ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846
Reader Mode
-A A +A

Farmers’ Choice of Milk-marketing Channels in India

Anjani Kumar ( is with the International Food Policy Research Institute, New Delhi. Ashok K Mishra ( is at the Arizona State University, United States. Shinoj Parappurathu ( and Girish Kumar Jha ( are with the Indian Council of Agricultural Research, Kochi and New Delhi, respectively.

Using nationally representative household-level data, the structure of milk markets is examined and the factors that determine the Indian dairy farmers’ choice of milk-marketing outlets are identified. The analysis of participation in various milk-marketing channels indicates that dairy farmers, irrespective of their asset-status, sales volume, and socio-economic status, prefer to sell their output through cooperatives and government agencies, even if these offer lower prices compared to the local traders. Concomitantly, of the various direct-to-consumer outlets, cooperatives are more inclusive and largely transcend the boundaries of caste and land size. Of the various economic factors that influence farmers’ choice, the access to institutional credit is critical in driving sales through the formal milk-marketing channels.

The authors are grateful to Ramesh Chand for his comments on an earlier draft of the paper. The comments of an anonymous referee for improving the paper are also thankfully acknowledged.

Since the 1970s, the contribution of the livestock sector to India’s agricultural gross domestic product (AgGDP) has been increasing, and this represents one of the most significant changes in the structure of India’s agricultural economy. Between 1970 and 2014, the share of the livestock sector in AgGDP increased from about 17% to approximately 29%, with livestock rearing (of which dairying is a major component) playing an important role in ensuring food and income security of rural households (Birthal et al 2014).

Dairying accounts for more than two-thirds of the value of total livestock output and is evidencing a consistent growth in milk production. Milk production in the country increased from about 21 million tonnes in 1970–71 to about 146.3 million tonnes in 2014–15, and the per capita milk availability increased from 112 grams in 1970–71 to 322 grams in 2014–15 (GoI 2016). Correspondingly, the share of milk and milk products in monthly per capita household food expenditures increased from about 12% in 1983 to about 18% in 2011–12 (GoI 2013).

India continues to be the world’s largest producer of milk; however, the overall milk productivity has remained low, about 4 kg/dairy animal/day, compared to about 34 kg/dairy animal/day in Israel and about 18 kg/dairy animal/day in the European Union. The annual milk yield per dairy animal in India is only half the global average (Kumar et al 2013). The low productivity is associated with the continued dominance of traditional channels for milk marketing apart from other factors such as scarcity of feed and fodder, high mortality rate and poor genetic potential of milch animals, high input costs, and constraints related to storage and post-harvest processing of milk (Kumar and Parappurathu 2014).

The dominance of traditional milk-marketing modes persists despite the emergence and expansion of several modern milk-marketing channels. Studies indicate that about 80% of the milk produced is sold through traditional channels that handle both raw milk and conventionally processed milk products (Staal et al 2006; Kumar and Staal 2010; Kumar et al 2011). However, the structure of milk markets and milk-marketing outlets is evolving with the emergence of several modern milk-marketing chains, and the share of formal milk-marketing channels (for example, cooperatives, processors and commission agents) has increased over time (GoI 2015). The private sector’s share in milk processing has also increased from 43% in 1996 to 60% in 2014–15 (GoI 2015). But, with weak institutions and governance, the livestock sector—and the dairy sector in particular—faces many risks with potentially large negative consequences for livelihood strategies, among other issues.

The slow integration of dairy farmers with organised and modern milk-marketing channels is troubling. Additionally, both dairy farmers and policymakers have raised concerns that socially and economically disadvantaged dairy farms are being systematically discriminated against by the modern milk-marketing channels. However, studies have failed to show the extent to which modern milk-marketing outlets have engaged dairy farms across the states and regions in India, whether milk-marketing patterns in India are scale-neutral, and whether the role of cooperatives in milk-marketing has increased. Moreover, most studies have focused on specific regions and/or milk-marketing channels. This could perhaps be due to a lack of nationally representative data. In this context, the current paper is an attempt to understand the evolution, functioning, and the structure of the extant milk-marketing channels in India. The authors also investigate the factors affecting a smallholder dairy farmer’s choice of a milk-marketing channel; and finally, assess the impact of access to institutional credit on the choice of a milk-marketing channel. The analyses are based on the data from the “Situation Assessment Survey of Agricultural Households in India” (GoI 2014).

Structure of Milk Markets in India

Milk production in India is dominated by smallholder1 dairy farms. The herd size is small, and it is reared in a system closely integrated with crop production. In 2003, about 75% of the milk was produced by marginal (<1 hectare of land) and small (1–2 hectares) farms, and together they contributed about 69% of the total milk production in India (Table 1). In 2013, the share of marginal and small dairy farms grew slightly, to about 76%, and their share of milk production increased to 73% (Table 1). Despite the continued dominance of smallholders, the positive relationship between the scale of milk production and land size has been preserved over time. The average annual milk production ranged from 403 litres for marginal farms to 1,375 litres for large farms in 2003. The range remains more or less unchanged with the data for 2013, with milk productivity of marginal farms being at 441 litres and that of large farms at 1,277 litres (Table 1).

The dominance of marginal and small farms in milk production has increased in most Indian states between 2003 and 2013, with only a few exceptions. In atypical states (Haryana, Karnataka, and Uttarakhand), though the share of smallholders2 in milk production has declined, the small/marginal farms themselves have registered an increase in the volume of milk produced (Table 2). However, the smallholders’ share in total milk production exhibits variations across states. For instance, the share of smallholder farms varied from 48% in Rajasthan to as high as 95% in Uttarakhand in 2003. While in 2013, it ranged from 55% in Rajasthan to 98% in West Bengal (Table 2). In other states, such as Bihar, Himachal Pradesh, Jammu and Kashmir, Jharkhand, Odisha, Tamil Nadu, Uttarakhand, and West Bengal, smallholders constitute 90% or more of the total number of milk-producing households and contribute more than 85% of the respective state’s total milk production. There is also significant variation in the annual milk production per household. The average annual milk production per smallholder household is the highest in Haryana (2,148 litres/household), followed closely by Punjab (2,118 litres/household) and Gujarat (1,397 litres/household). On the other extreme are states like Chhattisgarh (18 litres/household), Jharkhand (33 litres/household), Assam (129 litres/household), Odisha (142 litres/household), and West Bengal (200 litres/household). The differences in the scale of milk production can be attributed to the inequitable distribution of good breeding stock of bovines, variations in the resource base for feed and fodder, animal healthcare, and artificial insemination (Kumar et al 2011).

It is worth noting that rearing dairy animals is not necessarily a commercial activity in India. In some states, 40% of the milk-producing households do not sell milk (Vandeplas et al 2013). In fact, the intake of milk and milk products by Indian households has been increasing over time. Kumar et al (2014) estimated the consumption to increase from 45 kg/capita/annum in 1983 to about 61 kg/capita/annum in 2011–12. Anecdotal evidence shows that only about two-thirds of the milk enter the market. However, the marketed surplus is influenced by several factors, including herd size, household size, household income, food habits of milk producers, home demand, festivals, religious days, and family functions. Therefore, the marketed surplus exhibits a wide variation both within and across states. Since there is a lack of reliable, nationally representative estimates for marketed surplus milk, various studies show differentials in this estimate.

Estimates from the current data, however, show that on average 53% of the total milk produced nationally is sold in the market (Table 3). Rural Indian households retain the remaining 47% for domestic use. The scale of marketed milk surplus increases with farm size (Table 3). The quantity of milk marketed annually by an agricultural household ranges from 12 litres in Chhattisgarh to 1,052 litres in Gujarat (Table 4). Marketed milk as a percentage of milk production also reveals striking interstate differences, ranging from about 19% in Himachal Pradesh to 87% in Kerala.

Such differences arise from differentials in aggregate production across states, wherein states with higher production are able to supply greater quantities of marketed milk. Another reason is the disparities in food consumption pattern across states, wherein largely vegetarian states like Bihar, Haryana, Himachal Pradesh, Punjab, Rajasthan and Uttar Pradesh depend more on milk for meeting the protein needs of the household members, thereby limiting milk that is sold outside. Another reason is the relative status of development of milk value chains, as in the case of Gujarat, wherein dairy producers are better integrated with the milk-marketing system.

Choice of Milk-marketing Outlets

Milk marketing in India has largely remained an unorganised sector. Milk is marketed through the fragmented supply chains inclusive of the local milk vendors, wholesalers, retailers. The existing structure of milk marketing is an outcome of India’s demographics, traditions, infrastructural developments, socio-economic empowerment of milk producers, and institutional initiatives. Several types of milk-marketing chains presently operate in India. The dairies under the public sector, cooperatives, and organised private multinationals constitute the formal milk-marketing chains, and the remaining constitute the traditional or informal milk-marketing chains (Figure 1). Though there may not be uniformity in the estimates of marketed surplus milk, there is a consensus on the predominance of the informal sector in the Indian milk market.

The earlier estimates reveal that about 75% to 80% of the marketed surplus milk passes through the informal or traditional channels, and the remaining 20% to 25% is handled by formal or modern milk-marketing channels (Staal et al 2006; Kumar and Staal 2010; Kumar et al 2011). The overwhelming dominance of the informal sector can be attributed to several factors: (i) consumers’ preference for fresh milk; (ii) the perceived credibility of vendors to supply better quality milk; (iii) consumers’ unwillingness to pay an additional price for pasteurisation and packaging; (iv) uncertainties in the quantum of surplus milk; (v) the lack of alternative milk-marketing channels; (vi) the informal nature of business transactions; and (vii) the long-term relationship between vendors and milk producers. The much-accredited cooperative dairy development strategies in India are yet to significantly affect the dominance of informal milk-marketing agents, and most of the milk sales are still made through informal milk-marketing channels. However, recent data reveals an increasing role of the formal milk-marketing outlets in India, which is reflected in the structure of milk marketing.

Table 5 (p 61) shows that the cooperatives accounted for 34% of the marketed milk, with 26% of the milk-producing households selling milk to them. Local traders continue to be the dominant players among milk-marketing outlets, with 37% of the milk-selling households selling to them. About 38% of the marketed surplus milk was purchased by local traders. In fact, they are the preferred milk-marketing outlet for farms of all sizes. The third-most preferred milk-marketing outlet is direct sales to consumers. About 30% of the milk producers sell directly to consumers, and direct sales to consumers account for about 20% of the total marketed surplus. Finally, commission agents and processors play a smaller role in milk marketing in India, buying about 5% and 2% of the total marketed surplus, respectively.

Table 6 (p 61) shows state-specific data on milk-marketing outlets and broadly reflects the trends demonstrated in Table 5. There is a wide variation in the role of cooperatives as milk-marketing outlets across states. These cooperatives procure about 86% of the marketed surplus in Gujarat and 88% in Karnataka. The share of milk marketed through cooperatives is also significant in Andhra Pradesh (19%), Bihar (22%), Maharashtra (35%), Rajasthan (22%), Tamil Nadu (22%), and Uttarakhand (27%). Kumar et al (2013) found a significant variation in the performance of dairy cooperatives across states in India. About 70% of the total milk procurement is from the states of Gujarat (35%), Maharashtra (13%), Karnataka (13%), and Tamil Nadu (9%). However, the share of these four states in the total milk production in the country is only about 24%.

Overall, despite the rapid growth in the dairy sector and the increasing role of modern milk-marketing channels, the structure of milk-marketing outlets is largely informal. However, with increasing household incomes, rapid urbanisation, changing consumer preferences for processed foods, and the growing awareness among consumers for food safety and quality (Rajendran and Mohanty 2004; Singh and Datta 2010), India will see significant changes in its milk-marketing outlets. The structural transformation in the dairy value chain3 is largely driven by changes in demand patterns from traditional products to value-added and packaged products that are of better quality, and are produced more hygienically (Birthal and Negi 2012). The aforementioned drivers of milk production provide investment opportunities4 in technology and innovation in milk-marketing outlets from both private and international companies. The strong growth prospects of the dairy sector, driven by formalisation of the milk markets and growth of value-added dairy products, will encourage these investments.

The choice of a milk-marketing outlet is influenced by several factors—most importantly the price received by producers. Table 7 shows that the average price received by milk producers has no correlation with farm size. Furthermore, the prices offered by various milk-marketing outlets exhibit considerable variation. Contrary to general perception, the unit price received by milk producers from cooperatives is the lowest (₹ 26.50/litre) compared to the prices received from local traders (₹ 36.60/litre), processors (₹ 36.10/litre), and direct-to-consumer outlets (₹ 35.30/litre). This variation partly explains the continued dominance of informal milk-marketing channels, despite the emergence of several modern alternatives.

The state-specific data on prices received for milk reflects the national trend. With the exception of a few states, cooperatives pay a lower price for raw milk than local traders and consumers (Table 8, p 62). In Andhra Pradesh, Gujarat, Kerala, Rajasthan, and West Bengal, processors offer a higher price than other milk-marketing outlets. In Punjab, processors pay the highest price, and in Haryana and Uttar Pradesh, cooperatives offer the highest price. In these states, smallholders prefer to sell their output to cooperatives because they also provide livestock healthcare services, which are used by about 49% of farmers. The cooperatives also act as a source of breeding material and related services for the rearing of dairy animals. The role of cooperatives and other input dealers in providing additional inputs is limited, as most farmers source inputs such as seed, green and dry fodder, concentrates, and other feed from their own farm or buy from local traders.

Data and Methodology

The data for this study are taken from a nationally representative survey of agricultural households, conducted by the National Sample Survey Office (NSSO) in 2013. The survey was conducted in 4,529 villages spread over 566 districts in the country—about eight villages per district and eight farm households per village. The survey collected information on indicators of landholding, land use patterns, types of crop production and animal farming activities, seasonal variation in household farm activities, and ownership of livestock. The survey also collected information on social, economic, institutional, and organisational aspects of farming from a sample of 35,200 agricultural households, and on the production and disposal patterns of livestock products like milk. To get a perspective on developments in the dairy industry over time, we have also used data from the 2003 “Situation Assessment Survey of Farmers,” by the NSSO (GoI 2005).


The choice of a milk-marketing channel can be either supplier or producer-driven (Vandeplas et al 2013). The choice depends on a variety of factors, including ease of doing business, transaction costs and marketing margins, compliance with food safety measures, and prices received from different milk-marketing outlets, as well as a number of social and economic factors. There are six major milk-marketing outlets in India: (I1) direct-to-consumer outlets; (I2) commission agents; (I3) cooperatives and government agencies; (I4) processors; (I5) “others” (includes local shopkeepers and tea shops); and (I6) local traders. Our interest in this study is in determining the factors that affect smallholders’ choice of milk-marketing outlets and in examining the impact of farmers’ access to formal credit on their choice of milk-marketing outlets.

These milk-marketing outlets can be considered independent from each other and cannot be ordered in any logical way; therefore, the authors have used a multinomial logit (MNL) framework to model the smallholders’ decision on the choice of milk-marketing outlets. The probability that farmer i chooses milk-marketing outlet j is given by:

Prob (Yi = j|xi)=for

j=1,2,3,4,5,6 and ß0 = 0 ... (1)

x is a vector of explanatory variables associated with the farmers’ choice of milk-marketing outlets. ßs are parameters to be estimated. Thus, the log-odds ratio of farmer i selecting milk-marketing channel j is given by:

In ... (2)

where k = 0 refers to the base category.5 The odd ratios are calculated from the fitted MNL model by exponentiation of the estimated slope coefficients (Greene 2008). The interpretation of ßj is simplified even further by computing the marginal effects of xi´ on the probabilities of being in I1, I2, I3, I4, and I5 (Greene 2008):

... (3)

where is a vector whose elements are the averages of all estimated Pj. The sign of any particular and need not be the same.6 Although by definition which is done for facilitating the computation, the marginal effects of the attributes on the probability of a smallholder household choosing milk-marketing outlet category I6 are themselves not zero, and in fact are computed as

The explanatory variables used in the empirical model fall under seven broad categories: (i) physical capital of producers; (ii) human capital of producers; (iii) family and personal attributes of producer households; (iv) social hierarchy; (v) institutional support; (vi) income sources; and (vii) prices. Physical capital includes land and herd size. In the absence of data on herd size, marketed surplus of milk is used as a proxy variable. The hypothesis is that a higher marketed surplus of milk would lead to an increased adoption of modern milk-marketing channels. Human capital is proxied by the educational attainment of a farmer. A higher level of education enhances their capability for better management and, thus, makes them more likely to adopt modern marketing practices (Marenya and Barret 2006) and select better-paying marketing channels (Gong et al 2007).

The authors have used four dummy variables to represent farmers’ educational attainment—primary, middle, higher secondary, and above higher secondary level, with illiteracy as the base category. Farmers’ training in dairying and animal husbandry is represented by a dummy variable. Three variables were included to capture the influence of personal and household characteristics—the age of a household head, household size, and gender of a household head. The assumption is that young farmers would be better oriented to use organised milk-marketing channels like processors and cooperatives. Additionally, younger farmers are more in tune with modern needs and preferences of consumers.

In rural India, participation of females in dairying is higher than that of males in terms of labour contribution. Yet, their access to organised milk-marketing channels is limited due to social norms and traditions. To assess this gap in access, the gender of farmers was included as an explanatory variable. The household size, which represents potential labour supply, was another explanatory variable. Dairy is a labour-intensive enterprise, and family labour usually has lower opportunity costs; hence, the household members are likely to be employed at the farm.

Additionally, the social identity of a farmer in rural India has considerable influence on access to economic resources and thereby on outcomes. Certain groups like Scheduled Castes (SCs) and Scheduled Tribes (STs)7 are historically considered discriminated against in terms of social equality, ownership of resources, and access to modern technologies and capital. However, available evidence on the role of the caste structure in determining the benefits accrued from the dairy sector, particularly under milk cooperatives, is ambiguous. Therefore, to capture the effect of caste affiliation of dairy producers, three dummy variables were used: variables for STs, Other Backward Classes (OBCs),8 and the general caste, with SC as the base category.

To assess the effect of institutional support mechanisms such as access to institutional credit, as well as government-sponsored schemes like rural employment guarantee programmes (the Mahatma Gandhi National Rural Employment Guarantee Act [MGNREGA]),9 and public distribution systems (PDSs), three corresponding dummy variables were included as explanatory variables in the empirical model. For availing the facility of PDS, the possession of a below poverty line (BPL) card that ensures access to highly subsidised foodgrains was used as a dummy variable. It is assumed that access to government-sponsored programmes may discourage households from selling their milk to modern milk-marketing channels. Besides, other explanatory variables included social and economic variables, prices received by producers from different milk-marketing outlets, and the share of livestock income in the total income of a household. It is further assumed that farmers would prefer milk-marketing channels that offer them the best price. Similarly, farmers with a higher share of income from livestock, which may show specialisation in livestock production and higher milk production, may prefer to choose modern milk-marketing outlets.

Results and Discussion

Table 9 (p 63) presents the results from the MNL model, estimated by a maximum likelihood estimation method. The model was found to be significant at 1% level, and demonstrated a good predictive capability as indicated by a McFadden pseudo-R2 value of 0.05. Column 2 of Table 9 presents the results of the model, depicting the likelihood of selling surplus milk through direct-to-consumer outlets relative to selling surplus milk through local traders. The findings are consistent with expectations based on theoretical grounds and findings of previous studies on the choice of a marketing channel. Variable size of the operation, access to institutional credit, classification under the OBC category, and the share of livestock income have a negative and significant impact on the choice of direct-to-consumer milk-marketing outlet compared to sales through local traders. On the other hand, training in animal husbandry for smallholders, age of the operator, and the price of milk received are found to have positive and significant impact on the choice of direct-to-consumer milk-marketing outlet.

Educational attainment (primary level and above) and the household size of smallholders have a negatively significant effect on the choice of commission agents as a milk-marketing outlet compared to selling milk through local traders (Table 9, Column 4). However, the price of milk, size of operation, access to institutional credit, and participation in MGNREGA have a positive and significant impact on the choice of commission agents as a milk-marketing outlet.

Column 6 of Table 9 reveals the likelihood of choosing cooperatives and government agencies to sell surplus milk compared to selling milk through local traders. In this case, the results are significantly different than results for other milk-marketing outlets. For example, most variables are found to be significant at the 1% level of significance. The results reveal that milk price, educational attainment, access to institutional credit, participation in PDS, and affiliation to social groups and castes other than SC have a significant impact on the choice of cooperative and government agency as a milk-marketing outlet compared to the local traders.

Column 8 of Table 9 presents the likelihood of choosing processors over local traders for selling surplus milk. Again, in this case the size of operation, gender of the operator, access to institutional credit, and farmers’ caste and other group affiliations (STs, OBCs, and others) have a positive and significant effect on the choice of processors as milk-marketing outlets compared to selling milk through local traders. However, factors such as the attainment of secondary education, age of the operator, and milk price have negative and significant effect on the choice of processors as a milk-marketing outlet. Finally, Column 6 of Table 9 reveals that there are only a few factors affecting the choice of “others” as milk-marketing outlets compared to local traders, where “others” includes local tea shops and shopkeepers. Findings in Table 9 reveal that educated farmers and farmers with access to institutional credit are less likely to choose “others” as milk-marketing outlets. Additionally, the size of the operation, herd size, training in animal husbandry, classification under the ST category, and farmers’ participation in MGNREGA have a positive and significant impact on the choice of “others” over local traders.

Table 10 (p 65) presents predicted marginal changes in the probability of choice of a milk-marketing outlet due to a per-unit increase in continuous explanatory variables and a change in value from 1 to 0 for dummy variables. The results show that a 10% increase in farm size decreases the probability of using direct-to-consumer milk-marketing outlets by 0.15% and increases the probability of using commission agents, processors, and “other” milk-marketing outlets by 0.05%, 0.016%, and 0.026%, respectively. Compared to local traders, smallholders would prefer selling their milk through direct-to-consumer milk-marketing outlets, indicating that larger dairy farmers prefer selling their milk through commission agents, processors, and “others.” The bottom part of the table shows that, compared to local traders, a 10% increase in the price of milk increases the probability of selling milk through direct-to-consumer outlets and cooperatives, and government agencies by 1.6% and 1.4%, respectively. On the other hand, a 10% increase in the price of milk decreases the probability of selling milk (about 0.11%) through processors and “others.” The above findings indicate that cooperatives still function as a major anchoring point in dairy development in India and are vital for the transformation of the dairy industry as a venture for providing better livelihood options for farmers.

A positive and significant coefficient with respect to the age of a farmer indicates that with age smallholders are more likely to sell their milk through direct-to-consumer outlets and cooperatives. An additional year increases the likelihood of selling to direct-to-consumer outlets and cooperatives by 0.03% and 0.06%, respectively, compared to selling milk through local traders. Alternatively, an additional year decreases the likelihood of selling to commission agents and processors by 0.02% and 0.01%, respectively.

The household size, implying household labour supply, has a significantly negative impact on milk marketing through cooperatives and government agencies—10 additional members reduce the probability by 0.44%. This finding implies that, with larger labour resources, rather than sell to cooperatives, farmers would be able to either process milk, sell it, or search for and bargain with other milk-marketing channels and thereby increase their profits. As noted earlier, across all channels, cooperatives offer the lowest price for milk, yet constitute the most inclusive mode of milk marketing (Kumar et al 2011).

The analysis provides interesting results regarding the human capital of smallholder household heads. On the one hand, marginal effects in Table 10 reveal that educated farmers are less likely to sell their milk to commission agents compared to uneducated smallholder household heads, the probability of selling milk decreases between 1% and 3%. On the other hand, it is observed that the educated smallholder household heads are more likely to sell their milk to cooperatives and government agencies; the effect is between 3.5% for those who have finished secondary school or above, to 7.5% for those who have completed primary education, thereby indicating a “pro-poor” bias of cooperatives. The results reconfirm that cooperatives are more inclusive than the other three alternative milk-marketing channels. It is also seen that the probability of educated smallholders selling their output to processors decreases by 0.8% and to “others” by 0.7% (Table 10).

It was assumed that the caste of a farmer, an important social identity in India, would be a barrier for underprivileged producers in accessing organised milk-marketing channels. Compared to farmers belonging to the SC, farmers belonging to the ST, OBC, and general castes are more likely to use cooperatives and government agencies as milk-marketing channels at rates of about 20%, 10%, and 4.6%, respectively. These findings may provide evidence of the inclusive characteristic of milk cooperatives. Additionally, farmers belonging to the OBC (0.5%) and general castes (1.1%) are more likely to use processors as milk-marketing channels.

The results in Table 10 reveal that farmers with access to institutional credit are less likely to sell their milk through direct-to-consumer outlets (11%) and “others” (0.5%). However, the findings show that farmers with access to institutional credit are more likely to sell their output through processors (2%) and cooperatives and government agencies (6.5%). The farmers with formal credit-repayment commitments would plausibly seek a stable but lower income in order to pay off the credit, and milk-marketing channels like processors and cooperatives provide them with commitments to buy milk and, therefore, with stable incomes.

The estimates show that the farmers who participated in MGNREGA are more likely to use direct-to-consumer outlets (3.5%), commission agents (1.4%), and “others” (1.1%) as milk-marketing channels and are 3.6% less likely to use cooperatives and government agencies. The farmers with an Antyodaya10 or BPL ration card are less likely to use direct-to-consumer milk-marketing outlets (5.9%), but more likely to use cooperatives and government agencies (6.4%) as milk-marketing outlets. This finding is consistent with Vandeplas et al (2013), who found that the possession of a BLP ration card increases the chances of dairy farmers in Punjab joining cooperatives. Finally, a higher contribution of livestock income in the total household income increases the likelihood of using commission agents (2.6%) and processors (0.5%) as milk-marketing channels, and decreases the likelihood of using direct-to-consumer outlets (7%) as milk-marketing channels.


The study shows that the dominance of smallholders in milk production in India is increasing. The increase in smallholders’ share in milk output is a result of an increase in their number as well as per-household milk production. Dairying is, therefore, emerging as a livelihood option for smallholders, notably in states such as Maharashtra, where farm failure is often reported.

The average milk production in a dairy farming household is found to be directly proportional to the farm size associated with it. To be more specific, the average milk production per annum of a marginal farm is about 40% lower than a small farm, about half of that of medium farm and about a third of a large farm household. Notwithstanding this, the marginal and small farms together contributed about 70% of the total milk production in the country in 2013, thereby underlining their prominence in India’s milk basket. At an aggregate level, the share of marketed surplus of milk is about 53% of the total production. Interestingly, the share of marketed surplus is higher for smallholders than large farms. This finding is significant because it points to an increase in smallholders’ choice of dairying as a livelihood strategy, though possibly a considerable part of it is distress sale.

This study also investigates whether there is any systemic bias against the participation of resource-poor smallholders in organised milk-marketing outlets. Based on the land size, participation in cooperatives is size neutral and that large holders favour selling their surplus milk to processors, “others,” and commission agents. Considering the higher marketed surplus of milk by smallholders, this result can be viewed as consistent with the interests of small farmers. In addition, some broad generalisations can be drawn regarding the choice of a milk-marketing outlet.

The likelihood of choosing cooperatives and government agencies as an outlet is positively affected by (i) the age of a household head; (ii) educational attainment of a household head; (iii) caste or group affiliation (ST or OBC); (iv) whether a household is headed by a female; (v) access to institutional credit; (vi) training in animal husbandry; (vii) price of milk; and (viii) access to the PDS such as the BPL card. The exposure gained through better education and experience in dairying enhances farmers’ inclination towards participation in cooperatives as the educated and experienced farmers are in a better position to select marketing chains.

Another important finding from the study is that farmers with access to institutional credit are more likely to sell their output through formal milk-marketing channels like processors and cooperatives and government agencies. This may be attributed to the fact that famers having access to institutional credit can wait for payments,11 and they can also avoid advance payment by the milk traders.12

Overall, the evidence reveals that cooperatives are more inclusive and largely transcend the boundaries of caste and land size. This conclusion is strengthened by the fact that a unit increase in the price of milk, at the mean level, would enhance the probability of farmers choosing direct-to-consumer outlets and cooperatives and government agencies as milk-marketing outlets. Besides highlighting the importance of cooperatives, this paper raises important questions for further inquiry about why organised private entities (commission agents and processors) other than cooperatives have not emerged as preferred milk-marketing channels. It is probable that these milk-marketing channels need to provide incentives to attract smallholders, women, and underprivileged sections of the rural Indian society.


1 In India, largely crop production and dairy operations are integrated. Therefore, we use land area to define marginal, small, medium, and large farms. More than 67% of dairy animals are owned by marginal and small farmers, which constitute the core of the milk-production sector.

2 Smallholders include marginal and small farms.

3 Cheese, yogurt, ice cream, infant foods, and sweets will largely drive growth. A growing young population is leading to the increased consumption of branded, value-added dairy products. 

4 Investments towards developing a milk procurement infrastructure are needed to source superior-quality raw milk. This process entails setting up collection centers and milk chillers at the village level to directly engage with dairy farmers. This will also help milk processors secure greater control of milk sourcing.

5 Local milk traders represent the reference category.

6 This method of measuring marginal changes in probability due to a unit increase in the explanatory variable is appropriate when the variable is continuous. For a Kthdummy variable in the Jth milk-marketing outlet category, the marginal change in the probability for the ith household is computed alternatively as (see Greene 2008):

7 SCs and STs are backward, uneducated, poor, and officially regarded as socially disadvantaged people in India.

8 Other Backward Class (OBC) is a collective term used by the Government of India to classify castes that are socially and educationally disadvantaged. It is one of several official classifications of the population of India, along with SCs and STs.

9 Under the MGNREGA, the government has assured guaranteed wage employment to households BPL, for a minimum period of 100 days in a year, thus providing livelihood security to needy and resource-poor households in rural India.

10 Households below the poverty line are classified as BPL households and have BPL ration cards. The poorest among BPL families have Antyodaya cards.

11 Cooperatives generally make payment fortnightly or monthly.

12 Milk traders generally make payment in advance to ensure delivery of milk by dairy farmers.


Birthal, P S, D S Negi, A K Jha and D Singh (2014): “Income Sources of Farm Households in India: Determinants, Distributional Consequences and Policy Implications,” Agricultural Economics Research Review, Vol 27, No 1, pp 37–48.

Birthal, P S and D S Negi (2012): “Livestock for Higher, Sustainable and Inclusive Agricultural Growth,” Economic & Political Weekly, Vol 47, Nos 26 & 27, pp 89–99.

Gong, W, K Parton, R X Cox and Z Zhou (2007): “Transaction Costs and Cattle Farmers’ Choice of Marketing Channels in China: A Tobit Analysis,” Management Resource News, Vol 30, No 1, pp 47–56.

GoI (2005): “Situation Assessment Survey of Farmers: Some Aspects of Farming: NSS 59th Round,” National Sample Survey Office, Ministry of Statistics and Programme Implementation, Government of India.

— (2013): “Key Indicators of Consumers Expenditure in India: NSS 68th Round,” National Sample Survey Office, Ministry of Statistics and Programme Implementation, Government of India.

— (2014): “Situation Assessment Survey of Agricultural Households In India: NSS 70th Round,” National Sample Survey Office, Ministry of Statistics and Programme Implementation, Government of India.

— (2015): “Basic Animal Husbandry & Fisheries Statistics 2014,” Department of Animal Husbandry, Dairying and Fisheries, Ministry of Agriculture, Government of India.

— (2016): Economic Survey 2015–16, Ministry of Finance, Government of India.

Greene, W H (2008): Econometric Analysis, sixth ed, New York, US: Prentice Hall.

Kumar, A and S J Staal (2010): “Is Traditional Milk Marketing and Processing Viable and Efficient? Empirical Evidence from Assam, India,” Quarterly Journal of International Agriculture,Vol 49, No 3, pp 213–25.

Kumar, A, S J Staal and D K Singh (2011):Smallholder Dairy Farmers’ Access to Modern Milk Marketing Supply Chain in India,” Agricultural Economics Research Review, Vol 24, pp 243–53.

Kumar, A, S Parappurathu and P K Joshi (2013): “Structural Transformation in Milk Production in India,” Agricultural Economics Research Review, Vol 26, pp 209–21.

Kumar, A and S Parappurathu (2014): “Economics of Dairy Farming and Marketing: Micro-level Perspectives from Three Major Milk Producing States of India,” Indian Journal of Animal Sciences, Vol 84, No 2, pp 204–09.

Kumar, A, P K Joshi, P Kumar and S Parappurathu (2014):Trends in the Consumption of Milk and Milk Products in India: Implications for Self-sufficiency in Milk Production,” Food Security, Vol 6, pp 719–26.

Marenya, P P and B Barret (2006): “Household-level Determinants of Adoption of Improved Natural Resources Management Practices among Smallholder Farmers in Western Kenya,” Food Policy, Vol 32, No 4, pp 515–36.

Rajendran, K and S Mohanty (2004): “Dairy Co-Operatives and Milk Marketing in India: Constraints and Opportunities,” Journal of Food Distribution Research, Vol 35, No 2, pp 34–41.

Singh, S R and K K Datta (2010): “Understanding Value Addition in Indian Milk Sector: Some Perspectives,” Agricultural Economics Research Review, Vol 23, pp 487–93.

Staal, S J, I Baltenweck, L Njoroge, B R Patil, M N M Ibrahim and E Kariuki (2006): “Smallholder Dairy Farmer Access to Alternative Milk Market Channels in Gujarat,” Contributed Paper, 2006 International Association of Agricultural Economists Conference, Brisbane, Australia.

Vandeplas, A, B Minten and J Swinnen (2013): “Multinationals vs Cooperatives: The Income and Efficiency Effects of Supply Chain Governance in India,” Journal of Agricultural Economics, Vol 64, No 1, pp 217–44.

Updated On : 2nd Jan, 2019


(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top