The descriptive statistics presented in Appendix Table 7A.1 show that 66 per cent of all firms are involved in new product development while only 37 per cent carry out innovation in the hardware computer sector. Hence, the percentage of firms that are involved in new product development in the software sector is much higher than in the hardware sector. However, productivity, i.e. sales per employee (in millions of dollars), is higher in the computer hardware sector than in the computer software sector. In other words, sales per employee are on average about one million dollars in the whole computer sector and twice as much in the computer hardware. The figures for export intensity, i.e. the share of export sales in total sales, and increased net profit are on average similar for the computer hardware and software sectors. More specifically, export intensity is (on average) about 17 per cent in the whole sector and 13 per cent in the hardware sector, and net profit increased for 88 per cent of all firms and for 84 per cent of the computer hardware firms. In short, the propensity to innovate is far higher in software firms but many overall similarities exist in the two subsystems. The descriptive table also shows that 23 per cent of the firms are computer hardware firms and also have the lower percentage of staff with university or technical degrees (human capital) compared with the software firms. Not surprisingly, 73 per cent of workers in the whole sector have a university or technical degree while the percentage is only 55 per cent in the hardware sector. The figures for firm size, upgrade activities, technology source, government support, customer demand, technical capability and training in the whole sector are contrasted with those of the same variables in the hardware sector. On average, hardware firms are much larger in size than software firms. More specifically, the former are on average three times as large, in terms of employees, and four times as large, in terms of sales, as firms in the software sector. The percentage of firms that upgrade with reverse engineering and original design is on average
larger in the software than in the hardware sub-sector, while firms that upgrade with original brand is larger in the latter sub-sector compared with the former. Firms that upgrade with quality control are on average similar across the two sectors.2 And finally, when the figures on technology source of the whole sector are compared we find that software firms depend more for their technology on local expertise and in some cases on a combination of local and foreign expertise such as licensing from clients and buyers relative to hardware firms. Other sources of technology include hiring of skilled employees, collaboration with universities and public institutes, and reverse engineering. For hardware firms technology source is largely from foreign expertise and component suppliers. The two sub-sectors draw equally from joint venture partners, transfer from parent firm and suppliers of equipment.