The problem with this topic
Most published shrink figures are older than they look
Retail shrink is one of the most confidently misquoted statistics in commercial security. The reason is straightforward: the National Retail Federation's National Retail Security Survey used to publish an annual shrink rate and a headline dollar figure, and it stopped. The last one it published covers fiscal year 2022.
NRF still publishes research on retail theft: its current series is called The Impact of Theft & Violence, but that report measures incidents, tactics and reporting behaviour, not a shrink percentage. So there is no NRF shrink rate for 2023, 2024 or 2025, and any dollar figure attributed to NRF for those years did not come from NRF.
This page gives you the FY2022 figures with their year attached, the current FY2025 findings as NRF actually reported them, and the San Diego recorded-shoplifting series from SDPD's own data. Where a number does not exist, the page says so instead of filling the gap.
What NRF published, and when it stopped
Two real years, and three years where NRF published no figure at all.
| Fiscal year | Shrink rate, % of sales | Total shrink |
|---|---|---|
| FY2021 | 1.4% | $93.9 billion |
| FY2022 | 1.6% | $112.1 billion |
| FY2023 | Not published by NRF | Not published by NRF |
| FY2024 | Not published by NRF | Not published by NRF |
| FY2025 | Not published by NRF | Not published by NRF |
Source: National Retail Federation, 2023 National Retail Security Survey: nrf.com. Period: fiscal years 2021 and 2022. Type: ESTIMATE from a voluntary survey of 177 retail brands, representing approximately $1.6 trillion in annual sales and over 97,000 United States retail locations. Retrieved: 2 September 2026.
Also from the same survey: "Internal and external theft accounted for nearly two-thirds (65%) of retailers' shrink." NRF did not publish a separate split between employee theft and external theft in that release, so this page does not split it either. The remaining share is process failure and administrative error.
Methodology
How to read survey-based retail figures
Shrink numbers are not crime statistics. They come from retailers answering a questionnaire, and that shapes what they can and cannot tell you.
Participation is voluntary and self-selected
NRF's figures come from retailers who chose to respond. This is not a census and it is not government data. Retailers with an active loss prevention function are more likely to have the figures to hand and more likely to take part, which is a selection effect nobody can quantify from the published summary.
The sample changed size dramatically
177 retail brands responded to the FY2022 shrink survey; 66 companies representing 143 brands responded to the FY2025 theft report. A smaller sample is more sensitive to which particular retailers answered, and it is one reason the two reports cannot be treated as a continuous series.
Shrink is self-reported and self-defined
Shrink is the gap between recorded inventory and counted inventory. How a retailer classifies that gap between theft, damage, supplier error and administrative mistakes is an internal judgement, and it varies between businesses. That is part of why roughly a third of shrink is not attributed to theft at all.
NRF's series is not continuous
The National Retail Security Survey and The Impact of Theft & Violence are different publications asking different questions. There is no NRF shrink rate after FY2022, which means there is no basis for a shrink trend line through 2023, 2024 or 2025. Any chart showing one has invented the missing points.
Recorded offences are not incidents
The SDPD figures count offences reported to police and recorded under NIBRS rules. With 63% of retailers telling NRF they report fewer than half their incidents, recorded shoplifting counts are a fraction of what happens, and changes in them can reflect reporting behaviour as much as anything else.
Reporting rates come from a household survey
The Bureau of Justice Statistics estimates 30.5% of property crime was reported to police in 2024, but the National Crime Victimization Survey covers households and persons, not businesses. It establishes that recorded crime understates the total; it cannot tell you the commercial reporting rate. NRF's 63% is the better figure for retail.
Sources for this section: National Retail Federation, 2023 National Retail Security Survey: nrf.com; National Retail Federation, The Impact of Theft & Violence 2026: nrf.com; Bureau of Justice Statistics, Criminal Victimization, 2024, NCJ 310547, published 29 September 2025: bjs.ojp.gov. Retrieved: 2 September 2026.
Current
What NRF reports now, for fiscal 2025
NRF's current report covers fiscal year 2025 and was released on 30 July 2026. Its headline finding is one a security company is not supposed to lead with, so we will: retail theft went down.
| Measure | Reported figure |
|---|---|
| Shoplifting incidents, change on 2024 | Down 12.4% |
| Retail merchandise theft, change on 2024 | Down 8.1% |
| Retailers reporting fewer than half of incidents to law enforcement | 63% |
| Retailers citing repeat offenders | 50% |
| Retailers citing organised retail crime incidents | 40% |
| Retailers citing walkout or pushout theft | 37% |
Source: National Retail Federation, The Impact of Theft & Violence 2026, released 30 July 2026: nrf.com. Period: fiscal year 2025, compared with 2024. Type: ESTIMATE from a voluntary survey of 66 retail companies representing 143 brands and $1.7 trillion in fiscal 2025 sales, or 31.6% of total retail sales. Retrieved: 2 September 2026.
Do not compare these two surveys directly. The FY2022 shrink survey had 177 brands responding; this one had 66 companies. The questions differ, the sample differs, and the reports are not the same publication. A percentage from one and a percentage from the other do not form a trend, and we are not going to present them as though they do.
The number worth stopping on is the 63%. Nearly two-thirds of retailers told NRF they report fewer than half their incidents to law enforcement. That single figure explains why recorded crime statistics are a weak proxy for what a shop actually experiences, and it is the honest version of the argument this industry usually makes badly. If most incidents never become a police record, the record that exists is the one the business kept itself.
San Diego
Recorded shoplifting in the City of San Diego
Offences recorded by San Diego Police Department. Read the caveat below the table before you use any of these numbers.
| Year | Shoplifting | Theft from building | All larceny and theft |
|---|---|---|---|
| 2020 | 1,272 | 616 | 16,124 |
| 2021 | 1,179 | 518 | 18,113 |
| 2022 | 1,181 | 427 | 15,746 |
| 2023 | 1,423 | 440 | 16,144 |
| 2024 | 3,097 | 387 | 17,133 |
| 2025 | 4,221 | 370 | 16,399 |
Source: San Diego Police Department, Police NIBRS Crime Offenses, City of San Diego Open Data Portal: data.sandiego.gov/datasets/police-nibrs. Period: calendar years 2020–2025. Type: COUNT of reported offences. Retrieved: 2 September 2026. SDPD states its data is dynamic and records may change as investigations proceed.
Read this carefully. Recorded shoplifting rose from 1,181 in 2022 to 4,221 in 2025, while all larceny and theft stayed broadly flat and theft from buildings fell. A subcategory rising sharply while its parent category does not move is the pattern you would expect from a change in how offences are classified or recorded, not a change in behaviour. The dataset offers no explanation, and we could not verify one from a primary source, so this page reports the counts and attributes no cause. We are not going to tell you shoplifting tripled in San Diego, and nobody else should either without evidence of what changed.
Where we fit
What a camera system can and cannot do about shrink
Roughly a third of shrink, on NRF's own FY2022 figures, is not theft at all. It is process failure and administrative error: receiving mistakes, mispicks, markdown errors, damage. No camera improves any of that, and a security company that lets you believe otherwise is selling you the wrong thing.
What recorded video is genuinely good for is the part of the problem that is an event: knowing what happened at the till, at the back door, at the delivery bay and in the stockroom, and being able to produce that record afterwards. Given that 63% of retailers report fewer than half of their incidents to police, the practical position is that most of what happens in your shop will never be in a police file. It will only be in yours, if you kept one.
How that applies to a specific store (where the cameras go, which views actually have to prove something, what a stockroom or a back door needs) is set out on our retail security page, and the survey process behind it on commercial camera installation.
What we will not do is attach a shrink reduction percentage to any of it. There is no primary source establishing one, and the figures circulated in this industry do not trace to a study you can read. We would rather say that plainly than quote a number we cannot stand behind.
Two boundaries, as on every page here. We work on commercial premises only and not on private residences. And we do not hold a C-10 or a C-16, so we do not work on fire alarm systems; that is coordinated with your own licensed fire alarm contractor.