Guoli Chen, Ravee Chittoor and Bala Vissa thought that this embeddedness research that is focused on educational background could perhaps be especially valid in a Western context (where most of the research has taken place) but that in a different context, such as India, different types of affiliations might also play an important role. Specifically, they wanted to focus on the role of caste (i.e. people being of the same or different castes) and language (in terms of people sharing the same regional dialect).
Research setting: Equity analysts in India
To examine these different dimensions of inter-personal networks, they focused on a particular set of people and relationships, namely equity analysts. Firms listed on the stock exchange will often be followed and evaluated by analysts, as employed by banks, who make buy and sell recommendations to the public regarding the company’s stock.
Perhaps the most important task of such an equity analyst is to forecast – as accurately as possible – the future earnings of the firm. However, to make an accurate forecast, an analyst often has to at least partly rely on information received directly from the company; not seldom in the form of personal conversations with the Chief Executive. And Guoli, Ravee and Bala suspected that when the analyst happened to share the same background with the company’s CEO it would be much easier for him or her to get access to the CEO and his company information; making his earnings forecasts more accurate.
They tested this suspicion on a sample of 141 Indian firms, followed by a total of 296 equity analysts, between 2001-2010. First of all, they found clear evidence that equity analysts that are alumni of the same academic institution as the company’s CEO were indeed able to make much more accurate forecasts. But, in addition, the same was true for analysts who shared the same background in terms of caste, and in terms of regional language. In fact, the effects were roughly the same size, meaning that these old historical patterns (around caste and language) were just as important in India as the more contemporary ones (i.e. university affiliation).
They then examined the conditions under which these different types of informal ties mattered more or less or whether such ties were indeed always beneficial. They found that older CEOs – who could be expected to be influenced more heavily by traditional patterns – were more susceptible to issues of caste and language than younger CEOs. They were less influenced by joint academic affiliation. Hence, although these old historical patterns matter a lot in India; they matter less for younger people, who are relatively more susceptible to joint academic affiliation.
In addition, they found evidence that these informal relationships were particularly beneficial if it concerned a truly Indian firm (part of a traditional business group). In contrast, such informal ties hurted more than they helped, when the firm in question was an Indian subsidiary of a Western multinational corporation.
Overall, what Guoli, Ravee and Bala’s research shows is that, in a country like India, old historical social structures still matter a lot in the world of business, especially when it concerns firms that are part of a traditional business group. The effect of language (which is analogous to ethnicity) was particularly potent. These effects may begin to matter a bit less for younger people (i.e. since they were especially strong for older CEOs) but they still wield considerable influence on economic life.
Paper presented at the “Sumantra Ghoshal Conference for Managerially Relevant Research” at the London Business School.
Which old boy network matters? Basis of social affiliation and the accuracy of equity analysts’ earnings forecast of Indian firms. Guoli Chen (INSEAD), Ravee Chittoor (Indian School of Business), Bala Vissa (INSEAD)
This paper summary is published with permission from the authors.