Skip to content
Hugin
A metal spirit level lying along a sloping wooden plank in a dark workshop, its bubble pushed toward one end.

Journal

The Postal Service made its own deadline longer, made its own pass mark lower, and still missed. When an organization grades itself, 'missed the target' is a statement about two numbers, and only one of them is performance.

4 min read

Original editorial artwork generated for Hugin.

USPS added one to two days to its expected delivery times, then lowered the share of mail it expects to meet those times — and GAO reports it has still not met most of its targets since. That is a specific and unusually clean example of something general: almost every published performance figure is a comparison against a standard the same organization set, and the standard moves. Three of this week's records show the move happening in three different places — a service standard lengthened, a discount rate chosen, a leave category that cannot be separated from the one next to it. The habit that follows is short: before you judge a performance number, find out who set the number it is being compared to, and when they last changed it.

methodperformance-measurementuspsgaostatistics

The Postal Service publishes how much of its mail arrives on time. Two things have to be true for that sentence to mean anything: "on time" has to mean something fixed, and "how much" has to be measured against a target somebody set in advance.

Both moved.

The cleanest example you will get

From GAO's report this week, in order:

Effective October 2021, USPS added 1 to 2 days to its expected delivery times for certain products to accommodate its transportation change from air to ground.

In April 2025, USPS further changed expected delivery times by eliminating end-of-day or afternoon collection at the over 24,000 post offices that are more than 50 miles from a Regional Processing and Delivery Center.

Additionally, USPS lowered its service performance targets — the percentage of mail it expects to meet service standards — in fiscal year 2021 and has not met most targets since then.

Deadline extended. Pass mark lowered. Most pass marks still missed. GAO's own compression of it: "USPS has largely failed to meet even these reduced service standards."

Now imagine the counterfactual, which is the useful part. Suppose USPS had hit every target from 2022 on. The press release writes itself — service performance targets met across First-Class Mail — and every word of it would be true, of a target lowered in 2021, against a standard lengthened by up to two days. A reader who knew only the headline would conclude the mail got better. The mail would have got slower.

That is not an accusation about the Postal Service. It is a property of any number of the form percentage meeting the standard, and it is why the standard's history is part of the number.

The same move, in two other places this week

A discount rate. EPA's repeal of the 2024 power plant carbon standards puts the compliance cost savings at "$160 billion using a three percent discount rate and $95 billion using a seven percent discount rate" — same repeal, same years, same dollars. Publishing both is the correct practice. But the headline figure a reader carries away depends on which of two legitimate assumptions the writer picked up, and nothing in "$160 billion" carries that choice with it.

A category. GAO reports that OPM "does not know the actual costs of the paid administrative leave used for workforce reduction efforts" because that leave "is reported with other types of general paid administrative leave." Here the denominator has not moved so much as merged: a cost exists, it is real, and the only bucket it can be counted in also contains something else. GAO's sentence about what that prevents is the one to keep: "To calculate long-term savings, OPM needs to know short-term costs of paid administrative leave used for these efforts."

Three different mechanisms — a redefinition, an assumption, a merged category — and the same consequence. The comparison the number invites is not the comparison it supports.

Why this is not cynicism

None of the three is a lie, and treating them as lies is the wrong reflex — it produces a reader who dismisses every official figure, which is worse than one who over-trusts them, because it is unfalsifiable.

Every one of those moves is documented, by the organization that made it. USPS published the new standards. EPA printed both discount rates in the same table. GAO — auditing OPM — found and published the merged category. The information a reader needs is in the record. It is just never in the headline, because a headline is a single number and this is always two.

The habit, which takes one question

Before you judge a performance figure, ask: who set the thing it is measured against, and when did they last change it?

Three shapes to watch for, and where each hides:

  1. A standard that moved. Look for an effective date on the standard itself, not on the report. "Effective October 2021" is the tell.
  2. An assumption you could have made differently. Discount rates, baselines, inflation adjustments, which year is "year one." If the document gives a range, the range is the honest version and the single number is the press version.
  3. A category that changed shape. Merges, renames, and reclassifications break comparability silently, because the series keeps running under the same name.

None of that requires statistics. It requires reading one paragraph further than the number.

And the corollary for anyone publishing one

If you set the standard you are measured against, say when you last moved it, in the same place you report the result. USPS's own strategic-plan update is due in 2027, and GAO's recommendation is essentially that: communicate "about its ongoing and planned actions to address this issue, even as it seeks to cut costs, as well as about the challenges it faces."

A performance number published without its standard's history is not wrong. It is just not yet a fact about performance.

Source links