Step 2: Data collection
In March 2014, we collected qualitative data (step 2 of the framework) in semi-structured interviews with individuals as well as groups of respondents, according to feasibility and availability. To avoid being misguided by the initial hypotheses, the interviews were not explicitly structured around the hypotheses, but rather around the push and pull factors. We asked openly for push and pull factors and covered the push and pull factors known from literature. For each of the factors mentioned or confirmed by the respondent(s), we asked for an assessment of relevance and for explanations of how they worked to elicit the underlying mechanisms.
The employees were interviewed in three focus groups, two of which were single sex groups. In addition, we interviewed four managers and four staff members individually. Table
2 describes the respondents. Respondents were purposefully selected to form a varied, relevant and representative sample of the local health system, and we stopped selecting additional respondents when we reached saturation. Some respondents worked in Semera, the capital of the Afar Region; others in remoter areas. For safety reasons and travel constraints, the interviews did not cover all districts of the Afar Region. All respondents were from Administrative Zones 1, 2 and 4. The most Northern part of Zones 2 and Zones 3 and 5 was excluded. Interviews were conducted mostly in English. When required, the local researchers involved provided translation. All interviews were recorded with consent of the participants and transcribed.
Table 2
Overview of key informants and participants in focus group discussions
FGD 1
| 4 x M | 29.25 (avg) | E | Nurses |
FGD 2
| 6 x F | 26.8 (avg) | E | 5 nurses, 1 medical doctor |
FGD 3
| 4 x M 2 x F | 24.5 (avg) | E | 3 nurses, 1 medical doctor, 2 laboratory technicians |
Key informant 1
| M | 40 | E | District’s health office employee for prevention |
Key informant 2
| M | 29 | M | Head of a governmental health bureau |
Key informant 3
| M | 29 | E | Employee at the health bureau |
Key informant 4
| M | 42 | E | Employee at the Semera and Dubti local health offices |
Key informant 5
| M | 27 | E | Laboratory technician |
Key informant 6
| M | 38 | M | Project coordinator at an NGO |
Key informant 7
| M | 41 | M | In-share owner of a private healthcare clinic |
Key informant 8
| M | 32 | M | In-share owner of a private healthcare clinic |
Table
2 provides an account of focus group discussions and individual interviews, including the gender, age, role and occupation of the respondents. In addition, the type of data collection method (individual interview or focus group) is also indicated.
For reasons of transparency and verification, we now present an analysis of the collected qualitative data, to assess whether the proposed and other factors influenced turnover intentions in the Afar Region. To this purpose, we coded the transcribed interviews according to the factors listed above and labelled the newly proposed ones. We classified factors as dominant if they were mentioned by at least half of the respondents.
Turnover and turnover intentions
Respondents unequivocally confirmed that healthcare employee turnover and turnover intentions are common among healthcare professionals in the Afar Region. As turnover results from turnover intentions, we present a quote from the second focus group discussion (FGD 2) illustrating the omnipresence of turnover intention: ‘We are planning to leave, all of us. We have the intention to leave. Most of us don’t want to stay for more than two years.’ Indeed, most employee respondents admitted to thinking about leaving, and all respondents knew at least one direct colleague who had left.
While there is no official data on turnover rates, respondents agree that the rates are high. An employee at a district health office estimated the district turnover rate to be at least 50% in the last year. An employee at the local health offices in Samara mentioned that in the past year, seven co-workers had left. (The health office employed 10 health professionals at the time of study.) Another employee at the local health bureau said ‘I don’t know the exact incidence of turnover but it is a huge amount. Of the people that I know who left, one person left already after a month. Others left after a year. Even at the health bureau they are saying that there is a huge difficulty.’
The turnover mostly results in migration to other Ethiopian regions, especially to bigger urban areas such as the capital Addis Ababa. For example, a laboratory technician in Logia said ‘If someone can get a job in Addis or another city, he will go there to work. 90% of the people who leave go to another region.’ Switching jobs to the private sector was not uncommon, but clearly less frequent. Migration to other countries appeared to be rare.
Push factors: salary
Seven of the eight individual respondents and two of the three FGDs (focus group discussions) named salary as a main reason for employee turnover, and it was often the first reason mentioned. The head of the provincial health bureau explained that ‘The primary reason to leave in this area is salary. Employees leave public healthcare for private healthcare. For example, they go to work at the sugar factory, which has its own clinic. The difference in pay is very big. 5000 birr at the sugar factory and only 1400 birr in public healthcare.’ He added ‘The health office has payment issues. Both experts and managers are on the payroll, so most employees complain about this issue.’ Employees not only leave because of the low salary but also because they sometimes do not receive it.
Salary differences exist not only between the private and public sector but also between regions in Ethiopia. The Afar Region paid a top-up to the national salary, a policy now copied by many other regions, thus diminishing the reducing effect it may have had on turnover (intentions). In addition to the regular salary, supplements and other opportunities to complement it are important, for instance in the form of moonlighting as mentioned in FGD2: ‘In addition [...] the salary is a reason to leave, but also there is no access to private healthcare institutions to work overtime.’ (Earlier evidence also suggests that retention is high for employees who ‘moonlight’ at private organisations [
30].) If an area lacks this opportunity, employees might want to move to one where they can indeed work overtime at a private health organisation. This phenomenon is reinforced by the frequent and partly unpaid overtime demanded from employees in the public system: ‘The health professionals that still work here have to work sometimes 18 hours per day when they are on night duty. And they won’t get overtime payment. They get paid for the night work but not for six hours.’ The two respondents owning private clinics decided to open their own businesses because the salaries in the public healthcare organisations were too low.
Respondents confirmed the five dimensions of working conditions suggested in the literature: facilities, location of the workplace, management support, personal safety and health, and perceived workload. While the latter two appear to be less dominant, for ease of presentation, we consider them as five separate factors in Table
2 and in the text below.
Push factors: facilities
Employees at public healthcare organisations clearly worded their dissatisfaction with the physical conditions. One respondent in FGD stated ‘We need to have some standards. If the standards are not met, I want to move to another organization [...]It is likely that people want to leave for NGOs because of the assumptions that that kind of organisations are a little bit better in everything, accommodation, everything. And from government to private institutions, that’s also common.’
One of the participants of the first focus group, a medical doctor, talked fervently about her daily struggles due to the lack of facilities and materials: ‘Here in this health centre we don’t have enough equipment to help the patients. Because of that, we don’t have the moral or the interest to treat the patients. You learn how to treat patients in school but here it is impossible. You know their problem but you cannot help them. It is really boring and very frustrating. I might cry sometimes. You see patients suffering. [Respondent cries] Sorry. You know the patient’s problem, you don’t have the necessary medication. You don’t have the necessary equipment. Even if you send them to a hospital - Dubti is the nearest hospital - you know they can’t get the necessary treatment there. So you just stay there and stare at them. We have a health centre here, but it’s just a name and a building. It is not interesting to work here, I’m sorry. Sometimes I hate the time that I am here. If I was somewhere else, I could have helped someone with my education and my knowledge. Here, I am just wasting it. It’s not just about the money. This is really not good.’
Another participant confirmed that working in a workplace with poor physical conditions can be mentally draining: ‘Once I had a patient here who was vomiting blood, so her illness was very severe. The way to help her was to refer her to a hospital. But there is no ambulance here to transport patients, so this patient died. These situations also want to make us quit our jobs.’ In addition, respondents mentioned that the poor facilities sometimes prevented them from practising their knowledge and skills: ‘The main issue that turnover is very high is that we are lacking facilities. We don’t have equipment and facilities to get experience. So, we lose the knowledge that we had before.’
The inability to deliver health services may ultimately solicit aggression from patients and their families, thus endangering employee safety and health. Employee safety and health may be further endangered because of facility layout, capacity and use. This can take the form of having to work unsafely with chemicals, blood, or other laboratory services, or lack of hygiene in patient rooms. Moreover, facilities for nurses on night duty are perceived to be inadequate.
Push factors: location of the workplace
A focus group respondent compared location of the workplace to salary: ‘Salary is not the reason to migrate from rural to urban areas. It is the infrastructure. In rural areas there is no light, water. There is no access to some facilities, like shopping. That is the reason. It is not the salary; the salary is not that different.’ Workplace location also affects the lives of employees’ families. Respondents mentioned the lack of good schools, and affordability of and access to health services. ‘Even if we are working in a Woreda health office our organization didn’t offer us a free health service.’ Moreover, some respondents mention that life is expensive in the towns of Samara and Loggia.
Respondents consider the Afar Region to be very quiet: ‘This place is very boring. The whole area is boring. There is nothing to do here. Every day you go to work, you go home, you sleep. So everyone will get bored sometime.’ Some respondents explicitly mentioned the lack of recreational facilities. Added to the harsh and extremely hot desert climate of the region, it causes the regional living conditions to be viewed as unattractive. Many respondents mentioned the difficulty of the climate: ‘Only 15% of the people working here are born in the Afar Region. The people who are not born here have trouble with the environment, especially the people from the highlands.’
The culture of the Afar Region was also mentioned in relation to location as a cause of turnover (intention). First, because of the language barriers with patients and colleagues who speak the local language, which the majority of employees does not understand. In addition to the harsh climate, religious and cultural differences cause many employees to develop a limited commitment with the local community, and to be open to returning to their area of origin or moving on.
Push factors: management support
When respondents spoke about management support, it was often with dissatisfaction. Some respondents described the lack of management support in general terms ‘The management support is not that much satisfactory. The regional health office sometimes didn’t recognize our problem and especially leaders don’t consider our problems.’ Other respondents gave detailed accounts of dissatisfying practices, as already quoted under learning and career opportunities. Many respondents perceived a lack of interest and sometimes a lack of respect. ‘They had much attention for financial aspects, but not for the work……There is no motivation, recognition.’ Some respondents mentioned to have quit their jobs because of the lack of interest in operational problems. They said that management did not want to enter the laboratory because of the smell, when asked to attend the problems they had due to the lack of supplies.
Other respondents reported a lack of interest or intimacy as a main factor, even up to the point that it might outweigh salary: ‘It is not only about higher salary. Their management system should be good. It does not matter if the salary is high or not. I would not leave my institution. The management system is more important than salary.’ Another respondent added ‘For me, if the management is okay, I cannot leave. If not, then [laughs]. [...] If the management system has problems, then it’s not okay. That would be the main reason for me to leave.’ Managers were reported to have hit employees. Treatment differences between local and non-local employees were perceived as unfair.
Push factors: personal safety and health
While the Afar Region has a history of war and conflict, and presently is still not considered to be fully safe, respondents did not mention these aspects in relation to turnover. Violence did not play a role other than already mentioned above. The same holds for personal health.
Push factors: perceived workload
While not mentioned that frequently, some respondents mentioned that with every leaving employee the workload increases for the remaining ones, possibly causing other staff to leave their organisation or region in turn (see also [
4]). Moreover, the workload was perceived as high:
‘The health professionals that still work here have to work sometimes 18 hours per day when they are on night duty.’
We have already discussed how perceived workload may become even higher due to the emotionally demanding conditions and experiences. This appears to be further reinforced by the employees experiencing lack of autonomy or control. In fact, some respondents explicitly mentioned lack of autonomy and control to have caused them to quit their jobs in the past.
Pull factors
Respondents recognised the proposed pull factors of learning opportunities, career development and aggressive recruitment. Perceptions of better learning opportunities elsewhere have led to turnover to other jobs in the region or within Ethiopia, but not to migration to other countries. With the recent founding of Samara University, which includes a School of Health Sciences, the attractiveness of learning opportunities at alternative employment organisations was expected to lose some importance. Management, however, was perceived to favour local professionals over non-local professionals, who formed by far the majority of the work force. According to one respondent ‘The management was bad because they only wanted to have benefits for themselves and maybe for the native Afar people. For training and education, they wanted to give priority to themselves.’
Likewise, management has been reported to let personal relationships outweigh qualifications when filling positions: ‘Some managers give priority to people they know in assigning jobs. If someone they don’t know applies for the job, they don’t give it to him.’ Added to the early stage of development of public health organisations in Afar, employees perceived these practices as limiting their career development opportunities. Respondents explicitly mentioned that limited educational and career opportunities negatively affected their motivation and job satisfaction.
The migration flow appears not to be affected by active or aggressive recruitment strategies (see also below under salary) neither from within the country nor abroad. However, active recruitment by international organisations coming to the region has occurred on a modest scale and has been effective.
Other factors
When asked directly, the respondents did not suggest any other factors. However, the responses above suggest various alternative factors, which we have covered while discussing the factors on the interview topic list. One such factor is culture, which is covered under location. While culture is certainly closely related to geography (see for instance [
45]), we propose that it is not a locational geographical factor, but a distinctly independent sociological factor. To illustrate this point, we observe that the harsh desert climate is a push factor which clearly relates to geography, while the problems of living in a ‘boring place’ where locals speak another language, are cultural.
Likewise, the data suggest that the push-factor management support represents two factors. On the one hand, there is the operational management responsibility to manage scarce resources, e.g. salary payment, availability of equipment, occupational hazards. On the other hand, there is human resource management. Respondents asked for recognition, motivation, respect, autonomy, promotion on the basis of competence and so forth. The domains are interrelated, as priority setting in human resources management impacts health service operations and vice versa. Respondents indicated, however, that poor operational conditions can be excused as a push factor, when the management is ‘okay’. ‘Good’ management or ‘okay’ management appears to be associated with recognition, respect, etc., and with alignment of values and norms.
Step 4: Assessment and interpretation of the analysis
The iterative process to generate and identify mechanisms that are supported by the collected data and relate to existing scientific evidence and theory resulted in four mechanisms to explain the perceived turnover intentions among healthcare professionals in the Afar Region.
Lack of social and personal opportunities valued by healthcare employees (P-RE-M)
Clearly, our data imply a misfit between person and regional environment (P-RE-M, as inventoried in Table
1). Respondents indicate that the remote Afar Region offers poor perspectives for personal development, social life, and quality of family life. The climate is tough, the region is struggling to provide a reliable supply of water and electricity, there are very few leisure opportunities, the quality of schools is considered poor, and both local language and culture are different from the languages and cultures of the majority of employees. As the region is also considered expensive, the lack of moonlighting opportunities results in a misfit between the financial opportunities the healthcare professionals seek and the financial opportunities they find in the Afar Region.
At least until the opening of Samara University and School of Nursing, the perceived lack of educational development opportunities presented a misfit between personal development needs and the regional educational offering. The same holds true for career opportunities, which are perceived to be much better elsewhere, specifically in larger Ethiopian cities. While the two-factor theory classifies these factors as pull factors, the experienced lack of fit by respondents would cause them to be considered a push factors in the person-environment fit theory.
Dissonance between management logic and professional logic (Dis-Log)
Cultural differences may also drive the perceived lack of fit with management. The data indicate, however, that this lack of fit has other origins as well. Various professionals report that their professional logic—which emphasises quality of care, recognition of professional identity and autonomy—is hard to reconcile with the management logic. We propose the term dissonance to express that the difference in logic is not simply about rational priority setting. From the employee perspective, it seems rooted in management values and actions which conflict with their professional values and identity. While the literature suggests that various logics can co-exist in health service organisations, our data reveal that employees feel that the lack of recognition, respect and autonomy causes standing practices to result in dissonance with their professional identity.
The most vivid and direct illustrations of this turnover mechanism are the accounts of the use of violence by managers, and the unwillingness of management to enter the employee workspace because of bad smells. In addition, employees feel that professional competence is not considered in the case of promotion, hiring or selecting staff for education. Professionals also perceive that management does not share their priority for improving the quality of care for the patients; management is perceived to prioritise financial issues instead, often favouring locals over the majority of non-local employees. Employees feel unsupported in their continuous hard struggle to provide good care under high, draining workloads.
Standards of service operations are hard to accept (SOP)
The resource limitations, both in terms of facilities (rooms, ambulances) and medical equipment, cause employees to be disappointed about the services they can provide, lead to occupational hazards and cause them to lose professional competences because of lack of practice. Moreover, it is emotionally difficult not to be able to give patients the required services and have to watch patients suffer the (fatal) consequences. Altogether, these operating conditions can be considered so poor that despite the professional values to serve patients, they push employees to find alternative employers, where the operating conditions are acceptable for their professional values. There is also a misfit with personal values. The typically young employees strive to practice what they have learned and develop themselves, rather than lose their skills due to poor operating conditions.
Lack of financial improvement opportunities (Fin)
While public sector salaries in the Afar Region are not lower than elsewhere in Ethiopia, they may be less favourable in practice because of payment difficulties. Provincial budgets and treasury apparently cause problems for management to pay due salaries consistently on time. At the same time, employees indicate finding the Afar Region expensive and already have problems with the present salary. They are therefore open to opportunities to improve their financial conditions. One alternative is moonlighting; paid work in the private sector after office hours. There are, however, few private facilities in the region where moonlighting is possible, and their present jobs already demand long overtime hours. Alternatively, there are some opportunities to move to NGOs, start their own private facilities, or work for multinationals entering the region. In many cases, however, financial mechanisms will cause employees to move to other regions where the cost of living is lower and/or moonlighting opportunities are better. Thus, while this financial mechanism mostly pulls employees to other regions, some are pulled to non-public organisations inside the Afar Region.