7/2/17: There's Zero Growth in Irish Patentable Innovation & Research Outputs

In a recent post on Irish patents filings and applications with the EPO, I showed that:

1. Irish R&D and innovation performance - as reflected by patents data - is hardly impressive, with the country ranking 14th in the sample of 50 countries as an origin for EPO Applications;
2. There has been no material improvement in Irish standing in the data in recent years, compared to trends.

Some of the readers have taken me to task on the second point, despite the fact that my evidence (based on EPO data) shows no gains in Irish patenting activities with the EPO in terms of both applications and filings, and in comparative terms as a share of both in the total number of EPO applications and filings. 

So I took a different exercise, plotting a relationship between average levels of filings and applications (combined) across 2006-2008 period against the same for 2014-2015 data. 

Not surprisingly, Ireland comes smack in the middle of the distribution and right on the regression line, implying that:

1. Ireland's patenting performance is to the upper range of the overall distribution of 50 countries, but it is at the bottom of this sub-group of top performing countries. In fact, Ireland's position is statistically indistinguishable from 'mediocre' or 'average' group of countries. 
2. Ireland shows only tiny growth in applications between 2006-2008 period into 2014-2015 period (see Ireland's point position just slightly above 45 degree line), which is statistically indifferent from zero growth.
3. Once we control for the factors that drive global trend in patents (blue regression line), Ireland shows no statistically identifiable growth (Ireland's point is bang on the regression line).

Yes, patents are not the only measure of innovation and R&D, but, being the core part of STEM-focused research, they are the main measure of innovation and R&D, because patents data omits only one form of innovation - that linked to software. Now, software innovation is important, and Ireland may or may not be doing well in this sub-sector, but STEM research is based not on software innovations, but on 'hard' patents. And Ireland does not brand itself as 'Software-only Innovation Hub'. In fact, Ireland spends (as a State and economy) more on STEM innovation than on software innovation, so the key focus on Irish policies is, once again, measurable via patents.

Until we get 2016 data to update the above analysis, I rest this topic discussion.

5/2/17: European Patents and Ireland's 'Knowledge Economy' Myths

Irish policymakers are keen on promoting Ireland as a technology and R&D centre of excellence, often claiming the country is a ‘Knowledge Economy’, a ‘Data Island’, a ‘Europe’s Tech Capital’ and so on. While catchy, these tag lines are far from reality, and, in fact, represent an empirically dubious proposition. 

To establish this claim, consider the European Patent Office data on patent filings and approvals, with the latest data set covering the period of 2006-2015. 

As chart below clearly shows, Ireland is far from being a significant source of patent filing in Europe, despite the fact that many patents from Ireland are filed by the U.S. and other multinationals, including a score of foreign companies that choose to tax-invert into Ireland. The EPO data, in fact, fails to control for this distortion. Still, even with those companies filings counted as ‘Irish’ by origin, Ireland ranks 14th in a key metric of the rate of European Patent Applications per million of inhabitants. 

Worse, Irish rate of patent applications (119 per 1 million of inhabitants) is below the mean for the sub-sample of European states (including EU28 states and other countries within the EEA). Statistically, Irish rate of patent applications per inhabitant is not distinguishable from the rates filed by Italy and Slovenia, and is well below the rate recorded for France, Belgium, Austria, Germany, Denmark, Finland, Sweden and the Netherlands. Irish numbers are also statistically indistinguishable from the global average - the average that includes non-European states’ filings. It is worth noting that the data set includes other countries, that similar to Ireland serve as major tax optimnization locations for R&D and IP, e.g. Luxembourg and the Netherlands. However, even controlling for these states, Irish data does not shine beyond being average.

In absolute terms, Irish patent filings and applications with the EPO have trended up in 2006-2010 period, but have since flat-lined (on trend) for filings and declined (on trend) for applications. 

In addition to the chart above, combined filings and applications for EPO patents by Irish-origination stood at 1,325 in 2015, down on 1,364 in 2014 and down on 1,356 average for 2008-2011 period. While data can be interpreted in a number of ways, there is clearly no indication of an improving trend in either filings or applications over the recent years. This comes on foot of aggressive acceleration in tax inversions into Ireland in previous years - an acceleration that brought into Ireland a range of large R&D-intensive companies.

Consistent with the above, Ireland’s share of all European Patent Office filings and applications has declined, on trend, in recent years, as evidenced by data presented in the chart below:

As the chart above shows, Ireland accounts of just 0.266% of all EPO patent filings and 0.364% of all EPO patent applications. Separately reported data for patents approvals shows Irish share of all patents granted by EPO to be just 0.395%. The number is laughably negligible by Europe-wide standards and is massively out of line with Irish share of European GDP. However, these numbers are consistent with the simple fact, also highlighted by EPO data, that Ireland fails to register in the top tier of generators of patents in any of the sectors tracked by EPO. 

In summary, Ireland is far from being a powerhouse for R&D and knowledge economy activities as measured by a key research output measured by European authorities. 

5/1/16: What Aggregate R&D Spends Tell Us? Actually... little

In a recent comment on R&D Expenditure across the OECD countries, WEF has referenced Irish data on R&D spending as % of GDP at 1.58% which refers to 2012 full year results.

Which is surprising, given that we now have 2014 data available per Eurostat (http://ec.europa.eu/eurostat/documents/2995521/7092226/9-30112015-AP-EN.pdf/29eeaa3d-29c8-496d-9302-77056be6d586) which puts our R&D spending at 1.55% of GDP in 2014.

Irish GDP in 2014 in current prices terms was 16.07% above Irish GNP. The same gap in 2004 was 17.26%. Which means that adjusting for this gap, Irish R&D expenditure as a share of GNP was 1.38% of GNP in 2004, rising to 1.80% in 2014.

Thus, in 2004, Ireland ranked as 12th country in the EU in terms of R&D expenditure ‘intensity’ by GDP metric, and 11th by GNP metric, both metrics were at exactly the same ranking places in 2014.

Here is a chart showing longer evolution of the R&D expenditure series from OECD:

Overall, Irish R&D expenditures remain below the desired levels in absolute terms, both relative to the GDP and the GNP bases.

Eurostar provides a handy breakdown of R&D spending by origin across Private sector, Government sector, Higher education and non-Profit.

Few things stand out for Ireland:

• As a share of R&D spending, business enterprise sector appears to be carrying its weight in Ireland. 
• Government expenditure on R&D is extremely weak in Ireland, though one has to wonder what on earth can Irish Government research, given the quality of our state institutions.
• Higher education sector R&D spending in Ireland is ranked 20th in the EU - a ranking that is heavily influenced by a massive share of business enterprise spending of total R&D expenditure. 
• Apparently, there is no private non-profit spending whatsoever in Ireland.

Key to the above is, however, the nature of business enterprise spending. Per Government own statistics, in 2012, roughly 300 firms accounted for almost 70% of total R&D expenditure in Ireland. Just 107foreign firms spent more than EUR2 million on R&D per annum in Ireland and these account for 88% of the total R&D spent by MNCs in Ireland, or well over 70% of the total business enterprise R&D spend.

Here’s Finfacts take on the hype: http://www.finfacts.ie/irishfinancenews/article_1028789.shtml.

In other words, stripping out MNCs with their R&D activity booked through Ireland mostly reflective of tax optimisation rather than actual research, one wonders just how much exactly does R&D contribute to our GDP or GNP and just how much of the failures of Irish R&D spending are down to quantum of spend as opposed to quality of spend? Problem is: we do not know. All Government research on the matter, including research by the likes of the OECD (based on Irish Government-supplied data), is probably heavily biased by the insiders dominating analysis.

Take the following two charts from OECD latest report on Science and R&D (http://www.oecd-ilibrary.org/science-and-technology/main-science-and-technology-indicators_2304277x)

So in the first chart, Ireland is above EU and OECD averages in terms of researchers employment intensity, but in second chart, Ireland is below EU and OECD averages in terms of R&D output intensity (by one metric).

Which begs a question - is this difference down to quality of researchers or down to type of research (e,g. non-patentable fields of sciences and humanities) or down to classification by, say MNCs, of some business & admin personnel as research personnel for tax purposes and to create a smokescreen of ‘organic’ as opposed to tax channeling activity in Ireland?

Who knows… But in 2011, per OECD data, 71.1% of total R&D expenditure by enterprises in Ireland accrued to foreign affiliates (the MNCs).  Subsequently, we stopped reporting such data. It is worth noting that this does not include companies that redomiciled into Ireland via tax inversions, adding which to the pile would probably shift this number closer to 90 percent.

In simple terms, aggregate spending figures tell us very little as to the nature of Irish R&D activities or their effectiveness. The real data is being hidden from our view by commercial secrecy that conveniently obscures just what exactly is happening in the economy and in our research sectors. May be, the knowledge economy of Ireland is a de facto a convenient deus ex machina for the severe skews in the economy arising from the MNCs presence here. Or may be, it is all just fine and a crop of Nobel Prizes and research accolades for the country are only a matter of few more quid pushed into R&D line of private and public expenditure.

6/8/15: Irish Services Activity Index: June's Belated Sell-in-May

In previous post (link here) I covered Services PMI for Ireland for July.

To remind you: we are witnessing a massive boom (according to the PMI data) in Services, with overall sector activity readings at 108 and 109 months highs in June-July. In addition, based on quarterly averages, Services in Ireland should have been expanding at a break-neck speed non-stop from 2Q 2014 through 2Q 2015, with 2Q 2015 marking small acceleration in an already formidable speed on 1Q 2015. Effectively, over the last 3 quarters, PMIs have been signalling very high rate of growth in activity, with rate of growth being relatively stable over time.

Now, let's take a look at the latest quarterly data from CSO covering actual activity in the Services sector through June 2015.

Overall Services Sector activity index for 2Q 2015 rose 2.3% y/y, which is markedly down on 9.6% y/y growth recorded in 1Q 2015 and marks the slowest speed of Services sector expansion since 1Q 2014. This simply does not correspond to the PMI data readings. In fact, growth has been quite volatile over the last 5 quarters, and again, not consistent with the PMI signals.

As chart above indicates, Services sector growth fell sharply in 2Q 2015 falling below the period average (from 2Q 2014 on) and below the upper limit of statistical significance relative to the historical average rate. Contrary to the PMI signals, three out of six last quarters posted growth within historical averages and well below the period average when PMIs were hitting record highs.


Looking at the key sub-components of the index:

Domestic services sectors (Wholesale & Retail Trade, etc, Transportation & Storage, and Accommodation and Food, along with Administrative & Support services) posted an average rate of growth of 5.3% y/y in 2Q 2015, slower than both 4Q 2014 and 1Q 2015. Still, 2Q 2015 growth was the third fastest in 8 quarters. Over the last 6 months, domestic services managed to average expansion of 7.14% which is a major uptick on previous 6 months period when domestic services sub-sectors grew on average 5.40%.

Information and Communication services index posted a decline of 11.4% y/y in 2Q 2015, the first drop in the series since 4Q 2011 and the sharpest drop in the series on record. The sector is so skewed by activities of MNCs that not much can be determined out of these figures. Still, this drop brought past 6 months growth down to -1.8% against previous 6 months' growth of 6.5%.

In contrast to ICT sector, Professional, Scientific & Technical services sector posted a rise of 6.1% y/y in 2Q 2015, confirming yet again that there seems to be no serious correlation between activity in one side of our 'smart economy' and the other side of the same, despite endless droning on from our politicians and trade bodies about an alleged fabled link between the two sub-sectors through R&D and innovation.

It is worth noting that the sub-sector of Professional, Scientific & Technical services has been effectively whipped out by the crisis: over 2009, index of sub-sector activity averaged 118.88. This fell to 87.93 for the last four quarters - a decline of 26%. In a sense, our Professional, Scientific and Technical services 'did Greece', confirming yet again the deeply engrained culture of innovation and research in Irish economy. Of course, over the same period of time, Information and Communication services activity rose 35.1%. Go figure…

Despite all the issues highlighted above, the good news - as shown in the last chart - is that all three broadly-defined Services sectors have so-far been on a converging path prior to 2Q 2015 since roughly 1Q 2014 - as signalled by the compression of the period average lines. This, of course, reflects the belated return to growth in Professional, Scientific & Technical Services from 3Q 2014 on.