The training and visit extension system : an analysis of operations and effects

DISCUSSION PAPER Report No.: ARU 25 THE TRAINING AND VISIT EXTENSION SYSTEM: AN ANALYSIS OF OPERATIONS AND EFFECTS by Gershon Feder, Roger H. Slade and Anant K. Sundaram Research Unit Agriculture and Rural Development Department Operational Policy Staff World Bank November 1984 The views presented here are those of the author(s), and they should not be interpreted as reflecting those of the World Bank, THE TRAINING AND VISIT EXTENSION SYSTEM: AN ANALYSIS OF OPERATIONS AND EFFECTS Gershon Feder, Roger H. Slade and Anant K. Sundaram * * Messrs. Feder and Slade are World Bank staff members. Anant Sundaram is a graduate student at Yale University who at the time this paper was written was employed by the World Bank as a consultant. The World Bank does not accept responsibility for the views expressed herein which are those of the authors and should not be attributed to the World Bank or to its affiliated organizations. The findings, interpretations and conclusions are the results of research supported by the Bank (RP0672-29); they do not necessarily represent official policy of the Bank. The designations employed, the presentation of material and any maps used in this document are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates concerning the legal status of any country, territory, city, area or of its authorities or concerning the delimitation of its boundaries or national affiliation. -2- THE TRAINING AND VISIT EXTENSION SYSTEM: AN ANALYSIS OF OPERATIONS AND EFFECTS I. INTRODUCTION Agricultural extension has long been the handmaiden of agricultural research and usually a poor one. To some extent this has been due to the lack of conviction regarding the ability of extension to bring about the sharp increases in agricultural productivity heralded by the growing quantity of new or modified technology emerging from agricultural research institu- tions. Accordingly, increasing attention has been given, in recent years, to ways of improving the management and efficiency of extension systems. ne result has been the emergence of the Training and Visit (T&V) Extension System which was originally tested in Turkey in late sixties. This system which is comprehensively described in Benor and Baxter (1984) is being intro- duced in a number of developing countries often with the assistance of the World Bank. 1/ It has been most widely adopted, since 1977, in India where it has progressively replaced the older system of multipurpose village level ,orkers. Our purpose in this paper is to analyze several aspects of the opera- tion of T&V extension and some of the resulting effects. In the next section we delineate a number of questions worthy of study. In particular we discuss three related matters: (i) the supply of and demand for extension services, (ii) the relationship between extension agents and other sources of agricul- tural information, and (iii) farm productivity in relation to sources of information. The subsequent three sections of the paper address these issues -3- in order and in some detail, drawing on a rich data base provided by Monitor- ing and Evaluation reports of T&V extension issued by various states in India and a study undertaken by two of the authors in conjunction with the Haryana Agricultural University at Hissar, India. Finally, in section six we sum- marize our principal conclusions. II. THE FOCUS OF THE PAPER The early literature on T&V extension (von Blanckenburg et.al. 1980; von Blanckenburg 1982; Cernea 1981; Cernea et.al. 1983; Howell 1982a, 1982b; Jaiswal 1983; Moore 1984, Shingi at.al., 1982; Singh 1983) has been mainly qualitative in nature and more often than not, a review of first experiences. Hence, there are several issues worthy of research and we have chosen three. First is the supply of and demand for extension services as measured by the frequency of VEW visits. 2/ Specific questions are: (a) Is the supply of extension services close to its potential? Is demand as high as supply? (b) Is there a difference in the extension agents' interaction with farmers belonging to different farm-size classes? (c) Is there a difference in the pattern of visits between the two major cropping seasons, namely, kharif, (rainy season) and rabi (dry season)? (d) How does the pattern of interaction change as the new form of extension becomes more established? Regarding question (a), we presume that under the T&V system when it functions properly, there should be a high supply of extension services since the system allows inter alia for effective supervision. Technically, we define "high supply" as being close to the designed frequency of agents' -4- interaction with contact farmers. Demand, as measured by extension agent interaction with non-contact farmers, is bound to be lower than the deserved level of interaction with contact farmers, since such farmers are less aware of the availability of extension services. Demand should, however, be higher in areas with T&V extension than in non-T&V areas. As predicted by theory (Feder and Slade, 1984b), the cost to the farmer of information search and acquisition will be lower in an area of intensive extension coverage because agents are more numerous. There may, however, be a countervailing tendency because non-contact farmers under the T&V system are supposed to obtain information passed on from contact farmers and this could weaken their motivation to meet with the VEW (i.e., lower their apparent demand for exten- sion services). In areas without T&V extension however, there are no for- mally designated contact farmers and hence this demand reducing tendency may not exist. In relation to question (b), evidence suggests that extension agents are traditionally biased towards the more wealthy and influential farmers (see for example, Howell 1982a; p. 10). The factors and motivations gener- ating such a bias in the supply of extension could still be present under the reformed extension system. On the demand side, the economics of informa- tion-acquisition suggest that demand by smaller farmers will be less than that by larger farmers (Feder and Slade 1984b). Concerning (c) above, it is expected that the VEW will play a more significant role in the dry season (i.e., rabi) if there is a significant number of farmers with some access to irrigation. This is because agricul- tural research in countries like India has traditionally been directed -5- towards improving the technology in respect of rabi crops of irrigated rather than rainfed cropping and hence there is a greater quantum of proven technol- ogy available for delivery to farmers by the extension service. However, as we hypothesize below, the greater riskiness of rainfed agriculture could serve to increase the demand for accurate and proven information during the wet (kharif) season. In relation to question (d), it is conceivable that as initial enthusiasm and institutional support diminish, various aspects of extension operations, such as visits to farmers, slacken off. On the demand side, any favorable experience with extension advice and increased awareness of exter:- sion availability will tend to increase interaction with extension, while disappointment with recommended practices may diminish demand. These four questions are explored further in Part III below. The second set of questions concerns the VEW as an information source in relation to other sources of information (e.g., other farmers, radio, etc.) and the extent to which the VEW is a preferred source for more expen- sive or complex agricultural practices. Specifically, we ask the following questions. (e) How important is the extension agent as a source of information in relation to other sources in areas covered by T&V extension? How does this compare with the role of an extension agent in a non-T&V extension system? (f) What is the nature of the interaction between extension agents and other sources of information? (g) Is the T&V agent more important than other (non-personal and non-specialized) sources of information the more expensive/riskier an agricultural practice becomes? Concerning (e), we expect the VEW, if the training and upgrading of staff called for under the T&V system are effective, to be the most important -6- source of information, since he is a personalized and specialized means of information transmission to farmers while also being more readily available due to the schedule of frequent visits. We also hypothesize that the impor- tance of the agent under the T&V system outweighs that of an extension agent in a non-T&V setting, because of both superior training and higher availability and visibility. In considering issue (f), our approach is exploratory. We attempt to examine other information sources as either complements to or substitutes for personalized extension. For example, Orivel (in Perraton et al, 1983, p. 31) quoting another study related to India (Shore, 1980) states that although radio was the medium most equitably distributed, its use had no impact on the introduction of agricultural innovations. He suggests that it could, however, be complementary and if used as a means of transmitting information to VEWs it might also reduce training costs. Finally, regarding question (g), the expected answer is positive, since the value of specific and accurate information should increase for riskier, more expensive, or complicated practices. Hence, risk, cost and complexity should increase the importance of the VEW, since he is a special- ized and personal information source, with access to subject matter specialist for additional reference in case of doubts or additional inquiries on the part of farmers. The three questions reviewed above are explored further in Part IV below. Our third and final set of questions relates to farm productivity. Specifically: -7- (h) Are yields higher for farmers who report the extension agent to be.their main source of information? Does this hold for irrigated as well as unirrigated farms? If extension is delivering a flow of proven and acceptable technology then the adoption of that technology is likely to be greatest amongst those farmers who depend most heavily on extension for information. Hence such farmers should enjoy a correspondingly more discernible and positive effect on crop yields providing that the improved technology is not wholly cost reducing. Following the arguments above, we also expect this effect to be greater in irrigated than in rainfed farms. Moreover we hypothesize that this would hold even if the information provided by extension in rainfed and irrigated areas had equal expected value as the inherent riskiness of rainfed agriculture is greater than irrigated agriculture. This question is explored further in Part V. The data for our analysis pertain to India and are drawn from two main sources: primary data collected by the Haryana Agricultural University (HAU) in a case study in North India, sponsored and supervised by the World Bank between 1981 and 1983, 3/ and the Monitoring and Evaluation (M&E) reports produced by twelve of the thirteen states in India where T&V has been progressively instituted since 1977. Wherever relevant we indicate which source we have used. III. SUPPL7 OF AND DEMAND FOR EXTENSION AGENT VISITS The interactions of the VEW with contact farmers can be viewed as a system determined "supply" of extension services. "Supply" is a relevant concept because the T&V system requires the VEW to provide these services to contact farmers at fixed and regular times. Contact farmers, in turn, -8- are expected to disseminate this information to non-contact farmers. Thus for non-contact farmers, the interactions between farmers and extension agents are likely to be "demand" determined -- i.e., non-contact farmer meetings with the extension agent may reflect the farmer "demanding" the information, since it is not a system imperative that the VEW visit regularly farmers other than contact farmers. 4/ The agent is expected, however, to accommodate requests for information from all farmers. It is also expected that non-contact farmers will occasionally join the meeting between the agent and contact farmers. For the purpose of examining the broad differences in the quantum of visits by the extension agent to contact and non-contact farmers, data from M&E reports from seven states in India over a number of years are sum- marized in Table 1. These reports have been produced more-or-less regularly by the states were T&V has been instituted. The data are comparable across states because all M&E units use the same sampling design, definitions and questionnaire (for details, see Slade and Feder, 1981). The reference period for visit frequencies as reported in the Table should be one month, but it is possible that non-contact farmers, since they do not receive regular visits, have given their response with a longer horizon in mind, generating a fre- quency of visit per month higher than the actual. The critical indicator is the percentage of farmers reporting not seeing extension agents. For contact farmers, this ranges from 1.2 to 34.7 percent, while for non-contact farmers, it ranges from 21.4 to 59.2 percent. Across all seven states the average percentage of 'no-visits' reported by contact farmers is 15.4 percent (i.e., about 85 percent of contact farmers -9- Table 1: FREQUENCY OF VEW VISITS AND FARM SIZE m I CONTACT FARMERS I NON-CONTACT FARMERS I States/Seasong I Monitoring or Sample 1 No. of Visits Within I Sample I No. of Visits Within and Operated Monitoring I Size L_the Past Four Weeks I Size I the Past Four Weeks Farm Size cum Evaluation I 0 1 2 or morel I 0 1 2 or more --- ercqnt -- p --- ercegt Harvan4 Rabi 1981-82 M Small Farms g/ 251 17.1 14.0 68.9 333 48.6 27.9 23.5 Large Farms 202 15.9 12.8 71.3 138 45.7 29.7 24.6 Kharif- 1982-83 M Small Farms 232 16.4* 17.2 66.4 309 40.4 35.9 23.7 Large Farms 219 10.0 20.1 69.9 138 45.6 21.0 33.3* Kharif 1982-83 McE Small Farms 64 7.8 21.9 70.3 658 52.3 24.5 23.2 Large Farms 64 6.2 17.2 76.6 267 49.8 21.3 28.9 Karnataka Rabi 1981-82 M Small Farme 2024 15.4* 17.3 67.3 1482 30.2 21.8 48.0* Large Farms 1143 12.3 20.6 67.1 530 32.6 26.4 41.0 Rabi 1981-82 McE Small Farms 86 17.4 16.3 66.3 395 40.3 21.8 37.9 Large Farms 159 11.4 17.6 71.0 493 40.2 25.8 34.0 Kharif 1982-83 K Small Farms 1499 10.1 10.8 79.1* 1869 28.0 16.2 55.8* Large Farms 1133 10.9 14.7 74.4 760 25.8 25.8 48.4 Kharif 1982-83 McE Small Farms 307 13.0 24.1 62.9 2065 50.7 19.7 29.6 Large Farms 235 8.9 27.2 63.8 712 50.6 20.5 28.9 Rabi 1982-83 M Small Farms 1280 13.0 21.6 65.4 1621 28.2 24.6 47.2 Large Farms 988 14.0 15.0 71.0* 698 27.8 22.7 49.5 Rabi 1982-83 McE Small Farms 69 18.8 23.2 58.0 606 57.9* 18.0 24.1 Large Farms 168 11.3 17.3 71.4* 544 48.7 20.8 30.5* Kharif 1983-84 M Small Farms 1157 20.4* 13.9 65.7 1593 36.4* 18.8 44.8 Large Farms 944 15.4 18.6 66.0 648 31.3 19.4 49.3 Guijara t Kharif 1981-82 M Small Farms 503 33.2 19.9 46.9 - - - - - n. a . - - - - - - Large Farms 328 28.6 21.6 49.7 - - - - - n. a . - - - - - - Kharif 1981-82 McE Small Farms 298 8,0 19.5 72.5 - - - - - n. a . - - - - - - Large Farms 237 8.0 20.7 71.3 - - - - - n. a . - - - - - - Rabi 1981-82 M Small Farms 527 23.9 17.5 58.6 - - - - - n. a . - - - - - - Large Farms 308 20.1 14.0 65.9* - - - - - n. a . - - - - - - Rabi 1981-82 McE Small Farms 208 8.6 14.9 76.5 - - - - - n. a . - - - - - - Large Farms 184 12.0 20.1 67.9 - - - - - n. a . - - - - - - Kharif 1982-83 M Small farms 498 24.3 14.3 61.4 - - - - . a . - - - - - -. Large Farms 337 19.3 14.2 66.5 - - n. a . - - - - - - Kharif 1982-83 McE Small farms 394 12.4 15.0 72.6 - - - - - n. a . - - - - - - Large Farms 375 13.1 20.3 66.6 - - - - - n. a . - - - - - - Rabi 1982-83 M Small Farms 481 21.9 18.8 59.3 - - - - - n. a . - - - - - - Large Farms 354 20.2 22.1 57.7 - - - - - n. a . - - - - - - Assam Kharif 1982-83 McE Small Farms 394 27.3* 9.6 63.0 423 48.0* 22.7 29.3 Large Farms 332 11.2 18.0 70.8* 308 21.4 9.1 69.5* Maharashtra Rabi 1983-84 M Small Farms 1200 14.0* 12.2 73.8 - - - - - n. a . - - - - - - Large Farms 735 10.2 11.4 78.4* - - - - - n. a . - - - - - - Bihar Kharif 1983-84 M Small Farms 734 32.2 17.5 50.3 854 59.2 15.2 25.6 Large Farms 352 34.7 16.8 48.5 249 55.0 16.1 28.9 Tamilnadu Summer 1982 M Small Farms 347 2 4.6 93.1 - --- . a.------ Large Farms 83 1.2 8.4 90.4 -----n. a .------ Kharif 1982-83 N Small Farms 1317 3.4 2.8 93.8 - - - - - n. a . - - - - - - Large Farms 248 2.8 3.6 93.6 - - - - - n. - - - - - - * Significant at 5 percent probability level n.a. - not available Small farms were defined as less than 5.1 hectares in Haryana and Gujarat 4.1 hectares in Tamil nadu and Karnataka, 3.1 in Madhya Pradesh and Bihar and 2.1 hectares in Assam. - 10 - were visited at least once in the reference month), while 34.5% of the non-contact farmers reported no interaction with extension (see Table 2). The demand for T&V extension services as measured by non-contact farmers interaction with extensions thus appears significant. Considering that some share of 'no-visits' must be due to factors such as agent illness, unfilled vacancies, contact farmer non-availability, etc. (factors that Feder and Slade (1984a, p. 15) refer to as "normal friction") the actual supply of T&V extension services seems adequate relative to the potential supply. Table 2: FREQUENCY OF VEW VISITS TO CONTACT AND NON-CONTACT FARMERS: ALL STATES AND ALL CROPPING SEASONS /a Contact Farmers Non-contact Farmers One or More One or More No Visits Visits No Visits Visits --------------------------Percent--------------------------- 15.39 84.61 34.49 65.51 (3,321) lb (18,219) (4,268) (8,117) )a During the four weeks preceding the interview for contact farmers. For non-contact farmers the reference period may be longer. lb Figures in parentheses indicate sample sizes. Source: Various monitoring and evaluation reports of state governments in India. As expected, the demand for extension services (measured by agent's interaction with non-contact farmers) is significantly lower than the supply (measured by agent visits to contact farmers). But the actual supply avail- able to non-contact farmers must be less than that which is available to contact farmers, thus there is not necessarily a significant level of unused capacity. Further, the demand for extension services in a T&V area is far - 11 - higher than the demand in a non-T&V area -- data for a section of Muzaffar- nagar district in the state of Uttar Pradesh, which is not covered by the T&V system, show that between 89 and 97 percent of the farmers were not visited by (or did not seek out) the extension agent during the reference period (Feder and Slade, 1984c, p. 16). Because a distinction cannot be drawn between contact and non-contact farmers in non-T&V areas, this could be the result of either low demand or low supply. It is known however that the extension agent/farmer ratio is lower in non-T&V areas than in areas with T&V and that in the former areas agents have many duties other than extension. Hence it is possible thaz in areas without T&V extension the low supply of extension increases the cost to the farmer of acquiring information from extension and thereby reduces the amount of interaction between farmers and extension agents. As a partial check on the accuracy of the data from the M&E reports we compared the M&E results to those from the HAU/World Bank study. Specifi- cally, we compared the incidence of 'no-visits' to large and small, contact and non-contact farmers, in the whole of Haryana (all 12 districts) with those from the HAU/World Bank study in Jind and Karnal districts of Haryana for Rabi 81-82, the only season for which comparable data are available. The M&E reports in.dicate (see Table 1) th.at 17 percent of small contact farmers and 16 percent of large contact farmers are not visited, while the RAU/World Bank data (see Feder and Slade, 1984c, p. 16) show 17 and 14 percent respec- tively. For non-contact farmers, the M&E figures are 49 percent for small farmers and 46 percent for large farmers while the HAU/World Bank figures are 69 and 52 percent respectively. These are comparable results although in the - 12 - HAU/World Ban.k study area it seems that small non-contact farmers are visited significantly less frequently than their counterparts elsewhere. In addi- tion, the definition of the reference period was quite strict in the HAU/Bank study, while it may have been less so in M&E reports. Table 3 summarizes data on visits and 'non-visits' by farm size. Prima facie, there is remarkable similarity between large and small farms amongst both contact and non-contact farmers. Amongst contact farmers, 15.9 percent of the small farms and 14.5 percent of the large farms are not visited, a difference of 1.4 percent. Similarly, for non-contact farmers, the difference is only 3.2 percent. While these differences are statisti- cally significant at the 99 percent level (see Table 3), their size indicates that the bias in favor of large farmers is not great enough to warrant serious concern. 5/ Moreover, since non-contact farmers' interactions with extension agents are probably demand-driven, the difference between large and small farmers may merely indicate, as predicted by theory (Feder and Slade 1984b), the tendency of larger farmers to invest more in information gather- ing. - 13 - Table 3: FARM SIZE AND VEW VISITS TO CONTACT AND NON-CONTACT FARMERS: ALL STATES AND ALL SEASONS Ia Contact Farmers Non-contact Farmers One or More One or More No Visits Visits No Visits Visits --------------------------Percent---------------------------- Small /b 15.91 84.09 35.44 64.56 (2,098) /c (11,000) (3,038) (5,532) Large 14.49 85.51 32.17 67.83 (1,223) (7,219) (1,230) (2,594) /a See footnote /a in Table 2. lb See footnote 7 in Table 1. Ic Figures in pareatheses indicate sample sizes. Tests of Significance for 'No-visits' (Large vs Small Farmers) "t" Value Contact Farmers 2.823) ) Both significant at 99 percent Non-contact Farmers 3.535) Source: Various monitoring and evaluation reports of state governments in India. Further corroboration is provided by Feder and Slade (1984c). Using logic analysis, they examined factors that explain the probability that a contact farmer will be visited by an extension agent. Amongst others "area of land owned" was entered as an independent variable and was found to be a positive but not significant explanator of the probability that a contact farmer would be visited. Earlier, we hypothesized that visit frequencies would be greater in the dry (rabi) season compared to the kharif. Data for both contact and non-contact farmers (Table 4) indicate that the incidence of no-visits during the rabi season is significantly lower than in the kharif, although the absolute difference is small. This result is consistent with an analysis - 14 - conducted by Feder and Slade (1984c, pp. 30-32), which shows that knowledge diffusion rates tend to be higher for dry season crops than for rainy season crops. These findings support the hypothesis that the extension agent plays a greater role in the dry season although the cause may be more closely associated with the available technology and the riskiness of rainfed agriculture than the efficiency of the extension system. Another explanation may be that the rainy season creates problems due to more limiEed mobility of extension agents in bad weather. Table 4: CROPPING SEASON AND VISITS OF VEW TO CONTACT AND NON-CONTACT FARMERS: ALL STATES /a Rabi Kharif (dry season) (rainy season) - Percent----------- No visits 14.45 16.25 Contact (1,522) /b (1,799) Farmers One or more visits 85.55 83.75 (9,038) (9,281) No visits 32.60 36.00 (1,854) (2,414) Non-contact Farmers One or more visits 67.40 64.00 (3,836) (4,290) /a See footnote /a in Table 2. /b Figures in parentheses indicate sample size. Tests of significance for 'no-visits' (Kharif vs Rabi) "t" values Contact farmers: 1.44 (significant at 90 percent) Non-contact farmers: 2.31 (significant at 99 percent) Source: Various monitoring and evaluation reports of state governments in India. - 15 We next examine the trend in extension visits as experience with the T&V system increases. The results, summarized in Table 5 and Figure 1, form a mixed picture. The proportion of contact farmers not visited goes up significantly: amongst projects which are four or more years old nearly one in five contact farmers are not visited. This may, in part, be due to the VEW replacing those contact farmers who are deemed inadequate with other farmers without formally notifying the original contact farmers of the change. On the other hand, the proportion of non-contact farmers not visited declines equally significantly, from about 48 to 36 percent. This may be due to the fact that as the project gets established, knowledge about the availability of regular extension visits spreads and more non-contact farmers take advantage of the service. As mentioned earlier, it is expected in the T&V extension system that the agent respond to all farmers who approach with inquiries (Benor and Baxter, 1984). Recalling that contact farmers form only about 10 percent of the farming community, the most important result is that the abzolute number of visits to all classes of farmers is the greater, the older is the project. 6/ Figure 1 80% 60% 40% 4. Non-contact Percent of Fa8mers farmers not - visited 20% Contact Farmers 1 2 3 More than Year Years Years 4 Years Project Life ---> PROJECT LIFE AND FARMERS NOT VISITED BY VEW - 17 - Table 5: PROJECT LIFE, FARM SIZE AND VEW VISITS: /a ALL SEASONS AND ALL STATES /b Project All Life Visits Small Farms /c Large Farms Farms Sample Sample Percent Size Percent Size Percent Contact 0 5.01 96 7.50 40 5.74 > 1 94.99 1,819 92.50 493 94.26 < 1 year Non-contact 0 48.65 162 45.65 63 47.81 > 1 51.35 171 54.35 75 52.19 Contact 0 17.13 830 13.93 431 16.04 > 1 82.87 4,016 86.07 2,663 83.96 2 years Non-contact 0 32.34 607 36.27 371 33.83 > 1 67.66 1,270 63.73 652 66.17 Contact 0 14.26 592 14.06 448 14.17 > 1 85.74 3,560 85.94 2,739 85.83 3 years Non-contact 0 28.08 980 26.75 390 27.70 > 1 71.92 2,510 73.25 1,068 72.30 Contact 0 22.18 344 14.26 186 19.44 > 1 77.82 1,207 85.74 1,094 80.56 > 4 years Non-contact 0 38.83 783 28.14 269 36.09 > 1 61.17 1,233 71.86 687 63.91 /a See footnote /a in Table 2. /b Excludes Bihar because of great diversity in project initiation dates for different districts. For all other states, the most predominant starting date (in terms of number of districts covered) has been chosen. /c See footnote /a in Table 1. Source: Various monitoring and evaluation reports of state governments in India. - 18 - IV. EXTENSION AGENTS IN RELATION TO OTHER SOURCES OF INFORMATION It is reasonable to presume that farmers tend to prefer direct, specialized, personal and easily accessible sources of information, provided they see such sources as being reliable and professional. Therefore, in areas with a high supply of professional extension we expect the role of extension as a means of information dissemination to increase. Thus, we pose the fol- lowing questions (i) are T&V agents the most important information source in areas covered by T&V extension? and (ii) do extension agents play a more important role in such settings compared to those operating where a less intensive system of extension is present? We also wish to examine how the role of the extension agent changes in relation to other information sources as farmers gain increasing "access" to extension. Finally, we hypothesize that extension becomes more important to the farmer the riskier, more complex or more expensive an agricultural practice becomes. We first examine how important the extension agent is as a source of information, in T&V and non-T&V areas. Table 6 summarizes primary data from the HAU/World Bank study which was collected from geographically con- tiguous T&V and non-T&V districts in 1982. - 19 - Table 6: RELATIVE IMPORTANCE OF SOURCES OF AGRICULTURAL INFORMATION Karnal District (State of Haryana) T&V District Muzaffarnagar District Non-Contact (State of Uttar-Pradesh) Main Contact Farmers Farmers Non T&V District Information Source Small Large/a All Small Large/a All Small Large/a All (N=59) (N=101) (N=160) (N=93) (N=73) TN=166) (N=45) (N=45)_TN=90) -------------------Percent------------------- ---------Percent-------- Extension Personnel 42 45 44 11 16 13 0 3 2 Demonstra- tion Days 2 5 4 1 1 1 13 9 11 Other Farmers 27 20 22 48 43 46 49 42 46 Radio 16 17 16 18 22 20 20 25 23 Sales Personnel 9 7 8 15 11 13 12 8 10 Research Personnel 2 2 2 2 2 2 0 0 0 Other /b 2 4 4 5 5 5 6 13 8 All 100 100 100 100 100 100 100 100 100 /a Larger farmers were defined as those owning ten or more acres. lb Includes written materials, demonstrations, group meetings and miscellaneous. Source: uAU/World Bank Study, 1982. There is a dramatic difference between the two districts. In Muzaffarnagar (a non-T&V district) only 2 percent of the farmers (all large farmers) indicate that extension agents are the main source of information. In Karnal (a T&V district) 44 percent of the contact farmers and 13 percent of the non-contact farmers are of the same opinion. For non-contact farmers - 20 - in Karnal, "other farmers" are the most important source of information, followed by radio. In Muzaffarnagar, this was so for all farmers. "Other farmers" are an important source of information for contact farmers as well; slightly more so than radio. Sales personnel of firms marketing agricultural inputs also constitute a significant source of information for all farmers. Since other farmers are the most frequently cited source of agricul- tural information, and given the impracticality of attempting to reach all farmers directly by extension, it is logical to base an information dissemi- nation strategy on a two-step flow principle, whereby some farmers get con- tinuous and frequent extension visits. Through the natural process of infor- mation diffusion, these farmers may be expected to transmit information further to other farmers. These data also reveal the relative shares of different information sources (see Figure 2). Ranking farmer categories by access to the T&V agent we see that the share of the VEW goes up from 2 to 44 percent with increasing access to T&V extension. It is then of interest to know which sources decline in importance. Among contact farmers, as the importance of the VEW increases the importance of "other farmers" declines sharply. This is as expected, since contact farmers are in a position to obtain information first-hand rather than second-hand. Among non-contact farmers, however, "other farmers" continue to play a significant role, which is again as expected. This relationship between VEWs and "other farmers" as sources of information is consistent with the two-step communication flow characterizing the T&V system. - 21 - The share of radio remains more-or-less constant regardless of access to personal extension services. This is compatible with the hypothesis that radio is a source of information that complements the role of the VEW (as suggested by Orivel in Perraton et. al., 1983). The lowest ranked sources of information are "sales personnel" and "demonstration days". The former seem to be an important source for farmers in non-T&V areas suggesting that they serve as a partial substitute for visits by extension agents. These five sources of information account for a little over 90 percent of the information needs of all classes of farmers. To examine whether the-e may be differences related to farm-size, a similar analysis was conducted for large and small farmers separately. As can be seen from Figure 3, the results are almost exactly the same for both classes. There is a slightly higher preference among larger farmers for information from extension agents (irrespective of whether they were contact or non-contact farmers in T&V areas or farmers in non-T&V areas) but the differences are not statistically significant. Figure. 2 100% 100% Others Te- n rat onc Radio 50% -Other 50% Farmers " Ex tension Personnel T0 non-contact Non-T&V FresT&V contact Farmers Farmers CUMULATIVE SHARE OF INFORMATION SOURCES - ALL FARMERS - 23 - Figure 3 100% '10U% Others ) em as, e/Pers Radio Farmers Extension Personnel 0 ' 0 a) CUMULATIVE SHARE OF INFORMATION SOURCES - LARGE FARMERS 100% - - -- - - -100% Others - Dem. Days aePes Radio oter Farmers Extension Personnel 0 0 Non-TAV T&V in-contact TkV contact Farmers Farmers Farmers b) CMULATIVE SHARE OF INFORMATION SOURCES - SMALL FARMERS - 24 - We also examine data on the T&V system drawn from 7 states in India spanning 17 cropping seasons. These data are taken from the M&E reports produced by the State Governments and record the main source of information, namely the "VEW", "other farmers", "other sources", and "no advice" for contact and non-contact farmers. We presr.me that the category "no advice" means "minimal advice" or the acquisition of information through observation or casual conversation. From Table 7, we see that 80 percent of contact farmers and 54 per- cent of non-contact farmers claim extension agents to be their main source of information. Those who claim to take "no advice", comprise about 9 percent of contact farmers and nearly 19 percent of non-contact farmers. "Other farmers", as a source of information, are the main source for 20 percent of non-contact farmers, but less than 7 percent of contact farmers. These data support the finding that extension agents are the most important source of agricultural information in areas with T&V extension. - 25 - Table 7: FARMER'S MAIN SOURCE OF INFORMATION IN STATES WITH THE T&V EXTENSION SYSTEM Main Sources Contact Non-contact of Information Farmers Farmers ----------Percent----------- VEW 79.66 53.58 Other farmers 6.87 19.92 Other sources 4.66 7.81 No advice 8.81 18.69 All 100.00 100.00 Source: Various monitoring and evaluation reports of state governments in India. In contrast, data on sources of agricultural information from a study in a non-T&V setting on the socio-economic constraints to rainfed agriculture in Thailand (Hutanuwatr et. al., 1982) indicate that extension officers are the 4th ranked source -- the most important being "relatives and neighbors" (equivalent to our category of "other farmers") followed by radio programs and community leaders. The authors also report that "... more than half of the farmers sampled felt that extension officers could not help them solve any agricultural problems" (p. 25). We next examine the question of whether extension agents become more important as information sources the more expensive or complicated a practice becomes, by calculating "information source ratios" for two increasingly expensive categories of agricultural practices. 7/ "Expensive" means the opportunity loss resulting from wrong application of the practice as well as the simple financial cost. The information source ratio is an indicator of - 26 - the relative importance of two information sources: the VEW and "other farmers" (i.e., first-hand versus second-hand sources). Relevant data are shown in Table 8. We note that in areas without T&V extension the VEW plays a very minor role in relation to both groups of practices -- the ratios for less expensive practices and more expensive practices are 0.04 and 0.09 respectively. In areas with T&V extension the comparable ratios for non-contact farmers are 0.27 and 0.47, while those for contact farmers are much higher at 3.98 and 5.14. Reflecting the more favorable "supply" conditions, the ratios become higher as access to exten- sion increases. Table 8: WEIGHTED AVERAGE INFORMATION SOURCE RATIOS FOR WHEAT PRACTICES /a Non-T&V Farmers T&V Non-contact Farmers T&V Contact Farmers Less More Less More Less More Expensive /b Expensive Expensive Expensive Expensive Expensive Practices c Practices Id Practices /c Practices Id Practices /c Practices /d Small 0.00 0.0 0.20 0.28 3.51 4.43 Large 0.07 0.16 0.36 0.68 4.27 7.50 All 0.04 0.09 0.27 0.47 3.98 5.14 /a The "Source Ratio" is the ratio of the number of times that a VEW is cited as the main source of information to the number of times that "other farmers" is cited. /b "Expensiveness" refers to the opportunity loss resulting from the incorrect appli- cation of the agricultural practices as well as their cost, or complexity. /c Less expensive (variety choice, seeding rate and spacing). T More expensive (use of phosphate, potash and zinc, seed treatment against termites, seed treatment against disease, timing of nitrogen application). Source: HAU/World Bank Study, Rabi 1982/83. There is also a distinct pattern with respect to farm size -- the source ratios are consistently higher for larger farms. This may result from - 27 - larger contact farmers having somewhat greater access to the VEW (see Table 3) and larger non-contact farmers investing more in information acquisition. Moreover the VEW is likely to be a more expensive source (in terms of time taken to locate and meet.him) than "other farmers". In short, irrespective of farm size the data show that all classes of farmers prefer to receive advice about the more "expensive" practices from the VEW. Similar -views are expressed by Howell [1984, pp. 174, 175]. We also examined evidence contained in a detailed report on a study of T&V extension operations in the Indian state of Madhya Pradesh conducted by the National Council for Applied Economic Research (NCAER, 1983). Data in the report (pp. 49-51) generally support the contention that VEWs play an important role as sources of knowledge about recommended practices for both contact and non-contact farmers. Further, the NCAER data suggest that the role of the VEW increases with increasing riskiness or complexity of the agricultural practice. V. INFORMATION SOURCES AND FARM PRODUCTIVITY The process by which extension influences crop yields involves a wide range of intervening variables. The effect is indirect and not easily measured. However, if extension efforts are successful, this success must eventually result in increased productivity and/or reduced costs of produc- tion per unit output. Since the contact point between the extension system and the farmer is the village extension worker, it is essential that the VEW, as a first-hand information source, be "better" than other second-hand or non-personal sources of information. Otherwise, it would be hard to justify - 28 - the expense of an intensive agricultural extension system. A testable hypothesis is therefore implied, namely, that farmers whose main information source is the extension agent will have higher productivity than those who rely on other information sources, ceteris paribus. There may be some sys- tematic relation between the fact that some farmers are more inclined to utilize extension as a main source of information and other attributes which make them better farmers (e.g. intelligence), who obtain higher yields. Unfortunately, the data do not permit all other relevant attributes to be held constant and therefore the analysis below is suggestive rather than definitive. Drawing again on the state M&E reports in India, we use data on crop yields in the kharif and rabi seasons disaggregated by information source. For kharif, we use rice yields and for rabi, wheat, under both irrigated and unirrigated conditions. State average yields were calculated by applying weights based on the sample sizes for irrigated and unirrigated farms and contact and non-contact farmers. The resulting overall average-for each state was set equal to 100. Subsequently, each subset of yields was expressed as an index number relative to the overall state average. This conversion permits rice and wheat yields to be compared (since they differ in absolute magnitudes) and heavily damps differences in agro-climatic and socio-economic factors between states. The net result of this conversion is a series of index numbers that is comparable across states, crops and cropping seasons. A summary of the data for irrigated, unirrigated and all farms, disaggregated by main source of farmer information, is shown in Table 9. - 29 - Table 9: YIELD INDICES BY MAIN SOURCE OF INFORMATION Ia Main Source of Information Other Other No VEW Farmers Sources Advice 116.25 89.52 92.64 93.22 Irrigated (13) lb (13) (13) (13) 114.75 101.90 103.29 /c 79.88 Unirrigated *(13) (13) (13) (13) 114.50 99.08 95.77 86.11 All (15) (15) (15) (15) Ia The actual sample base in terms of numbers of farmers is more than 1,500; the sample sizes in the table refer to the number of average yield index figures and hence represent a mean of means. lb Figures in parentheses indicate sample size. Sample sizes differ because, for two states, data disaggregated by irrigated and unirrigated farms were not available. Ic One state, in one cropping season, had an unduly high yield figure and the sample base was extremely low in relation to the rest and hence was insignificant in the computation of weighted average yields for "all" farms. However, in computing the average across unirrigated farms in all states, this number receives equal weighting. Hence this particular figure should be considered an overestimate. Source: Various monitoring and evaluation reports of state governments in India. Farmers whose main source of information is the VEW have the highest yield index of 114.5. This is followed by those whose source is "other farmers" and their yield index is close to the average. "Other" sources (e.g., radio, demonstration days, sales personnel, etc.) have a lower yield index of 95.77 and those farmers who receive "no advice", 86.11. Prima ~~~~~ facie, it would appear that those using the VEW as the main source of ~~~~~ information have yields that differ substantially from all other sources, but the difference between "other farmers" and "other sources" is much less - 30 - marked. All three, however, appear to be better compared to those farmers receiving "no advice". We test these results more rigorously in a multiple regression framework (equivalent to analysis of variance since all explanatory variables are categorical), suppressing the variable against which we test for sig- nificance. We specify the dependent variable (yield) in its logarithm as conventionally done in production analysis. 8/ The first set of results, for irrigated and unirrigated farms and all farms, are shown separately in Table 10. In these results we have controlled for 'other sources', that is to say the coefficients of the remaining explanatory variables represent deviations from the effect of 'other sources'. Table 1.0: YIELD INDEX AND INFORMATION SOURCES, CONTROLLING FOR "OTHER SOURCES" Dependent Variable: Log (YIELD INDEX) Parameter Estimate Independent ('t' values in parentheses) Variable All Irrigated Unirrigated Intercept -0.061 (-1.419) -0.003 (-0.055) 4.584 (66.809) VEW 0.190. ( 3.119) 0.141 ( 2.140) 0.151 ( 1.551) Other Farmers 0.044 ( 0.721) -0.119 (-1.802) 0.010 ( 0.105) No Advice -0.107 (-1.750) -0.988 (-1.303) -0.223 (-2.296) F 8.418 6.425 5.059 R2 0.266 0.2953 0.2402 All three regressions have significant F values (at > 99 percent levels), suggesting reasonable overall explanatory power. As suraised, yields for farmers receiving information from VEWs are significantly higher - 31 - than those for farmers who depend on "other sources". There is significant difference at the 99 percent level for all farms; for unirrigated farms the difference is significant at a little less than 90 percent and for irrigated farms at 95 percent. "Other farmers" do not constitute a source of informa- tion that has a productivity effect significantly different from "other sources" for all farms and unirrigated farms. Receiving "no advice" is significantly worse than receiving advice from "other sources" for all farmers and unirrigated farmers, but not significantly worse (though the direction of the relationship is negative) in the case of irrigated farms. Similar results, when. "no advice" is the control variable, are presented in Table 11. In this set of regressions the coefficients of the explanatory variables represent deviations from the productivity effect of no advice'. VEW advice is significantly better than "no advice" in all three cases. "Other farmers" and "other sources" come out significantly better than "no advice" for all farms and unirrigated farms, and better, but not significantly so for irrigated farms. 9/ - 32 - Table 11: YIELD INDEX AND INFORMATION SOURCES, CONTROLLING FOR "NO ADVICE" Dependent Variable: Log (YIELD INDEX) Parameter Estimate Independent ('t' values in parentheses) Variable All Irrigated Unirrigated Intercept -0.168 (-1.536) -0.076 (-1.536) 4.361 (63.561) VEW 0.297 ( 4.870) 0.215 ( 3.193) 0.373 ( 3.848) Other Farmers 0.151 ( 2.471) -0.055 (-0.826) 0.233 ( 2.401) Other Sources 0.107 ( 1.750) 0.073 ( 1.072) 0.223 ( 2.296) F 8.418 6.641 5.059 R2 0.3039 0.3022 0.2402 VI. CONCLUSIONS This paper drew upon an extensive set of aggregate and farm level data mostly pertaining to T&V extension operations in India. The analysis implies the following conclusions. (a) Approximately 85 percent of contact farmers are visited at least once a month, suggesting that "supply" of extension services, taking normal "friction" into account, is reasonable. Amongst non-contact farmers 65 percent have interacted with extension workers at least once during the reference period, suggesting that demand is substantial. (b) The data indicate that there is a statistically significant bias in favor of visits to large farmers; however, the absolute size of this bias is very small. (c) T&V agents appear to be more active and in higher demand in the dry as opposed to the rainfed cropping season; this is probably explai- uble by the fact that research has traditionally emphasized dry-season cropping technology and this has resulted in more reliable advice in the rabi season. (d) Visits to contact farmers decrease with the life of the project, while they increase for non-contact farmers. With increasing project life, there is a sizeable increase in the absolute number of farmers meeting with extension agents. - 33 - (e) VEWs play a more important role in the dissemination of information in areas covered by T&V extension than they do in non-T&V settings; in both situations they are relatively more important to large farmers than to small. (f) The VEW and radio are probably complementary information sources. (g) "Other farmers" qua information sources appear to play a role consistent with a two-step communication flow. In a non-T&V setting they are the most important information source. In areas covered by T&V extension they are the major information source for non-contact farmers and a relatively minor one for contact farmers. These results hold for both large and small farmers. (h) VEWs become increasingly important information sources the more expensive or complicated an agricultural practice becomes; their role is somewhat greater in the case of large farms. (i) Yields in farms that rely on the VEW as the main source of information are higher than in farms that rely mainly on other sources of information. The other sources do not appear to differ greatly from one another, but any source of information appears to be better than receiving no advice. As more information starts to emerge from T&V projects in other countries, future research should focus on comparisons between expeciences in differing socio-economic and cultural environments. Such comparisons may provide insights into the ways in which extension systems may be further adapted and improved. - 34 - FOOTNOTES 1/ By 1984, it had been adopted by about 40 countries in Eastern and Western Africa, South and Southeast Asia, the Middle East, Europe, Central and South America (Benor and Baxter, 1984, p. 4). 2/ The terms "extent3ion agent" and "village extension workers (VEW)" are used interchangeably throughout the remainder of the paper. 3/ Two geographically adjacent districts - one with T&V and one without - in the Indian states of Haryana and Uttar Pradesh, respectively, were chosen. A sample survey amongst 972 farmers was conducted. For details, see Feder and Slade, 1984a, pp.7-8. 4/ In a recent case study of extension in India [Howell (1984)], a similar distinction between supply and demand is made. However, the indicators used for supply and demand are different, and no distinction is maintained between contact farmers and other farmers. 5/ The statistical test used is based on the large sample normality of the test statistic: Z (pi - ~P2)/ Vi(1-w)(1/Ni + 1/N2) where 7 is the proportion of farmers with a certain charateristic within sample 1, Ni is the sample sizes, .7 the population proportion.. 6/ This can be verified by calculating the weighted average for all farmers (contact and non-contact) of the proportion of farmers with "1no visit", using the weights .1 and .9 for contact and non-contact farmers, respectively. 7/ The "information source ratio" is merely the ratio of the number of times a VEW is cited as the main source of information to the number of times "other farmers" is cited. 8/ This transformation does not at all alter our results; the same regressions run for non-logarithmic data produced identical results. 9/ A test of the difference between the mean yield indices from farmers who depended on the VEW and farmers who depended on "other farmers" for advice produced a 't' statistic of 3.14, which was significant at 99 percent, suggesting that yields are significantly higher for farmers who depend on VEWs as opposed to "other farmers". - 35 - REFERENCES 1. Benor, Daniel and Michael Baxter, 1984, "Training and Visit Extension", Washington, D.C.: The World Bank. 2. von Blanckenburg, P., C. Sivayoganathan, M.W.A.P. Jayatilake and K.K. Navaratnam, 1980, "Some Observations on Agricultural Extension: The Training and Visit System in Sri Lanka", Faculty of Agriculture, University of Perdenia, Mimeo. 3. von Blanckenburg, Peter, 1982, "The Training and Visit System of Agricultural Extension - A Review of First Experiences", quarterly Journal of International Agriculture, Vol. 21, No. 1. 4. Cernea, Michael M., "Sociological Dimensions of Extension Organization: The Introduction of the T&V System in India", Extension Education and Rural Development Vol. 2: International Experience in Strategies for Planned Change, Bruce R. Crouch and Shankarian Chamala, eds., John Wiley and Sons, Ltd., Chichester, 1981. 5. Cernea, Michael M., J.K. Coulter, and J.F.A. Russell, eds., 1983 "Agricultural Extension by Training and Visit. The Asian Experience." A World Bank and UNDP Symposium. Washington, D.C. The World Bank. 6. Feder, Gershon and Roger Slade, 1984a, "Contact Farmer Selection and Extension Visits - The Training and Visit Extension System in Haryana, India", Quarterly Journal of International Agriculture, Vol. 23, No. 1. 7. Feder, Gershon and Roger Slade, 1984b, "Acquisition of Information and Adoption of New Technology", American Journal of Agricultural Economics, Vol. 66, No. 3. 8. Feder, Gershon and Roger Slade, 1984c, "Aspects of the Training and Visit System of Agricultural Extension in India: A Comparative Analysis". Staff Working paper No. 656, The World Bank, Washington, D.C. 9. Howell, John, 1982a. "Managing Agricultural Extension: The T&V System in Practice", Ovarseas Development Institute (London), Agricultural Administration Unit Discussion Paper No. 8. 10. Howell, John, 1982b, "Strategy and Practice in the T&V System of Agricultural Extension", Agricultural Administration Unit, Overseas Development Institute (London). Discussion Paper No. 10. - 36 - 11. Howell, John, 1984. "Small Farmers Services in India: A Study of Two Blocks in Orissa State", Overseas Development Institute (London), Agricultural Administration Paper No. 13. 12. Hutanuwatr, Narong, Suchint Simaraks, Krirkkiat Phipatseritham, Chalong Bunthamcharoen, Arnone Yamtree, 1982. "Socio-economic Constraints in Rain-fed Agricultural Production in the Lower North-East Thailand", Faculty of Agriculture, Kohn Kaen University, Thailand. 13. Jaiswal, N.K., 1983, "Transfer of Technology under T&V - Problem Identification", Background Papers: Workshop on Management of Transfer of Farm Technology under the Traning and Visit System, National Institute of Rural Development, Hyderabad. 14. Moore, Michael, 1984, "Institutional Development, The World Bank and India's New Agricultural Extension Programme", The Journal of Development Studies, Vol 20, No. 4. 15. National Council of Applied Economic Research (NCAER), 1983, "Evaluation of the Training and Visit Extension Service in Madlya Pradesh", New Delhi. 16. Perraton, Hilary, D.T. Jamison, J. Jenkins, F. Orivel and Laurence Wolff, 1983, "Basic Education and Agricultural Extension - Costs, Effects, Alternatives", Staff Working Paper, No. 564, The World Bank, Washington, D.C. 17. Reports on Monitoring and Monitoring-cum-Evaluation Surveys of T&V System. Various State Governments, India, 1979-1983. 18. Shingi, Prakash M., Sanjay Wodwalar and Gurindar Kaun, 1982, "Management of Agricultural Extension: Training and Visit System in Rajasthan." C.M.A. Monograph, No. 96, Indian Institute of Management, Abmedabad. 19. Singh, R.N., 1983, "T&V in Chambal Command Area (Kota District): Some Observations" in Background Papers: Workshop on Management of Transfer of Farm Technology under the Training and Visit System, National Institute for Rural Development, Hyderabad. 20. Slade, Roger and Gershon Feder, 1981, "The Monitoring and Evaluation of Training and Visit Extension in India: A Manual of Instruction." The World Bank, Washington, D.C. Mimeo DISCUSSION PAZ-ERS AGR/Research Unit Report No.: ARU 1' Agricultural Mechanization: A Comparative Historical Perspective by Hans P. Binswanger, October 30, 1982. Rport No.: ARU 2 The Acquisition of Information and the Adoption of New Technology by Gershon Feder and Roger Slade, September 1982. Report No.: ARU 3 Selecting Contact Farmers for Agricultural Extension: The Training and. Visit System in Haryana, India by Gershon Feder and Roger Slade, August 1982. Report No.: ARU 4 The Impact of Attitudes Toward Risk on Agricultural Decisions in Rural India. by Hans P. Binswanger, Dayanatha Jha, T. Balaramaiah and Donald A. Sillers May 1982. Report No.: ARU 5 Behavioral and Material Determinants of Production Relations in Agriculture by Hans P. .Binswanger and Mark R. Rosenzweig, June 1982, Revised 10/5/83. Report No.: ARU 6 The Demand for Food and Foodgrain Quality in India by Hans P. Binswanger, Jaime B. Quizon and Gurushri Swamy, November 1982. Report No.: ARU 7 Policy Implications of Research on Energy Intake and Activity Levels with Reference to the Debate of the Energy Adequacy of Existing Diets in Development Countries by Shlomo Reutlinger, May 1983. Reuort No.: ARU 8 More Effective Aid to the World's Poor and Hungry: A Fresh Look at United States Public Law 480, Title II Food Aid by Shlomo Reutlinger, June 1983. Reoort No.: kRU 9 Factor Gains and Losses in the Indian Semi-Arid Tropics: A Didactic Approach to Modeling the Agricultural Sector by Jaime B. Quizon and Hans P. Binswanger, September 1983, Revised May 1984. Revort No.: ARU 10 The Distribution of Income in India's Northern Wheat Region by Jaime B. Quizon, Hans P. Binswanger and Devendra Gupta, August 1933. Revised June 1984. ReDort No.: ARU 11 Population Density, Farming Intensity, Patterns of Labor-Use and Mechanization by Prabhu L. Pingali and Hans ?. Binswanger, September 1983. Recort No,: ARU 12 The Nutritional Impact of Food Aid: Criteria for the Selection of Cost-Effective Foods by Shlomo Reutlinger and Judit Katona-Apte, September 1983. -2- Discussion Papers (Cont'd.) Report No.: ARU 13 Project Food Aid and Equitable Growth: Income-Transfer Efficiency First! by Shlomo Reutlinger, August 1983. Report No.: ARU 14 Nutritional Impact of Agricultural Projects: A Conceptual Framework for Modifying the Design and Implementation of Projects by Shlomo Reutlinger, August 2, 1983. Report No.: ARU 15 Patterns of Agricultural Protection by Hans P. Binswanger and Pasquale L. Scandizzo, November 15, 1983. Report No.: ARU 16 Factor Costs, Income and Supply Shares in Indian Agriculture by Ranjan Pal and Jaime Quizon, December 1983. Report No.: ARU 17 Behavioral and Material Determinants of Production Relations in Land Abundant Tropical Agriculture by Hans P. Binswanger and John Mctntire, January 1984. Report No.: ARU 18 The Relation Between Farm Size and Farm Productivity: The Role of Family Labor, Supervision and Credit Conscraincs* by Gershon Feder, December 1983. Report No.: ARU 19 A Comparative Analysis of Some Aspects of the Training and Visit System of Agricultural Extension in India by Gershon Feder and Roger Slade, February 1984. Report No.: ARU 20 Distributional Consequences of Alternative Food Policies in India by Hans P. Binswanger and Jaime B. Quizon, August 31, 1984. Report No.: ARU 21 Income Distribution in India: The Impact of Policies and Growth in the Agricultural Sector, by Jaime B. Quizon and Hans P. Binswanger, November 1984. Report No.: ARU 22 Population Density and Agricultural Intensification: A Study of the Evolution of Technologies in Tropical Agriculture, by Prabhu L. Pingali and Hans P. Binswanger, October.17, 1984. Report No.: ARU 23 The Evolution of Farmining Systems and Agricultural Technology in Sub-Saharan Africa, by Hans P. Binswanger and Prabhu L. Pingali, October 1984. Report No.: ARU 24 Population Density and Farming Systems - The Changing Locus of Innovations and Technical Change, by Prabhu L. Pingali and Hans P. Binswanger, October 1984. Report No., ARU 25 The Training and Visit Extension System: An Analysis of Operations and Effects, by G. Feder, R.H. Slade and A.K. Sundaram, November 1984.