 |
 |
 |
I am not well versed in Keynesian business cycle theory. Therefore I have a very naive question for the Keynesian economists:
Here’s why I ask: According to what I take to be an orthodox Keynesian view, we are now in a liquidity trap. (My question does not apply to Keynesians, new or old, who believe otherwise.) That means that people want to hold lots and lots of money instead of spending it. Cool! We can provide money at almost zero cost. So it should be easy to make people very happy. What’s the problem?
Of course, people are working less, but that makes perfectly good sense in a world where people prefer to consume less. Why spend all day on an assembly line churning out widgets that people prefer not to buy?
A quick and obvious answer is that the people who are choosing to accumulate money and the people who are out of work are not the same people. In other words, to put this in slightly more technical language, you can’t address this question in a so-called “representative agent model” — a model that abstracts from interpersonal differences.
Still: The theory, as I understand it, is that vast numbers of people are choosing to hold vast amounts of money. Since money can be produced costlessly, this ought to count as a very good thing — which should offset a lot of very bad things, no?
Whatever answer there is might vary from one Keynesian economist to another, so let me subdivide my question into two:
- Why aren’t “old Keynesians” perfectly happy with the current state of the economy?
- Why aren’t “new Keynesians” perfectly happy with the current state of the economy?
An old Keynesian — a Keynesian in the tradition of, say, Sir John Hicks — would, as far as I can tell, be stricken mute here because old-style Keynesianism provides no yardstick for measuring the desirability of different outcomes. The model makes no clear assumptions about what people value (in more technical language, there are no utility functions in the old Keynesian model), so it offers no way to judge whether you’ve made people happier. In a recession, we consume less, which is presumably bad, and we work less, which is presumably good, and if the recession is accompanied by a liquidity trap, we voluntarily accumulate piles of money, which is also presumably good, but nothing in the model is capable of telling us whether the good outweighs the bad.
Maybe the old Keynesian will reply that he’s out to maximize some rule-of-thumb variable that can be described without regard to utility. But what could that variable be? Short-run consumption? But if that were really the target variable, the corresponding policy prescription would be to quash all saving. Surely that can’t be right. (On the other hand, I suppose it’s impossible to even contemplate quashing saving in a model where people automatically save a fixed fraction of their incomes.)
A new Keynesian, by contrast — a Keynesian in the tradition of, say, Michael Woodford — could at least address the question, because new Keynesian models do start by specifying people’s utility functions, so we know something about the trade-offs people are willing to make and can therefore talk meaningfully about whether times are (relatively) good or bad. But I’m still unclear on why, given that model, we should expect the current state of the economy to count as bad.
After all, when people hold money, they do it for a reason. That reason will vary from one new strain of new Keynesianism to another: Either money contributes directly to people’s utility, or it’s a prerequisite for transacting business. If we’re in a liquidity trap — that is, if people are hoarding money — there’s no problem transacting business, so (as far as I can see) they must be holding money for utility’s sake. In other words, if they’re hoarding, it’s because they like to hoard. Which brings me back to my question: Why, as the stock of money continues to grow, shouldn’t the joy of hoarding eventually compensate for the annoyance of not having food on the table?
I will be mildly surprised if there is a satisfactory “old Keynesian” answer to this question, but not at all surprised if there is a satisfactory “new Keynesian” answer. I’d like to understand that answer, and I’ll be very glad if someone can explain it.
Brilliant. Does your macroeconomics textbook contain these considerations? And can you point to other texts or articles with critiques along these lines? Thanks
Maybe I missed some deeper meaning of the question, but to me the answer pretty obvious, and I am not even a Keynesian. Only consumption of goods and services contributes to utility. Money is just an asset the holdings of which agents optimize in order to maximize their consumption.
I would construct the argument this way: the keynessians say, that people in such situation (liquidity trap) are unemployed involuntarily (therefore ‘not working’ cannot be constructed as ‘good’). Let’s say that the utility of hoarding comes only from the fact, that the others are doing it too (i.e. similarly as in your last ‘tournament’ question). Therefore, if you stop the hoarding, everyone would be better off.
“if they’re hoarding, it’s because they like to hoard”
I may have this totally and utterly wrong, but is not the whole crux that they are acting slightly irrationally? They are acting in a way that does not maximise utility because the joy of hoarding does not ovecome the annoyance of not having food.
It seems odd to say that money-hoarders are *happy* to hold it – surely they do so because they are averse to all the alternatives. I don’t *want* to hold money, but it seems preferable to every other option.
In a “trap”, of whatever kind, you have no good options, by definition. Thus, in this case, the hoarding of money is simply the least bad option, but it’s still bad as compared to being out of the trap — i.e., as compared to being reasonably sure of future income, and hence food. The trap is supposed to arise because people are increasingly not sure of their future income.
This is not, by the way, to endorse Keynesian economics, whether new or old.
Keynesians would agree that the obvious course of action is to print the money they so eagerly want to hoarde, thus making them “happy,” as you put it, or, thus “getting us out of the liquidity trap” (by causing inflation) as they say it. In other words, Keynesians only think that the current state of the economy is undesirable *given that we do not do what you suggest, which is to print money and give it to people*
You have an evil sense of humour. It’s one of your best traits!
“The theory, as I understand it, is that vast numbers of people are choosing to hold vast amounts of money.”
The number of people hoarding money (primarily corporations) is not a large portion of the population. Most of The money being hoarded is corporate profits–which is not being spent (by choice). The hoarded funds are invested in short-term investments, so the money is being used. Individuals do not have access to the same sources (or amounts) of funding or raising cash via debt offerings as corporations, so a comparison is not valid.
Andy, what does it mean to be unemployed involuntarily? More specifically, would you consider someone unemployed involuntarily if they voluntarily choose not work for an offered $10/hr when they desire $20/hr?
Josh, the Keynesians are to answer your question, not me…
They are hoarding money because they are scared about the economy, and this liquidity trap makes it impossible for monetary expansion to spur the economy, thus keeping them scared. It is a viscious circle. Who can be happy about that?
Josh, I think “unemployed involuntarily” means they would like a job at the going wage rate, but they can’t find it. (Then again, stated like that, a question arises: I too would like a quarterback job at the going wage rate for quarterbacks, but I can’t find it. So what?)
Well said, Steve_Landsburg! I’ve been wondering the same, but I haven’t been able to crystallize the question as well as you have. Keynesians seem to ignore the benefit of hoarding money — there is something that motivates people to do it, and until you can remove that reason, policies designed to blindly “stop the hoarding” are just going to make people worse off.
You could do helicopter drops so that people _have_ to spend or see their savings dwindle, but it’s unclear how that helps their utility. If people are hoarding because of uncertainty about the future, then you have to fix the uncertainty, not just squeeze them until they relent and buy crap they don’t want.
As I keep saying, spending is not good. *Good spending* is good. Spending that people engage in simply to insulate themselves from the latest government policy is not good.
I think Krugman’s answer to your question would be something like, “Yes, people want more money, and we can produce that money cheaply, but the Fed isn’t doing so.” That answer, of course, has its own problems: if people wanting money is simply a matter
Anecdotally, it seems that entities are hoarding money for multiple reasons. Large corporatations are conserving money because they anticipate a double-dip recessin, or something approaching it, and want to have cash on hand to fund future expenses.
Individuals are hoarding money because their investments in homes and retirement funds are below where they stood in four years ago, and they are saving to rebuild their cushions and retirement funds.
Perhaps the way out of this liquidity trap is to allow inflation to rise, so that people are encouraged to spend rather than save.
I know I’m about to sound like the Landsburg cheerleading squad, but posts like this are why you are my favorite blogger. The others make excellent points using rational analysis, but your ability to reach out and think clearly along lines EVEN OTHER ECONOMISTS would describe as ridiculous is extremely interesting and fun to read.
I am not a Keynesian, and not even a macro guy. So coming from my admittedly microeconomic perspective, I think of it as a kind of coordination game with multiple equilibria. Given that everyone else is holding money, it makes sense for you to hold money, and vice versa. But if everyone spent more money, then everyone would be better off. We are stuck in an inferior equilibrium, and Keynesian policy is supposed to goose us out of it, after which we’ll hopefully land in the superior equilibrium. (I’m not saying I agree this is what’s actually happening; this is just how I understand the Keynesian story.)
Very good question! I’d say that the question boils down to “What kind of market failure occurs in a liquidity trap situation?” New Keynesians should understand this question and be able to answer it. That’s not a trivial question since “liquidity trap” per se does not constitute a market failure – Friedman rule is optimal in many models.
Silas Barta pointed me tot his post, and this was my response on my blog (http://factsandotherstubbornthings.blogspot.com/2011/09/credit-where-credit-is-due.html?showComment=1316808780595#c3419822104092962465)
It seems to me Landsburg is trying to infer some sort of welfare implications from Keynesian theory. It’s not obvious to me that there is a particularly good way to do that. Keynesianism doesn’t explain happiness. It explains the path of employment, interest rates, output, etc. How happy that makes people or how we should evaluate that is a different question entirely.
The way I evaluate it is to say that (1.) people don’t seem to like being unemployed, and (2.) the model says they are unnecessarily being unemployed, ergo (3.) I’m not happy with the current situation.
This is a truly bizarre statement from Landsburg: “An old Keynesian — a Keynesian in the tradition of, say, Sir John Hicks — would, as far as I can tell, be stricken mute here because old-style Keynesianism provides no yardstick for measuring the desirability of different outcomes.”
Atomic physics “provides no yardstick for measuring the desirability” of detonating an atomic weapon in a densely populated area. But I can still adhere to what I perceive to be a correct understanding of atomic physics without being “stricken mute” on the question of the desirability of setting off the weapon. Landsburg seems to want an economic theory to be a normative and positive theory wrapped up in one. That seems dangerous to me, although obviously a lot of people talk about economics like that.
He then gets into New Keynesianism which does offer at least the temptation of a potential normative theory. I would think the answer is “yes, people get utility from holding money right now, and yes people get disutility from being unemployed, so why don’t we print more money and satisfy the utility of both. The normal costs associated with that (inflation, etc.) don’t seem like much of a threat”. So even if you want to force a normative frame on this, I don’t see why Keynesians should be “thrilled”. The situation still isn’t Pareto optimal, right? The utility that hoarders get from hoarding money comes at the expense of the unemployed.
Also – does anyone else think Hicks looks a lot like Michael Caine? Just saying.
Marek:
I’d say that the question boils down to “What kind of market failure occurs in a liquidity trap situation?”
Yes! This is my question exactly!
New Keynesians should understand this question and be able to answer it.
I expect that they can, and that they have — but I am not well educated enough to have heard them. I’m hoping someone will fill this gap in for me.
Daniel kuehn:
Atomic physics “provides no yardstick for measuring the desirability” of detonating an atomic weapon in a densely populated area. But I can still adhere to what I perceive to be a correct understanding of atomic physics without being “stricken mute” on the question of the desirability of setting off the weapon.
But economics, unlike atomic physics, is supposed to illuminate tradeoffs. In a situation where people want to hold a lot of money, and are able to do so, and also want to work, and are not able to do so, it seems to me that it is the job of economic models to give us some framework for weighing the good against the bad.
Karl Smith says “an economy that makes lots of people feel bad is, by definition, a bad economy”.
Perhaps I’m misunderstanding the argument, but it appears that you are arguing that a Keynesian analysis would recommend an inflationary policy to increase the amount of money held by individuals. Is this not a blatant confusion of real and nominal money? People are trying to maximize their real monetary holdings, not their nominal monetary holdings. Why would increasing people’s nominal holdings be beneficial? Or are you claiming that there is some “costless” method of increasing real money? If so, I sure would like to hear what it is.
Furthermore, given that you start your post by suggesting that Keynesians should approve of the current situation, I’m confused by the fact that you don’t give a clear reason for such a view, and in fact suggest that Keynesians should be unable to present a value judgment.
RF:
Is this not a blatant confusion of real and nominal money? People are trying to maximize their real monetary holdings, not their nominal monetary holdings.
In a liquidity trap, there is no distinction between supplying real money and supplying nominal money; demand keeps up with whatever you supply, so the price level doesn’t move.
I’d say that the question boils down to “What kind of market failure occurs in a liquidity trap situation?”
Yes! This is my question exactly!
The “market failure” just is the Keynesian idea of the “trap” — that is, the hoarding of money (the liquidity trap) is a result of economic uncertainty, and the economic uncertainty is a result of the hoarding of money.
Now, I agree that the notion that such intertwined causation actually is a trap is faulty. But not because people (whether consumers or investors) simply prefer to hold money — they’d prefer instead less economic uncertainty.
I think Krugman’s view is that people are balance sheet-constrained: they don’t really value money inherently, they just need a lot of it to pay off their debts. Once they do, they would love to spend lots of money, because there’s a lot of products and services that they want.
As to what the market failure is, I think Krugman has been fairly clear in his opinions: he thinks that wages are sticky.
A couple of points :-
Re : “It should be very easy to supply money at zero cost”
Maybe so, but the money isn’t actually being supplied – and our good Keynesian friend at the NYT is not happy.
Re : “If people are hoarding money, it must be because they like to hoard money”
No, it just means they prefer that option to all the others that seem to be available, eg, investing on infrastructure and education, building a better future, etc. Eg, if I need to spend $100,000 on medical bills, it doesn’t mean I’d like to do that. I’d prefer to be healthy and spend the money on a nice holiday – but that option has disappeared.
With the medical bills, Keynesian theiry doesn’t say much that can help. With the liquidity trap, however, the theory explains a clear route to quick economic recovery.
re: “But economics, unlike atomic physics, is supposed to illuminate tradeoffs.”
OK, you have a very different view of economics than I do.
Certainly we can learn more about the implications of trade-offs from economics, and that’s often the point. But I’ve never been under the impression economics was supposed to give us our ethics or value judgements. Indeed – our value judgements are the “black box” that we put into these models! It seems circular for you to demand that the models give us those value judgements.
Macroeconomic theory should give us insight into the behavior of the macroeconomy. We can use that insight to make judgement calls. That’s about all I’d say.
There’s a little problem of multiple equilibria happening here.
One equilibrium – everyone feels confident about the economy, so they happily spend, their employers have a lot of business, retain staff, give bonuses, and so forth, making everyone feel confident about the economy.
Another equilibrium – everyone feels pessimistic about the economy, so they hold their money, their employers find business is bad, lay off staff, withhold payrises, and so forth, making everyone feel pessimistic about the economy.
In the first equilibrium, there’s a lot of trade happening. In the latter, there’s much less trade. All economists, Keynesian or otherwise, know that trade enriches people. Therefore the first equilibrium is better.
Now, we are in the second equilibrium.
Anyone can improve the economy a little by spending. Almost nobody can spend enough to improve it enough to push it back to the first equilibrium. Governments often can. The US government, for example, could borrow enough to spend enough to push the economy to the first equilibrium (they’d be paying zero or negative real interest rates on the extra debt). After business and consumers are confident (and spending) enough to keep things rolling along, they can cut back their spending again. This would be sensible, but politically infeasible.
Mike H, Keshav Srinivasan, Larry, Marek, and others: I continue to agree with Marek that the question is: “What exactly is the market failure?”. But a prior question is: “What exactly is the model?”. For example: Does or does not money have to appear as an argument of the utility function in order for me to get a liquidity trap? Is a liquidity trap in a cash-in-advance model substantially different from a liquidity trap in a money-in-the-utility-function model? If so, what are the normative implications of that difference? Etc.
There’s a little problem of multiple equilibria happening here.
But there’s more than two, and that’s the fallacy of the “trap”. In fact, at any level of “confidence” there’s some level of spending (again, by either consumers or investors or both) — sometimes confidence is great, and spending is high; sometimes it’s middling, and spending is middling; sometimes it’s low and so is spending. And of course, the confidence/spending balance occupies all points in between. There is no real “trap” at any point, despite Keynes hand-waving nonsense about “animal spirits”, etc.
@Larry – yes, and most of these are not equilibria – except for certain levels of interest rates.
Reserve banks try, instead, to steer the economy towards an unstable equilibrium. When interest rates hit zero and the economy still doesn’t go up, their steering wheel falls off. That’s the “liquidity trap”. When interest rates have no effect, it turns out previously ineffective things (government spending/austerity, for example) have a big impact.
@Steve : I’m not an economist, so I can’t answer your question about what assumptions are needed in the model. However, here’s a 1999 paper by Krugman where he derives sufficient conditions : http://web.mit.edu/krugman/www/trioshrt.html – he blogged about it more recently here : http://krugman.blogs.nytimes.com/2008/11/15/macro-policy-in-a-liquidity-trap-wonkish/
Would writing down household debt count as “hoarding money” according to Keynes?
Mike H: That first Krugman reference might be exactly what I’m looking for; I’ll need to find a little time to study it.
I notice that he’s got a cash-in-advance constraint, rather than putting money in the utility function. That looks like it will help.
Mike H: Okay, the first Krugman reference helps. (Note that there’s a typo in his equation (3).)
First, there is a representative agent in this model, so my assumption that you couldn’t address these questions with a representative agent was wrong.
Second, what seems to be happening is that people are hoarding money because they expect future low prices (and because the nominal interest rate is higher than it “should” be), so they’re actually leaving consumption goods lying on the ground and refusing to pick them up (since you need money to pick them up).
The market failure, then, seems to come from the cash-in-advance constraint: It’s socially costless for me to pick up those unclaimed consumption goods, but it’s privately costly because I can’t pick them up without expending some money that I expect to be worth a lot in the future.
I’m pretty sure that’s the right intuition.
Next question: Is this the intuition underlying most New Keynesian analyses of the liquidity trap, or are there others?
@Mike_H: Anyone can improve the economy a little by spending.
Actually, for a multiple equilibrium model to be consistent (internally or with standard economic assumptions), it would have to be a worse situation than that. If “the economy” were improved by any small bit of spending, then it’s a simple matter of hill-climbing to get to the global optimum, because each instance of spending is also a local improvement. (Meaning that it’s also not truly a case of multiple equilibria either.)
What’s going on, rather, is that it would be a local _dis_improvement for people to make a purchase that they would otherwise not make, which is why you can’t switch to a better equilibrium from individual, uncoordinated decisions.
@Daniel_Kuehn: Economics is supposed to be able to identify Pareto (and KH) improvements that don’t hinge on a value judgment. Steve_Landsburg’s point is that these Keynesian models *don’t* do that — they give no way of estimating the social gain (however you aggregate individual preferences) from “making people not hoard money”. This is not an unreasonable thing to expect of it.
@Mike: yes, and most of these [confidence/spending balances] are not equilibria
Why not? I.e., what distinguishes one level of confidence/spending from another?
“and we work less, which is presumably good”
I’m no economist, but this statement is simply not accurate for millions of American’s who desperately want to work for various and sundry reasons.
Either your summary of their views is wrong, or their views are wrong, but what I don’t believe to be wrong is the drive for millions of unemployed to not worry about how they will feed their families
Tony Cohen: Very few people miss working, though quite a few miss the income they could have earned. We count the loss of income as a bad thing and the gain in leisure as a good thing. On balance, that might still be very bad, but we do have to count both components.
@Steve: “The market failure, then, seems to come from the cash-in-advance constraint”. I am not sure that’s a full answer. The “liquidity trap” happens precisely when the cash-in advance constraint doesn’t bind – people hold money not only for transactions, but also to store value. There must be something else to it.
One thing that Krugman thinks about is “price rigidity” (i.e. price is exogenous). That would explain why there are consumption goods lying on the ground and the price doesn’t adjust.
Now, if you think that price level is exogenous (and wrong), then it seems straightforward to conclude that there is a market failure. That’s the economics from 1960’s. A more interesting question is if this is still a market failure if the price level is rigid because of e.g. menu costs.
Marek: I now think I understand Krugman’s note less well than I thought I understood it this morning.
I’m no longer even 100% sure that they *do* leave consumption lying on the ground, though that seemed clear this morning.
I think you must be absolutely right about this: The “liquidity trap” happens precisely when the cash-in advance constraint doesn’t bind – people hold money not only for transactions, but also to store value. There must be something else to it. Thanks for this insight; it’s extremely useful.
Here are my specific points of confusion:
1) If people are holding “too much” money, it’s got to be because the private cost of holding money is less than the social cost of holding money. The social cost of holding money is surely zero, so the problem must be that the private cost is negative. In this model, exactly what constitutes the private cost of holding money and why is it negative?
2) With all the various markets that are failing to clear, I’ve lost track of what happens in the goods market. Does all the (exogenously provided) manna get consumed?
3) If we build in the microfoundations of price rigidity, is the outcome still suboptimal?
Finally, you are absolutely right about this also: A more interesting question is if this is still a market failure if the price level is rigid because of e.g. menu costs. Krugman doesn’t provide the microfoundations for his sticky prices, so it’s very hard to do a welfare analysis. I’m sure that more sophisticated new Keynesian models do provide these microfoundations, but I don’t know enough about them to know where the welfare analysis leads.
@Silas “it would be a local _dis_improvement” Yes, you’re right. My spending, at equilibrium, only improves the economy if I ignore the cost to me.
I think Keynes and Keynesians don’t regard a liquidity trap as an equilibrium though – there was a recent blog post by Krugman on the topic, and Keynes also wrote about the fact that obsolesence gradually forces the private sector to spend, even if the government does nothing.
@Steve “The social cost of holding money is surely zero” Why, exactly? After all, if we *all* stopped holding money and started spending it, wouldn’t the economy start roaring along again?
From what little I’ve read about these things, expectations are important. Krugman writes that even if the government won’t spend, the Federal Reserve could break the trap by “credibly promising to be irresponsible”, making people less willing to hold money. It’s much harder for the Fed to do this than for the government to spend, however.
Steve:
ad 2) Krugman seems to assume that output is “demand determined”. Since no resources are required to produce output, I’d say that it is equivalent to saying that a “full output” is produced, and some of the manna is unconsumed.
I wish I knew answers to 1) and 3)… Woodford and Eggertson have done some research on this: http://www.columbia.edu/~mw2230/BPEA.pdf I have not read this, but I’ve noticed that they assume Calvo pricing. So it looks to me to be just a more sophisticated version of Krugman. But I don’t know enough.
@Larry : At any level of “confidence” there’s some level of spending
there’s also a feedback loop, because current levels of spending affect future confidence. In other words, there are two graphs you can draw of spending vs confidence :
* today’s confidence vs today’s spending
* today’s spending vs tomorrow’s confidence
The equilibria only occur when these graphs meet.
In the absence of the reserve bank, there are, maybe, three equilibria : one where everyone expects the worst, so they don’t spend, one where the economy is roaring along, but inflation expectations has damped the economy enough so it doesn’t roar any faster, and an unstable equilibrium somewhere in between.
In normal times, the reserve bank tries to use interest rates to keep the economy rolling on somewhere in between the two stable equilibria. When interest rates hit zero, the reserve bank can’t do much any more, and other policies are required to keep us away from the lower stable equilibrium.
The lower stable equilibrium is the liquidity trap. Nobody likes it, but it is an equilibrium – hence it’s a trap.
The “market failure” here is, maybe, something like the “failure” of a used-car or health insurance market. People want to trade, but the optimal personal strategy, given the circumstances, is to not trade.
Mike H;
The “market failure” here is, maybe, something like the “failure” of a used-car or health insurance market. People want to trade, but the optimal personal strategy, given the circumstances, is to not trade.
If there’s a market failure leading to people holding too much money, then the private and social costs of holding money must diverge. I am not going to feel like I understand this model until I understand what the private cost of holding money is, and why it diverges from the zero social cost.
Marek:
You say (as I originally thought) that consumption is left lying on the ground, and I still think that’s the most likely interpretation. (If it’s not, then I don’t see what the problem is.)
But: If that’s true, then the Euler equation (3) is altered, because c_t shrinks. So it seems Krugman’s “IS curve” in Figure 3 must shift. Where does he account for that shift?
Or am I more confused than I think I am?
Seems to me this is Macro 100. The liquidity trap is not a market failure. People hold unlimited cash because they worry they will lose value holding something else. No market failure in that.If there is market failure, it is whatever is keeping unemployment high, presumably by keeping real wages too high (rigid money wages?)The liquidity trap is a problem because it makes monetary policy impotent as a means of offsetting this market failure by lowering interest rates and increasing spending by shifting demand from the future to the present.
Neil:
If there is market failure, it is whatever is keeping unemployment high, presumably by keeping real wages too high
There is no labor, and hence no real wage, in the Krugman note we’re trying to decipher.
Steve, here is a paper by Greg Mankiw which provides menu costs as the microfoundations of sticky prices, and I think he does the kind of welfare analysis you wanted:
http://www.economics.harvard.edu/files/faculty/40_Small_Menu_Costs.pdf
He concludes that a monopolist facing menu costs will lead to a socially suboptimal outcome. He also shows that even small menu costs will lead to a full Keynesian business cycle.
But why is the social cost of holding money equal to zero?
Suppose we could trade, but don’t because of some tariffs. Well, there’s surely some social cost to the fact that we hold our money instead of trading.
Now, suppose instead that we could trade, but don’t because we fear the future. Well, there’s surely some social cost to the fact that we hold our money instead of trading.
Mike H: But why is the social cost of holding money equal to zero?
Because money costs nothing to produce.
Keshav Srinivasan: Thanks. But I think I do understand how menu costs plus monopoly can lead to socially suboptimal outcomes. My specific question concerned the undesirablity of being in a liquidity trap.
Maybe the answer is that the liquidity trap is not bad in and of itself, but bad indirectly because it renders ineffective the policies that could improve on the Mankiw-like outcome. But I’d still like to understand this much better.
“What kind of market failure occurs in a liquidity trap situation?”
-Marek
“If there’s a market failure leading to people holding too much money, then the private and social costs of holding money must diverge.”
-Landsburg
Thank you both for putting this so succinctly. I’ve been trying to figure this out for a while now. My much fuzzier formulation of the question goes like this: “Keynesians want us to spend more than we individually want to spend. So spending and consumption must have positive externalities; saving must have negative externalities. Where do these externalities come from?” How does my consumption have any non-pecuniary externalities (of the kind that would justify Keynesian policies)? I understand that my spending might drive up prices and drive further production and employment, but I would think these kinds of “externalities” are entirely pecuniary. If we all spend more than we really want to (convinced perhaps that there is a social benefit to doing so), the additional production that results is the production of goods we don’t really want.
Sorry, I actually have no insights on this question. I’m just offering my own formulation of a similar question. I’ll be watching this thread (and watching for a follow-up post) closely for anything illuminating. I was kind of hoping a certain busy-trolling Keynesian would stop pretending he doesn’t understand Steven’s question and answer it.
@Mike: In other words, there are two graphs you can draw of spending vs confidence :
* today’s confidence vs today’s spending
* today’s spending vs tomorrow’s confidence
The equilibria only occur when these graphs meet.
Those two graphs, both of which slope up and to the right, don’t sound to me like they should ever meet, much less in two or more discrete points, and even if they did meet it doesn’t seem like the meeting should have anything to do with a stable equilibrium.
As opposed to this imaginary stability, what any real economy displays is volatility, or so-called unstable equilibria at all levels of confidence/spending. Just as no real economy “roars along” forever, so no economy stays rooted in recession/depression forever — both states (which are mere arbitrary labels after all) are as unstable as any other level of economic performance. This is not to say that the state is without effect on the economy — clearly it’s a (large) economic actor itself — but it’s no more able to divine the future than any other actor, and its actions are as likely to worsen rather than better the situation. And when the state worsens things, the very size or scale of its actions makes the problem a serious one indeed.
@Mike_H: Now, suppose instead that we could trade, but don’t because we fear the future. Well, there’s surely some social cost to the fact that we hold our money instead of trading.
I think there’s a difference between not-trading due to a tariff, vs. not-trading due to uncertainty about the future. The latter is the result of individual economic preferences, while the former is not (it’s instead the result of individual voting aggregated through various means independent of preference for certain goods).
If you can assume that consumer beliefs about the future are systematically wrong, then all bets are off — why not just assume they’re systematically wrong about the quality of foreign goods, thereby justifying a tariff? For the model to work, the lack of confidence would have be somehow self-fulfilling in a way that doesn’t require us to assume away individual utility curves.
If there is any “real” aspect to individuals’ hoarding — if, say, it is truly difficult to identify what should be done to restore the previous level of output (per a recalculation story) — then it wouldn’t be a case of self-fulfilling prophecies. Rather, people are hoarding because there are real costs to committing to certain modes of production, and the market *should* signal a high cost to doing so until better information comes in.
It all comes back to one thing: Profit.
If a business is currently making a profit (or it expects to do so), then it is interested in expanding and hiring more people only if the new hires will ADD to profit. Otherwise, they will not expand or hire. They will simply accumulate profits until such time as they choose to invest (goal: more profits) or distribute the money to the investors/stockholders.
The key point for a business is they are generally not dependent on one or two customers for their income–they are selling to a large economy. So, while some individuals may stop buying the company’s products–others will start. Thus, the business has a reasonably reliable source of income into the foreseeable future. That means the company only needs a workforce capable of supplying that level of output–which generates the current profit.
Individuals, on the other hand, are dependent on one or two employers for all income at any time. There is currently a surplus of workers in the economy. Plus, many growing markets are outside the US–so there is less need for US production/employment. The value of most employees has fallen in a monetary sense–as is seen by extended work weeks with no extra pay, less pay, etc. As companies closed and/or laid-off workers since 2001, there has been no significant additiional new-job creation in the US other than a few years. Obama manages to keep it positive, but just barely.
Because individuals essentially live “hand to mouth” their entire working lives, they have to allocate their current spending based on present AND future income expectations. In a good economy, they have expectations of being able to find new jobs reasonably quickly if their employer fails–or they see opportunities for advancement/ better pay elsewhere. So, they will not worry much about spending much of their current income or incurring reasonable debt into the foreseeable future–they expect to have the money to pay because they expect to be employed.
The problem occurs when private employers essentially stop hiring within the economy–but continue to sell within that same economy. Because production is now done overseas in many cases (i.e. outside the economy), the number of jobs needed within the economy is much smaller–as is the multiplier for service vs manufacturing employment. As the cost is typically now lower for that same production, the business profits go up–but employment (for that business) within the economy has gone down. The people who lost jobs now have to find jobs with other employers.
When people are not seeing jobs being available (especially in their own field), they tend to stop spending as much–as they are familiar with the problems associated with a bad job market and unemployment. Their spending will only pick up again when they feel comfortable with the job market and their future income worries have been substantially reduced.
As profit is what drives employment, that means reducing the cost of employment allows the company to reach profitabiity more quickly–and thus be able/willing to hire more people sooner. Thus, the temporary reduction/cut in the Social Security tax makes sense. It is a direct reduction of indirect employment costs incurred by business.
As stated earlier, because interest rates are so low, there is little incentive for people to invest in anything for the long term. Too much risk and not enough reward. For them, holding cash is the least-risky, most flexiable position. For business, they have the resources to invest–but they need to see “more profits” as the return before they will spend and hire. They are not seeing it at the moment, so they are not investing (at least, not in the US).
For the model to work, the lack of confidence would have be somehow self-fulfilling in a way that doesn’t require us to assume away individual utility curves.
Yes, I think I explained that, didn’t I?
I think there’s a difference between not-trading due to a tariff, vs. not-trading due to uncertainty about the future
But what? Either way, trades fail to occur, to everyone’s loss. NB – I didn’t say “uncertainty”, I said “fear” – meaning, the expectation that things will be bad in the future, leading to a desire to hold money as a store of value.
Mike H: But why is the social cost of holding money equal to zero? Steve : Because money costs nothing to produce.
Maybe it’s too late at night, but I just don’t get this. Especially in light of what I’ve described above, that as people hold money (ie, they fail to trade), they end up worse off. If, in aggregate, as everyone holds money, there’s a social cost, how then can there be zero social cost to holding money??
Mike H: Especially in light of what I’ve described above, that as people hold money (ie, they fail to trade), they end up worse off
If people make individually rational decisions to hold money and therefore end up worse off, that proves there’s a gap between the private and social costs of holding money. The social cost is zero, because we can supply people with all the money they want at essentially zero cost. Therefore the private cost must be negative.
The riddle, then, is: What, explicitly, is that negative private cost?
Pointing to the bad social outcome does not identify that cost; it only proves that it must exist.
This may be a stupid question, but why exactly is the social cost of someone keeping money and not spending it the same as the cost of giving that person that money?
Anyway, this probably isn’t relevant to your question, but here is Krugman’s explanation of why individuals saving more can lead to society collectively saving less:
http://krugman.blogs.nytimes.com/2009/07/07/the-paradox-of-thrift-for-real/
Keshav: You can choose from a variety of accounting systems (adding the constant of your choice to all the costs or all the benefits), but the conventional choice is to define the *cost* of an asset as its replacement cost.
@Steve: “The social cost is zero, because we can supply people with all the money they want at essentially zero cost.”
Like others, I really don’t understand your point here. People hold money not because they like its looks, but because it’s a store of value. It’s true you can supply people with an indefinite amount of paper money by just printing it, but that’s not the money they want. And the money they want — the store of value — costs what the value is worth.
I’d also say that the negative private cost of holding money is the present consumption/investment lost in holding onto the store of value. That cost is nevertheless rationally preferable in times of uncertainty, when future costs become harder to predict.
Larry:
It’s true you can supply people with an indefinite amount of paper money by just printing it, but that’s not the money they want. And the money they want — the store of value — costs what the value is worth.
You are confusing a social cost with a private benefit and hence are doubly confused! Unfortunately, this is not an instance where two wrongs make a right.
“You are confusing a social cost with a private benefit and hence are doubly confused! ”
No doubt. I think it would still be helpful, to others besides myself, if you could elaborate on how we supply people with all the money they want at zero “social cost”.
Larry:
No doubt. I think it would still be helpful, to others besides myself, if you could elaborate on how we supply people with all the money they want at zero “social cost”.
The cost of creating a dollar bill is a few cents worth of ink and paper, close enough to zero for our purposes. But most dollars exist only as arrangements of electrons in bank computers, and they cost even far less to create.
(You said something earlier about that “not being the money people want”, but any dollar is interchangeable with any other.)
Okay, I’m going to go out on a limb here and say some stuff that might look crazy in the morning….but it’s my best current understanding of what’s going on in Krugman’s note. If I still believe this in a few days I will blog on it.
In ordinary circumstances, with flexible prices, the social cost of holding a dollar is zero and the private cost (measured in forgone consumption) is positive. Therefore people hold too few dollars. That tells us there’s an externality. If we inquire into the nature of that externality we see that it’s this: Spending money drives up prices, which is good for sellers, bad for buyers, and bad for other moneyholders. The first two wash out (because everything that’s sold is also bought) so when people spend too little, we have a net negative.
With prices fixed, spending money can’t drive up the price level, so it drives up the real interest rate, which is good for lenders and bad for borrowers. Ordinarily that would wash out (because everything that’s borrowed is also lent). But in a liquidity trap, the average person is a lender, so the good outweighs the bad. Thus there’s an external benefit from spending, which, depending on what we arbitraily choose as the zero point, could be recast as an external cost of hoarding (as Larry has insisted).
As I said, I have no idea whether this will make sense in the moring. The one thing I’m sure of is this: If, in a liquidity trap, there is too much money-hoarding, then there must, in a liquidity trap, be some divergence between the private and the social cost/benefit calculus. Maybe this identifies it.
@Steve: The cost of creating a dollar bill is a few cents worth of ink and paper, close enough to zero for our purposes. But most dollars exist only as arrangements of electrons in bank computers, and they cost even far less to create.
But that’s just not what people want to hold. I know you keep thinkng that people want ink and paper, or arrangements of electrons in bank computers, or other forms that might be “interchangeable with any other”, but that’s where you go wrong. People want to hold stores of value. Since printing money undermines such stores, they want another form of such store, or they want adjustments to such stores, and such forms/adjustments are not cost free — not even “social cost” free.
(This is not to address your latest post, which is more complicated, and interesting.)
Mike H said:
“The lower stable equilibrium is the liquidity trap. Nobody likes it, but it is an equilibrium – hence it’s a trap.”
Perhaps I’m being a bit pedantic, but I’m a bit uncomfortable with how the word “equilibrium” is used. It seems to me that “equilibrium” is being used as shorthand for some other concept, such as “locally unique equilibrium”.
“Suppose we could trade, but don’t because of some tariffs. Well, there’s surely some social cost to the fact that we hold our money instead of trading.”
There’s a social cost to tariffs. The lack of trade is a reflection of that cost.
“The US government, for example, could borrow enough to spend enough to push the economy to the first equilibrium (they’d be paying zero or negative real interest rates on the extra debt).”
Perhaps I’m just showing my ignorance, but it seems to me that much of the apparent effect of government policy is illusory. Government can borrow money? From where? The only place they can borrow money is from investors who, by definition, have money they’re willing to invest in America. Money that they presumably would have found some other way to spend had the government not come along.
Steve Landsburg said:
“You said something earlier about that “not being the money people want”, but any dollar is interchangeable with any other.”
Well, no, a 2011 dollar is not interchangeable with a 1970 dollar. The point that I’m making, and presumably that Larry is making, is that if you just hand out money, you’ll just erode the value of the dollar, and people may end up with more dollars, but they won’t end up with any more (real) money overall. I just don’t see how nominal and real money can be the same. Suppose we were to introduce a new currency, and declare that we’ll hand out 100 of the new currency for every 1 of the old. Would the money supply then get multiplied by 100?
Some thoughts –
If, in a liquidity trap, “things” were locally optimal but not globally optimal, would that undermine anything you’ve said?
But in a liquidity trap, the average person is a lender
How can this be true?
And if there is
* an excess capacity for production,
* an excess of loanable funds
why would a marginal increase in spending drive up either prices or real interest rates?
If I consume now, I can’t consume later. While this is always true, for some reason now, the ‘later’ outweighs the ‘now’. There’s your negative cost of hoarding right there. Maybe the question is, why does ‘later’ outweigh ‘now’ in a liquidity trap? I guess people expect, on average
* no other store of value will produce better returns than just holding money
* future income will be less than current income
Not sure if you’ve seen this, and not sure if it would be helpful: http://www.princeton.edu/~pkrugman/optimalg.pdf … linked from http://krugman.blogs.nytimes.com/2008/12/29/optimal-fiscal-policy-in-a-liquidity-trap-ultra-wonkish/
Quote : “The bottom line here is that while we usually think of Keynesianism as the preserve of ad hoc models, in this case doing it “right” – using a macromodel with maximizing agents and a proper concern for intertemporal constraints – actually suggests a very strong case for big government spending in the face of a liquidity trap”
RF: I just don’t see how nominal and real money can be the same.
When I say “money”, I mean what you’re calling “nominal money”.
“Perhaps I’m just showing my ignorance, but it seems to me that much of the apparent effect of government policy is illusory. Government can borrow money? From where? The only place they can borrow money is from investors who, by definition, have money they’re willing to invest in America. Money that they presumably would have found some other way to spend had the government not come along.”
The govt sells debt–and once that debt is sold, it can not be redeemed until it becomes due. However, there is a large secondary market for govt debt, so the opportunity to sell that deal is real–but “at what price”? Therein lies the problem with buying debt if you do NOT intend to hold it for the long term. As rates are just about at rock bottom now, that means the price of debt is relatively high compared to the interest rate. But if interest rates go up, the “cash now” value of the debt (especially long-term debt) falls. Think about someone who bought 30-year US govt debt in the early 1980s–and held it. Interest rates were very high–so they got a great return for 30 years. Now, that debt will be retired (30 years have passed)–and what to do with the cash?
The groups buying govt debt are pretty much in two camps, as far as I can tell. Eitehr you are in it for the “long haul”–buy and hold (which sucks with today’s interest rates), OR you are buying strictly short-to-mid term debt (90 days to 2 yrs), with the idea being to move the money into other investments if the opportunities come up. If they don’t, just roll the money back into the govt debt and let ‘er ride again. But they keep looking for other options. They are not *afraid* to invest. They just do not see a lot of worthwhile investments right now (as far as they are concerned). The low-hanging fruit of overseas investment has already been picked, so those options are also gone. And the social/political/economic news is really only good for parts of South/Central America–and those countriesa are growing dramatically. Sure, they have their problems. But the profit potential is real–and it is being realized.
Steve: I like the explanation you gave. My intuition was that if the price level is “too high” and fixed, and some of the consumption good is left lying on the ground, the social cost of consuming additional unit of consumption is zero, but the private cost is the price which is positive. (Here too high means anything higher than zero, but that might not hold more generally).
There is a subtle difference, though. Your explanation relies on the fact that the price is fixed (at whatever level), while this explanation relies on the fact that the price level is too high. Or, maybe, your explanation also implicitly assumes that the price level is fixed at a level that is too high (so that consumption is smaller than the output, and an average person is a lender).
If it’s true that a price level must be too high, one might also ask the Keynesians of all vintages the following question: What makes you think that the current price level is too high? More precisely, what price level would you like to have to be happy?
Steven:
I have not read the other responses so I appologize if I am repeating.
The short answer is that in both cases is that people want money but we cannot give it to them. Therefore, the people are stuck in perpetual sadness.
The straight forward Keynesian response – in both versions – would be for the Central Bank to print money and give it to people. It could do this via checks or via burying bank notes in abandoned mine shafts to use an example from Keynes himself.
If you like I will write in more detail, but that is the core answer.
Old Keynesianism isn’t about people. Its about imagining one’self as a central planner of an economy as if one was playing a video game (like Civ) and tweaking knobs. It doesn’t try to answer this question because it doesn’t presuppose the utilitarian/individualist ethics that underpins most economics.
I have to agree with Larry that Steve_Landsburg’s points are non-responsive. When people hoard $X in money, the “real” thing that they actually want is something like, “An option on buying roughly the fraction f = X/Y of a future period’s economic output.”
Printing money does not satiate this desire! All it does is change what X people need to have to achieve the option on f. That is the sense in which people’s hoarding of nominal money is in pursuit of a real target.
While I agree that people are satisfying themselves by hoarding money, and that forcing people to spend money doesn’t make hoarders better off, it is wildly erroneous to claim that hoarders’ desires qua hoarders can be satisfied (costlessly) by printing money — this is exactly what they *don’t* want, except perhaps to the extent that it has side effects that coincide with things they want (like a “better economy”).
Silas Barta:
Printing money does not satiate this desire!
Nevertheless, we can still talk about the supply and demand for money.
Consider the case we *do* understand: When prices are flexible and nominal interest rates are positive. People care only about their consumption and their real balance holdings (i.e. M/P, where M is quantity of money and P is the price level). This does not stop every macro textbook in the world from studying the supply and demand for M.
I’m quite sure I’m not doing anything radical by applying this standard sort of analysis in the case where prices are fixed.
@Steve_Landsburg: I’m quite sure I’m not doing anything radical by applying this standard sort of analysis in the case where prices are fixed.
Sure, except to the extent that you suggested there is some kind of Pareto/KH improvement waiting to be realized by costlessly printing the money so desired by hoarders. I agree with your point that hoarders are just doing what moves them up their utility curve, which has to be weighed against the losses to others.
I disagree, however, with the flippant dismissal that, “hey, if they want money, we can just print more”. They don’t want money per se; they want a certain M/P, which is not easily manipulated by printing money.
Silas:
Sure, except to the extent that you suggested there is some kind of Pareto/KH improvement waiting to be realized by costlessly printing the money so desired by hoarders.
I suggested no such thing; I’m not sure why you think I did.
Hm, good point, Steve_Landsburg. It was probably when you said this:
“That means that people want to hold lots and lots of money instead of spending it. Cool! We can provide money at almost zero cost. So it should be easy to make people very happy. What’s the problem?”
and then followed it up with this:
“Still: The theory, as I understand it, is that vast numbers of people are choosing to hold vast amounts of money. Since money can be produced costlessly, this ought to count as a very good thing — which should offset a lot of very bad things, no?”
Silas: A more careful reading might have revealed to you that I was very much *not* asserting that money creation is a Pareto or K/H improvement, but rather that *given my understanding of the Keynesian model, it follows* that money creation is a Pareto or K/H improvement.
That leaves two possibilities: Either a) the Keynesian model actually implies this, or b) my understanding of the Keynesian model is flawed. My money is heavily on b), and I was asking for help with finding a better understanding.
We all (I think) can see quite clearly that printing money is *not* the road to an efficiency gain. We all can also see quite clearly (I hope) that printing money is socially costless. Therefore if there is a private benefit, there must be an offsetting external cost.
Repeating the obvious fact that there are no efficiency gains from money creation won’t help resolve this. It’s as if I had read a physics paper by a physicist I believe to be very smart, and had said: “Based on what’s in this paper, it looks to me like you could build a perpetual motion machine. What am I missing?”.
It is not helpful at that point to either a) keep repeating that there the second law of thermodynamics rules out perpetual motion machines or b) to imply that I had ever claimed otherwise.
My question is: In the Keynesian model, what is that offsetting external cost?
I’m not quite sure what Silas is arguing but I can’t think of a single Keynesian model -new or old – where money-financed government spending is not expansionary and welfare enhancing during a recession.
@Steve_Landsburg: A more careful reading might have revealed to you that I was very much *not* asserting that money creation is a Pareto or K/H improvement, but rather that *given my understanding of the Keynesian model, it follows* that money creation is a Pareto or K/H improvement.
Great! That was the context in which I was criticizing your point!
My answer is: The Keyesian model does not imply that it would be an improvement to print money and sell/give it to the hoarders, because people don’t have a *nominal* target when they save, but a *real* target. Yes, as you repeatedly remind us, they are saving in nominal dollars; however, they save until they hit a certain m/P (m = nominal savings). Therefore, you are hasty in concluding that the Keynesian sees a simple problem of people needing to be given money.
(I’m about the fifth person to point out this confusion between nominal and real, and this about your fifth glib dismissal of it on the basis that hoarders save nominal dollars. There is a good point to be made here, as I mentioned before … but this isn’t it.)
@Karl_Smith: I don’t recall saying otherwise; could you point to the comment you’re referring to?
I was thinking of where you said
Sure, except to the extent that you suggested there is some kind of Pareto/KH improvement waiting to be realized by costlessly printing the money so desired by hoarders. I agree with your point that hoarders are just doing what moves them up their utility curve, which has to be weighed against the losses to others.
I disagree, however, with the flippant dismissal that, “hey, if they want money, we can just print more”.
Why would that be a flippant dismal. As a Keynesian this is in fact what I would say. We should print more money.
There are some issues with distribution and expectation but the basic notion that liquidity demand can and should be satisfied by increasing liquidity I think is correct.
Silas Barta: You are overlooking (among other things) the fact that equilibrium in the market for real balances (M/P) can be recast as equilibrium in the market for money (M), as long as you are careful about transforming units properly, etc.
(You’re also overlooking the fact that in a model where P is fixed, there’s essentially no difference between the market for P and the market for M/P.)
Consider again the ordinary flexible-price positive-nominal-interest-rate situation. Every textbook agrees that M is provided at zero social cost; every textbook agrees that people care about M/P, not about M, and every textbook agrees that it is nevertheless useful to think about equilibrium in the market for M.
As long as M is supplied at zero social cost, and as long as money market equilibrium is suboptimal, there must be an externality to holding M. I believe (but am not sure) that in the Keynesian liquidity trap story the money market equilibrium is suboptimal. (I could be wrong; it could be that all of the suboptimality comes from the non-clearing of the goods market.) Therefore there must be an externality. Asking “Why can’t you just make people better off by printing money?” is equivalent to asking “What is that externality?”. It does not imply a belief that in the model you actually *can* make people better off by printing money; it is instead a way of asking why you can’t.
Now you’re going to say you’ve already told me why you can’t. That’s fine. But the argument from externalities says that in the absence of an externality, you *can*. So what I’m asking for a reconciliation of the two arguments.
See my earlier example: Physics paper A says blah blah. I say “But then I could do x, y and z and build a perpetual motion machine!” Silas Barta says: “No you can’t because of the second law of thermodynamics”. I say I know that, but it seems to me given blah blah, there is a way to build a perpetual motion machine. Silas Barta says: I’ve already explained to you why you can’t. I say: I think you are continuing to miss the point, Silas.
And yes, I do understand that people are demanding M/P, not M. To this there are two responses: First, P is fixed in this model, so increasing M and increasing M/P are the same thing. Second, even if that weren’t the case, you’d still be missing the point. I know why increasing M won’t work. I also see an argument (starting from the absence of externalities) that says it must work. (Zero social cost plus positive private benefit (revealed by the fact tht people forgo consumption to hold M plus no externalities implies Pareto improvement.) I’m making the latter argument in order to ask where the gap is.
I know why increasing M won’t work. I also see an argument (starting from the absence of externalities) that says it must work. (Zero social cost plus positive private benefit (revealed by the fact tht people forgo consumption to hold M) plus no externalities implies Pareto improvement.) [bracket added]
So then wouldn’t that suggest to you that there’s something wrong with the argument, at least as you’ve summarized it? And might that not be the fact that what people are forgoing consumption to hold is not really M but rather M/P, and M/P has positive social costs?
So SL did you still love your analysis in the morning?
One detail – did you mean to say that in ordinary circumstances (flexible prices) “people spend too MUCH”, i.e. “hold too few dollars” (because the resulting higher prices are “bad for other moneyholders”)?
Your (tentative) conclusion for the liquidity trap does seem to be the opposite of the standard Keynesian view that the real interest rate is too high due to concern over deflation (is that still a ‘fixed price’ model?) given the zero nominal rate floor, which is why they typically want the Fed to try to create a credible threat of inflation.
I can’t speak to the technical details of any particular models (and personally I think recent history has shown the downside of viewing the long-run as a series of short-runs that always require – and beget — more smoothing i.e. spending), but does it help address the disconnect at all to suggest that a Keynesian views the hoarding itself as providing little or no direct utility, because what people really want is to spend, but not until others start spending first? The intuition (whether or not you buy the premise) might be captured by a story I recall from Krugman’s book “Peddling Prosperity” about a baby-sitting coop, which encountered a situation where everyone wanted to increase their reserve of coupons so everyone ended up “sitting at home glumly” waiting for someone else who wanted to go out. Admittedly that was intended as an analogy for a standard recession and in that case printing more coupons did supposedly solve the problem. But the point is that presumably everyone would have been just as glum had the supply of coupons increased but people still weren’t redeeming them.
Apparently in that example the coupon printing also led to problems of inflation, so all that hoarding didn’t actually end up raising the real value of people’s reserves much. So maybe they were just “fooled” (they were specified as a group of lawyers after all).
Larry:
And might that not be the fact that what people are forgoing consumption to hold is not really M but rather M/P, and M/P has positive social costs?
I’ve explained several different ways why you’re off on completely the wrong track here. I can’t think of another way to explain it.
Edit: Nevertheless, I will try one more time.
1) Dollar bills exist. Some people create them and other people hold them. Therefore there are suppliers and demanders. Therefore we can talk about the market for those dollar bills.
2) I understand that the value of a dollar bill varies with the price level. Likewise, the value of a worker varies with his productivity. This doesn’t stop us from talking about the demand for workers. Likewise, the value of a winter coat depends on the climate. This doesn’t stop us from talking about the demand for winter coats.
3) The demand for money, like anything else, depends on many variables. One of those variables is the price level. In general, we expect that when the price level doubles, the demand for money will double. That doesn’t mean we can’t talk about the demand for money.
4) We can talk about the demand and supply for any given object quite independently of why people demand or supply these objects. We can do it even though demand might shift at any moment for any of a great variety of reasons. In the case of money, demand depends on the price level (usually in direct proportion). That doesn’t invalidate the machinery of supply and demand.
5) All of this applies to green pieces of paper exactly as well as it applies to workers, winter coats, or gorgonzola cheese.
6) There is a vast — I mean, really vast — literature on the demand for money. You are asserting that many hundreds of researchers have devoted their lives to studying a concept that they never noticed was meaningless. You might weigh the likelihood of that against the likelihood that you might be wrong.
The above is my last ditch attempt to address this confusion.
Steven:
I don’t know if you are not responding because my answer is only repeating what has already been discussed or because you do not find it interesting. However, you might like this quote from The General Theory which I think encapsulates the answer to your question:
Unemployment develops, that is to say, because people want the moon;–men cannot be employed when the object of desire (i.e. money) is something which cannot be produced and the demand for which cannot be readily choked off. There is no remedy but to persuade the public that green cheese is practically the same thing and to have a green cheese factory (i.e. a central bank) under public control.
http://books.google.com/books?id=xpw-96rynOcC&pg=PA213&lpg=PA213&dq=people+want+the+moon+keynes&source=bl&ots=WWimtteFBH&sig=KqXuyWMFB-djwVFcEUZrqltH0vY&hl=en&ei=vwyBTvLfMI-utwfr8q3MCQ&sa=X&oi=book_result&ct=result&resnum=4&sqi=2&ved=0CDgQ6AEwAw#v=onepage&q&f=false
Karl Smith: But my fundamental question has nothing to do with unemployment and makes perfect sense in a full-employment economy. I will post a clarification of this tomorrow (Tuesday) morning.
Iceman:
So SL did you still love your analysis in the morning?
All but the last bit of it, but the failure of that last bit caused me to focus in much more specifically on what it is I’m not getting. I’ll post about this Tuesday morning.
I’ve explained several different ways why you’re off on completely the wrong track here. I can’t think of another way to explain it.
No, fine, save your thinking for coming up with a way out of the dilemma you’ve posed yourself. Here are the alternatives: either increasing M does work, and your knowledge of why it doesn’t has to be corrected, or, as you’ve said, there’s a missing “externality”. Since you’ve ruled out the notion that there can be a social cost to holding money as “completely the wrong track”, I hope you’ll find that mystery ingredient.
And here’s my last ditch attempt, just to summarize what still seems to be the problem):
Earlier you said: The social cost is zero, because we can supply people with all the money they want at essentially zero cost.
And: The cost of creating a dollar bill is a few cents worth of ink and paper, close enough to zero for our purposes.
And now: There is a vast — I mean, really vast — literature on the demand for money.
I don’t think that literature refers to the demand for ink and paper, or fancy Monopoly money. And if not, then, with all due respect, your notion of money having a zero social cost might still be a source of error.
Larry:
I don’t think that literature refers to the demand for ink and paper,
Yes, actually it does — that and certain configurations of electrons that are also supplied at essentially zero cost.
Have you actually read any of this literature, or are you just making this up as you go along?
I’ve read some, and every one I’ve read distinguishes between nominal money (the ink and paper/electron configuration kind) and “real” money (nominal money discounted for inflation). And the demand for nominal money (i.e., nominal demand) is presented as a function of the demand for real money. In which case, merely increasing the amount of nominal money won’t satisfy the demand for money.
Larry: Of course there is a difference between the demand for money and the demand for real balances. They are both perfectly meaningful concepts, and it’s perfectly possible to discuss either of them. What’s relevant here is the demand for money.
Also, I have no idea what you mean by “satisfy the demand for money”. The market is always in equilibrium, where supply equals demand. If you double the quantity supplied, the price level will double, which doubles the quantity demanded, preserving equilibrium.
Look—how much you got in your pocket right now? Taking as given the price level, interest rate, etc, would you be equally happy with some other amount? You *chose* that number of dollars — that’s your quantity demanded. You demand green pieces of paper and the authorities supply them. There is necessarily an equilibrium in that market. I have no idea what you think is undefined in that scenario, but we are way off topic, distracting attention from the point of this thread, and you are just plain don’t get it, so I will not approve any further comments that just repeat the same confusion.
“Still: The theory, as I understand it, is that vast numbers of people are choosing to hold vast amounts of money. Since money can be produced costlessly, this ought to count as a very good thing — which should offset a lot of very bad things, no?”
Well sure, if you could easily get the money directly to people. However, the money doesn’t go directly to people who want to have more. It goes to banks who often don’t want to hold any more. (Look at banks now that are charging for deposits). The easy way to achieve this from an RE point of view is for monetary expansion and/or cuts in government spending.
I accept that we can imagine that people decide that leisure is more valuable and consumer goods less valuable and so they choose to work less and consume less. This would result in less current production, raise utility and be a good thing.
On the other hand, it is also possible for there to be a surplus of labor, so that quantity of labor supplied is greater than quantity of labor demanded. Given the amount actually employed, workers gain utility by having less leisure, working more, and consuming more.
With work sharing, this would be true of all workers. However, in the real world we can have unemployed workers who would gain utility by having less leisure and consuming more. It is even possible that those who remain employed have no interest in working more, and would value any additional consumer goods less than the leisure sacrificed. Only an unlucky minority of workers stand to gain. Or, even more realistically, only an unlucky minority of workers suffer the loss compared to their past position.
While lower real wages would be the usual answer here (with the unemployed workers undercutting the employed ones,) if there is a shortage of money, it is possible that the surpluses of labor reflect a surplus of consumer goods. In that case, lower prices and wages, leaving real wages unchanged, can allow the unemployed workers to enjoy less leisure and obtain more consumer goods. Real wages for the currently employed workers don’t need to fall.
In that situation, an expansion in nominal money balances (at zero cost of production) should have the same effect. The demand for consumer goods and labor both rise, and at the same level of prices and wages, the unemployed workers have less leisure and more consumer goods. They are better off. The already employed workers are left in the same situation.
If prices are flexible, and wages sticky, then this scenario is unlikely, we are back to the more standard scenario where labor markets aren’t clearing because real wages are too high. If the problem was a shortage of money, and output prices fell already, but wages didn’t fall, then real wages are too high. Raising the nominal quantity of money raises prices again, lowers real wages, and workers give up less valuable leisure for more valuable consumer goods. The “lucky” workers who benefited from the lower prices and higher real wages have their incomes fall back and the unlucky workers become employed and gain. The gains should be larger than the losses.
If the demand for money is extra high for any reason, the logic of these arguments is that it should be supplied. Not supplying it (say, because the level that should be supplied looks really, really high from a historical standard,) would be an error.
However, you need to consider that if we aren’t in a world where we just print money, spend it or give it away, and then who cares about the price level, but rather in one where stabilizing the price level, inflation, or nominal GDP is the goal, then all of this money might need to be withdrawn from circulation eventually. And so, they don’t create money and give it away. They purchase other assets with it. (Or it really forms a special kind of government debt.)
And that is when you get into these liquidity trap problems. Either you are back to lowering prices and wages to raise the real quantity of money to the demand, or else the monetary authority is working through the prices and yields of financial assets. Those accumulating the money spend less, and then somebody or other responds to higher prices of financial assets and lower yields by spending more.
If all money took the form of deposits, and paid interest, then this really wouldn’t be a problem. Reduce the yield on money, negative if needed, so that markets clear. Supply the amount demanded at a yield that reflects the cost of providing intermediation services, which is not the same thing as near zero cost of data entry. That only the sole cost when you print, spend, then not worry what happens to its future value.
Because hand to hand currency typically has a zero nominal yield, and making the nominal yield negative is really difficult, then issuing it a way where its future value is important, can sometimes be a problem. And that is what the liquidity trap is really about.
@Steve_Landsburg: I’m not saying, “You’re wrong because your argument leads to a false conclusion.” I’m saying, “You’re wrong (on a sub-argument) about what Keynesians believe, because their position doesn’t force them to come to that conclusion.”
Dispute my arguments on that point all you want; they were certainly responsive.
@Karl_Smith: I didn’t say Keynesians don’t advocate printing money. I said they don’t advocate it simply on the basis that it’s easy to produce and people want it. If that is actually what Keynesians believe, God help you all.
At the risk of a comment veto, Steve wrote:
“You demand green pieces of paper and the authorities supply them. There is necessarily an equilibrium in that market.”
What about an economy lacking sufficient specie? The smallest coin is worth more than the smallest transaction. This imposes extra costs and seems like unrequited demand to me.
Prof. Landsburg:
My reading of Krugman’s posts is the he is most upset that those economists/politicians who don’t accept that the new/old Keynesian model explains what is currently happening in our economy.
It is possible that these economists/politicians find it financially beneficial to refute a model that they know to be valid.
Perhaps, like yourself, Krugman has a very low tolerance for politically corrupt individuals (Obama, Bush, etc.).