In gender, the public service is almost evenly split between men and women. However, top management, defence, corrections and police are almost 90 per cent men. Women make up 90 per cent of nurses and two thirds of educational and clerical employees.
· For race, the public service is more or less representative overall and in the big professions. But almost all of the elementary workers are black. In contrast, black public servants constitute only around half of top and middle management, and about 70 per cent of police, defence and corrections.

Changes in Employment Distribution
Employment shifts between the occupations have had major implications for personnel costs. In 1994 to 1998, the public service shrank around 13 per cent, from over 1,2 million, to less than 1,1 million. But the job reductions fell mostly into the lowest salary category, while higher-paid professions -especially teachers and health workers - expanded significantly. As a result, average salaries increased, offsetting the reduction in overall numbers. In contrast, in the year to June 1999 alone, cuts in elementary and semi-skilled jobs led to net savings of about R750 million: Job losses saved the government R1 billion, but additional positions requiring university graduates increased employment costs by at least R250 million.

Since 1996, available information indicates that the largest job losses have been in the security services, unskilled rural and construction jobs and high-level medical professionals. Growth occurred among teachers, nurses, health therapists and social workers, administrative and personnel officials and other highly skilled positions.

Summary
Overall, to meet the diverse needs of the public service, remuneration policies must address specific requirements of each major occupational grouping. Core aims include:
· to maintain a decent wage for all employees, including elementary as well as more skilled employees;
· to support stable careers in the major professions - health, education and the security services;
· to encourage staff to fill vacancies in rural areas
· to find innovative ways to improve overall working conditions and career paths for top-level managers and smaller professional groups, while avoiding uncontrolled salary and benefit increases;
· in the longer run, to help raise skills at all levels in order to uplift the standards of the public service and society as a whole; and
· to enable and encourage all employees to improve productivity.

2.1. Why do people join the public service?
No public service has ever used money as the only incentive. It cannot hope to compete with business in that regard - after all, people join business just to make money. The public service has to offer other incentives. To understand the kinds of incentives needed, we also need to understand why people join the public service.

In this regard, several factors diverge from the private sector.
1. Many people, particularly at higher levels, join the public service because. they want to make a contribution to society.
2. Many people also remain in the public service because it gives a high degree of security combined with good retirement benefits. In reforming benefits, in particular, it is important to ensure that these incentives are not lost without adequate replacement.
3. In the major professions as well as some minor ones, the public service is the only real source of jobs. Once a student is trained as a teacher, nurse or police officer, their choices of employer become relatively limited.
4. For elementary workers, the public service is one of the few large employers that pays in line with formaI-sector norms, which is significantly higher than most rural or informal employers.

These factors point to the limited value of setting salaries based on comparisons with the private sector. It is more useful to decide if salaries are appropriate in terms of the ability to attract and retain people with the appropriate skills. A salary is then adequate if it can ensure adequate staffing; and conversely it is too low if turnover is unacceptably high. If we use seniority as a proxy for ability to retain skills, we find that virtually every profession has the same average seniority - just over ten years. The figure for some of the smaller professions, however, is significantly lower, indicating a need to review employment conditions (see below, section 7.1.4).

The public service ethos is the most intangible and hard to assess of the motives for working in the public service. Yet is critical to understand and maintaining this ethos - both to support service delivery, and to avoid spiraling wage bills.

The public service ethos survives only when public servants can see how their work contributes to social progress. That means
· the public service must develop a culture of appreciation, and
· work organisation must become less hierarchical and rigid.

If individual employees have some control over outcomes, they can appreciate the importance of their work. In contrast, employees who feel disempowered and discriminated on the job will not care how their efforts affect their communities. In the public service, disempowerment arises largely because rigid hierarchies persist in many establishments; employees lack the resources they need to do good work; and middle management is weak and therefore often repressive. Effective remuneration strategies must go hand in hand with measures to address these shortcomings.

The working conditions in many schools exemplify these problems. African schools, in particular, developed out of decades of discrimination and deprivation. Often, teachers still have virtually no say on the curriculum, lack adequate textbooks and classrooms, and have neither support staff nor strong management. In these circumstances, it proves difficult to maintain the morale amongst teachers.

It seems critical to understand all the motives public servants have for joining and staying with the public service, and a cost-effective remuneration system must rest on an appropriate synergy with these motives.

3. Changes in remuneration since 1994
3.1. The system before 1994
Pervasive inequalities characterised the remuneration system inherited in 1994. In black communities, public servants also suffered from poor working conditions caused understaffing, lack of infrastructure and other resources, and low levels of education and skills. Most public servants faced a highly repressive system of management and supervision. These circumstances caused demoralisation and poor productivity.

· Salaries and benefits discriminated by race and gender, in some cases explicitly and in others by differentiating between occupations and between racially divided administrations. Even after nominally equal pay for equal work was instituted in the early 1990s, discrimination in hiring, promotions and job design maintained inequities. Most benefits excluded elementary workers and married women, or gave them less.

· The security services had better benefits, especially pensions and medical aid, than other public servants.

· The public service was essentially exempt from legislation on labour relations, including minimum standards. The Commission for Administration set salaries without formal negotiations, although it consulted with selected staff associations.

· In the major professions, fairly crude incentives aimed to support long careers. Administrative personnel, police and nurses received notches of 3 to 4 per cent, and educators got just under 1 per cent. Benefits both increased greatly after ten years' service and provided heavy payments at retirement.

· Benefits were provided through inefficient systems, which raised the cost to the employer unnecessarily.

· Remuneration systems were complex and captured in confusing regulations. This gave managers considerable discretion and meant employees often suspected, and probably often suffered, unfairness and discrimination.

· Real salaries fluctuated dramatically, as Chart 1 indicates.[Chart 1 not included] This trend, which is found in many countries, resulted from crude attempts to subject
remuneration to fiscal imperatives. Typically, demoralisation and conflict resulted, with significant increases thereafter leading to fiscal problems.

3.2 Progress since 1994
After 1994, the democratic government made tremendous progress toward eliminating systemic and overt discrimination in the public service. It also established a much more transparent salary system, and began to apply national norms on labour relations. This section outlines major reforms between 1994 and the introduction of the new Public Service Regulations in 1999.

Rationalisation of the former apartheid administrations entailed the establishment of a unified system of salaries and benefits for the entire country. The national system permitted differences in pay based on occupation or sector, but only rarely by location.

From 1994, the government formally gave all employees - extending essentially elementary workers and married women - equal access to benefits including pensions, the homeowner allowance and medical aid. In the next two years, recorded payments for pensions, homeowner allowance and medical aid climbed 60 per cent. Thereafter, growth slowed to around 3 per cent a year in real terms. The rise was partly due to increased take-up, and partly due to better record keeping, since the former homelands often did not record benefit payments as personnel expenditure.

The democratic government also adopted a policy of closing the wage gap by increasing the minimum salary and holding the line on higher salary levels. Between 1993 and 1997, the minimum wage rose from 85 per cent of the Household Subsistence Level (HSL) to 125 per cent of the HSL. Salaries for senior managers grew less than average until 1996, when they received a 30 per cent increase.

After 1996, significant shifts in the salary scale continued. The lowest salary levels experienced a substantial real increase, salaries for the major professions remained virtually unchanged, and the highest salaries underwent a real decline. As a result of these measures, ignoring the large differences in benefits, the ratio of salaries from director general to unskilled employee fell from 23 to 1 in 1994 to 16 to 1 in 1999.

The policy of closing the wage gap reflected an explicit decision that the market would not dictate public-service wages at the top and bottom of the scale. That approach effectively assumed that:
· apartheid artificially reduced salaries for unskilled workers and inflated them at the top. This has resulted in South Africa having the highest income ratio between unskilled and skilled work, of all middle-income countries; and
· over time, with the end of apartheid policies, the economy would gradually generate more jobs and skills, leading to an increase in salaries for elementary work and a decline at the top.

In the event, the shift toward a more normal income distribution has taken more time. Departmental management argues that strains have emerged due to the compression of the public-service salary scale. In particular, they have faced soaring costs for elementary labour, and difficulties in attracting people at the top end of the salary scale that is, senior mangers and small professional groups. In these circumstances, pressure has grown to stop seeking to close the wage gap in the public service.

In 1996, an historic agreement between the public service unions and the State established the basis for a broad-band salary system. This system established 16 salary ranges on a single pay scale for the entire public service. It also incorporated most, although not all, occupational and sectoral allowance into salaries.

This system made possible a comparison of remuneration across all occupational groups. On this basis, a move to equal pay for work of equal value becomes possible. Before 1996, occupational differentiation of salary scales made these comparisons impossible. Now, the main obstacle is no longer to discover what different jobs pay, but how one should value work in different occupations.

Salary Levels and Occupational Groups
Under the public-service system, elementary workers occupy levels 1 and 2; skilled production workers, levels 3 to 5; professionals with tertiary degrees, supervisors and institutional management, levels 6 to 12; and top management, mostly in departments, levels 13 and up.

The introduction of the 1996 remuneration system involved an upgrade for health workers, police and to a lesser extent, educators. The average nurse received an increase of 30 per cent. The average policeman received 25 per cent, and teachers, 7 per cent. Thereafter, the agreement effectively stabilised real salaries. Subtracting the costs due to the upward drift in employment, the average salary rose by around 0,3 per cent a year after 1996.

Shift in Legal Framework
Underlying all these changes was a major shift in the legal framework. From 1994, the public service became increasingly aligned with national labour-relations requirements. Negotiations on salaries began in 1994. A central council, the Public Service Co-ordinating Bargaining Council (PSCBC) was established in 1998, and by 1999 had four sectors - education, health and welfare, the police, and one for the remaining employees.

In addition, the public service increasingly became subject to minimum standards. The Basic Conditions of Employment Act (BCEA) will apply from June 2000, requiring some substantial changes in working time and overtime payments.

Conclusion
The negotiation structure is currently in a state of flux. Until around I997, negotiations took place in independent sectors defined mostly by occupation -essentially education, police and a residual sector - combined with provincial councils. The current structures involve a co-ordinating council, the PSCBC, with a set of stronger sectoral councils defined by employer in three cases -health and welfare; safety and security; and the General Public Service Sector- and educators.

The process of changing the structure of negotiation forums arose because:
· sectors generally have more in common for negotiation than provincial departments, and
· unplanned differentials in remuneration based on geography seem likely, in a developing country like South Africa, to lead to brain drain and competition for scarce skills between government departments.

The current process leaves several questions open.

1. Currently, the PSCBC negotiates a single salary scale, while the sectors and departments negotiate changes in grade.

Should sectors establish separate salary scales? To ensure more appropriate salary systems for sectors will require considerable developmental work. A project is currently underway for educators, but not for other occupational groups.
If sectors establish separate pay scales, how should funds be allocated between them? The State as employer could allocate the funds between sectors as its sees fit, taking the negotiations at sectoral level into account. Alternatively, negotiations at the PSCBC could set a framework by setting the minimum and/or maximum salary; establishing an average increase for each sector; and/or approving agreements reached on pay scales at sectoral level.

2. As sectors establish provincial chambers, the scope will overlap with provincial councils. Two options emerge, that may be combined:
· the provincial councils can become the chamber of the General Public Service Sector, and
· provinces can establish a co-ordinating body in which representatives of provincial sectoral chambers can discuss cross-sectoral issues.

3. The LRA deems all provincial and departmental negotiation forums to be bargaining councils initially, subject to approval by the PSCBC. It seems more appropriate to establish sectoral chambers and departmental bargaining forums on agreement. That would give the parties greater flexibility in deterring the responsibilities of these forums in dispute settlement and negotiations.

4. The public service has increasingly moved towards industrial labour relations for all public servants. These procedures work well for large, homogenous occupations, such as elementary workers and the major professions. But they impose excessive rigidity on small, high-level professions and senior management. We need to explore options for more flexible treatment of these groups, reflecting the value of an employee's work to the employer.

Moreover, until 1999, procedures in the Public Services Act for dealing with misconduct and incapacity were unusually slow and legalistic. Even if an employee was performing extraordinarily poorly, most managers were reluctant to embark on these procedures, leaving few formal sanctions for poor performance. Between 1994 and 1999, only about I per cent of all employees who left the public service were discharged for misconduct or incapacity.

In 1999, the PSCBC agreed on much simpler disciplinary and misconduct codes. These codes should facilitate corrective action where poor performance warrants it.

Perhaps most seriously, the new system did not address the main shortcoming in performance management in the public service: the weak skills of lower-level supervisors and middle managers. Poor capacity at this level has been identified as a major shortcoming of the South African economy in general, and the public service is no exception.

Designing performance-management systems suitable to the responsibilities and capabilities of supervisors and institutional managers in the public service presents a major challenge. These positions range in grade from level 3 up to level 15. With 1,1 million employees, supervisors vary from semi-literate foremen to school principles, station commanders and medical and legal specialists.

Generally, supervisors receive virtually no training in performance management or indeed any other supervisory skills, and operate in hierarchical structures. Unsurprisingly, few provide constructive feedback and support to their subordinates. Managers of elementary workers, for example, typically argue that they cannot assess their performance at all.

In these circumstances, the assessment process is an arduous annual ritual. Often subordinates consider it fundamentally unfair or biased on the grounds of race, gender or union affiliation. Neither managers nor employees, learn much from the process, in order to improve their work.

The new regulations pave the way for improvements in performance management by
· requiring more systematic discussions of performance with employees, and
· allowing departments to design their own assessment procedures.

4.1.2 Rank promotions
Rank promotions have effectively become automatic Outstanding performance may shorten the waiting period, but poor performance will not prevent promotion.

Because they are virtually automatic, rank promotions mean that career progress no longer requires defined improvements in competencies or responsibilities. In the sectors that have a lot of rank promotions, such as health and police, managers argue that this overloads the higher ranks. As a result, they have tended to delay assessment processes, ending up with substantial backlogs.

In addition, rank promotions are very inequitably distributed between professions, which causes conflict between sectors as well as fiscal problems.

· Teachers do not qualify at all; elementary workers have only one promotion, from level 1 to level 2. But these constitute over half the entire public service. Their only real increase over time comes from the annual general increase for all public servants and to a lesser extent, the promotions into higher-level vacancies, based on the normal competitive processes. For unskilled elementary workers in particular, mobility is limited.

· In contrast, nurses, police, correctional officers, administrative personnel as well as some limited professional groups can receive between three and five promotions in the course of their careers. Even if the general increase falls below inflation, they will still gain over time.

As a result of these inequalities, wage drift varies tremendously by sector. Rank promotions add between 3 and 5 per cent a year to personnel costs in the above sectors, (although only about 1,5 per cent to overall public-service personnel expenditure). It has proven virtually impossible to budget appropriately for these different conditions.

Furthermore, the inequality between sectors and the three-year delay between increases means that in any given year, only 10 per cent of employees receive a rank promotion. For the remainder, the only real increase comes through the annual general increase. This situation leads to pressure for annual improvements above inflation, with no saving on turnover in the long run.

4.1.3 Notches and bonuses
Shortcomings in performance management also undermined the notch and bonus systems. Like rank promotion, neither system applies to teachers.

Each notch equals around 4,5 per cent. Because of the relatively high cost, departments often ration allocation, and even outstanding performance can fail to earn a notch. As a result, they fail as an incentive for many employees. Where departments do give notches more freely, the cost becomes high.

Central guidelines propose that 20 per cent of employees should be on notches above the minimum. However, the actual share ranges from over 50 per cent in some departments, down to less than 2D per cent in others.

Bonuses, which cost R300 million a year -just over half a per cent of salaries costs - are even more unequally distributed. The available evidence indicates that they are particularly inequitable by sector and level, although reasonably fair in terms of race and gender. The SARS alone accounts for R200 million in bonuses, for just 1 per cent of all public service employees. Generally, more skilled employees are much more likely to receive a bonus than employees on lower salary levels.

4.1.4 Summary
In sum, the performance incentives in the 1996 agreement failed in large part because they relied on a dysfunctional management and assessment system, without providing adequate competencies for supervisors and middle managers. In addition, as incentives, the salary systems were poorly designed.

· In key professions, rank promotion undermines the prospects of career progress as an incentive for improved competencies, responsibilities and performance, while notches tend to be inequitably distributed and cause considerable conflict,

· The 1996 system reduced salary drift from 3 to 4 per cent to around 1,5 per cent a year overall, but retained very high salary drift for health and police, and

· The 1996 system did not provide pay progression for half the labour force, and as a result made high general increases necessary every year.

4.2. Grading and job evaluation
A broad-band system should make it possible to ensure that work of equal value is on the same grade, which in turn translates to an appropriate salary level. Demands for regrading and by extension salary changes - emerge most explicitly within some of the smaller professions.

Generally, three factors tend to hinder changes in grading: delays in carrying out job evaluations, financial constraints, and the need to ensure some uniformity within occupations.

4.2.1 Grading and occupational groups
For smaller professional groups and departmental managers there is often agreement on the need for an upgrade - a move to a higher salary level. This situation reflects, in part, the very high incomes earned by comparable professions in the private sector, and in part the decline in real salaries at the upper levels of the public service.

In contrast, no one argues for an upgrade for any of the major professional groups as a whole, because the cost would be too high and because turnover and vacancies are not particularly burdensome. However, re-grading may be needed to address inappropriate grading of particular jobs (sub-occupations) within these groups. But in this instance, unless an upgrade is clearly justified by the specifics of a position, upgrades for sub-occupations within the major professional groups may cause considerable conflict.

Among administrative groups, the problem lies largely in the use of single job categories for very diverse occupations. In these cases, redefining jobs to take into account real differences could lead to shifts in grading for some employees.

4.2.2 Obstacles to regrading
The new Regulations and the associated agreements give managers considerable leeway to regrade jobs, if
· an upgrade is justified in terms of the needs of the public service,
· managers have the necessary funds and make public information on upgrades, and
· where necessary, the change is negotiated in the relevant forum.

Justifying regrading poses some difficulties.
First, employees and managers frequently lack information on vacancies and turnovers, basing calls for upgrades instead on general discontent with salaries or impressions - usually without actual data - that salaries in the private sector are higher. In the absence of evidence that the purported low salaries harm service delivery, justifying an increase proves difficult.

Second, to ensure equity within the public service, the new Regulations introduce a job evaluation system. Job evaluation systems essentially define broad characteristics that can be compared across very different jobs. The public service EQUATE system was developed by benchmarking existing jobs, in order to define what was historically considered appropriate. It now establishes an index essentially based on measures of competencies and responsibilities.

While the job evaluation system remains too new for a detailed assessment, some risks clearly exist.
· The system requires considerable capacity building. Given the size of the public service, no overall targets have been set for a grading exercise based on the job evaluation system. Instead, the regulations only seek to ensure a gradual improvement in grading by requiring the evaluation of new jobs and high-level vacancies.
· The system seeks to guide managerial discretion on salary setting. Managers can provide upgrades even if a job evaluation does not warrant one. The only control remains the requirement that they justify deviations in terms of high vacancies or turnover and report publicly on deviations. However, the discipline the system aims to impose may cause considerable resentment.
· The EQUATE system does not apply to educators. A current project aims to define a more consistent salary-grading system for them in order to remedy this shortfall.

A third problem that is emerging is the fragmentation of the public service and thus the budgeting and allocation systems. Certain smaller groups may be able to separate themselves from the public service (although remaining in state employment), and thus easier effect higher salaries or upgrades. This has happened, for example, with judges, magistrates and SARS. The cost, of course, is still carried by the same budget. However, it is more difficult from the budget side to allocate rationally between sectors and occupations, and to weigh up the costs and benefits of higher allocations to certain groups.

Financial constraints may prevent even justified upgrades. If departments need to upgrade an occupation, they have to find the funds on their own budgets, negotiate with the Department of Finance, or seek the funds set aside for the general salary improvement. Given current fiscal constraints, very little resources are available from any of these sources.

Finally, the issue of when and where to negotiate upgrades remains vague. The agreements associated with the regulations only provide that, where necessary, grading will be negotiated in the "relevant" bargaining forum.

Generally, grading for newly defined jobs and smaller occupations does not require negotiation. In contrast, the grades of existing occupations are still determined by the existing Personnel Administrative Standards (PAS), which have become collective agreements. Thus, the PAS for professional nurses has effectively become an agreement with (a) a scope - registered professional nurses - and (b) a set of grades for jobs that fall within that scope, associated with different levels of seniority.

4.2.3 Summary
As a whole, the 1996 system on grading provides greater flexibility, but it also poses considerable risks.
· If lack of capacity and public oversight means that the control systems established under the regulations do not function, the public service may face escalating costs for small professions and managers.
· The fact that managers may give upgrades does not in any way guarantee them the funds.
· Finally, if managers give upgrades to some sections of the large professions without clear justification, this may cause ratchet bargaining, job-hopping and conflict.

5. The major benefits: Pensions, medical and home-owner allowance
The formal extension of benefits to all employees did not go hand in hand with efforts to restructure the system to ensure greater efficiency and to target lower level and rural employees. As a result, most elementary employees still do not receive major benefits; security personnel enjoy better benefits than other public servants; and provisioning systems remain highly inefficient. In these circumstances, if take up increased, the public service would face severe financial constraints.

5.1. Pensions
The pension system aims to make provision for all employees' retirement. It should provide a reasonable and fair standard of living for retirees at all levels, without imposing excessive costs on the government. The government pension fund is called the Government Employees Pension Fund (GEPF).

Two fundamental problems emerge with the current system: the benefits appear to be inequitable; and the government pays a higher than average employer contribution. Particular problems that should be addressed include:
· The GEPF provides much higher pensions for some categories of
employees, and the effect of these incentives requires further investigation.
· Very small pensions for many former general assistants, who were not
allowed to join the normal pension fund until 1994. An agreement in 1998 aimed to backdate pensions for these employees according to their actual service, but has proven difficult to implement due to poor records. The backdating of pensions for former general assistants should be accelerated.
· The government contribution includes 21 per cent on employees' service bonus, an unorthodox and expensive practice.
· Group transfers of funds require complex legal manoeuvres, which prevents the shift to more innovative forms of work organisation.

5.2. Medical Assistance
Under the current medical assistance system, the government pays a monthly subsidy for employees outside the security services. The security services have very different systems, which provide heavily subsidised health care for all members.
The current system for the public service presents a variety of shortcomings. Distribution of Medical Assistance
About half of public servants outside of the security forces get medical aid. The system does not assist low-income employees, presumably mostly because of the relatively high employee contribution required by most medical aid schemes. Among elementary employees, only about one in five uses medical aid. These workers probably rely on the public health system. New legislation will require the provision of packages for all employees, with benefits for managing HIV /AIDS.

The available information suggests that the current system also does not serve rural employees. It provides insurance for treatment by private doctors and facilities, which are not common in more remote areas.

The system imposes high costs on the government as well as on employees who participate. Each individual employee engages a medical scheme separately (there are about 62 schemes covering public service employees) she or he does not benefit from the market power of the public service as a whole. This is despite the fact that 85% of public service employees are in fact covered by only ten different schemes.

Although the employer contribution rises with medical price inflation, the annual rise lags behind the actual increase in cost. Yet public servants account for about 20 per cent of alt medical aid membership, which should provide considerable leverage to improve benefits and control costs.

Because the employer contribution is linked to the medical price index, it rises much more quickly than overall inflation. Between 1995 and 1999, the medical price index rose 60 per cent - twice as fast as the consumer price index.

The cost of medical aid has doubled in real terms since 1994, both as a result of rationalisation and because the cost per member rises faster than consumer inflation. The current cost to the employer comes to R3 billion a year. If take up at all levels increased to 50 per cent under the current system, the cost would rise another R2 billion.

The current system allows public servants to join a medical scheme as late as one year before retiring, and then receive assistance after retirement, and the actual cost of post-retirement assistance is high in relation to the overall medical assistance cost. As with overall medical aid, this arrangement benefits mostly higher level employees and the security forces, and adds substantially to the cost.

Finally, it seems probable that the current medical aid system has some negative effects on the public health system and overall health costs. Private schemes have increasingly sought to redirect patients from the public facilities into private ones. This approach undermines the financial and social basis of the public health system. Furthermore, under the public-service system the government subsidises untrammelled choice of care for a small, relatively high-income group. This process can drive up the cost of health care throughout the society.

5.3. Housing
The homeowner allowance in the public service, which was introduced in its current form only in 1991, sought to ensure that on retirement, public servants would own housing. In effect, it acts as retirement provisioning.

This allowance provides the market interest on a bond of R70 000. With inflation, the amount of the allowance should decline gradually in real terms. In recent years, high interest rates have offset this potential saving. The allowance currently comes to around R800 a month, and is subject to normal income tax. In addition, the government provides guarantees on public servants' housing bonds.

A particular problem with the allowance is that many banks add on the subsidy after assessing gross income, allowing for bonds that exceed the usually accepted affordability-levels of the individual. (Many private sector employers have rather incorporated housing allowances within the salary.)

The homeowner allowance is not useful for township dwellers or rural people, and does not reach married women or public servants on lower salary levels.

Employees are eligible only if they are buying or building a house on credit. People in townships, however, rarely buy houses on credit - they either own their own houses due to earlier transfers, or rent them. Commercial banks often red-line low income housing, making it virtually impossible to obtain a bond. Low-level employees may also face difficulties in obtaining a loan and the down payment. This is evidenced by the fact that many employees request deductions for non-subsidised housing loans.

In the rural areas, houses are also not available for purchase and are relatively expensive to build. Moreover, because of uncertainties about the title to land in many areas, banks will not grant bonds.

Many women cannot access the homeowner allowance because their home is registered in their husbands' names. To obtain the allowance, however, employees must have title to their house. Transferring the bond is an expensive procedure.

The result of these conditions is that at every salary level, fewer Africans and women get the homeowner allowance than other public servants. Furthermore, very significant discrepancies emerge by level. Rural provinces also have much lower take up than relatively urbanised areas.

By province : % of employees getting home-owner allowance
Northern Province : 12%
Mpumalanga : 18%
Eastern Cape : 19%
North West : 21%
Northern Cape : 24%
Free State : 26%
KwaZulu Natal : 26%
Gauteng : 40%
Western Cape : 52%
National Departments : 29%
Total : 27%

As with the medical aid system, the home-owner allowance makes each public servant negotiate separately with providers - in this case, the financial institutions. In consequence, the public service reduces its managerial burden, but ends up with higher-cost services. The homeowner allowance meets the market interest, even though public servants represent a relatively stable and high-income group.
An additional cost arises because employees can extend the bond more or less freely. As a result, employees can receive the homeowner allowance long after they have effectively completed purchase of a house.

The cost of the home-owner allowance to the government doubled in real terms between 1995 and 1999, to reach around R2,5 billion a year. Take up rose by 34 per cent between 1994 and 1995, but growth slowed to 3 per cent in 1998/9. Still, if take up increased to 50 per cent, the cost would rise by another R2,5 billion. The guarantee system adds a relatively small sum, but a disproportionate administrative burden for departments.

Finally, at a time of housing shortages, the homeowner allowance skews the housing market toward a relatively high-income group. The government spends almost as much on homeowner allowances for public service employees every year as on subsidies for low-income housing. Public service homeowner allowances account for a substantial and rising share of the entire housing market. Between 1994 and 1998, the national share of bonds accounted for by subsidised public service employees increased from 10 per cent to 16 per cent. This increase reflected both the increase in public servants' participation and reduced demand in other sectors as interest rates rose.

5.4. Summary
In short, the major benefits now provided - pensions, medical and housing -remain skewed toward the security services and middle and upper income employees. Furthermore, if we take pension backlogs into account, none of them adequately reach rural employees. At the same time, in housing and medical aid, the reliance on procurement by individual workers has prevented the public service from using its market power to develop more cost-efficient (to both the employee and the employer) schemes. Finally, these two systems appear to have some negative effects on relevant markets - that is, medical care and housing.

6. Overtime, Leave and Miscellaneous Allowances
Overtime currently costs around R1,4 billion, or almost 3 per cent of salaries. But it is heavily concentrated on commuted overtime for doctors and dentists, which costs almost R700 million, and overtime for correctional officers, at around R450 million. These two occupations account for almost three quarters of all overtime payments in the public service. For doctors and dentists overtime constitutes some 40 per cent of salaries; for corrections officers, it comes to 25 per cent; for the rest of the public service it averages under 1 per cent of salaries.

At current levels, overtime does not impose a high cost. For most departments, it remains considerably under the manufacturing average, which is well over 5 per cent of working hours. Still, the rapid growth in overtime payments since 1996 has caused concern in some departments.

More than the level, the misuse of overtime causes concern. For some groups of professionals, overtime has been used in recent years as a substitute for salary upgrades. This practice is undesirable, since it can cause unpredictability and inequality. More generally, management often fails to implement control measures on overtime.

Finally, the current agreements do not align with the BCEA. They provide for time off equal to time worked on overtime, which is less than the legal requirement.

Leave should provide employees with time to recuperate and relax, in order to ensure higher productivity and permit normal family life. In addition, leave should meet special needs for time off, for instance for illness, maternity and to meet urgent family obligations.

From this standpoint, in the public service leave has been generous. The public service grants about a third more days of vacation leave, and half again as much sick leave as the BCEA requires. Furthermore, employees gain additional days for both kinds of leave after ten years' service. This increase fits in with the earlier approach of fairly artificial improvements in benefits for employees who stayed ten years with the public service.

Generosity in granting vacation leave has often translated into the accumulation of leave days. Employees receive a day's payment at the current salary - not at the salary when they earned the leave - for each unused leave day if they retire, take a VSP or are retrenched. Otherwise, they get nothing for accumulated leave. In 1998/9, normal leave gratuities cost R20 000 per retiring public servants, or a total of R250 million. Leave gratuities on severance packages probably cost another R100 million.

The accumulation system has the perverse effect of encouraging employees not to use normal vacation leave. Furthermore, it runs contrary to the BCEA, which requires the payment of leave accumulated up to six months after it falls due, if employees leaves their job for any reason.

In terms of sick leave, the current system effectively provides for disability and sick leave at the same time. As a result, most employees feel very little pressure to ration sick leave. This shows up in some obviously exploitative patterns. For instance, a very clear pattern emerges around the use of sick leave on weekdays, with very disproportionate days off on Mondays. If sick leave on Monday were equalised at the average for weekdays, the public service would save the equivalent of R70 million. It also appears that employees sometimes use sick leave instead of vacation leave, so that they can accumulate vacation days

The situation is worsened by extremely poor records on leave and the inclusion of weekends in the calculation of leave days. Records on leave vary tremendously by department and province. Thus, on average, around three quarters of employees use sick leave in a year - but records for the Northern Province shows 5 per cent use sick leave, and for Mpumalanga, less than 20 per cent. Inquiries indicate that the divergence is due to poor record keeping, not to differences in the quality of health.

Finally, the BCEA requires the provision of three days' family leave, and four months unpaid maternity leave for all confinements, where the public service now provides three months' paid leave for only two confinements.

The public service provides an array of smaller allowances and benefits in addition to the major provisions discussed in this document. Often, the information on these payments remains extremely poor. Table 6 indicates the major components.

7. Toward a remuneration policy for the public service
The proposals below arise from an analysis of the existing weaknesses, and aim to achieve the four principles of the remuneration policy:
· Support broad strategies to achieve reconstruction and development and improve service delivery.
· Continue to overcome the legacy of apartheid labour practices.
· Support effective public administration.
· Ensure cost-efficiency and effectiveness and synergy with fiscal policies.

7.1. Principles for the salary system and incentives
Salary-based incentives essentially work through pay progression - that is, individual movement against the salary scale. In contrast, the annual general increase serves to raise pay for all employees by shifting the entire salary scale upward, basically to compensate for inflation. Bonuses, which do not entail a permanent salary increase, can also give targeted incentives.

The system of pay progression and bonuses should meet several policy principles.

1. It should provide strong, easy-to-administer incentives for improved competency as well as acceptance of demanding responsibilities, good work and innovation. Competency should include skills learned through experience. The system should encourage multi-skilling, and on that basis more flexible work organisation. But it should not reward a person for skills that are not relevant to her or his job. It should ensure retention and attraction of people as required in the major occupational groups.

2. The system should give managers enough flexibility to take the particular
circumstances of occupations and individual employees into account. But it must also establish controls to ensure managers use their discretion only in ways that are fair and equitable.

3. The system should ensure equal pay for work of equal value both within and across occupations.

Pay progression has two aspects. The career path for an occupation takes the form of promotion through higher grades, which are reflected in ranges on the salary scale. Promotion to a higher grade should entail a substantial increase in competencies and responsibilities. Incentives for smaller improvements in competence or performance involve smaller increments within each salary range. In part, these incentives should reward experience in the major professions in order to ensure retention.

The new system must define more predictable and realistic career paths for the major professions and elementary jobs. Ideally, while employees would not be guaranteed automatic promotion, in every occupation they should know more or less the requirements and time frames required for a realistic chance to move up.

Since work organisation defines the nature and number of positions available in each grade, it effectively sets the parameters for career pathing and grading. Systems to support rational decisions on grading should
· require managers ensure value for money in setting the grade of a post, and
· ensure equal pay for work of equal value.
The systems of job evaluation and grading established by the new Regulations attempt to achieve these goals.

7.1.1 Performance management
In order for pay progression systems to provide appropriate incentives, an accurate and cost-effective performance management system is required. That in turn entails far more than just a link between money and assessment results. Performance management is about promoting a culture of performance and not simplistically about a more or less scientific system for measuring and evaIuating performance.

The core of performance management is the understanding that it is first and foremost the responsibility of management to ensure that employees can work productively. In that context, effective performance management needs:

1. assessment systems that reflect the real needs of the employer, do not impose disproportionate administrative requirements, and, to avoid unfairness, limit the discretion of individual supervisors;

2. mechanisms to ensure on-going feedback, so that both managers and employees understand problems and can correct them; and

3. well-defined links to promotion prospects and incapacity or misconduct procedures. If these fundamental performance incentives are not well structured, annual financial rewards, even if they are substantial, will not have much impact.

Over the coming year; the DPSA will support sectors in developing
· more appropriate assessment mechanisms and financial incentive Systems, and
· target training in performance management for middle managers.

7.1.2 A new system of career pathing and pay progression
The public service will replace the current rank promotions, notches and bonuses with a more equitable and targeted system of pay progression that accords with the basic principles defined above.

Under the new system, an average employee - that is, one with satisfactory work - will receive a 1-per-cent increment over the general increase every year. When the employee reaches the maximum salary in her or his grade, she or he will not longer receive an annual increment.

The specific systems for allocating increments, and their amounts, should be decided at the sectoral level, as long as the total cost comes to 1 per cent of their salary bills. For instance, a sector could agree on 3 per cent for high flyers, and no increment at all for an equivalent number of very poor performers.

By the end of 2001, sectors must define realistic career paths for major occupations, including elementary employees. In effect, they must determine
· what competencies and functions they will need in the next five to ten years,
· what grade they should give to people with those competencies and/or functions in different parts of the organisation, and
· how many positions they will need at that grade.

The newly defined career paths should involve post promotions, in which employees compete to fill vacancies, rather than automatic promotions as in the current system. Generally, organisational needs will limit the extent to which an employee can expect to rise up the salary scale. Thus, the commitment to defining career paths is not a promise that all employees will rise indefinitely on the salary' scale. Still, every public servant should ultimately have a reasonable understanding of the competencies they need to obtain and the obligations they must fulfil to be eligible for promotion.

In other words, the definition of career paths aims, not to guarantee promotions, but to ensure that opportunities and requirements for advancement should be defined and broadly understood. In that context, departments can design training plans that let employees take advantage of these opportunities. In particular, for elementary employees, career paths should be linked to the establishment of training opportunities, including ABET. Over time, career paths should be rooted increasingly in the National Qualifications Framework.

This average increase for good performance will initially cost 1 per cent a year over the general increase. Because senior employees rise above the starting salary, savings from turnover will emerge over time. With average seniority of ten years, after ten to fifteen years the wage drift will fall to around 0,3 per cent a year over the general increase.

In addition, sectors should negotiate appropriate agreements to determine the level of bonuses. They should establish guidelines for awarding bonuses that can ensure administrative justice, in accordance with the new regulations

7.1.3 Guidelines for general increases
The stabilisation in real wages over the past three years should continue. But the minimum salary has achieved an adequate level. In future, elementary employees should gain higher salaries primarily through improved training and promotions, not through the general increase.

By extension,
· the public service will no longer provide disproportionately large increases on the minimum levels, and
· unless severe fiscal difficulties intervene, general increases should reflect inflation or realistic inflation targets.

7.1.4 The small professions
Special problems emerge in defining grading and career paths for small groups of professionals. For some categories of these jobs, a pattern of high vacancies and turnover has emerged, indicating a shortage of skills.

The skill shortages reflect, in part, the change in demand. After 1994 government sought to equalise services for all communities. Instead of redefining work organisation to take into account the skills constraints, most departments tried to extend to the country as a whole the work organisation and skills found in historically white communities. High levels of vacancy, greater burdens of work and increased turnover began to emerge in many sectors.

In this context, strategies to manage scarce skills must
· review the need for skills and the possibility of substituting more easily available or lesser competencies,
· ensure people with the skills are deployed so as to maximise the benefits,
· consider more flexible forms of employment, such as part-time or contract work,
· improve the working environment,
· avoid upgrades in salaries until all alternatives have been explored, and
· avoid costly competition between departments or institutions for key groups of professionals.

Where a profession experiences high vacancies and turnover rates, managers must first conduct a job evaluation. If the salary seems appropriate according to the evaluation, they should conduct an in-depth review of working conditions rather than simply decreeing that market forces require higher pay.

Strategies to reduce the need for or cost of professionals and/or senior management include:

1. Ensuring that all high-level personnel are deployed properly to use their skills. In some cases, they undertake tasks that could be performed by less skilled employees reducing their time for their specialities.

2. Outsourcing or using contract or part-time personnel, although only if this will prove cheaper and supply the necessary services in both the short and long term. As the new Regulations suggest, the public service will establish guidelines for more flexible employment relationships.

3. In many fields, professionals treat the public service as a source of experience and training. In these cases, rather than trying to retain people for long service, it may prove more efficient to define clearly areas where relatively junior people can work effectively, while designing better packages for a much smaller number of senior professionals. Defined systems to support and mentor junior professionals should be established.

4. Developing a more attractive working environment for professionals by
· replacing tight and controlling hierarchies with team work and appropriately autonomous professional responsibilities
· providing perquisites appropriate to professionals, such as training, sabbaticals and study tours,
· setting up well-defined career paths that take professional development into account, and
· where appropriate, establishing starting positions at low pay, with rapid progression to higher levels.

7.1.5 Departmental management
Departmental managers also pose special problems for the employer. Senior managers - directors and up - constitute around a third of a per cent of all public servants; assistant and deputy directors are less than a per cent.

Senior managers, in particular, play a critical role in setting the strategic direction of the public service. For this reason, they should not be covered by normal collective agreements. Rather, the public service must design conditions and careers in accordance with their individual responsibilities and potential.

Specifically:
· all managers should be employed on performance contracts, although the term need not be fixed, and
· the DRSA must take greater responsibility for ensuring the appropriate deployment and redeployment of managers, in order to ensure an increasingly competent and skilled cadre.

The DPSA will sponsor an investigation by independent researchers into conditions of service for senior managers. The study will determine what the public service must do to retain and attract top level personnel for these positions. The analysis will relate to both the material and non-material aspects of managers' working lives.

7.1.6 Negotiating changes in grade
Sectors will replace the PAS with more modern and efficient agreements on pay and grading.

Guidelines for negotiating grading must ensure that where negotiations affect only one department or sector, they will be negotiated in that department or sector. If garding affects more than one sector or department, the negotiation forum will depend on whether the occupation is inherently coherent or in fact very diverse. Where an occupation is covered by more than one council, the employer side will have to ensure a co-ordinated position to avoid competition for skills.

Appendix A gives more extensive options for restructuring negotiations in the public service.

7.2. Benefits
In principle, benefits should aim to meet specific needs as cost-effectively as possible, rather than providing fairly artificial incentives for long service. Retention in the main professions will be better served by appropriate pay progression. Still, it may prove impractical to reform some existing benefits quickly or at all. In any case, all benefits will be reviewed to ensure that they contribute adequately to redeployment where restructuring requires it.

7.2.1 Pensions
Detailed proposals on pensions will be developed on the basis of more in-depth investigation under the leadership of the Department of Finance.

7.2.2 Medical assistance and the home-owner allowance
New schemes for medical assistance and the homeowner allowance should:
· involve collective procurement, in order to reduce the cost to both the employer and employees,
· reduce negative effects on medical care and the housing market,
· provide specifically for rural and lower-level employees and, in the case of housing, for township residents, and
· ensure that take-up and therefore overall costs more predictable.

The following schematic representation indicates key choices in designing benefits that can choose between external and internal providers. Given a large organisation like the public service, which can exercise market power and take advantage of economies of scale, greater individual choice also goes hand in hand with higher costs. In-house schemes will, however, impose considerable administrative burdens on the government, given the large numbers to cover.

In these circumstances, collective buying appears the best option. Middle and upper level employees would have a slightly more constrained choice of suppliers and products. But they could make considerable savings and would retain equal services. Efficiency gains and appropriately designed packages would let the public service direct additional resources to lower-income and rural employees.

To identify the needs in terms of health care for lower-income and rural public servants, the DPSA will undertake a survey in the course of the next six
months.

The new medical assistance programme will involve:
a. framework contracts with selected service providers, to ensure
· moderation in overall costs, effectively delinking medical aid from the medical price index. and
· maximum synergy with the public-health system,
b. defined packages for lower-level, rural and HlV-positive employees that accord with new legislation, and
c. an autonomous regulatory agency with a board consisting of employer and employee representatives, which would review the efficiency and
appropriateness of service providers.

The new system will apply to retired public servants as well, reducing the cost of post-retirement medical aid. In addition, employees will have to belong to medical aids for five years to receive post-retirement assistance.

The reformed homeowner allowance will negotiate a package with financial institutions to reduce the interest rate for public servants, and link the interest rate for the allowance to that rate rather than the market rate;
· shift guarantees on home loans for public servants to the pension fund, and eliminate government guarantees on This kind of loan;
· ensure that employee receive no more than the interest on a R70 000 bond over a normal term; and
· develop alternative schemes to address housing for low-income and rural employees.

7.2.3 Benefits and restructuring
In the past, primarily in an effort to retain professional groups, benefits hindered the mobility of employees. At the same time, the public service lacks benefits to enhance mobility. These benefits would include training and retraining provisions targeted to employees in vulnerable jobs, as well as intensive placement services.

To enhance mobility,
· Appropriate transfer mechanisms be de signed for pensions.
· Relocation costs will include only basic actual costs, which sectors will define within a national framework.

7.2.4 Overtime and the smaller allowances
1. Departments that spend over 5 per cent of their salaries on overtime must take steps to reduce overtime to normal proportions by re-organising work, giving upgrades if necessary, and defining appropriate numbers of positions.
2. The DPSA will help sectors define cost-effective ways to comply with the BCEA, and establish models for policies and controls on overtime.
3. Sectors will review smaller allowances and, where practicable, incorporate them into base pay.

7.2.5 Leave
Vacation leave should be a right and meet its fundamental aims - providing for necessary relaxation - and not act as a source of additional retirement income or an incentive for long service. Retention in the main professions should occur through appropriate pay progression. Similarly, sick leave should be used for illness, not extra time off. Generally, the BCEA will establish the minimum norms for leave.

These principles mean
· Leave will be calculated only in terms of working days, and no additional days will be granted for employees with more than ten years service.
· A ceiling will be placed on the accumulation of leave, although employees may keep the leave they have already accumulated.
· Sick leave would distinguish between normal sick leave - about seven working days a year- and long-term leave that would require greater certification.
· Employees will have to produce a certificate any time they take sick leave on Monday.
· Employees will receive family leave as required under the BGEA, and unpaid maternity leave in addition to the current paid leave to make up four months for every confinement.
· The DPSA will develop simple information systems to improve record keeping and controls on leave, set norms for cases when no records exist.

Appendix A. Some options for negotiations
Various debates have emerged about the appropriate structure and procedures for negotiations in the public service. In effect, they relate to the desirability and nature of co-determination in the public service. A separate policy process on labour-relations in the public service will address these questions in more detail. These notes serve mostly to point to problem areas and options.

1. What should be negotiated, what consulted, and what constitutes management prerogative.
The LRA does not define mutual interest, leaving considerable discretion to labour and management to define what issues to negotiate, and what to consult. In this context, negotiate refers to a process ending in a binding agreement or a dispute; consultation involves considerable discussion, but failure to agree does not lead to a formal dispute.

Areas that require consultation or negotiation on at least some aspects include:
· salaries, benefits and allowances,
· labour relations issues, such as the structure of negotiations forums, shop stewards, and so on, and
· work organisation, and by extension grading and possible job loss.

Most remuneration and labour-relations issues must be negotiated, although consultation may be possible on some aspects. Issues related to work organisation leave greater scope for managerial decision making and consultation.

In some parts of the country, managers negotiate trivial issues of procurement or implementation of agreements on benefits with employees. Yet they fail to consult labour about dealing with redundancies, where the LRA generally requires consultation.

In recent years, too, the government has negotiated some frameworks for work organisation at central level. These frameworks have ranged from fairly rigid - the rightsizing procedures and freeze on retrenchment included in the 1996 agreement, in particular - to the general; the 1998 service and skills audit process essentially only defined a process of research, with no requirement to act on the findings. Some observers argue that this kind of agreement undermines the ability of public-service managers to organise work as required to improve service delivery.

These experiences point to the need for guidelines to ensure that managers do not shift the burden of management onto labour, since that causes long delays and inappropriate decisions. But they must also define clearly when consultation is needed, and when managers must formally negotiate.

Ultimately, the extent of consultation on issues around work organisation largely defines the relevance of co-determination. If managers deem it expedient, they may consult on a broad range of issues - but, in legal terms, they do not have to. The LRA requires employers to consult on retrenchment, and generally regrading requires consultation or negotiation.

The decision on whether to consult must balance two conflicting imperatives.

· On the one hand, the democratic government must make and carry out policies in light of broad social mandates. From this standpoint, public servants form only one stakeholder. It makes no sense to subordinate policies on, say, staffing levels for schools primarily to the desires of employees. That kind of policy must involve the full range of stakeholders - in this case, including parents, children, other training providers and future employers - and the final say belongs to the democratically elected government.

· On the other hand, the realities of power mean that measures that require the support of employees for implementation may be more effective if their organisations agree to them. Agreement with organised labour, whether through consultation or negotiation, makes possible cheaper and more efficient measures to bring about changes in work organisation, and strengthens mobilisation in support for transformation. For instance, the recent redeployment exercise in education was vastly facilitated by the agreement on procedures at the Education Labour Relations Council.

A final consideration lies in the commitment of the government to participatory democracy. That means that where possible, broad-based orgadisations should be consulted.

2. How does the budget relate to negotiations?
Difficulties arise from the relationship between the budget process and negotiations. If the budget publishes a fixed overall amount for annual salary improvements, that effectively becomes the employer's starting offer. Yet increases above the budgeted amount will disrupt the budget process. On the other hand, if the employer keeps the amount secret in whole or in part, it undermines the transparency of the budget process.

Options include.
· maintaining the existing system, without publishing the full amount available for improvements in salaries every year, so that the employer effectively keeps a contingency reserve for negotiations;
· establishing the overall amount budgeted after representations by organised labour, and then negotiating increases at the sectoral level, without permitting an increase in the overall amount;
· setting salaries based on expert advice in each occupation or sector, after representations by organised labour, but within the amount budgeted; and
· eliminating central allocations for improvement in conditions of service, and leaving it to departments in sectors to define how much they can afford to raise salaries.

3. Where should issues be bargained or consulted?
Finally, the problem arises of deciding what should be negotiated centrally, sectorally, provincially, departmentally or institutionally.

The negotiation structure is currently in a state of flux. Until around 1997, negotiations took place in independent Sectors defined mostly by occupation essentially education, police and a residual sector combined with provincial councils. The current structures involve a co-ordinating council, the PSCBC, with a set of stronger sectoral councils defined by employer in three cases health and welfare; safety and security; and the General Public Service Sector and educators.

The process of changing the structure of negotiation forums arose because:
· sectors generally have more in common for negotiation than provincial departments, and
· unplanned differentials in remuneration based on geography seem likely, in a developing country like South Africa, to lead to brain drain and competition for scarce skills between government departments.

The current process leaves several questions open.

1. Currently, the PSCBC negotiates a single salary scale, while the sectors and departments negotiate changes in grade.
· Should sectors establish separate salary scales? To ensure more appropriate salary systems for sectors will require considerable developmental work. A project is currently underway for educators, but not for other occupational groups.
· If sectors establish separate pay scales, how should funds be allocated between them? The State as employer could allocate the funds between sectors as its sees fit, taking the negotiations at sectoral level into account. Alternatively, negotiations at the PSCBC could set a framework by setting the minimum and/or maximum salary; establishing an average increase for each sector; and/or approving agreements reached on pay scales at sectoral level.

2. As sectors establish provincial chambers, the scope will overlap with provincial councils. Two options emerge, that may be combined:
· the provincial councils can become the chamber of the General Public Service Sector, and
· provinces can establish a co-ordinating body in which representatives of provincial sectoral chambers can discuss cross-sectoral issues.

3. The LRA deems all provincial and departmental negotiation forums to be bargaining councils initially, subject to approval by the PSCBC. It seems more appropriate to establish sectoral chambers and departmental bargaining forums on agreement. That would give the parties greater flexibility in deterring the responsibilities of these forums in dispute settlement and negotiations.

4. The public service has increasingly moved towards industrial labour relations for all public servants. These procedures work well for large, homogenous occupations, such as elementary workers and the major professions. But they impose excessive rigidity on small, high-level professions and senior management. We need to explore options for more

5. Within the current structure of sectors at the PSCBC, the General Public
Service Sector effectively acts as a residual sector. As a result, it contains many small occupations and 200 departments. Around 70000 employees, or a quarter of the sector, are' non-educators working in education; 30 000 work for correctional affairs.

Subdivisions could be made
· between provincial and national departments, or
· according to the major productive sectors- that is, agriculture, water and forestry, public works, transport and housing, and the regulatory sector.

In either case, education employees should move to the education sector, possibly as a separate bargaining unit from educators. Similarly, correctional service employees should either join the safety and security sector or form a separate sector.