Weird economics

I wasn't able to blog yesterday about the US July CPI data, but it turns out that may be an advantage, because I now have the opportunity to focus on some of the downright weird ways that the numbers have been portrayed on social media.

First, though, a quick look at the numbers themselves. Headline CPI was unchanged month-on-month in July after rising 1.3 percent in June. This dropped the year-on-year rise to 8.5 percent from the 9.1 percent reported in June.  This result was slightly better than expected, and mainly reflected a sharp fall in energy prices.  Core CPI (ex food and energy) rose 0.3 percent month-on-month, the lowest increase for four months, resulting in a year-on-year increase of 5.9 percent, the same as in June.

Now, about the way some people have been interpreting those numbers. There seems to be an almost coordinated effort out there on social media to pour scorn on official data that don't correspond to the "we're doomed to stagflation and it's all Biden and Powell''s fault" narrative.  One genius on Twitter, whose mini-bio says he is an investment advisor, pontificated that "CPI isn't inflation", prompting one well-known economics blogger to respond "Jesus Christ! That's exactly what it is"! If your investment advisor thinks like this, it may be time to take your business elsewhere.

Then there's the crypto community, which seems to be ever more anxious to see societal collapse,  if it's a way of making money. One Bitcoin booster dismissed the July data with the argument that "inflation is cumulative".  This is of course true, but it might be pointed out that 0 percent per month cumulates in a rather different way from 1.3 percent per month. 

The same person also chose to comment that the 528,000 increase in non-farm payrolls reported last week was not a good number because more than 300,000 of the jobs reflected "seasonal adjustment", an opinion that suggests minimal or no understanding of how statistical measures are calculated. Misrepresenting both the inflation and the employment data in this way allowed them to declare that the US economy was indeed enduring "stagflation", an opinion that is frankly little short of gaslighting. 

Enough of the rant!  Does the July data indicate that the worst of the inflation spike has passed? One data point is never enough to draw a conclusion, but there are signs in energy markets, supply chains and shipping rates that allow at least cautious optimism.  One key example: the nationwide average cost of a gallon of gasoline, which topped $5 earlier in the summer, slipped back below $4 (well, to $3.99 to be exact) this week, which should help to keep the August headline CPI in check.

The bad news for Messrs Biden and Powell is that the decline in year-on-year CPI is likely to be painfully slow, thanks to the so-called base effect. Here's where the "cumulative" nature of inflation actually becomes important. The monthly numbers that will fall out of the index for the next few months (i.e. the numbers for August 2021, then September 2021 and so on) were relatively small; the big monthly increases really only came along after the turn of the year. This means that even if August and September see monthly CPI in line with July's -- and it's unlikely the actual numbers will be that good -- the year-on-year figure will very likely stay close to 8 percent.

That's not good news for the Biden administration as the mid-term elections loom. It also makes things tough for the Fed.  Even if low monthly prints for August and September indicate that what we might call the running rate of inflation, calculated by annualizing a few months of data, is falling back towards target, the year-on-year number that the media (and social media) tend to focus on will still be historically high. It's tough to figure out how to frame policy in those circumstances.