In an era where fake news is like a virus, media organisations need to be very careful that they are not adding fuel to the bonfire of fantasies. Even with the best of intentions journalists should be wary of arguing that government data is dishonest – such as recent suggestions that the real unemployment rate is much higher than what the ABS would have us believe.

Last week, the Australian’s Adam Creighton wrote two articles suggesting the unemployment rate is a basically a crock. One of the articles was titled “Be honest about unemployment — it’s above 15 per cent”, the other had the headline, “Real jobless rate well above 20pc” (pick the number you think sounds most scary, I guess).

Creighton argued that because the unemployment rate only counts people who have actively looked for work in the past four weeks it excludes people who would work but were not looking. He also added into the mix everyone who is underemployed and thus arrived at a total “well above 20%”.

It is not.

Creighton has form on unemployment trutherism. In 2014 he suggested “Australia’s unemployment rate is now just shy of 51%. That is not a misprint.” He came to that number by counting everyone not working – including children, stay-at-home parents and those who are retired.

Little wonder last week he was writing glowingly of Donald Trump, noting that he “at least has had the courage to slam the US unemployment rate”.

What Trump has done is suggest there are 94 million unemployed Americans rather than 7.9 million. He arrived at this figure by including everyone over 16 without a job – including students, retirees, stay-at-home parents (you see the pattern here).

Trump has also suggested the unemployment rate “is phoney” and that “the number is probably 28, 29, as high as 35. In fact, I even heard recently 42%”.

This is the problem with Creighton’s article – not that he is wrong to suggest that the unemployment rate may not tell the whole story (it doesn’t), but once you start suggesting the number is dishonest, then you are very quickly down the rabbit hole of making up your own facts.

And unfortunately many people will read such articles and be suckered into going down that rabbit role – and before you know it politicians are dismissing official data because it doesn’t suit their agenda. Heck, in Senate estimates on Monday, One Nation’s Malcolm Roberts referred to the article in a question to the Treasury secretary.

So let me say it clearly and plainly: the unemployment rate is not fake and it is not tricked up.

The unemployment rate definition used in Australia and around the world is one set by the International Labour Organisation – not an organisation all that disposed to hiding levels of unemployment.

Aside from being able compare across nations, the common definition also helps stop governments changing it to make things look better.

Creighton argues that the people who were available to work but weren’t looking account for around 22% of all people “not in the labour force”. If we included them, it would increase unemployment by around 1.5 million and raise the unemployment rate from 5.8% to 15.9%.

That might be nice if you want to come up with a big headline number to shock people, but would it really be “honest”?

Let’s look at the reasons why these people were not “actively looking for work”.

In September 2013 (the last time the ABS looked at the makeup of people not in the labour force) only 13% of those who wanted to work but were not actively looking were “discouraged job seekers”. They stopped looking because they felt their age, or lack of skills/education meant there were no jobs, or that the jobs weren’t in their region, or at suitable hours.

They had given up, and they certainly deserve attention.

The US Bureau of Labour Statistics includes such people in its “U-4” unemployment rate (it has six types of unemployment rate, “U-3” is the standard one) – and I’d happily agree the ABS should be funded to measure these people every month.

If we included them, the unemployment rate would be raised from 5.8% to around 7.3%. Higher, yes, but hardly headline-blaring higher.

What about the rest of the people who are “willing to work” but not looking?

Most are not in the labour force for family reasons – either looking after children or an ill relative. The next most common reason is health – short-term illness or injury, or long-term condition or disability – and being pregnant. And the third most common category (two-thirds more than discouraged workers) are those attending an educational institution.

Does it really make any sense to include such people as unemployed? Sure they might work if an attractive enough job came along, but they are happy enough if one doesn’t:

The other common complaint about the unemployment rate – and Creighton raises it as well – is that “anyone working more than one hour a week is considered employed”.

I get told this one almost every time I write on unemployment. For the love of God, please stop! It is not an issue. And what is more, you know it is not an issue.

Ask yourself how many people you know who work only one hour a week? You might know someone, but I’ll bet you know a heck of a lot more who work more hours than that.

We know this is the case because the ABS counts the number of hours worked by all employed. Unfortunately, they don’t count only one-hour workers, but we can use the number for those who work one to nine hours.

In March, just 4.9% of all workers worked fewer than nine hours – the lowest amount on record. But the total is skewed by youth workers. If we count only workers aged 25-64, the percentage falls to a mere 2.6% – again the lowest on record:

But even youth workers, although more likely to work short hours, are not doing so more now than in the past. The share of youth working fewer than nine hours has fallen since 2014 to 14.5%, and is now just above where it was in 2004:

But let us assume we counted such workers as unemployed. This would raise the unemployment rate, but rather than give a more accurate view, it might actually distort the picture.

Right now, the unemployment rate of 5.8% is the same as it was in April 2004, and yet an unemployment rate which counted one-to-nine hour workers as unemployed would be 10.3% – a rate actually lower than such a rate was in April 2004:

So excluding short-hour workers from the “employed” actually makes things look better now.

As I have written previously, the issue with underemployment isn’t a boom of one-hour workers, but part-time work replacing full-time work.

And this is another problem with coming up with broader rates of unemployment – to stop discounting people, you end up exaggerating others – whether it be people working short hours, or students not working, or stay-at-home parents. If you want to include them as unemployed then OK, but is anyone also arguing they should be eligible for Newstart?

Somehow I suspect not – especially when Creighton also complains about people on Newstart who don’t actively look for work and blithely writes that “about 55% of 770,000 disability support pensioners claim psychological problems or muscle and joint pain; a portion of these might not be genuine”.

The big issue of the labour force at the moment is underemployment.

I have written on it often – and have even argued that the unemployment rate is no longer the best indicator of economic health. But that is a long way from seeking to suggest the official figure unemployment is fake.

And as anyone who has been out of work knows, there is a very great difference between having a job but wanting more hours and not having a job at all.

We already measure both underemployment and underutilisation (unemployment and underemployment combined):

My two favourite measures (as readers would know) are the employment to population ratio:

And the per capita hours worked per week measure:

But there is no one measure that tells the whole story.

We should ensure our coverage reflects that. There is nothing wrong with creating new labour force measures, but we should be very wary of dismissing the official rates as dishonest.

Doing so only serves to reduce people’s confidence in the impartiality of the ABS. And we should discourage anything that would give succour to politicians – such as Trump – who seek to undermine institutions by suggesting any data inconvenient to their policies is fake.