Definitions of R & D : report with evidence.

Great Britain. Parliament. House of Lords. Science and Technology Committee.

Date:: 1990

Catalogue details

Licence: Open Government Licence

Credit: Definitions of R & D : report with evidence. Source: Wellcome Collection.

49/148 (page 47)

2 March 1989] [Chairman contd.] Frascati Manual to include the new technology- generating activities being carried on in large companies, not in R&D labs but in systems departments. There may be a case for adapting the manual in that respect. Secondly, there is an even stronger case for improving coverage of technological activities—call it R&D or whatever— not only in manufacturing but in service sectors, because the interesting point about software technology is that it is increasingly being developed in locations not previously associated with the generation of technology—banks, grocers, Marks & Spencer and so on. Lord Nelson of Stafford 36. You seem to be distinguishing between what is not innovative software? (Professor Pavitt) That is quite right. That is an extremely difficult problem. What is _ just programming and what is a real development of software technology? All I am saying is that there is a very important challenge, which no country has taken up, and which would require more resources if it was to be tackled properly. This is a point which the OECD may have taken on board already. It will require changes in habits, because it means looking for technology in places one has not looked at before. 37. Reverting to the controversial area of aerospace, like flight-testing, engine-testing, and so on, these activities require big sums which can have a dramatic impact on costs. Do you find that other countries involved in these activities, like France, Germany and the United States, interpret the term R&D in the same way as the Ministry of Defence? (Professor Pavitt) I cannot answer that question with regard to the statistics; I cannot say what inconsistencies there might be between government and industry in these different countries. 38. But it is important? (Professor Pavitt) Yes. I do know that, in terms of concepts, the United States Defense Department and other departments distinguish between research, development:and testing; they admit to distinctions between those activities. What the different habits and inconsistencies are I cannot say. Chairman 39. But Frascati say, for example, that the first prototype should be regarded as development and the second and third should not? (Professor Pavitt) Yes. 40. So that if you follow Frascati you should be fairly all right on this, should you not? (Professor Pavitt) You should. I become rather pessimistic about one’s ability to change it, at least in this country, because for nearly 30 years there has been this problem of inconsistency between what Government and industry say in aerospace. 41. And the figures are so large they distort the whole picture? (Professor Pavitt) Again this is where you have to be rather careful because although by Frascati definitions you are quite right that activity beyond the first prototype is not development, from a policy [Continued point of view lots of technical and financial resources and people are tied up in the subsequent stage, and it would be wrong to sweep that under the carpet one way or another. Lord Kearton 42. Let me take two examples—the modernisation of Polaris and the Chevaline programme—which were originally estimated at about £100 million and the final cost was over £1 billion. Was that R&D or was it classified and returned as R&D? (Professor Pavitt) 1 cannot answer that question. You can probably answer that question much better than I can. 43. If you take the Nimrod programme, is that R&D or product engineering, would you think? [Professor Pavitt) Again there are other people round this table who can answer that question better than I. Chairman: I would guess that actually the Nimrod programme was all within R&D; it did not reach production or order stage. Lord Chorley: This does bring one back to the question of useful ouput. You said you did not regard output as being a measure of activity, which I interpret as being a measure of expenditure. Are there any useful measures of output, in your view? Lord Kearton 44. Balance of trade in technical products? (Professor Pavitt) You get into problems there of how you define high technology. If you take certain definitions, Britain and the United States look better than Germany and Japan simply because you include aerospace and exclude mechanical engineering and automobiles. You can therefore get into all sorts of difficulties with how you define high technology industries. Let me be clear. What I have said about output is very much my own view. Some of my colleagues would say that, if you are careful, you can use patenting or counting discrete innovations as an output measure, provided you do it over a relatively short period, and for firms in the same industry. But you have to be careful. You cannot compare the ratio of patenting to R&D in fabricated metal products and in aerospace and conclude that the former is 100 times more productive than the latter. No, I do not see any easy, quick mechanical method of comparing inputs and outputs. The data always require interpretation, even on the input side. If you take small firms, you will find the numbers of innovations per unit R&D is extremely high. Does that mean small firms are more efficient than big firms? Answer, no. It is simply that R&D activity is only a small part of what is happening with small firms in terms of technology input. A very careful interpretation is necessary, and a mechanical system would be extremely dangerous. ~ Chairman 45. But to get to the basics, if there are all these difficulties in the adequacy of the statistics and the interpretation and everything else, what do we want the statistics for?