Insufficient Demand vs?? Uncertainty

It's a false dichotomy. Just a quickie, in response to Noah Smith and John Cochrane.

John Cochrane says: "John Taylor, Stanford's Nick Bloom and Chicago Booth's Steve Davis see the uncertainty induced by seat-of-the-pants policy at fault. Who wants to hire, lend or invest when the next stroke of the presidential pen or Justice Department witch hunt can undo all the hard work? Ed Prescott emphasizes large distorting taxes and intrusive regulations. The University of Chicago's Casey Mulligan deconstructs the unintended disincentives of social programs. And so forth. These problems did not cause the recession. But they are worse now, and they can impede recovery and retard growth."

OK. Let's assume that increased political uncertainty caused a reduction in willingness to "hire, lend, or invest". That sounds plausible to me. The sign is right, but I don't know about the magnitude.

A monetarist would say that would increase the demand for money, and that would cause a recession, unless the central bank took sufficient offsetting action.

A New Keynesian/Neo Wicksellian would say that would reduce the natural rate of interest, and that would cause a recession, unless the central bank took sufficient offsetting action.

Increased political uncertainty would reduce aggregate demand. Plus, positive feedback processes could amplify that initial reduction in aggregate demand. Even those who were not directly affected by that increased political uncertainty would reduce their own willingness to hire lend or invest because of that initial reduction in aggregate demand, plus their own uncertainty about aggregate demand. So the average person or firm might respond to a survey by saying that insufficient demand was the problem in their particular case, and not the political uncertainty which caused it.

But the demand-side problem could still be prevented by an appropriate monetary policy response. Sure, there would be supply-side effects too. And it would be very hard empirically to estimate the relative magnitudes of those demand-side vs supply-side effects. Looking (as Noah suggests) at whether inflation falls (demand-side) or rises (supply-side) might not tell us the whole story. Because anything which causes shocks to relative prices can, with nominal rigidities, shift the Short Run Phillips Curve adversely more than it shifts the Long Run Phillips Curve adversely, so keeping the economy on the LRPC might require a temporary increase in inflation. (We saw this for example in Canada when the GST (VAT) was introduced to replace a less efficient tax, and we see it whenever the relative price of oil rises.)

So it's not just an either/or thing. Nor is it even a bit-of-one-plus-bit-of-the-other thing. Increased political uncertainty can cause a recession via its effect on demand. Unless monetary policy responds appropriately. (And that, of course, would mean targeting NGDP, because inflation targeting doesn't work when supply-side shocks cause adverse shifts in the Short Run Phillips Curve.)

Comments

[Market Fiscalist]

Suppose a CB is targeting a specific level of nominal GDP growth. The govt introduces measures that cause uncertainty (random future taxes) and increased costs (more regulations).

The new govt measures cause AD to fall, and the CB, spotting this, increases the money supply to keep NGDP growth on trend.

It seems intuitive to me that the net result will be stable NGDP growth, but with an increase in the price level and a fall in RGDP to reflect the higher costs and greater uncertainty introduced by the govt. The govt policies have changed the equilibrium RGDP level downwards. It will probably also have changed the future inflation and RGDP growth trends that will result from hitting the NGDP target.

Sound monetary policy can minimize the effect of govt policy and stop it becoming a recession but I don't see how it can eliminate supply-side effects altogether.

[Ralph Musgrave]

Uncertainty ought to have no effect because any employer is happy to act in an uncertain environment just as long as the employer gets enough profit to compensate for the uncertainty. Employers are happy to do business in war-torn politically unstable countries just as long as they get 100% - 200% - 300% return on capital.

[Alex]

Nick,
John Cochrane has removed the post from his blog. Please fix the link.

[Jerry Brown]

How does a reduced willingness to lend cause a reduction in the rate of interest? And how would the reduction in interest rates cause a reduction in aggregate demand?

[marcus nunes]

The alternative link is to his WSJ op-ed which gave rise to the post. The paragraph above is there:
http://online.wsj.com/articles/john-cochrane-new-keynesian-macroeconomic-models-dont-support-more-stimulus-spending-1404342631

[Kevin Donoghue]

Odd that John Cochrane removed the post. Here's a link to the Google cached version:

webcache.googleusercontent.com/search?q=cache:WKnCONhl79gJ:johnhcochrane.blogspot.com/2014/07/macro-debates.html

[Vaidas Urba]

"Increased political uncertainty can cause a recession via its effect on demand."
True. There is another different and less important channel, where increased political uncertainty can increase the risk of recession via its effect on uncertainty about demand even when expected NGDP is on target.

[Nick Rowe]

Market Fiscalist: "Sound monetary policy can minimize the effect of govt policy and stop it becoming a recession but I don't see how it can eliminate supply-side effects altogether."

Yep. I agree. (Perhaps I should have made that explicit.)

Ralph: OK, but what would need to happen to real wages and employment in order to give employers the extra returns they would need to compensate them for the uncertainty? And would nominal wages fall instantly to enable that to happen? I doubt it.

Alex and Kevin: maybe it's because it was a Wall Street Journal piece also? And maybe WSJ didn't like it appearing twice? I'm trying to find the WSJ version to link, but failing. Or is my memory failing?

Jerry: Fair point. I think we need to distinguish safe vs risky interest rates. I interpret the reduced desire to lend being a reduced desire to lend to risky borrowers. That would increase the natural rate of interest on risky loans, but reduce the natural rate of interest on safe loans. The risk premium wedge has just increased. And if the central bank sets the safe interest rate in an NK model, it would need to reduce that interest rate to prevent a fall in demand.

Vaidas: Hmmm. Interesting point. I had to think it through to get it. So actual current NGDP would fall a bit even if the central bank kept expected future NGDP constant.

[Alex]

Nick,
http://online.wsj.com/articles/john-cochrane-new-keynesian-macroeconomic-models-dont-support-more-stimulus-spending-1404342631

[Majromax]

@Nick:
> I think we need to distinguish safe vs risky interest rates. I interpret the reduced desire to lend being a reduced desire to lend to risky borrowers. That would increase the natural rate of interest on risky loans, but reduce the natural rate of interest on safe loans.

That could go either way, though. Depending on what form the uncertainty takes, it could directly increase the desire to hold cash (as a liquidity hedge against unexpected things) compared to debt instruments irrespective of risk. We might even see an impact on term structure:

* Risky debt would have a higher interest rate, for the reasons you mention
* Long-term safe debt may have a slightly higher natural rate, since holding that debt reduces a firm's liquidity
* Short-term safe debt may move either way, depending on how "short term" compares with the perceived risks. If the government might do something crazy tomorrow then even 30-days is long-term; if the policy risk is months away then it might be fine.

[Nick Rowe]

Alex: Thanks! Link changed to WSJ.

Majromax: OK. Yep, when we talk about a reduced desire to lend, we really ought to specify what people want to do instead.

[Vaidas Urba]

Nick, see my old guest post at Lars' blog for some related points:
http://marketmonetarist.com/2012/12/22/guest-post-market-montarism-and-financial-crisis-by-vaidas-urba/

Note that VIX and other market measures of uncertainty are very low at the current moment...

[glasnost]

The causes Cochrane cites are completely witless. They boil down to 'the economy is suffering b/c the world hates democrats/the welfare state'. We have divided government, uncertainty about policy could hardly be lower. When President Clinton was being impeached, policy directions could hardly be more uncertain, but the economy didn't care. There is essentially zero uncertainty penalty for policy or leadership 'character' in almost any situation, short of massive violent unrest.

Economic actors respond to specific policy threats and debates that immediately threaten their welfare, but this is the kind of generic sawhorse not worth discussion.

Taking it seriously even for the sake of discussion does more harm than good.

[Tom Brown]

I like Benjamin Cole's comment here and foosion's comment here.

[Alex]

John Cochrane asks why recovery from the Great Recession is slow.

Perhaps, he should have tried to understand first why it was slow after the Great Depression. And then compare results here and there.

It looks to me that the causes of slow recoveries are somewhat different.

[Andrew_M_Garland]

Try to take an objective look at the last two paragraphs above starting "But the demand-side problem could still be".

These are a soup of possible actions and effects, and admitted ignorance about the resulting effects and how strong or weak the actions should be. It is not quantitative and it is not science.

It is much like a medieval doctor spinning various theories about the body, saying that he must do something, and settling on leaches, cathartics, or potions according to the most recent fad.

Bloodletting was a common therapy which usually killed the patient faster than the disease. To complicate things, bloodletting is actually useful when iron accumulates in the blood. Iron rich blood is removed and the body replaces it, lowering the iron concentration. The doctor must know exactly how much to remove over what time periods.

The point is, the activities of millions of people, their plans, and investments (the economy) are immensely complicated. How are broad interventions by a few "experts" supposed to help, when the experts have fragmented understanding and a few blunt instruments, facing political manipulation of the programs which are implemented? You can know many things with local good effects, but which interfere with healing and kill the patient faster than the disease.

The Soviet Union had planners in depth and multiple 5-year plans. They centrally planned themselves into poverty. Where is the evidence that our central planners in finance are any better?

[Ralph Musgrave]

Nick,

You seem to agree that uncertainty is no problem as long as employers are compensated by way of extra profits. And as you rightly say, that requires a fall in wages, or at least a fall in wages relative to profits. You then say “And would nominal wages fall instantly to enable that to happen? I doubt it.”

My answer is that I very much doubt there has been an “instant” increase in uncertain or unpredictable activities by politicians. Second, I suspect employers complaints about uncertainty are actually objections from politically right of centre people to politicians passing any legislation at all that affects businesses. Third, the employer surveys I’ve seen don’t put uncertainty high on the list of employers’ concerns. Their main concern over the last three years or so has been plain simple lack of customers, i.e. lack of demand.

Thus I suspect any increased uncertainly has been small in scale and has arrived slowly, thus there probably is or has been time for wages to fall relative to profits. But obviously I cannot prove that.

[Market Fiscalist]

"We want to get to the root of the problem of insufficient aggregate demand. We see aggregate demand as a monetary phenomenon, that only makes sense conceptually in a monetary exchange economy. And we see insufficient aggregate demand (just like an inflationary surfeit of aggregate demand) as due to an underlying failure of monetary institutions"

This is a statement that I can get 100% behind.

I'm just not sure that I draw the conclusion that I take to be implicit that monetary institutions should operate solely by swapping base money for other assets. Particularly as interest rates approach zero I question if such asset swaps are the most efficient way of managing aggregate demand. There is a worry that when the monetary authorities have to buy riskier assets in order to affect the real economy they will start to mess with people's assessment of risk and there will be a potential (if markets are not optimally efficient) to drive asset-price bubbles.

For this reason I think a rules-based aggregate-demand management policy needs to include a combination of "monetary policy" (assets swaps and expectations settings) and "fiscal policy" (adjusting the size of the deficit via rules-based adjustment to , for example, income and sales -taxes).

[Min]

"Increased political uncertainty would reduce aggregate demand."

Can you spell that out, please? Thanks. :)

[Glen Tomkins]

What these people mean by "uncertainty" is entirely uncertainty about rents. You might as well be writing about astrology in Sanskrit if you veer off their page and talk about uncertainty in demand.

[Min]

Glen Tomkins: "You might as well be writing about astrology in Sanskrit"

Sahib knows Jyotisha Vedanga?

[Susanto Basu]

Nick,

Agreed fully that increases in uncertainty can be thought of as adverse demand shocks. Brent Bundick and I wrote a paper a couple of years ago to make just that point: http://www.nber.org/papers/w18420

The mechanism in our setup works through precautionary saving: higher uncertainty causes people to save more. This reduction in consumption demand leads to lower output, labor input and even lower investment (that is, an increase in desired saving leads to a reduction in actual saving in our closed-economy model).

It is this desire to save more that causes the natural safe interest rate to decline; you are right that the risk premium will increase. Note that once past the period when output is driven by demand, higher uncertainty will tend to raise rather than lower output! That is, higher uncertainty often raises supply.

Bundick and I considered technological and demand uncertainty, not political uncertainty. But we found that the effects of uncertainty on the economy are much the same regardless of the source of uncertainty. This paper by Jesús Fernández-Villaverde, Pablo Guerrón-Quintana, Keith Kuester and Juan Rubio-Ramírez, who find similar results for fiscal uncertainty, makes me think similar results will hold across a broad range of causes for higher uncertainty: http://www.nber.org/papers/w17317

Cheers,

Susanto Basu

[Glen Tomkins]

Min,
Sadly, no. But I suspect that Aristophanes is probably more useful in understanding this dispute.

[Min]

Susanto Basu: "The mechanism in our setup works through precautionary saving: higher uncertainty causes people to save more."

Good point. But does that hold for political uncertainty of the ordinary kind? Do savings rates go up before close elections, for instance?