Dani Rodrik is right in claiming that in developing economies entrepreneurial activity, whose social value is much more than private costs, comes up against the issue of "first mover disadvantage" arising from uncertainty about the underlying cost structure - what can and cannot be produced profitably.
He writes about such entrepreneurs, "If they succeed, much of the gains are socialized through entry and emulation, whereas if they fail, they bear the full costs... One must presume that there are many more (discovery efforts or entrepreneurial activities) that could be taking place, but which are not, because it is difficult for pioneers to capture a large enough part of the social surplus they generate, even with the subsidy programs in place."
So how do we enable the pioneers to internalize a greater share of the social surplus? We cannot provide exclusive rights, like patents, because many of these are not patentable. Governments cannot assume the role of a venture capitalist of last resort. How can governments distinguish between "good" and "bad" ideas and entrepreneurs, and support the "good" ones? Can Governments avoid the moral hazard associated with picking winners, especially given the numerous examples of subsidies ostensibly given to support entrepreneurship, but plainly doled out to cronies and favorites? Nor can we expect the markets to step in and single-handedly perform the role of supporting entrepreneurship and picking winners.
There are those, like William Baumol and Co, who will say that Governments can play an in-direct role in enabling the development of entrepreneurial culture - lowering entry and exit barriers (start-up registration, flexible labour markets, accessing finance); institutional framework that rewards entrepreneurial activity (property and contract rights); discourage destructive entrepreneurship that divides the pie than increase its size( corruption, crime, excessive taxation, subsidies etc); and winners should be incentivized to continue their innovation and not stagnate (anti-trust laws, and openness to trade). But these are the necessary, but not sufficient, conditions for entrepreneurial activity and economic development.
Scanning the economic landscape of the past few decades, it is difficult to come across examples of such market driven rise of entrepreneurship in the developing economies. On the contrary, one can cite numerous examples of "guided bureaucratic" entrepreneurial activity - East Asia, China, India, the flower industry in Ethiopia etc. Even the economic history of Japan, US and Europe bears this out.
Take the example of the first flush of flower exporting entrepreneurs in Ethiopia. There was no way they could have emerged endogenously, through the dynamics of the market forces, given the inherent economic, commercial and technical uncertainties (not even "risks" - remember the "known unknowns and unknown unknowns"). An exogenous intervention, by way of subsidy through cheap land and tax holidays, was necessary to discover the underlying cost structure and break open the market. As Dani says and as the theory about positive externalities affirms, private sector will always be reluctant to step in where social benefits exceed private costs.
Given the crucial role of the Government in picking winners, the importance of a committed bureaucracy and enlightened political administrators assumes significance. This institutional framework, probably the most critical determinant, is rarely a consideration in standard economic principles of policy making for promoting entrepreneurship. Wherever acknowledged, it gets grudgingly marginalized into a corner, clubbed as part of the now fashionable "good governance".
I had blogged earlier on the similar chicken and egg problem facing the market for infrastructure financing and government sector software applications in India. Both remain trapped in a stalemate, and will continue to be so in the absence of active external (read government) intervention to break the dead-lock.