Migrants, urban poverty and the changing nature of urban–rural linkages in Kenya

Abstract

In the literature on rural–urban linkages in sub-Saharan Africa the focus has so far been predominantly on the urban dwellers contributing to the livelihood of the rural ones, usually through remittances from family members living in the city. Although acknowledged in the last two decades, the reverse flow, i.e. the extent to which town dwellers realise part of their livelihood from rural sources, remains poorly understood. Based on recent research in Nakuru town, Kenya, this paper demonstrates that urban–rural linkages are not only important for the rural households, but are becoming an important element of the livelihood (or survival) strategies of poor urban households.

¹Lecturer, Department of Geography & Environmental Studies, University of Nairobi.

1. INTRODUCTION

One response to increased urban poverty in Africa involves the strengthening and adaptation of the urban–rural linkages that have always been such an important part of urbanisation processes on the continent (Potts & Mutambirwa, 1990). Many urban households have rural components to their livelihoods and retain strong links with rural areas, while some keep part of their asset base in rural areas (Foeken & Owuor, 2001; Owuor, 2003). These combined urban and rural residences and livelihoods have been called ‘multi-spatial livelihoods’ by Foeken & Owuor (2001). Rural livelihood sources accessed by urban households are embedded in the linkages, interaction and reciprocity that are evident between them and their rural household members, homes or areas.

It is common for urban Kenyans to identify themselves with an ‘urban house’ and a ‘rural home’, which partly explains why the majority are never permanent residents in towns. A rural home is normally the ancestral land that is passed on from father to son. Traditionally, the daughter does not qualify to inherit her father's ancestral land, because she is expected to get married and make her ‘home’ with her husband's family. The urban migrant who identifies him/herself with a rural home is a well-established phenomenon of African migration (Oucho, 1996; Francis, 2000). In Kenya, as elsewhere in sub-Saharan Africa, the urban bias of development is symbiotic with the migrants' rural bias toward home – a place to visit periodically, to which they will retire and where eventually they will be buried (Owuor, 2004, 2005).

According to Tacoli (2002), rural–urban interactions include spatial linkages – flows of people, of goods, of money and other social transactions between towns and countryside and sectoral interactions – between ‘urban’ sectors in rural areas (e.g. rural non-farm employment) and ‘rural’ sectors in urban areas (e.g. urban agriculture). Through these interactions or linkages, households increasingly rely on both rural- and urban-based resources for their livelihoods – that is, many households straddle the city and village for their livelihoods (Satterthwaite & Tacoli, 2002).

Although urban dwellers have always maintained links with the rural areas, economic crisis and structural adjustment in the last two decades seem to have produced fundamental and interrelated changes in urban–rural linkages:

African urban residents have long maintained strong social and economic links with their rural ‘home’ areas, although the nature of those links has varied over time, as the nature of migration streams has adapted to changing economic and political circumstances, and from country to country with variations in factors such as colonial policy, urban history, and land tenure and land availability. The recent era of severe economic decline and structural adjustment has seen such linkages assume a new and vital significance. (Potts, 1997: 449)

First, new forms of migration have emerged or old ones have intensified and others have slowed (Tacoli, 1997). Research in the 1990s indicated that the rate of rural–urban migration was decreasing, while return migration, i.e. from the city to the rural ‘home’, was emerging (Tripp, 1996; Potts, 1997; Baker, 1997; Tacoli, 1998; Okali et al., 2001). Circular migration between urban and rural areas was also increasing (Smit, 1998). A review of recent empirical evidence on migration and urbanisation in francophone West Africa suggests that economic crisis may increase circular migration between towns and villages (Beauchemin & Bocquier, 2003). While acknowledging that urban out-migration is not a new phenomenon, Beauchemin and Bocquier (2003:10) argue that:

It seems to be increasing in importance. In addition to the traditional return flows of migrants, a new kind of urban-to-rural migration, linked to economic crisis, has appeared … since the early 1980s. The job market degradation and the deterioration of the standard of living created new relationships between migration, employment and education. In the past, people moved to town to attend school or to find a job. Today, the opposite is quite frequent. A large number of people who have been fired from formal sector jobs return to the villages. In addition, some urban residents with jobs, confronting their incomes to the urban cost of living, choose to return to rural areas where incomes are lower but where food and housing are almost free.

Secondly, rural links have become vital safety valves and welfare options for urban people who are very vulnerable to economic fluctuations (Gugler, 1991; Potts, 1997; Smit, 1998; Frayne, 2004). There is evidence of a significant shift in the nature of transfers of goods and cash between urban and rural households, in the sense that remittances from urban to rural areas are declining (Bah et al., 2003) and transfers of food from rural to urban areas are increasing, not only in amount and frequency but also in importance:

It appears that far more food is now being brought in from rural areas, which of course greatly enhances urban residents' vested interests in maintaining their social and economic rural links. These transfers can rely on surpluses generated by existing rural kin or on urban residents returning in the rainy season to cultivate, which they would probably not choose to do if they could afford urban food prices or could gain access to sufficient land in town to grow food. (Potts, 1997:466)

Urban households with limited social connections to rural areas are the most vulnerable to hunger. In contrast, those with active rural–urban linkages enjoy significant transfers of food from rural areas that offset hunger and vulnerability in the urban context (Frayne, 2004). For the large majority of urban dwellers, food from rural areas is important for the household's food security situation (Mulinge & Jayne, 1994; Mwangi, 1995; Tripp, 1996; Krüger, 1998; Muzvidziwa, 2001; Frayne, 2004). For others, rural produce (when sold) is an additional source of household income (Potts & Mutambirwa, 1990; Rakodi, 1995; Fall, 1998).

While migrants continue to rely on food from their rural homes, there is some evidence that remittances to rural areas are declining as urban dwellers find it harder to spare any money (Potts, 1997:466). The decline in remittances in amount and real terms is a consequence of the increasing employment insecurity and the cost of living in town. Greater economic hardship and other important aspects such as the increasing cost of education, health care, housing and food have led to migrants sending less in remittances than before. Despite the decline in urban to rural remittances, social links between migrants and their rural home areas remain as strong as ever (Tacoli, 2002). For many migrants, this is not only a part of their social identity but also a way of spreading their assets (and risk) across space and maintaining a safety net which helps in times of economic and social insecurity in cities (Bah et al., 2003:20).

Thirdly, to reduce household expenses, a husband may return his wife and all or some of the children to the village while he remains in town (Rakodi, 1995; Potts, 1997; Frayne, 2004). Also observed is a pattern where the wife goes to the rural areas for a substantial period during the main agricultural season and visits at other times of the year for shorter periods (Rakodi, 1995). With the wife returning to the rural home to engage in farming and children returning to attend school in the village, the family can face economic hardships better (Beauchemin & Bocquier, 2003) and at the same time the rural base is kept as a ‘safe haven’ (Bigsten, 1996).

Similarly, young people unable to find jobs in town may return to the rural home out of choice or be sent there by their parents. Fostering urban children at the rural home is also common among female-headed households. For example, Muzvidziwa (2001) and Nelson (2001) found that female-headed households in Zimbabwe and Kenya sent their children to stay at the rural home as a cost-cutting measure. The high costs of education in urban schools may push parents to return children to rural areas where schools and other related expenses are relatively cheap (Potts, 1997). Another reason for leaving family members behind in rural areas is a lack of housing in town (Potts & Mutambirwa, 1990). For Beall (2002), such arrangements support the argument that urban households' livelihood strategies cannot be seen in isolation from their wider context. In that case, the practice illustrates the role played by rural families in helping to reduce vulnerability for urban households (Frayne, 2004). Access to rural assets is therefore, according to Krüger (1998), at least a supplementary if not essential element for securing and stabilising the livelihood systems of many vulnerable urban households.

This article explores the hypothesis that with the present trends of economic hardships, increased cost of living and urban poverty, ‘falling back’ on rural areas and specifically reliance on rural food and income sources is, or has become, increasingly important in the livelihood of many migrants in sub-Saharan African cities. The article examines the nature, extent and direction of urban–rural linkages, including (1) social reciprocity between urban and rural areas, (2) flows of money and goods between urban and rural areas and (3) return migration from urban to rural areas.

2. SOCIAL RECIPROCITY BETWEEN URBAN AND RURAL AREAS

This article uses survey results and the in-depth case studies carried out in Nakuru town, Kenya, between 2001 and 2003. The first phase (in 2001) was a general survey of 344 households, using a standardised precoded questionnaire. The second phase (in 2002) consisted of in-depth interviews or case studies with 16 households drawn from the initial sample. Of these 16 households, five were further selected for the third phase – the rural visits (in 2002/2003) – a continuation of the in-depth studies, but at the respondents' rural homes.

Social reciprocity (or interactions) between urban and rural areas can be analysed within the context of regular visits that occur between urban-based and rural-based members. In the last quarter of 2001, during the general survey, all respondents were asked if the household head and/or spouse had visited their rural areas; 80 per cent had done so. Table 1 presents a summary of the characteristics of visits to the rural areas by Nakuru town households.

Table 1: Visits to the rural home or plot in the last quarter of 2001 (%)

Purpose of visit∗	See/visit rural-based family or relatives	71.1
	Attend to rural farming activities	55.9
	Attend to cultural ceremonies	29.5
	‘Holiday’ from town life	24.9
Number of visits	Less than 5 visits	54.7
	5–9 visits	16.0
	10 + visits	29.2
Frequency of visit	More often than monthly	17.5
	Monthly	21.2
	Every two to four months	37.8
	Once/twice a year	23.5
Source: 2001 survey (N = 349 plots).
∗Total > 100% due to combined answers.

There were various reasons for visiting the rural homes or plots. Typically of kinship and family ties, urban dwellers are obliged to occasionally visit their family members and relatives who live in the rural areas. Essentially, these visits are meant to maintain and foster kinship and family relations. It is for these reasons that about three-quarters of the plots were visited by the Nakuru town households ‘to see or visit’ the rural household or family members and relatives (Table 1). For both migrants and non-migrants, and especially the urban poor, this is not only part of their social identity but also an important safety net during periods of economic and social insecurity in the cities. So strong is this bond that urban households continue to visit the rural areas regardless of the distance from Nakuru and in spite of high transport costs.

For more than half the respondents, the second reason for visiting rural areas is to attend to farming activities. From time to time the household head and/or spouse visited the rural plot to supervise or participate in these activities. The third reason is to attend family functions and events such as weddings and funerals. Lastly, it is a practice for some townspeople to visit the rural areas for their annual leave, school holidays or at long weekends and on public holidays just to get out of the town environment.

The frequency of urban households' visits to the rural areas differs from household to household and is likely to be influenced by factors such as the distance to the rural area, the purpose of visit and the relationship of the household head to the rural-based household and family members. On average half the plots had been visited fewer than five times, while another quarter had experienced at least ten regular visits (Table 1). About two-fifths of the plots were visited on a monthly basis or more and another two-fifths every two to four months, while the rest were visited only once or twice a year.

The nearer the rural home or plot, the higher the interaction. Those who farm in the rural area also visit the rural plots more frequently than those who do not. Urban households practising rural farming are likely to frequent their rural plots to oversee and sometimes participate in farming activities. The frequency of visiting rural plots is also higher when the wife is living at the rural home (or at the plot). This is logical because the family bond between the urban and the rural parts of the household is maintained through such frequent visits. For example, one respondent explained that: ‘Since my wife and children went to live at home, I visit them any time I get an opportunity and there is no question about it. I cannot live in Nakuru town for more than two or three months without seeing them.’

The number of days spent in the rural areas per visit varied from one to seven days and, in a couple of cases, a month. Owing to ‘commitments in town’, the preferred length of stay in the rural areas for any given visit was between one and seven days. Even then, many spent only one or two days, preferably at weekends and on public holidays. Longer visits are made during annual leave, school holidays or at the Christmas and New Year breaks. The Christmas holidays are traditionally the time when people living in town return to the rural home to celebrate with their relatives. One middle-aged male respondent maintains his parents' tradition of going to the rural home in December:

Even though my father lived and worked in Nakuru town, it was tradition for us to go home once a year in December. My parents liked celebrating Christmas at home. I do the same even now. I prefer going home with my family in December.

That does not, however, prevent his wife from going to their rural home more than once in a year:

I go home more often than my husband. I prefer going home in April, August and obviously in December, all during school holidays. In many instances, my husband sends me to supervise and help in farming and to find out how the people at home are doing. I do not have a definite number of days I stay at home. It can be one day or two weeks, depending on what has taken me there.

Even though Nakuru townspeople wish to visit their rural areas as often as possible, they are limited by the high travel costs, especially for those whose rural plots are far from Nakuru. For half the households the frequency of visiting rural areas has decreased with time, for about a quarter there has been no change, and for another sixth the frequency has increased. People can no longer afford to make as many trips to their rural homes or plots as they did comfortably before. To reduce transport costs, some households no longer travel to the rural home with the entire family, except for once in every one or two years. However, it is still common for both the husband and wife to attend a close relative's funeral.

Others are simply unable to afford the high transport costs to their rural home frequently, such as the respondent quoted below, who nevertheless continues to maintain links with his wife at home whenever an opportunity arises:

Until the 1990s, I visited my family at home regularly, thrice or four times a year. I can no longer afford that because transport costs have become unbearable and life is more expensive nowadays than it was. As much as I would have liked to go home more often than I do now, I am restrained by the high transport costs. However, I rely on relatives who often go home while my wife relies on relatives who are coming back to Nakuru from home. Through that, she is able to send me some food and I am also able to send her some money.

Even then, there are circumstances when the rural plot may be visited more frequently for a certain period of time during the year. This could be because of the illness of a family member or a close relative at home or when one is building a house at home.

The flow of visits is not only from Nakuru to the rural areas but also the other way round. Eighty-six per cent of the respondents confirmed that their rural-based family members or relatives do visit them in Nakuru for various reasons. The reasons for the reverse flow are quite similar to urban dwellers' reasons for visiting the rural areas. Again, the main reason is the normal ‘to-see-them’ visits between close relations. Besides these, rural family members and relatives may come to Nakuru to obtain medical attention, to collect farm inputs, for family ceremonies or ‘when in a problem’ that needs the urban household's attention. Some bring money from the sale of farm produce while others come to obtain money for purposes such as farming activities or school fees. Generally they bring food from the farm with them and in return they go back with purchased (food) items from Nakuru.

Even so, the rural-to-urban visits are nowadays not common because urban households can no longer afford to accommodate their relatives in town for a long time:

These days we do not just entertain those unplanned visits from any relative like long ago. We cannot afford to accommodate them for long even if they should come. Life in town is expensive. We prefer planned visits with a specific purpose.

This is not the case with multispatial households where the wife and children live at home. In such cases, rural-to-urban visits are quite similar to the urban-to-rural visits described above. The rural part of the household, i.e. wife and children, are bound to visit the urban household, i.e. the husband/father, on a regular basis.

3. FLOWS OF GOODS AND MONEY BETWEEN URBAN AND RURAL AREAS

3.1 Urban-to-rural flows

Generally, the flow of goods between urban and rural areas was not easy to capture because of its complexity, involving as it does a wide range of goods and varied quantities in space and time. What emerged generally from the survey results is that when urban household members visit their rural homes or plots they bring purchased food items, non-food items or money with them. Similarly, when the rural household members or rural-based relatives visit Nakuru townspeople, they go back with similar items. Other goods find their way to the rural areas for specific purposes; for example, presents (e.g. clothes), building materials, farm inputs, or items for funerals.

Besides the flow of goods, further questions were asked about the urban-to-rural flow of money. The results presented in Table 2 show that about three-quarters of Nakuru town households contributed financially to their rural household or family members. The frequency of sending money and the amount vary. For example, the frequency is likely to be higher for multispatial households and for rural farmers. Exchange of money also occurs when urban household members visit their rural plots: 77 per cent of respondents reported that they normally give money to their rural household members, parents or other close relatives on most of their visits to the rural areas.

Table 2: Urban-to-rural flow of money (%)

Remitting money to rural areas (N = 327 households)	Yes	73.1
	No	26.9
Frequency of sending money (N = 239 households)	More often than monthly	5.0
	Monthly	35.1
	Every two to four months	41.4
	Once or twice a year	18.4
Reasons for sending money (N = 239 households)∗	General upkeep	90.4
	Farming purposes	57.3
	Payment of school fees	23.8
Change in frequency of sending money (N = 239 households)	Decreased	54.8
	Increased	19.7
	No significant change	25.5
Source: 2001 survey.
∗Total > 100% due to combined answers.

The reasons for sending money to the rural areas were basically for general upkeep, i.e. to support the family at the rural home; what respondents called ‘money to buy sugar’. Money is also sent for farming purposes or for paying school fees (Table 2). On specific requests, money may be sent for a festivity, a funeral or for medical purposes.

Given the tension between rising need and falling wages, it is not surprising that over half the households in Nakuru send money back home less frequently. A quarter have not experienced any change, while for a fifth the frequency has somehow increased (Table 2). Although they still continue to send money back home, the decline in remitting money to the rural areas is being felt slightly more by low and very-low-income households.

Nevertheless, as part of the household, the husband in town will always endeavour to support the rural household members. According to an elderly grandmother in the rural area, the exchange of money and goods between parents at home and their children in town and/or between a husband in town and his wife at home is mutual. For that matter, according to her, nothing is changing:

My son has always been very good to us. Although life is expensive in town, he can still manage to send us something small every now and then.

3.2 Rural-to-urban flows

Besides the urban-to-rural flow of money and goods, both food and non-food items flow from rural to urban areas as well, through the urban household members, rural household members or other relatives. First, even though urban households send money back home or to the rural plots, there are indications that the reverse is also true, albeit sporadically. In other words, some urban households also receive money from the rural plots: 5 per cent of the household heads (and/or spouses) who visited the rural plots brought money back with them, while 3 per cent of the rural-based household members who visited their urban households brought some money with them from the rural home or plot.

Secondly, as part of the complex socio-economic ties, it is common to observe exchanges of food items occurring between the urban and the rural parts of the household or rural-based family members. More often than not, when urban household members visit their rural counterparts, they take with them purchased items such as salt, sugar, milk, bread, tea, cooking fat, and so on. In return, they are given food from the shamba (plot or farm) to bring back to town for consumption. The most common items are green maize, local vegetables, sweet potatoes, cassava, maize or millet flour, groundnuts, fruit and chickens. These are ideally regarded as not being commonly available to the town dweller. Exchange is common between parents living at the rural home and children who have migrated to towns. Similarly, when the rural-based members visited the urban household, they brought food from the shamba with them and went back with purchased food and non-food items from town.

However, to increase their food security in town, a large majority of households nowadays rely on their own rural production. Through regular visits to their rural homes or plots, urban dwellers are able to practise rural farming, directly or indirectly, through the cooperation of rural household members, rural-based family members or through hired farmhands. The results presented in Table 3 show that for the large majority rural farming is an additional food and income source for the households concerned. Rural crop cultivation and livestock keeping constitute an additional source of food for about two-thirds of the urban households practising it and an additional income source for more than half of them.

Table 3: Importance of rural farming activities (%)

	Crop cultivation (N = 173)	Livestock keeping (N = 111)
Could not survive without it	74.6	39.6
Major income source	16.2	2.7
Additional income source	57.2	54.1
Major food source	23.7	0.9
Additional food source	66.5	67.6
Source: 2001 survey.
Total > 100% due to combined answers.

The importance of rural farming is stressed further by the fact that three-quarters of the crop cultivators indicated that they ‘could not survive without it’, with it being a ‘major food source’ to a quarter of them. Livestock is an important source of food and income only in situations where need arises, and therefore acts as a form of social security. The following four respondents explained the importance of rural farming for their households:

On average, we used to spend about 150 shillings on food, almost on a daily basis. Since we started cultivating our own food, the food budget has been reduced to about half that amount. In addition, we get a little money when some of the produce is sold.

The maize and beans in this house come from the shamba at home. With a good harvest, I rarely buy maize and beans in this house. I can therefore use that money for other things. I could not have survived without this. Life has become so difficult.

Our plot has been of great importance to us. The money we got when we sold our produce helped us to pay for our children's school fees.

When I have a financial problem, I sell one or two of my cows at home. For example, I have sold a number of my cows to pay our children's school fees. To me, having cows at home is an asset and a form of security because you can sell one whenever you are in trouble.

Food from own rural production reaches the urban household through rural–urban linkages. Urban households are progressively more involved in rural farming as a source of food rather than relying on food gifts and are intertwined within the social reciprocity between the migrant and his or her relatives at the rural home. Although reciprocity still exists in various forms, this particular kind is shifting from merely receiving food gifts to procuring own food from rural areas. In addition, there are signs that people are buying food items cheaply from rural markets for consumption in town or sometimes even for business.

4. RETURN MIGRATION

In addition to the reciprocity and economic flows, there is emerging evidence of new forms of urban–rural linkages, namely where part of the urban household – especially the wife of the male head and/or children – is sent to the rural areas. Data from the general survey reveal that in about one-third of the married households the wife and children had stayed for some period of time in the rural areas. Moreover, at the time of the survey 39 households could be considered multispatial, i.e. where the wives of the male heads were actually living at the rural home, mostly together with their children. Some had lived at the rural home from almost immediately after their marriage, while others went to live at their rural homes at various stages of their marriage.

The primary reason for sending some of the urban household members to the rural areas is economic (Table 4). Some of the urban household members were sent to there because there was not enough money to support them in town and because education for the children was cheaper in the rural area. Living in the rural area, the wife could concentrate on farming as a way of producing food for the family.

Table 4: Reasons for sending wife and children to live at home (%)∗

	Wife (N = 89)	Children (N = 73)
Not enough money in town	29.2	38.4
To concentrate on farming	73.0	–
Problem of housing	15.7	15.1
Education is cheaper at home	–	43.8
Source: 2001 survey.
∗Only for those whose wife or children had ever stayed at home. The reasons in each case reflect combined answers.

After 12 years of town life, the following respondent was forced to go and live at home because life in town had become harder and harder:

My husband does not have enough income to support us in town. His carpentry business has not been doing well for the past few years. This has been his main income source since 1989 when he married me. Since we had access to the rural plot and given that life is relatively cheaper at home, we decided to go and live there. My husband is now living in Nakuru alone.

For another wife, when her husband retired from the civil service, it was the loss of a regular source of income that induced her to move to the rural home. Instead of living in town during her husband's retirement, she decided to go and live at their rural home in 1991 and to concentrate on farming. For the same economic reasons another respondent, who earns a monthly income of KSh4500, was unable to take care of his wife and five children in Nakuru, let alone fit everyone into his one-room house:

A few years after our marriage, I suggested that my wife live at home because my salary was, and still is, too small to live with them here. I had a house at home and there was land for her to cultivate. She normally visits me at least once a year and I also visit them whenever I can. Without her agreement about living at home, I could not have survived with all of them in Nakuru. She gets most of their food requirements from the shamba. I am therefore left to concentrate on looking for school fees and once in a while sending them something small for sugar and salt.

In a few cases, the reasons were basically sociocultural: to build a homestead, to get to know the rural home and the parents-in-law, to take care of sick parents or to fulfil traditional rites. Even when the children are not accompanied by their mother, it is common for them to be sent to the rural area to stay with their grandparents when the household experiences economic stress and limited food availability. Being unable to provide for their unemployed son, this respondent decided to send him to the rural area, albeit temporarily:

At the moment, our youngest son is living at home with his grandmother. He recently did a course in Nakuru but was not lucky enough to get a job. Being unemployed, there is not enough money to take care of him here in Nakuru. He is better off at home helping there with farming for the time being, as we look for something for him to do here in Nakuru.

Through this split-migration practice, ties between urban and rural areas are enhanced, while at the same time reducing the food demand of urban households. The burden of providing the daily food requirements for a large household where few are in gainful employment is immense. This strategy of ruralising the urban household therefore helps to minimise the measures of meeting the household's food shortages.

5. POVERTY AND INTERACTION

Social and economic interaction occurs between urban and rural areas regardless of income category, the only differences being in its frequency and intensity. Irrespective of income, more than half of the household heads and/or their wives had visited the rural plot or home during the last quarter of 2001. However, owing to financial constraints, the poor (defined as those with a monthly income of KSh5000 or less) were not able to visit their rural plots as regularly as the non-poor. For example, almost all (91 per cent) the household heads and/or spouses in the non-poor group (those with incomes over KSh10 000 per month) had visited at least one of their rural plots, compared to two-thirds of the poor category. Moreover, the frequency of visits to the rural plot or home was much higher among the non-poor.

There is also some difference between the poor and the non-poor in terms of the purpose of visits. Farming-related visits are comparatively more important for the non-poor, while meeting the rural relatives is more often the reason for undertaking the trip among the poor. It is through such ‘just-to-see-them’ visits that the poor are able to maintain strong links with the rural areas and kin, and therefore a place to fall back on ‘in case of anything in town’.

The same trend can be seen with remittances. Although town dwellers send money back home regardless of income, a larger proportion of the non-poor did so and also somewhat more frequently than the poor. Even then, the poor and non-poor showed no clear differences as far as the change in sending money to the rural home was concerned. About 60 per cent of respondents in both groups held the opinion that the frequency of sending money had decreased over time. The same holds for the perceived decrease in frequency of visits to the rural home or plot. However, because these interactions are complex, it is not easy to draw conclusions about the changes that may be taking place between ‘town’ and ‘home’. For example, it may very well be that it is not the frequency of sending money that matters but how much is sent.

Finally, there is some difference between the two income groups as to whether the wife had ever lived in the rural home. For the poor, this was the case with 50 per cent of the households, against 31 per cent of the non-poor. The poor households do this as a coping strategy, i.e. because of lack of income to support them in town.

6. CONCLUSION

The impossibility of living on typical urban wages in Africa has been dubbed the ‘wages puzzle’ by Jamal & Weeks (1993). That people do survive is testimony to their ingenuity, determination and sheer hard work. A host of coping strategies have been developed. One of these is falling back on rural sources. Although the types of interaction examined here, as well as their effects, should not be over-generalised, a pattern can be discerned which may be used to gain an understanding of the emerging relationship between urban households and their rural homes or plots. Household-level food transfers from rural to urban areas have become more important with increased urban poverty, while the transfer of money and goods from town to the rural areas has decreased in real value. The flow of food resources from the rural kin to urban households therefore contradicts the common assumption that rural areas or kin act only as recipients of goods and services originating from town.

Returning one's wife and children to live temporarily or permanently in the village seems to be a survival strategy with two benefits: (a) it makes living in town less expensive, and (b) more food can be obtained from the rural home. Consequently, the economic balance of urban–rural linkages appears to have shifted in favour of the urban households. Although it is difficult to ascertain trends because of the complex nature of the interactions, there are indications that urban–rural ties, which have always been a vital part of African migration processes, have become more important for urban households. Urban residents are looking to rural areas as a subsistence fallback. Maintaining both an urban and rural base provides a safety net, especially for the urban poor in times of economic hardships.

However, rising transport costs in recent years have reduced the frequency of home visits for many people, particularly when long distances are involved. In addition, the quantities of remittances and goods (or gifts) have declined, as the cost of living in urban centres has soared. This does not necessarily imply weakening bonds. For many people who move to the city, the countryside and their native village still remain a reference point, both culturally and in terms of family life. The linkages that persist between urban and rural households are central to urban households' ability to survive.

Lastly, many development theorists and practitioners have, until recently, viewed rural and urban areas as two mutually exclusive entities with their own unique populations, activities, problems and concerns. However, this does not reflect the realities of multispatial livelihoods, which include both urban and rural elements. Interactions between urban and rural areas play an important role in the processes of rural and urban change. It is essential that policies and programmes reflect the importance of the ‘urban’ part of rural development and the ‘rural’ part of urban development. In other words, urban development strategies must take account of the rural links and context. The answer to urban poverty cannot be found in the urban areas alone. Policies which neglect this may increase poverty and vulnerability for those groups.

The author wishes to thank the Netherlands Foundation for the Advancement of Tropical Research (WOTRO) for financial support, and the African Studies Centre, Leiden, the Netherlands, for granting him visiting fellow status during this study.

Notes

¹Lecturer, Department of Geography & Environmental Studies, University of Nairobi.