submitted to the UN Emergencies Unit for Ethiopia

by Laura Hammond[1]

April 1995

Executive Summary and Background to Study

An objective of any repatriation operation is successful attainment by returnees of economic self-sufficiency. For Tigrayans who were forced from their homes by famine and war during the mid 1980s and who lived for nearly a decade in refugee camps in the Sudan, this goal is particularly important and at the same time extremely elusive. This report presents the findings of a study of household level resource flows carried out between Hidar and Lekatit 1987 E.C. (mid-November 1994 to mid-March 1995 Gregorian Calendar) in the Ada Bai returnee settlement, Humera wereda, Ethiopia.[2] The study follows closely the daily economic activities of ten households over the four month period. It is not intended to be an exhaustive socio-economic survey nor is it meant to be a definitive indicator of the overall economic condition of the returnee community. Rather, it is meant to provide longitudinal profile information from ten household types selected to represent the economic, social, and occupational and religious diversity within the community and to give a basic picture of economic activities and decision-making strategies being carried out within the community at the household level. The following data was solicited:

. household income levels

. amounts and patterns of household expenditures

. frequency of lending, borrowing and repayment

. patterns of gift giving

. levels and patterns of food consumption

Specifically, the study sought to find answers to the following questions:

. What is the cost of living for returnees in the Humera area?

. Is there significant economic stratification within the returnee community?

. Do returnees have access to enough cash?

. How do income levels relate to food consumption patterns?

. What are the relative roles of farm and nonfarm income to household economies?

. What access do returnees have to credit and to what extent is it a part of overall household budgeting?

An earlier report[3] gave preliminary findings of data collected in the first month of this study. This data suggested that returnee household economies in Ada Bai may be appropriately characterized as semi-urban, with significant reliance on the market as well as on farm production in supplying basic necessities. The first month of the study, Hidar (November), coincided with the start of the sorghum harvesting season (although most of the harvesting is usually done in Tahsas, or December). Main sources of income at that time were occasional waged labour (paid at a set daily rate) and sale of farm products at the local market or in Humera, 17 kms away. During Hidar, returnees' recorded rate of borrowing was extremely high. This was attributed to the need to cover the costs of harvesting and to repay debts incurred during the cultivation season. Expenses were limited to basic food and non-food items. Data from the first month also suggested that income earned at harvest time was insufficient to meet the cash and food needs of the sample households. Households reported selling their expendable assets in order to meet their cash needs.

The study was concluded at the end of Lekatit (mid-March 1994) which in 1994 was the period of greatest economic hardship for most households.[4] The final results of the study reflect seasonal changes in economic activity and resource flow, though ideally such a study would have to be carried out for the entire year to show all seasonal variations.

In the preliminary report, the hypothesis was posited that over time household resources are being eroded by increasing debt, sale of expendable assets, and in some cases, reduction of food intake. Verification of this hypothesis would raise the possibility that additional food aid might be needed in 1995. This final report seeks to deepen the analysis and to test this hypothesis. The results show that while the asset base erosion process does seem to be occurring, households have a variety of strategies at their disposal to mitigate losses and to recover from extreme levels of indebtedness without falling into life-threatening food shortage.

The report begins with an analysis of resource outflows, examining harvest and post-harvest season changes, with particular emphasis on food and non-food expenditures. A rough monthly per capita cost of living is computed on the basis of the data collected. Incoming resources are examined and compared, with attention to the relative roles of wage earning/sale of assets and borrowing over time. Sale of high-value items and harvested grain is also considered to show how such sales serve to reduce households' reliance on credit as well as to repay debts. Sources of credit from both inside and outside the community are outlined; these include formal and informal borrowing practices. Patterns of repayment are discussed; restrictions to timely repayment are identified. Finally, the use of labour migration as a supplement to overall household income is explored through case study interviews.


Humera wereda is located in the far northwestern corner of Ethiopia, bordered to the north by Eritrea and to the west by the Sudan. For the past thirty years, it has been a major center for commercial agriculture, the main crops being sorghum, sesame and cotton. In June 1993, 12,000 Tigrayan refugees were repatriated from camps in the Eastern Sudan to in the Humera area. They were joined in March 1994 by 2500 more Tigrayan returnees. The returnees are settled in three locations near Humera town. This research project is being carried out in Ada Bai, which with a population of approximately 7,500 is the largest of the returnee settlements.

Ada Bai is in many ways a large town rather than a rural village. Its population alone makes it one of the larger towns in the area. The town has a defined business district with shops, tea and sewa houses,[5] mills, tailors, weavers, metalworks, and a video cinema. There is a church, a mosque, a health centre, and a school for grades 1-5. There is an office for the baito[6] and a central place for public meetings. When shop owners are able to pay the rental fee, there is electricity in the business district from 6 to 9 p.m. every night provided by a private generator owner. (It is frequently shut down as there are not enough renters.)

While some returnees in Ada Bai have found occasional work as agricultural labourers and many cultivate food crops for at least a portion of their household food requirements, it is clear that the town is not strictly an agricultural settlement, but rather is a semi-urban community. Its residents are heavily reliant on the local market (and thus on cash) to supply their daily household food needs. Cross-border trade plays a major role in the local economy; most goods available in the market are brought to Ada Bai from the Sudan and Eritrea by traders. In addition, prices are very high due to the inaccessibility of the area from other Ethiopian trading centers (particularly Gondar and Mekelle).

The three returnee settlements were intended to become self-sufficient agricultural communities by farming their own land and working as daily labourers on the large commercial farms in the area. Each household was to be allocated between one and three hectares (depending on the number of its members) by the baito. For a variety of reasons, in each of the past two agricultural seasons (1993 and 1994) there have been large numbers of households who have not received any land at all or whose allocated plot is located too far from the settlement to be accessible. For the 1994 agricultural season, only 1947 of Ada Bai's 2540 households were allocated farmland.[7] The landless were disproportionately single women. Many of those who did not receive farmland entered into sharecropping arrangements with landholders from Humera town, providing all of the agricultural work needed on the land in exchange for one half of the harvest. Households which lacked available labour or resources to become sharecroppers (primarily female-headed households) or who were allocated land too late in the season to prepare it for cultivation did not farm at all in 1994 with the exception of their small home gardens.[8]

During the study period of November 1994 and March 1995, returnees in Ada Bai did not receive any external food aid, although oil was distributed in early April 1995 as part of a grain and oil relief package that had been promised the year before (the wheat having been distributed in May 1994). The last time a month's grain ration was distributed to the needy was in September 1994.

Selected households were assisted in farming their land with the provision of low-interest loans from the local government for ploughing and purchase of seeds. (See section on credit, below.) The entire community also receives free health care and prescriptions, available from the local clinic and (for referred cases) from Humera hospital.

Demographic Profile of the Respondents

The ten respondent households were selected to represent the economic and social diversity of the community. Of the households sampled, two were female-headed (one with a man who came and went periodically but did not share most resources with his wife). Two were Muslim; eight were Orthodox Christian (including one headed by a priest). Nine households had farmland that was either allocated to them or rented from another landholder. Two had regular supplementary income from a skilled trade (one was a tailor and one was a weaver). The average household size of the respondent group was five people. Table 1 provides a demographic profile of the ten households.

Table 1: Demographic Profile of the Ten Sample Households

Household   Number of   Male or        Religion       Land     Nonfarm       
  Number    Members     Female Head    (C=Christian;  Farmed   Sources of    
                                       M=Moslem)      (has)    Income        
    1       5           M              C              1        None          

   210      37          FM             CM             02       NoneWeaver    

Based on knowledge gained in the first year of this research project, a preliminary assessment was made of the relative economic strength of the households. Three of the households (HH 1, 2, 7) showed initial indications of representing the poorest stratum of the community. Four households (HH 3, 4, 5, 8) were estimated to represent the average household economic status in Ada Bai, and three seemed to have a slightly higher economic standing in the community than the others (HH 6, 9, 10). No very wealthy household was selected, as it was felt that such data would skew the data and give the impression that people are better off than they really are. At any rate, such households are very few in number in the Ada Bai community.

At the outset of the study, household assets were recorded. Four of the households (HH 1, 3, 6, 7) had no livestock at all save a few chickens. These households also had very little expendable property. The other households (HH 2, 4, 8) owned one donkey or goat with the exception of one household (HH 9) which had seven sheep and two goats and another that had two cows (HH 5).

Most of the households (except for the two that were female-headed) owned at least one metal bed (valued at approximately 200 birr each). Other durable property included metal storage boxes (100-150 birr), oil drums for holding water (150 birr), watches (100-200 birr), tape recorders (500-700 birr), radios (100-200 birr) and (in the case of a few of the women) gold jewelry (from Sudan, at 30 birr per gram). Two households (HH 5 and 8) reported at the start of the study that they had recently sold furniture or tape recorders to cover their house expenses.


All the participants in the study were known by the researcher prior to the beginning of the study. Indeed, the strong relationships developed in the first year of research in Ada Bai have helped to encourage honest reporting and to build trust in the research team on the part of the respondents. Confidentiality and anonymity were guaranteed to all respondents, as was the right of any household to withdraw from the study at any time, or to retract information already given. No household availed itself of this opportunity, but it is believed that having the freedom to do so encouraged people to be more forthcoming with the data, much of which is considered to be extremely personal and secret.

Returnees, no matter how great or small their resource base, go to great lengths to hide their resource holdings from both their extended relatives and their neighbours in order to discourage them from asking for loans or gifts. They also hide such information from local authorities so as not to be excluded from distribution of relief aid. Returnees commonly feel that decisions about who should receive aid are unfair and that those in need are often screened out if, for instance, they own animals, have what local officials consider to be expendable assets, or are engaged in work that is deemed more profitable (in many cases defined as more prestigious) than farming. Assurances of confidentiality and awareness that the research team is not directly associated with any aid organization (as well as the fact that the research assistant is a returnee himself who is known and trusted by the community) allowed for collection of data that is very likely to be more reliable than that obtainable by other aid actors present in the area.

The baseline questionnaire recorded such information as age, education and training of each household member; number of chronically ill household members, estimated monthly income; property (food, cash, animals, other assets); amount of any remittances from relatives working outside the community; resources shared with other households; size of landholdings, planted crops and yields; outstanding debts; schedules for repayment; and amount and date of last receipt of food aid.

As might be expected, households tended to under report their holdings at the start of the study, particularly with regard to the size of their cash and grain reserves. Such under reporting was revealed when households were forced to draw from their reserves in order to meet household expenses. Most respondents claimed that they had either "forgotten" to report the full amount of their resources, not realized that they had more than they initially reported, or thought that what they had was too insignificant to be worth mentioning. One woman who had under reported her grain holdings at the start of the study later said, "I consider anything less than a half a quintal to be insignificant. What is it? It is only enough to feed my family for a few days." Such sentiment was widespread, but after it was stressed that the researchers considered even one cupful of grain to be important, people were more forthcoming about their holdings, expenditures and consumption.

Household income and expenses from each of the households were recorded daily. Information was provided by an adult member of the household. In most of the male-headed households, the information was initially provided by the man, but later on the woman became the main respondent, as she was more likely to be in the house when the interviewers came and had a better idea of household expenditures than the man. In most cases the man deferred to the woman to give the information, except in some cases where the man earned income: the woman was not always aware of how much he had earned. Wherever possible, the man and woman in a household were interviewed together, though care was taken to administer the questionnaires privately, when neighbours or non-resident kin were not in the house. Data collected included wage earnings, money or goods obtained through sale, borrowing, or gifts, as well as all expenses (itemized), loans, or gifts given. Meals eaten during the previous day were also recorded.

Once the daily recall part of the study began, it appears that people reported their incomes and expenditures with more accuracy. Had there been serious levels of under or over reporting of either resource in- or outflows, they would have been identifiable in the data supplied in the questionnaire. It is likely that as people came to trust the researchers and the confidentiality of the study, they became more willing to reveal the full size of their asset holdings. In addition, data was continually monitored throughout the study period so that questions and inconsistencies concerning the data could be cleared up with respondents promptly. Most of these arose from misunderstandings about how to characterize certain transfers that did not neatly fall into one of the questionnaire categories.[9]

The small size of the sample and the variation between household types prevents the use of most forms of statistical analysis of the data. Where statistics are aggregated from the ten households as a group, the intention is merely to give the most basic indication of trends. In most cases, households are treated as independent entities, and this report seeks to elucidate and explain household decision making strategies as they relate to each of the selected types.

Expenditure profiles

In analyzing the data on resource outflows, distinctions were made between food and non-food expenses, loans extended and repaid, and gifts given. Special attention here is given to food and non-food expenditures over the four month period on an aggregated level, since no household reported extending loans to others and repayment is discussed later in the report. Figure 1 shows the levels of food and non-food expenses expressed as percentages of total household resource outflows.

Figure 1. Food and Non-food Expenses Expressed as Percentages of Total Household Resource Outflows

Most households in the community get nearly all of their non-grain food from the market. Sorghum is in most cases the only crop grown (though some people grow maize and a few small crops in their home gardens: these crops are usually consumed immediately upon harvesting). Because of this heavy reliance on the market, per person food expenditures do not change greatly from month to month unless there is an unusual event such as a celebration (HH 2 and 9 hosted funerals for extended family members) or a family member becomes sick and requires hospitalization (HH 3 had a sick child who spent three weeks in the hospital with his parents).

Food expenditures may also be influenced by changes in the non-food expenditure levels, which are subject to significant changes during harvest season. In all of the landholding households, non-food expenditures were much higher (recorded on an individual household level at between 20% and 70%) during harvest time than later on in the year when the harvest was finished. Most of these non-food expenditures can be attributed to the costs incurred by the harvest: purchase of grain sacks, repayment of loans incurred against the harvest during the growing season, and payment of workers. Other major expenses include clothing and shoes, items for which at other times of the year people do not have enough money.

Households whose non-food expenses were very high one month tended to reduce their food expenditure levels the next month. This reduction in food expenditure levels (in most cases manifest by consumption of cheaper and less nutritious kinds of foods rather than missed meals) reveals an important strategy used by households to recover relatively quickly from the increased cash outflow required by the harvest, a strategy which may well enable them to reach the next harvest season without encountering serious food shortage.

Cost of Living in Ada Bai

Despite the obvious limitations of such a small sample size, it is still worthwhile to examine the monthly per-capita cost of living as reported in the ten households. Table 2 shows average per person expenditures in the four major categories and as totals. Special notice should be taken of the increase during months 1 and 2 due to increased farm expenses and cost of such items as clothing.

Table 2: Average Per Person Cost of Living, Based on Resource Outflows in the Ten Survey Households[10]

Month   Total Food    Total Non-food  Repaid     Gift Given   Total Out      
  1    31.40          28.00           7.20       1.40         68.00          
  2    28.80          48.80           15.80      1.00         94.40          
  3    20.40          10.20           13.80      1.40         45.80          
  4    26.40          19.80           8.80       .40          55.40          
 Avg   26.75          26.70           11.40      1.05         65.90          

These totals should only be taken as crude estimates of the cost of living per person in Ada Bai. Actual per person expenditures reported varied widely, and are influenced, among other things, by a household's food stores prior to the start of the study, farming and nonfarming income generation activities, and the amount of available cash and credit.

Income Levels

Figure 2 below shows the aggregated sources of income for the ten households. Income is defined as a combination of wages earned and items sold. Harvest figures are for sorghum.[11] The table shows that after the first month, the level of borrowing drops dramatically. This is a reflection of the fact that people borrow early in the harvest season in order to cover their costs. Their income then rises during and immediately after the harvest, as farm products are sold quickly to pay off the household's debts and to purchase household items which people have gone without for a long time. For the most part, people are unable to store their grain until the price rises (usually several months after the harvest season) as their creditors demand immediate repayment.

Figure 2: Sources of Income of the Ten Survey Households

The sharp drop in borrowing after the first month of the study may not be entirely explained by a drop in demand for credit: lack of available credit may prevent some people from borrowing after the harvest season ends even if they are experiencing financial difficulty and would otherwise wish to borrow. Respondents reported that it was much easier to find people to borrow from during harvest time than in later months. Many lenders advance money against the promise of a certain amount of the farmer's harvest (usually at significant advantage to the lender). Those households who did not repay their debts promptly reported having difficulty finding people who were willing to continue to lend to them. In some cases, they were obliged to sell their high-value property as well as the grain from their harvests to cover household expenses.

Of particular concern is the sharp drop in earned income in the last month. This is largely attributed to the lack of available waged labour in the area in the months following the harvest. Those households who had enough income or harvest in the earlier months to get them through this period were able to draw on their reserves in the post-harvest season. Those who did not have enough either borrowed money or migrated to the Sudan in search of work.

The slight rise in gift-giving shown during Months 3 and 4 (Tiri and Lekatit) in Figure 2 can be attributed in large part to the holiday seasons for both Christians and Muslims. Christians celebrate Christmas (Lidet) and Epiphany (Timkat) during Tiri. The month preceding the Lenten fasting season (Lekatit) is traditionally a time in which many Christian weddings are held since people have more grain than at any other (non-fasting) time of the year. Guests are obliged to bring gifts of grain or cash. For Moslems, the Ramadan season is a period of increased gift-giving. In addition, all households contributed food and/or cash to the celebration of the anniversary of the Tigray People's Liberation Front (TPLF) on Lekatit 11.

Sale of Assets

Several of the households reported selling their assets in order to pay back loans or cover their basic expenses. Three households (HH 1, 8, 10) sold high-value property (2 tape recorders valued at 350 to 500 birr, a sewing machine for 2000 birr and a large lamp for 150 birr) to pay back their loans to the Women's Association (of 550 birr each). Another household (HH4) also sold a tape recorder to get enough money to support the household while the husband went to Sudan to look for work.

As common as sale of assets was, the sale of grain and other farm products was more commonly practised by the households to cover basic living expenses. Table 3 shows the size of the household harvests and the amount sold by each household in the first three months following the harvest. One household (HH 5) whose harvest was 10 quintals had sold 6.25 quintals by the end of Lekatit, and another (HH 8) with a harvest of 11 quintals sold 6.75 quintals.[12] Going into the driest part of the year, there was very little food left in these households to hold them over until the start of the rainy season, when wild plants become available.

Table 3. Levels of Harvest and Grain Sales (Sorghum)

   Household     Grain from       Grain Sold,          Stored in the House at   
    Number       Harvest          Hidar-Lekatit        end of Megabit (mid      
                                  (Nov-Mar)            Apr)                     
       1         5.0              1.25                 0.5                      

     2*10        0.08.0           -1.00                0.52.5                   

* These households bought sorghum from the market or borrowed from neighbors since they had no significant harvest.

As can be seen in Table 3, every household that had a harvest of more than one quintal sold grain to meet its household expenses. Those interviewed said that ideally they should have twice as much harvest as their basic food requirements (2.5 quintals per person per year) so that they could sell enough grain to cover their cash requirements, thereby becoming completely self sufficient by working on their own farms. Such large harvests are probably not realistic given present land tenure arrangements, grain prices, and investment costs. In most households there will always be a need to seek waged labour or supplement their income with nonfarm work. Six of the households surveyed reported having worked as daily labourers on the commercial farms near their homes; those who had supplementary income did not engage in waged farm labour. Nevertheless, people said that they would put their efforts first into their own fields and later into wage labour and other forms of economic diversification.

The preliminary report issued after the first month of data collection sounded a warning that if rates of borrowing continued at the rate observed, a severe cash and/or food crisis could ensue. While households are indeed slipping into indebtedness, further data shows that households are able to maintain at least some measure of control over their slide. The disturbing fact that they are selling high value goods to repay their debts shows that investments made with borrowed money tended not to be successful, but the household's ability to tighten its resource outflow through reduction of borrowing and expenditure levels, as well as resorting to less nutritious foods suggests that at least in the short term households are able to change their management strategies so as to effectively curb their slide into indebtedness. It remains to be seen whether in the next year they will have enough access to credit and will be able to manage their farming and investments differently so as to produce a more advantageous return on them, thereby recouping some of the 1993 and 1994 losses to their asset bases.

Sources of Credit and Patterns of Repayment

Access to credit and conditions for successful investment are among the biggest constraints to economic security in Ada Bai. Inability to invest adequately in farmland and businesses prevents reliable and sufficient returns on the small investments that are made. Credit sources can be divided into two groups: those from inside and those from outside the community. Outside credit plays an important role in providing larger loans to returnees, but comes with strict rules for repayment to which people are not accustomed and may not always take seriously. The failure of the community to generate its own credit coupled with its members' inability to repay larger outside loans highlights the economic vulnerability and slow economic reintegration following repatriation.

A. Intra-community Sources of Credit

In other parts of highland Ethiopia (in both Amhara and Tigrayan societies), mutual assistance networks help to create a safety net for peasant households. One formal association which provides assistance to member households in which there has been a death is known as iddir. Assistance may include preparing the body of the deceased for burial, covering the actual costs of the funeral, providing food during the seven day wake known as hazan, and contributing money to the survivors of the deceased for food, clothing, etc. Iqub is a savings association whereby members contribute money and distribute the funds in full each month or each week to a different member, whose name is drawn in a lottery. Mehaber is an association of neighbours which gathers on a particular saint's day to feast together, each month at a different neighbour's house. While not usually directly intended to distribute funds, people often turn to other members of the mehaber for credit or assistance in times of hardship. In the absence of a formally recognized iddir, members of one's mehaber may assist at a funeral.

In Ada Bai, most of these traditional associations are no longer practised. People say that iddir was for the most part practiced only in the towns, and that most rural people have never been part of one before. They report that while in the Sudan they tried to keep iqub and mehaber going, but had to abandon them upon repatriation because they did not have enough money to participate. There are reportedly only two iqub functioning now in Ada Bai among the traders who have enough money to participate. Mehaber were almost completely suspended during the first year after repatriation as people claimed that they did not have enough grain or money to participate. They operate occasionally now, particularly at harvest time or following the distribution of grain rations. They are usually on a much more informal basis, however, and the obligations that used to bind mehaber members to each other to provide mutual assistance is in some cases not as strong as it used to be.

By far the most common form of credit in the community comes from small loans of food, money and such goods as kerosene, charcoal, and labour. Usually such loans are so small that those who give them rarely expect repayment. To ask for repayment of a loan of one birr or two injera would be considered incredibly rude. Requests for small amounts of credit from one's relatives or immediate neighbours likewise cannot be politely refused. While respondents of the survey tended to report most cash that they "borrowed" from their neighbours, it is likely that the rate of borrowing of food items was much higher. In particular, Household 7 regularly borrowed salt and berbere from neighbours during Tiri (January) but did not report the full amounts that she borrowed. These formed the staple of her household's diet since she could not afford pulses, but she said afterwards that she did not consider them important enough to record daily since she borrowed in such small amounts and had no intention of returning them to the people who had given them to her. Neither did she consider them to be gifts. It is probably impossible to track such small inter-household exchanges with complete accuracy.

For larger amounts of credit, most people look to the town's business people for loans. These include professional moneylenders who offer loans at high rates of interest. The interest rate (harratsa) is 10%, but many people also borrow by pledging to give part of their crop (sesame or sorghum) to the lenders at harvest time. In this way, lenders are able to buy grain at extremely low prices and make large profits later in the year. Most borrowing from moneylenders is done secretly since they are not authorized to charge interest.

Probably more than they use professional moneylenders, local people turn to the town's shopkeepers for credit on purchased merchandise and cash loans. In most cases, the shopkeepers do not charge interest. They lend money out as a personal service, to the extent that they may even bring economic difficulty on themselves.

One shopkeeper who records all of the money that is owed to him says that in the three month period from Tahsas to Lekatit (mid December to mid March), he extended credit and cash worth 4115 birr. Of that, only 1757 birr was repaid in a short period of time. Another said that since Meskerem (mid September 1994) he has lent 4326 birr and only 1968 birr has been repaid. He described the situation in this way: "I feel upset that I lend too much. Some people owe me from two years ago. But some people can't pay and I can't ask them to pay. We were together for eight years (in the Sudan). If they have some problem, I am not going to force them to pay...This year is better than last year. Most people are repaying their loans. Last year hardly anyone paid me back. I went to their houses and I found nothing. So I canceled all the debts and started over." Indeed, the most successful traders are expected to lend money and if they do not people may not frequent their business as much as others who do extend credit.

Patterns of loan repayment to private individuals and shopkeepers within the community seem to follow an "ability to pay" trend. Debtors repay larger loans when they feel that they have enough money to spare. They are under no financial obligation to repay promptly, though they may do so in order to maintain friendly relations with the lender or to ensure that they will be able to borrow from the same person in the future. Shopkeepers very rarely insist upon immediate repayment, especially if they know that the debtor is experiencing financial difficulty.

B. Loans from Outside the Community

Loans from outside the community are usually sought to enhance a household's income generating power in a more substantial way than the small loans that come from within the community. In 1994, there were at least five external sources of credit.

Limited loans were made available to farmers by the regional government, with the baito acting as its distribution and collection agents. These loans were intended for seeds and tractor ploughing, though there was no mechanism for monitoring or controlling what people actually spent the money on. Many farmers say that they used the money to make their own farming arrangements. A large number of those who were allocated land said that it was overgrown with trees and there was insufficient time to clear it, so they rented or entered into sharecropping arrangements with Humera area landholders. Such arrangements did not result in a large enough profit to enable them to repay their loans. Local officials report that due to the low level of repayment they do not expect the loan programme to be renewed in 1995. Farmers will have to make their own arrangements this year for ploughing.

Hiwot Agricultural Company, the largest landholder in the area, has been offering tractor ploughing on a rental basis, the fee to be paid at harvest time with a 2% interest charge. It is not yet clear whether this offer will be made again to the returnee farmers for the 1995 agricultural season.

As mentioned above, the local women's association (through the regional Tigray Democratic Association of Women) extended loans of 500 birr at 10 percent interest to approximately 140 Ada Bai women. TDAW representatives in Mekelle said in October 1994 that women in other parts of Tigray have a very high rate of repayment (80% paying their loans back on time), which allows for the money to be loaned out again as a revolving fund. In Ada Bai, however, local women's association representatives say that the rate of repayment has been very low. Among the most common excuses for failure to repay:

"We bought goats (or other livestock) but they died."

"We sowed sorghum with it but are waiting for the price to go up before we sell our harvest."

"Our husbands are working in the Sudan and we will repay when they come home."

"We bought gold from the Sudan to sell here, but the quality is not as good so we were not able to sell it for a good price."

"We opened a sewa shop with the money but the location is too far from the central market so we have not been able to sell enough to repay the loan."

"We thought we would get a repatriation cash grant like the returnees who came after us did, and we were counting on that to repay the loan."[13]

One of the local women's association representatives responsible for collecting loan payments said, "If you look in their houses you see nothing. They have no hope of paying (the debt) back." Because of the low rate of repayment, the loans will not be offered to Ada Bai women in 1995.

While local people obtain small-scale credit from traders and shopkeepers, these local business people must look outside the community for credit for themselves. Several have taken out loans from the Guna Private Share Company but are having trouble repaying them. The Commercial Bank of Ethiopia has started to offer loans to business people, but is reluctant to do so unless the traders have no other outstanding debts. In both cases, the pressure to repay the loans is increased by the fact that the lessor is a large company situated outside the community. The usual intra-community custom of forgiving late or missed payments is not extended in any of the cases where credit comes from outside the community.

As outlined in the earlier report, those who borrow are not necessarily the households who have the best credit line. Very often it is the poorest households who borrow most. This is particularly true in the case of intra-community borrowing. People draw on "social capital" when they borrow from friends and relatives, knowing that the rules of society ensure that they will not be refused. They borrow knowing that they do not have the collateral to ensure that they can repay the debts. As the loan becomes more formal, and borrowers are forced to exhibit evidence of sufficient collateral, the poorest members of the community may find that they are unable to qualify for loans. Thus the more wealthy members of the community are the only people who can borrow from the bank or from the large companies in the area.

Labour Migration

Labour migration has been practised for a long time by Tigrayan farmers to supplement their incomes. Since its development as an agricultural center in the 1960s, Humera has been a magnet for labour migration. This community of former refugees is intimately familiar with labour migration, as many people during the war and famine of the 1980s first migrated to western Tigray in search of work before giving up and seeking aid in the Sudanese refugee camps. While living in the camps, many men and women worked in the irrigated fields in Eastern Sudan.

Since repatriation, returnees have continued to exploit labour opportunities on both sides of the Ethiopia/Sudan border. Indeed, labour migration has become an important form of supplementary income for many Ada Bai households. In October and November, men and women who are not busy with their own fields work for wages in the large sesame and sorghum plantations which surround the town. They are joined by migrants who come from the highlands of Tigray and Gondar, as well as a large number of Sudanese. From December to April, Ada Bai men return in large numbers to the Sudan to work harvesting the irrigated sorghum and cotton fields.

It should be recognized that while it is usually men who migrate to work, women engaged in the food, beer-selling, hotel and prostitution businesses (and these are usually not mutually exclusive) experience a dramatic increase in customers during the harvest season. Likewise, when many men leave the town to work in the Sudan, the women complain that business slows down dramatically.

Of the ten households, only one sent a member to work in the Sudan during the survey period. At the end of the survey (mid March 1995) he had not returned, though he had sent money to his wife and children. For the purposes of this report, other Ada Bai men were interviewed who had gone to Sudan in the last harvest season.

Berhe, 18-years old, reported going with six of his friends to the Gedaref area to work in Tahsas. They received a contract to cut and collect sorghum. For twenty days' work, they earned 230,000 Sudanese Pounds between them.[14] After that they were unable to find another contract because there were so many workers, so they worked for daily wages in one place for three days for 700 pounds per day (approximately 9 birr/day), and at another for five days at 600 pounds per day (approximately 8 birr/day). Before returning home, Berhe said that he bought new clothes for himself and 14 jerry cans to sell in Ethiopia. At the border he was taxed by the Sudanese police 200 pounds per jerry can, but he said that if he sells them for 22 birr each, he will still make a profit. After exchanging the money on the currency market in Humera, he said that he had 52 birr to bring home to his family (not including the proceeds from the jerry cans).

Girmay, aged 30, went to Sudan in Tiri (January). He also traveled with six other people to the Gedaref area. Because of his past experience working in the fields, he found work feeding a combine harvester (considered a skilled job). Workers on the combines are paid 80 pounds for every two quintals of sorghum fed into the machine. Girmay said, "I did the same work in previous years in the Sudan. This year was very good because the pay rate was high even though the exchange rate was low. I bought clothes for my family and brought the rest of the money home...I got about 300 birr in the end."

Those interviewed insisted that they preferred not to go to the Sudan to work. Such migration takes them away from their families for at least one month and often as long as four or five months a year. They insist that their priority is to make their own land productive enough that they will not have to go to the Sudan next year. In this sense, then, labour migration is still very much considered a strategy to be undertaken only in times of economic distress. Returnees' experience and continuing contact with the Sudan makes them well placed, however, to exploit migrant labour opportunities.


This report attempts to chart out the range of economic management strategies available to and practised by Ada Bai returnees, and to examine the ways in which households make decisions and allocate in order to maximize their scarce resources.

Not included in this report, but to be taken into consideration in the final analysis of the survey results, is information relating to food consumption levels and patterns. Of the ten households, it is suspected that the diet of at least two (HH 1 and 7) is seriously deficient in protein, energy, and micronutrients. This analysis will be carried out during the coming months.

Given the highly market and wage labour oriented nature of the Ada Bai economy, there is ample reason to consider the needs and opportunities available to returnees in the Humera as being qualitatively and quantitatively different from those of their kin in the highlands of Tigray. These differences must be taken into account in the planning and implementation of development strategies at the regional level.

To recount the findings of this report:

During harvest time, returnees' household budgets are strained by having to cover farming costs and repay debts incurred during the agricultural season.

In order to meet these commitments, returnees tend to borrow beyond their capacity from a variety of sources, both inside and outside the community.

To mitigate the effects of this over-extension of indebtedness, returnees employ a variety of strategies in the post harvest season, including constriction of spending levels, sale of high value assets, sale of grain reserves, and labour migration.

While these strategies help to slow the process of erosion of the household's asset base, an overall trend of diminishing property and resources can be detected.

This process indicates weakness in the ability of the community to generate its own sources of credit and lack of capacity on the part of individual houses to make substantial returns on their investments.

The need was therefore identified to help returnees to generate income internally (as a community) in order to increase number of credit options available, and to enable them to meet the repayment demands of such credit schemes.

The feeling on the part of returnees is that improved access to credit and ability to repay will come when there are adequate income generating opportunities within the community. To some extent, these opportunities will arise from within the community as its residents become more settled. Already, the number of mills and other businesses has increased over the 1994 levels. However, outside assistance can also help this process.

In early 1995, UNHCR funded the establishment of a small vegetable garden project to be implemented by the Women's Association of Ada Bai. This project, conceived of and planned by the individual members of the Women's Association, is designed to generate funds for the Association to be used as small scale loans for needy women. Still in its early stages, it is still too early to judge the effectiveness of the project, though initial indications are encouraging. There is a need for similar projects to be implemented in all of the returnee settlements so that returnees can generate the income to help themselves.

[1] This report is part of a larger study being conducted by the author for a PhD in anthropology from the University of Wisconsin-Madison, USA. Special thanks are due to Ato Abebe Bekele, without whose research assistance this study would not have been possible.

[2] E.C. is used to connote Ethiopian Calendar. Month names are given in Tigrinya.

[3] A report entitled Household Economic Survey, Ada Bai Returnee Settlement, Humera, was distributed in January. This report is available at the UN-EUE.

[4] Widespread food shortage in March-April 1994 reportedly resulted in the deaths of several adults in all three of the returnee communities and prompted emergency requests for food aid. The Relief and Rehabilitation Bureau (RRB) sent 5400 tonnes of imported wheat to meet the needs of 9000 people in the three settlements for four months (at 15/kg/person/month). See (Hammond: 1994).

[5] Sewa is local beer, known also as tella in Amharic.

[6] Local governing council.

[7] Figures reported by the baito.

[8] Each household is given a plot 16mx20m on which to build their houses.

[9] Some of the findings reported here differ from those presented in the preliminary report of the first month's findings. This is due to revision of certain data provided by respondents.

[10] These figures are derived by calculating the monthly resource outflows of each of the ten households during each of the months, obtaining an average household expense per month, then dividing by the average household size, which is five.

[11] Based on average price of 100 birr/quintal. Sesame is not included in the calculations, as only two of the households planted sesame, and both harvested it prior to the start of the survey.

[12] Based on cash received for sale of grain and average price of 100 birr per quintal.

[13] The 1993 returnees were repatriated by the Transitional Government of Ethiopia, with the Relief and Rehabilitation Bureau and Relief Society of Tigray implementing assistance (the United Nations was not directly involved). Due to lack of resources, there was no arrangement made to provide cash grants to these returnees. UNHCR signed agreements in 1994 an 1995 to fund assistance to returnees, including cash grants.

[14] 1,000 pounds=13 birr. 230,000 pounds=2990 birr, or 427 birr per person.

Back to document index

  • Disclaimer for EUE Materials

    Editor: Ali B. Dinar, (