American-Eurasian J. Agric. & Environ. Sci., 1 (1): 1610-1616, 01 ISSN 1818-6769 IDOSI Publications, 01 DOI: 10.589/idosi.aejaes.01.1.1.1918 Estimation of Technical Efficiency of Irrigated Rice Farmers in Niger State, Nigeria J. Ahmadu, and G.O. Alufohai Department of Agricultural Economics and Extension Services, Faculty of Agriculture, University of Benin, Benin City, Nigeria Abstract: The study estimated e technical efficiency of irrigated rice farmers in Niger State of Nigeria. A multistage sampling technique was employed for e selection of respondents for e study. A total of 60 respondents selected at random from e study area were administered structured pre-tested questionnaires. Data analysis was done using descriptive statistics, stochastic frontier production function and inefficiency model. The results showed at e irrigated rice production in e study area was practiced under small-scale production system. The average observed output of e farmers and frontier output were 3.63 and 4.11 metric tons per hectare, respectively. The presence of technical inefficiency effect in e irrigated rice production as indicated by e log-likelihood ratio test (7.044) was not significant. The insignificant presence of inefficiency in e farmers production was confirmed by e gamma coefficient (0.003) which indicated at only 0.3% of e deviation of output from e production frontier was attributed to technical inefficiency. The average technical efficiency of e farmers was 9%, leaving a gap of 8% for improvement. Despite e high level of technical efficiency of e farmers, eir average observed output and e output of e most efficient farmers per hectare given e production technologies employed were lower an e maximum potential yield of e irrigated rice by 3.37 and.89 metric tons per hectare, respectively, indicating low levels of production technologies utilized by e farmers. Thus, boosting irrigated rice production in e study area requires improvement in e levels of technologies used by e farmers. Key words: Stochastic frontier Technical Efficiency Irrigated rice Farmers Nigeria INTRODUCTION Rice is e second most important cereal in e world after wheat in terms of production [1]. It is an ancient Agriculture is still e most important sector in e staple food crop consumed by more an half of e world economy of Nigeria despite e fact at it is not accorded population - about 4.8 billion people in 176 countries. It is e required priority. Priority shifted to oil after its e most important food crop for over.89 billion people discovery in e early 1970s wi marked deterioration in in Asia, over 40 million in Africa and over 150.3 million e performance of e agricultural sector. However, over people in America []. In Nigeria, rice consumption has 70% of e livelihood of e Nigerian populace today increased tremendously (+10% per annum), emerging from depends on agriculture. Thus, emphasis is gradually being a food reserved for ceremonial occasions to a major being shifted back to boosting agricultural production. component of e people s diet [3]. Among e staple crops at are marked for massive Though Nigeria is e highest producer of rice in production by e Federal Government of Nigeria to e West African Sub-region, she is also e highest achieve food sufficiency by e year 00 are rice, maize, consumer of e commodity. Annual domestic production cassava and sorghum. Boosting rice production requires, of e commodity hovers around 3 million metric tons, among oer ings, e maximization of e limited while demand is as high as about 5 million metric tons, available land and oer productive resources of e leaving a huge gap of million metric tons annually which nation for dry season irrigated rice production. is often filled by importation []. Rice import stands at Corresponding Auor: J. Ahmadu, Department of Agricultural Economics and Extension Services, Faculty of Agriculture, University of Benin, Benin City, Nigeria. 1610
about 1,000,000 metric tons wi e country spending Data Collection: Data collection for e study was over US$300 million annually [4]. This worrisome situation achieved by e use of a well-structured pre-tested requires urgent attention, among which are expanding e questionnaire administered to e respondents and hectares under rice production including dry season complemented wi personal interview. Data were irrigation practice, improving e level of production collected on e socio-economic characteristics of e technologies for rice and utilizing e resources for rice farmers, farm size (ha), quantity of rice seed (kg), labour production efficiently to increase productivity. For (man days) fertilizer (kg), herbicides (litre) and output (kg) intervention in any of ese areas in order to boost rice realized. production, adequate information is needed. It is in view of is at is study seeks to estimate Data Analysis: The data collected were analyzed by using e technical efficiency of irrigated rice farmers in Niger descriptive statistics, stochastic frontier production State of Nigeria. The specific objectives of e study are function and inefficiency model. to: examine e socio-economic characteristics of e irrigated rice farmers in e study area; estimate e Descriptive Statistics: The descriptive statistics such as quantities of inputs used and output realized; e e means, percentages and frequencies were used to production function for e irrigated rice; e level of analyzed e socio-economic characteristics of e technical efficiency of e farmers; and examine e farmers, input and output variables and e distribution of factors influencing e farmers inefficiency, if any. efficiency levels of e farmers. MATERIALS AND METHODS Stochastic Frontier Production Function: The stochastic frontier production function using Maximum Likelihood ' Study Area: Niger State is located between latitude 8 15 Estimation (MLE) technique was used to estimate e ' ' ' and 10 45 Nor and longitude 5 15 and 8 15 East; and production function for e irrigated rice and obtained occupies land area of 76, 363km wi e population of e farmers specific levels of technical efficiency. 3, 950,49 according to e 006 population census. The Cobb-Douglas function linearized in log-form fitted The state has two distinct seasons - e wet season from for e estimation of e stochastic production function April to October and dry season from November to given by Aigner et al. [5] and Mecusen and Van den March. The mean annual rainfall is about 158mm wi Broeck [6] was adopted for e analysis. This is expressed high relative humidity. The study area falls wiin e as: guinea savannah zone of Nigeria wi ick vegetation cover wi tall grasses and trees. The predominant In Q = a 0 + a 1 ln X 1i + a ln X i + a 3 ln X 3i + a 4 ln X 4i + a5 occupation of e inhabitants of e area is agriculture ln X 5i + a 6 ln X 6i + (V i - U i) eqn. (1) and some of e major arable crops grown are rice, maize, sorghum, millet, groundnut, cowpea, cassava, yam and Where: tomatoes. It is e i farmer, Q is output of irrigated Sampling Procedure: A multi-stage sampling technique rice (kg), X 1 is farm size (ha), X is family labour (manwas employed to select e respondents for e study. day), X3 is hired labour (man-day), X 4 is rice seed The first stage of e sampling technique was e (kg), X 5 is fertilizer (kg), X 6 is herbicides (litre), V i is e purposive selection of all e ree (3) agro-ecological symmetric random error term responsible for variation zones of e state (Niger Nor, Central and Sou in output due to factors beyond e control of agro-ecological zones), because rice is produced in e farmers such as weaer, disease outbreak, virtually all e ree (3) zones. This was followed by a government policy, etc. and U i is e random error term random selection of one (1) Local Government Area (LGA) denoting e inefficiency effects, a 0, a 1,.. a 6 are from each zone giving rise to ree (3) LGAs and unknown parameters to be estimated and ln is natural communities from each LGA giving a total of 6 logarim. communities. Finally, 10 irrigated rice-based farmers were It follows at e MLE of e model in eqn. (1) also randomly selected from each community for e study yielded oer parameters such as sigma squared ( s ) and giving a total sample size of 60 respondents. gamma ( ) along wi a s earlier defined, as well as e 1611
log-likelihood function. These parameters, according to Where: Z 1 is age of e farmers (years), Z is gender Aigner et al. [5] and Mecusen and Van den Broeck [6], are (male = 1, female = 0), Z 3 is household size (in number), discussed as follows: Z 4 is educational level (years), Z 5 is years of farming The s gives an indication of e goodness of fit of experience, ei is e random error term and b s are e e model used. The 5 is defined as: unknown coefficients to be estimated. The computer program, FRONTIER version 4.1 s = vi + ui () proposed by Coelli [7] was used to jointly estimate e parameters of e stochastic frontier production function Where: vi and ui are e variances of e error term due and e inefficiency model jointly. to noise and technical inefficiency for e i farmer respectively. RESULTS AND DISCUSSION The gamma ( ) which gives e proportion of e deviation of output from e production frontier due to Socio-Economic Characteristics of Irrigated Rice e farmers technical inefficiency is given as: Farmers: The summary statistics of e socio-economic characteristics of e irrigated rice farmers are presented = u / s (3) in Table 1. The results indicated at irrigated rice production in Niger state was dominated by married males So at 0 1. If = 0, it means all deviations of who were relatively young (average age = 4 years) wi output from e production frontier are caused by noise high family size (13 persons). Majority (54%) of e and if = 1 all deviations are attributed to technical farmers had at least primary education wi long years of inefficiency. farming experience (19 years), corroborating e finding of A generalized log-likelihood ratio test was carried out Ugwuanyi et al. [8]. The dominance of males in e to ascertain wheer ere is significant presence of irrigated rice production confirmed e fact at males are technical inefficiency in e farmers production. The e household heads and erefore are in charge of e log-likelihood ratio (LR) test is given as: core farm production activities while women are mostly into processing and marketing [8]. The average age of e LR=- [L (H 0) - L (H 1)] (4) respondents showed at ey were still in eir active productive age. The large household size might have Where: L (H 0) and L (H 1) are values of log-likelihood positive implication on e farmers production due to e function under e null and alternative hypoeses family labour contribution from e household members, respectively. The LR test has a chi-square (X ) ceteris paribus. The high literacy rate and long years of distribution. arming experience acquired by e farmers might improve The measure of technical efficiency (TE) for e i eir ability to use resources more efficiently in farmer is us expressed as: production. TE = Q i/q i* =exp (- U i) (5) Inputs and Outputs of Irrigated Rice Production: The production inputs and output in e irrigated rice Given at 0 TE 1 production are shown in Table. The farmers were found to be small-scale farmers cultivating less an hectares Where: Qi is e observed output of e i farmer, Q i* of farm land. The implication of e low farm size is at is e frontier output of e i farmer and U i is as earlier e vast land area for rice production in e country [9] defined. which would have been exploited for irrigation during dry season is left unused, us contributing to inadequate Inefficiency Model: The random error term, U i accounting food supply and hence food insecurity in e country. for inefficiency effects in e farmers production is The total average family and hired labour usage per influenced by a number of e farmers socio-economic hectare was 86.84 man-days. This is lower an e mean factors and is defined by e inefficiency model as: labour usage of 99.14 and 144 man days per hectare reported by Idiong [10] and NCRI [9] respectively. U i= b 0+ b1z 1+ bz + b3z 3+ b4z 4+ b5z 5+ ei (6) This is expected because e labour usage by e irrigated 161
Table 1: Summary Statistics of Socio-economic Characteristics of Irrigated Rice farmers in Niger Stat Variable Unit Average statistics Gender: Male % 87 Female % 13 Age Years 41.78 Marital status: Single % 10 Married % 79 Widow/Widower % 11 Household size Number 13 Educational level: No formal education % 46 Primary education % 10 Secondary education % 13 Tertiary education % 31 Farming experience Years 19.18 Sample size Number 60 Source: Field Survey, 009 Table : Input and Output Variables in Irrigated Rice Production in Niger State Variable Unit Value/Farmer Value/ha Farm size Ha 1.36 1 Family labour Man days 66.60 48.97 Hired labour Man days 51.50 37.87 Tractor hiring Hours.04 1.5 Rice seed Kg 104.54 76.89 Fertilizer Kg 10.00 75 Herbicide Litres 4.31 3.17 Observed output Kg 4934.08 368 Frontier output Kg 559.93 411 Sample size Number 60 Source: Field survey, 009 Table 3: Maximum Likelihood Estimates of e Stochastic Frontier Production Function for Irrigated Rice in Niger State Variable Parameter Coefficient t-ratio Constant a0 11.877 18.804 Farm size (X 1) a1 0.39 1.08 Family labour (X ) a 0.141.434** Hired labour (X 3) a3-0.060-1.5 Rice seed (X 4) a4 0.717.114** Fertilizer (X 5) a5 0.111 1.000 Herbicide (X 6) a6 0.7 3.54*** Sigma squared s 0.113 5.147*** Gamma 0.003 0.88 Log-likelihood function (H 1) L(H 1) -16.685 - Log-likelihood function (H 0) L(H 0) -0.07 - Variance of error caused by noise v 0.117 - Variance of error accounting for inefficiency u 0.0003 - Log-likelihood ratio test LR - 7.044 *** Significant at 1%, ** Significant at 5%, Table values: Chi-square at 5% = 16.949, t 0.01 =.617, t 0.05 = 1.980, t 0.1 = 1.658 Source: Computed from Field data, 009 1613
Table 4: Frequency Distribution of Technical Efficiency of Irrigated Rice Farmers in Niger State Technical efficiency (TE) Frequency Percentage (%) 0.601-0.700 1 0.701-0.800 3 5 0.801-0.900 14 3 0.901-1.000 4 70 Total 60 100 Minimum TE 0.696 Maximum TE 0.998 Mean TE 0.90 Source: Computed from Field data, 009 Table 5: Estimates of Technical Inefficiency Parameters of Irrigated Rice Farmers in Niger State Variable Parameter Coefficient t-ratio Constant b0-0.171-0.681 Age (Z 1) b1 0.009 1.370 Gender (Z ) b 0.080 0.595 Household size (Z 3) b3-0.07-1.738* Educational level (Z 4) b4-0.00-0.07 Farming experience (Z 5) b5 0.009 0.501 t-tabulated: t 0.1 = 1.658 *Significant at 10% Source: Computed from Field data, 009 rice farmers was complemented by tractor hiring effect of herbicide on e output may imply increasing (1.50 hours/ha) for ploughing/harrowing and herbicides production efficiency by effective chemical weed usage (3.17 litres/ha) for weeds control. The family labour control. The farm size, hired labour and fertilizer, was higher an e hired labour by 11.10 man days per on e oer hand, were not significant probably due to hectare indicating at e large number of household eir low quantities used by e farmers. These members contributed to e farm production. The quantity results compares wi a number of findings. Idiong [10] of rice seed (76.89/ha) used by e farmers compares reported at labour, farm size and seed positively and favourably wi e recommended average seed rate of significantly related to swamp rice output while 80kg/ha by NCRI [9]. Fertilizer usage (75kg/ha), however, fertilizer was not significant. Similarly, e results of e was very low compared wi e rate of 375kg/ha Cobb-Douglas maximum likelihood estimate given by recommended for lowland rice [9]. This may imply low rice Backman et al. [11] showed at land, labour and seed, productivity, all ings being equal. Evidently, e among oers factors, positively and significantly observed output of e farmers and e frontier output per influenced rice production, while fertilizer had no hectare were lower an e maximum yield of 7.0 metric significant effect. tons per hectare expected from irrigated rice [9], leaving The coefficient of gamma (0.003) which indicated at wide gaps of 3.37 and.89 metric tons,respectively for 0.3% of e variation in e output of e irrigated rice was improvement. attributed to technical inefficiency was not significant. This means at 99.7% of e deviation of output Stochastic Frontier Production Function for Irrigated from e production frontier was occasioned by noise. Rice: The stochastic frontier production function for The log-likelihood ratio test confirmed at e presence irrigated rice in Niger State presented in Table 3 showed of inefficiency effect in e irrigated rice production was at wi e exception of hired labour, all inputs under not significant, implying at e Ordinary Least Squares consideration (farm size, family labour, rice seed, fertilizer (OLS) estimation technique which attributes random effect and herbicide), correlated positively wi rice output, in production to all factors beyond e control of e consistent wi a priori expectation. The coefficients of farmers can adequately estimate e production function family labour, rice seed and herbicide were significant for e irrigated rice. The sigma squared (0.113) was while oers were not significant. High quantities of family significant indicating e correctness of e specified labour and rice seed used (Table ) might have accounted assumptions of e composite error term. This finding is, for eir significant effects on output. The significant however, at variance wi e findings of Okoruwa and 1614
Ogundele [1] and Idiong [10] who established at rice CONCLUSION AND RECOMMENDATIONS production in Nigeria is characterized by significant presence of technical inefficiency effects. The study has established at irrigated rice production in Niger State was practiced under small-scale Technical Efficiency of Irrigated Rice Farmers: The production system. The presence of technical inefficiency technical efficiency (TE) estimates of e irrigated rice effect in e farmers production was found to be farmers are presented in Table 4. Technical efficiency of insignificant as only an average of about 0.3% of e e farmers ranged from 69.60 to 99.80% wi e average deviation of output from e production frontier was of 9%, corroborating e finding of Onoja and Achike accounted by technical inefficiency. The average TE of [13] who reported high technical efficiency of 95% for e farmers was 9%, leaving 8% gap for technical irrigated and rainfed rice production systems. The mean improvement. Despite e high level of TE of e farmers, TE indicates at given e level of technology of e e average frontier output based on e available irrigated rice farmers, little (8%) can be done to increase production technologies employed was.89 metric tons eir technical production capacity. The low level of lower an e maximum potential yield of e irrigated technical inefficiency (8%) is confirmed by e results of rice, implying at e levels of production technologies e gamma and log-likelihood ratio test (Table 3) where employed by e farmers were still low. only 0.3% of e variation in e output of e farmers was Based on ese findings, it is recommended at attributed to eir inefficiency and e presence of e irrigated rice farmers should intensify effort to technical inefficiency effect in e stochastic frontier expand eir farm size to maximize e use of e vast land production function was insignificant respectively. area for rice production during e dry season. The Despite e high level of technical efficiency of e farmers need to be empowered for is expansion. Thus, farmers, eir observed output as well as e output of e government, non-governmental organizations and most efficient farmers based on e available technologies financial institutions should provide adequate needed employed (Table ) was lower an e maximum potential capital for e farmers. Also, adequate irrigation facilities yield of e irrigated rice by 3.37 and.89 metric tons (e.g. dams) should be provided (existing ones respectively. This means at e existing levels of rehabilitated and new ones constructed) for e expansion technological practices employed by e irrigated rice of e irrigated rice production. Furermore, a detailed farmers were still low, a pointer to e need for study should be conducted to ascertain e levels of improvement. The low levels of e technologies raises a production technologies for irrigated rice in e study area question of e where about of e numerous improved wi a view to improving e standards of e technologies developed to boost rice production. It is technologies or transferring e technologies to e eier ere are lapses on e part of agricultural extension farmers. services in transferring e improved technologies to e farmers or e farmers could not afford e technologies. REFERENCES Technical Inefficiency Parameters of Irrigated Rice 1. Jones, M.P., 1995. The Rice Plant and its Farmers: The estimated technical inefficiency parameters Environment. West African Rice Development of e irrigated rice farmers presented in Table 5 showed Assocaition (WARDA) Training Guide, : 1-16. at only household size significantly (P<0.1) influenced. Biyi, D., 005. Government Policies and e technical inefficiency of e farmers. All oer Competitiveness of Nigerian Rice Economy. A paper variables (age, gender, educational level and farming presented at e workshop on Rice Policy and Food experience) were not significant, confirming e low level Security in Sub-Saharan Africa organized by of technical inefficiency effect in e farmers production. WARDA, Cotonou, Republic of Benin, 7-9 The negative sign of e household size indicated at November. increased household size decreased e inefficiency of e 3. Akande, S.O. and G. Akpokodje, 003. Rice Prices farmers. This is expected in view of e large household and Market Integration in Selected Areas in Nigeria. size (Table 1) which contributed higher family labour an Agriculture and Rural Development Department hired labour (Table ) for e rice production. Research Report. 1615
4. Akighir, D.T. and T. Shabu, 011. Efficiency of 9. National Cereals Research Institute (NCRI) (008). Resource Use in Rice Farming Enterprise in Kwande Training Manual on Rice Production and Processing, Local Government Area of Benue State, Nigeria. NCRI, Badeggi, pp: 15-7, 73-77. Int. J. Humanities and Soc. Sci., 1(3): 15-0. 10. Idiong, I.C., 007. Estimation of Farm Level Technical Available at: www.ijhssnet.com Efficiency in Small-Scale Swamp Rice Production in 5. Aigner, D.J., C.A.K. Lovell and P. Schmidt, 1977. Cross River State of Nigeria: A Stochastic Frontier Formulation and Estimation of Stochastic Frontier Approach. World J. Agric. Sci., 3(5): 653-658. Production Function Models. J. Econometrics, 11. Backman, S., K.M. Zahidul Islam and J. Sumelius, 6: 1-37. 009. Determinants of Technical Efficiency of Rice 6. Meeusen, W. and J. Van den Broeck, 1977. Efficiency Farms in Nor-Central and Nor-Western Regions Estimation from Cobb-Douglas Production Functions in Bangladesh. www. perpustakaan. depkeu. go. id/ wi Composed Error. Int. Econ. Review, 18: 435-444. FOLDERJOURNAL/sawah.banglades.pdf. Accessed 7. Coelli, T.J., 1994. A Guide to FRONTIER Version 4.1: on 4 August, 01. A Computer Programming for Stochastic Frontier 1. Okoruwa, V.O. and O.O. Ogundele, 004. Technical Production and Cost Function Estimation. Mimeo, Efficiency Differentials in Rice Production Department of Econometrics, University of New Technologies in Nigeria. www. csae. ox. ac. uk/ England, Armidate. conference/ 006-EO-RPI/papers/csae/okoruwa.pdf 8. Ugwuanyi, C.A., O.S. Balogun, O. Akinyemi, 13. Onoja, A.O. and A.I. Achike, 008. Technical G.F. Balogun, and A. Zungum, 008. Cost and Efficiency of Rice Production under Small-Scale Returns of Small-Scale Locally Milled Rice Marketing Farmer-Managed Irrigation Schemes and Rainfed in Uzo-Uwani LGA of Enugu State. Proceedings of Systems in Kogi State, Nigeria. Proceedings of 10 10 Annual National Conference of Nigerian Annual National Conference of NAAE held at Association of Agricultural Economists(NAAE) held University of Abuja, Abuja, Nigeria, 7-10 October, at University of Abuja, Abuja, Nigeria, 7-10 pp: 4-5. October, pp: 78-85. 1616