APPENDIX 7

THE REPRIORITISATION OF EXPENDITURE AND THE RDP FUND

A commitment to maintain fiscal discipline while effecting a reorganisation of expenditure in line with the objectives of the reconstruction and development programme has been a cornerstone of fiscal policy since 1994. To achieve this the government established the RDP Fund as a dedicated financing channel to prioritise key developmental projects.

Until its abolition, the RDP fund was a budget allocation scheduled to increase by R2.5 billion per year. In their first year projects appeared directly on the RDP fund vote, thereafter they were assumed under the relevant departmental vote.

The bulk of RDP funds have been allocated to social spending. In the area of urban development, the focus has been mainly on infrastructure and to a lesser extent housing, while rural development spending largely concentrated on land reform and water supply. Spending on health has been either in the form of bridging finance to provinces or in support of the primary health care initiative. The contribution to the education budget has been relatively modest, although R1 billion was set aside in the 1996/7 allocation for school building. The "other" category comprises a range of projects, often with a capacity building emphasis.

At a time of very tight budget constraints a significant increase in social infrastructure spending was achieved through the mechanism of the RDP Fund. It is indeed doubtful whether a switch in spending of this order would have been possible in its absence. The RDP Fund has now been abolished in recognition of the ability of the normal budget process to allocate resources to priority areas.

The extent of reprioritisation of the overall budget is shown in the table below which presents figures for the consolidated national and provincial budgets. The most notable change has been the decline in defence expenditure. However, the spending trends in other components of protection services has limited the scale of the relative reduction in spending within the overall function. The continued increase in the proportion of the budget allocated to interest payments, which has been compounded by falling inflation and high real interest rates, has reduced the amount available to other functions. Despite this, there has been an increase in the budget share allocated to social services. The increase in the social services allocation reflects the increase in spending on infrastructure, housing and welfare, and will increasingly come to reflect the impact of new education and health policies. The decline in economic spending has been achieved by a reduction in spending on transport, agriculture and export promotion.

APPENDIX 8

MANAGING THE PUBLIC SECTOR WAGE BILL

Fiscal discipline is synonymous with the proper management of wages and salaries within the public service, since remuneration comprises almost 50 percent of the non-interest budget. However, containment of the wage bill without maintaining adequate conditions of service for civil servants reduces the overall efficiency of government and is in the end counter productive. The Government's approach to restructuring the public service is aimed at:

• achieving equity and efficiency within the service;
• repositioning various skill categories higher up the salary scale;
• achieving parity between conditions of service in the public and private sector;
• simplifying the complicated grading system and the rigid occupational classification;
• increasing departmental autonomy in the determination of the personnel structure.

During 1996/7 the restructuring package will cost the exchequer an additional R7.4 billion, increased to R8.5 billion with savings to be implemented from July 1996. This will be financed with R6.5 billion set aside under the improvement of conditions of service vote, a supplementary appropriation of R950 million, R550 million of projected savings on the wage bill and R450 million from reduced contributions to the pension fund. The settlement takes the total wage bill from R56.7 billion in 1995/6 to R64.1 billion in 1996/7. For each of the subsequent two years the wage bill will increase by an amount of R6.5 billion, thus raising the wage bill to R70.6 billion in 1997/8 and to R77.1 billion in 1998/9. The real increment in the wage bill will be 3 percent in 1996/7, 0.5 percent in 1997/8 and decline in the following year.

The restructuring agreement provides for utilising any savings that result from the right sizing of the civil service to fund additional salary increases, although the overall increase in the wage bill must remain within the limits set out above. Right sizing will involve the reallocation of resources currently budgeted against unfilled posts, the elimination of posts that are no longer required and the removal of supernumerary officials. It is estimated that there are around 135,000 unfilled posts, of which 50,000 can be immediately eliminated resulting in a saving of R2 billion. The estimated 25,000 supernumerary officials in the provinces due to the incorporation of previous administrations will be offered severance packages. The defence force is scheduled to reduce its staff complement by up to 40,000.

The granting of severance packages is at the discretion of director generals in affected departments. Severance packages will be self financing through the application of savings from unfilled posts, the utilisation of accumulated rollovers, and other savings at the departmental level. The Departments of State Expenditure, Finance and Public Service will carefully monitor expenditure to ensure that the right sizing does not place any additional burden on the budget. Further right sizing is anticipated in 1997/8 and 1998/9 as departmental managers review their organisational structures.

APPENDIX 9

THE OUTLOOK FOR FISCAL POLICY

The fiscal component of the integrated package is a more rapid reduction in the budget deficit towards the targeted 3 percent of GDP. The intention is to reduce the deficit to 4 percent in the 1997/8 fiscal year, and thereafter at 0.5 percentage points per annum. In this appendix the expenditure implications of this fiscal strategy are analysed with the aid of the Medium Term Expenditure Model of the Department of Finance.

The results of two sets of expenditure projections are reported below, corresponding to the macroeconomic scenarios described earlier. In the modest-growth baseline projection the deficit declines to 4.5 percent of GDP in 1996/7 and at 0.5 percent in the subsequent years. The available resources for government expenditure increase at a slow 1 percent per annum from R161 billion to R167 billion in fiscal 2000/1. The faster deficit reduction associated with the integrated strategy places tighter limits on expenditure in 1997/8. However, the additional resources generated by the improved growth performance more than compensate from 1998/9 onwards, with expenditure over the period as a whole, increasing by an average of 2.3 percent per annum.

The cumulative additional resources available for public expenditure with 1995/6 expenditure as base of the integrated strategy amount to R42 billion. A continuation of current policies, on the other hand, would release a total R26 billion in additional government funds.

In order to determine the implications for line departments, a set of medium term expenditure projections broken down by economic category were generated. The results are summarised in the table below.

The remuneration items in each projection incorporate the effects of the recent agreement on public sector restructuring in term of which the nominal wage bill for 1996/7 and the subsequent two fiscal years has been fixed. Due to the lower inflation envisaged under the integrated package the real cost of remuneration will be higher than in the baseline. The differential is, however, compensated for by the lower anticipated debt servicing costs. The larger deficit and higher real interest rates incorporated in the baseline projections will mean that some 84 percent of the cumulative additional resources available for public expenditure will have to be spent on interest payments alone. This contrasts with a corresponding figure of 43 percent associated with the integrated strategy.

In both sets of medium term expenditure projections assumptions were made to effect a switch in expenditure from current transfers to capital expenditure. In each case significant spending cuts were assumed in respect of transfers to businesses (GEIS, agricultural subsidies, etc), extra budgetary accounts, local authorities, and to a lesser extent, households. In the baseline this would provide capital resources incrementing at approximately R1 billion per annum. In the integrated case, capital expenditure would be able to rise by R2.5 billion per year, sufficient to fund substantial housing, land reform and infrastructural development programmes.

The critical challenge facing fiscal policy is to ensure that the deficit target for the current financial year is not exceeded and to effect sufficient cuts in real spending in the short term to achieve a 4 percent deficit in 1997/8. To this end the Department of Finance has launched an audit of various expenditure programmes, beginning with those funded through the RDP Fund, to identify possible savings.

APPENDIX 10

THE EFFECTS OF RECENT TRADE LIBERALISATION

After a decade of gradually moving away from import-substitution policies, a far reaching programme of liberalisation was launched in 1995 aimed at increasing competitiveness. The key proposals were contained in an offer to GATT effective from January 1995 replacing all quantitative import controls on agricultural and industrial products with ad valorem tariffs, simplifying and rationalising industrial ad valorem tariffs and proposing a phased reduction of tariffs over periods varying between 5 and 12 years. This appendix considers the empirical evidence of the effect of the tariff reductions implemented thus far. The main concern is that subsectors with the largest tariff reductions would have faced the most competition leading to lower output and employment.

1. Change in tariffs between 1990 and 1995. It should be noted that clothing and textile tariffs were increased in many cases in 1992.

At the sectoral level the evidence is mixed. Among the sectors that faced the largest declines in nominal protection, some increased both output and employment, while others showed the opposite trend. However, these sectors are also not the major employers in manufacturing. The five sectors that faced the largest tariff reductions in 1995 account for only 17 percent of total manufacturing employment.

An analysis of trends among major employers - food, clothing, motor vehicles, electrical machinery and metal products - showed that restructuring need not be contractionary. Despite tariff cuts, a range of firms managed to expand output and employment.

The recovery has undoubtedly eased the potentially negative effects of liberalisation. Between 1994 and 1995 manufacturing output grew by 7.4 percent in real terms, while exports increased rapidly in almost all sectors despite a real appreciation of the currency. These trends have been accompanied by substantial private investment in manufacturing.

Firm-level responses to the pressures entailed by liberalisation are diverse. There is evidence of production processes being reorganised through the introduction of additional shifts, the consolidation of production lines and the alteration of product mixes. Increasing numbers of firms are seeking foreign linkages to acquire technology. In labour-intensive industries firms may resort to sub-contracting to increase flexibility and reduce adjustment costs.

Organised labour may experience the adjustment costs differently. Union employment is typically concentrated in larger traditional firms which may encounter higher adjustment costs, while union membership cannot be extended to new market entrants immediately. In fact, it is doubtful whether the data summarised above captures adequately the changes in employment over recent years.

In summary, preliminary indications are that the aggregate costs of trade liberalisation have been kept low, partly as a result of the economic recovery and a resurgence in exports. The recent depreciation of the rand should shift relative prices in favour of exporters. The integrated strategy is designed to maintain these benefits and limit their erosion through inflation.

APPENDIX 11

MEMORANDUM ON TAX INCENTIVES FOR INVESTMENT

It is proposed that an accelerated depreciation tax allowance scheme to enable existing manufacturing entities to expand in response to the challenge of globalisation, and a tax holiday scheme aimed at new projects in key regions and industries, designed to favour labour-absorbing manufacturing activities, should be introduced.

Accelerated depreciation tax allowance

The standard tax allowance for investment in manufacturing plant and equipment at present allows for depreciation at 20 percent per annum over five years. On buildings in the manufacturing sector, the allowance is five percent per year over 20 years.

For a limited fixed period, depreciation allowances in respect of plant and machinery used in a manufacturing process will be allowed at 33.3% per year over three years and in respect of buildings for manufacturing purposes at 10% per annum over a 10-year period.

The foregoing accelerated depreciation provisions will only apply in respect of qualifying plant and equipment which is both acquired and brought into use for the first time during the period 01 July 1996 to 31 September 1999.

Details will be made available by the Commissioner for Inland Revenue.

Tax Holiday

A tax holiday scheme is to be introduced to contribute to the achievement of South Africa's industrial development goals. It is structured as follows.

Projects which meet certain basic conditions will be awarded a tax holiday, the duration of which will depend on three criteria. The basic qualifying conditions, which will included a sufficient level of domestic value added (determined relative to sub-sectoral industrial averages) and evidence of a commitment to key economic goals including human resource development, foreign exchange conservation, and environmental responsibility. The conditions are not intended to be unduly restrictive.

The key criteria relate to labour absorption, regional location, and industrial priority. For each of the criteria satisfied, a project will be awarded a cumulative two-year holiday, up to a maximum of six years.

• Labour absorption will be determined by reference to the proportion of value added devoted to wages and salaries.

• Regional priorities will be determined according to an assessment of regional potential for industrial development and regional need. Qualifying locations, which may be defined as manufacturing development corridors or manufacturing development zones, will be designated by the Minister of Trade and Industry after discussion with provincial government and taking into account the re-evaluation of the Regional Industrial Development Programme that is to be completed in the near future.
• Priority industries will be defined by the Minister of Trade and Industry in the following categories: labour intensive industries, critical industries with strong economic linkages , sensitive industries affected by tariff reform, and future industries defined in terms of their potential to secure a larger share of global consumer expenditure.

Only new entities will qualify. Regulations defining value added requirements, labour coefficients, development corridors and zones, and priority industries will be released in due course. The tax holiday scheme will be introduced in the last quarter of 1996 and the window for entry into the tax holiday scheme will extend for a three year period. The tax holiday will come into affect after the entity becomes liable for tax after taking into account the accelerated depreciation allowances referred to above. However, all incentives granted under the tax holiday programme will expire in the tenth financial year after the project has been approved.

The tax holiday scheme will be managed by a board, which will operate according to clear and transparent regulations issued by the Minister of Trade and Industry.

APPENDIX 12

FOREIGN DIRECT INVESTMENT

Increased openness to international trade and capital flows is one of the hallmarks of the more rapidly growing developing economies worldwide. Foreign direct investment plays an important part in encouraging growth in several ways:

• modern technology is frequently transferred through investment flows;
• skills, management expertise and high level training accompany international investment projects;
• access to international sources of finance is enhanced; and
• access to global markets is facilitated.

Following the debt problems experienced in the early 1980's by many countries which had borrowed excessively and faced severe adjustment problems when interest rates increased and international trade slowed, the advantages of foreign direct investment are now widely appreciated. While short term capital and portfolio flows can meet balance of payments capital account financing requirements under certain circumstances, the risks associated with depending on internationally mobile loan financing are considerable. On the one hand, changed expectations which lead to capital outflows can precipitate serious balance of payments difficulties. On the other, excessive capital inflows associated, for example, with high interest rates and a strong growth performance, can put undue upward pressure on the currency and interfere with domestic financial management. International experience favours foreign direct investment as a more stable source of international finance and as a crucial element in a more outward-oriented growth strategy.

Accelerated investment in the South African economy requires international capital inflows to complement domestic savings and finance the increased imports of capital and intermediate goods which accompany faster growth. The growth model described in appendix four suggests that the current account deficit in period of high growth might reach some 3% of GDP, or perhaps US$ 5 - 6 billion. The strategy implies that perhaps 15% of the foreign savings required would be in the form of direct investment mainly in export-oriented manufacturing.

Since the 1994 elections, South Africa has received net capital inflows of some R30 bn. Most of the investment has taken the form of portfolio flows, which can be quickly reversed as has occurred in recent months. There has been little direct foreign investment. Redressing this imbalance requires attention to the fundamental determinants of international investment decisions and the underlying macroeconomic expectations which may be relevant. These might include:

• political and economic stability, including macroeconomic stability and clarity about economic policy;
• sustained high rates of economic growth;
• labour market stability and flexibility;
• investment incentives;
• the tariff regime;
• protection of property rights; and
• various determinants of expected investment returns.

Macroeconomic and other policies which aim to deal with these factors in the context of an export-oriented strategy can be expected in due course to lead to stronger foreign direct investment flows. International evidence suggests that growth itself is a powerful stimulus of foreign investment. The sustainability of the growth path can be weakened by excessive reliance on foreign capital, however.

Higher flows of foreign direct investment are more likely in the presence of a strong domestic savings performance, underpinning prospects for longer term industrial expansion.

Several developments are needed for South Africa to attract a more substantial volume of foreign direct investment:

• an overall macroeconomic environment conducive to growth and an expanding domestic and regional market;
• gradual removal of exchange controls;
• improved domestic savings and stead reductions in the fiscal deficit;
• a tax regime favourable to foreign investment;
• restructuring of state assets so as to create opportunities for equity investment in public corporations by foreign partners;
• improved labour market flexibility and increased training for a more skilled work force; and reduced crime and improved social stability.

APPENDIX 13

AN ANALYSIS OF LABOUR MARKET TRENDS

Recent trends in the labour market are discussed below. The analysis is based on the standardised employment series (SES), which remains the most satisfactory index of long term trends, although it is no longer fully compatible with available survey evidence.

These estimates indicate that commercial agricultural employment has been in decline since the 1970s. Employment in mining peaked in the mid-1980s at just over 800 000 and has since fallen to little more than 600 000. The contraction in these sectors, and the apparent decline in domestic service, are structural trends which will not be reversed by growth alone.

Manufacturing employment has remained at about 1.5 million since 1980, while employment in construction, electricity and transport has fallen significantly since their 1984 peak. Employment in trade reached just over 1 million in 1990 and has since fallen by 9 percent. The only sectors to record growth through the 1980s and early 1990s have been finance (up by 60 percent since 1980), private services (up 21 percent) and government (up 68 percent). Overall employment growth has fallen from 2.3 percent per year between 1950 and 1980 to just 0.7 percent a year between 1980 and 1990 and -1.1 percent annually since 1990.

Informal and unregulated employment has increased over the past two decades, accounting for perhaps 2 million people. Open unemployment has nonetheless increased from an estimated 19.6 percent of the labour force in 1980 to nearly one-third in the mid-1990s. Survey evidence indicates, furthermore, that another 2-4 million people, mainly in rural areas, are "discouraged" work-seekers who would be in the labour market were there realistic prospects of employment.

The main source of the decline in employment growth has been the secular deterioration in output growth of the economy. The elasticity of formal non-agricultural employment with respect to output growth has been comparatively high: output growth of 6.2 percent in the 1960s was associated with employment growth of 3.7 percent, while output grew at 3.5 percent in the 1970s and employment at 2.8 percent. The unemployment trend has been exacerbated, however, by a more recent decline in labour-output ratios. In agriculture, mining, and domestic service the downward trend in employment seems firmly established. During the course of the 1980s, the responsiveness of employment in several other sectors of the economy to output growth appears to have weakened. Sectoral trends between 1985 and 1994 are illustrated below.

In the absence of reliable evidence on the growth of informal and semi-formal employment, the total employment figures represented here take the 1994 estimates of the official October household survey as a point of departure and the difference between the 1994 SES figure and the OHS formal employment aggregate as a measure of unregulated employment. The emergence of this category of employment outside of the SES estimates is taken to be almhe post-1980 period

The links between real wage increases, output growth and employment over the past decade are complex. A simple logarithmic regression of non-agricultural formal private sector employment on the corresponding output and real wage series yields an output elasticity of 0.7 and a wage elasticity of -1.2, broadly consistent with international comparable evidence. Account should also be taken, however, of the effect of the price of capital and technology choice and employment.

Several general structural features of the domestic labour market can be noted:

• Alongside the regulated market, a complex unregulated market exists which has grown in importance. There are in reality no sharp lines between market segments: wage levels and working conditions vary along a continuum. Sub-contracting, substitution of casual for formal employment and considerable flexibility in the structure and organisation of smaller enterprises are important features of recent trends.
• Within the unionised sector, remuneration typically includes a growing non-wage element, which has contributed significantly to the overall rise in the costs of labour. Although the magnitude of wage effects on employment are subject to considerable uncertainty, it is likely that wage-bargaining in unionised sectors has contributed somewhat to the slowdown in employment creation and to a rising gap between remuneration in unionised and unregulated segments of the labour market.
• Apartheid-based migrant labour flows have evolved into more complex rural-urban linkages, both between urban and rural households or household members and involving longer term migration. The pace of urbanisation has been slower than observers were predicting a decade ago, which is consistent with the slow rate of formal job creation.
• Government has contributed disproportionately to employment creation over the past decade and public sector job creation has somewhat exceeded the average growth of the labour force since the early 1980s.
• As has occurred elsewhere, services have become increasingly important sources of employment, and indeed provide the only sector outside of government in which employment growth has continued during the 1990s.

Account must also be taken of global economic forces:

• Relative wages now play a major role in an international division of labour in which low wage products are almost exclusively produced in the developing world;
• International trade between developed and developing countries places downward pressure on unskilled wages and upward pressure on skilled wages;
• Greater labour market flexibility has emerged in many countries, improving prospects of the unemployed finding work but widening the wage distribution.

As South Africa proceeds with trade liberalisation and adapts to international competition, downward pressure will be placed on unskilled wages. If this is not accommodated by the labour market, then unemployment will rise and irregular, insecure forms of employment will increase. Employment of skilled workers and their wages are, by contrast, likely to rise as skills are in relatively short supply.

Institutions of collective bargaining are important vehicles for securing a stable and sufficiently flexible employment trajectory. Progress has been made in this regard, and there has been a marked reduction in the incidence of industrial unrest or of confrontational relations between business and labour. The collective bargaining system will need to strike an appropriate balance between regulating job security, working conditions and promoting training and productivity enhancement strategies, while at the same time ensuring labour market flexibility, especially at the lower end of the wage distribution.

In the context of a broader strategy aimed at accelerating growth, labour market measures which facilitate job creation for unskilled and semi-skilled work-seekers are needed to enhance the labour absorption associated with economic expansion.

APPENDIX 14

THE PROSPECTS FOR GOVERNMENT EMPLOYMENT

Some 1.5 million people are employed by general government, comprising national and provincial departments (1 162 000 employees), statutory bodies (70 000) and local authorities (26 000). Over the past decade, employment by general government departments and agencies has increased by nearly 20 percent, offsetting somewhat the decline in private sector employment over this period. However, the trend in more recent years has been more stable, with increases in education and police compensated for by declines in defence and local services.

The trend in real wages over this period is also of interest. Over the whole public sector, remuneration per worker increased by about 10 percent in real terms between 1986 and 1995, driven partly by improvements in wages at lower skill levels and partly by the sectoral shift in favour of educators and other higher-earning occupational groups. Average remuneration in the private sector increased by roughly the same margin over the period.

The police force currently numbers just under 145 000. Given the severity of the crime problem and the need to expand policing in under-serviced areas, there is a compelling case for some expansion in police services. Correctional services employ nearly 30 000 people at present, and some increase in staffing of prisons may be needed. The National Defence Force employs 120 000 people, and envisages a phased reduction in this number. Overall, total employment in the protection sector should remain fairly stable over the period to the year 2000, with increases in police personnel offset by reductions in defence.

The emerging primary health care strategy of the national health authorities envisages fairly rapid expansion of employment of health personnel in primary care facilities. Curtailment of growth and possible reductions in staff allocations to the hospital sector are implied by the proposed reorientation of health services. Growth in health services employment of 1 to 2 percent annually over the next five years should be possible, and would permit progress with the extension of the primary health network, while retaining appropriate personnel levels in hospitals.

Although there may be some scope for rationalisation of educational administration, total employment in this sector will need to grow in response to anticipated enrolment growth of 3 percent per annum. Increased employment of educators in public schools might feasibly be held to under 2 percent per year, while college and higher education expansion is likely to require a somewhat higher growth rate. Innovative financing arrangements and containment of real salary increases will be necessary if these growth targets are to be achieved.

Other government functions account for some 240,000 public service jobs. Organisational restructuring might permit cutbacks of perhaps 20-30 000 employees. Sustained employment reduction cannot be expected, however, unless accompanied by explicit decisions to terminate programmes, close regional offices, or devolve activities to the private sector. It must also be emphasised that in several crucial areas of general administration, serious personnel shortages have been reported.

Employment by local authorities has fallen over recent years. This trend should now be reversed, as consolidated authorities take responsibility for revised jurisdictions, municipal infrastructural development gets under way and service delivery is improved. The ongoing maintenance and service provision associated with improved services under local jurisdiction should be expected to lead to growth of between 50 000 and 100 000 jobs by the year 2000.

The main area for increasing employment is as a result of the expanded capital spending in the fields of agriculture and land reform, urban infrastructure and housing, social infrastructure construction and road construction and maintenance. It is estimated that capital spending of approximately R1 billion is needed in these areas for every 20,000 jobs, although the employment impact of these programmes will depend crucially on the extent to which labour intensive construction techniques are adopted. In the improved fiscal context of accelerated growth, and provided expenditure on current transfers and the normal government wage bill is contained, it would be possible to increase capital spending by an average of R2.5 billion per annum. A programme aimed at 250 000 jobs by 2000 would then be a realistic, if daunting target.

There are also compelling grounds for expanding the existing provision for special employment initiatives, while paying close attention to long term benefits and effective programme management. Rural land improvement and municipal infrastructure development are possible clusters of such activities. Government cannot undertake to absorb all potential work seekers through such projects, but can nonetheless provide jobs and limiting training opportunities to a significant proportion of the unemployed poor through well designed and carefully managed public works projects.

To conclude, in view of the need to contain public sector remuneration in GDP and the fact that the public sector restructuring process involves a realignment of salaries and wages in several occupational groups, careful management of the size of the public service is crucial. Although the scope for reductions in the size of the public service is uncertain, right-sizing initiatives should have the effect of reversing the growth in administrative personnel. There is more scope for increased employment at local government level associated with social and infrastructural development projects. These initiatives will include activities in which significant numbers of unskilled and semi-skilled workers could be employed, bringing down somewhat the average wage of total public sector employment. Government funded capital projects are an important source of demand for labour in the present context of widespread unemployment, and should be targeted at the end of the market where employment needs are greatest.

APPENDIX 15

THE PROSPECTS FOR EMPLOYMENT CREATION

A balance needs to be struck between encouraging high productivity employment and the expansion of lower wage jobs in order to reduce unemployment. As is illustrated below, sufficient job creation to reverse the deteriorating unemployment trend can be achieved if appropriate labour market reforms and public sector development initiatives are adopted.

A simple model of the determinants of sectoral employment trends is used below to compute future employment on the basis of projections regarding sectoral output growth and real wage movements. There are unavoidable judgement calls to be made in specifying the relevant elasticities to apply in an exercise of this nature, but it has been possible to utilise a range of evidence regarding the main assumptions.

The model takes as point of departure an employment structure based on the 1994 October Household Survey and Standardised Employment Series. The labour force is estimated to grow at 2.5 percent per annum initially, declining marginally thereafter.

The results of two employment projections are reported below, broadly consistent with macroeconomic scenarios described elsewhere. Higher economic growth in the integrated strategy contributes to a stronger employment performance. However, the model also incorporates changed assumptions about wage trends and the responsiveness of employment to output growth, consistent with the industrial and trade policy proposals of the integrated strategy, a more flexible labour market and a stronger emphasis on labour-based development projects in public sector spending.

Unemployment is projected to increase from 33.5 percent in 1995 to 37.8 percent in 2000 in the base run, in which GDP growth averages some 2.8 percent. The annual increase in job opportunities rises from 97,000 in 1996 to 134 000. Unemployment rises to 46 percent by the year 2020.

Unemployment in the integrated strategy reaches 34.3 percent in 1996 and then falls to 32.7 percent in 2000. By the year 2020, unemployment will have fallen to an estimated 8 percent. By 2000, new job creation has reached 409 000 per annum and is particularly strong in the manufacturing, construction, trade and private services, with important contributions from the public service, government administered development projects and the unregulated sectors of the economy. The increase of 833 000 new jobs created relative to the base run can be decomposed as follows:

1. Domestic service, semi-formal contractual employment and informal sector activities.

Close to 30 percent of the increase in employment in the integrated strategy arise from changes in the flexibility of labour markets and employment-enhancing shifts in other policies. Institutional and microeconomic reforms are the key to reversing the structural barriers to labour absorption in the economy.

Also evident in the sectoral trends is the critical role of growth in trade and services in ensuring a reversal of the unemployment trend. The primary sectors contribute little to employment growth. Manufacturing and construction are important growth sectors, but it is in trade and services such as tourism that there are the most favourable responses of employment to output growth. It is clear, furthermore, that labour-based development projects, which would mainly be the responsibility of local authorities, have a significant contribution to make to employment creation.

Base run assumptions

Non-primary formal private sector employment is assumed to respond to the real wage trend by a factor of 0.4 after a lag of one year, implying that a 1 percent rise in the real wage lowers employment by 0.4 percent. A smaller wage effect is assumed for mining and none for agriculture.

The secular declines in agricultural and mining employment are projected to continue at 1 percent and 2 percent per year respectively, which is somewhat slower than in recent years, while the employment elasticities with respect to output in both sectors are taken to be 0.2. The output elasticities in the manufacturing and transport sectors are set at 0.35, in construction and electricity at 0.5 and in trade and private services at 0.75.

Regular government service employment is kept constant in the first year and rises 10 000 per year thereafter. Labour-based projects as part of RDP development strategies add a further 25,000 jobs on average per year.

The model incorporates a positive response in the semi-formal and informal sectors to growth and to increases in formal sector real wages, reflecting the apparent trend towards casualisation in certain sectors. Urban contractual employment increases in response to growth with an elasticity of 0.8, and in response to formal private sector real wages with a positive elasticity of 0.3. The broader informal sector responds to GDP growth with an elasticity of 0.6 and to the private real wage trend with an elasticity of 0.1.

Integrated strategy

The integrated strategy model retains the same wage elasticity assumptions as the base run, but the lower private sector real wage trend in this strategy reduces the negative impact on formal employment. Stronger output growth improves formal, semi-formal and informal employment. In addition, the model allows for substantial increases in the responsiveness of employment to output growth from 1987 onwards due to greater labour market flexibility and to support measures for more labour-intensive sectors.

The secular declines in agriculture and mining are reduced to 0.5 percent and 1 percent respectively, while the output elasticities rise to 0.4. Output elasticities of 0.65 are assumed in manufacturing, electricity, construction and transport sectors and 0.85 in trade and private services.

In these projections, public service employment grows by 10 000 in 1996, rising gradually to 30 000 per year in 2000. Development projects administered by public authorities account for 25,000 additional jobs per year.

APPENDIX 16

DETAILS OF THE ECONOMETRIC MODELS

The quantitative assessment of the various policy proposals presented in the document was predominantly derived by means of simulations with several econometric models of the South African economy. The models concerned are the macro-econometric models of the SA Reserve Bank, the World Bank and the Bureau for Economic Research and the CGE models of the Development Bank of Southern Africa and the World Bank. Various cross-checks were performed on the different models in order to limit the effects of specific model design features or of specific parameter estimates on the simulation results. For purposes of consistency the final results were derived from a single model. The Reserve Bank model was selected for this purpose, but the results are broadly consistent with results obtained using all the other models. A brief description of the different models is presented below.

The SA Reserve Bank Model: The SARB model is based on the standard income-expenditure framework. It incorporates specific supply-side elements. These consist of an input-output conversion matrix, in which final demand components are transformed into value-added inputs per productive sector, and of a Cobb-Douglas production function which generates a measure of potential output and of the output gap. The latter feeds into imports, investment and into prices via a price expectation term in the determination of wages. The model also includes a monetary sector which determines interest rates and the money supply. The real Bank rate is treated as an exogenous policy variable.

The World Bank Econometric Model: The World Bank econometric model follows the income expenditure format. However, it has a number of unusual features. Firstly, by incorporating an explicit production function (a two-level nested CES with three factors of production) the model allows for switching between a demand constraint and a supply constraint. Secondly, it distinguishes between skilled and unskilled labour and provides for supply constraints in the skilled market. Thirdly, the model tracks the impact of changes in the distribution of factor incomes.

The Bureau for Economic Research Model: The BER model is similarly based on the income-expenditure approach and also deals with the impact of supply-side factors. As in the case of the SARB model, the supply-element is derived from a production function-determined potential output gap which affects prices, imports and investment. In addition, the BER model includes a policy reaction function for the Bank rate and an endogenous treatment of the R/$ exchange rate.

The Development Bank of Southern Africa Model: The DBSA model differs from the econometric models described above in that it is essentially a dynamic computable general equilibrium model resting on a social accounting matrix and various real-financial interactions. The model does, however, incorporate conventional macro-economic relationships.

The World Bank CGE Model: The World Bank CGE model is a fairly standard computable general equilibrium model with profit-maximising producers, utility-maximising consumers and flexible prices that clear all markets except the market for unskilled labour. The model is highly disaggregated with 94 productive sectors, 13 labour skill categories and 24 household types.