We're Still Driving Through the Data Fog

And it's getting extremely old.

The economist Alan Detmeister is widely consider to be a foremost expert on inflation. Here’s what he said about yesterday’s CPI report:

I think you largely just put this one to the side. Maybe this report gives a minor downward sign for overall inflation, but the vast, vast majority of this is just noise and should be ignored.

Meanwhile, over in the job market, we have the Chair of Federal Reserve telling us the monthly number is biased up by 60,000 per month, such that the average monthly gains, taking out the DOGE-infected federal gov’t, of ~40K is really -20K. Also, the labor market experts to whom I listen most closely tell us that we cannot reliably gauge the immigration numbers in the recent monthly employment surveys.

We should have third quarter GDP by now but we don’t. And the thing that’s most messing with our inflation prints—housing costs—isn’t going away. The downward bias will be with us until April, when we could see the shelter price spiking back the other way.¹

All of this is due to the government shutdown and the ensuing inability of the statistical agencies to collect data over that period (at least the current data problems are due to that; there’s a longer-term problems of insufficient funding). I didn’t quite appreciate how much that would lastingly mess up the timing and even accuracy of the data, and I don’t think I’m alone.

What does this mean for regular folks trying to get through their economic lives? Not a ton—just because me and my colleagues obsess about the high-frequency dataflow, doesn’t mean normal people are particularly burdened by the ongoing data fog. People know very well what things cost, what they earn, what’s in their bank account, how much their rent is, etc. But there are still both economic and political implications to these data issues.

First, the Fed is already in a confusing place, trying to sort out which side of the current stagflation to upweight in their interest-rate decisions. They’ll listen to Detmeister et al re yesterday’s CPI. The inflation measure they track more carefully, the PCE, gives less weight to shelter, so that’s good, but most of the PCE comes from the CPI, and shelter wasn’t the only price biased down in the Nov CPI (certain goods and services also fell prey). They’ll also track inflation expectations, but those too are a function of inflation prints, of course, and so this signal is jammed as well.

Given that our economic lives are always impacted by the Fed’s benchmark interest rate—it bleeds into mortgage, auto-loan, credit-card rates—added uncertainty there affects us all.

Politically, we have an administration that already is in full denial about the price impacts of tariffs, now with a biased-down inflation report in hand. They’re also quite anxious to declare mission accomplished when it comes to affordability pressures. Here’s one of their top folks responding to yesterday’s CPI:

I'm not saying ‍we're going to declare victory yet on the ​price problem, but ‌this is just an astonishingly good CPI report.

This is misguided in a small way and a big way. Re small way, it’s claiming “astonishing” goodness to an inflation rate that was biased down. Re big way, it once again confuses lower inflation with lower prices. The “price problem” is a price-level problem, not a price-growth-rate (inflation) problem. Telling people life is more affordable because inflation is slowing (which in itself is hard to know right now) is talking past them.

In sum, it’s very frustrating to be stuck in this data fog for so long, and it has some real life implications. It’s also yet another side effect of dysfunctional governance, something we’ve all had more than enough of, I’m sure.

1. Here’s how GS explained the source of the downward bias in rent and owners’ equivalent rent (OER) parts of the index. “Rent and OER are calculated based on a six-month rotating panel in the CPI, so the month-over-month increase in November is roughly the 6-month average monthly increase since May. This monthly rate is applied to the October index level—which the BLS assumed was flat relative to September—so the shelter components only increased by one month’s worth of rent inflation between September in November. …today’s reading could be partially offset by a rebound in the shelter components in the April CPI, six months after October.” The key points are that a) the BLS isn’t printing the “correct” level of the housing price in Nov. They’re growing the biased-down Oct level (biased down because they assumed no housing inflation in Oct) by 6-month avg noted above, and b) when the negatively biased Oct print comes out of the avg, we should get a jump (statisticians call this a base effect, explained here).