So we're told in this latest paper at least, the NOx emissions from diesel engines, over and above what emissions standards say they should be spewing out, cause 0.07% of deaths globally:

Diesel driven cars, lorries and buses churn out far more air pollution than standard testing procedures suggest, leading to many thousands of unreported deaths, scientists claim.

The excess emissions of harmful nitrogen oxide (NOx) exhaust gases can be linked to 38,000 premature deaths worldwide, according to the new research.

Mr. Google, our friend, tells us that there are some 54 million deaths a year globally. Thus that 0.07%.

At which point we've got to ask the obvious economic question - compared to what?

The heart of the problem here is that we have three things we'd like to have. Low diesel consumption for cost and CO2 reasons, low NOx creation for those health reasons and low cost engines just on the general basis that cheap is better than expensive. And as with the engineering wish list, faster, better, cheaper, we can only have two of the three.

As we crank up the combustion temperature in order to gain fuel and CO2 efficiency we end up creating more NOx. We can only avoid this by having expensive, not cheap, diesel engines - or adding the expensive urea systems to such vehicles. This is not an optional aspect of our universe. We cannot have all three, we can only have two of the three in pretty much any combination we desire. Low NOx is possible cheaply at the cost of general fuel efficiency, we can have low NOx and efficiency at high cost and so on.

So our 38,000 deaths are compared to what? To the extra cost of having no cheap diesel engines, only expensive ones.

No, we don't know the answer here either. But we do insist that this is the correct question to be asking. We've limited, scarce, economic resources. So, what is it that we'll have to give up to gain this purported reduction in deaths from NOx? For without that answer we cannot possibly hope to answer the question of whether we should do it or not.