Information about Asia and the Pacific Asia y el Pacífico
Journal Issue
Share
Article

India: Selected Issues

Author(s):
International Monetary Fund. Asia and Pacific Dept
Published Date:
February 2017
Share
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details

Economic Growth in the States of India1

India’s strong economic growth has been a striking feature of its economic development over the past several decades. At the state level, however, India’s economic growth appears unbalanced and income disparity has widened. This chapter analyzes developments in per capita incomes across Indian states, and tests whether cross-state income convergence exists. It finds some evidence of income convergence during the pre-1990s reform period, but disparities in Indian state income growth have widened in the post-1990 economic reform period.

1. Achieving fast rates of growth has been a striking feature of India’s economic development over the past several decades. The reform process in India traces back to the late 1970s, when the process of economic liberalization started. The subsequent reforms in the 1990s had significantly transformed the Indian economy into a more open and market-oriented economy, with a larger role for private sector participation. As these reforms began to bear fruit, Indian growth has accelerated and the poverty rate has declined. In particular, the average growth rate of real gross national income per capita during 2010–2015 was about nine times of that in 1970s. Similarly, the number of people in poverty as a share of the total population dropped from 45.3 percent in 1993 to 21.9 percent in 2011. Notwithstanding its high growth rate, India’s income per capita remains low when compared to other emerging economy peers.

2. At the state level, India’s economic growth appears unbalanced and income disparities have widened. While states’ growth rates of per capita income have risen, the dispersion of the growth rates across states has also increased. A number of initially-poor states have not been able to catch up with initially-rich ones, as the growth performance of those poor states was generally worse than average. Moreover, the ratio of the top-to-bottom net state domestic product per capita has increased to more than ten times in 2014/15 from about four times in 1980/81.

India: State Income Development

(Net State Domestic Product Per Capita in 1971-2014) 1/

Sources: CMIE States of India database and IMF staff calculations.

1/ See Table 3 for derivation of state acronyms.

India: Income Disparity in 2014/15 1/

Sources: CEIC, IMF staff calculations.

1/ Ratio of each states’ net domestic product per capita to median net state domestic product per capita. Data are at 2011/12 constant prices. Gray area reflects states where data are not available.

3. This study analyzes the developments in per capita incomes across Indian states and tests whether cross-state income convergence has occurred. There are two criteria for income convergence—where initially-poor states grow faster and catch up with initially-rich states (β-convergence) and where the dispersion of states’ per capita incomes narrows over time (σ-convergence). The analysis is based on traditional convergence tests and is complemented by panel unit root tests, and utilizes state-based measures of dispersion of per capita income for 32 Indian states and Union territories during 1961–2015 (see Annex 1).2

4. Some evidence of income convergence during the pre-1990s reform period is found. Neo-classical growth model regressions suggest there is little evidence of absolute convergence in per capita incomes across Indian states over the past five decades (Table 1). There is some evidence of absolute convergence during the 1961-1991 period, as found in Cashin and Sahay (1996) and Kalra and Sodsriwiboon (2010), but disparities in Indian state income growth widened in the post-1990 economic reform period.3 Conditional convergence across Indian states exists after controlling for various policy variables such as the economic sector composition, financial development and access, demographics and education (Table 2). Furthermore, the results of panel unit root tests confirm the above findings (Table 3).4 There is no evidence of convergence for the whole period (1961–2015) but is evidence of convergence in the sub-period of the pre-1990s (1961–1991).

5. State-based measures of income dispersion have also increased, especially since the 1990s. The analysis using state-based measures of income dispersion (Vuw, Vw, Mw, see Annex 1 for details) suggest cross-state dispersion of per capita incomes increased throughout the period studied (1961 onwards). State income dispersion widened sharply in the post-1990s, consistent with the above results based on neoclassical growth model regressions. In particular, income and growth in initially-poor states appeared to be more volatile than average and contributed greatly to the widening dispersion.

India: State Income Dispersion

(Unit, data include 18 states)

Sources: CSO, IMF staff calculations.

India: State Income Dispersion

(State contribution to Mw) 1/

Sources: CSO, IMF staff calculations.

1/ See Table 3 for derivation of state acronyms.

6. Differences in states’ policies and economic structure appear to contribute to the disparities in income and growth performance across Indian states. Conditional income convergence among Indian states, taking into account states’ policies and economic structure reflects wide gaps in individual states’ steady-state or long-run income potential (Table 2). In line with Purfield (2009), a larger share of higher value-added economic sectors, greater investment and access to credit, as well as higher spending on health and education, are found to have a positive impact on states’ per capita income growth.

7. Better functioning product and labor markets could help improve states’ income potential. Rigid product and labor market regulations appear to have a negative impact on growth. Based on OECD measures of the degree to which state policies are supportive of market competition, these Indian states that have a relatively less supportive regulatory framework for market competition would likely grow slower than those that have pro-competitive product market regulations. As an indication, the rank correlation between product market regulation and state growth is negative and high at -0.46. Likewise, more rigid labor market regulation could hinder job creation and therefore be detrimental to state growth. The rank correlation between a measure for the degree of labor market regulation burden and state growth is negative at -0.03.5

India: Product Market Regulation and State Growth

Sources: OECD, States of India Database, IMF staff calculations.

1/ Conway and Herd (2009) presents the indicator for product market regulation that measures the degree to which policies promote or inhibit competition in areas of the product market where competition is viable. They measure the economy-wide regulatory and market environments in 2 States of India. Data are as of 2009. Higher indicator indicates the less supportive regulatory framework for market competition.

India: Labor Market Regulation and State Growth

Sources: OECD, States of India Database, IMF staff calculations.

1/ Dougherty (2008) presents the OECD index of state-level labor reforms in India. The index identifies the areas in which Indian states have made specific discrete changes to the implementa and administration of labor laws. The index of labor reforms in the chart shows the reversed order Dougherty (2008) index. Higher score can be interpreted as the state is less advanced in implementing labor reform.

8. Continued sound macroeconomic policies and comprehensive structural reforms are needed to boost India’s long-term growth and lessen income equality. Greater investment, better credit allocation to the most productive sectors, as well as policies to facilitate the states’ transition into higher value-added services or industry could help create more jobs and boost growth. Spending on health and education by states is also important to help improve the quality of human capital. Wide-ranging product and labor market reforms will help boost competitiveness and productivity, thereby improving states’ long-term growth potential.

Table 1.Income Convergence for the States of India: Cross-Sectional Results
Growth rate of income per capita during
1961-19911961-20011961-20111961-2014
Log of initial income per capita−0.110.050.050.05
(0.08)(0.04)(0.02)***(0.02)***
Constant−1.01−0.47−0.46−0.49
(0.80)**(0.36)(0.18)***(0.19)**
Adjusted R-squared0.020.030.210.23
No. of observations30303030
States15151515
Source: IMF staff estimates.Note: Significance level is indicated by *** at 1% level, ** at 5% level, and * at 10% level respectively, with standard errors reported in parentheses.
Growth rate of income per capita during
1971-19911971-20011971-20111971-2014
Log of initial income per capita0.080.030.050.05
(0.12)(0.05)(0.02)*(0.03)*
Constant−0.76−0.26−0.40−0.42
(1.14)(0.50)(0.24)(0.26)
Adjusted R-squared−0.02−0.020.070.08
No. of observations36363636
States18181818
Source: IMF staff estimates.Note: Significance level is indicated by *** at 1% level, ** at 5% level, and * at 10% level respectively, with standard errors reported in parentheses.
Growth rate of income per capita over 20 years
1971-19911981-20001991-20102001-2014
Log of initial income per capita0.080.170.110.19
(0.12)(0.08)**(0.09)(0.12)
Constant−0.76−1.66−0.93−1.92
(1.14)(0.78)**(0.91)(1.23)
Adjusted R-squared−0.020.090.010.05
No. of observations36363636
States18181818
Source: IMF staff estimates.Note: Significance level is indicated by *** at 1% level, ** at 5% level, and * at 10% level respectively, with standard errors reported in parentheses.
Growth rate of income per capita
5-year window10-year window
Log of initial income per capita2.272.57
(0.32)***(0.38)***
Constant−18.92−21.46
(3.06)***(3.67)***
Adjusted R-squared0.220.32
No. of observations18397
States1818
Period1961-20151961-2015
Source: IMF staff estimates.Note: Significance level is indicated by *** at 1% level, ** at 5% level, and * at 10% level respectively, with standard errors reported in parentheses.
Source: IMF staff estimates.Note: Significance level is indicated by *** at 1% level, ** at 5% level, and * at 10% level respectively, with standard errors reported in parentheses.
Table 2.Income Convergence for the States of India: Panel Results 1/
Dependent variable: Growth of income per capita 2/
(1)(2)(3)(4)(5)(6)(7)
Initial income per capita−3.617−3.62−3.493−3.761−3.173−2.933−3.814
[4.16]***[4.00]***[2.35]**[2.47]**[2.10]**[2.14]**[1.80]*
Share of industry sector2.6484.16.2053.2672.0024.643
[0.77][1.07][1.46][0.73][0.43][0.76]
Share of services sector6.7517.8418.5937.8535.6497.259
[2.82]***[2.85]***[3.22]***[2.49]**[1.77]*[1.68]
Working-age population ratio, change0.8380.8040.5530.1880.546
[0.92][0.79][0.54][0.19][0.59]
Social spending0.0190.024
[0.29][0.34]
Credit growth0.067
[2.86]***
Net in-migration ratio0.127
[1.26]
Constant42.32238.11340.82748.37443.25942.70454.63
[4.36]***[3.61]***[2.59]**[2.96]***[2.44]**[2.56]**[2.27]**
Adjusted R-squared0.770.770.70.720.660.670.74
Fixed Effectyesyesyesyesyesyesyes
Time Effectyesyesyesyesyesyesyes
N15214610299757481
Source: IMF staff estimates.

Data are unbalanced panel of 32 Indian states and territories from 1961-2015, with data at annual frequency. Standard errors are robust and fixed effects are also included. Significance level is indicated by *** at 1% level, ** at 5% level, and * at 10% level respectively, with t-statistics reported in brackets.

Growth calculated as the differential of the logs of income per capita in the two periods divided by the time elapsed between the two periods multiplied by 100. This table presents the results for the sub-interval of ten-year periods, but analyses of different sub-intervals (for example, five years) do not alter the conclusions.

Source: IMF staff estimates.

Data are unbalanced panel of 32 Indian states and territories from 1961-2015, with data at annual frequency. Standard errors are robust and fixed effects are also included. Significance level is indicated by *** at 1% level, ** at 5% level, and * at 10% level respectively, with t-statistics reported in brackets.

Growth calculated as the differential of the logs of income per capita in the two periods divided by the time elapsed between the two periods multiplied by 100. This table presents the results for the sub-interval of ten-year periods, but analyses of different sub-intervals (for example, five years) do not alter the conclusions.

Table 3.Panel Unit Root Tests of Income Convergence for Indian States 1/
1961-20151961-19911991-2015
18 statesTest Statisticsp-valConvergenceTest Statisticsp-valConvergenceTest Statisticsp-valConvergence
Levin-Lin ADF-stat3.061.00No−0.840.20No1.420.92No
IPS ADF-stat (large sample adjustment values)1.630.95No−2.270.01Yes−0.140.45No
Bootstrapped IPS ADF-stat1.220.89No−2.520.01Yes−2.000.02Yes
Bootstrapped Fisher stat (MW method)31.350.69No52.670.04Yes46.090.12No
Source: IMF staff estimates.

The Table presents test statistics of panel unit root tests as described in Annex 1. Data are unbalanced panel of 32 Indian states and territories from 1961-2015. The results are presented for whole sample, sub-groups of Indian states, and sub-periods of pre- and post-1990s. Of which, 18 old Indian states and Union territories include Andhra Pradesh (AP), Assam (AS), Bihar (BH), Gujarat (GJ), Haryana (HR), Himachal Pradesh (HP), Jammu & Kashmir (JK), Karnataka (KA), Kerala (KL), Madhya Pradesh (MP), Maharashtra (MH), NCT of Delhi (DL), Odisha (OR), Punjab (PJ), Rajasthan (RJ), Tamil Nadu (TN), Uttar Pradesh (UP), and West Bengal (WB). In addition, 14 new Indian states and Union territories comprising Andaman & Nicobar (AN), Arunachal Pradesh (AR), Chandigarh (CH), Chhattisgarh (CT), Goa (GA), Jharkhand (JH), Manipur (MN), Meghalaya(MG), Mizoram (MZ), Nagaland (NG), Puducherry (PD), Sikkim (SK), Tripura (TP), and Uttarakhand (UT) are added to the analysis.

1961-20151994-2015
32 statesTest Statisticsp-valConvergenceTest Statisticsp-valConvergence
Levin-Lin ADF-stat3.161.00No6.401.00No
IPS ADF-stat (large sample adjustment values)1.050.85No3.351.00No
Bootstrapped IPS ADF-stat0.390.65No2.370.99No
Bootstrapped Fisher stat (MW method)62.260.54No57.480.70No
Source: IMF staff estimates.

The Table presents test statistics of panel unit root tests as described in Annex 1. Data are unbalanced panel of 32 Indian states and territories from 1961-2015. The results are presented for whole sample, sub-groups of Indian states, and sub-periods of pre- and post-1990s. Of which, 18 old Indian states and Union territories include Andhra Pradesh (AP), Assam (AS), Bihar (BH), Gujarat (GJ), Haryana (HR), Himachal Pradesh (HP), Jammu & Kashmir (JK), Karnataka (KA), Kerala (KL), Madhya Pradesh (MP), Maharashtra (MH), NCT of Delhi (DL), Odisha (OR), Punjab (PJ), Rajasthan (RJ), Tamil Nadu (TN), Uttar Pradesh (UP), and West Bengal (WB). In addition, 14 new Indian states and Union territories comprising Andaman & Nicobar (AN), Arunachal Pradesh (AR), Chandigarh (CH), Chhattisgarh (CT), Goa (GA), Jharkhand (JH), Manipur (MN), Meghalaya(MG), Mizoram (MZ), Nagaland (NG), Puducherry (PD), Sikkim (SK), Tripura (TP), and Uttarakhand (UT) are added to the analysis.

Source: IMF staff estimates.

The Table presents test statistics of panel unit root tests as described in Annex 1. Data are unbalanced panel of 32 Indian states and territories from 1961-2015. The results are presented for whole sample, sub-groups of Indian states, and sub-periods of pre- and post-1990s. Of which, 18 old Indian states and Union territories include Andhra Pradesh (AP), Assam (AS), Bihar (BH), Gujarat (GJ), Haryana (HR), Himachal Pradesh (HP), Jammu & Kashmir (JK), Karnataka (KA), Kerala (KL), Madhya Pradesh (MP), Maharashtra (MH), NCT of Delhi (DL), Odisha (OR), Punjab (PJ), Rajasthan (RJ), Tamil Nadu (TN), Uttar Pradesh (UP), and West Bengal (WB). In addition, 14 new Indian states and Union territories comprising Andaman & Nicobar (AN), Arunachal Pradesh (AR), Chandigarh (CH), Chhattisgarh (CT), Goa (GA), Jharkhand (JH), Manipur (MN), Meghalaya(MG), Mizoram (MZ), Nagaland (NG), Puducherry (PD), Sikkim (SK), Tripura (TP), and Uttarakhand (UT) are added to the analysis.

Annex 1. Tests of Convergence for the States of India

I. β-convergence

  • Neoclassical growth model as in Barro and Sala-i-Martin (1992), Cashin and Sahay (1996), Kumar and Subramanian (2012):
    Absolute convergence exists when all states converge to the same steady-state capital-labor ratio, output per capita and consumption per capita and have the same growth rate. On the other hand, conditional convergence allows states to differ in the their steady-state ratios, but as long as they have the same population growth rate then all their level variables, capital, output, consumption, etc., will eventually grow at that same rate.
  • Group-mean convergence based on panel unit root test, as in Pedroni and Yao (2006), Kalra and Sodsriwiboon (2010):
Panel Unit Root Tests
Levin-Lin-Chu testPooled within-dimensionH0: βi=0 for all iZLLCN(0,1)
H1: βi < 0 for all i
Im-Pesaran-Shin testGroup mean between-dimensionH0: β=0 for all iZIPSN (0,1)
H1: βi < 0 for some i
Maddala-Wu testAccumulated marginal significanceH0: βi=0 for all iPλ ⇒ χ22N
H1: βi < 0 for some i

The failure to reject the null hypothesis can be taken to imply that no subset of the members of the panel are converging toward one another.

II. σ-convergence

State-based measures of income dispersion follow Cashin and Strappazzon (1998):

  • Unweighted coefficient of variation, Vuw
  • Population-weighted coefficient of variation, Vw
  • Population-weighted absolute deviations of income relative to mean, Mw
References

    BarroR.J. and X.Sala-i-Martin1992ConvergenceJournal of Political Economy Vol. 100(2) pp. 223251.

    CashinP. and R.Sahay1996Internal Migration, Center-State Grants, and Economic Growth in the States of India,IMF Staff PapersInternational Monetary Fund Vol. 43 (March) pp. 12371.

    • Search Google Scholar
    • Export Citation

    CashinP. and L.Strappazzon1998Disparities in Australian Regional Incomes: Are They Widening or Narrowing?Australian Economic Review. Vol. 31(1) pp. 326.

    • Search Google Scholar
    • Export Citation

    ConwayP. and R.Herd2009How Competitive is Product Market Regulation in India? An International and Cross-state ComparisonOECD Economic Studies Vol. 45(1): pp. 149174.

    • Search Google Scholar
    • Export Citation

    DoughertyS.2008Labour Regulation and Employment Dynamics at the State Level in IndiaOECD Economics Department Working Papers No. 624OECD PublishingParis.

    • Search Google Scholar
    • Export Citation

    KalraS. and P.Sodsriwiboon2010Growth Convergence and Spillovers among Indian States: What Matters? What Does Not?IMF Working Paper WP/10/96 (International Monetary Fund: Washington).

    • Search Google Scholar
    • Export Citation

    KumarU. and A.Subramanian2012Growth in India’s States in the First Decade of the 21st Century: Four FactsEconomic & Political Weekly Vol. XLVII No. 3.

    • Search Google Scholar
    • Export Citation

    PedroniP. and J.Y.Yao2006Regional Income Divergence in ChinaJournal of Asian Economics Vol. 17 pp. 294315.

    PurfieldC.2009Mind the Gap—Is Economic Growth in India Leaving Some States Behind?IMF Working Paper WP/06/103 (International Monetary Fund: Washington).

    • Search Google Scholar
    • Export Citation

    SodsriwiboonP. and P.Cashin (2017) “Economic Growth in the States of Indiaforthcoming IMF Working Paper.

1

Prepared by Piyaporn Sodsriwiboon and Paul Cashin.

2

See also P. Sodsriwiboon and P. Cashin (2017).

3

The simple correlation between the log of per capita income in 1961 and the 1961-91 growth rate is -0.16, thereby reflecting β-convergence in state per capita incomes over this period.

4

See Table 3 for derivation of state acronyms.

5

The labor market regulation index is from the OECD, and indicates the reform progress on labor market regulation at the Indian state level and is the only set of indicators available. However, the index may not fully capture the core problems in the Indian labor market, particularly segmented labor markets and informality, and short samples from data limitation preclude a more formal analysis. Therefore, the negative relationship found between labor market regulation and growth is only for indicative purposes.

Other Resources Citing This Publication