Survey of worker shelters
The monthly average number of workers in each shelter ranged between six and ten. The majority (> 80%) of workers were males and most of them (> 90%) had stayed in the shelter the previous night. At the time of the study, none of the female workers reported being pregnant. The majority of the workers had either no education or only a primary level of education and this pattern was similar for each month.
The study team did not witness construction of new temporary worker shelters at the selected farms. Migrant workers stayed in temporary shelters established in previous years so there was no monthly variation of shelter construction. The roof of most (63%) shelters was grass while walls mere made of grass, iron sheeting, plastic sheeting or bamboo (> 90%). Most shelters had open eaves (94%). Surface water was the main source (75%) of drinking water (Table 2).
Table 2
Characteristics of migrant worker shelters (N = 16)
Shelter characteristics
|
Number (%)
|
Material of roof:
|
Grass
|
10 (63)
|
Thatch
|
1 (6)
|
Iron sheet
|
4 (25)
|
Other
|
1 (6)
|
Eaves:
|
|
Open
|
15 (94)
|
Closed
|
1 (6)
|
Material of wall:
|
Grass
|
7 (44)
|
Plastered
|
1 (6)
|
Other (Bamboo, iron sheet, plastic)
|
8 (50)
|
Source of drinking water:
|
Surface water
|
12 (75)
|
Protected well
|
1 (6)
|
Hand dug well
|
3 (19)
|
None of the shelters had window screening and all the floors of the shelters were made of mud. The number of rooms per shelter ranged from two to eight, but varied slightly between months due to modifications made to accommodate influx and outflux of migrant workers. Almost all rooms in a shelter were used for sleeping. The mean number of sleeping places per shelter was three and ranged from two to 11. No shelters had latrines, and firewood was used for cooking in all locations. All shelters had at least one individual with a mobile phone. Only two shelters had individuals with their own transportation (a donkey and a truck).
Out of 16 shelters, 12 (75%) owned at least one mosquito net in September. (Fig. 3). There was one mosquito net for every 7 to 14 migrant workers in each shelter, depending on the month. Three-quarters of the bed nets available in September were less than 36 months old (Table 3). Most of the bed nets were hung above elevated sleeping benches made from long poles. Only blue, rectangular bed nets were found in the shelters, with PermaNet as the dominant brand. Most bed nets had been obtained through a health facility or community agent free of charge. All bed nets were used by someone the previous night and the majority were shared by two or three people. No distinction was noted among farm types (large and small) and this applies to all sections of the result.
Table 3
Characteristics of available bed nets
Characteristic of bed net
|
Number of net (%)
|
September
|
October
|
November
|
December
|
Number of workers for each net:
|
9
|
7
|
10
|
14
|
Months since obtained:
|
≤36
|
10 (77)
|
14 (88)
|
10 (59)
|
1 (17)
|
>36
|
3 (23)
|
2 (12)
|
7 (41)
|
5 (83)
|
Sleeping place where net used:
|
Bed frame (finished)
|
2 (15)
|
0 (0)
|
1 (6)
|
0 (0)
|
Bed frame (sticks)
|
11 (85)
|
16 (94)
|
14 (82)
|
6 (100)
|
Ground
|
0 (0)
|
1 (6)
|
0
|
0 (0)
|
Reed mat
|
0 (0)
|
0
|
2 (12)
|
0 (0)
|
Brand of net:
|
PermaNet
|
12 (92)
|
16 (94)
|
16 (94)
|
6 (100)
|
MAGnet
|
1 (8)
|
1 (6)
|
1 (6)
|
0 (0)
|
Source of net:
|
Health Facility
|
4 (31)
|
4 (24)
|
6 (35)
|
1 (17)
|
Community Agent
|
4 (31)
|
6 (35)
|
6 (35)
|
3 (50)
|
Family or friends
|
1 (8)
|
4 (24)
|
2 (12)
|
0 (0)
|
Shop or supermarket
|
2 (14)
|
2 (11)
|
0 (0)
|
0 (0)
|
Market
|
1 (8)
|
1 (6)
|
3 (18)
|
2 (33)
|
Other
|
1 (8)
|
0 (0)
|
0 (0)
|
0 (0)
|
Paid money for net:
|
Yes
|
3 (23)
|
4 (24)
|
3 (18)
|
2 (33)
|
No
|
10 (77)
|
13 (76)
|
14 (82)
|
4 (67)
|
Don’t know
|
0 (0)
|
0 (0)
|
0 (0)
|
0 (0)
|
Net location at time of interview:
|
Hanging loose over bed/mattress
|
10 (77)
|
12 (71)
|
11 (64)
|
5 (83)
|
Hanging and folded up or tied
|
3 (23)
|
5 (29)
|
3 (18)
|
1 (17)
|
Not hanging but not stored
|
0 (0)
|
0 (0)
|
2 (12)
|
0 (0)
|
Stored away unpacked
|
0 (0)
|
0 (0)
|
1 (6)
|
0 (0)
|
Someone slept under the net the previous night:
|
Yes
|
13 (100)
|
17 (100)
|
17 (100)
|
6 (100)
|
No
|
0 (0)
|
0 (0)
|
0 (0)
|
0 (0)
|
Number of individuals using the net the previous night:
|
1
|
4 (31)
|
7 (42)
|
6 (35)
|
2 (33)
|
2
|
4 (31)
|
5 (29)
|
8 (47)
|
0 (0)
|
3
|
5 (38)
|
5 (29)
|
3 (18)
|
4 (67)
|
(Table 4 location: after line 251) |
Human behavior observations in the worker shelters
The number of migrant workers observed each month ranged from 162 in November to 84 in December. From approximately 20:00, there were always more people already indoors compared to outdoors or compared to those away from the shelter (Fig. 4). Nearly 50% were already outdoors or away from the shelter by 7:00 the next morning in September, which increased to 75% in November. The percentage of workers away from the shelter increased in later months compared to September (Fig. 4). During November and December, approximately 20% of workers remained outdoors during the whole night. Overall, an average of 87% of workers were indoors at midnight in all four months of observation.
More than 70% of migrant workers were not using bed nets during the observations (Fig. 5). The highest percentage of workers recorded sleeping under a net at any one time was 24%. Those with access to bed nets began using the bed net from 20:00 at night and the majority of those with nets were underneath a bed net by 22:00. The workers had all left the nets by 7:00 the next morning with the majority leaving the protection of the net at 6:00. This pattern was similar across all four months (Fig. 5). No worker was observed using other mosquito prevention interventions except for three individuals who tried to repel mosquitoes in the evening using smoke from firewood.
The most common activities among migrant workers after 17:00 were chatting, sleeping, cooking and eating. Up to 40% of workers were still away from the shelters until 20.00 (Fig. 6). Though almost all workers were sleeping by 22:30 in September and October, 10% of migrant workers were chatting or away from the shelter at 22:30 in November and December (Fig. 6). Overall, an average of 98% of migrant workers were sleeping at midnight and 98% were already awake at 08:00 the next morning during the four months of observation.
Human behavior observations in the fields
The most common field activity observed between 17:00 and 9:00 was harvesting (63%). Other field activities included caring for animals and storing harvest. The average number of farmers conducting field activities during the night was seven and ranged from two to 24. The duration of field activities throughout the study ranged from three hours to 16 hours with an average of 14 hours. Nighttime field activities started at 17:00 to 19:00 (usually at 17:00) and ended the next morning (8:00 to 9:00).
Almost all migrant workers who were working in the field were male adults. They wore shorts and short-sleeved shirts. They had sandals (‘’ergendo’’) as footwear and covered their head with a scarf. They wore head torches at night. None of the observed workers used any personal protection measures against mosquitoes while in the field.
Results of post observation survey
Sixteen shelter representatives responded to the survey following the night observations in September (Table 4), and the survey results were consistent across all months. Half of the respondents slept under nets and reported going to bed at 21:00 the previous night. The majority reported that they slept indoors with half of respondents inside by 20:00. Almost none of the representatives slept outdoors in any of the months. None of the representatives used preventive measures to prevent mosquito bites between 18:00 and 20:00, but three wore long-sleeved shirts after 20:00 to prevent mosquito bites in September.
Table 4
Post observation survey respondent behavior indicators for the month of September
Survey respondent behavior indicators
|
Number (%)
|
Slept under net last night:
|
Yes
|
7 (44)
|
No
|
9 (56)
|
Time went to bed previous night (hour):
|
20:00
|
3 (19)
|
21:00
|
8 (50)
|
22:00
|
4 (25)
|
23:00
|
1 (6)
|
Time woke up in morning (hour):
|
01:00
|
1 (6)
|
05:00
|
2 (13)
|
06:00
|
5 (31)
|
07:00
|
8 (50)
|
Slept indoor/outdoor:
|
Indoor
|
15 (94)
|
Outdoor
|
1 (6)
|
Location between 18:00 and 20:00:
|
Primarily inside
|
8 (50)
|
Primarily outside
|
8 (50)
|
Activities between 18:00 and 20:00:
|
Eating
|
8 (50)
|
Chatting
|
7 (43)
|
Cooking
|
1 (7)
|
Activities between 20:00 and the time you went to bed:
|
Chatting
|
6 (38)
|
Eating
|
6 (38)
|
Sleeping
|
1 (6)
|
Drinking alcohol
|
2 (12)
|
Watching TV
|
1 (6)
|
Bite prevention attempts between 20:00 and time worker went to bed:
|
Wore long sleeves/pants
|
3 (19)
|
None
|
13 (81)
|
Location between the time you woke up and 07:00:
|
Primarily inside
|
10 (63)
|
Primarily outside
|
5 (31)
|
Away from home
|
1 (6)
|
Activities between the time you woke up and 07:00:
|
Taking bath
|
2 (12)
|
Chatting
|
5 (31)
|
Washing clothes
|
1 (6)
|
Eating
|
2 (13)
|
Going to church and others
|
6 (38)
|
Bite prevention activities between the time woke up and 07:00:
|
Wore long sleeves/pants
|
3 (19)
|
Used a net
|
2 (12)
|
None
|
11 (69)
|
Qualitative Interviews
Almost all the workers responding to interviews came from West Gojam, East Gojam, Awi zone and South Gondar zones of Amhara region. Most arrived at the farms in June and returned home by December, although some reported they may stay later than December to harvest peanut and cotton. However, all respondents agreed that the conditions would be unfavorable for work after December when the weather becomes hotter and rivers dry out, making it difficult to obtain drinking water.
Workers find employment in two ways: either through an agent or on their own. Workers contact agents in their hometown or they travel to the small towns near the farms to arrange work with the agents. Recruitment through an agent is often based on prior written agreements between the agent and the farm employers, which help to safeguard the workers’ rights and benefits (in contrast to those that are self-employed). Those who search for work on their own travel directly to the farms or congregate at specific locations in the nearby towns waiting for employers to request labor. Only a few workers reported having a family back home and none brought families to the farms. All agreed that the living conditions at the farms were not suitable to accommodate a family.
Living conditions of migrant workers:
All workers reported staying in large, temporary communal shelters which were constructed by wooden frames covered with grass or iron sheeting (Fig. 2). The shelters have no doors, so the structures always remain open to the outside. Most workers try to cover the floor with plastic sheets to sleep while some workers prepared a bedding area to raise themselves above the ground to gain protection from snakes and scorpions. Most farms have a separate dedicated housing area for cooking. During peak times, the respondents reported that the number of inhabitants per worker shelter could exceed 100 workers each night. Sleeping spaces are distributed on a ‘’first-come first-served basis’’. Workers could be forced to sleep outside when shelter gets full, especially during the hot season. Some workers prefer to stay away from the shelters and sleep in the fields. One participant explained his reason to avoid sleeping in the shelters as follows:
’….. I sleep on the field. It is smelly inside. I enter the temporary shelter when it rains.
Most respondents worked every day of the week except Sunday and religious holidays. Working hours usually started at 07:00 and ended at 18:00, though some reported working at night (especially during harvesting). Work at night does not usually continue past midnight though some of the workers sleep in the field after nighttime work.
The farm workers were divided into two groups: temporary and permanent. Temporary workers (who were the majority) were either paid daily (fixed amount in a so-called labor scheme) or paid per hectare of weeding or harvesting (called a contract scheme). There were also groups of workers who arranged with farm owners (especially in small farms) to share in the profits from the crops. One worker explained a typical day at the farm as follows:
’….. When you do contract work (payment per hectare), you will have your own schedule. When you work as a laborer, you will have a fixed schedule. However, all farmers wake up in the morning (at 06:00), work till midday, rest for lunch and then work starting from 14:00 to 18:00. Some contract workers work in the night to avoid the overhead sun. We go to sleep between 22:00 and 23:00.
Sleeping times were variable. However, most workers reported going to sleep between 21:00 and 23:00. Most said they sleep earlier during times of harvesting, as they are aiming to arrive early to their worksite the following morning. The type of produce could also affect sleeping schedule. One worker said:
’….. During sesame harvesting, I sleep during the daytime and work at night. Sesame harvesting is done at night (you may get sick if you do it on sunny days and the seeds will fall off too). I work in the night till 03:00 and then go to sleep. We wake up at 10:00 and start cooking our food. During other times, we go to sleep between 20:00 and 21:00. We wake up at 06:00. Lunch time is from 12:00 to 14:00.
All workers agreed that they had better living conditions in their home towns but that they have to accept the difficult conditions working at the farms to earn money and improve their lives. However, some mentioned that they could not achieve their goals. The cost of living (including medical costs) was high on the farms, limiting their ability to save money.
Perceptions about malaria and its occurrence among farmers:
Most workers believe that malaria is caused by a poison from mosquitoes and that the mosquito transfers the poison from person to person (Fig. 7). Some workers also explained how the mosquito has naturally occurring poison similar to snakes. One worker described it as follows:
’….. The mosquito bites us and it introduces its poison into our body.
Exposure to cold, lack of food and poor personal hygiene were also reported as the other causes of malaria by workers. All workers assumed that they were at risk for malaria while employed on the farm. One worker elaborated about the risk of malaria while in the farms as follows:
’….. The risk is higher here. We take the malaria from here to our hometown. There is no malaria in our hometown.
Many workers reported several incidences of malaria in recent months and some of them were on antimalarial medication during the interview. Almost all workers reported febrile illness after arrival and they considered their hometowns to be safer locations. Some workers also said illnesses were more common among newcomers. All respondents agreed that malaria was the most common diagnosis among the migrant worker population, followed by diarrheal diseases.
Knowledge on malaria prevention and availability of prevention kits:
Most workers agreed that it was possible to avoid becoming sick from malaria while working on the farms. Using bed nets was mentioned as an important prevention tool and many hoped that obtaining one would bring some relief from the repeated attacks of fever and headache from malaria. Some workers described the malaria-free intervals they had enjoyed while using bed nets in the past (previous years or in other locations). They also pointed to the importance of other prevention measures such as eating good food, drinking clean water, keeping good personal hygiene and environmental hygiene (which was defined as draining swamps, cutting the grass or removing garbage) and prompt medical treatment. Other rarely mentioned prevention methods included the use of smoke from firewood and farm chemicals which were reported to help chase away the mosquitoes.
Few informants reported owning a bed net and none of the farms provided bed nets to workers. The workers were not allowed to get bed nets from the local health posts which required people to show identification to prove that they were from the local area. All respondents agreed that the only option for obtaining bed nets was by purchasing them from local drug shops or from local community members. Prices to purchase bed nets ranged from 40–80 ETB ($1.5-$3 USD) and purchasing bed nets usually involved traveling to nearby towns. In contrast, respondents said they had free bed nets provided at their original homes. The reason why they travel to the farms without their bed nets is that they usually share their bed nets with family members and did not want to leave them unprotected.
Most workers complained that they were never given a bed net while working on the farms. The few bed nets owned by some of the workers were old and torn and were either bought or borrowed from a friend. One worker explained his account as follows:
’….. I have a bed net but it is old. It has holes and doesn’t kill the mosquitoes when they land. I got it when I was working for an Indian company.
’...We can’t get bed nets here. They don’t give a bed net for migrants. They needed a local ID. We have been asking the health extension workers in my previous farm but didn’t succeed.
Respondents said they tried to cover their whole body with their cloth while sleeping to protect from the mosquitoes and the nuisance of biting insects which is more intense during and immediately following the rainy season. However, they still find their legs, hands and face covered with bites in the morning.
Farm managers and health workers did not deny that there was low utilization of bed nets at the farms, mostly due to lack of access. They confirmed that bed nets have never been distributed to migrant workers except when the district health administration rarely supplied bed nets to the farms based on a request from the farm managers. The health workers said the provision of community health and malaria prevention services were household-based and that this has contributed to the exclusion of the migrant population. In addition, many migrant workers were working far from community health services, making it logistically challenging to reach them.
Getting prompt and proper treatment:
Most workers agreed that malaria can be treated, although some workers seemed ambivalent. Their doubt was caused due to repeated malaria illness despite their attempt to find treatment in the past. Some workers tried to explain their stance as follows:
’….. It can be treated. But there is no complete cure. The treatment is temporary. The malaria will come again as long as we are here in the farm.
Fever and malaria illness are overwhelmingly common. Almost all respondents reported at least one malaria illness per month. Most workers knew the symptoms of malaria and seek prompt care though none of the farms have clinics for immediate care and treatment. When workers contracted fever they traveled to the nearby health centers and health posts (which could be up to 30 km away). Some farms occasionally arranged transport for ill workers, but other farms did not provide any transport and the ill workers needed to use public transport. Less commonly, some workers procured their own antimalarial drugs (mostly Coartem®). One worker explained his routines after fever as follows:
’…I have Coartem in my bag. When I feel the symptoms, I take it. I buy it from a local pharmacy or borrow from a friend. I mostly take the first four tablets, if symptoms improve, I stop. If I don’t improve, I take the next dose. If there is no change after this, I go to the nearby health facility. Most of the time, the first dose improves the symptoms. However, I usually fail to complete the whole treatment and the symptoms return after a week.
Some farms stock antimalarials (mostly Coartem) which they sell to workers without using a diagnostic test (price: 50–70 ETB/$1.8-$2.6 USD). Workers sharing antimalarials among themselves (or keeping some tablets for future illness) was a common practice. It was rare to find a worker using a traditional medicine for malaria and fever treatment. However, one worker did report chewing a bitter plant (locally called ‘’neem’’) when he could not afford to buy Coartem. He preferred Coartem to “neem” [referring to leaves of Azadirachta indica], but did say that ‘’neem’’ was helpful.
Drug stock-outs were commonly reported in the local health posts and health centers near the study area. As previously mentioned, some of these facilities request a local identification card for care (though respondents indicated that this practice may have decreased during the time of data collection). Some workers reported dissatisfaction (including poor reception, lack of appropriate medical examination and long waiting time) with the service provision at the public health facilities resulting in them visiting private clinics which cost more money. Some workers reported that they do not seek care when suffering from malaria symptoms due to these reasons. One worker talked about his encounter in a health center as follows:
’…I am sick today. I have been to the health center. I was diagnosed with malaria. I was given anti-pain for the headache. They told me they run out of the right antimalarial and informed me to buy from a private pharmacy. At the moment, I couldn’t afford to buy from a private pharmacy. And none of my colleagues have a tablet at hand. I am left with no options.
Health workers said that they do occasionally conduct fever treatment campaigns at farms which employ migrant workers. Health workers also admitted that they do not consider the influx and increased number of migrant workers during the planning of malaria control efforts and the calculation of commodities needed. One farm was reported to cover the medical costs that were accrued during a worker’s malaria treatment, but this was not common.