Nothing Was Hidden
Oct 28, 2026
On a summer evening in January 2025 I
walked the public boardwalk along the Auckland waterfront and stopped in front
of a shopfront. The name sat on the glass at eye height, in letters a person
could read from across the street. There was a doorbell. There was a small sign
asking visitors to ring for attention. I took three photographs from the
footpath and kept walking.
The firm is The Campaign Company. It was
incorporated on 4 November 2021 and classified in the companies register as a
public affairs consultant service. Its registered office is Unit 6, 143 Quay
Street. Jordan Williams founded it, ran it as chief executive, and was its sole
director from incorporation until January of this year. He co-founded the New
Zealand Taxpayers' Union in 2013 and remains its executive director, and he
co-founded the Free Speech Union and sits on its council. The firm's clients have
included two tobacco multinationals and a lobby organisation dedicated to
removing Māori collective rights.
None of that is secret. It took an
afternoon with a browser, a public register and a handful of interviews the man
gave under his own name. Most New Zealanders have never heard of the firm. That
is the product.
We have a vocabulary for this and it points
the wrong way. Dark money. Shadowy networks. The hidden hand. Each of those
words carries a picture of something buried, and each implies the same remedy,
which is to switch on a light. Disclosure law, lobbying registers and donation
thresholds all rest on the belief that a fact has gone missing and can be
recovered. In Aotearoa, across the campaigns I have spent the last two years
tracing, no fact had gone missing.
In April 2025 billboards appeared carrying
photographs of Green MPs beside the slogan Defund the Police. At the foot of
each one, in the small type the Electoral Act requires, stood an authorisation
statement giving an address: Unit 6, 24/143 Quay Street. That is the same unit
as the registered office of the firm on the waterfront. The requirement was met
in full. A driver stopped at the lights could have read the address, gone home,
opened the register and found the company inside ten minutes.
What that driver could not do was learn any
of it at the moment the billboard was working on them. Disclosure arrives in a
database, after the fact, for the handful of people who go looking. The message
arrives at the intersection, in three seconds, wearing the face of a political
party it does not belong to. Both things are true at once, and the law notices
only the first. Compliance is what makes the practice safe.
The sharpest demonstration came later that
year, and it did not require anybody to believe anything false. Parliament
restored the rule that councils which had established Māori wards must put them
to a binding poll. Forty-two councils went to the vote. Across all forty-two,
542,134 people voted to keep the wards and 467,923 voted to remove them.
Eighteen councils kept theirs, twenty-four removed them, and the result binds
until 2031. A national majority lost because it was never asked a national
question.
The objection to reading it that way has
real force, and it deserves stating properly. Local institutions are decided
locally, and a margin of seventy-four thousand votes spread unevenly across the
country is thin ground on which to override forty-two communities. Somebody who
holds that position is not arguing in bad faith. What the position cannot
account for is the timing. Those councils had established their wards under a
rule Parliament itself had written. Parliament rewrote the rule, required the poll,
and fixed the unit at which the question would be put. The unit was not a
feature of the terrain. Somebody chose it, and the outcome followed from the
choice.
Disinformation is the wrong name for what
happened here. A lie can be answered, and sometimes the answer wins. The shape
of a counting unit cannot be answered, because it is not a claim. It sits
upstream of speech and decides in advance what speech is permitted to achieve.
Every hour spent correcting the record on Māori wards was an hour spent on the
one part of the contest that had already been settled.
This is the part that should worry anyone
who works in communication for a living. We are trained to treat the argument
as the site of the struggle, and we are good at it. We write the rebuttal,
commission the poll, correct the misleading claim, run the campaign that tells
the better story. All of that assumes the question will be asked in a form
where the better story can win. Decide the form of the question and you have
decided the answer without saying anything at all.
The second reflex is to follow the money,
and it fails for a related reason. The Taxpayers' Union is a partner of the
Atlas Network, which lists around five hundred institutes across more than a
hundred countries. Jordan Williams completed Atlas's Think Tank MBA in 2015,
held its Smith Fellowship, and was profiled by Atlas as an alumnus in March
2018. He has said that Atlas funding never amounted to more than one or two
percent of the Union's annual income. He is probably right, and the figure
settles nothing. Critics hear it and picture a chain of command running from
Washington to Wellington. Defenders hear it and conclude the connection is
trivial. Both are counting the same wrong thing.
In 2008, tobacco retailers in the United
Kingdom ran a campaign against cigarette display rules under the banner SAVE
OUR SHOPS. In 2023, dairy owners in Aotearoa ran a campaign against smokefree
retail rules under the banner SAVE OUR STORES. The second was paid for by two
tobacco multinationals and fronted by a dairy owner introduced to the public
only as Bhavesh. When a journalist asked to speak with him, his photograph came
down from the site. The two campaigns share more than a name.
Money leaves a trail a disclosure regime
can follow. Training leaves none. A fellowship leaves none. The working
knowledge of how to assemble a public that does not exist, how to give it a
first name, a website, a petition and a convoy of tractors, travels between
countries inside people's heads, and no register on earth records the crossing.
Declared third-party spending in this country rose thirteen-fold across a
single electoral cycle, and every dollar of it was declared on time.
During that referendum year, billboards
went up carrying a photograph of a Rotorua kuia beside a message against Māori
wards. She had not agreed to any of it. She learned about it when her whānau
phoned her at twenty to two in the morning. I cannot show who produced those
billboards, and the white paper this essay sits alongside says so in as many
words. The argument does not need that link, and I will not manufacture one.
Asked about it, Ellen Tamati pointed to her
chin, named the two waka her moko kauae carries, and told the country to go and
vote for the wards. She had no budget, no agency, no domain registered quietly
in November, no creative director and no authorisation line to hide behind. She
had whakapapa, and she had a reporter in the room. The second of those is not a
quality of character. It is infrastructure, and it can be taken away.
Voice is not something a person has. It is
something a person is able to use, and the ability rests on ordinary material
that nobody defends until it is gone: a newsroom with somebody spare to send, a
council that answers the phone, a register that somebody is actually reading,
an electoral rule that asks the question at the scale the answer deserves. The
firm on Quay Street never hid, because it never needed to. It works in a
country where the register is public and unread, where the newsrooms that might
have sent somebody down to the waterfront have been shrinking for a decade, and
where Parliament can change the shape of a question between one election and
the next.
Correct the record and you have corrected
the smallest part of it. Voice is built, held and defended, or somebody else
buys it and uses it against the people it was taken from.
The five campaigns, the evidence behind
each of them, and the register of what is documented and what remains inference
are set out in Manufacturing
the Grassroots, CARE White Paper 36A, free and open access.