Profiling emerging contractors for effective transformation in the South African construction sector

Abstract

South Africa's construction industry is transforming. Its economy requires a wider base of contracting entities. Much of this growth is expected to be delivered by ‘emerging’ contractors. Yet these companies face significant obstacles. Targeted interventions that foster companies need a better picture of these firms and their core members. A survey of the civil engineering sector, targeting emerging contractors, showed that the owners of these companies lack training and experience, and that there are gender-related differences. Classifications according to company turnover, contract sizes or employee numbers alone do not reflect circumstances. Combining these data with the core members' backgrounds gives a clearer picture. Most companies hover between the ‘Existence’ and ‘Survival’ stages of business development. Development initiatives that assume companies require knowledge must accommodate large cognitive distances between sender and receiver. Initiatives thus need to focus on the owners, considering the status and background of their companies.

Keywords:

• construction sector
• industry development
• small
• micro and medium enterprises

1. Introduction

This study examined the capacity and capability building needs of small, micro and medium enterprise (SMME) contractors in the civil engineering sector. Our aim was to provide an overview and analysis of their make-up to inform future interventions fostering SMME contractors, which are required not only for an effective transformation of the industry but also for the delivery of much needed infrastructure.

The South African economy is currently experiencing a period of growth and transformation. From 2003 to 2008 the growth rate averaged 4.7% per annum. While South Africa's economy is not spared from the economic slowdown following the global financial crises of 2008 (SARB, 2009) and has recently experienced the first three consecutive quarters of negative GDP growth since 1992 (Stats SA, 2009), the construction industry itself remains ‘active and growing’ (CIDB, 2009a; Stats SA, 2009), although business confidence has decreased and demand for work has slowed down (CIDB, 2009b). In particular, government spending on infrastructure projects, and the work involved in the FIFA Soccer World Cup 2010, had their positive impact mostly on the civil engineering sector (Gillingham, 2009). The demand for new infrastructure remains, and construction industry development and capacity building stay on the agenda of government bodies such as the Construction Industry Development Board (CIDB, 2009a). The economic empowerment of historically disadvantaged individuals (HDIs) and the transformation of South African industry are part of the newly elected national government's plans (RSA, 2009), and the construction industry is at the centre of its programme for job creation and skills development (Magwenya, 2009). Construction in particular is believed to be one of the sectors suitable for skills and sustainable enterprise development (CIDB, 2009a).

However, official statistics on liquidations and insolvencies (Stats SA, 2008a) for 2000 to 2007 report a total of 2650 liquidated or insolvent entities in the construction sector – with the numbers expected to rise for 2008 and 2009 because of the economic slowdown (Gillingham, 2009). But it must be noted that the cited statistics only report on formally liquidated or insolvent companies – not on those that are no longer active though not formally liquidated or insolvent. The sustainable development of the many smaller contracting companies is thus questionable and reports indicate a rise in industry leavers (Gillingham, 2009). This echoes early doubts that development programmes aiming to empower emerging companies have managed to develop companies with an equity ownership of HDIs ‘into serious industry players’ (Kganyago, 2004). Recent data released by the CIDB suggest the same: the numbers of new entrants to the industry still exceed the numbers of contractors who have a track record of capability (CIDB, 2009b).

The preferred economic advancement of one racial group over time meant economic underdevelopment for all others. The low skills levels of HDIs (CIDB, 2007a) and the emigration of skilled workers (HSRC, 2004; Momberg, 2008) have resulted in a shortage of all technical and managerial skills in the South African construction sector. From as early as 2005 this sector, which has major infrastructure and housing projects under way, has been under pressure because of this lack of skills (Lawless, 2005), and there is evidence that the pressure continues despite the economic slowdown (CIDB & BER, 2009).

A wider base of contracting entities, able to deliver good quality built products, must be developed for the South African construction industry to relieve this pressure and provide the civil infrastructure needed to meet the country's remaining development challenges, which are massive. South Africa's informal sector – often referred to as the ‘second economy’ – is predominantly populated by the historically marginalised and is made up of informal enterprises that fail to comply with regulations (Wells, 2007), for example tax registration.

Companies owned by HDIs are in limbo between the informal and formal business worlds. They are beginning to move towards formality, yet are still partly based in the informal sector in terms of their type of customer and their informal labour relations (Wells, 2007). However, indications are that informal sector companies – commonly referred to as ‘emerging companies’ – are now actively seeking opportunities in the formal sector, encouraged by South African Government legislation such as the Broad Based Black Economic Empowerment (BBBEE) Act (RSA, 2003b) and the Preferential Procurement Framework Act (RSA, 2000). The aim of the BBBEE Act is to facilitate ‘a meaningful participation of black people in the economy’ (RSA, 2003b) by lowering barriers to entry for HDIs into the formal, historically white-dominated economy. For contracting companies, the promulgation of the Preferential Procurement Policy Framework Act of 2000 sets the rules for public procurement processes and has introduced, over and above price considerations, ‘specific goals’ that must be considered when awarding contracts. These goals include contracting with HDIs, such as blacks, women, or disabled persons. A tender that complies with these goals is awarded additional points in the evaluation process.

To tender for public projects, contracting companies must register with the Construction Industry Development Board (CIDB), which assists with the transformation of the South African construction industry in general, and fosters small businesses in particular. The CIDB register is divided into work categories (civil engineering, general building, electrical, and so on), and it groups contractors in these categories into nine grades according to their capabilities in terms of project size in South African rand. Public tenders stipulate the minimum grading required for tendering for contracts. To promote potentially emerging (PE) – read HDI-owned – contractors, the CIDB permits these contractors to tender for construction contracts one grade higher than their current grading (CIDB, 2009a). At the end of 2007 more than 55 000 contractors were registered with the CIDB, the vast majority (more than 50 000) being in the lowest four grades, suggesting that these businesses are micro to small enterprises as defined by the National Small Business Amendment Act, No. 26 of 2003 (RSA, 2003a) – an Act which aims to support small business in South Africa. Most of these 50 000-plus are in the lowest grade (grade 1), which gives the impression that many are opportunists rather than potentially sustainable businesses (CIDB, 2007b). Despite their large numbers, little is known about their make-up, business networks, level of technical expertise and experience, so it is difficult to target development initiatives towards them.

Thus our objective in conducting a broad-based survey of the registered contact persons of CIDB-registered HDI-owned civil engineering contractors was to capture data effectively, so we could explore key aspects of these companies, such as their interaction in the wider marketplace, their overall development status, and the characteristics of their core role players. We could then use this information to build up a picture of emerging companies in this sector, showing their internal composition and their surroundings. These company profiles could be used to design and implement tailor-made interventions.

The data this paper is based on are drawn from our project for assisting transformation in general and in particular supporting capacity building in the civil engineering sector – the sector of the construction industry responsible for providing the needed infrastructure. Our study explored how capacity might be built if skills and knowledge were transferred in situations where established and emerging contractors engage with each other, enabling the latter to become sustainable competitive entities that could tender for and work on bigger projects. The construction industry would benefit from the additional capacity and anticipated transformation. We present only the first part of the wider project here, describing the profiles of the ‘emerging’ companies and the background of their core members.

2. Motivation for study and profile characteristics used

Since the end of apartheid in 1994, government, NGOs and academics have undertaken a number of interventions to ‘uplift’ HDIs in the construction sector (Hauptfleisch et al., 2005; SAFCEC, 2007). Some quantitative (e.g. Cattell, 1994) as well as qualitative (e.g. Sebake & Sithole, 2005) studies of small contractors have been carried out. However, in a rapidly changing social and economic environment, profiling of the HDI-owned SMMEs is required in order to target development interventions correctly. Such interventions essentially provide the organisational learning needed for a business to improve and develop and keep up with ongoing changes in the environment (Senge, 1990) so as to survive in a competitive marketplace. This is particularly true for companies entering new markets, such as emerging contractors entering the formal sector with its different rules. The extent to which the developing companies are geared for change will determine their future path during economic downturns when the competition is tough.

If these SMME contractors are to achieve sustainable success and be competitive, they need to adapt to their environment, learn about their new domain and its underlying rules, and introduce systems that cater for this. The way they acquire or create this new essential knowledge is best described in the light of organisational learning theories. Such theories may focus on a conscious quality improvement objective (Argyris & Schön, 1978; Levitt & March, 1988), or may have a broader focus, as for example in the work of Nonaka & Konno (1998), who use the Japanese philosophical concepts of ‘Ba’ for the place where learning happens and ‘Basho’ for the bigger (market) place where people or entities interact.

The literature explains that certain characteristics of the individual, interacting within the Basho receiving knowledge, are important for any knowledge transfer. For example, Szulanski (2000) points out that the ‘existing stock’ of knowledge will determine the stickiness of knowledge as one characteristic. This follows closely what other authors describe as the antecedents of the individual (Simonin, 1999). The gap in knowledge, a further characteristic, defines the absorptive capacity (Cohen & Levinthal, 1990) and will determine the further path of knowledge transfers. Existing stock, antecedents and knowledge are difficult to assess; however, training received and relevant experience could stand in here as synonyms. From the SMME contractor's perspective, the main individuals interacting with other entities within the Basho are likely to be the owners or managers, who can be considered the core members of the SMMEs. It is these individuals who ‘play a critical role in assuring appropriate “redundancy” of information’ (Nonaka, 1991), which is crucial for any organisational learning.

Katz & Gartner (1988), drawing on the work of McKelvey (1980), address the difficulty of researching ‘organisations-in-creation’. They list four properties for identifying an emerging organisation: Intentionality – a new organisation is created with a goal in mind; Resources – a new organisation needs to possess some resources (financial, human, etc.); Boundary – a new organisation is defined through its boundaries with its environment, and the boundary is displayed through tax registration, phonebook entry, and so on; and Exchange – a new organisation needs to engage in transaction with its marketplace. They say a company has reached the stage of being a complete organisation, as opposed to a ‘pre-organisation’, if all four criteria are satisfied. Katz and Gartner's criteria make it possible to judge organisations at a very early stage. Churchill & Lewis (1983) introduce a more qualitative maturity model for small businesses, consisting of five stages a small business goes through in its development: Existence, Survival, Success, Take-Off and Resource maturity. Later work by Sullivan (2000) suggests that these ‘stages’ should in fact be called ‘phases’, since the boundaries between them are diffuse and development does not occur in a sequenced manner. Nevertheless, Churchill and Lewis's five-stage model is a good description of the process of developing a business and its implications for management. Using this model, we can place emerging contractors in the initial ‘Existence’ stage, where they struggle to find customers and deliver products – echoing Katz and Gartner's Resources and Exchange stages. At the next stage, ‘Survival’, the business shows signs of becoming an established company, though still simple in structure. At this stage it has enough customers and its products are found to be satisfactory (Churchill & Lewis, 1983), or, as Katz and Gartner say, the organisational properties of Resources and Exchange are established.

It is these situations and characteristics of the core members and their organisations that we tried to capture in our survey. Knowing these characteristics of emerging contractors helps us understand the current contexts in which they work. These contexts form the Ba and the Basho of any learning by the emerging contractors. To understand the way knowledge is being transferred from established to emerging contractors, and relate this to the Ba and Basho, we needed a thorough description of the make-up of these contractors. Using this description, we could describe these current mechanisms for transferring knowledge and make recommendations for improving the Bas and Bashos.

3. Survey method

As mentioned above, South Africa lacks a comprehensive overview of emerging contractors. However, to understand better how these companies learn, considering the Basho and learning contexts, we need first to map out the wider marketplace and the unknown emerging contractor itself. Since the wider research project takes a constructivist approach, aiming to describe how emerging contractors learn, we undertook a descriptive survey so as to give a better picture of the phenomenon ‘emerging contractor’. We did not aim to test cause-and-effect relationships or to define a typical emerging contractor but rather to create an informed base for discussion and understanding of emerging contractors, to aid in subsequent targeted interventions. The data were obtained via a broad-based computer-assisted telephonic survey. The questionnaire was developed in three phases. The first draft was based on a literature review. Then the questions were tested for flow and to eliminate ambiguity, using a convenience sample (involving an academic colleague and a friendly contractor). The questionnaire was set up for the researchers to read on a PC screen when they conducted the telephone interviews. The answers were recorded via a data-capturing interface to an MS Access database. The questionnaire was then pilot-tested on 10 contractors. Final improvements were then made to the questions and the software, resulting in the final questionnaire and survey tool.

The questionnaire had four sections: introduction and consent-to-participate, to ensure participants understood the purpose of the survey; particulars of the participant, to profile the respondents as key individuals in these small businesses; development status of the company; and information about cooperation with other entities.

Our proportional stratified sample comprised CIDB registered Civil Engineering contractors in grades 1 to 4 (work up to ZAR 3m). The CIDB denied us access to their full database, including demographic data, so we obtained the contractors' data through the publicly available CIDB internet portal (www.cidb.org.za). The data on registered contractors here include each company's work category and grade, and the name and details of the contact person. A ‘PE’ suffix to a contractor's grade signified ‘potentially emerging’. Our survey sample included both emerging and established companies. However, the number of established, i.e. white male owned, companies registered in these lower grades was expected to be low. At the time of the survey, September 2006, a total of 2638 contractors were registered in the lower four grades under the Civil Engineering work category.

To ensure that the survey was representative, we selected a sample size of 20%, which falls well within recommendations by other authors (e.g. Gay & Airasian, 2003). We thus randomly selected 528 contractors from across the whole of South Africa and attempted to contact them. A maximum of three attempts was made, on consecutive days at different times of the day. Approximately 25% of the contractors could not be reached using the registered telephone numbers and a further 22% were unwilling to participate in the survey. Five respondents were not HDI-owned companies, and were not described as ‘potentially emerging’ (PE) in the CIDB database either; these companies were not part of the target population, and were excluded from the analysis. A few unanswered questions were permitted, mostly because the participants were unable to answer them. The final number of valid completed questionnaires was 264, meaning that the response rate after adjustments was 50%.

The data collected were largely quantitative, supplemented with qualitative data stemming from some open-ended questions. We later grouped and categorised some of the qualitative data in order to analyse it, together with the quantitative data, using SPSS Version 15.0.

4. Analysis and results

As the questionnaire allowed for optional additional responses, the total number of variables collected exceeded 80 in some instances. This paper focuses specifically on data relating to the owners or managers and the development status of the companies.

4.1 The respondents

A Pearson chi-square test revealed a significant p-value of 0.000 when comparing the South African demographics (Stats SA, 2008b) with the survey sample. This indicates a representative distribution of all races among emerging contractors in the lower grades of the CIDB register and the construction sector. The race and gender distribution of the respondents is shown in Table 1.

Table 1: Race and gender of respondents

Gender		Race				Total
		Black	Coloured	Indian	White
Male	Count	162	7	7	6	182
	% of total	61.4	2.7	2.7	2.3	68.9
Female	Count	73	1	4	4	82
	% of total	27.7	0.4	1.5	1.5	31.1
Total	Count	235	8	11	10	264
	% of total	89.0	3.0	4.2	3.8	100.0

Of the 264 respondents who answered the phone, 93% were either the sole owner (237) or partners (9) in the business, 5% (14) were either administrative or managerial staff, and in two cases technical staff answered the phone and in one case a relative of the owner. It can be assumed that the respondents who were owners or managerial staff – the large majority of the sample – knew a lot about their company's operations and could thus be considered reliable sources of information.

The respondents were aged from 21 to 65, with approximately 60% being between 31 and 45. From the high number of female owners or partners observed (31%), it appears that the South African construction industry has more female contractors than other countries, for example female ownership of construction businesses in the UK is at 3% (CITB, 2005). This high level of female involvement may be a product of the ‘specific goals’ set by the Preferential Procurement Policy Framework Act (RSA, 2000), whereby women are considered to be HDIs and are thus preferred contracting entities who, because of their gender, can gain the maximum points in the preferential tendering process (DTPW, 2002).

The respondents had on average seven years of personal experience in the construction industry. More than a quarter (26%) had less than two years, and 30% had between two and five years. This somewhat contradicts the belief that Preferential Procurement encourages experienced and skilled black HDIs to set up their own enterprises (CIDB & CETA, 2005) and suggests that they may prefer to remain in established companies. Considering the learning processes of the various trades (e.g. Gamble, 2001) and the length of training programmes for artisans, the length of experience in the industry revealed by the survey appears inadequate. The surveyed companies' level of experience can be considered low, making these companies vulnerable to failure in a competitive marketplace. Further, a major gender-related difference in experience was observed (Table 2).

Table 2: Years of experience in construction

Gender	Years of experience in construction
	Mean	Std. deviation
Male	8.6	7.1
Female	3.5	3.0
Total	7.0	6.6

The participants' training level was also surveyed. There were six categories (Table 3). It must be noted that government programmes often offer short courses, such as the Contractor Learnership Programme of the Expanded Public Works Programme, so the distinction between government programmes and short courses as shown in the table may be deceptive, since the types of training may overlap. On-the-job training and cases where respondents felt they had received training were grouped together because of difficulties in clearly discriminating between the two. A large number of respondents had tertiary education, the highest qualification being a bachelor degree and the National Diploma the most common. The trend in reported training levels was similar to that of experience in the industry. A total of 55% had not received any formal training for work in the construction industry. Gender differences were also noted in the amount of training received: only 34.6% of the women had received relevant training, compared with 48.6% of the men.

Table 3: Training received

University programmes	Diploma level	Certificate courses	Workshops and short courses	Government programmes	On the job/other
12%	9%	14%	20%	19%	26%

When asked about their personal background, 37% said they had worked in administrative positions or had been shop managers. Technical backgrounds were less common (23%), and 18% of the respondents were school-leavers or students. A small percentage were teachers-turned-contractors (8%). The remainder was made up of unclear responses and those who said they had been unemployed. Overall, this suggests that many of them were not entrepreneurs in the strict sense of risking their own capital and relying on their own (managerial) competence and (creative) opportunism (Long, 1983). This is also reflected in the CIDB registration, as the grading of contractors is dependent on their ‘financial capability’ (RSA, 2004). Low levels of capital are a concern, as under-financing has long been recognised as the biggest problem faced by small companies elsewhere (Giritli et al., 1990).

On the positive side, the data on ownership and years of experience in the construction industry are noteworthy. The Preferential Procurement Policy Framework Act was promulgated in 2000, and its subsequent Regulations, clarifying and implementing some conditions, were published in late 2001. At the time of the survey, the rule about achieving ‘specific goals’ was five years old. One could thus hypothesise, given the above findings, that once the Act became known to the wider population it did succeed in attracting emerging contractors, especially women, into the construction industry. Noteworthy also is the fact that no relationship was found between the age of the respondents and the level of education.

However other findings from the survey paint a bleak picture for the South African construction industry, as the number of inexperienced persons with very limited training and knowledge who own and manage contracting companies is high. The South African construction industry is facing a precarious situation in respect of the sustainability of these entities. Poor management skills among contractors (Van Wyk, 2003; Sebake & Sithole, 2005), combined with low levels of training, may thus be one contributing factor to the often reported non-performance of SMME contracting entities and the ‘depletion of industry capacity’ (CIDB & CETA, 2005). In the light of clear gender differences in training and experiences, it would be interesting to assess the difference in success and sustainability between female-owned and male-owned companies.

As for development initiatives through knowledge transfer, the ‘existing stock of knowledge’ (Szulanski, 2000) is expected to determine the stickiness of new knowledge. It further determines the receiver's end of what Nooteboom (1999, 2000) calls the ‘cognitive distance’ between sender and receiver of knowledge in situations of knowledge transfer. It is thus crucial to know the existing levels of experience, skills, and knowledge when tailor-making mechanisms for knowledge transfer. In our survey data the levels of training and relevant experience are rather low. This increases the cognitive distance between knowledgeable senders, who are often project managers, engineers or staff from established companies, and the learning receiver, and this makes transfer difficult, as mutual understanding is difficult to achieve (Nooteboom, 1999).

4.2 The companies

The CIDB allows contractors to register in more than one category of work and 58% of the respondents had done so – jacks-of-all-trades-and-masters-of-none? This is probably because to submit a tender for a public contract the contractor must be registered in the relevant category. The multiple registration might reflect an attempt by these small businesses to diversify as a means to growth (Ansoff, 1957). The highest grade for an additional category registered for was grade 5. The validity of the sample selection was shown to be in order as the sample was homogenetic: 99% were not registered in any other category higher than grade 4.

The response rate differed slightly between the various grades (Table 4). Businesses registered in grade 4 in particular declined to respond to the questionnaire. This may be because the owners had heavy workloads. While companies registered in grade 1 might be opportunists with little in the way of steady workloads, those in grade 4 (capable of undertaking projects valued up to ZAR 3m) would be expected to be at a higher maturity level, have regular customers, and would have passed the ‘Survival’ stage in company development (Churchill & Lewis, 1983).

Table 4: Response rates across grades

Grade	Sample size and percentage	Responses	Proportion of total responses	Response rate within grades sub-sample
1	370 (70%)	194	73.5%	52.4%
2	79 (15%)	37	14.0%	46.8%
3	29 (5.5%)	14	5.3%	48.3%
4	51 (9.5%)	19	7.2%	37.2%
Total	529 (100%)	264	100%	49.9%

The size of the companies was evaluated in terms of employee numbers and turnover (Table 5). Three companies mentioned unusually high numbers of full-time employees. However, this may be because they were involved in labour-intensive projects at the time of the survey. Fifty employees is seen as the cut-off point between small- and medium-sized companies in terms of the National Small Business Amendment Act of 2003 (RSA, 2003a). In our sample 99% of the companies had fewer than 50 employees, and the average number of employees was 5.8, with a standard deviation of 7.6 and a median of 4. Twenty-six per cent of the sample had no employees at all, which supports a finding by Winch (1998) in the UK that ‘few of the self-employed employ others’.

Table 5: SMME classification according National Small Business Amendment Act (South Africa, 2003b) and responses

Size	Total annual turnover (less than)	Number of employees (less than)	As per turnover	As per number of employees
Higher*	n/a	n/a	1.1%	–
Medium	ZAR 26.0m	200	4.9%	1.1%
Small	ZAR 6.0m	50	7.2%	4.9%
Very small	ZAR 3.0m	20	36.4%	29.2%
Micro	ZAR 0.2m	5	49.6%	64.8%
*Two entities (0.8%) could not make any statement on the turnover.

The trend for reported financial turnovers was similar to that for number of employees, with 94% of the companies reporting a turnover of less than ZAR 6m – the benchmark set by the above Act in order to be classed as a medium-sized company. Furthermore, 50% reported turnovers below ZAR 200 000. It must be noted that when asked for their turnover many participants were either reluctant to state an amount, or reported they were not in the position to do so since ‘someone else is doing the books’. The company owners' inability to do the books in-house is a serious obstacle to sustainable development for these companies, and further points to their lack of systems.

The CIDB register grades are based on the companies' ability to fulfil the requirements of contracts of which a key criterion is the monetary value, i.e. the contract size. Table 6 shows the sizes of the contracts the participants were working on. The average was ZAR 1.68m, and at the time of the survey the large majority (75%) had thus far only been exposed to contracts under ZAR 1m. The high number of contractors without any previous record of contracts (17%) is remarkable, and is further evidence that many emerging contractors are in fact job seekers, and since there is no evidence of previous marketplace activity they must be considered pre-organisational entities (Katz & Gartner, 1988).

Table 6: Reported contract size

Contract size in ZAR (less than)	Responses	Proportion of total responses	Cumulative percentage
0	45	17.0%	17.0%
100 000	48	18.2%	35.2%
500 000	66	25.0%	60.2%
1 000 000	39	14.8%	75.0%
2 000 000	25	9.5%	84.5%
10 000 000	35	13.3%	97.7%
Higher	6	2.3%	100.0%
Total	264	100.0%

Table 7 shows the age of the companies. Almost all had been founded less than 11 years prior to the survey (97%). Forty-two per cent (112) had been founded less than three years previously, and a further 38% were between three and five years old.

Table 7: Age of company

Age in years	Responses	Proportion of total responses	Cumulative percentage
0 to 2	112	42.4%	42.4%
3 to 5	99	37.5%	79.9%
6 to 8	34	12.9%	92.8%
9 to 11	11	4.2%	97.0%
12 to 14	2	0.8%	97.7%
15 and older	6	2.3%	100.0%
Total	264	100.0%

Taking the quantitative data on employees, turnover, contract size, and age of the surveyed companies as a whole reveals a picture of a huge proportion of contractors situated at the ‘Existence’ stage (Churchill & Lewis, 1983). This is supported by the many statements by participants that they ‘struggle to get contracts’ – i.e. they struggle to find customers, as in Churchill and Lewis's criterion. We have noted above that 18% of the sample were school-leavers and students and thus would naturally be job seekers. This is also reflected in the significant relation (p-value of 0.00) between the core members' years of experience and the age of the company when calculated using ‘bins’ of four years. As Churchill and Lewis observe, ‘The owner is the business’.

While contract size is not by itself representative of a contractor's overall capability, when considered in conjunction with the data on employee numbers and turnover it is clear that the bulk of the sampled contractors are micro or very small businesses. Furthermore, the figures for annual turnover and average contract size do not correspond. This might be due to gross lack of knowledge about the company's financials, or multiple small contracts, or the fact that many small contractors struggle to complete projects in time and often experience major project delays (e.g. Sebake & Sithole, 2005).

Interestingly, no relation could be found between the number of employees and turnover. This could be due to typical labour-only subcontracting, the use of labour intensive contracts (both frequently found in South African public works projects), informal contracts with little or no book-keeping, or subcontracting arrangements. All this could distort the picture of company size and turnover. This makes it questionable whether a categorisation of companies based on these two features, as done by the National Small Business Amendment Act (RSA, 2003a), is an accurate reflection of the nature of these businesses.

An interesting finding was that six companies established prior to the end of apartheid in 1994 still considered themselves as ‘emerging’ and were registered as such. This raises the question of how long a company can continue to be considered ‘emerging’. A closer look revealed that five of these ‘old’ companies were micro to medium in size according to their number of staff and turnovers. The lack of relation between size and turnover can best be seen in one of these cases. This particular company could be labelled very small in terms of staff numbers (12 full-time employees), yet showed a turnover (ZAR 10m) of a medium sized company, possibly due to subcontracting arrangements. The official use of the term ‘emerging contractor’ could be misleading, as a company employing even six full-time employees (a very small entity according to the National Small Business Amendment Act) can have crossed the ‘Survival’ stage of business development, relying on existing knowledge, networks and systems, and not be ‘emerging’ in an economic sense any more. Yet in the official South African context, ‘emerging’ refers to HDI equity ownership rather than economic or organisational maturity; the classification is based on race rather than business maturity – as suggested by the term ‘HDI’ itself.

5. Conclusions and recommendations

Broadening its base of companies is essential for the South African construction industry if it is to assist the broader economy to meet its targets and meet the demand for construction services. To achieve this expansion, small companies must be integrated into the formal sector. However, the high number of failures among these companies overshadows this expansion effort. Targeted tailor-made development initiatives, taking emerging companies' core members and the overall status of the company into account, must be conceptualised using current information on these entities.

The data from the survey we describe in this paper provide an updated picture of emerging SMME civil engineering contractors in South Africa. We successfully contacted a total of 264 formal HDI-owned companies that were registered with the CIDB in the lowest four grades and ready to tender for public contracts. Our findings suggest that micro and very small companies in the South African construction industry are often formed by individuals with little experience in construction and no prior affiliation with the industry, often having the characteristics of job-seekers rather than entrepreneurs. This in itself is valuable information on the background of these companies. The low level of training among individuals, especially training in a relevant technical field, is of concern. Combined with the lack of even informal experience, this aggravates the inherent problem small contractors face in delivering the desired services. The main source of these new small companies appears not to be the informal economy, where relevant experience could have been gained; rather, they are new entities formed by individuals previously unaffiliated to construction. Individuals appear to be attracted by the preferential treatment provided under the Preferential Procurement Policy Framework Act (RSA, 2000). The high number of female-owned companies is particularly noteworthy, reflecting this attraction. Sadly, these attracted female owners are even less prepared for the tasks at hand than their male counterparts.

Thus, although a wide base of companies exists in terms of numbers, the lack of ‘deeper roots’ and a foundation in the industry means that they struggle to develop into more meaningful entities. Together with a lack of customers, the exposure to a high number of competitors registered in the lower CIDB grades is worsening their situation. Our surveyed companies' struggles could, however, in some cases also be attributed to a lack of entrepreneurial characteristics (cf. Long, 1983). This was also noted at a National Workshop held by the CIDB (CIDB & CETA, 2005). Combining the data on employee numbers, financial situation, evidence of lack of systems and formal planning and lack of customers with previous reports (Sebake & Sithole, 2005), we find that the bulk of the companies surveyed appear to be hovering between the ‘Existence’ and ‘Survival’ stages of their business development. Any interventions that aim to foster the development of these mostly micro and very small contractors must consider a low ‘existing stock of knowledge and skills’ (Szulanski, 2000) – essentially low experience and training of the core member – embedded in their organisational structure and reflected in their development status. In particular, interventions that aim to transfer knowledge and skills from established to emerging companies, or from professional teams (e.g. mentors) to emerging companies, will have to take into account the large ‘cognitive distance’ (Nooteboom, 1999) between these parties due to the low level of training on the receiver side. A similar observation has been made by Egbeonu & McCutcheon (2008).

A second set of companies, albeit a minority, appear to be fairly well established (their reported large contract sizes, high turnovers and fair sized workforces are evidence of this) yet are officially treated as emerging. The current South African definition of ‘emerging’ means that these companies fall within this definition and are thus profiting from preferential treatment in the awarding of public contracts. They are thus competing with genuinely emerging companies on equal terms. A revision of the definition of ‘emerging’ contractors is needed, keeping in mind the objectives of Broad-Based Black Economic Empowerment, and acknowledging economic circumstances. This could be combined with a review of the National Small Business Act of South Africa, in which the criteria for categorising construction businesses do not match observed patterns in our surveyed sample. This current mismatch might influence the same Act's intentions to support small businesses. For the purpose of initiatives to develop and enable emerging contractors, the current definition needs to be revised. A definition based on the development stage, such as that proposed by Churchill & Lewis (1983), possibly combined with an assessment of Exchanges and available Resources, as described by Katz & Gartner (1988), should be preferred to a definition based on quantitative characteristics and race alone. This would lead to a more broad-based approach to empowerment, enabling truly emerging companies to gain from empowerment programmes – with a broad-based transformed industry sector as the result.

As for development initiatives, it must be noted that our data on the companies and our data on the core individuals are highly congruent. This suggests that any efforts to uplift these HDI-owned companies will have to focus on the core individuals in the companies, since such an individual is the equivalent of the entire company. It is important to create suitable contexts for knowledge transfer from established companies to emerging contractors, or rather their core individuals, and crucial to consider the make-up of these emerging (i.e. HDI-owned) entities when designing programmes aimed at developing them through knowledge transfers. Only tailor-made environments that take cognisance of individual circumstances will form appropriate Bas and Bashos for the required learning processes. Matching the senders' and receivers' cognitive levels with an appropriate potential for transfers, yet not too low or too high, will be key for knowledge transfers. The cognitive distance needs to be optimised and revised throughout development interactions between emerging and established contractors. This could be done by periodically assessing the former's individual needs and matching them with the latter's possible offerings. For development initiatives to succeed, the core members of the companies must be considered and catered for on an individual basis.

The study described in this paper can inform and initiate future research. The impact of existing skills levels on the sustainability of companies has yet to be quantified. A special emphasis could be placed on the difference in skills levels between male and female contractors. Further, the relation between cognitive distance and what it takes to transfer knowledge should be investigated. It is further of paramount importance to maintain a comprehensive overview of changes within the population. Longitudinal studies seeking equivalent data on SMME contractors, and possibly mapping these against development interventions experienced by contractors over time, could lead to a better understanding of the dynamic person-dependent changes that affect emerging contractors.