Prices fell twice in June. That has happened 16 times before, and never like this.

In June the Consumer Price Index fell 0.4 percent and the Producer Price Index for final demand fell 0.3 percent. Both indexes, down, in the same month.

That sounds rare. It is not. Since December 2009, when the current producer price series begins, it has happened 17 times in 197 months. Roughly once every 11 or 12 months. In the years before the pandemic it ran closer to once every eight.

We are giving that away first because it is the fact most likely to be omitted by anyone selling you a story about June, including us. The event is ordinary.

What is not ordinary is the room it happened in.

The number nobody has put next to it

Here is every joint decline since the series begins, with the headline inflation rate at the time.

Read the fourth column down. It never goes above 2.2 percent. Not once in sixteen tries across seventeen years.

Then it goes to 3.5.

June 2026 is the only joint decline on record that happened in an economy with headline inflation above 2.2 percent. It is not close. The previous high-water mark, November 2018, sits more than a full point below it.

Every one of them is the same story

There is a second pattern in that table and it has no exceptions.

Every joint decline since 2009 is an energy event. Ten of the sixteen sit inside the 2014 to 2016 oil collapse and its aftershocks. Two are the Covid demand crater. The rest are smaller energy wobbles: the euro crisis slide, the fourth-quarter 2018 selloff. There is not one instance in this series of consumer and producer prices falling together for a reason that was not oil or a demand collapse.

June 2026 is the same mechanism. Final demand goods fell 1.4 percent. Gasoline fell 12.0 percent on the month, which is most of the move by itself.

So the mechanism is familiar. The setting is not. What is new is an energy unwind arriving on top of an inflation rate that no previous unwind has had to work against.

What is underneath is still hot

This is where the "inflation is beaten" reading breaks.

Strip out food and energy and the picture inverts. Core producer prices rose 0.1 percent on the month and are running 5.1 percent above a year ago. Core CPI was flat on the month and sits at 2.6 percent. Gasoline, for all that it fell 12 percent in June, remains 26.7 percent more expensive than it was in June 2025.

Compare that with 2020, the last time both indexes fell hard together. In April 2020 core CPI fell 0.5 percent and core PPI fell 0.8 percent. The whole structure went down at once, because demand had stopped.

Nothing like that is happening now. One input got cheaper. Everything else did not.

What this does not prove

The honest limits, because an analysis that only lists its findings is advertising.

The historical figures are current vintage, not what printed at the time. Both agencies revise seasonal adjustment factors annually. The table above says what the data looks like now, not what markets saw then. May 2026 PPI has already been revised down once.

Nine of the seventeen rest on a single -0.1 print, which is inside the noise and could round to zero. Only eight survive a stricter filter of -0.2 or worse on both legs. June 2026 is one of the eight, but do not read all seventeen as equivalent events.

Two months in the recent gap are unobservable. October and November 2025 have no computable CPI change, because the lapse in appropriations stopped collection. BLS footnotes this directly. Any claim about the length of the drought since the last joint decline contains two months nobody measured.

The series has a hard floor at November 2009. PPI final demand does not exist before that. We cannot tell you this is a first in decades, because we cannot see decades.

The correlation is mechanical, not remarkable. CPI fell in 21 of 197 months and PPI in 43. If the two were independent you would expect about five joint months. There are 17. That is not a coincidence to marvel at. It is one oil price hitting two indexes that both contain energy.

One correction we would like to make on someone else's behalf

While assembling this, we found a published claim that June was "the first negative month-over-month CPI reading since 2000."

That is false, and not marginally. Seasonally adjusted CPI has fallen in 21 separate months since 2009 alone, including a 0.8 percent drop in April 2020.

We mention it because it is the same failure this piece is trying to avoid. The interesting thing about June is genuinely interesting, and you do not need to invent a record to get to it.

Method, and the data

Every figure comes from the BLS public API, which needs no key: CPI-U all items seasonally adjusted (CUSR0000SA0), PPI final demand (WPSFD4), and the corresponding core and unadjusted series. June's figures are in the BLS releases for CPI and PPI.

Take it and check us:

• The table, as CSV: all 17 joint-decline months with both legs, core, and year-over-year
• The full working: every computation, every caveat, longer than the article
• The raw API responses: as pulled, unedited
• The script: run it yourself

One note for anyone reproducing it: the documented GET form of the BLS API silently ignores the year parameters and returns only three years, without an error. You have to POST. Our first pull hit that and the numbers looked wrong for a reason that had nothing to do with the economy.

We found no prior published comparison of joint declines as a series. If one exists, or if we have made an arithmetic error, write to corrections@moneyandworld.com and we will correct this and say that we did.

Correction, July 15, 2026: when first published, this section said the working and the raw data were published with the piece. They were not. They are now, and are linked above. See Corrections.