A statistical surprise

The New Zealand Association of Economists runs a poster competition at each annual conference - a good idea, encouraging concision and clear messaging, not always a forte of economists - and this year it was deservedly won by Alexandra Turcu of AUT's Work Research Institute, with her entry, "Underutilisation in the New Zealand labour force: Unused human capital, or an underpaid workforce?". The winner's decided by popular vote of the attendees, and I voted for it, too.

Before reading it, and (I'd guess) like everyone else, I'd always assumed that the "underemployed" were available to work more hours than they currently are, and were a reserve slush fund of unused labour availability. But Alexandra discovered something striking about the apparently "underemployed". 

She looked at the characteristics of two kinds of underemployed people - the full-time underemployed (full-time; available to work more; want to work more) and the part-time underemployed (part-time; ditto; ditto). Remarkably, she found that "Although the underemployed groups want to work more hours, and state that they are available to do so if more hours were available, our results reveal that they already work a similar amount to their fully utilised counterparts. The fully-utilised work only one hour more than the underemployed do per week": full-time employees, for example, who said they were "underemployed" were actually working 40 hours a week, virtually the same as the 41 hours of fully employed full timers, and it was the same story with part-timers where the "underemployed" ones were putting in 15 hours a week compared to the 16 put in by fully utilised part-timers.

So it is not at all obvious that these people actually form any substantial pool of increased labour supply - an important thing to know if, for example, you're the Reserve Bank wondering about potential output and full employment - and it's likely that what they're telling Stats has less to do with their availability to put in more hours and rather more to do with their relatively low incomes. As Alexandra put it, "This begs the question: Are underemployed workers truly underemployed, or are they just underpaid? When asked why they were underemployed, the majority of respondents said it was because there was not enough work available. However, it is important to note that the HLFS [household labour force survey] survey did not include "not enough income" as a potential answer".

So one practical lesson to take away is that there is less slack in the labour supply than you might have thought, if you had been relying on those apparently "underemployed" being available to step up to the plate and do more work. And if you were concerned about their low incomes, there's a positive to take away: fully 57% of those "underemployed" part timers transitioned into fully utilised full timers in the following quarter. They don't sit on the sidelines for long.

Here's the full poster.

The Bank's other stuff

Yesterday's Monetary Policy Statement went entirely as expected - no change to the Official Cash Rate (still 0.25%), no change to the Large Scale Asset Purchasing Programme (still capped at $100 billion), and the introduction of the signalled Funding for Lending Programme, intended to provide a new source of cheap funding to lenders (three year funding at the OCR rate). 

If used fully, the new FLP would amount to $28 billion. In Australia, the equivalent Term Funding Facility, which has been going since April, has a capacity of some A$200 billion, or about NZ$212 billion. Divide by 7 as a rough pro rata rule of thumb and the Aussie TFF would be a NZ$30 billion or so programme here, so the good news is that we've introduced a similar-sized stimulus.

There is a line of thinking that providing extra funding for lending might be the proverbial 'pushing on a string' if, in still unsettled Covid conditions, borrowers aren't much minded to take on more debt or banks aren't much minded to take on more risk, and it's true that (as of early November) the Aussie one has seen only A$83 billion taken up, or some 40% of the total available. But as the RBNZ said in the Statement, the actual take-up may not matter so much if the new FLP gets the cost of borrowing down, or as the Bank put it (p21)

The key success metric of the FLP will be whether it results in declines in funding costs, and encourages recent declines in these costs to be passed through to lower household and business borrowing costs. We could see a scenario where FLP funds are only drawn down in small amounts, but its availability encourages a broad decline in interest rates. We would consider this scenario successful, even though actual use of the FLP would seem minimal.

There's always interesting stuff in the body of the Statement and this time round I was especially interested in what the RBNZ had found as it went around the business traps, and in its comments on house prices.

"Many businesses", the Statement said (p18), "expressed concern about finding required staff. Some firms noted that re-deploying staff from one industry to another can be difficult, particularly for skilled jobs. Many firms rely on hiring skilled workers from abroad, which they have been unable to do because of the border closure". 

Granted, the large numbers at risk of losing their jobs in the most affected sectors - the Statement reckoned that pre-Covid international tourism and education made up 6% of GDP - won't always be a good fit for the vacancies available elsewhere, and labour market policy is always going to struggle to assist the transition. "Active" labour market policies (like these successful ones) try to ease the process, but you do wonder whether we've got enough of them. And the reported dependence on overseas skills again makes you wonder how well our labour market is working to match up the demand for skills with the supply of them.

House prices have grabbed everyone's attention, not least because the latest resurgence wasn't supposed to happen in a world where (supposedly) shell-shocked households were hunkering down, not trading up the house. It got a fair bit of attention at the media conference at the Bank after the Statement (tune in around the 29:25 mark, and stay with the rest of the video). But the Bank was quite right to say that it's not its ever-lower interest rates that are the big moving part in the house price increases. As it - completely correctly - said in the Statement (p28):

High house prices in New Zealand largely reflect structural and regulatory issues in New Zealand’s housing market. In particular, land use restrictions, such as urban planning rules, limit the land available for housing and how intensively it can be used. These land use restrictions impede the ability of the market to increase the supply of houses when demand for houses increases. As a result, house prices tend to increase more than otherwise in response to higher housing demand. Other supply-side issues include infrastructure  planning, the building consent process, and the cost of building.

You can berate the Bank all you like, and launch market studies into the building materials trade till you're blue in the face. They're all in the twopenny halfpenny place. Nothing's going to happen to high house prices until the supply side of the market starts to work a lot, lot better.

The OECD's take

Yesterday I had a go at estimating what the drop in New Zealand GDP had been during the June quarter, when the lockdowns were in place. Meanwhile, in Paris, the OECD was just putting to bed its Employment Outlook 2020 (best link here, others got on my wick with overenthusiastic use of dynamic charts and highlighted text), and they helpfully included this chart of what they think happened in June across the OECD. 

They have the New Zealand June decline at -15%: yet another sign that our June quarter wasn't quite as bad as it originally shaped up to be. It's also, despite the relative stringency of our lockdown, not too bad an outcome by OECD standards: the median decline across the OECD was -12.9%. If you'd like the raw data to play with they're here.

The OECD report also had this interesting chart on use of wage subsidy (and similar) schemes. We're right over there on the left hand side as a big user.

That's fine by me - it was IMHO a completely appropriate and even necessary fiscal response, and especially by New Zealand's chronic micromanagement standards a remarkably quick, simple and effective one. And it's interesting to see that nearly all of the other countries (ex France and Germany) went the same hassle-free 'ask no questions' route and approved virtually every application that came in.

As things stand, the wage subsidy scheme won't be available for much longer (last applications close on September 1): at the moment I can't see how it can sensibly be wound up without some replacement made available for the worst of the tourism, hospitality and accommodation sectors.

Beneath the calm surface

The Productivity Commission's been working on 'Technological change and the future of work'. Last month it came out with the second report, 'Employment, labour markets and income', in what will be a five-part series (press release here, whole thing here). The big headline takeaway was that it might be well worth looking at systems like Denmark's 'flexicurity', where people's incomes are supported through employment volatility. The idea is that the labour market needs to be able to be flexible, and jobs will come and go, but people's incomes will be cushioned against the volatility through, for example, an employment insurance scheme. All very sensible.

Along the way Chapter 4 looks at the case for 'active' labour market policies, things like retraining to help people find new jobs. The Commission is somewhere between agnostic and outright sceptical about their value: "There is a large gap between good intent and robust evaluation of the effectiveness of labour-market programmes. Few programmes are subject to robust evaluation.  Of labour-market programmes, ALMPs ['active' ones] have received more evaluation effort. The results of those evaluations are not encouraging (p78) ... Overall, the Commission cannot say whether New Zealand’s labour-market programmes are effective or not" (p80).

The Commission may have missed the latest bit of evidence, which I wrote about in 'Let's get more active', and which was more upbeat about their potential. It found that two kinds of programme appeared to be effective (wage subsidies and helping people to go out on their own as self-employed), vocational training wasn't too bad an option, but brokering services (helping match job seekers and recruiters) were a waste of space. So if I were the Commission I think I'd be taking a modestly more constructive view of the potential to make the labour market work better, particularly as it's a vital economic issue.

For one thing, governments in many countries (though not Denmark, obvs) have been failing to live up to the social compact underpinning an open, flexible, market-based economy. The core bargain is that the national gains from openness will create enough income for the winners to be able to compensate the inevitable losers and still come out ahead. But the redistribution to the losers hasn't been happening, and the resulting resentment in the world's Rust Belts is feeding tear-up-the-old-rules populists everywhere.

For another, virtually nobody outside the economics trade (and not always inside it, either) realises just how vast the flows in and out of the labour market actually are. We learn from Stats, for example, that total employment went up by 16,000 in the June quarter, and by 6,000 in the September quarter. That doesn't look like a lot of movement.

But what is actually happening is that huge numbers of people change jobs, get fired, and get hired each quarter. The 6,000 in the September quarter is the small net effect of enormous gross flows in, out, and between.

In recent years, roughly 155,000 new jobs are created each quarter, which has happily been ahead of the 145,000 or so jobs that have gone bung in the quarter. The 10,000 or so increase in employment in each quarter is the outcome of very large gross flows indeed. The data, by the way, come from Stats' Linked Employer-Employee Dataset ('LEED'), which you can play with yourself for free on NZ.Stat. I've done rolling four-quarter averages to take out the pronounced seasonality.

And the big levels of job creation and job destruction are only part of the wider ferment in the labour market. People are moving around from one job to the next in very large numbers. A bit over 350,000 people each quarter change seats.

Are we out on a market-turmoil limb here? Not at all. In the States, for example, the increase in jobs in any given month is around 200,000: in December it was 266,000, which was thought of as quite a large increase at this late stage of the long U.S. expansion. But that is absolutely tiny compared to the gross flows. According to the U.S. JOLTS data, which show us the underlying gross flows, in the month of October alone (the latest to hand), 3.5 million people voluntarily quit their job in the month. Another 1.75 million were laid off or fired. Employers hired 5.75 million people. In one - one - month.

Bottom line. There are two reasons we ought to be helping people a lot more to cope. One is that moral compact: for both efficiency and equity reasons, we need to have a dynamic but not painful labour market. And the other - acknowledging that a fair amount of it is entirely voluntary, with people quitting (especially in good times) to do better for themselves in a new job - there's far more turnover in the labour market than you likely thought. Flexicurity, and 'active' labour market programmes, aren't just for the unlucky few in the meat processing factories: they're for all of us.

Let's get more active

By far the best single way to get unemployment down is to keep up a good run of economic growth at a robust pace. Not only does the unemployment rate go down, but the longer and stronger the expansion, the more it succeeds in bringing more marginalised groups into employment.

I've shown previous versions of the graph below: here's the up to date version, which shows (yet again) that less favoured groups suffer distressingly high rises in unemployment when the business cycle goes pear-shaped, but conversely a long expansion works its magic on everyone, even on those who tend to be on the outer. It's not perfect - rates of unemployment for some groups, notably Maori and Pasifika, are still too high - but the long post-GFC expansion has seen big falls in the unemployment rate across all groups..

Because growth alone will not deal to everything there's always a role for 'active' labour market policies that try to make the labour market work better. Residual unemployment for example could be down to things like skills mismatches: employers are looking for people with skills they don't have. Or it could be down to regional immobility: who's going to move from Northland to a job vacancy in Auckland given the cost of putting a roof over your head in Auckland? And there can be other mismatches preventing people willing to hire from signing the deal with people wanting to work.

There's a brilliant new bit of research out looking at active labour market policies, and which ones work. They're not just any old active policies: they're ones where there was a control group experiment, where you can see how those who went through the programme fared compared to those who didn't, which is how the researchers can tell if it made any difference. Here's the more readable version for the intelligent public - 'Understanding what works for active labour market policies', on the excellent Vox site - and full-blown policy tragics can get the more academic version here.

Here's the key result in terms of the impacts on earnings and employment:

If we focus on the median impact on earnings, wage subsidies and independent worker assistance ['Support to micro-entrepreneurs and independent workers'] show the greatest impact relative to the control group, with improvements of 16.7% and 16.5%, respectively. Vocational training programs have a median impact of 7.7%, while employment services show an almost negligible impact. The median impact on employment outcomes exhibits a similar pattern

Here's the graph they drew to show the results, though to be honest it's hard to see the scale of the impact with the vertical scale they've chosen to use.

They also discovered a variety of other things that make complete sense: "Individualised coaching or follow-up of the participants, training exclusively focused on a specific industry, and the provision of monetary incentives to trainees all correlate with better outcomes in vocational training programmes (the most frequent ALMPs in our dataset)", and, unsurprisingly, they also found that you get better outcomes when you run these programmes in good times rather than in the pits of a recession.

They also found (noting that cost data isn't uniformly available) that you get what you pay for. As the graph below shows, employment services programmes are cheap, but useless, while the most effective option, helping people to do their own thing, costs the most.

There's a lot we could learn from this. Currently we're down the wrong end of the OECD league ladder when it comes to what we spend on active labour market policies (as I discussed here). This research gives us a pretty clear steer on what we should do to up our game.

It's also left me wondering about our policy institutions. This research works because all over the world people have invested in randomised control trials - experiments where you try something out and compare it with what happens where it wasn't rolled out. But I can't think of many home-grown examples where we've had a go at economic policy field experiments: charter schools, maybe, but they got sat on.

My feeling is that we're too fond of the over-dirigiste 'one size fits all' approach. And our chronic political oppositionism makes both the public service and the decision-making politicians far too wary of experiments that might go wrong - even when they might hold valuable information about what works and what doesn't. There's too high a political price to pay for what will be pilloried as 'failure', and it's hindering finding out what might or might not make a real difference to those hardest to get into work. We should move on: it's time to join the adult policy world.

Yet another reason why workable competition matters

A small but select audience turned out last night for the latest Auckland LEANZ seminar, featuring Bill Rosenberg, the Congress of Trade Unions' policy director and economist, on the topic of "Low Wages: Is Competition a Factor?". This was a reprise of a May presentation in Wellington which Bill gave in association with Peter Cranney, an employment lawyer at Oakley Moran: the May slides can be found here.

Maybe the topic is more of interest to a Beehive-focussed audience, but it's a pity more people didn't attend in Auckland, because Bill made an interesting case. Beforehand, I'd been afraid that he was going to have a go at attacking free (or at least workably competitive) markets, and I'd been sharpening my claws. Quite the opposite, as it happened: Bill's case might as easily have been titled, "Low Wages: Is Lack of Competition a Factor?". His argument was that businesses have accrued too much monopsonistic market power in the labour market, and for a variety of reasons, including as a side effect of increased market power in goods markets as 'super star' companies have emerged and (perhaps) merger policing hasn't been all that it might have.

Skewed labour markets are becoming A Thing in antitrust circles, and I suspect we're going to hear a lot more about them. Here for example is the peroration from a recent article (cited by Bill) in the Harvard Law Review:

Labor market power is ubiquitous and costly to society. It is bad for economic growth and equality, and fuels political conflict. Yet labor market power is generally ignored by antitrust authorities and never considered as a justification for subjecting mergers to scrutiny. This contrasts with the regulatory concern for product market power. We argue that this asymmetry is not justified by either legal doctrine or economic theory and suggest that the economic analysis of product markets regularly deployed in the scrutiny of mergers can easily be applied to the labor market (p600)

How you might address imbalances of market power in the labour market is the tricky bit. Maybe the generic competition law of the land has a part to play (on egregious use of no-poaching or non-compete agreements, for example), although as the Harvard Law Review article says, even in the litigious US it hasn't got very far. 

Bill's got a menu of other proposals (his slides 38-40). Two of them got the tick from me - raising productivity, and better support for people affected by disruptive job losses - and while I'm more ambivalent about the third route (a greater role for collective bargaining) it's possible that the Fair Pay Agreements Bill supports might deliver net benefits.

The real trick would be to pull off what Denmark's managed - retaining a high degree of employer hiring flexibility, while also providing a lot of support (eg through retraining) to displaced employees. We're not currently crash hot on the support side, as this OECD graph shows (original document here), whereas Denmark is tops.

These 'active' labour market policies also have something of a moral imperative to them. If you subscribe to the idea that freer global trade brings net national benefits (as it does), and you recognise that the overall benefits create winners and losers (as they do), then there's a good case for looking after those who have fallen in the overall victory. And there's also cold-eyed politics involved: if you don't help out,  the losers will turn on trade itself. Trump's election and the US ranking on the graph are no coincidence.

Well done Bill; thanks to Sasha Daniels from Spark who emceed the evening; and special thanks to James Craig and the team at Simpson Grierson who hosted the evening.

In their prime

Our latest monetary policy targets agreement requires policy to "contribute to supporting maximum sustainable employment within the economy", and the Reserve Bank now spends a fair bit of each Monetary Policy Statement reporting on where we are relative to the maximum sustainable level, using a suite of different labour market indicators.

In the latest Statement, the Bank said that "employment is near its maximum sustainable level" (p22). It might be above it: "Evidence reported by employers suggests the labour market is currently tight, and that employment is above its maximum sustainable level" (p22). Or it might be below it: "some other indicators of the labour market suggest that employment may still be below its maximum sustainable level. One example is the job-finding rate" (p23). But overall we look to be there or thereabouts.

While the Bank deploys a whole battery of perspectives on the state of the labour market, one line of attack that doesn't appear is what is happening to what the Americans call the "prime age labour force", those aged 25 to 54. They're the bedrock of the labour force - they're typically out of education and not yet contemplating retirement - and the argument is that it's the state of the core  labour force that matters most for things like wages growth.

Overseas what's happening to the prime age labour force consequently gets quite a bit of air time: here, for example, is the Wall Street Journal's graph explaining the November US jobs report (from 'Did the Job Market Slow in November? Here’s How It Compares', if you've got access).

Interestingly, despite the long post-GFC expansion in the US, neither the participation rate nor the employment rate for prime age people have got back to where they were just before the GFC, and there is a mix of structural and cyclical trends going on: there looks to have been a gentle trend downwards in participation even before the GFC hit.

The corresponding numbers for New Zealand don't get much focus at all (they weren't mentioned, for example, in Stats' news release on our latest employment data), so in the spirit of inquiry I went and dug them out to see what they might be able to tell us (they're easily calculated from the data in Infoshare). Here are the headline unemployment results.

The unemployment rate for prime age people is, as you'd expect, markedly lower than for the labour force as a whole. And it's certainly signalling a tight market: the latest unemployment rate (2.0%) is getting close to its all-time low (1.5% in late 2007). You'd imagine that a fair chunk of this low rate is frictional unemployment, and that there's precious little cyclical unemployment left.

Here is the participation rate picture. It's harder to interpret: we're in uncharted territory, as the prime age participation rate has been hitting record highs. Can it keep on rising? Is there still a large reserve of people who could be tickled out into employment? Who knows, but you'd guess that the reservoir must be getting lower. We're down to only 14% of prime age people who are not in the labour force, and who are doing things like looking after young or elderly family.

The prime age participation rate is usefully splitabble into male and female participation rates, and here they are.

The overall rise to record levels of prime age participation is largely driven by a sharp and ongoing rise in female prime age participation, and like in the US there are surely both structural and cyclical things going on. That drop in the male rate over 1986-2000, for example, may well reflect the post-Rogernomics shrinking of traditionally male-dominated activities like meat processing. Changing social attitudes about (for example) who should stay at home and look after the kids will be in there, too. So it's very hard to unpick the purely cyclical component. I don't have any good feel at all for where the prime age female participation rate might peak.

Overall, the data don't give any huge overlooked insights into where we are relative to maximum unsustainable employment, mostly because the grand sweep of history and the changing attitudes to who works, and works at what, overlap the more cyclical aspects you'd like to isolate. If there is one useful extra bit of data, it's the prime age unemployment rate, which is confirming the RBNZ's "there or thereabouts" assessment.

A little light on a mystery

There's been something quite odd happening for quite a while. Employers are saying they are having a lot of difficulty finding staff, as this graph from the latest Monetary Policy Statement shows.

But at the same time there are large numbers of people telling the Household Labour Force Survey (HLFS) that they are currently part-timers but are available, and want, to work more hours. So how can we have employers tearing their hair out that they can't get people, and yet a whole bunch of people are saying "Pick me! Pick me!"?

Quite apart from being a mystery that it would be nice to unpick, it's also a reasonably important macroeconomic issue. If there really is a reserve army of people ready to take up jobs, or at the very least more hours in their current jobs, then the labour market mightn't really be as strong as it looks. Big wage increases wouldn't look terribly likely either if there are lots of people standing ready to take jobs at the current prevailing rates of pay. Conversely, if this reserve army isn't really there, then maybe employers' concerns are the thing to watch as an indicator of likely pay increases down the track, as employers bid against each other for the limited staff availability.

The regular HLFS tables don't shed a lot on these 'underemployed' people. We know from Table 12 of the HLFS that they are disproportionately women: of the 117,000 (seasonally adjusted) who were recorded as 'underemployed' in June, two-thirds (78,000) were female, compared to the roughly 50:50 male:female split of employment. And we know from Table 13 that of the 112,800 underemployed (not seasonally adjusted this time) in June, some 48,000 of them are not 'actively seeking' work.

That might suggest that they're not, actually, awfully interested in taking on full-time jobs, and maybe the employers' view is more descriptive of reality. But I'm not sure about that. I don't think it's the right thing to do, in these days of online job ads, to say that someone is not actively seeking work if all they've done is read the ads. But that's how Stats views things: you have to do more than scan Seek (or wherever) to be counted as 'actively seeking'. Not how I see today's world, but there we go.

Other than those little nuggets, though,we know nothing from the headline HLFS about these underemployed folk. So I asked Stats if they could break out the underemployed by industry or by occupation. And the ever-helpful people at Stats came up with the goods. Here are the answers. Stats would like me to say "Source: Statistics New Zealand, customised report and licensed by Statistics NZ for re-use under the Creative Commons Attribution 3.0 New Zealand licence", so there it is.

I'd love to say, Aha! Solved it! But the data, interesting and useful though they are, don't crack the puzzle.

You'd wonder, for example, when the housebuilding industry is desperate for anyone who can carry a hod, how 21,400 labourers say they can't get as many hours as they'd like.  That kinda points in the "yes, there's slack available" direction.

On the other hand the highish number of underemployed represented by retail trade and accommodation points another way. No matter how strongly business picks up at the motel, the motelier is very likely not going to add another full-time person: more likely, it'll be another person to do the 8.00am to 2.00pm shift to clean up the 10 extra units being occupied. The economy can pick up all it wants and won't make a blind bit of difference to motel cleaner-uppers (or peak-time sales assistants) who'd like longer hours.

And then you start asking yourself, why don't the motel and shop people move to other lines of business where there might well be a full week's work on offer? Is it because they have low level skills that aren't in demand? Unlikely, since demand for unskilled and low skilled staff is actually stronger than it is for skilled or highly skilled, as MBIE's latest online vacancies survey shows. Plus you look at the 7,400 managers (who you'd think have transferable generic skills) and the 17,000 professionals (a fair proportion ditto, you'd guess), and wonder why they don't move.

Still at least as many questions as answers, I'm afraid, but at least we've now got more data to be ignorant about.

You're welcome.

Doom and gloom? Yes and no

As any number of recent headlines will tell you, there's a bunfight going on about a slump in business confidence and a rise in unemployment. There's the usual partisan point-scoring going on about the size of it, who caused it, and what comes next.

What's really happening?

Let's deal to the unemployment rate first, up from 4.4% in March to 4.5% in June. Should anyone be worried about that?

No, for at least three reasons. One, the statistic comes from a survey, which has sampling error. If I've read Infoshare right, and I've been known to get it wrong, the sampling error for the unemployment rate is 0.3%, which means there's a 95% chance the true unemployment rate is between 4.2% and 4.8%. A 0.1% rise may not even have happened. And second, even if it did, the economy is not an automaton, and you expect to find "noise", random fluctuations even in the middle of a longer-term trend. And third, and most important, the unemployment rate rose for a rather comforting reason: the participation rate went up.

The logic is that the participation rate goes up in good times. People aren't stupid, and can judge what's happening in the jobs market. Discouraged people lurk outside the labour force when they reckon there are few jobs to chase. They delurk when they think it's all on. Sure, there'll be the odd person who's forced by bad stuff - the mortgage getting out of hand, a redundancy in the family - to go hunting for a job, but overwhelmingly the evidence is that the participation rate going up is a signal of the labour market running in job seekers' favour.

Put that together with the all-time record employment rates for women and Maori and a strengthening in wage increases, and the marginal rise in the unemployment rate is neither here nor there. If you get a press release from a pollie banging on about it, mentally subtract a little from your previous estimate of their credibility. Negative numbers are allowed.

The business confidence slump is not so easily dismissed.

For a start, it's beyond any question of sampling error or random wiggles. It's large, and evident over several readings. There's a lot of focus on the ANZ Bank's latest and particularly glum survey, but it goes back further than that. The NZIER's June Quarterly Survey of Business Opinion showed more unhappy campers, too. It's true that you should focus on the 'activity' measures in these surveys rather than the 'confidence' ones, but the activity numbers are also in rapid retreat.

As the latest ANZ survey said, "Firms’ perceptions of their own prospects are a better gauge of economic outcomes, but the news wasn’t upbeat here either: it dropped 5 points to a net 4% expecting an improvement. This is the lowest reading since May 2009 and well below the long-term average of +27". There have also been growth slowdowns captured in the latest BNZ / BusinessNZ surveys of manufacturing and services.

Here's one graph that I think helps explain what's going on. On the trusted principle that in a market economy an analyst should follow the money, here's what businesses have been telling the ANZ survey what they think the outlook is for their profitability.

There's clearly (to my eye) a political component. The sharp drop in expected profitability after we got the new Coalition government might have been rational - "this lot aren't business friendly" - but likely also had some sort of political protest mixed in. But businesses appeared to have got over themselves by March or April of this year - only for expected profitability to drop to even lower levels than immediately after the election. So my guess it's no longer a "should have been National" two fingers, but a signal of something more real.

A good deal of it, I suspect, is pressures on wages and other costs (notably energy) which haven't been able to be passed on. As many others have commented, the rise in the minimum wage, from already high levels by international standards as a percentage of average earnings, and with more to come, is putting sectors like retailing under intense pressure. Retailing has large numbers of minimum wage workers, and bricks and mortar shops have little or no ability to pass the costs on when e-commerce is already stealing their lunch. I was walking around Newmarket today for the first time in a while, and while the area was generally busy, I was startled to see how many retail vacancies there are.

Nor are many businesses enthused about the cost - real or imagined - of having to go back thirty years and sign up again for collective agreements. And I suspect they, like everyone else, are wondering about the sort of generalised wage pressures that look like leaking from the public sector. Any public sector union worth its salt has sized up this government as an easy mark. And they're right: the chance of a Coalition Finance Minister actually getting to the finishing line forecast in this year's Budget is half of five eighths. As I said at the time, "The likelihood of the New Zealand political process actually leaving $7.3 billion unspent on the table is extremely low".

I'm not even there myself. With the infrastructure - where it exists at all - creaking all around us, I'd be using that money, too, especially as we've got not only cash in the government cheque account but also the opportunity to borrow at once-in-a-generation low interest rates and make a substantial and lasting difference.

Inaction on that front may be part of business malaise too, especially when contrasted with the readiness to spend on lower quality ideas (think boondoggle regional lollyscrambles). I was somewhat dismayed, reading MBIE's latest national construction pipeline report (summary here, full thing here) that "Infrastructure is forecast to remain relatively unchanged, increasing marginally to $7.3b in 2023" (p1 of the summary) and that (p4) " Infrastructure activity is lower than previously forecast". And although the summary also notes (p4) that " Pacifecon’s research data suggests that there is a high value of infrastructure construction scheduled to be initiated over the next six years", it always seems to be light rail tomorrow but traffic jams today.

The previous government, by the way, was just as bad as getting the facilities built that would enable all of us to get on with our lives more productively. The longer it goes on, the more likely we're going to hit capacity and productivity constraints that stop the economy growing at the rates we'd like. Whatever else may or may not be needed to be done (or undone) by the current government to move us forward from where we are now, a larger and earlier infrastructure spend has to be part of the answer.

Bottom line, some of the beat-up over the slowdown (actual or imminent) in the economy is exaggerated. But some of it is realistic, especially if you put some weight, as you've got to in this late stage of the post GFC global recovery, on the external environment hitting a bump. Recent surveys of global fund managers, for example, show that they are worried about the impact of trade wars on world economic activity, and with a buffoon pressing the protectionist policy buttons, they're right to be biting their nails. So the ANZ's take looks realistic: "with businesses in a funk, it’s fair to say that the road ahead is looking less assured, and risks of a stall have increased".

How competition benefits women's pay

Last year the folks at Motu came up with a great piece of research which, among other things, showed that wage discrimination against women in New Zealand was less when firms faced greater competition.

The logic is simple: you might try to discriminate if you could get away with it without repercussions, but you'd pay dearly for indulging in your sexist preferences when there are plenty of firms competing with you who'll scoop up the qualified women and do better than you. Ditto, sexism will cost you when the labour market is tight, since you can't afford to be unfairly picky when staff are harder to find than usual.

Here's the guts of the results, from my blog post at the time. The impacts of more competition and tighter labour markets are very large indeed:

That already large pay difference [19.2% against women] is doubled - doubled! - if there's lots less competition among businesses in the sector. However the large difference goes away completely - to be consistent, completely! - if there's lots more business competition. It also goes away completely if a tight labour market is holding employers' feet to the fire and forcing them to make gender-blind hiring decisions, which is what you'd expect.

There hasn't been a ton of research elsewhere in the world along the same lines, but thanks to a lucky accident another bunch of researchers have also shown how increased competition between businesses works to women's advantage.

The lucky accident happened in Portugal, where, as 'Product market competition and gender discrimination' shows, a new programme got rolled out which made it hugely easier to set up a new company. It increased the level of competition against established firms from new companies who had previously been kept out of the game by an expensive and lengthy company registration rigmarole. The authors show, for example, that rolling out the programme resulted in more firms being set up and in industries being less concentrated in a few hands.

The especially lucky part of the exercise was that the new "On The Spot Firm" initiative got rolled out in different places at different times. And that enabled the researchers to measure the difference in outcomes between places that had already rolled it out and those that hadn't yet. It also helped that Portugal has a huge linked employer-employee dataset (like the one in our own Integrated Data Infrastructure) so they could analyse what went on in very considerable detail.

Key results (bits in square brackets are my explanatory glosses):

We find the entry deregulation reduced the gender pay gap for medium- and high-skilled workers in affected municipalities [i.e. where On The Spot Firm had been rolled out]. The differential effect of the reform on women workers' pay is positive and statistically significant. Our estimates imply that for workers in high-skill jobs, while male wages increased by 1.5 percent, females' increased by 2.9% as a result of the deregulation, thus reducing the gender pay gap [by 1.4%]. Overall pay of medium-skilled males decreased by 0.6% in treatment municipalities [those with On The Spot], while those of females in the same skill category increased by 0.4% [a 1.0% narrowing of the gap] (p20)

We also find that the share of females in managerial positions increased following the deregulation, suggesting that discriminating employers kept women in lower positions than implied by their skills, and competition induced them to upgrade their occupational status (p20)

Deregulation and increased business competition wasn't a complete panacea:

these effects are not found for those in the lowest skill jobs or for CEOs. The labour market for the top executive still especially favours men and increased competition does not improve relative women's pay [in the CEO role] (p20)

But even so I was impressed by the size of the effect, given that these new firms would hardly be expected to be total giant-killers day one when they first started to go up against longer-established incumbents. Just making it easier to set up a business closed the wage gap by 1.0-1.4% for medium and high skilled women, and got them promotions they should have had before. That's a remarkably big pay-off.

Increased competition in the business marketplace is one of those ideas that neatly hit both equity and efficiency objectives. The equity outcome is obvious: the efficiency payback is that output rises as the previously underutilised female workforce gets to pull its proper weight. And what's true of women is very likely also true of other groups that would otherwise face relatively uphill going in the labour market.

People concerned with some of today's big social or environmental issues tend to be wary of markets, and rather inclined to see them as part of (or all of) the problem. What these New Zealand and Portuguese results show is that vigorously competitive markets can deliver more desirable social outcomes - provided firms aren't allowed the sheltered luxury of bias and slack.

Selective unemployment (yet again)

In response to my post yesterday about unemployment by educational level, a reader who knows their way around Stats' Infoshare database better than I do pointed me to a now discontinued series (it stopped at March '16). Thanks for that, mate: as the series goes all the way back to March '86, we can see how unemployment behaved through a couple of recessions. Here's the answer.

And it is indeed exactly as you might have surmised. In the worst episode - the combination of the costs of Rogernomic restructuring, the very tight monetary policy of the early days of the Reserve Bank Act, and the overseas 'Anglo-Saxon recession' of 1990-91 - everyone got battered, but those with no qualifications got battered most, with their unemployment rate peaking around 17%.

Conversely the good times of the pre-GFC 2000s rolled long enough to bring unemployment rates even for the less qualified down very significantly. By 2007-08 the unemployment rate for people leaving school with a qualification had got down to only 4% - lower than our overall unemployment rate today (4.4% in March). Sustained expansions do wonders for getting even the harder-to-place people into jobs.

The policy lessons stand. Sometimes - if you've let inflation get out of hand, if you've run big fiscal deficits for too long - you may find yourself in austerity mode. Best not go there in the first place, of course, but if you have, you'd better do something to alleviate the impact on the more marginalised groups in society, because they end up wearing the worst of the downturn. More positively, if you can contrive to keep an expansion going long enough, a good deal of social angst goes away as a progressively tighter labour market puts pay packets into far more people's hands.

Unemployment strikes selectively (revisited)

A reader looked at the cyclical pattern of unemployment by ethnicity which I posted about the other day ('Unemployment doesn't strike evenly') - Maori and Pacific people fare unusually badly in downturns and require a long spell of labour market strength before their unemployment rates get back down closer to those for European and and Asian people - and asked me if the same pattern applied by education level.

You'd think the same pattern would apply for less qualified people compared to more qualified people, but off the top of my head I couldn't recall whether our labour market data included educational qualifications. The good news is yes, they do, but the bad news is that they don't go very far back in time (they start in the middle of 2013).

Here's what's available. To keep it manageable I've just picked out three levels of qualification: post-grad, the better end of a secondary school qualification, and no qualification at all. If you're interested in more detail, head for Infoshare and have a fossick: go to 'Work, income and spending', then 'Household Labour Force Survey', and then 'Labour force status by highest qualification'.

Unfortunately there isn't a big enough cycle going on over this time period to see what happens to the less qualified in recessions: all we know (which anyone would have guessed beforehand) is that those with the highest qualifications have the lowest unemployment rate.

In the US, though, we can see a longer picture of cyclical history: here's what's happened to those at the top of the educational ladder (adults with a PhD) and those at the bottom (adults with less than one year of high school). The shaded areas are recessions. Again you can get more detail for yourself from the excellent (and free) FRED database.

You do indeed get the same pattern happening as for ethnicity. Those who find it hardest to get work in good times also get hit far worse in bad times, but if the labour market stays strong enough for long enough, even those with no formal qualifications at all will start finding jobs. Remarkably, the unemployment rate in the US for those with no qualifications is now down to under 5% - but it's taken the longest peacetime expansion on record to get it down to those levels.

Policy lesson: no matter how you cut it, the groups with less going for them suffer disproportionately when the economy turns down. 

MBIE's slick new collection of data

I'd headed over to MBIE's website to see if that inquiry into petrol stations had come out yet - it hadn't - but by happy accident I found that MBIE had come out with something else, its new Labour Market Dashboard. The announcement is here and the thing itself here.

It's a pretty impressive effort at gathering and displaying a variety of labour market data into the one spot. Here's just one interesting graph as a sampler: it shows the different ways companies recruit (there's another one showing it by firm size rather than by industry). It's in the 'Workplace' section, below the health and safety graphs.

Isn't it fascinating? Top of the list is 'Word of mouth', which in a small, informal, high-trust economy doesn't surprise me at all. And as in so many other areas, the internet has changed everything - TradeMe and Seek are now more commonly used than the traditional print job ads. Plus there's a useful self-help lesson here too: 'Candidate approaching us' is the fourth most common way jobs get filled.

I've wondered for a while whether we're any good at 'active labour market policies', programmes designed to make the labour market match up people better, usually with a focus on getting unemployed people back into the game. I'm leaning towards the view that we aren't, and the low prevalence of jobs sourced through Work & Income rather points that way. Not that other potential allies in the fight show up much better: on this showing, there are few school or university offices getting on the blower to employers and saying, "Listen, I've got this student who'd be just the right person for you".

I don't know who did the donkey work on the software, but it's pretty slick. I especially liked the automatic rescaling of the Y axis when you replace one X variable with another, which might be small beer to you pro dataviz types but wowed me (if there's a way to do it in Excel, I've not found it). And yes, whoever the uncredited developer is, I did notice your 'mbienz.shinyapps' heading for the site.

MBIE is looking for feedback. I've suggested the site could carry the unionisation data that the HLFS is now picking up, and I've just also had the thought it would be interesting to see the distribution of people on the minimum wage. You'll have your own ideas: why not help out this useful source and get in touch, the e-mail address is

LabourMarketDashboard@mbie.govt.nz

Just  noticed that this is the second nice thing I've written about MBIE this week. We'll be picking out curtains next.

An "ugly casserole" indeed

So we've tightened up our immigration settings, again. Opponents are saying it's fiddling at the margins. I hope the critics are right: while it's dismaying that they've felt the need to bend with the prevailing political winds at all - "an ugly casserole of prejudice, resentment, economic envy and xenophobia from which New Zealand is not exempt", as Vernon Small put it in his īnsightful Dom-Post article - I hope it's the absolute minimum the government could get away with.

Not that it's without its own political dangers.Throwing scraps of meat out of the sleigh as the wolves close in may delay the chase, but wolves soon learn to chase sleighs for reward, and will be after you fitter and faster than before.

And I certainly didn't like the "Kiwis first" messaging around the thing. That's more the sort of jingoism you expect from the Aussies, and indeed got this week in their own tightening of their immigration regime, which more obviously pandered to the likes of One Nation and is more likely to do real economic damage. Early reactions suggest they've shot themselves in the foot, making it harder for employers to fill hard-to-recruit vacancies.

Let's not pretend, either, that there's anything remotely linked to the economy behind this. It's pure politics.

For one thing, the cyclical state of the labour market is very strong. Wanna see a picture of a thriving labour market? Here's one from the ANZ's Bank's latest tot of job ads.

Wanna see another? This is MBIE's latest tot of vacancies - jobs available right now. High, and climbing steadily,

If you're concerned that unskilled emigrants are driving down wages and stealing the everyday battlers' jobs at the hard end of the labour market, the greatest rise in vacancies in recent years has actually been for unskilled and semi-skilled jobs.

So it's extraordinarily hard to make any case that immigrants have soaked up the available jobs. Despite a recent large increase in net migration, there are more vacancies available than ever, even at the low wage end.

As for the fable that immigrants are putting pressure on already stressed infrastructure, it's either a marginal effect or complete bollocks. Higher levels of net immigration are a relatively recent phenomenon - the last three years or so - whereas the infrastructure, especially in Auckland, was already stuffed every way to Sunday well before that. 

Here's the clincher: as an economist colleague helpfully pointed out when I was tweeting about this, there was very little difference between Statistics New Zealand's population projections for Auckland's population ten years ago and what has actually happened. We knew there was a surge coming, and didn't plan for it, and still aren't - that's the reality, under national and local administrations of all political hues. It's got very little if anything to do with migration, and everything to do with systematic underfunding of infrastructure and obstructive land-use and other regulation. Our politicians have preferred to run dairy farms, mine coal, deliver letters and keep prime housing land to grow onions, rather than deliver the social and physical infrastructure for a high-income economy.

And that's something else the anti-immigration crew are missing. Far from being a burden on the economy, many migrants are actually alleviating the roadblocks to growth we've inflicted on ourselves. Across the road from us, one of those terrible Asian property investors - probably got one of those suspicious Chinese-sounding surnames - has started to turn two houses into five. Round the corner another of these awful leeches has already redeveloped one house into three. Employment for two gangs of tradies, five extra houses. Not bad for two parasites.

Maybe I shouldn't be surprised about these latest turns of events here and in Australia. Liberal values and sensible economics are under attack pretty much everywhere, from the clowns in Washington to the Brexit-deluded in the UK through to the more sinister operators in (for example) France, Hungary and Poland. It's especially sad to see people "of the left", who might be expected to take a more progressive, internationalist view of the world, going along. 

But maybe there's hope around the corner. As this excellent article, 'In defence of liberalism', points out, the liberal values and better economic policies that created post World War Two prosperity are certainly under attack. But it concludes

the things we value can’t all be realised simultaneously and made compatible with each other. A liberal society is the best method yet devised of recognising this multiplicity of aims. It stresses value pluralism in the face of political and religious dogmatism, and of spurious appeals to national unity for the common good. I’ll doggedly stick to it.

So will I, even though in the short run the probability of a socially liberal, economically responsible government emerging from our next election looks to be half of five eighths of sod all.

Jobs and spin

Politicians everywhere spin the economic data, and the labour market data in particular. So it's no surprise that the Trump administration does it, too, nor that the Trump numpties do it with more more ineptitude and hypocrisy than most. Having claimed on the election trail that the official American employment and unemployment numbers were variously either bad measures or outright faked, the Trump team is now claiming the latest official data (a big rise in jobs and a fall in the unemployment rate in February) are mysteriously unfaked after all, and the right things to focus on.

Plus they've dissed a Federal regulation saying that Federal officials can't comment on official data until an hour after their release - a quid pro quo for the dubious practice of giving the President (via the Chair of the Council of Economic Advisers) advance access to the numbers, since it would be all too easy otherwise for incumbent Presidents to use that access to screw the first impression scrum.

We haven't followed the Americans down that path to privileged access for the executive, and good job, too. We do have pre-release 'lock ups' for the media for important statistics, which I see as a useful evening-the-playing-field device to ensure that independent, informed comment goes out at least as quickly as Executive spin. From that perspective it's also a shame that some thoughtless nerk led the Reserve Bank to cancel its lock ups for the Monetary Policy Agreements: not everyone agrees, but the Governor's take having the airwaves to itself for an uncontested hour is a step backwards, compared to the previous arrangements where journalists had a way to put out their own considered view at the same time.

Not that we're always saints and the Americans always sinners: if, for some great sin in a previous life, you listen to our Parliamentary Question Time long enough, you'll find the same smart-alecking going on about our own employment and unemployment numbers, though not to the shameless Trump standard. Some Minister will point to the low official unemployment rate, and some Opposition member will say "Aha! But if you look at the underemployment rate...". And they will of course swap positions on the data when they swap roles.

So as a bit of an air-clearer, here's how we've been travelling, based on our new and improved Household Labour Force Survey.

The graph shows three ways of looking at unemployment. The green line is the headline, official, unemployment rate. The orange line adds in people who are underemployed - people working part-time but who are available to work full-time, and would like to. And the red line adds in what Stats calls the 'potentially available' workforce, which is made up of two groups: people who aren't actually looking for a job right now but would be available and willing to take one on if it cropped up, and people who are looking and willing and will be able to take up a job in the coming month, but not right now.

All these are valid ways of looking at unemployment and underemployment, and depending on what you're most interested in, you'll follow one rather than another. But what's also very clear is that the trend is almost exactly the same across all three series. So next time you hear a politician saying, "Well, the official unemployment rate may be getting better, but the underemployment rate isn't", or some variation thereof, you'll know that you've got a gasbag talking his partisan position, just like the Yanks have been doing.

The other thing that occurred to me as I trawled my way through the data is that I'm left unsure about the solidity of estimates of the 'natural' rate of unemployment. That's the rate when the labour market has become so tight that employer competition for scarce staff starts to bid up pay rates and eventually the national rate of inflation.

There are sophisticated ways of measuring it: the latest I'm aware of is Weshah Razzack's Treasury Working Paper 'New Zealand Labour Market Dynamics: Pre- and Post-global Financial Crisis', which estimated the natural rate in recent years to be around 4.5%. But I wonder. Let's suppose that the official unemployment rate did indeed start going below 4.5%: is it really likely that wages/inflation would start accelerating at that point, when there are currently 115,000 people who are working part-time and would like to work longer hours? A more likely initial response is more employers wanting, and more employees agreeing, full-time hours.

Or another way of putting it is that I suspect the sky would not fall if our unemployment rate went into the low 4s or even below. Sure, institutional arrangements vary, and you can't always (or even often) say that what works in one country will work everywhere else. But there are countries, as you can see in the table below, which have got their unemployment rates down to lower levels than ours, without inflation starting to roar away (or even raise its voice much).

Don't be too 'asty - update

On Monday I wondered about the sense of the government tightening up on migration inflows at a time when businesses appeared to be finding it hard to get the people they need to keep growing.

Since then, while researching something else, I've discovered that there's even more evidence saying that the government ought to go carefully. It's the latest ANZ jobs ads count. As the chart below shows, the unemployment rate has been steadily falling while job ads have been steadily rising: you couldn't find a clearer illustration of how short businesses are of staff, and of the need for more people from overseas to keep the cycle going.

And you don't have to take my bleeding-heart-liberal instinctively-pro-free-movement-of-people word for it. Here's the ANZ's own take:

With firms in a mood to expand, skill shortages, rather than demand, are set to become the key brake on activity.

Those damn Chinese, keeping our economy going for us. Honestly.

I've also sat through today's Parliamentary Question Time. Sometimes it's moronic: patsy questions from government backbench MPs to their own side, or, even more egregiously, as we had today, patsy questions from one backbench opposition MP to another. Sometimes it's cynically oppositional: you know full well that the Opposition MP asking the question would, if in office, have done exactly what he is criticising the government for doing.

Today hit new lows, with ugly questions about immigrants supposedly taking jobs from our own youngsters. Never mind that the questions are hopelessly adrift of the true state of the labour market, where (as the ANZ data show) any sentient carbon based life form can get a job if it wants. But it's also ratcheted the immigration debate another moral notch lower. I didn't think, when I first came to New Zealand, that I'd ever see its elected representatives channelling Farage, Le Pen, or Trump.

In a race to the bottom, we've got some good runners.

Don't be too 'asty

The government has tapped the brakes on migration flows, for what commentators have assessed as essentially political reasons (see, for example, Duncan Garner on Stuff or Fran O'Sullivan in the Herald). 

But as Fran said, "The balancing act that Key has to apply is how he gets across the fact that New Zealand is dependent on immigration and ensures the doors to skilled people remain open". And that's getting a bit trickier, as some data today suggested we may be more imminently dependent on ongoing strong immigration flows than you might imagine.

The data were in the latest BusinessNZ/BNZ Performance of Services index, which showed the services still growing at a handy pace, though a tad less strongly than in August. In its commentary, the BNZ has flagged the possibility that difficulty in finding staff could be starting to hold firms back from continuing  to expand. In the graphs below, the bank has juxtaposed the composite index covering both services and manufacturing (the 'PCI') with GDP growth - there's usually a close fit - and said there's "the hint, from the PCI, that growth may be slowing". And below that, it put up the latest readings from the NZIER's Quarterly Survey of Business Opinion, which as the NZIER said, showed that "Firms report increased difficulty in finding labour, and this may have limited the extent to which firms could increase headcount over the past quarter. The difficulty in finding labour is particularly acute for skilled labour, with shortages at levels not seen since December 2007".

Sure, migration is a touchy issue in a lot of places right now, and one's that's backing generally economically responsible governments into places they wouldn't normally go: in the UK, for example, the Conservative government recently flirted with the tackily unpleasant idea of "naming and shaming" British companies that hire people from overseas (they've since had to backtrack). 

All I'd observe (as I've said before) is that the labour market is rather tighter than you would think from a 5.1% unemployment rate. Mucking about with the supply of staff from overseas might make political sense. But if it threatens to get in the way of the already limited ability of businesses to find the people they want, it's not likely to be economically costless.

Unions and collective agreements

Our revamped Household Labour Force Survey now collects information on union membership and on types of employment agreement, which sounded interesting, but when I went and downloaded the HLFS Excel files (here) the info wasn't there. So I got in touch with Stats and their helpful chap Ken Joe told me two things.

One was that Stats had put out a specific article about these very topics on August 25 (which for some reason I'd missed, and I may not be alone, as there doesn't seem to have been much media coverage of it). And the other was that the data is available, though not where I went searching for it first. If you'd like to look it up for yourself, it's on Infoshare: go to 'Work, income and spending', then 'Household Labour Force Survey' and scroll down to the P's where you'll find 'Paid employees by type of employment agreement' and 'Paid employees by union membership'.

The article said that "Statistics NZ collected this information only twice previously, in the Survey of Working Life 2008 and 2012": they didn't compare the new HLFS data for June 2016 with the previous years' estimates, however, so I've dug out the earlier data. If you missed the Stats release (like I did), and would like the wider historical sweep (such as it is), here is the result.

It's not a pretty picture if you're sympathetic to organised labour. The absolute numbers probably show it even more clearly: in 2008, an estimated 525,000 people were members of a union; by 2012, that had dropped a bit, to 501,000; but it's slumped in the past four years, to 379,000. The drop in numbers of people on collective contracts isn't as dramatic (467,000 to 440,000 to 410,000) but it's still substantial. If the declines continue at these rates in coming years, unions face an existential threat.

Especially if anything arises to weaken membership in the sectors where they are still important. As Stats said, "Some industries are far more unionised than others (see figure 2). Over 4 in 10 employees in both health care and social assistance (43.5 percent) and education and training (42.2 percent) belonged to a union. Given the size of these industries (231,100 and 203,500, respectively), this means half (49.2 percent) of the 379,300 union members worked in only two industries".

Here's that figure 2 Stats was talking about.

Outside what are largely public sector industries, union membership has fallen to insignificant levels - even in sectors such as finance, where unions were once a reasonably large player, but are now down to 10% of the workforce in the sector.

Originally I'd set out to look up some data releases I'd unaccountably missed (I'm on Stats' distribution lists for pretty much everything they put out), and not to do any policy research into the role of unions in a modern economy, and I'm not going to start now. All I'd observe is that the sharp membership fall of the past few years is a lot faster than I intuitively (econospeak for "at a total guess") would have expected to see, and I'd guess there are some worried people around the tables in unions' head offices.