QEC Autumn 2010

The latest QEC is here. Here’s the press release:

  • We expect that GNP will contract by 1½ per cent this year. For GDP, we expect a decline of ¼ per cent. For 2011, we expect GNP to grow by 2 per cent and for GDP to grow by 2¼ per cent.
  • We expect that employment will average 1.86 million this year, down 68,000 from 2009, a fall of 3½ per cent. We expect the rate of unemployment to average 13¼  per  cent. For 2011, we expect the number employed to average 1.85 million and the rate of unemployment to average 13½  per cent.
  • In the year ending April 2010, the CSO recorded net outward migration to have been 34,500. This was well below our forecast of 70,000. We discuss how this figure of 34,500 seems to be a conservative estimate of the rate of outflow when compared with estimates of migration contained in another CSO publication, namely, the Quarterly National Household Survey. We expect the net outflow in the year ending April 2011 to be 60,000. This is an increase of 10,000 on our earlier forecast for the year ending April 2011.
  • The General Government Deficit is expected to be 31 per cent of GDP this year, a truly dramatic figure. Of course, almost two-thirds of this is a one-off extraordinary item related to the banking bailout. For 2011, we expect the deficit to be 10 per cent of GDP, based on a budgetary package of €4 billion in savings.
  • In our General Assessment, we look at the budgetary challenges facing the country and in particular at the prospects of bringing the deficit down to sustainable levels in a reasonable timeframe. Using the “Low Growth” profile as published by the ESRI in July 2010, we assess what level of savings will be required to achieve a deficit of 3% by 2014. Our calculations suggest that savings of up to €15 billion could be needed, i.e., twice the sum that was under discussion at the time Ireland and the Commission agreed to the 2014 deadline.
  • We express a concern over the potential negative impact on the economy of this scale of adjustment over this period of time.
  • While the 2014 date strikes us as worryingly ambitious, we are mindful that an extension is highly unlikely and so we must operate within the constraints as presented. Although we have based our forecasts on a budgetary package of €4 billion of savings, it could well be that a higher amount will be sought. Whatever it is, the scale of the task is such that there will be a need for adjustments in current and capital spending and in taxation.

Don’t forget to read the articles in the Research Bulletin.

19 replies on “QEC Autumn 2010”

@Richard

Just over two weeks ago the Central Bank forecast GDP for 2010 to be +0.2%. What do you think has changed in the past couple of weeks to reverse that to -0.25%, do you think?

@Richard Tol

You quote from the QEC:

In the year ending April 2010, the CSO recorded net outward migration to have been 34,500. This was well below our forecast of 70,000.

A couple of weeks ago, on one of the other threads (link below), you wrote:

http://www.irisheconomy.ie/index.php/2010/10/08/downside-risks-and-up/

“There’s no egg on your face (referring to Kevin O’Brien). JtO is simply wrong here. Our annual forecasts are January-December.”

JTO again:

Will you now apologise and admit that I was totally correct about the inaccuracy of ESRI’s migration forecasts in the past couple of years. They ARE for the year to April. And they WERE wrong by 35,500. They are now admitting it.

@Richard Tol

You totally ignored my point about it being the year to April.

They are now effectively claiming that the CSO got it wrong, even though they have never got it wrong before. I wonder will the CSO be sueing ESRI. If I ran the CSO, I would.

“We discuss how this figure of 34,500 seems to be a conservative estimate of the rate of outflow when compared with estimates of migration contained in another CSO publication, namely, the Quarterly National Household Survey.”

In other words, they are blaming the referee for counting the scores wrong. But, the QNHS figures match very closely the population figures. Thus, the population figures measured in April shows a 9.9k fall in the population ages 15 plus. The QNHS for April-June shows an 11.4k fall in the population ages 15 plus.

I intend emailing the CSO to see what they say about this attack on their competence. However, I like to hear all the evidence, before I condem too harshly. Can you tell me where I can get hold of the full QEC, so that I can at least see what case ESRI are putting forward as to why they think the CSO got it wrong? It doesn’t seem to be available for the general public to read on the internet, just the summary that you gave above.

@Richard Tol

So, your answer to the statistical question of whether ESRI migration forecasts are for the year to April or the calendar year is “Q stands for quarterly”. Well, “B stands for b*ll*cks”, but it hardly answers the question.

However, no need. The answer is clear from ESRI themselves.

“In the YEAR ENDING APRIL 2010, the CSO recorded net outward migration to have been 34,500. This was well below our forecast of 70,000.”

Can you tell me where I can get the full QEC report? Is it on sale anywhere? I am quite happy to purchase it and add to ESRI coffers in so doing?

@JtO
Hold on a second. It is perfectly reasonable to think that there are multiple definitions in what constitutes being an emigrant, in the context of emigrating to the UK. Hence the disparities between various estimates

In the case of the US, Canada and Australia , it is presumably straight-forward to come up with a uniform definition of an emigrant (green card / work visa etc).
How do you make a uniform definition for the UK? not so easy to do.

@KO’B
It is an issue of measurement rather than definition. As Ireland does not have population register, we estimate the number of people (in between the census). The CSO produces two such estimates. JtO seems to think that one is accurate (atlhough the CSO itself would not make such a claim) while my colleagues think that the other is closer to the truth.

@Richard Tol

Can you tell me where I can purchase the full QEC?

Is it on the internet or for sale in a bookshop?

I would like to study why they say the CSO got it wrong?

The CSO estimates were deadly accurate in 2006 when the last census was carried out. No one has ever challenged CSO estimates before.

@JTO ths issue is not that the CSO got it wrong but that there are two separate sources of data from the CSO that do not seem to come up with the same answer. The QEC uses the QNHS data as the source for labour market data and hence it is natural to use the migration numbers implied by the QNHS, especially as the QNHS appears quarterly while the Population and Migration Estimates are annual.

By the way you are quite incorrect to say that the CSO Population and Migration Estimates are always totally correct – you should have a look at the revisions that were made after the last census (2006). These are estimates, so they are likely to be be revised.

Unlike the period of high immigration where it was possible to find the migrant households via surveys the problem now is that if a foreign national household leaves ireland there is no trace. If an irish national leaves there are likely to be relatives who as part of the QNHS will report on that emigrant. We will not know for certain what the level of emigration has been until we get the Census 2011 results.

Here is the passage on migation (note it is one paragraph out of 64 pages):

“On out-migration, the Population and Migration Estimates published in
September showed that there was a net outflow of 34,500 in the year
ending April 2010. This was well below our forecast of 70,000. However,
another source of CSO data on migration, (the QNHS), suggests that the
figure of 34,500 may be an under-estimate of the actual outflow in the year
ending April 2010 and that out forecast of 70,000 is likely to have been
more accurate than suggested by the Population and Migration Estimates.
According to the Population and Migration Estimates, the net outflow of
34,500 was made up of a net outflow of 14,400 Irish nationals and 20,100
non-nationals. However, if we compare the estimates for the number of
non-nationals living in Ireland in Q2 2009 and Q2 2010,13 we see that the
population is estimated to have fallen by 52,800. It should be noted that the figures in the Population and Migration Estimates refer to the full
population, whereas the QNHS counts people over the age of 15 years. It
is also the case that the time period covered differs slightly. Nevertheless,
the discrepancy between the two sources (20,100 versus 52,800) is certainly suggestive of a higher net outflow relative to the level reported in the Population and Migration Estimates.”

@Richard – I never claimed that it did. The Cenus will give us an accurate picture of the total population. Take that, the total population from the last census and births and deaths (from the Vital Statistics) and we can calculate total net-migration over the inter-censal period (the CSO actually publish this – Tables 1 and 2 in Volume 4 of Census 2006). That is the most accurate figure on net migration available. Those that arrived after 2006 and left before 2011 net out.

@Edgar Morgenroth

It never goes down well with juries to change your alibi when the first one has been rumbled. When I first started pointing the inaccuracy of the ESRI migration forecasts, I was told (not by you) that I was wrong about the dates. I was told (not by you) that the ESRI forecasts were for the calendar year, and not the year to April. This turned to be false, as is clear from the wording in the paragraph in the summary of yesterday’s QEC. It was only after this explanation broke down that ESRI came up with a new one, never mentioned before, to the effect that the CSO were under-estimating net emigration.

The CSO poulation and migration estimates were extremely accurate in the years leading up to the last census in 2006. These are the population estimates that they published on 14 September 2005, ie before census 2006 was carried out, for the years 2003, 2004 and 2005:

2003: 3,978.9
2004: 4,043.8
2005: 4,130.7

And these are their latest estimates (from their database) for those years, ie obviously after census 2006 was carried out:

2003: 3,979.9 (out by 1.0)
2004: 4,045.2 (out by 1.4)
2005: 4,133.8 (out by 3.1)

Very accurate indeed! What were ESRI’s population estimates for those years, made prior to the census of 2006?

The CSO estimate the population in April 2010 to be 4,470.7. What is ESRI’s estimate of the population in April 2010? If ESRI are still insisting that their net emigration forecasts for 2009 and 2010 are correct, then it must be 77.7k less, ie 4,393.0, ie back to its autumn 2007 level. But the report on the ’empties’ that was posted in the other thread yesterday, showing far fewer than was thought, and showing that around 77% per cent of houses built since 2007 are now occupied, would certainly indicate that the CSO estimate is more likely to be correct.

Regarding your statement ‘if a foreign national household leaves ireland there is no trace’, do you seriously suggest that the CSO don’t know this and therefore don’t know how to carry out a population enumeration accurately? That would seem to be the implication. But, the figures I gave above for 2003, 2004 and 2005 would indicate that the CSO are actually very good at population enumeration in between census years.

Regarding the QNHS, the paragraph you quoted from the QEC states that ‘in the QNHS the (foreign national) population is estimated to have fallen by 52,800 in the year’. But, in the same QNHS, the Irish national population is estimated to have risen by 41,400 in the same year, giving a net fall of 11,400, which is almost identical to the 9,900 given in Population and Migration Estimates. So, the figures for the total population are almost identical in both, although the distribution between Irish and foreign nationals is slightly different in Population and Migration Estimates as compared with the QNHS. But, ESRI can’t pick out one (foreign nationals), and not the other (Irish nationals), to support their case.

@JtO
The older CSO estimates are accurate because of post-census revisions.

As I have pointed out to you previously, the Quarterly works in quarters rather than in years.

@JTO – So you finally accept that the CSO produces estimates that are subject to revision – that took a while!!

‘if a foreign national household leaves ireland there is no trace……..’- as it happens this has been the topic of some discussion amongst those thinking about these numbers including the CSO. You should take a look at the methodology section on the CSO website and check out the QNHS questionaire and you will see that that is not going to find whole foreign national families that left as the question is whether someone living in the household previously has emigrated.

Finally, as I am neither involved in short-term forecasting or migration forecasting I think you should direct your queries to those that do.

Comments are closed.