Household Differentials and the Individual Decision to Migrate to South Africa. The Case of Gweru City in Zimbabwe

Zimbabweans of all walks of life have crossed over to South Africa since the late 1990s for long and short periods of time, some of them even remain there permanently. The increased amount of the migration is largely blamed on Zimbabwe’s socio-economic and political instability by most scholars. However, the actual decision to migrate is reached as a result of pressure and motivation from within a household. The study investigates the influence of the household on an individual’s decision to migrate to South Africa. This study was carried out using a household survey of 150 households from different wealth quantiles from the central city of Gweru in Zimbabwe. Basic descriptives were employed to generate the findings. Only 2% of the sample indicated that they did not have individuals who had migrated to South Africa or any other country. About 43.7% of all migrants were females and 45.7% of the female migrants had children younger than five years old staying with relatives in Zimbabwe. Most households tend to have a strong influence on the migration decision, and as a result the majority of the migrants send remittances back home.


INTRODUCTION
International migration among African countries was always influenced by economic push and pull factors as a result of the relatively faster economic growth of some countries as compared to others.In every African region, there is an economic powerhouse that offers better paying jobs, education and health systems.Hence studying, working and living in such a country is desired by neighbouring countries' citizens.In West Africa such countries are the Ivory Coast together with Gabon (Adepoju, 2005;Adepoju, 2006) and in Southern Africa it is South Africa (Kok et al., 2006).Political, economic and social instability in countries of origin also increase the migration potential as people seek more secure lives elsewhere.
Since the nature and volume of the movement of people varies a lot, it becomes a challenge to provide a conclusive and suitable definition of 'migration'.Some simple translocations such as moving a few blocks of buildings from one's former place of residence to another cannot be considered to be migration, because the distance involved is too small.Also, visiting a place for a day cannot possibly be termed migration because an apt definition of this concept should include a considerable duration of stay as well (Kpedekpo, 1983;Young, 1994).This is interesting because the length of stay of most international migrants is brief, and 53% of the visits by migrants from Zimbabwe to South Africa were previously said to last for less than a month (McDonald et al., 2000).
As a result, in Southern Africa, Kok et al. (2006) identify three types of migration namely, refugee, labour and permanent change of residence, all of which can be either through legal or illegal channels.Other important components in defining migration include crossing a political boundary and staying there for more than a month (Kpedekpo, 1983).Migration is either temporary or permanent depending on the duration of stay after the first month has been completed (Young, 1994).Hence for international migration, one has to cross a national boundary and stay in the other country for over a month, upon which they become a 'migrant'.

Literature Review
Since the beginning of the mining industry in South Africa, from the second half of the nineteenth century onwards, most labour migration between Southern African countries was driven towards that country (Crush et al., 2005).The Employment Bureau of Africa (TEBA) began to regulate migration flows in the 1920s and mining and agricultural colonies in Southern Africa, including present-day Zimbabwe, Zambia, Namibia and Tanzanian began to attract migrants (Crush et al., 2005).Most of the migrants went directly to South Africa and by the 1970s, there were over 270 000 migrants in that country coming from different parts of the continent (Crush et al., 2005).
Zimbabwe's political, social and economic situation deteriorated in the late 1990s and worsened at the turn of the century making the citizens more and more uncomfortable with life.This triggered an outflow of asylum seekers, and numbers of economic and political refugees increased over the years (Amnesty International, 2008;Bloch, 2008).Though the actual numbers of Zimbabweans crossing into South Africa are hard to estimate, the numbers being deported because of illegal migration suggest that huge numbers are crossing the border.
In July 2007, monthly deportations of Zimbabweans had reached 17 000 and between 2008 and 2009 up to 300 000 illegal migrants were deported from South Africa (Polzer, 2009).As a result of this high mobility, even though studies have indicated that Zimbabweans in South Africa cannot exceed 1.5 million, the general populace believe that there could be three million (Polzer, 2009: 3).This near-exodus of Zimbabweans into South Africa has influenced various South African Government policies aimed at stopping the tide (Landau and Wa Kabwe-Segatti, 2009).It has also influenced the continued social unrest in South Africa in the form of nation-wide Xenophobic attacks that turned violent in May 2008 (Landau and Wa Kabwe Segatti, 2009).
The challenges caused by Zimbabwe's outmigration are felt both at the destination and at the source.Household disintegration may occur when couples stay apart for long periods of time and numbers of pseudochild-headed households increase as parents stay in foreign lands where children cannot join them (Dinat and Peberdy, 2007).Women are unlikely to be accompanied by their children along because they work long hours as housemaids, farm-hands or in factories where the accommodation provided is not conducive for children (Dinat and Peberdy, 2007).
In the case of tricky challenges that include life threatening illegal migration and leaving vulnerable children behind, Zimbabweans continue to migrate to South Africa.The reasons for choosing to migrate may be the presence of a useful contact at the destination, or the desire to acquire better educational and job opportunities.However, literature has shown evidence that the household has central role of sanctioning the migration decision before a migrant can travel (Zinyama, 2000).Both the well-to-do households as well as the poor households are sending migrants to South Africa, making it necessary to investigate how and why a migration decision is being made.The demographic characteristics of household members chosen to migrate in different types of households, as well as their migration purposes would shed light on methods of reducing migration flows.
It has also been highlighted earlier in this paper that the characteristics of the households from where migrants come are barely discussed in most literature.Yet if the 'new economics of migration decision theory' is anything to go by, the household should have an influence on the migration decision (Stark and Bloom, 1985).This makes it important to understand households that have sent migrants to other countries as a factor that can be used to address international migration flows in the region.
The objective of this paper is to establish the characteristics of the Zimbabwean households and their members and how these influence an individual's decision to migrate to South Africa.Specifically, the study provides a set of demographic characteristics of households with migrant members who have moved to South Africa and it also assesses the contribution of household networks in the country of destination by a migrant.

Theoretical Framework: New Economics of Migration Theory
This paper is premised on Oded Stark's "new economics of migration theory" (Stark and Levhari, 1982: 191), which proposes that "…migration decisions are not made by isolated individual actors, but by larger units of related people -typically families or households…" (Massey et al., 1993).The theory brought an interesting dimension to understanding the decision to migrate arguing that, "migration decisions are often made jointly by the migrant and by some group of non-migrants" (Stark and Bloom 1985: 174).
As a result, "decisions to migrate are not always made by isolated individual actors per se, but sometimes by larger units of related people such as families, households and even communities" (Mafukidze, 2006).The relevance of the household in the overall decision to carry out the migration is as a result of a key part of the theory.Central arguments of the theory indicate links that pressure the migrant to maintain a "contractual agreement" with the household (Massey et al., 1993).

METHODOLOGY
Since migrants out of Zimbabwe to South Africa are currently coming from all parts of the country, Gweru was then selected because it is centrally located in Zimbabwe.Gweru is also home to both the Shona and Ndebele ethnic groups, which are the two major ethnic groups in the country (Central Statistics Office and Macro International, 2007;Mangizvo and Dzikiti, 2009).The study employed the quantitative approach so as to collect data from as many households as possible.Households and respondents were randomly selected using a systematic random sampling method (Cresswell, 1994) using a database of suburbs that was sourced from the city of Gweru's housing department.As the list was incomplete, data on seven suburbs could only be obtained indicating the number of households from each suburb.An exception was Mkoba, the biggest high density suburb in the city, which was included though only about a quarter of the suburb was accounted for in the data.
The suburbs selected were Mkoba, Senga, and Mambo high density suburbs, then there was also Irvine and Athlone middle density suburbs and lastly, Daylesford and Gweru east low density suburbs.The sampling interval was calculated as follows; Since: Sample size (n) = population size (N)K It follows therefore that: Interval (K) = Population size (N)/Sample size (n) K = 6930 / 150 Therefore K = 46.2Households In every household, adults aged 18 and above were interviewed, and efforts were made to interview the head of the household making a return trip to households where the head of household was not found at first visit.Where the household head was not found, the eldest person available, who was aged 18 and above was then interviewed.The data was collected over a period of seven weeks between June and July 2013.Respondents generally chose to discuss the less sensitive migrants, or those migrants who had gone to South Africa legally.Hence only a few illegal migrants were represented in the sample.The data collected was analysed using Strata 11 Intercooled to generate frequencies and cross-tabulations that describe the sample in line with the objectives of the study.

Data analysis
The migration decision among most Africans has been widely agreed by many scholars to be a complex one that involves a number of people (Massey et al., 1993;Stark and Bloom, 1985;Taylor and Barlow, 2000).These are usually household members who are involved in giving encouragement, approval, and providing the means to make migration feasible (Zinyama, 2000).It has also been discussed in literature that the selection of a migrant is done after considering a number of factors that include age, sex, skills and migration experience (De Jong, 2000).This study set out to investigate the involvement of household members in the decision-making process of migrants in Gweru.Of paramount importance in the data analysis is the influence of the household in the making of the decision to migrate.
From the data collected, 76% of the households were from high density areas.The biggest household in the sample had 10 members while the smallest household had two members.The households had a higher average number of women than that of men and the average number of people aged 18 and over was generally higher per household than that of those under 18 years.As was expected, the average number of people in formal employment (1.16 people per household) was less than the average number of people with informal employment (1.27 people per household).This is in agreement with the findings of the ZDHS, 2010-11, because the formal sector is performing poorly in the country.However, the evidence in Table 1 clearly shows that the informal sector, which includes cross-border trade, is being used to boost incomes in households since the country's formal sector has shrunk (Bloch, 2008;Mangizvo and Dzikiti, 2009).For some households in the sample, the informal sector was the main source of income, with 26.7% of the households in the sample relying solely on earnings from informal employment.Another significant factor is that 78.7% of the households in the sample had at least one member involved in the informal employment sector.Hence informal employment, including cross-border trading, is a key income earner for households in Zimbabwe.
A look at the challenges facing households in Zimbabwe indicates some valid reasons for the need to migrate out of the country.Water shortages are one of the challenges that could be pushing people out of Zimbabwe (Bloch, 2008).According to Table 2, up to 87.3% of the sample stated that they only have water supplies "sometimes" and one household stated that they do not have a water source within their household at all.This is not surprising especially because of water shortages caused by rampant pipe bursts and also due to reduced water supplying capacity by the national water utility ZINWA (Mangizvo, 2009).The water shortages are considered to be the main cause of the outbreak of the cholera epidemic that has hit parts of the country since 2008 (Mangizvo and Dzikiti, 2009).Hence some households now take the initiative to treat their water before using it to avoid cholera, and one of the common ways is to use chlorine tablets, most of which are sent home by migrants in other countries (Mangizvo, 2009).12.7% Source: Own calculations from data collected in Gweru.*The two parts with asterisks mean that an option is used when there are power cuts.
Table 2 also reveals that most Zimbabwean households surveyed no longer rely on electricity for cooking as a result of the rampant power outages that have affected the whole country.Up to 82% of the sample stated that they often have to use firewood for cooking as a result of the rampant power cuts.Some respondents admitted that they no longer attempted to use their electric stoves to cook meals such as dinner because they know that the power cuts would obviously disturb them.Therefore, these numbers indicate that the supply of electricity is at crisis levels and it makes people's lives difficult especially when firewood proves to be difficult to access, as is the case in most urban centres of Zimbabwe today (Mangizvo and Dzikiti, 2009).
The use of alternative fuels such as liquid propane gas is also on the increase as a result of the rampant power outages (Mangizvo, 2009).A challenge in this front is that the common suppliers of liquid propane gas, namely the service stations, also could not provide the services as a result of the country's economic situation.Hence, new types of stoves especially the Gel stove became common alternatives, notably because the fuel for these stoves is imported by migrants and small business people in the country (Mangizvo, 2009).Thus, migration is proving to be beneficial in all fronts for those households that have migrants.

Household Income
A number of households in the Zimbabwe are faced with enormous financial challenges and opportunities for earning an income within Zimbabwe are very slim under the current political and economic climate (Bloch, 2008;Mangizvo, 2009;Polzer, 2009).Migration has become a significant income earner for many households.Though Figure 1 suggests that the households in the sample had generally high incomes, figure 2 then shows that 77% of these households had their incomes increasing because of migrant remittances.This indicates that not only have remittances come as survival method but some migrants are managing to sustain their households.
The total monthly incomes of the households in the sample, shown in Figure 1, indicate that most households have a fair income.This is shown by 36% of the households having a monthly income of over R10 000, with another 28% of the whole sample earn between R7 000 and R10 000, hence 64% of the households have a total income that is above R7 000.However, as was also noted in a study by Bracking and Sachikonye (2006), it is interesting to note that households with higher incomes are not concentrated in the lower density suburbs, but distributed in all suburbs.The household monthly incomes also suggest that if migrants are the ones sending remittances back home to an extent that they manage to change the family incomes so well, then it is very likely that they intend to return home in future.This is supported by the fact that 92.9% of the migrants in the sample have decided not to cut ties with the household and they were reported to be sending remittances back home.Therefore results of remittance sending are positive for the households even though as Posel (2002) argues, the sending of remittances may be intended, by the migrants, to further personal interests such as investment for the future.
An interesting feature in figure 2 was that the 23% of households whose incomes did not change after the migrant left were either high income households or the migrants were students.The fact that some high income households did not show an increase in their incomes after a migrant left concurs with the view that the migrant's purpose for migrating could have been to enrich oneself and not the household (Posel, 2002).In cases where the migrants were students, it has been noted in most literature that this does not mean the households are rich, but they are trying to cushion themselves for the future when the migrant eventually gets a job (Bloch, 2008;Mangizvo, 2009).This view means that the migrant then has to repay the assistance provided by household members and in this study, some household respondents pointed out that they expected the migrant to repay them or that the migrant had promised that they will do so.These two responses came from 36.8% of the households with migrants in the sample, showing that there is an expectation from a significant number of households that migrants should repay the assistance that they receive from the household.
Ultimately, migrant remittances were evidently the biggest income earner for most households in the sample.Among all households, those that reported no migration at all had significantly lower incomes than other households, and this is shown by all of them earning less than R10 000 a month.On the other hand, 41% of households that had a migrant in any country other than South Africa were earning a monthly income above R10 000, and the same was true for almost 37% of households with migrants in South Africa.Therefore, though some of the households with migrants reported that they do not receive remittances, it is undisputable that remittances have had a huge effect on households' incomes and this is very likely to influence other individuals into migrating.

Migrant Demographics
The characteristics of migrants that were found in this study are very similar to those found in other studies (Crush et al., 2005;Zinyama, 2000).The findings show that there are various types of people migrating from Zimbabwe to South Africa and their decision to migrate is also influenced by a variety of reasons.In other words, migrants range from people with just primary school education to those with post-graduate degrees.And, though the majority (69.1%) migrated for job-seeking, others went to sell their wares (13.5%), or for schooling (13.5%), or to escape the political violence (1.6%), among other reasons.
The ages of the migrants in the sample ranged from 15 to 59. Almost 78% of them fell between the age groups of 20 and 40, confirming the assertion by Crush et al, (2005) that migrants are usually in their youth.On the other hand, migration is becoming more common in both males and females with 56.4% male migrants and 43.7% female migrants respectively, defying the traditional view about a migrant being a male.An analysis of the marital status of migrants surprisingly indicates that the majority of migrants (66.7%) are married.Only 26.2% had never been married and just above 7% were divorced or widowed.This finding suggests that migration is now determined more by the necessity to fend for the household than for personal gain since the bread-winners are the ones migrating.
The sample indicated that most migrants possess a good set of skills that can allow them to get jobs.Over 90% of the migrants within the sample had at least secondary school education and 6.4% had postgraduate degrees.What was striking is that 49.2% of the sample were reported to have a tertiary qualification.This supports the idea that households select the most capable individuals to migrate in a quest to generate a more lucrative income abroad (Crush et al., 2005).
A significant number of migrants were reported to be employed (47.6%) and another 42% were reported to be 'sometimes employed'.This is a very high rate of employment among the migrants indicating very positive gains for households with migrants.The positive returns to migration are more defined by the fact that 52.6% of the migrants that had initially left with the purpose of 'schooling' were reported to be 'sometimes employed' and an additional 10.5% of them were reported to be 'employed'.Therefore, 63.1% of people who migrate for the purpose of schooling end up finding a job, some of them being employed during their years of schooling.
As a result of the greater employment opportunities, almost all categories of migrants were in a position to remit some earnings back home.Of the 17 migrants that had left home for schooling, 15 were reported to be sending remittances back home.This meant that some of the migrants had finished school and managed to find a job.However, it was also reported that some of these people were still doing their studies but managed to find some part time work and they were sending remittances back home.This draws a very positive picture in the minds of the non-migrants back home about the employment opportunities in South Africa and puts pressure on them to find some way to also migrate.
Of all the professions that migrants seek to get into South Africa, teaching, domestic work and nursing are the most common with 19.2%, 10.8% and 10% of the migrants being in the three fields respectively.Other professions that were also significantly represented were cooks or waiters (8.3%), technicians (5%), then mechanics and accountants who each took up 3.3% of the sample.As a result of the wide variety of enterprises that migrants take up, most of which were unknown to the respondents, 30% of the migrant jobs were categorised as 'other'.
When asked to compare a migrant's current job to their former job in Zimbabwe, respondents gave some interesting responses.Only 24% were reported to be doing jobs that are similar to the ones they were doing in Zimbabwe.Another 24.7% were reported to be doing jobs that were better than the ones they did in Zimbabwe.This is significantly higher than the only 6% who were reported to be doing jobs that are worse than the ones they were doing back home.However, the fact that such a number of migrants chose to do lower earning jobs as compared to their former occupations indicates a great need to earn an income among Zimbabwean migrants.

Gender and Migration
Among the 126 households that had members who had migrated, 79.4% stated that the migrants had children.It was interesting to note that almost half of these migrants with children were females.Moreover, among these 46 female migrants, 21 (45.7%) had children below the ages of 5 years that they had to leave behind for some periods of time.For 17.4% of these women who worked as domestic workers, coming home was not very frequent, probably because they do not get much time off and also they earn very little as was also noted in a study by Dinat and Peberdy (2007).
Table 3 indicates some issues of concern as a result of migration by mothers.According to the Table, three of the migrants had children aged less than a year.To add to that the responses given stated that two of the mothers visited home only once every month and one visited once every three months.In another case, a migrant mother had left her 3 year old child more than two years prior to the study and did not get a chance to return since then.These few cases indicate some of the difficult situations that households have to bear while the breadwinners try to fend for their families.Source: Own calculations from data collected in Gweru.
When asked about the employment status of migrants in South Africa, most respondents indicated that a sizeable number of the migrants are employed at least 'sometimes' and at most 'all the times'.Of the 126 migrants, 60 were reported to be 'employed' and 54 were reported to be 'sometimes employed'.This means that 90.5% of the migrants in the sample were earning an income sometimes.For the people remaining at home, this is an incentive for migrating considering that unemployment in Zimbabwe is still hovering somewhere around 80% (Landau and Wa Kabwe Segatti, 2009).Hence, the pressure to migrate cannot be removed from Zimbabweans especially if such crucial issues as unemployment are not resolved, neither can migrants decide to return home seeing that even more people are trying to also leave home.
Marriage is a demographic characteristic that has been reported to have an influence on migration decision making (Dinat and Peberdy, 2007).According to table 4, among the migrants reported in this study, 66.7% of them were married.This indicates the greater need among married people to earn an income in an effort to provide for their families.What is more interesting are the gender differences between the married migrants.Among the male migrants, 77.5% were married and among the female migrants 52.7% were married.This agrees with the view that men tend to take up the role of breadwinner, but it also indicates an increasing involvement of married women in migration (Matshaka, 2008).
However, the effect of gender differences on migration decisions is clearer when we look at the involvement in migration by those who are in other nuptial categories.Historically, literature argued that women who are not in marriage are more likely to migrate than those within marriages (Dinat and Peberdy, 2007).It has been stated that among all female migrants, married women were the majority comprising 52.7%.Table 4 shows that among the 'never married', the fraction of female migrants was slightly more than that of male migrants at 30.9% and 22.5% respectively.Among those migrants who had either divorced or were widowed, there were no male migrants, while 16.4% of the female migrants fell in these categories.Therefore, women who are not in marriage are shown here to be more likely to choose to migrate than their male counterparts.4 brings out an interesting fact that married people in Zimbabwe, including women are more migratory than unmarried people.Among all migrants, there is an increase in the number of migrants from 26.2% among never married people to 66.7% among married people.This increase is shared among both sexes, though it is more pronounced among men who clearly migrate more when they are married possibly because women often take responsibility for the care of the children.As a result, while 77.5% of the male migrants were married, 52.7% of the female migrants were married.Another 16.3% of the female migrants were either widowed or divorced.

Household influence in migration decision-making
The household is in a good position to influence the decision of a potential migrant to eventually migrate by offering and providing various types of support for migration (Burns and Keswell, 2006).Respondents were asked if the household had helped the migrant in any way, and 94.4% agreed that they had helped the migrant to get through the process of migrating in one way or the other.The types of assistance provided to a migrant included financial, information, sourcing accommodation, networking the migrant with people at the destination and providing encouragement.
Table 4.5 shows a breakdown of the types of assistance that 119 households said they had provided to migrants.The other seven households that had migrants but are excluded here pointed out that they did not help the migrant in any way.The provision of encouragement is by far the most common among the households, with 34.5% stating that they had provided the migrant some encouragement, and another 31.9%had provided migrants encouragement together with other forms of assistance.With a total of 66.4% of households reporting to have encouraged their migrants to leave, this component of psychosocial support stands out as the most common form of assistance provided to a migrant.
As indicated in table 5, the second most common assistance provided by households is financial in nature.Up to 10.9% of households provided only financial assistance to the migrants.Another 47.9% again gave financial assistance together with another form of assistance to their migrants.This means that, in total, 58.8% of households provided some financial assistance to the migrant in order to help them to migrate.Though the financial assistance could have been given after the migrant had decided to move, it is obvious that this provision makes migrating more feasible.Providing a migrant with information about the destination is also important because it reduces potential Challenges in accessing desired resources (Zinyama, 2000).Migrants get this kind of help from the household members back home when they leave, as well as from other household members in the destination country, during their stay.Among the households providing assistance to migrants 4.6% stated that they only provided information.Another 47.9% stated that they provided information together with other forms of assistance.Thus households try to assists migrants by providing them with information to make their movement smoother.
Since households are mainly based in the country of origin, they are less likely to ensure that the migrant gets essential networks with people and also accesses accommodation at the destination.However, where households were in a position to do this, they did provide this assistance.From all households providing assistance to migrants, 12.6% and 20.2% reported that they provided links and assisted the migrant to get accommodation at the destination.These figures are supported by the fact that 46.03% out of 126 households had provided links to friends, relatives and even job providers.Therefore, such households have a big influence on the decision to migrate, as well as the destination of the migrant.

Remittance benefits for households
Zimbabwean households evidently have a lot to gain from sending some skilled members to other countries in search of jobs that would bring the households much needed incomes (Bloch, 2008;Bracking and Sachikonye, 2006).This study also found that remittances from migrants are an important source of income for many households in all suburbs in the city of Gweru.From the 117 households that reported they were receiving remittances, 26.5% stated that this income makes up over 75% of the total household income.Hence, this indicates that households are heavily reliant on remittances.
To add on to the heavy reliance on remittances, the study provides evidence that most of households' purchases for their various needs are done by migrants outside Zimbabwe.This has been a common occurrence since the near collapse of the country's economy made prices for all commodities unreasonably high and also most necessary commodities became scarce.Hence, as is shown in Figure 3, remittances from migrants come in all forms, with 87.2% of households with migrants stating that the remittances sent by migrants include money, food, clothes and assets.
A very significant proportion (94.1%) of the households with migrants indicated that money was usually part of the package of remittances that they received.This is not very surprising because the need for an income is one of the main reasons why households in developing countries urge their members to migrate (Stark and Fan, 2007).Though all suburbs had some households with migrants that either sent very little or did not send at all, it was interesting to note that the wealthier households seemed to have the most number of non-remitting migrants.This probably implied that these households were not really in need of a supplementary income, and even the purpose of migration by the migrants was largely for the migrant's personal benefit.
To illustrate the above, 38.5% of the migrants from wealthy households did not send any remittances back home.This agrees with the other findings that migrants from wealthier suburbs aimed at getting access to better education, retirement locations, tourism and fear of political violence which made up 3.8% of migration purposes.However, it has to be noted that another 30.8% of these households in richer suburbs received more than 75% of their income from migrants, suggesting that the migrant would have moved the households to its current place of residence.This is a trend in present day Gweru and more about it will be discussed in the conclusion.Source: Own calculations from data collected in Gweru.

Utilization of household networks by migrants
Households have been reported in other literature to network migrants with people living in the destination country (Zinyama, 2000).This study found that 46% of the households in the sample provided their migrants with helpful networks at the destination.According to

Remittances sent by Migrants
Percentage Frequency Though the provision of links is not very prevalent, it can be seen that Zimbabwean households have developed useful connections in South Africa.With 4.7% of households stating that they provided the migrants with links to job providers, this shows a rise in household networks into essential sectors in the destination country.Those migrants that are reported to have been provided links to job providers, are either "employed" or they are "sometimes employed".Hence, this proves the provision of links to job providers by households to be effective in securing jobs for the migrants.
Among all migrants that were provided with some form of links by their households, only two were reported to be unemployed.Hence it would seem that the household links that the migrants are furnished with, prove to be useful.However, it has to be noted that a large number of migrants were not provided any links, up to 53.9%, and from this group, only six of them were reported to be unemployed.The provision of links by the household also influences a migrant's feelings of altruism and increases their likelihood to send remittances back home.This is because a migrant would have to leave for an area where other household members or friends take care of them and they end up feeling that they owe the household.Evidence from this research shows that among the 46% of migrants that were provided links, all those that were employed and have been provided links at the destination by their households were sending remittances back home.On the other hand, 6.5% of the migrants that did not get links from their households were not sending remittances back home.Therefore, this comparison shows that the provision of links makes migrants to feel indebted to their households and influences them into sending remittances.

DISCUSSION
Though a few specifications can be made about the nature of Zimbabwean migrants that leave for South Africa, data from this study does not clearly identify typical migrant characteristics.This is largely because the socioeconomic and political situation in Zimbabwe has forced a wide variety of people to migrate for one reason or the other (Bloch, 2008).It seems as if the natural thing to do now for most households with migrants is to rely on remittance income and consider locally earned income a bonus.Hence, a better description can be found for a typical non-migrant household than for the opposite since it is now far less common to find a household without a migrant of some sort in Zimbabwe.This study found that up to 84% of the households surveyed reported that they had a migrant living in South Africa, while only 71.3% of the same set of households had migrants living in other countries.Thus it is evident here as well as in other literature that South Africa is receiving most of the migrants that come from Zimbabwe, though exact figures are not known (Bracking and Sachikonye, 2006;Crush et al., 2005;Polzer, 2009).This is also supported by the fact that most households have been reported in other studies to favour migration to South Africa more than to other countries (Zinyama, 2000).Hence, migration to South Africa has become very common and this makes it difficult to identify unique characteristics among the migrants.
Therefore, the findings here can be used to pave way for more detailed studies on Zimbabwean households and migration.Though this is possible, it would not be wise to generalize these findings to the greater populace of the city of Gweru, let alone the whole of Zimbabwe, because the sample used here is very small.This chapter will present the findings by responding to the objectives that the study set out to achieve, describing migrant households and the migrants, elaborating on the nature of the inter-dependency between them.
The study set out to address the following overall objective; to establish the characteristics of Zimbabwean households, as well as their members and how these influence the decision of individual household members to migrate to South Africa.Since the specific objectives of the study were developed from this overall objective, this discussion is going to focus on them (specific objectives), so that the aim can be covered in detail.Thus, the following sections will be organized in the same sequence as the specific objectives in the first chapter.

Not the ideal migrants portrayed in literature
The views found in most African migration literature are that migrants are usually the single and educated males and females (De Jong, 2000).This view portrays youthful people who are well educated as being the migrants while older and married people do not intend to migrate, probably because they have a role to care for the family.Younger females are also viewed as being more reliable migrants that are more likely to send remittances back home unlike their male counterparts (De Jong, 2000).However, the findings in this study do not bring out Zimbabwean migrants to South Africa in this way.
The findings indicate that large numbers of married people are involved in migration from the city of Gweru to South Africa.Up to 66.7% of the migrants reported in this study are married, which is unlike the views in most literature where married people are portrayed to be less migratory.This is obviously a result of Zimbabwe's very bad socio-economic and political situation that is forcing households to search for alternative sources of income.These findings echo the views of Zinyama (2000) who states that increasing numbers of married people are opting to migrate from Zimbabwe into South Africa, often after being encouraged by the household.
The importance of an income for the household is also indicated by the desperate migratory behaviour by Zimbabwean mothers.One woman with a child that was under 5 years of age had not visited home for over two years, but she was still frequently sending remittances back home.This suggests that the job she was doing did not give her enough time to be able to visit home, in spite of the fact that she had not seen her child for such a long time.Dinat and Peberdy (2008) found that most women who work as domestic workers in Johannesburg do not get much time off and they earn too little to frequently make the trip home so they prefer to only send remittances.Hence, such is the importance of earning an income for the household.
For the children who are left by their mothers in the care of grandparents or other household members (who may sometimes be children themselves), growing up is likely to be a difficult task.Many of them take up small jobs in order to add on to the household income, and these jobs may reduce their time in school.In one of the city's high density suburbs where a university campus is located, such children capitalize on the arrival days of university students and offer a service of ferrying luggage using wheelbarrows even during school hours.Therefore, it is apparent that these children skip school in order to wait for customers at the bus stops.
The study also indicated that 16.3% of the female migrants were either widowed or divorced but none of the male migrants were categorised as such.This highlights an interesting view by Gregson et al. (2002) that Zimbabwean females who lose their husbands usually find it difficult to remarry because of the stereotype that women who get out of marriages may be infected with HIV.This stereotype is not shared by divorced and widowed men who usually find it easier to remarry.Therefore, because most of the women would be dependent on their husbands' incomes, some of them are then forced to migrate in search of jobs or trade opportunities when they lose their husbands.
Therefore, studies on migration in Africa do not always clearly reflect the situation that Zimbabweans face.As a result of the harsh economic environment and the very high unemployment rates in the country, households often have to make the difficult decision of sending someone or some people to search for work in another country.The household head is often the most realistic option because of more work experience which makes them more competent in the job market.Some migrants are fortunate enough to take their families with them, and in this study 21.4% of the married migrants were reported to be living together with their spouses in South Africa.

The migrant households
The characteristics of Zimbabwean households that have migrants coming from them to South Africa are not clearly distinct.This is because all households in the country are faced with difficult circumstances within the country that force them to seek earning an income from other countries.As a result, this study found that Zimbabwean migrants in South Africa come from different backgrounds.However, some differences are found in the reasons for their migration with wealthier households sending migrants for schooling and leisure while poorer households do so in search of an income.
As has been mentioned earlier, it is easier to describe households that do not have migrants because in Zimbabwe, there are fewer of them than those that have migrants.This is especially true in this study where 84% of the households in the sample had migrants.Of the households without migrants residing in South Africa, only 12.5% had incomes below R5000.00 per month.This shows that most of those non-migrant households have relatively good sources of income either locally or from migrants in other countries other than South Africa.
Those households with migrants who leave for South Africa with the intention to get an education vary widely.Though 82.4% of households sending migrants for the purposes of education were earning more than R7000.00 per month, the other 17.6% came from households earning less.This indicates that households with little income are putting most of their earnings for the purpose of getting good education for their children, maybe with the hope for getting returns.
It is interesting however that the poorer households that sent family members to South Africa for schooling tended to state that they did not require repayment of costs from the migrants while the opposite was true for wealthier households.A total of 70.6% of the households with migrants that left for schooling either stated that the migrants promised to reimburse the costs incurred by the household or they expected the migrant to refund them.All of these households earned over R7000.00 per month.On the other hand, among the 29.4% of the households that did not require any such repayment, only one of them earned above R7000.00per month.
Therefore this is contrary to what one would expect that poorer households would seek to get financial returns from educating their children.
Among those households that had migrants leaving for the purpose of seeking work, 62.1% stated that their incomes were above R7000.00.This high income is significant because 87.4% of the households whose migrants left in search for work stated that their incomes increased after the departure of the migrant.Hence, households tend to benefit financially from labour migration by their members, and this would inevitably influence the perpetuation of the movement of prospective workers from Zimbabwe into South Africa.

Migrant Households and New Economics of Migration Theory
Interestingly, this situation of Zimbabwean households is in line with the views of Stark and Levhari (1982) in the new economics of migration theory.The central argument of the theory suggests that when there are market failures, and increased risks for a household such as drought and poor incomes then households send one member to another location to earn an income as a means of risk reduction (Massey et al., 1993).This is in line with the situation faced by households in Zimbabwe where economic crisis has caused distress and has increased risks of starvation for many households.
The fact that 46% of the households in this survey were able to provide links at the destination to the migrant also concurs with the theory.This is one way by which the household makes sure that the migrant will feel more obliged to send remittances back home, though this can hardly be regarded as a contract.The migrants also have the responsibility of providing the same help that they were given when they first migrated to other household members who migrate in future.Thus, the findings agree with the theory again in this aspect.Remittance sending is another key aspect of the theory, and the migrant sends them as a means of repaying the family's assistance.There is a strong suggestion of this in the findings of this study as 92.9% of the sample reported that they were sending some form of remittances back to their households.As a result, whether the migrant does it either to fulfil a contract or to further their own needs or because they feel pity for the household, it is clear that the households have effectively reduced their risks in Zimbabwe's economic woes.
It is also unquestionable that many households in Zimbabwe are faced by many risks including hunger, disease, and even death (Bloch, 2008).The government hospitals and clinics are usually ill-stocked with medical supplies and only the expensive private hospitals are reliable (Bloch, 2008).The cholera epidemic continues to infect people because of dilapidating water and sewerage reticulation systems (Mangizvo and Dzikiti, 2009).The agriculture sector, crippled by a never ending and chaotic land reform exercise fails to feed any significant proportion of the country (Mangizvo, 2009).Hence, households seek medical attention, water purification supplies and food from the neighbouring countries especially South Africa.
As a result, the migration from Zimbabwe is unique in that it has no clear selection of households that participate in it, its occurrence agrees with the new economics of migration theory that households seek to reduce risks.With this in mind, it is clear that in a time of crisis such as in Zimbabwe, all classes of society will seek to participate in migration as a response.As a result the migration from Zimbabwe to South Africa can more clearly be understood as survival migration.

The migration decision
The decision to migrate was investigated indirectly in this study.This was done through the use of questions that sought to find out the migrant's demographic characteristics especially age, sex and education attainment.However, a more direct question on the purpose of migration was asked in order to clearly inform the motivation for migration.The choice of the destination is sometimes also informed by other household members, which often determines the timing of migration.Therefore, the decision to migrate is made after a number of variables about the migrant are considered, and the household may inform the decision by providing links at the destination.
Migrants have always been considered to be youthful people, and their ages in this study reflect this.About 78% of the migrants in this study were aged between 20 and 49 years, indicating that they are mostly the working categories that are leaving the country.Though the young and the old also tag along in some instances, this is not widespread, suggesting that the migration is not intended to be permanent.Mostly the working class take part in the migration, an indication that the migration intention is to seek employment or other money making ventures.
The selection of a migrant seems to be largely in favour of heads of households.The majority of individuals that were selected to be migrants were heads of households.According to Table 4.4, 77.5% of the males and 52.7% of the females in the study were married.This is in agreement with the views of McDonald et al. (2000) that most household heads in Zimbabwe select themselves to migrate, probably because they feel obliged to provide for the family.Therefore, since they have the last say in household decisions, household heads tend to select themselves to migrate.
Among unmarried youths, females were found to be more likely to be selected by the household to migrate than males.According to Table 4.4, 30.9% of the female migrants and 22.5% of the male migrants were reported as "never married".This suggests a shift from migration historically being dominated by unmarried men, when the girls' place was at home.It also agrees with the findings by Gubhaju and De Jong (2005) that households in Thailand trust never married females in remittance sending and they encourage them to migrate more than males.Therefore, it seems as if females are no longer confined to the domestic arena and they are active participants in seeking incomes for Zimbabwean households.
However, it seems as if the migration decision is sometimes a desperate one because people in unsuitable situations, such as women who leave infant children behind, also tend to migrate.Women with children under one year of age take part in migration with two of them coming home once a month, and a third visiting home only once every three months.This suggests that the circumstances for their families would be very difficult, and they have little choice but to take part in migration, leaving behind their highly dependent babies.
The migrants identified through this study had high educational qualifications.With the exception of one migrant, the rest were reported to have at least secondary school education.Also, 27.8% of the migrants had at least a University Degree and at most a post graduate qualification.This indicates high levels of education among the migrants, which suggests that they were expected to be more likely to secure jobs in South Africa as compared to less educated individuals in the household.

Utilizing Migrant Networks
A sizeable number of the migrants received some initial assistance from people at the destination to whom they were referred by their households.Up to 46% of the migrants had assistance from some form of household links at the destination.Though the person that a migrant is referred to may not always be very helpful, the assistance that they give makes it possible for the migrant to settle down.Among the people to whom migrants were referred, about 29% were relatives to the migrant.They would provide shelter, food and advice among other things until the migrant gets a job or another opportunity.
As a result of these social networks, most migrants may end up living and working in areas where a number of their relatives are present.This is an advantage for the household because when their migrants are closer to each other, they would be in a better position to send assistance back home.This is because whenever one person plans to visit home, then all other migrants can send remittances through that person.
Coherently, sending remittances through relatives was one of the most common means of remitting found in this study with about 24% of the migrants using this method suggesting that such people will be living close to each other.This also influences future migration decisions, sometimes making more and more household members to move.This is because, according Massey et al. (1993), the costs of migrating would have been lowered by the presence of relatives at the destination.
As a result most migration decisions are influenced by the household.The migrant's destination, and sometimes the type of job that a migrant takes up are dependent on where the household's social networks are located.This is especially so if the migrant is female because households tend to be more protective towards them, and they are also easier to control for the household hence making it easier to tell them where to go than it is to tell male migrants where to go.

Altruistic and Contractual remittance sending
The remittance behaviour of the migrants is important in understanding how the decision to migrate was made.Some households may feel that a migrant has an obligation to send remittances back home, suggesting that the household played a role in the migration process and the migrant owes them.However, this may also be because the household expects the migrant to feel altruistic towards them.Questions were asked to differentiate between the two feelings by the household.Some of the households in the survey indicated a desire to have the migrant repaying them for the financial assistance that they provided.Up to 36.5% of the households stated that they desired the migrant to refund them, either because they expect it, or because the migrant promised to do so.The 15% of these households who stated that they expected repayment, seemed to be sure of their position as deserving recipients of remittances from the migrants.This suggests that some form of agreement was reached with the migrant assuring the household of remittance sending.This is therefore a strong suggestion of a contractual agreement.
On the other hand, the 21.4% of households that state that the migrant promised to send remittances still suggest a contractual agreement though it is weaker.The household may feel that the migrant owes them, but they accept that they do not have much power to influence such a decision and they depend on the migrant's love for them.This therefore suggests that the household does not only appeal to a contractual agreement, but also they hope for the migrant to feel altruistic towards them.
The larger group of households, about 65%, stated that there was no need for the migrant to repay them for their help.The interesting thing was that, the majority (95%) of the households in this group still received remittances from the migrant.This is a significant sign of altruistic remittance sending by the migrants from these households since the household members do not expect them to be sending anything.Therefore, the majority of remittance sending by migrants seems to be done out of love by the migrants, and with the knowledge that their households need the assistance.
Thus, though remittance sending is widespread, most of it seems to be done because of feelings of altruism by the migrants towards their households.Because the migrants understand the difficult circumstances under which their households are living back home, they tend to send remittances to assist them.However, some households expect remittances to be sent to them indicating the presence of some form of agreement between the migrant and the household.

CONCLUSION
This study suggests a large number of migrants from Zimbabwean households to other countries, with a significant number of migrants heading for South Africa.Considering the assistance that the migrants receive from their households, and the remittances that they send back to the same there is a strong suggestion that households have a big influence on the migration decision.While migration has obviously become the main source of income for many households, most of the remittances are sent by the migrants as a result of altruistic motivations.
The decision to migrate from Zimbabwe into South Africa is clearly made for the benefit of the household by most migrants.This is supported by the fact that most of the migrants are household heads, with their children and families remaining behind.Also, though some migrants fail to visit home themselves, they continue to send remittances to their households suggesting that their stay in the Diaspora is intended to improve the situation of their households.Therefore the migration decision largely seems to be for the benefit of the household.
The participation of the household in the migration act itself makes the process of migration feasible a lot of the times.Households provide migrants with people that can assist them at their destination, and in a few cases these people may be job providers.This makes the migrant's task easier, as well as enabling the migrant to quickly get employment.As a result of this, migrants tend to feel indebted to the household which in turn influences them into sending remittances.
Ultimately, it is apparent that the continued migration of Zimbabweans into South Africa is a result of the attractiveness of the option over all others.South Africa has the biggest economy in the region, hence it provides a lot more opportunities to migrants as compared to other destination countries.However, because a number of people have been migrating to that country since the early 20 th Century, there are more social networks in that country for Zimbabweans, hence they tend to favour it as a destination.Adepoju, A. (2005).Migration in West Africa.Lagos, Global Commission on International Migration: 1-23.Adepoju, A. (2006).The challenge of labour migration flows between West Africa and the Maghreb.Geneva, International Labour Organization.

Figure 3 :
Figure 3: Types of Remittances sent to Households by Migrants (N=117), 2010Source: Own calculations from data collected in Gweru.

Table 3 :
Visits back home by Female migrants with children (N=46), 2010

Table 5 :
Assistance provided to migrants by their households, (N=119), 2010 Table7, the most common type of links provided to a migrant by the household are links to relatives who stay in the destination country, comprising of 26.9% networks to family friends are less prominent with about 14.2% reporting to have provided these.The least prominent type of links provided are links to job providers which were reported by only 4.7% of the households in the sample.

Table 7 :
Networks provided to migrants against their reported employment status (N=126), 2010

Table 8 :
Cross-tab: Migrated to SA and Migrated to any other country (N=150), 2010