Thursday, August 16, 2012

Could Australia settle all the local refugees?

My question is simple: could Australia feasibly settle all the current refugees in the UN's South East Asian area over a period of, say, the next five years?

(First up, I'm not an expert in the field, there's almost certainly things I'm missing, and my method for translating DIAC's current budget into a costing for an expanded intake is simplistic to put it mildly. This is an exercise in back of the napkin estimates, except the napkin is a Google spreadsheet.)

There's precedent for far greater generosity in times of refugee crisis. Australia's intake has been a lot bigger than it is now. In 1980 our population was 14.5 million, our refugee quota was 22,000. That's roughly the equivalent of 35000 per year today, compared with total population. If you chucked in an adjustment for the assumption that our higher real GDP (we're not far off twice as rich as in 1980) means we can afford to support more refugees compared to total population, the possible quota goes even higher.

So, what does a refugee cost to process and settle? Let's go straight to the Department of Immigration's own budget statement. The DIAC's budget is handily divided 6 outcomes which are the different functions of DIAC:

1. Managed migration
2. Reffos
3. Borders
4. Visa compliance, status resolution detention, etc
5. Settlement services
6. Multicultural feelgoodery. Also citizenship decisions.

In addition there's also the Migration and Refugee tribunal, part of DIAC's ambit but with a separate budget.

By my maths and assumptions about these items, it turns out the direct cost, per settled refugee is roughly $43000 a year. Here's a link to the spreadsheet with my figures.

Just an explanation on my assumptions:

In addition to the obvious entirity of Outcome 2, Outcome 5 is settlement services, so needed to be apportioned. Half the entire budget item is free English classes, so I took a a stab at splitting that between refugees and family visa recipients. I came up with 35% being refugees, assuming 80% of our 13750 refugee placements needed classes (found a reference which said 64% needed interpretors) and a complete guess at 50% of the 40000 or so on family visas from non-English speaking countries (DIAC figures, table 1-18) needing classes.

For other some big items around settlement grants and schemes, I assumed 100% refugees and then split the remaining budget 50:50 because why the fuck not. Details in the spreadsheet.

The biggest dilemma is how to account for current offshore detention and its replacement under the new policy. It was $700m in 2010-11, for about 5000 asylum seekers, which is firstly insane when you think about it, and secondly you could somewhat spuriously say cost $137,000 per detainee. But that wouldn't scale or translate to the situation of a radically expanded intake. The Red Cross estimates it costs them $11,000 for community based aslyum seekers. I bumped that up a bit and called it $15000, since government is inefficient or whatever.

However that fairly obviously skirts the question of how the fuck people are to get here. Are overseas processing and transport costs already included in Outcome 2? Quite possibly. But maybe not. Maybe the UN pays that stuff but would be unable to pay for the expanded intake places. That'd be the biggest caveat I put on this guess - it might not cover the costs to the Australian government at the point of origin.

SO! $43000 per refugee, apparently! Plus getting people to Australia. That obviously excludes a bunch of other stuff from the napkin. Within the departmental budgets, does each refugee make the per-refugee cost higher or lower? Are there economies of scale as refugee numbers expand or greater bureaucratic overhead? What the fuck does it cost ASIO to do security checks anyway?

What I DO know is that there's about yay many refugees locally:

To accept every one of these 478000 refugees would cost about $21 billion, with the earlier discussed assumptions. Do it over 5 years (so a bit under 100 000 a year) and we're looking at $4.2b a year in direct settlement expenses.

That's just the direct costs of course. What about the costs for building accommodation? (Could $40 000 shipping container houses be the new fibro shacks?) What is the cost of laying down infrastructure to basically build a new Hobart over a few years? What about the the flow on effects to welfare or education or health of bringing many refugees here? Would it be an idea to cut other areas of migration for a while to keep population growth down. Who knows! But here's a starting point!

On the other hand, this is an upper bound estimate of the direct costs if every single one of these refugees was to be settled in Australia. But an undertaking like this would almost certainly be intended to be regional, and surely by showing some leadership Australia could drag some other countries into taking some refugee settlers. Skive a quarter off on New Zealand for a start, and see if "we'll take 75% of your refugees if you settle the other 25%" works on the government of Thailand or Malaysia.

But the huge wildcard here is those massive numbers on the borders of Myanmar. The UN says there are 400 000 Myanmar refugees in other countries. If they're all in Southeast Asia that's 4 out of every 5 regional refugee being from Myanmar. Essentially all of the refugees in Bangladesh are Burmese, as are most in Thailand. As fucked as the place is, would they all qualify as fearing persecution? I don't really know. And of course there's no reason to assume Myanmar would stay fucked forever - many refugees ultimately return and if the situation changes in Myanmar we might see that in this case.

Furthermore, there's also the argument that Australia doesn't need to take all the refugees. I think I've satisfied myself that it's vaguely feasible (we're talking what, half an NBN here?), but the current obsession is with stopping the boats. And I don't think we'd need to take every refugee currently residing in Bangladesh to do that. We could and probably would focus just on the 240 000 in Malaysia, Indonesia and Thailand and that drops us to $10 billion. Over 5 years, starting with the longest-exiled refugees first, that 2 billion a year to save lives and stop locking children up indefinitely starts to look pretty reasonable. (Also that gets us down close to an intake proportionately comparable to the early 1980s.)

Could such a policy cause Australia to be "swamped" by the desperate hordes? I really actually don't think so. Remember that not everyone can be a refugee just because they live in a poor or unstable country. There's two key limitations - the first is the fear of persecution has to be proven. Right now, Burma's our biggest regional arsehole government, and surely just about everyone who's going to flee there has probably done so, especially with things maybe starting to open up slightly over there. And there's no wars of note elsewhere locally. Numbers coming from further afield to seek asylum directly are inherently limited by geographical distance.

Just about the only situation which would result in a literally overwhelming swamping by refugees is something seriously fucked happening in Indonesia. And if that happens all the mandatory detention and tough rhetoric in the world won't stop the human tide. There are a million Afghan refugees in Iran right now. Several hundred thousand Vietnamese and North Koreans in China. You think they're a soft touch and just let them in? Didn't have tough enough borders or strong enough deterrents? Yeah right.

So in summary, I think it would cost about $10bn in direct settlement costs to settle all the refugees currently in Indonesia, Thailand and Malaysia and $21bn to take every one from here to Bangladesh.

No comments:

Post a Comment