Bt cotton, adopted as a solution to curb losses caused by bollworms and to reduce the use of pesticides, successfully brought about a decline in pesticide consumption and expenditure, increase in productivity along with a higher output-input ratio in Punjab in 2004-05. Despite increased productivity and reduced pesticide use, Bt cotton is expected to increase the incidence of primary bollworms, which could develop resistance and secondary pests. But the undesirable, indirect and unanticipated consequences of innovation go together, as do the desirable, direct and anticipated consequences.
REVIEW OF AGRICULTUREEconomic & Political Weekly december 29, 200773Attributes and Socio-economic Dynamics of Adopting Bt Cotton Rajinder Peshin, A K Dhawan, Kamal Vatta, Kamaldeep SinghWorld over, cotton is grown on more than 32 million hectares with approximately 71 per cent of the produc-tion in developing countries. In India, cotton is an im-portant commercial crop being grown on 8.9 million hectares, constituting 27 per cent of the world area under cotton. Despite the largest share in area and third largest share in total world production, India ranks 57th with respect to cotton productivity [TFAI 2002]. The major reasons for low productivity of cotton are multiple pest problems and the crop being grown largely on un-irrigated soil (almost 65 per cent area under cotton is unirrigat-ed). Insect-pests are estimated to cause losses up to the extent of 50 per cent in cotton productivity [CICR 1998]. The losses, in fact, have increased from about 18 per cent in the early 1960s [Prad-han 1964] to 50 per cent in 1990s [Dhaliwalet al2004]. In spite of the significant use of pesticides in cotton, the total crop losses amounted to Rs 287.6 billion (ibid). Chandra (1998) and Wahab (1997) estimate the annual damage by bollworm to the tune of Rs 120billion. Development of hybrid cotton varieties in 1970s, which occupy 45 to 50 per cent of area under the crop,1 gave boost to its pro-ductivity in the past. As a result, cotton productivity2 in India increased from 1.06 q/ha during 1970-71 to 2.65 q/ha during 1996-97. In Punjab, it jumped from 4.63 q/ha during 1990-91 to 5.02 q/ha during 1994-95 and then followed a downward slide and reached an all time low of 1.79 q/ha in 1998-99, when farmers were unable to recover even the cost of cultivation. This was due to the outbreak of American bollworm (helicov-erpaarmigera), which is one of the major insect-pests of cotton and has developed resistance to pesticides due to selection pressure. From about two-three insecticide applications in cottonin1970s, farmers were reported to have applied more than 30 insecticide applications in cotton3 during 2003 and 2004 [Peshin 2005]. To overcome this downward slide in the productivity of the cotton crop, many integrated pest management (IPM) pro-grammes such as the regional programme on cotton IPM by Commonwealth Agricultural Bureau International (CABI) in 1993,FAO-European Union IPM programme for cotton in 2000, National Agricultural Technology Project (NATP) for IPM in 2000and Insecticide Resistance Management Programme in 2002 were initiated. The focus of these IPM programmes was to reduce and rationalise pesticide use by encouraging farmers to adopt other pest management practices like cultural and manual-mechanical practices in the absence of any effective bio-agents.Rajinder Peshin (rpeshin@rediffmail.com) is in the Division of Agricul-tural Extension Education, Sher-e-Kashmir University of Agricultural Sciences and Technology, Jammu; A K Dhawan is at the Department of Entomology, Punjab Agricultural University, Ludhiana; Kamal Vatta is at the Department of Economics, Punjab Agricultural University, Ludhiana; Kamaldeep Singh is at the Department of Entomology, Punjab Agri-cultural University, Ludhiana.Bt cotton, adopted as a solution to curb losses caused by bollworms and to reduce the use of pesticides, successfully brought about a decline in pesticide consumption and expenditure, increase in productivity along with a higher output-input ratio in Punjab in 2004-05. Despite increased productivity and reduced pesticide use, Bt cotton is expected to increase the incidence of primary bollworms, which could develop resistance and secondary pests. But the undesirable, indirect and unanticipated consequences of innovation go together, as do the desirable, direct and anticipated consequences.
REVIEW OF AGRICULTUREdecember 29, 2007 Economic & Political Weekly74developed during recent decades are tolerant against pests and are the main alternative to the use of synthetic organic pesticidesfor crop protection. The first transgenic crop variety was developed in squash crop in 1994, which was resistant to vi-ruses [Norriset al2002]. During the same year, glyphosate-re-sistant transgenic varieties of maize and soybean crops were released. In the year 1995, insect-resistant cotton, maize and potatoes using Btgenes were released in theUS (ibid). The transgenic crops are grown in 17 countries on a total area of 81 million hec-tares [Zehr2006]. In India, Bt cotton was approved for cultivation in 2002. The Bt cotton acreage increased manifolds since 2002 (Table 1).There are two schools of thought; one, advocating theadop-tion of Bt cotton for reducing the use of insecticides, cutting farm-ers’ production costs and increasing yield [Bannett et al 2004; Perlak et al 2001; Huang et al 2002] andthe second criticising biotechnology and doubting its usefulness for small farmers in the developing countries [Mishra 2006; GRAIN 2001; Wang 2006].In theUS, in 1997, the farmers grow-ing Bt cotton achieved greater produc-tivity by $24.43 per acre including insect control costs and the increase inreturns rose to $39.86 per acre [Boulter and Hilder 2002]. Huang et al (2002) while reporting the potential benefits and impact of Bt cotton in China highlighted that the actual use of pesticides in Bt cotton was much less ranging from 11.8 kg/ha in 1999 to 32.9 kg/ha in 2001 as compared to that in non-Bt cotton, which varied from 48.5 kg/ha to 87.5 kg/ha over the same period. On an average, cultivation of Bt cotton helped re-duce pesticide use by 35.7 kg/ha (55 per cent reduction). How-ever, they failed to explain the significant and unwarranted rise in pesticide use in Bt cotton from 1999 to 2001. ‘Bt Technology Adoption, Bounded Rationality and the Outbreak of Secondary Pest Infestation in China’ claims that afterseven years of Bt cotton introduction in China (1996 to 2004), the expenditure on pesticides for Bt and non-Bt was identical in 2004 at $101 per ha and the earnings from Bt cotton werelower [Mishra 2006]. Narayanamoorthy and Kalamkar (2006) reported the economical viability of Bt cotton for Indian farmers (Maharashtra). Contrary to expectations, the total quantity of pesticides used in Bt cotton varietyMECH 162 was higher than non-Bt cotton varieties. The average net profit from Bt cotton was Rs 31,880 per ha, about 80 per cent higher thanthat from non-Bt cotton. There was no significant difference in pesticide use between Bt and non-Bt cotton varieties. However, Table 1: Rate of Adoption of Bt Cotton Year Total Area Area under Extent of under Cotton Bt Cotton Adoption of (Million ha) (Million ha) Bt Cotton wrt % Area2001-02 8.73 – –2002-03 7.67 0.07 0.912003-04 7.63 0.23 3.012004-05 8.50 1.3015.292005-06 8.96 1.4115.742006-07 8.96 3.0033.48Sources: Government of Punjab (2006), GoI (2005), Zehr (2006).The pest scenario also underwent a change since the introduc-tion of synthetic pesticides. The pesticide, which was considered “silver bullet” to overcome all pest-related problems, has in fact compounded the problems. Rachel Carson (1962) wrote a book Silent Spring that brought about the change in the mindset of the people associated with agriculture. However, it was not until 1975 that much attention was paid to the dysfunctional consequences of pesticide adoption in crop cultivation. Scientists and extension workers led farmers into “pesticide treadmill”. According to avail-able estimates, out of the total pesticide consumption, 50 per cent and more is used for the cotton crop amounting to Rs 160 billion, of which Rs 110 billion are spent only to control bollworms [Alagh 1988; Mayeeet al2002]. The share of insecticide cost in the total cost of cultivation of cotton in Punjab increased many folds from 2.1 per cent during 1974-75 [Dhaliwal and Arora 2001] to 21.2 per cent during 1998-99 [Sen and Bhatia2004]. In pesticide hotspots like Bhatinda district in Punjab, it was 50 per cent of the total cost of cultivation [Shetty 2004]. This was despite the fact that many IPM programmes were implemented in Punjab from time to time. Introducing Bt cotton for cultivation was another step to re-duce the pesticide use in the crop. Bt technology was found to have significantly reduced application rates of toxic chemicals, while significantly increasing yields [Qaim and Janvry 2005]. Bt cotton is grown in eight states, viz, Madhya Pradesh, Gujarat, Maharashtra, Andhra Pradesh, Karna-taka, Tamil Nadu, Punjab and Harya-na. In Punjab, the government of India approved the cultivation of Bt cotton in March 2005. However, the Punjab farmers had started cultivating Bt cot-ton in 2002, purchasing seeds from far off states like Gujarat. The six varieties recommended for cultivation in 2005 were: Ankur 2534, Ankur 651,RCH 134, RCH 317,MRC 6301 andMRC 6304. Out of these permission for two varieties, viz, Ankur 2534 and Ankur 651 was withdrawn by the Punjab Agricultural University (PAU), Ludhiana in 2006 due to low productivity. The present study attempts to investigate the socio-economic dynamics, attributes and rate of adoption of Bt cotton in Punjab. The study is divided in six sections. The next section briefly outlines the controversy surrounding Bt cotton. Details of the database and methodology used in the study are given in the second section. In the third section, awareness, adoption and attributes of Bt technology have been highlighted. Input use and productivity pattern of Bt cotton have been discussed in the fourth section. The economic structure of Bt cotton is high-lighted in the fifth section. The last section presents the concluding remarks. 1 Controversy Surrounding Bt CottonSpraying pesticide formulations based on bacillus thurigiensis (Bt) have been in limited use for the control of larvae of lepidopteran pests for more than 40 years. The transgenics Table 2: Socio-Economic Profile of the RespondentsParticular DistrictOverall BathindaFerozepurMansa(n=210) (n=70) (n=70) (n=70) A Pattern of education (% of respondent farmers) Illiterate 10 9 16 11 Primary 7 17 11 12 Middle 20 23 26 23 Matric 37 40 36 38 10+2 10 7 11 9 Graduate and above 16 4 0 7B Farmers with telephone connection (%) 67 64 56 62C Average size of operational holding (ha) 8.17 14.10 6.96 9.75D Distribution of operational holdings (%) Small (below 2 ha) 11 3 6 7 Semi-medium (2-4 ha) 20 11 27 20 Medium (4-10 ha) 39 37 51 42 Large (above 10 ha) 30 49 16 31
REVIEW OF AGRICULTUREEconomic & Political Weekly december 29, 200775it is too early to generalise in India, where four million small and marginal farmers have taken up cultivation of Bt cotton withestimated adoption rate of 50 per cent by the end of 2007 [Mishra 2006]. Illegal and spurious seeds coupled with non-maintenance of minimum 20 per cent refugia by these farmers may result in severe pest attack on Bt cotton due to selection pressure and outbreak of secondary pests like whitefly [Chari 2006]. The bollworm is expected to develop resistance in 2007-08, where it was introduced in 2002 [Kranthi 2006]. Cotton monoculture and non-adoption of IPM principles is the recipe for such failure.2 MethodologyThe present study was conducted in the cotton growing areas of Punjab. The state accounts for 10 per cent produc-tion from 5 per cent area under cotton in the country and the highest average productivity since 2000-01. The cotton growing districts in Pun-jab are Bathinda, Ferozepur, Mansa, Mukatsar, Faridkot, Sangrur and Moga with more than 70 per cent of cotton growing area fall-ing in the first three districts.4 These three districts were purpo-sively selected for conducting formative evaluation of insecticide resistance management (IRM) based IPM programmes in cotton. TheIRM programme is being implemented by Central Institute of Cotton Research, Nagpur since 2002 in 28 districts, distributed over 10 states in India, which account for more than 80 per cent insecticide use in cotton. Thirty villages (10 from each district) were covered under the IRM programme in 2004-05. From these 10 villages in each dis-trict, five villages were randomly selected for data collection, making a total of 15 villages being covered under programme. In addition, out of nine control villages (not participating in the IRM programme), six villages (two from each district) were selected randomly. Thus, out of a total of 21 villages, the data were col-lected from 210 farmers (10 from each selected village) for the present study. Most of the adoption research has worked on the socio-economic attributes of farmers to predict the causes of adoption or non-adoption. Rogers’ diffusion model has dealt comprehensively by generalising socio-economic characteris-tics responsible for adopting new technology in agriculture [Rogers 1995]. These generalisations are that formal education, income, level of living, landholding, social mobility, empathy, etc, have positive correlation with adoption, whereas age has no relationship with adoption. The personality varia-bles and communication behaviour have also been generalised. Due to the huge already existing body of knowledge about these socio-economic and other variables, this study does not explore the relationship with adoption/rate of adoption of Bt cotton, which is conventional input technology like the green revolution technologies of high yielding varieties, fertilisers, pesticides, etc. The diffusion research has put much effort on studying personal characteristics, but relatively little effort has been devoted to ana-lyse innovation difference/technological attributes which affect the rate of adoption (ibid). To overcome the pro-innovation bias, alternative research approaches to after-the-fact-data collection about how Bt cotton diffused, had to be employed. To study the diffusion of technology overtime, data are to be collected during the diffusion process at different stages and not af-ter completion of the diffusion proc-ess (ibid). This base-line study of Bt cotton adoption will serve as the base for further studies of Bt cotton adop-tion in Punjab. A semi-structured questionnaire, in the local language, was distributed amongst selected farmers before the start of the cotton-growing season in 2004-05. This was done in order to avoid the problems/discrepancies as-sociated with recall/recollection of in-formation by the farmers after the growing season is over. The respondents were revisited after regular intervals (with a time gap of not more than three weeks) during the crop growing sea-son, to ensure proper and correct recording of the required infor-mation and to dispel doubts emanating when recording the data. The questionnaire focused on the socio-economic characteristics of the respondents, extent and level of adoption, input use, cost of cultivation, production and returns.3 Awareness, Adoption and Attributes of Bt CottonMajority of the farmers (around 89 per cent) was literate with the proportion of illiterate being 11 per cent. The modal level of edu-cation was matriculation. Sixty-two per cent of these farmers had telephone connections. Average size of the operational holding was the highest in Ferozepur (14.10 ha), followed by Bathinda (8.17 ha) and Mansa (6.96 ha) with an overall average landhold-ing of 9.75 ha (Table 2, p 74). The average holding size was much higher than the average size of 3.61 ha in the state largely due to the progressive farmers (active farmers or larger landholders) be-ing trained under IRM project and the sample consisting largely of these farmers. The fact is further strengthened from pattern of distribution of the sample holdings, majority of which (more than two-thirds) belonged to the medium and large size categories. 3.1 Adoption of Different Cotton CultivarsCotton is the main ‘khraif’ crop in the study area. About 64 per cent of the total area was under cotton, with its proportion being the highest in Bathinda (67.92 per cent) and that being the least in Mansa (58.30 per cent). About 60 hybrid cotton, 24 non-hybrid cotton, Bt cotton and ‘desi’ cotton varieties are being cultivated in Punjab[Peshin2005]. The extent of adoption of Bt cottonand other varieties was measured in terms of percentage of area cov-ered under different varieties and percentage of farmers adopt-ing Bt cotton and other varieties. The extent of adoption of Bt cotton revealed a varied picture in the three districts. The extent Table 3: Adoption of Varieties of Cotton DistrictOverall BathindaFerozepur Mansa (n=210) (n=70) (n=70) (n=70) A Average size of operational holding (ha) 8.17 14.10 6.96 9.75B Average area under cotton crop (ha) 5.63 9.22 4.10 6.32 (67.92)(64.47)(58.30)(63.97)C Extent of adoption (% farmers)1 Bt cotton 93 42 83 72 Other hybrids 71 268 93 63 Othervarieties2 45 88 38 57D Per cent of total cotton area under Bt cotton 42 10 21 22 Other hybrids 29 8 58 25 Other varieties1 29822153 Figures in the parenthesis are % of cultivated area (i e, B as % of A); (1) Represents multiple responses. (2) It includes non-hybrid varieties and second progeny (F2 generation) Bt cotton varieties.All decimals have been rounded up to nearest whole numbers.
1Awareness knowledge Rate of adoption 56 20 41 30 12 4 8 00 0
72 16 80 39 9 0 40 Rate of adoption Awareness knowledge
REVIEW OF AGRICULTUREEconomic & Political Weekly december 29, 200777Table 5: Awareness, Attributes and Expected Adoption of Bt Cotton Percentage of Farmers BathindaFerozepurMansaOverallA Sources of awareness/information (multiple response) Other farmers 77 73 78 76 Representative of a company 21 20 16 19Commission agent 2 6 6 5Newspapers 13 10 10 11Extension officer 2 0 0 1B Sources of seed (2004-05) Authentic 52 42 33 43 Notauthentic 44 58 60 52Cannotsay 4 0 7 5C Attributes of Bt cotton (multiple response) Resistance to bollworms 96 100 98 97 Higherproductivity 83 87 93 87 Saving on pesticide expenditure 94 57 93 84 Easy to adopt 92 37 75 72 Compatible 77 23 68 60Not compatible 6 0 8 5 Facilitates timely sowing of wheat 0 7 10 2 High cost of seed 8 10 15 11 More water requirements 46 17 18 29 Higher fertiliser requirements 8 0 0 3 Susceptible to cotton leaf curl virus 0 0 3 1 Susceptible to tobacco caterpillar 6 3 0 3Nobenefit 4 3 8 5D Estimated adoption of Bt cotton in 2005-06 Area under Bt cotton (ha) 67.02 28.30 51.70 46.26 Farmers growing Bt cotton (%) 100 78 80 86agents (5 per cent). Interpersonal communication channels were the main source of diffusion of this innovation. The awareness pattern formed ‘S’ shaped curve overtime when cumulative number/percentage of farmers was plotted on a graph, while the frequency distribution of the number of farmers getting awareness per year approached normality (Figure 2, p 76). The rate of adoption of Bt cotton also formed an ‘S’shapedcurve,when cumulative number/percentage of farm-ersadoptingBtcotton overtime was plotted and approached normalitywhenthefrequency distribution of farmers adopting it over time was plotted on a graph. These results are in agree-ment with Rogers’ diffusion theory revealing that knowledge precedes adoption at a fasterrate. Bt cotton being similar to green revolution technologies, the diffusion theory fits in here, which is not the case with inter-dependent complex technologies likeIPM. Against 4 per cent rate in 2002 (the seed of Bt cotton was purchased by farmers from Gujarat or through com-mission agent as the government of India allowed cultivation of Bt cotton in Pun-jab only in 2005), the adoption multiplied to 16 per cent in 2003-04 and then jumped to 80 per cent during 2005-06.3.3 TechnologicalAttributes The attributes of technology are impor-tant variables that determine the rate of adoption. Diffusion studies have mostly tended to regard all technolo-gies/innovations as equivalent [Rogers 1995]. Technological attributes are more important to speed up the rate of adoption. The generalised attributes, which affect adoption are: relative ad-vantage, compatibility, trialability, ob-servability and complexity. Out of these, first four attributes are positively related totherate/extent of adoption, while complexity is negatively related to it. The attributes of Bt cotton as reported by farmers have been presented in Table 5.First, there has been a clear relative advantage of Bt cotton cultivation over the other varieties as reported by the farmers. Almost 97 per cent of farmers admitted to its resistance to boll-worms,87 per cent to relatively higher yields, 84 per cent for their potential to save expenditure on pesticides. Even 72 per cent of the farmers reported that Bt cotton varieties were easy to adopt. Further, almost 60 per cent of the farmers reported compatibility of Bt cotton with their farming systems and soils, etc. However, there were some reports of its non-compatibility (although small), Bt cotton being water-intensive (29 per cent) and fertiliser-intensive (3 per cent). Further, 11 per cent of the farmers cited Bt seed to be very costly, and 3 to 5 per cent reported their susceptibility to cotton leaf curl virus (CLCV) and tobacco caterpillar. Bt cotton being susceptible to CLCV and tobacco caterpillar needs to be researched. These non-compatibility issues can hinder the adoption of Bt cotton. Recently, the new state government has slashed the seed prices by almost 40 per cent to give a further push to the adoption. Finally, resistance to bollworms and higher yields were the benefits (observability) and the respondent farmers reported no complexity during Bt cultivation. All these attributes of Bt cotton have pushed up the rate of its adoption even before it was given an official go-ahead for cultivation in Punjab.4 Pattern of Input Use and ProductivityIn this section, based on the field study carried out during 2004-05, a comparative analysis of Bt cotton with other varie-ties has been done with respect to the input use pattern. The impact of Bt cotton in reducing the insecticide use and its productivity advantage as a result of reduced insect pest losses has also been discussed. 4.1 InputUseThe variety-wise input use pattern for cotton is depicted in Table6 (p 78). The recommended seed rate for hybrid and non-hybrid cotton varieties is 3.75 kg/ha and 10 kg/ha, respectively. In case of non-hybrid varieties, the PAU recommen-dation has changed over time. The seed rate recommended was 20 kg/ha in 1979, 10 kg/ha in 2004 and 8.75 kg/ha in 2005 [PAU 1979, 2004, 2005]. However, the ac-tual seed rate has differed from the rec-ommended levels as farmers used the seed on the basis of their own experience and feasibility analysis. In hybrid varie-ties of cotton, 97 per cent farmers had applied 1.87 kg to 2.25 kg seed per ha, whereas in case of non-hybrid varieties 80 per cent of the farmers applied higher seed rate (more than 10 kg/ha). The average seed rate in Bt cottonwas2.07 kg/ha and that in non-Bt hybrid varieties was 2.25 kg/ha. In case of non-hybrid varieties the average seed rate used was 13.65 kg/ha. It is clear that the farmerswereusinglessthanrecommended seedrate for the hybrid varieties, which was largely due to higher cost of hybrid seeds, and the seed companies recommending seed rate of 2.25 kg/ha, which is lower than the PAU recommendation of3.75 kg/ha. Fertiliser use was the highest in the case of Bt cotton, fol-lowedbyhybridcotton and was the least in the non-hybrid cotton varieties. The nitrogenous fertiliser use in Bt cotton was higher by 23 and 31 per cent when compared to the other hybrid and non-hybrid varieties, respectively. The respective phosphatic fertiliser use was higher by 17 and 50 per cent and the potashic fertiliser use was higher by 104 and 413 per cent. The use of zinc-sulphate was also higher in Bt cotton by 25 and 10 per cent, respectively. Although there wasnodifferenceinthe
REVIEW OF AGRICULTUREdecember 29, 2007 Economic & Political Weekly78number of insecticide applications in Bt cotton when compared to other cotton cultivars. 4.3 Productivity of Bt CottonInsect pest losses caused by the bollworm complex especially helicoverpa armigera (American bollworm) in Bt cotton were significantly reduced thus, resulting in higher productivity. The overall productivity of Bt cotton6 was 24.17 q/ha as compared to the productivityof23.39 q/ha and 18.51q/haofotherhybrid and non-hybrid varieties, respectively (Table 8). The seedcotton yield of recommended Bt cotton varieties (RCH 134,RCH 317, MRC 6,301 and MRC 6,304) was 25.19, 26.54 and 25.26 in district Bathinda, Fero-zepur and Mansa, respectively. In this field study, we have included as well as recom-mended other varieties of Bt cotton, non-Bt cotton hybrids and non-hybrid varieties to provide a comprehensive overview of the cotton growing scenario in Punjab. The average yield of different recommended varieties of cotton recorded by the PAUbased on re-search and farmer field trails is given in Table 9 (p 79). It is evident that the yields recorded by thePAU in case of recom-mended Bt cotton varieties are higher than the yield obtained by the farmers in the study area except in Bathinda. The Bt cotton varieties cultivated by the farmers were those approved by the GEAC and recommended by the PAU in 2005 and other unde-script Bt varieties like Tulsi, Navbharat 151, etc. But in case of non-Bt hybrid varieties the farmers had recorded higher yield than the average yield recorded by the PAU, but more than 50percent of the cotton growing area was under the varieties not recommended by the PAU like Ankur 2,226, Ankur 2,534, RCH 134,RCH 317 etc. ThoughPAU recommends non-Bt Ankur 651 for cultivation in Punjab, it has withdrawn the recommendation for Ankur 651 Bt cotton due to its lower yields (17.50 q/ha) compared to other Bt cotton varieties (Table 9). Some of the non hybrid and desi cotton varieties (F 1861, F846) have comparable or higher average yield potential than Bt cotton varieties but the losses caused by insect pests (es-pecially bollworms) neu-tralise that higher yield potential. pesticide5 use between the other hybridandnon-hybrid varieties, its use declined significantly by almost 137 per cent in caseofBtcotton. There was no such difference in the number of irrigations applied. Hence, there wasaclearcomparativeadvan-tage of Bt cotton in terms of pesticideusewhencomparedtothe other hybrid and non-hybrid varieties, translating ultimately into relatively lower costs of cultivation and hence more returns. 4.2 ReducingInsecticideUse There has been a significant decline in pesticide consumption in the case of Bt cotton when compared to the other hybrid and non-hybrid varieties. However, the impact on the insecticide use needs to be discussed. We have attempted to highlight such an impact in terms of the number of insecticide applications, quan-tity consumed (technical grade material) and the costs incurred. Bt cultivation has resulted in a significant decline in the number of insecticide applications, its average consumption as well as the costs incurred. The average number of insecticide applications in Bt cotton was 4.7 as compared to 10.5 in case of non-Bt cultivars (Table 7). In all districts, the number of applications reduced by more than half in Bt cotton as compared to non-Bt varieties. The consumption of insecticides was 2.57 kg/ha and 6.44 kg/ha in Bt and non-Bt varieties, respectively. The average cost of insecticide use reduced significantly by more than 60 per cent when com-pared to non-Bt varieties. The results followed the same pattern in all the districts although they varied in their extent. The highest numbers of insecticide application were registered in Ferozepur both in Bt cotton (6.51) as well as other cultivars (13.08). The farmers in this district were also using more number of cocktails of insecticides. The maximum number of insecticide applications by tank mixing of insecticides separately was 30 in non-Bt cotton, while it was the minimum in Bt cotton (just two). The Abohar area in this district is the hot spot of insecticide use in Punjab. In Bt cotton, farmers had mainly applied insecticides for sucking pest such as jassid, whitefly and in certain pockets for tobacco caterpillar due to its outbreak. A few farmers also applied insecticide for bollworms, either due to its attack in later stages of crop cycle (or fear of such attack) or owing to some incidences of spurious seeds of Bt cotton. Despite all this, there was lesser Table 6: Input Use Input Bathinda Ferozepur Mansa Overall Bt Cotton Other Non- Bt Cotton Other Non- Bt Cotton Other Non- Bt Cotton Other Non- Hybrids hybrids Hybrids hybrids HybridshybridsHybridshybridsFertiliser use (kg/ha) Nitrogenous 253.83 231.12159.29 305.66 239.54209.33236.54194.85146.02 260.59211.36199.50 Phosphatic 113.19 110.61 82.34 179.02109.32 80.77 93.48 100.63 87.84122.30104.78 81.47 Potashic 13.19 3.83 6.88 17.89 17.37 1.81 17.55 7.54 10.34 15.23 7.45 2.97Zincsulphate 2.165.091.611.33 02.816.42 03.413.012.402.73Seed (kg/ha) 2.07 2.28 10.80 2.00 2.26 15.04 2.13 2.25 12.54 2.08 2.26 13.65Pesticide*(weedicide+insecticide) 2.824.634.63 3.20 8.25 8.252.96 5.15 5.152.826.686.68Average no of irrigations 7.47 7.47 7.47 5.60 5.60 5.60 5.71 5.71 5.71 6.26 6.26 6.26*Average technical grade material. Figures in parentheses represent % difference wrt Bt cotton.Table 7: Pattern of Insecticide Use and Cost of Insecticide ApplicationsInput Bathinda Ferozepur Mansa Overall Bt Non-Bt % Bt Non-Bt % Bt Non-Bt % Bt Non-Bt % Cotton DifferenceCotton DifferenceCottonDifferenceCottonDifferenceAverage no of insecticide applications 3.78 7.79 106.08* 6.51 13.08 100.92* 4.98 10.01 101.00* 4.76 10.46 119.74*Insecticide use@ (kg/ha) 2350 7470 217.87* 3978 7934 99.45* 3258 7228 121.85* 2934 7546 157.19*Cost of insecticides (Rs/ha) 2.45 4.25 73.70* 3.08 8.13 164.12* 2.63 4.82 83.27* 2.58 6.44 149.86*@Technical grade material. *means significant at 1 % level.Table 8: Productivity of Bt Cotton, Other Hybrids and Non-hybrid Varieties in Punjab Districts Overall BathindaFerozepurMansa A Productivity(q/ha) Btcotton 27.89 21.20 21.37 24.17 Other hybrids 24.89 20.17 23.15 23.39 Othervarieties1 19.2318.3218.1018.51B Difference in productivity of Bt cotton wrt Other hybrids 12.05* 5.11 -8.33 3.33 Othervarieties1 45.03***15.72**18.07**30.58***It includes non-hybrid varieties and second progeny (F2 generation) Bt cotton varieties.***, ** and * represent significance at 1, 5 and 10 % levels.The yield difference between Bt cotton and other hybrids is not significant. This can be due the fact that farmers had sown undiscript Bt cotton, mostly in district Mansa.
REVIEW OF AGRICULTUREEconomic & Political Weekly december 29, 200779There was no significant difference between the productivity of Bt cotton and other hybrid varieties in the state. It was only in Bathinda that a significant difference (12 per cent) was recorded. In district Mansa, the Bt cotton yield was on the lower side than non-Bt hybrids. This may be due to the fact that in this district the Ankur varieties 651,2226, 2,534 and 655 were the most widely adopted varie-ties and the majority of farmers had cultivated undescript and Ankur Bt cotton varieties which are lower yield-ing (17.50q/ha) than other Bt cotton varieties (Table 9). The recommenda-tion of Ankur Bt cotton varieties was withdrawn for the same reason by the PAU in 2006. However, the productiv-ity of Bt cotton differed significantly from that of non-hybrid cotton varie-ties in the entire study area, the over-all difference being more than 30 per cent. The difference was the largest in Bathinda (45 per cent), while the dif-ference (although significant) was just 15.7 per cent and 18.1 per cent in Ferozepur and Mansa, respective-ly. The productivity of different cul-tivars of cotton was higher in Bathinda district than that in Mansa and Ferozepur districts. 5 Economic Structure of Bt CultivationThe relative advantage of a technology in terms of economic re-turns plays an important role in the rate/extent of its adoption. The cost-returns structure of different cultivars of cotton has been discussed below to highlight the economic benefits of Bt-cultivation in the state. The expenditure on cotton production was divided into six cat-egories, namely land preparation, sowing cotton seed, inputs, cost of spraying, hoeing and other expenditure (Table 10). The overall cost of cultivation (per ha) of Bt cotton, other hybrids and non-hybrids was Rs 22,431, Rs 24,259 and Rs 21,713, respectively. The cost of cultivation of Bt cotton was lower by 7.54 per cent than that of other hybrid varieties but was higher by 3.31 per cent than that of non-hybrid varieties. The major reason for lower costs in the former case is the significant reduction in the insecticide cost (including the cost of sprays), which even neutralised the higher cost of Bt seed. The higher cost of cultivation than non-hybrid varieties was due to the reason that the difference in the cost of seed was much higher, which could not be neutralised by reduced insecticide cost and the labour charges for the sprays. The average per ha cost of seed was Rs 4,621 for Bt cotton, Rs 1,914 for other hybrids and Rs 484 for non-hybrids. The cost of insecticides was Rs 3,005 and Rs 7,546 per ha and the labour cost of sprays was Rs 476 and Rs 1,046 per ha for Bt and non-Bt varieties, respectively. The cost of insecticides amounted to 17, 37 and 41 per cent of the total cost of cultivation of Bt cotton, other hybrids andnon-hybrids,re-spectively. It is thus clear that adoption of Bt cottoninthestate has helped in reducing the cost of cultivation of the crop. Table 10: Cost-returns Structure for Different Cultivars of Cotton in Punjab Districts Overall Bathinda Ferozepur Mansa Cost of Cultivation (Rs/ha) ALandpreparation Pre sowing irrigation1 472 375 496 448 Ploughing and planking2 2,3362,1722,0762,195 Sowing 2 415 415 415415Inputs B Seed3 Bt cotton 4,607 4,445 4,733 4,621 Hybridcotton 1,960 2,030 1,846 1,914 Other varieties (includes non-hybrid varieties and Bt cotton F2 generation) 512 458 512 484C Chemicalfertilisers4 Bt cotton 2,351 3,212 2,192 2,494 Hybridcotton 2,235 2,224 1,887 2,051 Other varieties 1,578 1,800 1,620 1,764D FYM5 150 64161 118E Insecticide6 Bt cotton 2,350 3,978 3,258 3,005 Hybrid cotton and other varieties 7,470 7,934 7,228 7,546F Weedicide7 306 134 261 234G Irrigation1 3,574 2,142 2,577 2,765H Cost of spraying (Rs ha–1) H1Insecticides Bt cotton 378 651 498 476 Hybrid cotton and other varieties 779 1,308 1,001 1,046H2 Weedicide 80 39 88 69 Hoeing (Rs ha–1) Manual 746 560 713 673 Tractorised 350 247 256 284J Other expenditure (Rs ha–1)3 J1Picking Bt cotton 4,741 3,604 3,633 4,109 Hybridcotton 4,231 3,429 3,936 3,976 Other varieties 3,269 3,114 3,077 3,147J2 Cleaningbunds 175 175 175 175J3 Cleaning of field after picking 350 350 350 350Total cost of cultivation (A to J) Bt cotton 23,434 22,519 22,009 22,431 Hybrid cotton 25,479(+8.03) 23,598 (+4.57) 23,466 (+6.07) 24,259 (+7.54) Other varieties 22,562 (-3.86) 21,287(-5.79) 21,006 (-4.77) 21,713 (-3.31)Output (q/ha) and returns (Rs/ha)A Average seed cotton yield Btcotton 27.89 21.20 21.37 24.17 Hybrid cotton 24.89 20.17 23.15 23.39 Other varieties 19.23 18.32 18.10 18.51B Gross returns 8 Btcotton 50,760 38,584 38,893 43,989 Hybrid cotton 45,300 (-12.05) 36,709 (-5.11) 42,133 (+7.69) 42,570 (-3.33) Other varieties 34,999(-45.03) 33,342 (-15.72) 32,942 (-18.07) 33,688 (-30.56)C Net income Bt cotton 27,326 16,065 16,884 21,558 Hybrid cotton 19,821(-37.86) 13,111 (-22.53) 18,667(+9.55) 18,311 (-17.90) Other varieties 12,437 (-119.77) 12,055 (-33.26) 11,963 (-41.14) 11,975 (-80.28)D Returns-costratio Btcotton 2.17 1.71 1.77 1.96 Hybridcotton 1.78 1.56 1.80 1.75 Non Hybrid cotton 1.55 1.57 1.57 1.55E Pesticide (including application) cost as % to total cost of cultivation Btcotton 13.29 21.32 18.65 16.87 Hybrid cotton 33.89 +20.6 39.90 +18.58 36.56 +17.91 36.67 +19.80 Non-hybrid cotton 38.27 +24.98 44.23 +22.91 40.84 +22.19 40.97 +24.10The figures have been given separately for Bt cotton, hybrid cotton and non-hybrid cotton for all those operations where the difference was estimated to exist. For rest of the operations the cost estimates are the same. The yield levels and returns have also been given separately. Figures in parenthesis are % difference over Bt cotton. The cost of cultivation includes the labour costs – family and hired labour (Rs 80 per man day). The cost of spraying pesticides Rs 100 per ha for both tractorised and manual spraying. The expenditure on account of hoeing was computed on uniform market rates. The average number of hoeings, tractorised and manual were multiplied with Rs 250 ha–1 and Rs 625 ha–1, respectively. The other expenditures included cost of picking (R s 170 q–1), cleaning of bunds (Rs 175 ha-1) and cleaning of fields after picking (Rs 350 ha–1). All these were calculated at uniform market price.(1) Irrigation as per expenditure given by a farmer for canal and tube well irrigation. (2) At uniform market rates. (3) Price of purchased seed or own seed at market price (Bt cotton: Rs 4,621, hybrid cotton: Rs 875-2,250, non-hybrid: Rs 312-500/ ha). (4) Fertilisers at market price plus cost of application. (5) Farm Yard Manure (FYM) @ Rs 150/trolley (40 tonnes). (6-7) Insecticides and weedicides at market price. (8) Returns are computed by using average actual market price (Rs 1,820/q).Table 9: Average Potential Yield of Different Varieties in PunjabVariety AverageYield (q/ha)A Bt cotton varieties RCH 134 Bt 28.75 RCH 317 Bt 26.25 MRC 6301 Bt 25.00 MRC 6304 Bt 25.25B Non-Bt hybrid varieties Ankur 651 17.50 Whitegold 16.25 LHH144 19.00C Non-hybrid cotton varieties F 1861 16.25 F 1378 25.00 F 846 27.25 LH 1556 21.25D Desi cotton varieties PAU 626 H 24.50 Moti 21.25 LD694 17.50 LD327 28.75Average lint yield for the state as a whole for cotton is 731 kg/ha.Source: PAU (2007).
REVIEW OF AGRICULTUREdecember 29, 2007 Economic & Political Weekly80As a result of such adoption, not only the proportion of insecti-cide use in the total cost has reduced but the net returns have also followed an increasing pattern. Increased attack of American boll-worm on the crop during the late 1990s seriously endangered the economic viability of this crop in the state resulting in fast depletion of this crop from the cultivation map. However, attractive econom-ics of cultivation of Bt cotton in the form of reduced cost of cultiva-tion and higher returns due to better yields have facilitated itsfast adoption by the state farmers ensuring its significant revival in the state. Owing to the difference in yields of these cultivars, the costof picking also varied accordingly and was Rs 4,109, Rs 3,976 and Rs3,147 per ha for Bt cotton, other hybrids and non-hybrids, re-spectively. On the productivity side, the average yield of seed-cotton was 24.17 q/ha, 23.39 q/ha and 18.51 q/ha for the respective culti-vars. As a result, the gross returns (per ha) were Rs 43,989 for Bt cotton, Rs 42,570 for other hybrids and Rs 33,688 for non-hybrids. Hence, at present, Bt cotton cultivation seems economically vi-able in Punjab. The economic viability of Bt cotton is reflected by the fact that the ratio between gross returns and cost of cultiva-tion was 1.96 compared to 1.75 in other hybrids and 1.55 in non-hybrids. The ratio is even better in Bathinda at 2.17, which might be due to the benefits of an earlier and hence, greater extent of adoption in the district. It is a positive sign and implies that the benefits can improve once the adoption picks up, spurious Bt cot-ton seed is eliminated from the market and the cultivation prac-tices based on IPM principles are standardised in the state. 6 ConclusionsThe base line study of adoption of Bt cotton in Punjab has rein-forced the diffusion theory that input intensive technologies follow the normal distribution curve of awareness-knowledge andrate of adoption with the latter following the former. Bt cotton cultivation has resulted in higher external input use, namely, fertilisers and irrigation water but reduced the insecti-cide use. The cost of Bt cotton seed and higher fertilisation cost has been offset by lower expenditure on account of insecti-cides. The higher gross returns due to higher productivity have resultedin improving the economic viability of Bt cotton cultiva-tion.This being a base line study, we need to follow the Bt cot-ton cultivation in Punjab over a period of time to study its functional (desirable) and dysfunctional (undesirable) con-sequences. The “silver bullet” technology for control of boll-worms needs to be studied without the innovation bias otherwise we may again prove wrong as we have been in the case of “pesti-cide” introduction in agriculture. Pesticides were termed as so-lution of all pest problems but it led the farmers onto a “pesticide treadmill”. The integrated pest management principles need to be applied for developing IPM to avoid development of resistance to bollworms against Bt cotton. Despite increased productivity and reduced pesticide use inUS, China and India, the monocul-ture Bt cotton is expected to increase the incidence of primary pest bollworms, which could develop resistance and develop secondary pests like caterpillars. The dying of Bt technology in China in seven years (1997-2003) has also been reported. The scientists and change agents/agencies over-emphasising adop-tion per se, assuming that the consequences of adoption of a technology will be positive and usual survey research methods being inappropriate to investigate consequences [Rogers 1995], may lead the farmers into transgenic-cum-pesticide treadmill. Hence, there is need to study short term, long term, desirable and undesirable consequences of Bt cotton cultivation over a period of time, to evaluate its sustainability.Notes 1 The hybrid varieties occupied 45 to 50 per cent of the total area under cotton before the introduction of Bt cotton. 2 It refers to the total quantity of lint produced per hec-tare. Lint is produced at the rate of one-third of seed cotton yield. 3 The cotton here refers to non-Bt varieties. 4 The total area under cotton in Punjab was 589 thou-sand ha during the period of study (2004-05), out of which the area in Bathinda, Ferozpur and Mansa was 141,133 and 82 thousand hectares, respectively. 5 Pesticide included both weedicides and insecticides. 6 Productivity is in terms of seed cotton. ReferencesAlagh, Y K (1988): ‘Pesticides in India Agriculture’, Eco-nomic & Political Weekly, 20: 1959-64.Bannett, R M, Y Ismeal, U Kambhampati and S Morse (2004): ‘Economic Impact of Genetically Modified Cotton in India’,AgBio Forum: The Journal of Agro-biotechnology Management and Economics,7:96-100.Boulter, D and V A Hilder (2002): ‘Transgenic Crops (Annual)’ in Pimental D (ed), Encyclopedia of Pest Management, Marcel Dekker, Inc, New York, pp 846-49.Carson, R (1962): Silent Spring, Fawcett Crest, New York.Chandra, M (1998): ‘Agriculture: Debt Drives Indian Farmers to Suicide’, http://www.oneworld.net.Chari, M S (2006): ‘Debate: Heavy Cotton’,Down to Earth: 15(6): 23-26.CICR (1998):Annual Report, Central Institute of Cotton Research, Nagpur, India.Dhaliwal, G S, R Arora and A K Dhawan (2004): ‘Crop Losses Due to Insect Pests in Indian Agriculture: An Update’,Indian Journal of Ecology, 31: 1-7.Dhaliwal, G S and R Arora (2001): Integrated Pest Management: Concepts and Approaches, Kalyani Pub-lishers, Ludhiana.Dhawan, A K, A S Sohi, M Sharma, T Kumar and K S Brar (2004):Insecticide Resistance Management in Cotton, Punjab Agricultural University, Ludhiana.GoI (2005): Biosafety Issue Related to Transgenic Crops: With Special Focus on Bt Cotton, Biotech Consortium India and Ministry of Environment and Forests, Gov-ernment of India.Government of Punjab (2006):Agriculture at a Glance: 2005-2006, Information Service, Department of Agriculture, Punjab, Chandigarh.GRAIN (2001): ‘Bt Cotton through the Back Door’, Genetic Resource Action International, Seedling 18(4). http://www.grain.orgHuang, J, R Hu, C Fan, C E Pray and S Rozelle (2002): ‘Bt Cotton Benefits, Costs and Impact in China’, Ag Bio Forum, The Journal of Agrobiotechnology Manage-ment and Economics, 5 :153-66.Kranthi, K R (2006): ‘Bt Retreat in China’, Down to Earth, 15(7): 20-21. Mayee, C D, P Singh, A B Dongre, M R K Roa and S Raj (2002): Transgenic Bt cotton, Central Institute For Cotton Research, India: 1-30.Mishra, S (2006): ‘Bt Retreat in China’, Down to Earth, 15:20-21. Narayanmoorthy, A and S S Kalamkar (2006): ‘Is Bt Cot-ton Cultivation Economically Viable for Indian Farm-ers? An Empirical Analysis’,Economic & Political Weekly,41 (26): 2716-24.Norris, R F, E P Caswell-Chen and M Kogan (2002): Con-cept inIntegrated Pest Management, Prentice-Hall of India, New Delhi.PAU (1979, 2004 and 2005):Package of Practices for Crops of Punjab-Kharif, Punjab Agricultural University, Ludhiana. – (2007): Package of Practices for Crops of Punjab-Kharif, Punjab Agricultural University, Ludhiana.Perlak, F J, M Oppenhuizen, K Gustafson, R Voth, S Sivas-upramanian, D Heering, B Carey, R A Ihrig and J K Roberts (2001): ‘Development and Commercial Use of Bollgard® Cotton in the USA-Early Promises Versus Today’s Reality’,Plant Journal,27:489-502.Peshin, R (2005): Evaluation of Dissemination of Insecti-cides Resistance Management Technology in Cotton Crop in Punjab, PhD Dissertation, Punjab Agricultural University, Ludhiana.Pradhan, S (1964): ‘Assessment of Losses Caused by Insect Pests of Crops and Estimation of Insect Population’ in Entomology in India, Entomological Society of India, New Delhi, pp 17-58.Qaim, Matin and A D Janvry (2005): Bt Cotton and Pesti-cide Use in Argentina: Economic and Environmental Effects’,Environmental and Development Economics, 10: 179-200.Rogers, E M (1995): Diffusion of Innovation, Free Press, New York.Sen, A and M S Bhatia (2004): State of the Indian Farmer: A Millennium Study-Cost of Cultivation and FarmIn-come, Vol 14, Ministry of Agriculture, Government of India, New Delhi.Shetty, P K (2004): ‘Social-Ecological Implication of Pesti-cide Use in India’, Economic & Political Weekly, 39: 5261-67.TFAI (2002):Fertiliser Statistics, The Fertiliser Associa-tion of India, New Delhi.Wahab, S (1997): Biotechnological Approaches for Manage-ment of Pests, Diseases and Weeds in Agriculture-Coun-try Status’, Pesticide World, September-October, 13-21Wang, S (2006): It’s Sad to See This Technology Die at the Age of Seven’ in Down to Earth, 15(7) :21.Zehr, U B (2006): ‘Focused Effort Needed’,The Hindu: Survey of Indian Agriculture, 2006:44-45.
REVIEW OF AGRICULTUREEconomic & Political Weekly december 29, 200773Attributes and Socio-economic Dynamics of Adopting Bt Cotton Rajinder Peshin, A K Dhawan, Kamal Vatta, Kamaldeep SinghWorld over, cotton is grown on more than 32 million hectares with approximately 71 per cent of the produc-tion in developing countries. In India, cotton is an im-portant commercial crop being grown on 8.9 million hectares, constituting 27 per cent of the world area under cotton. Despite the largest share in area and third largest share in total world production, India ranks 57th with respect to cotton productivity [TFAI 2002]. The major reasons for low productivity of cotton are multiple pest problems and the crop being grown largely on un-irrigated soil (almost 65 per cent area under cotton is unirrigat-ed). Insect-pests are estimated to cause losses up to the extent of 50 per cent in cotton productivity [CICR 1998]. The losses, in fact, have increased from about 18 per cent in the early 1960s [Prad-han 1964] to 50 per cent in 1990s [Dhaliwalet al2004]. In spite of the significant use of pesticides in cotton, the total crop losses amounted to Rs 287.6 billion (ibid). Chandra (1998) and Wahab (1997) estimate the annual damage by bollworm to the tune of Rs 120billion. Development of hybrid cotton varieties in 1970s, which occupy 45 to 50 per cent of area under the crop,1 gave boost to its pro-ductivity in the past. As a result, cotton productivity2 in India increased from 1.06 q/ha during 1970-71 to 2.65 q/ha during 1996-97. In Punjab, it jumped from 4.63 q/ha during 1990-91 to 5.02 q/ha during 1994-95 and then followed a downward slide and reached an all time low of 1.79 q/ha in 1998-99, when farmers were unable to recover even the cost of cultivation. This was due to the outbreak of American bollworm (helicov-erpaarmigera), which is one of the major insect-pests of cotton and has developed resistance to pesticides due to selection pressure. From about two-three insecticide applications in cottonin1970s, farmers were reported to have applied more than 30 insecticide applications in cotton3 during 2003 and 2004 [Peshin 2005]. To overcome this downward slide in the productivity of the cotton crop, many integrated pest management (IPM) pro-grammes such as the regional programme on cotton IPM by Commonwealth Agricultural Bureau International (CABI) in 1993,FAO-European Union IPM programme for cotton in 2000, National Agricultural Technology Project (NATP) for IPM in 2000and Insecticide Resistance Management Programme in 2002 were initiated. The focus of these IPM programmes was to reduce and rationalise pesticide use by encouraging farmers to adopt other pest management practices like cultural and manual-mechanical practices in the absence of any effective bio-agents.Rajinder Peshin (rpeshin@rediffmail.com) is in the Division of Agricul-tural Extension Education, Sher-e-Kashmir University of Agricultural Sciences and Technology, Jammu; A K Dhawan is at the Department of Entomology, Punjab Agricultural University, Ludhiana; Kamal Vatta is at the Department of Economics, Punjab Agricultural University, Ludhiana; Kamaldeep Singh is at the Department of Entomology, Punjab Agri-cultural University, Ludhiana.Bt cotton, adopted as a solution to curb losses caused by bollworms and to reduce the use of pesticides, successfully brought about a decline in pesticide consumption and expenditure, increase in productivity along with a higher output-input ratio in Punjab in 2004-05. Despite increased productivity and reduced pesticide use, Bt cotton is expected to increase the incidence of primary bollworms, which could develop resistance and secondary pests. But the undesirable, indirect and unanticipated consequences of innovation go together, as do the desirable, direct and anticipated consequences. 
REVIEW OF AGRICULTUREdecember 29, 2007 Economic & Political Weekly74developed during recent decades are tolerant against pests and are the main alternative to the use of synthetic organic pesticidesfor crop protection. The first transgenic crop variety was developed in squash crop in 1994, which was resistant to vi-ruses [Norriset al2002]. During the same year, glyphosate-re-sistant transgenic varieties of maize and soybean crops were released. In the year 1995, insect-resistant cotton, maize and potatoes using Btgenes were released in theUS (ibid). The transgenic crops are grown in 17 countries on a total area of 81 million hec-tares [Zehr2006]. In India, Bt cotton was approved for cultivation in 2002. The Bt cotton acreage increased manifolds since 2002 (Table 1).There are two schools of thought; one, advocating theadop-tion of Bt cotton for reducing the use of insecticides, cutting farm-ers’ production costs and increasing yield [Bannett et al 2004; Perlak et al 2001; Huang et al 2002] andthe second criticising biotechnology and doubting its usefulness for small farmers in the developing countries [Mishra 2006; GRAIN 2001; Wang 2006].In theUS, in 1997, the farmers grow-ing Bt cotton achieved greater produc-tivity by $24.43 per acre including insect control costs and the increase inreturns rose to $39.86 per acre [Boulter and Hilder 2002]. Huang et al (2002) while reporting the potential benefits and impact of Bt cotton in China highlighted that the actual use of pesticides in Bt cotton was much less ranging from 11.8 kg/ha in 1999 to 32.9 kg/ha in 2001 as compared to that in non-Bt cotton, which varied from 48.5 kg/ha to 87.5 kg/ha over the same period. On an average, cultivation of Bt cotton helped re-duce pesticide use by 35.7 kg/ha (55 per cent reduction). How-ever, they failed to explain the significant and unwarranted rise in pesticide use in Bt cotton from 1999 to 2001. ‘Bt Technology Adoption, Bounded Rationality and the Outbreak of Secondary Pest Infestation in China’ claims that afterseven years of Bt cotton introduction in China (1996 to 2004), the expenditure on pesticides for Bt and non-Bt was identical in 2004 at $101 per ha and the earnings from Bt cotton werelower [Mishra 2006]. Narayanamoorthy and Kalamkar (2006) reported the economical viability of Bt cotton for Indian farmers (Maharashtra). Contrary to expectations, the total quantity of pesticides used in Bt cotton varietyMECH 162 was higher than non-Bt cotton varieties. The average net profit from Bt cotton was Rs 31,880 per ha, about 80 per cent higher thanthat from non-Bt cotton. There was no significant difference in pesticide use between Bt and non-Bt cotton varieties. However, Table 1: Rate of Adoption of Bt Cotton Year Total Area Area under Extent of under Cotton Bt Cotton Adoption of (Million ha) (Million ha) Bt Cotton wrt % Area2001-02 8.73 – –2002-03 7.67 0.07 0.912003-04 7.63 0.23 3.012004-05 8.50 1.3015.292005-06 8.96 1.4115.742006-07 8.96 3.0033.48Sources: Government of Punjab (2006), GoI (2005), Zehr (2006).The pest scenario also underwent a change since the introduc-tion of synthetic pesticides. The pesticide, which was considered “silver bullet” to overcome all pest-related problems, has in fact compounded the problems. Rachel Carson (1962) wrote a book Silent Spring that brought about the change in the mindset of the people associated with agriculture. However, it was not until 1975 that much attention was paid to the dysfunctional consequences of pesticide adoption in crop cultivation. Scientists and extension workers led farmers into “pesticide treadmill”. According to avail-able estimates, out of the total pesticide consumption, 50 per cent and more is used for the cotton crop amounting to Rs 160 billion, of which Rs 110 billion are spent only to control bollworms [Alagh 1988; Mayeeet al2002]. The share of insecticide cost in the total cost of cultivation of cotton in Punjab increased many folds from 2.1 per cent during 1974-75 [Dhaliwal and Arora 2001] to 21.2 per cent during 1998-99 [Sen and Bhatia2004]. In pesticide hotspots like Bhatinda district in Punjab, it was 50 per cent of the total cost of cultivation [Shetty 2004]. This was despite the fact that many IPM programmes were implemented in Punjab from time to time. Introducing Bt cotton for cultivation was another step to re-duce the pesticide use in the crop. Bt technology was found to have significantly reduced application rates of toxic chemicals, while significantly increasing yields [Qaim and Janvry 2005]. Bt cotton is grown in eight states, viz, Madhya Pradesh, Gujarat, Maharashtra, Andhra Pradesh, Karna-taka, Tamil Nadu, Punjab and Harya-na. In Punjab, the government of India approved the cultivation of Bt cotton in March 2005. However, the Punjab farmers had started cultivating Bt cot-ton in 2002, purchasing seeds from far off states like Gujarat. The six varieties recommended for cultivation in 2005 were: Ankur 2534, Ankur 651,RCH 134, RCH 317,MRC 6301 andMRC 6304. Out of these permission for two varieties, viz, Ankur 2534 and Ankur 651 was withdrawn by the Punjab Agricultural University (PAU), Ludhiana in 2006 due to low productivity. The present study attempts to investigate the socio-economic dynamics, attributes and rate of adoption of Bt cotton in Punjab. The study is divided in six sections. The next section briefly outlines the controversy surrounding Bt cotton. Details of the database and methodology used in the study are given in the second section. In the third section, awareness, adoption and attributes of Bt technology have been highlighted. Input use and productivity pattern of Bt cotton have been discussed in the fourth section. The economic structure of Bt cotton is high-lighted in the fifth section. The last section presents the concluding remarks. 1 Controversy Surrounding Bt CottonSpraying pesticide formulations based on bacillus thurigiensis (Bt) have been in limited use for the control of larvae of lepidopteran pests for more than 40 years. The transgenics Table 2: Socio-Economic Profile of the RespondentsParticular DistrictOverall BathindaFerozepurMansa(n=210) (n=70) (n=70) (n=70) A Pattern of education (% of respondent farmers) Illiterate 10 9 16 11 Primary 7 17 11 12 Middle 20 23 26 23 Matric 37 40 36 38 10+2 10 7 11 9 Graduate and above 16 4 0 7B Farmers with telephone connection (%) 67 64 56 62C Average size of operational holding (ha) 8.17 14.10 6.96 9.75D Distribution of operational holdings (%) Small (below 2 ha) 11 3 6 7 Semi-medium (2-4 ha) 20 11 27 20 Medium (4-10 ha) 39 37 51 42 Large (above 10 ha) 30 49 16 31
REVIEW OF AGRICULTUREEconomic & Political Weekly december 29, 200775it is too early to generalise in India, where four million small and marginal farmers have taken up cultivation of Bt cotton withestimated adoption rate of 50 per cent by the end of 2007 [Mishra 2006]. Illegal and spurious seeds coupled with non-maintenance of minimum 20 per cent refugia by these farmers may result in severe pest attack on Bt cotton due to selection pressure and outbreak of secondary pests like whitefly [Chari 2006]. The bollworm is expected to develop resistance in 2007-08, where it was introduced in 2002 [Kranthi 2006]. Cotton monoculture and non-adoption of IPM principles is the recipe for such failure.2 MethodologyThe present study was conducted in the cotton growing areas of Punjab. The state accounts for 10 per cent produc-tion from 5 per cent area under cotton in the country and the highest average productivity since 2000-01. The cotton growing districts in Pun-jab are Bathinda, Ferozepur, Mansa, Mukatsar, Faridkot, Sangrur and Moga with more than 70 per cent of cotton growing area fall-ing in the first three districts.4 These three districts were purpo-sively selected for conducting formative evaluation of insecticide resistance management (IRM) based IPM programmes in cotton. TheIRM programme is being implemented by Central Institute of Cotton Research, Nagpur since 2002 in 28 districts, distributed over 10 states in India, which account for more than 80 per cent insecticide use in cotton. Thirty villages (10 from each district) were covered under the IRM programme in 2004-05. From these 10 villages in each dis-trict, five villages were randomly selected for data collection, making a total of 15 villages being covered under programme. In addition, out of nine control villages (not participating in the IRM programme), six villages (two from each district) were selected randomly. Thus, out of a total of 21 villages, the data were col-lected from 210 farmers (10 from each selected village) for the present study. Most of the adoption research has worked on the socio-economic attributes of farmers to predict the causes of adoption or non-adoption. Rogers’ diffusion model has dealt comprehensively by generalising socio-economic characteris-tics responsible for adopting new technology in agriculture [Rogers 1995]. These generalisations are that formal education, income, level of living, landholding, social mobility, empathy, etc, have positive correlation with adoption, whereas age has no relationship with adoption. The personality varia-bles and communication behaviour have also been generalised. Due to the huge already existing body of knowledge about these socio-economic and other variables, this study does not explore the relationship with adoption/rate of adoption of Bt cotton, which is conventional input technology like the green revolution technologies of high yielding varieties, fertilisers, pesticides, etc. The diffusion research has put much effort on studying personal characteristics, but relatively little effort has been devoted to ana-lyse innovation difference/technological attributes which affect the rate of adoption (ibid). To overcome the pro-innovation bias, alternative research approaches to after-the-fact-data collection about how Bt cotton diffused, had to be employed. To study the diffusion of technology overtime, data are to be collected during the diffusion process at different stages and not af-ter completion of the diffusion proc-ess (ibid). This base-line study of Bt cotton adoption will serve as the base for further studies of Bt cotton adop-tion in Punjab. A semi-structured questionnaire, in the local language, was distributed amongst selected farmers before the start of the cotton-growing season in 2004-05. This was done in order to avoid the problems/discrepancies as-sociated with recall/recollection of in-formation by the farmers after the growing season is over. The respondents were revisited after regular intervals (with a time gap of not more than three weeks) during the crop growing sea-son, to ensure proper and correct recording of the required infor-mation and to dispel doubts emanating when recording the data. The questionnaire focused on the socio-economic characteristics of the respondents, extent and level of adoption, input use, cost of cultivation, production and returns.3 Awareness, Adoption and Attributes of Bt CottonMajority of the farmers (around 89 per cent) was literate with the proportion of illiterate being 11 per cent. The modal level of edu-cation was matriculation. Sixty-two per cent of these farmers had telephone connections. Average size of the operational holding was the highest in Ferozepur (14.10 ha), followed by Bathinda (8.17 ha) and Mansa (6.96 ha) with an overall average landhold-ing of 9.75 ha (Table 2, p 74). The average holding size was much higher than the average size of 3.61 ha in the state largely due to the progressive farmers (active farmers or larger landholders) be-ing trained under IRM project and the sample consisting largely of these farmers. The fact is further strengthened from pattern of distribution of the sample holdings, majority of which (more than two-thirds) belonged to the medium and large size categories. 3.1 Adoption of Different Cotton CultivarsCotton is the main ‘khraif’ crop in the study area. About 64 per cent of the total area was under cotton, with its proportion being the highest in Bathinda (67.92 per cent) and that being the least in Mansa (58.30 per cent). About 60 hybrid cotton, 24 non-hybrid cotton, Bt cotton and ‘desi’ cotton varieties are being cultivated in Punjab[Peshin2005]. The extent of adoption of Bt cottonand other varieties was measured in terms of percentage of area cov-ered under different varieties and percentage of farmers adopt-ing Bt cotton and other varieties. The extent of adoption of Bt cotton revealed a varied picture in the three districts. The extent Table 3: Adoption of Varieties of Cotton DistrictOverall BathindaFerozepur Mansa (n=210) (n=70) (n=70) (n=70) A Average size of operational holding (ha) 8.17 14.10 6.96 9.75B Average area under cotton crop (ha) 5.63 9.22 4.10 6.32 (67.92)(64.47)(58.30)(63.97)C Extent of adoption (% farmers)1 Bt cotton 93 42 83 72 Other hybrids 71 268 93 63 Othervarieties2 45 88 38 57D Per cent of total cotton area under Bt cotton 42 10 21 22 Other hybrids 29 8 58 25 Other varieties1 29822153 Figures in the parenthesis are % of cultivated area (i e, B as % of A); (1) Represents multiple responses. (2) It includes non-hybrid varieties and second progeny (F2 generation) Bt cotton varieties.All decimals have been rounded up to nearest whole numbers. 
1Awareness knowledge Rate of adoption 56 20 41 30 12 4 8 00 0 
72 16 80 39 9 0 40 Rate of adoption Awareness knowledge 
REVIEW OF AGRICULTUREEconomic & Political Weekly december 29, 200777Table 5: Awareness, Attributes and Expected Adoption of Bt Cotton Percentage of Farmers BathindaFerozepurMansaOverallA Sources of awareness/information (multiple response) Other farmers 77 73 78 76 Representative of a company 21 20 16 19Commission agent 2 6 6 5Newspapers 13 10 10 11Extension officer 2 0 0 1B Sources of seed (2004-05) Authentic 52 42 33 43 Notauthentic 44 58 60 52Cannotsay 4 0 7 5C Attributes of Bt cotton (multiple response) Resistance to bollworms 96 100 98 97 Higherproductivity 83 87 93 87 Saving on pesticide expenditure 94 57 93 84 Easy to adopt 92 37 75 72 Compatible 77 23 68 60Not compatible 6 0 8 5 Facilitates timely sowing of wheat 0 7 10 2 High cost of seed 8 10 15 11 More water requirements 46 17 18 29 Higher fertiliser requirements 8 0 0 3 Susceptible to cotton leaf curl virus 0 0 3 1 Susceptible to tobacco caterpillar 6 3 0 3Nobenefit 4 3 8 5D Estimated adoption of Bt cotton in 2005-06 Area under Bt cotton (ha) 67.02 28.30 51.70 46.26 Farmers growing Bt cotton (%) 100 78 80 86agents (5 per cent). Interpersonal communication channels were the main source of diffusion of this innovation. The awareness pattern formed ‘S’ shaped curve overtime when cumulative number/percentage of farmers was plotted on a graph, while the frequency distribution of the number of farmers getting awareness per year approached normality (Figure 2, p 76). The rate of adoption of Bt cotton also formed an ‘S’shapedcurve,when cumulative number/percentage of farm-ersadoptingBtcotton overtime was plotted and approached normalitywhenthefrequency distribution of farmers adopting it over time was plotted on a graph. These results are in agree-ment with Rogers’ diffusion theory revealing that knowledge precedes adoption at a fasterrate. Bt cotton being similar to green revolution technologies, the diffusion theory fits in here, which is not the case with inter-dependent complex technologies likeIPM. Against 4 per cent rate in 2002 (the seed of Bt cotton was purchased by farmers from Gujarat or through com-mission agent as the government of India allowed cultivation of Bt cotton in Pun-jab only in 2005), the adoption multiplied to 16 per cent in 2003-04 and then jumped to 80 per cent during 2005-06.3.3 TechnologicalAttributes The attributes of technology are impor-tant variables that determine the rate of adoption. Diffusion studies have mostly tended to regard all technolo-gies/innovations as equivalent [Rogers 1995]. Technological attributes are more important to speed up the rate of adoption. The generalised attributes, which affect adoption are: relative ad-vantage, compatibility, trialability, ob-servability and complexity. Out of these, first four attributes are positively related totherate/extent of adoption, while complexity is negatively related to it. The attributes of Bt cotton as reported by farmers have been presented in Table 5.First, there has been a clear relative advantage of Bt cotton cultivation over the other varieties as reported by the farmers. Almost 97 per cent of farmers admitted to its resistance to boll-worms,87 per cent to relatively higher yields, 84 per cent for their potential to save expenditure on pesticides. Even 72 per cent of the farmers reported that Bt cotton varieties were easy to adopt. Further, almost 60 per cent of the farmers reported compatibility of Bt cotton with their farming systems and soils, etc. However, there were some reports of its non-compatibility (although small), Bt cotton being water-intensive (29 per cent) and fertiliser-intensive (3 per cent). Further, 11 per cent of the farmers cited Bt seed to be very costly, and 3 to 5 per cent reported their susceptibility to cotton leaf curl virus (CLCV) and tobacco caterpillar. Bt cotton being susceptible to CLCV and tobacco caterpillar needs to be researched. These non-compatibility issues can hinder the adoption of Bt cotton. Recently, the new state government has slashed the seed prices by almost 40 per cent to give a further push to the adoption. Finally, resistance to bollworms and higher yields were the benefits (observability) and the respondent farmers reported no complexity during Bt cultivation. All these attributes of Bt cotton have pushed up the rate of its adoption even before it was given an official go-ahead for cultivation in Punjab.4 Pattern of Input Use and ProductivityIn this section, based on the field study carried out during 2004-05, a comparative analysis of Bt cotton with other varie-ties has been done with respect to the input use pattern. The impact of Bt cotton in reducing the insecticide use and its productivity advantage as a result of reduced insect pest losses has also been discussed. 4.1 InputUseThe variety-wise input use pattern for cotton is depicted in Table6 (p 78). The recommended seed rate for hybrid and non-hybrid cotton varieties is 3.75 kg/ha and 10 kg/ha, respectively. In case of non-hybrid varieties, the PAU recommen-dation has changed over time. The seed rate recommended was 20 kg/ha in 1979, 10 kg/ha in 2004 and 8.75 kg/ha in 2005 [PAU 1979, 2004, 2005]. However, the ac-tual seed rate has differed from the rec-ommended levels as farmers used the seed on the basis of their own experience and feasibility analysis. In hybrid varie-ties of cotton, 97 per cent farmers had applied 1.87 kg to 2.25 kg seed per ha, whereas in case of non-hybrid varieties 80 per cent of the farmers applied higher seed rate (more than 10 kg/ha). The average seed rate in Bt cottonwas2.07 kg/ha and that in non-Bt hybrid varieties was 2.25 kg/ha. In case of non-hybrid varieties the average seed rate used was 13.65 kg/ha. It is clear that the farmerswereusinglessthanrecommended seedrate for the hybrid varieties, which was largely due to higher cost of hybrid seeds, and the seed companies recommending seed rate of 2.25 kg/ha, which is lower than the PAU recommendation of3.75 kg/ha. Fertiliser use was the highest in the case of Bt cotton, fol-lowedbyhybridcotton and was the least in the non-hybrid cotton varieties. The nitrogenous fertiliser use in Bt cotton was higher by 23 and 31 per cent when compared to the other hybrid and non-hybrid varieties, respectively. The respective phosphatic fertiliser use was higher by 17 and 50 per cent and the potashic fertiliser use was higher by 104 and 413 per cent. The use of zinc-sulphate was also higher in Bt cotton by 25 and 10 per cent, respectively. Although there wasnodifferenceinthe 
REVIEW OF AGRICULTUREdecember 29, 2007 Economic & Political Weekly78number of insecticide applications in Bt cotton when compared to other cotton cultivars. 4.3 Productivity of Bt CottonInsect pest losses caused by the bollworm complex especially helicoverpa armigera (American bollworm) in Bt cotton were significantly reduced thus, resulting in higher productivity. The overall productivity of Bt cotton6 was 24.17 q/ha as compared to the productivityof23.39 q/ha and 18.51q/haofotherhybrid and non-hybrid varieties, respectively (Table 8). The seedcotton yield of recommended Bt cotton varieties (RCH 134,RCH 317, MRC 6,301 and MRC 6,304) was 25.19, 26.54 and 25.26 in district Bathinda, Fero-zepur and Mansa, respectively. In this field study, we have included as well as recom-mended other varieties of Bt cotton, non-Bt cotton hybrids and non-hybrid varieties to provide a comprehensive overview of the cotton growing scenario in Punjab. The average yield of different recommended varieties of cotton recorded by the PAUbased on re-search and farmer field trails is given in Table 9 (p 79). It is evident that the yields recorded by thePAU in case of recom-mended Bt cotton varieties are higher than the yield obtained by the farmers in the study area except in Bathinda. The Bt cotton varieties cultivated by the farmers were those approved by the GEAC and recommended by the PAU in 2005 and other unde-script Bt varieties like Tulsi, Navbharat 151, etc. But in case of non-Bt hybrid varieties the farmers had recorded higher yield than the average yield recorded by the PAU, but more than 50percent of the cotton growing area was under the varieties not recommended by the PAU like Ankur 2,226, Ankur 2,534, RCH 134,RCH 317 etc. ThoughPAU recommends non-Bt Ankur 651 for cultivation in Punjab, it has withdrawn the recommendation for Ankur 651 Bt cotton due to its lower yields (17.50 q/ha) compared to other Bt cotton varieties (Table 9). Some of the non hybrid and desi cotton varieties (F 1861, F846) have comparable or higher average yield potential than Bt cotton varieties but the losses caused by insect pests (es-pecially bollworms) neu-tralise that higher yield potential. pesticide5 use between the other hybridandnon-hybrid varieties, its use declined significantly by almost 137 per cent in caseofBtcotton. There was no such difference in the number of irrigations applied. Hence, there wasaclearcomparativeadvan-tage of Bt cotton in terms of pesticideusewhencomparedtothe other hybrid and non-hybrid varieties, translating ultimately into relatively lower costs of cultivation and hence more returns. 4.2 ReducingInsecticideUse There has been a significant decline in pesticide consumption in the case of Bt cotton when compared to the other hybrid and non-hybrid varieties. However, the impact on the insecticide use needs to be discussed. We have attempted to highlight such an impact in terms of the number of insecticide applications, quan-tity consumed (technical grade material) and the costs incurred. Bt cultivation has resulted in a significant decline in the number of insecticide applications, its average consumption as well as the costs incurred. The average number of insecticide applications in Bt cotton was 4.7 as compared to 10.5 in case of non-Bt cultivars (Table 7). In all districts, the number of applications reduced by more than half in Bt cotton as compared to non-Bt varieties. The consumption of insecticides was 2.57 kg/ha and 6.44 kg/ha in Bt and non-Bt varieties, respectively. The average cost of insecticide use reduced significantly by more than 60 per cent when com-pared to non-Bt varieties. The results followed the same pattern in all the districts although they varied in their extent. The highest numbers of insecticide application were registered in Ferozepur both in Bt cotton (6.51) as well as other cultivars (13.08). The farmers in this district were also using more number of cocktails of insecticides. The maximum number of insecticide applications by tank mixing of insecticides separately was 30 in non-Bt cotton, while it was the minimum in Bt cotton (just two). The Abohar area in this district is the hot spot of insecticide use in Punjab. In Bt cotton, farmers had mainly applied insecticides for sucking pest such as jassid, whitefly and in certain pockets for tobacco caterpillar due to its outbreak. A few farmers also applied insecticide for bollworms, either due to its attack in later stages of crop cycle (or fear of such attack) or owing to some incidences of spurious seeds of Bt cotton. Despite all this, there was lesser Table 6: Input Use Input Bathinda Ferozepur Mansa Overall Bt Cotton Other Non- Bt Cotton Other Non- Bt Cotton Other Non- Bt Cotton Other Non- Hybrids hybrids Hybrids hybrids HybridshybridsHybridshybridsFertiliser use (kg/ha) Nitrogenous 253.83 231.12159.29 305.66 239.54209.33236.54194.85146.02 260.59211.36199.50 Phosphatic 113.19 110.61 82.34 179.02109.32 80.77 93.48 100.63 87.84122.30104.78 81.47 Potashic 13.19 3.83 6.88 17.89 17.37 1.81 17.55 7.54 10.34 15.23 7.45 2.97Zincsulphate 2.165.091.611.33 02.816.42 03.413.012.402.73Seed (kg/ha) 2.07 2.28 10.80 2.00 2.26 15.04 2.13 2.25 12.54 2.08 2.26 13.65Pesticide*(weedicide+insecticide) 2.824.634.63 3.20 8.25 8.252.96 5.15 5.152.826.686.68Average no of irrigations 7.47 7.47 7.47 5.60 5.60 5.60 5.71 5.71 5.71 6.26 6.26 6.26*Average technical grade material. Figures in parentheses represent % difference wrt Bt cotton.Table 7: Pattern of Insecticide Use and Cost of Insecticide ApplicationsInput Bathinda Ferozepur Mansa Overall Bt Non-Bt % Bt Non-Bt % Bt Non-Bt % Bt Non-Bt % Cotton DifferenceCotton DifferenceCottonDifferenceCottonDifferenceAverage no of insecticide applications 3.78 7.79 106.08* 6.51 13.08 100.92* 4.98 10.01 101.00* 4.76 10.46 119.74*Insecticide use@ (kg/ha) 2350 7470 217.87* 3978 7934 99.45* 3258 7228 121.85* 2934 7546 157.19*Cost of insecticides (Rs/ha) 2.45 4.25 73.70* 3.08 8.13 164.12* 2.63 4.82 83.27* 2.58 6.44 149.86*@Technical grade material. *means significant at 1 % level.Table 8: Productivity of Bt Cotton, Other Hybrids and Non-hybrid Varieties in Punjab Districts Overall BathindaFerozepurMansa A Productivity(q/ha) Btcotton 27.89 21.20 21.37 24.17 Other hybrids 24.89 20.17 23.15 23.39 Othervarieties1 19.2318.3218.1018.51B Difference in productivity of Bt cotton wrt Other hybrids 12.05* 5.11 -8.33 3.33 Othervarieties1 45.03***15.72**18.07**30.58***It includes non-hybrid varieties and second progeny (F2 generation) Bt cotton varieties.***, ** and * represent significance at 1, 5 and 10 % levels.The yield difference between Bt cotton and other hybrids is not significant. This can be due the fact that farmers had sown undiscript Bt cotton, mostly in district Mansa.
REVIEW OF AGRICULTUREEconomic & Political Weekly december 29, 200779There was no significant difference between the productivity of Bt cotton and other hybrid varieties in the state. It was only in Bathinda that a significant difference (12 per cent) was recorded. In district Mansa, the Bt cotton yield was on the lower side than non-Bt hybrids. This may be due to the fact that in this district the Ankur varieties 651,2226, 2,534 and 655 were the most widely adopted varie-ties and the majority of farmers had cultivated undescript and Ankur Bt cotton varieties which are lower yield-ing (17.50q/ha) than other Bt cotton varieties (Table 9). The recommenda-tion of Ankur Bt cotton varieties was withdrawn for the same reason by the PAU in 2006. However, the productiv-ity of Bt cotton differed significantly from that of non-hybrid cotton varie-ties in the entire study area, the over-all difference being more than 30 per cent. The difference was the largest in Bathinda (45 per cent), while the dif-ference (although significant) was just 15.7 per cent and 18.1 per cent in Ferozepur and Mansa, respective-ly. The productivity of different cul-tivars of cotton was higher in Bathinda district than that in Mansa and Ferozepur districts. 5 Economic Structure of Bt CultivationThe relative advantage of a technology in terms of economic re-turns plays an important role in the rate/extent of its adoption. The cost-returns structure of different cultivars of cotton has been discussed below to highlight the economic benefits of Bt-cultivation in the state. The expenditure on cotton production was divided into six cat-egories, namely land preparation, sowing cotton seed, inputs, cost of spraying, hoeing and other expenditure (Table 10). The overall cost of cultivation (per ha) of Bt cotton, other hybrids and non-hybrids was Rs 22,431, Rs 24,259 and Rs 21,713, respectively. The cost of cultivation of Bt cotton was lower by 7.54 per cent than that of other hybrid varieties but was higher by 3.31 per cent than that of non-hybrid varieties. The major reason for lower costs in the former case is the significant reduction in the insecticide cost (including the cost of sprays), which even neutralised the higher cost of Bt seed. The higher cost of cultivation than non-hybrid varieties was due to the reason that the difference in the cost of seed was much higher, which could not be neutralised by reduced insecticide cost and the labour charges for the sprays. The average per ha cost of seed was Rs 4,621 for Bt cotton, Rs 1,914 for other hybrids and Rs 484 for non-hybrids. The cost of insecticides was Rs 3,005 and Rs 7,546 per ha and the labour cost of sprays was Rs 476 and Rs 1,046 per ha for Bt and non-Bt varieties, respectively. The cost of insecticides amounted to 17, 37 and 41 per cent of the total cost of cultivation of Bt cotton, other hybrids andnon-hybrids,re-spectively. It is thus clear that adoption of Bt cottoninthestate has helped in reducing the cost of cultivation of the crop. Table 10: Cost-returns Structure for Different Cultivars of Cotton in Punjab Districts Overall Bathinda Ferozepur Mansa Cost of Cultivation (Rs/ha) ALandpreparation Pre sowing irrigation1 472 375 496 448 Ploughing and planking2 2,3362,1722,0762,195 Sowing 2 415 415 415415Inputs B Seed3 Bt cotton 4,607 4,445 4,733 4,621 Hybridcotton 1,960 2,030 1,846 1,914 Other varieties (includes non-hybrid varieties and Bt cotton F2 generation) 512 458 512 484C Chemicalfertilisers4 Bt cotton 2,351 3,212 2,192 2,494 Hybridcotton 2,235 2,224 1,887 2,051 Other varieties 1,578 1,800 1,620 1,764D FYM5 150 64161 118E Insecticide6 Bt cotton 2,350 3,978 3,258 3,005 Hybrid cotton and other varieties 7,470 7,934 7,228 7,546F Weedicide7 306 134 261 234G Irrigation1 3,574 2,142 2,577 2,765H Cost of spraying (Rs ha–1) H1Insecticides Bt cotton 378 651 498 476 Hybrid cotton and other varieties 779 1,308 1,001 1,046H2 Weedicide 80 39 88 69 Hoeing (Rs ha–1) Manual 746 560 713 673 Tractorised 350 247 256 284J Other expenditure (Rs ha–1)3 J1Picking Bt cotton 4,741 3,604 3,633 4,109 Hybridcotton 4,231 3,429 3,936 3,976 Other varieties 3,269 3,114 3,077 3,147J2 Cleaningbunds 175 175 175 175J3 Cleaning of field after picking 350 350 350 350Total cost of cultivation (A to J) Bt cotton 23,434 22,519 22,009 22,431 Hybrid cotton 25,479(+8.03) 23,598 (+4.57) 23,466 (+6.07) 24,259 (+7.54) Other varieties 22,562 (-3.86) 21,287(-5.79) 21,006 (-4.77) 21,713 (-3.31)Output (q/ha) and returns (Rs/ha)A Average seed cotton yield Btcotton 27.89 21.20 21.37 24.17 Hybrid cotton 24.89 20.17 23.15 23.39 Other varieties 19.23 18.32 18.10 18.51B Gross returns 8 Btcotton 50,760 38,584 38,893 43,989 Hybrid cotton 45,300 (-12.05) 36,709 (-5.11) 42,133 (+7.69) 42,570 (-3.33) Other varieties 34,999(-45.03) 33,342 (-15.72) 32,942 (-18.07) 33,688 (-30.56)C Net income Bt cotton 27,326 16,065 16,884 21,558 Hybrid cotton 19,821(-37.86) 13,111 (-22.53) 18,667(+9.55) 18,311 (-17.90) Other varieties 12,437 (-119.77) 12,055 (-33.26) 11,963 (-41.14) 11,975 (-80.28)D Returns-costratio Btcotton 2.17 1.71 1.77 1.96 Hybridcotton 1.78 1.56 1.80 1.75 Non Hybrid cotton 1.55 1.57 1.57 1.55E Pesticide (including application) cost as % to total cost of cultivation Btcotton 13.29 21.32 18.65 16.87 Hybrid cotton 33.89 +20.6 39.90 +18.58 36.56 +17.91 36.67 +19.80 Non-hybrid cotton 38.27 +24.98 44.23 +22.91 40.84 +22.19 40.97 +24.10The figures have been given separately for Bt cotton, hybrid cotton and non-hybrid cotton for all those operations where the difference was estimated to exist. For rest of the operations the cost estimates are the same. The yield levels and returns have also been given separately. Figures in parenthesis are % difference over Bt cotton. The cost of cultivation includes the labour costs – family and hired labour (Rs 80 per man day). The cost of spraying pesticides Rs 100 per ha for both tractorised and manual spraying. The expenditure on account of hoeing was computed on uniform market rates. The average number of hoeings, tractorised and manual were multiplied with Rs 250 ha–1 and Rs 625 ha–1, respectively. The other expenditures included cost of picking (R s 170 q–1), cleaning of bunds (Rs 175 ha-1) and cleaning of fields after picking (Rs 350 ha–1). All these were calculated at uniform market price.(1) Irrigation as per expenditure given by a farmer for canal and tube well irrigation. (2) At uniform market rates. (3) Price of purchased seed or own seed at market price (Bt cotton: Rs 4,621, hybrid cotton: Rs 875-2,250, non-hybrid: Rs 312-500/ ha). (4) Fertilisers at market price plus cost of application. (5) Farm Yard Manure (FYM) @ Rs 150/trolley (40 tonnes). (6-7) Insecticides and weedicides at market price. (8) Returns are computed by using average actual market price (Rs 1,820/q).Table 9: Average Potential Yield of Different Varieties in PunjabVariety AverageYield (q/ha)A Bt cotton varieties RCH 134 Bt 28.75 RCH 317 Bt 26.25 MRC 6301 Bt 25.00 MRC 6304 Bt 25.25B Non-Bt hybrid varieties Ankur 651 17.50 Whitegold 16.25 LHH144 19.00C Non-hybrid cotton varieties F 1861 16.25 F 1378 25.00 F 846 27.25 LH 1556 21.25D Desi cotton varieties PAU 626 H 24.50 Moti 21.25 LD694 17.50 LD327 28.75Average lint yield for the state as a whole for cotton is 731 kg/ha.Source: PAU (2007).
REVIEW OF AGRICULTUREdecember 29, 2007 Economic & Political Weekly80As a result of such adoption, not only the proportion of insecti-cide use in the total cost has reduced but the net returns have also followed an increasing pattern. Increased attack of American boll-worm on the crop during the late 1990s seriously endangered the economic viability of this crop in the state resulting in fast depletion of this crop from the cultivation map. However, attractive econom-ics of cultivation of Bt cotton in the form of reduced cost of cultiva-tion and higher returns due to better yields have facilitated itsfast adoption by the state farmers ensuring its significant revival in the state. Owing to the difference in yields of these cultivars, the costof picking also varied accordingly and was Rs 4,109, Rs 3,976 and Rs3,147 per ha for Bt cotton, other hybrids and non-hybrids, re-spectively. On the productivity side, the average yield of seed-cotton was 24.17 q/ha, 23.39 q/ha and 18.51 q/ha for the respective culti-vars. As a result, the gross returns (per ha) were Rs 43,989 for Bt cotton, Rs 42,570 for other hybrids and Rs 33,688 for non-hybrids. Hence, at present, Bt cotton cultivation seems economically vi-able in Punjab. The economic viability of Bt cotton is reflected by the fact that the ratio between gross returns and cost of cultiva-tion was 1.96 compared to 1.75 in other hybrids and 1.55 in non-hybrids. The ratio is even better in Bathinda at 2.17, which might be due to the benefits of an earlier and hence, greater extent of adoption in the district. It is a positive sign and implies that the benefits can improve once the adoption picks up, spurious Bt cot-ton seed is eliminated from the market and the cultivation prac-tices based on IPM principles are standardised in the state. 6 ConclusionsThe base line study of adoption of Bt cotton in Punjab has rein-forced the diffusion theory that input intensive technologies follow the normal distribution curve of awareness-knowledge andrate of adoption with the latter following the former. Bt cotton cultivation has resulted in higher external input use, namely, fertilisers and irrigation water but reduced the insecti-cide use. The cost of Bt cotton seed and higher fertilisation cost has been offset by lower expenditure on account of insecti-cides. The higher gross returns due to higher productivity have resultedin improving the economic viability of Bt cotton cultiva-tion.This being a base line study, we need to follow the Bt cot-ton cultivation in Punjab over a period of time to study its functional (desirable) and dysfunctional (undesirable) con-sequences. The “silver bullet” technology for control of boll-worms needs to be studied without the innovation bias otherwise we may again prove wrong as we have been in the case of “pesti-cide” introduction in agriculture. Pesticides were termed as so-lution of all pest problems but it led the farmers onto a “pesticide treadmill”. The integrated pest management principles need to be applied for developing IPM to avoid development of resistance to bollworms against Bt cotton. Despite increased productivity and reduced pesticide use inUS, China and India, the monocul-ture Bt cotton is expected to increase the incidence of primary pest bollworms, which could develop resistance and develop secondary pests like caterpillars. The dying of Bt technology in China in seven years (1997-2003) has also been reported. The scientists and change agents/agencies over-emphasising adop-tion per se, assuming that the consequences of adoption of a technology will be positive and usual survey research methods being inappropriate to investigate consequences [Rogers 1995], may lead the farmers into transgenic-cum-pesticide treadmill. Hence, there is need to study short term, long term, desirable and undesirable consequences of Bt cotton cultivation over a period of time, to evaluate its sustainability.Notes 1 The hybrid varieties occupied 45 to 50 per cent of the total area under cotton before the introduction of Bt cotton. 2 It refers to the total quantity of lint produced per hec-tare. Lint is produced at the rate of one-third of seed cotton yield. 3 The cotton here refers to non-Bt varieties. 4 The total area under cotton in Punjab was 589 thou-sand ha during the period of study (2004-05), out of which the area in Bathinda, Ferozpur and Mansa was 141,133 and 82 thousand hectares, respectively. 5 Pesticide included both weedicides and insecticides. 6 Productivity is in terms of seed cotton. ReferencesAlagh, Y K (1988): ‘Pesticides in India Agriculture’, Eco-nomic & Political Weekly, 20: 1959-64.Bannett, R M, Y Ismeal, U Kambhampati and S Morse (2004): ‘Economic Impact of Genetically Modified Cotton in India’,AgBio Forum: The Journal of Agro-biotechnology Management and Economics,7:96-100.Boulter, D and V A Hilder (2002): ‘Transgenic Crops (Annual)’ in Pimental D (ed), Encyclopedia of Pest Management, Marcel Dekker, Inc, New York, pp 846-49.Carson, R (1962): Silent Spring, Fawcett Crest, New York.Chandra, M (1998): ‘Agriculture: Debt Drives Indian Farmers to Suicide’, http://www.oneworld.net.Chari, M S (2006): ‘Debate: Heavy Cotton’,Down to Earth: 15(6): 23-26.CICR (1998):Annual Report, Central Institute of Cotton Research, Nagpur, India.Dhaliwal, G S, R Arora and A K Dhawan (2004): ‘Crop Losses Due to Insect Pests in Indian Agriculture: An Update’,Indian Journal of Ecology, 31: 1-7.Dhaliwal, G S and R Arora (2001): Integrated Pest Management: Concepts and Approaches, Kalyani Pub-lishers, Ludhiana.Dhawan, A K, A S Sohi, M Sharma, T Kumar and K S Brar (2004):Insecticide Resistance Management in Cotton, Punjab Agricultural University, Ludhiana.GoI (2005): Biosafety Issue Related to Transgenic Crops: With Special Focus on Bt Cotton, Biotech Consortium India and Ministry of Environment and Forests, Gov-ernment of India.Government of Punjab (2006):Agriculture at a Glance: 2005-2006, Information Service, Department of Agriculture, Punjab, Chandigarh.GRAIN (2001): ‘Bt Cotton through the Back Door’, Genetic Resource Action International, Seedling 18(4). http://www.grain.orgHuang, J, R Hu, C Fan, C E Pray and S Rozelle (2002): ‘Bt Cotton Benefits, Costs and Impact in China’, Ag Bio Forum, The Journal of Agrobiotechnology Manage-ment and Economics, 5 :153-66.Kranthi, K R (2006): ‘Bt Retreat in China’, Down to Earth, 15(7): 20-21. Mayee, C D, P Singh, A B Dongre, M R K Roa and S Raj (2002): Transgenic Bt cotton, Central Institute For Cotton Research, India: 1-30.Mishra, S (2006): ‘Bt Retreat in China’, Down to Earth, 15:20-21. Narayanmoorthy, A and S S Kalamkar (2006): ‘Is Bt Cot-ton Cultivation Economically Viable for Indian Farm-ers? An Empirical Analysis’,Economic & Political Weekly,41 (26): 2716-24.Norris, R F, E P Caswell-Chen and M Kogan (2002): Con-cept inIntegrated Pest Management, Prentice-Hall of India, New Delhi.PAU (1979, 2004 and 2005):Package of Practices for Crops of Punjab-Kharif, Punjab Agricultural University, Ludhiana. – (2007): Package of Practices for Crops of Punjab-Kharif, Punjab Agricultural University, Ludhiana.Perlak, F J, M Oppenhuizen, K Gustafson, R Voth, S Sivas-upramanian, D Heering, B Carey, R A Ihrig and J K Roberts (2001): ‘Development and Commercial Use of Bollgard® Cotton in the USA-Early Promises Versus Today’s Reality’,Plant Journal,27:489-502.Peshin, R (2005): Evaluation of Dissemination of Insecti-cides Resistance Management Technology in Cotton Crop in Punjab, PhD Dissertation, Punjab Agricultural University, Ludhiana.Pradhan, S (1964): ‘Assessment of Losses Caused by Insect Pests of Crops and Estimation of Insect Population’ in Entomology in India, Entomological Society of India, New Delhi, pp 17-58.Qaim, Matin and A D Janvry (2005): Bt Cotton and Pesti-cide Use in Argentina: Economic and Environmental Effects’,Environmental and Development Economics, 10: 179-200.Rogers, E M (1995): Diffusion of Innovation, Free Press, New York.Sen, A and M S Bhatia (2004): State of the Indian Farmer: A Millennium Study-Cost of Cultivation and FarmIn-come, Vol 14, Ministry of Agriculture, Government of India, New Delhi.Shetty, P K (2004): ‘Social-Ecological Implication of Pesti-cide Use in India’, Economic & Political Weekly, 39: 5261-67.TFAI (2002):Fertiliser Statistics, The Fertiliser Associa-tion of India, New Delhi.Wahab, S (1997): Biotechnological Approaches for Manage-ment of Pests, Diseases and Weeds in Agriculture-Coun-try Status’, Pesticide World, September-October, 13-21Wang, S (2006): It’s Sad to See This Technology Die at the Age of Seven’ in Down to Earth, 15(7) :21.Zehr, U B (2006): ‘Focused Effort Needed’,The Hindu: Survey of Indian Agriculture, 2006:44-45. 
