My preferred monetary stimulus
The commenter Benjamin suggested that I supplement my “moderate” proposal for Fed policy, with what I really believe. So here it is; I’ll just cut and paste from Bill Woolsey:
I favor some opportunistic disinflation from the Great Recession, shifting to a new, 3 percent target growth path for money expenditures, starting at the end of the Great Moderation, which I take to be the third quarter of 2008. The target for the second quarter 2010 would be $15.8 trillion, so the current value is 8.8 percent below target. The target for second quarter 2011 will be $16.3 trillion, so returning to target would require 13.4 percent growth in money expenditures [did he mean 12.4%?] over the next year. (That includes the already completed part of the year.) Of course, the targeted growth path of money expenditures would afterwards grow at 3 percent into the indefinite future.
My initial reaction was negative when I read Bill’s post, as I had just spent a lot of time criticizing opportunistic disinflation in a series of posts. But the more I thought about it, the more I liked it:
1. Getting back to the old 5% trend line is now a complete pipe-dream. Given that many contracts have now been signed reflecting the new world of low NGDP, it probably isn’t even wise any longer. After all, the Fed wasn’t promising level targeting of NGDP in 2008.
2. I’m a bit puzzled by the numbers Bill uses, but perhaps that’s because he uses nominal final sales instead of NGDP. I calculate the US as being about 5% below a 3% trend rate of NGDP growth, starting at 2008:2. Let’s split the difference and assume we are now 7% too low. In that case shoot for 10% NGDP growth next year, and 3% thereafter. Target NGDP futures contracts.
3. The negative associated with lower NGDP growth is the risk of liquidity traps. But under my ideal policy there would be no zero rate bound, as we’d be targeting NGDP futures prices. It would also eliminate the need for Josh Barro’s proposal to index taxes on capital, which is necessary if inflation is positive, but adds undesirable complexity to the tax code.
4. To avoid the “circularity problem” identified by Bernanke/Woodford, and also Garrison/White, you’d have to let the market actually determine the setting of the monetary base most likely to produce on-target NGDP growth. I described how in this post.
5. I have an open mind about the exact amount of catch-up, and whether it should be spread over one or two years. But the amount of catch-up should probably be in the 6% to 8% range.
Here’s what’s so clever about Bill’s idea. It gives us the more rapid recovery that people like me insist is possible and desirable, and yet it lowers long term inflation toward zero, a cherished goal of the Fed hawks. And it also eliminates any chance of future liquidity traps, which is what people like Paul Krugman worry so much about. What’s not to like?
PS: Just to be clear, this is my ideal proposal. I am not officially shifting from 5% to 3% NGDP targeting, as we don’t yet have an ideal NGDP futures targeting regime in place, which would prevent zero rate traps.
Tags: Circularity problem
3. September 2010 at 07:30
“Getting back to the old 5% trend line is now a complete pipe-dream.”
In July it was possible, in September it’s impossible. What’s changed?
3. September 2010 at 08:44
How does it ensure we keep our nose pointed towards 1% inflation? How much bigger does M2 have to get to have 10% NGDP? Over the medium term does 3% NGDP express the positive effect of the paying down of debts? Does the pump cause people to stop paying down debts, because they are being inflated away in the short term?
Assuming that all makes sense, can we now agree any newly printed money needs to touch the savers first? That our goal is embedding Creative Destruction into pump, rather than letting zombie banks, and dying companies live on another day.
3. September 2010 at 08:54
If however the Great Recession follows from 2000 on…
3% NDGP from $11T in 2003 puts us almost DEAD ON TREND.
If the entire housing debacle was a facade, we’re right where we need to be, we should stick to 3% NGDP.
3. September 2010 at 08:59
“Just to be clear, this is my ideal proposal.” Well, how “realistic” is your other proposal; perhaps half-and-half ideal/realistic? Do you also have a 100% realistic proposal (“Just do whatever you were going to do, anyway”)? How about a three-quarters ideal proposal? A one-quarter ideal proposal? A 17.3% ideal proposal?
It remains unclear to me what constraint you are imposing on yourself, to earn for yourself the prized epithet “pragmatist.” Might it be something like this: the bad old fundamentalists make proposals that have only a 0.0001% chance of being implemented; *you* make proposals that have a 0.01% chance? (Please revise those percentages as appropriate, or explain why I am completely off base.)
3. September 2010 at 09:00
Great post–the important thing is to keep talking about what monetary policy should be, in relatively concrete terms, and to get the other top economic bloggers (identified in recent Nunes link) discussing the same issues. When I have time, I will suggest to Scott Sumner some op-ed placements. (Scott, go to Amazon,com, see Benjamin Mark Cole, The Pied Pipers of Wall Street).
BTW I am no fan of zero inflation–see Japan. Think wage stickiness.
Consider real estate and equity investors, or small business start-ups, who have their “animal spirits” comforted if they think that 2-3 percent inflation is a backdrop.
If you borrow money to run a business or invest, zero inflation–dangerously close to deflation–is a dagger pointed at your financial heart.
I am not even sure what inflation means in world with rapidly evolving products and services, or rising incomes but fixed choice residential parcels. It may be that choice real estate has to increase in nominal and real value if incomes are rising, but that is reported as “inflation.” If rare oil paintings figured into the CPI,,,,,,,
And what is health care, or a cell phone, computer etc worth in 1950 dollars?
I have come to the conclusion that zero inflation is a dangerous utopian pipe dream.
3. September 2010 at 09:02
(You might *mention* a *bad old fundamentalist* proposal, and say it would work, but you wouldn’t *propose* it. Well, why is this self-restraint supposed to be important?)
3. September 2010 at 09:53
MW, It’s not impossible, but it is a pipedream. Won’t happen. And probably shouldn’t happen. The recession is almost 3 years old–with lots of contracts negotiated on the basis of reduced NGDP expectation.
The economy is even worse off than in July. Inflation expectations have fallen significantly.
Morgan, How should I know how much M2 we need? That’s for the market to decide.
Philo, The pragmatic proposal uses traditional Fed procedures and tools, and sticks to their traditional 2% inflation goal. That’s why they should be willing to consider it. It’s the one I pull out if they ever ask me to give a seminar there.
Not quite sure what the percentages are supposed to measure. It’s unlikely anything I propose is adopted 100% EXACTLY as I propose it. Yet it is very likely that they do at least a bit of what I propose.
This one uses unconventional tools, and goes away from inflation targeting entirely. It also lets the market run monetary policy, which I would think is a pretty radical suggestion. So I basically don’t follow where you are going with the question.
Benjamin, You said;
“BTW I am no fan of zero inflation-see Japan. Think wage stickiness.”
That’s a good point, but in terms of the GDP deflator, they’ve averaged about 1% deflation. And their NGDP growth has been near zero. So my proposal is much more expansionary than theirs, even taking into account the fact that US population growth is 1% higher.
Still, the wage stickiness argument suggests that 3% might be a bit low. Even so, this is all the least of our problems. It’s not the average rate that’s the problem, it’s the extreme instability.
You said;
“And what is health care, or a cell phone, computer etc worth in 1950 dollars?
I have come to the conclusion that zero inflation is a dangerous utopian pipe dream.”
Bingo! Which is exactly why I favor NGDP targeting.
3. September 2010 at 10:01
You believe in downward-sticky nominal wages, don’t you? You’re essentially advocating zero inflation (conditional on the use of NGDP futures, 3% RGDP growth as expected, and so on). With zero inflation, given the cross-sectional variation, workers in many markets will be facing downward wage pressure at any given time. If wages are downward-sticky, this would result in inefficiently high wages in those sectors and unnecessary unemployment. (This is the standard “greasing the wheels” argument associated with Tobin, Akerlof and others.)
I would go the other way and advocate even higher NGDP growth — maybe 7%. In my ideal, this would be combined with tax indexation along the lines of Josh Barro’s proposal. The extra grease on the wheels in the labor market more than compensates for the slight increase in the complexity of the tax system.
Also, I’m skeptical about the ability of the NGDP futures targeting proposal to fully eliminate problems associated with liquidity traps. All in all, I think futures targeting is a good idea, but it has the potential to create volatility under liquidity trap conditions, because the targeted path for NGDP may not be a feasible path when the equilibrium real interest rate is negative. See Adam P, Canucks Anonymous on that point. It’s not a problem if the targeted growth rate is high enough that you never get near a liquidity trap.
3. September 2010 at 10:37
Stop the wage stickiness crap, seriously.
It’s JAPAN – they get a job they work there forever.
Scott, I made a very clear point elsewhere:
Every month we create 4.5M jobs – you need to STOP fighting a rear facing action worrying about propping up the failing businesses.
If set policy to benefit the productive winners, we’ll see an increase in job turn over… we want to lose an extra 1M jobs and gain an extra 1.2M.
Velocity of job turnover ENDS the last vestiges of wage stickiness. Let’s target 6M+ plus new jobs and 5.7M job losses.
3. September 2010 at 11:03
Andy, I am skeptical of any indexing regime that forced me to keep receipts of every single time I pull money out of an ATM, and then calculate the real value of each deposit and each withdrawal. That seems like a complete nightmare to me. Josh said you can always use a tax preparer. I did that this year for the first time in my life, and it cost $610. How much would it cost if every single financial transaction had to be adjusted for monthly changes in the price level.
Actually, the first best policy is to eliminate all taxes on capital, and go with a progressive consumption tax. I know of no good arguments against that idea.
But let’s say we do index investment income. Then I favor a higher rate of growth in NGDP, for exactly the wage stickiness reasons you outline. Indeed your idea is defensible even without indexing.
I am less impressed by the anti-NGDP targeting post you link to. He seems to assume that the only way monetary stimulus boosts AD is by lowering the real rate of interest. But Robert King showed in 1993 that in a dynamic IS-LM model with ratex it is quite possible that monetary stimulus actually boosts real interest rates.
Mishkin’s textbook lists 10 different channels for monetary policy, above and beyond nominal interest rates. Some involve real rates, but others don’t. And I can think of other channels he missed.
The post is also vague on the nature of the problem. Is he claiming the Fed can’t peg NGDP futures prices? Is he claiming that future prices might differ from market expectations? Or that market expectations might be irrational? I can’t tell what the argument is, other than some sort of vague claim that things might get “volatile.” Yes, but in theory that could occur with money supply targeting or interest rate targeting or exchange rate targeting or any kind of targeting.
Japan has had roughly 0% annual NGDP growth since 1994, so I don’t see it as comparable to the 3% annual growth I propose. The wage stickiness argument is a good one, but the argument against NGDP future targeting makes no sense to me. The worst case that could occur if the equilibrium real rate fell to more than negative 3%, is the Fed would have to supply a lot of base money to satiate the public’s demand for cash and similar assets. Maybe they’d have to buy up all the T-bills, that doesn’t concern me at all. There will always be plenty of other types of assets out there.
One other problem with his “excluded middle” argument. It seems to assume growth rate targeting, whereas I am proposing level targeting.
Morgan, Who should target jobs? The central planner?
3. September 2010 at 11:27
“”I have come to the conclusion that zero inflation is a dangerous utopian pipe dream.””
“Bingo! Which is exactly why I favor NGDP targeting”
Scott,
Why is zero inflation a dangerous pipe dream? I can see why that trying to push inflation down might not be a great idea at the present time. But why, in 2015 with unemployment at 5.5%, would it be bad for the Fed to say “starting in 2019, we will target a broad price index which includes both consumer and asset prices and try to keep it in the range of 99.5 to 100.5”? Perhaps they can even target expectations for the price of future contracts in the index, so that expectations are always ~100.0 one year out?
I grant the nominal rigidity/sticky wage problem, but there are probably some offsetting benefits of zero inflation.
Take COLAs, either contractual or de facto. I know a lot of people received 3% (or 4%) raises at the end of 2007 / early 2008, right before the sharp downturn which certainly wasn’t fully offset with monetary policy. In a zero inflation regime, these COLA-type raises wouldn’t be around and the Fed would be committed to not letting the price level fall below 99.5.
If some measure of asset prices are also included in the index, it seems it would be more difficult for credit/asset bubbles to form. With fewer (hopefully zero) bubbles being blown, the need for post-bust heroic efforts by monetary policymakers should be signficantly reduced.
3. September 2010 at 11:49
The problem of sticky wages is very, very important, so you’ll have to come up with something better than that to justify zero inflation.
In a zero inflation regime, unexpected deflation would have exactly the same impact as the unexpected disinflation you mention – people would receive whatever % raises in real wages. And then we’d be stuck with the same problem. Zero inflation doesn’t solve the issue you mention.
3. September 2010 at 11:57
Scott,
I’m deeply skeptical that inflation rates below 1.5%-2% are optimal.
Akerlof, Dickens and Perry’s (1996) study on the US economy suggests the existence of an optimal non-zero inflation rate, between 1.5% and 4% because a moderate rise in the price level provides some “grease” to the price and wage setting process. Wyplosz (2001)found that unemployment seems to decline for rates of inflation above 2%. Due to nominal rigidities, Dickens (2001) argues that an inflation target below 2% could be very costly. etc. etc.
And what about Japan? Sure Japan has actual deflation, but doesn’t their experience suggest that price stability might also come at a cost, not just in the short run, but over the long run?
You wrote:
“Actually, the first best policy is to eliminate all taxes on capital, and go with a progressive consumption tax. I know of no good arguments against that idea.”
Unpublished cross sectional panel analysis by an unknown author named Sadowski suggests that consumption taxes are moderately and significantly growth enhancing. Theoretically progressive consumption taxes should be even more so.
My first best would be a 5% NGDP level target combined with a progressive consumption tax.
P.S. I really like Mishkin’s chart summarizing transmission mechanisms (Figure 3, Chapter 26, 7th edition). It’s a great teaching device.
3. September 2010 at 13:01
Pipe dream? Doesn’t monetary policy have a tendency to radically shift overnight?
Current track record of central bank ability to handle recessions suggests that we need to aim at 5% inflation, 2% inflation is bad but not yet disastrous, and 0% inflation is a Japanese style economic seppuku.
Hypothetically NGDP targeting might enable 0% inflation, but how willing are you to bet the entire world economy on something as untested? It’s dubious if it was net benefit or not even if it worked (wage and price stickiness vs shoe leather cost vs etc.), costs of failure is severe, and risk that it will fail is far too high.
3. September 2010 at 16:12
No Scott, the POLS. Tax cuts should be targeted towards increasing the turn-over of jobs.
So tax cuts that help crush dying companies – GOOD. Tax cuts that all companies benefit from – BAD.
The point is, you are wrong that the “best kind of fiscal stimulus” is a payroll tax, now to be right, you’ll need to explain why stringing along zombie companies is a good idea.
Are you really comprehending my point here?
4.5M+ jobs created every month. About the same give or take are lost.
You get that right?
My point is if the tax policy is weighted towards Creative Destruction, we get a higher rates of new jobs, and lost jobs.
Meaning gain 5.6M+, maybe only lose 5.4M (+200K). Gain 7M, lose 6.7M (+300K).
This on its face makes lots of sense. Newco’s do most of the hiring, dying companies rent seek and string themselves out death. Profitable companies find productivity gains, increase savings, increasing capital formation….
The point is you approach stimulus, like you are trying to stave off the bleeding – you are worrying about ALL companies equally.
You should be advocating tax policy, QE, etc. that builds in a fiduciary responsibility that favors profit makers, and disfavors money losers.
3. September 2010 at 17:18
Justin, Your question answers itself. We don’t know what inflation is, so why should we target it? The government CPI shows the price of housing rising between 2008 and 2009, the worst housing downturn in history. Indeed not just in nominal terms, but in real terms, as they had housing inflation exceeding the overall inflation rate. I don’t trust CPI numbers because they are pulled out of a hat.
And adding asset prices wouldn’t help. What if goods prices fell but the stock market rose? The Fed might do nothing, as the economy went into recession. There is precedent for rising stocks during a recession, it happened in 1991.
David, I agree, which is why I favor NGDP targeting.
Mark, I agree with everything you say. But the Japanese experience needs to be balanced against the US in 1922-29, when inflation was roughly zero. In the latter case NGDP growth was closer to 3%, whereas in Japan it was closer to 0%. Those 3% make a huge difference.
But it’s a moot point, as we aren’t going to have futures targeting anytime soon, so I still favor a 5% NGDP target fan. Indeed even Andy’s recommendation of 7% might be better, if the Fed doesn’t shape up. That’s closer to the Australian rate.
Tomasz, I hope you are right, but right now it seems a long shot to even convince the Fed to shoot for the 2% that is supposedly their target.
Morgan, You said;
“The point is, you are wrong that the “best kind of fiscal stimulus” is a payroll tax, now to be right, you’ll need to explain why stringing along zombie companies is a good idea.
Are you really comprehending my point here?”
No. Who mentioned saving zombie companies? I want to boost AD and AS.
3. September 2010 at 23:07
“No. Who mentioned saving zombie companies? I want to boost AD and AS.”
Scott, I want to boost productivity because I’m whip smart (you can be too!) and can delay gratification and seek the the correct first principles.
Productivity as our first order, MAXIMIZES of quality of life over the long term.
It does this because it makes things cheaper, and it makes them cheaper FAST.
I challenge you to deny that the incredible compression of reduction in price of technology has not done more for man (utility), than anything else since 1900.
The pace of improvement is mind boggling. And this happens in spite of government, not because of it.
Nothing is going to significantly pay our average worker more. It is mathematically impossible.
But we can make sure, that in 5 years, 10 years, what he can buy for his static pay is the maximum basket of goods achievable.
—–
OK then, AD/AS isn’t worth a bucket of spit. Because focusing on it, will often hurt productivity in the long term, and this is unacceptable.
Scott you should be smart enough to understand this: behaving as if we live on a single global currency CAN NEVER hurt our people in the long term. Imagining ourselves on a single global currency we cannot print more of – INSTANTLY we know how to behave: save, limit, export, MAKE QUALITY THINGS CHEAPER, let the world’s best minds live here, and sell them their morning coffee, cut their hair, sell them new expensive gadgets, so they get cheap quick.
If you are going to mention AD again Scott – I think you first need to explain clearly why you think – when in conflict – it trumps productivity, and why?
4. September 2010 at 11:14
Morgan, You said;
“Scott, I want to boost productivity because I’m whip smart (you can be too!)”
I despair at the thought that I will never be as whip smart as you.
All of your post describes long run forces, and I agree productivity is the key in the long run. But over the business cycle AD is what counts.
4. September 2010 at 13:50
“We don’t know what inflation is, so why should we target it? The government CPI shows the price of housing rising between 2008 and 2009, the worst housing downturn in history. Indeed not just in nominal terms, but in real terms, as they had housing inflation exceeding the overall inflation rate. I don’t trust CPI numbers because they are pulled out of a hat.”
Scott,
NGDP includes various imputations and implied transactions and is constantly revised, both from a data perspective and a composition perspective (adding software or R&D as investment, for example). There can also be debate as to which aggregate expenditure measure is better (NGDP, nominal final sales, etc). But the important point, as you have noted, is to stabilize expectations. Inperfections in how NGDP/NFS is defined and calculated are far less important than ensuring people expect NGDP/NFS to follow a moderate growth path, with any pertubations quickly offset by monetary policy action.
Don’t get me wrong, I think an NGDP futures targeting regime is a great idea and a huge improvement over the status quo (I’m in the 3% camp but I’d accept 5% too). From a nominal rigidities/AD perspective, I’ll allow that NGDP targeting offers advantages over targeting a 0% inflation path, especially starting today/near future. However, I don’t see how it would limit asset/credit bubbles from forming, which would cause real pain regardless of NGDP targeting (unemployment rose quite a bit before the 2008 nominal shock occurred). My particular area of disagreement is with the claim “zero inflation price level targeting is dangerous or foolish”. At worst, it may just be sub-optimal relative to a 3% NGDP price level expectations target.
I agree with you that inflation means different things to different people, and different baskets of goods/services/assets will show somewhat different rates of inflation. I’m not sure that means we can’t build a price index and then use futures targeting as you’ve described here to keep expectatations that index – and more importantly as a consequence expectations for the purchasing power of dollar – very well anchored.
Several potential guiding principles:
– Seek to end all but minor and transient changes in the cost of living as perceived by median income households, both over a given year and over generations (no more grandparents telling young kids that such and such used to cost $X.XX back in the day but now it costs $XX.XX or $XXX.XX).
– Create a basket of goods/services/assets which is based on what a typical median income family buys, set the base year to 100.0 and direct the Fed to target futures so that expectations for this index are 100.0 one year out.
– No non-cash imputations of whole categories (owners equivalent rent), hedonic adjustments, geometric weighting. This would be a price index, not a consumer satisfaction index.
– Include asset prices (such as home prices rather than OER). While not all rises in asset prices are bubbles, a monetary policy regime which targets asset prices should limit the occurence/severity of bubbles and hopefully asset price volatility as well. I can’t imagine the 2000s housing boom/bust occuring under a monetary policy regime in which house prices themselves are in the index.
I just can’t see how keeping expectations of such a price index stable would be disastrous. While less responsive to a sharp negative AD shock than NGDP targeting, I’d expect targeting a flat price level (including asset prices) to reduce the risk of a sharp negative AD shock as well.
5. September 2010 at 08:11
Justin, NGDP is far easier to calculate than prices.
I agree that housing distorts both indices. But housing is 39% of the core CPI, and a vastly smaller percentage of NGDP.
Yes, we could target the CPI, but why would we want to when NGDP is so much better? And I don’t think it is as easy to stop doing hedonics as you assume. What is “a computer”?
I don’t believe bubbles exist, so I don’t want the Fed trying to spot them, and then prevent them. They are likely to get it wrong, as in 1929, the last time they tried to pop a bubble.
9. September 2010 at 01:21
“MW, It’s not impossible, but it is a pipedream. Won’t happen. And probably shouldn’t happen. The recession is almost 3 years old-with lots of contracts negotiated on the basis of reduced NGDP expectation.”
And all these contracts were signed while you were on vacation?
9. September 2010 at 07:18
MW, A change in views from A to B is always a discreet event. It will be based on new information. If a few weeks worth of new information was never enough to change our policy views, then a few centuries also would not be enough. Isn’t this Xeno’s paradox?
By the way, even in July it was probably too late to return to the old trend line. If I said otherwise, it was probably a slip ,or I was oversimplifying to make a point. If we have level targeting, then we should always return to the trend line. If we don’t (and we don’t) then it is debateable as to what point you let bygones be bygones.
9. September 2010 at 09:12
Thanks for the clarification. I’m simply trying to understand your argument(s).
As previously, the evidence on ‘regaining the trend’ after a financial crisis and/or a housing crisis is mixed, but suggests the trend is NOT regained.
10. September 2010 at 13:28
In most financial crises, the currency falls sharply. In this case the dollar soared in value during the worst of the crisis. So I see this as a severe AD shock, which countries do tend to recover from, not primarily a financial crisis such as R&R studied, which countries do not tend to recover from.
13. September 2010 at 06:48
That’s a good point, but the US is a relatively closed economy and exports are a small share of GDP. Furthermore, to the extent that US exports are in high value-added goods/services (and I do realize the US is a large ags exporter), the exchange rate is much less important than trading partner growth (which is easily demonstrated, empirically).
13. September 2010 at 18:31
MW, That wasn’t my point. I know the trade effects are small. My point was that it was a sign that money was tight–an indicator. This is quite different from most financial crises.