<%BANNER%>

Convergence among the states of India in the time period from 1980 to 1999

University of Florida Institutional Repository

PAGE 1

CONVERGENCE AMONG THE STATES OF INDIA IN THE TIME PERIOD FROM 1980 TO 1999 By TIMIR PATEL A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2003

PAGE 2

TABLE OF CONTENTS Page LIST OF TABLES.............................................................................................................iii ABSTRACT.......................................................................................................................iv CHAPTER 1 INTRODUCTION........................................................................................................1 Literature Review.........................................................................................................1 Economic Background..................................................................................................5 2 THEORETICAL PASSAGE......................................................................................10 3 TEST OF HYPOTHESIS...........................................................................................16 Regression Specification............................................................................................16 Discussion and Interpretation of Empirical Analysis.................................................21 4 CONCLUSION...........................................................................................................35 LIST OF REFERENCES...................................................................................................37 BIOGRAPHICAL SKETCH.............................................................................................40 ii

PAGE 3

LIST OF TABLES Table page 3-1. Summary of Variables...............................................................................................19 3-2. Determinants of PCNSDP from 1980 to 1999..........................................................21 3-3. Determinants of PCNSDP from 1980 to 1989..........................................................22 3-4. Determinants of PCNSDP from 1990 to 1999..........................................................23 3-5. Determinants of GROSDP from 1980 to 1999..........................................................24 3-6. Determinants of GROSDP from 1980 to 1989..........................................................25 3-7. Determinants of GROSDP from 1990 to 1999..........................................................26 iii

PAGE 4

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts CONVERGENCE AMONG THE STATES OF INDIA IN THE TIME PERIOD FROM 1980 TO 1999 By Timir Patel May 2003 Chair: Bin Xu Major Department: Economics The present study uses the neoclassical growth model as a framework to study a sample of 14 states of India, in the period from 1980 to 1999. The empirical findings show a result of convergence among the states Per Capita Net State Domestic Product in the period from 1980 to 1999. The results do not show conclusively that the level of convergence in PCNSDP increased in the time period from 1990 to 1999 due to the change in economic and political policy during the early 1990s. The study finds six statistically significant determinants of economic growth. A state run by a high quality government and endowed with human capital as well as infrastructure has the foundations from which to grow. An increase in investment creates job openings, which stimulates economic growth. Non-developmental government expenditure going towards the military, social services and past interest payment on debt holds back a states growth potential. iv

PAGE 5

INTRODUCTION The paper uses a sample of 14 states of India in the period from 1980 to 1999 to find the determinants of PCNSDP (Per Capita Net SDP). The same sample is then used to determine whether the PCNSDP of the 14 states is converging or diverging. Having regressed GROSDP (Annual Growth Rate of SDP) against six explanatory variables, I find a result of convergence among the states PCNSDP in the period from 1980 to 1999. The results do not show conclusively that the level of convergence in PCNSDP increased in the time period from 1990 to 1999 due to the change in economic and political policy during the early 1990s. A state run by a high quality government and endowed with human capital as well as infrastructure has the foundations from which to grow. An increase in investment creates job openings, which stimulates economic growth. Non-developmental government expenditure going towards the military, social services and past interest payment on debt holds back a states growth potential. Literature Review An important economic question is whether poor countries or regions tend to converge toward rich ones. In neoclassical growth models for closed economies, as presented by Ramsey (1928), Solow (1956), Cass (1965) and Koopmans (1965), the per capita growth rate tends to be inversely related to the starting level of output or income per person. The assumption of diminishing returns to capital implicit in the neoclassical production 1

PAGE 6

2 function leads to the prediction that the rate of return to capital is very large when the stock of capital is small and vice versa. To understand how fast, and to what extent the per capita income of a particular economy is likely to catch up to the average of per capita income across economies, -convergence is used. There is absolute -convergence if the coefficient on initial income, denoted is negative and statistically significant. Absolute convergence is the natural rate at which the states PCNSDP are converging. When some additional variables are included in the regression to control for heterogeneities across states, there is conditional -convergence if the coefficient on initial income is negative and statistically significant. Conditional convergence in relation to absolute convergence shall have a larger value because conditional convergence is able to increases the rate of convergence by focusing on the significant variables. A second approach used in the literature to measure convergence, which is not used in this paper is -convergence. There is -convergence if the standard deviation across states tends to decline over time. To understand how the distribution of per capita income across economies has behaved in the past, or is likely to behave in the future, -convergence is used. Several studies of high-income economies undertaken during the 1990s, for the US, Japan, and regions within Western Europe, found evidence for strong convergence among regions and states (Barro and Sala-I-Martin 1995). When examining the simple relationship between the per capita growth rate and initial per capita product, the estimates of are around one percent for the sample of twenty OECD countries and near zero percent for the sample of ninety-eight heterogeneous countries (Barro and Sala-I

PAGE 7

3 Martin 1991). The technological and institutional differences across regions within a country or across similar countries are smaller justifying a test for absolute convergence. We find evidence of convergence for a sample of ninety-eight countries from 1960 to 1985 only in a conditional sense, that is, only if we hold constant variables such as initial school enrollment rates and the ratio of government consumption to GDP (Barro and Sala-I-Martin 1992). The literature studying growth rates between the Indian states has found evidence of both convergence and divergence. Dholakia (1994) analyzed 20 Indian states over the period of 1960-90 and found marked tendencies of convergence of long-term State Domestic Product (SDP). Interestingly, Dholakia found that 1980 was the year when several lagging states began growing and the leading states began to stagnate. Cashin and Sahay (1996) studied a longer time period from 1961-91 and found similar results to Dholakia of absolute convergence in a study of 20 states. Ahluwalia (2001) showed that whilst Punjab and Haryana were the two richest states in 1990-91, their growth rates of per capita SDP in the 1990s was not only lower than that in the 1980s, but also in both cases fell below the national average. Consistent with the convergence argument, he also showed that the poorer states per capita SDP growth rates did not lag behind, but rather performed well, particularly for the states of Rajasthan and Madhya Pradesh. Chaudhury (1974) studied state income inequalities between 1950-70 and found that the degree of state income inequality had remained unchanged. Majumdar and Kapoor (1980) was one of the first studies to suggest that there had been a steady increase in the inter-state income inequalities between the years of 1962-76. Recent studies by Rao, Shand and Kalirajan (1999) have found per capita SDP in the Indian states have tended to diverge rather than converge. Dasgupta et al. (2000) used a similar approach to Rao and

PAGE 8

4 also found a distinct tendency for the Indian states to have diverged during the period 1960-95. Cashin and Sahay (1996) argues convergence, whilst Rao et al. (1999) argues divergence for a similar time period. Rao et al. used a smaller sample of 14 major states, whilst Cashin and Sahay included the special category hill states of Assam, Himachal Pradesh, Jammu and Kashmir, Manipur and Tripura as well as the union territory of Delhi. Rao et al. calculated average annual growth rate for six time periods which where all fifteen year intervals. In contrast Cashin and Sahay calculated average annual growth rate for each year during his sample. The study presented here uses a sample of fourteen states, and calculates average annual growth rate. Basu (2002) studied government quality and how it affects growth rates between the Indian states. The significance of this paper comes from its creation of a Quality of Governance Index (QGOI) which shall be used here as an explanatory variable. Basus research was limited to creating a QGOI and showing that the quality of government does significantly affect growth rates among the sixteen states used in his study. Sterns 1991 growth determinants paper reviews previous theoretical growth work, and then puts forth a general model explaining economic growth. Part of Sterns explanation of economic growth was human capital, quality of infrastructure and government organization; from Sterns model the foundation of the paper presented here is built. A similar cross sectional regression equation used by Sala-I-Martin in his growth research is used in this empirical analysis. Chapter 2 puts forward a deductive theoretical argument as to what is causing the disparities in SDP. Chapter 3 is the empirical test of the theoretical argument and

PAGE 9

5 hypothesis put forward in Chapter 2. The empirical results are discussed and interpreted in Chapter 3 and a conclusion follows in Chapter 4. Economic Background Mahatma Ghandi led India to independence from the British rule in 1947. Having gained independence, India was given greater responsibility in making economic and political decisions. Jawaharlal Nehru pushed the National Indian Congress to implement Nehruvian socialism as the development strategy for India from the years of 1950 to 1980. The policy was aimed at achieving economic self-sufficiency by industrialization. The industrialization was different from that seen in other countries because a highly protected import-substitution approach was used. This approach was typified by high levels of protectionism against foreign competition, mistrust of foreign investors, public-sector domination of industry and a highly centralized allocation of resources. The resulting economy became bound in a web of bureaucracy, subsidies and price controls that effectively served to deaden innovation and progress. Protectionism isolated India from the rest of the world, and the countrys share of world trade declined from 2 percent in the 1950s to less than half of one percent in the late 1980s (http://www.worldbank.org). The policy approach employed between 1950 and 1980 brought the Indian economy on the verge of a great fall, and in 1984 the government began borrowing from the World Bank. India is one of the oldest members of the World Bank having joined in 1944. It is the World Banks largest single borrower, with cumulative lending of about $44 billion as of June 1998 in market-based loans from the International Bank for Reconstruction and Development (IBRD), and development credits from the International Development Association (IDA), the World Bank affiliate that provides interest-free loans to economies with low per-capita incomes (http://www.worldbank.org).

PAGE 10

6 The World Bank leant money on the condition that India adopts a moderate liberalized strategy. Support for liberalization grew all through the 1980s. Liberalization came up against powerful resistance from interest groups, and the lack of strong political leadership meant that little progress toward liberalization was seen. Central government borrowing exceeded five percent of GDP because public expenditure was growing faster than revenue in the 1980s. The government had to borrow to pay for the subsidies, which reached twelve percent of GDP in 1990. The subsidies were going to the agricultural sectors to pay the farmers who were receiving below market price for their produce. The budget deficit increased as the savings and investment gap widened. Finally, in 1991, the economy was on the verge of bankruptcy, with only two weeks worth of foreign exchange reserves and a current account deficit of minus three percent of GDP (IMF, 1995). The IMF had a standby arrangement facility with India in 1991 and leant $2.3bn to help India avoid an economic crisis. The IMF forced changes in the economic policies implemented by India since achieving independence in 1947. Cuts in public expenditure on export subsidies, health and education were seen to reduce the budget deficit. The economic policies used in the past forty years had failed and gradually a new approach to managing the Indian economy was being forced into place by mounting international pressures. Indias economic reforms began in 1991 when a newly elected Congress government, facing an exceptionally severe balance of payments crisis, embarked on a program of short term stabilization combined with a longer term program of comprehensive structural reforms (Sachs, Varshney and Bajpai 1999, 26). In 1991 the

PAGE 11

7 governments new liberalization approach attempted to curb the trade deficit, by attracting foreign direct investment and increasing exports. East Asia used export-oriented growth, which helped them grow much faster than India. East Asian GDP growth rates averaged around eight percent in the 1980s. Export oriented growth policies such as opening up their economies, specializing in exports, and attracting Japanese investment when the yen appreciated explains the East Asians increased GDP growth rate. India attempted to insert itself back into the world economy in 1991 through the expansion of its export trade. To boost exports the rupee was devalued by nineteen percent in 1991. The devaluation caused exports to accelerate from eight point six percent growth per year in 1986 to fifteen percent in 1995. The devaluation of the Rupee made Indian exports relatively cheaper and the demand for the relatively cheaper Indian goods increased. Those holding the Rupee at the time of the devaluation suffered because now their savings were worth less than before the devaluation. The Rupee now had less foreign purchasing power causing Indian imports to decrease. The IMF had recently helped India avoid bankruptcy, and now demanded India start changing the policies they had implemented post 1947. The states were literally forced by the IMF to relinquish control of some public sector enterprises and to encourage private enterprises to grow. Much of the industrial licensing system such as the license-permit Raj used to protect the undeveloped Indian economy was broken up. Electricity, power, steel and communications were now areas opened up to private sector investment. These new sectors became more efficient under private control and the owners of these newly privatized industries made profits.

PAGE 12

8 The changing approach of policy now allowed foreign investors in the 1990s to own major shares in companies. Red tape was cut making it easier to set up new businesses and allow foreign investment. Many aspects of business decision-making by the government were dismantled as more sectors became privatized. By 1998 direct foreign investment totaled $3.2 billion, which is an investment level twenty five times greater than that seen when India was a protectionist country. In 1993, a market-determined exchange rate was introduced. Up until 1993, India had used a fixed exchange rate for the rupee. The new floating exchange rate allows for greater control over the economy by the government using monetary policy. By 1993 capital markets had become liberalized and more transparent to encourage foreign investors. Import quotas were converted to tariffs, which opened up the Indian trade policy. The maximum tariff was lowered from four hundred percent in 1991 to thirty five percent by 1998 (WTO 1998, 21). The reforms of the 1990s had stabilized the Indian economy. Privatization helped end the balance of payments crisis. The Indian foreign exchange reserves in 1991 were at an all time low of $1bn and in contrast toward the end of 1996, the foreign exchange reserves rose to $20bn. The reserves increased because non-resident Indians began investing in India and the IMF had increased the capital flow. The end to protectionism and the rise of liberalization in the 1990s is happening slowly in India. Indias capitalistic private sector had been the most controlled by the government in the non-socialist world. Regulation still remains a problem in financial and agricultural sectors because India previously had a heavily regulated economy. Bound tariffs are maximum tariff rates committed in the WTO on agricultural goods. The Bound tariff

PAGE 13

9 ceilings remain high, ranging from 1one hundred and fifty percent to three hundred percent. In reality, applied rates for 1997/98 are considerably lower, averaging twenty six percent for the sector, with a peak of forty five percent, (WTO 1998, 12). The twenty six percent tariff on agricultural goods being imported artificially distorts the market and creates less incentive for the Indian farmers to become more productive and competitive. Indias increased openness and integration with the world economy have been important factors in explaining the healthy economic growth in the 1990s. (WTO 1998, 14). Economic stagnation between 1950 and 1980 has proven economic development under import-substitution industrialization is not a good path toward economic development. The Soviet Union collapse showed Communism was not an effective strategy for growth. The Austrian economists Mises and Hayek started the socialist calculation debate in 1920. The Austrian economists had predicted that socialism would not be sustainable in the long run. In the past seven years, liberalization and domestic restructuring has helped India develop at five to six percent per annum.

PAGE 14

THEORETICAL PASSAGE The theoretical passage is first used to show the origins of the model and equations later used in the empirical analysis. The work of Robert Barro and other economists in the growth field have identified a substantial number of variables that are partially correlated with the rate of economic growth. The basic methodology used by Sala-I-Martin (1997) to identify the determinants of growth consists of running cross-sectional regressions of the form: = + 1 x 1 + 2 x 2 + ..+ n x n + (1) The vector of rates of economic growth is and x 1 x n are the vectors of explanatory variables. Variables like the initial level of income, the investment rate, various measures of education, some policy indicators, and many other variables have been found to be significantly correlated with growth in regressions like (1) (Sala-l-Martin 1997, 178). The simplified neoclassical production function has only two inputs, physical capital, K(t), and labor, L(t). The production function takes the form: Y(t) = F[K(t), L(t), t] (2) The flow of output produced at time t is represented by Y(t). The neoclassical model assumes diminishing returns to capital, exogenous technological progress, full employment, a fixed relationship between the labor force and population, and exogenous growth of population. With respect to preferences, the model assumes that the saving rate derives from the choices made by utility-maximizing households over an infinite horizon (Barro and Sala-I-Martin 1991, 109). 10

PAGE 15

11 Economists have extended the notion of technology to include natural resources, such as geography, fertile land, and the availability of minerals, as well as governmental policies that effect property rights, the provision of infrastructure services, tax rates, and so on. Barro and Sala-I-Martin (1991) showed that the transitional growth process in the neoclassical model can be approximated as: (1/T) log(y it /y i,t T ) = x i + log(y^ i /y^ i,t T ) (1 e -T )/T + u it (3) The economy is indexed by i, t indexes time, y it is per capita output, x i is the steady state per capita growth rate, y^ it is output per effective worker, y^ i is the steady-state level of output per effective worker, T is the length of the observation interval, the coefficient is the rate of convergence, and u it is an error term. The higher the greater the responsiveness of the average growth rate to the gap between log(y^ i ) and log(y^ i,t T ), that is, the more rapid the convergence to the steady state. The model implies conditional convergence in that, for given x i and y^ i the growth rate is higher the lower y^ i,t T The second function of the theoretical passage is to show the relevance of each variable to be used in the empirical analysis and from these variables deduce a testable hypothesis. India has developed at a rapid speed over the past ten years, but still shows many signs of underdevelopment. The birth rate is high with the population growing at a rate of one point five percent a year. India ranks second in the world in terms of population, and seventh in terms of land area. Population growth generally initiates economic growth, but when population growth is at a very high rate, economic growth is hindered.

PAGE 16

12 Education has long been seen as a key concept in curtailing population and aiding economic growth. There is no public schooling system set up to ensure each child attains a minimum education level. CIA figures state the literacy rate to be thirty seven percent for women and sixty five percent for males. The literacy rate for women is very low, because tradition holds that women do not necessarily need to be educated. The norm for many Indian girls is to work around the house until marriage. The occupational structure of the labor force is related to the education level. India remains an agriculture-based society. CIA figures from 1995 show the distribution of the Indian work force: sixty seven percent agriculture, seventeen percent service and sixteen percent industry. Both the industrial and service sector is growing at a rate of six percent a year, but it is going to take decades for the majority of the Indian work force to move from farm work into industrial and service work. A lack of education leads to poverty because most uneducated people are paid very little for their work. More than a third of the population is too poor to be able to afford an adequate diet, and market surveys indicate that fewer than five percent of all households had an annual income equivalent to $2,300 or more in 1995-96 (http://www.cia.gov/cia). Data collected by the CIA in the year 2000 show that thirty five percent of the Indian population lives below the poverty line. Poverty figures like that seen in India in the year 2000 were seen in the U.S. before FDRs New Deal policy was set in place in the 1930s. Most of the population only makes enough to live day to day and therefore are not able to save money. Due to a lack of savings, India is not able to develop advanced credit facilities. The rapid growth of population in India has been

PAGE 17

13 associated with a worsening man-to-land ratio, the growth of landlessness, and the concentration of landholdings in the hands of a few (Jannuzi 1989, 111). For a long time it has been thought that having rich people was good for economic development, and that the government should actually help keep these people rich through such things as tax breaks. Investment stimulates growth in many ways, particularly by creating jobs. Rather than investing money, the rich actually carry out an act known as capital flight, which hinders the national economy. If a high percentage of the money, which is made in India is not spent or invested in the domestic economy, then many economic problems arise. Foreign direct investment (FDI) stimulates the economy by investing money earned abroad into the Indian economy. Growth begins in urban areas, which is where most of the well-developed infrastructure in a state is found. Cities are the focus of government, with great percentages of income collected in tax revenues being distributed in cities. This leads to better health care and education facilities in relation to that found in rural areas. The Indian government has not been creating jobs in the rural areas, which means the population in the rural areas have to move to where the jobs are being created. Individuals that work earn income that is eventually spent in the domestic economy stimulating economic growth. The perception among the Indian population is that cities have a lot of jobs. In reality there simple arent enough jobs in the city, and if the government were to create more jobs, then for each job created, ten more people immigrate to the city in search for that one job. The availability of transport and communication facilities is correlated with urbanized areas and hence is also an explanatory variable of growth. India has recently begun

PAGE 18

14 developing its poor infrastructure, which has aided in economic growth because infrastructure helps trade occur. When a manufacturing firm decides to build a new factory, the location and transportation system is a key factor driving any companys choice. Once a firm has built its factory in a particular state the manufacturing process stimulates SDP. Government expenditure on such things as education and infrastructure creates economic growth. Non-developmental and social services expenditure by the government hinders economic growth. Government debt caused by unnecessary and excessive spending on such things as the military hinders economic growth. Since achieving independence in 1947, India and Pakistan have been in dispute over Kashmir. Figures analyzed by the BBC set expenditure on the war at ten billion for India and 4 billion dollars for Pakistan a year. It is estimated that the bill to keep Indian troops on the remote Siachen glacier amounts to $700,000 a day (http://news.bbc.co.uk). As fighting continues, India may encounter some economic problems, as tax money is not invested in the country, but into the military. There are many alternate ways of spending the ten billion dollars allocated to the military each year. The quality of a state government effects economic growth. A high quality government is able to control crime rates, riots and industrial disputes, which in turn creates stability in the economy. One of the problems India faces is with its government and other people who enforce laws. The system appears to be democratic but often votes are bought or other forms of corruption occur. Low quality government is easily susceptible to corruption. Few people pay taxes in a poorly set up non-progressive tax system. India does not have the tax base it should for a country of its size simply due to

PAGE 19

15 the amount of illegal activity taking place. A larger tax base and a decrease in corruption would help India educate more of its population, and provide better health care. From analyzing the deductive theoretical argument, six hypothesized theoretical constructs of PCNSDP exist. A state run by a high quality government and endowed with human capital as well as infrastructure has the foundations from which to grow. An increase in investment creates job openings, which stimulates economic growth. Non-developmental government expenditure going towards the military, social services and past interest payment on debt holds back a states growth potential.

PAGE 20

TEST OF HYPOTHESIS Regression Specification The purpose of this statistical exercise is to explain the differing growth rates of a sample of fourteen states in the time period from 1980 to 1999. The hypothesized variables explaining the growth rates of SDP are: human capital, infrastructure, quality of government, government expenditure, investment and employment. The basic methodology used to explain PCNSDP consists of running regressions of the form: Y it = 0 + i + t + 1 X 1it + 2 X 2it +. + n X nit + it (4) This linear equation method is appropriate for running regressions and has been used in previous growth work (1) by many other economists. PCNSDP is represented by Y where i is the state and t is the year. X nit are the six explanatory variables for each state and year. The coefficients of each explanatory variable are n whilst is an error term. State and time fixed effects are represented by i and t respectively. The basic methodology used to explain GROSDP consists of running regressions of the form: Z it = 0 + i + t + 1 Y i0 + 2 X 1it + 3 X 2it +. + n X nit + it (5) GROSDP is represented by Z it The variable Y i0 denotes PCNSDP in the initial year. The coefficients of each explanatory variable are n whilst is an error term. The sample consists of fourteen of Indias states, which account for ninety-five percent of the total population. The fourteen states cover 2.7 million square kilometers, which amounts to eighty-three percent of Indias total land area. The smaller states such 16

PAGE 21

17 as Goa and Delhi were excluded from the analysis. The northern states of Jammu and Kashmir were excluded because the on going war makes it difficult for accurate data collection. Other states were excluded because of gaps in the data sources used in the empirical analysis. Using fourteen of the states shall capture a good cross section of all of Indias states because only five percent of the total population is not being included in this analysis. The dependent variable PCNSDP was taken from the Directorates of Economics and Statistics of Respective Governments, which were compiled in the annual Economic Surveys. The dependent variable GROSDP was calculated by: Z it = ln (Y it+1 ) ln (Y it ) (6) Two of the explanatory theoretical concepts use Census data. Literacy rate (LIT) has always been an accepted measure of education level by economists. LIT has always been an accepted measure of human capital in the field of economics. The variable WORK measures the percentage of the total population that works, and was also taken from the Census Reports. The production function used in this paper is comprised of six explanatory variables, which equal the left hand sides SDP. Dividing both sides by labor deals with the endogneity problem of the WORK variable and results in PCNSDP now equaling the production function. A states infrastructure is comprised of many things such as airports, ports and roads. The data used in the empirical analysis only measured roads and did not measure all possible types of infrastructure. ROADS capture the theoretical concept of infrastructure relatively well. The Ministry of Surface Transport in the 1996 and 1986 reports calculate the kilometers of highway in each state. Using this measure would not take into account

PAGE 22

18 different area sizes of each state and would be a useless measure. The area of highway was divided by the area of the state for all fourteen states to calculate the variable ROADS. The quality of government is measured using QGOI data for each state created by Basu (2002). The lower the quality of a states government, the higher the level of corruption found in that state. The QGOI scale ranges from 0 to 16, with 16 representing a government of the highest quality. QGOI is a function of: crime rates, riots, industrial disputes, strikes, Gini index and the debt to income ratio. The QGOI corresponds well to measuring the theoretical concept of government corruption. The annual Combined Finance and Revenue Accounts contain state government expenditure. EXPEND is a function of money spent on: social services, economic services, development and non-development. Total government expenditure was divided by NSDP (Net State Domestic Product) so as to avoid the effect of larger states having greater EXPEND values. Loans sanctioned by State Financial Corporations are used as a proxy for investment because it is difficult to find state level investment data. The total amount of loans given out in each state was divided by NSDP to create the variable INVEST. The data is created by the Industrial Development Bank of India and is compiled in the annual Combined Finance and Revenue Accounts. Loan data have been used in other similar studies and are a reasonable proxy for the theoretical construct of investment. To test whether Loans sanctioned by State Financial Corporations is a good proxy for investment, I calculated the correlation between INVEST and Total National Investment data for each of the twenty years, and found an R-squared value of 0.9494.

PAGE 23

19 Table 3-1. Summary of Variables. Variable No. Obs Mean Std. Dev Min Max GROSDP 266 0.1182838 0.0703775 -.0613505 0.3679864 PCNSDP 280 6231.386 5083.667 878 23398 LIT 280 60.217 13.33567 38.48 90.92 EXPEND 224 0.1785047 0.0331068 0.1175275 0.336864 QGOI 280 6.633714 2.726396 0 10 ROADS 280 0.8681745 0.994643 0.19334 5.18382 WORK 280 38.36339 5.019369 30.88 46.04 INVEST 196 0.0052828 0.0027581 0.0002606 0.0145365 INVEST has only 196 observations, whilst EXPEND has only 224 observations because the data is only available up to the year 1993 and 1995 respectively. All the variables have a positive value apart from GROSDP, which sometimes is negative representing a decline in a states growth rate. The regression equation to calculate the determinants of PCNSDP is: Y it = 0 + i + t + 1 LIT it + 2 EXPEND it + 3 QGOI it + 4 ROADS it + 5 WORK it + 6 INVEST it + it (7) Whilst the regression equation to calculate the determinants of GROSDP is: Z it = 0 + i + t + 1 Y i0 + 2 LIT it + 3 EXPEND it + 4 QGOI it + 5 ROADS it + 6 WORK it + 7 INVEST it + it (8) A low QGOI government is inefficient and holds back growth. As a state government increases its power and control, a more stable environment is created which should positively affect growth rates. Part of creating a more stable environment is developing human capital. As LIT increases the economy of a state can diversify into new sectors and grow. Increasing LIT has a positive impact not only on growth but the quality of life in each state. An educated population tends to move out of the rural areas into the urban areas where there are many more amenities and job opportunities. Factories tend to

PAGE 24

20 locate in urban areas because of the more developed infrastructure. As ROADS increase it is expected that more factories move into that area stimulating growth. Expenditure on items such as maintenance of the organs of the state, administration services, pensions and interest payments are types of non-development expenditure. Excessive government debt holds back the economy because greater percentages of the budget are spent on repaying debt rather than spending on the economy to stimulate growth. An increase in non-developmental expenditure shall impact growth rates in a negative way. Expenditure on education, health, various other social services and economic services are types of developmental expenditure. An increase in developmental expenditure should theoretically increase the growth rate. EXPEND is comprised of both developmental and non-developmental government expenditure. The balance of both types of expenditure that shall influence whether EXPEND has a positive or negative impact on growth. From looking at the EXPEND data, it can be seen that a greater percentage is spent on non-development, which implies that EXPEND should have a negative effect on growth rate. Previous analysis by Barro has found that the ratio of real government consumption expenditure to real GDP had a negative association with growth and investment (Barro 1991, 430). Increasing investment should theoretically have a positive effect on growth rates. As INVEST increases, it shall stimulate the economy and create jobs. WORK shall have a positive effect on growth rates because as more people work, more money shall be spent and circulated through the economy. The process of convergence is quickened by movements of people out of areas where ratios of capital to workers are low-and hence wage rates and levels of per capita income are also low-to areas where they are high (Barro and Sala-I-Martin 1991, 125).

PAGE 25

21 Discussion and Interpretation of Empirical Analysis The empirical research finds that the six explanatory variables upon which the hypothesis of this paper is based, do significantly explain the differing PCNSDP between the states for the three time periods, which are analyzed. All six of the explanatory variables have the expected hypothesized sign. Table 3-2. Determinants of PCNSDP from 1980 to 1999. (1a) (2a) (3a) (4a) LIT 142.96 144.21 21.29 0.3179 (10.001)*** (7.577)*** (1.825)* (0.012) EXPEND -15488.42 -17737.75 -13261.98 -9564.51 (-3.597)*** (-5.022)*** (-5.199)*** (-3.892)*** QGOI 305.53 430.31 256.34 264.48 (6.344)*** (5.146)*** (6.6)*** (6.102)*** ROADS 1446.29 1895.97 236.76 421.36 (6.126)*** (5.250)*** (1.501) (1.436) WORK 50.3 410.29 137.31 281.8 (1.812)* (5.818)*** (4.93)*** (6.952)*** INVEST 64593.69 91429.35 101100.8 103760.4 (1.291) (1.952)* (3.904)*** (4.314)*** Constant -5620.53 -22398.87 3038.88 -2790.46 (-4.936) (-10.339) (2.139) (-1.067) Time Effect Yes Yes State Effect Yes Yes No. Obs 196 196 196 196 R-squared 0.5142 0.6886 0.7867 0.955 t-value appears in parenthesis Significance at the 0.10 level ** Significance at the 0.05 level *** Significance at the 0.01 level All tests are two-tailed.

PAGE 26

22 Table 3-3. Determinants of PCNSDP from 1980 to 1989. (1b) (2b) (3b) (4b) LIT 76.71 93.17 107.1 132.93 (9.148)*** (5.325)*** (11.331)*** (11.529)*** EXPEND -0.5658 -0.6599 -0.0386 -0.0123 (-12.532)*** (-15.099)*** (5.716)*** (6.217)**** QGOI 285.35 430.8 125.44 38.78 (10.275)*** (6.17)*** (3.329)*** (2.373)** ROADS 820.94 624.47 1500.02 1813.84 (5.518)*** (2.017)** (8.334)*** (8.442)*** WORK 4.31 74.56 15.38 44.44 (1.305) (2.206)** (1.946)* (2.06)** INVEST 36178.21 4429.85 11080.2 12671.72 (1.399) (1.744)* (1.514) (1.985)** Constant -562.72 -3294.94 -208.22 -2769.25 (-1.087) (-1.834) (0.271) (-2.423) Time Effect yes yes State Effect Yes yes No. Obs 140 140 140 140 R-squared 0.7092 0.7127 0.887 0.894

PAGE 27

23 Table 3-4. Determinants of PCNSDP from 1990 to 1999. (1c) (2c) (3c) (4c) LIT 119.77 37.82 27.3 42.48 (5.424)*** (1.567) (1.749)* (1.964)* EXPEND -21248.57 -9308.67 -14449.2 -9582.65 (-3.348)*** (-1.996)** (-3.141)*** (-2.156)* QGOI 383.58 429.83 459.03 513.67 (6.149)*** (3.312)*** (4.334)*** 4.451)*** ROADS 562.8 586.7 237.3 103.19 (2.185)** (1.847)* (1.034) (1.931)* WORK 24.85 247.36 70.92 111 (1.661)* (2.717)*** (1.750)* (1.661)* INVEST 10369.23 238396.1 5375.73 28613.47 (1.991)** (2.289)** (2.081)** (2.405)** Constant 6519.71 -3133.82 1029.71 5236.64 (3.010) (-1.149) (1.102) (1.299) Time Effect yes Yes State Effect Yes Yes No. Obs 56 56 56 56 R-squared 0.4817 0.5557 0.8207 0.8445 Tables 3-2, 3-3 and 3-4 show the results of equation (7). Regression (1) uses the ordinary least-squares (OLS) method for fitting a regression line. The panel data set used in the empirical analysis replicates the observations across both states and time periods. A fixed-effects approach uses the variation in explanatory variables over time to identify regression coefficients. A fixed effect estimator controls for state-invariant determinants and is used in (2). Time fixed effects control for time-invariant determinants, such as cultural factors and is used in (3). Regression (4) controls for both time-invariant, and state-invariant determinants by using both time, and state fixed effects.

PAGE 28

24 Table 3-5. Determinants of GROSDP from 1980 to 1999. (5a) (6a) (7a) (8a) (9a) PCNSDP_0 -1.23e-06 -.000013 -.000027 -7.54e-06 -.00003 (-1.669)* (-4.032)*** (-5.964)*** (-1.345) (-3.316)*** LIT 0.002339 0.004384 0.001701 0.006683 (3.156)*** (3.374)*** (2.363)** (2.256)** EXPEND -.500067 -.860475 -.585777 -.91722 (-2.677)*** (-3.767)*** (-2.354)** (-3.189)*** QGOI 0.002378 0.011668 0.001244 0.013733 (1.058) (2.188)** (0.547) (2.532)** ROADS 0.031754 0.027279 0.027168 0.050273 (2.970)*** (1.183) (2.705)*** (1.514) WORK 0.001788 0.007578 0.002152 0.010319 (1.512) (1.663)* (1.898)* (1.987)** INVEST 1.32446 5.849269 0.992275 6.216752 (0.631) (2.029)** (0.501) (2.217)** Constant 0.110297 0.187723 0.677908 0.176682 0.928707 (15.834) (3.686) (4.177) (0.018) (2.651) Time Effect Yes Yes State Effect Yes Yes No. Obs 266 182 182 182 182 R-squared 0.008 0.1234 0.176 0.3203 0.3937 F-statistic 2.13 3.5 6.28 4.02 3.02

PAGE 29

25 Table 3-6. Determinants of GROSDP from 1980 to 1989. (5b) (6b) (7b) (8b) (9b) PCNSDP_0 -0.0000117 -0.0000252 -0.0000251 -0.0000288 -0.0000472 (-2.347)** (-2.655)*** (-2.055)** (-2.404)** (-2.671)*** LIT 0.002599 0.001504 0.002626 0.005209 (-2.201)** (0.609) (2.045)** (1.662) EXPEND -5.99e-06 -4.58e-06 -1.21e-06 -9.52e-06 (-1.834)* (-1.467) (-1.171) (-1.992)** QGOI 0.003202 0.002945 0.003346 0.004322 (1.765)* (1.303) (1.863)* (1.972)** ROADS 0.032314 0.010399 0.033985 0.05029 (1.783)* (1.993)** (1.731)* (2.005)** WORK 0.001889 0.000581 0.001272 0.000931 (1.218) (2.134)** (2.888)*** (2.238)** INVEST 3.850545 6.987681 1.476478 3.89648 (1.278) (1.719)* (1.510) (1.992)** Constant 0.073672 0.096471 0.13424 0.058191 0.204728 (-4.864) (1.687) (0.587) (1.024) (0.981) Time Effect Yes yes State Effect yes yes No. Obs 126 126 126 126 126 R-squared 0.0425 0.0919 0.1192 0.3193 0.3523 F-statistic 5.51 1.71 2.03 2.98 3.52

PAGE 30

26 Table 3-7. Determinants of GROSDP from 1990 to 1999. (5c) (6c) (7c) (8c) (9c) PCNSDP_0 -1.16e-06 -0.0000129 -0.0000247 -0.0000156 -0.0000551 (-1.985)** (-1.683)* (-1.762)* (-1.936)* (-2.693)** LIT 0.002737 0.003122 0.001712 0.004822 (1.77)* (1.338) (1.981)* (1.945)* EXPEND -2.79e-06 -.000021 -1.57e-06 -1.75e-06 (-1.682)* (-1.265) (-1.779)* (-2.094)** QGOI 0.004234 0.006734 0.005647 0.007134 (1.744)** (2.005)** (1.731)* (2.134)** ROADS 0.024 0.051503 0.019423 0.071636 (1.688)* (1.288) (1.843)* (2.250)** WORK 0.001785 0.004521 0.002562 0.007209 (1.753)* (1.705)* (1.747)* (1.755)* INVEST 1.427096 16.62934 3.277358 17.03527 (1.427) (1.798)* (1.699)* (1.936)* Constant 0.1403554 0.1295 0.610512 0.010339 0.573957 (10.625) (1.341) (2.382) (0.098) (1.035) Time Effect Yes Yes State Effect yes Yes No. Obs 140 56 56 56 56 R-squared 0.0067 0.0839 0.2613 0.2359 0.4165 F-statistic 0.93 0.63 2.12 2.62 Table 3-5, 3-6 and 3-7 show the results of equation (8). Regression (5) tests for absolute convergence. Regression (6) uses OLS to begin the tests for conditional convergence. A state fixed effect is used in (7) and a time fixed effect is used in (8). Regression (9) combines the approach of (7) and (8) by using both a state and time fixed effect. The natural rate at which the states PCNSDP are converging is seen in regression (5). Regressions (5a), (5b) and (5c) all have negative and statistically significant t-values, which demonstrate the existence of absolute -convergence. Comparing the regression

PAGE 31

27 coefficients of (5b) and (5c) leads to the conclusion that absolute convergence amongst the states is highest in the period from 1980 to 1989. The t-values of PCNSDP in regressions (6) through (9) are all negative and significant, which shows conditional convergence of the states PCNSDP levels. The coefficients of regression (6) through (9) are all greater than the coefficient of regression (5), which shows that adding additional variables increases the level of convergence in PCNSDP. Comparing the regression coefficients of (6) through (8) in Tables 3-6 and 3-7 leads to the conclusion that conditional convergence amongst the states is highest in the period from 1980 to 1989. Regressions (9b) and (9c) include both state and time fixed effects which shows conditional convergence is greater in the period from 1990 to 1999. The results do not show conclusively that the level of convergence in PCNSDP increased in the time period from 1990 to 1999 due to the change in economic and political policy during the early 1990s. Changes in economic and political policy do not have an immediate effect on PCNSDP because of time lags. Time lags may explain why convergence in PCNSDP levels is not always greater in the 1990 to 1999 time period when compared to the 1980 to 1989 time period. The empirical results show that human capital is one of the key determinants of growth. LIT had a positive coefficient and was statistically significant in all the regressions. Development data shows that ninety nine point nine percent of all U.S. children are enrolled in primary school education. The contrasting figure for children enrolled in primary school in India is only seventy seven point two percent. The education system available in Indias states is very good but expensive. Children are generally educated at private schools, where parents pay the tuition. Those parents in

PAGE 32

28 the lower income brackets are unable to give their children the opportunity to move up in social stature. The children of the poor are destined for a seemingly unending cycle of poverty they were born into. India does not have the capacity to create a publicly funded education system, so a school system of two layers needs to be created. The first layer shall comprise of the present schools that those who can afford attend. The second layer comprises of schools with not so many amenities, but is financed by the government. The focus of the government-funded schools would be to educate all children to a standard that allows young adults to enter the work force. Unemployment shall decrease, as the educated work force shall be able to use entrepreneurial spirit and enter markets previously not seen in India. If the work force becomes educated other firms apart from simple manufacturing firms shall want to locate in India, which will bring about a greater demand on the educated population. Kerala stands out for its relatively high level of development. In Kerala there are many universities and religion affiliated schools that educate the children at no cost to the parents. Literacy rates in Kerala are much greater than those found in the rest of India because every child has the opportunity to attend school. Educating the population is not sufficient by itself to create long-term economic growth. Kerala accounts for three point four percent of Indias population, however fifty percent of the countries migration occurs from the young educated population leaving Kerala. Without the educated, a country stands no real chance of advancing economically because they have a population unable to carry out the high technology jobs. Education is the key to decreasing the rate at which the states population is growing. Kerala has shown that education can bring population growth rates down to zero percent

PAGE 33

29 and this is what the other states need to focus on. Policies need to be created and implemented in India to control the growing population. World Development data shows the contraceptive prevalence rate in India to be only forty one percent, in contrast to the seventy one percent rate of the United States. The empirical analysis found that as crime rates, riots, industrial disputes and corruption decreased, a states growth rate increased. QGOI was significant in almost all the regressions and consistently had a positive t-value. Public sector employees maybe unproductive because of low and declining civil service salaries and promotions unconnected to performance. Workers become corrupt when wages are not paid fully or on time and this is the fault of the government. Budget funds needed to pay for wages maybe delayed because of the inefficient politicians. The motivation to remain honest may be further weakened if senior officials and political leaders use public office for private gain. Currently there is no employee protection system for those who report the corrupt employees. The cost of corruption cannot be measured, but is often interpreted. Recently the World Bank complied data predicting the level of corruption found in India. The report, according to informed sources said that, because of widespread corruption, two to three per cent of the country's $1.805 trillion GDP is lost annually. The report goes further by stating had there been some control on corruption, the country's present per capita income of $350 would have been increased to $750. The World Bank has formulated an anti corruption approach: GI & AC = F (KI, LE, CA) (9)

PAGE 34

30 Equation (9) shows that government improvement and anti corruption programs are a function of knowledge, information, political leadership and collective action. The World Bank has set out three primary functions to decrease corruption. For corruption to decrease the Indian population must take collective action to end it, and not carry on supporting corruption for their own benefits. Collective action is fundamental in putting new election laws in place. The population has to be educated and made to understand the long run negative effects and costs of corruption for them to begin demanding an end in corruption. The Indian political leadership and quality of government can only be changed through the ballet box. In the future, Madhya Pradesh should experience a larger increase in PCNSDP because of the states large step toward democracy compared to all the other states. A state that allows political freedom shall increase the quality of its government by allowing the population to use their democratic vote. In the long run Madhya Pradeshs change toward democracy should increase growth, but to what extent is extremely difficult to measure. In 1994 Madhya Pradesh was the first state to conduct elections to panchayats. The result was to take power away from the representatives and hand it to the people directly. The state government of Madhya Pradesh has moved from a directly administrative roll to a supervisory roll. Data on real annual average growth rates of per capita GSDP bear testimony to the fact that four out of five states that are more policy reform oriented (with the exception of Andhra Pradesh) are also the fastest growing states in India in the post-reform period (Bajpai and Sachs, 1999). Bihar, Uttar Pradesh, Madhya Pradesh and Orissa have lagged behind in both policy reform and growth of PCNSDP.

PAGE 35

31 The variable WORK was statistically significant in all regressions and had a positive t-value. With the large population India already has, something has to be done to create jobs. Addressing the issue of technology, India needs to focus on using labor-intensive rather than capital-intensive technologies. An un-educated person may have a job, but often it is not well paying enough to keep the individual above the countries nationally established poverty line. Financial and High-Tech services such as communications and information technology industries are reliant on a well-educated work force. These industries also require telephone and Internet connections, which are all characteristics of urban areas. Urban areas have a higher quality of life that attracts the highly skilled mobile workers. Maharashtra has the most well developed service industry among all fourteen states, with the countries financial center and other important IT based industries being located within the state. Tamil Nadu, Karnataka and Andhra Pradesh have relatively large service sectors in relation to the other ten states. As the population becomes more educated, Kerala has seen the majority of its population move from the primary sector to the tertiary sector, which is a similar kind of movement to that seen in developed countries. The Green Revolution dating from 1965 to the early 1980s created many jobs and initiated agricultural-ledproductivity growth. Irrigation and extensive fertilizer use helped increase wheat and paddy rice yields. The Green Revolution was primarily in the states of Haryana and Punjab, which have an ideal agro-climate. This short-lived burst in agriculture led growth has not occurred in the period from 1980-99. Agriculture jobs are seasonal and a village or town cannot survive on agriculture alone. In an attempt to revive growth in rural areas, China has adopted a policy trying to

PAGE 36

32 attract large manufacturing firms to these areas. China has remedied this problem by attracting big corporations to Chinese towns and villages so as to absorb the surplus labor. Corporations set up production plants in towns where the population can do both farming and manufacturing work. The best example of this type of export led growth is the arrival of Nike to the East Coast of China. In this situation the whole town is based around Nike; almost all residents in the surrounding areas work in the factory. Most of the villagers work at Nike during the day and go to the farm at night to make sure everything is all right. Having factories locate in small towns and villages has created the incentive needed to keep the rural population from migrating. Urban areas have relatively well developed infrastructure networks in relation to rural areas. ROADS hypothesized positive relationship with growth rate was supported in the results. Theoretically, the more money allocated by states to building up infrastructure, the faster a state should grow. Well-developed infrastructure attracts firms to locate in a particular state because infrastructure helps businesses. INVEST had a positive relationship with growth rate which was statistically significant in most of the regressions. The states giving out more loans grew faster because the money from loans was invested back into the state stimulating the states economic growth. The poor economic performance of Kerala can be attributed to insufficiency of investment, both private and public largely due to the failure of the government to implement policies effective in encouraging private investment. Private investment in Kerala is the lowest in comparison with all of the other states. Historically the states government has had a hostile attitude toward foreign investment. The lack of policy development by the state government results in very low FDI. Due to a lack of

PAGE 37

33 FDI, manufacturing only accounted for fifteen point five percent of Keralas SDP in 1991. The expected coefficient sign of EXPEND was difficult to predict because it depended on the proportion of non-developmental to developmental expenditure. The variable EXPEND was statistically significant in all regressions and had a constant negative impact on both PCNSDP and GROSDP. For EXPEND to have had a negative sign, it would imply that the majority of states spend more on non-development than developmental expenditure. The government spends almost half of its Non-Plan Expenditure budget on interest payments. This payment is only going toward paying interest on the debt and not the principal of the debt. In the long run economic growth is stifled because thirty three percent of the budget is not being put back into the Indian economy. The proportion of the budget allocated to subsidies is being put back into the economy. The amount of money allocated to subsidies has decreased since 1947, but it still remains a substantial portion of budget expenditures. The social services expenditure of 6,187 Crore Rupees represents money allocated to education and health. A comparison made between the money allocated toward defense and that allocated toward health and education, shows only one seventh of the money allocated to defense is allocated toward education and health. Money should be allocated to education instead of the military because allocating money to education would have a long run multiplier effect upon economic growth Rajasthan had an extremely high PCNSDP in the 1981-1990 time period, but after 1991 Rajasthan was no longer seen as the fastest growing state. Even though Rajasthan received the highest per capita transfers and grants from the union government among the

PAGE 38

34 four BIMARU states, and was one of the top recipients of federal transfers of all states, this alone does not account for the high growth rates. Orissa received the second largest per capita transfers, and was still the slowest growing state in the 1981-1990 period. The other factors that help explain the high growth rate of Rajasthan are the rapid electrification of the state and the building on the Command Canal in 1980 to help agriculture. Rajasthan was also a beneficiary of the Green Revolution, which contributed to the SDP just as the increased tourism did during this time. Rajasthans tourist industry has grown at the fastest rate among all fourteen states because of the popularity of visits to Jaipur and Udaipur and the states proximity to Delhi. Orissa is located on the coast but experiences some of the lowest PCNSDP. Orissas agricultural sector and economy suffer greatly from frequent flood devastation. The state is one of the most well endowed with resources in relation to the other states. Orissa is primarily an agricultural state despite its resource endowments. For this reason the poor agricultural performances year after year hinders the SDP and growth rate. An important factor that maybe explaining Orissas low PCNSDP is the fact that twenty five percent of the states population is tribal. Tribal populations are generally associated with lower social indicators in health and education, and suffer social and political exclusion.

PAGE 39

CONCLUSION India abandoned the protectionist approach advocated between 1947 and 1990. The liberalization approach has increased the rate at which India is developing. The new approach has set the foundations from which good economic and political policy can be created to help the states of India grow. Throughout the 1990s India has had many economic problems, which have been resolved using foreign aid. Theoretically, the relatively more urbanized states should experience economic growth faster than states that have a greater proportion of its population in rural areas. Urban areas have more of the variables that are both positively correlated and help initiate growth. Well-developed educational facilities are concentrated in urban areas. As the students finish their education and begin participation in the work force, they should theoretically find well paying jobs in urbanized areas. The job openings occur as the city and its infrastructure grows through government expenditure, attracting new businesses into the area. New businesses invest in the state economy, which stimulates economic growth. Urban areas cannot efficiently develop if the state government is of a low quality because of government inefficiencies. If government expenditure is not used to build roads, schools and other infrastructure, the growth process shall be hindered. The empirical evidence from a sample of 14 states, in the time period from 1980 to 1999 supports the theoretical argument. The six explanatory variables LIT, EXPEND, QGOI, ROADS, WORK and INVEST explain significantly the determinants of growth. A state run by a high quality government and endowed with human capital as well as 35

PAGE 40

36 infrastructure has the foundations from which to grow. The balance of investment and state expenditure are key determinants of both job creation and economic growth. The empirical results show that both absolute and conditional convergence among the states PCNSDP exists. The results do not show conclusively that the level of convergence in PCNSDP increased in the time period from 1990 to 1999 due to the change in economic and political policy during the early 1990s. The study implies the government role should be to develop human capital and infrastructure through development expenditure. Job creation is both the role of the government through developmental expenditure, and the population through investment. The study implies the primary role of the population should be to increase the quality of government by use of their democratic vote.

PAGE 41

LIST OF REFERENCES Ahluwalia, M.S. (2001). State Level Performance Under Economic Reforms in India. Working Paper No. 96, Center for Research Economic Development and Policy Reform, Stanford University, Stanford, CA. Bajpai, Nirupam, and Jeffery D. Sachs. (1999). Progress of Policy Reform and Variations at the Sub-National level in India. Development Discussion Paper No. 730, Harvard Institute for International Development, Harvard University, Cambridge, MA. Barro, R. (1991). Economic Growth in a Cross Section of Countries. The Quarterly Journal of Economics, Vol. 106, Issue 2, pp. 407-443. Barro, R., and Xavier Sala-I-Martin. (1991). Convergence Across States and Regions. Brookings Papers on Economic Activity, Vol. 1991, Issue 1, pp. 107-158. Barro, R., and Xavier Sala-I-Martin. (1992). Convergence. The Journal of Political Economy, Vol. 100, Issue 2, pp. 223-251. Barro, R., and Xavier Sala-I-Martin. (1995). Economic Growth. New York: McGraw-Hill, Inc. Basu, Sudip. (2002). Does Governance Matter? International Economics Department at University of Geneva, Geneva, Switzerland. BBC News. (1999). South Asia Analysis: The Cost of Conflict. Accessed 1 Feb 2003 at http://news.bbc.co.uk/1/hi/world/south_asia/370558.stm Cashin, P., and Sahay, R. (1996). Internal Migration, Center-State Grants, and Economic Growth in the States of India. IMF Staff Papers, Vol. 43, Issue 1. Cass, D. (1965). Optimum Growth in an Aggregate Model of Capital Accumulation. Rev. Econ. Studies Vol. 32, pp. 233-240. Census of India. 1991 and 2001. Provisional Population Totals. Office of the Registrar General, New Delhi. Central Intelligence Agency. (2002). The World Fact Book 2002. Accessed 1 Feb 2003 at http://www.cia.gov/cia/publications/factbook/geos/in.html 37

PAGE 42

38 Chaudhury, D. (1974). Behavior of Spatial Income Inequality in a Developing Economy: India 1950-76. Paper presented at the Ninth Conference of the Indian Association for Research in National Income and Wealth, Seoul. Dasgupta, Maiti, Mukherjee, Sarkar, and Chakrabarti. (2000). Growth and Interstate Disparities in India. Economic and Political Weekly, Vol. 1005, Issue 27, pp. 2413-2422. Dholakia, R. (1994). Spatial Dimension of Acceleration of Economic Growth in India. Economic and Political Weekly, Vol. 1009, No.35, August 21. Economic Intelligence Unit. (1997). Country profile on India. Economic Survey. (1980-2002). Mumbai: Ministry of Finance, Government of India. IMF. (1995). India: Economic Reform and Growth. Occasional paper 134. Washington: International Monetary Fund. Jannuzi, Tommasson. (1989). India in Transition. London: Westview Press. Koopmans, Tjalling. (1965). On the Concept of Optimal Economic Growth. The Econometric Approach to Development Planning. Amsterdam: North-Holland. Majumdar, Grace, and J L Kapoor. (1980). Behavior of Inter-State Income Inequalities in India. Paper presented at the Twelfth Conference of the Indian Association for Research on National Income and Wealth, Bombay. Ramsey, P. (1928). A Mathematical Theory of Saving. Econ. J., Vol. 38, pp. 543-559 Rao, Shand, and Kalirajan. (1999). Convergence of Incomes Across Indian States A Divergent View. Economic and Political Weekly, Vol. 1004, Issue 13, pp. 1-10. Report on Currency and Finance. (1980-95). Reserve Bank of India, Bombay, Vol. 2. Sachs, Varshney, and Bajpai. (1999). India in the Era of Economic Reforms. Oxford: University Press. Sala-I-Martin, X. (1997). I Just Ran Two Million Regressions. The American Economic Review, Vol. 87, Issue 2, pp. 178-183. Solow, R. (1956). A Contribution to the Theory of Economic Growth. Q.J.E., Vol. 70, pp 65-94. Stern, Nicholas. (1991). The Determinants of Growth. The Economic Journal, Vol. 101, pp. 122-133. The World Bank Group. (2003). The World Bank in India. Accessed 1 Feb 2003 at http://lnweb18.worldbank.org/sar/sa.nsf/India?OpenNavigator

PAGE 43

39 World Bank. (1997). India: Five years of stabilization and reform and the challenges ahead. Washington: World Bank. WTO. (1998). Trade policy review on India. Geneva: World Trade Organization.

PAGE 44

BIOGRAPHICAL SKETCH My family originates from the state of Gujarat in India. I was born and raised in England and moved to the United States in March of 1998. I graduated from the University of Texas at Dallas with a Bachelor of Arts in economics and finance in August of 2001. Currently I am at the University of Florida studying towards a Master of Arts in economics. 40


Permanent Link: http://ufdc.ufl.edu/UFE0000766/00001

Material Information

Title: Convergence among the states of India in the time period from 1980 to 1999
Physical Description: Mixed Material
Language: English
Creator: Patel, Timir ( Dissertant )
Xu, Bin ( Thesis advisor )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2003
Copyright Date: 2003

Subjects

Subjects / Keywords: Economics Thesis, M.A.
Dissertations, Academic -- UF -- Economics
Genre: bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract: The present study uses the neoclassical growth model as a framework to study a sample of 14 states of India, in the period from 1980 to 1999. The empirical findings show a result of convergence among the states’ Per Capita Net State Domestic Product in the period from 1980 to 1999. The results do not show conclusively that the level of convergence in PCNSDP increased in the time period from 1990 to 1999 due to the change in economic and political policy during the early 1990s. The study finds six statistically significant determinants of economic growth. A state run by a high quality government and endowed with human capital as well as infrastructure has the foundations from which to grow. An increase in investment creates job openings, which stimulates economic growth. Non-developmental government expenditure going towards the military, social services and past interest payment on debt holds back a state’s growth potential.
Thesis: Thesis (M.A.)--University of Florida, 2003.
Bibliography: Includes bibliographical references.
General Note: Vita.
General Note: Document formatted into pages; contains iv, 40 p.
General Note: Title from title page of document.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000766:00001

Permanent Link: http://ufdc.ufl.edu/UFE0000766/00001

Material Information

Title: Convergence among the states of India in the time period from 1980 to 1999
Physical Description: Mixed Material
Language: English
Creator: Patel, Timir ( Dissertant )
Xu, Bin ( Thesis advisor )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2003
Copyright Date: 2003

Subjects

Subjects / Keywords: Economics Thesis, M.A.
Dissertations, Academic -- UF -- Economics
Genre: bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract: The present study uses the neoclassical growth model as a framework to study a sample of 14 states of India, in the period from 1980 to 1999. The empirical findings show a result of convergence among the states’ Per Capita Net State Domestic Product in the period from 1980 to 1999. The results do not show conclusively that the level of convergence in PCNSDP increased in the time period from 1990 to 1999 due to the change in economic and political policy during the early 1990s. The study finds six statistically significant determinants of economic growth. A state run by a high quality government and endowed with human capital as well as infrastructure has the foundations from which to grow. An increase in investment creates job openings, which stimulates economic growth. Non-developmental government expenditure going towards the military, social services and past interest payment on debt holds back a state’s growth potential.
Thesis: Thesis (M.A.)--University of Florida, 2003.
Bibliography: Includes bibliographical references.
General Note: Vita.
General Note: Document formatted into pages; contains iv, 40 p.
General Note: Title from title page of document.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000766:00001


This item has the following downloads:


Full Text












CONVERGENCE AMONG THE STATES OF INDIA IN THE TIME PERIOD FROM
1980 TO 1999















By

TIMIR PATEL


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ARTS

UNIVERSITY OF FLORIDA


2003
















TABLE OF CONTENTS
Page

L IS T O F T A B L E S ........ ............ .. .. ...... .. ...................................................... iii

ABSTRACT ............... ................... ......... .............. iv

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

L iteratu re R ev iew ................................................................ 1
E conom ic B ackground........... ............................................................... ....... .... .... 5

2 THEORETICAL PASSAGE ......................................................... .............. 10

3 T E ST O F H Y P O T H E SIS ................................................................ ..................... 16

Regression Specification ................................ ......... ........... ............... 16
Discussion and Interpretation of Empirical Analysis .............................................21

4 C O N C L U SIO N ......... ......................................................................... ........ .. ..... .. 35

L IST O F R EFE R E N C E S ............................................................................ .............. 37

B IO G R A PH IC A L SK E TCH ..................................................................... ..................40
















LIST OF TABLES

Table p

3-1. Sum m ary of V ariables ........... ........................................................ ............... 19

3-2. Determinants of PCNSDP from 1980 to 1999. ................................. ...............21

3-3. Determinants of PCNSDP from 1980 to 1989. ................................. ...............22

3-4. Determinants of PCNSDP from 1990 to 1999. ................................. ...............23

3-5. Determinants of GROSDP from 1980 to 1999............ ...................... ...............24

3-6. Determinants of GROSDP from 1980 to 1989............................... ............... 25

3-7. Determinants of GROSDP from 1990 to 1999............ ...................... ...............26
















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Arts

CONVERGENCE AMONG THE STATES OF INDIA IN THE TIME PERIOD FROM
1980 TO 1999

By

Timir Patel

May 2003

Chair: Bin Xu
Major Department: Economics

The present study uses the neoclassical growth model as a framework to study a

sample of 14 states of India, in the period from 1980 to 1999. The empirical findings

show a result of convergence among the states' Per Capita Net State Domestic Product in

the period from 1980 to 1999. The results do not show conclusively that the level of

convergence in PCNSDP increased in the time period from 1990 to 1999 due to the

change in economic and political policy during the early 1990s. The study finds six

statistically significant determinants of economic growth. A state run by a high quality

government and endowed with human capital as well as infrastructure has the

foundations from which to grow. An increase in investment creates job openings, which

stimulates economic growth. Non-developmental government expenditure going towards

the military, social services and past interest payment on debt holds back a state's growth

potential.















INTRODUCTION

The paper uses a sample of 14 states of India in the period from 1980 to 1999 to find

the determinants of PCNSDP (Per Capita Net SDP). The same sample is then used to

determine whether the PCNSDP of the 14 states is converging or diverging. Having

regressed GROSDP (Annual Growth Rate of SDP) against six explanatory variables, I

find a result of convergence among the states PCNSDP in the period from 1980 to 1999.

The results do not show conclusively that the level of convergence in PCNSDP increased

in the time period from 1990 to 1999 due to the change in economic and political policy

during the early 1990's. A state run by a high quality government and endowed with

human capital as well as infrastructure has the foundations from which to grow. An

increase in investment creates job openings, which stimulates economic growth. Non-

developmental government expenditure going towards the military, social services and

past interest payment on debt holds back a states growth potential.



Literature Review

An important economic question is whether poor countries or regions tend to converge

toward rich ones. In neoclassical growth models for closed economies, as presented by

Ramsey (1928), Solow (1956), Cass (1965) and Koopmans (1965), the per capital growth

rate tends to be inversely related to the starting level of output or income per person. The

assumption of diminishing returns to capital implicit in the neoclassical production









function leads to the prediction that the rate of return to capital is very large when the

stock of capital is small and vice versa.

To understand how fast, and to what extent the per capital income of a particular

economy is likely to catch up to the average of per capital income across economies, P-

convergence is used. There is absolute P-convergence if the coefficient on initial income,

denoted 0, is negative and statistically significant. Absolute convergence is the natural

rate at which the states PCNSDP are converging. When some additional variables are

included in the regression to control for heterogeneities across states, there is conditional

P-convergence if the coefficient on initial income is negative and statistically significant.

Conditional convergence in relation to absolute convergence shall have a larger P value

because conditional convergence is able to increases the rate of convergence by focusing

on the significant variables. A second approach used in the literature to measure

convergence, which is not used in this paper is a-convergence. There is a-convergence if

the standard deviation across states tends to decline over time. To understand how the

distribution of per capital income across economies has behaved in the past, or is likely to

behave in the future, a-convergence is used.

Several studies of high-income economies undertaken during the 1990's, for the US,

Japan, and regions within Western Europe, found evidence for strong convergence

among regions and states (Barro and Sala-I-Martin 1995). When examining the simple

relationship between the per capital growth rate and initial per capital product, the

estimates of 0 are around one percent for the sample of twenty OECD countries and near

zero percent for the sample of ninety-eight heterogeneous countries (Barro and Sala-I-









Martin 1991). The technological and institutional differences across regions within a

country or across similar countries are smaller justifying a test for absolute convergence.

We find evidence of convergence for a sample of ninety-eight countries from 1960
to 1985 only in a conditional sense, that is, only if we hold constant variables such
as initial school enrollment rates and the ratio of government consumption to GDP
(Barro and Sala-I-Martin 1992).

The literature studying growth rates between the Indian states has found evidence of

both convergence and divergence. Dholakia (1994) analyzed 20 Indian states over the

period of 1960-90 and found marked tendencies of convergence of long-term State

Domestic Product (SDP). Interestingly, Dholakia found that 1980 was the year when

several lagging states began growing and the leading states began to stagnate. Cashin

and Sahay (1996) studied a longer time period from 1961-91 and found similar results to

Dholakia of absolute convergence in a study of 20 states. Ahluwalia (2001) showed that

whilst Punjab and Haryana were the two richest states in 1990-91, their growth rates of

per capital SDP in the 1990's was not only lower than that in the 1980's, but also in both

cases fell below the national average. Consistent with the convergence argument, he also

showed that the poorer states per capital SDP growth rates did not lag behind, but rather

performed well, particularly for the states of Rajasthan and Madhya Pradesh.

Chaudhury (1974) studied state income inequalities between 1950-70 and found that

the degree of state income inequality had remained unchanged. Majumdar and Kapoor

(1980) was one of the first studies to suggest that there had been a steady increase in the

inter-state income inequalities between the years of 1962-76. Recent studies by Rao,

Shand and Kalirajan (1999) have found per capital SDP in the Indian states have tended to

diverge rather than converge. Dasgupta et al. (2000) used a similar approach to Rao and









also found a distinct tendency for the Indian states to have diverged during the period

1960-95.

Cashin and Sahay (1996) argues convergence, whilst Rao et al. (1999) argues

divergence for a similar time period. Rao et al. used a smaller sample of 14 major states,

whilst Cashin and Sahay included the special category hill states of Assam, Himachal

Pradesh, Jammu and Kashmir, Manipur and Tripura as well as the union territory of

Delhi. Rao et al. calculated average annual growth rate for six time periods which where

all fifteen year intervals. In contrast Cashin and Sahay calculated average annual growth

rate for each year during his sample. The study presented here uses a sample of fourteen

states, and calculates average annual growth rate.

Basu (2002) studied government quality and how it affects growth rates between the

Indian states. The significance of this paper comes from its creation of a Quality of

Governance Index (QGOI) which shall be used here as an explanatory variable. Basu's

research was limited to creating a QGOI and showing that the quality of government does

significantly affect growth rates among the sixteen states used in his study.

Stern's 1991 growth determinants paper reviews previous theoretical growth work,

and then puts forth a general model explaining economic growth. Part of Stem's

explanation of economic growth was human capital, quality of infrastructure and

government organization; from Stem's model the foundation of the paper presented here

is built. A similar cross sectional regression equation used by Sala-I-Martin in his growth

research is used in this empirical analysis.

Chapter 2 puts forward a deductive theoretical argument as to what is causing the

disparities in SDP. Chapter 3 is the empirical test of the theoretical argument and









hypothesis put forward in Chapter 2. The empirical results are discussed and interpreted

in Chapter 3 and a conclusion follows in Chapter 4.

Economic Background

Mahatma Ghandi led India to independence from the British rule in 1947. Having

gained independence, India was given greater responsibility in making economic and

political decisions. Jawaharlal Nehru pushed the National Indian Congress to implement

Nehruvian socialism as the development strategy for India from the years of 1950 to

1980. The policy was aimed at achieving economic self-sufficiency by industrialization.

The industrialization was different from that seen in other countries because a highly

protected import-substitution approach was used. This approach was typified by high

levels of protectionism against foreign competition, mistrust of foreign investors, public-

sector domination of industry and a highly centralized allocation of resources. The

resulting economy became bound in a web of bureaucracy, subsidies and price controls

that effectively served to deaden innovation and progress. "Protectionism isolated India

from the rest of the world, and the country's share of world trade declined from 2 percent

in the 1950s to less than half of one percent in the late 1980s"

(http://www.worldbank.org).

The policy approach employed between 1950 and 1980 brought the Indian economy

on the verge of a great fall, and in 1984 the government began borrowing from the World

Bank. India is one of the oldest members of the World Bank having joined in 1944.

It is the World Bank's largest single borrower, with cumulative lending of about
$44 billion as of June 1998 in market-based loans from the International Bank for
Reconstruction and Development (IBRD), and development credits from the
International Development Association (IDA), the World Bank affiliate that
provides interest-free loans to economies with low per-capita incomes
(http://www.worldbank.org).









The World Bank leant money on the condition that India adopts a moderate liberalized

strategy. Support for liberalization grew all through the 1980s. Liberalization came up

against powerful resistance from interest groups, and the lack of strong political

leadership meant that little progress toward liberalization was seen.

Central government borrowing exceeded five percent of GDP because public

expenditure was growing faster than revenue in the 1980's. The government had to

borrow to pay for the subsidies, which reached twelve percent of GDP in 1990. The

subsidies were going to the agricultural sectors to pay the farmers who were receiving

below market price for their produce. The budget deficit increased as the savings and

investment gap widened. Finally, "in 1991, the economy was on the verge of bankruptcy,

with only two weeks worth of foreign exchange reserves and a current account deficit of

minus three percent of GDP" (IMF, 1995).

The IMF had a standby arrangement facility with India in 1991 and leant $2.3bn to

help India avoid an economic crisis. The IMF forced changes in the economic policies

implemented by India since achieving independence in 1947. Cuts in public expenditure

on export subsidies, health and education were seen to reduce the budget deficit. The

economic policies used in the past forty years had failed and gradually a new approach to

managing the Indian economy was being forced into place by mounting international

pressures.

"India's economic reforms began in 1991 when a newly elected Congress

government, facing an exceptionally severe balance of payments crisis, embarked on a

program of short term stabilization combined with a longer term program of

comprehensive structural reforms" (Sachs, Varshney and Bajpai 1999, 26). In 1991 the









government's new liberalization approach attempted to curb the trade deficit, by

attracting foreign direct investment and increasing exports. East Asia used export-

oriented growth, which helped them grow much faster than India. East Asian GDP

growth rates averaged around eight percent in the 1980s. Export oriented growth policies

such as opening up their economies, specializing in exports, and attracting Japanese

investment when the yen appreciated explains the East Asian's increased GDP growth

rate.

India attempted to insert itself back into the world economy in 1991 through the

expansion of its export trade. To boost exports the rupee was devalued by nineteen

percent in 1991. The devaluation caused exports to accelerate from eight point six

percent growth per year in 1986 to fifteen percent in 1995. The devaluation of the Rupee

made Indian exports relatively cheaper and the demand for the relatively cheaper Indian

goods increased. Those holding the Rupee at the time of the devaluation suffered

because now their savings were worth less than before the devaluation. The Rupee now

had less foreign purchasing power causing Indian imports to decrease.

The IMF had recently helped India avoid bankruptcy, and now demanded India start

changing the policies they had implemented post 1947. The states were literally forced

by the IMF to relinquish control of some public sector enterprises and to encourage

private enterprises to grow. Much of the industrial licensing system such as the "license-

permit Raj" used to protect the undeveloped Indian economy was broken up. Electricity,

power, steel and communications were now areas opened up to private sector investment.

These new sectors became more efficient under private control and the owners of these

newly privatized industries made profits.









The changing approach of policy now allowed foreign investors in the 1990's to own

major shares in companies. Red tape was cut making it easier to set up new businesses

and allow foreign investment. Many aspects of business decision-making by the

government were dismantled as more sectors became privatized. By 1998 direct foreign

investment totaled $3.2 billion, which is an investment level twenty five times greater

than that seen when India was a protectionist country.

In 1993, a market-determined exchange rate was introduced. Up until 1993, India had

used a fixed exchange rate for the rupee. The new floating exchange rate allows for

greater control over the economy by the government using monetary policy. By 1993

capital markets had become liberalized and more transparent to encourage foreign

investors. Import quotas were converted to tariffs, which opened up the Indian trade

policy. "The maximum tariff was lowered from four hundred percent in 1991 to thirty

five percent by 1998" (WTO 1998, 21).

The reforms of the 1990's had stabilized the Indian economy. Privatization helped

end the balance of payments crisis. The Indian foreign exchange reserves in 1991 were at

an all time low of $lbn and in contrast toward the end of 1996, the foreign exchange

reserves rose to $20bn. The reserves increased because non-resident Indians began

investing in India and the IMF had increased the capital flow. The end to protectionism

and the rise of liberalization in the 1990's is happening slowly in India. India's

capitalistic private sector had been the most controlled by the government in the non-

socialist world. Regulation still remains a problem in financial and agricultural sectors

because India previously had a heavily regulated economy. "Bound" tariffs are

maximum tariff rates committed in the WTO on agricultural goods. The Bound tariff









ceilings remain high, ranging from lone hundred and fifty percent to three hundred

percent. "In reality, applied rates for 1997/98 are considerably lower, averaging twenty

six percent for the sector, with a peak of forty five percent," (WTO 1998, 12). The

twenty six percent tariff on agricultural goods being imported artificially distorts the

market and creates less incentive for the Indian farmers to become more productive and

competitive.

"India's increased openness and integration with the world economy have been

important factors in explaining the healthy economic growth in the 1990s." (WTO 1998,

14). Economic stagnation between 1950 and 1980 has proven economic development

under import-substitution industrialization is not a good path toward economic

development. The Soviet Union collapse showed Communism was not an effective

strategy for growth. The Austrian economists Mises and Hayek started the socialist

calculation debate in 1920. The Austrian economists had predicted that socialism would

not be sustainable in the long run. In the past seven years, liberalization and domestic

restructuring has helped India develop at five to six percent per annum.















THEORETICAL PASSAGE

The theoretical passage is first used to show the origins of the model and equations

later used in the empirical analysis. The work of Robert Barro and other economists in

the growth field have identified a substantial number of variables that are partially

correlated with the rate of economic growth. The basic methodology used by Sala-I-

Martin (1997) to identify the determinants of growth consists of running cross-sectional

regressions of the form:

y= + Px1+ 02X2 + .....+ PnXn + (1)

The vector of rates of economic growth is y, and x...... Xn are the vectors of

explanatory variables. "Variables like the initial level of income, the investment rate,

various measures of education, some policy indicators, and many other variables have

been found to be significantly correlated with growth in regressions like (1)" (Sala-1-

Martin 1997, 178).

The simplified neoclassical production function has only two inputs, physical capital,

K(t), and labor, L(t). The production function takes the form:

Y(t) = F[K(t), L(t), t] (2)

The flow of output produced at time "t" is represented by Y(t).

The neoclassical model assumes diminishing returns to capital, exogenous
technological progress, full employment, a fixed relationship between the labor
force and population, and exogenous growth of population. With respect to
preferences, the model assumes that the saving rate derives from the choices made
by utility-maximizing households over an infinite horizon (Barro and Sala-I-Martin
1991, 109).









Economists have extended the notion of technology to include natural resources, such

as geography, fertile land, and the availability of minerals, as well as governmental

policies that effect property rights, the provision of infrastructure services, tax rates, and

so on.

Barro and Sala-I-Martin (1991) showed that the transitional growth process in the

neoclassical model can be approximated as:

(1/T). log(yit/yi,t T)= Xi* + 1og(yAi*/yAi,t T). (1 e-T)/T + uit (3)

The economy is indexed by "i", "t" indexes time, yit is per capital output, xi* is the

steady state per capital growth rate, yAi is output per effective worker, yA^* is the steady-

state level of output per effective worker, T is the length of the observation interval, the

coefficient p is the rate of convergence, and uit is an error term. The higher P, the greater

the responsiveness of the average growth rate to the gap between log(yAi*) and log(yAi,t

T), that is, the more rapid the convergence to the steady state. The model implies

conditional convergence in that, for given xi and yAi*, the growth rate is higher the lower

y i,t -T.

The second function of the theoretical passage is to show the relevance of each

variable to be used in the empirical analysis and from these variables deduce a testable

hypothesis. India has developed at a rapid speed over the past ten years, but still shows

many signs ofunderdevelopment. The birth rate is high with the population growing at a

rate of one point five percent a year. India ranks second in the world in terms of

population, and seventh in terms of land area. Population growth generally initiates

economic growth, but when population growth is at a very high rate, economic growth is

hindered.









Education has long been seen as a key concept in curtailing population and aiding

economic growth. There is no public schooling system set up to ensure each child attains

a minimum education level. CIA figures state the literacy rate to be thirty seven percent

for women and sixty five percent for males. The literacy rate for women is very low,

because tradition holds that women do not necessarily need to be educated. The norm for

many Indian girls is to work around the house until marriage.

The occupational structure of the labor force is related to the education level. India

remains an agriculture-based society. CIA figures from 1995 show the distribution of the

Indian work force: sixty seven percent agriculture, seventeen percent service and sixteen

percent industry. Both the industrial and service sector is growing at a rate of six percent

a year, but it is going to take decades for the majority of the Indian work force to move

from farm work into industrial and service work.

A lack of education leads to poverty because most uneducated people are paid very

little for their work. "More than a third of the population is too poor to be able to afford

an adequate diet, and market surveys indicate that fewer than five percent of all

households had an annual income equivalent to $2,300 or more in 1995-96"

(http://www.cia.gov/cia). Data collected by the CIA in the year 2000 show that thirty

five percent of the Indian population lives below the poverty line. Poverty figures like

that seen in India in the year 2000 were seen in the U.S. before FDR's New Deal policy

was set in place in the 1930's. Most of the population only makes enough to live day to

day and therefore are not able to save money. Due to a lack of savings, India is not able

to develop advanced credit facilities. "The rapid growth of population in India has been









associated with a worsening man-to-land ratio, the growth of landlessness, and the

concentration of landholdings in the hands of a few" (Jannuzi 1989, 111).

For a long time it has been thought that having rich people was good for economic

development, and that the government should actually help keep these people rich

through such things as tax breaks. Investment stimulates growth in many ways,

particularly by creating jobs. Rather than investing money, the rich actually carry out an

act known as capital flight, which hinders the national economy. If a high percentage of

the money, which is made in India is not spent or invested in the domestic economy, then

many economic problems arise. Foreign direct investment (FDI) stimulates the economy

by investing money earned abroad into the Indian economy.

Growth begins in urban areas, which is where most of the well-developed

infrastructure in a state is found. Cities are the focus of government, with great

percentages of income collected in tax revenues being distributed in cities. This leads to

better health care and education facilities in relation to that found in rural areas. The

Indian government has not been creating jobs in the rural areas, which means the

population in the rural areas have to move to where the jobs are being created.

Individuals that work earn income that is eventually spent in the domestic economy

stimulating economic growth. The perception among the Indian population is that cities

have a lot of jobs. In reality there simple aren't enough jobs in the city, and if the

government were to create more jobs, then for each job created, ten more people

immigrate to the city in search for that one job.

The availability of transport and communication facilities is correlated with urbanized

areas and hence is also an explanatory variable of growth. India has recently begun









developing its poor infrastructure, which has aided in economic growth because

infrastructure helps trade occur. When a manufacturing firm decides to build a new

factory, the location and transportation system is a key factor driving any company's

choice. Once a firm has built its factory in a particular state the manufacturing process

stimulates SDP.

Government expenditure on such things as education and infrastructure creates

economic growth. Non-developmental and social services expenditure by the

government hinders economic growth. Government debt caused by unnecessary and

excessive spending on such things as the military hinders economic growth. Since

achieving independence in 1947, India and Pakistan have been in dispute over Kashmir.

Figures analyzed by the BBC set expenditure on the war at ten billion for India and 4

billion dollars for Pakistan a year. "It is estimated that the bill to keep Indian troops on

the remote Siachen glacier amounts to $700,000 a day" (http://news.bbc.co.uk). As

fighting continues, India may encounter some economic problems, as tax money is not

invested in the country, but into the military. There are many alternate ways of spending

the ten billion dollars allocated to the military each year.

The quality of a state government effects economic growth. A high quality

government is able to control crime rates, riots and industrial disputes, which in turn

creates stability in the economy. One of the problems India faces is with its government

and other people who enforce laws. The system appears to be democratic but often votes

are bought or other forms of corruption occur. Low quality government is easily

susceptible to corruption. Few people pay taxes in a poorly set up non-progressive tax

system. India does not have the tax base it should for a country of its size simply due to









the amount of illegal activity taking place. A larger tax base and a decrease in corruption

would help India educate more of its population, and provide better health care.

From analyzing the deductive theoretical argument, six hypothesized theoretical

constructs of PCNSDP exist. A state run by a high quality government and endowed

with human capital as well as infrastructure has the foundations from which to grow. An

increase in investment creates job openings, which stimulates economic growth. Non-

developmental government expenditure going towards the military, social services and

past interest payment on debt holds back a states growth potential.















TEST OF HYPOTHESIS

Regression Specification

The purpose of this statistical exercise is to explain the differing growth rates of a

sample of fourteen states in the time period from 1980 to 1999. The hypothesized

variables explaining the growth rates of SDP are: human capital, infrastructure, quality of

government, government expenditure, investment and employment. The basic

methodology used to explain PCNSDP consists of running regressions of the form:

Yit = Po + 6i + 6t + 31Xlit + P2X2it + ... + PnXnit + Eit (4)

This linear equation method is appropriate for running regressions and has been used

in previous growth work (1) by many other economists. PCNSDP is represented by Y

where "i" is the state and "t" is the year. Xnit are the six explanatory variables for each

state and year. The coefficients of each explanatory variable are Pn, whilst E is an error

term. State and time fixed effects are represented by 6i and 6t respectively.

The basic methodology used to explain GROSDP consists of running regressions of

the form:

Zit = ao + 6i + 6t + laYio + 2Xlit + 3X2it + .... + anXnit + rlit (5)

GROSDP is represented by Zit. The variable Yio denotes PCNSDP in the initial year.

The coefficients of each explanatory variable are an, whilst ri is an error term.

The sample consists of fourteen of India's states, which account for ninety-five

percent of the total population. The fourteen states cover 2.7 million square kilometers,

which amounts to eighty-three percent of India's total land area. The smaller states such









as Goa and Delhi were excluded from the analysis. The northern states of Jammu and

Kashmir were excluded because the on going war makes it difficult for accurate data

collection. Other states were excluded because of gaps in the data sources used in the

empirical analysis. Using fourteen of the states shall capture a good cross section of all

of India's states because only five percent of the total population is not being included in

this analysis.

The dependent variable PCNSDP was taken from the Directorates of Economics and

Statistics of Respective Governments, which were compiled in the annual Economic

Surveys. The dependent variable GROSDP was calculated by:

Zit = In (Yit+i) In (Yit) (6)

Two of the explanatory theoretical concepts use Census data. Literacy rate (LIT) has

always been an accepted measure of education level by economists. LIT has always been

an accepted measure of human capital in the field of economics. The variable WORK

measures the percentage of the total population that works, and was also taken from the

Census Reports. The production function used in this paper is comprised of six

explanatory variables, which equal the left hand sides SDP. Dividing both sides by labor

deals with the endogneity problem of the WORK variable and results in PCNSDP now

equaling the production function.

A states infrastructure is comprised of many things such as airports, ports and roads.

The data used in the empirical analysis only measured roads and did not measure all

possible types of infrastructure. ROADS capture the theoretical concept of infrastructure

relatively well. The Ministry of Surface Transport in the 1996 and 1986 reports calculate

the kilometers of highway in each state. Using this measure would not take into account









different area sizes of each state and would be a useless measure. The area of highway

was divided by the area of the state for all fourteen states to calculate the variable

ROADS.

The quality of government is measured using QGOI data for each state created by

Basu (2002). The lower the quality of a states government, the higher the level of

corruption found in that state. The QGOI scale ranges from 0 to 16, with 16 representing

a government of the highest quality. QGOI is a function of: crime rates, riots, industrial

disputes, strikes, Gini index and the debt to income ratio. The QGOI corresponds well to

measuring the theoretical concept of government corruption.

The annual Combined Finance and Revenue Accounts contain state government

expenditure. EXPEND is a function of money spent on: social services, economic

services, development and non-development. Total government expenditure was divided

by NSDP (Net State Domestic Product) so as to avoid the effect of larger states having

greater EXPEND values.

Loans sanctioned by State Financial Corporations are used as a proxy for investment

because it is difficult to find state level investment data. The total amount of loans given

out in each state was divided by NSDP to create the variable INVEST. The data is

created by the Industrial Development Bank of India and is compiled in the annual

Combined Finance and Revenue Accounts. Loan data have been used in other similar

studies and are a reasonable proxy for the theoretical construct of investment. To test

whether Loans sanctioned by State Financial Corporations is a good proxy for

investment, I calculated the correlation between INVEST and Total National Investment

data for each of the twenty years, and found an R-squared value of 0.9494.









Table 3-1. Summary of Variables.
Variable No. Obs Mean Std. Dev Min Max
GROSDP 266 0.1182838 0.0703775 -.0613505 0.3679864
PCNSDP 280 6231.386 5083.667 878 23398
LIT 280 60.217 13.33567 38.48 90.92
EXPEND 224 0.1785047 0.0331068 0.1175275 0.336864
QGOI 280 6.633714 2.726396 0 10
ROADS 280 0.8681745 0.994643 0.19334 5.18382
WORK 280 38.36339 5.019369 30.88 46.04
INVEST 196 0.0052828 0.0027581 0.0002606 0.0145365


INVEST has only 196 observations, whilst EXPEND has only 224 observations

because the data is only available up to the year 1993 and 1995 respectively. All the

variables have a positive value apart from GROSDP, which sometimes is negative

representing a decline in a states growth rate.

The regression equation to calculate the determinants of PCNSDP is:

Yit = Po + 6i + 6t + P0LITit + p2EXPENDit+ p3QGOIit+ 04ROADSit + p5WORKit

+ 06INVESTit + Eit (7)

Whilst the regression equation to calculate the determinants of GROSDP is:

Zit = ao + 6i + 6t + aiYio + a2LITit + a3EXPENDit + a4QGOIit + asROADSit +

a6WORKit + a7INVESTit + rfit (8)

A low QGOI government is inefficient and holds back growth. As a state government

increases its power and control, a more stable environment is created which should

positively affect growth rates. Part of creating a more stable environment is developing

human capital. As LIT increases the economy of a state can diversify into new sectors

and grow. Increasing LIT has a positive impact not only on growth but the quality of life

in each state. An educated population tends to move out of the rural areas into the urban

areas where there are many more amenities and job opportunities. Factories tend to









locate in urban areas because of the more developed infrastructure. As ROADS increase

it is expected that more factories move into that area stimulating growth.

Expenditure on items such as maintenance of the organs of the state, administration

services, pensions and interest payments are types of non-development expenditure.

Excessive government debt holds back the economy because greater percentages of the

budget are spent on repaying debt rather than spending on the economy to stimulate

growth. An increase in non-developmental expenditure shall impact growth rates in a

negative way. Expenditure on education, health, various other social services and

economic services are types of developmental expenditure. An increase in

developmental expenditure should theoretically increase the growth rate. EXPEND is

comprised of both developmental and non-developmental government expenditure. The

balance of both types of expenditure that shall influence whether EXPEND has a positive

or negative impact on growth. From looking at the EXPEND data, it can be seen that a

greater percentage is spent on non-development, which implies that EXPEND should

have a negative effect on growth rate. Previous analysis by Barro has "found that the

ratio of real government consumption expenditure to real GDP had a negative association

with growth and investment" (Barro 1991, 430).

Increasing investment should theoretically have a positive effect on growth rates. As

INVEST increases, it shall stimulate the economy and create jobs. WORK shall have a

positive effect on growth rates because as more people work, more money shall be spent

and circulated through the economy.

The process of convergence is quickened by movements of people out of areas
where ratios of capital to workers are low-and hence wage rates and levels of per
capital income are also low-to areas where they are high (Barro and Sala-I-Martin
1991, 125).










Discussion and Interpretation of Empirical Analysis

The empirical research finds that the six explanatory variables upon which the

hypothesis of this paper is based, do significantly explain the differing PCNSDP between

the states for the three time periods, which are analyzed. All six of the explanatory

variables have the expected hypothesized sign.

Table 3-2. Determinants of PCNSDP from 1980 to 1999.
(l a) (2a) (3a) (4a)
LIT 142.96 144.21 21.29 0.3179
(10.001)*** (7.577)*** (1.825)* (0.012)

EXPEND -15488.42 -17737.75 -13261.98 -9564.51
(-3.597)*** (-5.022)*** (-5.199)*** (-3.892)***

QGOI 305.53 430.31 256.34 264.48
(6.344)*** (5.146)*** (6.6)*** (6.102)***

ROADS 1446.29 1895.97 236.76 421.36
(6.126)*** (5.250)*** (1.501) (1.436)

WORK 50.3 410.29 137.31 281.8
(1.812)* (5.818)*** (4.93)*** (6.952)***

INVEST 64593.69 91429.35 101100.8 103760.4
(1.291) (1.952)* (3.904)*** (4.314)***

Constant -5620.53 -22398.87 3038.88 -2790.46
(-4.936) (-10.339) (2.139) (-1.067)

Time Effect Yes Yes
State Effect Yes Yes

No. Obs 196 196 196 196
R-squared 0.5142 0.6886 0.7867 0.955
t-value appears in parenthesis
S Significance at the 0.10 level
S ** Significance at the 0.05 level
S *** Significance at the 0.01 level
All tests are two-tailed.











Table 3-3. Determinants of PCNSDP from 1980 to 1989.
(Ib) (2b) (3b) (4b)
LIT 76.71 93.17 107.1 132.93


EXPEND


QGOI


ROADS


WORK


INVEST


Constant


Time Effect
State Effect

No. Obs
R-squared


(9.148)*** (5.325)*** (11.331)***


-0.5658
(-12.532)***


-0.6599
(-15.099)***


285.35 430.8
(10.275)*** (6.17)***

820.94 624.47
(5.518)*** (2.017)**


4.31
(1.305)

36178.21
(1.399)

-562.72
(-1.087)




140
0.7092


74.56
(2.206)**

4429.85
(1.744)*

-3294.94
(-1.834)


Yes

140
0.7127


(11.529)***


-0.0386 -0.0123
(5.716)*** (6.217)****

125.44 38.78
(3.329)*** (2.373)**

1500.02 1813.84
(8.334)*** (8.442)***


15.38
(1.946)*

11080.2
(1.514)

-208.22
(0.271)

yes


140
0.887


44.44
(2.06)**

12671.72
(1.985)**

-2769.25
(-2.423)

yes
yes

140
0.894










Table 3-4. Determinants of PCNSDP from 1990 to 1999.
(Ic) (2c) (3c) (4c)
LIT 119.77 37.82 27.3 42.48
(5.424)*** (1.567) (1.749)* (1.964)*

EXPEND -21248.57 -9308.67 -14449.2 -9582.65
(-3.348)*** (-1.996)** (-3.141)*** (-2.156)*

QGOI 383.58 429.83 459.03 513.67
(6.149)*** (3.312)*** (4.334)*** 4.451)***

ROADS 562.8 586.7 237.3 103.19
(2.185)** (1.847)* (1.034) (1.931)*

WORK 24.85 247.36 70.92 111
(1.661)* (2.717)*** (1.750)* (1.661)*

INVEST 10369.23 238396.1 5375.73 28613.47
(1.991)** (2.289)** (2.081)** (2.405)**

Constant 6519.71 -3133.82 1029.71 5236.64
(3.010) (-1.149) (1.102) (1.299)

Time Effect yes Yes
State Effect Yes Yes

No. Obs 56 56 56 56
R-squared 0.4817 0.5557 0.8207 0.8445

Tables 3-2, 3-3 and 3-4 show the results of equation (7). Regression (1) uses the

ordinary least-squares (OLS) method for fitting a regression line. The panel data set used

in the empirical analysis replicates the observations across both states and time periods.

A fixed-effects approach uses the variation in explanatory variables over time to identify

regression coefficients. A fixed effect estimator controls for state-invariant determinants

and is used in (2). Time fixed effects control for time-invariant determinants, such as

cultural factors and is used in (3). Regression (4) controls for both time-invariant, and

state-invariant determinants by using both time, and state fixed effects.










Table 3-5. Determinants of GROSDP from 1980 to 1999.
(5a) (6a) (7a) (8a) (9a)
PCNSDP 0 -1.23e-06 -.000013 -.000027 -7.54e-06 -.00003
(-1.669)* (-4.032)*** (-5.964)*** (-1.345) (-3.316)***

LIT 0.002339 0.004384 0.001701 0.006683
(3.156)*** (3.374)*** (2.363)** (2.256)**

EXPEND -.500067 -.860475 -.585777 -.91722
(-2.677)*** (-3.767)*** (-2.354)** (-3.189)***

QGOI 0.002378 0.011668 0.001244 0.013733
(1.058) (2.188)** (0.547) (2.532)**

ROADS 0.031754 0.027279 0.027168 0.050273
(2.970)*** (1.183) (2.705)*** (1.514)

WORK 0.001788 0.007578 0.002152 0.010319
(1.512) (1.663)* (1.898)* (1.987)**

INVEST 1.32446 5.849269 0.992275 6.216752
(0.631) (2.029)** (0.501) (2.217)**

Constant 0.110297 0.187723 0.677908 0.176682 0.928707
(15.834) (3.686) (4.177) (0.018) (2.651)

Time Effect Yes Yes
State Effect Yes Yes

No. Obs 266 182 182 182 182
R-squared 0.008 0.1234 0.176 0.3203 0.3937
F-statistic 2.13 3.5 6.28 4.02 3.02











Table 3-6. Determinants of GROSDP from 1980 to 1989.
(5b) (6b) (7b) (8b) (9b)


PCNSDP 0


LIT


EXPEND


QGOI


ROADS


WORK


INVEST


Constant


Time Effect
State Effect

No. Obs
R-squared
F-statistic


-0.0000117 -0.0000252 -0.0000251 -0.0000288 -0.0000472
(-2.347)** (-2.655)*** (-2.055)** (-2.404)** (-2.671)***

0.002599 0.001504 0.002626 0.005209
(-2.201)** (0.609) (2.045)** (1.662)

-5.99e-06 -4.58e-06 -1.21e-06 -9.52e-06
(-1.834)* (-1.467) (-1.171) (-1.992)**

0.003202 0.002945 0.003346 0.004322
(1.765)* (1.303) (1.863)* (1.972)**

0.032314 0.010399 0.033985 0.05029
(1.783)* (1.993)** (1.731)* (2.005)**

0.001889 0.000581 0.001272 0.000931
(1.218) (2.134)** (2.888)*** (2.238)**

3.850545 6.987681 1.476478 3.89648
(1.278) (1.719)* (1.510) (1.992)**


0.073672 0.096471 0.13424
(-4.864) (1.687) (0.587)


126
0.0425
5.51


126
0.0919
1.71


126
0.1192
2.03


0.058191 0.204728
(1.024) (0.981)


126
0.3193
2.98


yes
yes

126
0.3523
3.52










Table 3-7. Determinants of GROSDP from 1990 to 1999.
(5c) (6c) (7c) (8c) (9c)
PCNSDP 0 -1.16e-06 -0.0000129 -0.0000247 -0.0000156 -0.0000551
(-1.985)** (-1.683)* (-1.762)* (-1.936)* (-2.693)**


0.002737 0.003122 0.001712 0.004822
(1.77)* (1.338) (1.981)* (1.945)*

-2.79e-06 -.000021 -1.57e-06 -1.75e-06
(-1.682)* (-1.265) (-1.779)* (-2.094)**

0.004234 0.006734 0.005647 0.007134
(1.744)** (2.005)** (1.731)* (2.134)**

0.024 0.051503 0.019423 0.071636
(1.688)* (1.288) (1.843)* (2.250)**

0.001785 0.004521 0.002562 0.007209
(1.753)* (1.705)* (1.747)* (1.755)*

1.427096 16.62934 3.277358 17.03527
(1.427) (1.798)* (1.699)* (1.936)*


LIT


EXPEND


QGOI


ROADS


WORK


INVEST


Constant


Time Effect
State Effect

No. Obs
R-squared
F-statistic


140
0.0067
0.93


56
0.0839
0.63


0.610512 0.010339 0.573957
(2.382) (0.098) (1.035)


56
0.2613
2.12


56
0.2359


Yes
Yes

56
0.4165
2.62


Table 3-5, 3-6 and 3-7 show the results of equation (8). Regression (5) tests for

absolute convergence. Regression (6) uses OLS to begin the tests for conditional

convergence. A state fixed effect is used in (7) and a time fixed effect is used in (8).

Regression (9) combines the approach of (7) and (8) by using both a state and time fixed

effect.

The natural rate at which the states PCNSDP are converging is seen in regression (5).

Regressions (5a), (5b) and (5c) all have negative and statistically significant t-values,

which demonstrate the existence of absolute 0-convergence. Comparing the regression


0.1403554 0.1295
(10.625) (1.341)









coefficients of (5b) and (5c) leads to the conclusion that absolute convergence amongst

the states is highest in the period from 1980 to 1989.

The t-values of PCNSDP in regressions (6) through (9) are all negative and

significant, which shows conditional convergence of the states PCNSDP levels. The

coefficients of regression (6) through (9) are all greater than the coefficient of regression

(5), which shows that adding additional variables increases the level of convergence in

PCNSDP. Comparing the regression coefficients of (6) through (8) in Tables 3-6 and 3-7

leads to the conclusion that conditional convergence amongst the states is highest in the

period from 1980 to 1989. Regressions (9b) and (9c) include both state and time fixed

effects which shows conditional convergence is greater in the period from 1990 to 1999.

The results do not show conclusively that the level of convergence in PCNSDP increased

in the time period from 1990 to 1999 due to the change in economic and political policy

during the early 1990's. Changes in economic and political policy do not have an

immediate effect on PCNSDP because of time lags. Time lags may explain why

convergence in PCNSDP levels is not always greater in the 1990 to 1999 time period

when compared to the 1980 to 1989 time period.

The empirical results show that human capital is one of the key determinants of

growth. LIT had a positive coefficient and was statistically significant in all the

regressions. Development data shows that ninety nine point nine percent of all U.S.

children are enrolled in primary school education. The contrasting figure for children

enrolled in primary school in India is only seventy seven point two percent.

The education system available in India's states is very good but expensive. Children

are generally educated at private schools, where parents pay the tuition. Those parents in









the lower income brackets are unable to give their children the opportunity to move up in

social stature. The children of the poor are destined for a seemingly unending cycle of

poverty they were born into. India does not have the capacity to create a publicly funded

education system, so a school system of two layers needs to be created. The first layer

shall comprise of the present schools that those who can afford attend. The second layer

comprises of schools with not so many amenities, but is financed by the government.

The focus of the government-funded schools would be to educate all children to a

standard that allows young adults to enter the work force. Unemployment shall decrease,

as the educated work force shall be able to use entrepreneurial spirit and enter markets

previously not seen in India. If the work force becomes educated other firms apart from

simple manufacturing firms shall want to locate in India, which will bring about a greater

demand on the educated population.

Kerala stands out for its relatively high level of development. In Kerala there are

many universities and religion affiliated schools that educate the children at no cost to the

parents. Literacy rates in Kerala are much greater than those found in the rest of India

because every child has the opportunity to attend school. Educating the population is not

sufficient by itself to create long-term economic growth. Kerala accounts for three point

four percent of India's population, however fifty percent of the countries migration

occurs from the young educated population leaving Kerala. Without the educated, a

country stands no real chance of advancing economically because they have a population

unable to carry out the high technology jobs.

Education is the key to decreasing the rate at which the states population is growing.

Kerala has shown that education can bring population growth rates down to zero percent









and this is what the other states need to focus on. Policies need to be created and

implemented in India to control the growing population. World Development data shows

the contraceptive prevalence rate in India to be only forty one percent, in contrast to the

seventy one percent rate of the United States.

The empirical analysis found that as crime rates, riots, industrial disputes and

corruption decreased, a states growth rate increased. QGOI was significant in almost all

the regressions and consistently had a positive t-value.

Public sector employees maybe unproductive because of low and declining civil

service salaries and promotions unconnected to performance. Workers become corrupt

when wages are not paid fully or on time and this is the fault of the government. Budget

funds needed to pay for wages maybe delayed because of the inefficient politicians. The

motivation to remain honest may be further weakened if senior officials and political

leaders use public office for private gain. Currently there is no employee protection

system for those who report the corrupt employees.

The cost of corruption cannot be measured, but is often interpreted. Recently the

World Bank complied data predicting the level of corruption found in India. The report,

according to informed sources said that, "because of widespread corruption, two to three

per cent of the country's $1.805 trillion GDP is lost annually". The report goes further by

stating "had there been some control on corruption, the country's present per capital

income of $350 would have been increased to $750".

The World Bank has formulated an anti corruption approach:

GI & AC = F (KI, LE, CA) (9)









Equation (9) shows that government improvement and anti corruption programs are a

function of knowledge, information, political leadership and collective action. The

World Bank has set out three primary functions to decrease corruption. For corruption to

decrease the Indian population must take collective action to end it, and not carry on

supporting corruption for their own benefits. Collective action is fundamental in putting

new election laws in place. The population has to be educated and made to understand

the long run negative effects and costs of corruption for them to begin demanding an end

in corruption. The Indian political leadership and quality of government can only be

changed through the ballet box.

In the future, Madhya Pradesh should experience a larger increase in PCNSDP

because of the states large step toward democracy compared to all the other states. A

state that allows political freedom shall increase the quality of its government by

allowing the population to use their democratic vote. In the long run Madhya Pradesh's

change toward democracy should increase growth, but to what extent is extremely

difficult to measure. In 1994 Madhya Pradesh was the first state to conduct elections to

panchayats. The result was to take power away from the representatives and hand it to

the people directly. The state government of Madhya Pradesh has moved from a directly

administrative roll to a supervisory roll.

Data on real annual average growth rates of per capital GSDP bear testimony to the
fact that four out of five states that are more policy reform oriented (with the
exception of Andhra Pradesh) are also the fastest growing states in India in the
post-reform period (Bajpai and Sachs, 1999).

Bihar, Uttar Pradesh, Madhya Pradesh and Orissa have lagged behind in both policy

reform and growth of PCNSDP.









The variable WORK was statistically significant in all regressions and had a positive

t-value. With the large population India already has, something has to be done to create

jobs. Addressing the issue of technology, India needs to focus on using labor-intensive

rather than capital-intensive technologies. An un-educated person may have a job, but

often it is not well paying enough to keep the individual above the countries nationally

established poverty line.

Financial and High-Tech services such as communications and information

technology industries are reliant on a well-educated work force. These industries also

require telephone and Internet connections, which are all characteristics of urban areas.

Urban areas have a higher quality of life that attracts the highly skilled mobile workers.

Maharashtra has the most well developed service industry among all fourteen states, with

the countries financial center and other important IT based industries being located within

the state. Tamil Nadu, Karnataka and Andhra Pradesh have relatively large service

sectors in relation to the other ten states. As the population becomes more educated,

Kerala has seen the majority of its population move from the primary sector to the

tertiary sector, which is a similar kind of movement to that seen in developed countries.

The Green Revolution dating from 1965 to the early 1980's created many jobs and

initiated agricultural-led- productivity growth. Irrigation and extensive fertilizer use

helped increase wheat and paddy rice yields. The Green Revolution was primarily in the

states of Haryana and Punjab, which have an ideal agro-climate. This short-lived burst in

agriculture led growth has not occurred in the period from 1980-99.

Agriculture jobs are seasonal and a village or town cannot survive on agriculture

alone. In an attempt to revive growth in rural areas, China has adopted a policy trying to









attract large manufacturing firms to these areas. China has remedied this problem by

attracting big corporations to Chinese towns and villages so as to absorb the surplus

labor. Corporations set up production plants in towns where the population can do both

farming and manufacturing work. The best example of this type of export led growth is

the arrival ofNike to the East Coast of China. In this situation the whole town is based

around Nike; almost all residents in the surrounding areas work in the factory. Most of

the villagers work at Nike during the day and go to the farm at night to make sure

everything is all right. Having factories locate in small towns and villages has created the

incentive needed to keep the rural population from migrating.

Urban areas have relatively well developed infrastructure networks in relation to rural

areas. ROADS hypothesized positive relationship with growth rate was supported in the

results. Theoretically, the more money allocated by states to building up infrastructure,

the faster a state should grow. Well-developed infrastructure attracts firms to locate in a

particular state because infrastructure helps businesses.

INVEST had a positive relationship with growth rate which was statistically

significant in most of the regressions. The states giving out more loans grew faster

because the money from loans was invested back into the state stimulating the states

economic growth. The poor economic performance of Kerala can be attributed to

insufficiency of investment, both private and public largely due to the failure of the

government to implement policies effective in encouraging private investment. Private

investment in Kerala is the lowest in comparison with all of the other states. Historically

the state's government has had a hostile attitude toward foreign investment. The lack of

policy development by the state government results in very low FDI. Due to a lack of









FDI, manufacturing only accounted for fifteen point five percent of Kerala's SDP in

1991.

The expected coefficient sign of EXPEND was difficult to predict because it depended

on the proportion of non-developmental to developmental expenditure. The variable

EXPEND was statistically significant in all regressions and had a constant negative

impact on both PCNSDP and GROSDP. For EXPEND to have had a negative sign, it

would imply that the majority of states spend more on non-development than

developmental expenditure.

The government spends almost half of its Non-Plan Expenditure budget on interest

payments. This payment is only going toward paying interest on the debt and not the

principal of the debt. In the long run economic growth is stifled because thirty three

percent of the budget is not being put back into the Indian economy. The proportion of

the budget allocated to subsidies is being put back into the economy. The amount of

money allocated to subsidies has decreased since 1947, but it still remains a substantial

portion of budget expenditures. The social services expenditure of 6,187 Crore Rupees

represents money allocated to education and health. A comparison made between the

money allocated toward defense and that allocated toward health and education, shows

only one seventh of the money allocated to defense is allocated toward education and

health. Money should be allocated to education instead of the military because allocating

money to education would have a long run multiplier effect upon economic growth

Rajasthan had an extremely high PCNSDP in the 1981-1990 time period, but after

1991 Rajasthan was no longer seen as the fastest growing state. Even though Rajasthan

received the highest per capital transfers and grants from the union government among the









four BIMARU states, and was one of the top recipients of federal transfers of all states,

this alone does not account for the high growth rates. Orissa received the second largest

per capital transfers, and was still the slowest growing state in the 1981-1990 period. The

other factors that help explain the high growth rate of Rajasthan are the rapid

electrification of the state and the building on the Command Canal in 1980 to help

agriculture. Rajasthan was also a beneficiary of the Green Revolution, which

contributed to the SDP just as the increased tourism did during this time. Raj asthan's

tourist industry has grown at the fastest rate among all fourteen states because of the

popularity of visits to Jaipur and Udaipur and the states proximity to Delhi.

Orissa is located on the coast but experiences some of the lowest PCNSDP. Orissa's

agricultural sector and economy suffer greatly from frequent flood devastation. The state

is one of the most well endowed with resources in relation to the other states. Orissa is

primarily an agricultural state despite its resource endowments. For this reason the poor

agricultural performances year after year hinders the SDP and growth rate. An important

factor that maybe explaining Orissa's low PCNSDP is the fact that twenty five percent of

the states population is tribal. Tribal populations are generally associated with lower

social indicators in health and education, and suffer social and political exclusion.















CONCLUSION

India abandoned the protectionist approach advocated between 1947 and 1990. The

liberalization approach has increased the rate at which India is developing. The new

approach has set the foundations from which good economic and political policy can be

created to help the states of India grow. Throughout the 1990's India has had many

economic problems, which have been resolved using foreign aid.

Theoretically, the relatively more urbanized states should experience economic

growth faster than states that have a greater proportion of its population in rural areas.

Urban areas have more of the variables that are both positively correlated and help

initiate growth. Well-developed educational facilities are concentrated in urban areas.

As the students finish their education and begin participation in the work force, they

should theoretically find well paying jobs in urbanized areas. The job openings occur as

the city and its infrastructure grows through government expenditure, attracting new

businesses into the area. New businesses invest in the state economy, which stimulates

economic growth. Urban areas cannot efficiently develop if the state government is of a

low quality because of government inefficiencies. If government expenditure is not used

to build roads, schools and other infrastructure, the growth process shall be hindered.

The empirical evidence from a sample of 14 states, in the time period from 1980 to

1999 supports the theoretical argument. The six explanatory variables LIT, EXPEND,

QGOI, ROADS, WORK and INVEST explain significantly the determinants of growth.

A state run by a high quality government and endowed with human capital as well as









infrastructure has the foundations from which to grow. The balance of investment and

state expenditure are key determinants of both job creation and economic growth. The

empirical results show that both absolute and conditional convergence among the states

PCNSDP exists. The results do not show conclusively that the level of convergence in

PCNSDP increased in the time period from 1990 to 1999 due to the change in economic

and political policy during the early 1990's.

The study implies the government role should be to develop human capital and

infrastructure through development expenditure. Job creation is both the role of the

government through developmental expenditure, and the population through investment.

The study implies the primary role of the population should be to increase the quality of

government by use of their democratic vote.















LIST OF REFERENCES

Ahluwalia, M.S. (2001). State Level Performance Under Economic Reforms in India.
Working Paper No. 96, Center for Research Economic Development and Policy
Reform, Stanford University, Stanford, CA.

Bajpai, Nirupam, and Jeffery D. Sachs. (1999). Progress of Policy Reform and
Variations at the Sub-National level in India. Development Discussion Paper No.
730, Harvard Institute for International Development, Harvard University,
Cambridge, MA.

Barro, R. (1991). Economic Growth in a Cross Section of Countries. The Quarterly
Journal of Economics, Vol. 106, Issue 2, pp. 407-443.

Barro, R., and Xavier Sala-I-Martin. (1991). Convergence Across States and Regions.
Brookings Papers on Economic Activity, Vol. 1991, Issue 1, pp. 107-158.

Barro, R., and Xavier Sala-I-Martin. (1992). Convergence. The Journal of Political
Economy, Vol. 100, Issue 2, pp. 223-251.

Barro, R., and Xavier Sala-I-Martin. (1995). Economic Growth. New York: McGraw-
Hill, Inc.

Basu, Sudip. (2002). Does Governance Matter? International Economics Department at
University of Geneva, Geneva, Switzerland.

BBC News. (1999). South Asia Analysis: The Cost of Conflict. Accessed 1 Feb 2003 at
http://news.bbc.co.uk/l/hi/world/south_asia/370558.stm.

Cashin, P., and Sahay, R. (1996). Internal Migration, Center-State Grants, and
Economic Growth in the States of India. IMF Staff Papers, Vol. 43, Issue 1.

Cass, D. (1965). Optimum Growth in an Aggregate Model of Capital Accumulation.
Rev. Econ. Studies Vol. 32, pp. 233-240.

Census of India. 1991 and 2001. Provisional Population Totals. Office of the Registrar
General, New Delhi.

Central Intelligence Agency. (2002). The World Fact Book 2002. Accessed 1 Feb 2003
at http://www.cia.gov/cia/publications/factbook/geos/in.html.









Chaudhury, D. (1974). Behavior of Spatial Income Inequality in a Developing
Economy: India 1950-76. Paper presented at the Ninth Conference of the Indian
Association for Research in National Income and Wealth, Seoul.

Dasgupta, Maiti, Mukherjee, Sarkar, and Chakrabarti. (2000). Growth and Interstate
Disparities in India. Economic and Political Weekly, Vol. 1005, Issue 27, pp.
2413-2422.

Dholakia, R. (1994). Spatial Dimension of Acceleration of Economic Growth in India.
Economic and Political Weekly, Vol. 1009, No.35, August 21.

Economic Intelligence Unit. (1997). Country profile on India.

Economic Survey. (1980-2002). Mumbai: Ministry of Finance, Government of India.

IMF. (1995). India: Economic Reform and Growth. Occasional paper 134. Washington:
International Monetary Fund.

Jannuzi, Tommasson. (1989). India in Transition. London: Westview Press.

Koopmans, Tjalling. (1965). On the Concept of Optimal Economic Growth. The
Econometric Approach to Development Planning. Amsterdam: North-Holland.

Majumdar, Grace, and J L Kapoor. (1980). Behavior of Inter-State Income Inequalities
in India. Paper presented at the Twelfth Conference of the Indian Association for
Research on National Income and Wealth, Bombay.

Ramsey, P. (1928). A Mathematical Theory of Saving. Econ. J., Vol. 38, pp. 543-559

Rao, Shand, and Kalirajan. (1999). Convergence of Incomes Across Indian States A
Divergent View. Economic and Political Weekly, Vol. 1004, Issue 13, pp. 1-10.

Report on Currency and Finance. (1980-95). Reserve Bank of India, Bombay, Vol. 2.

Sachs, Varshney, and Bajpai. (1999). India in the Era of Economic Reforms. Oxford:
University Press.

Sala-I-Martin, X. (1997). I Just Ran Two Million Regressions. The American
Economic Review, Vol. 87, Issue 2, pp. 178-183.

Solow, R. (1956). A Contribution to the Theory of Economic Growth. Q.J.E., Vol. 70,
pp 65-94.

Stern, Nicholas. (1991). The Determinants of Growth. The Economic Journal, Vol. 101,
pp. 122-133.

The World Bank Group. (2003). The World Bank in India. Accessed 1 Feb 2003 at
http://lnweb 1 8.worldbank.org/sar/sa.nsf/India?OpenNavigator.






39


World Bank. (1997). India: Five years of stabilization and reform and the challenges
ahead. Washington: World Bank.

WTO. (1998). Trade policy review on India. Geneva: World Trade Organization.
















BIOGRAPHICAL SKETCH

My family originates from the state of Gujarat in India. I was born and raised in

England and moved to the United States in March of 1998. I graduated from the

University of Texas at Dallas with a Bachelor of Arts in economics and finance in August

of 2001. Currently I am at the University of Florida studying towards a Master of Arts in

economics.