While SA is spending on R&D, this is not filtering through to patent level.
Dr Andrew Paterson, research director of the Education Science and Skills Development Research Programme at the Human Sciences Research Council, says research and development is happening in SA, but the country lags behind its peers in patented outputs and there is no clear idea of which skills are needed for what type of output.
Paterson, who posted a draft discussion document online in October, says there is also no indication of whether computer science graduates are contributing to research outputs.
In 2003/4, just over R944 million was spent on R&D in ICT, he says. SA intends boosting R&D output to 1% of gross domestic product by 2008. Two years ago, the country spent 0.81% of the country's revenue on R&D.
Just under 20% of all research and development spend by companies in 2001/2 - with a similar figure the following year - was spent on ICT. However, there is no indication of what value is added to the country through this expenditure.
No clear picture
Paterson's analysis "uncovered a disappointingly flat R&D performance in the South African ICT sector as measured in terms of domestic investment and international patent registrations, with some highlights in the sub-field of software development".
"This performance did not give strong reason to assume heightened demand for skills in the period from the early 1990s to about 2002," the paper says.
Paterson says there is no doubt that R&D is happening in the sector, but seems to be happening at company level and there is no resultant flow through into patents. "We should look at whether SA wants to encourage R&D that leads to income generation through patents."
Looking at the European Patent Office, Paterson notes in the paper that ICT-related patents are growing more rapidly than overall applications, with the notable exception of ICT patents locally and in Japan.
Which skills?
Paterson has not ascertained the level of demand for R&D, nor for the level of demand of post-graduate computer science professionals, as a result of most of the research taking place in-house. This, he says, is an area he intends researching further.
While innovation could occur outside of the strict computer science professional arena, the level of post-graduate professionals when compared to those in similar fields such as engineering is not encouraging.
Paterson says for every 100 undergraduate physics degrees, there are 40 post-graduate qualifications. This number drops to 29 in engineering, and only 18 in computer sciences. Looking at conversions from a bachelor's degree to a PHD, Paterson notes that engineering has a conversion ratio of 1.63%, while computer sciences has a ratio of 0.26%.
One conclusion that could be drawn is that R&D potential in local ICT enterprises could be constrained by a lack of suitably qualified professionals. However, investigating this angle was not feasible in the discussion document, it states, because "without reliable information describing demand, there is no means of assessing the adequacy of the supply of R&D skills from higher education".
Paterson says even through innovation can be derived from different sources, computer science as a discipline, is not contributing to the community of R&D at a level he would like to see.
An explanation for this, he says, is the skills shortage that results in these graduates being "snapped up" by the sector and not having time to continue their studies.
Share