[jvm]

The advantages of monetary stability come with a stable growth rate, because contracts are bets on the future growth rate of money and stability means those bets can be made reliably.

You have proposed a stable CPI growth rate of 0%. There are two problems with this: the first (more important) problem is that arguably stable money supply is more desirable than stable CPI. During a downward supply shock, tightening the money supply can lead to a financial crisis by causing contracts made before the shock to fail. In that instance, rising CPI is good because it reflects a real fall in supply (prices should be higher).

The second problem and the reason why they shoot for 2% CPI inflation is that many prices are sticky downwards so an average CPI growth rate of 0% leads to problems. For example, people are known to find a paycut of 1% under 0% inflation more aversive than a pay increase of 1% under 2% inflation. As a result, during deflationary periods wages are often frozen at levels higher than equilibrium, leading to greater unemployment.

Despite these reservations, I agree that 0% CPI growth would be an acceptable policy, particularly as opposed to erratic policy as we saw in 2008–9. But you should recognize that moving from 2% to 0% would itself be a strong downward demand shock that in most countries would likely cause a recession.

[goodcanadian]

<quote>But you should recognize that moving from 2% to 0% would itself be a strong downward demand shock that in most countries would likely cause a recession.</quote>

Any sudden policy change is going to lead to turmoil and a likely recession. I did not suggest the change should be sudden.

Also, a target of 0% does not prevent appropriate price increases or decreases any more than a target of 2%. It is a goal not a mandate. As long as everyone knows what to expect, plans can be made reliably.

[jvm]

> As long as everyone knows what to expect, plans can be made reliably.

Expectations are the flaw with CPI targeting.

Toy example: imagine NGDP is $100 and the CPI is 100. Fed guidance says 0% CPI growth. I have income $10 (note that NGDP is aggregate nominal income). I sign a deal with you for you to finance my house and promise you for $7 next year, assuming I"ll live off of $3.

Now a supply shock hits, with enough pressure to raise CPI to 110. The fed responds by dropping NGDP so now national income is $90 (I know the math here isn't exact but it's irrelevant to the point). If I'm the average person, I take a 10% income hit, dropping me to $9. I still owe you $7. My consumption now has to drop from $3 to $2, a 50% drop! I might not be able to do it so I might default on my loan. In any case I'm really screwed, and if I do default (and many other people also default) we'll have a financial crisis on top of our supply shock. One way to think about this is that the debtor has to eat the creditor's consumption haircut. Europe, cough cough.

In contrast, under NGDP targeting, NGDP remains the same, prices rise 10%. Now my income will still be $10, I will pay my creditor $7, and my only hit is the 10% inflation. My creditor has the same exposure.

I'm not trying to create sympathy for debtors, of course you could argue that they were over-leveraged and deserved what they got. But under NGDP targeting, debts will always be repayable in principle (because enough money will exist in the system) so systemic risk is minimized; minimizing that systemic risk is good for everyone.