Bryan Caplan  

"They'll Just Save It."

What Blinder and Zandi Don't K... Unemployed Houses: A Recalcul...

How many times have you heard the following argument in the last two years?

Tax cuts/helicopter drops of cash/whatever won't stimulate demand.  People are too nervous to spend.  Whatever you give them, they'll just save it.

The problem with this claim, like the analogous argument about reserves, is that it never considers the savers' motives.  Why are they saving in the first place?  Once again, there are two theories:

Theory #1: People just don't have anything they want to buy.  They're satiated, so no matter how much extra income they get, they'll just sit on it.

Theory #2: People want a buffer.  They aren't comfortable with their current asset cushion, so they're saving in order to return to their comfort zone.

Theory #1 is wholely implausible.  There's tons of stuff that people still covet.  The truth, then, lies in Theory #2: People will start spending again once they feel like they've got enough breathing room.

So what?  Well, if you believe in Theory #2, tax cuts/helicopter drops of cash/whatever do much more to stimulate demand than they appear.  Even if they don't persuade anyone to actually spend more, they move people closer to their financial comfort zone.  And once they reach that zone, they'll start spending again!  Even seemingly ineffective efforts to boost demand reduce the time we'll have to wait before demand begins to rise. 

If your goal is to stimulate demand, then, the right implication to draw from the "They'll just save it" mantra isn't fatalism.  Instead, you should ask yourself, "By how much do their savings need to rise before they will start spending again?" - and ramp up your tax cut/helicopter drop/whatever accordingly.

I'm well-aware that stimulating demand isn't everything, and has its own dangers.  My point is that a plausible account of savers' motives shows that, contrary to many, purely demand-based problems have been and remain easy to solve.

Comments and Sharing

COMMENTS (38 to date)
Piet le Roux writes:

Theory #3: People want a buffer. They aren't comfortable with their asset cushion and thus saves to get closer to it. Working against (in opposing force) to their successfully getting closer to their comfort level, is government action. Because they mistrust (i.e. expect bad consequences from) government's "[t]ax cuts/helicopter drops of cash/whatever" more than they are comforted by whatever cash injection they receive, they are not getting closer to their comfort level.

The situation continues and people become more and more uneasy, until the balance of forces changes through some crisis.

Reidar Foss writes:

But what if the stimulus increases people's comfort zone as well? If people anticipate that the increase in disposable income will be recouped with interest sometime in the future. If this is true than any stimulus will just increase the buffer, and not result in movement towards the comfort zone.

This is not an unlikely scenario for the US with its giant deficit. Sooner or later debts need to be settled.

Troy Camplin writes:

Why, what you say is such common sense that . . . Krugman will never believe it.

John S Cook writes:

At long last - he's said something sensible.

It makes sense also to direct it to people who may really need the money to purchase things - or even reduce debt and the interest they need to pay. There is reason to suspect that second and subsequent rounds of effects can then occur.

endorendil writes:

I think theory #1 should read
"People just don't have anything they need to buy. They aren't satiated, but they have no urgent need to get something "new" or replace something they already have."

It's no hardship to skip on the new iPhone when you have a perfectly good one that's only a year old. People are likely to reduce their reliance on credit, permanently changing their consumption behavior. I doubt that we're going back to an all-cash "save before you buy" economy, but it's quite possible that consumer credit will contract for a long time.

An added problem is that the boomers are about to retire en masse. They need to up their savings or face a dire retirement future. They can't rely on social security when deficits are so high. Since they have to compete in a labor market that's very tough, they can't really rely on their jobs either. So they will try to save whatever you throw at them.

Young workers are another problem. They are the most likely to permanently change their ways. A prolonged recession like this one is bound to make them more cautious than any post-war generation has ever been. On the bright side, they have many unfilled needs, so throwing money at them should lead to additional consumption, but it will bring them closer to the point where they start saving...

Throw money at people in the middle (not too young, and not too old) and they will pay down debts. That makes them feel better. Whether it brings them closer to the point where they start spending again is debatable. Suppose you've worked yourself into a financial hole you can't get out of, say 20 grand deep. If you get a couple of hundred dollars, you're likely to spend it and keep digging your hole until you are bankrupt - overspending all the way. But if you get 10 grand, you may take that as an opportunity to get out of the hole, by starting to pay down your debt and save money. Perversely, if you give people too much money, you may bring them closer to the point where they start saving.

The best way is to target the money at the poor and unemployed, who will be most likely to spend all of it. This is what a social safety net is for. If they want to really get the US economy on a safer footing, they should rebuild and expand the safety net.

P Ormosi writes:

And if people anticipate a fiscal policy that guarantees their buffer (and is credible enough) then they may start spending even before reaching the safety zone knowing that their respective safety zones are always within reach courtesy of the 'tax cuts/helicopter drops of cash/whatever'

Jeremy, Alabama writes:

I think I would prefer to see #2 include a reference to household debt. For many people, they are not "just saving it", they are paying off debt.

And in terms of savings, Spengler articles (on atimes) have convinced me that middle-age folks (say, late boomers/early X'ers) have suddenly realized that they are not going to reach their retirement goals by waiting for their house to appreciate or for the Dow to reach 25000. They have an "insatiable appetite for savings".

wintercow20 writes:

Color me awfully puzzled. How can it be the case that when Joe the Plumber gets a tax refund check he is inclined to save it, yet when Joe the Plumber is hired to build a $1 million outhouse in Glacier National Park, that same amount of money he receives will be spent?

The fact that G rises first in the latter is just an accounting trick.

Richard writes:

Don't we have to distinguish between saving and hoarding? Saving these days means putting money in the bank, not hiding it under a mattress. And more money in the bank can mean more lending, which increases GDP. Only when somebody (the person or the bank) actually hoards the money -- sits on the cash -- is the monetary expansion ineffective. Right?

Pat writes:

Who pays for the helicopter drop down the road? The taxpayers.

When the government forces you to borrow money from yourself, it doesn't help your asset cushion.

jb writes:

Earlier this year, I had a significant cash surplus, because I was saving up to buy a house.

If I had gotten helicopter money, I would have put it into two mental buckets - 1/3rd was reserved for future taxes, the rest would have gone to helping me buy that house.d

Of course, I don't have an account set up for "future tax burden" - I just mentally discounted the value of the helicopter money. And in my particular case, the helicopter money would have helped me buy a house sooner, which would have increased the amount of my mortgage tax deduction for the year, which would itself offset a fair amount of the helicopter money.

Right now, I have the house, and I ran the credit cards up a bit to keep some cash in the bank while I bought various things for the house, paid for movers, other unanticipated expenses, etc.

If I got helicopter money today, I would use it to pay down my credit card debt, so I could buy stuff sooner in the future.

In October, I'll have the credit card debt fully paid off. If I receive helicopter money then, I'll probably use it to buy christmas and birthday presents.

If it came in January, I'd pay off any credit card debt I accumulated to buy presents. If it comes any later than that, I might save it up for a vacation, or to buy that new big screen TV I've been eyeballing, etc.

The point is - there are all sorts of things that I would buy with helicopter money - if I save it, I would only save it for a few months as part of the process of saving up for some big-ticket purchase.

Will I have to pay for this helicopter money in the future? Of course, unless I die, or unless my income decreases drastically. Both of which are worse problems than paying back helicopter debt! But I won't have to pay it back all at once - it'll be leveraged over several years.

Morgan writes:

So we should concentrate the helicopter drop - say $100 million on each of just a few thousand households, or $500 billion in total.

For them it's like winning the lottery. Suddenly fabulously wealthy, and in a position to spend the vast majority of the drop.

Several other posters have already said this, and said it very well, so I'll just repeat it briefly before adding one minor additional point:

That "comfort zone" that people have is a moving target. If a dollar of helicopter-drop moves the comfort zone out a dollar then every penny of that dollar is going to be saved.

Now it is often said that people are not sophisticated enough to figure out that a dollar of helicopter drop today means (in present value) a dollar of additional spending tomorrow---and therefore people won't save all the money. But the "therefore" part doesn't follow. Unsophisticated taxpayers might just as well overestimate the effect of the helicopter drop (particularly if they're subjected to media bombardment over the upcoming deficit-fueled catastrophe) and so save more than a dollar.

To overcome Ricardian equivalence, it's not enough to assume that people are unsophisticated. You need a model of the particular way in which they're unsophisticated. Tweak that model and you can get a stimulus having a negative effect on spending.

Pat writes:

Morgan, those few lucky people would still save the bulk of their windfall for future consumption - consumption smoothing.

You'd need to attach some kind of expiration date on the windfalls - something like gift cards.

Yancey Ward writes:


+100. I have yet to see a single critic using the "they will just save it" argument actually address your completely logical point.

Pat writes:

Wintercow, Joe the Plumber's proceeds from a million dollar outhouse would be far more than his future tax liability. He'd be wealthier at all of our expense and he'd consume more.

But when the government sends me $600 and I expect my taxes to go up $600, I'm not wealthier and I don't spend more.

Silas Barta writes:

Excuse me, can we take a step back and go over why we need to "get people spending" in the first place?

How, exactly, does Bryan Caplan (or anyone) know that the efficient adjustment at this point isn't for people to save more and/or start providing services for themselves rather than buy them (i.e. cook their own food, etc.)?

Seems no one questions this basic premise...

Yancey Ward writes:


But most beneficiaries of government stimulus spending will accrue additional incomes that are not that different from the standard tax rebates we see. I simply don't see where the additional benefit comes from by spending the money as tax cuts vs spending to build outhouses, or whatnot, other than the fact that the second counts as GDP and the former doesn't immediately. Is one going to treat $1000 in tax cuts differently from $1000 in additional income from a government financed job? If so, why?

Tom Dougherty writes:

I think a lot of the comments are confusing monetary policy with fiscal policy. Bryan's mythical helicopter drop is a euphemism for monetary policy. If the Fed increases the money supply will people spend it or save it? With an excess demand for money, people are not spending much money, but as the supply of money increases the excess demand for money is reduced or eliminated and people will feel good about spending again.

Many commenters’ are worried about the effect on the deficit with the "helicopter drop". But this is a problem with fiscal policy. Government stimulus programs and spending cause deficits and future higher taxes. People may decide save today in order to pay for future higher taxes tomorrow. However, fiscal policy and deficit spending is not what is meant by a "helicopter drop".

Pat writes:

Yancy, I'm not arguing for tax cuts as stimulus either. I don't think any of it works. If everyone was given awarded useless government outhouse building jobs, then everyone would save the proceeds to pay for increased taxes. If your windfall is the same as your future tax increase, you're going to save it, whether it comes in checks, tax cuts, or government contracts.

Michael (Brady) writes:

Just as economists understand that there is no free lunch, people understand that there are no free "tax cuts, helicopter drops of cash, etc." As "Piet le Roux" explained, any increase in consumer savings will be offset by the percieved need for even greater savings as people anticipate and work against government action. Therefore, although consumers' savings may increase; there is a very disproportionate and negligible increase in demand. People understand that stimulus money is artificial in the sense that any resulting economic growth will be temporary and that the "drops of cash" they recieve will unnecessarily add to the national debt.

Tom Dougherty writes:

My mistake. Bryan does mention tax cuts to stimulate demand. The comments above are right to worry about the future tax burden due to increase deficits. This is also the reason why monetary policy should be the preferred method to increase spending at this time. Increasing the money supply to meet people excess demand for money will cause people to spend more WITHOUT increasing the deficit.

The important question is how to stimulate demand at this time - monetary policy or fiscal policy. The clear winner is with monetary policy. This can be done without an increase in future tax burdens.

Pat writes:

Tom, inflation is a tax burden too. Helicopter drops are good for preventing deflation, nothing more.

Philo writes:

Bryan talks of a "comfort zone," and he employs a simple dichotomy: spend/don't-spend. But actually these are matters of degree. Thus tax cuts and helicopter drops will promote *some* spending, and the more comfortable people become the more spending they will do. They won't spend 100% of the extra money they get, but the greater their comfort level becomes the greater the percentage they will spend. (Bryan's post assumes that tax-cuts/helicopter-drops *will* increase people's comfort levels; Pat and Landsburg, above, rightly question that assumption, but I am accepting it for the sake of argument.)

And the amount they don't spend they will save; why is that supposed to be bad (echoing Silas Barta, above)? I've noticed a lot of handwringing about Americans' low rates of saving. And savings are mostly invested, not hoarded (echoing Richard, above). Mightn't more investment be part of what the economy needs?

Tom Dougherty writes:


If there is an excess demand for money, then increasing the money supply to meet that excess demand is not inflationary. When the Fed fails to increase the money supply enough to meet the demand for money it is deflationary.

Tom Dougherty writes:


The Federal Reserve's goal should be to stabilize nominal spending. This should minimize macroeconomic volatility. I think one of the big reasons for the severity of this recession is that it failed to do that. Year-on-year percent change in aggregate demand had taken a nose dive in this recession not seen since the great depression. This being the case, what is the best way to boost aggregate demand - fiscal policy or monetary policy? Monetary policy can increase aggregate demand without large budgetary deficits that would be necessary trying to do the same thing with fiscal policy.

Silas Barta writes:

@Tom_Dougherty: But why is a) the low macro-volatility resulting from stable nominal spending, better than b) the *disadvantages* of stable nominal spending (SNS)? What if SNS requires people to perpetually engage in inefficient activities rather than endure a one-time, efficient readjustment to a lower SNS?

Extreme case (though more realistic versions of this happen all the time): What if technology makes it so that you only need to buy a few goods from other people, and some new gadget, plus your social network, give you everything else you need without having to make a money exchange?

Do you really think that the optimal thing for the Fed to do in that case is to engineer it so that those few goods become so expensive that they cost the same as people's *entire budgets* used to cost?

Is this focus on spending levels yet another case of what the Lucas Critique warned about, where a once-useful proxy is now no longer informative?

I have yet to see a satisfactory answer to these questions.

Tom Dougherty writes:


"Extreme case (though more realistic versions of this happen all the time): What if technology makes it so that you only need to buy a few goods from other people, and some new gadget, plus your social network, give you everything else you need without having to make a money exchange?"

I don't have time right now to answer all your questions, but for this one as I read it the demand for money is falling because people need to buy fewer goods and make fewer exchanges. In this case, if the demand for money is falling, the supply of money should fall as well to stabilize nominal spending. Otherwise, if the money supply does not fall then there will be a rise in the price level which will be inflationary. Inflation can create uncertainty and have negative effects on decision making.

Given that we have a Federal Reserve, what would you recommend they do in your example above?

Pat writes:


If we need a helicopter drop to prevent deflation, then great - go for it. (I don't want people to profit from stuffing money in mattresses either.)

I don't think of it in terms of directly stimulating the economy though. I think there are real reasons for the recession (Kling's recalculation story) and we shouldn't always just chalk them up to deflation.

Silas Barta writes:


If people are buying fewer things, how does decreasing the money supply maintain constant overall spending?

As for what *I* think the Fed should do, keep in mind the whole point is that the very question of "What should the Fed do?" makes assumptions about economic goals we have no reason to pursue. Things like "keeping up spending" are only good in very specific situations, and we have no reason to believe we're in one of those situations, nor that we wouldn't be amplifying inefficiencies by trying.

But if the Fed had to do something, I say it should just randomly enter bond auctions and buy a very small portion of the issued bonds so that it's not distorting economic signals about the value of those bonds. Any interest earned, if they want it sent back into the economy, should be given in equal shares to all adults.

But it's a really moot point -- the first question is to check the assumptions no one is questioning about *why* it's a good thing for spending to go up.

Steve Roth writes:

The entrance to the comfort zone is not a hard line. There's a "curve."

As people get more money they spend some and save some. The farther they move into their comfort zone (by your theory) the more they spend on consumption.

The proportion devoted to saving is larger now, but something like 95% still goes to spending/consumption. Give someone a dollar, they'll spend 95 cents. Even if it's 75 cents for "found" money from the government, it's still a big bump to AD -- a far bigger multiplier than, for instance, crediting banks with extra reserves, or businesses with tax credits, in hopes that it will flow through into "investment spending."

This is also an opportunity to comment on two old posts that I've been wanting to respond to for a while -- the latter, with a link to the former, is here:

Me: Employers spend a smaller percentage of marginal earnings on immediate consumption (which would increase short-term AD).

You: They may spend a smaller percentage on immediate consumption, but AD isn't just consumption spending. It's also investment spending - and remember that due to credit market imperfections, cash flow is a good predictor of business spending.

This confused me and seemed wrong initially, but I didn't really know why. Now I do.

You're essentially asserting that savings = investment. That there's a 1:1 (or close) and immediate (or close) correlation between the two. That there's a direct, friction-free pipe, which of course there ain't.

As it quite clear from current excess reserves, corporate cash balances, and capital investment rates, that assertion is false -- even falser in times like these.

Steve Roth writes:

IOW, there's money supply, and there's velocity. We're short on the latter.

Tom Dougherty writes:


"I think there are real reasons for the recession (Kling's recalculation story) and we shouldn't always just chalk them up to deflation."

I agree with that, however, I don't think we should make a bad situation worse by allowing secondary deflation.

Jayson Virissimo writes:

Bryan, another way to ensure that the money would be spent is to give it out on debit card accounts that decay over time. For instance, the money in the account might decrease by 10% every two weeks. How many people would wait around until their account lost the majority of its value?

floccina writes:

Presumably there will be a yield on the savings above the 1% that the Gov. is currently paying, particularly the saving is really the paying down of debt. So I would expect people to spend some of it.

Doc Merlin writes:

Another theory:
Total nominal Income dropped (NGDP dropped), this mean total ability to pay debt down dropped. This means that a higher % of income must go to pay down debts. This looks like an increase in the savings rate, but really is just the result of long term debt and income both being nominal.

What I like about this theory is that we don't have to think deeply about the motives of actors in the economy.

Derek writes:

I'll add another one.

Helicopter drops, fiscal stimuli, are a great idea in theory. But when reasonable people see a $1.5 trillion deficit, the Fed printing multiple trillions keeping madmen in business, the rational thing to do is to sock more away for the inevitable rainy day.

If the stewards act like they are insane, any wonder why everyone hunkers down?

I have heard that word used by reasonable people too often to discount the results.


Doc Merlin writes:

Right now "saving" means paying down debt... which I see as a good thing in the long term.

Comments for this entry have been closed
Return to top