Permitted wetland losses in this state are recorded in one system, by permit number, with an acreage. Compensatory wetlands — the ones created or restored to offset those losses — are recorded in another, maintained by a different program, with their own identifiers. The policy that connects them is called no net loss and has been in force for decades.
The two systems have never been joined. I have written about missing joins three times now and I decided that complaining a fourth time would be worse than useless, so this spring I tried to do it.
I got about two thirds of the way and I want to report both parts of that.
What worked: for permits issued after 2009, the permit record generally names the mitigation instrument — a bank, an in-lieu fee program, or a permittee-responsible site — and where it names a bank, the bank has its own ledger and the ledger is public. Matching those was tedious and mechanical and it worked. For that subset I could put a debit and a credit next to each other, and for the great majority of them the credit existed and the acreage was in the right neighbourhood.
That is a real finding and it is a reassuring one, and I want it said plainly because the version of this essay where everything is broken is more shareable and is not what I found.
What did not work is the rest, and the rest is not a remainder, it is a category.
Permittee-responsible mitigation — where the permit holder creates the wetland themselves, usually on or near their own site — has no central ledger. The obligation lives in the permit conditions as text. Whether it was performed lives in monitoring reports, which are submitted as documents, filed by permit, and not extracted into any field anybody can query. To check one, you read it. I read forty-one, chosen by a rule I wrote down in advance so I could not drift toward the interesting ones.
Of the forty-one: twenty-eight had monitoring reports on file covering the required period, and of those, twenty-two reported meeting their performance standards, which I have no basis to doubt and did not attempt to verify in the field. Six reported not meeting them, with a corrective action described. Thirteen had no monitoring reports on file that I could locate, which means either that they were not submitted, or that they were submitted and not filed where I looked, and I could not distinguish those two and I want to be clear that the distinction matters enormously and that I cannot make it.
So the honest summary of my spring is: the mitigation-bank portion of no net loss is auditable and broadly checks out; the permittee-responsible portion is not auditable by anybody outside the agency, by me, this year, with the tools available; and I do not know what is in it.
What I did learn is why nobody has done this, and it is not indifference. I spoke to four agency staff and every one of them was helpful, two of them sent me things I had not asked for, and one had clearly tried some version of this himself. The obstacle is that the monitoring reports are documents. Making them queryable is a data-entry project across decades of PDFs, and a data-entry project has no champion because its output is not a finding, it is the possibility of a finding.
That is the part I would fund. Not a study. Not an audit. A schema and a person, for two years, turning the required fields of the monitoring reports into rows — acreage, standard, met or not met, date — for new submissions going forward and backwards as far as the money lasts.
The state would then be able to answer, for the first time, the question its own policy is named after.
I have the forty-one in a spreadsheet with my rule written at the top of it. It is not a sample of anything and I have not published a number from it and I am not going to. What I will say is that I began this expecting to find a scandal, and what I found was a filing cabinet, and the filing cabinet is worse, because a scandal has somebody in it.
