TELECOM REFORMS AND STATE LEVEL ECONOMIC DEVELOPMENT IN INDIA

TELECOM REFORMS AND STATE LEVEL ECONOMIC DEVELOPMENT IN INDIA



Sudeshna Ghosh Banerjee
Europe and Central Asia Region, The World Bank
Sghosh1@worldbank.org
202-458-4150

Shreyasi Jha
Department of Public Policy, University of North Carolina at Chapel Hill
sjha@email.unc.edu
919-929-4994










Abstract: Telecom reforms have been among the most visible face of reforms in India in the past decade. In this paper we analyze the impact privatization and competition in the telecom sector on state level economic development using a comprehensive state-level data set from India. Though we do not find systematic evidence of the positive effect of telecom reforms on state level economic development, we do find moderate support in favor of telecom privatization on industrial productivity. Lack of a longer time series since NTP (1999) could be a possible explanation for this ambiguity in impact of telecom reforms.






Keywords: India; Telecom, Privatization; State-level economic development










INTRODUCTION
Telecom reforms’ across the world are energizing businesses and people. Long considered a natural monopoly, recent technological developments have facilitated competition in this sector leading to increased access to telecom services and gains in efficiency and quality of service. India has emerged as an international destination for processing and distribution of information. Availability of infrastructure for electronically transferring and assessing information are critical to maintaining the competitive advantage that it currently enjoys and embracing telecom reforms is a part of achieving that goal. Though the results of telecom reforms the world over have been positive on average; domestic political economy and institutions have impacted every country experience and India is no exception. In this paper, we examine India’s progress with telecom reforms and the impact of these reforms on state level economic development.

India’s economic liberalization program began in 1991. Aimed at raising the economy from low-growth equilibrium and putting it on a sustained growth path, these reforms targeted a wide range of sectors – from international trade to finance and infrastructure. Following these reforms, traditional sectors such as agriculture and industry that used to contribute nearly 70 per cent of the GDP now constitute an increasingly smaller share of economic output. Simultaneously, there has been rapid growth in the service sector that now contributes nearly 48 per cent to GDP (1998) and is growing at 8 per cent per annum.

Indian policy-makers also increasingly recognized the need for reforms and investment in telecommunication infrastructure in order to realize the forecasted economic and social growth rate of the country. It has been shown that investment in telecommunications infrastructure leads to economic growth in various ways. While telecommunication investment itself leads to growth by creating a demand for the goods and services used in their production, the economic returns on this investment are far greater than the returns from the investment alone. The multiplier effect of telecom investment on GDP is likely to be higher because of both the direct and indirect effect that this investment has on production . However, it has also been shown that since telecommunications infrastructure is characterized by network externalities the positive growth effects of investment in this sector are subject to having achieved a critical mass in a given country’s communications infrastructure (Roller and Waverman, 2000). Lucas (1988) sees familiarity with information technology as a capability-enhancing activity. In his view, an “important determinant of economic development is not just what people do but how much their activities enhance future growth through learning by doing.” Information technology, like other capability-enhancing activities help build capabilities for the future.

In sum, telecom reforms have spillover effect on economic and social development. Major reforms in the Indian telecom sector were initiated five years ago under National Telecom Policy, 1999 (NTP 99). Using India’s telecom reforms as a case study this paper examines the spillover effect of telecom reforms on state-level economic development. This is one of the first studies to rigorously assess the impact that telecom reforms in India on economic development in the states. In the rest of this paper, we review the institutional history of the telecom sector in India in section I, Section II presents the conceptual framework and the hypotheses and section III discusses the empirical strategy. In section IV, we present the results and in section V, we discuss and present some tentative conclusions.

I. INSTITUTIONAL HISTORY OF THE TELECOM SECTOR IN INDIA
I.1 Main policy organizations in telecom sector
The telegraph act of 1885 governed the telecommunications sector. Under this act, the government was in-charge of policymaking and provision of services (Dossani, 2002). Major changes in telecommunications in India began in the 1980s. Under the Seventh Plan (1985-90), 3.6 percent of total outlay was set aside for communications and since 1991, more than 5.5 percent is spent on it (Figure 1). The initial phase of telecom reforms began in 1984 with the creation of Center for Department of Telematics (C-DOT) for developing indigenous technologies and private manufacturing of customer premise equipment. Soon after, the Mahanagar Telephone Nigam Limited (MTNL) and Videsh Sanchar Nigam Limited (VSNL) were set up in 1986. The Telecom Commission was established in 1989.

(Figure 1 here.)

When telecom reforms were initiated in 1994, there were three incumbents in the fixed service sector, namely DoT (Department of Telecom), MTNL and VSNL. Of these, DoT operated in all parts of the country except Delhi and Mumbai. MTNL operated in Delhi and Mumbai and VSNL provided international telephony. (Singh, et al. 1999)

Given its all-India presence and policy-making powers, the DoT enjoyed a monopoly in the telecom sector prior to the major telecom reforms. However, subsequent to the second phase of reforms in 1999, which included restructuring the DoT to ensure a level playing field among private operators and the incumbent, the service-providing sector of DoT was split up and called Department of Telecom Services (DTS). DTS was later corporatized and renamed Bharat Sanchar Nigam Limited (BSNL). This meant separation of the incumbent service provider from the policy-maker. Broadly, DoT is now responsible for policy-making, licensing and promotion of private investments in both telecom equipment and manufacture and provision of telecom services. BSNL, a corporate body, is responsible for the provision of services (Singh, et al. 1999).

A crucial aspect of the institutional reform of the Indian telecom sector was setting up of an independent regulatory body in 1997 – the Telecom Regulatory Authority of India (TRAI), to assure investors that the sector would be regulated in a balanced and fair manner. TRAI has been vested with powers to ensure its independence from the government. The government has retained the licensing function with itself . The main issue with respect to licensing has not been whether it should be with the regulator but that the terms and conditions of licensing should involve consultations with TRAI to ensure transparency in the bidding process Some of the main functions of TRAI include fixing tariffs for telecom services, dispute-settlement between service providers, protecting consumers through monitoring of service quality and ensuring compliance to license conditions, setting service targets and pricing policy for all operators and service providers. (Gupta, 2002)

Further changes in the regulatory system took place with the TRAI Act of 2000 that aimed at restoring functional clarity and improving regulatory quality. TRAI can frame regulations and can levy fees and charges for telecom services as deemed necessary. The regulatory body also has a separate fund (called the TRAI General Fund) to facilitate its functioning. To fairly adjudicate any dispute between licensor and licensee, between service provider, between service provider and a group of consumers, a separate disputes settlement body was set up called Telecom Disputes Settlement and Appellate Tribunal (TDSAT) (DoT Annual Report, 2002).

I.2 Major policies shaping the telecom reforms – NTP 94 and NTP 99

Reforms in the Indian telecom sector have taken place within the framework of two Telecom Policy Statements, the National Telecom Policy (NTP) 1994 and the National Telecom Policy (NTP) 1999 . NTP 1994 marked a significant shift in the policy orientation towards telecommunications in that, telecom services were henceforth to be treated not as a luxury item but as necessity that should be available to all. While Universal Service Obligation (USO) was emphasized in the NTP 1994, the statement was designed with the approach that services should continue to be provided largely by a strong incumbent that faced little competition. Also, major targets for NTP 1994 were set without an accurate resource assessment.

The NTP 1999 sought to improve upon the 1994 policy by clearly specifying policy objectives and setting targets for the USO. NTP 1999 aimed for a teledensity target of 7 per cent per hundred by the year 2005 (and 15 per cent by 2015). For rural areas, the target was set to increase from 0.4 per cent in 1999 to 4 per cent during the same period. It therefore emphasized the achievement of targets along with a “transition to higher technologies” and an improvement in the spread of telecom services to all parts of the country. The strategies included a continuing role for the public sector particularly with regard to telecom service provision in rural areas and, recognizing the role of private investment in meeting these targets (Dossani, 2002). NTP 1999 allowed for multiple operators in the market with a level playing field between incumbents and private providers: DoT and MTNL now had to pay a license fee but DoT’s license fee was refunded because it had met the USO targets (Singh, et. al. 1999). NTP 1999 also introduced open competition for several services including basic services, national long distance and international telephony.

I.3 Progress of reforms
a. Private Participation in Telecom - For the provision of basic services, the entire country was divided into 21 telecom circles, excluding Delhi and Mumbai (Singh et. al. 1999) . With telecom markets opened to competition, DoT and MTNL were joined by private operators but not in all parts of the country. By mid-2001, all six of the private operators in the basic segment had started operating (Table 1). Table 2 shows the number of village public telephones issued by private licensees by 2002.

(Tables 1 and 2 here.)

After a recent licensing exercise in 2002, there exists competition in most service areas. However, the market is still dominated by the incumbent. In December 2002, the private sector provided approximately 10 million telephones in fixed, WLL (Wireless Local Loop) and cellular lines compared to 0.88 million cellular lines in March 1998 (DoT Annual Report, 2002). 72 per cent of the total private investment in telecom has been in cellular mobile services followed by 22 per cent in basic services. After the recent changes, the stage is now set for greater competition in most service areas for cellular mobile Over time, the rise in coverage of cellular mobile will imply increased competition even for the basic service market because of competition among basic and cellular mobile services.

b. Teledensity and Village Public Phones (VPTs) - India’s rapid population increase coupled with its progress in telecom provision has landed India’s telephone network in the sixth position in the world and second in Asia (ITU). The much publicized statistic about telecom development in India is that in the last five years, the lines added for basic services is 1.5 times those added in the last five decades! The annual growth rate for basic services has been 22 percent and over 100 percent for internet and cellular services (Parameswaran, 2003). As Dossani (2002) argues, the comparison of teledensity of India with other regions of the world should be made keeping in mind the affordability issues. Assuming households have a per capita income of $350 and are willing to spend 7 percent of that total income on communications, then only about 1.6 percent of households will be able to afford $30 (for a $1000 investment per line) (Dossani, 2002).

Teledensity has risen to 4.9 phones per 100 persons in India compared to the average 7.3 mainlines per 100 people around the world. Figure 2 shows the growth rate of fixed and cellular mobile subscription between 1998 and 2002. Although, the coverage is still much higher in urban areas - 13.7 in urban areas compared to1.4 in rural areas, the government has made efforts to connect villages through village public telephones (VPT) and Direct Exchange Lines (DEL). This coverage increased from 4.6 lakhs in March 2002 to 5.10 lakhs in December 2002 for VPT and from 90.1 lakhs in March to 106.6 lakhs in December 2002 for DELs. BSNL has been mainly responsible for providing VPTs; more than 84 percent of the villages were connected by 503610 VPTs with private sector also providing 7123 VPTs (DoT Annual Report, 2002).

(Figures 2, and 3 here.)

The overall telecom growth rate is likely to be high for some years, given the increase in demand as income levels rise and as the share of services in overall GDP increases. The growth rate will be even higher due to the price decrease resulting from a reduction in cost of providing telecom services. A noteworthy feature of the growth rate is the rapid rate at which the subscriber base for cellular mobile has increased in the last few years of the 1990s, which is not surprising in view of the relatively lower subscriber base for cellular mobile (table 3).

(Table 3 here.)

c. Foreign Participation – India has opened its telecom sector to foreign investors up to 100 percent holding in manufacturing of telecom equipment, internet services, and infrastructure providers (e-mail and voice mail), 74 percent in radio-paging services, internet (international gateways) and 49 percent in national long distance, basic telephone, cellular mobile, and other value added services (FICCI, 2003). Since 1991, foreign direct investment (FDI) in the telecom sector is second only to power and oil - 858 FDI proposals were received during 1991-2002 totaling Rs. 56,279 crores (Figure 4) (DoT Annual Report, 2002). Foreign investors have been active participants in telecom reforms even though there was some frustration due to initial dithering by the government. Until now, most of the FDI has come in the cellular mobile sector partly due to the fact that there have been more cellular mobile operators than fixed service operators. For instance, during the period 1991-2001, about 44 percent of the FDI was in cellular mobile and about 8 percent in basic service segment. This total FDI includes the categories of manufacturing and consultancy and holding companies (Figure 5).

(Figures 4 and 5 here.)

d. Tariff-setting - An essential ingredient of the transition from a protected market to competition is the alignment of tariffs to cost-recovery prices. In basic telecom for example, pricing of the kind that prevailed in India prior to the reforms, led to a high degree of cross-subsidization and introduced inefficient decision-making by both consumers and service-providers. Traditionally, DoT tariffs cross-subsidized the costs of access (as reflected by rentals) with domestic and international long distance usage charges (Singh et. al. 1999). Therefore, re-balancing of tariffs - reducing tariffs that are above costs and increasing those below costs - was an essential pre-condition to promoting competition among different service providers and efficiency in general.

TRAI issued its first directive regarding tariff-setting following NTP 99 aimed at re-balancing tariffs and to usher in an era of competitive service provision. Subsequently, it conducted periodic reviews and made changes in the tariff levels, if necessary. Table 4 shows the current level of telephone charges in India effective from January, 2003. Re-balancing led to a reduction in cross-subsidization in the fixed service sector. Cost based pricing, a major departure from the pre-reform scenario, also provides a basis for making subsidies more transparent and better targeted to specific social objectives, e.g. achieving the USO.

(Table 4 here.)

e. Service Quality - One of the main reasons for encouraging private participation in the provision of infrastructure rests on its ability to provide superior quality of service. In India, as in many developing countries, low teledensity resulted in great emphasis being laid on rapid expansion often at the cost of quality of service. One of the benefits expected from the private sector’s entry into telecom is an improvement in the quality of service to international standards. Armed with financial and technical resources, and greater incentive to make profits, private operators are expected to provide consumers value for their money. Telephone faults per 100 main lines came down to 10.32 and 19.14 in Mumbai and Delhi respectively in 2002-03 compared to 11.72 and 26.6 in 1997-98 (Figures 6 and 7). Quality of service was identified as an important reform agenda and TRAI has devised QOS (Quality of Service) norms that are applicable across the board to all operators (Singh et. al. 1999).

(Figures 6 and 7 here.)

II CONCEPTUAL FRAMEWORK
In this paper we examine the impact of privatization and competition in the telecom sector on state level economic performance using a comprehensive state-level data set from India. How do telecom reforms undertaken at the national level impact the economic performance of the states? As Ahluwalia (2001) notes, the final impact of reforms is a net of two forces – 1) positive efficiency effect of reforms that should raise the productivity and growth in all states and 2) inherent comparative advantage of states resulting in reallocation of scarce resources.

Telecom reforms’ are a relatively new phenomenon and have been among the most visible face of economic reforms around the globe. The introduction of privatization and competition in this sector has brought about substantial changes in output, employment and labor productivity. Early studies by Hardy (1980) and Norton (1992) and recently by Greenstein and Spiller (1996) find that telephones have a significant impact on GDP. In contrast to the single equation models used in these studies, Roller and Waverman (2000) employ a simultaneous equation model to take into account the endogeneity of telecom investment. Investment in communications infrastructure has significant impact on growth in their study. In their analysis of 21 OECD countries over a period of 21 years, they note that one-third of the growth in this period can be credited to communications infrastructure. But the reverse causality works too as growth impacts telecom investments through rise in telecom infrastructure and rise in income. Further, telecom infrastructure’s impact on growth depends on a critical mass, that is, the real gains accrue when the country has built up a substantial telecom infrastructure. The critical level in their study corresponds to a 40 percent telephone penetration rate. Developing countries still have a long way to go in achieving this and reap the potential gains from telecom investment (Roller and Waverman, 2000). This also underscores the need for investing in telecommunications for long-term sustainable growth in developing countries.

Privatization, competition and regulation have helped consolidate the reform gains with considerable impact on factor and labor productivity (Li and Xu, 2002). Wallsten (2001), in a study comprising 30 African and Latin American countries from 1984-97, found that competition (measured by mobile operators not owned by the incumbent) is associated with increase in network and connection capacity and price decline. In a similar vein, Fink, Mattoo and Rathindran (2002) found significant impact of privatization, competition, and regulation on sectoral performance in an analysis of 86 African, Asia, and Latin American countries during 1985-1999. They also found interesting evidence on sequencing of reforms; telephone penetration is lower if competition is introduced after privatization rather than simultaneously. In an earlier study of 12 Asian countries, Fink et al (2010) find higher levels of mainline availability, service quality, and labor productivity in countries where a comprehensive reform package of privatization, competition, and regulation has been implemented.

The role of regulation is critical; Wallsten (2002) finds that setting up regulatory agency prior to privatization has a positive impact on telecom investment, fixed and cellular penetration. Its presence also serves the purpose of a safety net; investors are willing to pay more for telecom firms in countries with regulatory agency. For instance, in Argentina, failure to establish a regulatory agency before privatization resulted in lower purchase price and raised the chances of buyers capturing windfall profits. Further, based on a case study on India, Dokeniya (1999) argues that attracting private investment for telecom development depends on setting up a regulatory regime that provides ‘credible commitment’. As evidenced in India, such a transition is difficult as it involves moving from a system of high level of discretionary power on the part of politicians and bureaucrats to a system of institutions based on transparency, accountability, and distance from the political process.

Though the impact of telecom reforms on sectoral performance has been assessed around the globe, their impact on individual states in a country-specific context is less than clear. This is because the effect that policies have on state-level economic development can be varied based on their respective initial conditions and institutional structures even though a broad policy is set in place and implemented by the central government. What are the conditions that allow a state to take advantage of the potential benefits of telecom reforms? Our paper addresses this gap in literature and it provides a unique opportunity to analyze how an exogenous policy change at the federal level impacts the development and productivity of individual states.

More specifically, we address the following hypotheses:
1. States with higher teledensity, networks, and exchanges will have higher state domestic product
2. States with higher teledensity, networks, and exchanges will attract greater FDI inflows
3. States with higher teledensity, networks, and exchanges have higher industrial productivity

We created a dataset consisting of information from 16 major Indian states for four years 1998-2001 , namely, Andhra Pradesh, Assam, Bihar, Gujarat, Haryana, Himachal Pradesh, Karnataka, Kerala, Madhya Pradesh, Maharashtra, Orissa, Punjab, Rajasthan, Tamil Nadu, Uttar Pradesh, and West Bengal. With cross-national analysis, country specific characteristics are often lost in generalizations and subjective indicators such as institutional variables. Sub-national analysis provides an opportunity to understand the regional impact of an exogenous policy change.

III EMPIRICAL FRAMEWORK
The following section briefly discusses the variables we have used in our analysis and the econometric specifications. To measure state level economic development, we include three dependent variables: 1) Gross state domestic product (in constant Indian Rupees); 2) Foreign direct investment (FDI); and 3) Industrial productivity as measured by net value added across states.

Factors that influence state level economic development are differences in infrastructure, labor market differentials, agglomeration effect, and institutional differences. Availability of telecom is an important aspect of infrastructural differences across states. The underlying assumption is that privatization of the telecom sector leads to more competition and greater availability of telecom infrastructure, thus leading to higher economic growth, inflows of FDI and industrial productivity. Privatization of telecom would have three effects: increase the number of people connected through telephones (teledensity); increase the total number of telephone connections (network effect) and increase in total of telephone exchanges (exchange effect). In our analysis, teledensity is measured as the number of telephones per 1000 population. State-wise number of telephone connections provided is used to capture the network effect. State-wise telephone exchanges set-up is used to capture the exchange effect. We also include planned expenditure on telecom to capture government commitment to the telecom sector.

Manufacturing wage is included in our analysis to capture the labor market differentials across Indian states. In addition to the wage rate, quality of labor force is also captured through the extent of work disruptions due to strikes and other labor disputes. We include the value of total manufacturing activity in the state to proxy for feedback or agglomeration effects whereby firms locate where hubs of economic activity already exist and therefore, affect state level economic development.

The quality and availability of infrastructure also affects economic development of a state. Power shortage (demand-supply) in a state is used to capture the effect of infrastructural differences across states. Planned expenditure on development captures a host of other factors such as the average quality of government across states. Literacy rate is included to capture of a host of factors such as quality of labor force and level of development. Summary statistics for our data are presented in table 5. Data for this study come from a variety of sources. Table 6 provides a list of variables used in this analysis and their data sources.

(Table 5 and 6 here.)

To test hypothesis 1, we examine whether state domestic product is higher in states with greater telecom privatization. We use state level data for 16 states from 1998-2001. The regression model is:
-----------------------------(1)
where GSDP is gross state domestic product in state s in year t; T is the set of variables (teledensity, network and exchange) that captures the effect of telecom reforms; and X is the set of other state characteristics that may affect GSDP. Given the small number of observations and only four year time period, we employed the pooled OLS regressions to estimate the equations after Hausman test suggested its appropriateness. If privatization of telecom sector did not have an effect on state domestic product, then we would find that β = 0.

Second, we measure whether privatization of the telecom sector had an effect on inflows of FDI into states in India. The regression model is:
-----------------------------(2)
where, FDI is inflow of FDI into state s in year t; T is the set of variables (teledensity, network and exchange) that captures the effect of telecom reforms; and X is the set of other state characteristics that may affect FDI. If privatization of telecom sector did not have an effect on inflows of FDI, then we would expect β = 0.

Finally, we measure whether privatization of the telecom sector had an effect on industrial productivity across states in India. The regression model is:
-----------------------------(3)
where, INDPROD is net valued added in s in year t; T is the set of variables (teledensity, network and exchange) that captures the effect of telecom reforms; and X is the set of other state characteristics that may affect INDPROD. If privatization of telecom sector did not have an effect on industrial productivity, then we would expect β = 0.

In addition to the above three specification, we also estimated a seemingly unrelated regression (SUREG) model because GSDP, FDI, and INDPROD may be simultaneously determined with correlated error terms. More specifically, the model we estimate is:


--------------------------(4)
where, GSDP, FDI, INDPROD, T, and X are as defined above. If privatization of telecom sector did not have an effect on state level economic development as measured by GSDP, FDI and INDPROD, then we would expect β = 0 for all three equations.

IV REGRESSION RESULTS
a. Impact of Telecom Privatization on GSDP
Table 7 presents the OLS estimates of the impact of telecom reforms on GSDP. Models (1), (2) and (3) include exchange as a measure of telecom reforms, models (4), (5) and (6) include network as a measure of telecom reforms and models (7), (8) and (9) include teledensity as a measure of telecom reforms. In models (2), (5) and (8) we include planned expenditure on telecom as an explanatory variable and in models (3), (6) and (9) we include regional dummies.

Teledensity positively impacts GSDP; 1 percent increase in teledensity has a 31 percent increase in GSDP (Model 7). Increase in the number of telephones can lead to rise in state domestic product. We also find that (lagged) planned expenditure on telecom (TELECOMEXP) has positive and significant effect on GSDP. The inference to be drawn is that government expenditure on telecom infrastructure has an immediate effect on GSDP compared to telecom output variables such as network and exchange. A possible explanation for this could be that the effects of privatization could be realized only after a time lag and four years is too soon to notice the effects of privatization on GSDP.

Power shortages have a significant negative impact on gross domestic product, reemphasizing the need for investing in public infrastructure. A one percent difference in demand and supply can mean 7 percent decline in GSDP. The value added in manufacturing (MANUFAC) appears to have a very significant negative impact on state domestic product.

(Table 7 here.)

b. Impact of Telecom Privatization on state-wise FDI
Table 8 presents the pooled OLS estimates of the impact of telecom reforms on state level FDI inflows. The specifications of models (1) – (9) are the same as in table 7. Interestingly, the coefficient on teledensity is negative suggesting a role of substitutability of public and private capital. One of the major driving forces of foreign investors is the population unserved; if the teledensity is high, foreign investors can shy away and invest in states where the demand-supply gap of telephones is higher. Value of manufacturing activity (MANUFAC) is an important determinant of state-level FDI inflows. There is some evidence that foreign investors would like to locate in states where population is more educated, one standard deviation rise in literacy rate can raise FDI by 0.4 (log) Indian Rs. (12.71*0.03).

(Table 8 here.)

c. Impact of Telecom Privatization on Industrial Productivity across states
Table 9 shows the results OLS estimates of the impact of telecom reforms on industrial productivity across states in India. We find that number of telephone connections (network effect) has a positive and significant effect on industrial productivity. A one percent increase in exchange leads to approximately 13 percent increase in industrial productivity. The value of manufacturing activity, which is used as a proxy measure of agglomeration effect in our specification, has a positive effect on industrial productivity across implying that industrial productivity is likely to be higher is states which are already hubs of industrial activity.

Not surprisingly, wages have a negative on industrial productivity but mandays lost due to labor disputes (MANLOST) has a positive impact. West Bengal leads the group of states with the highest number of mandays lost by far; we reestimated the model dropping West Bengal from the sample, even then the coefficient was positive. The workers overcompensate for the lost man days with higher productivity on the working days.

(Table 9 here.)

d. Seeming Unrelated Regression Results
Any inference based on tables 7, 8, and 9 can be erroneous because of possible simultaneity between the dependent variables. There is reason to expect that the GSDP, FDI and industrial productivity is simultaneously determined in a system of equation. Tables 10, 11 and 12 present the seeming unrelated regression (SUREG) results of equation (4) using telecom variables teledensity, network and exchange, respectively. Across all specifications, we find network and exchange have a positive and significant effect on industrial productivity. An increase in 1 percent in network leads to a 91 percent increase in industrial productivity and an increase in 1 percent in exchange leads to a 14 percent increase in industrial productivity. None of the other telecom variables have a significant effect on state economic development.

Value of manufacturing has a positive effect on FDI in tables 10 and 12. The agglomeration effect as captured by manufacturing activity is significant in table 10 and 12. Power shortage has a negative effect on GSDP, FDI and INDPROD.

(Tables 10, 11 and 12 here.)

V CONCLUSIONS
This paper has used state-level data to empirically analyze the impact of telecom privatization on economic development in the Indian context. If privatizing telecom provides a stimulus to state economic development, we would expect to see positive effect of telecom reforms on GSDP, FDI and industrial productivity. We find evidence to support the hypothesis that telecom reforms have a positive effect on industrial productivity as measured by number of exchanges and telephone network across states. The effect of network on industrial productivity is robust to different specifications. We also find that teledensity is a significant determinant of gross domestic product, that is, higher number of telephones facilitates communication and raises the state domestic product. However, our results on the effect of teledensity on state domestic product is not robust to all specifications. Telecom privatization does not appear to be an important determinant of state level FDI in India. Instead, agglomeration effect is a more important and significant determinant of FDI inflows.

A possible explanation for lack of systematic evidence of the impact of telecom privatization on state level economic development could be that it may be soon to notice any robust effect; only four years has elapsed since NTP 99. Macroeconomic variables such as GSDP and FDI tend to change in the medium to long run after a time lag. Changes in industrial productivity on the other hand, can take place in the short run. Therefore we are able to detect a positive effect of telecom privatization on industrial productivity.








References:
Ahluwalia, MS. 2002, “Economic reforms in India since 1991: Has gradualism worked?” Journal of Economic Perspectives. Volume 16(3) 67-88
Ahluwalia, MS. 2001, “State level performance under economic reforms in India” Working Paper No. 96, Center for Research on Economic Development and Policy Reform, Stanford University
Department of Telecommunications, Annual Report 2002-2003, Ministry of Communication and Information Technology, New Delhi
Department of Telecommunications “Indian Telecommunication Statistics: Policy Framework, Status and Trends”, Economic Research Unit (Statistics Wing), Ministry of Communications, New Delhi.
Dossani, R. (Ed.) 2002, Telecommunications reform in India. Quorom Books
FICCI, 2003, www.ficci.com/ficci/india-profile/ch5-pg1.htm
Fink, C, A. Mattoo, and R. Rathindran, 2001, “Liberalizing basic communications: The Asian experience” Policy Research Working Paper 2718, The World Bank
Fink, C, A. Mattoo, and R. Rathindran, 2002, “An assessment of telecommunications reform in developing countries” Policy Research Working Paper 2909, The World Bank
Greenstein, S. and PT Spiller, 1996, “Estimating the welfare effects of digital infrastructure” NBER Working Paper # W5770
Gupta, JR. 2002, “Enforcement and Dispute Settlement” Presented in APT Forum on Telecom Policy and Regulation.
Goswami, O. and Dollar, D. 2002. “Competitiveness of Indian Manufacturing: Results from a firm-level survey” Mimeo, The World Bank
Hardy, A. 1980, “The role of the telephone in economic development” Telecommunications policy, 4:278-286
Hill, A. and M.A. Abdala. 1993, “Regulation, institutions, and commitment: Privatization and regulation in the Argentine telecommunications sector” Policy Research Working Paper 1216, The World Bank
Norton, SW, 1992, “Transactions cost, telecommunications, and the microeconomics of macroeconomic growth” Economic development and cultural change, 40: 175-196
Roller, LH and L. Waverman, “Telecommunications infrastructure and economic development: A simultaneous approach” Center for Economic Policy Research Working Paper 2399, London
Singh, HV, A. Soni, and R. Kathuria, “Telecom policy reform in India” Mimeo, The World Bank
Li, W. and L.C. Xu, “The impact of privatization and competition in the telecommunications sector around the world” (Available from the first author’s website: http://faculty.darden.virginia.edu/liw/papers/Impact_of_Privatization.pdf )
Sachs, JD, N. Bajpai, and A. Ramaiah, 2002, “Understanding regional economic growth in India” CID Working Paper No. 88, Harvard University
Wallsten, S, 2002, “Does sequencing matter? Regulation and privatization in telecommunications reform” Policy Research Working Paper, The World Bank

TABLES

Table 1: Number of Licenses Issued and Number of Licensees for which Service Started by June 2001

Service Licenses Issued Service Started
Basic 6 6
Cellular Mobile 42 37
Radio Paging 137 92
Data Network 14 9
Voice Mail 41 2
E-mail 16 2
Internet 465 130
GMPCS 1 *
Source: Indian Telecommunication Statistics, 2002, DoT Economic Research Unit (Statistics Section), New Delhi


Table 2: Number of VPT Issued by 2002

Licensee and Licensed Circle/State No. of VPTs Provided
Tata Teleservices (Andhra Pradesh) 1314
Reliance Telecom (Gujarat) 2894
HFCL Infotel Ltd. (Punjab) 734
Hughes Telecom(India) Ltd. (Maharashtra) 1140
Bharati Telenet Ltd. (Madhya Pradesh) 348
Shyam Telelink Ltd. (Rajasthan) 693
Total 7123
Source: Indiastat.com


Table 3: Subscriber Base for Cellular Mobile Services

Category March '97 March '98 March '99 March '00 March '01
All Metros 325,967 551,757 519,543 795,931 1,362,592
A' Circle 9,698 176,954 354,799 585,653 1,165,778
B' Circle 3,000 138,309 288,321 460,094 932,685
C' Circle 366 15,296 36,915 42,633 116,040
All India 339,031 882,316 1,199,578 1,884,311 3,577,095
Source: Indian Telecommunication Statistics, 2002, DoT Economic Research Unit (Statistics Section), New Delhi


Table 4: Telephone Call Charges (new tariffs) in India ( as on effect from 01.04.2003, in Rupee, per metered call)

Existing TRAI Slabs Proposed by TRAI
Rural Urban Rural Urban
0-75 Free 0-60 Free 0-50 Free 0-30 Free
76-500 @0.80 61-500 @0.80 51-300 @0.80 31-300 @0.1.00
>500 @1.20 >500 @1.20 >300 @1.20 >300 @1.20
Source: TRAI

Table 5: Summary Statistics

Variable Mean Std. Dev.
Including 16 major state in India
Years: 1998-2001
Dependent variables
FDI (in million Indian Rupees) 12.60 8.81
Gross State Domestic Product (in million Indian Rupees) 12.30 17.80
Net Value Added (in 10,000 Indian Rupees) 33.35 29.82
Telecom variables
Network (in 10,000 Indian Rupees) 26.61 19.79
Teledensity 2851.25 1766.72
Exchange 126.30 159.79
Telecom Planned Expenditure (in million Indian Rupees) 18.20 18.90
Other control variables
Value of Manufacturing Output (in million Indian Rupees) 2.00 1.72
Minimum Wages (per day) 35.87 17.95
Man-days lost due to labor disputes 1599.16 2986.07
Power Shortage (demand - supply) -5.42 4.93
Literacy Rate 68.03 12.71
Planned Exp on Development (in million Indian Rupees) 14.50 7.70
Capital expenditure (in million Indian Rupees) 12.40 9.56
(Monetary values are in 1987 INR.)

Table 6: Description and source of Dependent and Explanatory Variables

Variable Definition Source
Dependent variables
GSDP Gross state domestic product India Subnational Database, The World Bank
FDI Foreign direct investment Infrastat (Indiastat.com)
VALUE_ADD Industrial productivity India Subnational Database, The World Bank
Telecom variables
TELEDENSITY Teledensity Infrastat (Indiastat.com)
EXCHANGE Number of exchanges Infrastat (Indiastat.com)
NETWORK Number of network Infrastat (Indiastat.com)
TELECOMEXP Telecommunications expenditure Infrastat (Indiastat.com)
Other control variables
MANUFAC Value added in manufacturing India Subnational Database, The World Bank
POP Population India Subnational Database, The World Bank
LITERACY Literacy rate Dept of Education, HRD Ministry, GOI
DEVEXP Development expenditure India Subnational Database, The World Bank
WAGERATE Wage rate Ministry of Labor, GOI
MANDAYS Man-days lost Ministry of Labor, GOI
ROADS Paved roads as share of total roads India Subnational Database, The World Bank
POWER Current Expenditure on Power India Subnational Database, The World Bank

Table 7: OLS results of impact of telecom reforms on GSDP

(1) (2) (3) (4) (5) (6) (7) (8) (9)
Dependent variable: Gross state domestic product
Exchange -0.011 -0.018 -0.127
(-0.11) (-0.18) (-1.42)
Network 0.003 0.024 -0.245
(-0.01) (-0.08) (-0.78)
Teledensity 0.307* -0.025 -0.314
(1.70) (-0.11) (-1.01)
Value of Manufacturing -0.395**
(2.16) -0.584**
(2.21) -0.573***
(2.80) -0.4
(-1.5) -0.605*
(1.87) -0.42
(-1.39) -0.463**
(2.46) -0.591**
(2.23) -0.539***
(2.74)
Power shortage -0.07* -0.076* -0.058* -0.066* -0.077* -0.065* -0.054 -0.079* -0.061*
(1.96) (1.87) (1.74) (1.83) (1.90) (1.97) (-1.50) (1.89) (1.95)
Capital Expenditure -0.005
(-0.03) 0.001
(-0.01) 0.051
(-0.34) -0.01
(-0.07) -0.001
(0.00) 0.063
(-0.44) 0.019
(-0.12) -0.004
(-0.03) 0.038
(-0.26)
Literacy 0 -0.01 0.001 -0.002 -0.01 0.004 -0.014 -0.009 0.01
(-0.01) (-0.95) (-0.11) (-0.19) (-0.96) (-0.38) (-1.36) (-0.88) (-0.95)
Man-days lost due to labor disputes 0.021
(-0.24) 0.006
(-0.07) 0.103
(-1.12) 0.03
(-0.30) 0

0.088
(-0.99) 0.025
(-0.28) 0.005
(-0.05) 0.072
(-0.85)
Wages 0.013 0.016 -0.015 0.013 0.016 -0.012 0.01 0.016 -0.01
(-1.54) (-1.57) (-1.48) (-1.65) (-1.55) (-1.31) (-1.40) (-1.6) (-1.22)
Lagged telecom Expenditure 0.403**
(2.37) 0.399**
(2.37) 0.405**
(2.32)
East 0.038 -0.043 -0.298
-0.14 -0.15 -0.71
West (0.445) (0.205) (0.280)
-1.15 -0.46 -0.72
North 1.617*** 1.414*** 1.434***
(3.96) (3.63) (3.57)
Constant 351.659 271.948 175.439 269.467 261.412 -15.304 230.687 252.641 20.32
(-1.52) (-1.14) (-0.81) (-1.06) (-0.86) (-0.07) (-0.99) (-0.95) (-0.1)


Observations 58 43 58 59 43 59 59 43 59
R-squared 0.21 0.39 0.42 0.19 0.39 0.42 0.22 0.39 0.42

Robust t-statistics in parentheses.
* significant at 10%; ** significant at 5%; *** significant at 1%.

Table 8: OLS results of impact of telecom reforms on FDI

(1) (2) (3) (4) (5) (6) (7) (8) (9)
Dependent variable : Foreign direct investment
Exchange -0.059 -0.135 -0.025
(-0.55) (-1.03) (-0.19)
Network 0.312 0.18 0.239
(-1.02) (-0.47) (-0.75)
Teledensity -0.497* -0.457 -1.023**
(1.81) -1.24 (2.50)
Value of Manufacturing 0.31**
(2.25) 0.506**
(2.47) 0.408**
(2.50) 0.123
(-0.57) 0.345
(-1.22) 0.222
(-0.92) 0.368***
(2.68) 0.473**
(2.51) 0.575***
(3.44)
Power Shortage -0.040 -0.044 -0.056 -0.035 -0.05 -0.046 -0.039 -0.059 -0.028
(-1.06) (-0.92) (-1.33) (-0.91) (-0.94) (-1.05) (-1.00) (-1.08) (-0.65)
Capital Expenditure 0.006 0.01 -0.039 -0.012 -0.011 -0.041 -0.109 -0.108 -0.152
(-0.03) (-0.05) (-0.21) (-0.06) (-0.04) (-0.21) (-0.55) (-0.41) (-0.77)
Literacy -0.006 0.001 0.001 -0.010 0.000 -0.004 0.017 0.022 0.035*
(-0.55) (-0.06) (-0.04) (-0.90) (-0.03) (-0.35) (-0.97) (-0.99) (1.80)
Man-days lost due to labor disputes 0.067 0.073 0.011 0.018 0.025 0.002 0.063 0.068 0.013
(-0.75) (-0.61) (-0.11) (-0.21) (-0.21) (-0.03) (-0.81) (-0.61) (-0.16)
Wages 0.007 0.006 0.012 0.01 0.009 0.011 0.009 0.011 0.009
(-1.02) (-0.74) (-1.12) (-1.22) (-0.90) (-1.18) (-1.38) (-1.14) (-1.05)
Lagged telecom Expenditure -0.044
(-0.26) -0.058
(-0.33) -0.039
(-0.22)
East 0.2 0.191 -0.599
(-0.71) (-0.70) (-1.54)
West -0.369 -0.201 -0.558*
(-1.13) (-0.62) (1.70)
North -0.227 -0.049 -0.099
(-0.38) (-0.10) (-0.21)
Constant 196.461 369.066 318.537 178.218 261.148 244.732 165.027 278.241 144.139
(-0.52) (-0.79) (-0.85) (-0.51) (-0.58) (-0.66) (-0.58) (-0.76) (-0.51)

Observations 57 42 57 58 42 58 58 42 58
R-squared 0.21 0.3 0.23 0.2 0.28 0.22 0.23 0.3 0.27

Robust t-statistics in parentheses.
* significant at 10%; ** significant at 5%; *** significant at 1%.

Table 9: OLS results of impact of telecom reforms on Industrial Productivity

(1) (2) (3) (4) (5) (6) (7) (8) (9)
Dependent variable : Industrial Value added

Exchange 0.005 0.001 0.01
(-0.23) (-0.01) (-0.44)
Network 0.136 0.147 0.131
(2.18)** (1.90)* (1.76)*
Teledensity -0.037 -0.069 -0.044
(-0.80) (-1.02) (-0.53)
Value of Manufacturing 0.918*** 0.882*** 0.91*** 0.994*** 0.968*** 1.00*** 0.928** 0.884** 0.911***
(28.36) (17.89) (28.01) (27.77) (17.35) (20.49) (30.04) (17.41) (28.22)
Power shortage -0.001 -0.003 0 -0.004 -0.006 -0.003 -0.002 -0.007 0.001
(-0.11) (-0.37) (-0.01) (-0.63) (-0.76) (-0.44) (-0.26) (-0.67) (-0.10)
Capital Expenditure 0.033 0.05 0.032 0.039 0.049 0.037 0.031 0.042 0.018
(-0.76) (-1.07) (-0.69) (-0.95) (-1.07) (-0.85) (-0.73) (-0.94) (-0.53)
Literacy -0.005* -0.008** -0.005** -0.003 -0.005** -0.003 -0.003 -0.005* -0.004
(1.92) (2.44) (2.08) (-1.54) (2.17) (-1.62) (-1.50) (1.96) (-1.63)
Man-days lost due to labor disputes -0.006 0.007 -0.002 0.015 0.031 0.013 -0.005 0.012 0.001
(-0.29) (-0.28) (-0.07) (-0.60) (-1.05) (-0.50) (-0.22) (-0.41) (-0.05)
Wages -0.004** -0.003 -0.002 -0.004*** -0.004** -0.003* -0.004** -0.003 -0.003*
(2.33) (-1.48) (0.98) (3.72) (2.38) (1.79) (2.62) (-1.18) (1.75)
Lagged Telecom Expenditure 0.024
(-0.69) 0.015
(-0.42) 0.038
(-0.94)
East -0.002 -0.009 -0.042
(-0.03) (-0.15) (-0.43)
West 0.05 -0.014 0.055
(-0.79) (-0.18) (-0.80)
North -0.085 -0.086 -0.07
(-0.87) (-1.01) (-0.81)
Constant -0.609 -0.681 -0.519 -0.437 -0.304 -0.521 -0.522 -0.522 -0.473
(-0.94) (-0.77) (-0.85) (-0.71) (-0.37) (-0.87) (-0.77) (-0.55) (-0.33)

Observations 62 45 62 63 45 63 63 45 63
R-squared 0.97 0.97 0.97 0.97 0.97 0.98 0.97 0.97 0.97

Robust t-statistics in parentheses.
* significant at 10%; ** significant at 5%; *** significant at 1%.

Table 10: SUREG results of impact of teledensity on state level economic development

(1) (2) (3)
Dependent Variable: INDPROD FDI GSDP

Teledensity 0.157 -0.492 0.248
(-0.26) (-0.35) (-0.27)
Value of Manufacturing 0.334* -0.262*
(0.17) (0.15)
Power shortage -0.069** -0.093** -0.065**
(0.03) (0.03) (0.03)
Development Expenditure (lagged) 0.448**
(0.22) -0.721***
(0.24)
Capital Expenditure -0.17
(-0.19)
Literacy -0.014 0.015 -0.016
(-0.01) (-0.02) (-0.01)
Man-days lost due to labor disputes 0.158*
(0.08) 0.1
(-0.105) 0.024
(-0.09)
Wages -0.001 0.013 0.004
(-0.006) (-0.008) (-0.007)
Constant 4.396 16.021*** 31.493***
(-3.42) (4.08) (3.59)

Observations 43 43 43

Absolute value of z statistics in parentheses.
* significant at 10%; ** significant at 5%; *** significant at 1%

Table 11: SUREG results of impact of network on state level economic development

(1) (2) (3)
Dependent Variable: INDPROD FDI GSDP

Network 0.911*** 0.153 -0.105
(0.18) (-0.35) (-0.32)
Value of Manufacturing 0.224 -0.192
(-0.26) (-0.21)
Power -0.036 -0.079** -0.068**
(-0.02) (0.04) (0.03)
Development Expenditure (lagged) -0.105
(-0.21) -0.628**
(0.269)
Capital Expenditure -0.036
(-0.18)
Literacy -0.022*** -0.008 -0.005
(0.01) (-0.01) (-0.01)
Man-days lost due to labor disputes -0.014
(-0.07) 0.062
(-0.12) 0.036
(-0.09)
Wages 0.001 0.012 0.006
(-0.005) (-0.008) (-0.007)
Constant 5.265* 11.725*** 31.18***
(2.72) (3.67) (3.69)

Observations 43 43 43

Absolute value of z statistics in parentheses.
* significant at 10%; ** significant at 5%; *** significant at 1%

Table 12: SUREG results of impact of exchange on state level economic development

(1) (2) (3)
Dependent Variable: INDPROD FDI GSDP

Exchange 0.152 -0.147 0.102
(2.08)** (-1.49) (-1.23)
Value of Manufacturing 0.329 -0.306
(1.87)* (1.92)*
Power -0.073 -0.081 -0.073
(2.73)*** (2.21)** (2.34)**
Development Expenditure (lagged) 0.444 -0.655
(2.17)** (2.77)***
Capital Expenditure -0.02
(-0.11)
Literacy -0.005 -0.01 -0.004
(-0.63) (-1.00) (-0.52)
Man-days lost due to labor disputes 0.123 0.122 0.009
(-1.50) (-1.16) (-0.1)
Wages 0.004 0.008 0.009
(-0.64) (-0.98) (-1.28)

Observations 43 43 43

Absolute value of z statistics in parentheses.
* significant at 10%; ** significant at 5%; *** significant at 1%




















FIGURES

Figure 1: Share of Communication in National Plan Outlays


Source: Infrastat (2003)

Figure 2: Mainlines and cellular lines per 100 inhabitants



Source: International Telecommunications Union

Figure 3: Urban and Rural Teledensity, 1995-2001


Source: Infrastat

Figure 4: Foreign Direct Investment in Telecommunications, in Rs. Crores




Source: Infrastat.com

Figure 5: Share of telecom service items in foreign direct investment, 1991-2002



Source: DOT Annual Report, 2002

Figure 6: Waiting list and telephone faults per 100 mainlines



Source: International Telecommunications Union

Figure 7: Fault rate in MTNL per 100 telephones (Mumbai and Delhi)


Source: Infrastat



APPENDIX 1
Changes in Telecom Policy Since the Early 1990’s

1992 --Bids invited for radio paging services in 27 cities and cellular mobile services in four metros
1994 --National Telecom Policy announced
-- Private service providers allowed to operate radio paging, Cellular mobile services in metros, V-SAT data services, E-mail services, video tex
-- DOT guidelines for private sector entry into telecom services in the country
-- Eight cellular licenses for four metros finalized
1995 --DoT invited proposals to operate basic services, cellular service circles and public mobile radio trunked (PMRT) services throughout the country
--Most cellular operators in circles sign license aggrements
1996 --After setting reserve price for circles, DoT invites fresh bids for basic services for 13 circles.
1997 --Telecom Regulatory Authority of India (TRAI) established and constituted.
--Five private basic service providers sign license
--Internet policy finalized
1999 --New Telecom Policy announced
--Private internet service providers start operations.
--conditions for migration to revenue sharing from fixed license fee regime issued.
--Commercial usage of spare capacity of public utilities like railways, GAIL, Power and gas permitted
--TRAI issues its first Tariff Order and begins tariff re-balancing
--TRAI issues its first order on interconnection charges and revenue sharing
2000 --Govt. allows internet service providers to set up international gateways
--Guidelines issued for license for new entrants for national long distance services, basic telecom services, cellular mobile service
--TRAI issues second phase of tariff re-balancing
--Policies announced for easier entry/operation of new service providers in the various sectors, e.g. VSAT, PMRT, radio paging, voice mail, etc.
2001 --Convergence commission of India Bill laid in Parliament
--Open competition policy announced for international telephony service.
--Usage of voice over internet protocol permitted for international telephony service.
--First license for national long distance service issued.
Sources: Singh et. al (1999), Indian Telecommunication Statistics (2002) and DoT Annual Report (2002)
APPENDIX 2
Cellular and basic service categories

Category State/Union Territory/City
CELLULOR CIRCLE CATEGORIES
Metros Delhi, Chennai, Kolkata
Category A Andhra Pradesh, Gujarat, Karnataka, Maharashtra (excl. Mumbai), Tamil Nadu (excl. Chennai)
Category B Haryana, Kerala, Madhya Pradesh, West Bengal (excl. Kolkata), Uttar Pradesh (west), Uttar Pradesh (east), Rajasthan, Punjab
Category C Bihar, Orissa, Himachal Pradesh, Andaman and Nicobar, Jammu and Kashmir, North East
BASIC CIRCLE CATEGORIES
Category A Delhi, Maharashtra (incl. Mumbai), Tamil Nadu (incl. Chennai), Karnataka, Andhra Pradesh, Gujarat
Category B West Bengal (incl. Kolkata), Madhya Pradesh, Punjab, Rajasthan, kerala, Uttar Pradesh (west), Uttar Pradesh (east), Haryana
Category C Bihar, Orissa, Assam, Himachal Pradesh, Andaman and Nicobar, Jammu and Kashmir, North East
Source: DoT Annual Report (2002)

ENDNOTES