Thursday, 5 July 2018

My tuppenceworth

Today I fronted up to the Transport and Infrastructure Select Committee to speak to my submission on the proposed new 'market studies' powers. The Commerce Commission by the way was two slots ahead of me in the queue: you can read its submission here.

Quick gist: I'm all for market studies, and have been for ages as regular readers (both of you) will know. We're the outlier these days in not allowing our competition authority to have a proactive look at potential competition problems, and we should join the international best practice pack. I'm in favour of the Commerce Commission having the power to initiate them off its own bat, as well as Ministers asking the Commission to do them. So I'm on board with the broad provisions of the Commerce Amendment Bill.

I have suggested three small improvements, though. As I posted about Tuesday's LEANZ panel on market studies, I think there's a consensus that the threshold for firing the starting gun on a market study is too low. At a minimum the Minister or the Commission should be required to tell us in more detail why they're bothering. You could add a formal requirement to consult on the terms of reference (though I'm pretty sure the Commission would take at least informal soundings before going live). One MP asked me if the requirement to consult on the terms of reference mightn't tie up the whole process in judicial review till kingdom come, and maybe that's right, but it has to be possible to design something cheap and cheerful and appeal-proof that would kick the tyres properly before going live.

The second improvement was providing for the possibility of a conference on a draft market study. I'm a big fan of the conference process: properly managed it can lead to useful debate and enlightenment. At the moment the Bill provides for submissions on a draft study: maybe implicitly (it's been suggested to me) it already allows for a conference. But in any event I've suggested making it clear.

And the final improvement is that Ministers would be required to respond yay or nay to Commission recommendations in a market study, within a tightish timeframe: I've suggested 60 working days. I don't see the point of spending perhaps $1.5 million a pop on these things only for the study to moulder in a Ministerial cupboard.

I got some questions. One was related to the Commissioners being both decision-makers and the board of directors: the logic was that there aren't many checks and balances on what they can get up to, and maybe it might be better to leave the instigating power with Ministers, who might face greater accountability constraints. I didn't see that as much of a risk. The Commission's governance arrangements are a bog standard way of running competition authorities, and if you look across the ditch at the ACCC, which has the same set-up, you don't see any difference between the kinds of studies the ACCC starts rolling and those that Ministers do.

I got an interesting question about whether the Aussie market studies had made any difference at the end of the day. Mostly yes, was my answer, for example in improving the bargaining position of small suppliers dealing with the market power of big buyers (eg dairy farmers at the bottom of a supply chain totem pole with the supermarket duopoly at the top and the reasonably concentrated dairy processing factories in the middle). But not always: the ACCC, for example, hasn't been happy that its ideas for getting beef farmers a better deal at the stockyard sales haven't got much traction. And in some areas (eg the ACCC's ideas on buyers of new cars getting more choice of repairers) it's too early to tell.

Bottom line, a scheme exactly like that proposed in the Bill has been running with no problems and some achievements across the ditch, and it's time we did the same. End of.


  1. First of all: thank you for this blog! Have been reading for a while and think you have a real knack for explaining stuff :)

    And here's a question: I was listening to a keynote from NZ Institute of Quantity Surveyors conference,, and in it Richard Joseph was saying how New Zealand's productivity per capita has been going down and down and down. Around a 4:00 mark.

    And I'm wondering: is the stuff he's saying correct? You've said before that depending on what graphs and statistics a person uses, they can pretty much "prove" anything. Sorry, it's not your exact wording, but I'm trying to explain the meaning :). That someone can use the data most "suitable" with their agenda and go with that.

    Richard said that "We've never worked longer hours or produced less" in New Zealand. And I've kept wondering today if I could ask what you think about this - and if it's true, what do you think causes it?

    I was thinking about the building sector and how, in the circumstance of skill shortage, someone may do longer hours because there just isn't someone else to do the job. And then gradually the whole industry moves to longer hours until builders think it's "normal" to do 48-hour weeks. And that's not even "overtime"! Because some work Saturdays on top of that. And I can see that their productivity goes down because... you can't really be a really good builder for 48 hours each week.

    But New Zealand as a whole?

    What do you think causes that? From an economical perspective? Or is it just a set of data-points that are very selective about what they measure?

    Thank you!

    1. Thanks for the comment Maria and apologies for slow response, I've only just picked it up.

      Richard Joseph hasn't quite got it right. The OECD data are correct but it's not true that "we've never produced less". GDP is going up, so we are producing more. If he meant on a per capita basis, which he likely did, he'd still not be right, as GDP per capita is also going up though slowly (only +0.6% in year to March '18). So the reality is that we are producing more, in aggregate and per person, not less.

      Are we working more hours than ever? Because the labour force is growing, that's true. But he presumably means "per person", and that's debatable. Stats' Infoshare is down at the moment, but on the data available on the main site back to March '16, there's not a lot of a trend. People work 32 to 32.4 hours a week on average or thereabout, there's not an obvious rising trend, but I'll have a longer historical look when I can get at the data.

      If you're interested in the productivity story, and it's important, a good place to start is the Productivity Commission, esp the speech referenced here

      Both Paul Conway and Patrick Nolan at the Commission are very approachable if you've got any questions.

    2. Infoshare is back up again and I've had a look at the longer-term trend in hours worked per person.

      Those numbers I quoted you earlier are not quite right (I think I mistakenly counted the unemployed in the calculation): the right number is total hours worked, divided by number of people employed in the labour force, which I've looked at going back to 1986 which is where the Household Labour Force data starts.

      The answer is, no, we're not working more hours than ever. There is no clear trend at all in the data towards higher numbers. And for the record the most recent average hours worked (36.2) have been exceeded in the strong growth periods of 1993-96 and again in the early 2000s. In strong growth cycles hours worked, like we've also been having at least until this year, unsurprisingly go up. But there's no obvious trend towards permanently higher hours.