In the current discussions about our duties (of justice) towards refugees, feasibility constraints are often invoked to justify the limits of what can be done: Austria has to close its borders and limit the number of daily asylum applications due to feasibility constraints, the feasibility limit of refugees admitted to the European Union as part of the current resettlement scheme with Turkey is set at 172.000 or mayors of various cities claim that it is not possible for them to shelter refugees.
Following these debates, I often gain the impression that infeasibility claims are invoked far too easily. Hardly ever they are based on feasibility studies and often they seem more of an excuse to shy away from our duties (of justice). At the same time, powerful arguments for the need to account for feasibility constraints when identifying our duties of justice in the non-ideal real world have been put forward in the literature on non-ideal theories of justice: what justice demands from us should not go beyond what we can possibly do. How is it possible to account for feasibility considerations in this later sense, without allowing them to become a cheap excuse in political debate to shy away from our duties of justice towards refugees?
In this blog post I would like to make a first attempt to address this question by drawing on the current migration debate. I do so by shortly discussing a number of features that seem to make feasibility arguments particularly vulnerable to misuse in political debate. In a second step I suggest some principles for public debate that could restrict a possible misuse of feasibility arguments in debates about justice.
The following features seem to make feasibility arguments particularly vulnerable to misuse in public debate:
- Feasibility is (often) uncertain: Especially in new situations the outcome of (policy) actions is often highly uncertain. There is limited experience in handling migration movements of the current scale, for instance, making it difficult to predict what is possible and what is not. This uncertainty feature of feasibility arguments can undermine claims of justice if risk-averse feasibility claims are advanced without proper justification.
- Feasibility is dependent on the time horizon: What is feasible in the short-run might be crucially different from what can be feasible in the mid- or in the long-run. For instance, it might be the case that current Austrian institutions can only handle a certain limit of daily asylum applications and that this limit cannot be changed from one day to the other. This does not mean, however, that this feasibility limit cannot be extended in the mid-run if one undertakes the necessary institutional changes. This feature of feasibility arguments can undermine claims of justice if only a short-term perspective is taken in the assessment of feasibility, without exploring the possibilities of altering the constraining factors themselves.
- Feasibility, ceteris paribus: Feasibility claims are often advanced taking all other things being equal. One example is the feasibility limit of the recent EU-Turkey agreement, that takes the existing European resettlement commitments as given.This feature can undermine claims of justice if feasibility arguments are employed to limit claims of justice taking the (institutional) settings as given, even if the source of injustice might lie precisely in these very institutional arrangements themselves.
The following principles for public debate are some first ideas how the above described risks of feasibility arguments undermining our duties of justice (towards refugees) could be countered.
- Not-too-cautionary principle: This principle consists in shifting the burden of proof towards those who claim that something is not feasible. In the current debate this would mean, for instance, that those who claim that Austria cannot let more people enter have to prove why this is not possible.
- Mid- & long-run feasibility principle: This principle demands short-run feasibility arguments only to be admissible if they come with a mid- and a long-run perspective. In case of the migration debate this would mean that immediate restrictions on the number of people allowed into the European Union could only be justified in the name of feasibility if they come with a mid- and long-run strategy as to how the limits of feasibility could be extended.
- Ceteris non-paribus principle: This principle demands a change in what should be taken as given in the assessment of feasibility. Instead of asking what is possible, taking other things (such a institutions or budgetary priorities) as given, we ask what would have to be changed in order to extend the limits of feasibility. In the current debate about migration, this would mean that instead of asking how many refugees can be admitted to a specific country or the European Union given current institutional structures (e.g. the Dublin III agreement), for instance, we would ask what needs to be changed in the current legal and institutional setting in order to extend the limits of feasibility as far as possible and necessary.
These principles are some first ideas as to how we could take feasibility constraints seriously in debates about justice, without allowing them to become a cheap excuse for not complying with our duties of justice.
Thanks for the great post, Constanze. I am sympathetic to pretty much everything you say here, so my comment is only to suggest another thread on feasibility considerations that might be of use. If I understand you correctly, the kind of feasibility constraints you mention here are ‘technical’ – they concern our logistical capacity. They can be separated from constraints of ‘political feasibility’ – concerning whether we can get policies past a set of MPs or an electorate.
I think you are right that there are often errors made in estimating technical feasibility, but I also think arguments sometimes make errors in estimating political feasibility and, indeed, in confusing the two (“we cannot increase our refugee intake because the public will never go for it”). And I think we should be very wary of claims about political feasibility, partly because, similar to technical feasibility, we ought to avoid common errors in estimating it, but also because claims about what policies MPs/electorates will accept speak so poorly to questions about what policies we ought to accept.
Thus, I might also be inclined to an additional principle along the lines of being specific about the kind of feasibility constraint being used and giving it appropriate weight/disregard in feasibility analysis.
Hi Andrew, many thanks for this very important point! I fully agree that a further principle about the kind of feasibility should be added. Whe reading your comment – I became aware that it is -among others- precisely the mixing up of different kind of feasibility constraints that confuses the current debate. However, I do not yet see clearly along which lines a fruitful distinction between different kind of feasibility constraints could be made and on which grounds one should identify their appropriate weight in a feasibility analysis, as you suggest. On which grounds would you assign more weight to say technical feasibility considerations than to political feasibility?
One way I see is to distinguish between different kinds of feasibility constraints dependent on how likely/difficult it is to overcome them. This is a route @Pablo Gilabert and Holly Lawford-Smith (2012) propose in a super inspiring article ( http://onlinelibrary.wiley.com/doi/10.1111/j.1467-9248.2011.00936.x/abstract ):They suggest a test for assessing (political) feasibility based on a distinction between hard (e.g. biological, technical (?)) and soft (e.g. institutional, legal or cultural) feasibility constraints: They then suggest a test to make comparative feasibility judgements based on soft constraints (particularly important for questions of political feasibility): It is more feasible for agent X to bring about outcome O than it is for agent Y to bring about outcome O if it is more probable, given soft constraints, for X to bring about O (given that (s)he tries) than it is for Y to bring about O (given that (s)he tries) (see p. 7).
This would allow to distinguish between different kind of feasibility constraints based on ones probability to overcome them if one tries. However, it would not allow to make distinctions on other grounds as far as I see (e.g. to give different weight to a claim as to what MPs/electorates will accept than to a claim about institutional feasibility if both constraints are equally well grounded and equally difficult to overcome (e.g. by combatting myths about asylum seekers as mentioned by David in his comment)). Do you think that -apart of how easy/difficult it is to overcome them- there are other reasons as to why different weights should be given to different kind of feasibility claims?
I don’t know if ‘feasibility’ is a fair term for what is going on here. Clearly if even poor small fragile countries like Jordan and Lebanon can shelter many hundreds of thousands of Syrian refugees then it is perfectly possible for Austria to accept at least as many. The word feasibility in itself seems to predetermine the outcome of your ethical reasoning – no analysis is really required.
‘Political feasibility’ may be more promising. What is going on here is a counting of the (opportunity) costs of admitting refugees and providing them with a minimally decent life (in Austrian terms). Clearly the more refugees are admitted the more services will be stretched (such as schooling, healthcare, and public housing) and the higher the direct (taxes) and indirect costs (e.g. health insurance) on current residents. The political debate is about what level of sacrifice by existing residents is appropriate. That debate turns on ethical values not logistics, whether one places a higher moral weight on the fact of ‘human need’ or the fact of ‘foreigner’. (As the percentage of government revenue spent on foreign aid suggests, most polities see the needs of foreigners as deserving much less from them.)
Hi Tom, many thanks for the important points you raise! I take them up one by one:
ad 1) It is not appropriate to use the term feasibility in case of Austria, given that the examples of Joradn or Lebanon clearly show that it would be perfectly possible for Austria (or other Western countries) to shelter more refugees: I agree on this point! Indeed it is this very point that triggered my thinking about these issues, i.e. that Austria denied to shelter more refugees claiming that it is infeasible/not possible (which seems so clearly not to be the case given the example of Lebanon or Jordan you raise). The question of my inquiry is precisely how we can distinguish such “cheap excuses” from “real feasibility constraints”.
One way to do so would be precisely along the lines your example indicates, which is the same as the sufficiency condition for feasibility Christiaan Broekman suggests in his Master thesis: something is feasible if it has been done/achieved somewhere else in the world (or previously in history).
ad 2) Again I fully agree! In fact I did not mean to restrict my discussion in any way to technical feasibility considerations (but I now realise that I did not make this point sufficiently clear). Moreover, I think that the point about (opportunity) costs you raise is very important in any analysis of feasibility:
It seems to me that many questions – as to whether something is feasibile or not- are relative to the costs (broadly understood) one is ready to pay to make something possible. Many infeasibility claims in the current debate seem to presuppose that certain costs will occur (e.g. sheltering further refugees would come with a tax increase) and that certain trade-offs cannot or should not be made.
Indeed, I think your suggestion to look at the (opportunity) cost side of the feasibility coin would bring the debate a huge step forward: it would allow to put these (implicit) assumptions ( about possible trade-offs and the costs we are willing to pay, underlying many feasibility claims) out into the open and to scrutinise and discuss them. It would allow to shift the debate towards the question ‘which costs we are willing to pay’ in order to comply with our duties towards refugees. Indeed, this could be a first step towards a debate about the relative importance we ascribe as a society to values of justice or humanitarian values vis-a-vis other values or interests. It seems to me that precisely such a debate is much needed and far too often cut off by infeasibility claims.
See Matthew Gibney’s discussion in THE ETHICS AND POLITICS OF AYSLUM (CUP 2004) which addressed just these issues – and also pointed to the political responsibility on states/politicians to expand what is feasible, technically and politically (not least by combatting myths about asylum seekers and refugees).
Many thanks for this reference, I am very much looking forward to reading it!
Correction: Please note that there is a typo in the post. The current resettlement scheme with Turkey is set at 72.000 (not at 172.000). Many thanks to Nick Baigent for making me aware of it!
For another great discussion on the topic see @Anca’s (Gheaus 2013) article ‘The feasibility constraint on the concept of justice’: https://sheffield.academia.edu/AncaGheaus/Published-Papers
I fully agree with her argument that feasibility constraints should not play a role in determining the concept of justice. Along my understanding, in the current debate about migration this would mean that even if it were indeed not feasible for Austria (or any other country) to admit more refugees, the situation can nevertheless be termed unjust. However, what I do not yet see is whether Anca’s argument has implications for our duties (possibly to extend feasibility constraints) in the current debate about migration as well.