Editorial Standards Policy Template
Two documents and two sheets that draft your policy from two years of real sourcing, anonymity, correction and conflict decisions, not a boilerplate.
Free download · No account needed
Every editorial standards template on the market gives you the same thing: a topic list, sourcing, anonymity, corrections, conflicts of interest, and a plausible-sounding clause for each one. The Online News Association's own builder works through 45-plus modules across seven categories the same way, letting you pick clauses off a shelf. None of it checks whether a single clause matches what your desk has actually decided when the topic came up for real, which is the gap a written policy exists to close in the first place.
This pack reads the decisions first. The Corrigan Sentinel, an invented regional daily, logged 208 sourcing, anonymity, correction and conflict decisions across 24 months: 74 anonymity requests, 96 corrections, 38 disclosed conflicts. Checked against a standard 40-clause policy, 24 clauses had a supporting precedent, 11 had never been tested at all, and 5 were directly contradicted. One absolute anonymity clause required no other way to obtain the information; 8 of 51 real grants did not follow it, because the desk's actual line is a reasonably safe alternative, not the mere existence of any alternative.
Open the pack in River and the agent reads your own sourcing, correction and conflict history before it drafts a single clause, or download the blank sheets and docs and build the register yourself. Either way, the Precedent Register holds every decision the policy is drafted from. The Standards Policy states which clauses are supported, untested or rewritten, and the Escalation Path names who actually decides each kind of call rather than routing everything to the top by default.
What's in the pack
Precedent Register
One row per real sourcing, anonymity, correction or conflict decision, with who decided it and the rationale given at the time. This is what every clause gets checked against.
Standards Policy
Drafted clause by clause from the register. Every clause is labelled supported, untested or rewritten, with the count of precedents behind the label stated in the text.
Escalation Path
Who actually decides each kind of call, tallied from the register, with the specific threshold that moves a case up a level and how fast it has to move.
Review Calendar
A per-topic cadence plus the trigger that moves a review earlier: a stated count of new decisions, or any single one a standing clause does not clearly cover.
The contradicted-clause check
A second pass over every drafted clause, searching specifically for the one precedent that did the opposite, so an absolute rule does not ship unchecked.
A space rule every prompt reads first
The policy is derived from precedent, not drafted from a boilerplate: a clause with no supporting decision is labelled untested, never adopted as settled fact.
Reconstructed-decision flagging
A decision with no paper trail is logged from memory rather than left out, flagged as reconstructed so the register states plainly how much weight it can bear.
How to use it
- 1
Open it in River, or download it
Open the pack and the agent reads your own decisions before drafting anything, or download the blank sheets and docs. The template library holds the rest of the newsroom's packs.
- 2
Send what you have on past decisions
Anonymity requests and their outcomes, your correction log, and any record of a disclosed conflict, however incomplete or informal.
- 3
Draft each clause against the register
Every clause is checked for a supporting precedent, silence, or a contradiction, and labelled accordingly before it ships as policy.
- 4
Set who decides, and when to check again
The Escalation Path is tallied from who actually decided past cases, and the Review Calendar sets when new decisions get checked against the standing policy.
Frequently asked questions
Is this template free?
Yes. Download the whole pack as Word documents and CSV sheets, no signup and no credit card. Edit with AI is a separate, optional path for anyone who wants the agent to read your own decisions and draft the labelled policy. Nothing happens until you send it something to check.
What if we have never written any of our past decisions down?
Log them from memory instead, flagged as reconstructed rather than left out. An editor's recollection of why an anonymity request was granted three years ago is thinner evidence than an email thread, but it is still a real decision, and an omitted one is indistinguishable from a decision that never happened.
How is this different from a generic code of ethics template?
A generic template gives you a plausible clause for each topic and asks you to sign it. This checks each clause against your own logged decisions first, and it labels the result: supported, untested because the desk has never faced it, or rewritten because a real decision contradicted it. The label is the product, not the clause alone.
Does a contradicted clause mean our desk did something wrong in the past?
Not necessarily. It usually means the boilerplate clause was written more absolutely than any real newsroom actually practices, such as granting anonymity only when a source has no reasonably safe alternative rather than literally no alternative at all. The point is to write the policy to match the real, defensible standard rather than an absolute one nobody follows.
Does this replace legal review of a specific story?
No. This drafts the standing policy from a pattern across many past decisions. Reviewing one specific piece before it publishes, and gathering the evidence counsel needs to assess it, is the pre-publication legal review prep tool. Checking the assertions in one draft against its sources is the fact-check pack.
Who should decide the labels: supported, untested, or contradicted?
Whoever holds editorial authority over the policy, using the register as evidence rather than memory. The register itself should be checked by more than one editor when possible, since a rationale one editor recalls clearly can read differently to the desk editor who was actually in the room for a given call.
What format are the downloaded files?
Two Word documents and two CSV sheets, zipped. The documents open in Word, Pages and Google Docs; the sheets open in Excel, Numbers and Sheets. Add a PDF query string to the download if you want to circulate the policy rather than fill it in.
Write down what your desk actually does, before a dispute
Download the blank Precedent Register, Standards Policy, Escalation Path and Review Calendar as Word and CSV, or let River check a clause against two years of your own decisions.
Edit with AI