Service boom in the Indian economy: an analysis of causal influences

Abstract

The article decomposes the influence behind the growth of service activities in the Indian economy over the last three decades into three components – the final demand effect, input structure effect and reallocation effect – and makes empirical assessments thereof. In terms of the influences, the behaviour of the group of services used basically as intermediate inputs, designated as category-I services, was distinct from the rest of heterogeneous services combined as ‘community, social and personal services’. The study finds that apart from the final demand effect, the other two influences too played very important but different roles in different phases of growth.

Acknowledgements

Searching questions from Mihir Rakshit, Amitava Bose, Pradip Maity and others present in a seminar at the ICRA Monetary Research Project, where a preliminary version of this article was presented, have helped to improve this article. Comments from the editor and a referee of Applied Economics have also helped. The author acknowledges his debt to all of them and accepts liability for the remaining inadequacies.

Notes

¹ For example, revolution in computing and telecommunications technology has not only given us several new products but also has revolutionized services in most fields.

² Consider, for example, the role of banks in foreign trade and mergers and acquisitions.

³ A study by Balakrishnan and Parameswaran (2007) shows that the growth of GDP for the Indian economy has two phases only. The break point is 1978–1979. The annual rate of growth accelerated by almost 2 percentage points in the second phase when it was 5.2%.

⁴ Growth acceleration in the 1980s in manufacturing in particular and in services in the later decade are well-documented in the literature (Rodrik and Subramanian, 2005; Srinivasan, 2005; Balakrishnan and Parameswaran, 2007).

⁵ The procedures in all these papers were flawed in that the inter-temporal comparison of Leontief technical coefficients was not based on corrections for price changes. Datta and Bhattacharya (2009) incorporate the correction and confirm the finding of rising service intensity. Also, see Francois and Reinert (1996) for a study on Organization for Economic Co-Operation and Development (OECD) economies.

⁶ See the interesting discussion on division of tasks between Central Administrative Office and Operating Manufacturing Establishments in Siegel and Griliches (1992).

⁷ Nevertheless, for quantitative analyses undertaken in this article, the price indices referred to here are probably the best available to researchers. We have used these indices in spite of the limitations.

⁸ It may be noted that a part of transport has a direct consumption demand while demand for trade services is entirely derived demand. Final demands for the two services, as shown in the IOTTs, result from demand originating from the need for distribution of final goods plus direct final demand for transport services.

⁹ Consider the bracketed part of the second term of Equation 10. [(s + m)e^μt /(ae^αt  + ne^μt )] = [s(1 + V · (1 + τ))e^μt /(ae^αt  + me^μt W · (1 + τ)]. Compare this expression with the corresponding part of Equation 6. The numerator and part of the denominator of the expression are multiplied by some factors in Equation 10. If V and W are equal, χ_t in Equation 10 is larger than that in Equation 6; and this is true even if W is larger than V by a reasonably wide margin.

¹⁰ Observations based on IOTTs for the Indian economy show that the share of intermediate TT in total TT services increased significantly from about 45% to 53% over the 10-year period from 1968/69 to 1978/79. Thereafter, the ratio declined somewhat. The value of the ratio was 48.5% both in 1993–1994 and 2003–2004.

¹¹ Further, growing complexities lead to emergence of new problem resolution mechanism. Emergence of firms such as APL Logistics and Maersk Logistics now provide integrated logistics service such as handling of congestions at ports and providing multimodal transport facilities for time-sensitive items within and beyond the East Asian region (Yusuf, 2004, p. 27). This provides an example of ‘spread’ of service activity though not specifically belonging to but facilitating trade and transport.

¹² See Coppieters (1987) in this context.

¹³ A very common consequence of technological progress in a sector is decline in the price of the relevant product. This leads to change in VA per unit of the product not only in the progressive sector but also in the relatively nonprogressive sector because of maintenance of wage parity. An interesting study on Spanish car industry is done by Matas and Raymond (2009).

¹⁴ Clearly,

Also,

Since Y ₀ = Y ₁ by assumption, we have w ₁ = (4/3)w ₀. So, distribution of VA between the sectors, which was w ₀:w ₀ (or 1:1) in the base year, is now 0.5w ₁:w ₁ = (2/3)w ₀:(4/3)w ₀ (or 1:2) in the current year. This means, (since X ₀ = X ₁).

¹⁵ An elegant formal model relating changes in relative shares and relative prices to differential elasticities of demand and rates of sectoral technological progress is given in Kongsamut et al. (1997).

¹⁶ Here, the assumption of a single quantifiable and homogeneous primary input makes the measurement of real VA straightforward. Otherwise, we could obtain the GDP deflator (based on the equality Y ₀ = Y ₁) and apply that to obtain sectoral real VA.

¹⁷ This allows us to use as the best available sectoral price deflators those provided by the CSO in spite of their limitations.

¹⁸ Estimates in Tables 1 and 2 do not tally perfectly with those given by IOTT. One reason is the treatment of indirect tax incorporated in inputs which we have included in the VA of the using sector for consistency of input–output analysis.

¹⁹ It, perhaps, provides a counter example to the notion that services are nonprogressive and, hence, rise of the sector will be accompanied by fall in productivity (Kaldor, 1975; Inman, 1985). In fact, Bosworth and Collins (2007) notes: ‘Unlike for China, India's impressive performance in services is largely reflected in a rapid improvement of TFP’. Also, see Goldar and Mitra (2008) in this context.