Last month, the Center for International Development (CID) at Harvard University made news with its prediction that India will be an important future growth pole for the world economy. One doesn’t need sophisticated analysis to make this kind of prediction at a general level, but the concepts and calculations behind the forecast are worth examining more carefully.
The CID report was based on the work of Ricardo Hausmann of Harvard and Cesar Hidalgo of MIT, and their index of economic complexity (ECI). The broad notion of economic complexity is tied to the diversity and sophistication of the products of a country’s economic activity. To measure this complexity in a manner that is sufficiently detailed and comparable across countries, the approach must use trade data, which is both disaggregated and standardised. A country will have a higher ECI if it exports a wider range of products (indicating diversity), and those products are exported by fewer other countries (indicating sophistication).
The report that predicted a rosy economic future for India was based on a favourable ECI calculation, giving India a higher ranking than its per capita output (GDP) would, and thereby suggesting current conditions that will favour growth. The last step in this chain of reasoning is based on analysis of recent data for many countries by the authors of the ECI, which suggests that the ECI is correlated with contemporaneous and future per capita GDP.
Several ancillary points support this line of reasoning. The ECI does better than another older and oft-used index, the Word Economic Forum Global Competitiveness Index. The ECI works as a predictor even when other variables such as the quality of a nation’s institutions are controlled for. And the ECI works similarly to more basic measures of product diversity that do not rely on export data. This last point might be used to make a case for using the simpler measure, as well as providing support for the robustness of the ECI as a predictor. The diversification measure is developed and analysed by Sabiou Inuoa, who also relates it conceptually to technological sophistication and innovation potential, which is presumably a major channel for economic growth.
So far so good. Other work by Hidalgo and several co-authors suggests that the ECI is also negatively related to income inequality, implying that ECI-driven growth can be more inclusive than that based on specialisation in a few products, especially natural resources. This makes the picture even rosier. Of course, the authors are careful not to suggest that the processes are automatic, and they still admit a role for policy making, with the ECI being a possible tool for that process of guiding the economy.
In practice, the task for India seems somewhat more challenging than the ECI might suggest. First, it is not clear how the aggregate measure of economic complexity can be translated into specific policy initiatives. Will producing more smart phones and tablets in India (beyond just final assembly of components made elsewhere) help growth more than making more of something else, say automobile components? Is producing many kinds of nuts and bolts better than making one type of sophisticated machine? Some of the underlying issues behind such questions pertain to classification of products, which often fails to keep up with, or fully reflect, variety and sophistication. Information technology products are notoriously under accounted for, in terms of the variety of software that is written. The distinction between products and services, apparent in the case of software, also crops up as an important complication for the ECI approach.
Thinking about such issues leads me to suggest that, while ECI may be a good summary statistic for how an economy is doing in broad terms, it has limited usefulness for policy making. Instead, a policy approach that tries to identify and ride on major trends in innovation and in demand makes more sense. For example, it is clear that the use of more sophisticated software tools across the entire economy, as encapsulated in the idea of “artificial intelligence,” will be a major driver of innovation and economic growth in the next decades. On the demand side, the needs of ageing populations will change the product mix in important ways as well, not just in high-tech health care and shopping, but also in food systems, and wellness and “anti-ageing” products. Indian policy makers might think about such possibilities, and the ECI will take care of itself.
Finally, despite the optimism engendered by India’s ECI, its poor competitiveness ranking, which relates to matters such as ease of doing business, as well as to industrial dynamics and innovation more generally, has to be a concern. Institutional reforms that improve regulation, delivery of public goods and services, and the competitive functioning of markets are still desperately needed in India. And, despite the importance of material well-being and progress, events in India continually remind us that there are other dimensions of well-being, such as human dignity and respect for differences, which also need to be nurtured. Neither the ECI nor GDP will capture that adequately.