The 2000 Khayelitsha/Mitchell’s Plain (KMP) Survey offers a unique view into the hindrances to self-employment activity in the KMP area. Respondents identify a lack of money/capital as the primary barrier to participation and hours worked in self-employment activities. Concerns over expected profit, while present, are not a dominant hindrance. A lack of skills, concerns over future access to formal jobs, and other “hidden” costs also play a role in limiting self-employment activities, though these are far less important than issues related to capital constraints. Further research is needed to identify whether capital constraints are tied to a lack of access to start-up capital or a lack of demand for borrowing due to ex-ante risk management strategies.
There is a tremendous need for a better understanding of how self-employed businesses operate and succeed in South Africa. An improved understanding would help analysts better explain why forty-one percent of the labour force remains unemployed rather than entering some sort of self-employment activity, survivalist or otherwise.2 More importantly, we could then help public policy makers, informal worker organisations, and NGOs design programmes or regulatory structures that increase the profit levels of those currently engaged in these activities and/or encourage more people to enter profitable selfemployment opportunities. Identifying and addressing barriers and hidden costs faced by self-employed businesses has the potential to lead to substantial poverty alleviation and increased economic well-being for a large segment of the population. This paper will attempt to illuminate some of these hindrances to self-employment activity in the Khayelitsha/Mitchell’s Plain area, on the outskirts of Cape Town.
Research addressing informal self-employment is currently underway, but more is needed. Multiple sector specific industry studies of the self-employed have been conducted in Durban and other areas. For example, Lund (1998) reviews a variety of quantitative and qualitative survey work concerning street traders.3 Additionally, the role of informal workers in the clothing industry has been examined, with particular interest in addressing how the formal and informal economies interact (Skinner and Valodia, 2000). Carr and Chen (2002) call for more such studies, identifying the global value chain of the particular goods, with an emphasis on the role of the self-employed in developing countries and an ultimate goal of identifying how these self-employed individuals might capture a larger share of the overall returns.
Such studies are particularly useful when they include analysis of policies that may be applied by governments to address the problems of the self-employed. Skinner (1999) lays the foundation for analysing local government policies that may benefit street traders in five South African cities. The work includes some specific problems faced by street traders as well as the administrative regulations and the institutional structures that such traders face. Additional quantitative work is needed in this area to determine which programmes/structures have been most successful over time.4
In some sense, this work represents another such study focusing on the selfemployed and attempting to identify issues and structures that might hinder their economic progress. Far and away the best such existing work that the author is
aware of is Skinner’s 2005 review of constraints faced by over 500 informal firms in the Durban metropolitan areas. This study sets itself apart due to the breadth of coverage of business activities surveyed, its extensive set of questions directly assessing and ranking limitations faced by these small firms, and its direct enquiry into the role government may play to assist these firms as judged by those actually engaged in these activities.
Like Skinner (2005), this study does not limit its investigation to those individuals (that is, firms) already engaged in one particular industry. It is inclusive of a large set of industries individuals may (potentially) be engaged in. In contrast to Skinner, I assess input from the potential self-employed as well as actual self-employed individuals. I pay particular attention to identifying what prevents individuals from choosing self-employment rather than unemployment. In short, I address this issue from a labour supply perspective.
The call for such work has been building. Cichello et al. find results from KwaZulu-Natal that are suggestive of barriers in the informal economy, including self-employment (Cichello et al., 2005). Chen (2004) specifically calls for further research to answer questions such as “Do barriers to entry into the informal economy still exist in South Africa?” As early as 2001, Kingdon
and Knight had a paper aptly titled, “Why high open unemployment and small informal sector in South Africa?” In that paper, and other accompanying research, Kingdon and Knight provide economic and psychologically based evidence from the nationally representative October Household Survey that suggests that many individuals are not “voluntarily” unemployed, but must,
instead, be facing serious impediments to entry to the informal sector (Kingdon and Knight, 2001, 2002).
Kingdon and Knight go on to provide a bountiful list of potential constraints. Many of these potential constraints, such as restrictive bye-laws, poor entrepreneurial skills, poor social/trading networks, are acute in South Africa due to the legacy of the Apartheid regime (Kingdon and Knight, 2001). Other potential hindrances include capital/credit/land constraints, a lack of
infrastructure in Black townships, the prevalence of violence and insecurity in the informal economy, and, for those employing non-family members, industry based wage and working condition mandates (ibid). Yet, there is no direct investigation of these constraints. At this point, the set of plausibly important hindrances is still unwieldy, making effective modeling of the self-employment decision difficult at best.
More empirical work is needed. Such work must allow for an expanded set of potential hindrances, before narrowing down the list of hindrances that policy makers should be most concerned about. The aforementioned industry specific studies have been helpful in identifying some important barriers/constraints, particularly for Durban street traders (Lund, 1998; Skinner, 1999). The room
for additional work in this area is extremely broad.
In this working paper, I use responses from affected individuals to directly identify potential barriers and/or hidden costs that may impact the decision process for individuals considering or currently engaged in self-employment activity in the Khayelitsha/Mitchell’s Plain area outside of Cape Town. I start with three obvious deterrents to participation in self-employment: 1) zero or negative expected profit (due to a lack of demand for products/ services or a high cost of complementary inputs); 2) capital constraints and/or risk concerns (that limit the funds available or desire to access funds for a venture with a positive expected profit); and 3) a lack of skills (either skills particular to a specific type of self-employment or basic entrepreneurial skills). These hindrances shall be referred to as profit barriers, capital barriers, and skill barriers, respectively.
I also refer to future-limiting barriers which captures the notion that selfemployment choices today may limit labour supply choices in future periods. This idea, found in Harris-Todaro models extended to include the informal economy, suggests that working in the informal economy (including selfemployment) in the current period limits one’s chances of accessing formal
employment in the next period.5
All other barriers and hidden costs shall be referred to as hidden cost barriers. Other “barriers,” might include formal restrictions (such as regulations that prohibit sales in an area) or informal restrictions (such as gangs that only allow those in their group to sell in an area). By “hidden” costs, I am referring to costs that an analyst may not typically include in his/her analysis, particularly when treating individuals as atomistic agents in static labour market models that assume well functioning labour markets.6 These include economic costs, such as the loss of profit due to theft or extortion, and non-economic costs such as the fear of and impact from violence that is related to their choice of employment.
They also include the loss of household resources one might lose access to after leaving unemployment and the additional payments to other household (or extended kin) one might make once self-employed. Finally, “barriers” may be thought of in terms of “hidden” costs, where a perfectly enforced barrier represents an infinite cost.7 I will use the terms barriers, hidden costs, and hindrances, interchangeably.
Given that analysts typically ignore such hidden costs, surveys are generally poorly designed for capturing them. The Khayelitsha/Mitchell’s Plain (KMP) survey, however, is an exception. From this survey, I make use of information describing past self-employment activity (including previous business failures) as well as current self-employment activities to identify such costs. Additional responses from the current unemployed also help the investigation into barriers limiting activity within the self-employment sector.
In following this empirical approach of analysing subjective answers from labour market participants, rather than modeling the decisions and testing for defacto constraints, two cautions are in order. First, the analysis is based on subjective answers from participants, which may be misleading for a variety reasons. For example, some hindrances may be so deeply rooted in the landscape and the mindset of respondents that they will not mention them. Or, they may have a myopic or otherwise tainted view of what represents a constraint on their activities. The former problem is a serious concern. The latter may actually provide extremely useful information to policy makers, as it is the perceptions of labour market participants is as important as reality. Another potential problem in drawing policy conclusions from this analytical approach is the fallacy of composition. While one individual might prosper if s/he was able to overcome a barrier, this does not logically imply that a large population would similarly prosper should the barrier be removed for all. Bearing in mind these cautions, I find that the KMP survey responses provide some important information for policy makers and researchers alike.
First, profit barriers do not appear to be keeping many of the unemployed from engaging in self-employment activities. Thus, there may be considerable room for welfare improvements if public policy and NGO organisations can help the unemployed engage in these profitable activities.
Second, capital barriers appear to be the primary deterrent to entry and a limitation on profits from self-employment activity. It is unclear whether this lack of money/capital induces lower participation and profits due to a) ex-ante risk management strategies that induce the unemployed to remain unemployed rather than face the variable income stream and/or downside risk associated with
self-employment; or b) capital constraints that limit the liquidity these individuals have for initial capital investments. Pursuing research to disentangle these effects and implementing public policy programmes and/or interventions from NGOs or informal worker organisations to overcome these hindrances are the primary action oriented recommendations of this analysis.
Finally, there is also some evidence that skill barriers and hidden cost barriers also limit self-employment activity. However, these hindrances appear to be of considerably less importance than the capital barriers.
The layout of this paper is as follows. Section II describes the data used in the empirical work. Section III provides an empirical description of selfemployment activity in the Khayelitsha/Mitchell’s Plain area. Section IV investigates hindrances to self-employment activity. Section V discusses how to best move forward in creating an appropriate economic model of labour supply
decisions and details why distinguishing an appropriate model will be important for public policy makers, NGOs and self-employed worker associations. The conclusion briefly summarises the main points. Appendix A provides suggestions for those who conduct future survey work in this area.
I gained valuable insights from conversations with and comments by Chris Barrett, Gary Fields, Chris Manolis, Nicoli Nattrass, David Newhouse and participants at the University of Cape Town’s CSSR Seminar series. I would also like to thank Caroline Skinner, Imraan Valodia, and Francie Lund for their thought provoking discussions (and tour of informal trading areas), and their encouragement in pursuing this important area of research. Any errors are my own.
The narrow and broad unemployment rates for South Africa were 27% and 41%, respectively, in March 2005 (StatsSA, 2005). At the time of the survey used in this study, the narrow and broad unemployment rates for South Africa were 26% and 36%, respectively, in September 2000 (StatsSA, 2001).
Lund (1998) also provides a brief, yet extremely informative, review of the major historical regulations affecting the informal economy in South Africa as well as the government policy and regulations in the 1990s.
Chen (2004) notes that Budlender (2000) and Budlender et al (2004) reviewed the impact that local (Durban) as well as provincial (KwaZulu-Natal) and national government budgets have on a variety of informal economy participants. These studies find the local governments to be responsive to the needs of the informal economy.
See Fields (1975) for one example of an extended H-T model that incorporates the informal sector. The limiting future barrier may include concerns beyond those implicit in H-T models. For example, costs may also include an option value if starting a business now precludes the right to gain access to resources to start another business in the future.
If the reader prefers, s/he can think of these as “often ignored costs” or “typically trivial costs.” This is obviously open to interpretation. The greatest shades likely come from those items that belong in the formulation of expected profit but are oft ignored. This could be anything from informal business taxes (i.e. extortion payments, etc.) to oft ignored
transportation costs. The key point is to be open to the possibility that issues we typically ignore may be important in this setting.
It would be more typical for a barrier to add to the expected value and the variance of such costs. For example, regulatory bye-laws may increase costs in periods where one happens to get caught. The barrier of a missing formal credit market may simply increase the cost of capital as an individual moves to money lenders.