The British people are wrong about everything. So says Ipsos MORI’s Bobby Duffy in a piece slamming our ignorance about immigration, Muslims, teenage pregnancy, benefit fraud, and foreign aid.

The piece got a mention on our Journalism and PR Facebook page, where we were all warned to ‘think on…’.

So I did.

And I’m not impressed. I don’t like the tone of the article, which is patronizing on an epic scale.

And I don’t like the substance either. Duffy’s problem is that he doesn’t believe us, but he does believe official statistics. Which are rubbish. As in …

Immigration
Duffy says we believe 31% of the population are immigrants, while ‘the official figure is 13%’.
But the official figure is meaningless.
Tony Smith, former chief executive of the UK Border Agency admitted in April 2013 that “we just don’t know who’s here and who isn’t“.
And MPs recently issued a warning saying that our immigration statistics are “unfit for purpose”.

Foreign aid
Duffy says we think we spend more than we do. But actually, the state says we spend less than we do. Our aid spending (£8.6bn) should hit 0.7% of gross national income this year, making us the first the G8 country to hit that target.
But private donations added another £1.1bn last year. For some reason, this figure isn’t included in our aid spending. The official figures just tell us how much of our money the state spends.

Your number’s up
But the real point is that politicians and policy wonks see stats as the big story. And they’re not. They’re just part of the story. And an unreliable part at that.
So think on this:

• There are lies, damned lies and official statistics;
• If you torture your statistics for long enough, they’ll tell you anything;
• 89.3% of all statistics are made up on the spot.

And finally …
The best way to handle statistics is to follow the example of the late Sir John James Cowperthwaite, a British civil servant.

As Financial Secretary of Hong Kong from 1961 to 1971, Sir John was tasked with finding ways the government could boost the then colony’s economy.

He rightly decided that the best thing the government could do was nothing. So he refused to allow his bureaucrats to collect economic statistics in case it encouraged them to meddle in the economy.

The result? Hong Kong became one of the most powerful financial centres in the world.

Sir John is my favourite civil servant, by some considerable distance.

I’d trade him for a hundred Duffys.

I’m looking for a starting point for some sessions on data-driven journalism – one that’s based on scepticism, rather than one that just launches us into the technology and the tools.

This is an interesting quote from the Data Journalism Handbook:

When information was scarce, most of our efforts were devoted to hunting and gathering. Now that information is abundant, processing is more important.

The key to the quote is the conflation of data and information. Yet it’s a truism of knowledge management that these aren’t the same thing. Information is data that have been processed.

So an abundance of information merely means that there’s more data-processing going on. This might be for a variety of reasons: easier tools; cheaper kit with more processing power; more people to do the processing …

In this context, of course, by more people, I mean more journalists.

So is the growth in data-driven journalism based merely on a (per)version of Parkinson’s law – the more people there are to process data, the more information there is?

And, if so, so what?

Here’s what: a key assumption behind data-driven journalism is that it has a value-free source as its basis: raw data.

But does it?

US journalism professor Mindy McAdams asked this question on Twitter yesterday (28/04/2013):

How is it okay that journalism students are able to graduate without ever taking a real statistics or mathematics class?

She was quoting from this post by Katie Zhu, who’s studying computer science and journalism at Northwestern University. Zhu has an equally good follow-up question:

Why is there not more room in the curriculum for computer science or statistics specialities?

… and a pointer:

the bigger issue is leveraging the existing resources (outside of the j-school) at universities to supplement journalism curriculums.

Any thoughts?