Market development and investment “bottlenecks” in the fisheries of Lake Kariba, Zambia - Ragnhild Overå

1. INTRODUCTION

When Lake Kariba was created through the building of a gigantic dam across the Zambezi river in 1958 for the production of hydro-electric power, the fishery that evolved in the new and large water body was anticipated to become a major benefit for the Tonga people in the area. The relocation of 36 000 Tonga that became necessary as a result of the inundation of the Gwembe valley created many problems.^[78] Nevertheless, a thriving fishery did develop, which created new livelihoods for a large number of both Tonga and migrant fishermen and fish traders on the Zambian side of Lake Kariba.

Four decades later, fish marketing continues to be in the hands of small-scale traders and fishing is conducted in unmotorised dugout canoes or “banana-boats”. The fishermen constantly adapt to the environment and improve their fishing methods according to their acquired experience. However, apart from a shift to nylon nets and smaller mesh sizes, very few technological or institutional changes resulting in more efficient and profitable fish production have taken place since the creation of the lake. Not even in the semi-industrial Kapenta fishery - opened in the early 1980s and largely owned by white businessmen - have investments and technological growth “taken off”. In light of recent research on the biology and ecology of Lake Kariba where no serious depletion of fishery resources is indicated (Kolding, Musando and Songore, 2003), this lack of investment may seem puzzling: given the relatively favourable resource situation in Lake Kariba one would, at least according to conventional common property theory, expect individuals to maximise profit through investment in technology and more efficient organization of production.

In order to understand the causes behind this apparent “lack of development”, this paper situates the fishery of Lake Kariba historically, socially and economically (i.e. in the context of its market), and thereby discusses the “bottlenecks” (Brox, 1990) that impede investments, institutional development, and technological growth in this fishery. As it will appear, a capital extensive and technologically simple fishery is a necessary and rational adaptation to the type of context, or “imperfect market”, in which Lake Kariba is located. Under the prevailing Zambian economic circumstances, over-capitalization of this fishery is not a threat. It is thus argued that rather than focussing on the regulation of fishing methods, public funding of fisheries management would be better spent on facilitating and securing a fair access to the lake’s resources.

1.1 Fisheries investment and contracts in an uncertain environment

In most African fisheries, fishing effort is characterised by often dramatic fluctuations over time. As Jul-Larsen et al. (2003) make clear, fluctuations in the fisheries of Lake Kariba have mainly been observed in what Ottar Brox (1990) calls horizontal changes in effort. This relates to changes in such factors as the number of fishermen, the number of nets, and the frequency and spatial extension of fishing trips. In Lake Kariba population-driven changes, as we prefer to call them, have largely been related to migration processes and developments over time in other economic sectors than fishing. Furthermore, Brox distinguishes population-driven change in effort from changes in factors such as fishing technology, capital investments, labour organization and concentration of gear units per owner. Such vertical (Brox, 1990), or investment-driven, changes in effort have hardly occurred in Lake Kariba fisheries. Though some capital-intensive enterprises have been observed, they are very few and they have not lasted long. It thus remains to be explained why effort dynamics in Lake Kariba continues to be largely population-driven.

FIGURE 1. Effort development: Number of fishermen and nets in the Zambian inshore fisheries of Lake Kariba 1960-1999. Source: Jul-Larsen et al. 2003.

When investment-driven growth has been observed in small-scale fisheries, an influx of investment capital from beyond the fishing units themselves has often proved necessary. Where credit from banks or other official sources are not available for fishermen, access to credit from actors in the fish market is therefore crucial. Traders (with the aim of increasing their fish supply) provide fishermen with credit in order to enable them to invest in more capital intensive gear. New organizational forms (i.e. ownership of fishing boats by traders) and new institutions (i.e. credit and price agreements) may evolve and result in increasing productivity of the fishery. Historically, such “endogenous dynamics” (Chauveau and Samba, 1989) have been observed in many small-scale fisheries in developing countries. In most cases the degree to which traders and fishermen are able to enter into binding and long-lasting contracts is crucial for such investment-driven growth to take place.

According to new institutional economics, the development of credit institutions and supply contracts in fisheries are aimed towards reducing the transaction costs in a situation characterised by market imperfections (inadequate flow of information, lack of infrastructure and access to resources and markets). Jean-Philippe Platteau (1989), for example, sees interlocking credit-supply contracts as an outcome of uncertainty. In order to overcome the obstacles created by market imperfections institutions that “link up credit with marketing relations by offering loans to owners of fishing vessels on the security of future catches” (Platteau and Abraham, 1987:480) will develop. This enables the actors to spread risk and to ensure access to and control over fish supply, capital and - importantly - labour. As Platteau (1989) notices, the institutionalization of such arrangements have been observed in fisheries world-wide, for example through the institution of marriage between fishermen and traders, as in many West African fisheries, or through long-standing relationships of trust, as observed between Malay fishermen and Chinese traders.

In the canoe fisheries of Ghana for example, women traders found a new investment object when the outboard motor was introduced in the early 1960s (see, Vercruijsse, 1984; Hernæs 1991; Overå, 1998). By extending considerable investment loans to fishermen for the purchase of outboard motors and new types of nets and purse seines, the traders were able to increase their fish supply. Some of the women also became canoe owners themselves, hiring men to fish for them. External aid in this innovation process was negligible: it happened from “within” the fishing communities. As a result, the degree of motorization of the Ghanaian canoe fleet increased from zero in the late 1950s to 20-25 percent in 1970 and to 57 percent in 1990. The annual fish landings increased from 20 000 tonnes in 1960 to 300 000 tonnes in 1992. This represents an increase in the productivity from 0.6 tonnes/fisherman/year to 1.4 tonnes/fisherman/year (Degnbol, 1992:215). Important preconditions for this investment-driven growth were (a) institutionalized credit links between fish producers and fish distributors through marriage and kinship, (b) a local gender ideology that allowed for female traders’ investment and control of male labour, and (c) the constant adaptation of culturally embedded local institutions to changing situations (male fishery leaders and female market leaders), through which credit and labour relations could be negotiated and sanctioned (Overå, 2001). Platteau’s assumption that such “traditional” institutions are more efficient than “modern” institutions in a market characterized by numerous imperfections thus seems well suited to explain the Ghanaian case.^[79] The question is, however, whether this approach is useful in explaining why a similar growth process, aided by the evolvement of mechanisms to reduce risk and enforce contracts, does not happen in the Zambian case of Lake Kariba.

The Zambian fish market must be said to be characterized by many “imperfections” that inhibit a free flow of goods, information and capital. The marketing of fish from Lake Kariba is thus neither particularly efficient, nor is the market well integrated with the fisheries through investments. Despite Lake Kariba’s location only three to five hours drive from the capital Lusaka where fish is in high demand, the efficiency and profitability of fish trade is highly variable. Likewise, investment by fish traders in boats and fishing gear in order to enhance fish supply, is limited: it is too risky, they say. Indeed, fish traders, Kapenta operators and other potential investors, such as local shop-owners or Lusaka-based fish distribution companies, constantly attempt to enter into contractual relationships with fishermen in order to secure a stable and reliable supply of fish. However, apart from a few exceptional cases, credit relationships and supply contracts are very unstable and tend to fail.

As this situation indicates, it seems difficult to establish institutions - broadly defined as “informal constraints (...) and formal rules (...) devised by human beings to create order and reduce uncertainty in exchange” (North, 1995:97) - that are legitimate among the various stakeholders on the shores of Lake Kariba. The few examples that exist of credit institutions or other long-lasting binding commitments between fishermen and traders have not been of such a magnitude that they have resulted in any significant technological or organizational changes in the fishery at an aggregate level. What is it, then, about this particular context that prevents mutually beneficial contracts and an investment-driven growth in effort from developing?

Sara Berry has pointed out that historically, conditions for the evolvement of “constellations of social interactions, in which people move, acquire and exchange ideas and resources, and negotiate or contest the terms of production, authority and obligation” (Berry, 1997:1228), have been unfavourable in Africa. In the case of Zambia and Lake Kariba communities, one can mention quite a few factors that have hampered the creation of what Bierschenk and Olivier de Sardan (1997) call arenas in which stakeholders could come to some sort of mutual agreement about rules for cooperation and for the sanctioning of non-cooperative behaviour. Among these factors are: a lack of “traditional” institutions before the creation of Lake Kariba with rules for the economic organization of commercial fishing and trade; the initial displacement of the Tonga people that created political tension and constrained peoples’ livelihood alternatives; subsequent macro-economic and demographic changes that resulted in a “fluid” situation with migration in and out of the fishery; land conflicts along the lake shore; destruction of infrastructure during the Zimbabwean liberation war; fluctuating lake levels and fish catches due to climatic variations; several shifts in administrative regimes and unclear authority structures at the local level. Thus, even if the relationship between fishermen and traders at Lake Kariba certainly is characterised by a recognition of the mutual benefit they have of each other, political and economic instability in the region often inhibit the development of enforceable contracts. When profit is made in fishing, it thus tends - as a perfectly rational strategy under the prevailing conditions - to be invested in a number of security networks (often with conflicting norms, rules and priorities) rather than in more efficient and capital intensive fishing gear.

Much research has been done on the importance of trust for the success of entrepreneurs and traders (i.e. Evers and Schrader, 1994). The Chinese business diaspora and Hausa trade networks are some of the classical examples. In Southern Africa, however, the development of long-term economic relations where trust goes beyond the short-term personal level has been constrained. In contrast to the durable networks of the famous Asian and West African cases, business networks in this part of Africa have been characterized (by the anthropologist Clyde Mitchell) as “instrumentally-activated personal networks” (MacGaffey and Bazenguissa-Ganga, 2000:12). The Zambian economy has thus in many ways come to resemble what Fafchamps and Minton (2001) call a “flea market economy”: high risks of theft and embezzlement and breakdown of legal sanctions make traders refrain from entering into contracts and most transactions therefore take a “cash-and-carry form” (ibid:230).

With the high mobility out of formal and into informal occupations (among them fishing), Zambians increasingly find themselves in a situation where they are “faced with the need to build economic relations from scratch in a world lacking both orderly state regulation and the segmentary political structure of their customary society” (Hart, 1988:178). In such a situation the possibility of building durable economic relations based on kinship (ascribed status) is limited, and so is the possibility of relying on contracts legally sanctioned by the state or civil society. There thus remains, as Hart puts it, “the zone of free-floating social relationships formed by the expectation of mutuality” (ibid.), a zone of association and friendship where trust plays a prominent role. Whereas kinship and contract offer a durable model for hierarchy and control (parental and legal sanctions respectively), trust is based on the negotiation of risk occasioned by the freedom of others. Trust is thus central to social life when neither traditional nor modern probabilities hold, but does not hold as a basis for industrial production and division of labour (ibid:191). Sanctions imposed by kinship and formal or informal legal contracts are in other words a “more durable basis for society” (ibid.), and for economic growth beyond individual enrichment.

The case of the fisheries of Lake Kariba sheds light on some of the constraints that inhibit the development of institutions (that require more than fragile economic relations based on trust and mutual personal interest) in an imperfect market. Clearly, one cannot take for granted (as institutional economics tends to do) that common norms and legal sanctions that would reduce transaction costs and thus would reduce the risks of investment in the fisheries, necessarily and “naturally” will evolve in all types of contexts. The case also illustrates the degree to which investment “bottlenecks” at the local level are intrinsically linked to macro-economic and political processes.

1.2 Methodology

Methodologically, the focus of this study is on the multi-local nature of the activities of fish traders, through which they connect the Lake Kariba fishery and communities to the national and urban economic context. The study is based on secondary sources like the anthropological studies of the Gwembe Tonga in the 1950s and 1960s that documented the consequences of the construction of Lake Kariba (see Scudder, 1960, 1972, 1985; Colson, 1960, 1962, 1971; Colson and Scudder, 1975, 1988). Furthermore, information has been gathered from evaluation reports of development projects in the area (see Brandt et al., 1973; Scudder, Colson and Scudder, 1982; Walter, 1988; Jul-Larsen et al., 1997), from reports by the Zambian ministries (Beatty, 1969; Chipungu, 1988), and in the National Archives of Zambia. In addition to information from the general literature, qualitative data have been collected during three periods of fieldworks in Lake Kariba communities and in Lusaka fish markets in 1998 and 1999 (four months in total). Interviews were conducted with fishermen, traders, transporters, fish traders’ organizations, Village Management Committees, Women’s Clubs, shop owners, Kapenta operators, chiefs, and with the staff of the Department of Fisheries, of the Ministry of Agriculture, Food and Fisheries, and the Councils in Sinazongwe and Lusaka.

The study provides a chronological description of how a marketing system based on fish from Lake Kariba developed in the 1960s leading up to a contemporary analysis of the socio-economic organization of fish marketing in the late 1990s, and of the major investment constraints.

2. AMAN-MADE LAKE - ANEW MARKETING SYSTEM

2.1 Trade and infrastructure in the Gwembe valley before the creation of Lake Kariba

Before the creation of Lake Kariba, the Gwembe valley with the Zambezi flowing through it was bypassed by most trade routes. For a short period at the turn of the twentieth century, Gwembe lay across the direct route (through Kalomo) between the European centres on the high plateau north and south of the valley: a few shops, missionary stations, schools and administrative posts of the British South Africa Company, which took power in Northern Rhodesia in 1898, were established (Colson, 1971:15). However, in 1906 the railway was built on the plateau, and trade and traffic bypassed the valley again (ibid.). Trade stores and missionary stations gradually disappeared in the 1920s. After the 1930s the valley had no white residents and few strangers visited the Gwembe District other than occasional traders who came to collect tobacco (Colson and Scudder, 1975:194). The tobacco trade was based on long-term bond friendships between the urban-based traders and the Gwembe valley producers (Colson, 1960:47).

The Valley Tonga did of course not live in complete isolation. From 1900 many men (hardly any women) participated in an extensive labour migration system which provided most of the cash income that supplemented the sales of tobacco and subsistence grain production (Colson and Scudder, 1975:194). One of the reasons for Tonga men to migrate was the need for cash when the “hut tax” was introduced in 1904. In 1956 as many as 42 percent of the District’s able-bodied men were absent as wage workers; mostly in the mines in Southern Rhodesia, especially Bulawayo, and in the Copperbelt in Northern Rhodesia. Thus the majority of the older men had at one time or the other travelled long distances, often walking on foot, in connection with labour migration (ibid.). Nevertheless, until 1950 there was no road in Gwembe (Colson, 1971:17). When colonial administrative officers toured the valley, they were carried by men who reluctantly were mobilized in the villages by appointed headmen. The first four cars owned by Gwembe Tonga were bought in 1956-57 (ibid: 18).

With the influx of money from migrant labourers and the building of roads, the establishment of trading stores followed, in addition to the emergence of “an increasing throng of itinerant traders” in the late 1950s (Colson, 1971:17). Fish trade, however, was not carried out on any significant scale. The seasonal surplus from fishing in the Zambezi was either bartered or sold locally (Scudder, 1960:42). As Colson (1960) put it: “a little dried fish was traded to the Plateau Tonga, but the Valley could not compete with the commercial fisheries of the Kafue and Luapula, handicapped as it was by its isolation” (ibid: 202).

With regards to infrastructure and commercialization, the above description provides a glimpse of the situation when the Kariba dam was sealed in 1958. Within five years a surface area of approximately 2 400 square kilometres would be covered by water. The dramatic and difficult relocation process of the Tonga has been described in detail in the above mentioned sources, and will not be discussed here. The main point is to say that when the new fishery in Lake Kariba emerged, there was neither a well developed infrastructure in the valley, nor a large body of Tonga traders with experience, interest, relevant knowledge or other resources, who stood ready to take care of the distribution of fish from within the new Tonga fishing communities.

2.2 Bumper catches and urban traders in the early years (1958-1962)

Until 1963 the Lake Kariba fisheries were reserved for the Tonga people, and its management was in principle under the control of the Gwembe Tonga Native Authority. The Northern Rhodesian government initiated training programmes. Bemba^[80] and Lozi fishermen with experience from other Zambian lakes were appointed to share with the Tonga their fishing skills, and credit for fishing equipment was supplied to encourage local men to turn to fishing (Colson and Scudder, 1988:29). Due to the abundance of nutrients in the water during the first years after inundation, fish stocks multiplied quickly and the catches were very good. The government provided fish markets along the lake, i.e. cemented structures with roofing and scales for the weighing of fish. Around these markets fishing camps were established, which became centres of economic activity and social change. During the first boom years when the lake was filling up, the number of young Tonga men joining the fishery reached 2 500. The Tonga fishermen made nice returns in cash from their sales of fish, and many women from the surrounding farming villages became incorporated in the market economy through the brewing of beer for sale in the fishing camps. Even the District Council responded to people’s new availability of cash by opening taverns and bottle stores near the fisheries: “In the early 1960s the camps were noisy with drumming and people making merry over beer” (ibid: 30). The income of the men from fish sales and the income of the women from beer and food sales (which indirectly gave women a share of the fishery profits), was mainly invested in cattle (oxen for ploughing), cotton production, bicycles, small grocery stores, and education of children. Obviously, the source of cash and other goods that led to these new patterns of consumption were the fish traders.

MAP 1. Lake Kariba fishing camps mentioned in the text and their main fish marketing region along the line of rail and Lusaka

Whereas the local population received training in fishing, no incentives through training or credit programmes were initiated to teach local people the skills of fish marketing. Moreover, “the Tonga were not traders by heart” (Colson, 1962:602). Thus, even though fishing was reserved for the Tonga, very few Tonga participated in the new fish marketing system. Fishermen felt that processing and marketing involved too much time and labour compared with disposing of the fish for cash and quickly go fishing again. Another important factor was the reluctance of most Tonga men to let their wives leave the immediate vicinity to be involved in economic activities and to be exposed to town life. Thus, since the fishermen were either too young to be married or, if they were married, did not allow their wives to become fish traders, they depended on distribution by fish traders from outside the Tonga communities. So who were these fish traders?

It does not appear that relocated Tonga in the resettlement areas entered fish trade (Colson and Scudder, 1975). Since very few of the fishermen or other Valley Tonga traded in fish in the initial years, it is likely that most fish traders who came to Lake Kariba lived in urban areas. Indeed, they were welcome: “Traders with vehicles were coming into the fishery from Lusaka and the Native Authority did not put any barriers to the export of fish”.^[81] It is also likely that these traders predominantly were male. In Nyirenda’s (1957) study based on a survey in the Lusaka markets conducted in 1954 it appears that only one fifth (18 percent) of all traders were women, and that almost all of the traders who dealt in fish were men. Three of the 122 fish traders in the survey were female, but these were wives who retailed the fish that their husbands brought to the market. In a follow-up of Nyirenda’s Lusaka market study, Miracle (1962) noted that the female proportion of traders had increased to almost one-third (29 percent). A major reason for the male dominance among urban traders was the colonial control on urban residence. Africans were mainly to reside in cities as migrant labourers, and it was assumed that the workers maintained families in the rural areas where they came from and that they would return to these upon retirement (Ferguson, 1999). However, restrictions on female residence in urban areas were never completely efficient, and increasingly the government provided housing schemes that included the worker’s family. Correspondingly, the proportion of economically active women in the urban areas, including female fish traders, increased from the 1960s onwards (Hansen, 1997).

None of the fish traders in Nyirenda’s Lusaka market survey were Valley Tonga. The Lusaka fish traders mainly came from the Northern and Western regions. It is thus possible that some traders with experience from the Copperbelt fish markets and from fish trade in Lake Mweru, Lake Bangwelu, the Upper Zambezi, and so on, tried out the new fisheries in Lake Kariba after having contact with the fishermen from these regions, who were appointed to teach the Tonga fishing.^[82] Fish traders in Lusaka were considered as occupying a pre-eminent position compared with traders in other food items, and fish was the only trade “profitable enough to compete with the other occupations educated men may follow” (Nyirenda, 1957:44). Considering the high proportion of fish traders with an education higher than Standard III (43 percent of fish traders as compared to 33 percent of all traders), Nyirenda assumed that these traders were hoping to “develop a large business rather than just to pass time” (ibid.). Nevertheless, almost all of them were small-scale middlemen, travelling themselves to the various fisheries: buying, processing, transporting and retailing fish in the market.

During the years of inundation and high productivity in the lake, more fish was exported fresh on ice than in a dried or smoked form from Lake Kariba (Colson and Scudder, 1975). In 1963, for example, 63 percent of all fish was marketed in a fresh condition. Since only a small proportion of the fish in the Lusaka market places was sold fresh (Miracle, 1962), this means that much of the fresh fish from Lake Kariba reached a “luxury market”: the urban white population and hotels and restaurants. It is not clear where all the ice to chill and preserve the fish with came from, but at least a large proportion must have come from the ice plant put up by a commercial company owned by white European businessmen in Sinazongwe. In fact, at one time Kafue fish traders travelled all the way to Sinazongwe to buy ice to put on the fish they bought in Namwala (Lower Kafue) to sell along the line of rail and in Lusaka (Beatty, 1969:23).

The European company in Sinazongwe had obtained rights to purchase fish from the whole Lake in 1963 (Malasha, 2003). The company also had plans of extending their operations beyond the production of ice and purchasing of fish from the Tonga fishermen; they also applied for fishing concessions. This was, however, effectively prevented by both the Gwembe Tonga Native Authority and the Northern Rhodesia African Fishermen and Fish Traders Association. The traders and fishermen were in favour of the company’s production of ice (which the traders needed) but threatened to boycott the company (by refusing to sell fish to them or to buy ice from them) if it was given any right to participate in the fishery. Rights to the fishery was a “hot potato” in the independence struggle, and the fish traders association was known to be connected with the African National Congress (ANC).^[83] The Northern Rhodesian Industrial Development Corporation, which had provided the company in Sinazongwe with a loan, therefore complained: “The system of marketing Lake Kariba fish is most inefficient in common with all other fisheries in the territory. This is mainly due to the fact that the entire control of the industry lies in the hands of the fish traders who use African Nationalist politics to obtain this control”.^[84]

The pre-independence political climate in Northern Rhodesia thus made an industrial investment (by Europeans) in the fisheries impossible. Furthermore, the high costs involved in ice production powered with a diesel aggregate and the problems that the company had with the fishing concession, coincided with a biological decline in the fisheries from 1963 onwards (to which we shall return below). An industrial venture where capital was planned to be pooled into motorized fishing gear, ice production and efficient transportation and marketing therefore never happened. Politically it was unacceptable, and with declining fish catches it also proved to be unprofitable. In 1964 the company pulled out of the area and moved their ice plant to the Kafue River (Malasha, 2003).

Fish marketing thus remained in the hands of the urban fish traders. Very little information is available on the extent to which they invested in the fisheries through credit-arrangements or directly in gear ownership. Beatty (1969:75) indicates that fishermen often tried to steer clear of entering into credit arrangements with the traders to avoid them determining an unfavourable fish price. However, with the declining profit margins on which the majority of fish traders operated (Obershall, 1972:120), it may have been a very small minority of the fish traders who had the financial capacity to extend long-term credit of any significance at all. The general impression from Colson and Scudder’s works is that the only credit-arrangement practised was the provision by fish traders of necessities on credit: mealie-meal (maize meal) and fishing nets were bartered with fish. This could probably work out well in the fishing camps around the fish markets. There, exchange between traders and fishermen was supervised by fisheries officers who were weighing the fish and calculating the price according to the gazetted fish price. Fish traders therefore had some security if they extended credit to fishermen. But any massive investment in gear by traders is not reported. Investments by the Tonga fishermen themselves tended to go in the direction of farming equipment and cattle (which represented their most valuable form of saving and wealth; lubono), rather than into more technologically sophisticated or capital intensive fishing gear.

2.3 Dispersion, migration, and new trading patterns (1963-1975)

Towards the end of 1962, Lake Kariba had reached its maximum level and the biological productivity declined drastically. This was not related to overfishing, but to natural processes in the “making” of a lake: nutrition levels in the lake sank, the catches and the size of fish decreased, whereas species diversity increased (Musando, 1996). Already by 1964 the catches had declined with 50 per cent, and the number of fishermen dropped from 2 500 in 1963 to 500 in 1969 (Walter, 1988:34). When fish catches in the vicinity of the fishing camps went down, most of the Tonga fishermen returned to farming; many of them producing cash crops (cotton and maize) (Colson, 1971:149). One important reason for a revived interest in farming was also the eradication of tsetse flies because all game had been shot in preparation for inundation of the Gwembe Valley (on the Zambian side) (Scudder, 1972). The end of sleeping sickness in combination with investment of profits from fisheries, led to a shift in the agricultural production system towards ox plowing.

Around the same time, in 1963, the ban on fishing by other ethnic groups than the Tonga was lifted. Whereas natural resources had been governed by Native Authorities (though under the indirect rule of the colonial government), rights to natural resources became centralized under the new government after independence in 1964, and should thus in principle be accessible to all Zambians. Native Authorities were replaced by Rural Councils. The fishery was now open to anyone, and predominantly Bemba men from the Northern Province started to move into the valley to start fishing. Some of these may have been fish traders who were familiar with Lake Kariba and who saw an opportunity in becoming fishermen when the ban on fishing by non-Tongas was lifted. Others were the appointed Bemba and Lozi “fisher trainers” and colleagues and relatives of theirs from their home regions.^[85] Since most Tonga were leaving the fishery in favour of farming and labour migration, the new non-Tonga entrants in the fishery were not viewed as competitors. Apparently, although little information exists on this relationship, Tonga “traditional” political leaders (chiefs and headmen) did not resist the entry of “foreigners” into the fisheries.

Demographic and biological changes led to a reorganization of the fishery. First of all, as a result of the lower productivity of the lake, the fishermen had to chase fish where it could be found. This meant that they had to adopt a much more dispersed fishing pattern. Since the migrants neither had rights to land, nor a close attachment to the land in the Gwembe valley as a livelihood and way of life (as the Tonga had), they probably felt more inclined to migrate freely in search of good fishing grounds. The fishermen thus frequently changed fishing locations while their families stayed for prolonged periods on the numerous islands in the lake. Sometimes a fishing camp on an island could consist of only one fisherman and his family.

The spatial dispersion of fishing, the declining catches and the diversification of the fishermen’s ethnic identity had far-reaching consequences for fish marketing. It became very difficult for traders to reach the fishermen. Only the harbours of Sinazongwe and Siavonga were accessible by road the whole year. Consequently, a shift from a predominance of trade in fresh fish to a reliance on drying and smoking as the main processing methods, occurred.^[86] It was simply impossible to get ice to and from the fishing camps quickly enough, and even many of the lakeshore camps were not accessible at all with vehicles. The transportation problems led to a decrease in the number of traders who frequented the fisheries of Lake Kariba, and the fishermen increasingly had to process and transport their fish to the market themselves. In the migrant fishing households, some integration of production and distribution through marriage and kinship therefore became common. Their trading networks often extended to relatives living in town, with whom they stayed while selling the dried fish, or in whose hands they would leave the market retailing of fish altogether.

During the late 1960s and early 1970s, more of the fish from Lake Kariba was marketed in the Copperbelt than in the early years. This can be attributed to (a) the longer durability of dried fish compared with fresh fish and (b) the origination of many of the “foreign” fishermen in the Copperbelt where they had relatives living in the mining towns who were important co-operation partners in the marketing of fish, and (c) the high demand of fish from the copper mine workers.

Some of the Tonga who remained fishermen also started processing and marketing their fish themselves, and some of the women started trading. However, even in cases where Tonga husbands allowed their wives to market the fish, the women often felt that the incentives for going on a strenuous trading trip were small. In many Tonga households the wife had to account for all the profit she made on a trading trip for the husband, and wives generally had very little control over what the money should be spent on. Hence, most Tonga fishing households continued to rely on traders, and the majority of traders were still male (Obershall, 1972). Often fishermen developed bond-friendships with dried-fish traders who spent as long as a month in the fishing camp buying and drying fish, during which time he would often be associated with one particular household. Traders finding their way to the far-away fishing camps portrayed themselves as “orphans” who were given shelter by the fisher families, rather than as potential absentee owners of fishing gear.

The role of the traders in bringing goods from town into the fishing camps was important. The need to purchase goods was thus as important as selling fish when fishermen made time-consuming journeys to towns. But there was also another reason why fishermen started to bring their own fish to the market: the government had introduced a price policy whereby the official producer price too was very low compared with fish prices in urban markets (Brandt et al., 1973:17). An organization of fish marketing like the National Marketing Board (NAMBOARD), a co-operative distribution system introduced to farmers, was thus impossible as long as the official producer price for fish remained unacceptably low (ibid.). In fact, deteriorating rural-urban terms of trade in general during the so-called “good” years of the urban copper economy (1964-73) meant that prices of agricultural goods declined by nearly 54 percent relative to the prices for the urban processed goods that villagers wanted to purchase (Colson and Scudder, 1988:125). Hence, fishermen either retailed what they produced themselves, which was time consuming and left them with less time for fishing (some of them sending their wives, which meant that they could fish more), or they relied on traders who were willing to pay a higher price than the official one. Another consequence of the price regulations was that it became even more preferable to fish from the islands and other far-away places where price agreements between fishermen and traders could not be observed by officials.

Declining profitability of fish trade and transportation problems were major bottlenecks in the development of the fishery. For example, the fish ferry service that the Gwembe Rural Council had been operating from Sinazongwe was shut down in 1968: the dispersed nature of the fishery had made it impossible to reach all the shifting fishing camps with the ferry. Neither did planners take the needs for improved infrastructure seriously. When a coal mine was established in Maamba in 1968, this opened up for a vibrant bicycle fish trade from the nearest fishing villages to supply the 700 coal mine workers. However, the potential benefits of the establishment of large-scale industry in the Gwembe valley were limited: the tarred road that was constructed from Batoka on the plateau down to Maamba was not extended to the lakeshore. And even if Maamba was electrified, no communities along the lake (the source of the electricity) were connected to electricity.^[87]

In such a situation, it was obviously difficult for the traders to invest in the fisheries and establish stable contracts with fishermen. The main strategy for fishermen to increase their number of nets and other fishing gear, was thus to purchase these items themselves when they brought fish to the urban market.^[88] With the low productivity of the lake and the amount of time the fishermen spent away from fishing while trading, it was unlikely that these investments would result in an increasing fishing effort. A report from 1972 probably describes the situation in a nutshell: “As the fishermen (...) suffer greatly from poor market integration, many of them have little incentive to catch more fish than is necessary for subsistence. High profit margins between producer prices and retail prices force the more enterprising fishermen into the fish trade which, due to bad transport conditions in Gwembe South, is very time-consuming” (Brandt et al., 1973:127).

A combination of ecological factors (declining nutrient levels in the lake resulting in declining catches and smaller fish), poor infrastructure and declining rural-urban terms of trade thus contributed to a continuing low level in fishing effort. In the early 1970s, the number of Tonga fishermen still remained around 500 with an additional 5-600 non-Tonga fishermen on the Zambian side of Lake Kariba.

2.4 War and collapse of infrastructure along Lake Kariba (1974-1980)

During the Zimbabwean war of liberation, several attacks by the Rhodesian army extended into the bordering Zambian areas. The Zambian shore of Lake Kariba was particularly hard hit, especially by land mines, and the dirt-road network along the lake that had existed (at least during the dry season) was blown up. Thus there were no longer any roads connecting the fishing villages. Only the main roads between the valley and the plateau remained. Rhodesian raids also destroyed all larger boats in Sinazongwe. No government development activity was implemented between 1975 and 1979 (Walter, 1988:17). On the whole, the war was the most serious setback since the relocation-process when the valley was inundated. After the attacks, shops in the Valley were few and poorly supplied. Whereas most homesteads had a bicycle in 1971, people again went on foot in 1981 as they had done in 1956-1957 (Colson and Scudder, 1988:35). In other words, a lot of the investments that the Tonga had been able to make from the first profitable years of the fishery, were destroyed.

Fishing did not come to a total standstill during the war years. The fishermen searched for fishing grounds in the safer areas, and they were allowed to cultivate some food for their own subsistence in the Tonga villages. However, as a result of the collapse of infrastructure and the danger involved in moving around in the Valley, fish traders almost completely stopped travelling to buy fish from Lake Kariba. Ahigh density of land mines also prevented safe movement by traders for many years after the war. Therefore, fishermen and their families processed the fish themselves, bartered it for food locally and brought dried fish to urban markets when possible. If fish marketing networks with established links between urban traders and Lake Kariba fishermen at all had evolved in the 1960s, most of them dissolved during the insecure years of the latter part of the 1970s. The fish traders who had established bond-friendships with particular fishing camps or households had to seek opportunities elsewhere.

2.5 Lake Kariba as an opportunity for the unemployed (1980s)

The Zambian economy was ruined when the copper prices on the world market fell as a result of the oil crisis in 1973. In 1983 Zambia, as part of an International Monetary Fund (IMF) Structural Adjustment Programme (SAP), devalued its Kwacha by 20 percent (Banda, 1991:12). Furthermore, Zambia had a decline in formal sector employment from 27 percent of the labour force in 1976 to 15 percent in 1988 (Pearce, 1989:6). As a result of these macroeconomic events numerous people all over Zambia, in particular in the Copperbelt towns, lost their jobs. Living conditions in general were deteriorating. When Lake Kariba became safe after Zimbabwe’s independence in 1980, people thus came from a variety of occupational backgrounds to seek their opportunity in the fishery that for long had been almost unexploited. The number of fishermen thus increased to 1 500 in the mid-1980s and reached 2 500 by the early 1990s (the same number as during the peak of the fishery in 1962). In a survey undertaken by the current project (Jul-Larsen, 2003) it appears that these fishermen were not, as fisheries officers have often claimed, coming from other (presumably overfished) lakes. It rather appears that the great majority had lost various kinds of wage employment in urban areas, particularly in the Copperbelt, and sought the Lake Kariba fisheries as an alternative source of livelihood.

One woman, for example, who arrived at Lake Kariba in 1980 at the age of 37 together with her husband, and who presently lives in the fishing camp Sinalilongwe, said: “My husband was a bricklayer and I was selling buns and other small things in Mansa. But then my brother-in-law, who had already gone to fish in Lake Kariba, told us that we were wasting our time in Mansa and that there was good money to earn in fishing. So we went.” Another woman moved from Eastern Province to Lake Kariba in 1982 at the age of 21. Her husband was a teacher and she was an untrained teacher: “It was my husband’s idea to come to Lake Kariba. He said there would be money in fishing, so he went and tried. Then I came after him”. Another fisherman had a background as a pharmacist who was not able to earn a viable income when people could no longer afford to buy drugs.

In the African context, the population of copper producing Zambia was one of the most urbanized. The rural-urban ratio had changed from 80:20 in 1963 to 60:40 in 1979 (Kaplan, 1979:86). However, this rapid urbanization trend slowed down for a while in the 1980s. Whereas it was expected that the urban rate would reach 46 percent in 1980, it landed at 43 percent, and between 1984 and 1988 there was in fact a modest reverse trend in the urban rate from 48.3 percent to 47 percent (Pearce, 1989:9).^[89] Seen in this perspective, migrants coming to Lake Kariba were responding to a situation of national economic decline, increasing unemployment in urban areas and a continuing population growth rate of 3.7 percent per annum (Banda, 1991). At the same time, the Zambian government repeated its “Go back to the land”-slogan from the 1970s and implemented relocation programmes in the latter part of the 1980s (ibid:84). The copper mines even held workshops giving workers practical information about how to settle in rural areas upon retirement (Ferguson, 1999:72). The decline in the copper economy and the urban crisis - a phase in Zambian history that Ferguson (Ibid:11) characterizes as “de-industrialization” and “counter-urbanization” - affected people from all ethnic groups. According to Walter (1988), there were both local and migrant Tonga, and migrants from a wide range of ethnic groups (though predominantly Bemba), among the “new” fishermen in Lake Kariba. The impression that the migrants came from urban employment and not from overfished lakes is strengthened by Walter’s finding that the migrants had a higher level of formal education than the Tonga (Ibid:40). Of those who had fished before moving to Lake Kariba, most of them had fished in Kafue (Walter, 1988:51).

Though both Walter (1988) and Chipungu (1988) point out that fishing and fish marketing were separate occupations, our data show that many of the fishermen started as fish traders. In 1988 Walter (1988:21) recorded a number of 800 fish traders out of which 180 were fishermen themselves. However, “the number of people seeking employment in fishing is higher than the number of jobs vacant” (Ibid:72). Many fish traders thus seem to have regarded fish trade as an entry ticket to the fishery, both economically and socially. A common career path could be like this: a man started trading, going to Lake Kariba buying and drying fish. After a while he might have established more stable contacts and could even be allowed onto the canoe of one fisherman to learn fishing. Sometimes this would be a relative who was already established as a fisherman. At this stage, the trader’s wife and children would join him. If he did not have a wife, it is very likely that he would marry one of the numerous female fish traders who were now coming to the lake.^[90] Through savings from his own and his wife’s fish trade, perhaps with a contribution from the fisherman who taught him fishing, he would invest in a canoe and a few nets to start fishing on his own.

During the 1980s, fish prices fluctuated and alternated between being regulated and deregulated (Chipungu, 1988:31). The profitability of fish trade was highly variable. Walter, (1988:91) found through his calculations of fish traders’ earnings that to sell fish as a wholesaler was comparably attractive in relation to wage employment. However, considering that 42 percent of the urban population subsisted below the poverty datum line in 1988 (Pearce, 1989:7) and therefore had a limited purchasing power - to the extent that many consumers had to skip meals (Ferguson, 1999) - fish trade was not particularly lucrative, even if it may have been regarded as one of the better alternatives. The price at which fish could be sold in a poor market remained limited. Considering the appalling state of infrastructure in the Lake Kariba area and the hardships involved in being a fish trader (like sleeping under a tree during the rainy season) it is perhaps not so puzzling that many traders preferred to become fishermen.

The poor profitability of fish trade can thus at least partly be seen as an explanation for the desire of traders to become fishermen. Many urban unemployed men first and foremost entered fish trade as a strategy to familiarize themselves with the Lake Kariba community and to establish personal relationships within the fishing camps in order to enter the fisheries at a later stage. Another strategy, which probably could only be employed after some contact with the fishing camps through trade, was to work as hired labour for the fishermen. Investments by traders in social relations in Lake Kariba fisheries was not, then, as in some other small-scale fisheries (i.e. West Africa and Asia), directed towards the creation of credit relations or ownership of the means of production. Zambian fish traders in the 1980s simply did not have any other capital to invest than their own labour. For women, their labour was not a valid currency in fishing, since it is considered a male activity, and their only niche in the fishing economy was thus in trade or in marrying a fisherman.

In the 1980s, fish processors at Lake Kariba largely dried or smoked the fish. One obvious reason was the problem of getting hold of ice. Lusaka did not have any large ice supplier until Kembe Coldstorage started producing ice blocks in 1989. Neither was electric power installed anywhere near the lake (except in Siavonga), so freezing of fish was not an option either. Traders thus had to stay for long periods in the fishing camps and on the islands buying and processing fish, and the practice of bartering and “subsistence” credit arrangements between fishermen and traders continued as before the war. Fishing camps were still extremely dispersed and variable in their duration. In a catch assessment report from 1986-87, only 26 percent of the fishing settlements identified were compatible with the settlements identified in a report by Beck from 1985 (Walter, 1988:34). Within a time span of two years, three quarters of the fishing camps must either have moved or been renamed. Furthermore, with the inadequate road system, unaffordable fuel prices, lack of accommodation and shelter both on the road and in the fishing camps, fish trade was extremely time consuming and exhausting. Both Walter and Chipungu report that because of their limited capital base, fish traders often did not purchase as much fish as they could have transported to the market. Most traders thus operated on a very “inefficient” level, making frequent trips trading in small quantities, and selling to poor customers in town.

2.6 New entrepreneurs in Kapenta fishing

Fishing techniques and gear types among the numerous new entrants in the inshore fisheries largely remained unchanged and - obviously - at a low cost level. However, another fishery expanded in Lake Kariba during the 1980s. The small sardine Limnothrissa miodon had been introduced in Lake Kariba from Lake Tanganyika in the late 1960s. In Zambia this fish is called Kapenta.^[91] The stocks increased with an amazing speed, but were not exploited until 1981 when the first Kapenta licenses were issued. The catching of Kapenta requires a particular type of vessel (a so-called rig) and strong light bulbs to attract the Kapenta in order to catch them with dip nets in the dark of night. The inshore fishermen were thus unable to utilize this resource from their small canoes.^[92] Neither did they have the capital nor technical or managerial know-how required to enter the Kapenta fishery. Hence, with the Kapenta fishery, a new group of entrants (in Zambia called Kapenta operators) with stakes in the lake and the land entered the Valley.

Unlike in the pre-independence days this fishery was not reserved for black Zambians, although 50 percent of the shares in each company had to be registered in the name of a Zambian. When the Zimbabwe war ended, a number of businessmen and commercial farmers thus stood ready to invest in the Kapenta industry. The new entrepreneurs were expatriates (for example Italian and Indian) who had worked for international companies and were resident in Zambia. But most of them were white Zambian commercial farmers or businessmen - many of them returning from years of “exile” in Southern Rhodesia and South Africa. Some Kapenta operators were South African or Zimbabwean (South Rhodesian) by birth. Capital for this new industry largely came from personal savings and loans from relatives abroad, and from the private sectors in Zambia, Zimbabwe and South Africa. No black Zambians have so far succeeded in entering the Kapenta industry on a permanent basis, except as workers employed on the rigs and in the drying and packing of Kapenta.

The first operators started fishing in 1981-1982, and several more established themselves in the following years. The Kapenta rigs were copies of those that already were in operation on the Zimbabwean side of the lake. On the Zambian side the rigs were built slightly smaller, and the technology in Kapenta fishing remains very similar today as when it started in the 1980s. Though some operators have diesel motors on their rigs, most of them are still operated manually with four workers hoisting the dip nets up from a depth of 20 metres with a crane.

Since Kapenta was a completely unexploited resource (at least on the Zambian side of Lake Kariba), the catches were tremendous during the 1980s. There was also a high domestic demand for dried Kapenta in the urban markets: Kapenta is ideal for the poor customer because it can be bought in very small quantities. The Kapenta operators thus made good money, not because prices were high but because Kapenta catches were good. Wholesaling and retailing of dried Kapenta became a new niche for urban traders. Walter registered a number of 200 Kapenta traders in 1988 (ibid:21). It does not appear that Kapenta traders combined trade in inshore fish with Kapenta trade, nor vice versa. Kapenta trade and inshore fish trade was - and still is - largely regarded as separate businesses by the individual traders.

TABLE 1. Overview of Kapenta companies in the Sinazongwe district.

Source: Own and Turid Bøe’s field data 1998.

In the early years of Kapenta fishing, the operators relied completely on traders coming to their premises to buy dried Kapenta. A system developed whereby the traders deposited the amount they intended to spend in the safe of the Kapenta company. On a first-come-first-served basis the traders are supplied with bags of dried Kapenta. The price per bag, however, is not decided at the point in time when the trader deposits the money, but when the operator supplies the bags. At times the traders will wait for many weeks, and in the meantime, the Kapenta operator gives them mealie-meal and Kapenta to eat. One may see this system as a kind of contract whereby the Kapenta operator gets an indication of how much Kapenta he will be able to sell. Since credit relationships between the Kapenta operators and traders very often fail (because an indebted trader can decide never to come back but to buy from other operators instead). He is also assured that the traders actually have enough cash to buy the volume of Kapenta that they have ordered. When the traders deposit money with a particular operator, he/she de facto commits himself/herself not to “run away” to buy from another operator even if his/her price is lower. The operator cannot prevent traders from leaving, of course, but traders rarely do this, since they come last in the queue for Kapenta if they switch to a new operator. Every month, a complicated price negotiation process is therefore going on: between the traders and each Kapenta operator, and between the Kapenta operators themselves.

As long as Kapenta catches were good, Kapenta operators did not worry too much about their marketing strategy. But as catches stabilised at a lower level after the first boom years, they had to make sure that they made enough profit to cover their costs. The paradox was, however, that the Kapenta operators were not able to co-operate amongst themselves in order to reach good prices. Instead, they underbid each other in the monthly negotiation process in order to ensure that the traders bought from themselves rather than from somebody else. The result of this lack of co-operation was that the prices of Kapenta remained low in the late 1980s, even if quantities of Kapenta on the market declined (due to stabilizing catch volumes). Despite this, the number of Kapenta traders remained quite stable: most of them just traded on a smaller scale than before. In my view, this indicates how “small” the Zambian market is: there is a clear limit to how much the Kapenta consumers, most of whom belong to the poorest segments of the population, can pay. In addition, competition from cheaper Kapenta, imported from Zimbabwe and Mozambique, has kept prices low.

Kapenta operators therefore increasingly had to try other marketing strategies (see Table 1), either they by-passed the traders by opening their own urban outlets, or they tried to get higher prices from a more wealthy segment of consumers, by pre-packing kapenta for distribution through supermarkets, a strategy which needed investing in freezing facilities. The Kapenta operators have faced many external constraints linked to macro-economic conditions in their strategies to increase the profitability of their business. But inevitably their inability to keep contracts (devised to create order and reduce uncertainty in exchange, to paraphrase North) has played a role in making expansion and investment difficult.

3. ECONOMIC LIBERALIZATION AND CO-MANAGEMENT IN THE 1990s

Continuously failing attempts by the United National Independence Party (UNIP) government at following its own version of structural adjustment after a break with the International Monetary Fund (IMF) in 1987, led to “food riots” among the population in the urban areas and to a reaction from the international donor community that decided to hold back bilateral aid (Banda, 1991). Eventually, the economic and political crisis led to a change in government, and the Movement for Multiparty-Democracy (MMD) gained power in 1991. Zambia embarked upon a new Structural Adjustment Programme (SAP). This involved liberalization of imports, trade and exchange rates, reduction of expenditures in governmental institutions, privatization of parastatal companies, removal of subsidies in the agricultural and transport sectors and introduction of user-fees in the health and education sectors. Unemployment rates increased. From 1990 to 1998 jobs in the formal sector fell from 543 000 to 465 000 (Rakner, van de Walle and Mulaisho, 1999:65). 26 000 jobs in mining and 15 000 jobs in construction were lost. The only alternative for the retrenched workers was to try to survive in the growing informal sector, one way or the other.

Considering this employment situation it is surprising, as Figure 1 showed, that the number of fishermen in Lake Kariba remained stable at around 2 200 from 1991 to 1994. Instead of increasing, the number dropped drastically in 1994 to 1 200 fishermen, a level that remained stable throughout the latter part of the 1990s. One explanation for the sudden decrease in horizontal effort could be environmental. Serious drought exasperated the economic crisis in Zambia in 1992. As a result Lake Kariba in 1992 reached its lowest water level since inundation (Karenge and Kolding, 1995), and this affected fish catches negatively. However, the lake level has risen again (it reached its maximum level in 2000). Consequently fish catches were good, but nevertheless the entry of newcomers into the fishery seemed limited. As local political access regulating mechanisms that reduced the growth in the fishing population are analysed in detail by others (Jul-Larsen 2003; Malasha, 2003), I will only briefly outline the major events as a background to facilitate an understanding of changes in the fish market.

3.1 Co-management through forced relocation

In the 1980s and 1990s there was increasing pressure on lakeshore land by migrant fishermen who needed to diversify their range of income generation. For example, many fishermen’s wives had small gardens where they grew vegetables, maize (for consumption and for beer brewing), and sometimes sunflowers as a cash crop. As increasing numbers of migrant fishing households settled in certain areas, Tonga farmers and village headmen found it difficult to accept additional clearing of land to make room for vegetable gardens. Tensions also arose because many fishermen and traders became involved in smuggling. Furthermore, Kapenta operators noticed increasing theft of fresh Kapenta from their rigs during the night. It was clear that the theft took place through transactions between rig-workers and traders in collaboration with fishermen who went out on the rigs with their canoes. Most of the stolen Kapenta was dried on the islands where many of the migrant fishermen lived. It was then smuggled out of the Valley through routes without police barriers. Kapenta operators also had interests in the development of wildlife tourism on several islands in the lake, and poaching of game by fishermen and others frequenting the islands was a thorn in their eyes.

In response to all these conflicts the Chiefs in alliance with the Kapenta Fishermen’s Association (KFA), Rural Councils and the Department of Fisheries (DOF) initiated a “co-management” plan that involved relocation of fishermen into designated villages in 1993. Using arguments about fear for overfishing and the need for fisheries regulations, these strong stakeholders aimed to get the theft and smuggling under control. In their view, the unregulated mobility of the fishermen had to be controlled. According to the new management plan, fishermen and their families were not to stay on islands, and they were - in many cases with force - relocated to the shoreline in 1994 (Jul-Larsen et al., 1997). The number of fishing camps was reduced from 278 in 1993 to 67 in 1995 (Jul-Larsen, 2003). Each camp elected a Village Management Committee (VMC), which in turn was represented in a Zonal Management Committee (ZMC), where fishermen, Kapenta operators, the Chief, the Rural Council, DOF, and the local business community (storekeepers) were represented. As part of the plan, levies were collected from both Kapenta traders and fish traders by the Rural Council. The idea was that this revenue, as a sort of motivation for compliance and as a compensation for the fishermen’s removal from their best fishing grounds, would be pooled back into the fishing communities in the shape of schools, health clinics, building and maintenance of roads. The Kapenta operators would be “rewarded” with better policing of smuggling and Kapenta theft.

All stakeholders with interests in the fishery itself or in land went to great lengths in negotiating and reaching solutions in their new communication fora, the Zonal Management Committees. However, problems were not easily solved. Temporarily, the Kapenta operators in cooperation with the police managed to catch many Kapenta thieves. As one “ex-smuggler” described the situation at that time: “It was like a war where you have to send out scouts before you fight the battle”. Gradually, however, the risk of being caught seemed to diminish again when funds collected by the Council through Kapenta levies failed to be allocated for the purpose of police patrols. The Kapenta operators also had problems amongst themselves in taking collective action. Finally, with accusations against the Council of embezzlement of funds, many Kapenta operators refused to collect the Kapenta levy on the Council’s behalf anymore.

In the fishing camps, people were equally disappointed with the outcome of the co-management plan. Firstly, the Council’s promises of improved infrastructure in the fishing communities never materialized. Secondly, the relocation of fishermen to the lakeshore led to an even higher degree of land conflict between fishermen and Tonga farmers. Many fishermen thus left the Lake Kariba fisheries in the mid 1990s. Some went back to the urban areas, others moved their equipment to other fisheries, like Kafue. The fishermen who remained, fished in the areas they could reach from their new locations.

Gradually catches within the “paddling-radius” of the fishing camps declined. Fishermen (usually without their families) started - at least periodically - moving back to the islands where catches were very good. With the high lake level and good catches, Tonga farmers also increasingly found fishing attractive again. Another reason for the Tonga’s resumed interest in fishing was that many cattle owners incurred great losses because of the “corridor” (foot and mouth) disease. Many Tonga farmers, especially young men, have therefore become part-time fishermen during the off-farm season. The number of people fishing may therefore be much higher than the number of full time fishermen registered by the Fisheries Department (Figure 1). Thus, even if many migrant fishermen left as a result of the relocation programme in the mid 1990s, fishing effort and total landings have probably increased in the late 1990s. This has provided numerous Zambians with a much needed possibility to earn a living at a time when other alternatives were hard to come by.

3.2 Unemployment-driven growth in the number of traders

Urban unemployed men and women continued flocking to Lake Kariba to buy inshore fish and Kapenta. In 1996, as part of the SAP, a large number of parastatal companies were privatized all over Zambia, a process in which 150 000 jobs were lost. Zambians were on the move in more than one sense: from formal to informal employment, moving between rural and urban areas, experiencing deteriorating living standards and moving down the social ladder (see Hansen, 1997 and Ferguson, 1999). A few examples of fresh fish traders’ backgrounds illuminate the enormous changes that Zambians were living through, and also the variety of actors that entered the fish market.

Mary, for example, was on her second fish-trading trip in Sinalilongwe in 1998. She had worked in the Ministry of Mines until 1996, when she lost her job. She had been trying to sell various things, but many other people were doing the same and she found it impossible to make a living that way. Mary took what she had left of the compensation from her former employer and tried out fish trade.

Charles, buying fish in Simuzila in 1998, lost his job in the Zambia Oxygen Company in 1994. Since his mother was a Kafue fish trader, he felt he had the competence he needed to become a fish trader. His mate, a local Tonga man lost his job in the Security Department of Zamtel, Lusaka, in 1994. He went home to farm, but sold most of his cows in 1998 in order to get capital to start up as a fresh fish trader.

Rose, whom the author met in Namafulu, had a husband who worked in the NAMBOARD. When he lost his job in 1996, Rose started fish trading. The husband joined her one year later. Since they have relatives who are fishermen, their aim is to settle and start fishing in Namafulu.

Peter, a young Bemba trader in Namafulu, was working in a bar in Lusaka. In order to make more money, he started selling “salaula” (second-hand clothes) in the Kamwala market (see Hansen, 2000). He left that trade in the hands of his brother and used his savings as well as his contacts with a fisherman uncle in Namafulu to start as a fish trader.

Anne used to work in the Ministry of Health until she quit her job in 1995. She felt that her salary was so low that she had to try something else. She therefore started buying bundles of “salaula” that she brought with her to Zimbabwe. She sold the clothes there and bought groceries that she took across the border to South Africa. She bought hardware for the profit and took it back to sell in Lusaka. This kind of cross-border trade has increased as a result of the lifting of import restrictions in Zambia and with the opening up of South Africa after apartheid. Through her international business, Jane managed to save enough to buy a pick-up, and she is now trading in fish as well as providing transport for other fish traders.

The fish traders try as much as possible to diversify income generating opportunities. This they have in common. Although some of them are aiming to become fishermen, this is not at all the case for all of them. Fish trade is just another source of many inadequate sources of income. On the lake they seldom buy fish from the same community for longer periods. Also they often shift to fisheries in other lakes and rivers when they hear that the catches and/or prices are better there. Most of the urban and semi-employed traders are thus neither interested in, nor capable of, investing in fisheries.

However, not only “retrenched urbanites” entered the Lake Kariba fish market. In particular, experienced traders with a long career in fresh fish trade in the Kafue (Lusaka’s nearest fresh fish source) began to shift their focus towards Lake Kariba. They faced more and more competition from Lusaka residents who attempted to buy fish in Kafue. This crowding of traders combined with declining catches in the Kafue River (caused by high fishing effort and drying up of the Kafue River especially after the draught in 1992), resulted in diminishing returns from fish trade. These two crises - one economic and one ecological - happened to coincide with an increasing availability of ice. Kembe Coldstorage in Lusaka started producing ice blocks in 1989. In 1995, Sinazonge finally got electricity, and many of those “connected” started renting out space in their freezers to fish traders. Also, many Kapenta operators established crocodile farms and acquired cold storage facilities for the storage of crocodile meat and skins. Others acquired freezing facilities in order to find new market niches through the sale of packets of frozen Kapenta. Though production of ice was not the main purpose of the new equipment, Kapenta operators began to sell ice to fresh fish traders who were on their way to the fishing camps. This made it possible for traders to travel from town to the lake without ice (by which they saved on transportation costs), and to buy supplies if their ice melted while they were queuing up in a fishing camp to buy fish.

Some of the more experienced Kafue traders therefore found it attractive to take up fresh fish trade in Lake Kariba, despite its location further away (than Kafue) from Lusaka. With their long experience in fish trade, and their established positions among traders in the market places in Lusaka, many traders in this “Kafue-group” became, as we shall see in a moment, quite influential in Lake Kariba fishing communities.

3.3 Chaos and order in new locations

Since the mid 1960s, the dispersed nature of the fisheries had made it difficult for traders to reach the fishing camps (especially islands) with ice. With the relocation operation in 1994, the new camps became more easily accessible to traders. Whereas it had been a problem in many locations to get traders to come at all, traders with ice now flocked to camps accessible by road.

The influx of fresh fish traders has been encouraging for the fishermen, but this trend has also resulted in conflicts in many households. Fishermen increasingly sell their fish directly to fresh fish traders, and household members seldom process fish anymore. Women thus have less control with their husbands’ income. Less is spent on food, and trading trips to town by wives also become less frequent. Women experience that men spend less on beer brewed by women locally, and spend more on factory-made beer on their fish-selling expeditions. Hence women not only lose track of how much their husbands earn and how they spend the income - they also lose an important source of income (see also Colson and Scudder, 1988). Not surprisingly, then, fish buyers report that wives enquire about the quantity of fish their husbands sell in order to get an idea about their income (whereupon husbands tell traders to underreport their catches).

There is little competition among fresh and dried fish traders. Dried fish traders still come, but mostly buy the fish that fresh fish traders are not particularly interested in: tiger fish (H. vittatus), bottle fish (M. longirostris), barbel (C. gariepinus), the largest and the smallest sizes of breams (O. mortimeri, S. condringtonii, and T. rendalli), and some other species.^[93] Among the fresh fish traders, on the other hand, competition can be severe. A trader has to adhere to many unwritten “laws”. His or her economic survival depends on knowing these rules, and success can be achieved when one knows how to use them favourably.

The first commandment for a trader is to build up a trustworthy reputation. Secondly, a trader must share information. By following these two thumb rules he or she can interact with other fish traders and acquire crucial resources: information, experience, credit and social relations - in the fishing camp, on the road and in the market place. Without these resources, it is difficult to gain economically, if not in the short run, then certainly in the long run. Unless a trader is substantially more powerful than the others, breaking of the rules (for example by hiding information or taking advantage of social relations in such a way that it economically harms others) will be sanctioned through the “setting of traps”, as one trader put it. These traps can consist in not helping the “immoral” trader with ice supplies or with punctures, refusal of credit, spreading of unfavourable rumours about a trader to the extent that they result in witchcraft accusations, and appealing to fishermen and VMCs to refuse a particular trader access to their fishing camp. Importantly, the market system extends spatially from lake and fishing camp to city and market place, and a trader’s reputation follows him or her in all these arenas. Information about a trader’s actions in a fishing camp easily reaches traders in Lusaka and vice versa.

Time is a much more crucial factor in fresh fish trade than in dried fish trade because of the lack of storage facilities for ice. The only way to slow down the melting process is to cover it with sawdust. A first-come-first-served institution is the “line system”. This institution is common in most Zambian fisheries, and is meant to prevent unfair competition among fresh fish traders. As one trader put it: “The line is our law”. Upon arrival in a fishing camp, the traders’ name is put on a list administered by the VMC. The first trader on the list (“in the line”) is allowed to fill up his or her container (usually an old freezer) with fish before the next one is allowed to buy. An important principle is also that the trader must have enough cash to fill up the freezer without buying it on credit. This system is supposed to prevent that some traders are stuck with melting ice while others are given “special treatment” by the fishermen. The breaking of these rules by both traders and fishermen are, however, a recurring source of conflict in the fishing camps.

Another institution in the market is the “cov-ice” system. This is a credit system that provides traders with labour during the trading trip, and provides newcomers or traders who have “fallen” (into debt), with an opportunity to (re)enter the market system. A trader who has purchasing capital, and has hired transport and purchased ice and sawdust, hires a person as a helper for the trip. They stay in a fishing camp until they have filled up the freezer with fish. When it is full, the fish is covered with a layer of ice. On top of this ice, the helper is allowed to put a layer of fish (“cover the ice”), say worth 50 000 Kwacha, paid by the trader. This credit is the payment of the helper. When they reach the market in Lusaka, the helper sells the top-layer of fish, and pays back to the trader the 50 000 Kwacha: “No more, no less”. This means that he or she can earn a small surplus, and may go on another trip and increase the volume of fish trade gradually.

As the examples of the line system and the “cov-ice” system illustrate, competition among small-scale traders combined with an increasing market preference for fresh fish, creates a situation where traders, in order to survive, create and adhere to “laws” or “trade regulations” that make it possible for a large number of people to find employment. Paradoxically, these mechanisms, which serve to limit the accumulation by some few traders at the expense of the majority, also make the establisment of credit-supply contracts between fishermen and traders extremely complicated. This may partly explain the limited investment in fishing equipment and absentee boat and net ownership by traders. Institutions like the line system thus enhance the security against “falling down” in the risky business of trading in a highly perishable good, but it also prevents individual traders and fishermen from entering into contractual agreements. A few exceptional cases exist of traders who have succeeded in establishing contracts with fishermen. The author will let them illustrate the strategies by which traders attempt to enter into binding contracts with individual fishermen and communities, but also the context-specific constraints that inhibit investments.

With more concentrated settlement of fishing households, some traders managed to establish themselves more permanently in particular camps. These were traders with long experience and wide networks of contacts in both fish market and fishing camps. Most of them belonged to the “Kafue-group” - usually a male trader organizing a core group of co-operation partners, including his wife, relatives, and traders known to be trustworthy. A few women were also leading traders. One of them is the “chairlady” of fresh fish traders in the largest open air fresh fish wholesale market in Lusaka.^[94]

Each of these leading traders established themselves in one fishing camp striving to be the only leading trader in that camp. As long as one trader did not trespass the “territory” of another, these leading traders co-operated rather than competed. They developed good relations with the community by offering transport, supplying goods (especially mealie-meal and nets), by helping in the maintenance of roads, by giving assistance during funerals, and they participated in VMC meetings. As a result, they were given land to build more permanent houses in the fishing villages. As one VMC secretary commented: “He [the trader] has been very good. He has been coming regularly for more than five years. The most important thing is that he [unlike the other traders] comes even when the road is inaccessible”. In this particular case, the trader has become a member of the VMC of “his” camp. He also owns nets that he entrusts in the hands of fishermen who fish for him, and he can buy fish on credit from fishermen. It takes many years to build up such relationships: “You must be very good to the fishermen. Time is the key”. This trader claims that by now he knows whom he can trust and whom he cannot trust among the fishermen. Importantly, his membership in the VMC also gives him a possibility to sanction those who cheat him, for example by reporting them for illegal fishing methods.

By making economic and social investments in the fishing camps, traders have a particular aim: to secure their supply of fish and to get privileges over other traders when supply is limited. In the established traders’ view, a trader who has made commitments and investments in a particular camp should be given precedence over the occasional “free rider” trader who seeks his or her fortune “here and there and now and then”, as one trader put it. One could say that the main difference is between those who attempt to institutionalize their relationship with the fishing community, and therefore have a long-term interest in this relationship, and those who do not have the means or skills to take part in the process, and therefore have an interest in defying the “traders’ ethics” in order to make short-term gains (see Evers and Schrader, 1994).

In order to exclude “free riders”, the established traders organize their fish supply according to the “swapping-method”. They have at least three freezers, one freezer by the lake, one on the road, and one in the Lusaka market where the fish is retailed. Through the “swapping” of freezers on the lakeshore - that is when one freezer is taken away from the village it is immediatly replaced - these traders always remain first in the line. However, to escape obeying the “law” of the fish market (the line system) the trader needs to be in alliance with the fishing community and be under the authority of the VMC. In some camps, the VMC has also allowed “their” trader to have his own purchasing spot independent of the line where other traders have to wait for their turn. One young man, who tried his fortune as a fish trader, told me that when he went to a fishing camp with such an established trader, “It’s just like fighting Mike Tyson. He will tell me: “Welcome my dear, but you will have to be number six in the line”.

In discussions with VMC-members, fears of tendencies towards monopsony by a limited number of traders were expressed, both because the fishermen felt that “their” trader kept the fish price too low, and because they were afraid of becoming too dependent on one particular trader. As it was put: “What if he dies?” Nevertheless, they acknowledged that as a VMC, and as a fishing community in the midst of Tonga farming communities, it was an advantage to be allied with prominent traders who in most cases act as spokespersons for the fishermen’s interests. One trader said: “We are part and parcel of the fishermen”, and the Tonga farmers clearly also see them as such. In land conflicts or in cases where suspicion is thrown on the fishermen because of their conspicuous (in the eyes of some farmers) consumption of beer and possession of “luxury” items like radios, it has happened that traders negotiate on fishermen’s behalf. When traders have tried to mobilize communal labour for maintenance of the dirt roads going through the farming villages, they are often met with very little enthusiasm and participation by the Tonga farmers, who see the initiative as primarily in the fishermen’s interest. In fact, the increasing conflict level after the relocation of fishermen to the “Tonga” lakeshore, is often mentioned by traders as a major obstacle in their attempts to establish permanent relations and supply contracts in fishing camps.

3.4 New actors in the market for frozen fish

In the 1990s there has been a trend towards a preference for fresh or frozen fish in Zambian urban markets. Partly this has to do with the emergence in the liberalized economy of a small (and increasingly health conscious) elite. However, also “average” Zambians who are used to eating smoked or dried fish, increasingly buy fresh or frozen fish, despite their low incomes and lack of freezers at home. Fish is still cheaper than beef or chicken. In September 1999 beef was K5 000 per kilogram and chicken was K4 000 per kilogram. In comparison, frozen large breams were K3 500 per kilogram. Especially the smaller breams “of the size of the palm of a hand”, are popular in poor segments of the urban market. Since breams of this size gives the trader high turnover, they are often called “money-makers”. In a fish shop, small breams were K3 000 per kilogram, which means that one such fish would be about K1 000. In comparison, a small cup of dried Kapenta - the “poor man’s food” - was also K1 000.

The growing number of fish shops in urban areas is an indication of the increasing demand for frozen fish. The shops prefer to have a variety of species and sizes of fish. This attracts customers of various ethnic backgrounds (urban dwellers often prefer the type of fish they can find in their home regions) and from various social classes. Even if imported marine fish from Namibia and South Africa, especially horse mackerel, is available in the supermarkets, Zambians still seem to prefer fish from rivers and lakes. Small-scale traders are therefore facing competition from investors who hope to make a profit on the trend among urban Zambians to eat fresh fish.

In the latter part of the 1990s, commercial companies have increasingly involved themselves in the fresh fish market. Chani Fisheries operating in Lake Mweru is the largest fish distribution company, and several smaller companies attempt to establish themselves in other fisheries, such as the Upper Zambezi and Lake Kariba. By “commercial”, I here refer to formally registered companies, whose activities are more capital intensive than in small-scale trade. They also have coldstorage facilities, trucks, shops with freezers, and telecommunication facilities. Companies normally use standardized measures, buying fish in kilograms instead of in the volumetric measure “heap”, which is an approximate kilogram that can be negotiated, and thus often preferred in exchange between fishermen and traders. There is little social interaction and face-to-face contact between the company owners and the fishermen, who sell fish to them through their buyers. These companies are often connected to other economic sectors than fisheries, as well as to important political and business circles.

In 1996, Chief Sinazongwe gave permission to a Lusaka company to establish itself on the condition that it “developed the area” (i.e. by connecting the site to electricity) and that it would employ local Tonga as workers. In addition to some Tonga workers, a young Bemba man with some experience from fish buying in Lake Mweru in Sinazongwe was employed as their manager in 1997. Headmen and VMCs in certain fishing camps were approached, and the company was allowed to put freezers with ice in their camps that would be filled with fish purchased by a local agent appointed by the VMC. The company’s buyers would regularly go around by boat to collect the fish and supply new provisions of ice. Fishermen could also sell fish directly at the company’s premises in Sinazongwe, and in the beginning, the company gave out free nets to certain fishermen in order to secure a steady supply of fish. Fish was stored in a freezer container and transported on a truck to the company’s fish shop in Lusaka. It has, however, been a problem for the company to financially sustain their capital-intensive operations, and they have not been able to compete with the small-scale traders price-wise. They have also had a problem of reliable supply of fish, both from the camps where they had contracts with the VMCs and from those of the fishermen they had supplied with free nets. By 1999, after a year’s on-and-off operation, the company’s scale of operation has decreased rather than increased.

Some of the local Kapenta operators seem to have more success in their attempts of entering the frozen fish market. At the same time as Kapenta catches showed a downward trend and the problem of theft continued, the operators observed that canoe fishermen had good catches of fine bream (because of the high lake level). Some of the Kapenta operators with freezing facilities therefore started buying fish from local fishermen on a small scale.

One Kapenta company started buying fresh fish for freezing on a larger scale, and opened fish shops in Lusaka and Kabwe. They made arrangements with the VMCs in two nearby fishing camps to purchase fish independently of the “line system”. The company started out on a small scale by sending workers on bicycles to buy bream (and to a limited extent other species). The bargaining process was cumbersome, so the company opened up a sales point on their premises and announced their wish to buy fish on posters in the health clinic, the church and in the Kapenta-workers’ living quarters. The word soon reached all the fishing camps in the vicinity, and fishermen started bringing their catch for sale. There have been many heated debates between the company and the fishermen about the price level, about whether to measure fish by the heap or by weighing it, whether payment could be received in the form of mealie-meal, and so on. The negotiation between the company and the fishermen may actually be seen as a discourse between two different market systems: one “European” based on fixed prices and standardized measures between “neutral” partners, and the other “African” based on negotiable prices and measures, and “personalized” economic relations. The buyers of the company therefore become crucial mediators between these two ways of thinking.

Despite winning the discussion about standardization of fish prices, measurement by weight and payment only in cash and not in kind, the company (as all other traders) discovered the problem of retrieving payment for nets given out on credit. Hence, they decided only to sell nets for cash. In addition, they mainly sold fishing nets with mesh sizes in the range between 3 and 6 inches in order to promote the supply of bream of 1-2 kilograms, which attract the best prices in the urban frozen fish market. To increase fish supply, the company also practiced a kind of lottery where fishermen who sell fish to them could win a net for free. Though there have been many obstacles in the relationship with the fishermen, the company was in 1999 buying almost all the bream above the “three-inch size”. The fishermen in the closest fishing camps have therefore become quite dependent on the company for a market. Moreover, it is common knowledge among traders in every market place in Lusaka now, that there is no point for a trader in travelling to those particular camps anymore: he or she will only be left with the small-sized fish that the company does not buy. Therefore, when the company stopped buying fish for a month because they had to change the staff in their Lusaka shop, the fishermen faced serious problems in getting traders to their camps.

The Kapenta company has succeeded in entering inshore fish distribution through their conscious strategy of tending of a good relationship with the local VMCs and by temporarily offering higher prices than travelling traders. Marriage by a white owner and a black Zambian woman also opened up a new universe of information about the Zambian fish market and consumer preferences, as well as access to contacts and trustworthy workers through kinship links. This case illustrates that despite considerable amounts of capital (compared with the individual urban trader) and connections in the white business community, success in the Zambian fish market by a commercial company is achieved through similar strategies as those employed by the small-scale traders: through negotiation with local leaders; through the building of trust over time; and through social interaction whereby access to information and favourable contacts and contracts are achieved.

So far, the commercial companies have steered far away from investment in fish production. As most other actors in the market, they find it too risky. The risk is mainly related to the problem of controlling the labour and deliverance of fish by fishermen. Moreover, unlike the traders who establish long-term relations in the fishing camps, companies that are “modern”, “commercial”, “white”, “industrial”, do not hold the type of knowledge, skills, and cultural and social codes that are required to establish trust, loyalty and control in relations with the fishermen. This would be the main obstacle if companies attempted to invest in modern fishing equipment. As long as the company can maintain a steady supply by the fishermen themselves, they avoid the problem of controlling labour. Whether an enterprise on this scale can be sustained in times of decreasing catches (i.e. if the lake level drops), remains to be seen. Economic and political factors, such as local politics (i.e. chief disputes), urban consumer’s purchasing power, or the future of the Kapenta industry, may actually be of far greater importance to fish distribution companies when they decide on whether or not to continue buying fish in the lake area than the condition of fish stocks in Lake Kariba. If the national economy of Zambia continues to decline and the profitability of fish distribution remains limited, individual small-scale traders will continue to play the main role in the fish market for many years to come. If this scenario becomes reality, no major investments in the fishery are likely in the near future.

4. CONCLUSIONS

Given the poor condition of the Zambian economy, the small scale at which traders operate, the inadequate infrastructure and the “fluid” power structure in lakeshore communities, population-driven rather than investment-driven changes in fishing effort will continue to be observed. Local access regulating mechanisms (in particular those regulating access to land) and political reforms (such as the regulation of fisher families’ mobility), limit the number of new entrants into the fishery and thus also the population-driven growth in fishing effort. In addition to the limited purchasing power of Zambian consumers, social mechanisms in the fish market prevent accumulation of capital among small-scale traders. As a result of these constraints, Lake Kariba traders’ activities remain small scale. Therefore, their possibility to exert control over labour and capital employed in the fishery, and their potential role as investors in new technology, are marginal.

Even if the fishery in principle is open to everyone, the same mechanisms that prevent a very large number of fishermen from entering the fishery also prevent those who do get access to the resources of the lake from maximising their output, and from expanding into more efficient fishing methods or distribution channels. Ever since the creation of the lake, the establishment of mutually beneficial credit-supply contracts between traders and fishermen has been economically, socially, culturally and politically complicated. Currently, a national economy in decline, local conflicts, and fragile stakeholder relations is not the kind of “environment” that enhances the evolution of common norms and rules that would make the outcome of contracts more predictable. Keeping investments in means of production at an absolute minimum - the strategy employed by fisher families and traders - may therefore be seen as their only choice.

Larger market actors, generally commercial companies with access to capital sources from other sectors of the economy, have to cope with constraints that are similar to those faced by the traders, since they operate within the same environment. Should commercial companies, despite all odds in the present macro-economic situation, be able to make substantial money on frozen fish marketing, it is nevertheless highly unlikely that they will make investments in fish production, let alone invest in technologically more sophisticated gear. Their main constraint is their inability to enforce labour contracts and credit-supply contracts. Over time, personal relations of trust and common interest can be established but only with a limited number of workers and middlemen. However, there is no shared legal framework beyond these small “trust circles” that can be activated when contracts are broken. Investments thus remain risky.

Turning back to the theoretical argument, the Lake Kariba case illustrates that even if credit institutions and technological development have been observed in small-scale fisheries in almost every corner of the world, it does not mean that such a modernization trajectory eventually will happen in all fisheries. The type of economic and institutional development required to set in motion a process of capital accumulation and increasing productivity must therefore be understood in particular historical, economic and cultural contexts. Contrary to Platteau’s argument regarding the “natural” evolvement of institutions in order to overcome market imperfections, this study of market development and investment on Lake Kariba shows that economic actors’ ability to reduce transaction costs through the establishment of risk-reducing rules for exchange and contracts is limited when the constraints are as many as described for lake Kariba. One may in fact conclude that the market is “too imperfect”: at present the constraints are so insurmountable that they hamper institutional development.

The location of this particular fishery within the constraints of the Zambian economy is thus a major explanation for the limited profitability of the fishery and thus for the limited investment in more efficient fish production. An investment-driven growth in effort is therefore unlikely to happen unless either external capital is invested or Lake Kariba producers link up to more profitable (export) markets. And as biological case studies elsewhere in this FAO Fisheries Technical Paper have shown, the fisheries of Lake Kariba are characterized by a high degree of ecological resilience. Therefore, the (unlikely) scenario of uncontrolled external or foreign over-investment in the fisheries most likely would not lead to a “tragedy” for Lake Kariba fish stocks, but it would be a tragedy for the fishermen and fish traders, whose economic alternatives for the time being are meagre.

5. REFERENCES

Banda, G. (1991), Adjusting to adjustment in Zambia. Women’s and young people’s responses to a changing economy, Oxfam Research Paper No. 4, Oxfam Publications, Oxford.

Beatty, D.M.F. (1969), Results of a fish marketing survey in Zambia, 1964-1965, Ministry of Rural Development/Department of Wildlife, Fisheries and National Parks, Zambia.

Berry, S. (1997), Tomatoes, land and hearsay: Property and history in Asante in the time of structural adjustment, World Development 25:1225-1241.

Bierschenk, T. &. Olivier de Sardan, J.-P. (1997), ECRIS: Rapid collective inquiry for the investigation of conflicts and strategy groups, in Human Organisation 56:238-245.

Brandt, H., Kerlen, H., Schumer, M. & Waller, P. (1973), Report on the development possibilities of Gwembe South Region (Zambia), German Development Institute, Berlin.

Brox, O. (1990), The common property theory: Epistemological status and analytical utility, Human Organization, vol. 49, no. 3, 227-235.

Chauveau, J.-P. & Samba, A., (1989), Market development, government interventions and the dynamics of the small-scale fishing sector: An historical perspective of the Senegalese case, Development and Change, vol. 20, no. 4, 99-620.

Chipungu, P.M. (1988), Utilization and marketing of fish in Zambia, Country Paper (UN/ECA/FAO), Department of Fisheries, Chilanga, Zambia.

Colson, E. (1960), Social organization of the Gwembe Tonga. Human problems of Kariba Vol. I., Manchester University Press, Manchester.

Colson, E. (1962), Trade and wealth among the Tonga, in Markets in Africa, Bohannan, P. and Dalton, G. (eds.), Northwestern University Press, Chicago.

Colson, E. (1971), The social consequences of resettlement. The impact of the Kariba resettlement upon the Gwembe Tonga. Kariba Studies IV, Manchester University Press, London.

Colson, E. & Scudder, T. (1975), New economic relationships between the Gwembe Valley and the Line of Rail, in Town and Country in Central Africa, Parkin (ed.), O. U. P., London.

Colson, E. & Scudder, T. (1988), For prayer and profit. The ritual, economic, and social importance of beer in Gwembe District, Zambia, 1950-1982, Stanford University Press, California.

Degnbol, P. (1992), Resource management and artisanal fisheries: Relevance and conditions, in Fishing for development. Small-scale fisheries in Africa, Tvedten, I. and Hersoug, B. (eds.), Nordic Africa Institute, Uppsala.

Evers, H.-D. & Schrader, H. (1994), Introduction, in The moral economy of trade. Ethnicity and developing markets, Evers, H.-D. and Schrader, H. (eds.), Routledge, London.

Fafchamps, M. & Minten, B. (2001), Property rights in a flea market economy, Economic Development and Cultural Change vol. 49, no. 2, 229-267.

Ferguson, J. (1999), Expectations of modernity. Myths and meanings of urban life on the Zambian Copperbelt, University of California Press, Berkeley.

Geisler, G. & Hansen, K.T. (1994), Structural adjustment, the rural-urban interface, and gender relations in Zambia, in Women in the age of economic transformation: Gender impacts of reform in post-socialist and developing countries, Pressman, S., Asangbeigui, N. and Sommerfield, G. (eds.), Routledge, London.

Hansen, K.T. (1997), Keeping house in Lusaka, Columbia University Press, New York.

Hansen, K.T. (2000), Salaula. The world of secondhand clothing and Zambia, The University of Chicago Press, Chicago.

Hart, K. (1988) Kinship, contract, and trust: The economic organization of migrants in an African city slum, in Trust. Making and breaking cooperative relations, Gambetta, D. (ed.), Basil Blackwell, Oxford.

Hernæs, P. O. (1991), Modernizing Ghanaian fisheries. The need for “Social Carriers” of technology, Ad Notam, Oslo.

Jamal, V. & Weeks, J. (1993), Africa misunderstood, or whatever happened to the rural-urban gap? The Macmillan Press, Basingstoke and London.

Jul-Larsen, E., Bukali da Graca, F., Raakjær Nilsen, J., & van Zwieten, P. (1997), Research and fisheries management; the uneasy relationship. Review of the Zambia-Zimbabwe SADC Fisheries Project, CMI Report R 1997: 1., Chr. Michelsen Institute, Bergen.

Jul-Larsen, E., Kolding, J., Overå, R., Raakjær Nielsen, J., & van Zwieten, P.A.M. (2003) Management, co-management or no-management? Major dilemmas in southern African freshwater fisheries. Part 1: Synthesis Report. FAO Fisheries Technical Paper 426/1. FAO, Rome.

Jul-Larsen, E. (2003), Analysis of effort dynamics in the Zambian inshore fisheries of Lake Kariba, in Jul-Larsen (ed.) Management, co-management or no management? Major dilemmas in southern African freshwater fisheries. Part 2: Case studies. FAO Fisheries Technical Paper No. 426/2, FAO, Rome.

Kaplan, I. (ed.) (1979), Zambia. A country study, Foreign Area Studies, The American University, Washington D.C.

Karenge, L. & Kolding, J. (1995), Inshore fish population and species change in Lake Kariba, Zimbabwe, in The impact of species changes in African lakes, Pithcer, T. and Hart, P.J.B. (eds.), Chapman & Hall, London.

Kolding, J., Musando, B., & Songore, N. (2003), Inshore fisheries and fish population changes in Lake Kariba, in Management, co-management or no management? Major dilemmas in southern African freshwater fisheries. Part 2: Case studies. FAO Fisheries Technical Paper 426/2, FAO, Rome.

Koranteng, K.A. (1996), The marine artisanal fishery in Ghana: Recent developments and implications for resource evaluation, in Fisheries Resource Utilization and Policy, Meyer, R.M., Zhang, C., Windsor, M.L., McCay, B.T., Hushak, L.T. and Muty, R.M. (eds.), Proceedings of the World Fisheries Congress.

MacGaffey, J. & Bazeguissa-Ganga, R. (2000), Congo-Paris. Transnational traders on the margins of the law, James Currey, Oxford and Indiana University Press, Bloomington and Indianapolis.

Malasha, I. (2003), Fisheries co-management: A comparative analysis of the Zambian and Zimbabwean inshore fisheries of Lake Kariba, D.phil. thesis, Centre for Applied Social Science, University of Zimbabwe.

Miracle, M. (1962), Apparent changes in the structure of African commerce, Lusaka 1954-1959, Northern Rhodesia Journal, vol. 5, no. 2, 170-175.

Musando, B. (1996), Inshore fish population changes in the Zambian waters of Lake Kariba from 1980 to 1995, M. Phil. Thesis, Department of Fisheries and Marine Biology, University of Bergen, Bergen.

North, D. (1995), Institutions, Journal of Economic Perspectives vol. 5, no. 1, 97-112.

Nyirenda, A.A. (1957), African market vendors in Lusaka, The Rhodes Livingstone Journal, no. 22, 31-63.

Obershall, A. (1972), Lusaka market vendors: Then and now, Urban Anthropology, vol. 1, no. 1, 107-123.

Overå, R. (1998), Partners and competitors. Gendered entrepreneurship in Ghanaian Canoe fisheries, Doctoral dissertation, Department of Geography, University of Bergen, Bergen.

Overå, R. (2001), Institutions, mobility and resilience in the Fante migratory fisheries of West Africa. Working Paper WP2001: 2, Chr. Michelsen Institute, Bergen.

Pearce, R. (1989), Food consumption and adjustment in Zambia, Working Paper No. 2, Food Studies Group UNICEF, Zambia.

Platteau, J.P. & Abraham, A. (1987), An inquiry into quasi-credit contracts: The role of reciprocal credit and interlinked deals in small-scale fishing communities, The Journal of Development Studies, vol. 23, no. 4, pp. 461-490.

Platteau, J. P. (1989), Penetration of capitalism and persistence of small-scale organizational forms in third world fisheries, Development and Change, vol. 20, no. 4, pp. 621-651.

Rakner, L., van de Walle, N. & Mulaisho, D. (1999), Aid and reform in Zambia country case study, Report commissioned by the World Bank, The Aid and Reform in Africa Project. Chr. Michelsen Institute, Bergen.

Scudder, T. (1960), Fishermen of the Zambezi. An appraisal of fishing practice and potential of the Valley Tonga, Human Problems in British Central Africa, vol. 27, June 1960.

Scudder, T. (1972), Ecological bottlenecks and the development of the Kariba Basin, in The careless technology: Ecology and international development, Farver and Mitton (eds.), New York Natural History Press, New York.

Scudder, T. (1985), A History of development in the twentieth century: The Zambian portion of the Middle Zambezi Valley and the Lake Kariba Basin, Institute of Development Anthropology and California Institute of Technology, Clark University.

Scudder, T., Colson, E., & Scudder, M.E.D. (1982), An evaluation of the Gwembe South Development Project, Zambia, Institute for Development Anthropology Working Paper, New York.

Vercruijsse, E. (1984), The penetration of capitalism. A West-African Case study. Zeed Books, London.

Walter, G. (1988), Lake Kariba (Zambia), Socio-Economic Baseline Study, Ministry of Agriculture and Water Development of Zambia/GTZ/Gwembe Integrated District Development Project