How efficient is public spending on R&D

In the context of current discussions about the government spending on R&D and more broadly on science, technology and innovation the key question should be how to use the public resources in the most efficient way. While the fact that markets do not provide sufficient incentives for individuals and firms to engage in socially optimal R&D and innovation is well established in theory and documented by existing empirical evidence (due to positive externalities private returns to R&D investment are lower than social returns), in some cases government action may not provide effectively appropriate incentives at sufficiently low administrative and compliance cost, and without creating further distortions (crowding out).

Given scarce public funds, assessing the efficiency and effectiveness of public spending on R&D and their determinants should be given a high attention in Ireland.

A recent study commissioned by the European Commission (“Efficiency of public spending in support of R&D activities”, by Michele Cincera, Dirk Czarnitzki, and Susanne Thorwarth, Economic Papers 376/2009, European Economy Series, European Commission) analyses the efficiency of public R&D expenditure with respect to fostering business R&D spending at the macroeconomic level in OECD countries over the past two decades (1980-2004). Two methods are used – the Stochastic Frontier Analysis (SFA) and Data Envelopment Analysis (DEA) – to relate business enterprise R&D expenditure (BERD) and alternatively R&D personnel employed in the business sector to three public R&D funding inputs: government financed R&D in the business sector (BERDBYGOV), expenditures on higher education R&D (HERD) and R&D conducted in public labs (GOVERD). Both methods estimate efficiency frontier and efficiency losses.

The efficiency performance of public R&D varies across countries. In the case of business R&D expenditure, the countries ranked as the most efficient by both methods include (in alphabetical order): Australia, Canada, Finland, Germany, Japan, Netherlands, New Zealand, Singapore, Switzerland, and the United States. Using the DEA method the following countries are ranked as most efficient in the case of both outcomes (business R&D expenditure and R&D personnel in the business sector): Australia, Canada, Iceland, Ireland, Japan, New Zealand, Portugal, Singapore, Switzerland and the United States. The following countries are found highly efficient irrespectively of model specification or method used: Finland, Germany, the Netherlands, United States, Japan and Switzerland. Japan, Switzerland and the United States are ranked in the first position in each of the five periods.

The differences in efficiency performance have been explained by framework conditions such as the nature of competition, the quality of the business environment, the IPR regime, access to markets and external financial conditions, institutional factors influencing the effectiveness of public R&D with respect to fostering R&D activities in the business sector. In particular the strength of the IPR system and the legal structure and security of property rights appear to have affected positively efficiency levels.

The paper can be downloaded from here:

13 replies on “How efficient is public spending on R&D”

This is taken from a comment I made on Ferdinand Von Prondzynski’s blog re the subject of the role of R&D. I hope it’s not tangential..

One has to distinguish between blue-sky research (MediaLab, Bell Labs etc..) and the ustilisation of technology in the evolution of business processes that will be more productive. Fetishing technology as the ’source’ of prosperity is one of the fallacies at the heart of Mao’s Great Leap Forward and our government’s talk of a ‘knowledge economy’. The innovative manager is looking for a cheaper and better way of doing things – period. Technology will support this. What stands in the way in Ireland is not lack of ‘expertise’ or lack of ‘technology’ (though if these are necessary it is only at the behest of the manager responsible). What stands in the way is the feudalisation of Irish public and (some) private organisations around rigid management structures where status and not achievement is the main output and refusal to adapt and not organisational innovation is the main strategy – usually driven passive-aggressively by professional bodies and trades unions. Take, for example, IT in hospitals. It should be very cheap to design a web-based, secure database of patient records universally available to all health practitioners. By ‘cheap’ I mean a couple of hundred thousand euro or even less. You could probably build it using open source tools for about 50k. Now then take the article in the Irish Constitution that says that any such attempt to innovate (I don’t need to spell out why this would be a huge innovation – a true great leap forward) would require that a tender be given to Accenture and a ‘consulting’ contract be given to PWC and that numerous committees of ’stakeholders’(i.e. all those who have nothing to gain if it works and everything to keep by pretending to be engaged in ‘consultative process’ whilst ensuring it doesn’t) spend years consuming the legitimacy of such a project to inflate their own importance (’major data protection concerns’ anyone?)….resulting in, oh I dunno, billions wasted for no product.

The solution is not more ‘expertise’ or ‘innovation’. The technology and expertise to revolutionise Irish public and private sector productivity is cheap and easily available. What’s lacking is the management will and power. What’s in the way, at every level of Irish life, is the politicisation of management decision making by ’social partnership’ and its baleful gravitational influence on the management culture of an entire generation of Irish managers. For example Brendan Drumm is paid more than Barak Obama – an enoumous salary – to run the HSE. With all his status he doesn’t even have the authority to fire anyone. If you can’t fire people you shouldn’t be doing such a job and, in fact, it is dishonest to remain within it. There is no danger of innovation here. More PhDs will not help.

Constantin Gurdgiev wrote about this in the Irish Times recently.

His argument was that the ‘knowledge economy’ is a distraction from the failure and continued unwillingness to take uncomfortable decisions, many of which would necessitate confronting self-serving ‘experts’ in the professions for whom productivity increases from innovation are a mortal threat.

Gurdgiev’s pointing to the impact of technology in retail is apposite. It’s a non-glamorous, mundane phenomena yet that is where it has worked.

At the same time, even, Gurdgiev unintentionally flatters those who think that technology is the only key. Wall Mart’s use of technology is better than our public sector because its organisation is adaptive. Historically the organisation is prepared to disrupt itself before it gets disrupted by its market.

Given where Ireland is at this focus on technology is a distraction from the fact that the entire public service is massively overpaid compared to the rest of Europe.

After a quick overview of the paper, I think a couple of points relevant to Ireland’s rating are appropriate.

1) The research covers the period 1980 to 2004. Substantial public expenditure on research only emerged in Ireland right at the end of this period. It can only have had a modest impact on input measures, and barely any impact on output measures.

2) A large part of Ireland’s business R&D activity within the period covered reflects software development and other IT development activity. With some notable exceptions, this activity had weak-to-non-existent links with publicly funded research. It is therefore not very meaningful that Ireland shows up under the DEA measure as having quite a large volume of business R&D activity relative to the volume of Government, HEd and publicly-funded business activity.

Sorry that last setence should read “It is therefore not very meaningful that Ireland shows up under the DEA measure as having quite a large volume of business R&D activity relative to the volume of Government, HEd and publicly-funded business *research* activity.”


In your first paragraph you mention the idea that “markets do not provide sufficient incentives for individuals and firms to engage in socially optimal R&D and innovation”, and that this is well established in theory and documented by existing empirical evidence. I am familiar with the theory – Paul Romer on “Endogenous Technological Change”, and “Too Much of a Good Thing? The Economics of Investment in R&D” by Jones and Williams.

From what I can gather, particularly from Jones and Williams, there is under investment in R&D. ( Are you referring to the calibration-work in Jones and Williams as the existing empirical evidence? If not, I would be keen to get an additional reference(s).

(As an aside, there is an article by Obama advisor Austan Goolsbee, from a few years back, which challenges the conventional wisdom that “the social rate of return to R&D significantly exceeds the private rate of return”, and he argues against the idea that “R&D should be subsidized.”



Thank you for your comment. Here are a few relevant references related to my point:

On the theory of endogenous growth

Aghion and Howitt (1992) is also relevant: “A model of growth through creative destruction”, Econometrica, vol. 60, 323-51.

Empirical evidence on R&D spillovers:

Griliches, Z. (1992) “The search for R&D spillovers”, Scandinavian Journal of Economics, vol. 94., 29-47

Jones and Williams (1998) “Measuring the social rate of return to R&D”, Quarterly Journal of Economics, vol. 113, 119-35

Griffith, R., Redding, S. and van Reenen, J. (2004) “Mapping the Two Faces of R&D: Productivity Growth in a Panel of OECD Industries”, The Review of Economics and Statistics 86(4), 883-895

This latter paper provides empirical evidence showing that R&D expenditure plays a double role:as an input into innovation and in assimilating the research discoveries of others. An interesting conclusion of this paper is that to the extent that in non-frontier countries R&D has the potential to generate TFP growth from both innovation and technology transfer, many existing studies based on US data (the frontier country) are likely to underestimatrate the world social rate of return to R&D.

I hope that these are helpful.


Maybe an additional factor influencing the efficiency of public R+D in stimulating of private R+D is the role of business and entrepreneurship education. Public and university R+D researchers might not know how to start up their own companies or how to engage the private sector in projects or understand its work culture. Thus basic research is not commercialized or transferred into applied research in the private sector. The possible reasons for the lack of understanding of business by public R+D researchers could be the absence of business and entrepreneurial education for Science/engineering/technology researchers and the absence of networking links between business and science faculties and their respective students and researchers.

@ Iulia,

I want to get back to reading your post above properly again, but my feeling on this matter, from my extensive research on ‘team work’ and so on, in terms of how you foster R&D, is that less is sometimes more. A lot of the breakthrough innovations happening by accident almost, often on very meagre budgets. Unix being a prime example. It was the result of a much larger, failed software project inside of the old AT&T. Basically, a couple of the guys who were disappointed by the failure of the project, decided they would do something ‘on the side’ and take the exact opposite philosophical approach to the AT&T main project.

Linux was the same, it was designed around the smallest chip, when everyone else was designing for the largest chips.

(I am not an academic or published anywhere, you will have to take my word, that I have an interest in this whole area)

But my instinct is that small amounts, of ‘pure’ R&D money injected into the bloodstream of Irish innovation, at the right times, are much better than the blanket approach of throwing money at it.


My key point is exactly that we need a change in the policy focus from how much to how efficient is public R&D expenditure. To this purpose, proper evaluations of R&D programmes should be part of the policy making process. If done with the state-of-the art methods these evaluations are extremely helpful not only to uncover inefficiencies but also successes.

I must thank Iula for her posts but I must say the original reports referred to in this post and the post of Sept 8 entitled – How efficient is public spending on R&D – were very heavy going with the EU report on efficiency particularly so. It is almost a study in how not to communicate.

I have for some time been looking at the research, development and innovation (RD&I) policies in Finland and I was pleased to see that the OECD report highlighted the role of TEKES in the Finnish response to their crisis of the early 1990’s. I have been tending for some time towards the conclusion that we, in Ireland, should radically restructure our approach to RD&I policy and reconfigure it on the Finnish model. Science Foundation Ireland (SFI), in particular, seems to have been “captured” by the “big scientists” who have a very distorted view of the relative importance of research in the basic sciences for economic development in the short and MEDIUM term at least. I put forward the proposition that an exhaustive examination of the information on the websites of TEKES; the Academy of Finland; and Statistics Finland should be compulsory reading for anyone interested in RD&I policy for Ireland. I set out just a few ideas on the Finnish system in the following paragraphs.

The two main conduits for government support for RD&I in Finland are the Academy of Finland and Tekes. The Academy of Finland is the prime funding agency for basic research. Tekes is the Funding Agency for Technology and Innovation. It supports wide-ranging innovation activities in research communities, industry and service sectors.

Tekes has a budget that is almost twice that of the Academy of Finland – 575 million euro versus 309 million euro respectively for 2009. The contrast with Ireland is stark. The budgets for basic research funding via the HEA based PRTLI (programme for research in third level institutions), and the SFI (Science Foundation Ireland) research budgets, are far bigger than the more applied research supported via Enterprise Ireland. The HEA and SFI may protest that they support applied research but, as best as one can establish from the very sketchy information they provide, they are both dominated by a very strong basic science orientation. This basic science orientation may be appropriate for the HEA but it certainly is not appropriate not for the SFI if it is to meet the objectives set for it by the government.

The type of RD&I supported with public funds in Finland is dramatically different to Ireland. Tekes promotes a broad-based view on innovation: besides funding technological breakthroughs, Tekes emphasises the significance of service-related, design, business, and social innovations. It supports innovation in processes and services as well as in the development of new products. This is very much in line with the OECD definition of innovation as consisting of four broad types: product innovation; process innovation; marketing innovation; and organisational innovation.

Tekes works with the top innovative companies and research units. Every year it finances some 1,500 business research and development projects, and almost 600 public research projects at universities, research institutes and polytechnics. Most of the Tekes projects are with companies of less than 50 employees. On the Networking section of the it’s website under Tekes programmes it lists 29 ongoing, very diverse programmes. Examples of success stories are also listed which range from the sublime to what some of our “big scientists” might even consider ridiculous. The range and nature of the programmes and projects is very instructive.
Tekes has also established Strategic Centres for Science, Technology and Innovation which are new public-private partnerships for speeding up innovation processes. Their main goal is to thoroughly renew industry clusters and to create radical innovations through long-term cooperation between industry and academia. International cooperation also plays a key role in the operation of the Strategic Centres. In the Strategic Centres, companies and research units work in close cooperation, carrying out research that has been jointly defined in the strategic research agenda of each Centre. The research aims to meet the needs of Finnish industry and society within a five-to-ten-year period. Six centres are in operation at present. These are (1) Forest cluster; (2) Information and communication industry and services; (3) Metal products and mechanical engineering; (4) Energy and the environment; (5) Built environment innovations; and (6) Health and well-being.

This is not to suggest that there is nothing of value in the Irish system. Isolated cases seem very similar to the Some Irish institutions operate in a manner very similar to the Tekes model. Finland has basic “science” programmes similar to Irelands. HOWEVER, IN IRELAND THE BALANCE SEEMS ALL WRONG AND FINLAND PROVIDES A TEMPLATE WHICH WE WOULD BE VERY FOOLISH TO IGNORE.

I have commented before on this and other sites on the background to these matters and I repeat some of these comments below.

However, I suggest that economists should be much more concerned about the influence that the research philosophy in institutions like UCD is having on Irelands approach to economic development. It is not at all clear that the emphasis on basic research, especially in the biomedical area, is correct for a small country like Ireland. I believe that exploiting cutting-edge developments in research that takes place worldwide and applying them for our benefit should be our priority, and that the ICT area is likely to have much more potential than the life-sciences area for us in terms of establishing new indigenous enterprises.
I include extracts from a memo I prepared some months ago dealing with these matters.

There have been a number of issues referred to recently dealing with science and innovation; the importance and the need for R&D for the services sector, especially traded services; the need for R&D for the building sector and the broad environmental sector; and the funding and the teaching priorities for third level institutions. As someone who has recently retired from UCD and who was very active in both teaching and research I feel compelled to say that we seem to be making a serious error in one aspect of our approach to research, development and innovation.

A fairly widespread but narrow, and dare I say, blinkered view of the role of basic/fundamental research is shown in the approach of SFI and some of the universities at least. I believe that a wide-ranging review is called for into the likely return that the people of Ireland will get from the large sums of the money that are currently being distributed on basic research by Science Foundation Ireland (SFI) and the Higher Education Authority (HEA). The emphasis on basic research in much of the funding is, I believe, misplaced.

The probability that any individual basic science research project will lead to some tangible benefit to society in the medium term is very low. Many projects will, in fact, contribute nothing of value except a few scientific papers that may occasionally be read by other academics and researchers. Because of the very low probability that any individual project will lead to anything of value it is reckless for a small country to invest a large proportion of its research budget into, of necessity, a limited number of basic research projects. While it makes perfectly good sense for the US government or the EU to invest in a large number of basic research projects, confident that some will give a good return, it is rash for us to invest in a small number in the (rather vain) hope that we might strike it lucky. If our researchers can get funding for basic research from the EU or from major international foundations so much the better, but the Irish government should strictly limit its funding of basic research.

University research in Ireland should probably be based on the Massachusetts Institute of Technology model. It is a world-class educational institution where teaching and research—with relevance to the practical world as a guiding principle—continues to be its primary purpose. High quality applied research which address real problems is much more complex and more challenging than a lot of basic research; and it is usually educationally more valuable than conducting basic research without reference to any real-world applications. In Ireland we will have to be selective but our emphasis should be on the applied research rather than basic research but using the best techniques and the most recent developments from anywhere around the world in carrying out our applied research. Basic research is usually freely available via the normal publication channels. However, establishing linkages with the best universities where the most relevant basic and applied research is being conducted should probably be a priority for us in Ireland.

The mission of the SFI is to build and strengthen scientific and engineering research and its infrastructure in the areas of greatest strategic value to Ireland’s long-term competitiveness and development. However, four of the five metrics which the SFI have set to measure its success deal with principal investigators, research fellows and postdoctoral researchers in underpinning disciplines; articles in international journals; membership of international academic societies; and patents and licenses filed (but not necessarily used by anyone). Only one metric refers to commercial start-ups and multinational and indigenous investments in R&D. These metrics are, I believe, indicative of a narrow approach which is not best suited to the mission of the SFI.

Might I also suggest that the metrics used to rank universities on a world scale may not be consistent with the objectives, explicit or otherwise, which Irish society has for our third level institutions? Very creative research on the optimum design of wind energy or wave energy systems for Irish conditions will hardly win a Nobel prize; or even get published in the highest ranked international journals. Individuals engaged in such valuable applied research are made to feel like second class academics, in some universities at least. Those engaged in SFI-funded projects are strongly in favour of basic/fundamental research as against applied research; and a small minority of PhDs and post-docs leaving SFI projects have gone into industry. The Indecon report on the SFI calls for (1) increased focus on effective industry collaboration and measures to enhance the commercialisation of research; (2) increased focus to align collaborations by SFI-funded researchers with the requirements of industry based in Ireland; and (3) the development agencies, including IDA Ireland and Enterprise Ireland, to intensify efforts to engage new and existing client companies with SFI-funded research teams/centres. Clearly, the approach taken in a major part of the SFI programme merits attention.

The activities of the SFI and our research funding agencies raise a number of related issues affecting our third level institutions.
(1) Are the (short-term) funding activities of the SFI distorting the (long-term) staffing decisions of the universities (and institutes of technology?) with long-term consequences for the university teaching programmes? Are specialised researchers valuable/useful for most undergraduate teaching? Has the amount of time devoted to teaching by staff been reduced while they pursue their research interests? Are the activities of the SFI and the HEA properly co-ordinated? Are the functions of the institutes of technology being deflected by the activities of the funding agencies?
(2) How is R & D supposed to lead to economic development? What models or precedents are being followed? If research is supposed to lead to the establishment of hi-tech SMEs, as in the Silicon Valley model, is there any likelihood that biomedical research will lead to the establishment of SMEs in the health sciences area? I would be very concerned about the usefulness of basic research in the life sciences; but less so about research in the ICT area.
(3) The World Bank stated that to rapidly increase R & D spending in an attempt to reach the EU target of 3% of GDP by 2010 may result in throwing good money after bad. Are our structures adequate to avoid making this mistake?

Could I point out that the Irish economy and our industrial development did brilliantly during the 1990’s. The graduates from our universities and institutes of technology performed very well. This was before the SFI was even established, and spending on research was minimal. The high growth rates in the economy since the turn of the millennium were greatly dependant on the unsustainable expansion of the building sector plus consumption driven by an explosion in personal debt. It probably has had very little to do with SFI-funded basic research. It would be instructive to examine what contribution a representative sample of Irish basic research over the last 10 years has made (or is about to make in the near future) to the Irish economy, or to examine how has Irish basic research in biomedical sciences has improved best practice in Irish health care.
These are just some of the issues which the politicians and the policy makers (and economists) must address. We need a wide-ranging debate with inputs from many quarters. Academics and researchers often have a very blinkered view. I have searched a great many sources for some insight into the issues raised – without a lot of success. There are no ready answers to these questions. We must be ready to question the conventional wisdom. The politicians and policy makers should demand clear and concise explanations for the various activities undertaken. They should demand a reasonable estimate of value for money for broad areas of activity and should hold people accountable for their “promises”. A lifetime in one of our universities has taught me to always be sceptical of the conventional wisdom, and that Hans Christian Anderson was correct in that the emperors frequently have no clothes.

Myles seems to have contributed something quite valuable above me there, which I will certainly return to when I have time.

My employer Liam Carroll, bit the dust for the second time around this evening. Not that I am surprised. I wrote something for the Sunday Tribune a couple of weeks ago, to describe a very unique kind of ‘skunkworks’ within the Irish construction industry. Something that sort of sprang up and blossomed, in the most unlikely of situations, Zoe developments.

Iulia wrote:

“If done with the state-of-the art methods these evaluations are extremely helpful not only to uncover inefficiencies but also successes.”


Do you think there could be many good templates, or good ideas out there, lodged in businesses somewhere, which are un-recognised?

Certainly, I believed that Royceton was one of those. It had to operate along with the other 250+ companies of the Zoe developments group. But I believe the Royceton company, if nurtured could have produced something valuable. Anyhow it is no more.

The big problem with Royceton, was it was a design, engineering and finance operation all-in-one, within a development and construction company. Basically, the external specialist companies who provide design, engineering and financial services to the construction industry didn’t like the emergence of Royceton at all. Because it threatened the status quo of their business model.

I enjoy a bit of ‘creative destruction’ myself, and I always endeavour to sniff out those green shots, where ever I find evidence of it. Even within my own area of construction and design.

I am reminded of the ‘Alta Vista’ search engine story, that is covered very well by John Battelle, in his book ‘Search’, where Digital Equipment Corporation had the saviour of the company, in that search engine technology, but their culture wasn’t able to recognise it.

Gordon Bell, the famous engineer wrote something about his opinion why Digital collapsed in the late 1980s.


Another quick memo, while I am here. It is well worth listening to Richard Sennett talk about the craftsman on BBC 4 radio in Feb 2008.

If you read the Sunday Tribune piece I linked above, before listening to the interview, you might identify some of the parallels in my thinking and that of Richard Sennett. I want to write a blog entry this evening or over the weekend, in relation to the taxation commission. They still seem to believe in the taxation commission that skills are mobile. I would like to counter that, with an argument for the opposite, in my future blog entry, when I get around to it. I don’t believe that skills are ‘mobile’ in the sense that everyone thinks.

There is obviously some sense to Richard Florida’s theory. But how much of that, is to do with the fact that places like Austin Texas, are places which enable the people to become more ‘creative and innovative’ when they get there. I mean, we can call Cherrywood science and technology park whatever we want to call it. But honestly, is it going to facilitate innovation in the way we hope? Heck no.

I wrote something about the McCarthy report, which is about getting more for less.

Rather than as Brendan Keenan submitted in today’s Indo, we might end up with less for more. What I am interested in, is to join up the dots – McCarthy, commission on taxation, NAMA, the whole bigger picture. I do know how Ireland should go about that in a joined up sort of way. Which is why I would be interested in having more involvement in this area.

The truth is, if you listen to some one such as Alan Kay talk about his days spent at Xerox Parc laboratory, he mentions the fact that after Parc he never had the quality of people to argue with again. He imagined he could go on by himself and be equally as innovative wherever he went. But that did not turn out to be the case. Hence, why it sticks in my throat to listen to the commission on taxation accounts speak of ‘skills’ moving in or out of the country. This isn’t the point. The point is, places such as Cherrywood science and technology park will never foster anything the way they are. That goes for a lot of other ‘innovation quarters’ also.

I was listening to an old 2002 MIT world webcast last weekend only.

A panel discussion in which they made the point, that at the time of Xerox Parc Labs, they possibly had 30% of the best talent available in the world at the time, to work on problems. Hence why nearly all ground breaking developments in other computers later, had originated at Xerox Parc. But no one is likely to ‘corner the market’ to the degree that Xerox did, ever again. Even in the top companies today, at best one could hope to get around a 3.0% of total market share of the grey matter.

It is about working best with that or a lot less.

I feel a lot about Royceton, in Zoe developments, the way that Alan Kay feels about the Xerox days.

Comments are closed.