After reading this Australian article, I looked for the relevant US diplomatic cable, largely because the paper cannot be assumed to quote things accurately or in context. I found something else that worried me though. Here’s two excerpts, with my emphasis.
Although the Board had intended to wait a few months for the new expansionary rate to flow through the economy, 1 Edey said the Board agreed that further deterioration in the global economy merited another aggressive cut in February. The Board decided to pause on cuts in March, in part to give the Reserve Bank enough ammunition down the road in case the economy spirals into deeper recession.
And later
Steven Kennedy, Chief Macroeconomic Advisor to PM Kevin Rudd, told us April 1 that he and his Treasury Department colleagues are increasingly frustrated with the RBA. Praising the Bank’s initial response to the global financial crisis, Kennedy said that the RBA’s aggressive approach helped reassure markets and complimented the Government’s stimulus packages. He complained that the RBA is “schizophrenic” in that the Banks simultaneously believes that Australia will be hit hard but that it should hold off on further rate cuts in order to “keep its powder dry” in case things get worse.
This is what disturbs me. I hope it’s just the diplomats projecting a typically American love of guns and thus gun analogies onto others. When it comes to monetary policy “keeping your powder dry” doesn’t make any sense.
Instead lets use a differing analogy of a central banker driving a car along an undulating road. Her task is to keep to a constant speed. In Australia under inflation targeting this speed is 2-3% CPI growth, but it could be steady growth in nominal GDP if you are so inclined. At boom times the car is heading down a hill and goes faster, with higher credit growth and inflation. The banker eases off the throttle. In bad times the car hits a climb and starts to slow, and the banker puts her foot down in order to maintain a constant speed. On any given slope there is an appropriate level of throttle for the desired speed.
Now imagine the car has hit a steep slope. She obviously responds by opening up the throttle. But she is unsure about whether the slope will get steeper still. Should she refrain from pushing the accelerator too much so she can push it harder if the slope does get steeper? Should she be saving accelerator just in case?
That wouldn’t make any sense. If the road gets steeper it won’t matter if she is able to push the pedal down from a higher starting point. If a pedal to the floor can’t make the car go fast enough up the hill, it doesn’t matter how early or late you pushed it there.
Similarly, there is no model I know of that would explain why, with a given macroeconomic climate, a given interest rate would be more stimulative provided it had followed a drop. If things got worse, why would you get better credit growth, consumption and investment by dropping the cash rate from, say, 2% to 1% than you would have seen if it was already 1%. Given the macroeconomic climate, it is either sufficient or it is not – whether people borrow, lend or invest enough at that rate isn’t related to whether it was a big change from last month.
As it turned out all this didn’t matter in 2009, thanks in no small part to a stimulus that remains an astoundingly successful piece of applied macro and its counterpart in China. More monetary easing wasn’t required.
Yet I’m still troubled by this cable. If the cables accurately reflect Edey’s and Kennedy’s words, why would a RBA staff member use the ammunition analogy to describe their method, and why would a government adviser think it was an accurate description of the RBA’s thinking. Recently I suggested that more recent policy decisions made sense under a “least regret” policy suggested by Stevens in 2003. This approach would also call for a policy setting based on uncertainty about the future, but in the circumstances of 2009 it would have meant the direct opposite of the gunpowder analogy’s logic. Lest things get worse it’s much better to drop rates too far than too little as getting into a liquidity trap is far easier than getting out.
It might make no sense, but the idea of higher rates as “more ammunition” is popular amongst media commentators. Some of whom were prepared to credit the RBA with foresight for having higher rates than the North Atlantic central banks . They may as well have credited Howard with foresight with his profligacy – a cunning plan to prod the RBA into higher rates in 2007 so they had ammo stores. If a bad recession had eventuated in 2009 would those same commentators have been assuaged that further cuts mean that the RBA was “doing something” and refrain from criticism?
I really, really hope that the RBA board weren’t allowing a political considerations about media coverage enter considerations of policy. But the papers had already been feral, and the alternative is that they actually believed the silly gunpowder analogy, and that’s more troubling.
Independence is meant to help stop this kind of thing, but it also creates its own kind of politics. At the end of the cable Kennedy is described as believing that the board is worried about not looking independent enough. But choosing policy settings based on whether they look sufficiently distant from elected government is as bad as choosing them to endear oneself the the government. In both cases it divorces the decision from it’s ideal objectives.
In 2009 this turned out not to matter. I hope it doesn’t end up mattering.
- Reserve Bank Assistant Governor for Economics[↩]
Nicholas made the same points as you about this silly “keep your powder dry” idea 2 years ago on this blog – see here.
As we all agreed last time, its silliness ought to be obvious even to journalists. However anyone who can do a little maths can quickly demonstrate its silliness (basically by mathematising your analogy); have the RBA staff forgotten everything their undergraduate lecturers taught them?
Was this type of bank activity once described using levers?
It always made me think of a railway switch control box with the bankers rushing off to pull that damned lever before the locomotive of economic action crashed or careened off onto a dead end line.
Levers are an acceptable analogy – just pull on them a little , get the tension just right and suddenly Goldilocks starts popping into conversations.
Yep.
Of course it would be possible to build a model in which ‘keeping one’s powder dry’ made sense. But at a time when the RBA is ‘keeping its powder dry’ you’d think that such models might be kicked around by the research department so we can see what their implications are and whether there is any evidence we’re in that kind of world. But not a bit of it.
It’s not as if you don’t have to fly by the seat of the pants. But making sure you’re also addle-brained is not necessary.
Alas, this is amongst a group of people who pride themselves on their rigour.
Very disturbing. Analogies have power, especially when oft repeated, and this one shouldn’t even pass the laugh test.
Murph – I’d have preferred to had an analogy where monetary policy was only a brake, given I think it can only constrain rather than stimulate (particularly outside Australia where variable rate mortgages are rare), but it complicated the main point.
Unfortunate that this post is accompanied by a reproduction of one of the most detestable newspaper front pages in recent Australian history.
I noticed that in declaring Melbourne, Sydney, Perth and Adelaide in their list of the world’s most livable cities recently, the Economist Intelligence Unit survey editor said “”Despite the rising cost of living, driven by the strong Australian dollar, these cities offer a range of factors to make them highly attractive.” Come again? While causing increased import competition, the high dollar drives inflation? Is the thinking here that capital inflow due to the high dollar is pushing up interest rates and the cost of living? Sounds like drawing a long bow.
things are a little more complicated. There are explicit models in which there is some rationale for ‘priming the pump’ (though you have to buy into particular goal functions, such as a desire to have no recession) and there is certainly a whole implicit model about the psychology of recessions and the animal spirits of markets in which an interest rate stimulus makes sense if done at the right time. Like debt, there is an implicit ceiling above which you dont want to raise interest rates. There is also a floor below which you cant go with interest rates (zero), giving a role for such ‘dry powder’ arguments.
There is an almost perfect analogy here with consumption and saving, whereby a temporary increase in consumption just means you have to save more after the consumption. There is only so much debt you can have. You can thus frown on a spending splurge too for almost identical reasons.
What is true is that central bankers tend to sound a lot more like psychologists when they are defending a stimulus. They are then experts in things like confidence, momentum, and optimism.
kevin1, the market for the Economists’ “liveability” index is corporations wanting to post executives overseas. They’re looking at the cost of living for an expatriate, and the high Aussie dollar means our cities are indeed expensive for them. Not so for those paid Australian wage rates in $AU of course.
I can’t see your point Paul. If things are so dire that you risk ending up at the zero bound, how does lowering more slowly help avoid that? In fact if you lower them slowly you’re far more likely to be in the grip of deflation before you get to zero. Unless you’re invoking the confidence fairy and just saying “no sudden moves – don’t scare the horses”. But by assumption the horses are already scared. Let’s invoke a little bit of rational expectations – if sudden moves are called for, the market will know it just as well as the RBA and thus are likely to lose more confidence if those moves are not forthcoming.
There is resilience with an economy though DD and in Australia’s current situation tremendous savings are being accummulated.
Lower rates will signal that expectations for future interest rates changes have changed and a surge of revisions to household budget planning and thinking will take a few months to show up as strength in domestic consumption.
So rather than just cutting rates in large steps I think I can understand why rates ratchet up and down in increments!
The sitution with zero bound rates and the liquidity trap and dire ecomonic circumstances is still bit mysterious to me though the explanations given seem very sensible.
murph, so that is indeed a “don’t scare the horses” argument – “don’t acknowledge we’re in trouble and no-one will believe we’re in trouble” seems to be what you’re saying.
I can see the case for “fine tuning” in stable times to probe just where the NAIRU is and in fact to gradually lower it. But that’s not what we’re talking about here – if every one knows interest rates are gonna have to be on their way down big time (eg 1991, 2008) then there is absolutely nothing to be gained, and potentially an awful lot to be lost, from resisting that.
In fact I’ve long beleived that the failure of the RBA to quickly lower interest rates to the level needed in the early 90s – that is, their determination in the name of “stability” and “signalling intent” to keep real rates high during the recovery – locked in long term unemployment through the 1990s. It was a more egregious error than the overtightening that caused that recession in the first place.
Of course they credit that with getting inflation down to its (now) target level.
I think there is a case to be made for the “ammunition” idea. Disclaimer: I have an honours degree in econ, not a phd. I have never worked for a central bank, I’m just a monetary econ nerd who loves DSGE models.
First, I agree with Paul that the zero lower bound is what drives the idea that we could “run out” of room to make conventional MP. Now I believe you and Matt Cowgill say this in response: But the level of interest rates is what matters for the effect of MP, and so the timing of when we got to that level doesn’t matter. So you are essentially saying, “Suppose crisis X occurs and we want a 100 basis point drop, to interest rate level Y, for macro reasons we all agree on. Well, it doesnt matter if we dropped 50 of those basis points 2 months before and the remaining 50 the week after the crisis, the point is we want to be at level Y after crisis X.” Please correct me if I mischaracterised your views.
I have two main issues with that response. First, the size of a change in interest rates at any given moment does matter. It’s not just getting to level Y, it’s the act of dropping 100 basis points. Why? Because the impact of MP in the short run is driven by relative changes in interest rates above what the market expected. If people expected a 50 basis point cut, then in frameworks like New Keynesian DSGE models, they would be pricing/factoring that in to their decisions (as best they can, though not perfectly) – but a 100 basis point cut is a 50 bp “surprise” and the surprise is what brings the big policy impact.
If you already lowered beforehand, you have a conundrum: to surprise people and get a big impact, you have to drop 100 bp on TOP of what you already dropped (and the resulting i level you’d get is lower than Y) — but you might not be able to do that if you are approaching the zero bound. Certainly you won’t be able to do that too many times (and from memory I think the RBA safely dropped 100 BP twice during the GFC…I might be wrong though).
Now while in the short run, NK DSGE models suggest that the magnitude of the rate change at a given time is important, in the long run, clearly the level of interest rates IS important in terms of inflation. That is my second concern: if you lower too early and keep rates low for a long time in anticipation that there may be a crisis in the future at some point, you are having greater-than-optimal inflation, and laxer-than-optimal lending standards (due to easy borrowing at these lower rates).
This is clearly part of what happened in the US. And lax lending standards are a key input of asset bubbles. So in many cases (but perhaps not all) long periods of pre-emptively lower interest rates are going to lead you to situations of crises, where you wish you could slice 100 BP off, maybe even more than once, without worrying about that lower bound.
Of course, I am not saying we shouldn’t cut interest rates when the situation calls for it, and sometimes that is before a crisis hits. I’m arguing simply that there is a tradeoff between lowering now and lowering later – and that’s what the ammunition analogy is trying to capture.
If I could edit my comment, I would add: I don’t think you have to even buy into the surprise idea to think it’s the change that matters. After all, even basic MP tools like the Taylor rule aren’t about set interest rate levels, but about the degree or size of the response to changing situations. Crises are prime examples of changing situations, and where we might be more concerned about credit transmission channels, but that’s not the only time when size matters. Even if we’re worrying about something as routine as changes in inflation due to aggregate supply shocks, the size of the response in the nominal interest rate is going to make a difference to the effect on the real interest rate, and that’s what drives outcomes.
Come to think of it there’s a lot of macro literature about MP responses being “big enough” to make sure the real interest rate is changing (look for example at Clarida, Gali and Gertler’s awesome paper on the topic, includes some good stuff about the US under Volker and Greenspan). They are explicitly talking about the size of the response!
I’m taking my time replying, something I’m sure you won’t object to, whilst looking through the C, G and G papers and clarifying some thoughts.
But for now I would like to point out that the surprise idea (whatever its virtues) probably isn’t the rationale that the RBA is using. That is of course if they really are using the ammunition metaphor[fn1]. After all a part of the reasoning behind inflation (or NGDP) targeting is to shape expectations not just in regard to future inflation but potentially central bank behaviour. There’s also the widely assumed practice of giving certain newspaper commentators a heads up to softly guide markets into the correct expectation. After the 2007 election they quickly seized the opportunity to release board minutes (which I understand Costello had prevented) to make decision making more transparent and statements intended to indicate future policy direction – none of which can be seen to be part of a bait and switch. Over the past 20 years they’re steadily moving ever in the direction averting surprise as much as possible.
If they were using the surprise idea, this would be the most alarming development of all. Coupled with a steady walk away from monetary mystique it would reveal a massive cognitive dissonance.
On the US example, I don’t think we can say that is was “clearly” case of preemptive easing. It’s hard to say anything was clear about Greenspan’s motivations, a great advocate of monetary mystique. I suspect his mystique was mainly a veil of sophistication over his own political self interest and spivery. That said, one thing he was fairly clear about is that he did not think crises were possible. Why pre-empt something you don’t think can occur? Using the metaphor in the post it’s the car driver speeding along the Nullabor plain – no suspicion of a hill at all, just too much accelerator. If there was a failure to raise rates against inflation, it cannot be due to uncertainty about a potential slow down. More likely it was due to baser political motives.
The issues of asset price bubbles are fascinating, but I think ultiamtely unrelated to the powder argument (somewhat discussed here).
[fn1] There’s only 2nd and 3rd hand evidence, and observed behavior is not conclusive.
Okay, so like I said you don’t need the surprise framework to argue that size matters. And I agree that central banks want their behaviour to be predictable – I don’t think that’s incompatible with the surprise idea, as they both come from standard New Keynesian macro (though in the literature the term used is “unexpected” not “surprise”)! So, if the RBA can be relied upon to cut interest rates by 50 BP if X happens, then people (imperfectly, slowly) build that into their decision making when X happens (or when they think X is likely to happen). That way, MP has a smooth, gradual effect before the RBA ever does anything, so inflation is way less volatile under this regime. That is really great and part of their general goals. But as a direct result, if the RBA wants a bigger effect at the time of changing rates, it has to surprise people at that time. (Mankiw and Reis’ 2002 paper has some stuff on this I think but it’s not the main point of that paper, and it was pretty technical…)
Also I think I did not make myself clear on the second point: I’m certainly not saying the Fed was pre-emptively lowering rates to prepare for a potential crisis. I’m saying that the Fed, for their own alchemical reasons, lowered rates too low for too long. This was a direct input into the US mortgage crisis. Did those low rates help them to manage the crisis once it happened? No. It did them no good at all to be that low. They needed to make cuts, so that they could change the status quo in the market from A to B (wrt lending or whatever). But they did not have enough room to move (and of course they also had a broken transmission mechanism).
I’m using this merely as an example of the things that can go wrong when you lower interest rates when you don’t actually need to, and that includes lowering them a long time before you think you will need to. I agree that we can debate how much time before a potential situation is a good time to do it, but I think we need to be conscious of the risks.
I think my argument stands: that there are good macro theory reasons to worry about change in rates, not just level of rates, and that having rates low when you don’t really need them that low comes with potential costs. I’m not using any metaphors or analogies here, as I think they’re misleading since we’re disagreeing about theory – so I’m trying to argue just based in the theory. :)