143. This Report has already documented the extent of abject poverty in South Africa. While the association between poverty and unemployment is widely recognised, it is less frequently realised that low pay and poverty are widespread even among South Africans with jobs. Ten percent of all regular, formal sector employees earn less than R250 per month. And 38% of South Africa's approximately 7.4 million regular workers earn less than the household Minimum Living Level (MLL), currently averaging R970 per month.

144. Very low pay is especially common in agriculture and domestic service. It is also widespread in the service and construction sectors and in pockets of mining, manufacturing and other industries. Black workers and women workers are more likely to be found in low paid employment.

145. In most countries mineworkers earn more than their manufacturing counterparts. This is because the work is dangerous, exhausting and dirty. White miners' pay in South Africa has historically complied with this pattern. However African miners earn less than their manufacturing counterparts. This is a continuing legacy of the migrant labour system.

146. Inequality is also deeply embedded in employment. A notable feature of remuneration in South Africa is the marked inequality in earnings between those at the top and those at the bottom of particular enterprises. In the public service, despite recent efforts to increase the minimum wage, the most senior civil servant still earns 20 times the salary of the lowest level employee. And differences in the private sector are also substantial. While it was difficult for the Commission to obtain detailed data, indications are that a typical mine manager earns more than 25 times what the lowest labourer earns; and even more if casual labour is included. Differentials are probably higher than in most other countries, developed and developing. These are a reflection on the one hand of structural, apartheid-based inequalities in access to education, training and jobs and the related shortages of skilled professional and managerial personnel, and on the other hand, of an overabundance of unskilled labour.

147. The Commission was mandated to take into account the provisions of the Reconstruction and Development Programme (RDP). This calls for a living wage for rural and urban workers and the reduction of wage differentials. It does not specify what monthly figure constitutes a "living wage".

148. The MLL, calculated by UNISA's Bureau for Market Research, is the most commonly used measure of basic living costs. It calculates living expenses for an average size household in various cities and medium-sized towns. The MLL is not an extravagant estimate and includes little more than the cost of rent and basic services, provision for the family to meet the minimum dietary requirements recommended by the Department of Health, schooling expenses, some clothing and washing materials, and travel expenses to and from work. A slightly more generous calculation emerges from the Supplementary Living Level (SLL) which adds items such as pension, unemployment and medical aid contributions, some extra toiletries and washing materials, provision for non-work transport and some recreational expenses. In August 1995 the MLL for the average African family stood at R970,38 with figures for specific cities or towns varying within a range of approximately 10% above or below this. The SLL for an average African family was estimated at R1 307,84. The average family size used in the UNISA calculation was between four and five people.

149. The MLL calculations suggest that the cost of living for workers, the poor and the unemployed in South Africa is high for a developing country. Food prices and transport costs are widely held to be a major culprit. The poorly developed informal sector is also thought to inadequately serve low-income consumers relative to their counterparts in countries at a similar stage of development. The relatively high cost of living places definite obstacles on the downward flexibility of unskilled wages, even with the high level of unskilled unemployment in the country. Arguments which emphasise South Africa's relative wage costs or which suggest that wages are "too high" need to take into account the relatively high income required to keep a low-income family from starving.

150. For a variety of reasons it is difficult to make accurate and conclusive international comparisons of wages and productivity and, hence, unit labour costs. But the available evidence does suggest that average unit labour costs in South Africa are higher than comparable middle-income developing countries. This obviously impairs our ability to compete internationally. However, the policy conclusions to be derived from this are by no means clear cut.

151. In analysing aggregate wage levels it is important to note that the macro and industry-level data covers all remuneration from the lowest blue-collar wage to the highest managerial salary. The aggregate figures will therefore disguise and distort the reality of what are effectively First World living standards and salaries at the top end, combined with developing country wages and living standards at the bottom end. In short, the impact on average unit labour costs of depressing unskilled wages may be relatively small. However, this will vary across the economy depending on labour-intensity and on the contribution of wages (and especially unskilled wages) to total costs. The fact that increased international competition has placed a strong burden of adjustment on South African firms utilising a high proportion of unskilled workers indicates that unskilled wages in certain segments of the South African economy may well be comparatively high relative to productivity.

152. Some analysts have blamed "high unskilled wages" for South Africa's lack of competitiveness and its high level of unemployment. The issue is clearly more complex than common sense orthodoxy assumes. This important caveat notwithstanding, the level of unskilled wages ­ and, of course, the productivity of unskilled workers ­ cannot be omitted from a discussion seriously concerned with mass unemployment amongst this layer of the labour force. We accept the argument that the higher-skill, higher-wage industrial sectors are unlikely to generate large numbers of new jobs; future formal sector employment growth is most likely to occur in lower-wage sectors such as tourism and clothing. We have elaborated an industrial strategy that seeks to achieve this. We have already emphasised that, in these sectors wage increases which over time exceed productivity improvements are not sustainable. However it would be simplistic to conclude that wages can simply be lowered, or that job growth would automatically follow a reduction in wages. Wage repression may simply breed resistance and result in increased industrial and social unrest, which would in turn discourage both growth and new employment. Such an approach would also fly in the face of government's attempts to build a consensual social partnership with organised labour and business.

153. One of the most complex issues that the Commission had to consider was the potentially negative impact of wage levels and wage growth on job creation. Two recent studies, one undertaken by the World Bank and one prepared for the Commission, have attempted to calculate the wage-employment elasticity ­ the correlation between wage increases and job creation or destruction. Such studies are subject to considerable debate on econometric and methodological grounds, and are sharply criticised in the ILO Review. However the broad thrust of their findings should be noted. Both studies, using significantly different methodologies, estimate long run elasticities of approximately -0.7. This figure is not particularly extreme by international standards, but is worrying in a context of high unemployment.

154. What these figures mean is that wage increases impact negatively on employment to the extent that, over the long term, a 10% increase in real wages is associated with a seven percent decline in employment levels. It should be noted that the data: embrace all remuneration (both wages and salaries) in the definition of wages; measure the relationship between two factors, using historical trends and holding other factors constant; and (in one of the studies) measure African wages only. While the econometric estimation of elasticity makes it the same for wage increases as for wage decreases, in practice wage increases and decreases may have employment effects of different absolute magnitudes. There is thus no guarantee that a 10% decline in wage levels will lead to a seven percent increase in employment. The studies do however suggest that, in the long run, employment levels are sensitive to changes in wage levels. And it implies that a range of measures are needed to contain increases in unit labour costs. The study prepared for the Commission concluded: "... [I]t would be unwarranted to conclude that factor price strategies for expanding employment are the only effective approach [nevertheless] it would not be prudent to undertake policies without taking account of the likely employment impacts of wage changes".

155. While there is therefore, to some extent, a wage-employment trade-off it does not follow that unemployment is simply caused by "high" wages. First, we have already elaborated the role played by macroeconomic and industrial policies in low employment growth. Contractionary and uncoordinated macroeconomic policies and industrial policies that reward capital-intensive investments will impact negatively on employment growth independently of the operation of any wage-employment trade-offs. Second, anecdotal evidence suggests that many employers are reluctant to employ large numbers of unskilled workers because of the "hassle factor" ­ the time and management energy involved in dealing with a demanding workforce aware of its rights. This reinforces the importance of a structured set of institutions able to enhance shopfloor stability. Third, in some parts of manufacturing and services, technological imperatives underpin the slow growth of formal employment with wage levels being a relatively minor player in determining choice of techniques. Fourth, shortages of skilled labour are a recognised constraint on economic growth. Labour-intensive production also requires skilled employees (especially managers) and shortages may act as a bottleneck here as much as in the more capital-intensive sectors.

156. In short, while wage levels in some sectors may be discouraging job creation, a range of other factors are equally important. The most obvious source of sustainable employment growth is strong economic growth. But, as already elaborated at length in earlier chapters, this in turn requires a range of associated policies and institutions in order to maximise the employment generated.

157. In general then, wage levels are influenced by a complex variety of factors including market forces, the nature of specific sectors, the standard and cost of living expectations of employees, the strength of unions, technology, and the nature of work organisation, to name some of the most obvious and important. In South Africa, of course, the wage structure has been greatly affected by apartheid which significantly determined the racial and gender composition of employment in the various grades and occupations. The earnings assigned to grades and occupations were then crucially influenced by the race and gender of those employed in the grades and occupations in question.

158. Wages are also determined through a variety of institutional mechanisms with differences depending on the industry, region and occupational categories involved. Internationally, bargaining is commonly analysed at three major levels ­ national, industrial and enterprise ­ although in practice the picture is always more complex. In South Africa there is no national-level cross-industry wage bargaining between trade union and employer federations. Local and enterprise negotiations are widespread. Beyond this there is a veritable patchwork of bargaining arrangements with dramatic differences between industries, companies and regions.

159. A number of industries ­ including clothing, textiles, auto, engineering, furniture, and recently the public service ­ have industry-level bargaining through industrial councils (to be known in future as "bargaining councils") or similar bodies. This bargaining sometimes takes place at national level and sometimes in regional industrial councils. Agreements vary in occupational and geographical scope, with minimum rates or actual increases or percentage increases generally extended to non-parties within the defined area of jurisdiction of the council. Some bargaining councils accommodate supplementary plant-level bargaining through house agreements. Others determine actual rates and discourage further enterprise-based bargaining. No agreements cover the whole country; the former TBVC and SGT "border areas" such as Botshabelo or Dimbaza are generally excluded and in practice pay significantly lower wages. This may change with the integration of labour law. In some industries, notably construction, industry bargaining has been undermined by the increasing use of labour brokers, casuals and sub-contractors. In construction differential conditions for casuals have been introduced recently in an attempt to regulate the situation.

160. In most of the service sector, transport and telecommunications, finances, and in large parts of manufacturing (such as food, printing, paper and chemicals) wages are determined largely through enterprise or company-level bargaining. In some parts of these sectors, such as chemicals, security services and goods transport, industrial bargaining has recently emerged or is about to do so. For those gold and coal operations organised by the National Union of Mineworkers (NUM) or other unions active in the mining industry, and falling under the Chamber of Mines, an industry-level agreement provides the framework within which actual wages are negotiated at the company or mine-specific level.

161. Wages and conditions in non-unionised enterprises and sectors take the form of individual contracts, often determined unilaterally by management. This is the predominant pattern in agriculture, domestic service and in smaller enterprises in the services sector, as well as among casual and unregistered employees. Remuneration packages for higher-level occupations and white collar employees generally take the form of individual contracts.

162. The unevenness of institutional arrangements governing wage determination inhibits the possibility of co-ordinating labour market policy with industrial policy and macroeconomic policy. However, this problem will not be confronted overnight by legislative edict. It is the product of complex historical and social processes. This Report will recommend that strong encouragement be given to sectoral bargaining as a mechanism for establishing minimum rates and conditions and that actual rates be established through enterprise-level bargaining.

163. Unions, not surprisingly, affect wage levels through bargaining with employers. Union organisation and collective bargaining are a necessary counterbalance for workers in a market economy where employers have greater economic and social power. The Commission strongly supports free collective bargaining between employers and employees.

164. However, for all the vaunted power of the South African unions and the distorting impact that they are alleged to have on wage structures and patterns, these exhibit many of the hallmarks of market determination, manifest, for example, in significant wage differentials between sectors and between grades. Indeed, all too frequently, the unions appear to have been markedly unsuccessful in overcoming apartheid-based differentials. Some studies have argued that a measure of the union's impact is in a large differential between workers covered by statutory bargaining arrangements, on the one hand, and those who are not covered, with the latter used as a measure of what workers are "willing" to accept, the former a measure of the unions' ability to extract a premium on the "market rate". While there is much about this analysis that is questionable (differences may reflect the unionisation of higher skilled workers or the gains of increased productivity) and much that is unexceptionable (unions explicitly aim to bargain increased wages), it does direct the Commission to focus on the impact of the extension of bargaining council agreements to non-parties.

165. The Commission believes that labour market policy, and this naturally includes the collective bargaining system, should reinforce the national goals of enhancing employment and productivity, and tackling poverty and inequality. In doing so there are difficult policy choices to be made. Poverty cannot be eliminated by legislative edict, nor can productivity be easily enhanced when extremely low wages are paid or when earnings inequalities are manifestly extreme. There may be trade-offs between promoting greater equity and industrial peace for those in employment and increasing access to employment by those without jobs. The Commission's recommendations in relation to wage determination are intended to be practical and incremental, balancing the various goals of post-apartheid development policy.

166. The Commission strongly believes that collective bargaining should be encouraged. Bargained agreements between unions and employers can generally lead to realistic and agreed minimum conditions being determined since they involve the parties directly affected. Collective bargaining is, therefore, always the first-preference system. The Commission also believes that there is some justification for the judicious use of legal intervention - using the mechanisms of the Wage Act and the LRA's provisions allowing the extension of bargaining council agreements - to plug gaps in the bargaining system. At the same time, a collective bargaining system that damages employment will have failed as dismally as one that produced poverty among the employed. The Commission was particularly concerned that the extension of agreements to non-parties does not afford those affected by this arrangement the opportunity to influence it. For this reason the Commission has reflected deeply on this aspect of the collective bargaining system.

Collective Bargaining

167. The Commission recommends that collective bargaining be promoted. We accept the framework of the new LRA which encourages bargaining at both industry and at enterprise (or company) level. The new labour law envisages an increased role for industry-level bargaining councils. The Department of Labour should encourage these and, if necessary, facilitate the establishment of such councils in industries where they do not exist. It should also support moves aimed at rationalising the bargaining system. Bargaining councils can play a vital role not only in determining minimum wages and conditions and promoting industrial relations stability, but also in building consensus around longer-term training and industry policy needs and in developing appropriate economies of scale for the administration of various benefit funds. Importantly they can enhance certainty by establishing defined standards.

168. The approach by the Commission outlined above is premised on the assumption that individual employers and firms are also, in their own right, accorded full bargaining rights and that these are, and remain, enshrined in law.

169. Enterprise- or company-level bargaining is a crucial supplement to industry-level bargaining. It allows actual wages and conditions to be determined (above the minima) in line with the particular circumstances prevailing in that company or enterprise. And it enables some conditions, such as the arrangement of working hours or "spreadover" times to be adjusted by agreement. The Commission recommends a bargaining system in which basic minimum conditions are set at industry level, and supplementary bargaining takes place at enterprise/company level. This latter bargaining would improve on minima by agreement and would take place in enterprises/companies where the workforce is significantly unionised. Whether the most appropriate level is the industry or the enterprise depends on the issues being negotiated and the traditional practices in that sector. It should be noted that some issues ­ such as benefit funds or industry training arrangements ­ may lend themselves to negotiation at industry level. Other issues, such as productivity bargaining, are best handled at enterprise level. And yet others are best handled through supplementary negotiations aimed at improving on or adapting bargaining council minima. Some items, such as actual wages, lend themselves to variation at enterprise level and some items do not. It is best left to each bargaining council to decide where various items are bargained, and which items lend themselves to supplementary bargaining.

170. The new labour law envisages a role for NEDLAC and the social partners in demarcating industries. In this chapter, as in common South African usage, the terms "industry" and "sector" are used interchangeably. This imposes an obligation to discuss the type of industry-level bargaining arrangements envisaged. The Commission believes a number of principles should inform a rational approach to demarcation. First, the aim should be to bring together in one bargaining forum broadly similar producers or service providers. The product market must be assessed to ensure that the industry scope is neither too broad nor too narrow. Second, it is particularly important to take into account the labour-intensity of the component parts of the industry to ensure that the same minimum conditions do not automatically apply to vastly different situations, possibly acting to discourage job creation Third, account should be taken of the actual or planned structure of training arrangements in the industry concerned. Fourth, the number of employees covered should be sufficiently large to allow economies of scale in relation to, for example, benefit funds, while not being too large such that sub-sectors with little in common are bunched together.

171. In practice it is not envisaged that, say, biscuit-making be treated as a stand-alone industry. On the other hand an all-embracing food industry could be too broad in its coverage. Where bargaining councils do cover a broadly-defined industry, it will be more important to set variable minima, where appropriate, for the different component parts and for different regions if necessary. In general, the more extensive the defined scope of an industry, the less will be the ability of its bargaining council to set uniform conditions and vice versa. This could result in bargaining councils establishing sub-councils as in the case of the textile industry.

172. The Commission believes that industry-level bargaining can play an important role in providing a stable bargaining framework and, in line with the LRA, should be encouraged. However, changing global conditions mean that in many sectors there is a need for companies to be able to respond rapidly to changing markets and technology. Therefore the Commission supports some structured decentralisation of bargaining; that is decentralisation, preferably within guidelines to promote such supplementary bargaining, set by the relevant bargaining council and monitored by it.

173. Bargaining councils should aim to set minimum conditions and encourage further, supplementary, bargaining at enterprise or company level. The council should set the rules of the game, dealing with some items centrally and channelling others to enterprise or company-level negotiations. It might, for example, set minimum conditions and outline the parameters within which supplementary bargaining could take place. The intention is, for example, to ensure compliance with minimum hours but enable the actual scheduling of hours to be bargained at enterprise or company level. This approach would also allow a bargaining council to determine the minimum hourly rate and percentage increase but enable supplementary increases linked, say, to a productivity or profit-sharing arrangement. This could promote enterprise efficiency, grant workers a share in the fruits of such gains, and set realistic minima for the less profitable enterprises, while at the same time moderating undue wage drift between enterprises within one sector.

174. Bargaining councils also have a crucial role to play in determining industry policy, ensuring appropriate training (in conjunction with the appropriate training board), encouraging the development of exports and that industry's competitiveness, and agreeing on employment-generating strategies for the industry.

175. The Commission heard arguments for and against the extension of bargaining council agreements to non-parties. Opponents of extension argue that where wage agreements are extended to non-parties who cannot afford to pay the minimum wage then employment may be damaged. Where the imposition of higher minimum wages on less productive firms threatens jobs, there is a clear conflict of interest between higher wages for some and employment for others. Because bargaining councils tend to represent workers and employers from larger, predominantly capital-intensive (often higher wage) firms, the parties to agreements do not have to carry the consequences, including the employment consequences, of extensions. From this argument it follows that wages appropriate to capital-intensive industries must not be generalised across labour-intensive industries which are more vulnerable to increases in labour costs. It has also been argued that a central tenet of democracy is that those who perform public functions (including bargaining council representatives) should be capable of being held accountable for the consequences of their decisions; and that this does not occur when agreements are extended to non-parties.

176. Proponents of extending agreements justify it largely by the stability this brings to the bargaining system and the industrial relations environment. They argue that it takes wages out of competition between firms and rules out very-low wage production or service provision, thereby penalising highly inefficient firms and encouraging better practices; both, it is argued, socially desirable objectives. Further, the extremely low wages and atrocious conditions faced by many workers (generally workers outside of a collective bargaining framework) are often caused as much by power imbalances rooted in the apartheid era, as by an oversupply of unskilled labour.

177. The new LRA allows the extension of agreements by the Minister to non-parties. The extension of agreements aims to ensure that all employers are bound by collective agreements reached in representative bargaining councils and/or where extensions are needed to encourage more stable and representative sectoral bargaining. The Minister should treat applications for extension with circumspection and exercise caution in extending agreements where there are grounds for concern that significant job destruction might ensue. The Commission believes the approach contained in the new LRA needs to be carefully monitored. The LRA virtually obliges the Minister to extend an agreement reached by over 50% of the industry, but discourages him from extending agreements where this threshold is not met. The Commission recommends that the Minister should have greater discretion in deciding whether or not to extend agreements. The Minister should take less account of the representivity of the parties to the agreement and more account of whether the agreement reached is sensitive to the problems of non-parties and the job-creating goals of the RDP. This means the Minister should also consider extending agreements reached in less representative councils if these criteria are met.

178. The debate over extensions is largely associated with government policy in support of labour-intensive production in general and small business in particular. The Commission believes there are four basic options for dealing with the special problems faced by small and medium enterprises (SMEs). One approach involves exempting firms below a certain size from all minimum standards and collective agreements. However this would run counter to the trend of establishing one legal framework for all employees and employers, would effectively limit the basic labour rights of a significant section of the workforce, and is an approach rejected by the Commission. A second approach involves the extension of collective agreements to all employers and employees in the industry regardless of firm size. This is the current practice and individual SMEs wanting exemptions can thereafter apply to a bargaining council. Currently some bargaining councils grant exemptions relatively easily, but others do not. In practice many small enterprises simply evade the regulatory net or break the applicable minimum standards. A third approach is to have only a simple, minimalist agreement (or no agreement) at industry level, leaving it to enterprise bargaining to top up low minima. A fourth approach involves setting a simpler, less onerous schedule of minimum conditions for SMEs. In general the Commission favours the fourth option since it realistically acknowledges the different circumstances of many SMEs and the dualism within the labour market, but simultaneously keeps all employers within a basic, regulatory net. Importantly, this approach recognises that we have one basic labour law covering all workers and employers.

179. If, however, the second approach is adopted by government then the variations from the minima laid out in bargaining council agreements should be relatively easy for small enterprises and start-up businesses to obtain. The new LRA gives bargaining councils the duty to define fair exemption criteria and provides for exemptions to be considered by an independent body in which small business interests are accommodated. The Commission believes that, in relation to exemptions, three points should be taken into account when applying the LRA. First, the viewpoint of employees of a small business wanting an exemption should be heard and not only, as the law currently provides, the viewpoint of the relevant employer. Second, there should be some effort to co-ordinate (possibly through NEDLAC) the criteria set by bargaining councils to ensure that these are relatively consistent across industries and that the intentions of the LRA (to remove possible obstacles to small, new businesses) are not undermined. Third, bargaining councils should define small enterprises in their sector through an appropriate turnover-related measurement rather than on the basis of employee numbers. An employment-size measurement, while easier, may discourage small businesses with relatively low turnovers from employing more people.

180. In considering how to regulate and accommodate small business three general points should be remembered. First, micro-enterprises and survivalist employment is not, by and large, susceptible to regulation by bargaining councils. Second, the definitions of small and medium enterprises will, necessarily, differ from sector to sector. The recently-released draft bill on small business provides useful guidelines in this respect. Third, exemptions and extensions will need to be sensitive to the real distinctions which exist in practice among businesses classified as SMEs.

181. It is helpful to think of exemptions as variations from the norm rather than as permission to ignore norms and agreements. The aim of an easier exemption/variation policy is to remove obstacles to small business and, hopefully, to facilitate job creation. The Commission believes variations should be partial and should exempt the applicant from aspects of the agreement for a specific period. For example, a new small business might, if appropriate, only be required to pay a percentage of the minimum wage. As the business becomes established there could be a rising floor of compliance. This approach could be used whether option two above (bargaining councils granting exemptions) or option four (a simpler, less onerous schedule for small businesses and start-ups) is adopted. An added advantage of option four is that it would reduce the need for exemptions.

182. Public support for small business by state agencies should be linked to these businesses complying, or being assisted to comply, with basic labour standards. These would include the minimum wage rates and basic conditions established by the relevant bargaining council, wage determination or other law. Similar considerations should form part of the state's procurement policies.

183. The notion of a less onerous schedule for certain categories of companies could, if agreed, also apply to trainees. The preponderance of structural unemployment and the high rates of youth unemployment is of particular concern, and is reason enough for encouraging mechanisms that would provide incentives for employers to hire and train these workers, even if there is no expectation of this leading to permanent employment. A young worker who fails to find employment becomes increasingly less employable as the years pass by. This proposal is potentially open to abuse if employers simply replace permanent workers by younger "trainees" earning a lower rate. To avoid this, appropriate mechanisms can be built into the design of such schemes. And if abuse was extensive then such youth/trainee programmes would need to be reviewed.

184. As the law provides, bargaining councils and the Ministry of Labour are responsible for enforcing agreements and wage determinations. In doing so, two points should be taken into account. First, sanctions and inspections should be considered only on receipt of a complaint regarding breach of standards from the employees concerned. Unions may help such employees to draft their complaint. This is a significant departure from current practice and aims to prevent either workers or employers in competitor establishments from initiating an essentially protective action. Second, the focus should be placed on facilitating employer compliance with minimum standards. Punitive measures should only be considered as a last resort. Where clear job destruction possibilities can be anticipated it is preferable to encourage a negotiated and incremental application of minimum standards. Wherever possible this should be done by facilitating agreements between the employer/s and the workers concerned.

185. New approaches to enforcement require appropriate training and retraining for bargaining council agents and Department of Labour inspectors. Knowledge of conciliation techniques and possible wage-employment-training trade-offs is essential.

186. The Commission's support for sectoral bargaining and structured flexibility in the application of agreements requires strong, representative unions and employer organisations. These are essential for a stable industrial relations system. The Commission recommends that the Department of Labour give financial support to a trade union-controlled education and training institute aimed at improving the skills of unionists, including shop stewards.

187. Evidence before the Commission suggests that some employer organisations are weak, underfunded and play a limited role in industrial relations. The new LRA encourages, and its success is heavily dependant on, the formation of strong employer associations. At present employer associations generally perform limited representative roles in terms of advice and annual negotiations. The Commission believes it is essential for these associations to expand their operations to include the representation of members, assisting members with conciliation and arbitration conducted under the new LRA, HRD planning and providing much of the advice on dispute resolution procedures, recognition agreements and Workplace Forums which is currently provided by consultants.

188. Some sectoral employer associations find it difficult obtaining membership from the small businesses community. When employer associations become dominated by the larger business concerns, the danger looms that they will be insufficiently sensitive to the interests of small business. Although the new LRA makes provision for the representation of small business on bargaining councils, employer associations should make a special effort to recruit small business. Employer associations on bargaining councils should also appoint a small business co-ordinator to improve the quantity and quality of small business participation. The Commission recommends that bargaining councils agree to a modest industry-wide levy on all employers within its jurisdiction to be used to achieve the above objectives. Government should also consider providing financial and other resources to these programmes. The recently established Small Business Council (SBC) established under the auspices of the Department of Trade and Industry is to be welcomed and could be approached by employer organisations to identify small business counterparts in their sectors.

189. While the Commission acknowledges that collective bargaining will occur primarily at sectoral and enterprise/company level, the growing requirement to co-ordinate economic policy suggests a likely role for national agreements involving business and labour and, where necessary, government. This is, to a large extent, the function of an Accord for Employment and Growth ­ dealt with elsewhere in this Report. Developments in this direction will inevitably draw into the bargaining arena issues such as industrial policy, human resource development strategies, training objectives, social insurance and other issues which may impact on labour costs. As noted in Chapter 10, the bargaining agenda may be expanded by complementary sectoral and regional accords.

190. The Commission believes that innovative ways need to be found to reduce unit labour costs without reducing the standard of living of ordinary lower-paid workers or compromising job creation. We have not had time to explore this fully, but submissions to the Commission suggest that indirect wage costs often add significantly to the wage bill without a commensurate improvement in living standards for employees. For example, more cost-effective provision of retirement, medical, unemployment and accident benefits could be explored. Where these exist they may add as much as 30% to total labour costs, indeed sometimes even more. Again, the range of social wage issues will be more effectively incorporated into collective bargaining through the mechanism of broader based accords.

Minimum Wages

191. The Wage Act was introduced in 1925 as a companion to the Industrial Conciliation Act. It established a Wage Board, a body appointed by the Minister to make recommendations on the determination of minimum wages and working conditions. The original aim was to provide a minimum floor predominantly for white workers in those sectors and areas where there were no industrial councils.

192. As documented in the ILO Review, the Wage Board currently sets minimum wages and other basic minimum conditions for approximately 730 000 workers, including both unionised and non-unionised employees. In practice the Board has become less active in recent years. Its coverage has decreased, its determinations are generally updated infrequently, the minima it sets are usually low, and determinations are rarely enforced by the Department of Labour's inspectorate. There are no minimum determinations in the lowest paying sectors ­ agriculture and domestic service.

193. The Basic Conditions of Employment Act (BCEA) stipulates the minimum conditions of employment for the economy as a whole excluding the public service. Conditions covered include hours of work, overtime pay, Sunday work, annual leave, sick leave and public holidays. These conditions are fairly rigid in their formulation although the Act provides for variations in three ways. First, an exemption may be given through an administrative procedure handled by the Department of Labour. Second, a promulgated collective agreement concluded by an industrial council can trump the BCEA. Third, to the extent that a Wage Board determination deals with matters falling under the BCEA, these also trump the Act.

194. The Commission gave extensive consideration to the desirability of setting minimum wages and received numerous submissions on the issue. Arguments against were mainly that a minimum wage would lead to increased unemployment. Firms unable to afford the minima would either go out of business or make do with fewer workers; employers would be encouraged to use more capital-intensive techniques; and there would be increased inflation which would in turn undermine the minimum. The thrust of the argument against minimum wages was that poverty is best alleviated by creating employment, even low-wage employment.

195. In submissions received from opponents of a minimum wage, a more complex picture emerged. Farmers' organisations argued that a minimum wage in agriculture would be impossible to enforce and could lead to substantial job losses. They stressed the enormous differentiation in conditions between crops and regions. They conceded that agricultural employment was declining, even in the absence of minima, and attributed this to mechanisation, uncertainty around land reform and state assistance, as well as pressures to become more internationally competitive. Organised industry and commerce stressed the needs of small business and the disemployment effect of minimum wages. They did, however, accept ILO protective minimum standards conventions and the need to avoid exploitation. They argued that the real issue would be the level at which minimum wages should be set, although they did not recommend a figure. Some stressed that the focus should be on regulating the non-wage components of minimum conditions. In a written submission the South African Chamber of Business stressed that if minimum wage setting was adopted "it would need to be highly differentiated and narrowly targeted and vary between sectors and regions of the country". Employer submissions generally preferred to see minimum conditions being set through collective bargaining.

196. The case for a minimum wage relied on a variety of arguments ­ it would assist low paid workers; it would prevent firms undercutting each other solely on the basis of low wages; it would increase consumption demand in the economy especially in labour-intensive sectors; it would reduce labour turnover and unrest; and spur employers to make productivity increases. Those in favour contested the argument that significant unemployment effects would necessarily result. The primary aim, argued minimum wage proponents, is to prevent vulnerable workers from being exploited.

197. Minimum wages were strongly supported by organised labour. One farmworkers union reported on the poor wage and living conditions faced by farmworkers and called for a minimum of R750 p.m. The union argued that minima would need to be varied by sub-sector (crop) and accepted both the phasing-in of minima and the need for some type of exemption system. The major mineworkers' union called for a national minimum targeting the bottom 10% of workers and argued for a minimum of R650 p.m. A domestic workers' union spokesperson complained not only of low wages (and even instances of no wages) but also of long working hours, child labour, sexual harassment, payment in kind and the general lack of employment contracts. She called for a minimum of R600 p.m., conceding that it could lead to some job loss. Two major union federations declined to put a figure to their minimum wage call, stating that any determination should take into account both the need to end exploitation and the need to maintain macroeconomic balance. In general they stressed the need to avoid sweatshops and argued that any minimum wage should be used to reinforce collective bargaining.

198. There was surprisingly little research to assist the Commission in assessing the potential impact of a minimum wage in South Africa. Studies of wage employment elasticity (discussed above) suggest a significant negative correlation between the two. But these measure average relationships and do not necessarily hold for wages at the lowest level.

199. Internationally, recent studies suggest that the evidence against minimum wages is "surprisingly fragile", even in less developed countries. At best the argument that there is a negative relation between employment and minimum wages is inconclusive. And some studies have even shown employment rising in response to an increase in minimum wages. In sum the international evidence suggests that minimum wages need not be employment-destructive provided they are set at realistic levels and do not attempt to regulate the conditions of too large a percentage of the workforce.

200. On balance the Commission is in favour of the principle of determining minimum wages and conditions, whilst modernising and adapting the systems used and the scope and extent of coverage.

201. Setting minimum wages and conditions can prevent socially unacceptable levels of exploitation. Practices such as child labour, unfree labour, and pitiful wages and working conditions must be legislated against. They have no future in a democratic society. Wage costs at the low end of the spectrum are only one component, and often a small component, of total production costs. In agriculture, for example, total wage costs (including both low-paid and high-paid employees) comprised only 21% of total costs on average in 1995.

202. It must be stressed, however, that minimum wages are not the solution to poverty in South Africa. Minimum wages are simply one, relatively modest instrument in any comprehensive multi-pronged attack on poverty.

203. The Commission believes the Wage Board should set minimum wages and working conditions in sectors (or parts of sectors) where no effective collective bargaining exists. The setting of minimum conditions should not be used as a substitute for collective bargaining. Indeed minimum standard setting should facilitate the transition to collective bargaining wherever possible.

204. The Commission does not believe it is realistic to set one national minimum wage. Rather, sectoral minima should be set. And while there is a need to avoid too complex a set of minimum determinations, there may at times be a need to differentiate geographically or to set a cross-sectoral minimum for a specific locality. In practice, during the transition to industry bargaining, the Wage Board will of necessity hold hearings as the need arises, on the request of the Minister. The outcomes will not always be, nor should they be expected to be, neat and tidy.

205. The aim of setting minima should be to prevent extremes of exploitation and to assist workers who may otherwise be at the mercy of their employer. In general there is a need to focus on setting minimum wages and conditions rather than making detailed or complex determinations regarding all aspects of work and all categories of employees. In cases where minima have the potential simply to drag down the wages of more skilled employees in the sector then minima for a few key employment categories should be established ­ for example specific minimum rates for general farmworkers, tractor drivers and supervisors.

206. There may be scope, at some future point, for the social partners to agree at national level on the parameters of national and regional wage policy, and on a national minimum wage. This is discussed elsewhere in the context of the Accord for Employment and Growth.

207. The Wage Board should be the instrument for recommending appropriate minimum wages and working conditions. These should be forwarded to the Minister of Labour for approval prior to being gazetted. The Minister should gazette all recommendations timeously or publicise his/her reasons for not doing so.

208. The Board should ensure interested parties are informed prior to holding hearings. The Board should involve regional or multi-stakeholder institutions where these exist in its information-gathering exercises and in its deliberations. It should also encourage written and verbal submissions, hold its hearings in public and make its proceedings accessible to ordinary workers and citizens by providing translation in an appropriate official language. The Board should publicise its recommendations and give reasons for its findings.

209. In reaching its decisions the Board should be guided by the need for a fair wage and decent conditions, the financial circumstances of the employers, and the social conditions and economic prospects prevailing in the sector/locality concerned. It should be particularly mindful of the employment implications of its recommendations.

210. The Commission does not wish to make detailed recommendations regarding the structure and operations of the Wage Board, other than making a few general points and observations. The existing Board needs to be modified and modernised in a number of respects. Its scope should include all sectors of the economy, although in practice it would not operate where effective collective bargaining exists. And since it deals with wages and working conditions, the Commission endorses the proposal in the Green Paper on Employment Standards to consolidate the Basic Conditions of Employment Act and the Wage Act into one piece of legislation, and to give the Board a more appropriate name. It is not envisaged that minimum conditions should apply to highly skilled, managerial and professional employees, or those earning above a specified income. These employees can rely on individual or collective bargaining to protect their interests.

211. The Wage Board should consist of full-time members appointed by the Minister, after consultation with NEDLAC. Appointment criteria would need to include some or all of the following ­ economic and social policy expertise, ability to communicate in a number of official languages, and an appropriate degree of independence from unions and employer organisations.

212. The Wage Board should at all times facilitate collective bargaining, encouraging parties to reach agreements rather than rely on Wage Board determinations. Individual Board hearings for a specific sector/locality should be convened by one or more Board members. Wherever possible an employee and an employer representative (and where appropriate an elected community representative) drawn from the sector/locality under review should act as assessors. The instrument of a wage order can be used to ratify agreements reached between employers and employees, as has been done recently in parts of the retail industry. A Wage Board determination might also take the form of an extension of a bargaining council agreement.

213. In general the Wage Board needs to be made a stronger, more efficient body. Its role should not simply be that of setting minima. It also has a role to play in encouraging good employment practice by companies.

214. It will be the duty of the (reformed) Wage Board to determine the actual minimum wages and conditions. The Board will need to apply its general mandate to the specific circumstances prevailing in that sector and/or locality. However the Commission notes that, in practice, minimum wages will need to be set at modest levels if employment generation is not to be harmed. Regulations which, in practice, raise the wages of substantial numbers of employees will lead either to job losses in both the short and long run, or to evasion of the minima by both employers and employees.

215. In the absence of sufficient evidence it is advisable that minimum determinations be introduced incrementally, both in terms of the levels set and the number of employees covered. This will enable the Board to feel its way towards appropriate minimum levels. It will be essential to monitor the impact of determinations on both employment levels and living conditions and to publicise these. Relatedly, the Board will need to have some capacity to collect basic data or conduct fact-finding investigations.

216. The enforcement of agreements should be handled in a manner similar to that outlined above. Enforcement should be triggered primarily by employee complaints rather than complaints from other employers; it should focus on facilitating employer compliance; and punitive measures should be used only as a last resort.

217. Given the modest and cautious approach to minimum wage setting outlined here, it is not envisaged that exemptions be granted often from minimum conditions. Where exemption applications need to be considered this should be done in a Wage Board hearing following the procedures already outlined. If the exemption is of a minor nature, a simpler administrative procedure might suffice.

218. There may, however, be a need for variations in wages and conditions to be contained within specific Wage Board determinations. Here too the Commission does not make detailed recommendations. However it is advisable for legislation (through a revised Basic Conditions of Employment Act) to define the broad categories of non-variable minima and variable minima. Non-variable minima might include a ban on child labour, prohibiting forced overtime work, setting minimum premiums for overtime hours, or ensuring that basic safety standards are adhered to. Variable minima might involve setting the parameters of certain conditions but leaving the details to negotiations at enterprise (plant) level. These might include limited, but meaningful, regulation of part-time and casual employees. They might involve lower wage minima for certain categories of enterprises or for trainees especially young trainees, or to regulate basic conditions of short-term workers on public works programmes.

219. Special thought needs to be given to the conditions of farmworkers and domestic workers, two categories of employees likely to be most affected by the introduction of minimum wages. Approximately 1,3 million people work on South Africa's farms, forests and fisheries. An October 1993 survey found that 73% of the regular, formal sector agricultural workforce earned less than R590 per month (expressed in 1995 rands). The gross monthly cash wages of the median male worker, expressed in 1995 rands, were just R303, while women received just R223 per month. Furthermore, 20 000 farmworkers received no pay at all. Many workers also receive payment in kind, and some get limited grazing rights, both of which are difficult to measure in monetary terms.

220. While the statistics may be unreliable in detail, it is safe to say that earnings in agriculture are generally extremely low. This generalisation holds true despite sectoral differences ­ between the low-paying maize sector and the higher-paying fruit sector ­ and despite variations according to farm size. Most disturbingly, agriculture is capital intensive. The apartheid government subsidised credit and capital for white farmers. This, in large part has contributed to an estimated 20% decline in total agricultural employment between 1970 and 1994, a decline which occurred in the absence of labour legislation or minimum wage regulations.

221. Wage levels in agriculture are frequently determined by the power imbalance between farmers and farmworkers, and not only by issues such as affordability. Anecdotal evidence suggests that there are many instances where similar farmers in one locality pay vastly different rates, with wealthier farmers sometimes paying less and poorer farmers paying more. Minimum wages set at appropriate levels, combined with a de-emphasis on capital intensity in most sectors of agriculture, may eliminate extremely low wages, encourage the more productive use of labour, and increase total employment in agriculture. One effect of a minimum wage in agriculture (and one with ambiguous implications) may be to reduce the number of non-South African, frequently unpaid, farmworkers.

222. Domestic workers also find it difficult to negotiate their terms and conditions. This sector is comprised overwhelmingly of black women who generally deal with their employers on an individualised basis. South Africa is estimated to have 874 000 domestic workers employed predominantly in white households but also, increasingly, in black households. Their conditions vary markedly and appear to depend heavily on the income of the employer, with those working in wealthier suburbs generally earning more. Many domestic workers live on the premises but are often not allowed to have their children or partners visit them. Many receive some payment in kind, often in the form of food or rations. Hours of work are frequently long and the span of hours where the domestic worker is expected to be present is often excessive. Overtime hours are common although overtime pay is rare.

223. Average earnings for domestic workers are low. While the Commission favours setting a minimum wage for domestic workers it is aware that there are many difficulties associated with doing so. These include enforcement, potential disemployment effects, and the possibility of realistic minimum wages being seen as maximum or desirable levels. The vast difference in average domestic earnings between Sandton and Sandspruit may be impossible to contain in one set of minimum wage regulations. In the circumstances the Wage Board might consider a number of minimum determinations with local authorities making recommendations as to which determination is most appropriate for their area. The Commission recommends that local authorities consider establishing advice offices where domestic workers can receive assistance in formalising their contracts and ensuring they are paid at least minimum rates.

224. While we support the determination of a minimum wage for domestic work, enforcement problems and the very low starting point for a determination, coupled with fear of disemployment effects, will probably mean that the impact of this measure on alleviating the poverty of a great many domestic workers will be slight. Creative mechanisms need to be found that will encourage employers to pay higher wages to domestic workers. One suggestion, although not unanimously supported within the Commission, is to include monetary payments to domestic workers as a tax deductible expense. Such a claim would only be accepted if accompanied by formal documentation establishing that the payments claimed had indeed been made. This may then not only encourage employers to pay somewhat higher wages to their domestic workers, it would also encourage the drawing up of formal contracts and the use of payslips, the payment of UIF, etc., thus contributing to a movement away from the peculiar set of relations that characterise this work.

225. Payment in kind is frequently a component of income for farm and domestic workers. The Wage Board will need to attach a monetary value to these and decide on the maximum amounts which can be discounted by employers. Such payments are often useful for workers. A guaranteed bag of maize provides an inflation-proof form of income and on-premises accommodation may be useful given the country's housing shortage. However, payment in kind perpetuates paternalistic relations and binds the employee to the employer. The loss of a job, for example, frequently means the loss of accommodation too. The Commission, with the exception of one Commissioner, supports the gradual phasing out of payments in kind in favour of straight-forward relationships based on monetary remuneration. However, the transition should be handled sensitively and realistically. The transition must not disadvantage the employees concerned. A distinction should be made between less problematic aspects such as a worker receiving some rations (often a small part of the farm's produce), and items such as housing and access to land where structural solutions need to be found well beyond what can be achieved by monetisation.

Pay Differentials

226. The Commission believes that pay differentials between the highest and lowest paid in South Africa are extremely high both by developed and developing country standards. The ILO Review contains more information in this respect. It is insufficient to compare wages and productivity with workers in other developing countries without also comparing the remuneration and productivity of management. In general there is a tendency for professional and managerial employees to compare their wages and living conditions with their counterparts in industrialised countries as this group tends to be increasingly mobile, while less skilled workers earn wages comparable with those in other middle or low income countries.

227. South Africa's less-skilled workers inevitably contrast their earnings with those of the middle- and upper-management with whom they are in daily contact. Large pay differentials are, therefore, not only socially undesirable in a society committed to equity, but also reinforce upward pressure on wages at the bottom.

228. Large pay differentials are rooted in apartheid inequality where whites expected a standard of living that could only be sustained if income distribution was highly unequal. Once based almost entirely on race, these differentials are now being deracialised rather than narrowed. The Commission believes South Africa must adopt the long-term goal of reducing earnings differentials substantially, not simply deracialising them.

229. However, this goal cannot be achieved by artificially implementing policies which impede productivity, employment and economic growth. Much of the distortion relates to shortages of skilled, professional and managerial personnel, enabling these categories to command far higher remuneration than is appropriate in a developing economy. In practice, little can be done to restrain the salaries of certain categories of managerial and professional personnel whose skills are tradable internationally. But the same is not true of all skilled, technical and professional personnel, particularly at lower levels.

230. The shortage of skilled labour is a recognised constraint on economic growth in South Africa. The Commission believes that sustained efforts must be made to increase the supply of skilled people in higher-level occupations where shortages clearly exist. The various ministries should thus give consideration to a number of primarily supply-side measures which would tend to reduce pay differentials (some of which are already in process) including:
     • concerted support for the reorientation of our education and training system to ensure an adequate supply of skilled personnel, especially in the managerial, financial and technical fields;

     • removing almost all restrictions on issuing work permits to skilled non-citizen personnel;

     • removing current veto powers on the issuing of work permits which many professional associations still have;

     • actively recruiting skilled personnel from other developing countries where appropriate. Many developing countries (such as India and Cuba) have an excess of skilled people in certain occupations and could supply recruits on fixed contracts at reasonable rates of pay;

     • providing guidelines to high-level professional training institutions (such as medical and engineering schools) regarding the selection of candidates and the repayment of loans, to discourage the practice of state-funded training followed by emigration and

     • pressing organised business to practice salary moderation (especially at middle management level) as part of a broader national Accord.

---