While Asian software parks have attracted foreign software firms and created job markets for technology workers, ultimately they have not become the hubs of innovation that policy-makers hoped they would be. These are the main conclusions of a new report entitled Parks and valleys: Growing Asia’s software industries published this week by the Economist Intelligence Unit, the business-to-business arm of the Economist Group, which publishes The Economist magazine. The report was sponsored by the Business Software Alliance (BSA).

The report looks at the efforts of several Asian countries to emulate the success of hardware manufacturing parks by establishing their equivalent for the software industry. Focusing on four Asian software parks (Cyberjaya in Malaysia, Dalian Software Park in China, Nankang Software Park in Taiwan and Quang Trung Software City in Vietnam), the research examines the extent to which they have been successful in promoting start-ups and domestic software firms, as well as fostering innovation.

The report compares the development of the software industry in Israel and Ireland and says in relation to Ireland: Despite efforts to promote R&D, Ireland deals mainly in mature technology that others have developed elsewhere. By 2000, Ireland had become the world’s largest exporter of software products, largely because companies like Microsoft had located their regional packaging, localisation, distribution and logistics processes there. Ireland’s indigenous companies have had a marginal role in this rapid growth in exports of high-end software products, with foreign firms accounting for 90% of the exports.

• The report concludes that while technology parks may work in encouraging hardware manufacturing, they generally do not have the attributes necessary to foster the growth of thriving software development communities. Software developers need access to vibrant environments where ideas, capital and people move quickly—something isolated parks some distance from central commercial districts do not offer. Rather than using software parks as the main route to development, governments should aim to create and sustain a policy environment that makes their entire country attractive to those who build or finance technology businesses. Based on in-depth interviews with software park administrators, executives of companies based within the parks, and venture capitalists in Asia, the report’s key findings include:

• Start-up firms have higher survival rates inside parks than outside them. Each of the software parks examined in this report has an incubator to help launch start-up companies. Facilities offered by the incubators include shared R&D resources, low rent, tax breaks and business administration support. The survival rate of start-ups after two years ranges from 43% at Cyberjaya in Malaysia to 80% at Nankang Software Park in Taiwan—considerably higher than survival rates outside the parks.

• Asia’s software parks have not become innovation hubs. Despite providing opportunities for knowledge sharing, software parks have not lived up to their promise of promoting technological innovation. Tenant companies often focus on outsourcing or software services. Poor intellectual property protection makes multinational software firms reluctant to locate core R&D abroad, which presents a barrier to innovation and technology transfer. Unlike Silicon Valley, the parks examined for this report are far from commercial centres and this can make them unattractive for the best and brightest software engineers.

• The focus on exports can pose long-term growth issues. Building an export-focused software industry generates investment and economic development, but an approach tipped too far towards exports may pose long-term growth issues. Increasing domestic demand and the growth of indigenous IT companies should reduce Asia’s vulnerability to an economic slowdown in Western markets.

• Asian governments can do more to promote venture capital. Venture capital is a key component in the development of software industries. In Asia, restrictions around ownership, governance and capital repatriation can dissuade quick investments in the technology sector, in which financiers typically seek large stakes in small firms. Asian governments have often used public funds to stand in for venture capital. In some cases this practice can be helpful in jump-starting software industries, but it risks crowding out private investors.

• Fast and cheap telecoms services are critical to software development and are best promoted through liberalisation. There is a tendency for Asian policy-makers to treat telecoms services as a component of public infrastructure. However, to increase penetration and reduce costs, a market-driven approach through liberalisation is more effective than a public planning approach.

• Education policy needs to be directed to create sufficient numbers of highly-skilled graduates. The quality of surrounding academic institutions is a central consideration for companies weighing up whether to relocate to a software park. The focus on teaching of some Asian universities means they lack the kind of research capabilities that foster innovation in the software industry.