Election campaigns compress everything. Deadlines shorten, editors ask for more, and the urge to write the loud version of your argument becomes hard to resist. Most of the mistakes below are not dishonesty. They are shortcuts: the tribal line that saves a paragraph of thinking, the unattributed claim that saves a phone call, the policy-shaped hole where a policy should be. Individually they look small. Across a six-week campaign they add up to columns that inform nobody, persuade nobody, and read badly the morning after polling day. Here are the ones worth watching.
Tribal rhetoric: cheering rather than arguing
Tribal writing is recognisable by what it lacks: an argument. "Only one party understands people like us." "The other lot would sell off the health service." These are chants, not claims. They feel satisfying to write and they attract warm approval from readers who already agree, which is precisely the problem — they ask nothing of anyone and give nothing back.
Try this test before filing. Write one sentence describing your opponent's position that they would accept as an accurate summary. If you cannot, you do not yet understand the position well enough to attack it. Then state what you want to happen, who benefits, and what it costs. A column that names a trade-off is much harder to dismiss than one that names an enemy.
Unsourced claims and borrowed certainty
Election copy is full of phrases that look like evidence and are not: "it is widely accepted", "critics say", "experts agree", "the numbers speak for themselves". Each one shifts the burden of proof onto a crowd that never turns up.
Name the source. A published document, an institution, a named spokesperson speaking on the record, a page in a manifesto. If a figure comes from a think tank, say which one, and remember that think tanks have positions. If the claim is your own inference — and much of the best opinion writing is inference — label it as such: "my reading is", "the plan implies". Readers forgive a stated judgement. They do not forgive a fact-shaped guess.
Check the vintage of your numbers, too. Announced figures get revised, costings get updated, and a statistic from last autumn's fiscal event may be several versions out of date by the time you file. A pledge is not a costed plan, and a party's aspiration is not a policy until someone has decided who pays for it.
Treating policy detail as boring
Campaign coverage drifts towards gaffes, leadership and the theatre of the stump, because those things are new every day. Policy is slower. But policy is what a government actually does, and it is what most readers are deciding about: waiting lists, school places, rents, energy bills, the tax taken from a payslip. A column that never names a measure has nothing to be wrong about and nothing to be useful about.
Follow one policy properly across the campaign. What does it do? Who pays? Who delivers it? When does it bite? If a party promises more appointments, ask who staffs them. If another promises to cut a tax, ask what is cut instead, or deferred.
Three questions that save a column
- Who pays? Taxpayers generally, a named group, or a borrowing line that someone else inherits?
- Who delivers it? A department, a council, a private provider, or a workforce that does not currently exist?
- When does it start? In the first year, later in the Parliament, or "in time"?
Reporting the race and ignoring the stakes
Polls, tactics and the mood in the bubble are legitimate subjects, but they crowd out everything else because they are cheap to produce and easy to argue about. If you use a poll, say what it measured, who paid for it, and whether the movement sits inside the margin of error. One survey is not a trend, and a by-election result in one corner of the country is not a national verdict.
The stronger version of a horse-race piece tells the reader what the numbers would mean if they held: which policies become possible, which promises quietly die, which parts of the country end up holding the pen. Otherwise you have written a weather report about a storm nobody has to stand in.
Writing only for people who already agree
Assume a reader who is intelligent, undecided and busy. That rules out nicknames for party leaders, in-jokes from political feeds, sneer quotes, and acronyms that mean nothing outside Westminster. Explain the mechanism, not just the conclusion: not "the plan is unfunded" but "the plan commits spending that no budget line currently covers".
One useful form is the conditional address: "If you are considering voting for this because of X, here is what you are also choosing." It respects the reader, sets out the trade-off in plain terms, and lands far harder than an insult. It also survives being read by someone who ends up disagreeing with every word.
A checklist before you file
- Does every factual claim have a source a reader could check?
- Have I named a policy, or only a party's attitude towards one?
- Would an opponent accept my description of their position?
- Are my figures the most recent version available?
- Have I separated fact, forecast and my own judgement?
- Have I included the strongest version of the other side?
- Is there a line somebody who disagrees could still use?
- Have I resisted predicting the result?
Write for the morning after
Election columns age faster than almost any other kind. A piece that reads well on a Thursday mid-campaign can read as noise by Friday morning, so the closing question should be simple: would you be content for this to be the version of your thinking that survives the result?
Practically, that means filing earlier than usual, because desks are stretched and subeditors have less time to catch a shaky claim. Keep your sources in one place so you can answer complaints quickly and calmly. If you get something wrong, correct it plainly and at roughly the length of the original error rather than burying it. And keep one honest paragraph in reserve for the outcome you did not expect, because the column written entirely in advance is the one that ages worst of all.
Photo: webandi / Pixabay
