Trapped in decline : a sociological analysis of economic life in Mgeta, Uluguru Mountains Tanzania

The research for this thesis was carried out in Tanzania during the period 1985- 89 and focuses on the Mgeta division in the Uluguru mountains, Morogoro rural district. Research was also undertaken among migrants from the area living in Dar es Salaam where they traded in foodstuffs. I made a return visit to the area in November 1991 to look again at questions which arose during writing up. The research also reflects seven years' employment at the University of Dar es Salaam (1982-89). The Tanzanian economy changed drastically during these years. Attempts to regulate the economy through government intervention were prevalent at the beginning of this period; at the end, the Tanzanian economy was to a large extent a free trade economy. This change resulted undoubtedly to a large extent from internal forces, but external pressures from the World Bank and the International Monetary Fund also played an important role. The dominant mode of thinking in these institutions is a belief in the free market: if market mechanisms are freed from the shackles of control, then rational decision making will lead to optimal fulfilment of society's needs. Such forces were, in their opinion, severely disturbed in the Tanzanian economy because of government intervention, and this reasoning was supported by academics who documented, for example, how government's parasitic behaviour suppressed peasant initiatives to produce.

The examination of government intervention was thus considered crucial to any understanding of rural Tanzania. The objective of the present research was then to make a detailed study in a rural area of how such government intervention influenced local society. An actor-oriented approach was adopted, society being seen as structured by actors. An actor-oriented approach was expected to uncover creative responses to macro-economic influences and government policies. The idea was to avoid bias towards an assumption of passivity on the part of the citizen, since the farmer is easily seen as a victim of government policies.

Economic life in Mgeta, as well as among those people who had migrated from the area to Dar es Salaam, appeared in the course of the research to be characterised by stagnation. The appropriate word to describe agriculture in Mgeta is decline. Mgeta has become increasingly a food importing area, and the opportunities to raise cash through trade and through the cultivation of crops for sale are limited. Nevertheless, petty trade pervades socio-economic life because of people's search for cash. Petty trade is the most common way of making a living among migrants from Mgeta in Dar es Salaam. Their trading careers are characterised by frequent bankruptcies, and accumulation is very limited and brittle. This pattern of socio-economic life cannot be explained by government intervention; the role of government in people's life is, as a rule, marginal.

Socio-economic life in Mgeta is better understood as socially constructed by local actors than as shaped by outside forces. This particular society is to a large extent atomised: enduring, strong and multistranded social relations are absent. The construction of bonds of solidarity is troublesome and often fails. For example, people avoid credit as much as possible since it requires dependable social relationships. If they use credit it is on a short term basis linked to one particular transaction. This example illustrates that social practices shape the basic elements of economic analysis: scarcity, preferences and utility. Such utility preferences are thus constructed by actors, but in turn constrain behaviour. Credit provides another illustration: credit is scarce because people do not trust each other. The resulting scarcity of capital limits the possibilities for entrepreneurship. An epistemological critique of a particular type of economistic reasoning emerges from such analysis. This broadens into a wider critique of rational choice approaches.

Rational choice theories assume a simple, linear and direct relationship between an actor's interpretation of reality and actual social practice. The relationship between socially-formed discourse on economic life and social practice is a major theme of the study. The research material shows that these relations can be much more complicated than rational choice theory assumes, and there appears to be a range of possible relationships between culture and behaviour. Actor-oriented approaches will therefore have to avoid the pitfall of assuming, a priori, a knowledgeable and capable actor who acts straightforwardly according to perceived strategic interests.

The results of the research are reported in six interconnected scientific articles.

'The continuing trial of development economics: policies, prices and output in Tanzanian agriculture' in Journal of International Development Vol. 5, forthcoming, 1993.

This article deals in general terms with Tanzania's rural areas. The research in Mgeta was a source of inspiration for writing the article, and the article makes it possible to see these research results against a wider background. It is an epistemological critique of certain kinds of economistic reasoning. It resulted from critically following the debates on the Tanzanian economy in the nineteen eighties and is based on secondary literature, official statistics and newspaper cuttings.

Decline in agricultural production in Tanzania is usually explained in terms of macro-economic variables and parasitic government intervention. These analyses use nationally aggregated figures in long time-series. Regional disaggregation of these figures and attention to erratic scores in time-series falsify such centrist explanations. A great variety of regional reactions to centrist influences appears possible. Production initiatives appear as characteristic for Tanzanian agriculture as decline. There are also other reasons for assuming that external influences are less linear and more ambiguous than the literature assumes, especially with respect to government intervention., The last part of the article hypothesizes and searches for alternative explanations which leave more room for the forces from below and for differential changes in agricultural production. Such alternatives see elements of analysis as used in economics - like scarcity, supply, demand as socially constructed in a framework of struggles between men and women, old and young, rich and poor in particular settings.

'Agricultural decline in Tanzania: the case of the Uluguru mountains' in African Affairs (91, 1992; pp.73-94).

This article illustrates the thesis that changes in patterns of agricultural production have to be understood from below. It also provides the background to the succeeding articles, which describe more detailed patterns of action.

The empirical material in this article was collected through a simple household survey of neighbourhoods in different parts of Mgeta. The survey collected primarily demographic material. The population structure of Mgeta appeared unbalanced in comparison with the national population, and even in comparison with Morogoro Rural District of which it is a part. This imbalance is caused by the emigration of young adults, especially to Dar es Salaam, in response to pressure on land because of population growth, erosion and depletion of the soil.

Mgeta is increasingly encapsulated in wider economic networks. A major driving force in this is the increasing need of households to buy maize - their staple food. The only significant local source of cash is the cultivation of, and trade in, vegetables. Vegetable cultivation is only possible, however, if one has access to suitable land, which is scarce. Many households are therefore dependent on income which is remitted to them from outside. This spread of cash usage has not resulted in a differentiation into rich and poor in the area: the pattern can more accurately be characterised as a general spread of poverty without significant accumulation. Agriculture has thus become a reserve activity, connected with economic life outside the area.

In contrast to the population structure in Morogoro Rural District as a whole, Mgeta's population structure is also unbalanced because of a female surplus. Women are said to guard the land there; land ownership is often bitterly contested. Claims on land are often made on the grounds of matrilineal descent, which is central to the dominant kinship ideology.

Earlier descriptions of social life in Mgeta portrayed social organisation on the basis of matrilineal descent; uxorilocal marriage was the norm; and authority in these matrilineal residential groups was described as being in the hands of a male head elected by the women. Matrilineal descent and uxorilocal marriage are still important, but the position of lineage head has disappeared. Social life appears in important respects to be atomised. Marriage bonds are weak and unstable. Many households consist of a single woman and children. Pubescent children are especially important as labour.

Marriage within Mgeta is not attractive for men from an economic perspective. The strong matrilineal ideology and the prevalence of uxorilocality denies a man claims on land belonging to his wife's clan or a house in his wife's neighbourhood in the event of the death of, or divorce from, his partner. The woman may have, given the kinship ideology and the scarcity of land, easier access to land in this situation, but this land does not guarantee a prosperous existence since it is often eroded and exhausted. Contacts outside Mgeta are much more important because they give access to people with a cash income. If a household does not have land on which to cultivate vegetables and does not succeed in establishing reliable access to cash from outside the area, life in Mgeta can be life in a poverty trap.
 

This particular pattern of economic life is described primarily as constructed by local actors rather than as a result of either government intervention or incorporation into wider processes of modernisation or commoditisation. There is no evidence that government plays a significant role in shaping people's lives. Elements of modernisation or commoditisation are found in socio-economic life. The eclectic adoption by people in Mgeta of parts of such models of behaviour indicates an active construction of socio-economic life. This construction has to be explained in the first place as a reaction to pressure on land, because actors cannot avoid this fact of life. Pressure on land is common in other mountainous areas in Tanzania, and a comparison with reactions there shows that the response in Mgeta is distinct. There are thus many possible reactions to pressure on land, illustrating again that socio-economic life in the area takes its form as a construction by local actors.

'Waluguru traders in Dar es Salaam; an analysis of the social construction of economic life' in African Affairs (91, 1992; pp. 181-205).

This article is the result of research among Dar es Salaam traders who originate from Mgeta. These traders have a dominant, but not an exclusive, position in the trade there in foodstuffs, especially vegetables.

A number of traders were asked to keep a diary in which they were asked to record their social contacts. These diaries were discussed every two weeks. The discussions revealed a clear classification of trade: trade between Dar es Salaam and Mgeta; illegal street selling in Dar es Salaam; trading from legalised market stalls in Dar es Salaam; wholesale trade between Dar es Salaam and areas upcountry; brokerage of supply spontaneously offered from outside, and demand in Dar es Salaam. This is a hierarchical classification, but only a minority climb this ladder and many fall back after making some progress. Economic life among these traders is very insecure, largely because their trade is in perishables, and price movements can be unexpected. This insecurity can also be seen as resulting from the way social life is structured. For example, their products, especially vegetables, can come from various parts of Tanzania and there is no information on the probable supply. Traders could organise networks which would provide reliable information about supply, but they do not. Bankruptcies due to sudden price fluctuations are frequent, but these fluctuations never hit the whole trade. Only those who at a given point have bought dear and must sell cheap suffer. Traders could insure each other therefore through mutual arrangements, but this does not happen. They travel with large supplies of cash, thus making themselves targets for robbers. The use of credit would reduce the need to carry large amounts of cash. These traders prefer, however, to avoid credit as much as possible. If credit is given then it is for a short duration and tied to one particular transaction.

This is one aspect of a general trait in their economic life. Social networks in trade are unstable, limited in content and of short duration. Traders prefer to cope with risk in their economic activities in ways which do not demand cooperation with others. If they are successful, they will move into less perishable goods which are less risky but demand more capital. Their favourite investment, however, is real estate in Dar es Salaam. If they manage to rent out houses or rooms then they have ensured themselves of an income which will continue if they are ill or if they are unlucky in business. This makes them also independent of Mgeta. As stated above, existence on the land is precarious.

In this manner, the article describes how social organisation is a formative influence on economic life. Such social organisation is, in its turn, a result of social practices structured by actors. The article illustrates thus that supply, demand, investment decisions and the like are not the outcome of economic laws operating outside time and place, but have to be understood as socially constructed and embedded in the ways in which a particular society operates.

'The arbitrary state: state legal arenas and land disputes in Mgeta, Uluguru mountains, Tanzania' in The Journal of Modern African Studies (31,3, 1993, pp. 1 - 1 g).

It is striking in the two articles summarised above that government does not play a direct role in the construction of economic life in Mgeta. There is, however, one field of life where government is of great importance: the judiciary.

Land is scarce in Mgeta and it is therefore not surprising that there are many land conflicts. It is possible to settle such conflicts in local councils for reconciliation, but many are brought before the Primary Court in Mgeta. Disputes can even reach the highest courts in the country. The virulent nature of these conflicts contrasts with a local culture which stresses the avoidance of open conflict. This culture also permeates the court procedures: motives, circumstantial evidence and the history of the conflict before it reaches the court are not discussed. The procedure is, therefore, extremely formal and concentrates on material facts. It is considered a disaster if people become involved in such court cases. Not only is it a threat to property which may be a source of livelihood, but it also costs much in terms of time, money and energy. In addition, recourse to the law does not guarantee resolution of the conflict. Judicial procedures take an extraordinarily long time; judgements concern themselves mostly with procedural considerations that are not essential to the conflict. Preferably, judgements avoid a decision and do not lay down principles which could lead to the formation of precedents. To get caught up in such a court case over land is to enter an arbitrary universe.

The description of a series of seven court cases, all concerning the same land, illustrates this. The first case was brought before Mgeta Primary Court in 1986 and it came before the High Court in Dar es Salaam in 1989. In the meantime, a large number of divergent judgements were produced by the Primary Court in Mgeta and in appeal by the District Magistrate in Morogoro. Reasons given in the judgements do not refer to one another and often no reasons are given at all. In the end, the case foundered at the Supreme Court for administrative reasons: one file was lost.

The behaviour of litigants in such cases is therefore not easy to rationalise. Both parties seem to have an unambiguous interest in keeping the matter out of the government's arena of justice, but that does not happen. These conflicts cannot be comprehended as conflicts between 'modern' and 'traditional' because both parties usually formulate their claims in terms of the matrilineal ideology of the Waluguru. Nor are these conflicts to be understood as straight-forward conflicts between rich and poor. There is little such class differentiation in Mgeta; these conflicts cost a lot of money and, in the event of winning, the gains are in no way commensurate with these costs. Often these conflicts end because parties have outspent themselves. Government cannot be portrayed here as beiing in the service of a particular dominant economic group; neither can it be shown as being the guardian of a capitalist economic order as is claimed in Marxist doctrine on the autonomy of the state. These conflicts should not be seen however as a mere aberration, unique in time and place. Simmel's sociology has aptly described the nature of similar conflicts. Conflict can become an end in itself when the reason for conflict is no longer the disputed object (in this case land) but the mere fact that someone else possesses it. The parties in dispute will then try to legitimise themselves by appealing to a third party. The latter can then without restraint play one off against the other, like the judiciary does here, as the object of dispute is lost from sight.

'Legal insecurity and land conflicts in Mgeta, Uluguru mountains, Tanzania' in Africa (63,2, 1993, pp. 197-218).

This article analyzes why land conflicts in Mgeta take such a virulent form leading to involvement in court cases which give no prospect of resolution of the conflict. The basic premiss is that all reality is subject to social definition and such social definition has a legal aspect. No set of abstract rules can foresee the situations which will crop up during social change. It is therefore inherent in law that new situations have to be interpreted in the light of existing rules. Society can more or less succeed in this process, which has to be carried by a binding authority to be effective. The article, through the analysis of land cases which came before Mgeta Primary Court, describes how and why society in Mgeta does not succeed in this with respect to land conflicts.

Virtually all litigants in Mgeta legitimate their claims in terms of the matrilineal ideology. Land is in the possession of individuals, but they do not own it. They have residual property rights. It is supposed to belong to clans, or segments thereof, which are structured by descent in the female line. Sons can, however, inherit land from their mother, and these sons can pass this land on to their children. If the latter die, then the ideology maintains that the land has to return to the original matrilineal clan.

People refer to this ideological construction during disputes, but it only partially structures social practice. Memories of the past are often vague and there are no written sources. The past in such a situation is primarily a social construction and structural amnesia therein plays a large role. If all vague memories of the past were considered as valid sources of law then extreme legal insecurity would result. Land can easily pass out of the control of particular clans, e.g., through inheritance from the father; or land can pass into the hands of a spouse after divorce or death. This gives rise to situations wherein the hegemonic ideology is no longer applicable. Land in dispute can have a clan identity which differs from that of the parties in dispute.

Land is seldom claimed on purely individual title; claims are usually legitimised by larger groups, and parties present themselves as representatives of such groups. Such groups are, in their turn, often again social constructions. It can happen that people claim solidarity in corporate groups in ways that do not fit the matrilineal ideology.

These legal conflicts do not, therefore, originate from laws promulgated from above, but result from social practices formed in local society.

Such a situation is not surprising. Discrepancies will emerge in every society between ideology, which may be legally defined, and social practices which are continuously formed and reformed. The legal aspect of the social definition of reality can, given these discrepancies, create the definitions of situations within which behaviour is acceptable. This can preclude the situation in which social life is unpredictable because of arbitrary behaviour. What is needed is a binding authority which may be vested in the state, or in institutions outside the state, or it may be manifested by group consensus. There is a vacuum of such authority in Mgeta.

Most land conflicts originate from the distribution of an inheritance. Such matters are usually settled at a beer party, forty days after the death. There, an executor is appointed, usually chosen from among the matrilineal relatives of the deceased. This can give legal security when the dispute concerns the estate of somebody who is recently deceased, but many land conflicts refer back to a very distant past. The executor, who is generally an older person, is often by then deceased himself as well. Reconstruction of the past in such a situation becomes a struggle to mobilise as strong as possible a group which can dictate a dominant interpretation of the past.

The use of written documents seems to be a way out in such situations. The court accepts these as valid evidence only if all parties agree to have been witness to their writing. If there is conflict, then this is of course denied. These documents are also extremely unsophisticated, and it is easy to make the accusation that they have been fabricated.

Earlier descriptions of social life in Mgeta mention the clan or lineage elder as the source of authority in land conflicts. This position, however, either had disappeared or must have been a reification in earlier work on the subject. The mother's brother (mjomba) is another source of authority in the matrilineal ideology. This is a clearly designated person in many matrilineal societies, but in Mgeta the term is interpreted so widely that it can apply to every male person related in the female line. Hence, a large number of people can claim this position, and parties in dispute can search for a mjomba who supports them. The courts recognise some people as legitimate representatives of the larger corporate group and others they do not recognise. The courts avoid, however, defining rules designating the authorities within these kinship groups who can make authoritative decisions about land.

In these land cases, the past has to be reconstructed in situations where relevant witnesses have died; written evidence is, for one reason or another, generally unacceptable; and there is no establishment of definitive authorities who can give a binding declaration. This results in great legal insecurity. A recourse to individual ownership of property through buying and selling land would seem to be a means of ensuring more legal security. That is not the case however. Such a transaction is only considered valid in law if representatives of the clan have given permission. This again begs the question as to who are the legitimate representatives of the group, and potential malcontents within this large group can challenge the sale at any point. The use of written documents gives rise to the same problems as mentioned above. Land can be held on individual title as a result of a cash transaction, but, in the event of it being inherited, its legal status can once again change and property rights are then again dispersed within large corporate groups. This situation cannot be called one in which there are dual legal systems, but legal systems intermingle with social practices. Individual title gives thus no protection in this situation.

These conflicts cannot, therefore, be explained in terms of the nature of the legal rules in use. There is no 'cultural lag' where new legislation based on individual title clashes with so-called 'traditional' law. Social change does not necessitate such a change in law, a fact also borne out in a survey of the literature. A multitude of legal forms governing land ownership are compatible with social change. The situation in Mgeta is the result of a failure within society to set limits to the situations in which claims can be made. Law then no longer controls the powers of envy, and these wasteful, long-lasting conflicts become a threat in daily life.

'Life chances, life worlds and a rural future; life expectations of school leavers in Mgeta, Uluguru mountains, Tanzania' (submitted for publication).

The focus of research shifted over time increasingly to factors internal to Mgeta. External factors like government intervention or commoditisation did not provide adequate explanation for the particular trajectory of social change in the area. Attention shifted more and more to the way in which society was structured by actors. It was therefore a logical step to study how actors interpreted the world, assuming that this would structure their behaviour and consequently shape the particular pattern of social change in the area. That appeared to be a simplistic assumption.

The article mainly discusses the responses to a questionnaire administered to school children just before leaving school. They were asked to give judgements about the future they preferred; the future they expected; and the expected remunerations of these possible futures. They were given options such as agriculture, further education, wage labour in town, trade in town and trade between Mgeta and Dar es Salaam. There were slight differences in the responses to these different questions, but the major pattern was the same. The most highly valued existence was that of agriculture, narrowly followed by further education, while trade received a low valuation in all categories.

This is surprising because it runs contrary to social practices in the area. Agriculture is a declining sector, and prospects are not good because of land shortage, erosion and soil exhaustion. There is massive emigration to the urban areas where trade is the migrants' main way of making a living. Very few of these pupils have any chance of further eduction. Consequently, further education is simply not an option for them.

The school leavers were also asked to rank a number of professions according to prestige. The professions were selected on the criterion that they were present in their environment. The answers were remarkably consistent with the values expressed in their expectations for the future. Overwhelmingly most prestige was awarded to farmer and teacher. (The latter is hardly surprising in a classroom situation.) Trade was very lowly valued. There was no clear judgement on professions such as village secretary, medical assistant and priest. Images appeared to be vague in the options other than trade and agriculture.

The questionnaire contained also a number of statements on issues in local discourse, and pupils were asked to indicate their agreement/disagreement in terms of five options: very true, true, do not know, untrue, a lie. Further education was valued highly, independently of the possible advantages it gives with respect to professional status or money. The image of agriculture appeared to be less unambiguous, but it was striking, for example, that a majority denied that there was a problem of land shortage in Mgeta. Agriculture was, however, clearly much more highly valued than trade. Trade was considered to be an insecure existence that provided an income at the expense of farmers.

These answers can, of course, merely show that the school is successful in ideological indoctrination. Therefore, an attempt was made, using more qualitative methods in a follow-up visit, to see whether these values and images existed also among wider segments of society. That appeared to be very much the case. Either problems such as land shortage, erosion and depletion of the soil were denied, or the output of science (mostly fertilizer) was expected to bring the solution. Government should then disseminate these good things. These interpretations had an outspoken social character and were vocally articulated in public performances. Social practices in Mgeta, which did not fit this ideology, were obscured through such mechanisms.

This last article is exploratory. Many questions remain about the ways in which lifeworlds influence behaviour when there is a discrepancy between culture and social practices. It shows unequivocally however that sociological explanations need to allow room for discrepancies. For example, it is a simplification to reduce the differential outcomes of social change to results of intended and unintended consequences of action based on differential social interpretations of the world. If people and societies are caught in interpretations of the world which are in discord with their social practices, then such a linear perspective on thought, interpretation of reality and social practice does not make much sense.

A direct correspondence between culture and social practice is inherent in the idea of the rational actor: socio-economic behaviour is thereby seen as structured by actors who calculate within a means-goals paradigm based on preferences and expected returns given market relationships of supply and demand. Such interpretations can easily be made after the fact: the returns from agriculture in Mgeta are meagre and therefore men opt for making a living from trade in town. This interpretation ignores crucial elements of the local culture and is at variance with the interpretation of behaviour from the point of view of the actor. It results in reification: observers project their ideas on a society because of the assumption of a rational actor.

The rational actor is central in the image of the market as found in the policies of the International Monetary Fund and the World Bank. Actors can act rationally, but they do not automatically do so. Irrationality is caused by external factors, notably government intervention, in the mindset of the IMF and the World Bank. The evidence presented in this thesis does not agree with their image of rational people who are caught in the irrationalities of government intervention. Economic behaviour cannot be understood as following immanent and impersonal laws of rational choice that should not be disturbed; rather, scarcity, utility and preference are shaped by social practices within particular cultures. The idea of the rational actor operating according to market forces can best be seen as an ethnocentric social projection which blinds the mind to cultural diversity.