Research purpose: The purpose of this article was to analyse and compare the effects of, and level of compliance with, the regulation of the domestic worker sector in two very different residential areas in Pretoria.

Motivation for the study: Researchers have conducted all previous micro-study investigations of the topic in Langenhoven Park, Bloemfontein. We need more micro-level studies to investigate the effects of regulating this labour market because areas with different socioeconomic conditions may yield different results.

Research design, approach and method: The researchers followed a quantitative micro-research design using structured questionnaires. They used the research methodology applied in similar micro-studies as the basis of the survey to make the results comparable. They used the criterion sampling technique. Respondentscompleted 87 questionnaires in Orchards and 89 in Soshanguve.

Main findings: Evidence suggests that areas in close proximity to one another in the same metropolis yield significant differences in the wages or salaries and non-wage working conditions of domestic workers. A blanket approach to identifying and monitoring the effects of the legislation for this sector is not an appropriate one.

Practical/managerial implications: The sector needs micro-studies over an extended period and in different areas to form a more nuanced picture of this multifaceted labour market. This study emphasised the necessity for improved monitoring of the existing legislation.

Contribution/value-add: This is the first micro-study to compare the effects of regulating the domestic workers sector of two residential areas with different socioeconomic characteristics. The results will give authorities a better understanding of the level of compliance with, and effects of, the regulation of domestic workers. It will also guide the monitoring of, and enforcement decisions for, this labour market.

However, there has been a slight improvement in the figures. In 2007 just over a million (1.003 million) employees worked in this sector (Statistics South Africa, 2009, p. 15). The figure remained at about one million in 2010 (Department of Labour, 2010).

Data from the International Labour Organisation (2010a, p. 6) show that domestic work in developing countries makes up between 4% and 10% of total employment. In industrialised countries, this figure is much lower and ranges between 1% and 2.5%.¹

Domestic work is one of the oldest occupations for women in South Africa and provides work for a large number of them. In many countries, domestic workers were’particularly devoid of legal and social protection‘ and ’singularly subject to exploitation‘ (International Labour Organisation, 2010c, p. 12). International studies on attempts to protect domestic workers include studies in Brazil, India and Namibia (International Labour Organisation, 2010c, p. 14).

South Africa is no exception. In September 2002, the ’Sectoral Determination 7: Domestic Worker Sector‘ came into effect and granted formal labour market protection to South African domestic workers in the form of minimum wages, working hours and other conditions of employment (Department of Labour, 2002a). Despite these regulations, employers in the South African labour market generally ignore minimum wages. The biggest losers, because authorities do not enforce the regulations effectively, are security guards, farm workers, domestic helpers and some retail workers (Bhorat, Kanbur & Mayet, 2010). This calls for better monitoring of the compliance to, and effects of, these regulations. However, a lack of micro-level research in this field inhibits our understanding of the dynamics and behaviour in this labour market.

The study of Bothma and Campher (2003) predicted a decrease in the already declining demand for domestic workers. Hertz’s macro-study showed that the minimum wage legislation benefited domestic workers in terms of real wage increases and improved non-wage terms of employment (Hertz, 2004).

The macro-study by Hertz (2004) uses labour force surveys from September 2001 to September 2003. The data from these surveys include two waves before the effective date of the regulations and two thereafter. The surveys cover between 2400 and 3100 domestic workers per wave (Hertz, 2004, p. 2). This study also reported a significant reduction in the hours of work. The micro study of 2006, by Blaauw and Bothma (2010), confirmed this. They reported that employers are employing domestic workers on a part-time basis. Blaauw and Bothma (2010, p. 9) attributed this increase to the increased regulation of the sector. All studies reported a decrease in the employment levels of domestic workers.

Although the results of these studies are valuable for monitoring the compliance with, and effects of, these regulations, it leaves significant gaps in our understanding of the micro-economic effect of these regulations at a cross-regional level. Because of prevailing socioeconomic conditions, different areas may yield different results because of this legislation. We need cross-regional micro-studies to fill the gap in the current knowledge on this topic.

The structure of the rest of this article follows. Firstly, the researchers reviewed the literature on the regulation of the domestic workers sector. A description of the research methodology the researchers used in the study follows. The researchers then present the results of the study in terms of the stated research objectives. The article concludes with a discussion of the main findings, conclusions and recommendations as well as the limitations of the study and suggestions for future research.

Most domestic workers work in the informal and unregulated sectors. Most of them are unskilled workers (Matjeke, 2008, p. 18). Barker (2003, p. xix) defines the informal sector as the unorganised and unregulated sector with mostly legal but unregistered economic activities.

The wages and employment conditions of domestic workers have raised worldwide concern for many decades. As far back as 1965, the International Labour Conference adopted a resolution on the conditions of employment of domestic workers. This resolution recognised the ’urgent need‘ to establish minimum living standards that are ’compatible with the self-respect and human dignity which are essential to social justice‘ for domestic workers in developed as well as developing countries (International Labour Organisation, 2010a, p. 1).

In various parts of the world, domestic workers engaged in collective action trying to seek their rights. In Brazil, the first domestic workers’ organisation formedin 1936 in São Paulo. After 1988, Article 7 of the Brazilian constitution protected their right to do so (Brazil Constitution, 1993). In India, its supreme court decided in 1977 that it would not recognise domestic workers as ‘organised labour’ and attempts at legislative reform failed. In Namibia, the Namibian Domestic Workers’ Union (NDWU) had recruited about one-third of the 12 000 domestic workers in the country (International Labour Organisation, 2010c, p. 14).

In 2009, the government of Cambodia convened a national consultation on decent work for domestic workers. The government, workers’ and employers’ organisations in the country agreed that they needed new international labour standards that extended social protection to domestic workers (International Labour Organisation, 2010b, p. 7).

In 1999, the minister of labour of South Africa announced his intention to introduce minimum wages for domestic workers. An investigation by the Department of Labour into the minimum wages and conditions of employment of domestic workers followed (Mdladlana, 2001, p. 2). In 2002, the government promulgated the ’Sectoral Determination 7: Domestic Worker Sector‘. In 2003, it amended the Unemployment Insurance Act to include domestic workers (Hertz, 2005, p. 1). The legislation provides benefits for unemployment, illness, dependants, maternity and adoptions.

The ’Sectoral Determination 7: Domestic Worker Sector‘ sets, amongst other things, minimum wages, maximum working hours and deductions for accommodation (Department of Labour, 2002a). The minimum wage regulation applies to all domestic workers, whereas the other conditions of employment apply only to domestic workers who work more than 24 hours a month for an employer (Department of Labour, 2002b, p. 6).

It set minimum wages according to two criteria. These were the area of work and the number of hours worked a week. It divides areas of work into metropolitan and rural according to municipal boundaries (Department of Labour, 2002a, p. 8).

The number of hours worked per week criteria distinguish between domestic workers who work for more than 27 ordinary hours a week and those who work for 27or fewer ordinary hours a week. The wage or salary for domestic workers in metropolitan areas, who work more than 27 hours a week (for the period from 01 December 2006 to 30 November 2007), amounted to R5.57 per hour, R246.21 per week and R1066.83 per month (Department of Labour, 2006, p. 1). This timeframe is important as it coincides with the timeframe of the fieldwork for this study.

The minimum hourly wages for domestic workers who work fewer than 27 hours a week are slightly higher than for those who work more than 27 hours a week to compensate the worker for not having a full-time job. For the same period, these wages and salaries were R6.46 per hour, R174.50 per week and R756.09 per month(Department of Labour, 2006, p. 1).

Therefore, the ’Sectoral Determination 7: Domestic Worker Sector‘ regards workers who work for 27 hours or fewer as part-time workers. The minimum wages for domestic workers in rural areas are also generally lower than the prescribed minimum wages in metropolitan areas.

These prescribed minimum wages are to increase annually at the inflation rate, as the Consumer Price Index (CPI) measures it, plus 2%. The Government Gazette must publish any subsequent wage increase, which exceeds 8%, as an amendment to the ’Sectoral Determination 7: Domestic Worker Sector‘ (Department of Labour, 2002a, p. 7). If the annual increase in the CPI is 10% or higher, domestic workers are entitled to receive a wage increase equivalent to the increase in the CPI (Department of Labour, 2002a, p. 7).

However, some economists argue strongly against setting minimum wages and suggest that minimum wage policies impede job creation and employment in the sectors where they are implemented (McConnell, Brue & Macpherson, 2010, p. 398). High minimum wage levels can discourage employers from engaging new workers because it increases the relative cost of labour to capital. Employers will shift their choice of production methods to relative cheaper capital. This argument holds specifically for sectors where the minimum wage is above the existing wage bands, especially for unskilled labourers. However, one expects governments to create enabling environments, in terms of labour legislation, to encourage employers of unskilled workers to employ more of them (McConnell et al., 2010, p. 398).

Domestic workers may work a maximum of 45 ordinary hours a week, or nine hours a day if they work five days a week and a maximum of eight hours a day if they work for more than five days a week. When domestic workers work for longer hours, they are eligible for overtime pay (Department of Labour, 2002a, p. 15).Overtime payment amounts to one- and-a-half times the hourly wage (Department of Labour, 2002a, p. 16).

Deductions for accommodation should not exceed 10% of the total wage or salary. Paid accommodation should also meet conditions. Rooms should be weatherproof and in good condition. They should have at least one window, a door that one can lock as well as a toilet and a bath or shower. Alternatively, tenants should have access to a bathroom (Department of Labour, 2002a, p. 36).

In April 2003, the government also amended the Unemployment Insurance Fund (UIF) and employers had to register their domestic workers with the UIF for the first time. Domestic workers and their employers had to contribute 1% of the domestic workers’ wages or salaries to the fund. The fund excluded domestic workers who worked fewer than 24 hours a month from this provision. After April 2003, domestic workers were able to claim unemployment benefits (Hertz, 2005, p. 1).

The researchers will now analyse the effects of, and compliance with, the regulations in Orchards and Soshanguve.

The researchers traced domestic workers in the two areas through observing and interviewing people informally in these areas. The observation showed that most households in Orchards employ a domestic worker. Although Soshanguve is a very large residential area compared to Orchards (roughly three to four times the size of Orchards), the number of households that employ domestic workers is very limited.

The researchers interviewed residents from blocks C, G, F, L, K, CC and WW in Soshanguve. The number of observations in each suburb corresponds to the number of observations researchers used in the micro-studies in Langenhoven Park, Bloemfontein.

The respondents completed the questionnaire in structured interviews rather than by themselves. According to Bless and Higson-Smith (1995, p. 111), ‘… this method can help to prevent misunderstandings and misinterpretations of questions‘ because researchers can explain exactly what information they need from the respondents and can immediately clarify uncertainties about the questions. It is also valuable for overcoming language barriers, especially when dealing with illiterate and uneducated samples.

Kingdon and Knight (2001, p. 2) also stated that investigating unemployment in South Africa requires reliable micro-economic data that comes directly from people or about the people from their households. An investigation of the dynamics in any labour market relies on reliable micro-economic data (Blaauw & Viljoen, 2009, p. 51; Burger & Von Fintel, 2009, p. 1).

The researchers used purposeful sampling to complement criterion sampling. The main advantage of this method is that it allows researchers to target those places where there is abundant information to ensure that the investigation is comprehensive (Jacob, 2007, p. 34).

Furthermore, the sampling process also involved some degree of convenience sampling. In cases where domestic workers or employers were not present at the time of the visit to that household, the researchers would substitute that household with another one in the same geographical block.

Adherence to accepted ethical standards was also an integral part of the research procedure. In order to comply with the ’doing no harm‘ principle in research, the researchers trained fieldworkers to identify themselves clearly, explain that the research was for academic purposes only and that it was not linked to any government department (Bless & Higson-Smith, 1995, p. 102; Rubin & Babbie, 1997, p. 59–63).

At all times, respondents had the right to refuse an interview. Therefore, the study adhered to the principle of voluntary participation. Confidentiality and anonymity is another important ethical principle and can become a weakness in any survey research (Bless & Higson-Smith, 1995, p. 103). Therefore, the researchers protected the identities of the respondents and did not disclose their names. The researchers are confident that they have complied with all the ethical requirements.

The researchers recruited and trained two final-year students enrolled for a National Diploma in Accounting and Auditing at the Tshwane University of Technology (TUT) as fieldworkers to help with conducting the interviews. The study benefitted from the close proximity of TUT to the areas the researchers chose for the interviews. The fieldworkers speak several of the official languages prevalent in the two areas of the field study. They have extensive experience in structured oral and written student awareness, attitude and prevalence survey projects. The researchers trained both fieldworkers before the start of the fieldwork. The training included theoretical coaching as well as simulation exercises.

Earlier research in other geographical areas had field-tested the core sections of the questionnaire. Therefore, the researchers decided that a fit-for-purpose pilot study was not necessary.

The fieldworkers conducted the interviews during June and July 2007. They distributed the questionnaires in person to the chosen respondents.

The fieldworkers always informed respondents that their participation was voluntary and that they could suspend the interview at any time. Furthermore, they assured the respondents of anonymity and that the research was for academic purposes only.

The fieldworkers conducted structured interviews with domestic workers and their employers in the two residential areas to complete the questionnaires, to achieve the objectives of the study and to compare its results with other micro-studies. Respondents completed 87 questionnaires in Orchards and 89 questionnaires in Soshanguve.

The first author held regular debriefings with the fieldworkers to check on their progress. He compared the data in the completed questionnaires to his own informal interviews with employers and domestic workers. The training that the fieldworkers received also focused on the importance of reliable research results.

The researchers cleaned the data and eliminated all the questionnaires that contained possible misinterpreted responses. In cases where domestic workers were not present (especially during weekends), they asked employers to respond to the questions about the demographics of their employees. When an employer gave a suspicious response, the researchers regarded the questionnaire as spoiled and excluded it from the survey.

As a result, one questionnaire from Orchards and four from Soshanguve were spoiled and the researchers eliminated them. This brought the total number of completed questionnaires in Orchards to 86 and to 85 in Soshanguve.

The level of education of the domestic workers in Orchards and Soshanguve in 2007 was relatively low. Most of the respondents (63.3% in Orchards and 67.1% in Soshanguve) reported having no formal education. The percentage of respondents with some primary education was almost the same in Orchards as it was in Soshanguve – 8% and 10.6% respectively. The figures for secondary education differ somewhat in the two areas – 28.7% in Orchards compared to only 22.3% in Soshanguve.

One can attribute the slight differences between the education levels in Orchards and Soshanguve to the differences in the social status of the two areas. Orchards is a more elite area and one expects its residents to employ more workers than will those in Soshanguve, which is a residential area with larger variances in the socialstatus of its residents.

However, the number of domestic workers without formal education was much lower in Langenhoven Park and declined from 25% in 2001 to 19.7% in 2006 (Blaauw& Bothma, 2010, p. 3). The total percentage of domestic workers with some secondary education was much higher in Langenhoven Park and increased from 32% in 2001 to 46.2% in 2006 (Blaauw & Bothma, 2010, p. 3; Bothma & Campher, 2003, p. 198).

Laundry, which includes washing and ironing, is the duty that most domestic workers in both Orchards and Soshanguve performed. Washing occurs far more frequently in Orchards and Soshanguve than it does in Langenhoven Park, but ironing has about the same incidence in all three areas. Cleaning was part of the job descriptions of a significant number of domestic workers (91.9%) in Orchards compared to the 61.9% in Soshanguve. Cleaning was the most important duty for domestic workers in Langenhoven Park as well (Blaauw & Bothma, 2010, p. 4; Bothma & Campher, 2003, p. 198).

Cooking was more common in the job descriptions of domestic workers in Orchards. Cooking was part of the responsibility of 25 respondents (31.6%) in Orchards compared to only seven respondents (9.1%) in Soshanguve. In Langenhoven Park, the importance of cooking changed over the five-year period. It decreased from20.7% in 2001 to 15% in 2006. This possibly reflects the part-time nature of domestic worker employment in Langenhoven Park.

The demand for child- and elder-care, at 61 respondents (70.9%), is highest in Orchards. This is far higher than that of the 17 respondents (20.7%) in Soshanguve. Although the figure in Langenhoven Park doubled between 2001 and 2006, it was still lower than it was in Soshanguve. One can attribute this to the fact that Orchards is a new and relatively more youthful area compared to the others.

The average for Soshanguve was lower than that in Orchards, lower than the 6.9 hours a day recorded in 2001 and lower than the 6.65 hours recorded in Langenhoven Park (Blaauw & Bothma, 2010, p. 4; Bothma & Campher, 2003, p. 198) in 2006.

Most employers in Orchards (40%) employ domestic workers for eight hours a day. Only 17.4% employ them for seven hours a day and another 17.2% employ them for nine hours a day. In Soshanguve, the percentages of domestic workers employed for seven, eight and nine hours per day are 26.6%, 21.5% and 26.6% respectively.

The minimum number of days a week employers in Orchards employed domestic workers is two days, whilst the maximum is seven days. Most (83.9%) employers used their domestic workers for five or more days a week. In Soshanguve, the minimum number of days a week employers used domestic workers is one day and the maximum is six days. Only 39% of the employers in Soshanguve employed their domestic workers for five or more days a week.

In Langenhoven Park, most (41%) employers employed domestic workers for only one day a week in 2006. This is much higher than the figure of 33.6% in 1998 (Blaauw & Bothma, 2010, p. 5).

In Soshanguve, 51.2% of the respondents, who work for more than one employer, work for two employers, whilst 43.9% work for three employers.

In Orchards, the incomes of domestic workers varied between R300 and R1899 a month in 2007. In Soshanguve, the lowest-paid earners earned R60 and the highest-paid domestic workers earned R1500 a month during the same period. The average income in Orchards was R833.68 a month and in Soshanguve it was R555.29 a month. The largest percentage of domestic workers in Soshanguve (70.6%) earned an income below R700 a month. Most domestic workers in Orchards (81.43%) earned more than R700 in 2007.

Table 6 shows the average wage per day or week a domestic worker earned in Orchards and Soshanguve in 2007.

The average daily wage of domestic workers, who work for only one day a week, amounted to R98.21 in 2007 compared to the R35 average for domestic workers who work for seven days. Therefore, it may be more beneficial to work for many employers. However, working for more than one employer demands more effort and is not sustainable in the long run.

Daily wages show significant variations between different regions. The average daily wage of R98.21 in 2007 for working only one day is considerably higher than that recorded in 2001 and 2006 in Langenhoven Park in Bloemfontein. In 2001, it amounted to R45.68 (Bothma & Campher, 2003, p. 199). In 2006, it was R65.47 (Blaauw & Bothma, 2010, p. 5). Some employers also contribute towards their domestic workers’ travelling costs. These contributions increase their earnings.

In Orchards, only 21% of the employers supplemented their employees’ income with travelling allowances. In Soshanguve, this percentage is much higher at 32.5%. At first glance, this is surprising. However, the tendency for most domestic workers in Orchards to sleep at their place of work may explain this. Several employers in Orchards provide accommodation at the place of work, negating the need to pay transport allowances. Only 24.4% of domestic workers have to travel, and 21% of these do receive compensation for travelling (see Table 7).

In Soshanguve, on the other hand, 89.4% of the domestic workers have to travel and only 32.5% receive additional compensation for travelling. Therefore, domestic workers in Orchards are better off than those in Soshanguve are.

Table 8 presents the monthly average salaries, adjusted for transport, which employers in Orchards and Soshanguve paid in 2007. The ‘Sectoral Determination 7: Domestic Sector’ does not regulate these. Therefore, they are open to negotiation between the parties (Department of Labour, 2002a, p. 33).

On average, domestic workers’ salaries (adjusted for transport) in Orchards are still 35.63% higher than those in Soshanguve are, although the percentage of employers who contributed to their employees’ transport costs is lower in Orchards compared to those in Soshanguve.

FIGURE 1: Domestic workers’ monthly salaries in Orchards and Soshanguve (2007).

Slightly more than 47% of employers in Orchards signed service contracts with their domestic workers in 2007. This figure stood at only 10.6% for Soshanguve. In Langenhoven Park, this figure improved from 30.3% in 2001 to 75.8% in 2006 (Blaauw & Bothma, 2010, p. 5; Bothma & Campher, 2003, p. 199).

This study shows that areas in close proximity to each other show significant differences in wages and the non-wage working conditions of domestic workers in South Africa. This is a manifestation of the differences, for different areas in the same metropolitan region, in complying with the legal requirements of the Department of Labour in South Africa.

The analysis of the demographic characteristics, employment structure and characteristics, wage or salary levels and working conditions of domestic workers in Orchards and Soshanguve reaffirms the results of the previous research of Blaauw and Bothma (2010) as well as that of Bothma and Campher (2003). Domestic work tends to be an occupation for middle-aged women in South Africa. One may attribute this to the fact that young people are becoming more skilled because of compulsory school attendance. In addition, domestic work has always been an employment opportunity for the unskilled, as the data on the educational level of domestic workers suggests. The large number of domestic workers in Orchards and Soshanguve without any formal education confirms this.

A pleasing aspect of the analysis of the employment characteristics of domestic workers in the two areas is that most employers adhere to the legislation that specifies the maximum number of working hours a day. Most domestic workers in Orchards and Soshanguve work for seven or more hours a day. They work a maximum of nine hours a day if they work for five days a week and a maximum of eight hours a day if they work more than five days a week. In Langenhoven Park, the number of hours worked per day did not change significantly between 2002 and 2006.

Almost all employers in Orchards employed their domestic workers for five or more days a week, with an average number of 7.13 hours worked per day. This suggests that domestic workers in Orchards work full-time. On the other hand, Soshanguve shows a greater demand for part-time domestic services. This corresponds well with the findings in other areas of the country. In Langenhoven Park, Bloemfontein, there has also been a significant shift from full-time to part-time domestic work since 2001 (Blaauw & Bothma, 2010).

A significant difference between Orchards and Soshanguve is that most domestic workers in Orchards worked for only one employer in 2007. The figure in Soshanguve is almost half that in Orchards, while in Langenhoven Park it is almost a third. The reason for this might be that the affordability of domestic workers differs between the areas. If households cannot afford to employ domestic workers full-time, domestic workers try to supplement their income by working for more than one employer. On the other hand, domestic workers might prefer to work part-time because the minimum prescribed hourly wages for part-time workers are higher than that for full-time workers. This finding is also consistent with that of Blaauw and Bothma for Langenhoven Park (2010).

The average monthly salary in Orchards was much higher than that in Soshanguve. If one compares this with the prescribed minimum salary for 2007, it is clear that the average monthly salary in Orchards (of R833.68) falls somewhere between the minimum prescribed salaries for part-time and full-time workers of R756.09and R1066.83 respectively. Given that 83.9% of workers in Orchards are full-time employees, one would expect that they would receive incomes of at least R1066.83 per month. The results show that only 64.4% of employees received an income higher than R756.09 and only 9% received R1066.83 or more. Most workers earned salaries of R900 a month.

In Soshanguve, where most workers are part-time, only 15.3% received an income above the minimum prescribed monthly salary of R756.09 for part-time workers. Therefore, most employers in these areas did not comply with the minimum wage legislation at the time of the study. It is also clear that the average daily wage and the number of days that a domestic employee worked for a particular employer correlate negatively. The minimum wage prescribed for domestic workers, who work fewer than 27 hours a month, is higher than for those who work more than 27 hours a month. This may explain the tendency.

In Soshanguve, most workers have to travel to and from work. However, only a few of them received additional compensation for travelling.

Domestic workers’ salaries, adjusted for transport costs and accommodation deductions in Orchards, were still higher than those in Soshanguve. According to these adjusted salaries, a maximum of 15.1% of the domestic workers in Orchards received a monthly income of R1066.83 and higher a month. A maximum of 14.1% of workers in Soshanguve earned more than R756.09 a month in 2007. If one compares the prescribed minimum salaries for 2007 with these adjusted average monthly salaries, the differences are not that large. The overall conclusion is that very few employers in Orchards and Soshanguve complied with the minimum income regulations at the time of the study.

These results contrast starkly with the situations in other areas like Langenhoven Park (Blaauw & Bothma, 2010). Compliance levels here are significantly higher than in the areas surveyed in this study. Furthermore, this study shows a lack of compliance with the UIF registration requirements.

The results of this study have several important implications for policy makers. Most important is the realisation that distinguishing between rural and urban areas in the sectoral determination of minimum wages for domestic workers may very well camouflage significant differences, even in the same urban area. The results suggest that there are important drivers, like socioeconomic characteristics and the income levels of the employers in different suburbs in the same metropolis, which influence the levels of compliance with legislation.

Therefore, policy makers must carefully consider these factors when determining the levels of minimum incomes for domestic workers. Added to this is the recognition that any blanket approach to assessing the effects and outcomes of minimum income legislation for domestic workers will not be the most appropriate monitoring strategy.

There is always the possibility that some of the sampling blocks in the two areas may have been slightly over- or under surveyed when the fieldworkers conducted the interviews. Furthermore, the results do not apply outside of the geographical areas the study covered. One cannot generalise the results to other geographical areas.

This limitation is common to all micro–studies. In fact, it suggests that studies should state that their aims are to investigate the differences in the effects of the legislation in areas that are close each another.

The practical consequence of this limitation is the realisation that researchers should conduct more micro-studies in different areas over a longer period. This will help policy makers to form a more complete picture of the changes occurring in this labour market because of the government’s legislative efforts. This is an important area for further and ongoing research on this topic.

Future research of this nature can provide the data for important panel data analysis in order to test the robustness of observed changes in this labour market over time. In fact, previous studies on this topic have identified the unavailability of successive waves of data from the same respondents as limiting factors in microanalyses of this labour market.

The results of this study also open the door to qualitative research in this field. Focus group research amongst domestic workers and their employers can provide valuable insight into the decision-making behaviour of the economic agents in this important labour market. This data is necessary for investigating aspects like the well-being of those people who are the intended beneficiaries of the minimum wages the government has imposed.

One cannot over emphasise that a blanket approach to setting and assessing the effects of minimum incomes in this important labour market will be inappropriate.

The differences in the employment characteristics, wages or salaries, non-wage working conditions and the compliance levels the researchers observed in this study emphasised the necessity and importance of conducting further micro-studies of this nature in the various metropolitan and rural areas of South Africa.

Further micro-studies like this one can provide a better perspective on the effects of these policies on the lives of those who are their intended beneficiaries.

1.All countries in the world fall into two broad categories. These are industrial countries (also referred to as developed countries in the literature) and developing countries. Industrial countries are high-income countries and are mostly located in Europe and North America. Developing countries are heavily concentrated in Africa, Asia and South America. South Africa is a middle-income developing country (Mohr, Fourie & associates, 2008, p. 50).