Day: September 26, 2017

Immigration: numbers and options

Michael Reddell Immigration, NZ politics

On the off chance that anyone thinking about negotiations with New Zealand First might also be considering immigration policy options, I thought it might be time for a refresher on the numbers (as well as yet another dig at MBIE for not making accessible data readily available on a timely basis).  Since much of the accessible data MBIE do release is for June years, for this post I’ll mostly use data for the year to June 2017.

Recall that the headline writers focus on net permanent and long-term migration, calculated from the declared intentions of those (New Zealanders and foreigners) crossing the border.  If you are leaving and expect to be away for at least 12 months, or are a non-resident arriving and expect to be here for at least 12 months, you are in the PLT statistics.   Plans do change, but the new 12/16 data I wrote about a few weeks ago suggests that during the current cycle the PLT numbers have been capturing pretty well not just declared intentions but what actually happened.    In the year to June 2017, a net 72,305 people arrived as PLT migrants.   Just slightly more than that number of non-New Zealand citizens arrived, and 1284 New Zealanders (net) left.

PLT sept 17

As people often stress, a lot of the variance in the net PLT series is typically accounted for by changes in the choices of New Zealanders (net outflows have fluctuated between around 0 and around 40000, and there have been quite big fluctuations –  hard to predict –  every few years).  The choices of New Zealanders are not a matter of immigration policy.

But policy has pretty full control over the number of non-citizens arriving (Australians are allowed in without advance specific approval, although the numbers typically aren’t large).   And sometimes you will see this chart, which uses PLT arrivals data (gross, not net) to show what sort of visa people were on when they crossed the border as PLT arrivals (the “not applicables” are New Zealand and Australian citizens).

PLT arrivals by visa

But this chart doesn’t tell us anything much about immigration policy.  In the year to June 2017, 16711 people arrived on residence visas.  But during that year, MBIE granted 47331 residence visas, the overwhelming proportion to people who were already here (and who typically will have entered first on a student or work visa).  Perhaps it is worth noting, for all the talk of the success of the export education sector, by far the biggest increase in arrivals in recent years (absolute and percentage) has been in people with various types of work visas: around 24000 in the year to June 2012, and around 45000 in the year to June 2017.

If we want to look at immigration itself, it is much better to turn to the administrative data on the numbers of people approved for various classes of visas.  Unfortunately, unless you like playing with spreadsheets with half a million lines, MBIE only produce data annually, for June years, and the data for the year to June 2017 hasn’t yet been released.   Having said that, it doesn’t look as though there will have been big changes when the data do finally emerge.

Here are the numbers for visas granted to new workers under various policies (ie excluding renewals etc).

Number of new workers by policy
2011/12 2012/13 2013/14 2014/15 2015/16
Study to work 9,319 9,131 6,259 9,610 16,097
Essential skills 6,197 6,247 7,885 7,709 8,334
Work to residence 1,653 1,558 1,426 1,483 1,717

and there has been a big increase in the numbers granted working holiday visas

2011/12 2012/13 2013/14 2014/15 2015/16
Working holidays 41,561 47,168 53,131 59,742 63,230

Fortunately, Education New Zealand don’t seem to mind the half million line spreadsheets, and produce a nice monthly product on student visas.   Here is the chart of outstanding valid student visas by class of institution for the last few years.

vsv

Numbers are growing, but in the last year or two there has been quite a switch from private training enterprises (which will have included some of the more questionable institutions/courses) towards universities in particular).

What of residence approvals?  I did download the huge spreadsheet for that subset of the data to get an overview of the 2016/17 numbers.  Here are residence approvals in the last few years.

Number of residence visas approved
2011/12 2012/13 2013/14 2014/15 2015/16 2016/17
40,448 38,961 44,008 43,085 52,052 47331

Recall that (a) there is a “planning range” (in effect, a target) for the number of residence approvals granted. That range was 45000 to 50000 per annum, but was cut to 42500 to 47500 late last year.  Actual approvals fluctuate around the target, rather than being mechanically managed to meet it month by month or year by year.  The 2015/16 approvals were high, but the numbers have been cut somewhat in the most recent year.

Recall that most of those getting residence visas were already living here (on work, study, or related visas).

In terms of nationality, in 2015/16 these were the top source countries

China 9,360
India 8,498
United Kingdom 4,934
Philippines 4,614
South Africa 2,970
Fiji 2,230
Samoa 2,156
United States 1,288
South Korea 1,125

I didn’t calculate all the numbers for 2016/17, but the patterns looked pretty similar.

I hadn’t seen this data in the published MBIE summaries, but I was a little surprised to find that among the residence approvals 1937 were for people in a category of

Uncapped Family Sponsored Stream Dependant Child

These aren’t the children of principal applicants who are themselves getting residence visas (as those children are approved with the parents).   Around half of all these “dependent children” were Samoan, and of them 242 were aged 20-29, not typically what one thinks of when one hears of “dependent children”.   I’m not sure how or why such a policy exists, but when I get time I might have a dig around.

So that is the numbers.  Perhaps the key thing to keep in mind is that the residence approvals planning range –  the centrepiece of the immigration programme –  has been pretty stable for a long time (modest cut last year).  Much of the variability in the headline PLT numbers is New Zealanders, and most of the variability in the non-NZ net inflow relates to policy streams other than the residence approvals programme.

Of course, variability is only part of the picture.  The striking thing about the residence approvals programme is its sheer size: equivalent to almost 1 per cent of the population each year, and in per capita terms three times the size of the US “green card” issuance (under both recent administrations).   We have a very large number of legal temporary foreign workers here by international standards, but most of them will eventually go home.  What really marks us out is the size of the residence approvals programme –  bigger per capita than in almost any OECD country, and far bigger than most.   I’ve argued for cutting programme back to, say, 10000 to 15000 per annum (a similar size, per capita, to the US programme.)

As I’ve noted here previously, if one looks at the New Zealand First website there isn’t much specific on immigration policy.   Winston Peters has sometimes talked of lowering the annual inflow to something like 10000 to 15000, but quite what is meant by that hasn’t been clear.  Most naturally he may have wished to suggest a net PLT inflow of around those numbers.  If so, it would have to be treated as an average over time, since annual PLT flows are almost wholly unpredictable (given the variability in the net flow of New Zealanders).

Having said that, one could make some estimates of a trend net outflow of New Zealanders, likely to resume as the Australian labour market improves.   Assume that outflow is 20000 on average over the cycle (a bit less than in the past), and you might lower the residence approvals target to 30000 to 35000 per annum (the net of the two flows on average producing something like a 10000 to 15000 inflow per annum).  That doesn’t sound terribly radical, and frankly there looks to be plenty of room to (a) drop off the lower-skilled portion of the current approvals, while (b) removing the sort of absurd bureaucratic hassles really skilled people (eg the teachers profiled in the Herald the other day) can face.

One of the other, rather general, strands of New Zealand First’s immigration policy is

Ensure that there is effective labour market testing to ensure New Zealanders have first call on New Zealand jobs.

I’m sceptical of the practical means to do this, even if I’m somewhat sympathetic to the concerns that motivate it.  I don’t think bureaucrats should be trying to decide which job is really in excess demand, let alone try to reach Soviet-type judgements on which regions should be favoured, or whether wages for those particular skills should just be left to rise.  But in various recent presentations, I have included an option for reforming the work visas system (in addition to substantially tightening up on student work visas and post-study visas, for those with lower level qualifications)

Institute work visa provisions that are:

a) Capped in length of time (a single maximum term of three years, with at least a year overseas before any return on a subsequent work visa), and

b) Subject to a fee, of perhaps $20000 per annum or 20 per cent of the employee’s annual income (whichever is greater).   [To limit risks of exploitation, require the employer to prove that the employee has been paid at least $10000 above the mimimum wage, with no “fees”  allowed to be paid back to the employer or related entities.]

The key element is the second one.  If your firm really needs a highly-skilled person (surgeon, lawyer, CEO or whatever, earning say $200000 or more), and can’t find one on market in New Zealand, the annual fee is unlikely to be prohibitive given the key short-term such a person is like to be playing.   But, equally, there aren’t many of those sorts of people/roles, and many won’t want to stay here forever.  So I’d make it easy to recruit them, but with a strong emphasis (because the visa is non-renewable) on the need to identify a local permanent person.   At the bottom end of the labour market, if the business your firm is doing is really so valuable you can afford the $20000 annual fee on top of the annual salary, that might be a reasonable pointer to serious scarcity.  But it seems unlikely that we’d be granting many visas to lower-end chefs, or dairy workers, or aides in rest homes.  And that would, over time, be a good outcome for New Zealanders.

 

(And MBIE could you please please make the monthly data more easily available in an accessible format, as Statistics New Zealand and other agencies do.)

 

 

A problem awaiting the new government

Michael Reddell Exports, New Zealand economic performance

Whichever party, or group of parties, gets to form the next government will face the same facts about our disappointing economic performance.  As I noted a few weeks ago, based on the recent PREFU projections, not even Treasury seemed to rate very highly the chances of meeting the National-led government’s export objective.

Here is the share of exports in GDP, showing actuals for the last decade or so, and Treasury’s projections for the next few years.

x to gdp

By the end of that forecast period, there will only be four more years until the goal of a much-increased export share of GDP was to be met.  On these numbers, exports as a share of GDP would by then be at their lowest since 1989, 32 years earlier.  So much for a more open globalising economy.

One of the indicators I like to use is a rather rough and ready decomposition of real GDP into its tradable and non-tradable components, first developed by an IMF staffer looking at New Zealand a decade or so ago.    It assigns the primary sector and the manufacturing sector components of real production GDP, and the exports of services component of expenditure GDP, to the tradables sector –  the bits where New Zealand firms are competing with the rest of the world.  The rest of GDP is classed as non-tradable.    It isn’t a precise delineation by any means: some local manufacturing isn’t really tradable (due to high weight and low value, and thus transport costs) and, for example, the electricity generated for Comalco is, in effect, tradable.   But, broadly speaking, it seems to capture something meaningful about the New Zealand economy.  In the early days of the National government, then Minister of Finance Bill English was quite keen on it.

All economies need firms in both tradables and non-tradables sectors.  So one sector isn’t inherently better or worse than the other.  But countries that are catching up with the world-leading economies tend to be ones in which the tradables sector (exports and import-competers) lead the way.  In such economies, firms are finding more and better, more lucrative, ways to tap the much larger global market.  Of course, we also gain when the non-tradables sector is becoming more productive –  both directly as consumers, and as a reduction in the input costs of tradables sector firms.    But there is a limit to how many cafe meals we can serve each other.  There isn’t really a technical limit on, say, how many smart ideas, translated into appealing products, that firms in a small country could sell to the rest of the big world.

As well as dividing real GDP into tradables and non-tradables components, I’ve also expressed both components in per capita terms.    Over long periods of time, most real economic series trend upwards, and actually it is something like per capita production or value-added that matters most in looking at gains in material well-being.   Here is the latest version of my chart, updated for last week’s GDP release.

t and nt components to jun 17

The series do bob around a bit.  The tradables sector, for example, had a very good June quarter on the back of a couple of tourism one-offs (the World Masters’ Games and the Lions tour) but then it had had a poor year last year.   But what I try to draw attention to is that (a) the peak in the tradables series was as long ago as 2004, and (b) real per capita tradables sector output is now no higher than it was at the end of 2000, almost 17 years ago.  Across the whole terms of two governments, one National-led and one Labour-led, there has been almost no growth at all in the real per capita GDP of the tradables sector. None.

Some economists really don’t like the chart.  So lets look instead at each of the components that make up the tradables sector measure.

tradables components 2

Services exports, in real per capita terms, did very well in the 1990s, growing quite strongly until around 2002.  But, overall, almost no growth since.   The mining sector briefly did very well around 2007/08 when a new oil well came on stream.   And, in per capita terms, the agriculture, forestry and fishing component of production GDP, and the manufacturing component, have gone almost nowhere over 25 years, again in real per capita terms.

What changed 15 years ago?  Well, one of the things that has changed a great deal is the real exchange rate.  Here is a chart of the Reserve Bank’s index, showing an average for the last 15 years (as well as one for the previous few years).

rer to july 17

It is unlikely –  all but inconceivable in fact –  that if we keep on doing what we’ve been doing for the last 15 years or more, in terms of economic policy settings, that we’ll see any sustained per capita growth in our tradables sectors.  It is that old line about a definition of insanity being doing the same stuff over again and expecting a different result.

Even to sustain those sectors at the sort of flat levels –  no growth at all – we’ve had over the last 15 years or more has involved the significant subsidies of (a) unpriced pollution externalities especially around water, (b) significant direct subsidies (to, most notably, the film industry) and (c) significant effective subsidies to the export education industry (by offering a bundled product where students can pay for an education –  in some cases an “education” –  and get preferred access to work and residence visa entitlements too –  that benefit being provided free to the providers by the New Zealand government).

I’d be very happy for a new government, of whatever stripe, to deal directly to any or all of those distortions.  But they, and their advisers, need to bear in mind that exchange rate chart.   Unless the real exchange rate falls quite materially, it is difficult to envisage much growth in other tradables industries to replace the shrinkage in the subsidised industries.  (It was exactly the same issues policy advisers faced when we started liberalising, and stripping away earlier subsidies, in 1984.)   Real exchange rates can’t be managed directly, but they can be materially influenced by removing the sorts of other policy distortions that put intense pressure on domestic resources, and drive up the prices of non-tradables relative to tradables, skewing the economy away from the tradables sector.

I’m not optimistic about prospects, but the good thing about pessimism is that one can, just occasionally, be pleasantly surprised.