Sure it is only statistics?

 Last week I posted a blog about comparative data relating to asylum seekers including numbers deported or ordered to leave.  Coincidentally, some breaking news came to my attention after posting the blog that EUROSTAT had released data apparently showing that Ireland had the highest rate of ‘illegal’ non-EU migrants per 1,000 of population.

This was odd because the data that I had carefully extracted from Eurostat, checked and published on this blogsite showed the complete opposite!  Surely, I was wrong and Eurostat was right? I rechecked and rechecked data sources, formulae and the published Eurostat document cited by a number of media outlets and jumped on by a few politicians in Ireland to make the case that not only was immigration ‘out of control’ but that we had more ‘illegal’ immigrants per 1,000 of population than any other EU Member State.  All my raw data were the same as those used by EUROSTAT.  I was using a figure of 1,485 ‘illegals’ (I would prefer the term irregular but EUROSTAT use the term ‘illegal’).  I got a rate per 1,000 of population of 0.28 or 0.36 if you average the data over 10 years.  I realised that EUROSTAT had published completely wrong data based on correct raw data for population and numbers of ‘illegals’. Someone, somewhere had miscalculated.

EUROSTAT, once it was drawn to their attention, corrected the data and updated the relevant document here.

This sort of thing really should not happen and I believe it is rare. However, mistakes can happen and I firmly believe that EUROSTAT and the national statistical offices across the European Union are institutions of integrity practicing their profession by the highest standards of public service and professional code (a declaration – I was a Government statistician for many years but am now happily retired writing blogs among other things).

So far so good.

Now for the hard bit.

Statistics are not just numbers.  Statistical information is vital in the current debates about immigration, climate change, economies and societies. Misused or misunderstood by politicians, media or the general public we can face serious trouble. Moreover, a statistic or set of statistics may be correct but may not show the full picture about a complex and fluid reality. Data are not always comparable especially through time and across countries. But, that is beside the point.

Was the Eurostat error, which was thankfully corrected within 24 hours of it being noticed, just another blip on the radar? I suggest not for the simple reason that the error was picked up and used by political forces to prove a point. In the case of Ireland the narrative was that Ireland is soft on immigration and we have more ‘illegals’ than anyone else. Within 24 hours the recalculated figures showed more or less the opposite. I reproduced, below, two screen grabs – one false and one corrected.

First, the incorrect data as of yesterday morning.


And now for the correct data as of late afternoon yesterday



In the follow up to the correction made mid-afternoon on Monday 13th May, a number of media outlets mentioned about a decimal point being shifted.  This might suggest a computational error that shifted Ireland by a few decimal places. However, the dramatic change in country rankings tells a different story.  In the incorrect version published by EUROSTAT and which was on their website up until yesterday, Ireland had the highest rate per 1,000 of population while today it has the 5th lowest.  However, the rank ordering for other countries shows huge shifts: for example Croatia went from having the 4th lowest rate yesterday to the highest rate today while Germany went from 20th position to 11th position and Czechia went from 2nd to 17th over night.  Was anyone in Germany or France or Italy paying attention to this in recent weeks?  Were the Sweden Democrats jumping up and down when they were showing the 3rd highest rate of ‘illegals’ and today are at position 24 (fourth lowest)?

My point is twofold:

  1. This is not a critique of the statistical agencies involved.
  2. The error was potentially (and actually in the case of Ireland) a politically explosive  one misleading (unintentionally on the part of the agencies involved) the general public.

My conclusions:

  • EUROSTAT needs to say something about this publicly
  • Media and politicians who have moved on from yesterday’s story need to acknowledge that this was not a small error for one or two countries.

We need to be mindful of how statistics can be misused by the multiple re-posters (and not the original posters who were just using the story) who have no conscience or propriety when it comes to matters of international migration. In this case, the error and the claimed situation has become well entrenched in the cesspit that is much of social media to the detriment of public welfare and informed political discussion.

 

Comments

Popular posts from this blog

The benefits and costs of immigration in Ireland

Are asylum seekers scammers?

To Rwanda or Connacht?