Transcript of Question & Answer Session Uncertainty

Right thank you, Guy. In keeping with the theme, one of the things my iPhone tells me, which I wasn't able to do a few years ago, was predict the weather. So, it seems we have a storm heading our way in the next 15, 20 minutes. So I think we'll take many 10 minutes or so of questions, if that's okay. And while we had planned to bring you down, it's a bit of a steep lecture theatre, so we'll bring a mic to you. So we have a question here, and then we have Marty here at the front. One, two and a third by …

Hi, I was at Sydney University, economics graduate, 1987. Very significant time in the world with one of the great stock market falls. Thank you for your great lecture on forecasting uncertainty. But I'd like to ask a practical question relating to 2007/2008. Obviously, a global financial crisis caused by excessive debt, government stepped in to save banks and then they couldn't handle it and central banks stepped into save governments. And all we saw happen was quantitative easing, to the point where from 2008 to 2017, total quantum of debt increased by $58 trillion. Let alone derivative debt, that's unquantifiable and all we see that's happened right now is a huge disjoint in the world between rich and poor. To a point where we're seeing asset bubbles in both the stock market and property markets and therefore leading to political instability due to this big disjoint.

The question I want to ask is, given this total amount of debt, it appears that everything is fine. Unemployment seems to be great. Why wouldn't it be with the monetary expansion? Inflation seems to be under control because it seems again, that no one's spending any money because they're worried about their employment. So the question is, is there a day of reckoning? Is there a doomsday to come? Or am I just imagining this? Or we just keep on expanding debt. Now, in business I understand, no problem with expanding debt, provided at some point, income is increasing faster than the rate of debt.

I suppose I wouldn't quite subscribe to all those degrees of causation you said on the way through. But, I mean one thing, one point to note as you said, unemployment is low and it's nice that you attributed that to monetary policy. But I think it's important to note that the alternative is that you don't have a policy response and then you'd have a hell of a lot worse outcomes in terms of unemployment and that's a hell of a lot more damaging to income distribution than what we've seen subsequently.

So the first order impact on inequality, on income distribution is whether or not people have a job. That is a much more material impact on that than almost anything else and so the fact that unemployment in a lot of the major economies is, as I was saying earlier, about as low as it's been for a few decades, is something which is going in the opposite, is something which has a first order effect on income distribution. Yes, some of the other things you're talking about go in the opposite direction. That's absolutely true. But one thing that has a first order impact is whether or not people actually have a job or not.

You can debate about the efficacy of the various policy actions and that's a debate which is certainly worth having and has been going on for quite some time. But in the end, if you look at where things are today, in most of the major developed economies as I said, they've got unemployment rates about as low as they've been for a long time. That's not a bad outcome. You can debate the issues around inequality, I don't think I'd completely attribute them all to the policy response in 2007-08. There are a few other things going on as well. Both in terms of say, technology and technological change and the differential impact that's been having. So I think that's a much more nuanced question.

But, as I said, you can debate the degrees about the policy response but in the end, some of the proof of the pudding has been in the eating and there have been decent macroeconomic outcomes in a number of economies and the counterfactual, I think, is unambiguously a lot worse than what we've actually seen.

One of the things you said Guy, was that it would be nice to have monthly inflation figures as many countries around the world do. As we move forward into the era of the Internet of Things and increasingly big data, I'd be interested to know whether … our problem may be that we have too much data and we don't have good tools for analysing it. So I'd be really interested to hear your view or the central banking view at the moment on how you think this might be incorporated usefully. And I accept that academics like myself still have a lot of work to do in working out the tools that might make these useful. But what is the central banking view on whether you think this is going to be useful information for you?

I think I sort of alluded to it, too. There's a trade-off between signal and noise, right? So, more and more timely data is not always and everywhere unambiguously good. So there is that trade-off. As you say, you can have more information but it's not much good if you can't actually process it, for one. That said, I don't think we're quite in the right point of that trade off where we are currently with the information we have. So I do think we can shift to a better point by, for instance, utilising monthly CPI and it partly comes back to something I said before: we don't just want better information just because we can have it. There's a question of is it actually material for our decision making I think is another way of looking at it. Being economists, we're well aware of the universality of budget constraints. The ABS, like everyone else, is operating under a budget constraint and they've got to make trade-offs and you've got to work out what is the usefulness of the information which could be obtained versus the cost of obtaining it.

Now, as you alluded to, we're both seeing an increase in the volume of data but we're also seeing an increase … there's a technological race in terms of the tools to be able to process that data. That's always been, to some extent, always been thus. I would say that we look at a hell of lot more series now than we would have even two decades ago, partly because we can. So both they exist, but also we've got a greater capacity to process them; to try and extract that signal. And again, we're operating under a budget constraint. We've got only so many analysts we can throw at analysing the economy. What's the most useful way of analysing the data that we have? There's any amount of data that's out there, we could be out there collecting all the data from all the scanners at the supermarkets, is that actually useful? The answer may well be yes, because now we may well have the tools to be able to process that in a way that just absolutely wasn't true a few decades ago.

So, I do, I suppose I would frame it in that way, that we're operating under this budget constraint but with an objective function which is what is the usefulness we can extract from this in terms of the signal or is it just going to give us … is the signal noise ratio too low in this respect, that we're really just mostly getting noise and not a hell of a lot of signal? I suppose the point I was making, at least on the CPI, I don't think we're quite at the right point of that trade off, where we are currently.

My question is about inflation. I think you talked about tenth of a point not mattering here or there and that's definitely true. On average, you said with the re-weighting, I got to the methodology of it, the re-weighting of the CPI in the past. If also you think that maybe these global things that we haven't got a good feel for in terms of downward pressure on inflation, are we picking up Amazon? Are we picking up things like that in our inflation measures, probably not. So maybe even if inflation is even lower than that quarter of a point possibly over the last five years. So when we troughed, I think it something like one and a half for underlying inflation. We're now at a bit below two. Do you think now that actually matters more?

Your predecessors and you have talked about getting a feel for the starting point of the economy, for the outlook for monetary policy, just how hard that is. If the starting point for inflation in recent years has actually been say, a quarter to a half a point below what it's actually being printed at, how important is that? Do you think that could explain why wages growth around the world has looked to where it's actually in real terms, maybe not so bad? And I want to give you the opportunity to talk about the currencies. Is jawboning still off the table for the RBA, and does it work, does it not? There's been a bit of momentum behind the currency here at the moment so here's your chance to have a crack …

Which way, up or down. Yesterday it was down. The inflation point's sort of interesting. I certainly don't think a quarter point here or there is consequential for wage outcomes. I don't think that everyone ever said … "Damn, the inflation rate is a quarter point lower than I thought, I'm going to ask for a quarter point lower wages than I thought I was going to ask for." That, I don't think is consequential for decision making. So, it's below the decision-making threshold.

I think one thing which is interesting is as I said, we have all these models including the Phillips curve i just stuck up there earlier, which links inflation to concepts of spare capacity, right? And that's based on a historical data which has those biases in the CPI in there throughout the whole series. So you have to think about what that implies for your estimation. So just saying that it's lower isn't quite true because also it has an impact on the relationship you've estimated in trying to work out where you are at the moment.

It's a big global central banking industry at the moment trying to work out why wages and inflation are lower than you might expect them to be. Here, actually it's not quite so much of a puzzle in the sense we are sitting here with what we think is spare capacity. So that does at least line up with inflation a bit on the low side and wages a bit on the low side. We're not like the US where, whatever you thought spare capacity was before, you're now below that. But you've still got low wages and inflation. So the dynamic here, in Australia, is different from what it is elsewhere. On Amazon, I think it's quite interesting that it'd have to be about the most effective advertising campaign in the history of advertising campaigns. Don't spend a dollar, get everyone else to talk about you, all your competitors to talk about you, before you're even here in the country, pretty effective advertising campaign. Get them to adjust their prices before you even get here, it's an interesting advertising – well, non-advertising campaign, I suppose.

But there are all these dynamics going on out there. I think one thing which is interesting is looking at what, as I said, this is a global phenomenon, not a local phenomenon and therefore the question is, does it require a global explanation or is it just a whole bunch of idiosyncratic local explanations? I mean there probably is something global going on out there. The fact that it is so ubiquitous across the developed world and even the emerging world too, actually. But to some extent, as I said, coming to Mardi's point about where resources are being devoted, that is one area where a hell of a lot of resources are being devoted to, right round the world. But to date, haven't come up with a particularly good explanation yet.

Just a few more questions. Here and then …

I'm just wondering, if we haven't seen what we want to see at the moment in terms of inflation and wages, is it possible that we won't get the pick up in inflation and wages for years to come? Have you got any evidence to suggest that that won't happen? Are we going to get a lift in wages next year, the year after? When is it actually going to happen?

Well, if you're on a minimum wage, you actually got a wage increase bigger than you were expecting last month. So for some people, it is happening. But not economy wide, as you said. Similarly, on inflation, I suppose the other thing, inflation's higher than it was a year ago. Not by a lot but it has actually gone up, not down. And it's sitting just below two, as the number we got yesterday. We're going to publish a new set of forecasts in two weeks, yeah? But at least judging on the ones we had previously, yeah, we expect it still to go up. It's pretty gradual, both on the wage front and on the price front. But …

Can you understand the frustration, though? Because there's a big proportion of the Australian community that hasn't had a decent pay raise, but we're getting that evidence. That's becoming more and more frustrating for people.

Well, the Governor has said on a number of occasions that is frustrating but it is what it is. From our point of view, it would be good if wages growth was higher, obviously for the people who are getting those wages, it would be good if their wages were higher. Partly as I was saying a minute ago, understanding why exactly that is the case is somewhat of a mystery. It's not something which is unique to Australia. It's a global phenomenon that we're seeing this unexpectedly low wage and price outcome. The only thing I'd mention on the price side. So it's lower, but as I said, it's only just below two. In the US, it's not quite where they want it, it's about a quarter of a percent lower. So we're not talking on the inflation front at least, really big misses but on wages, it has been surprisingly low, globally, not just locally and it is a mystery as you say. It is frustrating. It would be nice if it was higher.

Last question.

You gave quite an elegant speech on all the approaches and the methodologies and the actually called the complexities of modelling, macro aggregates and key variables that the Reserve Bank undertakes, and the inherent uncertainty of it. I found it quite illuminating. However, what you didn't mention was from the practical point of view amongst economists and retired economists and lay people and consumers that we're confronted with, not just the Reserve Bank, we're confronted with the Commonwealth Treasury who undertake rigorous macroeconomic modelling. We're confronted with numerous Commonwealth departments that undertake macroeconomic modelling, we're confronted with numerous state governments that do their own macroeconomic modelling.

And surprise, surprise, we're also confronted with numerous private sector agencies and institutions that undertake macro econometric modelling. So in the jungle of econometric modelling, who is the King Kong? Who beats their chest the loudest? From the consumers point of view, us and the community, do we have more confidence in the Reserve Bank's more sophisticated approach? Or do we have more confidence in the Commonwealth Treasury's alleged sophisticated approach? Or even our colleagues in the universities when they come out? I'm sure you're aware of that Sydney Morning Herald 's Once a Year outline of we'll pick 25 economists or groups and we'll see how all those forecasts narrowed out. So my simple question with a rather long comment was the uncertainty also is just as important to us in the community because who the hell do we trust? The politicians will always choose the forecast they want to push, they want to justify.

But from the community's point of view, who do you trust in terms of what degree of confidence do we have, whether it's the Reserve Bank or whether it's our colleagues in Treasuries right across the board or whatever.

One thing I would say is that our forecasts and the Commonwealth Treasury forecasts are very similar. So if you're choosing one or other of those, you're pretty much going to get the same information there. One thing, I referred to Tetlock earlier. So one thing, it is useful that there is a bit of competition in the forecasting space. So that it's actually, you can learn from what other people are saying and you can discount it if you want. So I take your frustration from the public consumption point of view, but from at least, our point of view, it's actually useful that there's a few others out there that are doing this and we're not just, because that at least, challenges us to justify why we have these forecasts and someone else has those forecasts.

So that's a useful discipline to be able to have. We obviously think our forecasts are okay otherwise we wouldn't put them out there. So if you want to use them, go right ahead. They're our best guess as to what we think is going to happen. But as I say, and partly to that point, there's not, most of the time, there's not a huge variation and to some extent, and it makes for good media to highlight the differences rather than similarities, right. And if you do see a big variation then that, to some extent is interesting in and of itself, because that probably means that there is something probably pretty big going on there, if you're able to get that big of a range of outcome.

So to some extent, there is information in the distribution or the spread of forecasts that you actually see. But I would say, notwithstanding all the different sources of forecasts that you said, by and large, the variation is not all that much and it's challenging but, to some extent, you're actually sourcing the wisdom of crowds by just looking at the range there and saying, "Okay, this gives me some idea of the spectrum." Most of the time that spectrum's actually not that large and again, coming back to something I said earlier, probably not that consequential for your own decision making. When that spectrum is large, I think that's interesting in and of itself. Except as I said, that probably means something's going on and we may well be at some sort of large turning point in terms of the outcome. So, actually, I think it's better to have that multiplicity of forecasts rather than just have one because I think there's actually information in that variation which is useful.

All that's left for me to say is Guy Debelle, thank you, very much.