This story is generating lots of reaction. Which is good! I think it's an important story. But also a nuanced one. And so a few points I think it's worth making, in a Saturday evening 🧵:

1. I'm getting lots of comments about how this is obviously political manipulation of the economic numbers. I don't think this is as cut-and-dry a case as it might look to some people.

For one thing, there are very real issues with the CPI's legal services index. See this screenshot -- the index hasn't been published regularly since early 2023. So looking for an alternative data source is hardly crazy on its face.

It's also important to understand that BEA has always applied more judgement to its estimates than BLS or Census. It is an aggregator and synthesizer of data, not a collector of data like those other agencies.

Given that context, I don't think it's implausible that BEA's economists have been skeptical of the CPI legal services estimates for a while, and so when they saw a big, unexplained jump in January, decided to use the PPI data instead. This is more or less what BEA's economist told me yesterday.

Whether that was the *right* decision is a different question. You can certainly argue that PPI is an inferior measure of the legal prices faced by consumers; that if they make a switch, they should adjust the historical data for consistency; that making ad hoc adjustments is bad practice; etc.

But I suspect that if this decision had been made on the normal calendar, disclosed ahead of time, etc., no one would be making a big deal of it. It's the kind of methodological decision stats agencies make all the time.

2. On the other hand, when I've talked to experts on the statistical system over the past year, I've often asked them, "If someone were going to try to interfere in the data, how could they do it?" And the answer is basically: stuff like this.

I think outsiders often imagine that interference would look like a president or one of his appointees dictating the unemployment rate or CPI. For a whole host of reasons, that would be basically impossible (or certainly impossible without it being discovered).

But could political appointees apply pressure (subtle or overt) on agency staffers to favor methodological choices that would lead to lower inflation? That's more plausible. Especially if, as in cases, like this, the methodological choices aren't on their face illegitimate.

To be clear: I do not have any evidence to suggest that is what happened here, or that it has happened at any point in this or past administrations. Indeed, I've spoken to lots of people inside the system, or who have recently left, who say explicitly that they are NOT feeling this kind of pressure.

But this is the *kind* of decision that you could imagine being influenced by political considerations. The fact that is was an ad hoc decision, not a permanent change in methodology, adds to the concern: Once you start making one-off judgment calls it's easy for bias to seep in, even unconsciously.

3. It's not a coincidence that this change got noticed literally within minutes of the report coming out. Eagle-eyed forecasters like @fcastofthemonth pore over every detail of these reports and go deep into the methodological weeds.

That isn't a reason to ignore concerns about the data. But it should increase our confidence that manipulation would be detected quickly, and also that the numbers to date have been reliable (or at least free of bias).

4. The statistical agencies need to understand that they are operating in an environment of heightened scrutiny and behave accordingly. Maybe a few years ago, no one would have noticed or cared about a change like this, but that simply isn't true today.

Failing to disclose decisions like this (or better yet, announce them ahead of time) fuels suspicion. Credit to the BEA economist who responded quickly to economists' emails asking about this and who took my phone call. But the official press office was much less forthcoming.

Ever since Trump returned to office, and especially since Erika McEntarfer was fired, I've been deluged with people asking how I can trust the numbers coming out of this administration.

My answer has always been: 1. The agencies are using the same procedures as always, and being transparent about their operations; 2. The people who know these numbers best still have faith in them; 3. I am confident that we will hear from insiders if they are feeling pressure.

Decisions like this one are a hit to #1 (change in procedure, not transparently disclosed), and to #2 (people who have been staunch defenders of the agencies have expressed at least some concern about this move). #3 still holds for now.

Which brings me, as always, to: If you work inside any of these agencies, I'd love to talk to you. That's true whether or not you've seen anything questionable. I'm on Signal at bencasselman.96 -- anonymity guaranteed. And with that, back to March basketball.