ESRI QEC; Understanding GDP / GNI(GNP) data

The new ESRI QEC is out –  executive summary here.

John Fitzgerald has a note on the impact of re-domiciled firms on Irish national accounts data here.  This relates to the phenomenon of international firms locating the headquarters office here for legal/tax reasons, even if all of its activities are elsewhere. A feature of balance of payments accounting is that the retained earnings of these firms are counted as part of Ireland’s gross national income (GNI)  [GNI used to be termed GNP until it was accepted that income was a better term than product, since the concept does not relate to domestic production].

John’s article shows that these retained earnings have been very large in recent years, so that the level of GNI has been boosted (with a large impact on the current account). This is why it it is important to also keep an eye on the international investment position (IIP), since the liabilities to the foreign investors that own these redomiciled firms rise in line with retained earnings and this will be reflected in the IIP.

If the pool of retained earnings is depleted by the payment of dividends to the shareholders, then GNI will decline –  so the initial boost to GNI is eventually reversed.

Also, some of these redomiciled firms are now shifting HQ back to the UK, which will also undo the effect.

So, the Irish national accounts now features two unusual elements:

(a) the large-scale operations of the affiliates of foreign multinationals mean that there is a large gap between GDP and GNI due to the high recorded profits of these firms. Moreover, the high import content of the exports of these firms means that there are analytical issues in understanding the dynamics of valued added in Ireland. To the extent that transfer pricing means that true imports are understated as a means to boost recorded profits, it also means that the trade surplus is overstated (but one-for-one net factor income is understated, so the current account is unchanged)

(b) the more recent feature is the impact of redomiciled firms which are counted as Irish firms, since the headquarters are here and even though these have virtually zero domestic activities and the ownership is entirely foreign.  As shown in John’s note, this sharply alters the interpretation of the GNI data –  but has no impact on the GDP data (which relates to domestic production).  It also sharply alters the interpretation of the current account data, which is a key variable in the European Commission’s “macroeconomic imbalances” scorecard.

So – both GDP and GNI data need to be handled with care. In particular, the traditional short cut of interpreting GNI as a better measure of true domestic activity is not wise  –   it is an income measure, not an activity measure.

It is important to emphasise that these are interpretation issues rather than measurement issues (the CSO implements the globally-agreed rules).

8 replies on “ESRI QEC; Understanding GDP / GNI(GNP) data”

Good work by John FitzGerald and I argue that the 2012 GDP data are also unreliable. However, as during the bubble, sailing against the winds of consensus when it’s seen necessary to endorse questionable data, is not easy.

The QEC says: “The volume of service exports exceeded that of goods in 2012. With continued flows of service sector FDI into Ireland this extra capacity is expected to result in continued strong growth in service exports, with volume growth of 5.7% forecast for 2013 and 8.0% in 2014.”

Is this true or a delusion?

However, computer services or the two business services categories in 2012 services exports, did not jump by 15% and 13% because of FDI inflows. The facts show that these surges are tax-related.

David Duffy and Kevin Timoney should present the evidence that FDI inflows were responsible for the surge in services exports not tax-related revenue diversions to Ireland. Google’s revenues now account for 41% of Irish computer services exports.

The renowned American baseball player Yogi Berra (b. 1925) once quipped “It’s déjà vu all over again.” It’s true!!! 

It would sure be all grand if we could monetise fairytales.

I sent an email to David Duffy on Tuesday and copied it to Prof Frances Ruane, ESRI director. I did not get a reply.


I will not be able to attend tomorrow’s press briefing as I am based in Kuala Lumpur.

However I hope that there will be more honesty than what’s forthcoming from elsewhere, about the surge in tax-related services exports resulting from revenue diversions by US companies such as Google, Apple, Microsoft and Facebook.

Official claims that the result of Google diverting almost 50% of its revenues to Ireland reflect competitiveness are lies and against the public interest.

Google’s global web revenues grew by 29% and 21% in 2011 and 2012 respectively. When the Google Ireland’s 2012 accounts will be published this year, they will show revenues of at least €15 billion or 41% of Irish computer services in 2012.

Google, Microsoft, Apple, and commercial aviation leasing, employing about 6,500, account for 52% of Irish services exports.

A surge in headline exports but full-time employment in exporting sectors lower than in 2000 is the reality on the ground.

I see that John FitzGerald has determined that the GNP decline has been understated in recent years. In 2002, it was only possible to report GDP growth because of tax-related accounting transactions at US multinationals..”

@ Michael Hennigan

What are the implications if what you say is the case and a large part of Ireland’s services export growth is attributable to the activities of Google and a few other large firms? Is it that you feel this is fundamentally unstable activity which may pull out of the country in a few years and hence damage economic growth and the exchequer balance?

@ Carson

The biggest issue is that there is an illusion of progress when there is none.

That fantasy is important for the people responsible when there should be a greater public recognition of the gravity of the jobs challenge.

Richard Bruton can move from one jobs announcement to another hyping success and of course he will always have the spiel for the journalists.

Tom Healy of the Nevin Institute proposes a stimulus program but that could only make sense if there was the prospect of a significant recovery in a few years but there is going to be no opportunity to see Noonan’s rocket in orbit.

Google’s global web revenues grew by 29% and 21% in 2011 and 2012 respectively and in 2011, 45% of the total was booked in Ireland. Google Ireland reported revenues of €12.4 billion in 2011; payroll costs were €218 million; corporate tax in Ireland on trading activities was €3 million; total tax charged at €22.2 million including foreign withholding tax. When the Google Ireland’s 2012 accounts will be published this year, they will show revenues of at least €15 billion or 41% of Irish computer services in 2012.

Google, Microsoft, Apple, and commercial aviation leasing, employing about 6,500, account for 52% of Irish services exports.

Up to half of the services exports total of €90 billion in 2012 may be effectively fake i.e. unrelated to Irish economic activity.

Google began at zero in Ireland in 2003, just as Apple was starting its rapid ascent – – just two firms’ bookings in Ireland would dwarf the total exports of the whole indigenous sector.

There is little of the big sums left in Ireland.

It is very like a rerun of pre 2008 with the same fake explanations coupled with timidity from the mainstream media, including The Irish Times.

In the Forfas Annual Employment Survey 2011: the following are the reported employment levels;

Information, Communication and Computer Services –  – Full Time Employment (foreign and indigenous) Page 18

2007: 65,064
2011: 65,642

So no increase in head count in period 2007-2011.

‘Computer services’ exports were valued at €19bn in 2007 and €36.5bn in 2012 — up 92% and 15% in 2012

It’s similar for 2 categories of business services.

@ MH-ff: – or any other sentient being:-

This is the bog standard, ECON 101, expression for aggregate economic activity:-

Y = C + G + I +Nx.

Am I correct in believing that Nx is now a completely bogey value? That it is actually (really) smaller? If so then (assuming the values of C, G and I remain the same) – Y has to be lower. Yes?

Now we appear to have some intellectually challenged folk close by who keep encouraging us to – “Stick with the ‘consolidations’ – and we will ‘grow out of it!” – whatever ”it’ means. This is despite the published empirical evidence refuting this mantra (eg: that we have to shrink G to ‘grow’), which axiomatically means C must also shrink somewhat, whatever about I.

A secondary, but by no means an insignificant issue, is how to account for both existing debt (private, corporate and sovereign), and the annual change in these debts, in the above expression. Debt is a charge on future income. The more one has, the greater the deduction on your future income – when that income materializes.

What about I? Apparently this includes debt paydown – which is the destruction of part of the existing stock of money. This axiomatically leaves less money in the economy? This will have NO impact on the value of C then?

Did I hear this morning that Y would ‘grow’ both this year and next? Are these intellectually challenged folk, seriously challenged in their math departments as well?

The overstating occurs on both side of the net exports figure. Computer services exports are a facade but their are large service imports on the other side.

Irish service exports are massively overstated but the balance of services in 2012 was less than €3 billion. If the false exports disappeared so too would a lot of false imports.

Computer service exports were €36.5 billion in 2012. Counter to this royalties/license imports were €32 billion. The pharmaceuticals will account for a chunk of that but miscellaneous business services imports including inter-affiliate management charges were a further €23 billion.

@ Michael Hennigan

This is a quote taken from Davy’s economic commentary this morning which is relevant I think. I’m not sure which CSO release this data was taken from.

“Yesterday, official Central Statistics Office data on the artificial positive impact of re-domiciled PLCs on Ireland’s current account balance was published for the first time. The data show that incoming net profits from re-domiciled PLCs have grown exponentially since 2009, up from €1.5bn to €7.4bn in 2012. The underlying current account surplus, excluding re-domiciled PLCs, was 0.4% of nominal GDP, well below the headline 5% measure.”

@ Seamus Coffey/ Carson

Conall Mac Coille’s counterpart at Goodbody, Dermot O’Leary, says today;

“It is widely accepted that Ireland is enjoying an export-led recovery brought about by an improvement in competitiveness…Services exports have been taking up the slack over recent quarters, with growth of 8% yoy in Q4 2012. The Irish Exporters Association has stated that services exports grew by 8% yoy in Q1.”

Widely accepted? Maybe but not true.

Until 2012, the services giants and also presumably Apple, left very little in Ireland to tax e.g. Microsoft Operations Ireland in its fiscal 2011 accounts showed an 18% rise in revenues to €13.37bn but reported profits fell 58% to €593m as cost of sales grew from €770m to €1.15bn, while administrative expenses jumped from €9.13bn to €11.61bn. Net income before tax fell to 4.4%.

At a global level, the situation was calm and unchanged: In 2011, Microsoft Inc. reported net income before tax of $28.1bn on revenues of $69.9bn – – a ratio of 40%, unchanged from 2010.

The services revenue shifting differs from the experience of profit shifting in respect of goods manufacturing.

So when a forecaster predicts a jump in services exports in 2014 and 2015, with the exception of 2012, there is no record to assume that the headline surge will result in a net exports benefit for GDP.

Comments are closed.