Yesterday’s post on deflation prompted a flurry of comments and emails remarking on how much economists disagree. This, I think, misses the point. Indeed, what I was trying to emphasize was that we all agree on the advantages of deflation as spelled out in Milton Friedman’s essay on The Optimum Quantity of Money. (I’ve put a quick summary of the key points here.) Therefore, the commentators who are currently worried about deflation must fall into two categories: First, there are the ignoramuses, of whom there are plenty on all sides of all issues. Second, there are those who are clearly not ignoramuses. Those in the latter camp have surely digested Friedman’s analysis, and understand the upside of deflation, but believe it is outweighed by some downside. It is frustrating to me that many of those commentators have failed to explain exactly what downside they have in mind.
I’ll say this again: We agree about the upside of deflation. We might or might not agree about the downside. I can’t tell because I can’t figure out what downside these people are talking about. In this regard, it is instructive to read the comments on yesterday’s post by Sandy, who clearly understands these issues extremely well. Following an incisive and articulate summary of some key points, he’s left with this conclusion:
I presume that what commentators are worried about is a deflation that represents a long, drawn-out period of adjustment in response to a severe contraction of money and credit.
In other words: a) The commentators in question have left us in the position of having to “presume” that we know what they’re worried about, as opposed to, say, telling us. And b) Our best presumption is that it has something to do with a “long drawn-out period of adjustment”, though the causes and nature of this adjustment period are vague and unspecified.
Edited to add: After I wrote this, but before I posted it, our commenter Steve Reilly pointed me to this post by Paul Krugman, which I had not seen. Here Krugman spells out three specific concerns about deflation. The first (“people become less willing to borrow”) confuses me, because I can’t figure out what’s being held fixed here; by itself, deflation should make people more willing to hold money and hence less willing to lend. In any event, Krugman indicates that this first concern kicks in only when the nominal interest rate is trying and failing to go below zero, whereas the Friedman prescription is for a deflation rate that pushes the nominal interest rate to zero and no further.
Krugman’s primary concern seems to be nominal wage rigidity—the prospect that wages will not be able to fall fast enough to keep pace with deflation. I am probably more optimistic than Paul about employers’ resourcefulness when it comes to cutting wages. But this does count as a clear and legitimate concern.
Krugman says people are “less willing to borrow” because “the loan will have to repaid in dollars that are worth more than the dollars you borrowed”.
I suppose he means something like this: Suppose under normal circumstances you could borrow $100 to build a factory, and then at the end of the year, sell 30 widgets at $4 apiece and gross $120. You should be willing to borrow at up to 20% interest to make the investment.
But now suppose everyone expects prices to fall by 50% in the next year. You still have to borrow $100, but at the end of the year your 30 widgets only bring in $2 apiece for a total of $60, so you don’t take out the loan.
Steve you said deflation should make people less willing to lend, but are you sure? Certainly it should make them less willing to *spend*, because you can get more for your money if you wait. But you don’t have to actually have the money in your possession. You lend it out, get repaid with interest a year later, and now you can buy more with your money just like you planned all along. Am I missing something?
In other words there could be two different meanings to “holding money” — “holding” it as opposed to spending it, and “holding” it as opposed to lending it out. It seems deflation could encourage “holding” in the first sense but not the second one.
As above, but in two words, discourages investment.
Surely the potential for a deflationary spiral is enough of a reason to avoid it?
Why are we concerned about about there being too little money holding? What are the benefits of people lending less money and holding more? I understand that it costs money holders more than it costs society, but why do we care? People who punch themselves in the groin repeatedly bear more costs than society – do we get too little self-inflicted groin punches?
I rolled my eyes at your post yesterday on deflation when it didn’t make mention of the increased burden on the indebted (as what they need to repay is less easily repaid with their falling incomes). It seems to me that in America, a country with particularly bad problems with respect to its public and private debt levels, deflation is a very unwelcome thing whatever its other merits.
Well I don’t think you have even addressed what I take to be the “traditional” concern about deflation: that a RAPID significant deflation bollixes up the normal flow of credit. Overnight we see 50% deflation. All demand loans will be called in; few loans will be rolled over. Businesses will close. In a fractional reserve system this can worsen. Can markets clear fast enough in response? Not always.
Personally I worry more about inflation in the next few years. But I have not seen you address this question: can the speed at which the money supply can contract (or expand) exceed the speed at which markets can absorb the effect?
Pat:
People who punch themselves in the groin repeatedly bear more costs than society – do we get too little self-inflicted groin punches?
Why would you say this? People who punch themselves in the groin bear all of the costs, so I expect we get pretty much the right number of self-inflicted groin punches.
You are, I think, overlooking the fact that the self-punchers are part of society, so their private costs are included in the social costs. In the case of moneyholding, the private cost to you of holding money is included in the social cost, but so is the offsetting (negative) cost imposed on others via the change in value of their money. In the case of groin-punching, I think you’ll be hard pressed to find an analogue to this.
Steve, you’re right – that’s exactly what I was overlooking.
Pat: Let me add one more thought to this. Again, you said:
People who punch themselves in the groin repeatedly bear more costs than society – do we get too little self-inflicted groin punches?
Your assumption is: People who punch themselves in the groin repeatedly bear more costs than society. The only plausible way this assumption could hold is if people really enjoy seeing other people punch themselves in the groin (so that the social cost, which equals your pain minus my enjoyment, is less than the private cost, which is your pain alone). In that case, yes, we absolutely get too few self-inflicted groin punches.
Pat: I posted my “one more thought” comment before seeing your lastest brief comment (which appeared pretty much simultaneously). Sorry for the overkill.
Perhaps I am missing something but – in response to trot’s comment on deflation hurting the indebted – doesn’t this depend on what kind of deflation? If deflation is due to increased production, why wouldn’t income levels remain steady (because companies have increased profits but can charge less for products)?
Isn’t that how standards of living rise – steady wages/income – but gradual deflation?
Then these income earners can spend less on food and MORE on paying off debt. Deflation might even be a good thing for the indebted.
Feel free to correct me where I might be wrong.
Steve: Trot is certainly correct that an unexpected deflation is bad for people who owe money (assuming the debts are denominated in dollars, as opposed to, say, apples). Offsetting this, of course, is that the same deflation is good for lenders. (Though Trot, I think, would point out that if the average American is currently a borrower, then there’s a net loss to Americans here.)
Increased production is of course a good thing. And yes, given a constant supply of money, steadily increasing production leads (probably) to a steadily increasing demand for money and hence to deflation. But first, that kind of deflation is unlikely to be unanticipated, and second, even if it were, and even if we didn’t like the consequences, there’s an easy fix, which is to increase the *supply* of money at the same rate as the demand.
I wouldn’t think that deflation would be uniformly spread across the economy. For example, over the last 20 years, we’ve had steady inflation (well, not completely steady, the rate of inflation varies), and yet the prices of some products has declined. I can get a cheap laptop now for around $200 that is more powerful than one I could get 10 years ago for $2000. The same is true for mobile phones – 5 years ago I paid $100 for a phone, today I can get that same phone basically for free.
So when we talk about a deflationary spiral, wouldn’t some products continue to appreciate in price even while many products deflate? And if so, wouldn’t consumers purchase the inflating products, which they can get now more cheaply than in the future, and defer purchasing products that are declining in price? I would think deflation would skew the economy toward certain products and away from others.
@Steve, in yesterday’s column you wrote “But the cost to society of your holding a dollar is zero, because it costs us essentially zero to print that dollar up for you.” Are you arguing that when an economy enters a deflationary trap, that the government should print money to keep the economy moving? That seems reasonable, as the normal cost of printing money is inflation, but if there is little risk of inflation printing money is, well, free money. Has Japan printed money during its lost decades?
Let’s say you’re a central banker or treasury department official who wants to supply the right amount for society. The right amount accordng to you is the “demand.” How is this measured? The quantity demanded will depend on quantity supplied it seems.
Slightly OT, perhaps, but Friedman’s deflation prescription arises as the only way of “paying interest” on money. On currency that is the case, but currency is only a small part (less than 10%) of a broader definition of money, say MZM (money zero maturity). And much of MZM is (or, was) interest bearing.
Neil:
Friedman’s deflation prescription arises as the only way of “paying interest” on money. On currency that is the case, but currency is only a small part (less than 10%) of a broader definition of money, say MZM (money zero maturity). And much of MZM is (or, was) interest bearing.
Point extremely well taken.
Josh: To maintain a steady price level, you’d want to increase supply at the same rate at which demand is increasing.
Al V: So when we talk about a deflationary spiral, wouldn’t some products continue to appreciate in price even while many products deflate?
Relative prices change all the time, but this has nothing to do with deflation or inflation, which is the rate of change of the price level. If a laptop used to cost you a week’s wages and now costs you a day’s wages, you’ll now be more likely to buy a laptop, independent of what the price level is doing. So I don’t see the basis for your statement that “I would think deflation would skew the economy toward certain products and away from others.”
Are you arguing that when an economy enters a deflationary trap, that the government should print money to keep the economy moving?
No.
Wage-stickiness, caused by various psychological and sociological tendencies, is often offered as a contributor to involuntary unemployment. Do you have an alternative explanation for involuntary unemployment?
It seems to me that the first concern Krugman points out is that deflation leads people to expect more deflation which causes them to spend too little. Less spending, aka less consumption & less trade, means less wealth is created, and keeps the economy depressed. So, it’s not just the deflation itself, he’s suggesting – even if there’s a level of deflation that’s otherwise optimal, people’s expectation of more deflation can lead to a spiral of less and less consumption.
> In any event, Krugman indicates that this first concern kicks in only when the nominal interest rate is trying and failing to go below zero, whereas the Friedman prescription is for a deflation rate that pushes the nominal interest rate to zero and no further. <
But how do you prevent the deflation rate from going past that point?
Secondly, you write:
> Krugman’s primary concern seems to be nominal wage rigidity—the prospect that wages will not be able to fall fast enough to keep pace with deflation. I am probably more optimistic than Paul about employers’ resourcefulness when it comes to cutting wages. But this does count as a clear and legitimate concern. What this means is that in general economies don’t manage to have falling wages unless they also have mass unemployment, so that workers are desperate enough to accept those wage declines. See Estonia and Latvia, cases of. <
So what he's saying is that yes, employers can cut wages: by laying off employees. For wages to go down to keep pace with deflation, we have to create more unemployment, and this is another reason why deflation is bad.
You're suggesting that employers' resourcefulness would not only allow them to cut wages without creating more unemployment, but that they'd actually do so – that is, other solutions (besides layoffs) would be more attractive to employers. Why do you think so? What do you think these other solutions might be?
Ugh, I’m sorry, you’re blog software swallowed part of my text and turned it into invalid HTML tags. I’ll try that last comment again (and feel free to delete the earlier duplicate)…
Secondly, you write:
>> Krugman’s primary concern seems to be nominal wage rigidity—the prospect that wages will not be able to fall fast enough to keep pace with deflation. I am probably more optimistic than Paul about employers’ resourcefulness when it comes to cutting wages. But this does count as a clear and legitimate concern.
Krugman writes:
>> What this means is that in general economies don’t manage to have falling wages unless they also have mass unemployment, so that workers are desperate enough to accept those wage declines. See Estonia and Latvia, cases of.
So what he’s saying is that yes, employers can cut wages: by laying off employees. For wages to go down to keep pace with deflation, we have to create more unemployment, and this is another reason why deflation is bad.
You’re suggesting that employers’ resourcefulness would not only allow them to cut wages without creating more unemployment, but that they’d actually do so – that is, other solutions (besides layoffs) would be more attractive to employers. Why do you think so? What do you think these other solutions might be?
Cos:
But how do you prevent the deflation rate from going past that point?
By decreasing the money supply at the desired rate and no faster.
Cos:
You’re suggesting that employers’ resourcefulness would not only allow them to cut wages without creating more unemployment, but that they’d actually do so – that is, other solutions (besides layoffs) would be more attractive to employers. Why do you think so? What do you think these other solutions might be?
Since workers’ productivity hasn’t changed, and since the employers wanted them around before, they’ll still want them around. If workers won’t accept (nominal) wage cuts, I expect employers to get pretty creative about making them understand that the choice is between those wage cuts and unemployment. If that doesn’t work, there are all sorts of other ways to cut wages in effect without doing so explicitly—you cut fringe benefits, coffee breaks, on the job training, etc.
The fear of deflation is that everybody wants to get money, nobody wants to spend it, and the economy stops.
Your proposed solution, govt. buying things with new money, would keep the econ. moving AND prevent deflation. Therefore by assuming deflation you’ve also assumed that the govt is NOT implementing your solution — which is pretty close to what is happening in reality, and is scary.
As an aside, I really hope the govt never spends newly printed money, which is the ultimate form of fiscal policy. This is too politically addictive and if it ever gets started I can’t picture any politician stopping it — you think our guys are any better than Zimbabweans? Better to do the same thing by simply increasing the money supply with QE.
Doesn’t Friedman’s essay implicitly assume that the natural real interest rate is positive? Given that we live in what is mostly a service economy, and most services are not storable, and given that most investments are risky and therefore require a risk premium over and above the natural risk-free interest rate, is there any reason to make the assumption that that rate is positive? Given that the actual short-term nominal interest rate is near zero, even though the expected inflation rate is still positive, and given that we have historically high levels of unemployment in this circumstance, doesn’t this suggest that the assumption is in fact wrong when applied to the present situation?
In other words, the case where “the nominal interest rate is trying and failing to go below zero” is one that can happen even without deflation, one that is apparently already happening, and one that will be exacerbated if actual deflation arises.
Didn’t Milton Friedman argue that the Great Contraction was a stock market panic exacerbated but tight money?
I read The Optimum Quantity of Money decades ago (for an Econ class while working on my MBA), so I may not be remembering it correctly. One of his arguments was that with a low (i.e. negative) inflation rate, people would hold more money and this would result in greater total wealth for society. My reaction at the time was that that wouldn’t accomplish anything: printing up a bunch of money and persuading people to hold it wouldn’t increase society’s wealth. I remember from Finance class that while individual portfolios may be adjusted arbitrarily, for society as a whole financial assets must equal financial liabilities and read assets must equal net worth. Money is its own liability: adding more of it to the economy just adds an equal amount to the financial assets and financial liabilities, leaving net worth unchanged.
The only way that the wealth of society can be increased is by increasing the real assets. Changing the holdings of financial assets doesn’t have any effect. Do you agree with this analysis?
Tom Mahon: What you are overlooking is that people’s lives are more convenient when they hold more money (fewer trips to the ATM, fewer discoveries that you don’t have the cash on you for something you want to buy, etc.) That extra convenience has value, and it is created at no social cost.
jj, yes, the current reality is scary. The problem is that QE, as shown in Japan, does nothing. Banks don’t lend without creditworthy borrowers. As lending is not constrained by bank reserves, making them bigger doesn’t do anything. The Fed, even headed by a Zimbabwean, has less power than people think. The only thing that can get us and the rest of the world out of deflation/depression is “printing money” (whether it is “borrowed” – i.e. whether bonds are sold, is secondary) to support fiscal policy; the most urgent one being to immediately employ the unemployed. Steve’s example to fund government operations like Social Security and Medicare is very good.
“This is too politically addictive and if it ever gets started I can’t picture any politician stopping it” Deficit spending is necessary to a modern economy. It’s what got the US out of the Great Depression – the New Deal and the war. The nations that hadn’t forgotten Keynes e.g. Australia, China and Southeast Asia did much better than the Austerians of Europe or half-hearted stimulators like the USA. The main problem now is that the politicians refuse to spend enough, rather than too much, out of misguided fears of imaginary fiscal problems.
I will say this cash seems to me to be more efficient than debt. If I borrow money to buy a house or car, the bank and I need to do more work.
Your deflation posts confuse two different issues – unexpected deflationary AD shock and long term steady and expected deflation. Milton Friedman spent all his life advising us about the dangers of the deflationary AD shocks, while also explaining about the long term benefits of deflation.
Start with ratex and Bob Lucas. When central bank unexpectedly departs from the long term trend of nominal GDP (aka “Great Moderation), there is lot of damage as labour and debt contracts were not optimized for such a shock. If you advocate deflation, you should also think about the optimal speed for such a transition so people can adjust to your proposed reforms.
Money Demand Blog: I agree with you that it’s important not to confuse these issues, but I’m not sure why you think I’ve confused them. I was addressing those pundits who are warning about the prospect of long-term, ongoing (and presumably fully anticipated) deflation, and I think I confined my comments to that issue.
No. Most of the pundits are objecting to the adjustment costs of unexpected deflation. For example take Krugman, Delong etc. Their Neo-Keynesian models are illustrating the adjustment costs to unexpected deflation that was caused by low AD, these models have nothing to do with long-term ongoing fully anticipated deflation.
Only once has Paul Krugman mentioned the obscure research by Fortin :
“What this literature argues is that, probably due to bounded rationality, there’s some downward inflexibility in prices and wages even after expectations have had time to fully adjust.” Validity of this research is controversial and it is very rarely brought up when discussing deflation.
How can you say that deflation is fully anticipated, if most mortgages that are currently valid were signed before 2009? The same story applies to labour market, Scott Sumner (a libertarian leaning macroeconomist) has more on labour market maladjustment in his excellent blog.
Money Demand Blog: But Krugman et. al. are warning us of the consequences of a deflation that they currently foresee and that they worry will last a decade or more. How can that be unforeseen and/or short-term?
Krugman’s worry about a 10 year deflation is certainly not a market forecast. The price of treasury inflation protected securities indicates that markets are expecting 1.8 inflation for the next decade. Krugman is worried about a new AD shock, with markets converging on his pessimistic scenario. If Krugman is right and a new AD shock arrives and drives 10 inflation expectations below zero, we will get a new crisis. In the short term (i.e. for the next four years) his New-Keynesian models will give you a rough answer about the size of the real GDP drop. In long run (for the next five-ten years) we will get an excellent natural experiment and we will find out if Fortin was right about bounded rationality and long-term non-adjustment.