The Optimal Rate of Inflation
The Bank of Finland asked me to respond to a survey about the optimal rate of inflation. (This was the sort of survey that is sent to those who might have a professional opinion about the optimal rate of inflation.) I thought I’d share my answers here. I give their questions in bold. I give my answers in italics.
Should the central bank have an explicit inflation target? If so, what rate of inflation should it seek to achieve, given the current longer-term, structural economic trends?
Yes. 0%.
My answer of 0 inflation as the best target is assuming what I consider the appropriate strategy of being willing to use deep negative interest rates. I have written about that here:
All the pieces flagged here provide context:
Imagine a hypothetical scenario in which the central bank had previously not adopted an inflation target but now decides to adopt one. What rate of inflation should the central bank target, given the current longer-term, structural economic trends?
0%.
The argument is the same. The readiness to use deep negative rates (which is the appropriate strategy) makes an inflation target of zero optimal. But it is also important to do the transition from a higher inflation rate to zero in a low-cost way. Distinguish between paper currency and bank money ("electronic money"), allowing a nonpar exchange rate between them. While most prices are sticky in the paper unit of account, mechanically make e-money have a zero rate of inflation. Having established credibility in that way, it is likely that gradually more prices will be set in terms of the e-money unit of account. That can be encouraged by shifts in the literal accounting rules. As more and more prices are set in terms of the e-money unit of account, zero inflation in that e-money unit of account needs to be maintained by standard monetary policy tools. That can no longer be done mechanically when almost all prices are set in terms of the e-money unit of account.
What should the central bank's objective(s) be?
Please choose only one option.
Here, ‘central bank’ refers to the central bank responsible for monetary policy in your country of residence.
Price stability only
Price stability and other objective(s) with equal weights.
Price stability and subordinate objective(s). Please feel free to specify the secondary objective(s)
No opinion
Price stability and other objective(s) with equal weights.
Please feel free to specify the other objective(s):
Keeping the output gap equal to zero
Although not literally true, I consider the "divine coincidence" a useful rough approximation. Price stability and keeping the output gap at zero are the same objective when the "divine coincidence" holds, but to the extent that the divine coincidence does not hold, a central bank should worry about both objectives.
Among the options below, what specific observable variable(s) would be the most preferable target(s) for the central bank in the conduct of its monetary policy?
Please choose only one option.
Here, ‘central bank’ refers to the central bank responsible for monetary policy in your country of residence.
The inflation rate
The price level
The inflation rate and the unemployment rate
The growth rate of nominal GDP
The level of nominal GDP
Other, please specify
No opinion
The level of nominal GDP
What specific price index should the central bank use in the conduct of its monetary policy?
Here, ‘central bank’ refers to the central bank responsible for monetary policy in your country of residence.
Headline consumer price index
Core consumer price index (excluding food and energy prices)
Headline personal consumption expenditures index
Core personal consumption expenditures index (excluding food and energy prices)
GDP deflator
Other, please specify
No opinion
A price index with a higher weight on investment goods prices and other durables prices than the GDP deflator
How likely is the central bank to achieve its inflation target over the next three years?
Here, ‘central bank’ refers to the central bank responsible for monetary policy in your country of residence.
Relatively unlikely
The only reason I think average unemployment would decrease is that I believe unemployment is convex in inflationary/disinflationary pressures.
Given the current longer-term, structural economic trends, do you think that the benefits of increasing the inflation target of the central bank would outweigh the costs of doing so?
Costs clearly outweigh benefits
The right solution is readiness to use deep negative rates when called for, not increasing the inflation target. The appropriate paper currency policy for deep negative rates engineers inflation relative to paper currency, but not relative to e-money. Better to have comparatively innocent inflation relative to paper currency only when needed than to ever have the quite damaging inflation relative to the e-money unit of account.