6. Production Performance of Indian Agriculture in the Era of Economic Reforms

See, for example, Singh (1994), Ahluwalia (1996), Gulati (1998), Gulati and Kelley (1999).

Recent research has shown that the elasticity of poverty with respect to agricultural GDP is greater compared to elasticity of poverty with respect to non-agricultural GDP. In other words, expansion of agricultural GDP reduces poverty more compared to expansion of non-agricultural GDP. See Cristiaensen et al. (2006, 2010a), Cristiaensen et al. (2010b), de Janvry and Sadoulet (2009), Ravallion (2010).

See Chand et al. (2007) and Chand and Parappurathu (2012).

This is not to say that this issue is unimportant. It is indeed important if our objective is to understand the fluctuations in agricultural performance in the era of reforms.

This is also the approach adopted by Bhalla and Singh (2001, 2012) in their studies on Indian agriculture.

We used the price data available from Singh (2007). To construct all-India price series for individual crops, Singh considered the data on value of production of various crops as available from National Accounts Statistics (published by the Central Statistical Organization) which have been divided by the physical quantities of respective crop-production levels as available from Area and Production of Principal Crops in India (published by the Ministry of Agriculture, Government of India). To overcome the yearly fluctuation, an average of 3-year prices (for 2001–2002, 2002–2003, and 2003–2004) has been considered.

On the basis of visual inspection of data series for all-crop production, we assumed a ‘break’ between the crop years 1994–1995 and 1995–1996.

It, however, needs mention that the ‘kinked exponential model’ suffers from the limitation that it considers the breakpoint in a given time series as exogenously determined. To overcome such a limitation, Quandt (1960) and Andrews (1993) developed a methodology (popularly known as the Quandt–Andrews methodology) to endogenously determine the breakpoint in the time series. This methodology has been further extended by Bai and Perron (1998, 2003) to examine presence of multiple breakpoints in the time series. In this study, we have not applied these methodologies to determine the break-year for the following reasons: First, the methodologies for endogenous determination of breakpoints are ideally suited for sufficiently long (historical) data series. Second, these methodologies when applied to different crop-production series do not produce identical result with regard to the presence of the breakyear. This causes difficulty in dividing our study period into two sub-periods (pre- and post-reform). However, our assumption regarding presence of a ‘break’ in 1995 is broadly in conformity with Chand and Parappurathu (2012) who applied Bai-Perron methodology to a 60-year long data series on value of agricultural GDP and found that 1995–1996 was one of the six breakyears for such a series. Third, in a study such as this, the year of break should not be construed as one that merely fulfills a statistical criterion; rather, it should also correspond to some definitive shift in economic policy regimes. Viewing from this later perspective, our choice of 1995 as the break-year seems justifiable as efforts toward globalization of Indian agriculture formally started from early 1994 the effect of which on agricultural production performance is likely to be reflected after a time-lag.

The annual compound growth rates of population in India during 1991–2001 and 2001–2011 were 1.95 and 1.65 %, respectively.

Our discussion here is too broad and synoptic, and needs to be expanded substantially. We also admit that the list of constraints provided by us is not exhaustive. Actually, the purpose of this section has been to initiate discussion on this vital issue. In any case, more detailed research should be undertaken to complete our understanding about the obstacles faced by the agriculture sector in different states/regions of India in the ongoing phase of globalization and liberalization.

Cropping intensity = (Gross cropped area/Net sown area) * 100.

For calculation of growth rate of an input (X), the following formula has been used: where ‘r’ is the rate of growth and ‘n’ is the time length between two periods (i and j). Using this formula, ‘r’ when multiplied by 100 provides growth rate in percentage terms.

Several scholars have attempted to explain Gujarat’s success on the agricultural front in the era of economic reforms. They observed public investment in infrastructure like irrigation, power, roads, watersheds, check dams, technology like BT cotton, and diversification in agriculture played crucial roles in raising agricultural growth rate in Gujarat since mid-1990s (see Gulati 2009; Shah et al. 2009).

The rural credit-deposit ratios of commercial banks for all-India were 57.98, 59.98, 39.04 and 56.26 in 1980–1981, 1990–1991, 2000–2001 and 2005–2006, respectively (Chatterjee 2006).

Chand (2009) compared India’s yield levels for some crops with the same for China. He found that average yield of paddy in India is less than half of the yield level in China; for wheat, it is about two-thirds of that in China.