A World Bank vice president has noted that about half of young Bangladeshis ageing up into the job market have not found a job.
If we look at other official, internationally-recorded numbers like the International Labour Organization we see that youth unemployment is about 14% -- double the country's more general unemployment rate.
It cannot be that both numbers are true -- so which is wrong?
Well, the first thing to note about such economic numbers is that all of them are wrong.
It's simply impossible to be absolutely correct about any such large number and it's not because we run out of fingers and toes to count them upon.
It's just that anything as large and complex as an economy is difficult to count.
My own insistence about economic numbers -- and do note this is me, not something generally agreed -- is that we're lucky if we can get the right number of digits and also the first one correct.
So, one thousand, six hundred thousand, five million, those might be about right. Anyone who gives us a number of 13.27% for a society of 150 million people is just proving they have a sense of humour.
Which is, of course, why detailed planning of anything to a specific and accurate target doesn't work.
We don't know where we're starting from, we'll not know where we get to, and we'll not know how well we're doing either -- we can deal only in generalities.
However, even with that insistence, we're always talking about vagaries; those two estimates are so different that we do need an explanation.
Here it's that the two estimates are using different definitions.
The World Bank is talking about formal jobs. With a contract, within the law, at or above minimum wages, and so on.
The ILO is talking about people going out to work and getting paid.
These are different things as anyone who has spent three minutes looking at the world around them knows.
We all know people not working very hard at formal jobs and we all also know people working extremely hard at what isn't a job but is work.
My own view is that -- I am one of these extreme free market people after all -- the distinction shouldn't matter. That most of the world doesn't agree with me is just one of those things.
But if it is formal jobs that everyone thinks should be achievable -- with all those protections over wages, holiday time, being fired, and the rest -- then the existence of huge numbers of people able to find work but not a job tells us something.
It tells us that the costs we're imposing upon the job -- those protections of minimum wages, holiday time, being fired -- are too high.
Too high for the work to be transformed into a job which is subject to all of those costs.
Loosen and reduce the costs of employing someone formally and more of those informal jobs will become formal jobs.
It has long been a standard analysis of the European labour markets that there's a division between north and south. The “Latin” (Italy, Spain etc, those who speak derivatives of Latin itself) economies have very strong protections, very high costs, for those who have a formal job.
The Nordics (so, the Scandinavians, Norway, Sweden etc) do it differently. Few of those formal protections -- there is no minimum wage for example -- but an excellent welfare system to retrain those who do get fired.
The people who suffer most from that Latin system are the young -- they're untrained, untried, no one knows how they will work out if they are hired.
But if they are and they don’t work out, then it costs a fortune to fire them.
So, they don't get hired -- youth unemployment rates in the Latins are hugely higher, two and three times as a percentage, than in the Nordics.
It's even possible to see it changing in real time.
In my native Britain, recent governments have increased the wages and protections offered to the young in formal jobs along with the taxation of those employing them.
As some of us would expect, this has led to a surge in youth unemployment rates, so much so that the British government has recently announced a tax subsidy for employing young people. The very young people their previous actions made more expensive to employ.
It's possible to think that perhaps the government should work better than this.
But back to our World Bank and ILO differences. One is measuring formal jobs, with contracts and legal protections, the other is measuring work being done without them.
One of those little things that economics teaches us is about supply and demand. Here, it's that increasing the cost of formal employment leads to less demand for it.
Exactly the thing we find out by noting that employers will demand more labour if they've not got to pay those costs of those formal employment contracts.
Or, to really shout about that supply and demand idea, if you want employers to demand more formal labour then you've got to make that formal labour cheaper.
No, this doesn't mean -- necessarily -- reducing wages. Reducing all the non-wage costs works too.
At which point, here at the end, it's necessary to point out that economics has no view on this. We are being purely descriptive, not suggestive.
We're not saying anyone should do this or that, only that if people do this or that then this will be the outcome.
If you want more youths to get a formal job then the price difference to an employer between a formal and informal job must reduce.
That's just the way this universe works.
Tim Worstall is a senior fellow at the Adam Smith Institute in London.
