His critics blame Alan Greenspan for setting interest rates too low, which caused the house-price bubble, which then burst and caused the financial crisis. As I argued back in February, the critics are typically confused between interest rates that are low, and interest rates that are low relative to the natural rate.
The topic is making the rounds of the economics blogs again, with Brad DeLong cogently making the same point about the natural rate. I want to add a Canadian twist. Canada also had a big increase in house prices. But since average Canadian inflation over the same period was almost exactly equal to the Bank of Canada's 2% target, we know that the Bank of Canada did not on average set interest rates below the natural rate.
If you can't blame the Bank of Canada for the boom/bubble in Canadian house prices, how can you blame the Fed for the boom/bubble in US house prices?
If the central bank sets an interest rate below the natural rate, inflation will rise relative to expected inflation; and if it tries to keep it below the natural rate, actual and expected inflation will rise without limit. If the central bank sets an interest rate above the natural rate, inflation will fall relative to expected inflation; and if it tries to keep it above the natural rate, actual and expected inflation will fall without limit. That's what the "natural rate" is best defined to mean.
It is economic nonsense to blame a central banker for keeping interest rates too low, unless you mean to blame him for keeping interest rates too low relative to the natural rate. Because using monetary policy to try to set interest rates above the natural rate would cause a deflationary spiral, and that is not what central bankers are supposed to do.
So, if we are to make any sense of Greenspan's critics, they must be saying that Greenspan set interest rates below the natural rate. What evidence is there for such a statement?
We do not observe the natural rate of interest. We cannot tell, in real time, if the actual rate is above or below the natural rate. But with hindsight, things may become clearer.
In Canada, things definitely become a lot clearer in hindsight. That's because Canada has a well-defined inflation target for monetary policy. If the Bank of Canada gets monetary policy right, expected inflation is supposed to stay anchored at the 2% inflation target. So if inflation rises above 2%, it is above what expected inflation is supposed to be, and we know, with hindsight, that the Bank of Canada had previously set interest rates too low, relative to the natural rate, and given the inflation target. (The only ambiguities concern when precisely the Bank of Canada had interest rates too low, and whether it could reasonably be expected to have known in real time what we all now know with hindsight).
We know, for example, that Canadian inflation (both total and core CPI) was above the 2% target from mid-2002 to mid-2003, so we can say with hindsight that the Bank of Canada had set interest rates too low about a year or two earlier. But at other times, inflation was below the 2% target, and we know with hindsight that the Bank of Canada had set interest rates too high.
The US does not have a well-defined target for monetary policy. And in particular, the Fed does not have an inflation target against which we can judge what actual and expected inflation are supposed to be. So we cannot say, even with hindsight, whether the Fed set interest rates "too low", without making some additional judgment about what inflation and expected inflation should have been.
The US had a big increase in house prices. So did Canada. The Teranet-National Bank composite index (Canada's version of Case-Shiller) shows an 85% nominal increase from July 2000 to the peak in July 2008. (The data begin in January 2000, but I have chosen July 2000 to get rid of seasonal effects.)
From July 2000 to July 2008, Canadian core inflation averaged just slightly less than 2%, and total CPI inflation averaged just slightly more than 2%. Since 2% is the target for both actual and expected inflation, we can say, with hindsight, that during the house price boom in Canada, the Bank of Canada, on average, had set interest rates almost exactly equal to the natural rate.
You can't blame the Bank of Canada's setting interest rates too low (relative to the natural rate) for Canada's house price boom/bubble, because we know that the Bank of Canada did not set interest rates too low, on average, relative to the natural rate, during the period when Canadian house prices were rising.
Greenspan's critics need to do two things to make their case: first they have to show that the Fed set interest rates too low, relative to the natural rate (and that is not easy without a clear inflation target); second they need to argue some kind of "American exceptionalism", because their argument that the house price bubble needed low interest rates, relative to the natural rate, doesn't seem to work elsewhere.
Brad rightly writes:
This is indeed the old Austrian definition of the natural-rate, but there is no reason to expect that this is the same as the neutral-rate--or the rate at which the price-level is constant.
Posted by: Jon | July 04, 2009 at 04:11 PM
Jon: I would argue that it is the same, and that it is also the same (in real terms) as the rate at which actual inflation equals any expected target rate of inflation. It could only be different if you believe there is some non-super-neutrality of money. http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/06/nonsuperneutralities-an-open-invitation-to-austrians-especially.html
Posted by: Nick Rowe | July 04, 2009 at 04:42 PM
There must be an argument that the issue of housing prices is completely separate from this natural rate business. After all, house inflation is reflected as "owners equivalent rent" in most central bank inflation measures, not as inflation in the asset price. Is the natural rate the issue here, or the measure of inflation?
Posted by: anon | July 04, 2009 at 06:35 PM
Let's imagine two scenarios, in the first businesses will borrow X dollars and spend it productively causing a lower price level, in the second businesses will borrow X dollars and spend it frivolously causing a higher price level (demand for whatever they are buying with no corresponding increase in productivity). The same rate of savings and the same demand for investment by businesses (ie. the same natural rate) but two different price level outcomes. If you were only watching price levels you would have missed the point, how the money is spent matters. Is that wrong?
Posted by: pointbite | July 04, 2009 at 06:35 PM
Is the natural rate the rate the market would set if there were no central bank?
Posted by: anon | July 04, 2009 at 08:25 PM
If house prices were included in CPI the way they used to be we would have registered over 5% inflation. Central banks got it wrong by focusing too much on narrow metrics like Core CPI and missing the big picture. The housing bubble was obvious *when it was happening*. Any graph of credit vs. real wages showed that. Savings rates went to zero. Please stop making excuses for these people. They had one job to do, and they got it wrong.
Posted by: rp | July 05, 2009 at 02:17 AM
rp: " Please stop making excuses for these people. They had one job to do, and they got it wrong."
Please read my post. The Bank of Canada did indeed have one job to do: to keep CPI inflation at the 2% target. And it got it almost exactly RIGHT (on average, over the period in question when house prices rose).
And as [email protected] says, house prices are indirectly included in the Canadian CPI, via owners' equivalent rent. Whether the owners equivalent rent term captures house prices as fully as it should is another question; but the Bank of Canada was targeting successfully that agreed-on measure of inflation.
[email protected]: If I remember correctly, Wicksell originally defined the natural rate of interest as the rate that would prevail in a world without money. The trouble with that definition however is that a world without money would be very very different in many ways, so it is very doubtful if the real equilibrium rate of interest would be the same without money. It's almost impossible to imagine the relevant thought experiment.
Perhaps I should do a post on the meaning of natural rates.
pointbite: Let me change your example a little if I may. Imagine two scenarios: in A business investment is in productive assets; in B business investment is wasted in digging useless holes in the ground. Many things will be different between A and B. After a few years, A will have a much higher supply of real goods than B. But whether the price level will be different depends on how monetary policy responds. Under a monetary policy of inflation targeting, A and B will have the same inflation rate and price level, because the central bank would make it so.
[email protected]: I'm not sure if I follow you there. Sorry.
Posted by: Nick Rowe | July 05, 2009 at 08:00 AM
Nick,
I am not sure this always holds. If productivity growth suddenly increases and the other factors driving the natural rate--intertemporal preferences, population growth--don't change, then one could have an increase in the natural rate with deflationary pressures. Now if the central bank drops its policy rate to offset these deflationary pressures so it can reach its positive inflation target one could have a situation where (1) the market rate is below the natural rate and (2)a low inflationary target is being met.You state the following:
Given the above scenario, it seems to me that looking to expected and realized inflation is not very informative as to whether the policy rate is below the natural rate.
Posted by: David Beckworth | July 05, 2009 at 09:36 AM
Nick wrote: "Whether the owners equivalent rent term captures house prices as fully as it should is another question; but the Bank of Canada was targeting successfully that agreed-on measure of inflation."
Given owners' equivalent rent understated inflation in the U.S., it can be argued that the FOMC was not successful targeting inflation if we are to assume their acceptable inflation range is close to 2%. If adjusting CPI to more accurately depict housing prices produces a much higher CPI, would this not be evidence that interest rates were low relative to the natural rate?
Posted by: HB | July 05, 2009 at 10:01 AM
I more or less agree with the original post. The only things I really fault Greenspan for are:
1) blaming the low natural rate on the asian/opec savings glut (correctly), but not attempting to push for any solution like what the Chinese & co. are now pushing for. He just went with the flow.
2) pushing hard for financial de-regulation, helping to kill Brooksley Born's recommendation for regulating OTC derivatives, etc.
I think he is somewhat culpable for this disaster, but not because of the low interest rates.
Posted by: Jeff | July 05, 2009 at 11:03 AM
Nick, but is it desirable that A and B have the same price level, even though A is much more productive? Maybe prices need to better reflect how money is invested.
Posted by: pointbite | July 05, 2009 at 11:17 AM
David: I both hoped and expected you would join in on this argument!
"If productivity growth suddenly increases ..[other real things constant].. then one could have an increase in the natural rate with deflationary pressures."
That's an interesting point, and well-above the standard for Fed-critics. (Actually, I had you specifically in mind when I inserted the weasel-word "typically" in my opening paragraph "...the critics are typically confused between interest rates that are low, and interest rates that are low relative to the natural rate.")
But I am not sure it's right.
First, I would question your assumption that there was an increase in productivity growth.
Second, I would question your assumption that the natural rate of interest increased (regardless of what caused it).
Third, even if I grant both those assumptions, it is not clear to me that an increase in productivity growth is deflationary. That's the point that really interests me theoretically. You see that I messed with your quote above, by inserting my own words "other real things constant". I did that deliberately, because I don't think we can say whether an increase in productivity growth is deflationary or not without specifying monetary policy.
If there is a downward-sloping AD curve, in {P,y} space, and if monetary policy holds that AD curve constant (or holds its rate of movement constant), then an increase in productivity growth causes the AS curve to move right at a faster speed, and will indeed be deflationary. But is that a correct depiction of monetary policy?
If monetary policy targets inflation (as in Canada), the AD curve is horizontal, and moving upwards vertically at 2% per year. The rate at which the AS curve is shifting rightwards should have no impact on inflation. And if the Bank of Canada gets it right (and it did over this period) the actual rate of interest should stay equal to the (higher) natural rate throughout.
And if monetary policy targets the (real) rate of interest, we get a vertical AD curve. If the central bank gets it exactly right, and raises the actual rate when the natural rate increases (due to faster productivity growth), then the vertical AD curve should be exactly on top of the vertical LRAS curve, and the price level (and inflation rate) is in neutral equilibrium (like a ball on a flat table), with an undetermined tendency to either deflation, inflation, or neither.
Now it may be possible to re-state your argument to get around my criticism above. Suppose we argue that wages, or wage inflation, are sticky, or intertial. In the short run, output price inflation is determined by wage inflation (which is temporarily pre-determined by interial forces) minus productivity growth. The optimal monetary policy in such a world would be to target wage inflation, not price inflation. (Standard argument: you target the stickiest prices, that don't want to change, so they don't need to change). But if the central bank targets price inflation instead, you would get a short-run disequilibrium in the labour market, with excess demand for labour, and unemployment below the natural rate of unemployment.
I think that argument could work in theory, but not sure whether the assumptions and predictions would match the particular historical episode.
HB: The trouble is, we don't know what the Fed was supposed to be targeting, and so we don't have a good idea which of the many different definitions of inflation people should have expected the Fed to hold constant. The Canadian CPI leaves stuff out too. But we know exactly what the bank of Canada is targeting, so we know exactly what definition of inflation the public should expect to be at 2%, and so we know for sure whether monetary policy caused actual and expected inflation to deviate.
If Americans expected the Fed to target "broad" inflation, and so expected "broad" inflation to be constant at 2%, while the Fed actually targeted "narrow" inflation to 2%, and if actual "broad" inflation turned out to be higher than "narrow" inflation, then yes, that's a short-run monetary policy shock. But that's also a lot of assumptions. And why did Canadian house prices rise, without a Canadian monetary policy shock? Maybe the US house prices increased for some other )non-Fed) reason too, if Canadian ones did?
Posted by: Nick Rowe | July 05, 2009 at 11:20 AM
Jeff: As far as my knowledge lets me judge, yours is a much more reasonable position.
I would like to see some sort of comparative (across countries) analysis of the (de)regulation question too. For example, if you blame the US problems in the lack of regulation A, but it turns out other countries had A, but had the same problems as the US, it looks less plausible. But I just don't know enough about regulation.
Correct me if I'm wrong on this, but I think I remember that Greenspan did push for one solution to the low natural rate problem. Didn't he push Bush to have a loose fiscal policy, which should have raised the natural rate (assuming, reasonably, that the US is either a large open economy [sorry, but Canadians have to explicitly remind ourselves NOT to think in small open economy terms sometimes] or has imperfect capital mobility). But I don't hear any of Bush's critics (and there's no shortage of them) complaining that Bush had fiscal policy too tight!!
In fact, if the "permanently indignant" critics were at all consistent (they aren't) they should be praising Bush for his loose fiscal policy, which helped keep the natural rate of interest high, and so should have helped prevent any housing bubble!
pointbite: the whole (and very old) doctrine of the neutrality of money says that the overall level of prices, measured in money, does not matter. It's relative prices, not absolute prices, that need to be right, to give the right rewards and incentives to real investment decisions.
Posted by: Nick Rowe | July 05, 2009 at 11:41 AM
Nick:
You know how to bait me in! I, an assistant professor seeking tenure, should be working on a paper and instead I get pulled into a Nick Rowe trap from which I probably cannot extract myself! Anyhow, here goes my response.
First, most U.S. productivity measures show an increase in early-to-mid 2000s. Here is a figure that shows the y/y growth rate of non-farm business labor productivity.
Second, here is a paper from the ECB that shows rigorously that the Fed did push its policy rate below the natural rate.
Third, I agree with your point that productivity does not necessarily create deflationary given certain types of monetary policy. But in a sense that is part of my point: in the absence of such monetary policies there would have been deflation from these productivity gains. And to the extent that monetary policy has short-run real effects on the economy, this added monetary stimulus that prevents the deflation from ever emerging can temporarily push real economic activity beyond full employment in a hidden way (no inflationary warning signs emerge).
By the way I was in Canada for my first time last week. I was attending the Western Economic Association meetings in Vancouver. Lovely place, lots of good food.
Posted by: David Beckworth | July 05, 2009 at 12:51 PM
David: yes, sorry, I know how easy and dangerous is it to get sidetracked. Long time since I was in Vancouver. I was in Texas much more recently, and had a great time on Padre Island!
I disagree with your point that preventing deflation from productivity growth requires monetary stimulus in the sense of setting the actual rate below the natural rate. I thought I had dealt with that "Neo-Austrian" fallacy (it's not I think in Mises or Hayek) in my non-superneutralities post. But leave that aside.
I was really interested to read the Lombardi and Sgherri paper you linked. It's a serious bit of work. The econometrics is way over my head of course, but something in the results really troubles me.
Look at the pictures on page 43. Look first at the episode in 1974-1980, when the actual rate of interest was set well below the estimated natural rate. The results are exactly what one would expect: a positive output gap (output above the natural rate): an increase in inflation above target: and an increase in the estimated inflation target as well. (They assume inflation inertia in the Phillips curve, so any increase in inflation, whether expected or not, requires the actual rate below the natural rate of interest). So far so good.
Now look at 2000-2006, when there was (according to their estimates) an almost equally large deviation of the actual rate below the estimated natural rate of interest. OK, we see the same positive output gap. But look at inflation, and the target!! The estimated target rate of inflation rises a little, but inflation does not, and inflation stays roughly equal to target!
These results make little sense theoretically, and are very different to the 1974-1980 results, which do make sense theoretically. If the actual rate is a long way below the natural rate, we ought to see a big increase in target inflation (due to inflation inertia), and/or a big rise of actual inflation above target. We don't. What is going on in that model??
It came to me while cutting the lawn; they have got a CLOSED economy model. Any fall in the natural rate can only come from lower household time preference or a decline (yes) in output growth. It can't come from China, or OPEC, or anywhere else outside the system. So the standard story for the fall in the natural rate is ruled out by assumption. Their econometrics is desperately trying to resolve the contradiction between strong US consumption with no big uptick in inflation. It's forgetting imports, and borrowing from abroad.
Posted by: Nick Rowe | July 05, 2009 at 03:28 PM
Nick:
I missed your post dealing with the Neo-Austrian fallacy but will be sure to take a look. I may well be missing something here. For now I will note that I am puzzled as to why it must always be the case that inflation must rise if the actual rate of interest is below natural rate. Why? If this assumption is not true, as I believe, then the results in the paper you cite are not so mysterious. A positive output gap could easily emerge without inflationary pressures emerging. I think this can be shown most clearly in terms of nominal spending.
Consider the following two scenarios.
(1) A central bank has a 2% inflation target and the economy's natural rate of growth is 3%. Here we have nominal spending or GDP growing at 5%. Now imagine the central bank generates a positive monetary policy shock that increases nominal spending and pushes inflation temporarily to 4%. Now nominal spending is growing at 7% and if there are any nominal rigidities (i.e. upward slopping SRAS curve) this increase in nominal spending (or AD) should push should push output beyond its natural rate. Hence, a positive output gap is created and there is an uptick in inflation. This scenario describes the 1970s experience you reference above.
(2) A central bank has a 2% inflation target and the economy's natural rate of growth is 3%. Once again, nominal spending is growing at 5%. Now assume a permanent productivity innovation pushes the natural rate of real economic growth to 5%. Assume also that surge in productivity in the absence of any new accommodation or changes in monetary policy--that is, the central bank is still increasing money supply at rate that would have created a 2% inflation target under the old steady state of 3% real growth--would have pushed inflation down to 0%. If the central adopts this approach and does not accommodate the increase in productivity, nominal spending will still be at 5% (0% inflation + 5% real growth). Note, there has been no change in AD (still growing at 5%) and thus no movements against the SRAS by which to create an output gap.
Now assume the central doesn't sit idly by but accommodates the productivity shock so that its inflation target is maintained. It will have to stimulate nominal spending such that the potential 2% drop in inflation is avoided. Now nominal spending jumps to 7% from its previous value of 5%. Here, we have an increase in nominal spending (or AD) that in the face of an upward-slopping SRAS will temporarily push output beyond its natural rate. In other words, an positive output gap will emerge. But here there is no change in observed inflation or inflation target! This latter scenario seems to describe what happened in the early-to-mid 2000s.
Now these scenarios don’t mention the natural rate vs. policy rate issue. But it is not difficult for me to believe that in the latter scenario, where monetary policy temporarily pushes output beyond its natural rate it also pushes its policy rate temporarily below the natural rate as well. If the natural rate of output can be surpassed on the upside with no change in inflation then surely the natural rate of interest can be surpassed on the downside by the policy rate with no change in inflation.
I hope this makes sense.
As an aside, note that had the monetary authorities been targeing nominal income at 5% there would be no positive output gap created when the productivity shock occured. By targeting inflation, however, they set themselves up for macroeconomic instability.
Posted by: David Beckworth | July 05, 2009 at 10:06 PM
David: I nearly followed that. You just lost me on the last bit, where the central bank is targeting inflation: "Now assume the central doesn't sit idly by but accommodates the productivity shock so that its inflation target is maintained....."
Let's simplify. Zero growth, and zero inflation target. Start in equilibrium. There is a one-time real shock that shifts the LRAS curve 2% to the right.
Key question: what happens to the SRAS curve? Does it shift right too? And by the same 2%? Or more, or less?
If the SRAS shifts right by the same 2%, there is no problem. The central bank just shifts the AD curve right by the same 2%, and we stay at the new natural rate of everything (on the LRAS curve), and the price level stays the same (we keep to the 0% inflation target).
Now, if we have a New-Classical misperceptions story of the SRAS curve, so that output supplied deviates from the natural rate if and only if the actual price level deviates from the expected price level, that is exactly what will happen. The SRAS and LRAS shift right by exactly the same amount, provided the expected price level stays the same.
But if we have a New Keynesian sticky wage or price story, it depends. The SRAS and LRAS may shift by different amounts, holding expected P constant, and it depends on the nature of the real shock. With sticky wages, for example, a labour supply shock does not shift the SRAS right at all. A productivity shock may shift the SRAS right either more or less than LRAS, depending on how much it shifts the labour demand curve.
What did you imagine the SRAS curve to be doing?
Posted by: Nick Rowe | July 06, 2009 at 08:09 AM
This might be relevant:
http://economistsview.typepad.com/economistsview/2009/07/a-bubble-mystery.html?cid=6a00d83451b33869e2011570b7652e970c#comment-6a00d83451b33869e2011570b7652e970c
Posted by: reason | July 06, 2009 at 08:32 AM
Nick:
I am assuming the SRAS moves proportionally with the LRAS, though it does not have to be exactly 1 for 1 for my story to hold. Where we differ, however, is in your assumption that that monetary authorities can shift can shift the AD curve (or increase nominal spending) to the LR equilibrium point where all three curves intersect and they can do this without encountering any nominal rigidities along the way.
I believe there will be nominal rigidities encountered along the way, especially if the central bank quickly and aggressively responds to the positive productivity shock. How can the central bank create a sudden increase in nominal spending without running up against some frictions? For this reason, when I draw the AD-AS model in this scenario I draw the evolution of this change in nominal spending in two steps.
First, I have the SRAS curve shift out proportionally and have the new LR equilibrium taking place where the new LRAS, new SRAS, and the old AD curve intersect. (This would be the point had there been no monetary accomodation.), Second, I then shift the AD curve out and here it climbs up the SRAS to a point that is beyond the natural rate of output. To make this easier, I hastily graphed this process here.
It may be helpful if I restate my view of the SRAS in terms of a sticky input price story. Imagine there is a positive productivity shock that affects all firms. This productivity shocks causes per unit costs of production to fall and increases profit margins. Now given there is some competitive pressures, firms will start to cut their output (sales) price to gain market share. They comfortably can do this because of the decline in per unit costs. Economy-wide this should lead to a decline in the price level and stable profit margins. Now what if the central bank increases nominal spending to offset (or prevent) this decline in the price level? If successful, the central bank will have kept output prices from falling and this, in turn, would swell profit margins so long as input prices are sticky. If firms respond to these swollen profit margins by increasing production then output could go above its natural rate for a time. Eventually, input prices will adjust and profit margins will be restored to appropriate levels. In this story there would be a positive output gap and no change in inflation (or the inflation target).
Okay back to work!
Posted by: David Beckworth | July 06, 2009 at 11:52 AM
Thanks David:
I wish I could figure out how to post links to drawings. But I expect you need an Ipod, a Blackberry, and other young people stuff and skills to do it!
Your diagram of "Nick's Approach" is a correct picture of my view (at least as a benchmark case).
But I really do not follow your approach. The shift in the SRAS curve out to be independent of the shape of the AD curve. Only by sheer fluke would SRAS2 intersect LRAS2 at the same point as AD1. And I don't understand why you are doing it in two steps (even though I realise it's just a device to explain it, rather than your saying in moves in 2 steps in real time).
Nevertheless, I can reproduce something like your result under certain assumptions.
Assume labour is the only variable input. Assume the nominal wage W is fixed for one period at the expected market-clearing level. Assume that in the short run actual employment is determined by labour demand (and not by min{labour demanded, labour supplied}). In other words, standard sticky-wage New Keynesian assumptions.
Then assume an unexpected increase in productivity, that shifts the production function upwards, and increases its slope. Since the slope of the production function is the MP of labour, and is the labour demand curve, this will shift the labour demand curve too.
In this case, the SRAS curve will shift horizontally to the right by more than the LRAS curve shifts. Exactly as you have drawn it. (Proof: because if P and W stay the same, then W/P stays the same, but LR equilibrium W/P must rise, because MP of Labour curve shifts up, so actual W/P would be below equilibrium W/P, so labour demanded and employment would exceed LR equilibrium employment, therefore y must also exceed LR y).
But the central bank would have shifted AD by more than you have drawn it in Figure 2, in order to keep the price at P1.
Anyway, we are basically on the same page, in the end, theoretically.
BUT. That all assumed sticky wages. The model you linked to had perfectly flexible wages, and sticky output prices. If my intuition is correct, a sticky price/flexible wage model ought to give exactly the same results, roughly speaking, as your diagram of "Nick's Approach". Prices (or the rate of inflation, given inflation interia) tend to stay where they are, unless Y is greater or less than the natural rate. That means a productivity shock that shifts the LRAS curve to the right will also shift the SRAS curve to the right by exactly the same amount.
Posted by: Nick Rowe | July 06, 2009 at 04:56 PM
So does this mean the key issue boils down to whether input or output prices are the most sticky?
Posted by: David Beckworth | July 06, 2009 at 05:29 PM
David: I think that's one of the key issues. Probably the most important one. If the labour supply curve (or the input supply curve) were very elastic, it wouldn't make much difference whether wages or prices were sticky. Or if the productivity shock didn't change equilibrium real wages much (for any reason), it also wouldn't make much difference whether wages or prices were sticky, and you would get approximately my results. If the productivity shock had a big effect on equilibrium real wages, and wages were sticky, we would get your results.
Posted by: Nick Rowe | July 06, 2009 at 06:38 PM
Aren't all good "neo-Keynesian" models closed economy models?
Nick writes:
"It came to me while cutting the lawn; they have got a CLOSED economy model."
Posted by: Greg Ransom | July 07, 2009 at 02:20 AM
Time to name names, Nick. I don't know anyone who is making this mistake. Who's making it, and why?
Nick writes:
"the critics are typically confused between interest rates that are low, and interest rates that are low relative to the natural rate."
Posted by: Greg Ransom | July 07, 2009 at 02:25 AM
Greg: I would call it a "New Keynesian" rather than a neo-keynesian model. Early new keynesian models were closed economy, which made sense initially, because they wanted to focus on price setting and monopolistic competition, not on aggregate demand. But the Bank of Canada's new keynesian models, for example, are open economy models. They have to be, given what they are used for.
There are two reasons I don't name names: first, because it overly personalises into a dispute between people what is at root a dispute between theories; and second, because I have a really bad memory for names, and would probably get them all muddled.
Posted by: Nick Rowe | July 07, 2009 at 08:36 AM
This would make perfect sense if banks operated "rationally" - that is if their behaviors corresponded to the market model. But just imagine if, just if, they were able to take advantage of their position (oh, say because US regulations were lightened so that commercial and investment banks could join up). Then you would have the possibility that asset prices and consumption prices would decouple. At that point, you have set an "unnatural" interest rate - with a great feedback loop to ensure you stay there - any change toward the "natural" interest rate will crater asset prices AND bring deflationary pressures (via crashing employment).
Canada was a little less corrupt (or better at hiding it, I'm not sure) than the US. I can't imaging anyone is seriously discussing the effect of interest rates as if the economy has been operated as a pristine economic model over the last 15 years, instead of the Ponzi scheme it is finally exposing itself to have been.
Posted by: pebird | July 07, 2009 at 11:17 AM
pebird: if the banking system were implicitly 'subsidising' interest rates to borrowers, then that would have required the Fed to set a higher interest rate to compensate. It raises the natural rate of interest (for the Fed funds rate). Now that may indeed be part of the story of house prices. But why did we not see a bigger increase in US inflation above the Fed's (implied, implicit) target, if that were the whole story? And why did Canadian house prices rise too, by roughly the same sorts of amounts?
Posted by: Nick Rowe | July 07, 2009 at 02:34 PM
Nick -- if I'm not mistaken, the authors themselves call their model "neo-keynesian". I was really teasing about the original Keynesian heritage, more than anything. I think that the best advice would be to give all of these models new names, dropping the "Keynes" stuff.
Nick writes:
"Greg: I would call it a "New Keynesian" rather than a neo-keynesian model"
Posted by: Greg Ransom | July 07, 2009 at 10:36 PM
Greg: I hadn't noticed the "neo". "Neo" sounds just so, well,...dated!
Yes, some of these models are a very long way from JM Keynes.
Posted by: Nick Rowe | July 07, 2009 at 11:28 PM
It seems to me that people are assuming that debt levels should be changed via interest rates to maintain 2% price inflation.
What happens when there is too much debt?
Is the "fungible" money supply off?
How are central bankers achieving 2% price inflation? Is cheap labor via free trade, legal immigration, and illegal immigration and productivity growth producing price deflation and central bankers are using debt (future demand) on the lower and middle class to prevent it?
Posted by: Too Much Fed | July 08, 2009 at 12:18 AM
I would like to know what would have happened if people hadn't been suckered by the greenspan fed into a housing bubble.
Instead of focusing on the monthly payment, what if people would have kept housing prices nearly the same and housing sizes nearly the same? In other words, people should have increased their monthly payment and got out of mortgage debt sooner.
How would greenspan employ all that cheap labor he loves to brag about?
Posted by: Too Much Fed | July 08, 2009 at 12:25 AM
Was the point of the housing bubble to extract past savings and asset gains from homes and use it to create a trade deficit?
Posted by: Too Much Fed | July 08, 2009 at 12:27 AM
If housing prices start going up 6% to 10% a year and someone can get a mortgage with a 3 year teaser rate of 3%, should they "speculate" in housing?
If housing prices start going up 6% to 10% a year, should people use home equity loans/credit to pay off credit card debt, promise not to run up the credit card debt, break the promise, and keep on extracting the asset (housing) price gains via debt until there are no more asset (housing) price gains?
Posted by: Too Much Fed | July 08, 2009 at 12:32 AM
If debt levels via interest rates produce a desired price inflation (2%) but produce asset bubbles, is there at least one imbalance somewhere else?
Posted by: Too Much Fed | July 08, 2009 at 12:39 AM
"I did that deliberately, because I don't think we can say whether an increase in productivity growth is deflationary or not without specifying monetary policy."
Define monetary policy. Is it attempt to create more private debt, especially on the lower and middle class to turn them into debt slaves, so that price deflation does not occur?
Posted by: Too Much Fed | July 08, 2009 at 12:46 AM
"In other words, people should have increased their monthly payment and got out of mortgage debt sooner.
That's not right. It should be "In other words, people should have kept their monthly payment the same, got out of mortgage debt sooner, and not bought a more expensive house."
Posted by: Too Much Fed | July 08, 2009 at 12:50 AM
Too much Fed: Have a read of an old post of mine, about the relationship between monetary policy and debt: http://worthwhile.typepad.com/worthwhile_canadian_initi/2008/12/hair-of-the-dog.html
Basically, monetary stimulus is not about increasing debt.
Posted by: Nick Rowe | July 08, 2009 at 06:03 AM
I've really can't think of a bank or mortgage "shop" that wants to lend less when interest rates go down as long as they think they will get paid back, the spread is good enough, or they can take the collateral if any.
Posted by: Too Much Fed | July 08, 2009 at 07:47 PM
"Financial innovation looks far more plausible as a cause of the increased gross debt. Low interest rates were not the cause of the problem in the first place."
I would say they both were. Try to create a housing bubble and mortgage debt extraction with a 5% or higher fed funds rate. I believe even geekspeak (greenspan) has said asset bubbles don't tend to happen with interest rates above 5% (not sure about which interest rate he was talking about).
Posted by: Too Much Fed | July 08, 2009 at 08:04 PM
"Low interest rates cure a recession both by increasing the demand for loans and by reducing the supply of loans; they encourage an increase in spending through both channels, and can be effective even if gross debt stays exactly the same."
I seriously doubt if anyone at goldman sachs has ever has saved less and spent more just because interest rates went from 6% to 3%. Is it more likely they will try to leverage their excess savings from too much income to speculate in financial assets to attempt to create even more "income" and more excess savings?
Posted by: Too Much Fed | July 08, 2009 at 08:08 PM
"Basically, monetary stimulus is not about increasing debt."
IMO, it is. Whenever the fed has cut rates in the somewhat recent past, they attempt to create more debt. That is especially true about the lower and middle class consumers. They also want the "investment banks" to lever up to prop up the stock markets. When consumer debt hits retail sales, stocks go up. They profit on their leverage bets and eventually sell. Everyone thinks all is well until there is too much lower and middle class consumer debt.
Links:
http://www.calculatedriskblog.com/2009/03/business-cycle-temporal-order.html
http://www.itulip.com/forums/showthread.php?p=86995
Title: "Debt Deflation Bear Market: First Bounce" talks about the credit wheel
Posted by: Too Much Fed | July 08, 2009 at 08:28 PM
It is the effect of interest rates on ultimate lenders' and ultimate borrowers' consumption/saving and investment decisions that matters. That's how interest rates affect the demand for newly produced goods and services. Banks merely intermediate between those ultimate lenders and borrowers.
Posted by: Nick Rowe | July 08, 2009 at 09:50 PM
Nick,
I am late getting back to you on this, but here goes. I believe it is reasonable to expect whether input or output prices are the most sticky depends on the type of the shock. With aggregate demand shocks I suspect output prices would be more sticky given the standard stories (menu costs, aggregate demand externalities, etc.). However,when it comes to positive productivity shocks these standard stories don't apply and firms would be more willing to change their output prices downwardly. As I mentioned above, with positive productivity shocks, firms have lower per unit costs of production and thus, they can cut output prices without harming profit margins.
If the above is true it lends support to my claim that Fed did push its policy rate below the neutral rate. For during the time in question, productivity did suddenly accelerate (i.e. there were positive productivity shocks). Here is another hastily drawn sketch of this development. Note, I have drawn a horizontal SRAS curve which indicates output prices to be the most sticky with respect to AD shocks. However, I let the SRAS drop in response to the positive productivity shock similar to the way I did it above.
Posted by: David Beckworth | July 10, 2009 at 11:13 AM
David B, I think you are getting closer to the truth.
Here is the question:
Should productivity shocks and/or cheap labor be met with debt (future demand) from interest rate cuts by the fed to prevent price deflation?
Posted by: Too Much Fed | July 16, 2009 at 12:42 AM
David B, imo your (2) from the sketch should say
"fed accomodates prod. shock and/or cheap labor by lowering interest rates and attempting to sucker the lower and middle class consumers further into debt"
If that is true, there are now excess corporate profits for the banks (they make money producing the debt) and stocks (they have Q growth, P growth, and margin expansion). Since the few mostly own the stocks and the banks, there is wealth/income inequality.
Posted by: Too Much Fed | July 16, 2009 at 01:06 AM
David: And I am even later getting back to your comment above! (The reason is I missed seeing your comment, what with all the spam).
If there is a positive productivity shock, in the face of menu costs for output prices, then the LRAS curve should shift right, exactly as you have drawn it, but I think the SRAS curve should shift horizontally right too, so the new SRAS curve will, if it's horizontal, lie on top of the old SRAS curve, so the SRAS curve would appear not to move.
There's a difference between "sticky" as in "not wanting to move", and "sticky" as in "not needing to move". I can understand how the type of shock (and the monetary policy response to the shock) might affect which prices are "sticky" in the second sense, but not in the first sense.
I mean, if you really do believe the menu cost story, for example, and that it applies to final output prices, then firms do not want to change those final output prices. The restaurant has a fall in costs, and would otherwise cut its prices, but it doesn't, because printing the new menus would cost too much.
Too much Fed: As I argued in my old "Hair of the Dog" post, if monetary policy cuts interest rates, borrowers will want to borrow more, but lenders will want to lend less. Real spending increases, and creates the increase in income to finance the increase in spending.
You must distinguish between the individual and the world. If the individual wants to spend more than his income, he can, by running down his assets or increasing his debt. But for the world as a whole, one person's spending is another person's income. If the capacity to produce goods increases (as in David's example), then a cut in interest rates leads people to want to spend more than their income, and income expands to match the higher capacity to produce income. So the extra spending is financed by extra income, not by debt. We can't borrow from Mars.
Posted by: Nick Rowe | July 16, 2009 at 02:55 AM
test
Posted by: Too Much Fed | July 17, 2009 at 05:35 PM
"and income expands to match the higher capacity to produce income."
Could you expand on that? At first glance that does not seem correct to me.
Posted by: Too Much Fed | July 18, 2009 at 01:28 AM
Too much Fed: Yes, absolutely. This is the key point behind many of our disagreements.
Let me give an example:
Assume one individual, working half-time and producing and selling $50,000 of goods per year. He wants to work full-time, and has the capacity to produce $100,000 of goods per year, but he only works half-time because he can only find buyers for $50,000. (He is 50% involuntarily unemployed). Assume he initially has no debt, and is spending $50,000 per year on consumption, so is neither saving nor dissaving.
Suppose our individual now decides to spend $60,000 on consumption each year. His income stays at $50,000, he dissaves $10,000, and goes steadily deeper into debt, and as the interest payments mount, his income net of interest gets steadily lower.
Suppose instead our individual now decides to spend $40,000 on consumption each year. His income stays at $50,000, he saves $10,000 per year, and accumulates assets, earns interest, and his income inclusive of interest slowly grows over time.
Now assume a closed economy with one million identical individuals, just like the one above.
Suppose every individual now decides to spend $60,000 on consumption. Each individual now finds that demand for his goods has increased, and he can sell $60,000 of goods each year. So he works more, and increases production and sales to $60,000. Which means his income from producing and selling goods is also now $60,000 also. So he is not dissaving, and not going into debt.
Suppose instead every individual now decides to spend $40,000 on consumption. Each individual now finds that demand for his goods has decreased, and he can only sell $40,000 of goods each year. So he works less, and decreases production and sales to $40,000. Which means his income from producing and selling goods is also now $40,000 also. So he is not saving, and not accumulating assets.
This is the "Paradox of Thrift". In certain special abnormal circumstances (unemployed resources and a demand-constrained economy) what is folly for the individual (consume beyond your income) becomes wise for the aggregate, and what is prudential for the individual (consume less than your income) becomes folly for the aggregate.
Posted by: Nick Rowe | July 18, 2009 at 08:42 AM
OK.
"Assume one individual, working half-time and producing and selling $50,000 of goods per year. He wants to work full-time, and has the capacity to produce $100,000 of goods per year, but he only works half-time because he can only find buyers for $50,000."
Can you add a corporation to that? The corporation wants to pay workers the same or less, maximize margins, increase quantities, and increase prices all to maximize corporate profits.
Posted by: Too Much Fed | July 18, 2009 at 11:34 PM
"Now assume a closed economy with one million identical individuals, just like the one above."
Can you add a private fed to that? It controls the "fungible" money supply and its composition (assume it has considerable influence over the amount of gov't debt).
At some point I might want two groups, 10,000 in one (1%) and 990,000 (99%) in another.
Posted by: Too Much Fed | July 18, 2009 at 11:53 PM