Growth, decomposition and yield convergence of paddy production in India: Evidence from state-level panel data, 2000-01 to 2021-22

Introduction

Rice, grown in India as paddy, is the staple food for nearly half the country's population. It is also the largest user of net sown area among foodgrains. India produced an estimated 150 million tonnes of rice in 2024-25 from over 500 lakh hectares of land across 614 districts, overtaking China as the world's top producer. This achievement hides two key concerns. First, the balance between area expansion and yield improvement has changed since the Green Revolution era. In a land-limited economy, future growth relies on how quickly yields improve. Recent farm-level evidence shows that average rice yields across major producing states remained 1.7-2.4 tonnes per hectare below attainable levels, leaving considerable scope for productivity-led growth (H. S. Nayak et al., 2024) [14] . Second, the technology that boosts yields has not been shared fairly. Punjab, Haryana, and parts of Andhra Pradesh and Tamil Nadu saw high yields early on, while much of eastern and central India fell behind, despite having good agro-climatic conditions. Whether this gap has closed in the last two decades is important for food security, fairness between regions, and the direction of investment. A large body of research has looked at rice growth and its components in India. (Laitonjam et al., 2018) [11] found that yield effects accounted for over 92% of production growth between 1980 and 2014, a result supported by several state-specific studies like (Faldu et al., 2024) [7] . (Banerjee & Kumar Kuri, 2014) [2] and (Kumar et al., 2014) [10] reported mixed and fluctuating evidence of convergence, while (Chatterjee, 2017) [3] found beta-convergence, but not sigma-convergence, in per-hectare agricultural income from 1967 to 2010. At the crop level, (Mukhopadhyay, 2021) [12] found that convergence in cereal yields was stronger for wheat than for rice.

(Balaji & Gopinath, 2023) [1] revealed that farm incomes were converging across Indian districts, but at a slower pace in recent years, while (Sinha, 2021) [17] reported robust district-level convergence in rice yields particularly in Bihar, but wheat and maize showed limited and divergent trends. The present study aims to: (i) estimate compound annual growth rates of paddy area, production, and yield across Indian states for three periods: 2000-01 to 2011-12, 2012-13 to 2021-22, and the entire period; (ii) break down production growth into area, yield, and interaction effects; (iii) measure inter-state yield inequality using measures like CV, Gini, and Theil's T while tracking its trend; and (iv) test sigma-and beta-convergence and categorize states into catch-up, moderate, and stagnating groups.

Materials and Methods

Data and Study Period

The study uses secondary, state-wise annual data on paddy area ('000 Hectares), production ('000 Tonnes) and yield (Kg. /Hectare) for 2000-01 to 2021-22 from the Directorate of Economics and Statistics (DES), Government of India. The period was split into Period I (2000-01 to 2011-12) and Period II (2012-13 to 2021-22) and the full 22-year period.

Compound (Exponential) Growth Rate

Compound annual growth rates of area, production and yield were estimated for each state using the log-linear growth model (Dandekar, 1980) [4] :

where Yₜ is the area, production or yield in year t; a is the intercept; b is the regression coefficient; t is the time variable (year); and uₜ is the random disturbance term. The compound annual growth rate, expressed as a percentage, was computed as: CAGR (%) = (eᵇ -1) × 100 ………………………….…(2)

Decomposition Analysis

The contributions of area, yield and their interaction to production growth was examined using the additive decomposition model of Sharma (1977) [16] where the total change in production is decomposed as:

With A = area, Y = yield, P = production (P = A × Y), and subscripts 0 and n denoting the base and terminal years, respectively. Each component was expressed as a percentage of ΔP:

Component share (%) = (Component ⁄ ΔP) × 100 ………………………………………(4)

Yield Inequality Measures

Inter-state inequality in paddy yield in each year was measured using three broad indicators. Coefficient of variation:

Where, σ is the standard deviation and μ is the mean of state-wise yield in that year.

Gini coefficient (Gini, 1921) [8] , computed from the Lorenz curve of state-wise yield: G = 1 -Σₖ₌₁ⁿ (Xₖ -Xₖ₋₁)(Yₖ + Yₖ₋₁) …………………… (6) where Xₖ and Yₖ are the cumulative proportions of states and of yield respectively, ranked from lowest to highest yield, and n is the number of states. Theil's T index ((Theil, 167 C.E.)):

where yᵢ is the yield of the i-th state and ȳ is the mean yield across all n states.

Sigma-and Beta-Convergence

The Sigma-convergence tests were conducted to see whether the interstate inequalities decline over time, by regressing each inequality indicator on a linear time trend:

where Iₜ is the inequality indicator (CV, Gini, or Theil's T) in year t. A negative, statistically significant β indicates sigma-convergence. Beta-convergence (Barro and Sala-I-Martin, 1992) tests whether low-performing states cadtch-up with the high-yield states, using cross-sectional regression:

where gᵢ is the compound annual growth rate of yield for state i over the relevant period, and Yᵢ,₀ is state i's yield in the initial year. A negative, statistically significant β indicates beta-convergence.

State Clustering by Convergence Performance

Each state in all the three periods was classified based on the Beta-convergence results into: (i) Catch-up (Low initial yield, High growth), (ii) Moderate/Neutral, and (iii) Stagnating (High yield, Low growth). All the maps in the paper were produced using R software and Survey of India Shapefiles.

Results and Discussion

Growth in Area, Production and Yield of Paddy

The compound annual growth rates (CAGR) of area, production and yield for Period I (2000-01 to 2011-12, 'P-I'), Period II (2012-13 to 2021-22, 'P-II') and the overall period were reported in Table 1. The area growth remained stable at the all-India level (0.00% in P-I, 0.58% in P-II, 0.18% overall), while production grew at the rate from 1.82% to 2.39% per annum, and yield growth remained steady (1.82%, 1.80%, 1.84%). It is evident that the significant growth in the paddy production in India has been largely contributed by the yield improvements, rather than area expansion during the study period. The growth of area under paddy in Telangana was 10.08% per annum after becoming a separate state in 2014, largely due to new lift-irrigation schemes such as Kaleshwaram, which helped to increase the state's irrigated command area severalfold (Panjala et al., 2024) [15] . Rajasthan (3.48% overall, and 5.49% in Period II) and Gujarat (2.03%) also experienced irrigation-led area expansion story. On the other hand, Kerala (-4.48% in Period I, -2.76% overall), Goa (-2.40%), Puducherry (-2.22%) and the Andaman & Nicobar Islands (-3.60%) have been lost paddy land -a mix of urbanisation, rising labour costs, and farmers shifting to more profitable non-paddy uses (Jose & Padmanabhan, 2016). Andhra Pradesh's area actually grew in Period I (2.00%) but sharply declined after bifurcation (-3.09% overall). Telangana recorded the highest production growth (11.51%), largely due to irrigation-led area expansion (Panjala et al., 2024) [15] . Madhya Pradesh (8.17%) and Rajasthan (7.78%) also recorded significant production growths, largely due to yield enhancements (Faldu et al., 2024) [7] . The Andaman & Nicobar Islands (-4.78%) recorded negative growth, still recovering land lost to the 2004 tsunami, which left paddy soils saline and waterlogged in many low-lying areas (Raja et al., 2009). Madhya Pradesh's 6.48% yield growth -could be achieved largely due push into hybrid and short-duration varieties and expanding canal and tube-well irrigation (Faldu et al., 2024; A. K. Nayak et al., 2023) [7,13] . Rajasthan, Chhattisgarh, Bihar and Jharkhand all recorded more than 3% yield growth, a catch-up pattern similar to district-level convergence evidence for rice in many eastern states. Punjab (0.79%) and Haryana (1.06%) recorded yield growths well below the national 1.84% average, owing to decades of intensive paddy cultivation and falling water tables (Dangar & Mishra, 2024) [5] .

Decomposition Analysis of Production Growth

The growth in production of paddy was decomposed into area effect, yield effect and interaction effects and the results are presented in Table 2 and Table 3. At the all-India level, the yield effect accounts for 91.7% of production growth, area for only 8.3% in the entire time period, confirming that growth was largely due to productivity gains. In period I, area decreased at the all-India level, and the yield effect (107.5%) more than offset a negative area effect (-7.5%); and in period II, area was a large contributor (38.2%) alongside the yield effect (61.8%). Across most states, the yield effect clearly outweighs the area effect as the source of production growth over the full period. Madhya Pradesh (85.03%), Rajasthan (84.25%), Punjab (62.36%), Haryana (63.63%) and Assam (228.83%) all show yield as the dominant contributor. This confirms that productivity gains have led to growth in production in these states rather than area enhancements. Area effect was relatively modest in almost every state. Gujarat (29.98%), Haryana (36.37%) and Manipur (61.04%) experienced a larger area contribution than other states.

Inter-State Yield Inequality, and Sigma-Convergence

The inter-state yield inequalities were measured over the period of time and results are presented in Table 4. Average paddy yield across all the states increased by almost 41% from 1,843 kg/ha to 2,603 kg/ha during the study period. However, all three inequality measures declined during the entire period. CV decreased from 43.6% to 25.4%, the Gini coefficient from 0.25 to 0.14, and Theil's T from 0.09 to 0.03. This highlights a continuous fall in yield inequalities across states in the country. The above trend of decrease in yield inequalities across the states was statistically confirmed using the Sigma-convergence tests of the inequality measures (Table 5). The slope coefficients of all the inequality measures (CV: -0.0237, Gini: -0.0241, Theil's T: -0.0493) were negative and statistically significant at 1% level providing strong evidence of sigma-convergence in paddy yield across the states. The results are stronger than the mixed sigmaconvergence evidence reported for broader agricultural income measures (e.g., Chatterjee, 2017) [3] which reflects a more standardised, rapid spread of rice-specific technologies in the country. 5 where initially low-yield states (Chhattisgarh, Madhya Pradesh, Rajasthan, Gujarat) cluster in the high-growth region indicating a strong catch-up effect, while initial high-yield states (Punjab, Haryana, Andhra Pradesh, Tamil Nadu, Kerala) cluster in the low-growth region indicating a yield stagnation situation.

The above findings have revealed that paddy production growth in India over 2000-01 to 2021-22 was largely contributed by yield improvement rather than area expansion. The yield effect accounted for almost 92% of the production growth across the country. inter-state yield inequality declined steadily, with significant sigma-and beta-convergence. Madhya Pradesh, Chhattisgarh, Gujarat and Rajasthan were the strongest catch-up performers, while Punjab, Haryana, Uttar Pradesh, Andhra Pradesh, Kerala and Tamil Nadu stagnated in terms of yield growth. Accordingly, policy formulations should focus on yieldenhancing investments (improved varieties, soil health, precision inputs) over further area expansion as there are chances of an increase in yield gaps due to climate change effects (Debnath et al., 2021) [6] . Extended irrigation, seed replacement, and extension support should be improved for the catch-up and moderate-cluster states to sustain the growth. Yield stagnation and resource sustainability in Punjab, Haryana and Uttar Pradesh should be focused through diversification incentives and water-saving technologies. The state-level yield performance shall be judged by decomposing growth into its area and yield sources instead of only focusing on the production numbers.

References

1. Balaji SJ, Gopinath M. Spatial growth and convergence in Indian agriculture. Agricultural Economics. 2023;54(6):761-777. https://doi.org/10.1111/agec.12807
2. Banerjee A, Kumar Kuri P. Agricultural Growth and Regional Disparity in India: A Convergence Analysis. Sri Lankan Journal of Agricultural Economics. 2014;16(1):77. https://doi.org/10.4038/SJAE.V16I1.4605
3. Chatterjee T. Spatial Convergence and Growth in Indian Agriculture: 1967-2010. Journal of Quantitative Economics. 2017;15(1):121-149. https://doi.org/10.1007/s40953-016-0046-3
4. Dandekar V. Introduction to seminar on data base and methodology for the study of growth rates in agriculture. Indian Journal of Agricultural Economics. 1980;35:1-12.
5. Dangar S, Mishra V. Groundwater sustainability in India through nonrice-dominated cropping pattern. PNAS Nexus. 2024;3(9). https://doi.org/10.1093/pnasnexus/pgae378
6. Debnath S, Mishra A, Mailapalli DR, Raghuwanshi NS, Sridhar V. Assessment of rice yield gap under a changing climate in India. Journal of Water and Climate Change. 2021;12(4):1245-1267. https://doi.org/10.2166/wcc.2020.086
7. Faldu SM, Bhopala UD, NJ A. An analysis of growth and instability in area, production, yield and price behaviour of rice in India. International Journal of Statistics and Applied Mathematics. 2024;9(1):06-14. https://doi.org/10.22271/MATHS.2024.V9.I1A.1566
8. Gini C. Measurement of Inequality of Incomes. The Economic Journal. 1921;31(121):124. https://doi.org/10.2307/2223319
9. Jose M, Padmanabhan M. Dynamics of agricultural land use change in Kerala: a policy and socialecological perspective. International Journal of Agricultural Sustainability. 2016;14(3):307-324. https://doi.org/10.1080/14735903.2015.1107338
10. Kumar S, Lala KA, Chaudhary KR. Agricultural Growth and Economic Convergence in Indian Agriculture. Indian Journal of Agricultural Economics. 2014;69(2):1-18. https://doi.org/10.22004/AG.ECON.206377
11. Laitonjam N, Singh R, Yumnam A, Kalai K, Meena NK. RICE PRODUCTION IN INDIA: DECOMPOSITION AND TREND ANALYSIS. Plant Archives. 2018;18(1):435-438.
12. Mukhopadhyay D. Beta-convergence of crop yield across countries: A Modern Panel Data Analysis. Vidyasagar University Journal of Economics. 2021;27:1-18. https://doi.org/10.21203/rs.3.rs-530931/v1
13. Nayak AK, Tripathi R, Dhal B, Nayak AD, Vijayakumar S, Satpathy B, et al. Eco-efficiency and technical efficiency of different integrated farming systems in eastern India. International Journal of Agricultural Sustainability. 2023;21(1). https://doi.org/10.1080/14735903.2023.2270250
14. Nayak HS, McDonald AJ, Kumar V, Craufurd P, Dubey SK, Nayak AK, et al. Context-dependent agricultural intensification pathways to increase rice production in India. Nature Communications. 2024;15(1). https://doi.org/10.1038/S41467-024-524486
15. Panjala P, Gumma MK, Mesapam S. Geospatial assessment of cropping pattern shifts and their impact on water demand in the Kaleshwaram lift irrigation project command area, Telangana. Frontiers in Remote Sensing. 2024;5:1451594. https://doi.org/10.3389/FRSEN.2024.1451594/TEXT
16. Sharma KL. Measurement of the effects of area, yield and prices in the increase of value of crop output in India. Agricultural Situation in India. 1977;32:349-351.
17. Sinha R. Crop yield convergence across districts in India's poorest state. Journal of Productivity Analysis. 2021;57(1):41-59. https://doi.org/10.1007/S11123-02100618-9
18. Theil H. Economics and Information Theory. NorthHolland Publishing Company; 1967. https://www.scirp.org/reference/referencespapers?refere nceid=1620420