The Karnataka government is set to release a new industrial policy next month with the goal of encouraging investment in tier-II cities. As it has been in the past, this goal is likely to be framed in zero-sum terms i.e. achieved by pushing IT companies to move away from Bengaluru and in other cities instead. For instance, Karnataka Chief Minister BS Yediyurappa recently said that the government will announce a revised IT policy that will focus on attracting investments and employment to Tier-II and Tier-III cities.
We will limit this article’s focus on what such a policy direction would mean in high-tech sectors such as biotech, aerospace, and IT. , this push towards creating an alternative of centre gravity for the high-tech industry seems to be an intuitive answer for achieving balanced regional growth. And yet, this view is wrong because it doesn’t square with the empirical experience of high-tech clusters elsewhere in the world.
Agglomeration and Clustering
For instance, the high-tech industry in the US tends to be concentrated in a small number of expensive markets. Recent research by economist Enrico Moretti finds that more than half of all inventors in the US in three high-tech fields – computer science and information technology (70 per cent); semiconductors (79 per cent); and biology and chemistry (59 – work in clusters of just 10 cities across the vast country. Moreover, this concentration has only become stronger over the last five decades, pointing to the benefits of increasing geographical agglomeration for inventors.
Investigating the reasons for clustering, the research finds that an inventor moving from a small cluster to a large cluster enjoys an increase in annual productivity, as measured by the number of patents produced in a year or the number of citations. Had the same number of inventors been uniformly distributed across the country, the industry as a whole would have been less innovative.
There is ample evidence for this clustering effect in other countries as well. For instance, economists Carrincazeaux and others found that six regions in France accounted for 75 of all corporate R&D workers, compared with 45 per cent of the production workers.
The most likely explanation for clustering is that technological progress builds upon itself, with the expansion of one domain propelling future work in linked fields. For example, Daron Acemoglu and others in a 2016 study found that a stable “innovation network” acts as a conduit of this cumulative process of technological and scientific progress.
Agglomeration effects occur in cities also due to the sharing of common inputs (from high-speed broadband to computer science engineers), of tacit knowledge from one firm to another, and the development of institutions and universities that create and sustain the technology clusters. The high density of firms and highly skilled in the better quality matching of the right person for the right job.
Implications for Bengaluru
This research suggests that Karnataka faces a paradox not unique to the state. Even though the high-tech sector around Bengaluru is a key driver of Karnataka’s economic growth, it will operate efficiently only if it remains spatially concentrated.
Given this reality, it is unwise to expend efforts with a view of driving away high-tech industries from Bengaluru in the hope that they will set up shop elsewhere. In fact, Karnataka should be doing quite the opposite: Make Bengaluru an even bigger and better high-tech hub than it already is.
Karnataka needs to embrace the narrative that under-governance, not over-population, is Bengaluru’s fundamental problem. The focus should be on increasing the productivity of Bengaluru, which requires meeting three to the eminent planner Alain Bertaud: One, firms and households should have the freedom to stay put or migrate at will; two, travel within the city should remain fast and cheap, and three, real estate should be sufficiently affordable that it does not distort the allocation of
If policies are created with these outcomes in mind, we can deepen the innovation network and prolong its march towards prosperity for more Kannadigas.
What about balanced regional development?
The view that economic clusters get built because governments wish so is a commonplace planning fallacy. Policymakers around the world have been attempting to build the ‘next Silicon Valley’ and unfortunately, these attempts have been unsuccessful.
Picking a geographical location and targeting the growth of clusters is usually a high- approach due to three reasons: One, there is no fundamental reason why policymakers are better than entrepreneurs in assessing the future economic potential of a venture, including the development of a cluster. France, for example, and created elaborate high-tech policies in the 1980s to develop a microelectronic cluster and openly admitted their failure five years later.
Second, the possibility for areas to reap benefit from high-tech clusters may be limited. High- tech clusters offer fewer opportunities for mass employment. Further, it would mean that a newly ‘selected’ high-tech cluster must compete with well-established ones. In Karnataka’s case, a second cluster that tries to mirror the cluster model of Bengaluru would have to compete against Bengaluru. However, left to itself, markets would develop in a way that would be differentiated from established centres.
Finally, in trying to pick a winner, policymakers often ignore the question of whether the necessary preconditions for the successful development of the cluster are present in the area. Differences in starting-position, economic structure, and institutional particularities are important. The target location may not have the absorptive capacity for new technologies or it might not have the necessary conditions for the development of universities. The costs associated with such policy failures can be quite large as exemplified by the ghost Siberian ‘city of science’ Akademgorodok or the failures of ambitious tech policy in Southern Italy and Ruhr in ‘cathedrals in the desert’ or projects built in the middle of nowhere that hardly anybody would use.
The best form of industrial policy for a backward region is to ensure that the basic requirements for any industry are present. This includes good educational institutions, good infrastructure, and transport facilities, rule of law and property rights, development of markets, and real ease of doing business. Once these conditions are present, the market is equipped to pick out development ventures and the agglomeration effects naturally develop. The government can develop policies around this to encourage or facilitate the development, but attempting to pick winners at scratch is counterproductive.
(Anupam and Pranay Kotasthane are researchers at the Takshashila Institution)
The views expressed above are the author’s own. They do not necessarily reflect the views of DH.