Methodology
Where these numbers come from, and what's wrong with them
628 rates sit in this dataset. 77 rows came from public threads and carry a link to the thread. 551 rows arrived through the form on this site, anonymously. Nothing here has been independently verified against a remittance or a contract, and the site has no way to do that.
Nothing is verified, including the rows that look official
Every page carries the word "reported" because every figure is someone's account of their own rate. The dataset has a verified field and it is false on all 628 rows. If a figure is ever confirmed against a primary document, that field changes and the page says so. None has been yet.
What the linked rows have that the submitted rows don't is a public thread you can read yourself. That's provenance, not verification.
The high end, row by row
A comment on the thread that brought most of this data in said the top of the range "must be bad data or miscategorized", singling out a $284 report from Nebraska. That's a fair challenge and it deserves the actual rows rather than a reassurance. 8 rates in the set sit at $200 or above:
| Rate | Payer | St | License | % of their Medicare | What the submitter said |
|---|---|---|---|---|---|
| $284 | BCBS | NE | PhD/PsyD | 170% | Increased from 277 last July |
| $227 | BCBS | IN | PhD/PsyD | 136% | — |
| $223 | Quartz | IL | LCSW | 178% | — |
| $220 | Medicaid | NM | — | 176% | LMSW, group practice |
| $210 | BCBS | WY | LCSW | 168% | — |
| $202 | Dean | IL | LCSW | 161% | — |
| $201 | Medicaid | OR | LPC/LMHC | 161% | Used to be 184 until billed rate raised to 220 |
| $200 | Medicaid | OR | LCSW | 160% | — |
The last column measures each rate against the Medicare schedule that applies to that clinician: $167 for a psychologist or physician, $125 for everyone else. Read that way these run from about 120% to 170% of Medicare, which is high, and which commercial contracts do reach. The Nebraska row is a doctoral-level clinician on the full schedule, and the submitter volunteered a prior-year figure for the same contract. Two are Oregon Medicaid, which other reports here independently put among the highest-paying Medicaid programs in the country.
None of that makes them right. It makes them consistent with the rest of the data rather than obviously broken, and there's no honest basis for deleting a figure because it's surprising. If one of these is a typo, the person who filed it is the one who can say so.
Why the outliers stay in
The range is the finding. A clinician looking at an offer needs to know both what the middle looks like and how far the ends stretch, because the ends are what a contract negotiation is arguing about. Trimming them would produce a tidier median and a less useful one. Every page shows the count alongside the figure so a small sample is visible as a small sample.
What's missing
License is blank on 53 of 628 rows and state on 24. Blank fields are left blank rather than guessed. 12 rows carry a CPT of "Other" — one of them is a $22 rate that the submitter noted was a 90853, a group session, which is not comparable to an individual hour and is kept out of every per-code median because of it.
Correcting a row
If a figure here is wrong, the form takes a correction the same way it takes a rate. Rows are changed or removed when there's a reason to, and the CSV carries every field this page describes so the data can be checked without taking any of it on trust.
One more caveat for anyone modelling this: reports arrive in bursts after a discussion thread, so states and payers where those discussions travelled furthest carry more weight than they would in a random sample.