Trade in reed-based craft products in rural villages in the Eastern Cape, South Africa

Abstract

Selling traditional craft products made from fibrous plants is an important source of income for economically vulnerable rural women. In the Eastern Cape Province of South Africa, Cyperus textilis and Juncus kraussii have been used for centuries to make products of functional and cultural importance, such as sleeping mats and baskets. In the former Transkei village clusters of Mpozolo and Ntubeni, female crafters harvest the raw material and make and sell the products in their communities and in nearby towns. Interviews with 40 of them revealed what the trade contributes to their livelihoods and what enhances or limits their success. The findings show that crafting contributes vital income to vulnerable households, on average 26  ±  4 per cent of annual household cash income, over 40 per cent for the poorest households and 5–15 per cent for wealthier households. Lack of access to non-traditional markets was identified as the main constraint on the trade.

1. INTRODUCTION

Plant fibres are used worldwide to make products of functional and cultural importance, such as baskets, brooms, mats and rope. In places such as Central and South America (Ugent, 2000) and the Caribbean (Berman & Hutcheson, 2000) this activity dates back to ancient times. These products are still made and used today, and often sold to generate income within the communities in which they are made and also to non-traditional and tourist markets for their decorative and novelty value. The commercialisation of craftwork has been viewed as a development option for poor rural communities in several southern African countries, where the sale of baskets and other craftwork to tourists and the export market has increased rural incomes (Cunningham & Milton, 1982; Terry & Cunningham, 1993). This research deals specifically with the contribution of trade in woven, reed-based craft products to the livelihoods and household economies in two rural villages in the Transkei region of the Eastern Cape, South Africa.

The amaXhosa people of the Eastern Cape have been making craft from fibrous plants, such as palms, grasses and sedges, for centuries (Kepe, 2003). Species such as Cyperus textilis, Juncus kraussii, Phoenix reclinata and Flagelleria guineensis (Heinsohn & Cunningham, 1990; Cawe & Ntloko, 1997; Gyan & Shackleton, 2005) are used to this day to make products of functional and cultural importance, such as mats and baskets (Mathe, 2001, Kepe, 2003). For day-to-day use, most households have sleeping mats which children and sometimes women sleep and sit on, baskets which are used for holding harvested crops or animal dung (used to fuel fires) and brushes for functional and cultural use. Crafted products are sometimes used as payment for services or goods received from neighbours, and in this way contribute to the relative wealth of households. For cultural purposes, these products are often given and received as part of traditional ceremonies such as weddings and the initiation of boys and girls into adulthood, and they symbolise the starting of a new life (Kepe, 2003; Cocks & Dold, 2004). Food mats, or isithebe, are used to serve meat at community feasts and are such an integral part of this tradition that the gatherings have also come to be known as isithebe (Kepe, 2003). The commercialisation of these products has increased in the last few decades, as private entrepreneurs and non-governmental organisations (NGOs) have taken to buying and reselling craftwork (Cunningham, 1987; Krüger & Verster, 2001), and as rural people have migrated to urban areas and are no longer able to access the resources needed to make the products for themselves. Local-level trade in these products remains undervalued by most governmental policy makers. As a result, institutional support for these activities is lacking (Scherr et al., 2003; Shackleton & Shackleton, 2004).

Internationally, there has been an increased acknowledgement over the last ten to 20 years of the role that biodiversity and natural resource products play in contributing to the wellbeing and livelihoods of the rural poor (Arnold & Ruiz-Pérez, 2001; Koziell & Saunders, 2001; Campbell & Luckert, 2002; Kaimowitz, 2003). Numerous studies have shown that millions of rural people all over the world depend on a diversity of natural resource products for their livelihood and income needs (Byron & Arnold, 1999; Narendran et al., 2001; Scherr et al., 2003; Shackleton, 2005). Trade in traditional craft specifically often represents a critical source of income for very vulnerable members of society, especially rural women with few options for cash generation. For example, Shackleton et al. (2002) report on findings from a study conducted in northern KwaZulu Natal, South Africa: here, trade in products such as baskets, mats and hats woven from plant fibres served as a safety net for the poorest, most vulnerable households, for whom craft trade was a last resort in the absence of any other source of income (Shackleton & Shackleton, 2004). As craft economies are often based on traditional skills and make use of locally available natural resources, production can be home-based, requires very little or no capital input and is therefore ‘an important entry point into the economy’ (Rogerson, 2000) for people with few or no employment alternatives.

For these very reasons, trade in natural resource-based products such as traditional craft seldom reaches a scale at which poverty eradication can be achieved (Kaimowitz, 2003), and the initial enthusiasm about the potential for these products to provide pathways out of poverty has come into question (Neumann & Hirsch, 2000; Wunder, 2001; Belcher et al., 2005). Nevertheless, the direct-use value and commercial value of small-scale local trade in such products has been shown to be substantial in certain contexts and for certain individuals, contributing over 25 per cent of household income streams (Cavendish, 2000; Dovie et al., 2002; Shackleton, 2005).

There is therefore currently a mixed picture of the actual importance of natural resources for rural livelihoods and poverty alleviation (Ros-Tonen & Wiersum, 2005), and there are gaps in understanding. This situation has led to a call for further empirical case studies (Arnold, 2002), not just of market chains, but of contribution to total livelihoods (Luckert et al., 2000, Shackleton, 2005). In South Africa, several studies have reported on the contribution that natural resource-based craft trade makes to rural livelihoods, both economically (Krüger & Verster, 2001; Mathe, 2001; Kepe, 2003; Gyan & Shackleton, 2005) and through non-economic benefits such as the maintenance of tradition (Rogerson & Sithole, 2001; Cocks & Dold, 2004). However, none have considered the contribution of this kind of trade relative to other livelihood sources, nor quantitatively differentiated its relative importance to people and households of varying vulnerability within heterogeneous rural communities (Shackleton, 2005; Shackleton & Shackleton, 2006).

The primary objective of this study was therefore to add to a growing body of knowledge, and to address gaps in current understanding, by holistically characterising the contribution that the local trade in traditional reed craft products makes to crafter livelihoods in two clusters of rural villages in the Transkei, South Africa. A secondary objective was to identify the specific factors which either enhance or limit the economic success of individual crafters. The key questions to be answered were therefore: (i) what is the socio-economic profile of the crafters; (ii) what contribution does the trade make to local livelihoods relative to other sources of income, and for households of varying socio-economic status; (iii) what are the perceptions of the sustainability of the trade; and (iv) what are the opportunities for and constraints on trade?

The study focused on local trade in craft products made from two species of wetland reeds: Cyperus textilis, known locally as imizi, and Juncus kraussii, known locally as uduli. The culms (flowering stalks) are split and woven, or rolled into twine (Mathe, 2001). Traditional reed-based products traded in the area include sleeping mats (amakhukho), food mats (izithebe), baskets (iingobozi) and beer strainers (iintluzo).

2. STUDY SITES

This research was carried out in Mpozolo (32° 10′ S; 28° 46′ E), and Ntubeni (32° 12′ S; 28° 58′ E) in the former Transkei, Eastern Cape Province, South Africa (Figure 1). Mpozolo and Ntubeni are administrative areas comprising fourteen and seven villages respectively. These villages fall under the Mbhashe Local Municipality and the Amatole District Municipality.

Figure 1: The study areas and nearby towns

This study included a comparison of the trade in the two areas which differ primarily in their distance from urban centres and the Dwesa Nature Reserve and in the extent to which external interventions have had an influence (Table 1). These villages were selected because of the presence of an NGO called REHAB in Mpozolo, which is working towards helping improve trade in traditional crafts in the area.

Table 1: Differences between the two study sites

	Mpozolo village cluster	Ntubeni village cluster
Coordinates	32° 10′ S; 28°46′ E	32° 12′ S; 28°58′ E
Distance from Willowvale	20 km	45 km
Distance from Idutywa	45 km	70 km
Distance from Dwesa Nature Reserve main gate	25 km	1–5 km
NGO intervention	REHAB involved in training and marketing since July 2003	None

The climate in this region is subtropical and the mean annual temperature near the coast is 20°C (Van Wyk & Smith, 2001). Mean annual rainfall in the area is in excess of 1000 mm, with most rain falling between October and April. The vegetation in this area can be characterised as Coastal Forest (Low & Rebelo, 1996), consisting mainly of shrubland, with indigenous forest patches isolated to protected riverine gorges and coastal dunes (Van Wyk & Smith, 2001; Kepe, 2003). The area is characterised by high levels of biological productivity and biodiversity.

The population of the Mbhashe Local Municipality is estimated at 253 379 people (Statistics South Africa, 2001). Residents of the Eastern Cape, particularly of the former Transkei, have seen significant political upheaval but little economic development since 1994. The Eastern Cape is one of the poorest, least developed regions of South Africa, with the highest provincial unemployment rate, at 49 per cent (Dold & Cocks, 2002). The majority of the population of the Eastern Cape is rural, and 73.5 per cent have a monthly household income of less than US$250 (Office of the Premier, 2001). Most make a livelihood by arable and livestock farming (mainly subsistence) and collecting and using a range of natural resources, and many depend on pensions and migrant remittances (Office of the Premier, 2001; Kepe, 2003).

The village clusters in which this research was conducted are close to the Dwesa Nature Reserve, a provincial nature reserve of approximately 57 km² (DEAE&T, 1999). There is a history of conflict over the rights of local people to harvest resources within the boundaries of the reserve (Palmer et al., 2000). At the time of this study, however, they were permitted to harvest thatching grass and weaving reeds from inside the reserve once a week for seven months, between April and October, but were restricted to harvesting only as much as they could carry.

3. APPROACH AND METHODS

3.1 Data collection

Data were collected via replicate interviews with 40 crafters, 18 in Mpozolo and 22 in Ntubeni, as well as engaging in less formal group discussions with them and accompanying them on harvesting trips. Interviews and discussions were conducted in isiXhosa, with the assistance of a translator. All interviewees were involved in the manufacturing and trading of at least one of the four types of products mentioned above.

To identify any external factors that might influence the crafters' success, Mpozolo and Ntubeni were compared (Table 1). In Mpozolo, interviewees were selected on the basis of their involvement with the recently launched crafting project facilitated by REHAB. This project was initiated as a result of the community identifying traditional craft as a focus for development efforts, along with community gardening and the training of health workers. Participants in the crafting project were required to have already been involved in crafting on their own initiative, and the project involved further training and assistance with making unique products, the teaching of basic business skills and the purchasing of craft products to be sold at markets in urban centres such as Port St Johns and East London. The interviewees from Ntubeni were selected using the snowball sampling technique (Neis et al., 1999), whereby an initial crafter identified other individuals involved in the crafting trade, and so on. Analyses were performed on both the data set for the individual samples and the combined data set.

The interview schedule was designed to collect data in the following main categories: individual crafter characteristics; the crafter and the trade; household characteristics; harvesting and the resource; processing, marketing and selling; and income. The questionnaire-based approach to data collection is limited in that it relies on respondents' memory, and the assumption is made that their recollections are reasonably accurate. Informal group discussions were used to obtain information about the manufacturing processes, the species used and institutional requirements for harvesting.

3.2 Data analysis

For continuous data, simple descriptive statistics were used to summarise the data. For data expressed in temporal terms such as daily, weekly or monthly income from trade, total household income, amount of resource harvested, etc. all quantities were converted to annual figures. In terms of seasonality of the trade, ‘busy months’ refer to June and December, when many culturally significant events such as weddings and circumcision rituals take place, and at which time traditional craft is in high demand. ‘Normal months’ refers to the remaining ten months of the year.

The calculation of costs included only direct costs, such as the cost of travelling to harvest or to trade centres, and the cost of such items as twine and scissors. The opportunity costs of labour were not calculated, because crafting is carried out on an ad hoc basis, in between or at the same time as other activities, which makes it difficult to compare to other forms of wage labour. However, returns per hour were calculated in order to investigate the relationship between times spent crafting and returns from crafting at the two study sites. In 2004 the exchange rate between the US dollar and the South African rand was US$1.00 = R6.08.

Because of the high variability and hence non-normal distribution of the data, non-parametric Mann–Whitney U-tests were used to test for significant differences between incomes at the two study sites. The sample was divided into income classes in quartiles, based on total annual household cash incomes. Differences in the contribution of crafting to these different income classes were checked using Kruskal–Wallis tests. Relationships between categorical variables were investigated using χ² tests.

4. RESULTS AND DISCUSSION

4.1 Harvesting, weaving and trading

All the crafters interviewed were involved in producing and selling their own craft; 85 per cent of interviewees harvested the resource themselves, and the remaining 15 per cent bought bundles of reeds from others. All the Ntubeni crafters and 34 per cent of the Mpozolo crafters used C. textilis and J. kraussii that had been harvested from the Dwesa Nature Reserve. Outside the reserve, crafters harvested C. textilis from nearby rivers and wetlands and the gardens of neighbours and relatives, or their own gardens. One-quarter of interviewees grew C. textilis in their own gardens. The institutional arrangement identified by Kepe (2003), whereby there is private ownership of C. textilis in the communal wetland areas as well as in home gardens, applies to this study area.

The most commonly traded craft products were large sleeping mats – 80 per cent of interviewees sold these items (Table 2). Compared to other craft products, large sleeping mats also used the largest amount of resource, took the longest time to make and sold for the highest price. A simple weaving frame can be used to make sleeping mats; 42 per cent of the crafters who made mats used such a frame. Large sleeping mats are important culturally as well as functionally: they are favoured wedding gifts, and are placed in graves to symbolise ‘resting in peace’ (Kepe, 2003). In group discussions, the crafters said that sleeping mats were popular among urban people who wanted to remain in touch with their culture and there was therefore a ready market for them. Small sleeping mats sold in the highest numbers, followed by large sleeping mats, large baskets and large beer strainers. Large baskets and beer strainers were more popular for traditional purposes, such as household use and wedding gifts, while the smaller sizes were more popular with non-traditional markets such as tourists and urban dwellers (personal communication, A Mostert, Manager of the REHAB programme, 2004). At the time of this study, traditional local markets were still of primary importance in these areas, as reflected by the higher sales of larger products. Some crafters focused on making only one or two products, while others made the whole range.

Table 2: Characteristics of products and processes of production

Type of product	Percentage of interviewees making product	Approximate time taken to produce (hours)	Numbers sold per year (mean  ±  SE) by crafters making these specific products	Price of product in rands (mean  ±  SE)	Gross returns per hour in rands (mean  ±  SE)
Large sleeping mats	80	9	19 ± 2	48 ± 9	5.3 ± 1
Small sleeping mats	63	3	21 ± 3	19 ± 4	6.3 ± 1.4
Large food mats	70	3	16 ± 2	14 ± 2	4.7 ± 0.6
Small food mats	35	1.5	14 ± 2	11 ± 2	7.3 ± 1.4
Large baskets	43	6	19 ± 1	22 ± 1	3.7 ± 1
Small baskets	40	3	15 ± 2	13 ± 1	4.3 ± 0.4
Large beer strainers	15	3	19 ± 2	15 ± 1	5 ± 0.3
Small beer strainers	10	1.5	12 ± 1	12 ± 1	8 ± 0.7

In group discussions, the crafters named beer strainers as the most difficult product to make. These were made by only 10–15 per cent of crafters (Table 2). However, those crafters who are able to make them do not face much competition, and are therefore able to sell relatively high numbers of these items, which are popular gifts for young men undergoing initiation into manhood. Gross returns per hour were calculated by dividing the mean price of the product by the approximate time taken to make it (Table 2). Small beer strainers, which are traded mainly in urban markets and to tourists, gave the highest return per hour worked, at R8.00 per hour.

4.2 Socio-economic profile of crafters

The crafters' profiles were very similar for Mpozolo and Ntubeni. All the crafters but one were female; the one male crafter interviewed justified his involvement in women's work, or izinto zabafazi (Kepe, 2003), by saying he was using technology in the form of a weaving frame to manufacture sleeping mats. The crafters' mean age was 56 (Table 3). The majority were heads of households: 70 per cent across both zones, over 80 per cent in Mpozolo and 59 per cent in Ntubeni. Only one had received tertiary education. On average they had not completed primary school. On average, households were made up of five people, with an average of two children, often grandchildren who crafters were supporting. This profile is very similar to that of palm bush vendors elsewhere in the province (Gyan & Shackleton, 2005).

Table 3: Socio-economic characteristics of crafters and reasons cited for entering the trade

Attribute	Measurement	Combined n = 40	Mpozolo n = 18	Ntubeni n = 22
Age	Mean ± SE	55.9 ± 1.8	56.69 ± 2.4	55.39 ± 2.3
Years of education	Mean ± SE	5.89 ± 0.5	6.89 ± 0.7	4.99 ± 0.8
No. of people in household	Mean ± SE	5.49 ± 0.4	5.69 ± 0.7	5.19 ± 0.5
No. of children in household (under 16)	Mean ± SE	2.39 ± 0.3	2.49 ± 0.5	2.29 ± 0.3
Households headed by female crafters	%	70	83	59
Households receiving at least one pension	%	45	50	41
Households receiving at least one child support grant	%	28	39	18
No. of household livelihood sources other than the craft trade	1 (%)	13	22	5
	2 (%)	33	28	36
	3 (%)	28	28	27
	4 (%)	25	22	27
	5 (%)	3	0	5
Most important sources of household income identified by respondents	Pension (%)	40	44	36
	Crafting (%)	25	33	18
	Child support (%)	10	17	5
	Household member's salary (%)	15	6	23
	Other (%)	10	0	18
Reasons for entering the trade	Struggling, household needed money (%)	55	50	59
	Husband left or died (%)	20	34	9
	Extra income (to supplement existing livelihoods) (%)	25	16	32

More of the Mpozolo crafters than the Ntubeni ones had only one livelihood source other than crafting (Table 3). More Ntubeni interviewees had income from employed household members, of whom 9 per cent were employed by Dwesa Nature Reserve, 9 per cent were involved in Working for Water projects in Dwesa Nature Reserve and 9 per cent were domestic workers in coastal cottages. Also, 18 per cent of Ntubeni households were involved in ‘other’ informal livelihood activities, such as selling poles, thatch and bricks; this could be due to their proximity to the coast and to the protected area of Dwesa Nature Reserve, which gives them easier access to resource rich forests. Pensions were ranked as the most important source of household cash income by 40 per cent of interviewees across both study sites. Crafting was ranked as the most important source of income by 33 per cent of crafters at Mpozolo, and by 18 per cent of crafters at Ntubeni (Table 3); overall, 70 per cent of crafters ranked crafting as their second most important source of cash income, showing that crafting is generally a crucial livelihood source for participants in the trade, rather than being merely a casual hobby. The main reason for entering the trade, given by 55 per cent of respondents, was the need for more money; either there was no alternative source of income, or existing sources were insufficient for households to cope. The second most common reason, at 20 per cent, was the loss of a husband, the main breadwinner. For these crafters, the trade acted as a safety net when other livelihood sources were unexpectedly lost.

Crafters can therefore be described as, generally, a highly vulnerable group of people, who have come to rely on income from crafting largely out of desperation. This mirrors the findings of others (Shackleton & Shackleton, 2004; Gyan & Shackleton, 2005), who provide evidence that traders in natural resource based craft are extremely vulnerable, the poorest of the poor, and that the trade is a safety net. However, although the mean returns are low, the returns per household are directly proportional to the amount of time and effort expended (Gyan & Shackleton, 2005), such that for full-time traders the returns are better than local unskilled wage labour rates.

4.3 Contribution of the trade to household livelihoods

Mpozolo crafters were earning significantly more per annum than Ntubeni crafters (Mann–Whitney U = 113.0; P = 0.02) (). Monthly incomes in the busy seasons (June and December) were similar, but Mpozolo crafters earned significantly more than Ntubeni crafters in ‘normal’ months (Mann–Whitney U = 103.5; P = 0.01) (Table 4). One explanation for this difference between the two sites is the presence of the NGO in Mpozolo, which helps the crafters produce unique products (e.g. mats modified into conference bags) to appeal to a wider market, and buys from the crafters throughout the year. This departure from the traditional situation of earning most income from the craft in only two months of the year, as do Ntubeni crafters, makes a substantial difference for Mpozolo crafters.

Table 4: Incomes and costs, hours of labour, returns per hour and contribution to household incomes

	Combined n = 40	Mpozolo n = 18	Ntubeni n = 22	Mann–Whitney U -test results for comparison of the Mpozolo and Ntubeni samples
Net monthly income from craft: normal months (R)	759 ± 11	1209 ± 19	439 ± 8	MW-U = 103.5
	0–245	0–245	0–137	P = 0.01
Net monthly income from craft: busy months (R)	3639 ± 38	3659 ± 44	3619 ± 59	MW-U = 173.5
	50–937	95–775	50–937	P = 0.505
Net annual income from craft (R)	12469 ± 144	16499 ± 241	9169 ± 141	MW-U = 113.0
	7–3757	217–3757	7–2301	P = 0.02
Total annual costs of crafting (not labour) (R)	2009 ± 27	1549 ± 31	2379 ± 40	MW-U = 148.0
	13–823	33–423	13–823	P = 0.17
Costs as a proportion of net annual income from craft (%)	199 ± 3	119 ± 3	269 ± 5	MW-U = 91.0
	1–94	1–36	4–94	P = 0.003
Total annual hours spent harvesting, manufacturing, travelling and trading (hours)	3249 ± 35	4059 ± 57	2579 ± 40	MW-U = 118.0
	24–998	28–998	24–864	P = 0.03
Returns per hour (R)	49 ± 0	49 ± 0	49 ± 1	MW-U = 135.0
	0.13	1.5–10	0.3–13	P = 0.09
Total annual household cash income (R)	918399 ± 1 712	82989 ± 1 250	99079 ± 2 967	MW-U = 176.0
	615–68 887	1287–19 405	615–68 887	P = 0.598
Contribution of net craft income to total annual household income (%)	2699 ± 4	349 ± 8	199 ± 5	MW-U = 125.0
	0.01–100	5–100	0.01–100	P = 0.079
Notes: Means ± standard errors are given, as well as ranges. The Mpozolo and Ntubeni samples were compared, and the Mann–Whitney U-test values are given, as well as P values. Significance at the P < 0.05 level is highlighted in bold type.

The average, net, annual income from craft across both study areas was R1246  ±  144. This compares favourably to the findings of a study conducted in the Bushbuckridge region of South Africa, where the net annual income from woven mat sales was R998  ±  119 (Shackleton, 2005) (these data come from another case study in the suite of case studies funded by SANPAD, looking at locally initiated trade in natural resource-based products in southern Africa). It was found that crafting contributes on average 26  ±  4 per cent of annual household cash income across both areas in this study, making it clear that this trade is very important for crafter households in these villages.

The significant difference in mean total annual costs as a proportion of mean net annual income from the craft (Mann–Whitney U = 91.0, P = 0.003) (Table 4) can be attributed to the higher transport costs incurred by Ntubeni crafters, who have to travel further to the markets in Idutywa, Elliotdale and Willowvale. The mean cost of travelling to markets for Ntubeni crafters was R140  ±  31, and for Mpozolo crafters it was R70  ±  24. Ntubeni crafters did not have the option of selling from their own villages in ‘normal’ months as did Mpozolo crafters because of the presence of the NGO.

There was a statistically significant difference between Mpozolo and Ntubeni crafters in terms of hours spent on the trade, with Mpozolo crafters spending significantly more time overall (Mann–Whitney U = 118.0, P = 0.03). Considering this difference in more detail, the largest difference was in the mean number of hours spent manufacturing, with Mpozolo crafters spending significantly more time manufacturing craft products than Ntubeni crafters. Returns per hour were very similar in the two study areas (Table 4). This indicates that the higher numbers of hours Mpozolo crafters spent manufacturing was not the reason for their higher incomes but rather the result of their receiving more orders.

The interviewees' total annual household cash income was divided into quartiles. The income from the trade contributed significantly more to the total annual cash income of the poorest households – over 40 per cent – than to relatively wealthier households – between 5 per cent and 15 per cent (Kruskal–Wallis-H (3.39)=15.26, P = 0.0016) (Figure 2a). In other words, it is the poorest who rely most heavily on this trade. There was no significant difference in the amount of income earned from the trade across the different income classes (Kruskal–Wallis-H (3.40)=2.41, P = 0.4924) (Figure 2b); however, crafters from the wealthiest income class did generate, on average, higher incomes from crafting than crafters from the poorest income class. The difference might have been statistically significant if the sample size had been larger. A trend that has emerged in other studies of the contribution of natural resources to rural livelihoods, e.g. Angelsen & Wunder (2003), has been that wealthier households receive greater benefits from such activities than poorer households, as they are able to invest more capital in obtaining the raw materials and manufacturing and marketing the products.

Figure 2a: Contribution of crafting for different household income classes in quartiles (in rands).

Figure 2b: Net income from craft for different household income classes in quartiles.

Income from the craft was used primarily to buy groceries, pay for school fees and buy school uniforms. The fact that one of the peak income-generating periods is in December is important for timing, as Christmas is a time when many family members come home and money is spent on festivities. It also coincides with the need for money for school fees and school uniforms at the start of the new school year in January. Using income from crafting to pay for education could be seen as intergenerational poverty alleviation. In addition to the economic benefits and contributions of the trade, there are some important non-economic benefits. These include the cementing of relationships between women that occurs on the highly sociable harvesting trips and during the group crafting sessions, and the maintenance of the tradition of Xhosa culture. Involvement in crafting represents a ‘space’ in which women can experience autonomy and independence from the men who usually dominate when it comes to community decision-making, but who avoid involvement in the crafters' affairs since they see crafting as women's work.

4.4 Perceptions of sustainability of the trade

There were large differences in perceptions of the economic sustainability of the overall trade in Mpozolo and Ntubeni (Table 5), with a significantly higher proportion of Ntubeni respondents referring to an increase in the trade in the last five to ten years (χ² = 13.8; df = 2; P = 0.001). This could be linked to the opening of Dwesa Nature Reserve to reed harvesters, of which many of the women of Ntubeni take advantage.

Table 5: Perceptions of changing trends

	Percentage of crafters
	Combined n = 40	Mpozolo n = 18	Ntubeni n = 22	χ^2 test results for comparison of the Mpozolo and Ntubeni samples
Overall trade in the last 5–10 years∗
Increased	63	29	94	χ^2 =13.8 df = 2 P = 0.001
Decreased	27	50	6
Same	10	21	0
Number of years in the trade
< 10	27	33	23	χ^2 =3.2 df = 5 P = 0.674
10–20	7	5	9
21–30	23	17	27
31–40	10	17	4
41–50	23	22	23
> 50	10	6	14
Trends in individual crafter's income
Increased	45	44	46	χ^2 =0.57 df = 2 P = 0.752
Decreased	25	22	27
Same	30	34	27
Prefer to do something else to generate income?
Yes	72	65	88	χ^2 =1.02 df = 1 P = 0.312
No	28	35	12
Notes: The Mpozolo and Ntubeni samples were compared, and the χ^2 test values are given, as well as the degrees of freedom (df) and P values. Significance at the P < 0.05 level is highlighted in bold type.

Most of the crafters interviewed had been involved in the trade for more than 20 years (Table 5). There is therefore an established history of trade in these products. Many crafters had taught their daughters crafting skills, as they said it was important for them to know how to do this in case they needed money, although they said they would prefer their children to be well educated and employed rather than involved in the craft trade. The trade was seen as an important option to fall back on, but certainly not a first choice as a livelihood source. Over 40 per cent of crafters said their income from the trade had increased since they had started trading, and most said this was because they were putting more effort into crafting now because they needed the money more desperately. Most responded that they would prefer to be doing something else to generate an income, suggesting that crafting is a last resort.

With regard to environmental sustainability, most crafters believed that resource availability had increased or remained the same over the last five to ten years (Figure 3), and 90 per cent of interviewees believed there was enough C. textilis or J. kraussii for everyone who wanted to harvest.

Figure 3: Perceptions of changes in resource availability in Mpozolo and Ntubeni

The high number of Mpozolo crafters perceiving an increase in resource availability (Figure 3) could be due to the increased cultivation of C. textilis in gardens. There are no quantitative data on this pattern from this study, but in the group discussions respondents said that more people were now growing C. textilis in their gardens. C. textilis can be grown only in gardens which border on rivers or wetlands; the plant has also been grown alongside the streams and rivers in communal areas, and access to this stock is controlled by the people who originally planted it there. According to Mathe (2001), C. textilis has been domesticated not only because of the popularity of the products manufactured from the plant, but also because the plant takes root easily and requires minimal attention or horticultural input: ‘the more we cut it, the more it grows’. The motivation to domesticate this plant has been found to be related to perceptions of scarcity (Mathe, 2001). Local people are allowed to harvest from Dwesa Nature Reserve free of charge. This is a valuable opportunity for the villagers of Ntubeni, but for crafters from Mpozolo the time, distance and physical exertion associated with harvesting from Dwesa means that this is not a popular option. Only those Mpozolo crafters who do not have C. textilis gardens, or are too poor to buy reeds from other villagers, harvest from Dwesa Nature Reserve occasionally. The reasons given for the increase in C. textilis cultivation in Mpozolo were that ‘there are more things to do with imizi now’, more people buying, and local people relying more heavily on crafting. There is a commonly held perception that there is plenty of C. textilis and J. kraussii inside the Dwesa Nature Reserve, and that even if the resource were to become scarce in the villages themselves, crafters would have the option of harvesting from Dwesa to fall back on.

4.5 Opportunities for and constraints on the trade

The most frequently cited problem with trading in both areas was that local people buy on credit and do not pay back the crafters – this was mentioned by 30 per cent of interviewees. Crafters reported that when this happens there is nothing they can do to force the issue – ‘just pray that they will pay one day’. The second most commonly cited problem is that customers haggle for lower prices, and that if they do not lower their prices the customers will just not buy from them. Trading from home, as many of these crafters do, to the people nearby, can be problematic; so much hinges on one's reputation in the village and on maintaining good relationships that failure to pay will often be overlooked. Crafters are not able to insist on set prices or to refuse to give products on credit. The third most commonly cited problem was the fear of criminals in towns stealing money and products.

The seasonality of the trade is another factor limiting the economic returns from crafting; crafters who rely on local markets are restricted to selling in two peak months, June and December, when it is traditional to buy craftwork to give as gifts. Crafters who have access to wider markets throughout the year are able to generate consistently higher incomes (Figure 4). Almost 30 per cent of Mpozolo crafters were able to generate more than R200 in ‘normal’ months, whereas over 60 per cent of Ntubeni crafters made less than R50 in ‘normal’ months.

Figure 4: Net monthly cash income from crafting in Mpozolo and Ntubeni in ‘normal’ months.

While the amount of resource was not a constraining factor for the crafters in this area, access to it was a problem for some. The cultivation of C. textilis in home gardens requires that the gardens border on streams or wetlands; therefore this was not an option open to all. The cultivated C. textilis stands in communal wetland areas were also not for communal use, and in the group discussions it emerged that, although there was an abundance of the resource growing in the village, many crafters felt uncomfortable about asking for permission to harvest from the people who had originally planted in these areas.

In Mpozolo, where the REHAB project has been training and buying from crafters, crafters have expressed excitement about and renewed enthusiasm for crafting, saying ‘We already knew the skills but REHAB put us upwards’ and ‘We are more experienced than before’. The Mpozolo crafters have taken the initiative of linking with the Mbashe Crafter's Association, and are reportedly receiving more orders than before. It therefore seems that business training, skills training and the forming of crafter cooperatives represent opportunities for increasing income from the trade.

It is important to note that the higher net annual incomes from the craft for Mpozolo crafters – relative to Ntubeni crafters (Table 4) – were recorded in the year in which REHAB was most intensively involved with the Mpozolo crafters, and was buying from them on a regular basis. This greater success may be because Mpozolo crafters had access to a new, non-traditional market, buying products consistently and from within their own home villages. This situation solved all the problems mentioned above, but unfortunately it was a short-term intervention. It could, however, be used as a model for providing opportunities to local craft traders.

5. SUMMARY AND CONCLUSIONS

Trade in natural resource-based products is of particular importance to the poorest of the poor (Angelsen & Wunder, 2003), and trade in traditional craft can generate income for highly vulnerable rural women. Such trade is therefore of magnified importance, since it is a safety net and helps ameliorate poverty (Shackleton & Shackleton, 2004). This study provides empirical evidence of this, demonstrating that the people involved in the reed-based craft trade in the villages of Mpozolo and Ntubeni are vulnerable and heavily reliant on income from the craft. It also shows that income from the trade in natural resource-based products can be high relative to other sources of income, something not shown by previous resource-specific market chain analyses. Furthermore, it shows clearly that this kind of trade is of the highest relative importance to the poorest households, corroborating the results of Shackleton (2005).

The main factor limiting the success of craft traders in this area is the lack of access to markets. The conclusion drawn by research into rural handicraft projects in South Africa, especially KwaZulu Natal, during the 1980s (Rogerson, 2000) was also that, assuming ecological sustainability, access to markets is the primary factor determining trader success. What is needed, therefore, is the opening-up of markets for rural crafters, beyond the confines of local trade. Local trade is important for easing the hardships of poverty (Shackleton & Shackleton, 2004) and for maintaining tradition in terms of the cultural use of these products, but it should be combined with trade in wider markets. If external markets can be brought to the source of the trade, transport costs for rural crafters can be reduced and the risk of local producers being exploited by ‘middlemen’ can be avoided. In the case of the rural crafters of Mpozolo and Ntubeni an opportunity for tapping into the tourist market exists, owing to the proximity of the Dwesa Nature Reserve. At present, visitors to the Dwesa Nature Reserve drive past these communities without being exposed to the prevalent craft trade; if a stall selling locally produced craft were set up in the Reserve, the crafters' incomes might be improved. In addition to this, the continued support of NGOs such as REHAB, buying local craft and circulating it to wider markets, could provide valuable benefits.

A secondary limiting factor is the fact that women throughout the Transkei make very similar, almost identical, products. Manufacturing and marketing unique and innovative products which will appeal to urban and tourist markets would increase crafters' income. Traditional items are culturally important, and will continue to be made for functional use in the homes of rural people and sold to urban dwellers who want to maintain links to their rural customs. However, the difference between the Mpozolo crafters and the Ntubeni crafters is that the Mpozolo crafters were selling more in the off-season, and the lesson to be learnt from this is that non-traditional markets need to be accessed. Therefore skills development and training to encourage the production of distinctive craftwork should be facilitated. The formation of crafter cooperatives, which could improve the accessibility of C. textilis grown in communal areas and promote the sharing of skills, could assist in targeting new and wider markets.

When thinking about promoting and supporting local craft trade it is important to consider that there are a number of factors that provide ‘poor producers’ with an advantageous competitive edge over other producers (Scherr et al., 2003). If these are identified and action is directed at enhancing these factors, the low-income local crafters' chances of success will be greatly improved. Some examples of the kinds of advantages held by rural traders are: indigenous knowledge about the resource and about local markets, superior monitoring and protective control over the resource, as the managers are local people living close to the resources and with their personal interests at stake, and access to specialist ‘socially/environmentally responsible’ target markets (Scherr et al., 2003). For both rural craft economies and natural resource based trade the chances of poverty elimination are extremely low, but the contribution is high enough to justify recognition and support for these activities. With some intervention, greater economic benefits could become a reality.

Additional information

Notes on contributors

Sheona Shackleton

Respectively, Researcher; Professor and Head; and Research Associate, Department of Environmental Science, Rhodes University, Grahamstown, South Africa. The authors wish to thank Annegret Mostert and the staff of REHAB for their help and for the important work they do; the crafters of Mpozolo and Ntubeni for sharing their experiences; Dylan McGarry, Loyiso Fiyane, Eliecer Diaz, Brett Scott and Monique Pelser for help with the field work; and the South Africa–Netherlands Research Programme on Alternatives in Development (SANPAD) for providing the funding for this study.

Notes