PPS Numbers and migration estimates: The CSO's view

[civic_critic]

Apropos of our discussion a while back about the discrepancy between the CSO's estimates of immigration and the number of PPS numbers issued I decided to ask the CSO themselves about some of the points raised during that debate and see if they could clarify matters a little.
I had initially written to them some months ago about the apparent discrepancy between the PPS numbers and the migration estimates, at which point they had replied that they basically thought the excess numbers of people had gone home. I outlined their response from that time during our recent discussion here.

And I left it at that. Until our recent discussion that is. I decided to ask the CSO some more specific questions to try to get to the root of these things and so took up where I had left off. I asked them last week about the announcement of an increase in net migration since 2002 of 200,000 and about the discrepancy between that number and the number of PPS numbers issued over the last 6 years. They replied as follows:

As I said in my original email to you, the allocation of PPS numbers to non-Irish persons is consistently higher than the immigration of non-Irish persons into Ireland. We believe that this is because a large number of people arrive and obtain no work or else for for a short period of time and then return home. Analysis of Revenue data supports this view. Such persons would not be captured in in the annual population estimates as immigrants.

It is not possible to add/subtract estimates of immigrants/emigrants from the population estimates and migration release with numbers of PPSN's issues to non-Irish nationals as they are different types of measures.
The former measures actual inflows/outflows over a 12 month period whereas the latter measures people who apply for a PPSN but we have no idea of where they subsequently live.

Hope this helps you,

Regards,...

I replied:

many thanks for your reply. I'm afraid I'm still not entirely clear on this and given the importance of the issue and the enormous discrepancy between the figures I'd really like to be absolutely clear. The PPS numbers verify and record the full numbers of people who have actually come here to live and work whereas the figures for gross immigration used by the CSO are derived from the QNHS, is this so? The figures derived from the QNHS are based on a question asked once annually in April about the migration pattern of the respondent and the figures arrived at are 'grossed up' to cover the entire country?
My question therefore is, what is the definition of 'gross immigration' used by the CSO and how does it differ from 'net migration' in any given year? To my mind gross immigration suggests those who came anytime throughout the year for the purposes of living and working in the country and net migration would be those who remain after a year. 'Gross immigration' therefore, I would have thought, is precisely reflected by the PPS numbers. I am not a statistician so perhaps you might briefly explain the difference for me if you have a moment (I fully appreciate you have plenty of other things to be doing). Most ordinary people would think that gross immigration means those who come here to work, which to their mind would be the same as the PPS numbers.

Given that 183,500 PPS numbers were issued last year yet the CSO figure for net migration of non-nationals was around 70,000, does this mean that the CSO believes that all 113,500 of the other people have left the country and were in fact 'transient workers'? If so, why has their departure not been reflected in the figures for gross emigration, which numbers 17,000 of both Irish and non-nationals in total? How can the QNHS, by asking a question once in April, be justifiably used as an accurate and reliable source of data for 12-monthly migration flows yet the PPS numbers - gathered throughout the year - are excluded from such results? And why, if the PPS numbers are excluded from the results for migration flows, does the CSO state in its background notes to 'Population and Migration Estimates 2006' that "other migration indicators are also used", including "the number of PPS numbers allocated". At the very least one would have thought these numbers would appear in the figures for gross immigration since they are the most verifiable data on migration inflows that we have.

And finally, are there any differences in the way the QNHS is administered and analysed from the way the Census is adminstered and analysed, other than in terms of scale? I ask because I wonder is there a significant methodological difference between the two such that the QNHS may miss 113,500 in one year yet we are to accept that the census figures are more accurate? If both use the same fundamental methodology for collecting their data and one has consistently failed to pick up over 100,000 people in one year, then why should we believe the accuracy of the other? In short, how accurate does the CSO itself believe its estimates of non-national immigration are from the census of 2006? From my own research I have found the following info for the gross immigration of non-nationals:

CSO estimate of gross immigration 2001 - 32,700; PPS numbers issued 2001 - 94859
CSO estimate of gross immigration 2002 - 39,900; PPS numbers issued 2002 - 88476
CSO estimate of gross immigration 2003 - 33,000; PPS numbers issued 2003 - 82542
CSO estimate of gross immigration 2004 - 33,200; PPS numbers issued 2004 - 124534
CSO estimate of gross immigration 2005 - 51,000; PPS numbers issued 2005 - 185639
CSO estimate of gross immigration 2006 - 67,200; PPS numbers issued 2006 - 183,500

These are extremely big discrepancies between the estimates for gross immigration and the actual verifiable numbers of people arriving in the country seeking work according to the issuing of PPS numbers. Is the CSO saying that 502,550 more people have immigrated into Ireland over the last 6 years than have been recorded in official estimates and that all of these people have subsequently left, again without being recorded within official estimates for gross immigration and emigration? Or is the CSO saying that the estimates have been inaccurate? If the estimates have been inaccurate, by how much does the CSO believe the data is inaccurate and how many people does the CSO believe may still be in Ireland over and above the official figures given both by its estimates and by the latest census?

Lastly, you said in your email that "Analysis of Revenue data supports" the view that most people return home. I presume this means that tax returns and the discontinued use of PPS numbers confirms that people are no longer in employment and therefore have presumably left. I wonder can you point me to that data - I have asked the Department of Finance for precisely this information but they have not been forthcoming.

I hope I haven't been too long-winded. Thanks for your patience and your time and I look forward to your reply.

Kind regards,...

The complete reply I received this week (minus the names) was:

...

The population estimates release does not use the term gross immigration.

Net migration is defined as the numbers of immigrants less emigrants.

As I have stated before, it is not possible to add/subtract estimates of immigrants/emigrants from the population estimates and migration release with numbers of PPSN's issued to non-Irish nationals as they are different types of measures. It would be like trying to add apples and oranges.

Regards,...

I replied thus:

thank you for your reply, though frankly it's so curt as to border on the fatuous if you don't mind my saying. 'Gross immigration' and 'gross emigration' are terms frequently used in the literature to describe migration inflows and outflows and to differentiate these from net migration over any given period. Obviously these were what I was referring to, especially since I differentiated them from 'net migration', a phrase used within the literature and also by the CSO. However, to be clear let me rephrase my question from the previous mail: what is the CSO's definition of 'immigrant' and how does it differ in fact from someone who applies for a PPS number? Does the CSO use the QNHS to measure 12-monthly flows and does the CSO believe that this on its own is adequate to accurately record these flows? Does the CSO use PPS numbers as an indicator of migration inflows, as it says in its reports? Does the CSO use revenue data as a indicator of outflows, as you suggested in your last mail? If so, can you point me to that data as I have asked the Department of Finance for information but they have not been forthcoming. What margin of error does the CSO employ in its estimates and how accurate overall does the CSO believe its estimates of net migration are year-on-year for the last 6 years?

I don't wish to appear facetious about this; I am genuinely looking for enlightenment about this serious matter but a lack of information as well as apparent discrepancies in the data mean that the public debate over this issue is frequently confused and contentious. I am attempting to get to the bottom of the matter but simply saying that it is all a difference between 'apples and oranges' does no credit to the reasoning outlined in my prior email nor, to be frank, to my intelligence as a citizen seeking clarity on these issues from a department of government.

I trust you understand my position and might give me some answers to at least some of the questions I have outlined.

Again thanks for your time and I look forward to your reply.

Kind regards,...

The full reply (minus only the names) that I received this morning was:

...

I think that I have been very clear in my email correspondence with you
to date and I believe that I have nothing further to say which could be usefully added to this discussion,

Regards,...

Make of that exchange what you will.

[dub006]

hi civic civic

your analysis is brilliant.The replies you received say to me that they dont know the answers to your questions or they are afraid to give them to you.
Their reticence in answered very good questions should make everyone suspicious.The CSO is suppposed to be independent.If its not doing a good job in gathering the facts or if they have they are severely challenged in communicating it.

[ibis]

Funnily enough, I think the CSO's answer is both clear and sufficient. They've stated that they can't capture short-term 'transient workers', and that that's why the PPS issue numbers are much higher than the net immigration numbers. They've also explained that you can't simply compare PPS issue numbers with immigration/population figures because they're totally different types of measure.

Oddly enough, I seem to remember saying exactly the same on the thread you refer to. But, well, you clearly don't like it, so I can't imagine you'll do anything but stick with your own "figures". I suspect this is why the CSO got a little curt towards the end - you were making it pretty clear you thought they were talking out their holes.

[Pidge]

Have you asked a TD to table a question? That might be the best way to proceed.

[dub006]

I recently read but have been unable to confirm it that the CSO does not collect data on self employed people.
does anyone know if this is true.?

also a perusal of Dept of Trades website catalogs work permits issued by nationality.It does not coerrelate in any way with the numbers recorded in
the census or with PPS numbers issued.
For instance 14 work permits were issued to romanians since January 1.
Yet the CSO indicates that approx. 6,500 PPS numbers were issued to romanians in that period.

[FutureTaoiseach]

Well even the self-employed are supposed to have PPS no.s. I agree there is a problem with the bogus self-employed among Romanians however. On the discrepancy between the Census and the PPS no.s I sortof accept that a lot of the Eastern Europeans did go home - after all why would someone from the EU-10 feel the need to hide from the census enumerators? They are not here illegally so I am broadly satisfied with the estimates of the numbers of Poles, Lithuanians etc. here - though I think it has certainly risen to much higher since the 2006 Census - remember 11,000 a month were coming in since EU Enlargement. However I still think the overall non-national population is higher than 10%. Why? Well look at the tens of thousands (around 40,000) who refused to tell the census enumerators their nationality. Obviously they may have feared deportation. Furthermore - and especially striking - is the tens of thousands of properties deemed unoccupied. The basis for them being labelled thus seems in party related to noone coming to the door. I am convinced that in many of these properties there were occupants who were illegal immigrants hiding from the authorities. In some cases I believe there to be collusion between exploitative employers who use cheap illegal foreign labour and the illegal immigrants themselves in terms of hiding them from the authorities. Both parties gain from this - the illegals by escaping deportation and the employers by still having cheap labour.

[dub006]

In the US it is accepted that a a large number of illegal aliens are not counted in their cencus for the reasons that the previous coontributor stated.
They estimate there are between 12 and 20 million illegals.
I think the illegal numbers here are much higher than 40,000

For instance reports in the Irish times refer to a chinese population of 100,000.I have seen an interview with chinese who maintain the figure is higher.The number of work permits issued to chinese since record began is less that 6000,the number of student visas is around the ssame number.

The Irish times reported last week from Gort in Galway on a substantiaal illegal Brazilian population.I have noticed the same inconsistencies occurring with other nationalities such as nigerians,russians,south africans

[Hana]

Solution:
Right lads I have the solution to all this immigration malarky. We'll take all the immigrants in Ireland & trade them for all the Irish immigrants in the USA. I'm thinking about 2 boats leave New York & Cork at 3pm next wednesday (subject to high tides).
Hey presto we get our missing dispora back & no more non irish nationals.
Now who's going to tell my cousin in Chicago he has to leave his lucrative contractor business with 42 employees and come home to dublin to work in the toilets of copper face jacks.......

[feenxp]

Surely the CSO should know how many PPS numbers are inactive?

You need your PPS number for work, Benefits etc, so surely it would be easy enough for them to see how many PPS cards issued over a period of 5 years are not in use. That would then give us a real figure!

[Catalpa]

The CSO are out of their depth in monitoring the numbers of immigrants.

They have not got the moral courage to admit this though so instead try to persist with figures that make no sense.

For instance if we take the figure of the population of the State in 2002 as 3,917,203 and the figure for 2006 as 4,239,848 and an officially recorded increase of immigrants of the order of 200,000 or so over those 4 years we have an annual average increase of immigrants of 50,000 per annum.

So if these figures were to be believed we have experienced fractional increases in immigration over those years of the order of 1/76th for 2002/2003 to close to 1/84th for 2005/2006. They might not be 100% precise but I’m sure you will see my point.

I don’t think any sane person could believe that the changes we have seen in those years were that low.


BTW and while we are here:

I posted this up last week but in case any of you missed it:

CSO admits its figures can’t be trusted.

The CSO has finally admitted that its ability to conduct an accurate Census is flawed.

On the front page of this weeks Metro Eireann it states:

CSO ADMITS IMMIGRATION STATS ARE ‘NOT DEFINITIVE’.

The CSO has admitted that the figures it released concerning the number of immigrants in Ireland [sic] may not be accurate. The figures, released last week as part of its principal demographic results, fall well short of unofficial estimates for the number of people from outside Ireland currently living here…

A spokesperson for the CSO, the organization tasked with providing the statistics that inform Government policy on immigration and other issues, told Metro Eireann: ‘I don’t think these statistics give a definitive figure.’ She continued, ‘If someone doesn’t want to have their details appearing on a census form then there’s not a lot the enumerators can do about it.’

Well there you have it folks – even the CSO admits they can’t do their own job.

BTW Wednesday I seriously hope you will now see the light on this one and stop being so naïve about the CSO and what they come out with!

The report concluded there were just 11,218 Chinese people in Ireland on census night, a figure far fewer than estimates which generally put the number of Chinese people living in Ireland [sic] at between 80,000 and 100,000.