Who will build the houses?

One of the Prime Minister’s campaign lines has been “who will build the houses?” if immigration numbers are cut back.   It is a curious line of argument, for a variety of reasons.

But it takes on a particular air of unreality when used –  as I heard in a debate last week –  to attack the Labour Party.    After all, Labour is campaigning on a policy that will (a) leave the current 45000 per annum residence approvals target unchanged, and (b) reduce students visa numbers quite substantially (resulting in a one-year reduction in the net PLT migration inflow).  On their own numbers, the changes they are proposing won’t make that much difference to the number of work visas issued, and where those numbers do change, the intention is to focus reductions at the lower-end of the skill spectrum.  (Their document is here, and my post on it is here.)   For what it is worth, Labour even proposes a Kiwibuild visa, designed to ensure that any reductions in work visa numbers don’t interrupt a flow of construction workers.

The student aspect aside (and even that isn’t part of the 100 day plan, although it isn’t that long until the new academic year starts), one might reasonably doubt whether Labour is serious at all about reducing ongoing immigration pressures.   Their policy, if implemented, won’t materially alter the net inflow over time. And I heard this morning an extended interview with Jacinda Ardern on Radio New Zealand in which she declared that she would have no problem at all with a net 70000 migration inflow per annum if only the houses were there, and actively endorsed some recent strongly pro-immigration comments made by Helen Clark.    Labour, like National, is still a “big New Zealand” party –  despite the economic damage that strategy has been doing over decades (remember how bad our productivity record has been) and will continue to do (ever more people and a heightened priority on improving water quality and meeting climate change targets is a recipe for severely undermining our productivity prospects.)

But this post isn’t about Labour’s proposals, but about (a) what has actually been happening over the last few years in the construction sector and related migrant inflows, and (b) more briefly, how the economy might adjust if there was to be a sustained material cut to target levels of non-citizen immigration.

In his weekly column in last Friday’s Herald, Brian Fallow touched on some of the first of those topics.  He went to the latest annual MBIE Migration Trends and Outlook publication (for the year to June 2016 –   MBIE could you please make data easily accessible on a more timely basis), looked at the data on who had been granted Essential Skills work visas in recent years, and concluded thus:

The conclusion has to be that the impact of net migration flows on the housing market and the construction industry is overwhelmingly on the demand, not the supply, side.

There has been a big increase in construction activity in New Zealand in the last few years.  Some of that is driven by the Christchurch repair and rebuild process, but increasingly the key influence has been the unexpectedly rapid growth in the population.  Each of those people needs a roof over their head.

And so employment in the construction sector has increased rapidly.    Here is the data from the HLFS, showing the percentage increase in people employed from calendar 2013 to calendar 2016 for each of the sectors employing more than 100000 people.

HLFS by sector

The construction sector has had by far the biggest increase in employment over the last three years.  Around 56000 more people were employed in construction in 2016 (on average) than in 2013 (the current total number of people employed in the sector is around 240000).

What contribution has non-citizen immigration (the bits our policy controls) made to this employment?

As Brian Fallow noted, on MBIE’s own numbers, this is how many Essential Skills visas were granted for construction trades and construction labouring roles in the year to June 2016.

And a startlingly low proportion – 7 per cent, or 2233 to be precise – were classified as construction trades workers like carpenters, plumbers, plasterers, tilers and painters. If you include scaffolders and builders’ labourers, the proportion rises to nearly 10 per cent.

And here are the corresponding figures for the previous couple of years.

Essential skills visas granted
2013/14 2014/15 2015/16
Construction trades workers 2090 2123 2233
Construction and mining labourers 399 546 831
Construction sub-total  2489  2669  



Total 26502 28548 31766
Construction as % of total 9.4 9.3 9.6

Those might look like quite large numbers but:

  • at last report, construction jobs made up 9.3 per cent of all employment (and yet in this really rapidly growing sector only around that share of Essential Skills visas –  suggesting that immigration was hardly easing sector-specific pressures), and
  • as Brian Fallow also pointed out, most of these Essential Skills visas were being granted to people who were already in New Zealand (eg renewals).  Of 31766 Essential Skills visas granted in the year to 2016, only 8334 (or 26 per cent) were new workers (the proportions are similar in the earlier years).
  • people arriving and taking up first-time work visas need to be offset against people leaving.    In the three years I’m looking at here, MBIE tells us that the total stock of people here on essential skills visas increased by only 10062.   If the patterns were similar for construction jobs as for other roles, construction would account for about 1000 of that increase.

Of course, some people will have moved from work visas and obtained residence visas.  Based on the 2015/16 residence visa approvals numbers, that might have been around 450 people working in construction roles per annum.  Over three years, perhaps as much as 1500 people.

In other words, in a construction sector where total employment has increased by 56000 in three years, perhaps only 2500 (or less than 5 per cent) of that increase will have been met by the immigration of non-citizens.

So in that sense the answer to the Prime Minister’s question is easy.  Who will build the houses if immigration is cut back?  The same people who always overwhelmingly have, people who were already here.

But perhaps more importantly, if immigration were to be sharply cut back, the number of people needing accommodation would fall.  At one extreme, if the population is growing by 100000 per annum (as it has in the last couple of years), that suggests a “need” for around another 35000 houses each year (on top of the small number that would need replacing each year with a static population).  With net non-citizen migration at present in excess of 70000, the non-citizen immigrant flows alone create a “need” for perhaps 23000 additional houses each year.     Even if we go back to Brian Fallow’s original numbers of gross approvals of Essentials Skills visas, 3000 construction workers cannot build 23000 houses a year.   So the way immigration policy is actually being conducted it is exacerbating pressure on the construction industry, not relieving it.  That additional pressure is substantial.

(It needn’t be that way of course.  In the short-term immigration will almost always in increase economywide demand more than it increases supply.  But composition of the immigrants afffects where the pressures are most felt.  At the extreme, if all the migrants were builders (and related occupations), they’d probably just about keep up with the additional demand for housing (building, in effect, to house themselves).  Then the demand pressures would show up more severely in other sectors.    But when there is a big increase in the population, and hence in construction activity, immigration policy certainly isn’t relieving construction sector constraints when only around 10 per cent work visas are going to construction workers, when almost a quarter of the new jobs are in construction.)

So what would happen if, say, the 45000 residence approvals target was cut to, say, the 10000 to 15000 per annum I’ve been advocating (still, in per capita terms, around the rate of permanent approvals in the United States), and issuance of work visas was also tightened up, so that the stock of people on temporary work visas was no longer growing?

Overall, growth in domestic demand would weaken, and with it the pressure on domestic resources.  The notion that the short-run demand effects of immigration outweigh the supply effects shouldn’t really be controversial.  It has been that way in New Zealand for many decades.  But, given the huge scale of the pressures that new people put on the construction sector (not just houses, but roads, schools, offices, shops etc), and the fact that immigration policy as actually run has not seen us bring in many construction workers (10 per cent of the visas, when 25 per cent of the new jobs have been in construction), such a policy change would greatly ease resource pressures in the construction sector specifically.  In some other sectors it is quite conceivable that resource pressures could increase (one could think of export-oriented sectors such as tourism or dairying) if such an immigration policy change was made.  But on the construction side of things –  one of the most politically and economically pressing areas of our economy – the gains (the relief of pressure) would be substantial and almost immediate.  Not only would construction sector resource pressure ease, but land prices could also be expected to fall back to some extent (due to a reduction in expected future demand).

More generally, across the economy one would expect to see  interest rates falling (both market interest rates and the OCR) and with them the real exchange rate.    A lower real exchange rate would help secure the overdue resource-switching towards the tradable sectors. It would also provide the additional margin that would enable employers in those sectors to bid up wages to the extent required to attract existing residents to take up jobs in those sectors.    Plenty of people would be freed up from the construction sector –  a country with a modestly growing population wouldn’t have 10 per cent of total employment in construction –  and they’d be looking for jobs elsewhere.  Most of them would be long-term residents or citizens –  something we know with a high degree of confidence because the government’s own data tell us not many visas have been issued in recent years to people in construction, whether skilled workers or labourers.

But I guess the Labour Party can’t really use these arguments to push back against the Prime Minister because they aren’t actually planning a material and sustained reduction in non-citizen immigration at all.  That’s a shame.

(And if you wonder why all this discussion has used visa numbers up to June 2016, that is because MBIE only release more recent numbers in massive (600000 line) unwieldly spreadsheets.  It is possible that patterns in the last year have been a little different, but it seems unlikely –  given the similarity in each of the previous three years.  But debate would be better-informed, and more timely, if MBIE would make  timely data available in more readily accessible formats, as happens for almost all other important economic data released by Statistics New Zealand, the Reserve Bank or whoever.)