Since 2019, California's Highway Patrol organized retail crime task force has run 4,577 investigations, made 5,134 arrests, and recovered 1,620,803 stolen items worth $74.6 million.
That is the flagship enforcement program in the largest retail market in the United States, over roughly seven years. Through April of this year it had produced 119 arrests. In November 2025 the governor's office was announcing 800 in a year. The run rate is falling.
Meanwhile the Council on Criminal Justice reported in July that shoplifting in the first half of 2026 was up 4% against 2025 and 5% against 2019 — the only one of thirteen offense categories still above pre-pandemic levels.
And meanwhile again: shrink is improving on the P&L, materially.
This edition is about why those three facts cannot all mean what they appear to mean, and what that costs you at budget time.
Three facts that do not fit the story
Take the earnings calls at face value.
Dollar General, June 2026: a 28 basis point reduction in shrink year over year, while lapping a 61 basis point improvement from the prior year — on top of more than 80 basis points delivered in 2025.
TJX, February 2026: "shrink is essentially back to our pre-COVID level," with 20 basis point improvements in each of the last two years.
Ulta, June 2026: 100 basis points of gross margin improvement attributed to lower shrink, with reductions "across every category and every region."
Home Depot, Q1 2026: one of the loudest voices in the 2023 ORC debate did not mention shrink or theft once on the call.
So enforcement is producing fewer arrests, reported shoplifting is up, and shrink is down across multiple large retailers. That combination does not fit the story this industry has told for three years.
The reconciliation nobody wants to run
There are two explanations, and neither requires anyone to be lying.
The first is reporting behavior. The Council on Criminal Justice states it directly: "if retailers increased reporting, then that would increase the count of shoplifting offenses even if there was no actual increase." CCJ documented one major retailer whose reporting rate went from 8% of incidents in 2022 to 20% in 2023 — a 2.5x change in reporting propensity, with no necessary change in theft at all.
Spend three years lobbying for laws, funding task forces and telling stores that police will finally act, and stores report more. The line goes up. The behavior underneath it may not have moved.
The second is value per incident. CCJ's Marc Levin, in July: "The number of shoplifting incidents has been declining in the last several years, but the amount stolen per incident has been increasing." Higher-value incidents cross felony thresholds more often, which mechanically produces more chargeable cases and more legislative urgency — from a smaller number of events.
Both point the same way. The metrics used to justify the campaign are metrics of reporting and charging, not metrics of loss.
A perspective from the field
I supported most of these laws and still think several were correct. What I object to is that we built a public case on numbers we did not control, and now that shrink is finally improving, we cannot demonstrate that any of it was caused by what we asked for.
That is not a policy failure. It is a measurement failure, and it is ours. The CFO who reads two more quarters of improving shrink will conclude the problem solved itself — and then ask why the budget grew. I have watched that conversation happen. Nobody wins it with a legislative scorecard.
Where the machine actually stands
The state count is real and it is large: nine states and Washington DC passed 23 ORC bills in 2024, fifteen states passed 19 in 2025, and more than a dozen have acted in 2026.
This year's crop is not symbolic. Oklahoma HB 3244, effective November 1, carries 10 to 20 years and fines of $20,000 to $100,000, and extends ORC to operations spanning two or more municipalities or combined with online and phone fraud. Wisconsin's Act 106 allows aggregation of separate thefts. Wyoming lowered the felony threshold for repeat offenders and created an "absconding for criminal purposes" offense.
California's Proposition 36 deserves an honest look. Roughly 40,000 charges were filed under it in 2025, about 70% theft. Around 900 people went to state prison in its first year, and jail populations rose 2,600 between November 2024 and November 2025. The legislature requested $250 to $400 million a year to implement it, against roughly $127 million available.
Federally, the picture is thinner than the press releases suggest:
The Combating Organized Retail Crime Act has been introduced in three consecutive Congresses. It passed the House on May 12 with 348 votes and has 47 bipartisan Senate cosponsors. It is now being attached to the National Defense Authorization Act because it has not moved standalone.
The one federal law that did pass drew this from its own authors. Senators Durbin and Cassidy on the INFORM Consumers Act, in force since June 2023: "there has been no indication that the FTC has commenced any investigations or enforcement actions under the Act."
Three practical moves for the next 90 days
Stop reporting incident counts as performance. They measure your reporting rate. Edition #09 made this case about apprehensions; this is the same error at policy level.
Instrument your own reporting propensity. Track incidents reported to police as a share of incidents recorded internally. If that ratio is moving, every external number you cite is contaminated by it — including the ones that flatter you.
Attribute your own shrink improvement, in writing, before someone else does. Process, markdown discipline, self-checkout changes, staffing, enforcement. If you cannot decompose your own result, you cannot defend your budget when another function claims credit for it.
Closing note
The industry got what it lobbied for. The next three years are about whether anyone can prove it mattered.
That is a measurement problem, which means it belongs to us — and it is a better problem than the one we had, when we were arguing about whether the loss was real at all.
If your organization has decomposed its shrink improvement into causes, I would like to see the method. Reply with anything you can share, anonymized always.
Forward this to one LP or AP leader who should be reading it.
— Gabriel
The LP Brief is a weekly intelligence read for senior loss prevention and asset protection leaders. Free. No vendor noise.
Not yet subscribed? thelpbrief.com