TRANSITION REPORT 2014 Innovation in Transition

Box 5.3. Assessment of links between industry and science in transition countries

The previous box reviewed various schemes for fostering links between industry and science. To assess the effectiveness of such initiatives, we can look at various indicators of links between the two.

One such indicator is based on patent information. It is available for all countries over a long period of time, but it covers only a fraction of the links between industry and science. It looks only at the information contained in patents, so it is unlikely to capture the majority of the links between industry and science, particularly in countries that are in the process of catching up with the technological frontier.36

Read more

Chart 1.15 in Chapter 1 showed that, when compared with the United States or Germany, a remarkably large percentage of patents in transition countries are applied for by universities or public research organisations. This is particularly true of Russia, Poland and Ukraine, where more than a third of all patents are held by universities or research institutes. In countries such as Estonia, Slovenia and the Czech Republic, academic patenting is much closer to the levels observed in the United States and Germany. Meanwhile, Turkey has very low levels of academic patenting.

Furthermore, while co-patenting involving academia and industry is relatively rare everywhere, its incidence in the transition region is relatively high compared with the United States. Russia stands out in this regard, accounting for 62 per cent of all co-patenting in the transition region. This suggests that universities and research institutes have a high degree of involvement in the development of technology, especially in Russia. Consequently, links between industry and science in transition countries such as Russia mostly involve the scientific community supplying new technology to industry.

The picture is dramatically different when looking at such links from the perspective of corporate demand – in other words, when looking at how often corporate patents refer to scientific literature (scientific non-patent references) as prior art for patented inventions.

When assessed on the basis of this indicator, Russia and Ukraine score very poorly (see Chart 5.3.1). Latvia, Slovenia, Hungary and Estonia, on the other hand, score much better than other major patenting countries in the transition region, albeit they still lag behind the United States. Israel scores almost as highly as the United States on this indicator. With few patents being used as prior art for further patenting by firms, the impact of academic patenting in the transition region is likely to be limited in practice.

Thus, these data suggest that what is lacking in most transition countries – including those where the scientific community develops a lot of new technology – is a corporate sector that actively uses its links with science to innovate. There is a policy bias in this regard, with countries stimulating the supply of new technology – rather than demand for it – without much regard for the country’s level of innovation.

Overall, this assessment raises the question of whether industry actually needs the results of scientific research conducted by local public institutions and whether it has sufficient capacity and incentives to take those findings on board. It also indicates that policies need to pay greater attention to industry demand in the area of innovation.

CHART 5.3.1

Source: PATSTAT and authors’ calculations.
Note: Data cover all patent applications reported in PATSTAT that originate in the relevant country. The figure for the transition region is an unweighted average, including only countries with at least 1,000 patent applications. This leads to the exclusion of Albania, Armenia, Azerbaijan, Bosnia and Herzegovina, Egypt, FYR Macedonia, Georgia, Jordan, Kazakhstan, the Kyrgyz Republic, Mongolia, Montenegro, Tajikistan, Tunisia and Uzbekistan. Croatia is not included because its sectoral allocation is not reliable.