Introduction

In this paper, which is a situation and trend assessment of wheat (Triticum spp.) production in India, we examine the regional differences and temporal changes that have occurred in the wheat sector between 2000 and 2018. Cultivated over an area of 214 million hectares, wheat is the largest crop with respect to acreage in the world (FAOSTAT 2021), and is of high strategic importance to ensure food security in developing countries (Shiferaw et al. 2013). The primary status of the crop as a major source of calories and protein for both urban and rural consumers of the Global South has remained unchallenged in the post-Green Revolution era (Shewry and Hey 2015; Lantican et al. 2016). People across the world consume wheat more often than any other cereal grain. Also, wheat consumption has been increasing in regions where it has not been considered a major food item traditionally – Africa, for example (as indicated by Shiferaw et al. 2013) – and the increasing disconnect between wheat consumption and wheat production has heightened import-dependence in some countries. For example, in the Low-Income Food-Deficit Countries (LIFDCs) – 32 countries including India, Kenya, and Uganda, where most of the global poor reside – wheat import is 134 times wheat export in quantity (Krishna et al. 2020). An increase in wheat prices in international markets could therefore negatively affect the diets of the poor in these countries (Cudjoe, Breisinger, and Diao 2010).

The second reason for the global relevance of wheat is that millions of marginal and resource-poor smallholders are dependent on wheat cultivation for subsistence consumption and income generation. Although the productivity of the crop has increased dramatically in the post-Green Revolution period, there exist significant regional heterogeneities, with the Global South lagging much behind the industrialised world (FAOSTAT 2021).

India is one of the largest wheat producers in the world. In 2018–19, wheat production in India registered an all-time high productivity of 3.4 tonnes/hectare, and the country’s share in global wheat production stood at 13.6 per cent (IIWBR 2019). The area under wheat cultivation was about 29.6 million hectares in 2018–19, and 102.2 million tonnes of grains were produced in that year. More than 90 per cent of the area under wheat was sown with bread wheat (T. aestivum), which is used to make chapati or roti. Durum or macaroni wheat (T. durum) accounts for eight per cent of the sown area. Shewry and Hey (2015) have shown that there was a remarkable increase in wheat availability in India between 1961 and 2011, from 12 per cent to 20 per cent of total kilocalorie (kcal) consumption.

Over the last few decades, wheat has become an integral part of the daily diet in regions of India where the crop has never been cultivated (e.g., Kerala). Wheat has become the second major energy source after rice, and the most important source of protein for Indian households from cereal consumption (Figure 1). In the foreseeable future, the per capita consumption of wheat is projected to remain unchanged at around 5 kg/month (Alae-Carew et al. 2019). These average values, however, conceal the prevailing inequality in wheat consumption, particularly in rural India, where the poor struggle to meet the required quantities of cereal consumption despite the targeted public distribution system (PDS) (Mukesh et al. 2019).

Despite scientific advances and barring the recent productivity hike, the trend in wheat production in India remained sluggish during the 1990s and the 2000s, especially in the less developed States (Bhushan 2005; Rada 2016; Tripathi and Mishra 2017). Some of the wheat-producing agro-ecologies face the threat of global warming (Newport et al. 2020). The terminal heat stress during grain-filling currently reduces wheat yield by 5 per cent, and this negative impact is expected to become even more serious in the coming decades (Ortiz et al. 2008; Lobell et al. 2012; Dubey et al. 2020). Another challenge is mounting biotic stresses. For instance, wheat blast disease caused by the fungus Magnaporthe oryzae pathotype Triticum, which affected more than three million hectares in Latin America in the first decade of this century, has now emerged and spread in Bangladesh, and poses a significant threat to Indian wheat (Islam et al. 2016; Mottaleb et al. 2018). Some researchers view the stagnating yield levels and decline in total factor productivity in the Northwestern IGP as unsurprising since the States have reached their attainable yield limits, making further improvement difficult (Ramadas et al. 2020). However, there is another line of thought that holds that by targeting less productive States and districts for technology and policy interventions, the wheat productivity of the IGP can be increased substantially (Jain et al. 2017; Jain et al. 2019). Our paper explores the rationale behind this argument by studying the spatial heterogeneity in wheat production and profitability in the IGP States.

Because of the high spatial heterogeneity in the type of cultivation systems and farmer characteristics, there is a need to strategically place potentially useful technologies in specific regions in order to maximise their effectiveness. Such targeting of technologies requires detailed information on the characteristics of wheat production systems across the country.

In this paper, we carry out a situation assessment of Indian wheat cultivation, and identify the sources of spatial heterogeneity in wheat productivity in the IGP, the breadbasket of India. For this, we analyse secondary data from different official sources, including the Agricultural Census (GoI 2019a), Statistical Year Books (GoI 2019b), and publications of the Directorate of Wheat Development (GoI 2020a) and other government agencies (GoI, various years).

Spatial Heterogeneity in Wheat Productivity, Yield Gap, and the Diffusion of Promising Technologies

The IGP of India, dominated by the rice–wheat cropping system, has played a significant role in boosting the food security of the country since the early years of the Green Revolution. The region contributes to 50 per cent of national food production from a mere 27 per cent of net sown area (Chauhan et al., 2012). Wheat is a winter (rabi) crop that spans an area of 18.1 million hectares in the IGP. For wheat, the IGP States are 52 per cent more productive than the rest of India, making the relative contribution of these States to national wheat production higher (67 per cent) than the area share (61 per cent) (Table 1). Wheat productivity in the IGP States, on average, was 3.4 tonnes/hectare in 2017, which is high as compared to the other major producers not only in the Global South (e.g., Pakistan with 3 tonnes/hectare, Turkey with 2.8 tonnes/hectares), but also the industrial world (e.g., the USA with 3.1 tonnes/hectare, Australia with 2.6 tonnes/hectare) (FAOSTAT 2021). However, except for the year 2019, when wheat production set a record in India, productivity has remained stagnant over time (Mukherjee and Huda 2018).

Despite high productivity, the sustainability of wheat production systems of the IGP has been questioned in the post-Green Revolution period. Concern has been raised about the negative environmental externalities associated with unsustainable intensification practices in the rice–wheat systems (Krishna et al. 2012; Pingali 2012), and vulnerability to climate change (Ortiz et al. 2008; Gupta, Somanathan, and Dey 2017; Mukherjee and Huda 2018) and biotic stresses (Mottaleb et al. 2018). Productivity growth has remained sluggish across the States post-Green Revolution, with the quinquennial growth rate in wheat yield remaining below two per cent for most of the period (Tripathi and Mishra 2017; Ramadas et al. 2020).

Agronomists and breeders have been concerned about the stagnation of wheat yield in India and have designed and disseminated various technological interventions. The most prominent of these interventions was the development of varieties with either high-yield potential or tolerance to diseases and pests (or both). The impact analysis of wheat breeding research has shown the important role played by international research institutions in developing countries, including India, which continue to produce high rates of return (Heisey, Lantican, and Dubin 2003; Lantican et al. 2016). The varietal technologies are popularised along with agronomic interventions under the umbrella term “sustainable intensification” to produce more output while using potentially fewer resources on existing agricultural land and reducing adverse environment or ecosystem impacts (Pretty and Bharucha 2014). Central to this concept is improved managerial practices that reduce production costs (Giller et al. 2015; Kotu et al. 2017). Zero tillage with residue retention for soil cover is one such technology that has been found to generate considerable agronomic and economic benefits while improving the environmental footprint of the production systems (Krishna and Veettil 2014; El-Shater et al. 2016; Keil, D’souza, and McDonald 2017).

As compared to productivity enhancement, less attention has been paid to reducing East–West disparity in wheat productivity through targeted innovations. We have shown in Table 1 that the high average productivity of wheat conceals a high regional disparity among the IGP States. The Northwestern States, such as Punjab and Haryana, have the highest land productivity. The largest producer, Uttar Pradesh (contributing one-third of national production), has 35 per cent lower wheat productivity than Punjab, while the States in the Eastern IGP (Bihar, Jharkhand, and West Bengal) have the lowest productivity in the region. As shown by Keil et al. (2020), the yield difference between the Western and Eastern IGP States has been prevailing from the Green Revolution period, and has only widened over time. However, the regional differences are not only in productivity growth rates but also in the variability of growth rates. The Eastern IGP States show, alongside lower productivity, higher fluctuations in growth rates in the post-Green Revolution period. On the other hand, in a highly productive State like Punjab, fluctuation in yield growth has been low (Figure 2).

During the years 2012–15, there was pronounced regional targeting of international R&D investment directed towards Eastern India, to enhance cereal production and reduce poverty in the Eastern Indo-Gangetic Plains (Keil et al. 2016). For example, the Cereal Systems Initiative of South Asia (CSISA) – a project funded by the Bill and Melinda Gates Foundation and the United States Agency for International Development (USAID), and led by the Consultative Group on International Agricultural Research (CGIAR) centres, especially the International Maize and Wheat Improvement Centre (CIMMYT) and the International Rice Research Institute (IRRI) – shifted the concentration of its R&D activities to Bihar and Odisha from Punjab and Haryana after 2012 (CSISA, various years). The attention of the Indian government also turned towards the East through spatially targeted programmes such as “Bringing the Green Revolution to Eastern India” (Subramanian 2015).

An important question that arises here is whether the inferior performance of the Eastern region is due to a comparatively low yield potential, or inadequate and incompatible technological change. Our analysis focuses on the district-level wheat yield data (Figure 3), which reveal a discernible gradient for wheat yields from the Western IGP to the Eastern IGP. Yield disparity has been in the range of 1.5–5.5 tonnes/hectare across the IGP districts. A smooth transition in yield across State boundaries depicts a relatively lower impact of State-specific interventions on wheat yield and the significance of agroclimatic factors. However, the Haryana–Uttar Pradesh and Jharkhand–West Bengal borders show discontinuous yield patterns. Similar patterns were also observed in remote sensing (Jain et al. 2017). These discontinuities imply the (in)effectiveness of agricultural policies of the State and associated institutions to help farmers achieve the potential yield. The results also show significant within-State variation, especially for Uttar Pradesh and Bihar, which could correspond to varying levels of technological progress, infrastructure, and market development within the State. However, the spatial variation could also arise from differences in the attainable yield due to differing agroclimatic congeniality. We estimated the yield gap (i.e., attainable yield minus actual yield) across districts.

The yield gap across IGP districts is depicted in Figure 4. Agroclimatically attainable yields (potential yields) are obtained using the Agro-Ecological Zones (AEZ) methodology, jointly developed by the International Institute for Applied Systems Analysis (IIASA) and the Food and Agriculture Organisation (FAO) of the United Nations.^1,2 Baseline results of climate and agroclimatic analysis are based on mean climatic data for the period 1961–90. Individual district-level mean values were obtained from respective potential yield images. Extreme yield gaps (>100 per cent) were capped to 100 per cent and negative yield gaps to 0 per cent for ease of depiction. The district-wise yield gap analysis shows a gradient corresponding to that of wheat productivity. The Western districts have attained full production potential, while the Eastern districts have scope for further yield improvement. Even within the leading wheat producer State, namely Punjab, there are districts where production can be improved to fully meet the agroclimatically attainable yield, although the State was considered to have saturated its production potential. On the other hand, several districts of West Bengal, despite being low productive areas, have near-zero yield gaps. More importantly, there is high potential to increase India’s wheat production by targeting R&D investments towards the districts of Jharkhand, Bihar, and Southern Uttar Pradesh, where the scope for yield improvement is in the range of 50–100 per cent. Traditional farming techniques and age-old wheat varieties may no longer be able to harness the wheat yield potential of these districts.

Increasing the yield potential, achieving yield stability, and reducing the per-unit cost of wheat production have been addressed through the development of an array of locally adapted farming technologies and managerial practices in India. Several micro-level (farm household-level) studies have documented the crucial role of these technologies in enhancing wheat productivity and profitability in India (Krishna and Veettil 2014; Meena et al. 2016; Ali et al. 2018; Meena et al. 2018; Bhardwaj et al. 2019; Kumar, V. et al. 2019; Gahlot et al. 2020; Keil et al. 2020; Ghosh et al. 2021). However, there is little information on the on-farm adoption of these technologies, which is due to limited R&D investment in adoption research (Krishna et al. 2020). At the same time, new technologies are being developed to address the problems faced by wheat farmers.

1. Varieties with greater yield potential and nutritional value. A good example is DBW 303, released in 2020. This wheat variety, developed by the ICAR–Indian Institute of Wheat and Barley Research, Karnal, is rich in protein (12.1 per cent, as compared to 8–10 per cent in popular varieties). The yield potential of this variety is also high, at 8.1 tonnes/hectare (Yadava et al. 2020). On the 75th anniversary of FAO in 2020, the Prime Minister of India dedicated to the nation 17 biofortified varieties of eight crops, and one of them was DBW 303.
2. Varieties tolerant to biotic stresses. Wheat is susceptible to several diseases, such as leaf and stem rusts and blast (Aktar-uz-Zaman et al. 2017; Mottaleb et al. 2018). Marasas, Smale, and Singh (2003) estimated that genetic resistance to controlling leaf rust in the developing countries – measured in terms of the yield losses avoided by research – was adding US$ 5.36 billion in 1990 prices, with a benefit–cost ratio of 27:1. Byerlee and Moya (1993) had previously estimated that maintenance breeding in wheat contributed to slightly more than one per cent of yields each year by making it possible to avoid yield losses. In irrigated areas alone, this represented an annual contribution of about US$ 150 million in 1990 prices. No recent estimates are available, and there are not many micro-studies. However, it has been shown that biotic stress tolerance has spurred varietal turnover in the face of epidemics replacing susceptible varieties (Jaleta et al. 2019). Researchers expect revolutionary changes in wheat breeding with the new CRISPR technology for targeted genome editing, and thereby a significant increase in disease tolerance.³
3. Combinations of managerial and seed-based innovations. One of the typical examples is the “October sowing of wheat” in India with the help of newly developed varieties in the Northwestern Plains Zone. Conventionally, wheat sown in November is considered to be on time, while wheat sown in December is considered late (Jat et al. 2018). Through early sowing, farmers could reduce the yield loss from terminal heat stress at the end of the crop season and attain higher water-use efficiency, particularly in the Eastern IGP. Based on coordinated multilocational trials, three wheat varieties (DBW 187, DBW 303, and WH 1270) were identified for release in 2020 for October sowing under irrigated conditions in Punjab and Haryana (personal communication with CIMMYT breeders).
4. New machinery for conservation tillage. Second-generation conservation tillage machinery called Turbo/Happy Seeder is spreading in the Northwestern IGP, which facilitates sowing under heavy stubble conditions and thereby avoids the need for residue-burning (NAAS 2017). A recent study (Keil et al. 2021) indicates that even in the absence of notable yield increase, the technology has high potential to reduce the cost of production and increase farm profitability.
5. More efficient management of weeds, especially for increasing the benefits of Conservation Agriculture practices. While Conservation Agriculture (CA) has several advantages, weed management emerges as a particularly problematic area that can benefit from such technological gains (Ghosh et al. 2021). The use of new pre-emergent and post-emergent chemical herbicides has been standardised for better weed management in Indian wheat cultivation. Development of herbicide resistance in wheat, especially against its most serious weed menace – Phalaris minor Retz. – leads to a total failure of the crop; hence, new herbicide chemicals are continuously developed, tested, and recommended for farmers (Raseed et al. 2020).

In addition to increasing and maintaining crop productivity, efforts are being made to enhance the nutritional value of wheat grain through both breeding and agronomic practices. Some studies have noted a decline in the nutritional quality of wheat grain during breeding for higher yield. In the last decade, significant efforts have been made to achieve biofortification and bio-availability of wheat grain for micronutrients, especially iron and zinc (Gupta et al. 2021). These varieties are expected to curtail hidden hunger to a large extent. The recent release of three biofortified wheat varieties for Indian farmers – Zinc Shakti (Chitra), WB-02, and HPBW-01 (Singh and Govindan 2017) – are examples of technological innovations in this direction. One of the recently released iron-enriched wheat varieties in India is HD 3171 (Gaikwad et al. 2020). The rate of diffusion and impact of these varieties among farmers and consumers are yet to be estimated.

Economic Implications of Yield Heterogeneity and Market Access

The stagnation in wheat productivity despite increased system intensification results in a reduction of total factor productivity and profitability, which poses a serious threat to food security in the region (Ladha et al. 2003). Against this backdrop, we analyse the economics of wheat cultivation in the IGP States to understand the trends and patterns in input use, cost of cultivation, and profitability. The average cost composition of wheat for 2004–10 and 2011–17 are shown in Table 2. The most noticeable differences – both spatially and temporarily – were concerning labour (human/animal/ machine) use. In the Western IGP, a significant proportion of wheat cultivation cost was incurred to hire and operate machinery (15 per cent in Punjab in both periods), while in the East, the highest share of cost went towards hiring human labour (24 per cent in Bihar in 2011–17). Use of machine labour was low in the East and use of human labour was low in the West. It is true that the concentration of machinery in agriculture in the Western IGP is well-documented, even from the pre-Green Revolution period (Bergmann 1963). The changes in cost composition present other patterns that are more relevant for R&D investment in wheat, as pointed out below.

1. While there was a general reduction in animal labour in wheat production, its share in the cost declined drastically in the Eastern States. For example, in Jharkhand, the cost share of animal labour reduced from 16 per cent to 3 per cent between the two periods.
2. Changes in the human labour share showed a mixed pattern. No big changes were visible in the Western States of Haryana and Punjab. In Bihar, Uttar Pradesh, and West Bengal, the cost share has increased over time, despite being already high as compared to the Western IGP in 2004–10. In Jharkhand, a slight reduction in cost share was observed (from 24 per cent to 22 per cent).
3. In Jharkhand, the reduction in cost share of human labour and animal labour was matched by an increase in the share of machine labour cost (from 9 per cent to 19 per cent). The share of machine labour has remained very low in West Bengal (7 per cent in 2011–17), as most of the cultivation activities are carried out with human labour (accounting for 36 per cent of the cost).

Unlike in the case of labour, there was no striking temporal or spatial difference in the cost share of material inputs such as seeds and chemical fertilizers. Due to heavy government subsidies, irrigation charges were low in Punjab in comparison to the rest of the IGP. Because a large share of the wheat is cultivated on leased-in land, the share of lease value in total cost was high in Punjab but negligible in other States. One may expect a relatively high cost of cultivation in the West to realise and maintain high land productivity. The costs and returns of wheat cultivation over 2004-05 to 2016-17 are shown in Figure 5. Indeed, the cost of cultivation of wheat, especially the fixed cost component, was high in Punjab and Haryana (Rs 60,000–70,000/hectare) as compared to Bihar and Jharkhand (Rs 30,000–40,000/hectare). However, the returns from wheat cultivation were also high in the West (Rs 90,000–100,000/hectare), making wheat cultivation a profitable enterprise for farmers. On the other hand, the difference between returns and total cost was small but positive in Uttar Pradesh and Bihar, and negative in Jharkhand and West Bengal. The negative returns occurred despite the negligible fixed cost component.

Farmers of Punjab and Haryana could make profits from wheat throughout the study period, partly because of better access to and availability of technologies, and the resulting higher grain yield. However, one should not underestimate the role of market access. To understand the role of output prices, we plotted the cost of production of wheat alongside the grain price for wheat from 2004-05 to 2016-17, in Figure 6. These graphs show that the cost of production of wheat was almost the same in Bihar and Haryana. However, the profitability of wheat cultivation was much higher in Haryana, due to better grain prices. Access to the government grain market (mandi) and availability of the minimum support price (MSP) were important determinants of profitability of wheat cultivation.

The cereal-centric policy of the Government of India, which started in the Green Revolution period, introduced MSP for wheat and other cereals with a government-backed procurement guarantee to safeguard the interests of farmers and ensure grain availability at low prices to consumers through the PDS. However, in several States of India, farmers are not aware of the MSP, the procurement system, or both. According to Aditya et al. (2017), only 33 per cent of wheat farmers in India are aware of the MSP, and this awareness is higher in the West (e.g., Punjab with 53 per cent awareness among farmers) than in the East (e.g., Bihar with 22 per cent awareness). Ideally, all cereal farmers should obtain an output price equivalent to the MSP if the open market price falls below it (Negi et al. 2018). For wheat, the system is highly efficient in Punjab and Haryana, where the farmers’ selling price is equal to the MSP (Figure 6). In the East, farmers often sell at a price below the MSP.

From Figures 5 and 6, one may deduce that the difference between returns and costs is reflected in the difference between selling price (SP) and cost of production (CoP), and that the condition that SP=MSP has great relevance for farm profitability. The share of farmers obtaining MSP varies across States. The government procures at MSP about 25–30 per cent of the total wheat production in the country, depending on the available buffer stock and demand from the PDS (Negi et al. 2020).

Wheat procurement by the government is highly skewed in favour of the major wheat-producing States of the Western IGP. For example, 67 per cent of the wheat annually produced in Punjab was procured in 2011–18 (Table 3). In the Eastern States, only a negligible share of the wheat produced is procured, and the local markets provide farmers with lower prices than the MSP. Coupled with the heightened risk of climate change, the lack of market incentives for wheat likely reduces crop acreage and production in the Eastern IGP, which can have a detrimental effect on the food security of the country. It might also place an additional burden on the natural resources of major wheat-producing States like Punjab and Haryana, where the unsustainable exploitation of natural resources is already affecting the vigour of the system (Shah et al. 2009).

How would low levels of wheat procurement affect farmer livelihoods in the Eastern IGP? A recent study by Negi et al. (2020) indicates that in the absence of government intervention in the cereal market in the form of procurement, the cropping pattern would shift from cereals to cash crops such as oilseeds, fibres, and vegetables. Depending on the market conditions of cash crops, the implications of this change for farm income could be positive or negative. One cannot be sure of how the rural food security situation would change as the subsistence consumption of cereals declines. Nonetheless, this scenario would present a great challenge to sustaining, let alone increasing, national wheat grain production.

Technological Interventions and Targeting

Across the IGP States, there exist significant differences in the cost of wheat cultivation per land unit. However, the per unit CoP (cost incurred to produce one kg of wheat grain, measured as Rs/kg) does not vary significantly between Punjab, Haryana, Uttar Pradesh, and Bihar. The CoP peaked in these States in 2016–17, at around Rs 11–12/kg, but lay below the supply price and the MSP (Figure 6). On the other hand, strong fluctuations were observed in the CoP of wheat in Jharkhand and West Bengal. Here the CoP often surpassed the supply price, resulting in financial losses to farmers. The CoP figures show structural differences in wheat production across the IGP. While farmers of Punjab and Haryana benefited from economies of scale, farmers of Bihar benefited from low wage rates (Table 4). Punjab farmers invested heavily in wheat to get higher yields, as indicated by their higher total cost of cultivation (Rs/hectare) in Figure 5, whereas Bihar farmers spent less on wheat and harvested less. From an ecological perspective, the Bihar scenario would generate fewer environmental externalities due to the low dependence on chemical fertilizers and herbicides, although the low input application was due to the credit constraints of farmers (i.e., the lack of access to formal credit), ill-developed input supply chains, and bottlenecks in agricultural extension. Nonetheless, the yield and gross margin generated per land unit were significantly high in Punjab, which is important for increasing national wheat production and farm income under irrigated conditions. In this section, we examine the differences in the pattern of technology changes across the IGP, taking the case of improved varieties (a divisible factor) and mechanisation (an indivisible factor) in wheat.

New, Improved Wheat Varieties

Since 1965, about 448 wheat varieties have been notified, of which 378 were bread wheat varieties (Gupta et al. 2018) and almost all were open-pollinated varieties (OPVs).⁴ Certified seed production and distribution in India, backed by the National Seed Policy (2002) and the Seed Bill (2004), is organised into three generations of seed multiplication to ensure quality and demand. Breeder seeds form the first generation, which are used for producing foundation seeds and certified seeds sequentially (Krishna et al. 2016). Indents for breeder seeds are submitted by the State governments to the Directorate of Agriculture and Cooperation (DAC), which is under the Ministry of Agriculture. The DAC, in turn, compiles the whole information crop-wise and sends it to the Indian Council for Agricultural Research (ICAR) and State Agricultural Universities (SAUs) for producing breeder seeds (Seednet India Portal 2020). State-wise demand for certified seeds of new varieties thus reflects in the indents placed by the respective State governments.

In wheat, the overall seed replacement ratio (SRR) increased from 34 per cent in 2012 to 50 per cent in 2019 in India (GoI 2020a).⁵ However, many farmers still cultivate age-old varieties, and the improvements in genetic gain and tolerance to adverse biophysical parameters are reaching the growers at a much slower pace (Krishna et al. 2016; Ray and Maredia 2016). The mean age of varieties in the breeder seed indents across the IGP States is plotted in Figure 7. Varietal turnover is found to be occurring at a slow pace in the East as compared to the Western States. Innovations in genetic improvement of wheat are reaching wheat farmers of the East at a relatively slow pace. The need for rapid dispersion of wheat varieties that can avert the terminal heat conditions is high in the East, and hence the low demand for new varieties in the region is an important concern for the R&D institutions. Unfortunately, only a few studies have addressed the determinants of farmer adoption of modern varieties of wheat in the developing countries (Krishna et al. 2020).

Mechanisation

The second technology category that we include to estimate the spatial heterogeneity in wheat productivity is the degree of mechanisation of crop production. As shown in Table 4, irrigation facilities, size of landholding, and wage rate of human labour are relatively high in the Western IGP, and all these factors prompt farmers to adopt mechanisation in farming. Through collaborative national and international research, new machinery has been developed and disseminated in the wheat production sector, viz. the Laser-guided Precision Land Leveller (Aryal et al. 2020), Turbo/Happy Seeder (Keil et al. 2021), Combine Harvester (Krishna et al. 2012; Kumar et al. 2017), etc. Several of these machines have both private and public value (e.g., increasing farm profits while reducing the incidence of residue burning and air pollution by adopting the Turbo/Happy Seeder for wheat sowing).

Recent studies have shown that Conservation Agriculture, which requires unique machinery, is more popular in the Western IGP than in the East (Bhargava et al. 2017). As we move from the West towards the East, we see that the share of machine labour in total operational cost declines (Figure 8). To a certain extent, the lower wage rates in these States might have been slowing down the replacement of human labour by machine labour. One may argue that the necessity for resource-conserving technologies have a higher demand in the West, where scarcity of natural resources is severe. However, this argument does not explain why Precision Land Levelling is more prevalent in Punjab, where irrigation water is affordable, than in Bihar and Jharkhand. Hence, an equally or more important reason is that farm endowments (e.g., land owned) and the status of general agrarian development (e.g., presence of machine manufacturers) are responsible for the extent of mechanisation in a State. Due to an array of socio-economic and institutional differences between the West and the East, a generalised policy on farm mechanisation might not be feasible in the IGP.

Conclusions and Policy Implications

The Sustainable Development Goals (SDGs) of the Agenda 2030 adopted by the United Nations General Assembly in September 2015 emphasise the importance of achieving food security and improved nutrition while minimising the ecological footprint of food production and distribution. This multidimensional nature of anticipated developmental outcomes puts the future of R&D investment on the rice–wheat systems of the IGP at a crossroads. One solution is to identify and act on the high spatial heterogeneities of production in the region. Rice–wheat systems of the Western IGP districts are characterised by high input use, land productivity, and returns as compared to the Eastern IGP districts. Stagnating yield levels result in an additional burden on natural resources and increased costs of production in the West, and therefore the R&D focus will be on a reduction in input use and cost, while sustaining yields. Resource-conserving technologies, such as the Turbo/Happy Seeder for zero tillage in wheat and precision land levelling, are popularised to reduce excessive dependency on external inputs and the natural resource-base (NAAS 2017; Aryal et al. 2020). Farmers of the Eastern IGP districts, on the other hand, have the disadvantage of low yields and low profits per hectare, although their cost of production is also low. The development and dissemination of locally adapted high-yielding varieties would be one of the R&D interventions suitable for this area. To sum up, we argue that R&D projects and policies in the Western IGP must adopt a paradigm shift towards more eco-friendly agriculture, while those in the Eastern districts require a strategy to enhance farmer access to input markets and information, for bridging the yield gap in wheat and for escaping the productivity plateau.

One of the key findings is the relevance of output price in wheat production. Sale price has emerged as the most important contributing factor for spatial heterogeneity in wheat cultivation across the IGP States. The mere opportunity to sell grain at the MSP would be enough to make wheat profitable for thousands of farmers of Bihar and Jharkhand. Furthermore, assurance of a reasonable output price would facilitate a more efficient production process through increased mechanisation and higher varietal turnover, and reduce the per unit cost of production. Higher and non-volatile output prices would act as an incentive for farmers to improve wheat productivity (Haile, Kalkuhl, and Braun 2016). As Das (2020) and Aditya et al. (2017) noted, farmers’ access to public procurement agencies is determined by various socio-economic factors and the extent of their economic marginalisation. The probability of access is low among smallholders, women farmers, and farmers belonging to socially oppressed castes. An improvement in market access, therefore, can have a long-lasting and pronounced impact on economic inequality and rural livelihoods in India.

Our study has also shown a high level of spatial heterogeneity of farm characteristics and- differential technology adoption across the IGP districts. To make technological interventions more effective in increasing wheat productivity in the Eastern IGP, these differences must be taken into consideration while designing R&D policy. For example, the small landholding size and capital constraints in the East necessitates a more widespread custom-hiring system for agricultural machinery (Keil, D’souza, and McDonald 2016). In States like West Bengal and Jharkhand, where most of the wheat area is rainfed, breeding for drought-resistant and stress-tolerant varieties will attract more R&D investment and policy focus. Our results indicate a gap in the varietal dissemination process in the East; one reason for this could be the unavailability of new varieties adapted to local agro-climatic conditions. Furthermore, going beyond spatial targeting, future research on Indian wheat may explore the causes and implications of social targeting in technology development to reduce inequalities.