Compensation to Affected Individuals Appears Twice in 587 Files

This file totals what the corpus can say about money reaching affected people.

One organisation reimbursed customers at a scale that reflected their loss: Coinbase, at 25-0530, roughly $2,600 to $5,700 per person. It was possible because the harm was a transfer of a known amount on a known date.

Beyond that, there is litigation — $6.60 per person at 25-0703, $3.81 at 25-0410 — and credit monitoring, offered at notification, offered again in settlement, and addressing almost none of the exposures this database records as serious.

Everything else is nothing. No compensation for the 6.5 million Co-op members at 25-0501, the 5,556,702 at Yale New Haven at 25-0308, the students at 25-0105, the crash-report subjects at 25-0612, the employees at 25-0613.

Compensation happens where loss is quantifiable and does not happen where it is not. That is the whole rule.

It explains why the financial sector compensates and the healthcare sector does not, why a fraudulent transfer is made good and a leaked diagnosis is not, and why the settlements in this section price litigation risk rather than harm. It is the evidentiary vacuum at 25-1219 expressed as an economic outcome.

The obvious response — require compensation for exposure — runs immediately into the measurement problem. Pay everyone a fixed sum and you have a tax bearing no relation to harm, funded ultimately by the same population it pays. Pay the harmed and you must identify them.

This desk does not have the answer and will not pretend to. What it can say is that the current position — where the consequence of losing eight figures of medical records is a monitoring subscription and a few dollars — is not the product of anyone deciding that is the right amount. It is what happens when nobody can measure the thing being compensated.

Every file that ends with credit monitoring should be read as ending with nothing. The apparatus of notification, remediation and settlement produces motion rather than restitution.

Graded medium: this is a synthesis across the database rather than a finding from a source, and the corpus’s own coverage is shaped by what gets disclosed.