CSO made look silly for sticking to the rules
Krugman missed the target with his remark, the leprechauns are in Luxembourg, not here, writes Colm McCarthy
The Irish economy, on all credible indicators, grew by about 5pc last year. The GDP growth rate could even have been a little higher. But it could not have been 26.3pc, the amazing figure computed by the Central Statistics Office.
The CSO explained that it followed faithfully the rules laid down by Eurostat, the statistical arm of the European Union and it appears that the CSO has indeed calculated the figures by the book. National statistical offices do not make these numbers up, it follows harmonised procedures laid down by international agencies. In Ireland's case the rules are contained in a manual called ESA 2010, binding on EU members and authored by Eurostat in Luxembourg.
Paul Krugman, the Nobel prize-winning economist, tweeted an instant crack on Tuesday about 'leprechaun economics', providing the world's journalists with the hook for their coverage of these extraordinary figures. The impression created is that the Irish state cannot organise a statistics office to count economic activity properly. Krugman missed the target with his instant smart remark. The leprechauns are in Luxembourg, not in Ireland.
The Eurostat rules require that the assets of numerous Irish and foreign corporations based in this country be counted as part of Ireland's capital stock, even if they are not deployed producing output in Ireland. These assets include intellectual property, such as patents and computer codes. The annual depreciation of the capital stock is added on to economic activity to calculate Ireland's GDP - that's why the term 'gross' appears in Gross Domestic Product, to distinguish it from net product.
With depreciation left out, net product rose only modestly in 2015, in line with what every sensible person felt was happening to the real-world economy. The peculiarities in the Eurostat rules have been known for years, and several Irish economists, including Seamus Coffey and John FitzGerald, have drawn attention repeatedly to the potential distortions, but the issue became really serious only recently.
There were large transfers of intangible assets on to the books of companies based here last year, driven by tax features, including provisions in the US corporate tax code and its interaction with Ireland's arrangements. This contributed to a huge, and utterly implausible, increase in the capital stock and hence in depreciation and GDP. This phantom increase in the capital stock does not reflect extra buildings and machinery actually brought into use in Ireland last year.
There are other oddities: companies based here which arrange for contract manufacturing abroad can see the corresponding output counted in Irish production and exports, even though the goods concerned never enter Ireland at all. Companies which redomicile their worldwide operations here through the takeover of Irish-based firms can also boost the apparent activity rate. These are the so-called tax inversions, since curtailed by unilateral action of tax authorities in the US but too late to avoid some big distortions in 2015.
A further weakness in the Eurostat rules concerns aircraft-leasing companies. They retain on their books in Ireland planes leased to airlines around the world which never fly into this country. Bizarrely, if the leasing companies were classified as banks (which in economic substance is what they really are) this anomaly would disappear. Under a revision to international accounting rules due to come into effect in 2019, leased assets will have to be located on the balance sheet of the client company (somewhere in Asia, quite likely) rather than on the balance sheet of the leasing company here in Ireland, which happens to have a large aircraft-leasing industry.
To illustrate just how strange the Eurostat rules are, a new airplane which leaves Boeing in Seattle for China, and spends the rest of its life flying around Asia, never visiting Ireland or even Europe, contributes disproportionately to Irish output as measured because it appears on the balance sheet of a Dublin-based company. After 2019 it will cease to do so, when new accounting rules kick in, so the 2015 distortion gets undone (and GDP will appear to fall).
The Irish CSO has been made to look silly because it has followed, it would seem diligently, Eurostat rules which yield daft results. It would have been better, and would have avoided some reputational damage which the country could do without, if the CSO had published the revised 2015 numbers it actually believed in, alongside the incredible Eurostat version, explaining why the latter should not be taken seriously. Instead it appears (courtesy Paul Krugman) that Ireland has an incompetent statistics office staffed by little green jokers, which is not the case at all.
The anomalies which yielded the weird figures on Tuesday have come to light only recently. In April, the CSO published estimates for the final quarter of 2015 which showed plausible numbers for last year. But the revised quarterly figures on Tuesday show that industrial output, for example, almost doubled in the single quarter from Q4 2014 to Q1 of 2015. This is poppycock and could not possibly have happened. It would appear that the CSO first heard about last year's phantom increase in the capital stock from companies reporting to them since April, and dutifully computed the massive depreciation charge and hence the absurd GDP growth rate, all in accordance with Eurostat's ordained methodology.
The CSO sought to repair the damage with a statement on Wednesday, noting that some key indicators such as consumer spending and employment had risen in 2015 at healthy but plausible rates. Consumption was up 4.5pc, retail sales 8.2pc and employment 2.6pc. These numbers point to an overall growth rate around the 5pc figure settled on by independent economic observers.
The CSO went on to announce its intention to establish a committee of experts to help with the interpretation and presentation of Irish economic statistics. This committee is not a priority. The statistics needed to get a handle on what is really happening are already available. The problem is the procedures laid down by Eurostat and the impossibility of producing plausible estimates for Ireland within these rules. The CSO should instead have convened an international committee of statisticians to recommend changes to ESA 2010 designed to ensure that EU countries can produce meaningful macroeconomic statistics.
If taken literally, the 26.3pc growth rate announced on Tuesday would have some laughable implications. On the one hand it makes the economy seem much larger and hence able to sustain a relaxation of budget rules, something ministers were quick to dismiss. But equally it implies that the economy has been overheating dramatically, pointing to a tightening of policy. The next budget can hardly ease and tighten simultaneously.
Another absurd implication of Tuesday's aberration could be an extra cost to the Exchequer. EU members pay into the community budget in line with a measure called Gross National Income, which is also inflated by the phantom depreciation provision on the imaginary capital stock. This could cost an additional €300m per annum, double the threatened revenue from water charges.
International statistical conventions should provide an intelligible framework for countries to produce economic statistics which reflect their true level of economic activity. Eurostat has failed to deliver to this simple standard.