Bad Ideas For Ending Global Poverty (Part II)

It is hard to evaluate the role of technology parks in sustaining growth.

Indian IT exports from outside IT parks rapidly increased along with the rest of the industry (from $164 million in 1991 to $1,248 million in 1999).

The most successful parks are those near engineering colleges such as Bangalore and Hyderabad, while parks set up in areas otherwise unattractive to the IT industry such as Gandhinagar have floundered (only around a fifth of initially registered firms in the park remain).

It is difficult to evaluate the percentage of firms in the parks that established themselves in parks rather than not establishing in India at all as compared to the percentage that established in parks rather than elsewhere in India.

Prior to the growth of the industry, the government did relax a number of regulations governing ICT infrastructure and business regulation. These reforms continued during the 1990s, including simplified export procedures, and waived limits on foreign ownership and foreign exchange purchase.

The government also trained thousands of highly skilled graduates in Indian Institutes of Technology. Because most of the earliest graduates from the 1960s and 1970s were forced to look abroad for work due to a miserable business environment at home, the Indian diaspora ended up running a considerable chunk of Silicon Valley.

By 2000, Indians headed 972 Silicon Valley firms accounting for $50 billion in sales and 26,000 jobs.

When the Indian government finally got around to improving the business climate back at home, there was a ready-made army of Indian entrepreneurs and managers with good qualifications, contacts and experience in the United States. Many were happy to return back home and found companies or manage subsidiaries.

The Indian experience suggests a novel role for government — focus on high quality tertiary education in fields with a labor shortage in the developed world while ensuring a poor home business environment and/or exiling anyone with a tertiary degree.

Wait 20 years, and then reform the business environment and rescind the exile orders. But this is not the usual strategy that ICT proponents envisage.

If government subsidies and tax breaks for ICT production have often floundered in the more economically successful parts of Asia, they have to be a particularly high-risk strategy for a Rwanda or a Laos to follow.

It would, of course, be an even riskier strategy for Laos to follow if Rwanda was already following it, because of the potential for a “race to the bottom” in terms of competing subsidies and tax breaks of the sort that have frequently destroyed the economic benefits of export processing zones.

The “fallacy of composition” (that what works for one country will work as well if every country tries it) begins to bite hard if the whole developing world decides to set itself up as a haven for programmers — as Russia, Eastern Europe and South Africa, also attempting to build IT industries, may soon discover.

There is only a finite demand for new computers or programs or ICT-enabled services. If the whole world starts to produce them, their price will inevitably fall. And less efficient countries (where the broader institutional environment is not as favorable, for example) will be priced out of the ICT market.

In turn, they will either have to subsidize the IT industry to an ever-greater degree, or see all of their earlier subsidies come to naught. Such a race to the bottom would be a costly mistake.

Malaysia’s and India’s experience suggest this race to the bottom has already commenced. Indeed, despite (surely) having achieved the scale of a successful IT cluster, manufacturers in technology parks in India are still provided with infrastructure, core computer facilities, ready-to-use office space and the right to duty — and license-free imports in addition to tax exemptions, all at a significant cost to the government.

Furthermore, the benefits of this expenditure accrue to an economic elite — subsidizing the IT industry means tax breaks to Bill Gates, not to landless farmers.

We have seen that clusters are likely to form where there are concentrations of highly educated people, venture capital and other factors of success. Meanwhile, the spillover effects of clusters on poor communities are unclear — poor people don’t produce much needed by ICT firms, and they are not major consumers of software.

The Nobel Prize-winning economist Amartya Sen noted that even 100 of India’s Bangalore technology hubs would not solve India’s poverty and deep-seated inequality.

However successful the Indian ICT model, challenges of replication, limited evidence of spillovers and the highly regressive nature of subsidizing jobs for the highly educated and the investments of multinationals and elites — all this suggests that ICT industry promotion is a comparable pop-gun in the armory for the global fight against poverty.