Working out refugee policy

Australian refugee policy has a lot of moving parts, and it's easy to get lost in them. If you listened to the Guardian politics podcast this week, for instance, you might have heard some complaining about the lack of logic in Malcolm Turnbull pledging to accept 5000 refugees from Costa Rica, while continuing to refuse resettlement to the roughly 1200 refugees on Nauru and Manus Island. On the flipside, it's become a regular line on the right that we can't be too generous towards refugees without encouraging more to get on boats and so causing more deaths at sea.

Both of these arguments are confused and misleading. The reality is that Australia is fairly unique among countries in our relationship to the refugee crisis and policies responding to it. For the last year and a half, Germany has operated a generous policy towards asylum seekers which involves letting people who arrive at its borders enter the country to claim asylum, treating them quite well while their claims are processed, and allowing a large number of them to settle permanently in the country. Because there are huge numbers of people arriving at Germany's borders, and because there's not much Germany can do about that (at least by itself), those steps are intimately linked. In Australia that's not the case.

The upshot is that there are three strands to Australia's refugee policy which are, to a reasonable degree, independent of each other. What you want to do about one of them need not determine what you want to do about the others. So it's important to understand the differences between the three issues. Once you get clear on them, you can see the way to criticisms of our refugee policy - and alternative approaches - which fit the logic of the situation.

Here are the three questions that we have to answer in formulating a refugee and asylum policy. I mainly want just to explain how they are and aren't related, and lay out some of the lines you could take. But I'm not going to pretend to have no views on what should be done. In fact I think that working through the logic of the policy question makes it clear that a lot of current Australian practice is even more indefensible than it might appear.

1. What should we do towards addressing the world's general refugee problem?
According to the UNHCR, there are just over 65 million forcibly displaced people in the world, of whom 16.1 million are refugees on the move internationally. At their peak in the first half of 2013, boat arrivals to Australia were running at around 26 000 per year. The point is not that this is a small number of people that we could easily have handled - it's just that the global crisis is vastly bigger than Australia's direct exposure to it. That's true of most countries, with the exception of those (like Turkey or Jordan) immediately bordering crisis states.

Australia has a humanitarian intake - we can decide how many recognised refugees we will accept for resettlement, and this is a policy choice that has nothing to do with boat arrivals. What's unusual is that - unlike Canada, for example, which also has a resettlement program but doesn't have any boat arrivals - we also have people arriving by boat, and need to make some policy choices about that. But they're not related. Flying in and resettling refugees from camps in other parts of the world would do nothing to either encourage or discourage people from trying to reach Australia by boat. (That's a bit hasty; I'll come back to it.)

So the choice is about how many of the world's millions of refugees we want to give the chance to settle in Australia. The kinds of arguments relevant to this choice are basically generic ones about immigration. You might argue that we should accept fewer refugees, because large numbers put economic strain on infrastructure and labour markets or because they threaten the country's cultural stability. You might argue we should accept more, because the harms to Australia are small or non-existent and the benefits to the refugees are enormous. As I already said, there's no room in this strand of the debate for the 'gotta be cruel to be kind' arguments that are now ubiquitous in Australia's refugee discourse. Accepting fewer refugees from a camp in Yemen or Kenya does nothing to help either those refugees or any others. Either you want to help more refugees, or you think the costs of doing so are too high.

I think that once you face up to the fact that the choice about our humanitarian intake is one we have to make on this basis - how much do we want to help, what will it cost us, how much are we willing to pay - it becomes pretty clear that we are doing way too little on this score. As of last week, Australia will accept 18 750 refugees per year from 2018 onwards. There is a constant back and forth about where this puts us on a world ranking of generosity. It's extremely confusing what the appropriate measure or point of comparison might be.

But that argument is pretty uninteresting. It seems almost certain that all countries are failing to pull their weight and Australia's ranking tells us nothing about this. The UNHCR's estimate of resettlement need in 2016 is 1.1 million, and the total resettlement places being offered are around 80 000. Matthew Gibney thinks (or thought a while ago) that at least a third of a developed country's total immigration intake should be composed of refugees, which for Australia would involve more than tripling the current target.

Anyway, the bottom line is that you have to think about (a) what the costs of resettling more refugees would be, (b) how much those people would be helped by being resettled, and (c) what moral duty Australia has to those people. Nothing else is relevant.

2. What should we do about refugees travelling directly to Australia?
Here, things start to get a bit more complicated. The policy we adopt towards people who travel to Australia by boat affects their welfare, but also the welfare of future refugees who might consider the same action. There are more elements to the calculus. And there are also more varied levers to pull than the single choice about the humanitarian quota: should we settle these refugees? how should we process them? how should we treat them while they're being processed?

First things first: how does this question connect to the previous one? Extremely loosely. By and large, you can take any position from 'resettle them all' to 'deport them all' about boat arrivals, irrespective of what you think about the appropriate size of the humanitarian intake. In 2012, the number of asylum seekers arriving by boat exceeded Australia's humanitarian quota. That's never happened before or since, but if it did - more likely if you think the quota should be much lower, or if boat arrivals spike for some reason - then you might need to rule out the 'resettle them all' option. Otherwise, you can combine these however you like. At the poles, for example, you might think that we shouldn't settle any boat arrivals but should have a very large humanitarian intake;  or that we should have no humanitarian intake other than settling all refugees who arrive by boat. (This would be an odd position for Australia, but it's roughly Germany's policy: you can get settlement if you come to Germany and apply, or if you're a family member of someone who's done that, but there's not really a more general resettlement program.)

What bears on the choice about policy on boat arrivals? You might want to reduce the number of asylum seekers trying to reach Australia by boat. That could be for Howard-esque reasons about sovereignty and border security, not wanting people who we haven't screened in advance to enter the country. (By the way, Australia is quite unusual in requiring a visa application, in advance, for any length of stay - including transits - so I guess this is at least consistent with our general border policy.) Or it could be because you don't want refugees to attempt the crossing, since so many drown while doing so.

Whatever the goal, there are three ways of discouraging boat crossings. Firstly, you can make them impossible or very difficult. There's a grab-bag of interventions in this vein which directly disrupt the availability of boats or the willingness of migrant smugglers to take people on the journey. The much-maligned policy of buying boats from smugglers or would-be smugglers is one such idea. You could also pay people smugglers not to make the crossing. (This has happened; the form it took in reality stopped people reaching Australia but did nothing at all to avert drownings. So it matters what your goal is. But you can imagine a version of this policy which makes payments at an earlier point and averts any dangerous boat travel happening.) Then there are more straightforward law enforcement operations to identify and arrest smugglers, which directly stops some boat travel and discourages other people from being willing to facilitate it.

Beyond that, you can make refugees not want to make the journey. (Think of this as reducing demand, rather than blocking supply.) One way of doing this is to make it much easier for people who might consider travelling by boat to come to Australia another way. Boat travel is not great at the best of times, and the operations run by people smugglers from Indonesia to Australia are not the best of times - even before you count the risk of drowning, of which refugees are obviously well aware. People won't do this if they don't need to. So you could open some centres in Indonesia at which refugees can apply to travel to Australia for humanitarian reasons, and then bring them by plane if and when their applications are approved. Clive Palmer once loosely proposed this, and otherwise I've mostly heard it discussed by university debaters; neither are well-known for their political realism, and this is mostly dismissed as utopian.

But you should give it another thought! Haiti has been degenerating into political and economic crisis for years; one consequence was that many Haitians were fleeing to Brazil. A large number did this by the 'jungle route' - travelling by plane to another South American country where they didn't need a visa, then overland and illegally into Brazil. This was extremely dangerous for the migrants, and Brazil didn't particularly want migrant smuggling networks infiltrating its borders. So it set up a special Humanitarian Visa Application Centre in the Haitian capital and made it extremely easy for Haitians to apply for such visas and then travel to Brazil by a regular route. The number using the irregular route dropped precipitously. This might not work for the Indonesia-Australia sea route, but this is an option that at least deserves to be thought about more.

Thirdly, you can reduce refugees' willingness to make the trip by giving those who arrive by boat worse treatment. There are various forms that could take: you could make them wait longer to receive settlement, or give them only temporary settlement status, or refuse to settle them at all. You could make them wait in detention while decisions are made about their settlement status, so that they can't enjoy Australian life in the meantime. You could put the detention centres offshore, so they can't even set foot on Australian territory. You could reduce the service provision in those centres.

All those measures come with their cost to the asylum seekers involved. That's the point, of course. But you need to think carefully about how those costs weigh against the benefits you're trying to realise. The starting point has to be that you inflict as low a cost as possible on refugees consistent with your policy goal. I'm not exactly sure how you quantify or weigh the benefits of border security. For the humanitarian motive, this means thinking about how some improvement in how we treat refugees would affect likely boat arrivals, and how likely boat arrivals would affect likely deaths at sea. It's also important to factor in here that refugees will generally have quite low-quality information about how they'll be treated if they make the boat journey. So some changes in treatment might not have any impact on the number of boat arrivals; and if there are ways in which we can treat boat arrivals better which don't mean more people try to come by boat in future, there is simply no excuse for not taking them.

There used to be a debate about whether Australian policy, in particular the deterrent policies of mandatory detention, offshore processing, and so on, had any real impact on the number of boat arrivals. That might kick up again at some point, but the starkness of the declines in 2002 and 2013-14 when harsher policies were introduced makes it seem pretty clear that the impact is sizeable. Still, there's argument about which policies are crucial, and this matters a lot. If it's the policy of turning back boats which discourages boat arrivals, and offshore processing doesn't contribute, then there is no justification for offshore processing. Even if offshore processing does make some further contribution, it might not be enough to justify the costs. Back of the envelope: about 900 refugees died between 2007 and 2013, out of 44 000 people making the journey. If you assume that proportion would stay fairly constant, then a deterrence policy which stops 100 people per year from making the journey - roughly two boats - would save two lives. (This is not a good assumption, because it's not like one person on each boat dies - the distribution comes in large clumps; but I'm not sure how to do better for back-of-the-envelope purposes.) There are plenty of costs that it's worth inflicting on refugees to save two lives per year, but it's not worth any cost. It might not, for instance, be worth keeping 1500 people indefinitely in offshore camps without proper medical care. (By the way, it's not a nice thought, but you might also think it's not worth the financial cost. The government spends hundreds of millions of dollars each year on offshore processing, for what seems to be an average of 150-200 lives saved. There are, to be understated, a lot of opportunities to spend millions of dollars saving non-citizens' lives which Australia passes up. Why insist on taking this one, which also involves being cruel to people in need of protection?)

To round out: absolutes are dangerous, but I don't think there are any other reasons bearing on this choice. What we do with (or to) refugees who arrive by boat does not determine the size of our refugee intake, let alone our total immigration intake. So there's just no way that concerns about immigration - economic, cultural, or any other - can give any reason for having harsher policy towards boat arrivals. The fact that we so often hear these concerns trumpeted as justification for harsh deterrence of boats is just evidence that the public debate is almost exclusively driven by xenophobia - as a stand-in for other countries' xenophobic debates about regular migration - rather than by anything to do with the actual issue.

3. What should we do about the refugees on Nauru and Manus Island?
This question is the most unique to Australia, and in some ways the most puzzling. That's because its relationship to the other issues - the second one, anyway - is inscrutable. Recap: there are about 1200 refugees on Nauru and Manus. I haven't been able to find good information confirming this, but as I understand it all or almost all of these people arrived under a previous policy regime - before the start of aggressive turnbacks and the total ban on settlement for boat arrivals.

In theory, that means Australia can treat these refugees as well or badly as we like, without affecting whether future refugees will come by boat. We have an established policy towards people who try to reach Australia by boat: their boats will be turned around and pushed back to Indonesia. If that's for some reason not possible, they'll be taken to Nauru until their asylum applications are processed and a country can be found to resettle them. This creates a very clear set of costs and incentives for people who are considering making the boat journey to think about. What we do with people who made the journey in 2013 and earlier, when this policy was not in place, under a different set of rules and with different incentives, shouldn't make any difference.

If that's the case, then there is no excuse for keeping people in these detention centres, especially in the horrific conditions they currently endure. Harsh treatment of refugees is only justified by its effect on reducing the number of boats; if how we treat the people on Nauru and Manus has no effect on future boat flows, then it's entirely unjustifiable to treat them badly at all, let alone as badly as we currently do.

Unfortunately it's not quite that simple. Australia's refugee policy has changed quite frequently, so when an asylum seeker decides whether to make the boat journey or not, they're unlikely to take the current policy regime as a fixed law of nature. If all the refugees currently in detention were resettled in Australia, that might convince you that Australia can no longer tolerate asylum detention and so wouldn't really put you in it, if your boat wasn't turned back. Or it might at least make you think that detention would only last a while before the government gave in to public pressure and resettled all those in detention again. Add into the mix that most asylum seekers in Indonesia are unlikely to have very good information about Australian policy, so if the detention centres are emptied it's easy to see how that could be understood as an end to the offshore processing regime. The argument here is broadly similar to ones in the US about amnesty or legalisation for illegal immigrants. This policy is generally proposed as a one-off: we will give legal status to (some) current illegal immigrants; but there will be measures to stop future illegal immigration and no legal status for people who migrate illegally after the amnesty date. This is meant to ensure that the policy creates no incentive for more illegal immigration, but if there's been one amnesty, maybe there'll be another one...

The big question here is how you think our treatment of refugees currently in detention will filter through into the decision-making of future people considering making the boat crossing from Indonesia. The less impact you think it has on that decision, the less justification there is for anything short of resettlement. A secondary question is whether different ways of modifying the way we treat these refugees have different effects. For example, suddenly resettling everyone from Nauru and Manus in Australia and closing all the processing centres might be a dramatic move that would encourage future refugees to board boats; but maybe quietly and more gradually resettling people to a range of other countries, without any clear announcement that you're aiming to resettle anyone, wouldn't. And that latter policy is obviously preferable to keeping people in detention indefinitely, so if there's no difference in the incentive effects, there's no justification for not doing it.

Can do better
That last line is really the refrain of this post. There are some really difficult areas in refugee policy, where decent people are faced with a dilemma about whether we can afford to offer more generous treatment without endangering the lives of other vulnerable people. But because the debate is so wide-ranging and contentious, these hard areas haven't been disambiguated from other areas, where it's either definitely or quite possibly the case that we can treat refugees much better without any downside at all. That we aren't taking these steps, or even looking into them, is a scandal.

Here's a more upbeat conclusion. I've written before about the sleight of hand that lets politicians be ambiguous between whether they want to stop the boats to save lives or to protect the border. The good side of that is that, whether they're motivated by border protection or humanitarianism, everybody has to agree that harsh treatment which doesn't contribute to the goal of stopping boats is pointless and indefensible. So if we can identify ways of being kinder and more humane to refugees which won't make boats more likely to arrive, nobody has much reason - moral or political - to resist that. None of them are going to look like the total overthrow of the current policy regime which is probably warranted. But it's at least a way of, step by step, starting to make things better.