understanding regional economic growth in india...major driver of economic growth, and the coastal...

Under standing Regi onal Economic Gro wth in IndiaUnder standing Regi onal Economic Gro wth in India

Understanding Regional EconomicGrowth in India

Jeffrey D SachsDirectorThe Earth Institute atColumbia UniversityColumbia UniversityB-16 Hogan Hall2910 BroadwayNew York NY 10027USASachscolumbiaedu

Nirupam BajpaiSenior Development Advisor ampDirector India ProgramCenter for Globalization andSustainable DevelopmentThe Earth Institute atColumbia UniversityColumbia UniversityB-16 Hogan Hall2910 BroadwayNew York NY 10027USANb2046columbiaedu

Ananthi RamiahCenter for InternationalDevelopment (CID)Harvard University79 John F Kennedy StCambridge MA 02138USAananthi_ramiahyahoocouk

Asian Economic Papers 13 copy 2002 The Earth Institute at Columbia University and the Massachusetts

Institute of Technology

AbstractThis paper aims to explain the growth experiences of 14 majorstates of India between 1980 and 1998 Using two measures ofconvergence s-convergence and ucirc-convergence we examinewhether per capita incomes in the states have been converging ordiverging By both standards of convergence India demonstratedoverall divergence during 1980ndash98 as well as during both the pre-reform and post-reform subperiods Interestingly the richer statesexperienced a degree of convergence during the post-reform pe-riod whereas the poorer states did not Divergence was most no-table within the poorer group of states A remarkable 82 percentof the cross-state variation in growth is explained by just the ur-banization variable in India with no hint of any conditional conver-gence after controlling for the degree of urbanization The regres-sion estimate shows that a 10 percentage point higher rate ofurbanization is associated with 13 percentage points per yearhigher rate of annual growth

1 Introduction

India accounts for a meager 24 percent of the worldrsquos sur-face area yet it sustains a whopping 167 percent of theworldrsquos population amounting to a little over 1 billionpeople residing in 28 states and 7 union territories Thevariation in physical geography culture and economicconditions across these states and territories is enormousSome states have achieved rapid economic growth in re-cent years whereas others have languished This paperattempts to explain the differential economic performance

This paper was prepared for the Asian Economic Panel meetingheld in Seoul Korea on 25ndash26 October 2001 and was presentedto the Prime Minister of India on 25 December 2001 in NewDelhi

of Indiarsquos states especially under the forces of globalization in the 1990s The papermay be read most pro tably as a companion to Deacutemurger et alrsquos recent work (2002)on regional differences in Chinarsquos economic performance

We focus on the 14 most populous states in India which excludes the Himalayanstates the northeastern states and the 7 union territories (UTs) The included stateshave a combined population of 897 million (approximately 90 percent of Indiarsquospopulation) and cover 27 million km2 (83 percent of Indiarsquos total land area) Someimportant data for the 14 states are listed in table 1 and the statesrsquo locations areshown in gure 1 Economic performance varies greatly among the 14 states Thegross state domestic product (GSDP) per capita ranges from 1261 rupees per monthin 1997ndash98 in the poorest state Bihar (population 82 million) to 5690 rupees permonth in the richest state Maharashtra (population 96 million) Some states haveachieved rapid economic growth in recent years such as Gujarat with growth instate per capita income at 78 percent per year from 1992ndash93 to 1998ndash99 whereasothers have languished such as Bihar at 202 percent per year over the same period

The differential economic performance across states raises important policy ques-tions To what extent are the differences a manifestation of global economic forcesacting upon India especially during a period of economic liberalization and towhat extent do they re ect differences in economic policies at the local and nationallevels Will market reforms tend to make the rich states richer in relative terms orwill they lead to economic convergence across states Speci cally are the pooreststates (especially the so-called BIMARU states of Bihar Madhya Pradesh Rajasthanand Uttar Pradesh) condemned to fall farther behind the front-runners at least inrelative terms

Deacutemurger et al (2002) found that the underlying drivers of economic growth inChina and hence the tendencies toward convergence or divergence differed mark-edly across subperiods as a result of major shifts in the economic policy regime In1978ndash84 during the rst phase of Chinarsquos market reforms the dismantling of thecommunes and the partial liberalization of food production gave a great boost tomajor food-producing regions By the late 1980s international trade became themajor driver of economic growth and the coastal regions outpaced the interiorprovinces

Policy regimes have also affected economic growth in India In the state planningperiod up to 1991 international trade played only a minor role and industrializa-tion was heavily affected by state investment plans which attempted (at leastmildly) to promote the laggard regions The Green Revolution which introduced

33 Asian Economic Papers

Understanding Regional Economic Growth in India



Tab

le1

Ind

ian

stat

esM

ain

char

acte

rist

ics

GS

DP

per

cap

ita

1997

ndash98

(rup

ees

per

mon

th)

Gro

wth

inG

SDP

per

capi

ta

1980

ndash98

()

Coa

stal

acce

ss(

ofp

opu

lati

onw

ithi

n10

0k

mof

coas

t)

Mai

ncl

imat

ezo

ne(p

erce

nta

geof

land

area

)

Maj

orpo

rtci

ty(p

opul

atio

ngr

eate

rth

an5

mil

lion

)Tr

opic

alA

rid

Subt

ropi

cal

And

hra

Prad

esh

252

12

942

5445

1Bi

har

126

11

00

50

95G

uja

rat

450

54

371

1782

0H

arya

na4

516

31

00

8713

Kar

nata

ka3

109

36

1940

600

Ker

ala

282

33

11

10

0M

adhy

aP

rade

sh2

286

25

035

956

Mah

aras

htra

569

04

534

6139

0M

umba

iO

riss

a1

871

17

5667

033

Punj

ab5

079

30

00

7624

Raj

asth

an2

621

38

07

8013

Tam

ilN

adu

345

44

365

964

0C

henn

aiU

ttar

Prad

esh

202

32

00

58

82W

estB

enga

l3

308

33

4367

033

Kol

kata



Figure 1 Indian States

Source Registrar General and Census Commissioner (2001)

Note P = Pondicherry

certain high-yield grain crops to India led to sharp increases in grain productivityin regions such as Punjab and Haryana which were suited for the improved crops(mainly wheat) In the market reform period after 1991 market forces and interna-tional trade played a larger role in Indiarsquos economic growth although Indiarsquos en-trance into the global economy has been much less dramatic than Chinarsquos Wewould expect the coastal regions to have advantages over the interior regions after1991 because they face much lower transactions costs in participating in global tradeand investment

Several studies of high-income market economies undertaken during the 1990s forthe United States Japan and regions within western Europe found evidence forstrong convergence among regions (see Barro and Sala-i-Martin 1995 chap 11) Sim-ilar to the ndings for China we found little evidence of comparable convergenceamong Indian states In China and India geographical variation across regions ap-pears to block or slow the convergence of incomes

This paper is organized as follows Section 2 reviews the results of previous studiesof convergence among Indian states during the period 1980ndash98 discusses measuresof convergence and summarizes our conclusions on interstate convergence Insection 3 we account for the lack of convergence discussing the social and demo-graphic factors that affect the economic performance of the 14 states under study Insection 4 we attempt to unravel some mysteries at the state level Policy implica-tions and directions for future research are discussed in section 5

2 Convergence among Indian states (1980ndash98)

21 Previous studies of convergence across Indian states

Several studies covering different time periods examine whether per capita incomelevels have been converging or diverging in India and most nd a tendency towarddivergence rather than convergence Nair (1971) nds no noticeable reduction in in-terstate income differentials between 1950 and 19601 and Chaudhury (1974) con-cludes that the degree of state income inequality did not change between 1950 and1970 Majumdar and Kapoor (1980) suggest that interstate inequalities of income inIndia steadily increased during 1962ndash76

Rao Shand and Kalirajan (1999) suggest that per capita state domestic product(SDP) in the Indian states has tended to diverge rather than converge Per capita



1 Nair concludes that neither changes in the degree of industrialization nor changes in laborproductivity helped reduce income disparities

SDP growth is positively related to the initial levels of income States with better in-frastructure and human resources have been able to attract more investment in thepost-reform era Dasgupta et al (2000) also report a distinct tendency for per capitaSDP in Indian states to have diverged during 1960ndash952 Kurian (2000) nds widen-ing regional disparities among the Indian states and a clear dichotomy betweenwhat he calls the forward and backward states The forward states have higher lev-els of per capita income better infrastructure higher per capita resource ows andprivate investment and better social and demographic indicators

In contrast Dholakiarsquos (1994) analysis of 20 Indian states over the period 1960ndash90 nds marked tendencies of convergence of long-term GSDP growth rates3 This ap-pears to be a result of including the 5 special-category Indian states4 and Delhi withthe 14 major Indian states Cashin and Sahay (1996) also reach similar conclusions nding absolute convergence in a study of 20 states over the period 1961ndash91

Ahluwalia (2001) analyzes the economic performance of the Indian states during thepost-reform period and suggests that not all the richest states got richer relative topoorer states for example Punjab and Haryana were the two richest states in 1990ndash91 but their per capita SDP growth rates in the 1990s not only were lower than inthe 1980s but actually fell below the national average Ahluwalia also points outthat two poor states Rajasthan and Madhya Pradesh had performed well in termsof growth over the 1980s but he does not offer an explanation for this result

In general the studies that focus on the most populous states uncover little evidenceof convergence although there may be some convergence of the small northeasternstates with the rest of the country Our study which is based on data from the 14most populous states and leaves aside the question of the northeastern areas simi-larly nds evidence of overall divergence rather than convergence

22 Measures of convergence

There are two standard ways of examining the presence or absence of unconditionalconvergence (see Barro and Sala-i-Martin 1995 383ndash87) The rst measure is s-conver-gence We measure the standard deviation across regions of the logarithm of the realGSDP per capita There is s-convergence if the standard deviation across states tends



2 In terms of the shares of the different sectors within each statersquos SDP Dasgupta et al (2000) nd a tendency for increasing similarity across states in sectoral composition

3 Dholakia identi es 1980 as the turning point when several of the lagging states started grow-ing and the leading states began to stagnate

4 Assam Himachal Pradesh Jammu and Kashmir Manipur and Tripura are classi ed as spe-cial-category states by the Planning Commission because of their mountainous terrain

to decline over time The second measure is b-convergence Here we regress the pro-portionate growth in per capita SDP on the logarithm of initial income There isb-convergence if the coef cient of initial income denoted b is negative and statisti-cally signi cant

In addition to looking at s-convergence and b-convergence across the 14 states wealso divide the states into two groups based on GSDP per capita and examine con-vergence within these two subgroups5 Group I includes the high-income states ofMaharashtra Punjab Haryana Gujarat Tamil Nadu West Bengal Karnataka andKerala Group II includes the poor states of Rajasthan Andhra Pradesh MadhyaPradesh Uttar Pradesh Orissa and Bihar Table 2 shows that the states in group Iranked within the top eight positions in GSDP per capita in all the years inquestion6

-convergence measure Figure 2 shows the standard deviation across the statesof the log of real GSDP per capita We can see that the 14 states as a group show anincreasing standard deviation between 1980 and 1990 and an increased rise in thatstandard deviation in the post-reform period There was an overall rise in inequalityof 240 percent per annum between 1980 and 1998 Of the two periods within thattime frame the 1992ndash98 period experienced the most divergence 214 percentper annum compared to 124 percent per annum in 1980ndash90 Thus there was nos-convergence during either of the periods in question

Group II states exhibited much greater volatility in dispersion than group I states inaddition to exhibiting the greater absolute dispersion (even though group II statesstarted off with considerably lower standard deviations) This is the case for bothperiods From 1980 to 1990 group I states experienced an increase in dispersion ofa mere 124 percent per annum compared to 251 percent per annum within thegroup II states During 1992ndash98 group I states exhibited a decrease in dispersion of035 percent per annum whereas group II states witnessed an increase of 461 per-cent per annum Thus the richer states experienced a degree of convergence duringthe post-reform period whereas the poorer states did not



5 We have used only states with populations above 10 million Delhi is not included (despitehaving a huge population) because it suffersbene ts from factors related to being the capi-tal city Assam is not included because of incomplete data

6 All data in this paper (unless otherwise indicated) are calculated from the Economic and Po-litical Weekly Research Foundationrsquos National Accounts Statistics of India 1950ndash51 to 1996ndash97(EPWRF 1998) GSDP per capita levels for 1997ndash98 and 1998ndash99 are calculated fromAhluwalia (2001)

-convergence measure The lack of s-convergence is mirrored by a lack ofb-convergence Thus there is a positive coef cient on initial income in table 3which shows the results of regressions of the growth during each subperiod on thelog of initial income Regressions are given for 1980ndash98 1980ndash90 and 1991ndash98

During 1980ndash90 growth patterns were divergent The state with the highest GSDPper capita in 1980 was Punjab (Rs3020 per month) and the state with the lowestGSDP per capita was Bihar (Rs1062 per month) From 1980ndash81 to 1990ndash91 Punjabrsquosgrowth rate was 378 percent per annum and Biharrsquos was 294 percent per annumPunjab was the 6th-fastest-growing state and Bihar the 10th In general the richeststates had the highest growth rates (see table 4) The only notable exception isRajasthan which had the second-lowest initial GSDP per capita level but was thefastest-growing state during the 1980s Apart from Rajasthan all states grew in amanner that perpetuated divergent trends during the pre-reform period

Do states exhibit convergent trends during the post-reform period The fastest-growing state is Maharashtra but it also has one of the highest GSDP levels of the 14states Overall there is a signi cant positive relationship between initial income in1990 and growth during the 1990s indicating divergence The regression evidencesuggests that b-divergence was especially marked for group I states in the post-reform period



Figure 2 Standard deviation (StDev) of log real per capita GSDP for group I group II andall states



Table 2 Fourteen Indian states ranked according to GSDP per capita levels (highest income= 1 lowest income 14)

1980ndash81 1991ndash92 1996ndash97 1998ndash99

1 Punjab Punjab Maharashtra Maharashtra2 Maharashtra Haryana Punjab Punjab3 Haryana Maharashtra Haryana Gujarat4 Gujarat Gujarat Gujarat Haryana5 West Bengal Tamil Nadu Tamil Nadu Tamil Nadu6 Kerala Karnataka West Bengal West Bengal7 Karnataka West Bengal Karnataka Karnataka8 Tamil Nadu Kerala Kerala Kerala9 Andhra Pradesh Andhra Pradesh Rajasthan Rajasthan

10 Madhya Pradesh Rajasthan Andhra Pradesh Andhra Pradesh11 Uttar Pradesh Uttar Pradesh Madhya Pradesh Madhya Pradesh12 Orissa Madhya Pradesh Uttar Pradesh Uttar Pradesh13 Rajasthan Orissa Orissa Orissa14 Bihar Bihar Bihar Bihar

Table 3 -convergence regressions (Dependent variable growth rate)

1980ndash98 1980ndash90 1991ndash98

Log of initial per capita real GSDP(All) 20 (254) 123 (109) 291 (191)Log of initial per capita real GSDP(group I) 202 (018) 132 (086) 934 (347)Log of initial per capita real GSDP(group II) 44 (151) 2066 (011) 2174 (062)

Table 4 Rates of growth of per capita GSDP in Indian states

State (rankedfrom highest in-come in 1980ndash81)

Growth rate of percapita GSDP from1980ndash81 to 1990ndash91



Punjab 38 Punjab 28Maharashtra 43 Haryana 26Haryana 46 Maharashtra 68Gujarat 38 Gujarat 78West Bengal 23 Tamil Nadu 50Kerala 25 Karnataka 35Karnataka 36 West Bengal 48Tamil Nadu 50 Kerala 46Andhra Pradesh 29 Andhra Pradesh 37Madhya Pradesh 29 Rajasthan 44Uttar Pradesh 30 Uttar Pradesh 16Orissa 10 Madhya Pradesh 39Rajasthan 58 Orissa 16Bihar 29 Bihar 202Unweighted average 35 Unweighted average 38

In table 5 we decompose the interstate variance of income as equal to the variancewithin groups I and II and the variance between the two groups7 The method usedfor calculating variance decomposition is described in the technical appendix From1980 to 1990 variance increased within groups I and II as well as between the twogroups From 1992 to 1998 there was a smaller increase in dispersion in overall vari-ance of 202 percent compared with 355 percent in the earlier period8 Over thewhole period there was an increase in dispersion within group I of a mere 08 per-cent per annum whereas the dispersion within group II increased by 1717 percentper annum There was also an increase in the dispersion of average incomes be-tween the two groups of 383 percent per annum Thus we conclude that the in-crease in overall dispersion during the second period was driven by increased dis-persion within the group II states

Conclusions on interstate convergence By both standards of convergence Indiademonstrated overall divergence during 1980ndash98 and during the pre-reform andpost-reform subperiods Divergence was most notable within the poorer group ofstates This nding is consistent with the experience of China in the post-reformperiod but differs from the ndings for the United States Japan and Europeanregions

3 Economic performance of the states

We hypothesize that regional differences in growth re ect regional differences in themarginal productivity of investments by subsector To some extent the relative re-turns to investment in each subsector depend on the general business environmentbut to an important extent they also depend on speci c geographical factors Someof these geographical considerations are listed in table 6

Agriculture can occasionally be a leading sector in economic growth as a result ofeither a spurt in agricultural productivity or increased cash crop exports In Indiaagricultural productivityndashled growth occurred in one major historical period theGreen Revolution dating from the mid-1960s to the early 1980s The Green Revolu-tion originated from the introduction of short-stemmed high-yield wheat and to alesser extent paddy rice These crops depended on irrigation and intensive applica-



7 The variance decomposition is calculated as follows overall variance = a (within group Ivariance) + b (within group II variance) + c (average between group I and group II vari-ances) The constants a b and c are weights that depend on the number of states in eachgroup (see technical appendix for a more detailed explanation)

8 We omitted 1991 because it represents a structural break in the data resulting from the eco-nomic crisis in India

tions of fertilizer9 The epicenter of the Green Revolution was in Punjab andHaryana and to a lesser extent other states of the North Indian Plains (as far east asBihar) and southward to Rajasthan Gujarat and Maharashtra High-yielding ricevarieties made the most impact in West Bengal and Tamil Nadu Note that Chinalike India experienced one short-lived burst of agriculture-led growth during thedismantling of the communes food output jumped during 1978ndash84

Almost all the regions among the group II states (except perhaps much of westernRajasthan and parts of western Madhya Pradesh and southern Uttar Pradesh) havethe agro-climatic potential to yield high returns in agriculture because of reasonableto high rainfall and availability of perennial river waters The poverty in these statesis a consequence of human failure rather than of natural factors Although it may beuseful to identify states with high or low incidences of poverty there are states withhigh variations within them owing to historical and economic antecedents and agro-climatic factors This is typically true of the larger states though such variations ex-ist in smaller states as well

A more disaggregated National Sample Survey (NSS) region-wise picture of poverty(head count ratio) shows that there are signi cant heterogeneities in poverty in eachstate except perhaps Bihar which is uniformly poor Sharp contrasts are witnessedin Andhra Pradesh Karnataka and Maharashtra though variations can be seen insmaller states like Haryana and Punjab as well The regions can be segregated ac-



Table 5 Variance decomposition within and between group I and group II states

Within group I Within group IIBetween group Iand group II Overall variance

1980 000576 000131 000187 0008941981 000618 000162 000206 0009861982 000672 000185 000226 0010831983 000683 000162 000223 0010671984 000608 000084 000183 0008741985 000765 000096 000228 0010891986 000748 000060 000213 0010211987 000699 000125 000218 0010411988 000765 000158 000244 0011671989 000784 000184 000256 0012241990 000744 000214 000253 0012111991 000661 000194 000226 0010811992 000698 000320 000269 0012861993 000624 000307 000246 0011771994 000624 000374 000264 0012621995 000606 000427 000273 0013071996 000648 000488 000300 0014351997 000654 000523 000311 0014871998 000664 000559 000323 001546

9 Most rice farmers were too poor to introduce the high-yield varieties of rice which requiredsubstantial applications of fertilizer as well as irrigation Also the taste of the rice was unfa-miliar and public acceptance was limited Therefore rice became the ldquoorphanrdquo of the GreenRevolution (Tirtha 2000 268)

cording to their poverty levels low (up to 20 percent) medium (21ndash40 percent) high(41ndash60 percent) and very high (more than 60 percent) Southern Bihar southernOrissa southwestern Madhya Pradesh and southern Uttar Pradesh fall in the very-high-poverty bracket These regions are composed of the districts in Chotanagpurand Santhal Parganas in Bihar Koraput and Phulbhani districts in Orissa the Jhansiregion in Uttar Pradesh and its adjacent regions in Madhya Pradesh includingBetul Khandwa and Hoshangabad These regions are either mainly tribal (exceptJhansi) or rocky and dry yet densely populated because of their agro-climatic fea-tures The main inference drawn is that tribal areas are predominantly and distinctlypoor

The areas of high poverty are in Bihar portions of Madhya Pradesh inlandMaharashtra northern Tamil Nadu eastern and central Uttar Pradesh and parts ofWest Bengal These are generally tribal thickly populated semi-arid areas andthose areas that have been neglected historically are poor West Bengal has madestrides in poverty alleviation in some areas Medium-level poverty persists in re-gions of the western states and a few regions have made more progress than otherscompared to the eastern ones where there is uniform poverty Typical examples areMadhya Pradesh Maharashtra Tamil Nadu and Uttar Pradesh Pockets of low pov-erty include the western coastal regions all of Andhra Pradesh Punjab and parts ofMadhya Pradesh and Rajasthan which form a north-south belt that was affected bythe Green Revolution

The manufacturing sector is a much more consistent engine of growth that haslikely played a growing role since 1991 with the opening of the economy As Chinarsquosexperience demonstrates trade liberalization in a low-wage surplus-labor environ-



Table 6 Some regional factors in sectoral growth

Key determinantsfor future growth Favored states Time period

Currentcontribution togrowth

Agriculture Climate agriculturaltechnologies

Punjab Haryana Green Revolution1970s to 1980s

Low

Manufacturing Urban coastalmajor port facilitiesattraction of foreigndirect investment

Maharashtra WestBengal TamilNadu Gujarat

1980s onwardwith risingimportance in the1990s

High

Tourism Historical culturaland natural attractionsproximity to majorports of entry such asDelhi and Mumbai

RajasthanMaharashtra

1980s onward Moderate

High-techservices ( nanceand ICT)

Urban skilled laborforce universities

Maharashtra TamilNadu KarnatakaWest Bengal

1990s High

Note ICT = information-communication technology

ment permits a rapid expansion of export-oriented industry that can absorb a largenumber of workers to provide goods for the world market Indiarsquos insertion into theworld economy has been much less dramatic and successful than Chinarsquos The shareof exports of goods and services in GNP was stable at 7 percent in 1980 and 1990 itrose to 11 percent in 1999 after the reforms (World Bank 1997 2001 table 13) InChina the comparable share rose from 6 percent in 1980 to 18 percent in 1990 and to22 percent in 1999 Several factors may account for these differences Chinarsquos re-forms were bolder in promoting both foreign direct investment (FDI) and manufac-tured exports (Bajpai and Sachs 2000) and China bene ted from the vast in ows ofFDI from overseas Chinese investors especially from Hong Kong Taiwan andSoutheast Asia

The most likely site for sustained manufacturing growth in India as in China isalong the coast especially at the four large port cities of Mumbai (Maharashtra)Kolkata (West Bengal) Chennai (Tamil Nadu) and Kandla (Gujarat) Coastal ur-ban-based industry can serve both the internal market and the international marketand can more readily make logistical links with foreign suppliers and customersthan can interior-based enterprises New export-oriented units (EOUs) are thereforeheavily concentrated on the coast (table 7) Manufacturers in interior regions can ofcourse service the domestic market particularly in consumer goods such as pro-cessed foods but the potential for rapid growth based on the internal market tendsto be more limited than the potential growth based on exports to the world marketFor this reason we have seen much faster growth in coastal China than in theinterior

The tourist sector can also be a source of export-led growth but in a country the sizeof India it is likely to play a secondary role except in some local niches Tourist po-tential is very much geographically determined as it depends on the physical envi-ronment the presence of historical sites and easy access to transport nodes espe-cially international airports Rajasthan has been the state with the most signi cantgrowth and scale of the tourist industry because of the popularity of Jaipur andUdaipur and its proximity to Delhi

High-tech services such as information- and communications-based industry (egsoftware production) or nancial services usually rely on a network of universitiesand an urban labor market These sectors are much less dependent on coastal accessbecause much of their business can be transacted by telephone or via the Internet Ahigh quality of life of the location as an attraction for highly mobile skilled workersis probably more important for these service sectors than for other sectors of theeconomy The most important state for service-sector activities is Maharashtra it



combines the countryrsquos nancial center with an important information technology(IT)-based industry Other key states include Tamil Nadu Karnataka Delhi and to alesser extent Andhra Pradesh (see table 8)

Foreign investors have various motivations to service the domestic market to ex-ploit site-speci c natural resources (eg mining) and in low-wage countries to es-tablish export platforms in labor-intensive goods in labor-intensive stages of theproduction process or in standardized technologies that are easily transferable tolower-wage settings Coastal access is a huge bene t for all export platform manu-facturing as we have seen clearly in the case of China FDI is also attracted to urbanareas and to regions with natural-resource deposits Interior cities (such asBangalore and Hyderabad) may be suitable for IT-based activities that do not de-pend on coastal access Although the data on state-by-state FDI are spotty table 9shows the cumulative FDI approvals by state on an aggregate and per capita basisfor 1991ndash2001 The following simple regression con rms that FDI as a percentage ofGSDP owed mainly to the urbanized states and to the states with large mining sec-tors (especially Orissa and to a lesser extent Madhya Pradesh)10

FDI per capita 5 24682 1 220 ( urban) 1 459 ( mining in GSDP)(404) (203)

N 5 14 R2 5 062

These considerations suggest that urbanization is likely to have been a key determi-nant of economic growth in the 1980s and 1990s because existing urban areas were



Table 7 Distribution of export-oriented units (August 1991 to May 2001)

Stateunion territory Number of unitsPercentage of totalEOUs

EOUs per millionpopulation (2001)

Andhra Pradesh 404 110 53Bihar 7 02 01Gujarat 458 124 91Haryana 206 56 98Karnataka 376 102 71Kerala 72 20 23Madhya Pradesh 136 37 23Maharashtra 563 153 58Orissa 41 11 11Punjab 127 35 52Rajasthan 205 56 36Tamil Nadu 547 148 88Uttar Pradesh 206 56 12West Bengal 98 27 12

Source Secretariat for Industrial Approvals (SIA) Newsletter June 2001 Ministry of Industry Government of India

Note Group I states are in boldface type

10 The dependent variable is cumulative approvals of FDI in rupees per capita

the preferred location for new investments in manufacturing and services As wesee in table 10 the extent of urbanization varies widely among the states from a lowof 13 percent in Bihar and Orissa to 39 percent in Maharashtra (as of 1991) with therelative proportions of urbanization by state relatively constant over the past 30years The degree of urbanization itself depends on underlying geographical factorsespecially the location of the main national ports and the productivity of agriculturein the region Regions of high agricultural productivity tend to support a larger pro-portion of the local population in an urban setting whereas regions of low agricul-tural productivity tend to have a high proportion of the population in peasant sub-sistence agriculture Empirically we nd that two factors account for two-thirds ofthe variation in urbanization rates across the 14 states having a major port (eg in



Table 8 Software exports by state 1995ndash96

StateValue of software exports(Rs millions)

Software exports per capita(Rs millions)

Andhra Pradesh 9313 129Bihar mdash mdashGujarat 551 12Haryana 6299 347Karnataka 72784 1509Kerala 387 12Madhya Pradesh 25 00Maharashtra 70854 821Orissa mdash mdashPunjab 90 04Rajasthan mdash mdashTamil Nadu 31167 533Uttar Pradesh mdash mdashWest Bengal mdash mdash

Source httpwwwmaharashtragovinenglishinvestsoftwhtm (accessed as of 4 April 1999)


Table 9 Flows of foreign direct investment to Indian states

Foreign direct investment approved (August 1991 to July 2001)

StatesInvestment(Rs in millions)

FDI per capita(1991ndash2001 per 2001 population)

Andhra Pradesh 124701 1646Bihar 7395 89Gujarat 168556 3331Haryana 31948 1515Karnataka 208156 3947Kerala 14361 451Madhya Pradesh 91542 1515Maharashtra 456286 4716Orissa 82289 2241Punjab 19519 803Rajasthan 25917 459Tamil Nadu 222804 3587Uttar Pradesh 42048 253West Bengal 84235 1050

Source Secretariat for Industrial Approvals (SIA) Newsletter August 2001 Ministry of Industry Government of India

Maharashtra Tamil Nadu West Bengal and Gujarat) and having a dry steppe cli-mate (Bs) suitable for wheat production11

Urbanization () in 1991 5 194 1 111 (major port) 1 126 ( population in Bs climate)(99) (41) (28)

N 5 14 R2 5 068

The rate of growth of GSDP per capita is highly correlated with the extent of urban-ization at the beginning of the period in question The results of six regressions areshown in table 11 In a regression of growth for 1980ndash98 on initial income in 1980and urbanization as of 1981 the urbanization coef cient is highly signi cant with acoef cient of 013 and a t-statistic of 53 (regression 1) A remarkable 82 percent ofthe cross-state variation in growth is explained by this variable alone with no hintof any conditional convergence after controlling for the degree of urbanization Theregression shows that a 10 percentage point higher rate of urbanization is associatedwith 13 percentage points of higher annual growth A simple bivariate regressionwithout initial income shows the same results (regression 2) The relationship be-tween 1981 urbanization [e(urban81|X)] and growth [e(gr8098|X)] during 1980ndash98is shown in gure 3

During 1980ndash90 the Green Revolution continued to play a role in growth differen-tials across states To capture the effect of the Green Revolution we construct adummy variable equal to 10 in Punjab and Haryana (the epicenter of the areas



Table 10 Rates of urbanization

1971 1981 1991

Andhra Pradesh 193 (7) 233 (7) 268 (7)Bihar 100 (13) 125 (13) 132 (14)Gujarat 281 (3) 311 (3) 344 (2)Haryana 177 (8) 219 (8) 248 (9)Karnataka 243 (5) 289 (4) 309 (4)Kerala 162 (11) 187 (11) 264 (8)Madhya Pradesh 163 (10) 203 (10) 232 (10)Maharashtra 312 (1) 350 (1) 387 (1)Orissa 84 (14) 118 (14) 134 (13)Punjab 237 (6) 277 (5) 297 (5)Rajasthan 176 (9) 211 (9) 229 (11)Tamil Nadu 303 (2) 330 (2) 342 (3)Uttar Pradesh 140 (12) 180 (12) 199 (12)West Bengal 248 (4) 265 (6) 274 (6)

Source Registrar General and Census Commissioner (1981 1991)

Note Group I states are in boldface type Figures in parentheses indicate relative rankings

11 The major port variable takes the value of 1 for the states Maharashtra West Bengal TamilNadu and Gujarat and 0 otherwise The Bs variable is the proportion of the population(0 to 100 percent) living in the Bs climate zone of the Koeppen-Geiger climate classification

where the high-yield wheat is grown) 025 in Gujarat Maharashtra and Rajasthanand 0 elsewhere This variable has a coef cient of 354 (t = 401) in the growth re-gression for the period 1980ndash90 holding constant the initial income and the degreeof urbanization in 1981 Rajasthan is an outlier in this regression with growth morethan one percentage point per year faster than is otherwise explained (table 11 re-gression 3) This may be attributable to the boom in tourism or to the rapidelectri cation of the state in the 1980s or to a more signi cant effect of the GreenRevolution than is captured by the value 025 or to some other unmeasured effectInterestingly if one holds constant the urbanization variable the Green Revolutionvariable and the Rajasthan dummy variable there is evidence of conditional con-vergence with the slower states achieving faster growth than the richer states Thisis the only regression result in which we nd this conditional convergence

By the 1990s the Green Revolution effect has disappeared entirely as has the condi-tional convergence and the fast growth of Rajasthan (regression 4)12 The only vari-able that accounts for cross-state growth in the 1990s is urbanization as of 1991 (re-gression 5) with a point estimate of the urbanization coef cient that is somewhathigher than the coef cient for the 1980s (030 compared with 013) A simplebivariate regression of growth in the 1991ndash98 period on urbanization in 1991 is alsoshown (regression 6) with this single variable accounting for 71 percent of the varia-tion Economic liberalization may have added to the growth-promoting bene ts ofurbanization especially for the coastal cities and the main cities engaged in IT ex-ports (eg Bangalore Chennai Hyderabad and Delhi)



Table 11 Growth equations 1980ndash98

Regression 1 2 3 4 5 6

Period 1980ndash98 1980ndash98 1980ndash90 1991ndash98 1991ndash98 1991ndash98

Ln (initial income) 2004 2468 008 2183(008) (329) (002) (135)

Urbanization 013 013 019 025 030 023(530) (731) (530) (209) (475) (538)

Green Revolution 354 2127(401) (048)

Rajasthan dummy 156 161(243) (124)

Constant 039 005 3317 2329 1025 2233(009) (012) (337) (012) (109) (198)

R2 082 082 088 079 075 071

Note Figures in parentheses are t-statistics For the equations beginning in 1980 urbanization in 1981 is used for the regressions be-

ginning in 1991 urbanization in 1991 is used

12 The coefficient of the dummy variable for Rajasthan is still positive but is no longer statisti-cally significant

Consistent with our view of urban-led (and export-led) growth during the past de-cade we nd that the gap between urban and rural expenditure per capita has wid-ened in the past decade Table 12 shows the urbanrural expenditure ratios by statefor 1987ndash88 and 1999ndash2000 and the percentage change of these ratios in the two peri-ods Simple regression analysis (not shown) con rms that the urbanrural ratio in1999ndash2000 is higher in the more urbanized states and lower in the states thatbene ted most from the Green Revolution (Punjab and Haryana) The increase inthese ratios is highest in the urbanized states although Bihar is an outlier because ithas a large increase in the urbanrural ratio but a relatively low level of urbaniza-tion This result has also recently been reported by Radhakrishna (2002 248) ldquoThebene ts of better income growth during the 1990s seem to have bene ted the urbanareas more than the rural thus aggravating the urban-rural dividerdquo Radhakrishnaalso cites Bhallarsquos (2000) result that employment growth during the 1990s was con-centrated in urban areas

Since the initiation of economic reforms in 1991 Indian states have competed witheach other to attract private investment both domestic and foreign Within statesthe investment has tended to ow to the urban areas State-level data on FDI ap-provals (aggregate FDI approvals between 1991 and 2001) suggest that the relativelyfast-moving reformers have attracted higher levels of FDI (see table 9) Gujarat (pop-



Figure 3 Relationship of urbanization (1981) and economic growth (1980ndash98)

ulation 50 million) received over a fth of the private investment proposals whereasBihar (population 83 million) barely managed a 5 percent share of such proposalsMaharashtra and Gujarat accounted for 37 percent of the total investment proposalsbut Bihar Madhya Pradesh Orissa Rajasthan and Uttar Pradesh together were ableto attract only 26 percent of such proposals

We have not been able to analyze the role of public sector investment in cross-stategrowth especially the central-government funds allocated to the state governmentsin Indiarsquos ve-year investment plans The data we have collected to date are toospotty for a serious analysis There are hints that such investments have made a dif-ference at least in some places and times The rapid electri cation of Rajasthanrsquos vil-lages in the 1980s and 1990s (table 13) probably helps account for the otherwise un-explained growth of that state

31 Accounting for the lack of convergence

It is surprising but robustly the case that after controlling for urbanization alonethere is no evidence whatsoever of conditional convergence We did not nd anycandidate explanatory variables that once they were controlled for allowed signs ofconditional convergence to emerge This poses a major issue of interpretation Whyis it that the US states displayed unconditional convergence in most decades of UShistory as did Japanese prefectures and European regions but India and China donot show signs of conditional convergence much less unconditional convergenceThere are several possible hypotheses for the lack of unconditional convergence

1 The geographical differences are larger in India and China than in the UnitedStates Europe and Japan



Table 12 Ratio of urban to rural expenditures 1987ndash88 and 1999ndash2000 (total expenditures)

State TE UR 1987ndash88 TE UR 1999ndash2000 Percentage change

Andhra Pradesh 1438 1705 18567Bihar 1365 1563 14505Gujarat 1493 1617 8305Haryana 1173 1277 8866Karnataka 1494 1823 22021Kerala 1259 1218 23256Madhya Pradesh 1662 1727 3910Maharashtra 1739 1959 12651Orissa 1766 1657 26172Punjab 1106 1210 9403Rajasthan 1338 1450 8371Tamil Nadu 1613 1890 17173Uttar Pradesh 1458 1479 1440West Bengal 1664 1905 14483All India 1581 1759 11259

Source For 1987ndash88 the data are from National Sample Survey Organization (1991) For 1999ndash2000 the data are from National Sam-

ple Survey Organization (2001)

Note TE UR total expenditure urbanrural ratio

2 Population movements in the United States Europe and Japan more readily ar-bitrage differences across regions

3 Policies of the national or regional governments have prevented convergence4 Economic convergence is easier at levels of economic development higher than

those present in China and India

We nd some merit in each of these possibilities Certainly the intrinsic economicadvantages or disadvantages of Japanese prefectures and western European regionsare much smaller than those found in the different regions of either India or ChinaConsider coastal access for example (table 14) In Japan 97 percent of the popula-tion lives within 100 km of the coast In the European Union 51 percent of the popu-lation lives within 100 km of the coast and 89 percent lives within 100 km of thecoast or a sea-navigable waterway (eg the Rhine or the Danube)13 A surprisinglyhigh proportion of the US population 65 percent lives within 100 km of the coastor sea-navigable waterway

One reason that the United States has such a high proportion of the population atthe coasts and along navigable waterways is that it has highly ef cient agriculturewhich can feed the entire population (and much more) with just 2 percent of the la-bor force For this reason few people in the United States are ldquobound to the landrdquoin the economic sense of needing to be in the place where food is grown With muchlower food productivity in China and India (a re ection of the long history of muchhigher manland ratios in Asia) a much larger part of the population is needed toproduce food This means that populations are ldquostuckrdquo in the interior of the country



Table 13 Percentage of villages with electricity

1985 1998 Change


Source 1985 data are from the Seventh Five Year Plan (1985ndash90) Planning Commission Government of India p 164 1998 data are

from Economic and Political Weekly January 13 2001


13 By sea-navigable waterway we mean that seagoing vessels may proceed up and down thewaterway to and from the sea

and much less able to participate in international trade and globalized productionsystems (eg outsourcing for multinational rms) Large numbers of poor near-subsistence farmers live in the hinterland of China and India in regions that are notpart of convergent growth except to the extent that households migrate in largenumbers

Climatic variability is much lower in Europe the United States and Japan than it isin India and China (see table 15) Substantial proportions of Indiarsquos population livein tropical arid subtropical and highland ecozones whereas the overwhelmingproportion of US European and Japanese populations reside in temperateecozones China also has large variations in climate but only a small proportion ofChinarsquos population lives in tropical ecozones which have proved most dif cult fordevelopment in other parts of the world (including India) Similar to the variation inaccess to the sea the climatic variation most likely puts a brake on cross-regionalconvergence

This brings us to the question of migration In China migration is limited by thehousehold registration system which has blocked the legal migration of familiesfrom the hinterland to urban areas In India migration is not restricted yet poorfamilies without social safety nets apparently face such high costs and risks associ-ated with migration that internal labor ows are not powerful enough to createforces of convergence14 In the case of China the policy regime in the 1980s and1990s favored the already-favored coastal provinces and this accounted for part ofthe continuing divergence between coastal and interior regions In India no suchpreferential policies are readily discernible

32 Social and demographic factors in convergence and divergence

We examined whether social and demographic factors could account for cross-stategrowth patterns The results were surprisingly negative The states vary consider-



Table 14 Proximity of population to coast

RegionPercentage of populationwithin 100 km of coast

Percentage of populationwithin 100 km of coast orsea-navigable waterway

India 22 38

China 19 45Japan 97 97United States 35 65European Union (EU 15) 51 90

Source CID Geography Database (2001)

14 Exactly why this is the case is beyond the scope of this paper but is certainly worthy ofmuch closer investigation

ably in social indicators such as infant mortality rates (tables 16 and 17) Generallythe southern states have outperformed the northern states by a wide margin andKerala has outperformed all states However this superior performance has nottranslated into discernibly higher rates of economic growth at the state level15

Kerala has reaped the returns of improving the literacy and health of its populationmigration has increased and large amounts of remittance income have owed backto the state This remittance income is counted as part of the statersquos income but notas part of its GSDP which is the measure of economic development used in thisstudy Thus Keralites have a higher income standard than is measured by the pro-duction within the state We speculate on why this is so in the next section

4 Unraveling some mysteries at the state level

This section explores four unusual aspects of state-level performance in India(1) the mediocre growth of Kerala despite its excellent social indicators (2) the rela-tively fast growth of landlocked arid Rajasthan (3) the improved growth perfor-mance of landlocked Madhya Pradesh and (4) the poor growth performance ofcoastal Orissa

41 Kerala

During the period in question Kerala has moved from sixth to eighth place in percapita GSDP rankings it grew at 25 percent between 1980 and 1990 and 46 percentbetween 1992 and 1998 Although its growth performance has improved making itone of the biggest bene ciaries of the reform period (moving from the 12th- to the6th-fastest-growing state) it is unclear why Kerala is not among the top states ingrowth performance given its outstanding results in health and education



Table 15 Percentage of population in ecozones 2001

Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)

Temperate(Cf Cs Dw Df E)

Highland(H)

India 33 32 28 0 7

China 0 10 36 30 24Japan 0 0 0 100 0United States 0 0 19 70 10European Union (15) 0 0 1 98 1

Source CID Geography Database (2001) using Koeppen-Geiger climate classi cation

Note Aw = tropical wet and dry Am = tropical monsoon Bs = dry semiarid (steppe) Bw = dry arid (desert) Cf = humid subtropical

coast Cs = humid subtropical mediterranean Cw = humid subtropical dry winters Dw = continental mid-latitude dry winters

Df = continental mid-latitude wet all seasons E = polar H = highland

15 The simple correlation of growth with literacy is positive but it disappears once we controlfor urbanization which is correlated with growth and the degree of literacy

Kerala has had a very small manufacturing base In 1981 manufacturing as a per-centage of GSDP was 139 percent by 1991 that gure had risen to a meager 155percent Subrahmanian (1990) argues that Keralarsquos poor economic performance canbe attributed to a limited focus on the commodity production sector The secondarysector areas that have seen growth are construction and power rather than areas in-volving actual manufacturing activity Resource-based industries do not generallyprovide the growth potential of manufacturing (capital goods or demand-based) in-dustries in part because the latter are likely to lead to intersectoral linkages andtechnical progress whereas the former have a tendency to stagnate The lopsided in-dustrial structure of Kerala is a symptom however and not a cause of its low levelof growth the cause is insuf cient public and private investment and a lack of effec-



Table 16 Infant mortality rate (deaths under age 1 per 1000 live births)

State 1990 1998


Source EPWRF (2000)


Table 17 Life expectancy at birth (years)

State 1981 1993

Andhra Pradesh 598 628Bihar 515 580Gujarat 593 620Haryana 590 640Karnataka 620 639Kerala 715 756Madhya Pradesh 519 546Maharashtra 621 658Orissa 530 562Punjab 636 684Rajasthan 538 594Tamil Nadu 574 644Uttar Pradesh 485 560West Bengal 580 628All India 598 628

Source Central Bureau of Health Intelligence Ministry of Health and Family Welfare Government of India 1994


tive government policies to encourage private investment16 Keralarsquos private invest-ment is a mere 177 percent of GSDP the lowest in India Kerala also receives verylittle FDI (see table 9)

The lack of FDI and private investment in Kerala is also the consequence of the rela-tively hostile attitude of state government to private investment and of the highlymilitant labor force The Communist Party which governed the state for many yearsuntil 2001 had part of its base in the unions and hence did not resist the labor mili-tancy This reputation for aggressive labor tactics has discouraged private investorsboth domestic and foreign over the years The change in government to Congressparty rule may lead to an improved investment climate in the state

It should be noted that in recent years Keralites working abroad (mainly in the Gulfcountries) have remitted about Rs6000 crore17 annually which is about one- fth ofKeralarsquos domestic product and is three times greater than the budget support thestate receives from the central government Between 1980 and 1995 more thanRs31350 crore owed into the state from the Gulf18 Since remittance income iscounted as part of the statersquos income but is not included in its GSDP Kerala may notbe as poor a state as its GSDP gures suggest

42 Rajasthan

Rajasthan recorded the highest overall growth in the pre-reform period (table 4) Itsagriculture tourism construction and service sectors grew the most among all 14states and it was among the top three group II states in manufacturing growthWhat spurred Rajasthanrsquos incredible growth spurt in the 1980s One part of the an-swer is that Rajasthan received the highest per capita transfers and grants from theunion party government among the four BIMARU states However other states(eg Orissa) also received large transfers but did not experience such successful eco-nomic growth Four additional factors probably contributed to Rajasthanrsquos highgrowth (1) the bene ts of the Green Revolution in the wheat-growing areas of thestate (2) the positive impact on agriculture from the construction of RajasthanrsquosCommand Canal in the early 1980s (3) the tremendous increase in tourism duringthe 1980s and 1990s and (4) the rapid electri cation of the state that took place inthe 1980s



16 There are also historical reasons for Keralarsquos pattern of industrial development but they arebeyond the scope of this work

17 One crore equals 10 million rupees

18 M G Radhakrishnan India Today 1 May 2000

43 Madhya Pradesh

Madhya Pradesh increased democratic participation in the 1990s making it the statewith the highest degree of direct democracy Madhya Pradesh was the rst state tohold elections to panchayats in 1994 (Behar 2001)The goal was to give power to thepeople directly rather than to their representatives Kumar (2001) explains how em-powering the people has enabled the government to overcome dif cult situationssuch as water shortages The state supplies the panchayats with funding and techni-cal support in order to ldquowiden de-silt and deepen village ponds dig new wells andbuild dam checksrdquo (Kumar 2001 18) The state government has moved toward a su-pervisory rather than a directly administrative role

The extent to which enhanced democratization has aided growth is unclear Onewould expect the reforms resulting from direct democracy to operate in the longterm rather than the short term Possibly increased growth occurred because the re-forms were well targeted and effectively implemented

44 Orissa

Orissa has traditionally been one of Indiarsquos poorest states It had the third-lowestGSDP per capita in 1980 and was also the slowest-growing state in the 1980s (1 per-cent per annum) which is partly attributable to its even lower agricultural growth(072 percent) It is unclear why agricultural production has fared so poorly giventhat its soil quality and irrigation are on par with those of the more successful statesOrissarsquos mines and quarries are the most productive in India and their output grewby 156 percent during 1980ndash90 adding to the mystery of poor performance

In the post-reform period Orissa grew somewhat more rapidly than in its pre-reform phase (16 and 10 percent respectively) but it was near the bottom of thestates in growth performance (12th during 1991ndash98 and 14th during 1980ndash90) Tosome extent Orissa became an innovator in economic reforms it was the rst state toreform its power sector it has strongly set out industrial policy promoting privatesector investment and offering scal and other concessions and it was the rst stateto announce new agricultural and tourism policies

In the post-liberalization period Orissa ranks sixth in foreign investment This is theresult of a change in the statersquos economic policy which has directed investors to-ward Orissarsquos abundant natural resources As Singh (1997) observed ldquoOrissa has 90percent of Indiarsquos chrome ore and nickel reserves 70 percent of bauxite and 24 per-cent of coal reserves With no other state having such abundance of natural reourcesthe big business houses have no option but to set up steel alumina and coal based



power projects in Orissa onlyrdquo Orissarsquos post-reform ability to attract both privateand foreign investment suggests that its pre-reform failure was primarily a result ofineffective or nonexistent industrial policies to exploit its mineral wealth

It is unclear why Orissarsquos agriculture sector has grown so poorly In the post-reformperiod the sector declined further experiencing a contraction of 05 percent Oneexplanation may be Orissarsquos vulnerability to oods and devastation from tropicalcyclones As Orissa is predominantly an agricultural state its poor agriculturalperformance augurs very badly for its future income growth and income distri-bution

5 Policy implications and directions for future research

The main nding of this study is that the forces of convergence (absolute and condi-tional) are very weak in the 14 Indian states examined We expect that growth willcontinue to occur in those states in which urbanization is already high perhaps be-cause of coastal access or the relatively high productivity of agriculture There is lit-tle evidence to suggest that growth will equalize across regions This assessment ishardly a pessimistic one however for several reasons First there is much more po-tential for growth in India than has been achieved to date Whereas the per capitagrowth rates in Indiarsquos states have varied between 2 and 8 percent per annum theper capita growth rates in Chinarsquos provinces have ranged between 8 and 13 percentper annum during 1992ndash98 (Deacutemurger et al 2002 table 4) Many Indian coastal cit-ies have not begun to attract foreign direct investment for export-led growth Themost stunning example is Thiruvananthapuram (Trivandrum) Kerala which boastsa skilled labor force and a natural harbor but has almost no FDI With a new stategovernment declaring its intention to begin attracting FDI it is quite likely that FDI-led exports from Kerala will begin to grow Other coastal cities that could be muchmore dynamic include Cochin also in Kerala Bhubaneshwar the capital city ofOrissa and Vishakhapatnam the major port of Andhra Pradesh

The reasons why these other coastal port cities have not rapidly developed are nu-merous The continuing power of the central government over regional infrastruc-ture (airports major highways power and telecommunications) has certainly re-duced the capacity of state governments to implement rapid economic reforms InChina provincial governments have had ample leeway to make key infrastructureinvestments in India the powerful monopoly state enterprises in key infrastructuresectors have resisted competition especially from potential foreign investors This ischanging but gradually and not without continued resistance



India like China but unlike the United States boasts several cities with populationsgreater than 1 million that are far from the coast or navigable waterways19 Lucknow(Uttar Pradesh) Kanpur (Uttar Pradesh) Hyderabad (Andhra Pradesh) andBangalore (Karnataka) are 100ndash500 km from the sea and Delhi Jaipur (Rajasthan)Bhopal (Madhya Pradesh) and Nagpur (Maharashtra) are 500ndash1000 km from thesea These Indian cities can readily support industry for the internal market as wellas IT-based services Moreover with a new high-quality internal highway systembetween these major cities even these inland urban areas could become export ori-ented Establishing improved transport and communications networks (including ber-optic cables) among the major cities is a high priority

India will likely continue to face the same problems as China in the inland areasparticularly the inland rural areas Even with faster overall growth the inland areasare likely to grow more slowly than the coastal areas which is likely to provoke po-litical pressures and to increase internal migration from rural areas to cities andfrom the interior to the coast India however may have an advantage over China inthat the westernmost regions of China are much farther from the coast than Indiarsquosheavily populated interior regions of the Gangetic valley China has four cities ofmore than 1 million that are more than 1000 km from the coast (Chengdu LanzhouUrumqi and Xian) whereas India has none

As in China a careful balance will have to be struck between two kinds of invest-ments in the rural hinterland (eg in Uttar Pradesh and Bihar) physical infrastruc-ture such as roads rail airports and telecommunications to bring these regionscloser to the international markets and investments in human capital mainly edu-cation and health to raise the productivity of the rural population The latter invest-ments may serve to attract new rms eager to bene t from an increasingly skilled la-bor force or they may provoke large-scale migration to more economically vibrantcoastal regions Either way the currently impoverished populations would bene tfrom rising living standards wherever in India they are enjoyed

Appendix

Data sources and description

Our data cover the period 1980ndash98 in 14 major states of India The states not in-cluded are Arunachal Pradesh Assam Chattisgarh Delhi Goa Himachal Pradesh



19 Eleven US cities with populations of 1 million or more are remote from the coast but in allcases except Dallas they are close to a navigable waterway These large interior cities areBuffalo Chicago Cincinnati Cleveland Dallas Detroit Milwaukee Minneapolis Pitts-burgh Portland and St Louis

Jharkhand Meghalaya Mizoram Nagaland Pondicherry Uttaranchal and SikkimData are not available for these states over a large number of years The data usedfor analyzing income convergence or divergence among Indian states are from In-dian of cial statistical publications Various volumes of the Central Statistical Orga-nization (CSO) publications provided the estimates of SDP The concept of SDP usedin the paper is based on income originating in the state rather than income accruingto the state

The primary sector consists of agriculture and allied activities and the shery for-estry and mining and quarrying subsectors The secondary sector includes manu-facturing construction electricity gas and water supply The tertiary sector com-prises transport communication storage hotels and restaurants nance and realestate banking and insurance public administration community and personal ser-vices and other services We have used the share of the primary sector in SDP tomeasure the changes in the economic structure and their effect on growth By totalpopulation we mean the total number of people in the state as of the end of a partic-ular year CSO data on state-wise real capita SDP are arrived at by using total popu-lation along with SDP de ated by the base year prices

Some of the data are from the Sample Registration System (SRS) of the RegistrarGeneral India The SRS is a large-scale demographic sample survey that providesreliable annual estimates of the birth rate death rate and other fertility and mortal-ity indicators at the state and national levels and for rural and urban areas By pro-viding age-speci c mortality rates SRS data also facilitate the construction of life ta-bles at birth and at selected ages for the above disaggregated categories includingthose for males and females separately The age composition of the population fromtwo successive censuses provides data for the construction of decadal life tables butthe SRS alone provides trends in life expectancy at more frequent intervals (annuallyin recent years)

The SRS uses a dual-record system for collection of data First continuous enumera-tion of births and deaths is undertaken by a part-time resident enumerator Secondan independent retrospective survey every six months is conducted by a computersupervisor The data obtained through these two processes are matched The un-matched and partially matched events are re-veri ed in the eld and thereafter anunduplicated count of births and deaths is obtained This procedure is said to elimi-nate errors of duplication and allow a quantitative assessment of sources of distor-tion in the two sets of records that is it ensures a cross-check on the correctness andcompleteness of the data collected The SRS was introduced in 1969ndash70 after the ab-



sence of dependable data from the Civil Registration System began to be felt in themiddle of the 1960s

Data on the level of investment in individual states comparable with the investmentdata at the national level obtained from the national accounts are simply not avail-able

Technical appendix

There are 14 states (N = 14) 8 are in group I (n1 = 8) and 6 are in group II (n2 = 6)Variance20 is calculated as follows

Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N

where xi is the log of real GSDP per capita We proceed to add and subtract themeans within group I and group II denoted x1 and x2 respectively Some of theterms disappear because the sum of the deviations from the mean is zero leaving uswith

=-

- + - + - + -aring aring aring11 1

21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var

In our sample the constants yield an equation of the form

0538Var(group I) + 0385Var(group II) + 0264Var(average between groups)

References

Ahluwalia Montek Singh 2001 State Level Performance under Economic Reforms in IndiaWorking paper no 96 Stanford Center for Research on Economic Development and Policy Re-form Stanford University

Bajpai Nirupam and Jeffrey D Sachs 2000 Indiarsquos Decade of Development Working paperno 46 Cambridge Center for International Development Harvard University

Barro Robert J and Xavier Sala-i-Martin 1995 Economic Growth New York McGraw-Hill

Behar Amitabh 2001 Madhya Pradesh Gram SwarajndashndashExperiment in Direct Democracy Eco-nomic and Political Weekly 36 (22) March 10



20 The source of these equations is Jian Sachs and Warner (1995)

Bhalla Surjit 2000 Behind Poverty The Qualitative Deterioration of Employment Prospectsfor Rural Indians Working paper no 7 New Delhi Institute for Human Development

Cashin Paul and Ratna Sahay 1996 Internal Migration Center-State Grants and EconomicGrowth in the States of India IMF Staff Papers 43 (1)123ndash71

Central Bureau of Health Intelligence Ministry of Health and Family Welfare Government ofIndia 1994 Health Information of India New Delhi

Central Statistical Organization Department of Statistics Government of India Estimates ofState Domestic Product Various issues New Delhi

Chaudhury Mahindar D 1974 Behaviour of Spatial Income Inequality in a Developing Econ-omy India 1950ndash76 Paper presented at the ninth conference of the Indian Association for Re-search in National Income and Wealth

CID Geography Database 2001 Cambridge Center for International Development HarvardUniversity

Dasgupta Dipankar Pradip Maiti Robin Mukherjee Subrata Sarkar and SubhenduChakrabarti 2000 Growth and Interstate Disparities in India Economic and Political Weekly 35(27) July 12413ndash22

Deacutemurger Sylvie Jeffrey Sachs Wing Thye Woo Shuming Bao Gene Chang and AndrewMellinger 2002 Geography Economic Policy and Regional Development in China Asian Eco-nomic Papers 1 (1)146ndash97

Dholakia Ravindra 1994 Spatial Dimension of Acceleration of Economic Growth in IndiaEconomic and Political Weekly 24 (35) August 21

Economic and Political Weekly Research Foundation (EPWRF) 1998 National Accounts Statis-tics of India 1950ndash51 to 1996ndash97 Mumbai India EPWRF

Economic and Political Weekly Research Foundation (EPWRF) 2000 Data provided to the au-thors by the EPWRF Mumbai India

Jian Tianlun Jeffrey D Sachs and Andrew Warner 1995 Trends in Regional Inequality inChina Development discussion paper no 518 Cambridge Center for International Develop-ment Harvard University

Kumar Venkatesh B 2001 Madhya Pradesh Panchayats and Water Scarcity Economic and Po-litical Weekly 36 (29) July 21ndash27

Kurian N J 2000 Widening Regional Disparities in IndiamdashSome Indicators Economic and Po-litical Weekly 35 (7) February 12ndash18538ndash50

Majumdar Grace and J L Kapoor 1980 Behavior of interstate income inequalities in IndiaPaper presented at the twelfth conference of the Indian Association for Research on NationalIncome and Wealth

Ministry of Finance Government of India 1995 Economic Survey 1994ndash95 New Delhi

Ministry of Planning and Programme Implementation Department of Statistics Governmentof India Annual Survey of Industries Various issues New Delhi

mdashmdashmdash Statistical Abstract Various issues New Delhi



Nair K R G 1971 A Note on Inter-state Income Differentials in India 1950ndash51 to 1960ndash61 Jour-nal of Development Studies 7 (1)441ndash47

National Sample Survey Organization Department of Statistics Ministry of Planning Govern-ment of India 1991 Sarvekshana 15 (1) issue no 48 JulyndashSeptember

mdashmdashmdash 2001 Report no 457 May New Delhi

Radhakrishna R 2002 Agricultural Growth Employment and Poverty A Policy PerspectiveEconomic and Political Weekly 36 (3) January 19243ndash50

Rao M Govinda Ric T Shand and Kali P Kalirajan 1999 Convergence of Incomes across In-dian StatesmdashA Divergent View Economic and Political Weekly 34 (13) March 27769ndash78

Registrar General and Census Commissioner Census of India 1981 and 1991 Provisional Popu-lation Totals New Delhi Of ce of the Registrar General

mdashmdashmdash 2001 Census of India 2001 Provisional Population Totals New Delhi Of ce of the Reg-istrar General

Singh Kalwajit 1997 Corpwatch (on the Web) MNC Masala August httpwwwcorpwatchorgtracfeatureindiaglobalizationorissahtml

Subrahmanian K K 1990 Development Paradox in Kerala Analysis of Industrial StagnationEconomic and Political Weekly September 15

Tirtha Ranjit 2000 Geography of India Jaipur and New Delhi Rawat Publications

World Bank 1997 World Development Report 1997 The State in a Changing World New York Ox-ford University Press

mdashmdashmdash 2001 World Development Report 20002001 Attacking Poverty New York Oxford Univer-sity Press



of Indiarsquos states especially under the forces of globalization in the 1990s The papermay be read most pro tably as a companion to Deacutemurger et alrsquos recent work (2002)on regional differences in Chinarsquos economic performance

We focus on the 14 most populous states in India which excludes the Himalayanstates the northeastern states and the 7 union territories (UTs) The included stateshave a combined population of 897 million (approximately 90 percent of Indiarsquospopulation) and cover 27 million km2 (83 percent of Indiarsquos total land area) Someimportant data for the 14 states are listed in table 1 and the statesrsquo locations areshown in gure 1 Economic performance varies greatly among the 14 states Thegross state domestic product (GSDP) per capita ranges from 1261 rupees per monthin 1997ndash98 in the poorest state Bihar (population 82 million) to 5690 rupees permonth in the richest state Maharashtra (population 96 million) Some states haveachieved rapid economic growth in recent years such as Gujarat with growth instate per capita income at 78 percent per year from 1992ndash93 to 1998ndash99 whereasothers have languished such as Bihar at 202 percent per year over the same period

The differential economic performance across states raises important policy ques-tions To what extent are the differences a manifestation of global economic forcesacting upon India especially during a period of economic liberalization and towhat extent do they re ect differences in economic policies at the local and nationallevels Will market reforms tend to make the rich states richer in relative terms orwill they lead to economic convergence across states Speci cally are the pooreststates (especially the so-called BIMARU states of Bihar Madhya Pradesh Rajasthanand Uttar Pradesh) condemned to fall farther behind the front-runners at least inrelative terms

Deacutemurger et al (2002) found that the underlying drivers of economic growth inChina and hence the tendencies toward convergence or divergence differed mark-edly across subperiods as a result of major shifts in the economic policy regime In1978ndash84 during the rst phase of Chinarsquos market reforms the dismantling of thecommunes and the partial liberalization of food production gave a great boost tomajor food-producing regions By the late 1980s international trade became themajor driver of economic growth and the coastal regions outpaced the interiorprovinces

Policy regimes have also affected economic growth in India In the state planningperiod up to 1991 international trade played only a minor role and industrializa-tion was heavily affected by state investment plans which attempted (at leastmildly) to promote the laggard regions The Green Revolution which introduced





Tab

le1

Ind

ian

stat

esM

ain

char

acte

rist

ics

GS

DP

per

cap

ita

1997

ndash98

(rup

ees

per

mon

th)

Gro

wth

inG

SDP

per

capi

ta

1980

ndash98

()

Coa

stal

acce

ss(

ofp

opu

lati

onw

ithi

n10

0k

mof

coas

t)

Mai

ncl

imat

ezo

ne(p

erce

nta

geof

land

area

)

Maj

orpo

rtci

ty(p

opul

atio

ngr

eate

rth

an5

mil

lion

)Tr

opic

alA

rid

Subt

ropi

cal

And

hra

Prad

esh

252

12

942

5445

1Bi

har

126

11

00

50

95G

uja

rat

450

54

371

1782

0H

arya

na4

516

31

00

8713

Kar

nata

ka3

109

36

1940

600

Ker

ala

282

33

11

10

0M

adhy

aP

rade

sh2

286

25

035

956

Mah

aras

htra

569

04

534

6139

0M

umba

iO

riss

a1

871

17

5667

033

Punj

ab5

079

30

00

7624

Raj

asth

an2

621

38

07

8013

Tam

ilN

adu

345

44

365

964

0C

henn

aiU

ttar

Prad

esh

202

32

00

58

82W

estB

enga

l3

308

33

4367

033

Kol

kata








































































1980 000576 000131 000187 0008941981 000618 000162 000206 0009861982 000672 000185 000226 0010831983 000683 000162 000223 0010671984 000608 000084 000183 0008741985 000765 000096 000228 0010891986 000748 000060 000213 0010211987 000699 000125 000218 0010411988 000765 000158 000244 0011671989 000784 000184 000256 0012241990 000744 000214 000253 0012111991 000661 000194 000226 0010811992 000698 000320 000269 0012861993 000624 000307 000246 0011771994 000624 000374 000264 0012621995 000606 000427 000273 0013071996 000648 000488 000300 0014351997 000654 000523 000311 0014871998 000664 000559 000323 001546












Low




High







1990s High











N 5 14 R2 5 062




























N 5 14 R2 5 068






1971 1981 1991













Urbanization 013 013 019 025 030 023(530) (731) (530) (209) (475) (538)



Constant 039 005 3317 2329 1025 2233(009) (012) (337) (012) (109) (198)

R2 082 082 088 079 075 071






























1985 1998 Change
















India 22 38








41 Kerala





Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)


Highland(H)

India 33 32 28 0 7











State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References












































Tab

le1

Ind

ian

stat

esM

ain

char

acte

rist

ics

GS

DP

per

cap

ita

1997

ndash98

(rup

ees

per

mon

th)

Gro

wth

inG

SDP

per

capi

ta

1980

ndash98

()

Coa

stal

acce

ss(

ofp

opu

lati

onw

ithi

n10

0k

mof

coas

t)

Mai

ncl

imat

ezo

ne(p

erce

nta

geof

land

area

)

Maj

orpo

rtci

ty(p

opul

atio

ngr

eate

rth

an5

mil

lion

)Tr

opic

alA

rid

Subt

ropi

cal

And

hra

Prad

esh

252

12

942

5445

1Bi

har

126

11

00

50

95G

uja

rat

450

54

371

1782

0H

arya

na4

516

31

00

8713

Kar

nata

ka3

109

36

1940

600

Ker

ala

282

33

11

10

0M

adhy

aP

rade

sh2

286

25

035

956

Mah

aras

htra

569

04

534

6139

0M

umba

iO

riss

a1

871

17

5667

033

Punj

ab5

079

30

00

7624

Raj

asth

an2

621

38

07

8013

Tam

ilN

adu

345

44

365

964

0C

henn

aiU

ttar

Prad

esh

202

32

00

58

82W

estB

enga

l3

308

33

4367

033

Kol

kata








































































1980 000576 000131 000187 0008941981 000618 000162 000206 0009861982 000672 000185 000226 0010831983 000683 000162 000223 0010671984 000608 000084 000183 0008741985 000765 000096 000228 0010891986 000748 000060 000213 0010211987 000699 000125 000218 0010411988 000765 000158 000244 0011671989 000784 000184 000256 0012241990 000744 000214 000253 0012111991 000661 000194 000226 0010811992 000698 000320 000269 0012861993 000624 000307 000246 0011771994 000624 000374 000264 0012621995 000606 000427 000273 0013071996 000648 000488 000300 0014351997 000654 000523 000311 0014871998 000664 000559 000323 001546












Low




High







1990s High











N 5 14 R2 5 062




























N 5 14 R2 5 068






1971 1981 1991













Urbanization 013 013 019 025 030 023(530) (731) (530) (209) (475) (538)



Constant 039 005 3317 2329 1025 2233(009) (012) (337) (012) (109) (198)

R2 082 082 088 079 075 071






























1985 1998 Change
















India 22 38








41 Kerala





Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)


Highland(H)

India 33 32 28 0 7











State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References

















































































































1980 000576 000131 000187 0008941981 000618 000162 000206 0009861982 000672 000185 000226 0010831983 000683 000162 000223 0010671984 000608 000084 000183 0008741985 000765 000096 000228 0010891986 000748 000060 000213 0010211987 000699 000125 000218 0010411988 000765 000158 000244 0011671989 000784 000184 000256 0012241990 000744 000214 000253 0012111991 000661 000194 000226 0010811992 000698 000320 000269 0012861993 000624 000307 000246 0011771994 000624 000374 000264 0012621995 000606 000427 000273 0013071996 000648 000488 000300 0014351997 000654 000523 000311 0014871998 000664 000559 000323 001546












Low




High







1990s High











N 5 14 R2 5 062




























N 5 14 R2 5 068






1971 1981 1991













Urbanization 013 013 019 025 030 023(530) (731) (530) (209) (475) (538)



Constant 039 005 3317 2329 1025 2233(009) (012) (337) (012) (109) (198)

R2 082 082 088 079 075 071






























1985 1998 Change
















India 22 38








41 Kerala





Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)


Highland(H)

India 33 32 28 0 7











State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References




















































Low




High







1990s High











N 5 14 R2 5 062




























N 5 14 R2 5 068






1971 1981 1991













Urbanization 013 013 019 025 030 023(530) (731) (530) (209) (475) (538)



Constant 039 005 3317 2329 1025 2233(009) (012) (337) (012) (109) (198)

R2 082 082 088 079 075 071






























1985 1998 Change
















India 22 38








41 Kerala





Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)


Highland(H)

India 33 32 28 0 7











State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References



















































N 5 14 R2 5 062




























N 5 14 R2 5 068






1971 1981 1991













Urbanization 013 013 019 025 030 023(530) (731) (530) (209) (475) (538)



Constant 039 005 3317 2329 1025 2233(009) (012) (337) (012) (109) (198)

R2 082 082 088 079 075 071






























1985 1998 Change
















India 22 38








41 Kerala





Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)


Highland(H)

India 33 32 28 0 7











State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References



























































N 5 14 R2 5 068






1971 1981 1991













Urbanization 013 013 019 025 030 023(530) (731) (530) (209) (475) (538)



Constant 039 005 3317 2329 1025 2233(009) (012) (337) (012) (109) (198)

R2 082 082 088 079 075 071






























1985 1998 Change
















India 22 38








41 Kerala





Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)


Highland(H)

India 33 32 28 0 7











State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References


















































Urbanization 013 013 019 025 030 023(530) (731) (530) (209) (475) (538)



Constant 039 005 3317 2329 1025 2233(009) (012) (337) (012) (109) (198)

R2 082 082 088 079 075 071






























1985 1998 Change
















India 22 38








41 Kerala





Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)


Highland(H)

India 33 32 28 0 7











State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References




































































1985 1998 Change
















India 22 38








41 Kerala





Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)


Highland(H)

India 33 32 28 0 7











State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References




















































India 22 38








41 Kerala





Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)


Highland(H)

India 33 32 28 0 7











State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References














































41 Kerala





Tropical(Aw Am)

Subtropical(Cw)

Arid(Bs Bw)


Highland(H)

India 33 32 28 0 7











State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References














































State 1990 1998


Source EPWRF (2000)



State 1981 1993







42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References













































42 Rajasthan







43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References










































43 Madhya Pradesh



44 Orissa
















Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References




















































Appendix














Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References


















































Technical appendix


Var( ) ( ) ( ) ( )xN

x xN

x x x xi i i i

n

Nn

=-

- =-

- + -aelig

+

aringaring11

11

2 2 2

12

1

1

egrave

ccedilccedil

ouml

oslash

dividedividearing

1

N


=-


21

2

12

22

21

2 2N

x x x x x x x xi

n

in

N

n

N

( ) ( ) ( ) ( )1

1n

aringaelig

egraveccedilccedil

ouml

oslashdividedivide

This simpli es to

( )( )

( )( )

( ) (nN

xnN

xn N n n N

i i1

12

21 1

221

111

--

+--

+- +

Var Var-

--

nN N

x xi2

2

22

1)

( )( ) Var



References










































understanding regional economic growth in india...major driver of economic growth, and the coastal...

Documents