Editorial standards

How I check what I publish

Hope only helps if it’s true. So I show my working, I grade my own claims, and I’m as careful with the bad news as the good.

How sure am I?

Every main claim on this site gets one of five plain labels, so you know how much weight it can bear:

• Well established. Many independent sources agree, and the trend is large and long.
• Likely. Good evidence, with some gaps.
• Contested. Serious researchers disagree. I explain both sides.
• Speculative. A possibility or forecast, not a finding.
• My idea. A way of seeing things that’s mine, not a measurement. I’m proud of these, but I won’t dress them up as data.

Where the numbers come from

I trace every figure back to the original dataset, not the article that quoted it, and I open every source myself. Official statistics, large datasets with published methods and systematic reviews carry the headline claims. Single studies and expert books support them. Press releases and one off polls only ever add colour, and anything I can’t trace doesn’t appear at all. The central claim of each article rests on at least two independent sources.

Some numbers are easy to bend without lying, so I check the usual traps: whether a different start year would change the story, totals against per person figures, what happens to a global average with and without the biggest country in it, and whether the definition quietly changed along the way. Very old figures (before about 1950) are often careful reconstructions rather than counts, and I label them that way. Where the choice of poverty line changes the picture, I show more than one.

The Evidence panel

Every article ends with a short, folded away Evidence panel. Open it and you’ll find the main claims and how sure I am about each, the strongest objection and where I answer it, what would change my mind, the sources grouped by claim, any corrections, and the date I last checked it all.

Bad, and better

Progress is real, but it isn’t the whole story, and an optimist who only reports good news isn’t worth listening to. Every article gives real space to the strongest case against its own argument. At least one article in six starts from something that’s going wrong. And each year, alongside the annual progress report, I publish a Still Loading ledger of what got worse, graded with the same care.

When researchers disagree, I set out what they agree on, where they part company and why, and what evidence would settle it. I name critics, give them their best case and concede what they get right. I don’t mock anyone or label them by their politics.

Trends, forecasts and hopes

A trend is what has already happened. A projection is what a named organisation expects, and I say who. My own guesses about the future are kept on separate pages, split into likely, plausible and speculative, each with a date by which we’ll know if I was right. I’ll score them in public.

Real stories, real people

• I never invent a person, a scene, a quote or a number.
• Real people’s stories are told with their permission, or with their name and identifying details changed so nobody can be recognised. When I change a name, I say so.
• Quotes are short, accurate, credited and traced to where they were first said.

Keeping it fresh

Facts go stale at different speeds. Anything about energy, AI or conflict is rechecked every quarter, health and wellbeing figures after each major annual release, and long run history every couple of years. The “last checked” date in each Evidence panel tells you where things stand.

When I get it wrong

I will, sometimes. Every article has a line inviting you to report an error, and I aim to reply within 14 days. Mistakes are corrected in place with a short, dated note saying what changed. If a correction changes an article’s main claim, it also goes at the top of the piece and in the next newsletter. Every correction is listed in one public log, and a “Where I was wrong” page keeps the bigger ones. I never change a publication date to make something look new.

Independence

Everything here is free to read. There are no adverts. I take no money from political parties, campaign groups or advocacy think tanks, and a funding page lists every source of income over £250 a year. If I link to a book and might earn a small commission, I say so next to the link. Sponsors, if there are ever any, never touch the articles.

What I stay away from

• Party politics. I follow long trends, not elections.
• Health advice. I report research carefully, and nothing here replaces advice from a professional.
• Spiritual claims presented as science. Articles about the inner life use the psychology evidence, graded like everything else.