Getting real

We don't have an adequate grasp on the real-time state of the economy. We need much more close-to-real-time data.

That's it. The rest of this post is a riff on the same theme.

As covid wreaks its damage, it would be good to know, now, what that damage is: what's the initial hit to GDP? Where are the worst impacts? How are the knock-on ripple effects going? Among many other things, it would inform how hard fiscal policy needs to fight back. As it is, we - and many other countries - are at the "Oops, did I say $10 billion? I meant $40 billion" stage of Finance Ministers' groping for what the fiscal response needs to be.

Crises apart, we should always have had better high frequency cyclical data than we've actually had. As Michael Reddell said recently in the 'Measuring the slump' post on his Croaking Cassandra blog, "We and Australia are the only two OECD countries without a monthly CPI, our GDP estimates (quarterly only, as with most countries) come out only with a very long lag, we don’t have a monthly industrial production series, and we still don’t have an income-based measure of GDP". These deficiencies have been known for a long time - I can remember discussions over the years at Stats' Advisory Committee on Economic Statistics (since disbanded) - but the debate never got as far as prising open Treasury's chequebook.

Michael had several constructive ideas on how to improve things, including having Stats publish the monthly results from its rolling quarterly Household Labour Force survey, and having Stats also "look at hosting some sort of dashboard pulling together, and making openly available, all manner of formal and informal economic indicators. There have to be lots".

The good news is that at least two parties (though not Stats, yet) have had a go at dashboards.

First out of the blocks was Sense Partners: you can access their latest six-variable dashboard here. It's good, isn't it? I especially liked the electricity generation graph (below). You'd think that it must be reasonably close to the fall in GDP, and is suggesting something like a drop of close to 20% in output since late March (though there may be seasonal stuff going on, too).

They were closely followed by Treasury, who've just put out theirs (media release here, access to the pdf dashboard here). Personally I felt the Sense set gave me more of a real-time feel, though Treasury's information was interesting in its own right, especially the traffic data and the data on uptake of the job subsidy scheme (below).

The next thing that would be really useful would be some composite indicator of all these glimpses of the overall underlying reality. As it happens economists have a nifty way of devising one: it's called 'principal components', and works by assembling a lot of data, hopefully all related (positively or negatively) to some underlying common influence, and then analysing it econometrically to see if you can identify a common background factor.

Treasury's dashboard helpfully included some international data, and one of the series shown was, indeed, one of those composite indicators, for the US. It looks like this.

If you want to follow the series yourself it's here at the Fed of New York site, and there is a (short) write-up of how it all works here. The devisers of the index have calibrated it so that percentage changes in the index can be (give or take) read as percentage changes in GDP, or as they put it, "a reading of 2 percent in a given week means that if the week’s conditions persisted for an entire quarter, we would expect, on average, 2 percent [GDP] growth relative to a year previous". So we know that if early-April conditions persisted for the whole of the June quarter, US GDP would be some 9% lower.

Can we be sure that the weekly economic index does in fact track US GDP pretty well? Yes, we can, as you can see below (this is from Chapter 1, 'US economic activity during the early weeks of the SARS-Cov-2 outbreak', in Issue 6 of the Centre for Economic Policy Research's covid economics series, well worth following).

Two things to finish with.

One, obvs, wouldn't it be nice if someone did the same number crunching on New Zealand data and gave us close to a real-time reading on where GDP has got to? And yes, I'm happy to be part of the effort if anyone feels the urge to get it done.

Two, the wider point about the limited range of our current official cyclical data, and the speed with which what we have gets published, needs to be addressed. You can't go to a competition or regulation conference these days without people blathering on about how 'big data' enables market power, or threatens privacy, but you don't see anything like the same focus on how the torrents of big data could be used to generate near-real-time cyclical gauges.

It's fine to have the business as usual, industrial strength, quality-assured things like the quarterly national accounts. But as Grant Robertson and the lads at Treasury - and the rest of us - are finding out, there's a big role for the cheap and cheerful but timely and informative indicators, too. We've shelled out some $9 billion on wage support: in the greater scheme of things a couple of mill to zero in on where we actually are would be money very, very well spent.

Workshop takeaways

Last weekend was the latest annual workshop of the Competition Law and Policy Institute of New Zealand (CLPINZ). Wellington turned on its loveliest weather, all the speakers and discussants performed well, there were sausage rolls at the meal breaks, and during my absence in Wellington the Warriors broke their home game hoodoo and took the two points. An excellent week-end all round, and special thanks to Chapman Tripp for providing the excellent facilities.

Many of the usual suspects were at the workshop, but if you're in the game and had to miss it, head over to the CLPINZ website, sign up as a member, and you'll get access to the papers (they're not up yet, but will be). You'll also get access to previous years' presentations.

The papers speak for themselves: here are some personal thoughts I took away.

Cartels - the cause célèbre du jour is the Lodge case. This is the one where a whole bunch of real estate agencies pleaded guilty to price-fixing, but for the eponymous hold-out in Hamilton, who went to trial and to widespread surprise beat the rap in the High Court. At the time I wrote about the fines for those who had pleaded guilty and said that "in the context of small to medium provincial businesses, even to my unsympathetic eye they were looking down the severe end". Having listened to the cartels session, and to the session on coming to a negotiated settlement with the Commission, I'm more of that view now. I never thought I'd feel that way, but as David Blacktop's presentation said, there can often be some "precipitating event" that causes otherwise well-meaning, normally competitive businesses to lapse into a concerted (rather than independent) response to the event - in this case to Trade Me's attempted jack-up of real estate listing fees. No, they shouldn't collude on a response, and in principle it doesn't matter that they would have come to the same decisions independently, but all the same they got backed into a corner and were anything but the cartoon cartelists in the proverbial smoke-filled room. Time for more understanding, in my view, of the reality they found themselves in. The goss, by the way, is that the Commission stands a good chance of having Lodge overturned in the Court of Appeal, but stranger things have happened. And while we're on the topic...

The law - okay, I'm an economist, and anyone who gets their legal advice from an economist deserves what happens to them, but I've got some questions. Is there anyone on the look-out for where Australian and New Zealand competition law might be diverging? We don't want, for example, the situation John Land described, where the legal approaches to price fixing may be going down different roads. Is there any kind of trans-Tasman body that keeps a weather eye out and acts to harmonise on best practice? And speaking of harmonising on best practice, it seems from what Minister Faafoi said at the workshop dinner that reform of s36 - anti-competitive abuse of market power - will be back on the agenda next year. Good: the Aussies have fixed their equivalent, and we should get in behind. I've also got some disquiet (partly stemming from Lodge but also more generally) whether behaviour at the lower end of culpability will be prosecuted under the forthcoming criminalisation regime, rather than the 'hard core' cartels that should be its target.

Regulation - I found myself in strong agreement with Ross Patterson's response to Sasha Daniel's paper on 'The future shape of telecommunications regulation'. Ross argued that there is no lack of competition in access to fast broadband and hence no case for regulation, especially when you're regulating one technology (fibre) but not another (fixed wireless) and with a - I think he said 'clunky', but if he didn't I am - a clunky form of 'building block' price cap regulation.

The internet - our new digital economy is going to be a minefield from both competition and consumer law perspectives, and I suspect we'll be making both Type 1 and Type 2 errors for some time before we get it right, if we ever do. That was my overall impression from the keynote 'Collusion without the smoke-filled room: from public statements by wetware to algorithmic pricing by software' from Professor Joseph Harrington and 'Consumer Analytica: NZ consumer law application to international developments in privacy and use of data' from Sarah Keene. I suspect there's likely to be behaviour that is anticompetitive or unfair/misleading that will not be pinged, and behaviour that's legit that risks being rapped. Joe Harrington is surely right that we likely need jurisprudence and new guidelines to distinguish between the two, but we're still a long way from being able (for example) to "develop rules for how a platform can intervene in the setting of prices" or to "define the class of prohibited pricing algorithms".

Market studies - the papers presented at the session I chaired were absolutely on the money. If you're thinking about how the Commission should use the powers it's (more than likely) going to get, you've got to read the excellent presentations from Mike Tilley and Richard Meade. They've both had first-hand experience of doing these studies, and it showed. Market studies are a great idea, but there are more process issues to think through than you (or I) might have imagined. I'll just chuck in one final thought for MBIE's consideration: I suggest that any company that attempts to invoke our 'anti-dumping' provisions should automatically trigger a market study into its industry.

Quantification - James Mellsop and his NERA colleague Kevin Counsell gave a very good presentation on 'Mergers: exploring the economic tool box', and walked the attendees through unilateral effects in auction markets (using a pathology merger example), vertical arithmetic (a version of critical loss analysis) using the AT&T/Time Warner example, and then some Cournot modelling of a wool scouring merger (using made-up data, by the way, if anyone involved in any of those cases is wondering). Good stuff, and they got it across in a user-friendly way that - my keyboard nearly inserted 'even' - lawyers could understand. My feeling is that we are, finally, on the brink of a new more data-driven and more quantitative approach to competition analysis, after a long period when the tide had gone out a very long way indeed on playing with the numbers. As I've said a few times before (eg here or here) there's far more empirical data becoming available, and better (and often more robust) ways of interrogating it. The Commission's Reuben Irvine said that some of these quantitative techniques, like the auction and vertical arithmetic tools James and Kevin mentioned, are already in use, if somewhat behind the scenes, at the Commission, and about time, too. In my stint there, applicants and opponents very rarely reached for even the more basic econometric methods (regression, differences-in-differences), and you could go years without tripping over a correlation coefficient. We've become an immensely data-rich world: time to start using it, rather than making anecdotal guesses about (for instance) the degree to which products are or are not in the same market.

If all else fails, crunch the numbers

There's a very interesting article by Jonas Björnerstedt and Frank Verboven in the July issue of the American Economic Journal: Applied Economics, 'Does Merger Simulation Work? Evidence from the Swedish Analgesics Market' (here's a link but to get further than the abstract you'll need to be a member of the American Economic Association or otherwise have access to its journals).

Normally, merger simulation - formally modelling what might happen (particularly to prices) if competitors merge - is done before the event by competition authorities (and also, sometimes, by economics consultancies, generally in support of the merging parties). What the two authors of this article have done, however, is turn the process on its head: instead of asking, before the merger, what post-merger effects does the model predict, they ask, after the merger, what model would have best predicted the effects that actually happened.

Usually, this isn't doable. If a merger is cleared, and it doesn't substantially lessen competition, generally there are no obvious price effects to see. Fortunately for the authors - but rather unfortunately for the Swedish competition authority - they've got data for a clearance that went badly wrong. Prices were jacked up substantially and immediately post-merger, and to some degree by third party producers.

Here are the facts in approximate brief. Analgesics are painkillers: there are three main ones, paracetamol, ibuprofen and aspirin. The only two makers of paracetamol in Sweden applied in late 2008 to merge, and the competition authority allowed it in April 2009. The regulators relied on a broad market definition (the other two painkillers would be good substitutes, constraining any rip-off on paracetamol) and on the prospect of greater competitive constraints following an adventitious deregulation of the pharmacy sector (up to late 2009 there had been a state-owned pharmacy monopoly).

In fact, the price of paracetamol went up by some 40% immediately post-merger, and the aspirin makers cashed in too, with price rises from 7% to 18%. Ibuprofen prices didn't change much (no, I don't know why, either). Here's the graph ('ASA' is aspirin).

Clearly, the painkillers weren't in fact good substitutes for each other (the facts speak for themselves, and when they ran their models the authors also found low cross-price elasticities) and the paracetamol makers were able to coin it. There were lucky, in that an anti-drug-overdose measure was implemented around the same time, and they were required to sell 20-tablet packs instead of 30-tablet ones: they were able to smuggle in higher per-tablet prices by not reducing packet prices proportionately. The authors allow for the fact that it's more expensive to make smaller packs, but even so that would have accounted for only around 15 percentage points of the 40 percentage points price increase, leaving 25% as pure lower-competition gravy.

The authors were also able to go a long way with their main research interest - which models of producer behaviour and consumer demand best fitted the facts (answer: none perfectly, but some pretty impressively).

Okay, that's the specifics of their work, but I think their research also makes some more general points.

First, it suggests that (at least some) merger simulation models are well worth running as part of the clearance process. Our own Commerce Commission used to have a formal model, but it may have been put out to graze: a search today of the website for "merger simulation" or "Bertrand" came up empty, and the five "econometrics" results weren't relevant. This Swedish study, however, suggests there's a dance in the old dame yet. I wouldn't push the argument too far: as Stephen King, at the time one of the ACCC Commissioners, said in 2005 in 'The use of empirical methods in merger investigations'

Because of its complexity and sensitivity to particular assumptions, merger simulation is generally contentious and, at best, provides ‘back up’ input for a more complete merger analysis

That may be true, but it's beginning to look to me like a low-ball estimate of the potential value of merger simulation modelling.

Secondly, I think it says something about the potential use of more econometrics in competition and regulation analysis more generally. So far, it's been a bit of an uphill struggle:

Lawyer: So, your model rests on quite a specific set of assumptions?
Economist: Yes, but...
Lawyer: And if those assumptions do not hold exactly, the results may not be reliable?
Economist: No, but...
Lawyer: And they haven't held exactly, have they?
Economist: No, but...
Lawyer: No further questions of this witness, m'lud.

But we're in a new world of big data where crunching the numbers with better tools is getting easier and more reliable. It's time, I reckon, to fire up more models, more often. As I've said before

we may be getting closer...to being able to do a better job of taking a more robust empirical approach to measuring things like demand curves, and own- and cross-elasticities of demand. If, using things like scanner data, improved econometric methods, sophisticated consumer choice testing, and clever analysis of 'natural experiments' - what happened, say, after a fortuitous interruption to one source of supply - we can get a more scientific handle on the extent to which products are or are not substitutes for each other (and so are or are not likely to be in the same market), why wouldn't we use that information to derive empirically grounded market definition?