The Role of Agricultural Extension Services on Rice Production in Bangladesh

The actual situation of extension contact is examined with a view to evaluate the agriculture extension in Bangladesh. Ten villages by taking two villages from five sub districts of the Gazipur district of Bangladesh were selected for the study. Of the two villages, one village is located comparatively nearer to the sub district headquarter while the other one is located comparatively away from the sub district headquarter. Total 1000 farmers by taking 100 from each village were surveyed through pre structured questionnaire. The study is done through regression analyses. It is found that the extension contact coefficient of 1-2 times is significant and positive in the nearer villages. However, the roles of extension contact coefficients are unclear in the far villages since their coefficients are insignificant. Evidences show from the study that 1-2 times number of times extension contacts is contributing positively and significantly for the rice production. It is concluding that there is enough scope to employ more extension agents in the villages because many farmers are still being neglected from the extension contact.


Introduction
Bangladesh saw increased food production and relatively stable prices of rice and vegetables for almost the entire 2014. It was projected that overall production would rise because of increased acreage of hybrid varieties [2] though such expectation yet been achieved. Recently food ministry says that the food grain production to fall due to a drop in acreage [3]. For example, boro a well known rice crop cultivated on 46.85 lakh hectares, the lowest since 2008-09. Another crop aus was targeted to produce 24.75 lakh tons while it fell down 22.89 lakh tons during 2016 [3].
The above dual picture of the rice production in Bangladesh shows a grim picture of rice production which is the chief food of the people of Bangladesh. There are many reasons for the inconsistency of rice production in Bangladesh. It is said that farmers fail to recover their costs of production owing to the low paddy price [3]. The department of agriculture extension can ensure farmers in order to increase rice production and price of the rice. It is because agriculture extension has one essential object is to raise farm production by providing necessary counseling to the crop producers. It is possible by increasing the number of times extension contacts between the extension agents and farmers [7].
Previous studies investigated the possible number of times of extension contacts on the crop productivity in terms of income [8]. Compare to this, the study of extension impact on rice income and rice production can be recognized [7][8]. However, results of these studies are unclear because it failed to indicate the possible number of times extension contacts that may be useful to generate higher rice productivity either in terms of income or physical quantities. It is also seen in [7] that the impacts of extension contacts are particularly great in the farms which are comparatively nearer to the urban areas. It means agriculture extension in Bangladesh fail to disseminate the extension supports to the all farmers. [9] and [12] also observed that extension services covered farmers are benefitted than non covered farmers of extension services. But it is also not clear from those studies about the possible number of times extension communication which may be required by the farmers. The object of the present study is to identify the actual number of times extension contacts which may able to increase the higher rice productivity.
Like many other countries, the food prices of Bangladesh may double [11]. Food prices should be decreased if it is possible to raise the food production. At the same time, it may possible to export food crops such as rice to rice deficit country by producing more rice. This is possible since the growth rate of population in Bangladesh gradually decreases. Therefore, the present study can able to assist the policy makers about the role of agriculture extension in order to increase rice production.

Sampling Design
The selection of the Gazipur district, upazilas (sub districts), villages and sample respondents were done purposively. There were some salient features in the selection procedure. First one, the selected district includes some important infrastructures, such as Bangladesh Agricultural Research Institute and Bangladesh Rice Research Institute, etc. Secondly, total number of selected villages was ten by selecting two villages from five upazilas, namely: Sadar(central), Sreepur, Kapasia, Kaliakoir and Kaliganj. Of the two villages in each upazila, one village is selected comparatively nearer to the upazila headquarters and the other one is selected comparatively away from the upazila headquarters. The selected nearer villages were Samantapur (Sadar), Bagnahati (Sreepur), Dushya Narayanpur (Kapasia), Katalia (Kaliakoir), and Poinlanpur (Kaliganj). The selected villages which were comparatively away from the upazila headquarters, namely, Bara Bhabanipur (Sadar), Saitalia (Sreepur), Noyanagar (Kapasia),PoshimChandpur (Kaliakoir), and Bhatgati (Kaliganj). Thirdly, the total households were more than one hundred in the selected villages [5]. It was then decided to collect one hundred samples from each village. The total numbers of investigated farmers were one thousand (2 villages x 5 upazilas x 100 farmers). Primary data was collected by using survey method and personal interviews were conducted through pre-tested questionnaires with a view to collecting data. The survey was administered with the help of staff of BARI in 2002. Lastly, each upazila has some characteristics: Sadarupazila is completely urban type; Sreepur, Kapasia and Kaliganjand Kaliakoirupazilas are rural type and headquarters of these upazilas are the only urban areas [5].

Conceptual Framework
Many of the previous researches used the productivity index representing the amount of production per unit of farm land, the value added of production, which is found by deducting production costs from gross income. By using that index, it is possible to convert the specific quantities of products into given amounts of money to add up; therefore it represents a considerable analytical benefit. The method of settling the type of variables from which the index is determined, expected to be discussed [8].
As is commonly used in analyzing production, chemical fertilizer, farm buildings, irrigation facilities, family and hired labours should be considered as important investment functions [6]. [8] considered crop income per unit of land as dependent variable and chemical fertilizer cost per unit of land, irrigation cost per unit of land, experience of farmers, farm area, number of times extension contact as independent variables. Therefore, it summarized the model, Ln crop income=f (ln chemical fertilizer, ln irrigation, ln experience, ln farm area, lnlabour, extension contact dummy 1 , extension contact dummy 2 ). [7] interpreted rice yield as dependent variable, while age of the farm household head, number of family earners in the household, number of times extension contact, proportionate effect of flood to crop land, distance from farm land to market, actual size of cultivated land, per unit cost of chemical fertilizer, per unit land cost, per unit irrigation cost, village dummy were taken as independent variables. The production function was solving by applying the ordinary least squares.

Empirical Model
The model applied here is the input-output model. The heart of the input-output model is the concept of the production function [Y=f (Capital, Labour)] which helps us in understanding the role of important variables like capital and labour in determining the crop productivity. But only two factors have no reflection on the productivity of any crop. Therefore, based on related past studies [6][7]and logical analysis, some important explanatory variables are considered in this study namely age of the farm household head (ag), years of schooling of the head of the household (ed), number of family members (Fm), number of family earners in the household (fea), 1-2 number of times extension contact received by the farmer for the sample crop season (etdummy1),3 times and above number of times extension contact received by the farmers for the sample crop season(etdummy2), proportionate effect (%) of flood to crop land (fec), distance from farm land to market in miles (mr), actual size of cultivated land in acre (fs), total money spent for irrigation (irr), total cost spent for chemical fertilizer(ch), total labour cost (lab), village dummy (vdummy) = 1 if near village; otherwise = 0. Dependent variable is the rice production per unit of farm land (Ric). Table 1 represents the summary statistics of variables.
Data have been analyzed through frequency, percentage, mean, standard deviation and regression analyses. The productivity expressed in terms of rice production per unit of farm land is as follows, Ric=f(ag,ed,fm,fea,fec,mr,fs,lab,ch,irr,etdummy 1 ,etdummy 2 , vdummy).
To understand the quantitative relationship individually between rice yield and some selected inputs namely extension contacts (et), years of schooling of farm household head (ed), irrigation cost (irr) and cost of chemical fertilizer (ch), data were fitted with a linear regression model of the form: Y= a+bx, where Y=rice yield, b=coefficient of the input variable, x=relevant input and a=constant [4].
In the objective of this research, the most important independent variable is that of the activities of the agricultural extension services. In Bangladesh Training & Visit (T&V) system, farmlands are divided into blocks and the T&V workers (Sub assistant agriculture officer) target the representative farmers of the different blocks, who are referred to as "contact farmers" [8]. Although the T&V workers can directly get in touch with ordinary farmers, they mainly train the contact farmers, who afterwards transmit the training results to the other farmers, in a progressive system [8]. Considering this situation in Bangladesh, the current paper used the frequency of contacts on the basis of actual number of times contacted between ordinary farmers and T&V workers or contact farmers. Note that the combination of T&V workers and contact farmers is hereinafter referred to as "extension agents" (ibid). Most of the farmers of Bangladesh are either illiterate or unskilled. Thus with the knowledge derived from extension services through extension contact, farm operators may increase their farm productivity [8] and [10]. Explanations of the selection background of other explanatory variables can be found in relevant literatures [8].

Results and Discussion
The results of regression are presented in Table 2. The results suggest that all the selected variables have positive effects on the yield of rice except few variables.
It is found that the extension contact coefficient of 1-2 times (0.110) is greater than extension contact coefficient 3 times or above 3 times (0.023). The coefficient for extension contact 1-2 times of nearer villages is also particularly great (0.197), compared to villages away from the sub district headquarters (-0.006). According to [10], frequency of one and two operations per year between extension agents and farmers generate a high contribution to yield. [8] found that the extension contact coefficient for three or more than three times per year was 0.353 while the coefficient for extension contact of 1-2 times per year was 0.234. [7] suggests that in Bangladesh, the higher number of extension contacts between extension agents and farmers seem to be more effective in the case of nearer villages of urban areas. Conversely no such clear effects are found in the comparatively far villages. There is a shortage of extension workers in Bangladesh. Usually, extension workers in the country visit the farmers of progressive villages due to better road communication and transport facilities than in the non progressive villages. So, farmers of the non progressive villages fail to receive the extension contact adequately. This result agrees with [10] but disagrees with [7] and [8].
It is found ( Table 3) that the use of extension contact resource is particularly lower (AR 2 0.6%) compared to the irrigation and chemical fertilizer with an exception. Therefore, it is cleared from the study that many farmers are being neglected from the extension support in the country.

Conclusion
It is evident from the study that farmers of Bangladesh can increase rice yield per year by receiving 1-2 times extension contact from the extension agents especially in those areas which are nearer to the urban areas. However, the role of extension agents of the remote type areas is confusing since their extension coefficients are insignificant. Further investigation is necessary for this.
Though the overall impact of extension contact coefficient of 1-2 times is positive and significant, policy should be formulized so that farmers of remote type areas can receive extension contact reasonably.
It should be investigated by the researchers and policy makers of the country about the expected number of times extension contacts between the farmers and extension agents in the remote areas of Bangladesh. This is very important with a view to increase the rice production per unit of farm land for the whole country.
It is also observed from the study that most of the farmers are being neglected from the extension contact. Recently many extension programs are being operated by the department of agricultural extension in Bangladesh such as school of farmers with a view to provide training, technical knowledge on farming ton increase productivity [1]. But it is assumed that such programs are covered within the farm rich areas [1]. Therefore, it should be covered by employing many extension agents regarding out the type farm rich and farm poor areas.
Policy makers would rejuvenate the agriculture extension policy by consolidating the outcomes of the study either in Bangladesh or elsewhere.