I have many people to thank. My enormous gratitude goes to Professor Andrew Sayer, my supervisor, who provided encouragement, intellectual stimulation and wit. In particular, his attention to detail combined with his editing skills were useful. Dr. Wendy Olsen and Dr. Barbara Harriss-White also gave valuable support and guidance. I thank my colleagues in the Sociology Department for providing a congenial and stimulating environment. Special thanks to market traders and other interviewees for sharing their fruitful insights. My deepest thanks to my mum.

This research project was funded by the Economic and Social Research Council (studentship award R00429334303).

The thesis is a study of markets and how market processes are ‘embedded in social relations’. Taking the example of fruit and vegetables market trading as its empirical focus it examines the way in which all aspects of the business are embedded in their respective social milieux. This embedding has cultural and political dimensions, invoking interpersonal, private and public spheres, all of which make a difference to the economic activity.

This leads to a rather different view of markets from that of textbook economics or indeed popular rhetoric and understandings of markets or ‘the market’, in which markets are the spontaneous products of competition between rational individuals pursuing their self-interest. While economic forces such as price movements undoubtedly have an effect, we shall argue that real markets are affected by many more influences than these.

Perhaps surprisingly given the importance of markets in modern society they have attracted surprisingly little attention from economic sociologists, who have generally preferred to study organisations. However, the rise of the ‘new economic sociology’ (Swedberg, 1990; Granovetter and Swedberg, 1992; Smelser and Swedberg, 1994) with its focus on ‘embeddedness’ offers a fresh perspective for examining markets. Briefly, this claims that economic processes are always (and necessarily) embedded in ongoing social relations in ways which make a difference to outcomes. On the one hand these new economic sociologists might consider that markets are too important to leave to economists. On the other hand, it is important to be aware of the usual sub-text of rival disciplinary imperialism in such arguments and take one’s distance from them. We therefore need to be aware of the limitations of sociological approaches too, and take due account of the economic literature. Although mainstream economics remains notably insular, there have been interesting developments around institutional and evolutionary economics and political economy which offer more possibilities for dialogue with economic sociology (Hodgson, 1988; Campbell et al. 1991; Sayer and Walker, 1992; Sayer, 1995). Indeed, the concept of ‘embedding’ has been one of the main areas of overlap between these new literatures.

Why choose fruit and vegetables market traders? There are several reasons. Firstly, it is relatively simple market trading straightforward, familiar products, and hence one which appears to be close to the economists’ ideal type. Unlike many markets it invokes a very direct form of competition - including price competition - among traders who are literally close enough together to know what each is doing and for customers to make comparisons. Such a market represents a strong test of the embeddedness hypothesis. It would more easily be accepted by economists that market relations involving complex, customised products like weapons and cooked-chill food would be strongly embedded in wider supply chains. However, spot markets with arm’s length relations would appear to be less favourable for the development of embedding.

Secondly, it is a familiar, relatively accessible phenomenon for researching, especially given that as we note later, a quasi-ethnographic approach is needed to study something like embedding. In particular, the domestic and regulatory contexts are more accessible than for many other kinds of market.

Thirdly, it offers an opportunity to explore issues of marginality and ethnicity, as we shall show. Self-employment is the focus of the study, a subject which usually is not adequately covered by existing literatures on family business and ethnic minority business.

Fourthly, in examining fruit and vegetables retailing, the study counters conventional preoccupations on manufacturing activity. Other kinds of market traders (e.g. clothes traders) were not included in this study firstly because their markets do not have to be cleared so quickly as the fruit and vegetables market, and hence are less close to the textbook model. There were also practical reasons of what was feasible to cover in the time period.

This, then, is our empirical focus. In order to use the concept of embeddedness to analyse, we need to take a sceptical look at what the concept actually implies. Various writers (such as Werbner, 1984; Granovetter, 1985; Smelser and Swedberg, 1994) have argued that exchanges are embedded within culture, rules, habits, trust and other forms of non-contractual elements of exchanges. Embeddedness should be regarded as an on-going process and setting that is in part endogenous to economic functions but partly a result of wider influences. For example, in one of the seminal works by Granovetter on labour markets, apparently weak ties had a major influence on how people got jobs (Granovetter, 1974).

However, despite the importance of this insight, there is a danger that ‘embeddedness’ has become a catchall category that allows one to gloss over the constituent processes. In other words, ‘embedding’ is treated largely as an unchanging and unproblematic ‘factor’, something given, lacking richness, variability, and contestation. As Lindberg et al. note:

Secondly, embeddedness arguments are weak in analysing how social relations can become disembedded and ‘lifted out’ from their contexts (Giddens, 1990). As we shall see, the twin processes of ‘disembedding’ and ‘re-embedding’ suggest that far from moderating economic rationality, embedding can be a more effective and economic way of doing business (e.g. large retailers’ close relationships with ‘listed’ suppliers). At the same time, embedding is costly and social relationships face the ultimate sanction of economic viability; if they cease to be useful and become economically dysfunctional they may be threatened.

Thirdly, embedded relationships are unlikely to involve equals, and power differentials and economic marginality are significant for how social relations are managed and on what and whose terms. As Doel (1997) notes, in economic analysis power is usually subsumed beneath efficiency imperatives. Although sociological approaches allude to the question of power, their interpretations of it are frequently both superficial and peripheral to explanatory accounts. In particular, our understanding of economic marginality has not been helped by the verbiage of embeddedness and references to trust and cultural ties. While the latter are important for smoothing exchanges, they usually take place in the context of uneven relationships. For example, Ram (1994a) points out how Asian employers in medium-sized firms emphasise the importance of trusting co-ethnic workers while paying them very low wages and giving them minimal rights.

Fourthly, embeddedness arguments have tended to ‘de-politicise’ and remove the role of regulations and state policy by focusing mostly on market actors involved in the immediate economic relationships. However, as we shall see regulators are very much involved in structuring entry and conduct, including competition.

Thus, economic sociologists usually fail to carry through their own project by not incorporating market actors into wider social spheres. We may select out markets for study and expect behaviour in them to be shaped by market considerations and market forces alone, but of course in practice there are always many things going on at once in real markets, and the fact that actors are always also involved in other spheres is usually significant. For example, market actors are embedded in a ‘domestic setting’ which can make a difference to their ability to start up and develop economic relations.

Several writers (such as Werbner, 1984; Ram, 1994b) have suggested minority ethnicity makes a difference to how cultural ties are bound to economic activities. Their work on ethnic minority businesses shows that ethnicity can be an important aspect of the embedding of economic processes. An important consideration here is the avoidance of essentialist characterisations of ethnicity in which members of ethnic minorities are treated stereotypically as having enduring, common characteristics which determine their behaviour regardless of external circumstances and internal change and variation, for example inter-generational change and gender differences. However, while there were some ethnic minority traders, ethnicity was found to be of relatively limited significance in our study. This serves to underline several points about ‘embedding’. It is variable and contested and depends very much on particular context. Paradoxically, sometimes embeddedness arguments take insufficient account of the ‘dynamics of context’, and a single aspect, such as ethnicity, is offered as the key, effectively ignoring the nature of the concrete and complex social milieu. In offering a partial explanation that abstracts from concrete circumstances it inevitably makes essentialist claims. However, I suggest that general claims have limited power, since in practice, several processes interrelate with one another in contingent ways so making for particularities, and not merely ones of parochial interest. This suggests the need for a concrete analysis of markets.

Throughout, this thesis endeavours to explore the nature of the processes surrounding the initiation and development of ‘embedding’ within market trading. One manifestation has been traders ad hoc and opportunistic economic actions (e.g. spot exchanges at wholesale market places). At the same time, market trading also depends on other contexts which constitute different forms of ‘embedding’ (e.g. domestic relations and regulatory system). This means that it involves an on-going process of negotiating positions with others which may take the form of non-market exchanges. In market trading most actors are socially, and sometimes culturally, marginalised. Thus, market trading is a process of struggling for economic survival and seeking social recognition.

The first aspect of embedding examined is the labour process of traders and how this is socially embedded, including their interactions with suppliers and customers. Existing labour process literature tends to ignore retailing and exchange, thereby tending to underestimate the importance of face-to-face relations and negotiation among independent market actors. (I.e. not all workers operate within the technical divisions of labour of firms. Some operate in buying and selling across the social division of labour.) Moreover, it offers insights into the lives of self-employed, generally low income workers, instead of employees as usually studied in the literature.

The second aspect explored is how the domestic relations interrelated with the business of managing to survive. In trying to make ends meet, ‘self-employed’ businesses relied on family members to subsidise the business through unpaid and underpaid labour, giving capital to start-up the business and doing paid work so as to secure a family wage, as the household attempted to ‘make out’.

The third aspect of embeddedness is the ways in which markets are constructed by regulations. Indeed, regulators are necessarily involved in starting-up and managing market places, as we shall see. Regulators bring buyers and sellers together, and have a dual role of disciplining market traders and representing and protecting their interests. They also pursue their own interest and operate within their regulatory system.

The thesis also throws light on issues of marginality, in particular showing the links between economic marginality and social and cultural marginality. Not only is market trading itself marginal in relation to food retailing, but many of those involved are themselves economically marginal; in turn some of these are culturally marginalised as members of ethnic minorities. Finally some customers are economically and socially marginal too. At the same, this marginality implies that actors are not only trying to make ends meet but to improve their social standing and win respect.

The broad themes sketched in this introduction are elaborated in Chapter Two to provide a conceptual framework for the remainder of the thesis. The Chapter brings together several literatures on family business, minority ethnicity business, economic sociology and industrial organisation. On their own the literatures fail to account for the diversity and complexity of the processes surrounding how market traders set up and survive. In part, this shortcoming reflects weak conceptualisations of the role of the domestic economy and the regulatory system. It is argued that the power of market traders should be analysed in relation to the different ways their activities are embedded and disembedded. Like most economic activities market trading involves a mixture of competition, negotiation, domination, custom and regulation.

Markets such as that for fruit and vegetables have a history and their current situation can only be understood in the wider context of development of the food retailing industry. Chapter Three addresses the historical evolution of competition between different types of retail outlet within the fruit and vegetables sector. Supermarkets, convenience stores, greengrocers and market traders essentially vary in the ways they integrate across the social division of labour. This is critical to understanding how competition is both a medium and an outcome of such a matrix of social relationships within the food industry. Competition is not spontaneous and neutral in its implications but highly structured.

Market traders are in large part hostages to the competitive nature of the market places. Chapter Four examines the labour process and how it is largely shaped by the nature of the product and volume and timing of custom. Insights are drawn from several literatures, but Simmel’s (1955) argument that competition produces civilising effects is particularly pertinent. Bearing in mind the highly competitive context and the close physical proximity of stallholders to each other we examine how traders’ sociability helped to ease tensions between them and to pass away the time in what was inevitably low productivity work.

Competitive advantages accrue to market traders in the way they buy fruit and vegetables at the horticultural wholesale market places. But the potential benefits cannot in themselves provide a sufficient explanation for arm’s length relationships which dominate such situations. In Chapter Five, it is argued such a pattern of buying is unusual and exceptional, and owes much to the highly perishable nature of the product, and the marginal market positions of market traders and wholesalers in the industry. It is also argued that spot market exchanges are not between equals and are weighed unfavourably against market traders.

In Chapter Six, the emphasis shifts from weakly embedded contexts to stronger ones. Market traders have had to become to become able at managing their domestic division of labour. However, the manner in which this occurs is variable, depending principally on the structure and composition of the household and the income from market trading. The ways in which traders acquire new and established businesses are critical to their competitive performance. Gender differences are also central since power differentials within households affect what and when economic opportunities are grasped. The manner in which ethnicity influences the domestic relationship is also explored.

Competition operates in a political and regulatory context, indeed markets such as these are actually constructed through regulation. The interactions between the processes of competition and regulation are explored in Chapter Seven. The councils responsible for market trading have become ever more elaborate and complex, and their actions in this sphere are strongly influenced by their other responsibilities and by external pressures. As we shall see, the survival of market places owes much to this context, and indeed to the history and geography of market trading. The food industry, market places and councils have also had to deal with the twin processes of ‘de-regulation’ and ‘re-regulation’. The manner in which councils, in particular their market departments, both represent and discipline market traders is addressed, especially in light of the recent attempts to ‘de-regulate’ the delivery of public services. This process of de-regulation is far from neutral in its competitive implications. Moreover, we note in what ways market traders have had to become more ‘responsible’.

Finally, in Chapter Eight, we draw out the main conclusions and implications of the study.

The study aimed to get detailed, concrete and critical pictures of how market traders ‘made out’ and of the contexts in which they did so. Prior to formally beginning the research project, I worked for a year for 4 market traders, doing participant observation as a pilot study for my research. This enabled me to get an insider’s view of market trading, and to assess what would be required for the formal project. I kept a daily journal and noted the interactions between market traders and customers, traders and the regulators and exchanges between traders. I also noted the changing products, product prices and customers’ queues at the fruit and vegetables market stalls. The traders that I worked for sold clothes and were male Indians, of which two were born in the U.K. I worked at the Bull Ring Open Market, the Birmingham Row Market, the West Bromwich Outdoor Market, the Rugby Outdoor Market and several Sunday market places. I also attended the monthly meeting of the Bull Ring Open branch of the National Market Traders’ Federation (‘NMTF’) and committee meetings of the Birmingham City Council Commercial Services Department, which managed the retail and wholesale market places.

During this period I tried giving 100 traders at the Bull Ring Open Market questionnaires seeking information on inter-traders’ networks. Despite providing a pre-paid envelope with each form I received only 20 back, of which 15 were poorly completed. It became clear that questionnaire-based research would not be viable given the fact that traders lacked the time and inclination to respond. In any case it would not have been possible to do justice to the richness and subtlety of the subject using such methods. I therefore concluded that I needed a ‘quasi-ethnographic approach’ using methods such as participant observation and unstructured interviews. The interviews needed to be in a ‘natural’ setting to maximise information flow to unearth the substantial and causal relationships and processes. I needed to study several things: first, the close analysis of the labour process of market traders; second, the relationship between wholesalers and market traders; third, the role of regulation and relevant official bodies in shaping and responding to industrial changes and non-economic considerations; fourth, market traders’ biographical account and family relationships; and fifth, and cross-cutting all this, I needed to identify minority ethnicity groups to evaluate in what ways they made a difference.

As regards to research sites, it was already clear that West Midlands had a variety of types of market places. The selection of sites was not a representative sample of all market places. Rather they were chosen for possible theoretical interest, for example according to differences in size, regulatory framework and ethnic mix, as well as for practical reasons of access. The Birmingham city centre retail market places were key sites because they were significant to the urban and retail landscape, had indoor and outdoor market places, and had 32 fruit and vegetables market traders, of which 3 were African-Caribbeans and one was West African. There were many Indian and Pakistani traders who sold clothes, and many Anglo traders who sold house plants, meat and poultry and dairy products. It was also easily accessible to my home and I already had a working knowledge of the site.

The Wolverhampton city centre and the Bilston town market places were chosen because they were economically significant to the local population, had indoor and outdoor market places, and had 32 fruit and vegetables market traders, of which 2 were Indians, one was Pakistani, one was East African Indian, and 2 were African-Caribbeans. There were many Indian and Pakistani traders who sold clothes and many Anglo traders who sold general grocery items. The Wolverhampton city centre and the Bilston town market places were also chosen as a way to check and corroborate results found in the Birmingham city centre market places and to highlight similarities and differences.

The Coventry city indoor retail market place was included because it was in a middle of a re-development project - redevelopment schemes were present in most major market places but less advanced. The West Bromwich retail market places were chosen because they contrasted with other retail market places; the former were smaller and less frequent, and being part of the Sandwell District Sandwell Council did not operate a horticultural wholesale market place. Later, I looked at the Telford small town retail market places because they were contracted out to a traders’ co-operative but still owned by the Telford City Council. I also looked at the Bearwood and the Smethwick market places because they were privately managed market places and made to pay an ‘acknowledgement fee’ to the Birmingham City Council in order to operate. Finally, I looked at the Poolway outdoor market place because it was newly opened and managed by the Birmingham Council Commercial Services Department, however it proved to be short-lived.

Several different kinds of information were needed for the formal study. First, background information on the structure of the retail and wholesale fruit and vegetable sectors, was obtained mainly from secondary literature including trade journals, council minutes and other archival material, and independent reports. Second, information on the basic dimensions of fruit and vegetables market trading in West Midlands was sought through a survey of what types of market places existed and who worked in them. Third, information was required on the day-to-day operations of the traders, and their family situation and of other key actors. This was obtained partly through unstructured interviews with market traders, horticultural wholesalers, and different types of regulators, and partly through participant observation. Direct but non-participant observation was also needed of relevant official meetings.

The people chosen to observe and interview were not randomly selected but were selected for likely theoretical interest; e.g. according to factors such as market traders’ ethnicity, gender and age, wholesalers’ economic size, product niche and ethnicity, and the different types of regulators. However, I feel confident that I have interviewed a sufficient traders in sufficient depth to have built up a representative picture of fruit and vegetables trading. The ‘sample’ had 41 retail fruit and vegetables market traders, of which six were African-Caribbeans, one was West African, two were Indians, one was Pakistani, and one was East African Indian. Table Appendix (‘A’) 1.1 shows the distribution of ethnic identity of traders in the selected retail market places; 96 were ‘white’ traders and 14 were ‘minority ethnicity’ traders, or in other words 14.5% fruit and vegetables traders were of minority ethnicity. Seven of the traders were women, of which one was African-Caribbean, one was West African and one was Indian. The overwhelming majority were white males.

I also observed and interviewed 20 horticultural wholesalers, of which two were East African Indians, one was East African Pakistani, and one was Pakistani. Table A1.2 shows the distribution of ethnic identity of wholesalers in the selected wholesale market places; all 8 ‘minority ethnicity’ wholesalers operated in the Birmingham Wholesale Market. Four of the wholesalers were local managing directors of two national wholesale companies, and the rest were large and small regional independent companies. The overwhelming majority were white small wholesalers.

In order to study the regulation process, I selected interviewees on the basis of the different types of role they played and according to differences of context. Thus, two were horticultural wholesale market place inspectors from the Ministry of Agriculture, Fisheries and Food (‘MAFF’), three were council environmental officers who were responsible for regulations of food, safety and hygiene controls of retail food outlets, two were council officers from the planning and architecture department, two were council officers from the finance department, ten were council officers from the commercial services department who among other things regulated retail market places, one was an officer from the Ministry of Local Government at the Department of Environment and Planning, one was a market trader who managed retail market places in small towns in Telford on behalf of the council, and one was a market place supervisor who managed a private market place.

In introducing myself I told the market traders, wholesalers and regulators that I was a student doing a study on market trading and I just wanted some information on how they did things. Normally, I began by showing them a table illustrating the changing market shares of retail fruit and vegetables through different types of retail outlets between 1974-1994. Invariably the interviewees described how the multiples had increased their share of the market and said what were the problems and opportunities of greengrocers and market traders. This was an effective way of beginning because it quickly established a topic of common concern and helped to overcome initial distrust or suspicion.

The interviews with market traders were conducted behind the retail stalls and were frequently interrupted so as to serve customers and court trade (see ‘Appendix Interviews’ for an example of a transcript of an interview with a market trader and observational notes). Usually, the traders would serve and talk. Often the information sought could best be gained by observation rather than questioning. For example, I witnessed the different degrees of turnover between traders and during the day and week, and who did what kinds of tasks at the stall. The on-site interviews were also open-ended so giving traders more freedom to set and challenge my implicit frameworks of ideas and helping me to make causal linkages and suggesting other lines of enquiry. For example, I quickly learnt and felt the intense competitive pressures and close proximity of market trading. Furthermore, the ‘natural’ setting meant interviewees were less likely to hold anything back especially on sensitive issues. For example, in one case an Anglo trader described how he did not enjoy selling and serving customers in particular to African-Caribbean ones who were very fussy what they bought, but he thought about the money he would get so easing his manners towards them and bringing a smile to his face. Another Anglo trader described how Anglo customers did not like Indian and Pakistani customers fingering fruit and vegetables and had to yell ‘Don’t Touch’ to the latter. Such responses also indicate that although, like any interviewer, my identity almost certainly had some influence upon the responses, it did not mean that the interviewees only told me what I wanted to hear. Furthermore, it is worth noting that the market setting did not inhibit the traders from talking about their domestic situations. For example, woman traders often described the pressure of time doing a multitude of different kinds of tasks from getting the children ready to school to coming to the market place and then going back home to cook the evening meal.

Similarly, interviews with wholesalers were on-site and unstructured combining observation and discussion with questions (see ‘Appendix Interviews’ for an example of a transcript of an interview with a wholesaler and observational notes). This proved very informative for three reasons. First, I observed how and when market traders moved around the wholesale market place. Second, through ‘being there’ I could appreciate the extent of the volume of perishable stock of produce that needed to be sold to independent retailers. Third, I saw the different types of wholesaler and how they varied in terms of size and product niche - a point which might not have been apparent had the interviews taken place in an artificial setting.

The interviews with regulators were more interactive and provided me with an opportunity to learn the degree of complexity and detail in managing market places and to understand the multiplicity of agendas (see ‘Appendix Interviews’ of an example of a transcript of an interview with a retail market council officer). Before and after the interviews I spent a lot of time reading council minutes and other archival material and attending council meetings so as to maximise the information flow. Sometimes the information required corroboration through second interviews and checks with other regulators and archival material. The unstructured nature of the interviews meant regulators were more willing to use data, charts and other visual aids and sometimes they invited me to watch them working with people they ‘regulated’. Further, they suggested new lines of enquiry that I had not considered and which questioned my misconceived ideas and prejudices. Through ‘snowballing’, I followed up other regulators and was able to build up a picture of the role of the regulators.

The ‘triangulation’ of techniques of direct observation, discussion and secondary literature were combined in carrying out on-site interviews. The interviews with traders, wholesalers and regulators were not taped recorded or written at the time as this would have inhibited them. After the interviews I would jot down the key information and write up the interviews in greater detail onto my computer file. It was undesirable and impractical to do otherwise. I benefited from the richness of tacit knowledge and from the more ‘natural’ and interactive interviews which gave interviewees an opportunity to question my fallible ideas and to set and challenge lines of enquiry.

As already noted, the research study involved collecting data from different sources and using different research techniques. This generated an immense volume of data. Data analysis was inevitably therefore a complex task. There is no ‘right’ procedure in analysing and organising the data but nonetheless the techniques employed involve careful consideration at each stage of analysis. As McCracken (1988) and Sayer (1992) note, the failure to identify the ‘right’ way should not be an excuse for an amateurish analytical approach; Patton notes:

On the other hand, the process also involves practical considerations and learning by doing.

The analysis of interview transcripts (see ‘Appendix Interviews’ for some examples), archives, and observational notes involved reading, cross-checking and re-reading, and further clarifications and explorations in later interviews and observations and further search into the archives. In this way, the research project was an ongoing process that generated more data as I gradually built up a picture of the different processes involved. The process also ensured that the quality of my analysis and interpretation was carefully controlled. Initially the information was recorded and listed in a chronological order, but was later recorded and re-listed according to theoretical themes. This helped to make the data more manageable and facilitated moving across the various identified themes as the research project began to take shape. Indeed, through an interpretative process the theory and the empirical data came to inform and structure each other, so that a theoretically informed empirical narrative of the processes of how markets were socially embedded emerged.

To be sure, the transcription, analysis and writing are not discrete stages in the research process. Indeed, the process involves a lot of hard work, careful attention to details and re-writes. Extracts from transcripts, archives, observational notes were not chosen to illustrate pre-conceived theoretical points, but rather causally implicated in the various issues involved. In fact, the ‘evidence’ did not fit easily with any single issue, but rather highlighted how different themes were interrelated in contingent ways. In this way, the explanations exhibited fertility, capturing not just the particulars but also the general properties of markets.

fruit and vegetables market traders in selected retail market

places in 1994

selected wholesale market places in 1994

I shall discuss the relevant aspects of several theories illuminating the various dimensions of marketing trading. While I recognise that there are overlaps between some of them, I shall discuss them initially one by one. I will suggest how the theories need to be combined and in some cases modified but the details of this are pursued in the substantive chapters. For example, many traders can only survive because of the support from the domestic economy, relying on second income and unpaid labour - a claim that I elaborate in Chapter Six (‘Market Traders: Making Out’). While industrial organisation and regulation are treated separately, they are strongly inter-related as the latter affects the former, as we shall see.

There are several theoretical themes: industrial organisation, embedding and trust, regulation, domestic economy and gendered roles, marginality and ethnicity. Normally they are conceptualised separately, but in practice they combine together, and in some cases produce unexpected and surprising findings.

While market trading is an unusual form of economic activity, it nonetheless provides a useful example and context by which to evaluate some theoretical themes in industrial organisation, economic sociology and ethnic minority retail business literatures.

The conventional industrial organisation theorists like Coase (1937) pose the question of why all production is not incorporated within a single big firm or distributed among millions of individuals exchanging through spot markets. In reality there are many other forms of organisations and the question is why? Table 2.1 highlights the key social relationships in different sub-markets in the retail industrial sector and indicates how ‘markets’ are co-ordinated through competition, regulations, embeddedness and trust, reciprocity, planning, networking and control. The industrial organisation approach is a useful start to explain why all fruit and vegetables trade is not done through shops and supermarkets? Why are there many traders in each market place rather than just one large trader? What influences the size of the sector?

Table 2.1 The social relationships in the retail industrial sector.

The nature of any one sub-market depends on what is happening in other sub-markets. Significant economies of scale and scope can be achieved by firms that organise production and distribution along the supply chain. ‘Real’ economies of scale refer to cases of technical indivisibilities, sharing of technical know-how and working skills, coordinative economies and social domination and sub-ordination of workforce and suppliers (Storper and Walker, 1989:130-1; Doel, 1994). However, informational, technical and material problems associated with products and networking along the supply chain lead to diseconomies of scale and scope. For example, supermarkets do not know precisely the retail demand for fruit and vegetables and the perishable nature of fruit and vegetables creates problems of distribution and storage. Also suppliers cannot control harvest crops so gluts and shortages develop in the market. Small businesses take advantage of the unsophisticated nature of products to make suitable market arrangements (Williamson, 1975); for example, buying groups may be set up to achieve economies of scope and lower transport costs. However, variations in local retail demand, lack of shopfloor space and geographical dispersal of businesses mean that diseconomies of scale and scope of buying groups will set in, so limiting the number of small businesses that can benefit from them.

The diseconomies of scale and scope mean that an economic space exists for fruit and vegetables market trading and for micro-businesses, but always on a small scale. The impossibility of a fully centralised production and distribution organisation, and the limitations of buying groups provide the pre-conditions for the existence of fruit and vegetables market trading. The perishable and unprocessed nature of the product makes the spot market an important consideration to explaining how microbusinesses compete.

Microbusinesses cannot achieve ‘real’ economies of scale and scope through spot exchanges, though some economies of scale are quickly realised over a small volume of trade. Any cost-savings are based on how quickly the traders learn to trade and get to know of mis-matches of demand and supply in the market, as we shall see in Chapter Five (‘Market Traders’ Relationships with Suppliers’).

The structure of microbusinesses in the market is influenced by the level of competition, which is affected by market entry and start-up costs. Microbusinesses are a simple form of retailing that require little capital and knowledge to start-up, so making them attractive to working class people who aspire to be ‘entrepreneurs’. Furthermore, the ‘open’ market means that microbusinesses are formally free to enter and exit. Combining the points about low economies of scale, low start-up costs and ‘open’ market it would suggest that there would be many microbusinesses in the market place, all operating on low profit margins, as standard economic models would suggest, rather than just one or two large traders for each market place. However, the equalisation process of competition leading to low profits and tiny businesses need not happen because ‘first-mover’ competitive advantages also create ‘innocent’ but significant barriers to entry: in other words, history matters (Chandler, 1990). Nevertheless, any tendency for market traders to concentrate into larger businesses is always limited by competition from marginal traders who offset their higher costs per unit of sales through self-exploitation and use of unpaid or underpaid family labour and capital.

The industrial organisation approach has paid scant attention to differentiation of customer demand as an explanation of the structure of the retail industrial sector. Customers differ not only in their incomes, and hence the kind of products they can afford, but in their tastes, according to their social and cultural characteristics, as demonstrated by Bourdieu (1984) and Miller (1987). Customers do not just buy products but also experience shopping and live out a cultural lifestyle, and different kinds of retailer cater for different kinds of customer. For example, even within the market trading sector, fruit and vegetables market traders differentiate and specialise their product-lines to sell to certain sub-groups of the market shoppers on the basis of quality, fruit and salads, vegetables, traditional English produce, and minority ethnicity produce. This underlines a general point that microbusinesses compete and survive by differentiating retail demand as well as existing upon an already-differentiated retail sub-market, and that microbusinesses do not always or only compete by wooing the same sub-groups of market shoppers. In such cases, microbusinesses do not directly compete among themselves ‘head-to-head’ and threaten to drive each other out of the market.

In addition to the differentiation of customers, there are different kinds of shopping (even undertaken by the same person), which also influence the structure of the sector. Thus, roughly speaking the supermarkets cater primarily for weekly car-based shopping often paid for with credit card, and small grocers for top-up shopping paid in cash. Market traders serve either type of shopping but mainly for customers using public transport and paying by cash.

Finally, there is the simple but nevertheless important point that the size of these various market segments is limited by the volume of customers using each segment so affecting the number of businesses each segment can support and their profit levels.

In addition to differentiation by customer and kind of shopping, there is also a spatial differentiation of markets, such that there is usually an element of spatial monopoly in competition (Guy, 1994; Wrigley and Lowe, 1996). Thus, supermarkets compete with one another for prime sites with maximum degrees of spatial monopoly and corner shops depend on a local customer base. In the case of market traders, as shall see, there is a degree of artificially created spatial monopoly as a result of regulations governing the spacing of markets.

‘Embedding’ refers to how and in what ways economic relations are also social relations, and indeed are required by them in some shape and form. In economics textbooks formal models ignore embedding or see it as a market imperfection or irrationality. Yet, as Richardson (1959) points out, ‘imperfections’ are necessary for competition and price mechanism to work, otherwise no investments or commitments could be undertaken in a market of perfect information and costless entry and exit. However, though Richardson’s claim seems correct, these effects of imperfections are unintended and hence cannot explain the existence of embedding. A better explanation of the latter is given by Hodgson (1988:184-6) who notes that unwritten rules, routines, norms, habits and conventions are necessary parts in economic life for interpreting information, ignoring ‘noise’, and coping with complex information. In real markets, individuals and organisations need trust and ways of judging trust to achieve mutual understanding (e.g. Baker, 1984). Embeddedness and ways of interacting provide means of checking, achieving and building trust, and ensuring that it is not misplaced. Embedding may not always be primarily economically functional but can be dysfunctional, i.e. causes problems and is costly (Sayer, 1995). Embedding can also be non-functional as actors pursue activities for other purposes (e.g. socialising) which may inadvertently help or hinder the business.

At the same time, the term ‘embedding’ is used with ‘disembedding’ and ‘re-embedding’ to refer to the dynamic economic and social processes. In capitalist economies, market forces have powerful disembedding effects (i.e. ‘lifting-out’ economic relations from existing social relations); as Marx notes:

In other words economic processes do not break permanently free from embedding. Rather they are re-embedded in new ways and places, perhaps in more instrumental and oppressive ways.

Several writers (e.g. Granovetter, 1985; Hodgson, 1988; Giddens, 1990) suggest that social relationships are embedded, disembedded and re-embedded but do not describe how and in what ways the processes actually operate. Thus, while it is true that economic processes can have disruptive and anti-social effects on established forms of embedding, and hence can be associated with disembedding (e.g. Polanyi’s (1944) representation of ‘self-regulating markets’), they have to be re-embedded somehow. So recognition of disembedding processes is not an excuse for reducing concrete economic processes to ideal types (e.g. ‘markets’ and ‘bureaucracies’). Concrete economic activities are always multi-dimensional and combine several modes of coordination in ways which are specific to those activities (Sayer, 1995).

Embedding varies in degree and kind depending on how complex the transaction is in terms of information and material process, as Wilkinson (1996) notes. For example, market trading in fruit and vegetables is weakly embedded because market traders sell generic and standard products where only price and simply quality information matters. However, supermarkets selling specialised goods liked cooked and chilled meals require strongly embedded technical relationships with suppliers to ensure safety, quality and regular supplies (see Doel, 1994). In such cases, the complexity of the production and distribution (and the high risks associated with food hygiene problems) necessitate considerable collaboration. While strongly embedded relations can induce innovation and quality improvements, embedding is costly if incentives for change are weak, and actors fail to be dynamic and competitive in their pursuit for new markets and lower costs (i.e. become stagnant). In such cases, relationships can be characterised as ‘voice without threat of exit’. As Hirschman (1970) argues, while economists tend to assume ‘exit’ is the only response to dissatisfaction, there is also the possibility of ‘voice’. In different ways, both voice and exit can be expensive; voice, if it fails to be effective, and exit because it involves risks and change-over costs. But voice and exit do not have to be considered separately; in particular voice can be strengthened by threat of exit. The relative importance of voice and exit again depends on the nature of the business. Several writers (such as Dore, 1983; Hodgson, 1988; Storper and Walker, 1989; Sako, 1992) have implied that strong and weak nature of embedding are necessary for capitalists to be more competitive and profitable.

Several writers (e.g. Granovetter, 1985; Williamson, 1985; Werbner, 1984; Swedberg, 1990) note that economic relations are embedded but do not examine issues of power and domination. The writers represent embedding as a negotiated process of equal parties freely entering a formal relationship leading to a fair outcome. However, embedding is far from being cosy (nor could it be as already noted above), and can be oppressive and exclusionary, and represent a form of domination furthering economic exploitation. Sometimes this domination takes racist and/or sexist forms (e.g. Phizacklea, 1990; Ram, 1994b) though domination and subordination is just as likely to occur regardless of ethnicity and sex since, in the abstract, the ‘market’ remains fundamentally neutral to the social characteristics of the actors. Combining these points about domination through ‘embedding’ and ‘disembedding’ processes, it needs to be recognised that market forces can liberate people from socially oppressive relationships, as Simmel (1955) points out, though of course ‘re-embedding’ can bring new kinds of domination and subordination. Later, we shall see how concrete markets are marginalised, and what differences ethnicity and gendered roles make.

Any business activity requires actors to trust in others and to be worthy of others’ trust, and this depends both on the honesty and commitment of actors and their competence and reliability. In additional to this interpersonal trust, there is also inter-organisational trust which might depend on reputations of institutions (e.g. Sainsbury’s had a reputation of paying on time and in cash to market horticultural wholesale firms). There is also system trust, such as in the use of money (‘symbolic tokens’) and the legal system and other expert systems (Giddens, 1990:29-36). There is always something ambivalent about trust relations. Trust is called into question when actors have grounds to doubt others’ trust claims or ways of doing business, i.e. trust is fragile and requires attention and building up.

As Wilkinson (1996) notes, different degrees and kinds of trust are required depending on the nature of the product (e.g. familiar or unusual product), the business (e.g. manufacturing, trading and retailing), and type of business outlet (e.g. supermarket, shop and market place). Furthermore, the nature of trust varies with different economic relationships: landowner-tenant, employer-worker, intercompetitors, business-regulator, business-family members, buyer-supplier and seller-customer, all of which vary across the types of retail businesses (see Table 2.1). However, it is implicitly assumed by some authors (e.g. Williamson, 1985; Granovetter, 1985) that trust is earned and given in situations where actors have equal power (i.e. are free from domination). Yet where actors lack power, they may have no option but to accept the offer of the dominant party, in which case trust hardly enters into it. Trust should not be confused with (reluctant) acceptance of domination.

As Werbner (1984) and Ram (1994b) note, social and economic marginality can make a difference to the kind and degree of trust required and accepted for doing business. Trust in economic matters often depends on creditworthiness, and the economically marginal businesses may be deemed untrustworthy simply because they are high risk debtors. Economically and socially marginal actors, more so than others, have to accept others’ trust claims and ways of doing business despite the trust relationship being low. As a result, the economic relationship is unlikely to develop further.

In standard economics textbooks (e.g. Lipsey, 1983; Samuelson, 1975), market exchange is seen as simply as a transfer of goods, services or money between free agents. But since the act also involves exchange of property rights in any developed system of commodity exchange, there must be a legal system inscribing, proscribing and protecting property rights, whose evolution is historically specific. Thus, contrary to popular myth, expansion of market relations necessarily involves regulators to extend and enlarge property rights. Property rights also imply certain structures of incentives and disincentives for individual actors to internalise externalities. In this way, the individual agent is treated as if he/she were the best judge of his/her own welfare. However, property rights over a thing and its ownership and control imply not only a relationship between persons and things, but social relations too (see Hodgson, 1988; Sayer, 1995).

Rules and regulations specify what can be sold, when and where, and by and to whom. In market trading, rules and regulations lay down opening hours, the manner in which services are rendered, and a system of fees, charges, penalties and fines under which licensed retailers have to operate. These are supposed to protect customers from harm and deception, ensure stability of competition but also, as we note below generate economic rent. The exact details of regulations vary according to the material nature of products, types of retail outlets and local economic and social conditions (see Wilkinson, 1996). For example, fresh meat products require tighter controls than fresh fruit and vegetables because meat products pose greater health risks to customers; and public retail market places are subject to more public order bye-laws than private retail market places. Regulation should not be thought of as necessarily counter to the wishes and interests of the regulated, for it is usually in the latter’s interest and is often actively sought by them. Indeed regulators to some extent rely upon actors to discipline themselves, both because its in their interest to do so and because they have the necessary concrete knowledge of the processes (Sayer and Walker, 1992). Moreover, in consequence, the relationship between the regulator and the regulated is not fixed but varies over time through an on going process of negotiation.

Regulators and the regulated have both shared and different interests for extending, stabilising and limiting competition. The nature of entry controls reflect concerns for possible economic effects on existing retailers and desire to offer shoppers qualitative choice and variety. For example, planning authorities assess possible harm and benefits to local retailers, shoppers and residents and consult with them in deciding whether to grant supermarkets permission to trade. Regulators assess how and in what ways a new entrant would enhance or disrupt the balance and mix of different trades that already exist in the retail sub-market; i.e. will competition generate a better deal for shoppers? Retail sub-markets need managing to make them attractive places to shop offering variety and choice. For example, retail market managers are careful to avoid having too many fruit and vegetables and jeans market traders so as to leave stalls free for less profitable but desirable tradelines like dried flowers and house plants and craftwear.

The nature of entry controls also relates to the suitability of retailers for undertaking competitive trading in a trustworthy manner without requiring too much supervision, regulations and controls; i.e. to act in honest and competent manner. Vetting procedures differ with different types of retail outlets. For example, for supermarkets reputation matters, whereas for small retailers checks on past businesses, banking details and referees may be required, and in the case of market trading a new market trader is given a trial period before he/she can apply for a permanent position. Clearly, entry is a negotiated process between regulators and the regulated having significant implications for how competition between retailers and between sub-markets works in practice.

Again as Wilkinson (1996) notes, the material nature of products and specificities of retail outlets affect the kinds of quality, food hygiene and safety control present in the retail industrial sector. For example, fire services demand greater controls for closed retail outlets like supermarkets, shops and indoor market places than for open retail market places. Regulations and controls are more particular for irregular retailing, unfamiliar products and economic activities that are very casual and insecure, such as car boot sales, Sunday markets and seasonal and occasional markets, as I will show in Chapter Seven (‘Market Rules and the Regulatory System’).

In practice, regulation is frequently about more than the regulation of markets; there are often other agendas impinging on it (see White, 1993; Harriss-White, 1993). In particular, by having monopoly control over who can trade what, where and when, regulators are able to extract economic rent by granting licences to trade. This is an important organising principle behind the relationship between regulators and the regulated, explaining their actions towards different retailers and sub-markets, and how they sometimes lead to perverse effects, as we shall see. Regulators may also be able to extract rent where they are landowners in relation to retailers as tenants, where they are service providers for retail sites, and where they are council authorities able to levy business rate taxes. The fusion of these different roles often creates conflicts of interest, which, for good or ill, impact on market traders.

Regulators are not free-floating but are themselves embedded in social structures, in particular the structures of the local authorities (see Mackintosh, 1997). Their agendas and the interests they pursue therefore differ according to their position within these structures. This means that the power balance between the regulated and the regulators depends not only on the size and influence of the former, but on the kind of regulator and their position within these larger structures and their relations to other concerns. For example, in spite of concerns raised by retail markets council department and market traders, planning and architecture and local economic development council departments approved plans for large shopping developments to give their city prestige and industrial and retail competitive edge over neighbouring cities, as in the case of the Dudley ‘Merry Hill’ Shopping Centre in the West Midlands.

Moreover, regulators need to be regulated themselves, and in the case of market trading there are indeed tiers of regulation. These tiers may not be coordinated entirely bureaucratically and some regulatory activities may be put out to competitive sector (see Ascher, 1987; Parker, 1990; Carnaghan and Bracewell-Milner, 1993). As we shall see, not surprisingly given this complex multi-level structure and its divergent interests, regulation does not follow a single logic and can even produce irrationalities.

The main questions on regulation that I shall address are to what extent regulatory practices ossify markets and sectors, to what extent they accommodate to structural changes within the sectors, and how far these reflect the widely differing powers of interests involved.

At first sight it seems only reasonable that studies of industrial organisation should deal simply with the formal economy, ignoring its relation to the domestic economy. However, there are many areas of production and market relations that depend on hidden subsidies from underpaid/ unpaid domestic work and undervalued household capital (see Phizacklea, 1990; Ram, 1994a; Adkins, 1995). The nature of the domestic economy and its articulation with the organisations of the formal economy therefore make a difference to the latter, and of course, the converse is also true. Similarly, much of the ‘new economic sociology’ literature (e.g. Swedberg, 1990; Granovetter and Swedberg, 1992; Smelser and Swedberg, 1994) also overlooks the relevance of domestic economy in understanding and explaining how and in what ways individuals are ‘domestically embedded’, so influencing their choices and resources for pursuing a career, taking up a job or starting up a business. Furthermore, in reality, domestic relations take certain gendered and class formations that have further implications for domination and subordination within the households and wider links to business, work, kinfolks and community (e.g. Finch and Mason, 1994; Skeggs, 1997). The combination of different types of work (i.e. self-employment, paid and unpaid work and domestic work) depends on various mixtures of customs, negotiation, power and know-how (see Pahl, 1984; Anderson et al., 1994).

On the other hand, the literature on small businesses and small Asian businesses has paid attention to the importance of the family. Yet it has often failed to state explicitly what they have abstracted the family from; i.e. the specific nature of the business. On the whole, the literature has examined well-established and middle-class businesses (e.g. Goffee and Scase, 1987; Metcalf et al., 1996). As we shall see, microbusinesses have both similarities and differences in relation to these.

Microbusinesses are a more economically marginal and simple form of retailing than small family retail businesses. Domestic relations and households assets are significant in providing start-up capital and a second income, especially as microbusinesses may lack support from banks and may struggle to build up customer trade. One implication of this is that microbusinesses are very vulnerable to failure in the event of domestic break up, as we shall see.

The exact nature of the relationship between family and business is contingent since how and in what ways family members are involved depends on career and job aspirations and work exclusion, the structure of the family in terms of age and health of family members, pursuit of self-interest and own goals, legal and moral family commitments and responsibilities, and economic position, security and performance of the business. Hence, the link between families and their businesses is an ongoing product of custom, negotiation, resistance and domination.

The domestic economy is invariably gendered and takes different forms according to class. For example, well-established businesses can absorb family members (usually wives and sons, sometimes both) whereas working class and more economically marginal businesses may require family members to do paid work to make the family wage. Further, as Skeggs (1997) notes, the male ‘breadwinner’ and female ‘carer’ are regarded as respectable domestic ideals and powerful ideals for working class households to aim for. For these reasons we consider it crucial to consider industrial organisation and domestic economy as related matters.

It is tempting to think of our object of study as being characterised by ‘marginality’. Yet this could mean several different things. First, it could refer to the economic position of market trading in the retail industrial sector. The economic position is peripheral to the industrial sector largely shaped and dominated by large capitalist organisations (see Smith and Williams, 1986; Guy, 1994; Wrigley and Lowe, 1996). That market traders continue to exist at all owes much to contingent conditions which stabilise them such as regulations on entry, underpaid and unpaid labour, and the continued existence of a retail niche market.

Secondly, marginality can refer to individuals who have an income which is low and insecure. Yet, as we shall see, the income may or may not be marginal since it depends on the competitiveness of their businesses, the size and stability of the sub-market, and on other sources of income (e.g. paid work), to which occupants may have access.

Thirdly, marginality also refers to individuals’ position in a wider environment in relation to cultural and political powers (i.e. ethnic and racial identities and gendered roles). Several writers (such as Murray, 1990; Morris, 1994; Jones, 1993; Ram, 1994b) have rightly or wrongly associated ‘minority’ ethnicity with marginality. Other associations can also leak into the discussion that seemingly give support to the second usage (e.g. culture of welfare dependency, cultural deprivation, entrepreneurial values, culture of poverty and racism). However, in order to do justice to ‘ethnicity’, it needs to be considered in general and in its own right.

These three kinds of marginality need not go together. A marginal economic position may provide an income which is not marginal for an individual who is not culturally marginalised.

‘Ethnicity’ is a theme which runs throughout the thesis rather than having a specific chapter on its own. In view of this, we shall say rather more about this at this juncture than we have said about other topics dealt with in specific chapters.

Ethnicity refers to collective groups with a shared culture, which may be defined (or defined themselves) on the basis language, religion or nationality. The idea of common origin may also be important and this origin may be mythical or real. The crucial point is that ethnicity binds the members of the group together in a sense of belonging and constructs boundaries between them and other groups.

The concept of ethnicity also highlights the diversity of experiences and interests of previously broadly defined groups such as ‘blacks’, ‘Asians’ and ‘whites’. Minority ethnic groups acquire certain social status, and each attempts to distant others and protect their own social standing - Anthias and Yuval-Davis (1992) represent this as the ‘ethnic hierarchy of difference’. Furthermore, as Mac an Ghaill (1988) and Ram (1994b) note, emphasising and celebrating dissimilarities between ethnic minorities may conceal shared or different responses to patterns of disadvantages, in particular white racism. For example, in Handsworth where there are many minority ethnicities the term ‘blacks’ can refer to a strategic alliance between them to fight against white racism, yet in Tower Hamlets ‘Bangladeshis’ may be a more suitable term: context matters.

Frankenberg (1993) argues that ethnicity can refer to white majority ethnicity, and describes how members normalise and naturalise historically contingent ways of thinking, practices and ideas. The majority ethnicity can create structural disadvantages for blacks as the latter’s practices and ideas may be regarded with deep suspicion, distrust and hostility. (Ethnicity also refers to white ethnic minority groups, and members’ greater appreciation of sense of common origin (mythical or real) based on religious texts, historical events and/or the idea of a (sometimes lost) homeland; e.g. Welsh and Scottish ethnic groups.)

Ethnicity is not fixed but depends on a changing mixture of customs, negotiation, power and know-how as ethnic groups re-interpret and mix cultures for practical purposes. In other words, to employ a recently introduced term ethnicity is characterised by ‘hybridity’; i.e. ethnicities are diverse, mixed and relational. Ethnicity involves both inclusion and exclusion mechanisms, and these vary between ethnic groups and depend on external conditions. As Williams (1991) notes, the process is filled with silences and discretions as well as cries and shouts, and can involve assertion and power as well as negotiation. The patterns of inclusion and exclusion are also unstable: for example white German shopkeepers in the U.K. were attacked by white English youths when England lost to Germany in the semi-finals of the European Football Cup; and Islamic mosques were targeted and firebombed by white fascist organisations during and after the Gulf War. While ethnicity is not fixed, challenges to boundaries can be difficult, and the concept of ‘hybridity’ can underestimate the amount of work and compromise minority ethnicities have to do to participate in what is natural and familiar to majority ethnicity (see Williams, 1991; Frankenberg, 1993; Gillespie, 1995).

Hybridity also refers to ways in which minorities within minority ethnic groups (such as women and lower castes) seek re-interpretations and mixtures of cultures to resist and liberate themselves from oppressive practices of their own culture. At the same time they may become more embedded in fending off vulgar aspects of majority ethnicity (see Gillespie, 1995). Indeed, both points are often overlooked by writers and researchers who tend to idealise and romanticise minority ethnicity and celebrate diversity (such as Mirza, 1992; Metcalf et al., 1996).

The hybridity of cultures can also refer to very particular and selective cultural practices. Commodities and practices can be easily re-interpreted and mixed (e.g. ‘curry and chips’; a pub crawl followed by a Chinese takeaway). By comparison, non-economic, social and personal relationships, such as inter-racial marriages, require greater work and compromise, as Frankenberg (1993) notes.

Traditional approaches to ethnicity were ‘essentialist’, that is, based on the assumption that a particular social category is marked by unchanging qualities, a common essence shared by all members of the category (e.g. all Asians are united by common characteristics, experiences and interests). However, empirical studies uncovered specific ethnic experiences and interests marking out distinctive ethnic groups (e.g. Ward and Jenkins, 1984; Mac an Ghaill, 1988; Metcalf et al., 1996). But, as Bradley (1996) notes, new forms of essentialism may emerge that imply that all members of a specific ethnic minority have similar experiences and interests. This overlooks other forms of stratifications. Indeed, society is a patchwork of internally diverse ethnic groups, and individuals may occupy different positions in relation to formal and domestic economies that run counter to the common experiences and interests of a specific ethnic group. Furthermore, second and third generation members of ethnic groups will have different experiences and interests to immigrants and first generation members, reflecting each generation’s distinctive challenges and access to resources: age matters. The second and third generation or ‘British Asians’ are more comfortable in crossing-over ethnic boundaries and having wider interests. Presumably, the virtue of the concept of hybridity is that it highlights this. There may also be scope for alliances and reconsolidation; e.g. black youth gangs and community elders can come together to fight against white racist gangs and organisations. In other words, the identities of individuals are relational: person A’s identity with respect to person B is different from that with respect to C. But in either case, the other relational identities are not wholly absent: A’s identity is structured by both B and C.

The new forms of essentialism may lead researchers and writers to wrongly attribute concrete events to ethnic identity that are actually produced by other social structures (Bradley, 1996). As Sayer (1992) notes, it is a common and easy mistake when researchers and writers fail to pursue qualitative questions after undertaking empirical research surveys involving generalisations. Generalisation surveys are mistaken for casual analyses, and usually researchers and writers will seek out cultural explanations to ‘explain’ variations in associations between concrete events without reference to causal mechanisms. For example, Mac an Ghaill (1988) notes, at one time it was common for empirical studies to cite cultural values and ethnic stereotypes in order to explain away the different academic achievements of each ethnic groups, thereby concealing other factors such as sub-standard school teaching practices and under-resourced schools.

In combining new forms of essentialism and generalisations surveys, researchers and writers may unwittingly produce new forms of racism that draw on and re-produce cultural and racist stereotypes. However, essentialism is not merely an academic problem. Essentialist ways of thinking about ethnicity or ‘race’ are endemic in everyday society and they structure social action and can tend towards racism. As Goldberg (1993) notes, racism need not always be purposeful and hateful, but emerge unconsciously where members of majority ethnicity idealise and romanticise minority ethnicity in ways that give the members of minority ethnicity unwarranted reputations, which are hard to shake off, and which may restrict their options, for example, African-Caribbeans as entertainers and Asians as businessowners, both ethnic groups serving society in useful ways. This is not to suppose that everyday essentialism goes unresisted: it is continually brought into question in the subtlest ways. Sometimes, however, the challenges are too subtle to be recognised.

How far and in what ways does minority ethnicity make a difference to retail businesses?

In some ways markets are neutral with respect to ethnicity and other differences while in other ways, particular to do with targeting customers, they are highly sensitive to them.

In the abstract, the neutrality of markets suggests that ethnicity should or will not make a difference because the market mechanism operates through unintended consequences of market choices, success may have nothing to do with merit or fairness and much to do with luck, and the logic of the market is that all that matters is what is offered for sale and its price (Roemer, 1988). The social characteristics of actors are irrelevant since in advanced economies markets operate largely under conditions of ignorance of how products are produced and sold, and who works and trades for who. Most importantly, money is abstract and neutral - one person’s money is as good as the next’s. Moreover, as Sayer (1995) notes, under competitive conditions actors have an incentive to be neutral for fear of losing sales and bargains to others. As Simmel notes:

However, in concrete cases of markets other social structures and contingent conditions give retail sub-markets historical, spatial, economic, political and cultural specificity (see Mackintosh, 1990, White, 1993; Sayer, 1995). While retail sub-markets are always embedded, they are not fixed but remain contested and ever-changing. As noted in earlier sections, market trading operates under contingent conditions that are controlled, marginal and unequal (e.g. entry regulations, underpaid and unpaid labour, and second income), yet the conditions can also be absent and be replaced by others. To be sure, market trading takes advantage of such conditions and in doing so may reinforce them, but they are produced elsewhere (though as we shall see, market trading does produce its own marginality and inequalities). For example, kinfolks and Asian workers, who are excluded from the primary labour market, will be poorly paid by market traders. Furthermore, as already noted, the social structures and conditions embedding the market place can create irrationalities (e.g. relaxed entry regulations and intense ‘head-to-head’ competition can further reduce quality of trading and lead to more empty stalls). The social embeddedness of the market place can also be economically non-functional (e.g. need to socialise), though not necessarily dysfunctional.

For markets to operate well, knowledge of customers’ tastes and preferences is important. Ethnicity may matter since some products carry significant cultural symbolic meanings (e.g. mangoes and breadfruit as desserts in African-Caribbean meals). In such cases, ethnic minority market traders will have a competitive advantage over others. Furthermore ethnicity may matter since some social interactions carry significant symbolic meanings. For example, market traders may act in racist ways to appeal to racist customers (though note a market trader need not always be an Anglo appealing to co-ethnic customers but can Jewish, West African or a Muslim targeting different sub-groups within the sub-market). However, ethnicity may not be significant when trading in mundane grocery items, and when competitors are in close physical proximity since traders will see what is selling well and copy others: the nature of the product and type of outlet matter. Furthermore, moral considerations and the economic need to woo non-racist customers may over-ride temptations to act in racist ways.

Ethnicity may also make a difference in terms of trust. As Wilkinson (1996) notes, the nature of the product and business make a difference to production and market relationships. Market trading requires trust, and ethnicity can make a difference in deciding who to trust. Visible (sometimes discrete) ethnic markers can act as trust credentials for similar individuals, but they can be offputting for certain others; e.g. a Muslim trader wearing a prayer cap may not offend Anglo, African-Caribbean and Muslim customers but offend Hindus. Furthermore, in the case of market trading, products are largely familiar items bought and sold at arms’ length under competitive and regulated conditions so that trust is not necessarily imputed to social characteristics of actors but rather to the social milieu of market places, and nor is there need for much trust in simple market transactions.

Ethnicity may be significant for access to material resources to start up a business and to survive (e.g. tapping into ethnic minority networks (Werbner, 1984), facing racism from banks (Deakins et al., 1995; Ram and Deakins, 1995)). However, in making the claim of a cultural and ethnic style of doing business, researchers and writers may unwittingly produce new forms of cultural essentialism and hence racism, as noted earlier; to see ethnicity as fixed underrates change and hybridity (Waldinger, 1995).

While some black people may take up market trading because of racism in the labour market and exploit racialised resources (cp. Ram, 1992), this does not mean that ethnicity makes a significant difference to how retailers behave. For sure, racism, sexism and other kinds of exclusions from the labour market encourage people to consider self-employment in occupations like market trading. Yet, this only suggests that market trading is affected by inequalities and marginality produced elsewhere. Once people are in the market place, the main thing is price. Market trading is not inherently racist since the former can survive without the latter. Furthermore, how and in what ways actors and resources come into sub-markets does not tell us how they will be actually used since the nature of the product and type of retail outlet make a difference, as we shall see.

The recent literature on ethnic minority businesses has unintentionally created new forms of essentialism and racism, and has created problems of identifying causality.

Many researchers and writers have conducted empirical generalisations associating many discrete events and measuring key variables after undertaking extensive research, using quantitative analysis, semi-qualitative questionnaires and in-depth interviews with sample size varying from 50 to 1,100 businesses (e.g. Ram, 1992; Rafiq, 1992; Jones et al., 1994). Researchers use either ‘objective’ indicators to measure discrete events such as education qualifications, number of employees, profit levels and number of hours worked (e.g. Jones et al., 1994), or get to know the ‘subjective’ nature of business through views, opinions and attitudes on issues such as racism, trust and entrepreneurial values and business aspirations (see Ram, 1992), and sometimes researchers do both (e.g. Metcalf et al., 1996). However, the research methods do not pursue qualitative questions or go into intensive research techniques and methods.

In most cases, researchers and writers conflate different types of business outlets across different industrial sectors, and then distinguish them on the basis of ethnicity of owners (e.g. Patel, 1988; Jones, 1993). Sometimes broad ethnic categories are used such as African-Caribbean, white and Asian (e.g. Jones et al., 1994), or more refined ethnic categories such as Muslims, Sikhs and Hindus (e.g. Baker, 1981; Rafiq, 1992) and Pakistanis, Bangladeshis, Indians and East African Indians (e.g. Metcalf et al., 1996). Some researchers (such as Ward and Jenkins, 1984; Blackburn, 1994) have examined businesses owned by white ethnic minorities such as Irish, Italians and Greek and Turkish Cypriots. The empirical studies conflate businesses in various ways, for example newsagents, restaurants and other services (such as taxi-drivers, insurance and estate agents) (e.g. Metcalf et al., 1996), retail shops, restaurants and wholesalers (e.g. Rafiq, 1992; Basu, 1995), and clothing and engineering manufacturing firms and professional services (e.g. Ram, 1992; 1994b). Moreover, most research studies involve businesses already well-established and middle-class (e.g. Werbner, 1984; Srinivasan, 1995) and middle-sized (e.g. Ram, 1994a).

Not surprisingly, different and conflicting issues and conclusions emerge given the broad nature of the research and the vast numbers of different businesses. Many researchers and writers (such as Werbner, 1984; Waldinger, 1990) suggest that there is something culturally distinctive about ethnic minority businesses. The distinctive style of doing business enables owners to survive and succeed. Other researchers (such as Ram, 1992; 1994b; Jones et al., 1994) do not deny the distinctive characteristics of ethnic minority businesses, but suggest that the styles reflect owners’ ways of coping with racism in the labour, capital and product markets.

Recently, some authors (such as Rafiq, 1992; Basu, 1995; Metcalf et al., 1996) have noted a greater diversity among black and Asian businesses, and have suggested a distinctive cultural style of doing business for each ethnic group that can be linked to their rate of survival and success. Yet, Metcalf et al. (1996) acknowledge that these different experiences may reflect different disadvantages each ethnic groups face from the white majority ethnicity, as well as from difference in their internal resources.

Several researchers and writers (such as Srinivasan, 1995; Basu, 1995) suggest that many Asian businesses are far from being economically marginal, as portrayed by earlier researchers such as Baker (1981), McEvoy et al. (1982) and Jones et al. (1994). Some businessowners seek and achieve upward economic mobility and social status.

In evaluating the literature we must remember that researchers and writers are addressing different questions and audiences, yet most commit the same flaws. As Wilkinson (1996) notes, how businesses link to families, kinfolks, industrial sectors, banks and regulators will depend on the size and nature of business, whether manufacturing or retailing, or selling unfamiliar or mundane goods. By conflating products and businesses and failing to pursue qualitative questions, researchers and writers may wrongly attribute success and failure to ethnicity when it is actually due to something else. Furthermore, as Cater and Jones (1991) note, in attributing causality to ethnicity, researchers and writers may unwittingly produce new forms of cultural essentialism and racism. To pre-empt critics who may accuse me of ignoring the significance and importance of ethnicity and racism, I simply point out that materiality matters, and it is not a weakness but a strength to give a better appreciation and understanding of the role of ethnicity.

The fruit and vegetables market trading provides a ‘window’ onto relations between several themes which are normally analysed separately but which actually intersect. The thesis will enable us to assess the relative significance of these different relations, and by bringing together ideas from different literatures avoid the dangers of mis-attributing causality and giving a one-sided account, and in the process correct false lay and academic knowledge.

In order to understand the position of retail fruit and vegetable market traders we need to situate them in a wider picture and understand the major elements and trends within the environment of retail marketing. The changes are becoming both more rapid and more diverse, with increasingly mobile, demanding consumers and with growing competition between retail formats, and between different supply channels.

The first section considers changes in consumers’ expenditure and the second section looks at changes in shopping patterns. Attention then turns to the main types of retail organisation (i.e. large multiples, co-operatives and convenience stores, greengrocers and market traders).

The total market for fruit and vegetables (‘produce’) in the U.K. in 1993 was worth around £6.2bn (retail prices); of this around £5bn was in-house consumption, £0.75bn catering and £0.45bn for food processing. In terms of volume (‘000 tonnes), the sector was 12,400; of this around 6,300 was in-house consumption, 2,480 catering and 3,620 processing.

Fresh produce is the third most significant household food purchase, accounting for 13.3% of the household food expenditure in 1990. While volume consumption of fresh fruit for in-house consumption has risen from 28.54 kg per person per year in 1983 to 31.76 kg per person in 1992 (see Table Appendix (‘A’) 3.1), volume consumption of vegetables has declined partly due to continued decline of potatoes from 97.6 kg per person per year to 84.7 kg per person per year between 1992-83 (see A3.2).

One of the main features of fruit and vegetables trade is the broadening of the range of fresh produce. Produce which would have been considered rare 10 years ago, such as mangoes and sweet potatoes, are now commonplace. This is partly due to improved transport so that products from distant producers can reach the U.K. quickly for quality to remain unimpaired. Better transport has helped to increase the all-year availability of produce. To a greater extent than the vegetables sector, fruit consumption has benefited from an increased willingness of many people to experiment with food. The fresh produce sector has also benefited from its healthy image and convenience. Greater affluence has also meant certain consumers will pay a higher price for produce that has an unusual new flavour and appearance. Further, foreign marketing boards have been active in promoting their soft fruit and ‘exotic’ produce; e.g. the New Zealand kiwifruit. Demand from minority ethnic groups in the population has also been a factor. The higher value/volume ratio of new produce over traditional produce has spurred large multiples to develop ways to strengthen and build up exotic and branded fresh fruit and vegetables trade. But exotic fruit and vegetables sector is still a minor sector, and traditional produce, such as potatoes and apples, remain dominant (Table A3.3 and A3.4).

These developments have meant a rise in imported produce relative to home-grown. After potatoes, mushrooms are the most valuable U.K. vegetable marketed commodity. Apples remain the most valuable U.K. fruit marketed commodity (predominantly Cox and Bramley) but strawberries are of increasing importance. British fresh vegetables (excluding potatoes) provided 66% of total value of supply in 1988. Fruit imports, in contrast, dwarf domestic sources, accounting for 81% of supply. Fresh produce imports are worth a total of £1.5bn and market penetration is increasing depending on the nature of the product. This has implications for the different types of retailers, wholesalers, growers, importers and overseas marketing broads, as we shall see.

The growth in car ownership has had a considerable impact on retail marketing; shoppers are more willing and able to select retail outlets over a wider area. Two main effects emerge. First, retail outlets no longer enjoy such high degrees of spatial monopoly and are forced to find ways to compete and survive. Second, large outlets with an appropriate retail mix of product-lines can achieve increased sales from a very wide catchment area.

Most shoppers who buy from large multiples come by car, whereas public transport remains important for many shoppers who buy at a nearby town centre. The weekly major shopping trip for groceries is the norm for those who shop at large multiples due to the convenience of having a car. Moreover, those buying at large multiples tend to be in full-time employment and relatively affluent.

Reduced shopping frequencies can be related to increased ownership of refrigerators which have increased the storage cycle. Shoppers have also become more selective and more demanding in their choice of retail outlet and what they buy, and are therefore willing to undertake longer trips to get the standard of shopping required. This invariably means they will buy most of their groceries there and then.

The increased number of women in paid employment has also reduced the time available to shop frequently. Late shopping and weekend shopping have become important for households where both partners work. Invariably, this has meant shopping has had to become more time-efficient, and more questionably to be a more pleasurable activity.

This section examines the changes in the retailing environment in terms of retail outlets. The physical structure of retailing has evolved rapidly in recent years through a combination of economic forces, consumer trends, competitive initiatives and relaxation of certain planning constraints.

The retail grocery sector is highly uneven; it is populated with a small number of a very large corporations and a very large number of smaller businesses. Table 3.1 shows that large multiples (e.g. Sainsbury and Tesco) had 43% of total retail sales in fresh fruit and vegetables products in 1990; co-operatives, symbol stores, small multiples and independent convenience stores had their share almost halved between 1980-90 to 7%; and greengrocers, market traders and farm shops saw their total share fall by 16% in 10 years.

Outlets 1980-90

The rise of large multiples is analysed and comparisons are made with convenience stores. Attention then turns to greengrocers. Finally market trading is examined.

Large multiples are defined as retailers operating supermarkets with selling area of under 25,000 sq. ft and superstores of selling area between 25,000-50,000 sq. ft. By building more and large stores and increasing sales turnover of existing stores, multiples attracted custom from rival businesses. Large multiples apparent lack of price competition between themselves reflected their attempt to establish spatial monopoly. Each new and large store could establish a loyal market of shoppers living closer to that store than other stores; as Moir (1990) notes:

Figure 3.1 shows how since 1980 Sainsbury, Tesco and Safeway increased their pace of stores openings: Tesco from 12 large stores in 1987 to 25 large stores in 1993, and Sainsbury from 15 to 23. However, Asda went into a sharp decline. Several factors made the pace of new store expansion possible.

First, regulation on planning permission was relaxed to allow large multiples to build supermarkets and superstores out-of-town, which offered advantages in terms of delivery access, car parking for staff and shoppers, flexible internal layouts and significant space to display the 20,000+ food and grocery lines.

Second, large multiples had huge investment and differential access to capital to fund the programme of new superstores and supermarkets. Wrigley (1992) points out that large multiples often paid in the region of £2m per acre for superstore development sites. In 1992 the ‘big three’ (Sainsbury, Tesco and Safeway) pumped over £2bn into new store expansion and raised £1.4bn of new capital via rights issues to fund this expansion.

Third, the relationship between gross floor area and sales area improved as less space was required for storage and preparation. Whereas before individual suppliers sent their goods to local stores, supply chains became more centralised and longer; goods came from deports already customised to the needs of local stores, with additional suppliers from local areas to top-up or to satisfy local demand niches. Furthermore, ‘cool/chilled chains’ meant fresh produce could be transported with quality unimpaired. Primary wholesalers have become less dependent on sales of produce and are more important as handlers, checkers and physical distribution of produce, especially for multiples and large retail mixed businesses. Equally, some symbol groups and greengrocers with large chains often have arrangements with growers and importers and directly source significant proportion of their produce. Nevertheless, a vast majority of symbols, convenience stores and greengrocers buy indirectly at the horticultural wholesale market places. Also, large multiples and large retail chains may buy in the region of 5-10% from local wholesale market places for ‘top-up supplies’. Table 3.2 shows the extent to which distribution of fresh produce changed; in 1980 34% of fresh produce was direct and this increased to 50% in 1990. Table 3.3 shows the extent to which large multiples centralised their supplies; e.g. in 1991 Tesco had 97% of its grocery supplied through the centralised distribution system. Multiples increased share of fruit and vegetables is not as dramatic as that for food in general. Table A3.5 shows the proportion of fresh produce sales to general sales of different types of retailer, and multiples’ share rose from 3.6% to 6.5% in 1991.

(% of retail sales values)

Distribution

The overall effects can be seem from Table 3.4 which shows the degree in which a large multiple achieved real economies of scale, scope and speed; operating margins improved from 66 under 15,000 sq. ft store size to 114 to over 25,000 sq. ft store size. Wrigley (1987:1286) notes that large multiples typically generated no less than two thirds of their annual increases in sales from the new stores which they have opened in that year.

Margins by Size of Store

However, the strategy of new store expansion has been recently revised in two ways. First, in 1993-4 central government revised regional policy guidance on retailing and transport and gave local authorities more discretion to protect and enhance the vitality, variety and sustainability of town centre retailing offered by smaller and specialist retailers.

Second, the capital market has become concerned by the inflated land prices of retail sites as multiples bid against each other for the best location. Further, recent warnings and fall in profits and increase in sales turnover has marked the beginning of the end of multiples as ‘glamour’ stocks (The Independent 1994). With the areas of new store expansion intermeshing (e.g. Sainsbury expanding northwards from its core south and Asda expanding southwards from its core north), with competition for key sites becoming increasingly intense, and with saturation levels for grocery superstores being approached, it has compounded into the capital market thinking that multiples can no longer achieve annual levels of turnover and profits on the basis of new store expansion programmes, and multiples are looking into other ways to sustain past level of performance (e.g. diversification into out-of-town D.I.Y., in-store new services like post office, medicine and books, and strategic buys of regional small multiples and overseas retail chains).

A convenience store is defined as one that stays open seven days a week and has extended hours of trading, sells a wide range of products in addition to ‘core’ grocery items, and is under 3,000 sq. ft. in sales area. Convenience stores are therefore much smaller than large multiples and their main role