India’s Services Revolution Amidst Worldwide Structural Change

According to Engel’s Law of income elasticity of demand, as income increases, the proportion of income spent on food grains decreases and proportion of income spent on manufacturing and services increases.

According to Kuznet’s categorization, A sector includes Agriculture and Allied activities, I sector includes Mining and Quarrying, Manufacturing, Construction, Electricity, Gas and Warehouse (EGWs) and Transport, Storage and Communication (TSC) and S sector includes Trade, Hotels and Restaurants (THR), Banking and Insurance (B&I), Dwellings, Real Estate and Business Services (DRB), Other Private Services, Public Administration and Defence (PA&D) etc.

Here we are taking the data of 10 OECD countries (including Japan), viz., Denmark, Italy, France, Japan, Netherlands, Sweden, Germany, United Kingdom, United States of America and Spain, for the period 1950–2010. Now, we have taken the shares of industry, service and manufacturing as the relevant dependent variables for our analysis. We are interested to know whether a significant structural break exists around 1970 for the OECD countries (including Japan) (i.e., DCs).

The type of the underlying series is shown in Table 2 for sectoral shares of Industry, Service and Manufacturing for each of the 10 nations in our study. The study shows that for Industry share in Value Added (Table 2(a)), Denmark, France, Japan, Germany and USA are showing Crash Model to be appropriate. Italy, Netherlands, Sweden, UK, Spain are showing Mixed Model to be appropriate. Correspondingly, according to the critical values corresponding to the asymptotic distribution of Zivot and Andrews (1992), Denmark, Italy, France, Japan, Netherlands and USA are showing unit root, i.e., these series are Difference Stationary (DS) and remaining four countries, viz., Sweden, Germany, UK and Spain are showing Trend Stationary (TS) series. Further, the study shows that for Service share in Value Added (Table 2(b)), Denmark, France, Japan, Sweden and Germany are showing Crash Model to be appropriate. Italy, Netherlands, UK, USA and Spain are showing Mixed Model to be appropriate. Correspondingly, according to the critical values corresponding to the asymptotic distribution of Zivot and Andrews (1992), Denmark, France, Japan, Netherlands, Germany and USA are showing unit root, i.e., these series are Difference Stationary (DS) and remaining four countries, viz., Italy, Sweden, UK and Spain are showing Trend Stationary (TS) series. Again, the study shows that for Manufacturing share in Value Added (Table 2(c)), all the 10 countries are showing Mixed Model to be appropriate. Correspondingly, according to the critical values corresponding to the asymptotic distribution of Zivot and Andrews (1992), Italy, Japan, Netherlands, Germany and Spain are showing unit root, i.e., these series are Difference Stationary (DS) and remaining five countries, viz., Denmark, France, Sweden, UK and USA are showing Trend Stationary (TS) series.

The type of the underlying series is also shown through the methodology adapted by Sen (2003) in Table 3 for sectoral shares of Industry, Service and Manufacturing for each of the 10 nations in our study. The study shows that for Industry share in Value Added (Table 3(a)), according to the critical values corresponding to the asymptotic distribution of Sen (2003), for the period 1950–2010, Denmark, France, Netherlands, Sweden, Germany, USA and Spain are showing unit root, i.e., these series are Difference Stationary (DS) and remaining three countries, viz., Italy, Japan and UK are showing Trend Stationary (TS) series. However, for the period 1950–2007, Italy, Japan and Germany are showing unit root, i.e., these series are Difference Stationary (DS) series. Further, the study shows that for Service share in Value Added (Table 3(b)), according to the critical values corresponding to the asymptotic distribution of Sen (2003), for the period 1950–2010, Denmark, Italy, France, Japan, Netherlands, Germany, USA and Spain are showing unit root, i.e., these series are Difference Stationary (DS) and remaining two countries, viz., Sweden and UK are showing Trend Stationary (TS) series. However, for the period 1950–2007, Japan is showing unit root, i.e., this series is Difference Stationary (DS) series. Again, the study shows that for Manufacturing share in Value Added (Table 3(c)), according to the critical values corresponding to the asymptotic distribution of Sen (2003), for the period 1950–2010, Italy, Japan, Netherlands, Germany, Sweden, USA and Spain are showing unit root, i.e., these series are Difference Stationary (DS) and remaining five countries, viz., Denmark, France, and UK are showing Trend Stationary (TS) series. However, for the period 1950–2007, Italy and Japan are showing unit root, i.e., these series are Difference Stationary (DS) series.

A look at Table 4 will make it clear that Category-I Services-intensity (in terms of indirect effects) is by far the highest in the Secondary Sector. Furthermore, this intensity has increased substantially from 1973–1974 to 1993–1994 and to 2006–2007 relative to 1973–1974. This deepening of Service intensity is present in other sectors also. This fact indicates to a change in the structure of final demand from material goods to Services (Table 5 in appendix) will change the sectoral shares in GDP in favour of Services; and consequently it will also change the relative weights of the two categories of Services. But this is not the only source of change, the change in relative weights will also be affected by the increase in intermediate Service-intensity of production which should raise the relative share of Category 1 Service further.

Normal practice is to classify the I-O framework into 3x3 matrix based upon three major sectors, viz., Agriculture and Allied activities, Industry and Services. However, here, it may be noted that we have divided the Service Sector in two broad parts, viz., Category-1 Service and Category-2 Service. Category-1 Service is basically of intermediate nature and Category-2 Service is basically of final use in nature. Hence, instead of using the terms intermediate and final, we use neutral terms –Category-1 and Category-2 Service – as Category-1 has some final use while Category-2 has some has some intermediate use. Hence, we have classified our four major sectors as ‘Agriculture and Allied’ (A&A), ‘Industry’ (I), ‘Category-1 Service’ (S1) and ‘Category-2 Service’ (S2) (Datta et al. 2008).

As Datta (2011) has shown,Then, \(\underline{\Delta {\varvec{\chi }}} = \underline{\mathbf{V}_{\mathbf{0}}(\mathbf{I-A}_{\mathbf{0}})^{-\mathbf{1}}\Delta \mathbf{F}} + \underline{\mathbf{V}_{\mathbf{0}}\Delta (\mathbf{I-A})^{-\mathbf{1}} \mathbf{F}_{\mathbf{1}}} + \underline{\Delta \mathbf{V}(\mathbf{I-A}_{\mathbf{1}})^{-\mathbf{1}}\mathbf{F}_{\mathbf{1}}}\)

where, \(\Delta \chi \) implies Value Added Effect; \(\hbox {V}_{0}(I-A_{0})^{-1} \Delta \hbox {F}\) implies Final Demand Effect ; \(\hbox {V}_{0}\Delta (I-A)^{-1}\hbox {F}_{1}\) implies Input Structure Effect ; and \(\Delta \hbox {V}(\hbox {I-A}_{1})^{-1}\hbox {F}_{1}\) implies Reallocation Effect.