Scott Sumner  

Aggregate demand is not consumption

Robert Murphy on Economic Calc... "The Credit Card Companies Wer...

I occasionally see commenters talk about how monetary stimulus and/or low interest rates would merely shift demand from the future to the present. That claim never made any sense to me. Now I see that Paul Krugman is equally perplexed:

[Mervyn] King is, however, having none of it. Under his leadership, the Bank of England was aggressively engaged in monetary easing by keeping interest rates low--the bank was as aggressive in this respect or even more so than the Bernanke Fed. Now, however, King seems to condemn his old policies:
Monetary stimulus via low interest rates works largely by giving incentives to bring forward spending from the future to the present. But this is a short-term effect. After a time, tomorrow becomes today. Then we have to repeat the exercise and bring forward spending from the new tomorrow to the new today. As time passes, we will be digging larger and larger holes in future demand. The result is a self-reinforcing path of weak growth in the economy.
Is this argument right, analytically? I'd like to see King lay out a specific model for his claims, because I suspect that this is exactly the kind of situation in which words alone can create an illusion of logical coherence that dissipates when you try to do the math. Also, it's unclear what this has to do with radical uncertainty. But this is a topic that really should be hashed out in technical working papers.
I wonder if people making this argument aren't confusing aggregate demand with consumption. Recall that Keynesians often speak as though demand is "spending". At that point many non-economists get confused, equating "spending" with consumption, as in "spending, not saving".

But that's not as all what Keynesians mean by "spending." This is easiest to see in a closed economy context where S = I, by definition. Keynesian aggregate demand is not C, it's C+I+G. Rather than think of that concept as spending, it makes more sense to view it as production. It is the total dollar value of output, in a given year. If people save more (ex post), then they also "spend" more on investment goods.

In contrast, in our daily lives we think of there being a choice between spending today and deferring spending into the future. But in that case we are using the term 'spending' in a very different way from how it used by Keynesians. We are referring to consumption, not total nominal output. If we consume more today, we will have less wealth available to purchase consumption goods in the future.

At the macro level, however, things look very different. If low interest rates lead to an economic boom, then investment spending tends to actually increase as a share of GDP. Not only does the higher level of current consumption not come at the expense of future consumption, it actually leads to more future consumption, as our capital stock is larger. Of course it must come at the expense of something, and that something is leisure time. Our sticky wages (or money illusion) fool us into working more.

King would have been better off talking about "bringing forward" labor effort. If we work harder today we might burn out, and not want to work as hard in the future.

I think this is part of a general problem in macroeconomics, where people try to visualize things using common sense analogies from their daily life, when thinking about macroeconomic policy. But it just doesn't work. Thus many people visualize monetary stimulus as low interest rates leading to more "spending". But monetary stimulus is more likely to raise interest rates, and the actual effect occurs due to a combination of higher NGDP and sticky nominal wages. Unfortunately, those concepts don't relate as easily to everyday life, and hence people use false analogies, such as, "I'll be more likely to qualify for a mortgage if Janet Yellen cuts rates".

Comments and Sharing

COMMENTS (8 to date)
bill woolsey writes:

The models typically are consumption only models.

They are based on inter-temporal optimization, and lower interest rates do bring desired consumption forward. That is an individual experiment.

But there is now growing "hole" in the future.

If done right, the interest rate is set such that desired saving is zero and consumption equals potential output in each and every period.

This period everyone consumes their entire income which. Next period everyone does the same.

In these economies there is no borrowing or lending. If the interest rate is too high, everyone wants to lend. If it is too low, everyone wants to borrow. If it is just right, there is no lending, no borrowing, no saving, no dissaving, and everyone consumes all of their income.

Suppose you assume that lower interest rates increase spending by causing people to go into debt to fund consumption. Everyone is dissaving. They are bringing consumption forward and will have to save in the future to pay down the debts. There is a growing hole.

What this ignores is that there are lenders. They are saving and accumulating assets. We are building up this use potential increase in future demand when they sell off those assets and spend the funds.

Finally, it is probably true that low short term interest rates do move investment forward. If you were going to buy a capital good anyway soon, if rates are low now, moving up the timing is advantageous. Of course, it is better to think about it as long term rates being low now because the relevant period includes a time--now--when short term rates are low. Start the project now when rates are low.

However, more investment projects become profitable at lower rates. I don't see how that is a move of demand for capital goods to the present. They do increase output and income in the future. So, there will be a need for extra demand in the future too. But the higher income leads to more demand.

Anyway, it is better to understand that interest rates coordinate saving and investment (and in the consumption only economy, coordinate savers and dissavers so that on net, there is no saving and consumption equals income.)

If interest rates are too low, sure enough, we are creating an imbalance in the future. But that is a situation where demand is too high for productive capacity right now.

If interest rates are too high, then the problem is that people are pushing consumption demand into the future and not building a capital stock sufficient to produce the extra future consumer goods. Lowering interest rates is solving this problem, not creating the opposite problem. And, of course, people not consuming now and not buying capital goods now means that demand is below productive capacity.

Benjamin Cole writes:

Excellent blogging.

Scott Sumner writes:

Bill, You said:

"They are based on inter-temporal optimization, and lower interest rates do bring desired consumption forward."

Of course it depends why interest rates are lower. Any model that doesn't account for that is reasoning from a price change. I really wonder why someone would even develop a model like that. What do they think they are explaining?

This would all be much simpler if people would think in terms of the time path of production, not "spending".

Thanks Ben.

José Romeu Robazzi writes:

Aren't Mervyn King's words being mentioned out of context? Isn't he talking about government spending?

Patrick Trombly writes:

The main problem with this analysis is that it relies on aggregates. All investments are not equal, and they do not all "add to the capital stock." Artificially low interest rates have a history of fueling bubbles - malinvestments - which are just wastes of capital. Bubbles don't build on each other. The malinvestment in housing stock and, on the part of worker-consumers, in construction skills, did not create something on which the economy could build - the subsequent oil bubble, brought about by QE as banks' trading desks sought an inflation hedge and then banks' lending arms increased lending to oil drillers, was not built using the wreckage of the housing bubble as a foundation. That's why "kick starting the economy" never "gets it going" for a long enough time for all that "multiplier effect" income to repay the debt that gave rise to it - - the boom never leaves the sector in which it is created - you may get follow on booms in feeder sectors (e.g., housing leading to a boom in copper production), but the GDP growth is just the bubble. The "positive GDP response" to Greenspan's low rates WAS the housing bubble, for example - to call that response a positive is to take someone who is having trouble gaining body mass, prescribe a carcinogen, and then refer to the resulting tumor as "body mass growth."

The mechanics by which artificially low rates fuel investment bubbles is the WACC vs IRR equation - simply put, consumers save less and spend more, thus increasing current consumption, which may be confused for higher potential IRR, while Rf falls, causing both cost of debt and cost of equity to decline. But cost of equity also has a volatility component - beta. Over time, serial boom-bust cycles result in an increased beta, which offsets the rate effect in all but the most rate-sensitive sectors (like construction). But the evolution of consumer finance over the last century, coupled with the impact of beta on supply side malinvestments, has resulted in low rates affecting consumption more than anything else - which is part of the reason that "mature economies are consumption-based" - it's no intrinsic; it's the result of policy.

In the 2000s, artificially low rates first inflated house prices - by increasing the amount of debt in the hands of home buyers by 50% - and then, on a low volume of home sales, produced a consumption boom. Americans doubled their mortgage debt, largely through cash-out refis. They financed current consumption with 30 year mortgages. And an increasing share of what they bought was made in China using materials imported from Brazil, Africa, Australia... Today's "insufficient aggregate demand" and "China slowdown" reflect the simple fact that that one-time pulling-forward of consumption through the printing and lending of new money was financed with 30 year mortgages that are still being paid off.

Britmouse writes:

José, no, that is definitely his view on AD, he has expressed the same view in numerous speeches when BoE Governor. e.g.

Governor Shirakawa of the Bank of Japan also made similar arguments.

Silas Barta writes:

Sorry to hear about your frustration with this misconception. I guess that's what happens when these very same Keynes-citing advocates are blending such advocacy with factoids like "70% of 'the economy' is consumer spending, so we have to keep that up to keep up aggregate demand". I hope you allocate some of your efforts to refute this similar equivocation.

James Alexander writes:

I also think that part of the experts vs ordinary folk confusion arises from the idea that a recession is caused by an (unmet) increase in demand for money. Where does that money go? Isn't it saved? And if it's saved, isn't it invested, too?

Comments for this entry have been closed
Return to top