CAN I SELL ONE OF MY COWS? INSTITUTIONS ASSETS AND GENDER BASED POVERTY

This paper examines the linkages between institutions, assets and the existing gender differentials in levels of production and poverty in peasant economies in Northern Ghana from an institutional economics perspective. Quantitative and qualitative empirical data were collected from eight study communities in four major ethnic groups in the Upper East Region of Ghana. The results from the analysis show that the institutionalised age old norms regarding the ownership, access and control 0/ assets have affected the degree to which men and women in traditional societies actually own and control critical productive assets like land, cattle, irrigation water and even their own labour. Women were found to have minimal control over these assets, since the accepted norm is that the male head 0/ household has ultimate control over these assets. The direct consequences of these institutionalised ownership and control norms are the low levels of production of women and the relative poverty of women in the region. An outstanding merit of the study is the unravelling of the complex relationships between institutional set-ups, assets, production and poverty. These complexities point to the fact that the issue of poverty in Northern Ghana and how it can be addressed is a complex subject and one will need to appreciate the institutional context. The findings also provide useful in/ormation to both policy makers and practitioner on how to address the complex issues related to gender and development.


Introduction
It has been argued that the ownership and control of assets, especially productive assets, has a significant effect on the level of production and consequently on poverty (IFAD, 2001). "Lack of assets is an effect as well as a cause of poverty in terms of income opportunities, consumption and capacity building of people ... " (ibid). In all societies, assets are unevenly distributed among different social and economic groups. The glaring differences in the levels of both wealth and poverty can be attributed largely to the central issue of the ownership and control of productive assets. In peasant economies, the prevailing cultural tenets with regards to ownership and control tend 2 Senior Lecturer in the Dept. of Planning, Land Economy and Rural Development 25 to affect the assets portfolio of individuals within the household, and consequently their ability to produce. This aspect of the effects of the traditional set-ups in Northern Ghana has not been adequately documented. An understanding of these relationships will inform both policy and practice in the fashioning of development interventions. This paper focuses on the prevailing institutions, i.e. cultural norms, taboos and rules regarding the issue of ownership and control of productive assets and how these affect gender based production and poverty.

Conceptual framework
The major concepts employed in this study are discussed within the frame of reference of institutional economics.
Integrative social research today can no longer rely on narrow subject based theories to address the rather complex and intricately intertwined social, economic, politicaland environmental issues and their context specific nature. In the realm of planning today, the institutional economics theory provides a broad and integrative perspective (Nabli & Nugget, 1989). The integrative nature of the theory makes it possible for different lines of investigation to intersect. Essentially, the institutional economics theory is not a monolithic theory per sc. It "encompasses a wide range of issues, which take into consideration the social aspects, as well as the economic aspects of human phenomenon" (Bacho, 2001: 50). The issue of how socially crafted institutions affect gender based ownership, access and control of assets is more of a social character, albeit not wholly. The ownership, access and control of assets and the concomitant effects on gender based production and poverty straddles between the social and economic realms. One cannot, therefore, analyse adequately these intertwined issues from a purely economic perspective or from an institutional perspective. One key tenet of the institutional economics theory, which is of relevance to this study, is that institutions affect economic activities just as economic activities affect institutions. The key concepts: institutions, assets, production and poverty are discussed in the frame of the institutional economics theory.

Institutions
Institutions are nothing but the collectively agreed upon social arrangements that govern the interactions among members of a given group of people. This is often referred to as the rules of the game (Olson, 1965;Hardin, 1977). Institutions can evolve overtime, thus reflecting the time and place experiences of the group (Olson, 1965;Smith, 1968;Ostrom, 1990). These types of institutions originate in small groups, such as kinship groups; informal groups linked by some interest who tend to interact frequently. On the other hand, institutions can also be deliberately created based on a specific social, economic, political and physical context of the given group. Pejovick (1995: 31) defines such formally created institutions as: ... constitutions, statutes, common laws and governmental regulations which are externally enforced. They define the political system, (hierarchical structure, decision-making powers, the individual rights), the economic system (property rights in scarce resources, contracts) and protection system (judiciary, police, military).
This concept of institution has a wider and more uniform applicability, applied usually in national or supra national context (Bacho, 2001). In reality, evolved institutions in small groups may, with time, develop into the formal model presented by Pejovick (ibid). The type of institution not withstanding, all institutions have some common functions. All institutions help members of the group in the internalization process; maintenance of the group structure and relations; definition of expectations, rights and obligations; minimization of conflicts, transaction cost, free rider problems; and in the regulation of behaviours as they affect the every day interactions in the group as well as censorship of deviants. Individual actions or inactions are then judged either legitimate or illegitimate within the given institutional context. Through a process of internalization, the individual becomes acquainted with what is the established order. The internalization process results in a fatalistic acceptance of the group norms as given. Shifts of this position can only be brought about by either a strong internal revolt or an external force or influence. Group pressure, by way of rewards and sanctions, ensures individual compliance. The fear of sanctions, especially sanctions by the divine supernatural forces· or the spirits of the ancestors compels compliance. Institutions therefore, have a controlling effect on the individual. In most traditional societies in Africa, it is still a commonly accepted view that men are the decision makers. It is argued in this paper that the institutionalization of gender based assets ownership and control explains the gender differences in levels of production and poverty in Northern Ghana.
A central issue in the on-going debate on the co-coordinating functions of institutions and production is that of incentives. How do institutions either motivate or de-motivate individual performance? Institutions establish assets ownership, but very often "no connection is made between the ownership of resources and the motivation that will result" (Leibenstein, 1984, 75). One glaring example is institutionalized socialism which addressed ownership and coordination problems without addressing the individual motivation problem.

Assets
The term asset is used in this paper to connote any valuable productive resource, which "can be used without being used up" (IFAD,200 1: 72). In the context of peasant economies, assets can vary widely from farmlands, water resources, bullocks, donkeys and horses, structures such as a barn, sty, kraal and farm houses and even skills and health. All these endowments can be acquired or inherited. Assets can also be owned either individually or collectively. In farming communities in rural Africa, it is common knowledge that individuals can own certain categories of livestock such as poultry and small ruminants, but may not own water bodies or grazing land. Assets can be acquired over time through savings and investments, such as in the acquisition of skills through time invested in education. On the other hand, one can also lose assets over time through damage or simply depreciation such as in the case of a hoe or a plough, which wears out through continuous use. It is also common knowledge that through the predatory activities of others such as fraud, theft or excessive draws by a large number of dependants, one can also be dispossessed of his assets.
The critical importance of assets lies in the fact that the individual or collective group can use them for wealth creation, as source of security against risk, or convert them into cash or other valuable assets in times of need. In peasant households, cattle constitute an important source of security just as they are used for productive purposes. Cattle can also be converted into other valuable assets such as a com mill or house. The extent to which an individual benefits from an existing stock of assets depends largely on the issue of ownership, access and control. Ownership, access and control are defined by the prevailing time-place institutions crafted by the given group.
Ownership or property rights ensure the security of assets and rights of access (Acheson, 1995;North, 1990; Pejovick opt. cit.). These are necessary for the stable and sustained use of assets for productive purposes. In the case of individuals for instance, "access to land through ownership or secure tenure is the sine qua non of improving agricultural productivity" (FAO, 1998: 9). Without secure tenure, land cannot be used as collateral to access credit from formal institutions. Ownership or property rights also act as a check on the misuse of common pool resources. Without the appropriate institutions, a communal asset like a small-scale dam for dry season irrigation will encounter excessive free-rider problems like excessive usage of water resources without the commensurate contribution (Hardin, 1977, Bituro, 2000. Similarly, without institutions, a bullock plough that belongs to a poor man can be forcibly expropriated from him and thereby jeopardizing his otherwise stable livelihood system. In reality, no one person can acquire and own all the assets necessary for productive purposes. This raises the issue of access. Assets can be accessed either through rent, hire, or employing one's influence or authority. For example, the village chief, by virtue of his authority and influence, can access a tractor service from the rich tenant farmer residing in his community. One may also own or have the right of use of an asset, yet may not be able to access it when it is most needed depending on who ultimately controls the asset. In rural communities in Northern Ghana, agricultural land is communally owned (Kasanga, 1988) yet access to good farmlands by poor families and women is problematic. In most of traditional African societies, the use of family or clan assets must be done with the permission of the head of the family or clan. While this serves as a check against possible abuses, it could also impede production. In some ethnic groups in Africa, women own substantial assets like cattle, but may not dispose of them directly without the consent of their husbands. This theme is central to the issues raised in this paper where age old traditional institutions continue to define rights of ownership, access and ultimate control of productive assets such as land, livestock and labour time along gender lines (Songsore & Denkabe, 1995).

Production
One salient feature of peasant economies is the mixed bundle of small scale productive activities such as subsistence crop and livestock farming, small scale non farm economic activities such as handicrafts, metal works and extractive industries like small scale surface mining, charcoal burning, stone quarrying and sand wining. All these economic activities depend on the asset portfolio of the individual and/or farm family. The level of production is very much affected by the type and quality of productive assets. one is able to access. "An owner occupied agricultural land offers better opportunity for a stable and sustainable level of output than probably a hired land of the same size" (IFAD, opt. cit). The sustainability of the production system is low since there will not be a strong incentive for the tenant farmer to invest in long term land conservation practices (Spring, 1994).
Another critical aspect is the gender based differential access to productive assets. In most traditional societies, women are allotted marginal lands for cropping. The problem is further compounded by the low access to inputs such as bullock plough services, extension services, fertilizers and pesticides. These consequently affect their levels of production.

Poverty
The perceptions, dimensions and causes of poverty have not only changed overtime but also differ from one socio-cultural group to another. Poverty is often perceived as a multidimensional phenomenon, which translates into one's inability to meet adequately his/her basic survival needs like clothing, shelter, food, health, and education among others. In a sense, all these translate into the concept of assets based poverty. Assets based poverty constitutes an important thread in the continuing academic debate on poverty.
The perceptions and causes of poverty have changed dramatically overtime since the time of Malthus. Over the years, the discussions on poverty have also become more and more prominent among academics, policy makers and practitioners alike, albeit the wide range of views. The Malthusian theory perceives, rather simplistically, the cause of poverty essentially as an issue of resource constraint under increasing population or demand (Ahaoet. aI.,. 1998). The role of institutions and assets distribution was, therefore missed in his theory. Similarly, the nineteenth century economists argued that poverty is caused by low capital (assets) accumulation and can be eradicated through technological progress that enhances the assets of both individuals and the state (ibid). It was argued that both improved physical and human capital enhance productivity and consequently reduces poverty. Having well trained human resources not only stimulates individual initiatives but also increases productivity, which have positive effects on poverty reduction (Shultz, 1981; IFAD, 200 1). Here again, the institutional dimension is missed. While recognizing the role of skills as a productive asset, society at large has tended to ignore the yawning gender gap in the prevailing formal educational and training systems, especially in sub-Saharan Africa where these gender differentials in access to skills based education are partly the result of the culturally determined investment decision making regarding whether a girl or boy should go to school.
The rights and opportunities to access both physical and human assets contribute to the types and levels of production and consequently on the nature, dimensions and magnitude of poverty ·among different socio-economic groups and regions. One can also talk of institutionalized poverty, whereby socially defined norms, rules, regulations and laws minimize or exclude some segments of society such as minority groups from the ownership, access, control and/or use of productive assets. The case of settler farmers not owning land on a permanent basis in SOmeparts of Ghana is one case that hinders productivity (Kasanga, 1988;Spring, 1994). The institutionalized share cropping system, whereby the tenant farmer takes two-thirds of his/her own output, renders him/her perpetually poor (Spring, ibid).
Traditional gender roles define what type of crop a man and a woman should cultivate. In most ethnic groups in Ghana, men tend to cultivate cash crops, major cereals and root crops while women cultivate vegetables and legumes mostly for household consumption. A survey of root and tuber crops farmers in the Northern Region of Ghana revealed that yam was particularly a "male" crop (Anamoh & Bacho, 1995). This accepted gender based farming system of peasants in northern Ghana explains the relative poverty of rural women.
While assets-base and institutionalized poverty may be prominent, inherited poverty, (that is when one is borne into poverty) is a wide spread phenomenon. A person born into a destitute family, environmentally fragile environment or protracted periods of widespread catastrophe, economic malaise or socio-political upheaval is at risk of becoming poor.
The review shows the conceptual difficulties in isolating a few variables to establish causal relationships between the three key concepts. Each of the key concepts can be both a cause and an effect of the other two.

Research approach and data collection
It is argued here that the relationship between institutions, assets, production and gender-based poverty is one of circular causation. The types and nature of the institutional set-ups determine the ownership, access and control of assets, just as the very distribution of assets empowers different groups of people to influence institutional set-ups. Assets determine one's productive capacity and output levels, and therefore one's degree of vulnerability. Poverty is, thus both a cause and an effect of the ownership, access and control of assets. Ones ability to access the needed productive assets is crucial to avert the risk of being poor. Similarly, poverty with its multiple dimensions reduces one's productive capacity physically, psychologically, socially and financially. These perceived relationships informed the research approach.
The study, therefore, adopted a triangulatory approach to collect a combination of both qualitative and quantitative data. The study was conducted in two major interrelated phases. In phase one, qualitative data was collected using different interactive tools comprising mainly, focus group discussions, community fora and key informant interviews. The rationale of this approach in the first phase was to enable the investigator gain insights into the specific perceptions and meanings of the issues under investigation in order to progressively understand and appreciate the complex nature of the issues. This stage of the research focused on the institutional issues such as the context .specific norms, beliefs and taboos governing gender based ownership, access and control of assets, decision-making processes and gender roles with regard to production, and the control and use of produce. Data on the perceptions and the manifestations of poverty and how poverty affects different social groups was also gathered through these interactive processes.
The second phase of the investigation focused on, essentially, the quantitative data. The emphasis was on collecting data that will facilitate quantitative measurements. A questionnaire was developed and administered to sampled respondents from the four' ethnic groups in the Upper East Region -Kasena-Nankana, Grune, Kusasi and Builsa. The main issues covered include: household assets and distribution, sources of income and expenditure, types of economic activities by gender, production levels, sizes of holdings, quantities produced; and division of labour time. A sample of eight .(8) communities, two from each of the four major ethnic groups, was randomly selected. These eight communities formed automatically the first stage of the sampling process. The second stage was the selection of respondents. A total of 146 women and 143 men were sampled and interviewed, using trained interviewers who had direct contacts with 33 the respondents. This approach yielded 100 percent coverage of the intended sample of respondents.

Presentation and discussion of findings
In this section, the findings based on the empirical data collected are presented and discussed. This is done under three main themes: the effect of institutions on the ownership, access and control of assets; the relationship between ownership, access and control of productive assets and production; and consequently how the two interact to affect gender based poverty.
Data from the preliminary interactive studies show that in all the four ethnic groups the categories of productive assets that playa vital role in 'the production process of the peasant households are land, livestock, farm/implements (like hoes, axes and ploughs), water, and labour time. The focus of this section, therefore, is on these four categories of productive assets -land, livestock, water and labour time.

Institutions, assets and gender differentials in production
One major asset island. Land generally was observed to be scarce in the eight (8) study communities. The average land comprising both compound and bush lands, available to a farm household, vary widely from one community to another. In the Zanlerigu and Dorongo communities of the Grune, average holding is 4.3 and 3.4 acres respectively. Households in the relatively land rich area in Kayoro-Wuru, Nyangnia and Kunkwak of the Kasena-Nankana and Builsa respectively, have much higher average holdings of 7.5, 6.8 and 7.6 respectively. There is, indeed, a wide variation in the sizes of land available for household use in the Upper East Region. In those communities with acute land shortage, the problem of access, acquisition, ownership and control of land becomes even more critical, despite the culturally established procedures.
The available data also shows that among all the four ethnic groups, except for minor details, the land acquisition procedure is the same. Traditionally, the first settler in each community is regarded as the landlord (Tindana among the Grune and Kusasi, Tengyono among 34 the Builsa and Tigatu among the Kasena). The Tindana or Tigatu allocates lands to each family. or individual. Land, so allocated, does not imply a transfer of property rights. The family or individual can use the land for as long as that family or individual requires it, but it still belongs to the community. After communal lands have been allocated to families, other subsequent settlers, usually from the same clans or any person wishing to acquire land will then have to approach the family or individual whose portion of land helshe intends to use to request for it. If the family or individual agrees to assist, the Tigatu, Tengyono or Tindana, as the case be, is approached for the formal allocation of the land for the specified period of use. This procedure described above is also similar to what pertains among some other ethnic groups like the Dagaaba and Dagomba. This general procedure not withstanding, the culturally accepted procedure for the acquisition of communal land by women is different. If a woman wants to acquire her own piece of farmland, she may have to do so through her husband, a male relative or guarantor as the case may be. In the case of family land, the woman will need to ask her husband or family head for a parcel, usually to crop vegetables for both household consumption and sale. The reasons for these institutionalized procedures are varied but one common rationalization is that women are usually regarded as "strangers" in the clan or family they have married to and can leave at any time in the event of a divorce. This procedure is, therefore, regarded as a check against the possible loss of land and conflicts since the husband or male guarantor will be held accountable in the event that the woman leaves the community. This customary procedure places some impediments on the way of women farmers since they do not have direct access to land except through their husbands or a male guarantor. Out of the 146 women farmers interviewed 36.9 percent farm on land that is not owned by them or their families. These women followed the procedure outlined above to acquire the land for temporary use.
One other dimension of ownership, access and control is the established procedure for sharing family land. Family land is usually sub-divided by the head of family and allocated to the males in the family. In polygamous families, where a woman hasno male child, her chances of access to the family land are slimmer especially when her husband dies. Even in the monogamous families where the widow has no son the brother or male relative of the deceased is entitled to take over the farm lands of their deceased brother, since by the patrilineal inheritance system they are entitled to both the property and even the wife of their deceased brother or relative as the case may be. They are also by custom responsible for the upkeep of the widow and the off spring of their deceased brother. This explains why the widow will not directly inherit the lands of her deceased husband. Another reason is that the widow could. remarry outside the clan and that will then mean a loss of family land. These institutionalized patrilineal inheritance systems hinder women's access and control of the family farmlands, especially after the husband's death, albeit the intended safety guards.
The procedure described above shows that communal property rights still remains the underlying guiding principle for land ownership and management. The culturally accepted procedures, whereby women will still need to go through their husbands or male guarantors to acquire land, remains an undue encumbrance as will be seen in the ensuing sections.
These general patterns not withstanding, available data revealed that in the study area land is becoming a highly valued commodity and, therefore, attracting high unofficial rents from landlords. This means access to land is increasingly dependent on the ability to pay. This emerging trend is likely to aggravate the plight of peasant women. The diminishing status of the Tindana/Tigatu vis-a-vis that of the chiefs and Assemblypersons is beginning to pose a challenge to the traditional land acquisition procedures.
Increasingly, some chiefs ·have arrogated the right of land allocation to themselves as was observed in Basyonde, where the chief intimated he allocates land. The elders, however, protested against the chiefs claim thus underscoring the emerging conflicts in land issues. Women are apprehensive about these emerging trends with regard to the issue of who controls land and the rising costs since these will worsen their plight in terms of access.
On the whole, it was found that women were allocated marginal lands that are usually no longer used for the cultivation of high value crops like the major staples by their husbands, household heads or the individuals from whom they have acquired the land. Where the land is fertile, the sizes of holdings tended to be comparably smaller as shown in Table 1.1 below.   Traditionally women do not cultivate major staples like millet and sorghum, which are used mainly for household consumption and other customary purposes like funeral performance and rituals. Increasingly, however, women now cultivate these major staples on their own. These are mostly widows, women whose husbands have either migrated, are incapacitated or irresponsible.
Another interesting point worth noting is the level of productivity. On the whole, the level of productivity with regard to the production of major staples is higher for men than the women. These major staples are cropped in the compound farms, which are more fertile because of the continuous use of organic manure such as animal droppings, household refuse and lately compost. These compound farms are controlled by the head of household, who invariably are men. Where a woman, for some reason wishes to cultivate her own staples, she is allotted the marginal fringes of the compound farms away from the fertile zones. The reason is that the compound farms provide better opportunities for cropping the major staples like millet, sorghum and legumes for the household's food requirements. The men view this as their responsibility to see to it that these requirements are met to avert starvation. From the point of view of the men, therefore, they are justified in allotting the marginal fringes to the women to enable them satisfy their individual wishes.
On the other hand, the productivity of women, for crops like rice and vegetables, are high. In the case of rice the reason is the intensive care which leads to effective weed control and water management.
From the data presented and analyzed here, one can draw the conclusion that one of the salient causes of low levels of output and productivity of women farmers is the institutionalized land ownership, access and control, which tends to look at the women as external to the land owning community and can, therefore, not be allocated land of their own on a permanent basis. Secondly, the institutionalized gender roles, which recognize the man as the bread winner legitimizes the practice of allocating the fertile portions of the land, specially compound farms to men as heads of household.

Livestock ownership and control
The data shows that among all the four ethnic groups' livestock rearing is an important component of the livelihood systems of the household. Livestock is kept for a variety of reasons. These were categorized into: household food security, direct and indirect investment, and the fulfilment of social obligations. Livestock, therefore, constitutes one of the valuable assets and serves as a basis for social differentiation. The ownership, access and control of livestock in all the four ethnic groups are, therefore, important issues.
Livestock rearing, just like in every sphere of economic activity in a largely peasant economy is affected by gender role perceptions. Among the four study ethnic groups, livestock ownership is, traditionally, vested in the head of households.
So, even though women and children may keep and own livestock they must do so with the consent of their head of household. Where women keep and own livestock, tradition requires that they place them in the care of their husbands or their grown up sons or the appropriate male head of household. Except for strictly female-headed households, married women will usually inform their husbands before using their livestock for whatever purposes. This cultural norm implies that the women do not have absolute control over their livestock since they need the consent of their husbands before they use it for whatever purpose. The evidence given by women in Basyonde community confirms this (See Box 1.1).
Although we women can own cattle, tradition demands that we inform our husbands or who ever is the head of household before we can sell one or use it for· any other purpose. Usually it is the man who drives the cattle to the market and sells on behalf of his wife. As intimated by one of the women during the focus' group discussions, "you need to go to the man and ask: CAN I SELL ONE OF MY COWS?" (Focus Group session with women in Basyonde -April 2002.) This traditional norm regarding resource ownership has both positive and negative consequences. Where harmonious relationships exist in the household, it can lead to a productive use of the combined stock of the household. On the other hand, an autocratic husband or male head of household can squander the livestock, and worsen the plight of the woman, where there is no harmony.
Women, especially the older generation, have accepted these norms regarding livestock ownership and control as normal. The reasons women have given vary, although they all centre on two cultural tenets: the idea that the man is the head of the household, and also through the dowry system, the woman is perceived as equally the "property" of the man. During the focus group discussions with women in the eight sampled communities, one often hears the statement "the man owns me and my property" thus confirming the strong culturally ingrained belief that the man is the ultimate owner of all that is in the household. These are reflected in the responses in Table 1.2 below. Respect for husband 03 The man owns the house 04 Men feed women 05 Women usually acquire animals in their marital homes 06 The man may need the animal for other uses The women were also of the view that culturally, the ultimate , responsibility of feeding the household, meeting the health needs of the family, educating children, meeting social obligations and providing other essentials like the spiritual well being through soothsaying and sacrifices to the gods rest squarely on the man. Consequently, they found it appropriate for men to use available household resources to fulfil such obligations. In the event where household member is taken ill, the head of household will need to consult the soothsayer who will prescribe the types and number of livestock to use to make the necessary sacrifices to appease the offended god that is causing the illness or problem. It is also the responsibility of the head of household to perform such other social responsibilities as dowering the wives of the younger males in the house and performing the funeral rites of dead members of the household as well as supporting that of his in-laws. All these multiple responsibilities require some livestock. The household head by custom has the right to dispose of any livestock to meet any of the varied needs as shown in Table 1.3. As can be seen from Table 1.3, the most important reasons for taking the woman's livestock are to purchase extra food to meet the household requirements and meeting funeral expenses. It was found that households experience severe food shortages during the months of February to June, when the meagre harvest is exhausted. At such times the sale of livestock becomes inevitable to avert starvation. Funeral rites also form a critical component of the household expenditure among all the four ethnic groups. The inability to perform a funeral is regarded as a disgrace to the family. It is also mandatory for married men to support their in-laws to perform the funerals of dead relatives, which will usually require the donation of animals, foodstuff and other gifts.
Another dimension of the gender differences on livestock ownership is the type of livestock kept by men and women. The survey results show that on the whole cattle and donkeys are kept mostly by men. This explains why women rarely own bullocks and donkeys. The reasons given for this trend are that cattle and donkeys are expensive and generally beyond the means of women, and secondly these two categories of animals are generally viewed as serving purposes closely linked with the traditional roles of men. Cattle are traditionally used for dowries, funerals, sacrifices to the gods, household insurance, a form of savings. Traditionally, women own cattle under the shadows of their husbands, sons or male head of household. These institutionalized norms with regards to the livestock ownership patterns explain why women have low access to bullocks and donkey ploughs for farming purposes. Women usually rely on their husbands' ploughs or hire the services of some other peoples ploughs. In the absence of the above alternatives their farm work is done physically as shown in Table 1.6.
These cultural norms regarding livestock ownership not withstanding, there is growing gender awareness. The survey results also show that a few married women who own livestock use them without informing their husbands. This is a reflection of the gradual breakdown of the cultural barriers regarding female ownership and control of agricultural produce, especially livestock rearing, which , was primarily regarded as a male dominant farming activity.
Apart from the gender differences, it was also found that livestock rearing is not aimed primarily at economic profitability but rather social security. A variety of livestock are kept to meet the nexus of socio-cultural demands. They also insure the household against unforeseen risks and the chronic seasonal food shortage.

Access to productive water
Que to the annual long spells of dry seasons, spanning usually the months of November to June, dry season gardening is a supplementary source of household income and food in the study area. In those communities in the Upper East Region with well-developed dams/dugouts with accompanying irrigation facilities, irrigation farming is an important activity. Those other communities without such facilities rely on natural sources of water like shallow hand dug wells; waste water from bore holes, streams and rivers. Water, especially in the dry season is an important asset in this fragile environment. Beside Dorongo which has an established small scale irrigation scheme the other seven communities use either waste water from boreholes, shallow -hand dug wells, natural streams and rivers. In all cases, access to such agricultural water resource by women is lower than men as reflected in Table 1.3. The percentage of women farmers who gain some access to productive water ranges from a low of 19.4 percent in Sarabongo to 38.0 percent in Kayoro where the existence of natural sources from the Sissili River and its numerous streams provide avenues for vegetable cultivation during the dry season.
Developing water sources for dry season gardening is capital intensive. It also requires a lot of labour. Besides, once the system is developed the inputs required, for dry season gardening are prohibitive, and beyond the means of most rural women. The survey information reveals that an average expenditure on farm inputs and labour is about 22.3 percent of total annual expenditure of households.

Labour time
In labour intensive economies a critical asset is labour time. The amount of time allotted to productive economic activities as against accomplishing routine household chores and other social obligations affects one's level of income. Some critical issues were observed from the data on labour time. The first is that women spent an inordinate proportion of their labour time on routine household chores. Secondly women have definite roles in the household production system. Although no activity was specifically limited to either sex among the four ethnic groups especially in households without the supporting spouse, there are certain activities that are generally regarded as men or women roles as shown in Table 104. The gender role differentiation was found to be based on the following three (3) culturally determined criteria: physical ability, skill, care and diligence, and household stability/security. These three issues are reflected well in Table 1.4. Land preparation which involves activities such as slashing, digging and ridging are physically taxing tasks and are therefore undertaken by men. On the other hand, processing activities such as winnowing, milling and cooking, using traditional technologies, are activities that also require skill, patience and sometimes a great deal of endurance and are therefore undertaken by women who are said to have such qualities. The issues of storage and control of the major cereals is done mostly by men to ensure that the household does not ran out of food before the next harvest. The rationale is that since women are in close touch with their children and do the cooking, if they are allowed to control the food the tendency will be to overuse the food to satisfy their children. It is also argued that, if women are made to know the food situation of the household, they may ran away when they get to know the household has ran out of food. This rationalisation aside, there is a high risk with the sole management of food by men since they could also sell out the food for their personal use and drink.
These socially accepted gender roles in the production system of peasant households demand that women divide their labour time between working on the family farm and their own farms. This is not the case for men, who do not usually work on the women's farms, as women usually farm "non male crops". The complaint by Basyonde women in Box 1.2 is a reflection of the general situation in the study area.
Box 1.2: Complaint of Basyonde women "Even though we are marginalizedwhen it comes to ownership ofland and other properties, we do a lot Of farm work. We work along side with our husbands on their farms -weeding, sowing, and harvesting. We also do all the work on our own farms single handily. By the time we finish the work on the men's farms our own farms are overgrown with weed". The complaint of the Basyonde women applies well to women in the other study communities. Despite their complaints they find it culturally acceptable to invest their labour time in the generation of collective resources for the upkeep of the household. Here, one can see clearly the role of institutions in controlling the behaviour of women. They are compelled to accept the laid down rules although they know they do not benefit much by complying.

Conclusion
The study results show that in the peasant economies of the Upper East Region of Ghana, like in other parts of northern Ghana, land and livestock constitute major assets that have an important bearing on the livelihood system of households. Given their critical role in the livelihood system, elaborate culturally determined institutions have been crafted to govern the ownership, access and control of these assets. It was found that in the main, household heads, usually men, have ultimate control over these assets, although ownership rules appear rather liberal to the extent that even children can in principle own especially livestock. To the extend that the ultimate decision as regards use will have to be sanctioned by the male head of household, access in the real sense remains encumbered. The systematic socialization process has resulted in the acceptance of the status .quo, as the views often openly expressed by women during the study demonstrate the recognition and acceptance of the fact that it is socially unacceptable for a woman or a younger member of the household disposing of a livestock without informing the head of household or a woman acquiring a piece of land by approaching the Tigatu, Tengyono. Herein lies the critical role of institutions in determining the critical issue of women's real access, control and use of important productive assets necessary to liberate them from their poverty.
The study also reveals an obvious disparity between the production levels of men and women leading consequently to the relative poverty of women in the eight study communities. What these findings point. to is that among the multifaceted causes of poverty in the study region the socially determined ownership, access and control of productive assets contribute in a significant way. These findings are also confirming the theoretical position that institutions do have definite impacts on assets distribution especially with regards to ownership, access and control, and ultimately on production and poverty levels among different social groups, in this case women.
The above findings have profound implications for policy and practice. Both development practitioners and policy makers have, over the years, regarded gender based poverty as a simple issue of the lack of assets and the means to acquire these essential productive assets. The policy thrust has been towards providing vulnerable groups, especially women with these assets. The proliferation of income generation activities and credit schemes testify to this. Not until recently, when women activist embarked on advocacy for gender equally, governments have done little about influencing the institutional framework that determines how assets are distributed 46 among different social groups. If women are to gain equal access to vital productive assets, one will need to examine the institutional question and design appropriate policies to effect changes. The education of women is one such viable strategy that can be employed to effect institutional changes. Besides, the introduction of appropriate laws on property ownership will protect vulnerable groups, especially women against possible abuses. Development practitioners on the ground can also embark on vigorous education and awareness creation. One needs, however, to conclude on a cautionary note that social issues are difficult to change overnight, hence the need for a gradual approach.