Essay

The List That Does Not Say

A Wi-Fi failure destroyed Washington's bar exam in July. The state's answer was to admit some of the stranded without it — and then to order that the public record of admissions not show who was which. The second decision is the expensive one.

Pedro D. Miranda · 27 August 2026 · 8 min

On the morning of 28 July 2026, several hundred people sat in the Yakima Convention Center in central Washington, waiting to become lawyers.

They had come for the first live administration of the NextGen Uniform Bar Examination — a new two-day, online-only test built by the National Conference of Bar Examiners to replace the exam American lawyers have sat for decades. Washington volunteered for the first wave. Ten jurisdictions ran it that week across sixteen sites, for about 2,643 applicants in all.

In Yakima the network failed. Applicants waited most of the day; only a fraction ever connected. The first day was abandoned, then the second. The Washington State Bar Association cancelled the administration outright. The NCBE, in its statement of 28 July, called it a "venue-specific infrastructure issue" and said the platform itself "performed as designed". Roughly 700 applicants had been scheduled to test in Washington; press accounts put the number who could not start at about 645 to 650.

On 6 August the Washington Supreme Court entered Order No. 25700-B-753. It is three pages long and it is public. It is worth reading slowly, because the coverage did not.

What the order actually sorts

The headline version is that Washington rescued the people its own vendor had stranded. That is true, and it is not what the document does.

What the document does is sort them into three groups.

The first group is applicants who could not sit, "and who have not previously sat for and failed to receive a minimum passing score on a bar exam in any United States jurisdiction". They were offered a choice: sit a make-up exam, or "accept a waiver of the requirement to pass the bar exam in order to be licensed in Washington and be admitted by court order".

The second group is applicants who could not sit and who had failed a bar exam somewhere before. They "are not eligible for waiver". They may re-sit in September or February, with the fees covered.

The third group is the minority who did get through the exam. They are ineligible for the waiver too. They will receive a score.

Elections were due by noon on 14 August and became binding on receipt. So the line that decided who became a lawyer in Washington without an examination was not competence, or readiness, or which school anyone attended. It was whether they had ever failed before.

Then there is a second sentence, in the Court's public statement of the same day, which none of the coverage I can find quotes.

The WSBA will publish an admission list on or after October 12, 2026, that does not reflect the method of admission.

The exam was waived for one group. The record of who was waived was withheld from everybody.

The three marks

Barrier. Something has to be able to say no. For the waived group, after noon on 14 August, nothing could. The order leaves the MPRE at a score of 85, the Washington Law Component, the oath and character-and-fitness review in place — real gates, and worth saying plainly, but not the gate anyone means by "the bar". Washington's last completed cycle refused 222 of 780 candidates: a failure rate of 28.46 per cent, comfortably inside the fifteen-to-forty band that marks a live standard. For one group that number is now zero. For the repeat takers it stands at full height, and they were sent back through it a second time.

Crossing. Everyone paid. The year of study, the fees, the drive over the pass, the hotel, the hours in a hall watching a router. But a crossing is not what you spend; it is what you are measured against. The waived paid in full and were never measured. The repeat takers paid in full and were told to pay again. Only the third group — the ones who got through on a broken day — completed the crossing, and they are the group nobody is writing about.

Witness. This is the mark the Court could have preserved at no cost to anyone's livelihood, and it is the one it removed by rule. The list published on or after 12 October will not say how anybody got in. Not for the waived. Not for those who will sit at UW Tacoma on 1 and 2 September and pass a real exam under a contingency plan. The register has been instructed to forget, for the entire cohort.

The Court knew it was close

This was not a unanimous court doing the obvious thing. The order records the dissents in its own margins: four justices dissented from the first paragraph, one from the second, three from the third with a fourth abstaining. The Court's statement says so in the open — "There is no perfect resolution and certainly none that satisfies everyone, including the members of the Court."

It also records the options it turned down. WSBA staff had recommended conditional admission through 31 January 2028, with a real exam or a new licensure pathway to follow. The deans of Washington's three law schools recommended diploma privilege — admission on the strength of the degree — for their graduates. The Court took neither.

Then the pressure came from inside. On 13 August the WSBA Board of Governors held an emergency meeting and formally asked the Court for equal treatment across all applicant categories. On 14 August the Court declined to modify the order. Its resolution, it said, "was not intended to be individually tailored to each applicant's circumstances".

An applicant campaign had formed around the same request. It is an interested party and says so; its argument is one line long. "We shared the same exam. We faced the same system failure." A retaker who signed it, Jorge, from Orlando, put the objection in the form the Court did not answer: a remedy is not equitable if the people who suffered the failure carry all of its consequences.

The man who was going to take it anyway

Jack Despain Zhou is thirty-one. He spent six years in the Air Force before law school, graduated from Temple, drove from Seattle to Yakima, took a hotel room, and spent 28 July in the convention centre in what he described to reporters as an immense amount of uncertainty. He was a first-time applicant, which put him in the eligible group. He could have had the licence without the exam.

Here is what he told Bloomberg Law about that: "It would feel wrong to me that I never showed that I was capable of completing that rite of passage."

He told the same reporter he was leaning towards accepting the waiver, and I do not know what he finally did. That is the part worth sitting with: he was going to take it, and he said that sentence anyway. It is not modesty and it is not a young lawyer being careful in front of a reporter. He is naming a loss, on the day he was handed a win, and he is naming it precisely: not that the licence is worth less on the market, but that he never got to find out.

That is the part institutions consistently fail to price. When you remove a barrier to spare someone the pain of it, you also remove the only mechanism by which they could have learned something about themselves that nobody can hand them. The waived cohort will practise law. Many will be excellent. And every one of them will carry a small unfalsifiable question for the rest of their career, and the Court's own list has made certain they can never settle it in public.

Which is the reverse of what the list was for. Anonymising the method of admission was plainly meant to protect the waived from suspicion. Its actual effect is to spread the suspicion evenly across everyone Washington admits this cycle — including the ones who sat in Tacoma in September and earned it. You cannot tell from the list. That is what the list says.

The best argument against all of this

It is not a weak argument and it should be stated properly.

The Court was handed several hundred innocent people and a failure none of them caused. Making them wait punishes the wrong party. Prior failure is a defensible, administrable proxy for risk to the public, and Washington's own figures support it: in the July 2025 cycle, first-time candidates from ABA-accredited schools passed at 85.44 per cent against 79.44 per cent for all such candidates. Repeat takers do carry more measurable risk. And anonymising an admission list is ordinary privacy practice that costs nothing, since employers can ask an applicant for a portable score directly.

I accept most of that. I accept the triage. Under a fortnight's pressure, with a broken vendor and no clean option, waiving the exam for the group least likely to have failed it is a defensible emergency act, and I would not want to have made the call.

The register is a different kind of decision. It is not triage. It has no deadline, no emergency, and no expiry. It was taken calmly, and it solves a problem belonging to the profession rather than to the applicants — the awkwardness of a public document recording that in 2026 Washington made lawyers without an exam. That awkwardness is real. It is also the profession's to carry, and it was moved onto the people it was meant to protect.

So the verdict, and it is the contestable part: Washington spared roughly 650 people a stigma by ensuring that none of them can ever disprove it.

I have no standing in this. I am not admitted to practise law anywhere in the United States and I do not practise there; I hold a law degree and a master's in law from Brazil, and no stake whatsoever in the Washington bar.

But the sentence about the list is not a legal problem, and it is not rare. It is the most common administrative reflex there is, and you have almost certainly signed one. It sits at the end of a policy, after the part everyone argued about, and it reads like housekeeping: the record will not distinguish. The promotion list that does not say who went through the panel and who was slotted. The certification register that does not separate the cohort who sat the assessment from the cohort grandfathered in when the standard changed. The alumni page that does not mark the honorary ones.

Go and find yours. It will take you twenty minutes. Then ask the only question that matters about it, which is not whether it protects anybody, but who was in the room when it was written — and whether a single person who would have to live under it was there.

Sources

Figures carry the date on which they were true. The counts of affected applicants differ between sources — the NCBE gives about 700 scheduled in Washington, press accounts about 645 to 650 unable to start — and that divergence is reported here rather than resolved.

This essay is one application of the method. The book is where it is built from the ground up.

One email when the first chapter is ready, and one when the book is. Nothing else. No sequence, no funnel.