World Bank Reprint Series: Number 301 Gershon Feder and Roger Slade Contact Farmer Selection and Extension Visits The Tralning and Visit Extension System in Haryana, India Reprinted with permission from Quarterly Jornltll of Inleri17tiiiial Agriculture, vol. 23, no. I (January/March 1984), pp. 6-21. World Bank Reprints No. 260. Pradeep K. Mitra, 'A Theory of Interlinked Rural Transactions," Journal of Plublic Ecotnomlics No. 261. David L. Lindauer and Richard H. Sabot, "The Public/Private Wage Differential in a Poor Urban Economy," Journal of Developmient Eco.zonolms No. 262. J. B. Knight and R. H. Sabot, "Labor Market Discrimination in a Poor Urban Economy," Journal of Developmienf Studies No. 263. Carl Dahlman and Larry Westphal, "Technical Effort in Industrial Development: An Interpretative Survey of Recent Research," The Econom.ics of Neuw Technologyi in Developing Coutn tries No. 264. Michael Bamberger, "The Role of Self-Help Housing in Low-Cost Shelter Programs for the Third World," Built Environment No. 265. Bela Balassa, "The Adjustment Experience of Developing Economies after 1973," IMF Conditionality No. 266. Bela Balassa, "Outward Orientation and Exchange Rate Policy in Developing Countries: The Turkish Experience," The Middle East Journrd No. 267. Dipak Mazumdar, "Segmented Labor Markets in LDCs," Amteriican Economlic Rezviewo No. 268. Stephen P. Heyneman and William A. Loxley, "The Effect of Primary-School Quality on Academic Achievement across Twenty-nine High- and Low-Income Countries," The American Journial of Sociology; No. 269. James R. Follain, Jr., Gill-Chin Lim, and Bertrand Renaud, "Housing Crowding in Developing Countries and Willingness to Pay for Additional Space: The Case of Korea," Jotrnmal of Devielopmitenit Econoomics No. 270. Bela Balassa, "Policy Responses to Extemal Shocks in Sub-Saharan African Countries," Joutnial of Policy Modeling No. 271. Jaime de Melo and Sherman Robinson, "Trade Adjustment Policies and Income Distribution in Three Archetype Developing Economies," Journal of Developmtellt Economnics No. 272. J. B. Knight and R. H. Sabot, "The Role of the Firm in Wage Determination: An African Case Study," Oxford Economiic Papers No. 273. William G. Tyler, "The Anti-Export Bias in Commercial Policies and Export Perfor- mance: Some Evidence from Recent Brazilian Experience," Weltwirtschaftliches Archii No. 274. Ron Duncan and Emst Lutz, "Penetration of Industrial Country Markets by Agricul- tural Products from Developing Countries," World Developnment No. 275, Malcolm D. Bale, "Food Prospects in the Developing Countries: A Qualified Optimistic View," Thle American Econoinic Rev7iew (with Ronald C. Duncan) and "World Agricultural Trade and Food Security: Emerging Patterns and Policy Directions," Wisconisini International Law Journall (with V. Roy Southworth) No. 76. Sweder van Wiinbergen, "Interest Rate Management in LDCs," Journal of Monetarty E-ccionomics No. 277. Oli Havrylyshyn and Iradj Alikhani, "Is There Cause for Export Optmism? An Inquiry into the Existence of a Second Generation of Successful Exporters," lJ .ltzaJfliches Archlit) No. 278. Oli Havrylyshyn and Martin Wolf, "Recent Trends in Trade among Developing Countries," Eutropean Ecotionoic Review No. 279. Nancy Birdsall, "Fertility and Economic Change in Eighteenth and Nineteenth Century Europe: A Comment," Pouiilbtion anid DU'v'loptlieot Rezview Quarterly Journal of International Agriculture, Volume 23, No. 1, January -March, pp. 6 - 21 Contact Farmer Selection and Extension Visits: The Training and Visit Extension System in Haryana, India by Gershon Feder and Roger Slade * Introduction With dwindling unexploited land and water resources in many less developed coun- tries, there is a growing realization that the major potential for increasing agricultural output and improving the lot of the rural population lies in increasing the productivity of existing land and water endowments. A significant gap exists in many countries between known and feasible agricultural technologies, and what is actually practiced by the majority of farmers. Attempts to close this gap have included programs that focus primarily on the delivery of physical inputs - fertilizer, water, seeds - as well as those aimed at improving farming practices. It has been asserted by some that even when equipment and material-dependent techniques are excluded, there are still sub- stantial gains totbe realized if cultivation practices are improved in accordance with the results of well directed agronomic research tailored to specific agro-climatic and loc31 situations. A major channel for the dissemination of information on known feasible technologies is the agricultural extension system. Extension agents are an important link between agricultural research institutions and the ultimate clients of agricultural know-how, the farmers. Accordingly, improving the efficiency and quality of extension services is con- sidered by many to be an indispensable element in any agricultural development strategy (4, 6, 1 2). The present paper deals with some aspects of the interaction between extension agents and farmers within a specific system of extension, commonly known as the Training and Visit (T & V) method. This system has been adopted in many less developed coun- tries, often with the support of the World Bank, and it incorporates several features designed to enhance the effectiveness of the extension service. One particular aspect of this system is the selection of individual farmers to act as regular contact men with ex- tension workers. The analysis in this paper focuses on the selection of these farmers and on the extent of their interaction with extension agents. The paper is organized as follows. The next two sections describe the T & V system and the area of the study respectively. An analysis of the characteristics and selection of contact farmers and their exposure to extension is follGwed by a discussion of the overall coverage achieved by extension agents through their visics. The last section pro- vides a summary and some conclusions. 6 1 The Training and Visit Extension Systo-mi The T & V extension system is designed to overcome some of the major problems afflicting earlier forms of extension organization. Its structure and objectives, which are fully elaborated in several published sources (1, 2, 3, 6), are briefly summarized below. Under the T & V system the extension organization deals exclusively with extension work, eschewing all other activities (for example, direct involvement with agricultural research or input supply). Personnel are organized in a unique line of command, such that each officer at each hierarchical level has a sufficiently small number of staff under him to allow him to provide effective and personal guidance, supervision and training. At the bottom of the hierarchy are village level extension agents, who cover areas typically containing 700 - 800 farming families, divided into about eight groups' In each group, about 10 percent of the farmers are selected as "contact farmers" and their names are recorded at local extension headquarters. The extension agent visits each of the eight farmers' groups once every two weeks, on a prespecified and fixed date. These visits mostly take place in the fields of contact farmers, but other farmers are expected to participate. Obviously, since non-contact farmers can freely join in extension agent visits, they can become de-facto contact farmers if they persist in attending visits in a relatively re- gular manner, and if they experiment with the practices recommended by the ex- tension worker. Over time the composition of the formal contact farmer group may be changed, and it is likely that farmers who have shown interest in extension visits will actually become formal contact farmers while others will be dropped. During the visit, the agent passes on information regarding improved practices which are relevant at the particular stage in the season. Occasionally, the extension worker may organize a large group meeting, but as wiil be shown later, in the area covered by the present study, most farmer-extension contacts take place in small groups or individually. The fixed schedule of visits is advantageous both for the effective supervision of agents' activities by their superiors and for maintaining farmers' interest and confidence in the exten- sion agent. In order to have a significant impact, the T & V system, when operating properly, focuses advice on the main crops grown and on the most important practices. Simple, non-cash requiring practices are promoted first, so that all classes of farmers are able to benefit. Contact farmers are expected to adopt recommended practices (or at least try them out) and transmit them to other farmers. Accordingly, the selection of con- tact farmers should not be arbitrary, as they are an important link between the primary source of knowledge and the farming community as a whole. 2. The Study Area Haryana is part of one of the relatively more advanced agricultural regions in India and benefitted substantially from the Green Revolution of the late sixties and seventies. The Training and Visit extension system was introduced in Haryana in 1979 and by 7 August 1981 (two months prior to our survey) about 95 percent of village level ex- tension agents and 80 percent of subject-matter specialists were in position. However, about 25 percent of the positions for officers directly in charge of supervising and guiding village level agents had been vacant since project inception, thus diminishing the effectiveness of supervision and the quality of the extensive lower level coverage. Additionally, District Technical Committees, Zonal Workshops and the State Tech- nical Committee had either not been convened or had not functioned properly over the same period. These institutions were expected to play an important role in gene- rating locally-specific and timely recommendations and useful adaptive research work. These problems continue to hinder the development of the reorganized extension systern. The two districts covered by the study are geographically adjacent, but there are some agro-climatic differences between them. Jind district, with average annual rainfall of 592 mm, is dryer than Karnal district (803 mm), and has less irrigation facilities (58 percent of net cropped area irrigated compared with 74 percent in Karnal). Karnal is somewhat larger in area (3,766 square km versus 3,224 for Jind) and has a larger po- pulation (about one million compared to 750,000 in Jind). The differences in popula- tion density are not substantial (267 persons per square km in Karnal versus 233 in Jind). The present analysis is based on data from a recent sample survey of 972 farmers in the two districts. Random samples of contact and non-contact farmers were selected. While contact farmers constitute approximately 10 percent of the farming community, samples of approximately equal size were chosen (430 contact farmers and 542 non- contact farmers) in order to generate reliable statistics for each group. 3. Characteristics of Contact Farmers As stated in Section 1, one of the basic principles of the T & V extension method is the reliance on the "Two-Step-Flow" of communications (11, p. 219) whereby new ideas and innovations spread first to opinion leaders, and from them to their followers. Contact farmers are expected to act as opinion leaders. They receive new and timel'1 agricultural information through the regular two-weekly visits of extension agents. This information is then communicated to other farmers, either through the initiative of the contact farrners, or in response to inquiries from other farmers who may have observed the new practices being used by contact farmers. The T & V method stresses that extension agents should not exclude non-contact farmers from direct interaction with them. Such farmers should be made aware of the extension agents' regular visits to the group (or the village) and of their freedom to attend. They should be encouraged to attend visits by extension agents in contact farmers' fields, and their queries should be answered (Benor, personal communica- tion). Indeed (see section 5), non-contact farmers do report significant direct exposure to extension'. Contact farmers receive a continuous and regular flow of information from extension agents, and it follows that the way in which they are selected is a matter of considerable importance. While their potential for opinion leadership on matters of crop husbandry 8 is the key criteric-, there are additional considerations which should be kept in mind. For instance: "Contact farmers must be willing to try out practices recommended by the extension workers and be prepared to have other farmers visit their fields. But they should not be the community's more progressive farmers who are usually regarded as exceptional and their neighbors tendi not to follow them. On the other hand, very weak farmers tend to be slow in adopting new methods. Furthermore, the contact farmers must be of good standing in their community so that their views on new practices will be respected by other farmers." (1 , pp. 13, 14). Evidence suggests that there is a considerable overlap between the attributes which characterize individuals who tend to be opinion leaders and the characteristics of inno- vators, or fast adopters (8, pp. 1 55 - 160). If, however, the contact farmers are mostly wealthy individuals with greater resources than other farmers, the diffusion of innova- tions may be hampered. Smaller and poorer farmers may not be convinced that the new practices are better, perceiving a better performance by contact farmers to be a result of their advantageous resource position. In practice, overcoming this problem may prove difficult. Traditionally, extension agents have tended to focus attention on the well-to-do farmers, because this ensured an immediate and visible impact and because wealthier farmers could offer them personal rewards (meals, accomodation, produce). This bias has been frequently noted in the literature (2, p. 8; 3, p. 233; 4, p. 7). Moreover, it has been argued that efficiency considerations require that scarce extension resources should be used to achieve maximum impact, implying a bias toward larger, and usually richer, farmers (13). Thus, there is a trade-off between choosing those farmers who will adopt innovations most speedily and those who are somewhat less suitable (from the point of view of potential for fast adoption), but whose resource position is more typical of the majority of farmers and hence, their behavior more readily imitated. A further complication is introduced by the fact that the extension agent is sometimes obliged to nominate influential individuals as contact farmers even if they are not suitable as communicators of innovations. This is prompted by the need to avoid hosti- lity from such powerful individuals, who may otherwise make extension work diffi- 3 cult to carry out The discussion above has highlighted some of the considerations involved in selecting contact farmers. It is not, therefore, surprising (see below) that in some respects con- tact farmers as a group are representative of the farming community as a whole, while in other ways they are significantly different4 . The area of land owned is an indicator of wealth, which, in turn, is related to both opinion leadership and the propensity to adopt innovations (8, p. 160). The data, as presented in Table 1, indicates clearly that contact farmers (who comprise approxima- tely 10 percent of the population) inctude a significantly higher proportion of larger land owners than the rest of the population>. Such a pattern has been observed else- where in India6, but as indicated above, this element in the set of trade-offs is complex and we have not attempted to determine whether the bias in favor of larger farmers observed in our sample is counter-productive. However, the proportion of very small farmers (less than 5 acres of owned land) in the contact group is not negligible. 9 Table 1. Distribution of Contact a,id Non-Contact Farmers by Size of Owned Holding Size Class Contdct Farmers Non-Contact Farmers Acres of Owned Land (N = 430i (N = 542) (%) (%J Less than 5.0 13 27 5.0 - 9.9 22 28 10.0- 19.9 36 31 20.0 plus 29 14 Another characteristic of contact farmers, which is of obvious relevance in the Indian context, is caste. Traditionally, leadership in Indian villages has been monopolized by higher castes (8, p. 73). If this pattern is repeated in the choice of contact farmers it may hamper communications between contact and non-contact farmers. Table 2 presents the caste composition of both groups of farmers, and shows that there is practically no difference between contact and non-contact farmers. Table 2: Caste Composition of Contact and Non-Contact Farmer Groups Contact Farmers Non-Contact Farmers Caste (N = 436) (N = 545) (.) (%) Brahmin 7 8 Rajput 6 6 Jat 58 55 Sikh 1 1 10 Harijan 1 2 Others 17 19 Additional characteristics of contaQt farmers are described in Table 3, which enume- rates several indicators relevant to farmers' socio-economic status and their potential as opinion leaders and fast adopters. It has been argued, for instance, that education is an important factor in explaining willingness to adopt innovations (7, p. 195). Our data confirm that the contact farmer group is characterized by higher education levels compared to the rest of the popula- tion. The difference is statistically significant and is observed even when contact and non-contact farmers are compared across size of holding 10 Table 3: Socio-Economic Characteristics of Contact and Non-Contact Farmers Contact Farmers (%) Non-Contact Farmers 1°) Criteria Smaller Larger Smaller Larger Farmersa Farmersb All Farmersa Farmersb All (N = 151) (N = 279) (N = 430) 1N - 298) 1N - 244) (N = 542) No formal education 70 56 61 74 69 72 Membership in village institutions 21 25 24 9 18 13 Participation in short-term agricultural training 7 12 10 5 3 4 aLess than 10 acres owned. bTen or more acres owned. Membership in village institutions is an indicator of status, and is hypothesized to characterize innovativeness (9, pp. 101 - 102). Status is clearly related to wealth, and it is not surprising that the percentage of contact farmers who are members of village institutions is almost double that of the rest of the population. This also holds across farm size classes, suggesting that status in the community is a selection criterion independent of wealth when the latter is measured by the area of land owned. Participation in a short-term agricultural training course (a service provided by the Haryana Agricultural University) is an additional indicator of agricultural knowledge and, therefore, the farmer's potential for opinion-leadership on agricultural matters. It may also be characteristic of innovativeness, suggesting that the farmer is willing to make extra efforts to gain new knowledge. The data shows that a significantly higher proportion of contact farmers, when compared with non-contact farmers, have under- taken such training. Access to reliable irrigation is an important factor affecting farmers' crop and tech- nology choices. In the two districts covered by this study, canal irrigation is available to most farmers. However, investment in tubewells is much more common in Karnal district. Access to tubewell irrigation enables better control of irrigation timing and tends to increase yields. Tubewells also reduce yield variability and thus diminish the risks faced by the farmer. Farmers with tubewells may, therefore, be more receptive to innovations and thus may be suitable for selection as contact farmers. However, if contact farmers are characterized by a pattern of tubewell ownership which is sub- stantially different from that of other farmers, their pedagogic role may be diminished. For instance, if only a small fraction of the population owns tubewells while most 1 1 contact farmers are tubewell owners, non-contact-farmers may see the achievements of contact farmers as an outcome of their favorable situation (rather than be im- pressed by the benefits of adopting improved practices) and the diffusion of innova- tions may be adversely affected. Table 4 shows the distribution of tubewell ownership in the sample. In both districts access to a tubewell is clearly related to wealth (approximated by land owned). In Jind district the selection of contact farmers is biased in favor of tubewell owners within each size class (the differences are statistically significant for each sub-sample). In Karnal district practically all large farmers have access to tubewells, and there is little difference between large contact and non-contact farmers. But the selection of small contact farmers is significantly biased in favor of those owning tubewells. Table 4: Ownership of Tubewells Contact Farmers Non-Contact Farmers District Farm Size Sample Size Owning Tubewella Sample Size Owning Tubewella Smallb 90 28 189 15 Jind Large 169 48 155 37 All 259 41 344 25 Small 61 92 109 72 Karnal Large 110 97 89 99 All 171 95 198 84 aOutright or shared ownership. bFarmers owning less than 10 acres are defined as small farmers In the discussion above, each socio-economic characteristic of contact farmers was considered independently, except for some classifications by farm size8 . While the cor- relations between the different variables are quite low (see apl-ndix), it is appropriate to consider their relevance to the selection of contact farmers within a framework where all elements enter simultaneously. Multiple regression is not an appropriate methodology in t'ie present case because the dependent variable is dichotomous (con- tact or non-contact), thus violating the normality assumption required for statistical test. We have, therefore, chosen to use logit analysis, which is well suited to problems involving qualitative response variables (10). The Logit model posits that the probabi- lity of a farmer being selected as a contact farmer is a function of the socio-economic explanatory variables with the following specific formulation: (1) P(C) - eB x,, 11 i eBx] 12 where P(C) = probability of being selected as contact farmer e = the exponential operator x = a vector of explanatory variables B = a vector of parameters The probabilities P(C) are not observable. The estimation procedure requires that all observations of contact farmers are assigned a probability of 1 , while the other obser- vations are assigned a probability of zero. The estimated parameters (B) are asympto- tically normally distributed, and tests of statistical significance can be performed. In order to avoid bias which may result from thie differences in irrigation infrastructure in the two districts covered by the study, separate estimates were obtained for each district and are reported in columns 2 and 3 of Table 5. The results corroborate the Table 5: Logit Analysis of Factors Affecting the Selection Probabilities of Contact Farmers Coefficient Jind Karnal Pooled Criteria District District Sample (2) 13) (4) 11, FarmSize .0275, .0339t 0317 131431) b (341) (5.154) '2! Education 'years) 0397t 0573' 453' 11 6371 (2.227) (2 5931 Dummy Variables 1 3X Ownership of tubewell 0570' 8987' .3956' 3.0161 (1.949) 12.250) 41 Membership in .3416 ,4374" 4090' village institutions (1.443) (1.5241 (2.260) 5) Previous participation 1.-042' 8262 .81 72, in Agric. training l2.180) 12.213) (2.851) Likelihood ratio statistic 60.55 44.96 99.23 No. of observations 599 361 960 aNo estimate of the constant is presented; the sample is not random with respect to the pro- portions of contact non-contact farmers, and the estimated constant is therefore biased. The other coefficients are not affected, however (McFadden, personal communication). bNumbers in parentheses are asymptotic "t" values. 'Significant at a 5 ( 'one tailedi level of signiffcance "Significant at a 7 'one tailedl level of significance 13 conclusions of the earlier discussion, as almost all the indicators which were pre- viously found to differ among contact and non-contact farmers have a significant co- efficient with the expected sign. The coefficients are quite similar across districts, with the exception of the dummy variable for ownership of a tubewell. The likelihood ratio statistic for each estimated equation has a x. distribution, and it can be shown that the estimated equation is a relevant model since the critical x?' value for a 5 per- cent test is 14.07. Given the similarity of coefficients across districts, the model was re-estimated utiliz- ing a pooled sample. The results (see column 4 in Table 5) indicate that all the va- riables are significant at the 95 percent confidence level. Parameter values are only slightly affected by the pooling of observations, except for the coefficient for tubewell ownership. The results confirm that variables such as tubewells and membership in village institutions which are usually related to wealth, play an independent role in affecting the selection of contact farmers. 4. Coverage of Extension Visits As explained in Section 2, urder the T & V extension system agents make a regular two-weekly ,;sit to each of the farmers' groups assigned to them. They visit the con- tact farmers, but other interested farmers attend as well. The day of the visit is fixed throughout the-year, and should be known to all farmers. The regularity of these visits is important for the effectiveness of the extension system. Farmers who are assured of continuous and regular contact with extension may be expected to demonstrate in- creased receptiveness to new practices. Uncertainties and doubts may be diminished by the prospect of future opportunities to discuss problems and queries with the agent. The continuity of interaction with extension also tends to strengthen farmers' confi- dence in the agent, provided, of course, that this advice is generally useful. If, however, visits are irregular and unpredictable, farmers may rapidly lose interest and will not make efforts to maintain their contact with extension. Visits to contact farmers are to a large extent initiated by the extension agent but they require an interest and a willingness to listen on the part of the farmer. The siutation is reversed in the case of non-contact farmers. Whilst the agent should make sure that knowledge about his vicits to the village is widespread, and he is expected to encourage other farmers to be present when he visits contact farmers, it is left mostly to the non- contact farmers to take the initiative and participate in the agent's visits to the fields of contact farmers, or to approach the agent with queries. The analysis in this section focuses on extension visits as reported oy both contact and non-contact farmers. The primary hypotheses are (i) that contazt farmers will report a much higher incidence of contact with extension and (ii) that the pattern of their contact will be more regular than that reported by non-contact farmers. Because ex- tension agents' visits are frequently observed to be biased in favor of larger farmers, it is also of interest to test the hypothesis (iii) that extension agents' visits to larger con- tact farmers are more frequent and more regular, 14 Table 6: Number of Visits to Farmers by Extension Agents During the Past Four Weeks Contact Farmers Non- Contact Farmers District Farm Size Sample No. of Visits Sample No. of Visits Size Size 0 1 2+ 0 1 2+ in percent in percent Jind Smalla 90 31 18 51 189 80 1 1 9 Large 169 18 1 1 71 155 72 15 13 All 259 22 13 65 344 76 13 11 Karnal Small 61 20 21 59 109 58 24 17 Large 110 17 18 65 89 52 29 19 All 171 18 19 63 198 56 26 18 aFarmers owning less than 10 acres are defined as small farmers Table 6 records the incidence of extension visits during the four weeks preceding the interview. Recall that within such a period, contact farmers are supposed to receive two visits9. Coverage is not expected to be perfect for a variety of reasons which may be termed "friction". Some agents may miss visits owing to illness or leave; similar- ly, some farmers may miss visits as a result of personal circumstances. The data, classified by district, reveal that approximately 80 percent of the contact farmers were visited at least once during the reference period and about two-thirds received two or more visits. These data also indicate that in Jind district, unlike Karnal, there is a (statistically) significant bias in favor of visits to contact farmers with large farms. This may reflect the tendency of extension agents to associate with larger farmers, a feature frequently noted in the literature. However, it is also possible that some of the smaller contact farmers are less interested in participating in visits because they perceive the information delivered to be irrelevant to their needs. As expected, non-contact farmers report a substantially (and statistically significant) lower rate of direct interaction with extension. But it is quite remarkable that 44 per- cent of the non-contact farmers in Karnal district have had, during the reference period, at least one direct contact with extension. The rate of direct exposure, at 24 percent, is much lower in Jind district but is not negligible. The reasons for the differences between the two districts cannot be analyzed with the avaiiable data. A later survey in a neighboring state not covered by the Training and Visit extension system revealed a much lower rate of direct extension contact (only 11 percent of the sampled farmers met an extension agent in the reference period). Although the data indicate a possible farm-size bias in the participation of non-contact farmers in ex- tension visits, the differences between different farm size classes are not statistically significant. Results similar to those reported above were obtained when the analysis was replicated using data pertaining to the whole agricultural season (Kharif, 1981) but excluding the month before the interview. 15 Table 7: Regularity of Extension Visits to Contact and Non-Contact Farmers Contact Farmers Non-Contact Farmers Farm Size Class Sample Size Reporting Sample Size Reporting Regular Visit Regular Visit Ml% (%l Srmiall Farmersa 110 62 82 28 Large Farmers 228 73 87 25 All 338 69 169 27 aFarmers owning less than ten acres are defined as small farmers. The majority (about 70 percent) of contact farmers who received at least one visit during the reference period reported that the visit always took place on the same day of the week' 0. A comparison of the responses across farm size classes (see Table 7) re- veals a statistically significant difference between larger and smaller farmers. This is consistent with the farrr;-size related bias in visits discussed in the preceding section. Such bias may reduce the effectiveness of extension work, since an undue focus on larger contact farmers could limit the possibility of imitation by small farmers. As hypothesized, non-contact farmers in both districts report a much higher inci- dence of visit irregularity (about 73 percent of those who had any visit during the reference period). This may reflect their lack of knowledge regarding the availability of regular extension visits in the village as well as the actual irregularities in visit schedules which can be inferred from the responses of contact farmers. Unlike the case of contact farmers, there is no evidence of significant farm-size related bias in the reported incidence of these less regular visits. An issue of some interest is whether extension workers' interaction with farmers is mostly through individual (or small group) meetings, or through large group meetings. It is reasonable to expect higher effectiveness in small group meetings, but then the coverage of farmers is not as extensive. Our data suggest that the extent of large group meetings in the area covered by our study is fairly limited. As Tat *e 8 shows, only a miniscule proportion (less than 4 %) of non-contact farmers surveyed recently (May 1983) indicated that they have participated in group meetings. The proportion of contact farmers reporting attendance in group meetings is much higher (about one- third in Jind district and 15 percent in Karnal district), but given the responses of non-contact farmers, the implication is that the "groups" in the meetings are rather small, and composed mostly of contact farmers. Corroborating evidence is contained in farmers' response to a question regarding the primary sources of agricultural infor- mation, where less than 5 percent of non-contact farmers in Jind and less than 12 per- cent in Karnal mentioned group meetings as an important source of information. 16 Table 8: Participation in Extension Group Meetings a Jind District Karnal District Farmer Type Farm Size Sample Size % Attending Sample Size % Attending Meetings Meetings Contact Small b 80 29 59 7 Contact Large 116 34 101 20 Contact All 196 32 160 15 Non-Contact Small 183 2 93 4 Non-Contact Large 158 1 73 3 Non-Contact All 341 1 166 4 aResults are derived from a survey conducted in April 1983. b Farmers owning less than 10 acres are defined as small farmers. Summary This paper has focused on two important aspects of the "training and visit" extension method, namely, the selection of contact farmers and the coverage of extension visits. For the particular districts covered by the present study, the analysis shows that con- tact farmers manifest several characteristics which are commonly hypothesized to indi- cate the propensity to lead opinion and potential receptiveness to innovations. Such attributes are essential if contact farmers are to convey agricultural information to the majority of farmers. It is inevitable, however, that contact farmers, as our results show, will not be representative of the farming community in many respects. Thus farmers who are wealthier, more educated, more favorably endowed with irrigation facilities and of higher social status than the majority are more frequently selected as contact farmers. But in other respects which may be important for the ability to communicate and to induce imitation (e.g., caste) contact farmers are fairly representative of the general population. While it is difficult to judge whether the selection of contact farmers is optimal, the evidence of significant farm-size related bias in visits to contact farmers in one of the districts studied warns that some of the traditional weaknesses afflicting extension work can occur under the training and visit system as well. The overall coverage ob- tained under the T & V system is quite satisfactory: most contact farmers meet ex- tension agents at the prescribed frequency. Although visited much less frequently than contact farmers, a significant proportion of non-contact farmers report direct interaction with extension: a favorable result when compared with the much lower rate of direct farmeriagent interaction in non-T & V areas. A majority of the contact farmers also indicate that their interaction with extension is of a regular nature (as it 17 should be). It is not clear, however, whether most non-contact farmers are aware of the regular visits to their villages by extension agents, The picture portrayed by the findings of this study is mixed. Some aspects of T & V extension operations in the study area still require improvement, while other aspects are broadly satisfactory. Other features of T & V extension which were not covered by the present study should be the subject of future research, including the effective- ness and reliab lity of agent training, the relevance of advice offered and the extent to which it is adopted. Only after these features have been reviewed can a judgement on the effectiveness of the extension system as a whole be made. Zusammenfassung Dieser Artikel befaBt sich mit zwei wichtigen Aspekten der Beratungsmethode "Train ing & Visit", namlich mit der Auswahl von Kontakt-Bauern und der Streuungsdichtf von Beratungs-Besuchen. Die vorliegende Studie ergibt fur bestimmte Distrikte, daB die Merkmale von Kontakt bauern vor allem Fuhrungsbegabung und potentielle Obernahmebereitschaft fur Neue rungen sein sollen. Die genannten Eigenschaften sind zur Weitergabe landwirtschafi licher Informationen an die Mehrheit de, Bauern wesentlich. Wie die Ergebnisse der Untersuchung zeigen, sind die Kontaktbauern in vielen Hinsich ten jedoch nicht reprasentativ fur die Bauerngemeinde. Die ausgewahiten Kontak bauern sind meist wohlhabender, besser ausgebildet und mit Bewasserungseinrichtui gen ausgestattet, und sie verfigen uber einen hoheren Sozialstatus als die Mehrhei In anderer Hinsicht aber, z.B. der Kasten-Zugehorigkeit, die fur die Fahigkeit, zu kon munizieren und Nachahmung anzuregen, wichtig zu sein scheint, sind Kontaktfarm( etwa typisch fur die gesamte Bevolkerung. Es ist schwierig zu beurteilen, ob die Auswahl der Kontaktbauern optimal ist, aber wird deutlich, daB der Besuch dieser Bauern in einem der untersuchten Distrikte ei deutig betriebsgroBenbezogen war. Dieser traditionelle Schwachpunkt scheint alk auch die Beratungsarbeit im T & V-System zu beeinflussen. Die breite Streuung, welche dieses System erreicht, ist recht zufriedenstellend: C meisten Kontaktbauern treffen die Berater so haufig wie vorgesehen. Obwohl ein t trachtlicher Anteil der Nicht-Kontakt-Bauern weit weniger oft besucht wird, berichtt auch diese von direkter Interaktion mit der Beratung. Dies ist ein gunstiges Ergebn verglichen mit dem sehr viel geringeren Bauer-Berater-Kontakt in Gebieten ohne d T & V-System. Auch zeigt die Mehrheit der Kontaktbauern, daB sie mit der Beratur wie es sein soil, regelmaBig zu tun hat. Es wird jedoch nicht klar, ob sich die meist Nicht-Kontakt-Bauern dessen bewuBt sind, daS die Berater ihre Dorfer regelmaK besuchen. Das Bild aus den Ergek-`sen dieser Studie ist gemischt. Einige Teile der T & \ Beratungsarbeit 'm Untersuchungsgebiet bedurfen der Verbesserung, wahrend ande' weitgehend zufriedenstellend sind. Wiederum andere Punkte der T & V-Beratung, c von dieser Studie nicht beruhrt wurden, sollten Gegenstand kunftiger Forschung sei 18 so die Effizienz und Verlal3lichkeit der Berater-Ausbildung, die Bedeutung der Be- ratungsinhalte und der Grad ihrer Obernahme. Erst nachdem diese Dinge untersucht worden sind, ist eine Beurteilung der Leistungsfahigkeit des Beratungssystems als Gan- zes m6glich. Notes The views expressed in this paper are those of the authors. They do not necessarily represent the view of the World Bank or its affiliated institutions. In sparsely populated areas this ratio may fall considerably in order to keep the time extension agents spend travelling within reasonable bounds. 2 Contrary to the frequently expressed view that the contact farmer system operates at the ex- pense of the more disadvantaged: for example, see (6, p. 12). 3For instance, in some places in India, large land owners are formally announced as contact farmers, even though they are not directly involved in cultivation. Similarly, we have been in- formed of instances in Sri Lanka where the village teacher (who is not a farmer) is considered a contact farmer. See also evidence cited in 5. 4Farmers included in the sample survey were classified as either "contact" or "non-contact" on the basis of extension agents' written records and farmers' statements. In 93 percent of the cases, the results were identical. For the rest, the farmers' declared status was used. The statistical test used here and in other sections of the paper is based on the large sample normality of the test statistic Z - fP: - P. i v, zr( Z) [N, i + ( 1 rr N, where P, is the proportion of farmers with a certain characteristic within sample 1, N1 are the sample sizes, and i. Is the proportion of farmers possessing the characteristic under investiga- tion within the population. 6 "The selection of representative contact farmers has often proved difficult restilting in an un- warranted proportion of large and progressive farmers facing very different problems from those of the majority" (14, 0D. 17 - 18). However the difference is statistically significant only for the larger farmer class Isee Table 3). One of our referees pointed out that in Haryana, the geographical dispersion of contact farmers within the village is an important criterion in the selection of contact farmers. The objective Is to select contact farmers with fields widely dispersed, so as to increase the demonstration effect. We do not have empirical data on this aspect. Since no. correlation is expected between this variable and other explanatory variables, the econometric work reported below is not likely to be affected by an omitted variable bias. 9 Some contact farmers may have reported only one visit because the interview took place just before the second visit. I0 It should be borne in mind, however, that the agent may sometimes conduct visits on week- ends to make up for missed visit days. Similarly, if the agent's headquarters is located in the village, farmers may see him on days other than the visit day. Such encounters do not consti- tute formal visits. 19 References 1. Benor, Daniel and James 0. Harrison: Agricultural Extension: The Training and Visit Systern Washington D.C., World Bank, 1977. 2. von Blanckenburg, Peter: "The Training and Visit System in Agricultural Extension: A RE view of First Experiences", Quarterly Journal of International Agriculture, vol. 21, no. 1 January/March 1982, pp. 6- 25. 3. Cernea, Michael: "Sociclogical Dimensions of Extension Organization: The Introduction o the T & V System in India", in Extension Education and Rural Development, Vol. 2; Inter national Experience in Strategies for Planned Change, edited by Bruce R. Crouch and Shan karian Chamala (Chichester: John Wiley and Sons Ltd., 1981), pp. 221 - 235. 4. Chambers, Robert: "Two Frontiers in Rural Management: Agricultural Extension and Manag ing the Exploitation of Communal Natural Resources", Institute of Development Studie Communications Series, No, 1 13, 1976. 5. Dubey, Sutton and Gallup: Village Level Workers: Their Work and Result Demonstration INational Institute of Community Development, Government of India - Manager of Publi cations. Delhi, 1962). 6. Howell, John: "Managing Agricultural Extension: The T and V System in Practice", Over seas Development Institute (London), Agricultural Administration Unit Discussion Pape No 8, May 1982. 7. Jamison, Dean and Lawrence Lau: Farmer Education and Farm Efficiency (John Hopkins University Press: Baltimore and London, 1982). 8. Klvlin, Joseph, Frederick Fliegel, Prodipto Roy and Lalit Sen: Innovation in Rural Indit (Bowling Green State University Press: Bowling Green, Ohio: 19711) 9. Llonberger, H.F,: Adoption of New Ideas and Practices (Ames: Iowa State University Press 1960). 10. McFadden, Daniel: "Conditional Logit Analysis and Qualitative Choice Behaviour", in P. Za rembku (ed.) Frontiers in Econometrics (New York: Academic Press, 1973), pp. 105 - 142. 11 Rogers, Everrt M.: Diffusion of Innovations (New York: The Free Press of Glencoe, 1962). 12. Sinha, P.R.K., T.K. Chakravaraty and H.P.S. Arya: Village Level Workers: A Study of Factors influencing Their Performance (National Institute of Community Development, Hyderabad, 1976). 13 Welch, Finis: "The Role of Investments in Human Capital in Agriculture", in Theodore W. Schultz (ed.) Distortion of Agricultural Incentives (Bloomington: Indiana University Press, 19791, pp. 259 - 289. 14 World Bank, Agricultural Extension in India, Agricultural Technical Note No. 5, South Asia Projects Department, August 1981. 20 Appendix: Correlation Matrix of Socio-Economic Indicators Tubewell Membership in Previous Farm Size Education Ownership Village Institutions Agric. Training Education .101 Tubewel I Ownership .179 .200 Membership in Village Institutions .150 .148 .147 Previous Agricultural Training .061 .145 - .002 .068 Con tact Non-Contact .217 146 1 60 .132 .126 21 No. 280. Walter Schaefer-Kehnert and John D. Von Pischke, "Agricultural Credit Policy in Developing Countries," translated from HaJdbucl clder Latndwirtschlaft wnd ErnVt'ihrmg in deni Entwicklntnigsl1iindern (includes original Gerrnan text) No. 281. Bela Balassa, "Trade Policy in Mexico," World Development No. 28 Ia. Bela Balassa, "La politica de comercio exterior de Mexico," Comercio Exterior No. 282. Clive Bell and Shantayanan Devarajan, "Shadow Prices for Project Evaluation under Alternative Macroeconomic Specifications," Thle Qiuarterlu Jouirnal of Economnlics No. 283. Anne 0. Krueger, "Trade Policies in Developing Countries," Handbiook of Internationial Economics No. 284. Anne 0. Krueger and Baran Tuncer, "An Empirical Test of the Infant Industry Argument," Anmerican Econiomic Review No. 285. Bela Balassa, "Economic Policies in Portugal," Econiomlita No. 286. F. Bourguignon, C. Michel, and D. Miqueu, "Short-run Rigidities and Long-run Adjustments in a Computable General Equilibrium Model of Income Distribution and Development," Journal of DeE'elopment Econiomics No. 287. Michael A. Cohen, "The Challenge of Replicability: Toward a New Paradigm for Urban Shelter in Developing Countries," Regioial Development Dialogue No. 288. Hollis B. Chenery, "Interaction between Theory and Observation in Development," WVorld Development No. 289. J. B. Knight and R. H. Sabot, "Educational Expansion and the Kuznets Effect," The Americant Econonmic Revieuw No. 290. Malcolm D. Bale and Ulrich Koester, "Maginot Line of European Farm Policies," The WVrid Econiomiiu No. 291. Danny M. Leipziger, "Lending versus Giving: The Economics of Foreign Assistance," World Dez)elooiieild No. 292. Gregory K. Ingram, "Land in Perspective: Its Role in the Structure of Cities," Vbrled Congress on Land Policq,, 1980 No. 293. Rakesh Mohan and Rodrigo Villamizar, "The Evolution of Land Values in the Context of Rapid Urban Growth: A Case Study of Bogota and Cali, Colombia," World Congress oni Land Policyl, 1980 No. 294. Barend A. de Vries, "International Ramifications of the External Debt Situation," The AMEX Banik Review Special Papers No. 295. Rakesh Mohan, "The Morphology of Urbanisation in India," Economic anid Political Weeklu No. 296. Dean T. Jamison and Peter R. Moock, "Farmer Education and Farm Efficiency in Nepal: The Role of Schooling, Extension Services, and Cognitive Skills," V'4'rld De.clopmiiel! No. 297, Sweder van Wijnbergen, "The 'Dutch Disease': A Disease after All?" Thle Ecoionomic Jourlnal No. 298. Ame Drud and Wafik M. Grais, "Macroeconomic Adjustment in Thailand: Demand M4anagement and Supply Conditions," Jolurnial of Iolicu Modeling No. 299. Shujiro Urata, "Factor Inputs and Japanese Manufacturing Trade Structure," The Review of Ecotnomics and Statistics. No. 300. Dipak Mazumdar, "The Rural-Urban Wage Gap Migration and the Working of Urban Labor NMarke: An Interpretation Based on a Study of the Workers of Bombay City," Indian Ecojionoic Revieuw Issues of the World Bank Reprint Series are available free of charge from the address on the bottom of the back cover.