This is the first of two posts where I deal with the issue of turnover in humanitarian organizations. I introduce the problem and some of its possible drivers here. Here you can find the second post, focused on solutions.
Humanitarian workers don’t like long commitments. Obviously, I am talking about work. Chances are that a work experience as project manager lasts no longer than 12 months (and often shorter than that). For senior staff the situation is slightly better (and for local staff also different dynamics apply), but still: it is uncommon for aid workers to stick long around.
Part of this is structural – at the end of the day, emergency response is supposed to be a temporary endeavour (although in fact the situation is very different for complex emergencies). At the same time, the nature of the work does require some staff rotation, in order for staff to avoid burnout and for organizations to maintain some dynamicity. Still, there is some wide consensus among practitioner that the current rate of turnover experienced in the sector makes for a very inefficient system.
Although there is not an extensive literature about the issue, some papers and dissertations have dealt with the problem. In this and in a following post I will start from the most relevant papers I found, to try to find an answer to the following questions:
- Why are turnover and failure to retain employees problematic?
- What is the scale of turnover within humanitarian organizations?
- Why this happens?
- Which measures have been suggested or are in place to counter the issue? Do they work?
A point that I will stress is the general lack of quantitative analysis and of reliable data on the issue, as most of the studies are qualitative in nature and rely on impressions and surveys of practitioners. Furthermore, the frameworks used in these surveys differ between a study and the other, making cross-study comparisons quite difficult.
Preamble – what are we talking about?
Turnover is generally quantified as:
Total number of leavers over period / average total number employed over period * 100
This figure includes all leavers but doesn’t usually take into account those leaving at the end of a fixed-term contract, which makes it of limited use to understand specifically the problem in a sector where fixed-term contracts are very widespread. As noted in a study by David Loquercio (p. 4), it makes more sense to distinguish between involuntary turnover (action taken by the employer), and voluntary (action taken by the employee), and in this between dysfunctional turnover (exit of effective workers) as opposed to functional (exit of substandard workers). Further, dysfunctional turnover can be either unavoidable, when it is due to force majeure reasons like serious illness, childbirth, etc. and avoidable (a totally voluntary decision of the worker). The key focus for Loquercio is hence on avoidable turnover, an understanding shared also by other authors– even though the reality and focus of the analysis are often more blurred.
Given the confusion that can stem from the use of the term “turnover” without further specification, I prefer to talk about “failure to retain” rather than “(avoidable) turnover”, implying that in this expression there is an interest from the organization to keep the employee and/or renew his/her contract. This also helps to avoid some of the confusion that can stem from the fact that humanitarian agencies, despite their use of temporary contracts, are quite often interested in retaining an employee, offering however another temporary contract. I will go back to the issue, as this seems to me a central aspect that appears to be quite often ignored by the studies.
Further, in these posts I deal only with HR issues pertaining expat staff. Failure to retain and turnover generally are issues involving also local staff, but the dynamics that apply to them are somehow different, and the literature about the issue is even less developed.
Why is this a problem?
As said, turnover per se is not problematic, and there is widespread consensus that a certain level of staff turnover is healthy and necessary. Failure to retain an employee (when an organization wish to do so), however, is in fact problematic, and while it is hard to quantify the present turnover rate within humanitarian organization (more on this later), there is wide consensus on the fact that it is not functional.
According to the study of Loquercio et al. published by ODI’s Humanitarian Humanitarian Practice Network in 2006, dysfunctional turnover present direct costs, in terms of separation costs (debriefings and administrative procedures), recruitment of new staff, and their induction. According to the estimates available in 2006, recruiting a new program manager (external to the organization) can cost between 5-6.000 GBP (estimate from Save the Children UK) and 15.000 GBP (estimate for a ICRC delegate). In terms of indirect costs, high turnover has a negative impact on the quality of a program, in terms of loss of institutional memory, additional work and stress for the remaining staff and consequential drop in performance. Continuity in program management is jeopardized, and most likely the new manager will spend a significant amount of time before finding his or her footing in a new project (in my personal experience, lack of a proper handover has often been a major issue).
This understanding of the issue is basically shared (often verbatim) by all the studies I came across (see e.g. the paper by De Calan (p. 4), and the dissertations of Sciberas (p.9) and Tirfe (p. 8)).
An indirect cost worth mentioning is the amount of strain that this can create on HR departments, working around the clock trying to fill gaps and creating hence a sort of emergency mentality, with few time left to identify and implement solutions to the root problem or to other issues related to staff welfare and performance.
An aspect that I didn’t find mentioned in terms of negative impact on the quality, but I believe is also relevant (and should be further investigated), concerns the lack of accountability and sense of ownership towards the endeavour that can stem from a temporary commitment to a project or a program. In other words, it is unlikely that a manager who knows that s/he will leave a project before its end will put the same amount of effort in the endeavour as one who will cover it from beginning to end (and even more if s/he has been the one designing it), as s/he will probably not be held accountable for it. A basic rule would be that the same manager opens and close a project (ideally, design and implement the project) – this often happens not to be the case. Of course, this problem applies mostly in the context of longer-term emergencies.
What is the scale of turnover in humanitarian organizations?
That is a good question. Although there is a widespread consensus that the present rate of turnover in humanitarian agencies is excessive, this has only seldom been quantified, and only a few organizations are tracking it.
Moreover, as said above, the usual rate of turnover doesn’t take into account the fixed-term contracts, which are the norm in humanitarian endeavours. An MBA dissertation (by Synzi Dadié, p.12) attempts a comparison between the median turnover rate in the not-for-profit sector (hence not only humanitarian aid sector) and other sectors based on data collected by CIPD in the UK for 2012, but such an exercise is of limited use, given the constraints mentioned above and the louse definition of not-for-profit.
Loquercio (pp. 4-5) warns about the risks of comparisons based on the sheer turnover rate, suggesting instead to use other indicators such as average duration of field missions, vacancy rate and seniority of staff. According to his research, the average duration of a mission for an expatriate ranges from 5,2 months (MSF-France) to 10,1 months (ICRC).
The most comprehensive review of estimates is contained in Tirfe’s dissertation (p. 22), to which I would refer for a more detailed perspective. Although the data available is not systematic, the picture that emerge from the few numbers available and the impressions of practitioners is pretty clear in designing a dysfunctional system, although further research is needed and other indicators should be devised. Alongside those suggested by Loquercio, and keeping in mind an understanding of the problem in terms of failure to retain an employee when the agency wish to do so (or to recall him or her when the agency need him/her), possible indicators could be the percentage of workers who do not renew a contract that when this is offered or that the agency would have offered, and percentage of workers who accept a second mission after a few months (as I will note in the next paragraph, there is already a good example of a study based on this indicator).
Why this happens?
This is the question that most of the studies published so far have dealt with.
Generally speaking, both push (dissatisfaction with present conditions) and pull (attractiveness of alternatives) factors play a role in the decision of quitting an organization, with push factors usually considered more significant than pull factors.
These dynamics, according to Loquercio (pp. 8-9) and Loquercio et al. (pp. 6-9), are also at play in humanitarian agencies, with push factors like divergence of values between employee and employer, lack of support, dissatisfaction with management. These factors can be considered alongside others specific to the sector, like (lack of) quality of life, challenges in having a family while on expatriate duty, lack of career opportunities. An important aspect on which I will come back lies in the short-term funding cycles that push agencies into offering short-term contracts (although doesn’t necessarily force them to do so) (Loquercio et al., p.7), this problem seems particularly relevant for small organizations (whose lack of resources affect retention on several levels).
Most of the qualitative analysis available also agree that the main factors leading employees not to renew is a mix of dissatisfaction with the management, lack of opportunities with the present organization and better opportunities with others, and reasons linked to quality of life and family prospects.
Dubey et al. identify 24 variables potentially impacting turnover, grouped in three areas: external, personal, and work-related factors), with a framework that is in fact quite different from Loquercio’s. From an exploratory survey delivered to aid managers in India (including development organizations) they conclude that turnover is mainly a function of internal factors like employment perception and job satisfaction, as well as readiness of the individual for humanitarian work (as the data and the tool used for the study are not available, it is quite difficult to further discuss these results).
In an interesting analysis of the rate of aid workers who accept a second mission with MSF Holland (so-called “re-enlisted”), Korff et al. identify family status, nationality and job prospects as the key predictors for the chance of re-enlistment with the organization, while other factors like security of the mission, age and gender don’t seem to have strong correlation with re-enlistment. These conclusions are hardly universal, given the peculiarity of the sample – MSF, as recognized by the authors themselves, is quite a unique organization, for identity and business model. Moreover, some possible predictors like (perceived) quality of management are not taken into account (and some others, like security, could have been gauged in a more precise way). Furthermore, focusing on only one organization, the study doesn’t take into account those workers that goes back to humanitarian work (or have already worked) with another organization. However, as one of the few (if not the only) study available at the moment based on quantitative analysis rather than surveys and qualitative analysis, it presents a promising way forward for further investigation.
A factor that seems not to be really taken into account (apart from Tirfe’s dissertation (p.27) based on the impressions of some aid workers, and as mentioned in Loquerico et al. ), is the role played by fixed term contracts in shaping the decisions of humanitarian aid worker. In other words, has not been clearly investigated yet the role that a final term to a contract, as opposed to an open end (or a short term vs a longer term contract) can play in the decision of an aid worker not to renew.
The possible reason for this is that short term contracts are considered the norm, and a peculiar aspect of the way humantiarian work is managed. However, if we consider failure to retain as the main problem, we can see how short/fixed contract are worth of investigation as a major driver – no matter the rationale behind their use. I will come back to this point in the next post, where I deal with possible solutions to the problem.
Sciberras Dimitri, Why is staff turnover an issue for Humanitarian Organisations and what are the strategies to this matter?, Bachelor Project submitted for the Bachelor of Science HES in Business Administration, Haute école de gestion de Genève, 2015