ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Growing Services in India

This study is an explorative, inter-sectoral analysis of state domestic product data to understand the determinants of services sector growth in India during recent years. It is a demand side analysis where the services sector output of a specific state is not only a function of the outputs of a state's own agriculture and industry but also the output of the commodity-producing sector of the rest of the Indian economy. The findings suggest that while a state's own industry turns out to be the most important determinant of services sector growth, the commodity-producing sector of the country outside the state does play a significant role as well in determining services sector performance under certain conditions that basically relate to the supply side.

Growing Services in India

An Inter-Sectoral Analysis Based on State-Level Data

This study is an explorative, inter-sectoral analysis of state domestic product data to understand the determinants of services sector growth in India during recent years. It is a demand side analysis where the services sector output of a specific state is not only a function of the outputs of a state’s own agriculture and industry but also the output of the commodity-producing sector of the rest of the Indian economy. The findings suggest that while a state’s own industry turns out to be the most important determinant of services sector growth, the commodity-producing sector of the country outside the state does play a significant role as well in determining services sector performance under certain conditions that basically relate to the supply side.

DEEPITA CHAKRAVARTY

IIIII
IntroductionIntroductionIntroductionIntroductionIntroduction

T
he services sector has in recent decades been performing much better than agriculture and industry in India and in many other developing countries as well. According to standard literature, services experience an accelerated growth only after a certain level of development in agriculture and then in industry. In this regard, the Indian experience is somewhat different. Even today India cannot be considered as an industrially developed economy. This paper tries to see how far industrial development is responsible for the expansion of services.

In the development literature much attention is given to theorising the evolution of agriculture and industrial growth, as also to the relationship between the two. A theoretical treatment of the services sector and its growth, in a macroeconomic context, on the contrary, is hard to find.1 However, it may be assumed that in a three-sector economy consisting of agriculture, industry and services, the demand for services in a closed economy is a function of the outputs generated in the commodity producing sectors of agriculture and industry. In an open economy, domestic services can expand either directly through external demand for specific services or indirectly through the boost in local incomes provided by remittances from emigrant labour. States within India are fully open with respect to other states in the country. Thus, growth in incomes elsewhere, especially in neighbouring regions can promote the expansion of services in any region under favourable demand and supply conditions.

With this brief introduction we now turn to the hard facts. The second section deals with the issues of data and the related empirical literature to justify a new study in this area. The last two sections are devoted to the analysis. We look at the state level data to get insights into regional specificities.

IIIIIIIIII
DataDataDataDataData

The two variables we concentrate on are output and employment. We begin with sectoral outputs. National Accounts Statistics (NAS) provide time series sectoral output data for the country as a whole, as well as for the states. The output figures, or for that matter, the figures for gross state domestic products (GSDP) are given under the broad heads of the major nine industrial categories with some disaggregation. For our convenience we have clubbed the categories of agriculture, forestry, livestock and fisheries together under the agricultural sector. Mining, manufacturing, electricity, gas and construction come under the industry category and the rest are grouped under services. The information for GSDP is available both in terms of current and constant prices. The period for analysis chosen in this study is 1980-81 to 2002-03. Currently, the data for state domestic product are available in 1980-81 prices for the period 1980-81 to 1997-98. Another series is available from 1993-94 onwards till 2002-03 in 1993-94 prices for all states and till 2004-05 for a few. In order to get continuous data researchers have converted the data with 1980-81 as base to conform to 1993-94 prices [Bhattacharya et al 2004]. However, the definitions used in constructing the 1993-94 wholesale price index are considerably different from those used for the construction of the index numbers with 1980-81 as base. As this difference of definitions can influence the numerical values to a considerable extent, we decided not to combine the two series.

Our intention is to understand the services sector boom in the 1980s and 1990s. We work with two periods: initial phase of liberalisation from 1980-81 to 1992-93 with 1980-81 as base and the later phase of liberalisation from 1993-94 to 2002-03 with 1993-94 as base: we expect to capture changes through the relevant comparisons.

What about the employment figures? The decennial census gives us the sectoral employment figures for all states once every ten years. As a consequence a time series analysis was not possible and we decided to do a cross section analysis of the output elasticities of services sector employment and productivity with respect to the outputs of the commodity producing sectors of the economy. There are still two more difficulties with this data source. First, for the year 1981 employment figures are not available for the state of Assam. Therefore, our 1981 cross section analysis does not contain the output figures for Assam. Secondly, we had to limit ourselves to the early 1990s because for the year 2001 data for workers under the nine-fold classification are not yet available. The Census data for 2001 gives a four-fold classification where the fourth category consists of the non-household industry and the services sector employment together. For the analysis of employment we have considered the main workers only.

The services sector boom in India has evoked the issue of sustainability in a major way in the recent years. Studies have basically approached the question either through an input-output framework [for example, Hansda 2001] or through econometric modelling [Sastry et al 2003] at the all-India level. Though the input-output framework allows us to look at the interdependence between different sectors, it has serious limitations too.2 In their attempt to estimate and forecast the state domestic products for three states in India, Bhattacharya et al (2004) do shed some light on services sector growth at the state level in recent years. But they worked basically with a closed economy, where rest of the country’s commodity producing sector has no role in determining a single state’s service sector performance.

The present study approaches the problem from the demand side. Our specification includes not only the outputs generated in the commodity producing sectors within the state but also in the rest of the Indian economy. In order to quantify the “rest of the Indian economy” with respect to a specific state, we deduct the specific state’s commodity producing sector’s output from that of the total 16 state’s combined output of agriculture and industry. We calculated the output elasticities of demand for services output using a simple double log specification with three independent variables, namely, the state’s own agriculture and industry and the output of agriculture and industry in the rest of the Indian economy. We have also looked at the elasticities without introducing the independent variable “rest of the Indian economy” in the model. The equations are estimated for the initial as well as the later period. Apart from these, trend rates are calculated with a semi log specification for the relevant variables.3

It is worth mentioning here that our specification does not address the demand for services generated outside the country. This is important, for the Indian economy is in the process of opening up itself to the world market since the early 1980s. The indirect effect of the external economy mentioned above has been taken care of by the fact that we are using the “gross domestic product” for measuring output.4 The most visible and well known dimension of the expansion in the services sector in India, especially in the late 1990s, is IT-based services, including exports. However, Gordon et al (2004) argue that growth in services in India has been much more broad-based than IT. The study shows that although IT exports has a profound impact on the balance of payments, the sector remains a small component of GDP. Even in the year 2003, total IT related services accounted for only 3 per cent of the total services output.

IIIIIIIIIIIIIII
Estimation ResultsEstimation ResultsEstimation ResultsEstimation ResultsEstimation Results

Table 1 indicates a fairly high rate of growth of the services sector for a majority of Indian states in the decades of the 1980s and 1990s. In the initial phase, the maximum is about 8 per cent and the minimum hovers around 4.5 per cent. While the minimum remained more or less the same in the later phase of liberalisation, the maximum touched 9.5 per cent. On the basis of the rate of growth of the services sector in the initial phase of liberalisation we have classified the states into three groups respectively with per cent growth rates (1) above 6.5; (2) between 6.5 and 5.5;

and (3) below 5.5. Interestingly, the composition of the three groups has changed considerably in the later phase of liberalisation with more states moving up to the first group. The most significant change is noted in the cases of Bihar and West Bengal; both these states have moved up in terms of services growth.

Let us now turn to the essential question asked in this paper: what are the factors responsible for the performance of the services sector in different states? Tables 2 to 4 give services output elasticities with respect to the state’s own agriculture, industry and the rest of the Indian economy in two periods. We begin our discussion with the first group of states with services growth rates above 6.5 per cent. Table 2A presents the results for the period 1980-81 to 1992-93. First, note that as indicated by Durbin-Watson statistics, serial correlations are not present. In all the states industry is a significant variable. (The addition of the “rest” variable makes industry insignificant in the case of Rajasthan. In this case we ignore the specification with three independent variables.) The corresponding elasticity varies from

0.76 in Haryana to 1.17 in Maharashtra. During the initial phase of liberalisation, the variable “rest of the Indian economy” plays a significant role only in the cases of Haryana and Gujarat.

This trend persists only in the case of Haryana in the later phase. For Gujarat, addition of the “rest” variable turns industry insignificant in the second period. The presence of serial correlations for both the specifications in Karnataka in the later phase ruled out the possibility of any discussion. However, Maharashtra shows an interesting trend of insignificant elasticity for industry in this period with a significantly high elasticity for the “rest”. In all these cases while agriculture shows an insignificant elasticity, the responsiveness with respect to industry is generally significant in both periods.

Incidentally, the value of the elasticities with respect to the rest of the Indian economy is quite high and above one in all the cases. It is interesting to see why in some cases the rest of the Indian economy becomes significant. We come back to this issue shortly. Marginal changes can be seen in the elasticities with respect to industry in the later phase of liberalisation when compared to the initial one in the states of Rajasthan and Gujarat,

TTTTT
ableableableableable
1:1:1:1:1:
Annual Rate of Growth (in per cent) of the ServicesAnnual Rate of Growth (in per cent) of the ServicesAnnual Rate of Growth (in per cent) of the ServicesAnnual Rate of Growth (in per cent) of the ServicesAnnual Rate of Growth (in per cent) of the Services
Sector in Different States of IndiaSector in Different States of IndiaSector in Different States of IndiaSector in Different States of IndiaSector in Different States of India

States Ranked in 1980-81 to 1992-93 1993-94 to 2002-2003 Descending Order on the Basis of the Growth Rates Achieved Growth Residual Growth Residual in the Period Rates Mean Rates Mean 1980-81 to 1992-93 Square Square

Rajasthan 8.1* (0.41) .003 7.4*(0.44) .002
Maharashtra 7.4* (0.29) .002 7.0*(0.24) .000
Haryana 7.2* (0.28) .001 9.3*(0.23) .000
Karnataka 7.1* (0.12) .000 9.5*(0.16) .000
Gujarat 6.6* (0.24) .001 7.8*(0.24) .000
Himachal Pradesh 6.4* (0.28) .001 7.7*(0.39) .001
Tamil Nadu 6.3* (0.23) .001 8.0*(0.36) .001
Madhya Pradesh 6.2* (0.19) .001 5.5*(0.41) .001
Uttar Pradesh 5.9* (0.22) .001 4.6*(.002) .000
Orissa 5.9*(0.29) .002 6.5*(0.16) .000
Andhra Pradesh 5.6* (0.22) .001 7.0*(0.13) .000
Bihar 5.1* (0.16) .000 7.5*(0.57) .003
Kerala 5.0* (0.17) .001 5.5*(0.41) .000
West Bengal 4.9* (0.10) .000 9.3*(0.14) .000
Assam 4.9* (0.21) .001 4.7*(0.43) .002
Punjab 4.4* (0.12) .000 6.0*(0.31) .001

Notes: (1) * indicates significance at 5 per cent level.

(2) Standard errors are given in the parenthesis.

Economic and Political Weekly July 8-15, 2006

but none of them are statistically significant. During the last two decades, the industrial sector of both these states have performed reasonably well and comparatively much better than the agricultural sector (Tables 5 and 6). However, even in the case of industrial growth rates, it is difficult to trace a statistically significant change in the later phase when compared to the initial one. This indicates no qualitative change in industrial growth leading to change in the responsiveness for the services sector output. However, results for Haryana show a statistically significant improvement in the elasticity with respect to industry. Moreover, though Haryana has also performed well in the industrial sector during recent years, there is a statistically significant decline in industrial growth rate in this state in the second period. A probable qualitative change in the industrial initiatives in the state of Haryana might have played an important role in determining the elasticity. It is, however, essential to look at the composition of industrial growth in order to establish this hypothesis.

Among the states in the second group, i e, with services growth between 5 and 6.5 per cent, the equation for Tamil Nadu in the initial phase of liberalisation shows indications of multicolinearity for the specification with three independent variables.5 For this state the same specification indicates a strong relationship in the later period between the state’s industry and the rest of the Indian economy. Therefore, we decide to consider the specification involving the state’s own agriculture and industry alone for Tamil Nadu. A similar relationship is also suggested in the case of Madhya Pradesh for the specification involving three independent variables for the concerned periods. While the Orissa results suggest the existence of multicolinearity in the initial phase of liberalisation, the later phase of liberalisation indicates the same for Himachal Pradesh and Uttar Pradesh for the same specification with three independent variables.

Apart from the above-mentioned problems, Tables 3A and 3B suggest quite similar patterns depicted by Tables 2A and 2B. States in the second group show insignificant elasticities with respect to agriculture for both periods as well except Andhra Pradesh in the later phase. Moreover, the elasticity with respect to industry shows significance for the last two decades with significant increase in the numerical value only for the state of Orissa in the later period when compared to the first. The declining rate of growth of industry in Orissa together with the above result is suggestive of some diversification in the industrial initiatives in the later phase in this state.

Finally, let us turn to the last group of low services growth (below 5.5 per cent) states. Incidentally, services sector in most of these states performed reasonably well in the later phase of liberalisation, except in Assam. It is clear from Tables 4A and 4B that the industrial sector once again turns out to be significant in most of the states in generating demand for services in the state economies. However, in some of these states, agriculture also does play a significant role but not consistently in both the periods. Incidentally, the rest of the Indian economy is significant in both the periods only in the case of Bihar. Yet the difference between the two numerical values of elasticities is statistically insignificant.

While, in the initial phase of liberalisation agriculture was insignificant in generating demand for services in Bihar, it became significant in the later phase. There is a marginal improvement in the rate of growth of agriculture in Bihar in the second period over the first from a low base of 1.6 per cent per annum. However, the change is statistically insignificant. Incidentally, the improvement in the industrial growth rate in Bihar is considerable and statistically significant in the later phase of liberalisation leading to no significant change in corresponding elasticities. Can it be

Table 2A: Services Sector Output Elasticities with Respect toTable 2A: Services Sector Output Elasticities with Respect toTable 2A: Services Sector Output Elasticities with Respect toTable 2A: Services Sector Output Elasticities with Respect toTable 2A: Services Sector Output Elasticities with Respect to
Agriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture Plus
Industry of Rest of the Indian Economy for the High GrowthIndustry of Rest of the Indian Economy for the High GrowthIndustry of Rest of the Indian Economy for the High GrowthIndustry of Rest of the Indian Economy for the High GrowthIndustry of Rest of the Indian Economy for the High Growth
States during 1980-81 to 1992-93States during 1980-81 to 1992-93States during 1980-81 to 1992-93States during 1980-81 to 1992-93States during 1980-81 to 1992-93

States Agriculture Industry Rest R-squared DW Stat

Rajasthan -0.06(.118) 1.11*(.095) – 0.97 2.457 -0.26(.145) 0.45(.338) 1.42* (.707) 0.98 2.797 Maharashtra -0.01 (.121) 1.17* (.089) – 0.98 1.040 -0.05(.118) 0.79*(.271) 0.62(.424) 0.98 1.325 Haryana 0.21 (.201) 0.76* (.116) – 0.97 1.591 -0.23(.227) 0.52*(.128) 0.94*(.354) 0.98 2.066 Karnataka 0.21 (.192) 0.96* (.091) – 0.99 1.270 0.22(.201) 0.85* (.281) 0.17 (.403) 0.99 1.200 Gujarat -0.01 (.073) 0.88* (.060) – 0.96 2.334 -0.06 (.045) 0.33*(.126) 1.00*(.218) 0.99 2.316

Notes: (1) Standard errors are given in the parenthesis.

  • (2) ‘*’ indicates significance at the 5 per cent level.
  • (3) The specifications used: (A) ln (serv) = a+ ln(agri) + ln(ind), (B) ln (serv) = a + ln(agri) + ln(ind) + ln(rest). Where ‘agri’ stands for output of the state’s own agricultural sector, ‘ind’ represents output of the state’s own industrial sector. ‘serv’ stands for output of the state’s services sector and Rest stands for output of the commodity producing sector of rest of the Indian economy. Ln represents natural log.
  • Table 2B: Services Sector Elasticities with Respect toTable 2B: Services Sector Elasticities with Respect toTable 2B: Services Sector Elasticities with Respect toTable 2B: Services Sector Elasticities with Respect toTable 2B: Services Sector Elasticities with Respect to
    Agriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture Plus
    Industry of Rest of the Indian Economy for the High GrowthIndustry of Rest of the Indian Economy for the High GrowthIndustry of Rest of the Indian Economy for the High GrowthIndustry of Rest of the Indian Economy for the High GrowthIndustry of Rest of the Indian Economy for the High Growth
    States during 1993-94 toStates during 1993-94 toStates during 1993-94 toStates during 1993-94 toStates during 1993-94 to
    2002-032002-032002-032002-032002-03

    States Agriculture Industry Rest R-squared DW Stat

    Rajasthan -0.03 (.151) 0.98*(.099) – 0.94 2.230 -0.11 (.111) 0.18 (.297) 1.83*(.659) 0.97 2.088 Maharashtra 1.27 (.843) 0.83 (.513) – 0.65 0.925 -0.17 (.562) -0.22 (.370) 2.13*(.495) 0.91 1.968 Haryana 0.25 (.435) 1.39*(.121) – 0.99 1.216

    0.37 (.337) 1.91*(.230) -1.02*(.417) 0.99 1.966 Karnataka 0.37(.430) 1.13*(.235) – 0.92 0.747 0.36(.419) 0.49 (.590) 1.38 (1.167) 0.94 0.608

    Gujarat -0.16(.117) 1.09*(.101) – 0.94 2.012 -0.18*(.077) 0.37 (.233) 1.532*(.474) 0.98 1.676

    Notes: (1) Standard errors are given in the parenthesis.

  • (2) ‘*’ indicates significance at the 5 per cent level.
  • (3) The specifications used are same as Table 2A.
  • Table 3A: Services Sector Elasticities with Respect toTable 3A: Services Sector Elasticities with Respect toTable 3A: Services Sector Elasticities with Respect toTable 3A: Services Sector Elasticities with Respect toTable 3A: Services Sector Elasticities with Respect to
    Agriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture Plus
    Industry of Rest of the Indian Economy for the Medium GrowthIndustry of Rest of the Indian Economy for the Medium GrowthIndustry of Rest of the Indian Economy for the Medium GrowthIndustry of Rest of the Indian Economy for the Medium GrowthIndustry of Rest of the Indian Economy for the Medium Growth
    States for the Years 1980-81 to 1992-93States for the Years 1980-81 to 1992-93States for the Years 1980-81 to 1992-93States for the Years 1980-81 to 1992-93States for the Years 1980-81 to 1992-93

    States Agriculture Industry Rest R-squared DW Stat

    Himachal 0.05 (.210) 0.80* (.084) – 0.95 2.467 Pradesh -0.09(.181) 0.39*(.176) 0.85*(.333) 0.97 2.744 Tamil Nadu 0.36 (.239) 0.95* (.215) – 0.94 2.299

    0.23(.230) 0.11(.530) 1.10(.644) 0.95 1.975 Madhya -0.05 (.502) 0.92* (.190) – 0.92 0.917 Pradesh -0.22 (.364) 0.22 (.261) 1.19*(.377) 0.96 1.953 Orissa 0.05(.132) 0.79*(.052) – 0.96 1.861 0.12(.131) 0.43 (.243) 0.61 (.399) 0.97 1.732 Uttar Pradesh 0.30 (.404) 0.72* (.156) – 0.99 1.965 -0.39(.327) 0.58*(.109) 0.67*(.180) 0.99 2.057 Andhra 0.14 (.103) 1.05* (.063) – 0.99 1.723 Pradesh 0.06(.099) 0.70*(.174) 0.46*(.216) 0.99 1.762

    Notes: (1) Standard errors are given in the parenthesis.

    (2) ‘*’ indicates significance at the 5 per cent level. The specifications used are same as Table 2A.

    because of the decline in diversification in the industrial base failing to generate fresh demand for services?

    The state of West Bengal, on the contrary, shows a significant elasticity for agriculture in the initial period and an insignificant one in the later period. Though the rate of growth of agriculture has declined significantly in the second period compared to the first in the state of West Bengal, it is still around 3.4 per cent. The question of lack of diversification may once again arise here.

    West Bengal (WB) shows a significantly higher elasticity for the industrial sector in the second period compared to the first. This state has also achieved a significantly high rate of growth in the services sector in the second period. Over the last three decades, WB was not doing well at all in the industrial sector. However, a significant improvement in industrial performance can be traced in the later years of our consideration. A notable improvement over a low base might have generated significant demand for services in the state, for the rest of the Indian economy has no role to play in this state consistently over the years of concern. We come back to these issues shortly.

    While both agriculture and industry of the state are insignificant in generating demand for services in Kerala during the 1980s the commodity-producing sector of the rest of the Indian economy plays a significant role for the concerned period, but the other two variables turn out to be insignificant. If we try to drop the rest of the Indian economy from our analysis, the specification suffers from a problem of positive autocorrelation for the initial period. Thus, there is a possibility that rest of the Indian economy does play an important role in the case of Kerala, while the state’s own economy does not. But it requires further research to say anything conclusive. The situation, however, changes somewhat in the second period when the state’s own industrial growth rate improves marginally and the elasticity for services with respect to industry turns significant.

    IVIVIVIVIV
    Some InferencesSome InferencesSome InferencesSome InferencesSome Inferences

    The behaviour of the elasticities in different states leads to certain observations. First, the variable, “rest of the Indian economy” is significant only in some states. This calls for an explanation. Secondly, industrial activities of a state turn out to be the most important factor in determining the state’s services sector growth. But there are a few exceptions too, which demand attention. One important point to be mentioned here is that there is no significant change in the elasticities with respect to industry in the second phase of liberalisation when compared to the initial phase with some exceptions. This is interesting for it suggests that the services sector growth is a mere translation of the high growth in the industry in the last two decades. Incidentally, it is hard to find any significant improvement in the industrial sector performance of the industrially developed states in the country in the later phase when compared to the initial one. It is some of the industrially less developed states that have shown significant improvement in the industrial growth rates during the liberalised trade regime.

    The rest of the Indian economy becomes significant for a specific state in terms of generation of demand for services for three dominant reasons. First, in a specific state, labour cost can be higher for certain services as a result of strong trade union practices compared to its neighbouring states. If the transportation cost is not higher than the cost of getting the services at home, entrepreneurs are likely to prefer the services from the neighbouring states. Secondly, industrially more developed states are likely to have better infrastructure for industry related services, activities which are not available in a newly industrialising state. Moreover, infrastructure in general, in a more developed state is usually much better than in the less developed ones. Here infrastructure assumes the nature of a public good and requires heavy expenditure that is usually taken up by the government. This includes roads, electricity, etc. All these three factors can be clubbed under the supply side determinants of services sector

    Table 3B: Services Sector Elasticities with Respect toTable 3B: Services Sector Elasticities with Respect toTable 3B: Services Sector Elasticities with Respect toTable 3B: Services Sector Elasticities with Respect toTable 3B: Services Sector Elasticities with Respect to
    Agriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture Plus
    Industry of Rest of the Indian Economy for the Medium GrowthIndustry of Rest of the Indian Economy for the Medium GrowthIndustry of Rest of the Indian Economy for the Medium GrowthIndustry of Rest of the Indian Economy for the Medium GrowthIndustry of Rest of the Indian Economy for the Medium Growth
    States for the Years 1993-94 to 2002-03States for the Years 1993-94 to 2002-03States for the Years 1993-94 to 2002-03States for the Years 1993-94 to 2002-03States for the Years 1993-94 to 2002-03

    States Agriculture Industry Rest R-squared DW Stat

    Himachal 0.22 (1.077) 0.81* (.184) – 0.90 1,204 Pradesh 0.95(1.239) 0.02 (.721) 1.75(1.554) 0.92 1.143 Tamil Nadu 0.13 (.336) 1.68* (.218) – 0.89 1.498

    -0.02(.223) 0.45 (.402) 1.67*(.513) 0.96 1.248 Madhya 0.03 (.138) 0.76* (.064) – 0.95 1.658 Pradesh 0.01(.105) 0.32(.178) 0.92*(.364) 0.98 1.945 Orissa -0.06(.294) 1.29*(.148) – 0.92 1.776 -0.06(.241) 0.72*(.298) 0.84*(.399) 0.95 1.453 Uttar Pradesh 0.52 (.375) 0.69* (.194) – 0.93 1.118 0.03(.449) 0.03 (.444) 1.22 (.750) 0.95 0.880 Andhra 0.30* (.107) 1.07* (.068) – 0.99 1.637 Pradesh 0.31*(.105) 1.25*(.173) -0.33 (.285) 0.99 2.081

    Notes: (1) Standard errors are given in the parenthesis.

    (2) ‘*’ indicates significance at the 5 per cent level. The specifications used are same as Table 2A.

    Table 4A: Services Sector Output Elasticities with Respect toTable 4A: Services Sector Output Elasticities with Respect toTable 4A: Services Sector Output Elasticities with Respect toTable 4A: Services Sector Output Elasticities with Respect toTable 4A: Services Sector Output Elasticities with Respect to
    Agriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture Plus
    Industry of Rest of the Indian Economy for the Low GrowthIndustry of Rest of the Indian Economy for the Low GrowthIndustry of Rest of the Indian Economy for the Low GrowthIndustry of Rest of the Indian Economy for the Low GrowthIndustry of Rest of the Indian Economy for the Low Growth
    States for the Years 1980-81 to 1992-93States for the Years 1980-81 to 1992-93States for the Years 1980-81 to 1992-93States for the Years 1980-81 to 1992-93States for the Years 1980-81 to 1992-93

    States Agriculture Industry Rest R-squared DW Stat

    Bihar -0.00 (.143) 0.93* (.081) – 0.95 1.825 0.04(.116) 0.48*(.183) 0.56*(.216) 0.97 2.001 Kerala -0.08 (.574) 1.26* (.475) – 0.86 0.668 0.10(.270) -0.00(.304) 1.08*(.178) 0.97 2.193 West Bengal 0.32* (.096) 0.78* (.126) – 0.98 1.422 0.29*(.104) 0.63*(.216) 0.19(.219) 0.98 1.355 Assam 1.60* (.306) 0.29 (.180) – 0.95 1.786 0.86(.437) 0.17(.165) 0.52*(.245) 0.96 1.363 Punjab -0.09 (.178) 0.70* (.122) – 0.99 2.763 -0.09(.191) 0.68*(.178) 0.03(.159) 0.99 2.789

    Notes: (1) Standard errors are given in the parenthesis.

    (2) ‘*’ indicates significance at the 5 per cent level. The specifications used are same as Table 2A.

    Table 4B: Services Sector Output Elasticities with Respect toTable 4B: Services Sector Output Elasticities with Respect toTable 4B: Services Sector Output Elasticities with Respect toTable 4B: Services Sector Output Elasticities with Respect toTable 4B: Services Sector Output Elasticities with Respect to
    Agriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture PlusAgriculture and Industry of a State, and Agriculture Plus
    Industry of Rest of the Indian Economy for the Low GrowthIndustry of Rest of the Indian Economy for the Low GrowthIndustry of Rest of the Indian Economy for the Low GrowthIndustry of Rest of the Indian Economy for the Low GrowthIndustry of Rest of the Indian Economy for the Low Growth
    States for the Years 1993-94 to 2002-03States for the Years 1993-94 to 2002-03States for the Years 1993-94 to 2002-03States for the Years 1993-94 to 2002-03States for the Years 1993-94 to 2002-03

    States Agriculture Industry Rest R-squared DW Stat

    Bihar 0.56* (.115) 0.68* (.063) – 0.97 1.711 0.51*(.092) 0.53*(.080) 0.47*(.200) 0.99 1.609 Kerala -0.28 (.193) 1.52* (.123) – 0.97 2.597 -0.55*(.200) 0.74 (.375) 0.92*(.428) 0.99 2.249 West Bengal -0.33 (.408) 1.74* (.241) – 0.99 1.412 -0.04(.408) 1.95*(.254) -0.69 (.423) 0.99 2.964 Assam 1.72 (1.342) 1.59* (.755) – 0.56 1.013

    0.69 (.953) -1.01(.973) 1.55*(.494) 0.83 2.964

    Punjab 0.71* (.224) 0.87*(.094) – 0.99 2.363 0.67*(.266) 0.80*(.214) 0.12(.359) 0.99 2.280

    Notes: (1) Standard errors are given in the parenthesis.

    (2) ‘*’ indicates significance at the 5 per cent level. The specifications used are same as Table 2A.

    Economic and Political Weekly July 8-15, 2006

    growth and need to be addressed at the state level separately. The complex interplay of all these factors outside and inside a state finally determines the ability of the domestic services sector to attract demand from both inside and outside the state.

    Let us take up the example of West Bengal. We have noticed that the elasticity with respect to the rest of the Indian economy is consistently insignificant for this state. On the contrary, the elasticity with respect to rest of the Indian economy for the three neighbouring states of Bihar, Orissa and Assam is significant at least in one period under consideration. It is well known that labour is historically much more organised in West Bengal and likely to be more costly compared to its neighbouring states. Incidentally, the government-stipulated minimum wages in West Bengal are much higher than that of the three other states mentioned above. In this context, the concept of rest of the Indian economy in our analysis looks too global. It may make more sense if we consider the impact of the commodity-producing sector of the neighbouring states on the services of a specific state in a region.

    Interestingly, our analysis shows while rest of the Indian economy does not have any role to play in the case of Maharashtra in the initial phase of liberalisation, it does matter in determining the services sector growth in the state in the later phase of liberalisation. On the contrary, the rest of the Indian economy plays a significant role in determining services growth in the state of Gujarat in the first period. These two neighbouring states are alike on several counts, having similar kinds of social and physical infrastructure facilities [Ghosh et al 2004]. Moreover, labour conditions are also more or less the same in these two states [Chakravarty 2003]. Is it then a historically determined factor which explains the different behaviour of state services sectors towards the rest of the Indian economy in the first period?

    What happens in the second period? In the later phase of liberalisation Maharashtra shows a significant decline in the rate of growth of industry when compared to the initial phase. This might have thwarted the usual demand generation for services in the industrial sector in this state. However, Maharashtra is historically one of the most industrially developed states of India. Consequently the service facilities related to the industrial sector are expected to be highly developed in this state as well. A significant decline in the state’s own industrial sector demand for services might have led to a fall in the prices of those services in the state. This, in turn, probably generated a fresh demand in the rest of the Indian economy for the services of the state of Maharashtra. However, this is only a speculation; a detailed state level analysis of the cost and price structures can alone confirm this hypothesis. Unfortunately, it is not possible to say anything about the impact of the rest of the Indian economy in the case of Gujarat in this period for the reasons discussed above.

    The second issue emanating from our analysis is the industrycentric evolution of the services sector. To this end, Tables 5 and 6 may provide us some primary insights. The combined facts from Tables 5 and 6 reveal that the rate of growth of agriculture in every state is considerably lower compared to the industrial sector. It is likely that if the rate of growth of a sector is not high enough, the demand generated by that sector for another will not be significant, especially when the interlinkage between the two sectors is not very strong. However, the rate of growth of agriculture is not so low in all states. West Bengal has achieved the highest rate of growth of agriculture in the first period and consequently shows significant output elasticity of services with respect to agriculture.

    Table 5: Annual Rate of Growth (in Per Cent) of the AgriculturalTable 5: Annual Rate of Growth (in Per Cent) of the AgriculturalTable 5: Annual Rate of Growth (in Per Cent) of the AgriculturalTable 5: Annual Rate of Growth (in Per Cent) of the AgriculturalTable 5: Annual Rate of Growth (in Per Cent) of the Agricultural
    Sector in Different States of IndiaSector in Different States of IndiaSector in Different States of IndiaSector in Different States of IndiaSector in Different States of India

    States Ranked in 1980-81 to 1992-93 1993-94 to 2002-03 Descending Order on the Basis of the Growth Residual Growth Residual Services Sector Rate Mean Square Rate Mean Square Growth Rates Achieved in the Period 1980-81 to 1992-93

    Rajasthan 4.6* (1.19) .026 0.9* (1.70) .024 Maharashtra 3.5* (0.91) .015 1.7* (0.57) .003 Haryana 4.3* (0.57) .006 1.5* (0.35) .001 Karnataka 2.9* (0.39) .003 3.1* (0.82) .006 Gujarat 0.5 (1.82) .061 -0.89 (2.09) .036 Himachal Pradesh 2.2* (0.66) .008 1.3* (0.27) .001 Tamil Nadu 4.0* (0.69) .006 1.1* (1.04) .009 Madhya Pradesh 2.1* (0.39) .003 -0.8 (1.14) .011 Orissa 0.5 (0.01) .013 -0.6 (0.83) .006 Uttar Pradesh 2.7* (0.18) .001 2.2* (0.44) .002 Andhra Pradesh 2.3* (0.63) .007 2.7*(0.87) .006 Bihar 1.6* (0.78) .011 3.0* (1.28) .014 Kerala 2.8* (0.52) .005 -1.9* (0.81) .005 West Bengal 5.2* (0.47) .004 3.4* (0.34) .001 Assam 2.2* (0.20) .001 0.51 (0.29) .001 Punjab 4.6* (0.21) .001 2.1* (0.31) .001

    Notes: (1) * indicates significance at 5 per cent level.

    (2) Standard errors are given in the parenthesis.

    Table 6: Rate of Growth (in Per Cent) of the Industrial Sector inTable 6: Rate of Growth (in Per Cent) of the Industrial Sector inTable 6: Rate of Growth (in Per Cent) of the Industrial Sector inTable 6: Rate of Growth (in Per Cent) of the Industrial Sector inTable 6: Rate of Growth (in Per Cent) of the Industrial Sector in
    Different States of India in the 1980s and 1990sDifferent States of India in the 1980s and 1990sDifferent States of India in the 1980s and 1990sDifferent States of India in the 1980s and 1990sDifferent States of India in the 1980s and 1990s

    States Ranked in 1980-81 to 1992-93 1993-94 to 2002-03 Descending Order on the Basis of the Growth Residual Growth Residual Services Sector Growth Rate Mean Square Rate Mean Rates Achieved in the Square Period 1980-81 to 1992-93

    Rajasthan 7.5* (0.31) .002 7.1* (0.85) .006 Maharashtra 6.2* (0.31) .002 2.9* (0.94) .007 Haryana 8.0* (0.53) .005 6.4* (0.19) .000 Karnataka 6.7* (0.31) .002 6.7* (0.77) .005 Gujarat 7.2* (0.52) .005 6.6* (0.72) .004 Himachal Pradesh 7.5* (0.59) .006 8.5* (0.70) .004 Tamil Nadu 4.8* (0.37) .003 4.2* (0.61) .003 Madhya Pradesh 6.2* (0.63) .007 6.9* (0.71) .004 Orissa 7.0* (0.57) .006 4.6* (0.01) .002 Uttar Pradesh 7.0* (0.34) .002 4.5* (0.63) .003 Andhra Pradesh 5.0* (0.19) .001 5.8*(0.24) .000 Bihar 5.2* (0.42) .003 8.0* (1.13) .010 Kerala 3.6* (0.51) .004 4.6* (0.30) .001 West Bengal 4.1* (0.25) .001 5.9* (0.19) .000 Assam 3.5* (0.55) .005 1.4* (0.36) .001 Punjab 6.8*(0.20) .001 5.2* (0.43) .002

    Notes: (1) * indicates significance at 5 per cent level.

    (2) Standard errors are given in the parenthesis.

    Table 7: Services Sector Employment and ProductivityTable 7: Services Sector Employment and ProductivityTable 7: Services Sector Employment and ProductivityTable 7: Services Sector Employment and ProductivityTable 7: Services Sector Employment and Productivity
    Elasticity with Respect to the Outputs of the CommodityElasticity with Respect to the Outputs of the CommodityElasticity with Respect to the Outputs of the CommodityElasticity with Respect to the Outputs of the CommodityElasticity with Respect to the Outputs of the Commodity
    Producing SectorsProducing SectorsProducing SectorsProducing SectorsProducing Sectors

    Year Agriculture Industry Agriculture R squared

    + Industry

    1981 Employment 0.292 (.184) 0.640*(.148) – 0.91 – – 0.981*(.099) 0.89 Productivity -0.022 (.126) 0.078 (.101) – 0.11 – – 0.069 (.061) 0.09 1991 Employment 0.334 (.207) 0.558*(.168) – 0.88 – – 0.915*(.099) 0.87 Productivity -0.126(.120) 0.159 (.095) – 0.21 – – 0.069 (.060) 0.09

    Note: The calculations are based on the Census data for employment (see text for details).

    Economic and Political Weekly July 8-15, 2006

    3066 Economic and Political Weekly July 8-15, 2006

    What about the other states showing comparatively high growth rates in agriculture? Again, it seems that in order to generate demand for services in agriculture, while a high rate of growth is necessary, it may not be sufficient always. What is likely to be important too is the pattern of growth. Let us take up the case of Rajasthan and Maharashtra. The residual mean squares for the agricultural growth rates of Rajasthan shows the value of

    0.026 for the first period. In the case of Maharashtra it is 0.015. On the contrary, West Bengal shows a much smaller value for residual mean squares in both the periods, indicating a more stable growth experience in the state. Incidentally, it is worth mentioning here that the numerical values for the residual mean squares for the industrial growth rates are generally much lower when compared to the agricultural growth rates. However, by this logic it is difficult to explain the cases of Punjab, Haryana and also WB for the later phase of liberalisation. It is likely that the lack of diversification in agriculture in the case of West Bengal and Haryana thwarts the dynamism between the sectors of agriculture and services. Therefore, it is again essential to look at the nature and composition of agricultural growth to track down the sources of demand for services.

    Finally, let us now turn to the employment question. It has been noted that services sector growth has failed miserably to achieve a significant shift of employment out of agriculture in India. An obvious question comes to mind: Is it because of capitalintensive technological modernisation as it has happened in the case of manufacturing? As a result of non-availability of time series data for employment in India, we did a cross-sectional analysis for the years 1981 and 1991. We have tried to estimate the output elasticities of services sector employment and labour productivity with respect to the commodity-producing sector of the economy. In other words, we try to understand the effect of the demand generated in the commodity-producing sector on employment and labour productivity in services. The elasticities are reported in Table 7.

    Table 7 clearly shows that the increase (decrease) in the demand for services in the commodity-producing sector of the economy has no effect on labour productivity. This is suggestive of no significant technological improvement. On the contrary, the output elasticity of employment turns out to be significant for industry, separately as well as for the commodity producing sector as a whole. However, both sets of elasticities have experienced a decline in the year 1991 when compared to the year 1981. Incidentally, just like the output elasticities at the state level in the time series exercise, the output elasticity of employment for agriculture in the cross section analysis turns out to be insignificant.

    To conclude, leaving some loose ends aside, this analysis does give a handle to begin to understand a vast and complex problem of the Indian economy. While services sector growth in the recent era is a common experience of all the Indian states, the nature and determinants are not exactly the same. The industrial sector turns out to be the most important determinant of services sector growth in different states. However, the analysis suggests that it is necessary to have a high and stable rate of growth as well as diversification over time in the commodity-producing sector to foster growth in services. Under these conditions, agriculture is also likely to generate substantial demand for services. Interregional divergences in the development experience thus become highly influential in this context. The rest of the Indian economy’s commodity producing sector may play an important role in determining the services sector performance of a specific state depending upon inter-regional disparities in labour cost, infrastructure facilities, etc, that basically relate to supply side issues. Incidentally, except for the services sector, hardly any significant change can be identified in the growth performances in the other two sectors in the post reform period, including industry, when compared to the early reform era of the 1980s. This is true for the output elasticities of services output as well. If this is so, it is essential to take a hard look at the macro policies that promoted a more open and liberal economy, particularly after the early 1990s.

    m

    Email: deepcv@yahoo.com

    NotesNotesNotesNotesNotes

    [ I am primarily grateful to N Krishnaji for helping me conceptualise this paper. I am also grateful to C S Murty for his observations made in the CESS faculty seminar.]

    1 The theory of stages of economic growth [Fisher 1939; Clark 1940 and Rostow 1953] is based on two straightforward assumptions on production and consumption in order to link economic growth to the development of a service economy. As far as production is concerned, productivity gains are assumed to be higher in industry than in the service sector. As for consumption, the income elasticity of demand for services has to be greater than that of demand for goods. These propositions explain the forms rather than the causes of economic growth.

    2 Preparation of these tables involves a huge amount of data collection and makes it impossible to acquire this information annually. As a consequence, the results based on these tables are generally static and relate mainly to the reference period alone.

    3 It is worth mentioning here that this study is basically an explorative analysis and has not attempted any precise econometric estimation. 4 Gross domestic product of a country refers to the value of final goods and services produced within the geographical area of the country. 5 All the coefficients are statistically insignificant with a high R squared

    value of 0.95.

    ReferencesReferencesReferencesReferencesReferences

    Bhattacharya, B B, N R Bhanumurthy, Sabyasachi Kar, S Sakthivel (2004): ‘Forecasting State Domestic Product and Inflation: Macro-econometric Model for AP, Karnataka and UP’, Economic and Political Weekly, Vol 39, No 31, pp 3541-50.

    Chakravarty, Deepita (2003): ‘Technological Change, Organisation of Work and Employment Contracts: Emergence of Decentralised Bargaining in the Indian Textiles’, Unpublished PhD Dissertation, Jawaharlal Nehru University, New Delhi.

    Clark, C (1940): The Conditions of Economic Progress, Macmillan, London, revised edition, 1957. Fisher, A G B (1939): ‘Production, Primary, Secondary and Tertiary’,Economic Record, 15, June, pp 24-38.

    Ghosh, Buddhadeb and Prabir De (2004): ‘How Do Different Categories of Infrastructure Affect Development? Evidence from Indian States’, Economic and Political Weekly, Vol 39, No 42, pp 4645-57.

    Gordon, James, and Poonam Gupta (2004): ‘Understanding India’s Services Revolution’, IMF Working Paper No wp/04/171, International Monetary Fund, Washington DC.

    Hansda, Sanjay K (2001): ‘Sustainability of Services-led Growth: An Input-Output Analysis of the Indian Economy’, Reserve Bank of India Occasional Papers 22, Nos 1, 2 and 3, pp 73-118.

    Rostow, W W (1953): The Process of Economic Growth, Oxford University Press, Oxford.

    Sastry, D V S, Balwant Singh, Kaushik Bhattacharya and N K Unnikrishnan (2003): ‘Sectoral Linkages and Growth Prospects: Reflections on the Indian Economy’, Economic and Political Weekly, Vol 38, No 24, pp 2390-97.

    Dear reader,

    To continue reading, become a subscriber.

    Explore our attractive subscription offers.

    Click here

    Comments

    (-) Hide

    EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

    Back to Top