In a recent post, Tyler Cowen offered the following criticism of my views on inflation:
More seriously, Scott seems to dismiss the price level concept altogether. For instance he once wrote: “In the past, I’ve frequently argued that inflation is an almost meaningless and useless concept. I’m not even aware of any coherent definitions of the concept.” I don’t think this is a defensible point of view, and you have to compare Scott’s criticisms of the o1 model to his own approach, which is fairly nihilistic. And I think wrong. If inflation were higher and someone offered Scott an inflation-indexed contract to sign, would he be unable to evaluate such a transaction? Obviously not.
In a previous post, I provided a fairly detailed response to this criticism. More recently, Bryan Caplan directed me to a 2004 post by Tyler that seemed almost as nihilistic:
Peter Gordon looks at a 1902 Sears Roebuck catalog and asks whether money was worth more back then.
Of course it depends how much you are given. $5.00 back then goes a longer way, but I would rather earn $100,000 a year today, and yes that is not adjusting for inflation. For Peter modern pharmaceuticals are the clincher:Would you want their best 1902 camera for $7.90? Probably not. High-end cutlery for 6 for $1.79? Why not? A great western saddle for $8.95? Sure.
It’s the Sears “Drug Department” that is the real eye opener. “Fat Folks, Take Rose’s Obesity Powders and Watch the Result … $4.20 per dozen boxes.” Herb laxative teas for 16 cents a box may be OK. Dr. Rose’s Arsenic Complexion Wafers 35 cents a box may have few takers today. Vin Vitae for 69 cents (“Not a Medicine … Not Merely a Tonic”). The “White Ribbon Secret Liquor Cure” went for $2.50 a box. The list goes on and does focus the mind.My question for today: Does this mean that we should adjust the gdp deflator series to show ongoing deflation for the 20th century?
At the time the post was written, official price level data showed the cost of living to be nearly 20-fold higher than in 1900, and yet Tyler expresses doubt as to whether there had actually been any inflation at all!
Perhaps I expressed my views with more over-the-top language than Tyler, but I believe he shares my skepticism as to whether adjusting wage income by government price indices allows us to ascertain how living standards have evolved over time.
READER COMMENTS
Matthias
Jan 6 2025 at 12:07am
Just use the one and only Big Mac index go measure the price level.
Arqiduka
Jan 6 2025 at 4:40am
I think we may be confounding two related but still different concepts: cost of living vs technological improvement.
That is rather live on 100k today than 100k in 1950 (and forego innumerable modern gadgets I’m used to) reflects technological improvement.
That, of those things available and comparable across both periods, all are far more expensive today, reflects cost of living.
The two concepts are to be separated. The former is interesting but we have no way of measuring it. The later is also useful and allows us to say that inflation has eaten up what giant gain’s technology has awarded us.
Even if we could measure the true standard of living by allowing people to time travel, I’d be on guard against what the government would infer from the data: “let’s inflate away until people are indifferent between 100k now and 100k 50 years ago” they would say. No mate, that’s not your job.
Todd Ramsey
Jan 6 2025 at 10:33am
Scott, Mercatus could be the means through which we lobby the government to implement NGDPLT and a NGDP security.
I hope we can stop nitpicking with Tyler over definitions of inflation and help him remember that he fundamentally agrees with your models. He was among the first to promote your insights during the 2008-09 financial crisis.
I hope we can keep our eyes on the prize: an NGDP security market.
Scott Sumner
Jan 6 2025 at 10:42pm
I view the main prize as NGDP level targeting, not an NGDP linked asset.
Todd Ramsey
Jan 7 2025 at 10:00am
A NGDP asset, with a large thick market, would provide a timely, well-informed, low-cost estimate of NGDP expectations that would be difficult to game.
Rajat
Jan 7 2025 at 2:24am
You two have had a lengthy dialogue so I may have lost sight of its genesis, but isn’t there an important difference between the measurability and usefulness of inflation for the purposes of macroeconomic stability & policy (which focusses on the short-medium term, say 3-5 years), and for the purposes of comparing living standards over long periods of time and/or across countries? Tyler seems to focus on the first issue whereas you seem to focus on the latter?
Scott Sumner
Jan 8 2025 at 2:12am
Yes, I’m claiming that it’s not useful for short term macro analysis–we have much better nominal indicators. That’s where I most strongly disagree with Tyler.