Impact of Farmer Fields Schools on Technical Efficiency of Tobacco growers: A Case Study of District Swabi Pakistan

Mubassir Zubair¹, Nasrullah² and Robina Karim^3*
*Correspondence: Robina Karim, Department of Agriculture and Applied Economics, the University of Agriculture, Peshawa, Pakistan, Email:

Keywords

FFS • Tobacco • Efficiency • Stochastic frontier production • Propensity score matching

Introduction

Agriculture contributes significantly in meeting the basic needs of individuals. The general role of agriculture extension in crop productivity and the specific role in enhancing the livelihoods of people in rural areas are recognized by different international organizations and agencies [1]. Tobacco is one of the agricultural commodity which is widely produced around the world however the major tobacco producing countries are India, Brazil, China, USA, Indonesia, Pakistan, Turkey and Zimbabwe. Pakistan ranks 6th in the production of tobacco (GOP, 2014). In Pakistan amongst all food and nonfood crops, the tobacco crop has features of its own. Tobacco is an important cash crop not only in Pakistan but throughout the world. The reason being that about 30% of the government revenue comprises of receipts from the CED (Custom and Excise Duties) on tobacco [2]. Out of all the provinces, KPK is famous for tobacco production and particularly district Swabi for the production of the Flue Cured Virginia tobacco; as its agronomic and environmental conditions are suitable for its production. In agriculture sector the improvements in efficiency and introduction of new technology can enhance productivity. Agricultural productivity in the short-term can be enhanced by improvements in efficiency as the acquisition rate of new technology is quite low in Pakistan [1]. In order to attain the prolong growth in efficiency, productivity and agriculture differentials have to be diminished by enhancing the managerial skills of the farmers communities and development of infrastructure [3,4]. In this scrim, agricultural production’s efficiency measurement is the important sketch in developing nations.

A small number of studies have been conducted on estimating the technical efficiency of tobacco crop in Pakistan [5]. Therefore there is a need to conduct study on investigating the technical efficiency of Pakistan’s tobacco crop. The present research will assist the farmers to locate factors that influence the technical efficiency of tobacco production. Moreover for increasing the output it is necessary that the tobacco growers have better knowledge and skills about farming and proper use of machinery, sowing seed, fertilizer, irrigation and harvesting etc. Extension methods play a vital role in transmission of skills and knowledge.

In so far as the history of Pakistan’s agriculture extension is considered, the Training and Visit Programme was the first Programme to be used. Many studies have shown that the Training and Visit Programme is more effective than other Programmes for agricultural extension [6,7]. The traditional agricultural extension system was facing many issues because of limited coverage, reliance on the contact farmers, sampling biasness, inadequate management and technical skills of the extension staff [8,9]. As a result of all these drawbacks; the managers of the extension system had to substitute with another strategy in order to solve the issues of farmers [10].

The Farmer Field School (FFS) methodology of agricultural expansion was the most appropriate one given the state of agricultural system globally. Owing to the success of FFS in Indonesia, many other South Asian countries introduced such a Programme [11]. FFS was introduced in Pakistan in order to educate the wheat and cotton growers about the Integrated Pest Management (IPM) practices [12].

Due to its success, the Government of Punjab also adopted it by the name of “Fruit and Vegetable Development Project”, the purpose of which was to train the vegetable and fruit farmers through the approach of FFS [13]. The Programme was adopted in all the districts of Punjab where fruit and vegetables were produced including Sargodha, Vehari and Sheikhupura that specializes in the production of citrus. Despite the large investments in Asia that have been made in FFS few studies have been conducted on examining their impacts [14-17].

Given this backdrop, the study at hand is conducted to know about the effect of FFS on efficiency of tobacco growers in the district Swabi because FFS program has not been evaluated for tobacco crop in Pakistan.

Materials and Methods

This research was conducted for the district Swabi in KPK province of Pakistan. In the underlying study primary data was employed. The unit of primary data was tobacco producers. Data collection was done through a well-designed questionnaire and the various inputs like seed, labor, fertilizer, tractor and labor hours were used to obtain the output. Socioeconomics characteristics were also taken in the study.

Model specification

Technical efficiency is measured with 2 types of techniques specifically parametric (Stochastic Frontier Analysis) and nonparametric (Data Envelopment Analysis). These two approaches have their own pros and cons. The main benefit of DEA is that apart from input and output quantities no other information is required.

Nevertheless, the estimates of DEA are sensitive to noise and errors because it characterizes all deviations from the frontier to inefficiencies and is deterministic. The strong point of SFA is that noise in data is also taken into account and statistical testing of hypothesis regarding degree of inefficiency and production structure. Therefore Stochastic Frontier Analysis (SFA) is utilized for formulating the efficiencies of tobacco farmers. In the stochastic frontier model, it is assumed that the stochastic production function bounds the output.

The stochastic frontier function has a composite error term that is farmer’s inefficiency and random errors [18]. Stochastic frontier production function that is applied to measure the efficiencies of tobacco growers is described mathematically as below:

i=1………n

Whereby; Yi is the production, Xi denotes the inputs, β is the production function’s parameter whose value is unknown and ε represents composite error term that consists of 2 components as given below (Figure 1), Source: Adopted [19].

where;

vi = is a symmetric element that depicts the output changes caused by factors which are out of farmer’s control like plant disease, earthquake, breakdowns and climatic conditions.

μi = is an asymmetric element that captures the deviation in the output caused by the inefficiency factors of farmers.

For the estimation of technical efficiency, following stochastic frontier production function of Cobb-Douglas form was used:

Where,

Seed = No of seed/acre

Labor = Labor days used/acre

Irrigation = No. of irrigation for whole season

Pesticides = Pesticides used in ML/ acre

Fertilizer = Amount of fertilizer (Kg/acre)

FYM = Amount of Farm Yard Manure (Kg/acre)

Tractor = No of tractor Hours/acre

In order to investigate the factors that contribute to technical inefficiency, the stochastic frontier and inefficiency model was estimated mutually through the use of one stage maximum likelihood estimation (MLE) method utilizing frontier version 4.1 as follows [20].

whereby Xi is a vector of variables consisting of the Age of Farmers (Years), Farmers’ Experience (years), Farmers education level (years), Tenurial status, number of extension visit, Land of farmer, Income of farmer, Occupation of farmer and access to credit.

Farmer field school and efficiency

OLS and propensity score matching: OLS is the basic regression design. However there are few issues with it for instance biasness of omitted variable. The coefficient will be biased if relevant variables are omitted. Secondly, controlling for variables that are affected by the variable of interest will produce prejudiced coefficient. Hence instead of doing an OLS regression, matching methods can be used. Matching method is desirable to be used when the variable takes on only two values. OLS and matching method are similar if the treatment effects are constant. However if treatment effects are distinct, they will vary as distinct weighting schemes are employed. OLS is efficient under the assumption that the treatment effect is constant, so observations are weighed by the conditional variance of the treatment status [21-25]. The treatment group and the control groups may be very different in matching and OLS. Since there are imbalances between the groups hence logistic regression is employed to cater for these discrepancies. The main advantage of a logistic regression is that it can be used to control for many variables at once. An additional method to control for imbalances is the propensity score, which is the conditional probability of a subject’s receiving a particular treatment given the set of variables. For calculation of a propensity score, the variables are used in a logistic regression to predict the exposure of interest, without the inclusion of outcome.

A program that is implemented to some groups while other groups receive no treatment is known as treatment evaluation. Unlike the control group, the treated group receives the treatment. The objective of treatment evaluation is to evaluate the effect of a treatment on the treated group while using control group as a benchmark. There are 2 types of studies, first one is controlled experiments where assignment into treated and control groups is random [26-29]. However we usually have observational studies where the assignment into treated and control groups is not random i.e. some individuals decided to participate in the program while others don’t. Since people who participate in the program are different than those who did not hence their outcomes cannot be directly compared. Thus it is necessary to first match them as much as possible in order to compare their outcomes. This leads us to the propensity score matching methodology.

Firstly the observations are assigned into two groups: the treated group that received the treatment and the control group that has not. In our study those who are registered in the FFS program are the treated group and those who are not registered with the FFS program are the control group. There are various types of propensity score matching such Kernel, Nearest Neighbor Matching, Radius and Stratification or Interval Matching, Inverse-probability Weights and Inverse-probability Weight Regression Adjustment. After matching we found the treatment effect (The variation amid the outcome of control and treated observations is called the Average Treatment Effect (ATE)) and compare the outcome y between treated and control group.

Conclusion and Recommendation

The main theme of this study was to estimate the technical efficiency of tobacco production and to identify the factors that influence the technical inefficiency and the impact of Farmer Field School (FFS) on technical efficiency. Maximum likelihood estimation technique was used for the stochastic frontier Cobb-Douglas production function to estimate the technical efficiency. The regression analysis of tobacco yield shows that seed quantity, farm yard manure, Total Tractor hours, Labor man days, total pesticides and total seed quantity significantly affects tobacco production. The means difference between treated and control group as well as the probit regression analysis revealed that occupation, tenure status and access to credit affect the likelihood of the farmer being registered in FFS. Moreover different matching techniques were used in order to estimate the propensity score. It was found that the extension visits which is a source of knowledge dissemination among the farmers, greatly contributed in enhancing the technical efficiency.

Since FFS has been found to be effective in promoting the technical efficiency of tobacco production in the study area, thus the study suggests that the government should promote the FFS programme. In other words, the farmers should be encouraged to register in FFS. Moreover, workshops and seminars should be conducted to enhance the exposure of the farmers and ease their mindsets to remove the prejudice of the farmers regarding the third-party interference. In addition, the FFS programme should be made part of the agriculture policy in the province. The improvements in the knowledge on scientific cultivation of tobacco growers through their participation in field days, trainings and contacts with extension workers can help in achieving higher productivity through correct adoption of the recommended production technologies and thereby high gross income.

References

1. Javed, Mohammad Ishaq, Adil Sultan Ali, Javed Muhammad Siddique and Hassan Sarfaraz “Efficiency Analysis of Rice-Wheat System in Punjab, Pakistan.” J Agric Sci 45 (2008): 95–100.
2. Rahman, Qazi Latif Ur, Sajjad Muhammad, Khan Noor and Shah Shen et al. “Costs and net returns of tobacco production in district swabi (Khyber Pakhtunkhwa) pakistan.” Interdiscip J Contemp Res Bus 3 (2011): 160-171.
3. Ghura, Dhaneshwar and Richard E. “Just, Education, infrastructure and instability in African agriculture: Implications for structural adjustment programmes.” Econ Dev 1(1992): 85-105.
4. Pingali, Pingali L and Heisey Paul W. “Cereal crop productivity in developing countries.” Res Agric Econ (1999): 99-03.
5. Saddozai, Khurram Nawaz, Nasrullah Muhammad and Khan Noor P. “Stochastic frontier production analysis of tobacco growers In district mardan, pakistan”. J Agric Res 28 (2015): 346–53.
6. Hussain, Syed Sajidin, Byerlee Derek and Heisey Paul W. “Impacts of the training and visit extension system on farmers’ knowledge and adoption of technology: Evidence from Pakistan.” J Agri Econ 10 (1994): 39–47.
7. Lodhi, TE. “Need for paradigm shift from top down to participatory extension in the Punjab, Pakistan: perception of farmers.” (2003).
8. Davidson Andrew P, Ahmad Munir and Ali Tanvir. “Dilemmas of agricultural extension in pakistan :Food for thought.” Agric Res 116 (2001).
9. Ashraf, Ijaz, Muhammad Sher and Chaudhry Khalid Mahmood. 2007. “Effect of decentralization on linkage among research, extension and farming community.” Pak J Agri 44 (2007): 660–663.
10. Jurgen, H, Chuma E, Murwira K and Connolly Mike. “Learning together participatory extension. A guide to an approach development in harare.” (2000).
11. Braun, Arnoud Jiggins Janice, Röling Niels and Berg Henk van den et al. “A global survey and review of farmer field school experiences table of contents.” Review Literature And Arts Of The Americas (2006).
12. Duveskog, D and Friis-Hansen Esbern. “Farmer field schools: a platform for transformative learning in rural Africa.” (2008).
13. Government of Punjab Agriculture department. (2012).
14. Tripp, Robert, Wijeratne Mahinda  and Piyadasa V Hiroshini. “What should we expect from farmer field schools? A sri lanka case study.” World Dev 33 (2005): 1705–1720.
15. Habib, Mubashir, Zafarullah Muhammad, Iqbal Mahmood and Nawab Khalid et al. “Effect of farmer field schools on sugar cane productivity in malakand agency.” Sarhad J Agric 23 (2007): 1133–1137.
16. Saddozai, Khurram Nawaz, Saboor Abdul, Azeem Khan Muhammad and Jan Dawood et al. “Analyzing the impact of farmer field school on technical efficiency of cotton growers in southern districts of punjab- pakistan : stochastic production frontier approach.” Sarhad J Agric 29 (2013): 291-298.
17. Butt, Tahir Munir, Gao Qijie and Hussan Muhammad Zakaria Yousuf. “An analysis of the effectiveness farmer field school (ffs) approach in sustainable rural livelihood (srl): the experience of punjab-pakistan.” Agric Sci 6 (2015): 1164–1175.
18. Battese, GE and Coelli Timothy J, “A model for technical inefficiency effects in a stochastic frontier production function for panel data. Emp Econ 20 (1995): 325-332.
19. Kumbhakar, Subal C and Sarkar Subrata. “Deregulation, ownership, and efficiency change in indian banking: an application of stochastic frontier analysis.” Marketing Science 23(3): 419–428.
20. Coelli, Timothy J and Battese GE. “A model for technical inefficiency effects in a stochastic frontier production function for panel data.” Emp Econ 20 (1996): 325–332.
21. Huluka, Admassu Tesso. “The impact of farmer field school training on farmers technical efficiency: evidence from small holder maize farmers, oromia, ethiopia.” EJTDS 2 (2015): 1-28.
22. Davidson, Andrew and Alex Rivera. “privatization and the crisis of agricultural extension in pakistan: caveat emptor.” JRS 2(2004): 35–41.
23. Goldey, Patricia, Breton, Shelley Le, Martin adrennie and marcus rachael. “approaches to address gender specific needs in relation to access to technological change.” Agri Syst 55 (1996): 155-172.
24. Pakistan economic survey. Government of pakistan. (2014).
25. Hasan, Md Faruq, Imai Katsushi, and Sato Takahiro.2012 “Impacts of agricultural extension on crop productivity, poverty and vulnerability: evidence from uganda.” (2012):1–33.
26. Iqbal, Mahmood, Nawab Khalid, Sajjad Mahammad and Pervaiz Urooba. “Farmers field schools and rice productivity an empirical analysis of district malakand.” S J Agric 28 (2012): 143-148.
27. Mahmood, Iqbal, and Nawab Khalid. “Farmers field schools and bitter gourd productivity: an empirical analysis of district charsadda, khyber pakhtunkhwa - pakistan.” S J Agric 29(2013): 599–605.
28. Phillips, Joseph M and Marble Robert P. “Farmer education and efficiency : a frontier production function approach.” Econ Educ Rev 5(1985): 257–64.
29. Srinivas, T, Rizvi Sjh, and Kadian M. “Technical efficiency of seed potato farmers of badakshan province of afghanistan.” J Potato Res 39(2012): 118–27.