Abhijit Banerjee explains the mechanics of growth:
The problem is that for a business to rise beyond its many competitors—the thousands of fruit vendors in Chennai---it has to have something special about it: The product (P) must be different or the quality (Q) must be especially high, or the firm must have special reputation for reliability (R) or the scale of operations (S) must be large enough to generate significant cost savings. And each of these requires a combination of special skills and substantial amounts of money, both beyond the reach of all but a few poor or even not so poor business owners.
It is these PQRS businesses that generate the good jobs that other aspire to, and the earnings that come out of them lead to other businesses and so on. This, to a first approximation, is my vision of the process of how growth happens. It is what China has managed to do very successfully and Africa will have to find a way of doing.
Consider this particular feature of self-employed, self-sustaining, (and stagnant at lowest equilibrium) business model in the developing countries (“follow-the-herd” business model):
It turns out that the businesses of the poor are also poor businesses: The typical business has zero paid employees and no machines in almost every country where we have data and where we have the information to be able to calculate this, what the household earns from the business is less than what they would earn on the lowest end of the labor market. They are in effect buying a job and not particularly good job at that.
Isn’t this just a reflection of the fact that they do not have enough capital to run a proper business? Yes and no. These businesses are certainly undercapitalized, but the businesses of those who are significantly richer (those who live on $6 to $10 a day, for example) really do not look all that different from these. Moreover the amounts of money invested in these businesses are so tiny that a family living on three or four dollars a day per capita, could easily double or treble their capital stock in a year by simply halving what they spend on tea or cigarettes.
Dean Karlan and Sendhil Mullainathan, in a recent paper, put this point rather starkly. They study fruit vendors in Chennai, India, who make about two to three dollars a day by buying fruit in the morning on credit and paying it back at night. It turns out that the interest rate they pay is 5% per day and at that rate, saving the ten cents they spend on tea for just one day would allow them to pay back their entire loan in six months (the power of compound interest) and add a dollar a day to their earnings. Yet most of them seem to be permanently stuck in their business model.