chaptertwenty

Guy Stubbs

FORECASTING THE DEMAND

FOR SCARCE SKILLS, 2001-2006
Ingrid Woolard, Philip Kneebone and Deborah Lee
In this chapter

Introduction 459

Background 461

Demand Forecasting Methodology 463

Results 468

Skills Shortages 470

Conclusion 473

Access the data for this chapter at http://hrdreview.hsrc.ac.za

458 HRD REVIEW l 2003

 abstract | This chapter describes a study of the demand for
high-level human resources for the period 2001 to 2006, and
forecasts the number of new positions arising and the
replacement demand over the period. The research uses a labour-
demand model to estimate the number of new positions that
will arise as the result of industrial growth. The authors show
that the demand for engineers, natural scientists and computer-
related professionals will be strong (i.e. more than 2 per cent),
reflecting the increasing use of new and more sophisticated
technology in all spheres of life.
At the same time, the continued fiscal restraint by government
will limit increases in professional employment in the public sector,
with educators and lecturers showing no growth, and medical
practitioners and nurses showing only 1 per cent growth each.
Nevertheless, replacement demand should be strong, as the effect
of the HIV/AIDS epidemic escalates and adds to the vacancies
created by retirements, emigration and other causes of mortality.

INTRODUCTION
The South African economy is undergoing gradual structural change. This
change is largely the result of the dramatic political and social changes
that have been occurring since the transition to democratic rule. Some
changes, however, can be attributed to more common factors which
accompany the evolution of a ‘new’ economy. The highly protected and
insulated domestic economy of the 1980s is something of the past, and
businesses are learning to compete globally in order to survive. Export
opportunities have opened up and protective tariffs have been slashed,
making the adoption of new technologies vital to South Africa’s remaining
internationally competitive. These economic changes have had a significant
impact on the profile of the workforce, with an increase in the demand
for highly skilled workers and a decline in the number of unskilled jobs.
There exists a double imbalance in the South African labour market
with an oversupply of some skills and an undersupply of others. Despite
there being 7.8 million unemployed people in South Africa (Stats SA 2002),
the South African Institute of Race Relations (SAIRR) reports that in 2001,
South Africa had a shortage of between 350 000 and 500 000 people in
the managerial and technical sectors (SAIRR 2001).
Van Aardt (2001) asserts that too few members of the South African
workforce have the educational qualifications necessary to sustain high

2003 l HRD REVIEW 459

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

economic growth in the modern sector. According to the Labour Force Survey in
February 2002, there were 928 000 people of working age with degrees, and a further
1 308 000 with post-secondary diplomas in South Africa in that month (Stats SA
2002). This equates to only 8 per cent of the working age population falling into the
high-level human resources category. Significantly, the unemployment rate among
those with degrees was less than 5 per cent, in comparison with a rate of 46 per cent
among those with Grade 11, and 32 per cent among those with Grade 12. Among
diplomates the unemployment rate was 17 per cent, which is low relative to the national
average, but still quite high. Thus, even some people with post-secondary qualifications
have skills which are in oversupply.
As a starting point for our exploration into scarce skills, this chapter applies a
consistent methodology for forecasting the demand for the high-level human resources
discussed in other chapters in this Review, and discusses how the results coincide with
or differ from those obtained by the other authors. The methodology used here builds
on a study of labour market trends and workforce needs conducted by the HSRC in
1999 (Whiteford, Van Zyl, Simkins & Hall 1999). The 1999 model was updated and
modified slightly to produce forecasts for the period 2001 to 2006.
The overall changes in aggregate occupational structure arise through a combination
of shifting patterns in industrial employment structure and the changing occupational
composition of employment within industries. The former can be regarded primarily as
a reflection of the way in which the changing pattern of demand for commodities impinges
on occupational structure. The latter is more a reflection of technological and
organisational changes affecting the manner in which goods and services are provided.
The level of employment in a particular occupation can change for two reasons: either
the industries in which it is concentrated grow or decline, or because of changes in
occupational composition within industries. The former can be termed the ‘industrial
effect’ and the latter, the ‘occupational effect’.
The occupational effect may arise for a number of reasons. In the medium term,
technological change has a major effect on the demand for certain skills. Demand may
also change in response to changes in the relative wages associated with certain trades,
which may in turn be affected by the supply side of the labour market. Finally, the short-
term level of employment in each industry may depend on the cyclical position in which
the industry finds itself.
In this chapter, we use a labour-demand model to estimate the number of new
positions that will arise as a result of industrial growth. (The sectoral growth rates are
exogenously specified.) We consider only the ‘industrial effect’ and not the ‘occupational
effect’, but, given the relatively short period over which we are forecasting (i.e. five
years), this should not bias the results significantly.
Net changes in occupational employment are only one indicator of future demand.
We also need to consider the replacement demand needed to offset outflows due to
retirements, emigration and inter-occupational mobility. Even in those occupations where
employment levels are expected to decline substantially, there is likely to be a need to
train new individuals, simply to maintain the existing stock of skills at the required level.

460 HRD REVIEW l 2003

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

We calculate replacement demand using the particular characteristics (demographic and

skills level) of each occupation.
The forecast period 2001 to 2006 was selected for two reasons. First, it is notoriously
difficult to forecast industrial growth rates beyond the medium term. Second,
technological changes occur so rapidly that it is unrealistic to assume that the occupational
structure of employment within sectors will remain unchanged over the longer term.
The five-year forecast period used is consistent with the period chosen by Canada and
the Netherlands, both recognised leaders in the field of human resources forecasting. It
reflects a general reduction in forecast periods from the long term to the medium term
to enhance the accuracy of projections.

BACKGROUND
Over the years there has been increasing demand for information pertaining to human
resources forecasting. These concerns arise from the observation that there exists
‘structural’ unemployment, meaning that unemployment is not just a temporary
misallocation, but also a longer-term problem. This is one of the main motivations for the
need for information on future trends in various occupations in the labour force.
In 1999, the HSRC undertook a study of South African labour market trends and
workforce needs in respect of formal employment, for the period 1998 to 2003 (Whiteford
et al 1999). The study incorporated eight of the nine economic sectors of the South
African economy (the agricultural sector was excluded). Detailed forecasts of future
demand were made at subsector level of the 68 professional and ten artisan occupational
categories. From the outset, the authors were aware of the limitations of this type of
research, and on the very first page, pointed to the many unresolved debates around
this type of high-level human resources forecasting:

Despite the popular demand for this information, human resources forecasting,
better known as ‘manpower forecasting’, has always been and still is a
controversial activity. Criticism against forecasting relates to certain labour
market analysts’ beliefs about the need for such efforts, the accuracy with which
forecasts can be made, the context within and reasons for which forecasting is
done, and the methodology used in the process (Whiteford et al 1999: 1).

The 1999 study commenced with a survey of employers. Information was gathered on
current as well as expected employment, skills shortages, and possible changes in future
skills needs, by means of structured questionnaires from 273 randomly selected
companies. An integrated demand-forecasting model for the period from 1998 to 2003
was then developed. The model provided highly disaggregated estimates. Demand
forecasts were made for the total number of people employed in each of 81 occupational
categories in 36 sectors.

2003 l HRD REVIEW 461

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

Changes in total employment arise from two major contributors: changes in output
and changes in labour productivity. Estimates of the change in these two factors and the
derived change in total employment in each subsector were based on the expectations
of people active in each sector. In the survey, respondents were asked to estimate the
change in output and the change in employment within their sector. From this, it was
possible to calculate the change in employment relative to the change in output or,
more technically, the elasticity of labour demand relative to sector growth.
The HSRC study (Whiteford et al 1999) concludes that fewer than 50 000 jobs would
be created over the period 1998 to 2003, despite an estimated growth in output of 2.7
per cent. The trade sector was expected to be the largest creator of employment, followed
by the finance and construction sectors. In terms of occupational demand, the highest
growth was expected to be in the information technology (IT) field. Commercial
occupations such as accounting and financial professions were also expected to show
strong growth. Demand for engineers – especially electrical and chemical – was expected
to be robust.
In 2001, the European Union, the Department of Labour and the Department of
Trade and Industry commissioned Carel van Aardt, from the Bureau of Market
Research, to investigate key skills shortages and the fast-tracking of skills development
(Van Aardt 2001). His study used a mixture of methodologies: qualitative (interviews
and workshops), quantitative (questionnaires, demographic analysis) and meta-
analytical (secondary data) methods. The study shows that the pool of people in the
South African labour market appointed in high-level human resources occupations
increased dramatically from 296 000 in 1965 to 1.11 million in 1994. The fastest
growth in demand over this period was for engineers and engineering technologists,
engineering technicians, accountants and auditors, specialist managers, computer
programmers, systems analysts and software engineers.
Van Aardt’s study concludes that the South African higher education sector needs
to produce more IT specialists, electronic engineers and specialist managers. He also
argues that the educational system needs to be more closely aligned with the needs of
employers; in other words, greater emphasis should be placed on teaching the specific
skills that industry requires.

462 HRD REVIEW l 2003

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

DEMAND FORECASTING
METHODOLOGY
For the study on which this chapter is based, we created a spreadsheet model following
the steps outlined below.

Baseline estimates
The first step was to estimate employment by occupation and sector for 2001. To obtain
the occupational structure of employment within each sector, the previous study relied
on the Manpower Survey, which was last available for 1995. Consequently, we elected to
use the data sets from the Labour Force Surveys conducted in February and September
2001 (Stats SA 2001a, 2001b). The two surveys were based on independent cross-sections,
making it possible to pool the two data sets in order to obtain a reasonable sample size.

Sectoral forecasts
The second step was to obtain sectoral growth forecasts for the period 2001 to 2006.
These sectoral growth rates are important exogenous variables as they play a large role
in determining the model. We relied heavily on ABSA’s growth predictions for the forecast
period (ABSA 2001). ABSA forecasts the performance of 37 sectors of the South African
economy. The approach used is ‘top-down’, meaning that the analysis begins at the
macroeconomic level. The various components of demand are forecast using a Keynesian
demand model, with a supply-side element that relies, in turn, on a variety of international
and domestic assumptions. The projections of the components of demand then serve as
inputs into an input/output-based sectoral model.
ABSA’s sectoral forecasts are thus grounded on a set of underlying assumptions about
the macro-environment for the next few years. These can be briefly summarised as follows:
l
The rand is not expected to strengthen in the medium term, implying the continuation
of a highly competitive currency and stronger exports (aided by tariff reductions).
l
Improved government and domestic savings rates are expected.
l
Low inflation (around 6 per cent), coupled with lower and stable interest rates, is
expected.
l
Continued job losses are expected to occur in the formal sector.

The sectors expected to grow the fastest are transport, communications, and finance,
insurance and business services. Transport and communications have been growing
steadily for some time, largely as a result of the rapid expansion in the telecommunications
sector. The demand for cellular technology in South Africa has been phenomenal and
Internet and e-mail usage continues to rise. Finance, insurance and business services
have embraced technological developments, which has allowed for the provision of a
range of new services. This sector contributed 17 per cent to national value added in
2000, but only employed 7 per cent of the workforce. Despite the evidence of capital

2003 l HRD REVIEW 463

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

intensity, this sector is important for our modelling purposes, as much of its workforce
is highly skilled.
Recent growth in the government sector has been constrained by the fiscal austerity
measures implemented by the Finance Minister, Trevor Manuel (ABSA 2001: 2). ABSA
predicts, however, that government expenditure as a percentage of the gross domestic
product will level out, and the share of expenditure on interest payments will start
declining. This will make more resources available for government services.
The model inputs are summarised in Table 1. Where the ABSA forecasts were too
aggregated for our purposes, we looked at the historical trends in the subsector and made
our own estimates of sectoral growth, such that the (weighted) average of the disaggregated
subsector growth rates still summed to the aggregate sector figures provided by ABSA.

TABLE 1 Sectoral growth rates, 2001-2006

Sector Average annual growth rate (percentage)
Coal mining 0.1
Gold and uranium mining 0.7
Other mining 0.4
Food manufacturing 0.3
Beverage and tobacco products -1.8
Textiles -0.3
Clothing -2.6
Leather and footwear -1.7
Wood and wood products 4.4
Furniture -0.1
Paper and paper products 1.1
Printing and publishing 0.3
Chemicals 0.7
Rubber and plastic products 2.4
Non-metallic mineral products -0.6
Basic metals 1.9
Fabricated metal products 0.0
Machinery 0.4
Electrical machinery 2.9
Motor manufacturing 4.2
Other manufacturing 0.6
Electricity, gas and water 3.0
Building 0.6
Civil engineering 0.6
Retail and wholesale trade 1.7
Motor trade 1.5
Accommodation and catering 3.0
Transport 5.2
Communications 7.0
Banking 5.4
Insurance 5.4
Business services 5.4
Government 1.2
Local government 1.2
Community and social services 1.2
Recreational and cultural services 1.2
Source: Authors’ estimates

464 HRD REVIEW l 2003

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

Demand forecasts
The sectoral forecasts were then fed through a matrix of elasticities to obtain the changes
in occupational demand within each sector. (As explained above, these elasticities
represent the change in employment relative to output growth.) The change in demand
for a particular occupation was obtained by multiplying the sectoral forecasts by the
elasticities and then adding up the occupational demand across all sectors. This gave an
estimate of the increase in the number of positions available for each occupation.
The elasticities used in the previous study (HSRC 1999) were estimated through a
complex process heavily reliant on the information provided through interviews with
companies and industry experts. A repetition of this survey process was beyond the
scope and budget of our task. Thus the elasticities obtained in the previous study were
assumed to still hold for all sectors, with the important exception of IT.
In the case of IT, we were of the opinion that domestic and global circumstances
had changed sufficiently to require a re-assessment of the elasticities used in the 1999
study. Interviews were conducted with several of the major IT companies, and this
information was used to recalibrate the elasticities for computer-related professionals.
In general, we found that IT companies were far more cautious in their predictions of
the demand for IT professionals than they had been three years previously. IT companies
indicated that they did not expect their workforce to grow by more than 1 or 2 per cent
per annum over the next five years. In addition, they indicated that very few ‘entry-
level’ positions were being created. If new positions were being created, these were for
experienced staff or specialised programmers.

Replacement demand
In addition to the new demand for labour, we need to consider replacement demand
arising from retirements, net migration, movement into other occupations, and in-
service mortality. Indeed, given the relatively low rate of economic growth in South
Africa, coupled with high mortality, it can be expected that replacement demand will
exceed the creation of new positions.
Estimating replacement demand is not straightforward, especially in a country like
South Africa where the data on the mortality effects of HIV/AIDS remain uncertain and
controversial. To estimate replacement demand, one ideally needs the following
information:
l
data on the age and gender structure of occupational employment;
l
data on the rates of outflow due to:
l
retirement (and other reasons for people leaving the workforce);
l
emigration;
l
inter-occupational mobility; and
l
mortality.

Information on the age and gender structure is required because many of the flows,
especially retirements and mortality, are age- and gender-specific. Age structures vary

2003 l HRD REVIEW 465

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

significantly by occupation – for example, a higher proportion of managers than IT

professionals are likely to be nearing retirement age. Differences in age structure across
occupations will clearly influence exits, with older people retiring, but younger people
changing occupations. Age structure also affects mortality. While older people are more
likely to die of non-HIV/AIDS-related causes, the data suggest that younger people are
more likely to succumb to HIV/AIDS.
Because the data for our model were obtained from a household survey, we had the
distinct advantage of having detailed information about the demographics of each
occupation. This made it possible to estimate specific rates of retirement and mortality
for each occupational class.

Retirements
For the purposes of modelling retirements, we considered the fraction of the occupational
class that was aged 50 to 60 in 2001, and assumed that 10 per cent of this group would
retire each year (Table 2). We chose a fairly wide age category because, in some cases,
the occupational samples were quite small.

TABLE 2 Number of workers retiring, 2001-2006

Occupation1 Number retiring, Percentage retiring, Average percentage
2001-2006 2001-2006 retiring per year
Academics 3 314 8.9 1.7
Doctors 698 2.2 0.4
Nurses 12 025 8.3 1.6
Computer-related professionals 1 833 2.8 0.6
Scientists 228 5.4 1.1
Science technologists 242 5.3 1.0
Educators 7 281 2.2 0.4
Engineers 1 508 5.4 1.1
Engineer technologists 752 2.6 0.5
Managers 14 843 5.5 1.1
Source: Authors’ estimates

According to the study, the occupational classes most likely to be affected by retirements
over the next few years will be nurses and academics, as these have a high proportion of
older members. In comparison, there are relatively few doctors, educators, IT
professionals and engineering technologists of retirement age.

Mortality
It is expected that HIV/AIDS-related deaths will outnumber non-HIV/AIDS-related deaths
by 2003 (Abt Associates Inc. 2001). Among the adult population, AIDS deaths are mainly
concentrated among the 25 to 50 age group (Abt Associates Inc. 2001) which means that
the labour force is disproportionately more affected than the overall population. Among
the highly skilled, the HIV infection rate is forecast to peak at 13.1 per cent in 2005,
compared with 22.8 per cent for the skilled and 32.8 per cent for the semi- and unskilled.

466 HRD REVIEW l 2003

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

Annual HIV/AIDS-related deaths among the highly skilled are forecast to peak at 1.2 per
100 workers in 2008, while deaths among the semi- and unskilled groups are expected to
peak at 3.4 per 100 workers in 2009 (Abt Associates Inc. 2001). Thus, while the less
skilled categories are disproportionately affected by HIV/AIDS mortality, the effects on
the high-skills occupations are indisputably also severe.2 To estimate replacement demand
arising from deaths in each occupation we used the race, age and gender-specific mortality
rates contained in the Actuarial Society of South Africa’s (ASSA’s) ‘AIDS demographic
model 2000’. The ASSA model was developed by the ASSA AIDS Committee, which was
set up in 1987 to assist the actuarial profession (and later the wider public) in estimating
the impact of the HIV/AIDS epidemic in South Africa. This committee has produced various
papers and monographs on the epidemic and has constructed various models (the ASSA500
model released in 1996, the ASSA600 model released in 1998 and most recently the
ASSA2000 suite of models) to predict the impact of the epidemic at both the national and
regional levels (ASSA).
Using the mortality assumptions of the ASSA2000 model, we calculated the death
risk for each member of an occupational class in the survey data, and used this to arrive
at the number of people that will be required to replace those dying over the five-year
period (Table 3).

TABLE 3 Mortality rates by occupation, 2001-2006

Occupation Mortality numbers Mortality percentages Percentage average
2001-2006 2001-2006 annual mortality rate
Academics 2 048 5.5 1.1
Doctors 1 995 6.2 1.2
Nurses 12 040 8.3 1.6
Computer-related professionals 3 205 4.9 1.0
Scientists 136 3.2 0.6
Science technologists 184 4.0 0.8
Educators 36 086 11.0 2.1
Engineers 1 234 4.5 0.9
Engineer technologists 1 798 6.2 1.2
Managers 16 288 6.1 1.2
Source: ASSA2000 (ASSA)

The mortality rates differ considerably, with mortality rates lowest among scientists
and highest among educators. Considerable replacement demand arising from ‘in-service
deaths’ will need to be met in the case of educators (36 000) and nurses (12 000) over
this period.

Migration and mobility

The migration of skilled professionals from the country has been an increasing source
of concern over the past few years. As is pointed out by Bailey in Chapter 10, this is not
a situation unique to South Africa, nor is it a new situation in the history of the world,
but it is increasingly viewed negatively, as can be seen by the use of the word ‘drain’
when reference is made to the phenomenon.

2003 l HRD REVIEW 467

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

Skills migration can be seen as a positive outcome – gaining, rather than losing,
intellectual insight and knowledge if the move is not permanent. Increasingly, the trend
is for these moves to be temporary, with permanent emigration becoming rarer.
However, the migration of skills is not evenly spread around the globe and only a handful
of countries appear to be benefiting from these human-capital flows. As reported by
Bailey in Chapter 10, the biggest benefactors of South African professionals are European
countries (especially the United Kingdom), which receive 45 per cent of all those who
opt for emigration. The next highest is Australasia, receiving 24 per cent of South African
professional emigrants, most of whom go to Australia. The third most popular region
for South Africans to emigrate to, is North America (14 per cent) with a fairly even
spread between Canada and the United States. Between 1994 and 1997 approximately
20 000 professionals emigrated. An increasingly larger proportion of those emigrating
are professionals.
Skills migration is clearly a reality and must be taken into account, as these losses
also impact upon the supply and demand for higher-level human resources. For the
purpose of modelling we use the estimates of emigration provided in Chapter 10, and
assume that 0.2 per cent of the working adult population in South Africa emigrate
annually. This figure is probably too conservative, but the dearth of adequate migration
statistics makes it impossible to fix the number more accurately.

Validation of model results

The final step was to verify the plausibility of the results by means of interviews with
industry experts. It was decided that these interviews needed to be conducted by the
authors of the scarce-skills chapters themselves, as they had a deep and nuanced
understanding of the specific occupational areas.
Where the results of the model were regarded as significantly out of line with the
expert views, we revisited the matrix of elasticities for that occupational group and
considered whether it was reasonable to make small adjustments within the industrial
sectors that seemed to be most in question. We did not, however, arbitrarily change the
elasticities to manipulate the results to what the experts expected.

RESULTS
The model produces forecasts for 81 occupations, but here we present only some of the
aggregated results. In many cases, the original cell sizes in the survey data were too
small to allow us to present the disaggregated occupational results with any statistical
significance. Consequently, we group some of the occupations and only present those
that will be discussed further in the scarce-skills chapters that follow. Tables 4 and 5
present the forecasts for the occupational demand as well as the new and replacement
demand for high-skills occupations.

468 HRD REVIEW l 2003

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

TABLE 4 Additional occupational demand for specific high-skill occupations, 2001-2006

High-skill occupations Number Percentage average annual change Total new positions arising,
in 2001 in number of positions 2001-2006 2001-2006
Academics 37 237 0.5 914
Doctors 34 370 1.2 2 191
Nurses 155 516 1.2 9 934
Computer-related professionals 75 841 2.5 9 900
Scientists 4 647 1.6 388
Science technologists 4 729 0.5 126
Educators 354 469 1.4 26 417
Engineers 29 824 1.4 2 095
Engineering technologists 32 132 2.1 3 132
Managers 280 298 0.8 11 298
Source: Authors’ calculations

TABLE 5 Number of people needed to meet new and replacement demand, 2001-2006
High-skills occupations Number in 2001 Number of workers required
to meet new and replacement demand over five years
Academics 37 237 6 651
Doctors 34 370 5 207
Nurses 155 516 35 461
Computer-related professionals 75 841 15 600
Scientists 4 647 795
Science technologists 4 729 599
Educators 354 469 73 077
Engineers 29 824 5 116
Engineering technologists 32 132 5 973
Managers 280 298 45 130
Source: Authors’ calculations

We expect the demand for engineers to remain strong, as new and more sophisticated
technology is introduced into all spheres of life. Similarly, the demand for natural scientists
(e.g. physicists, chemists and geologists) is expected to increase as more emphasis is
placed on research and development.
In the case of computer-related professions, our outlook remains more positive
than that of most people interviewed from IT companies. We would argue that, while
the IT companies are not currently expanding employment, this does not necessarily
reflect the demand for IT professionals by other (non-IT) companies. For example, the
financial services sector – which relies increasingly on good computer systems – is
expected to grow significantly over the forecast period, which should have positive
spin-offs for IT professionals.
The two largest professional occupations, namely educators and nurses, are heavily
concentrated in the public sector, and growth in demand for these occupations will be
dictated by the government’s fiscal stance. We have assumed that there will be some
relaxation of the tight fiscal restraint shown in recent years. This assumes that savings

2003 l HRD REVIEW 469

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

from debt servicing will be channelled into increased provision of social services, including
health and education. We have assumed that output in these sectors will grow by 1.2 per
cent per annum, which is still well below the overall expected economic growth rate of
2.6 per cent. But even if there is no growth in professional employment in these sectors,
there will still be huge demand for new workers as a result of replacement demand.
More than two-thirds of the demand for educators and nurses comes from replacement
demand arising from mortality, retirements and emigration.
A small increase in the demand for managers is expected as firms continue to re-
orientate their business systems to survive in the global economy. The HSRC study
(Whiteford et al 1999) specifically identified a need for more managers with financial
skills coupled with ‘people skills’.

SKILLS SHORTAGES
It is very difficult to accurately quantify the supply of high-level human resources. The
data are patchy; for example, there is virtually no information on the output of the
private college sector. This sector enrols substantial numbers of students, both at the
further and the higher education levels, for a wide variety of courses ranging from
week-long training to three-year courses (see Akoojee in Chapter 17). Additionally, the
composition of technikon enrolments has been volatile in the past, making it very difficult
to predict the future pattern of enrolment.
In the case of some fields of study (such as nursing, engineering or medicine), it is
fairly likely that a graduate will practise the profession in which he or she is trained. But
there are other professions – such as management – where the link between qualification
and occupation is far more tenuous (see Horwitz and Bowmaker-Falconer in Chapter
26). Similarly, in Chapter 21, Crouch and Perry point to the significant number of
educators that are in non-teaching occupations.
In this section, we nevertheless try to draw together the supply- and demand-side
figures to determine whether there is strong evidence of future skills shortages. We
have not included managers, as there is no accurate way of estimating their supply.

Academics
In the chapter on the academic labour market (Chapter 22), Koen reports that the SAPSE
data show that there are 14 700 permanent academic staff employed at the country’s
universities and technikons. Our estimate from the Labour Force Surveys is slightly
more than double that number, which is not entirely implausible, given the proliferation
of private colleges and the move towards contract appointments within the tertiary
institutions.

470 HRD REVIEW l 2003

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

Even if we assume that our figure is slightly high, it is evident that roughly a thousand
new academics will be needed each year. This is before we factor in the evidence of a net
movement out of academia into the private and government sectors.
Koen argues that there is clearly a domestic supply problem. He cites two pieces
of evidence: the existence of vacancies at many institutions and the increased reliance
on foreign nationals. In particular, he points to a shortage of staff with PhDs in the
tertiary institutions.
It seems likely that the shortage of highly qualified academics will persist over the
next five years.

Doctors and nurses

The model suggests that about 35 000 additional nurses will be required over the next
five years. Currently about 6 500 nurses are trained each year, which will be almost
enough to meet demand. This assumes, however, that emigration will remain at 0.2 per
cent per annum and that nurses will not move into other non-nursing occupations. In
Chapter 23, Hall and Erasmus assert that poor working conditions, low salaries, the
constant outflow of skills and the impact of HIV/AIDS will restrict the supply of nurses,
in which case a skills shortage is likely.
The model suggests that about 1 000 new doctors will be needed each year over the
forecast period. It should be noted, however, that our estimates of mortality among
doctors are much lower than those obtained by Hall and Erasmus (Chapter 23) who
assume that the demographic profile of the medical profession is identical to that of the
overall population.
Hall and Erasmus report that 1 194 doctors qualified in the year 2000, which is
roughly the number that we estimate will be required. However, anecdotal evidence
suggests that a significant proportion of newly qualified doctors are migrating overseas,
albeit only temporarily in some cases. If this migration is not halted, a shortage of
doctors could arise. In addition, if the public sector were to significantly improve the
doctor/patient ratio, this would necessitate the training of many more doctors.

Computer-related professionals
In the model we looked only at computer-related professionals rather than at the
information and communications technology (ICT) sector as a whole. Consequently, it
is not possible to reconcile our demand figures with the supply figures provided by
Moleke, Paterson and Roodt in Chapter 27. Van Aardt (2001) uses a definition closer to
ours, however, and reports that about 2 500 IT specialists graduate each year. This
suggests that demand will continue to exceed supply, although the result needs to be
treated with caution, as this is not a homogeneous labour market. The interviews that
we undertook in the IT sector led us to the same conclusion as that reached by Moleke,
Paterson and Roodt (Chapter 27), namely that there is sufficient entry-level supply, but
a shortage of highly skilled IT professionals.

2003 l HRD REVIEW 471

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

Scientists and science technologists

While the chapter by Walwyn on biotechnology (Chapter 25) looks at the supply of a
particular, highly qualified group of scientists, this group cannot be seen as representative
of scientific fields in general. The term ‘scientists’ includes many different types of
professionals, only some of whom are in scarce supply. Unfortunately, we were unable
to model occupational demand separately for chemists, physicists, botanists, etc. as the
sample sizes for these occupations were too small.
The 1999 HSRC study (Whiteford et al 1999) suggests that demand for natural
sciences technologists and technicians is strong, and that supply is not growing fast
enough since it is constrained by a legacy of inadequate attention to mathematics and
science education. New industrial processes need to be implemented, maintained and
monitored. On the other hand, the 1999 study does not find a skills shortage in the
biological science occupations (e.g. zoologists and botanists) as demand in these fields
is significantly constrained by the level of government spending.

Educators
We estimate that more than 70 000 educators will need to be trained over the next five
years. This number is clearly not overestimated, given that, in Chapter 21, Crouch and
Perry suggest the figure could be as high as 20 000 per year. Crouch and Perry show
that, in the medium term at least, supply will be insufficient to meet the demand.
Immediate and significant interventions will be required to entice young people into the
teaching profession.

Engineers and engineering technologists

Employment in engineering has been declining, but it is not obvious whether there is a
genuine decline in demand or a lack of supply. The evidence points to an oversupply of
certain types of engineers and technologists (especially in mining and metallurgy) and a
shortage of electronic, chemical and industrial engineers. Van Aardt (2001) and Steyn and
Daniels (Chapter 24) emphasise a shortage of engineers from previously disadvantaged
groups in all disciplines. These members of designated groups are in short supply and high
demand. In addition, Van Aardt suggests a shortage of experienced engineers who could be
employed in management positions, and a shortage of engineers with IT-related skills.

472 HRD REVIEW l 2003

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

CONCLUSION
This chapter has set out a forecasting methodology based on exogenous sectoral growth
variables and a matrix of occupational demand elasticities. Occupational and sectoral
data from two independent labour force surveys in 2001 (Stats SA 2001a, 2001b) were
aggregated to create baseline estimates of occupations by sector, which were fed through
the model to derive estimated demand by occupation for the year 2006.
Additional demand for high-level human resources is expected to be low over the
forecast period of the next five years. Only in the case of engineering technologists and
computer-related professionals will the increase in annual demand exceed 2 per cent
per annum. Replacement demand will be strong, however, as the effect of the HIV/AIDS
epidemic escalates and adds to the vacancies created by retirements, emigration and
other mortality.
The importance of the possibility of a high-level skills shortage warrants rigorous
attention in a country that is perceived to have a lower growth rate than other similar
middle-income countries. There is a strong case for the assertion that this shortage of
high-level human resources is a limiting factor in the economic growth and development
of South Africa. A variety of supply-side interventions are required if South Africa is to
unblock the current skills shortages. This would include interventions such as increasing
the quality of mathematics and science education in the schooling system, a determined
effort to stem emigration, and the introduction of policies to entice young people into
the teaching and medical professions.

2003 l HRD REVIEW 473

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

Biographies
Ingrid Woolard [BSc Natal, BAHons South Africa, PhD (Economics) Cape Town] is a Senior Research
Specialist in the Employment and Economic Policy Research Programme of the Human Sciences Research Council. She
also teaches postgraduate econometrics at the University of Port Elizabeth. Ingrid has extensive experience in social
research, especially in the microeconometric analysis of household surveys, and has published widely in the fields of
labour market analysis, income inequality and the measurement of poverty.

Human Sciences Research Council, P O Box 40191, Walmer, Port Elizabeth 6001
Tel: (041) 581-7825 Fax: (041) 581-7826 e-mail: [email protected]

Philip Kneebone [BCom Port Elizabeth, BComHons Port Elizabeth] is a Lecturer in Economics at the
University of Port Elizabeth and is currently completing his master’s degree at the same institution. He has consulted to
various organisations in several areas of social science research, including social security, international trade issues and
the Southern African Development Community.

10 Kuruman Kloof, Summerstrand, Port Elizabeth 6001

Tel: 083 466 7141 e-mail: [email protected]

Deborah Lee [BCom, BComHons Port Elizabeth] is a Lecturer at the University of Port Elizabeth in the
Economics Department and is currently doing her master’s degree at the same institution. She is a Research Assistant
in the Employment and Economic Policy Research Programme of the Human Sciences Research Council, based in
Port Elizabeth.

Human Sciences Research Council, P O Box 40191, Walmer, Port Elizabeth 6001
Tel: (041) 504-2209 Fax: (041) 581-7826 e-mail: [email protected]

474 HRD REVIEW l 2003

 SCARCE SKILLS | CHAPTER TWENTY | FORECASTING THE DEMAND FOR SCARCE SKILLS, 2001-2006

Notes
1
The distinction between a professional and a technologist in an occupation, as outlined here, is based on
Stats SA’s application of the International Standard Classification of Occupations.
Computer-related professionals are defined as computer programmers, computer systems designers
and analysts, computer assistants, computer equipment operators, industrial robot controllers and
other computer science occupations.
Science technologists refers to physical science technologists, for example, chemistry and physics
researchers, or individuals who work in such fields, and technicians in the life sciences.
Engineer technologists refers to all technologists in all spheres of engineering, i.e. chemical, electrical,
mechanical, mining and civil fields.
2
For more information on HIV/AIDS and its prevalence among specific occupations, refer to Chapter 8.

References
Abt Associates Inc. (2001) Impending Catastrophe Revisited: An Update on the HIV/AIDS Epidemic in South
Africa, Parklands: Henry J Kaiser Family Foundation
ABSA (2001) South African Sectoral Outlook 2001-2006, Johannesburg: ABSA Group Economic Research
Akoojee, S (2003) Private further education and training, In Human Resources Development Review 2003:
Education, Employment and Skills in South Africa, Human Sciences Research Council, Cape Town: HSRC Press
ASSA (Actuarial Society of South Africa) <www.assa.org.za> Accessed 17 May 2002
Bailey, T (2003) Skills migration, In Human Resources Development Review 2003: Education, Employment
and Skills in South Africa, Human Sciences Research Council, Cape Town: HSRC Press
Crouch, L & Perry, H (2003) Educators, In Human Resources Development Review 2003: Education,
Employment and Skills in South Africa, Human Sciences Research Council, Cape Town: HSRC Press
Hall, E & Erasmus, J (2003) Medical practitioners and nurses, In Human Resources Development Review
2003: Education, Employment and Skills in South Africa, Human Sciences Research Council, Cape Town:
HSRC Press
Koen, C (2003) Academics, In Human Resources Development Review 2003: Education, Employment and
Skills in South Africa, Human Sciences Research Council, Cape Town: HSRC Press
Moleke, P, Paterson, A & Roodt, J (2003) ICT and associated professionals, In Human Resources Develop-
ment Review 2003: Education, Employment and Skills in South Africa, Human Sciences Research Council,
Cape Town: HSRC Press
MRC (Medical Research Council) (2001) The Impact of HIV/AIDS on Adult Mortality in South Africa, Cape
Town
SAIRR (South African Institute of Race Relations) (2001) South Africa Survey 2000/2001, Johannesburg
Stats SA (Statistics South Africa) (2001a) Labour Force Survey, February 2001, Pretoria
Stats SA (2001b) Labour Force Survey, September 2001, Pretoria
Stats SA (2002) Labour Force Survey, February 2002, Pretoria
Steyn, G & Daniels, R (2003) Engineers and technicians, In Human Resources Development Review 2003:
Education, Employment and Skills in South Africa, Human Sciences Research Council, Cape Town: HSRC
Press
Van Aardt, C (2001) Key Skills Shortages and the Fast-Tracking of Skills Development, Pretoria: Bureau of
Market Research, UNISA
Walwyn, D R (2003) Biotechnologists, In Human Resources Development Review 2003: Education, Employ-
ment and Skills, Human Sciences Research Council, Cape Town: HSRC Press
Whiteford, A, Van Zyl, E, Simkins, C & Hall, E (1999) SA Labour Market Trends and Future Workforce Needs
1998-2003, Pretoria: Human Sciences Research Council

2003 l HRD REVIEW 475