Monday 17 December 2018

Brendan Keenan: Fantasy figures must go if we are to avoid policy errors

currency appreciation illustration - Irish economy - 3D render
currency appreciation illustration - Irish economy - 3D render
Brendan Keenan

Brendan Keenan

It's not much use saying the Emperor has no clothes, if you are then going to hold up his non-existent gilded coat for our admiration. Yet that is the situation in which our statisticians and economic commentators find themselves.

The imaginary coat is, of course, GDP (gross domestic product) - but most of the undergarments are pretty illusory as well. With hardly anything real on display, they are reduced to explaining why what is being published is really not much use.

The last few weeks have been particularly troublesome, with the publication by the CSO of the national accounts for 2017, followed last week by the much-watched quarterly commentary from the ESRI. Earlier, the Central Bank faced the task of explaining why its forecasts could not be taken at face value - something we may presume the bank hates having to do.

The reasons are well enough known - the distortions caused by the activities of the large multinational corporations in Ireland. Which means it is much more than a problem of statistical aberrations. The role of the selfsame corporations and the tax they pay is now second only to Brexit as a pressing political and economic question.

The simple souls who report on such matters in the media have also been drawn into the morass. The CSO figures showed GDP growing by almost 8pc last year - even faster when comparing the last quarter with the same period in 2016. Reporters duly wrote and broadcast the figure, and then explained that it was not to be taken at face value.

For their own professional reasons, they hate doing that as much as central bankers. The ephemera of mass media does not take kindly to subsidiary clauses and footnotes. Almost certainly, a greater than ever number of readers, viewers, listeners and scrollers were left with little idea of what was actually going on.

They probably came away with a general impression that the economy was doing quite nicely, thank you, but proper political debate requires a bit more precision and precision is hard to come by.

The CSO is encouraging the use of its 'Modified Domestic Demand' which excludes obvious multinational distortions such as aircraft leasing and 'imports' of R&D.

On that basis the economy grew by 4pc last year - about half the GDP number and less than the 5pc expansion in national product (GNP) which we used to use quite happily but is also now hugely distorted. Simon Barry at Ulster Bank Economics reckons a realistic figure for last year would be even lower than 4pc.

Like other private analysts, he tries to strip out more difficult multinational activities. These include intangible investment, particularly imports of software, which increased by around €3bn last year. That represents a quarter of the growth in the CSO's measure. The bank suggests true growth in the domestic economy may have been between 2.6pc and 3.7pc.

And even that could be worse than it looks. When output is segregated, Barry reckons the economy is operating in 'two-speed mode,' with the multinational dominated sectors registering growth of more than 10pc last year, while the more domestic "indigenous" sectors expanded by less than 3pc.

That is a completely different world from the one portrayed by GDP. It is still a volatile one, with what we might call ordinary decent investment rising and falling rapidly because of major projects, such as the completion of the Dublin cross-city tram line.

In addition, the indigenous economy is boosted by the rapid expansion of construction - welcome, but not part of the tradeable economy. All this is exacerbated by the unpredictable spending on intangible assets by multinationals and aircraft purchases by leasing firms.

It is very difficult to bring these kinds of complication centre stage. The Budget shenanigans are already under way and will be unusually intense, given the prospect of a post-budget election and the potential spending pot of some €3bn. Around which of those two worlds - that of GDP or of genuine domestic activity - will debate take place?

The different measures do not particularly affect the 2019 budget arithmetic - although one advantage of the more realistic measures is that the forecast government surplus of €2bn next year represents almost 1pc of national income, versus 0.6pc of GDP. The issue, though, is not next year, but the longer term.

A debate about spending and taxation based on GDP will lead us further down a blind alley - and that is before considering the growing risks to the corporation tax revenues on which we depend so much.

Even if - and it is a big if - there are politicians and interest groups willing to use the more credible figures, they would have to struggle with the fact that the figures are not easy to find.

In its commentary, the ESRI called for a comprehensive overhaul of the national economic statistics. It suggested parallel figures, excluding the more esoteric multinational activities, could be published alongside the conventional ones.

Their frustration is understandable. Most economists, including those at the central bank, prefer the new national income statistics, GNI* (modified Gross National Income). The last available figure for that is 2016, when it stood at €190bn. GDP was put at a mighty €276bn; clear proof that it is a weapon of mass information destruction which should be decommissioned as soon as possible. 2016 is too long ago for analysts and market researchers. They also want quarterly figures which adjust for seasonal variations. These are not yet available for GNI*. The ESRI's summary forecasts have to make do with GDP and GNP, and we are left with little more than the 2pc rise in employment next year as a guide to the economy's performance.

The CSO says in reply that it is working hard on the new measures but points out that they are already in use in some places, most notably in assessing national debt. Quarterly figures and seasonal adjustments do not matter much here and the IMF, EU Commission and credit ratings agencies are all applying GNI* estimates rather than GDP.

Unlike budget surpluses, this is a disadvantage when it comes to the optics. Irish public debt is second only to Italy's in the EU when GNI* is used. When adjustments for the cost of living are made to get debt per person - as the OECD does in its most recent report - Ireland is marginally worse than Italy and bottom of the EU table.

In such a situation and in view of the clearly identifiable risks - Brexit, threats to corporation tax from both the EU and US, general global uncertainty - a surplus (at last!) of 1pc of national income looks hardly enough. Yet the only people it might suit to use even this figure would be opposition politicians attacking the government for worryingly loose fiscal policies.

No such politicians are likely to make an appearance. It may therefore be time, at least until the CSO completes its work, for even official bodies like them or the central bank, to put the modified, more realistic estimates, not parallel as the ESRI suggested but at the top of their statements, with GDP and GNP further down. As the old sub-editors' adage has it: who reads the second paragraph?

Indo Business

Business Newsletter

Read the leading stories from the world of Business.

Also in Business