# Impact of Irrigation Ecology on Rice Production Efficiency in Ghana.

1. IntroductionSmallholder farmers in sub-Saharan Africa including Ghana, like those in other developing regions of the world, face a number of constraints that limit their farm productivity and farm income [1]. In many parts of SSA cereal crop yields are estimated to be <1.5 ton/ha while the actual potential is more than 5 tons/ha. The low yields are largely attributable to low use of organic and mineral nutrient resources, which has also resulted in negative nutrient balances [2, 3]. The reasons for these poor yields also include lack of sufficient information about production methods and practices and market opportunities. Their ability to access credit and inputs is limited by the lack of collateral which is an important requirement by most financial institutions.

Studies have shown that technical efficiency measures for Ghana's agriculture are low. Reference [4] found that average profit efficiency for rice farmers in Northern Ghana is 63%, with profit efficiency ranging between 16% and 96%. Reference [5] provided evidence to show that smallholder rice farmers in the Upper East region of Ghana produce, on average, 34% below maximum output. The estimated technical efficiency for smallholder irrigators and nonirrigators was 53% and 51%, respectively, using a simple t-test to compare the significance of their means. The authors of [6], in their study of rice farmers under irrigation in Tono, also concluded that mean technical efficiency estimate for irrigation rice farmers was 0.81, which is an improvement of earlier studies. All these studies used the SFA. These studies did not take into consideration the issue of self-selection that could potentially bias their predictions. This study will use the endogenous treatment-regression model to account for possible endogeneity and self-selection bias.

Efficiency measurement is very important because it has a direct effect on productivity and economic growth. Efficiency studies help firms to determine the extent to which they can raise productivity, incomes, and profit by improving their efficiencies, with the existing resource base and the available technology. Such studies could also support decisions on whether to improve efficiency first or to develop a new technology in the short run. More importantly, enhanced efficiency will not only enable farmers to increase their yield and profit but also give direction for the adjustments required in the long run to achieve food sustainability. The main objective of the study is to assess the impact of irrigation ecology on farm household technical, allocative, and economic efficiencies in Ghana.

2. Materials and Methods

2.1. The Study Area. This study covered Northern, Upper East, and Volta regions of Ghana basically because of their rice production potential in the country, which is mainly savanna. About 80% of total rice production in Ghana comes from these three regions. In Northern region three districts were selected based on their involvement in rice production, namely, Savelugu Municipal, Kumbungu, and Tolon districts. The communities covered are Golinga, Gbuli, Vogu, Kushebo, Zangbali, Kprim, Yipelgu, Wuba, Dalung, Kumbungu, Tarikpaa, Duko, Dinga, Gbanga, and Nakpanzo. In the Upper East, the district covered were Bolga Metro and Kassena Nankana East. The communities are Pungu, Kajelo, Yogbania, Chuchulga, Kogwania, Korania, and Wuru. Volta region had only one district participating in the study which is the South Tongu district with Sogakope as the only community; the details are shown in Table 1.

2.2. Sampling Strategy and Sample Frame. Multiphase (purposive and probability) sampling was used to sample the representative smallholder rice farmers in Northern, Upper East, and Volta regions. The three regions were purposively selected because their production constitutes about 80% of national rice production, and hence results can be generalized to be representative of national situation. Each region was classified into the two production ecologies (irrigation and rain fed ecologies). This procedure allowed us to take a representative sample with characteristics that can be generalized for the entire population which it represents. The population of interest for the study included rice smallholder farmers (SHF) working under irrigation and rain fed production systems in Northern, Upper East, and Volta regions of Ghana; see Table 2.

2.3. Types, Sources, and Methods of Data Collection. The study used different data collection tools. These included both quantitative methods (questionnaires) and qualitative (participatory rural appraisal tools; focus group discussions, key informants' interviews) methods. Besides that, literature was obtained on existing studies already done on similar subject. Household survey, focus group discussion (FGD), and key informants' (KI) interviews of the smallholder farmers and other actors were carried out. Focus group discussions were carried out with randomly selected rice FBOs working within the project district. This was aimed at collecting qualitative data to support the data gathered by the survey and also serve as a means of triangulation to ensure that the data is of good quality. This was guided by a preprinted checklist tailored to meet the information needs of the study.

Key informants' interviews were also conducted, basically engaging in a conversation with key stakeholders in the district such as MoFA crop officers, scientists from SARI, processors, and aggregators. This was guided by a preprinted checklist. Semistructured questionnaires were administered to multistage purposively and randomly selected farmers within the project district to enable us to obtain data on their livelihoods, which includes production, marketing, credit access, adoption of good agronomic practices, income status, food security situation, farm and farm household demographics, and rice production status.

3. Analytical Framework

The study employed both descriptive and inferential statistical analysis. Descriptive statistics (e.g., mean, minimum, maximums, standard errors of the mean, and standard deviation) were used to summarize and describe 350 rice farmers' survey results. Inferential statistics were used to arrive at conclusions based on probability. Some of the results are presented in tables and others are presented in graphs. SFA was used to investigate and measure rice farmer's production efficiencies, and endogenous treatment effect regression was used to estimate the impact of irrigation ecology on production efficiency.

3.1. Stochastic Production Frontier Analysis. Stochastic production frontier was first developed by [7, 8] and is now widely used and reported in literature to measure farm performance [9-13]. The specification allows for the decomposition of the error term into a nonnegative random variable, uj, associated with the technical inefficiency (TE) of the ith farm, as well as the normal error term, v;, which represent the random variation in output due to factors beyond the control of farmers, such as variation in weather patterns, measurement error, or any unspecified input variable. The random error term can be positive or negative, and thus the frontiers vary about the deterministic part of the model, exp([x.sub.i][beta]).

In Figure 1, Rice farm D uses [X.sub.2] inputs to produce output. If there are no inefficiency effects, the frontier output could be [D.sub.1]. This is the deterministic part of the frontier (point B); therefore the noise and inefficiency effects are negative. The distance between point D and point represents inefficiency, while the distance between and point B represents variation due to random events.

3.2. Specifications of SFA Model for Production Efficiency Estimation

3.2.1. The Production Frontier Function. The production frontier function is specified as

[mathematical expression not reproducible] (1)

where Y denotes rice output (paddy), X denotes the factor inputs, the subscript i identifies the rice farm, [beta] represents the parameters to be estimated, and e is the error term representing both inefficiency[ .sub.ui] and noise factors v;. The rice production frontier shows the relationship between farm inputs (labour, fertilizer, seed, etc.) and farm output (rice yield), and the value of [beta] indicates the relative importance (propensity) of each input in influencing the rice production process [11]. A parametric production frontier needs to assume a functional form, and two forms that are relatively easy to derive and commonly used in efficiency analysis are the Cobb-Douglas and the translog production functions.

A Cobb-Douglas stochastic frontier, using the terminology of [14], is defined by

ln([y.sub.i])=[X.sub.i][beta] + [v.sub.i] -[u.sub.i] I = 1,2 ... n (2)

where

ln([y.sub.i]) is the logarithm of the rice output of the ith sample farm (i = 1,2 ... 350),

[X.sub.i] are the logarithms of the input quantities used by the ith farm,

[beta] is a column vector of unknown parameters to be estimated for each covariate,

[U.sub.i] is the technical inefficiency (TE) of the ith farm and in this study it is assumed to be an independent and identically distributed (i.i.d.) half normal random variable,

[V.sub.i] is the random error term, assumed to be i.i.d. normal random variable with zero mean and constant variance, [sigma][v.sup.2], independent of the [u.sub.i].

The technical efficiency of the ith rice farm, in time period t, is given by the ratio of observed output to the maximum potential output, as defined by the frontier.

TE = [Y*.sub.i]/[Y.sub.i] = exp (-[u.sub.i]) (3)

Where [Y.sub.i] = the total production frontier, [Y*.sub.i] = the stochastic production frontier

3.2.2. The Rice Production Cost Frontier Function. The rice production cost frontier function is specified as

[mathematical expression not reproducible] (4)

where C denotes the total production cost observed for ith farmer, [X.sub.i] is the output quantity for farmer i (rice produced), [P.sub.i] is the input price vector used for the ith farmer, [beta] is the parameters to be estimated, and e; is the composite error term representing both inefficiency, [u.sub.i], and noise factors, [v.sub.i].

AE = [C*.sub.i]/[C.sub.i] = exp ([u.sub.i]) (5)

where [C.sub.i] is the total production cost frontier and [C*.sub.i] is the stochastic cost frontier. This will give us the allocative efficiency from which economic efficiency will be estimated as

EE = TE x AE (6)

where EE is the economic efficiency.

The maximum likelihood estimation technique is used to estimate the inefficiencies. In addition to estimating the levels of technical efficiency among farmers, the factors influencing efficiency are also being examined under the endogenous treatment effect model.

The endogenous treatment effect model will be used to assess the impact of irrigation ecology on production efficiencies while examining the determinants of production efficency and also irrigation ecology choice decision which is explained in Section 3.3.

3.3. Endogenous Treatment Effect Regression Model (ETERM). The endogenous treatment effect regression model is a linear model that allows for correlation between unobservable factors affecting the treatment equation and those affecting the outcome measures. The idea is to model the treatment effect on the outcome measure as in [15]. This model assumes a joint normal distribution between the errors of the treatment equation and the outcome equation.

3.3.1. The Endogenous Treatment-Regression Model (ETRM) Specification. Estimation of endogenous treatment effect model is a common feature in empirical studies in economics. When the treatment can be categorized by a dichotomous indicator function, its effects are typically estimated via instrumental variables or variants of the control function approach motivated by [16,17].

The endogenous treatment effects model is a linear model that allows for correlation between unobservable factors affecting household irrigation ecology choice decision and those affecting the household production efficiency measures (technical, allocative, and economic efficiencies). The household technical, allocative, and economic efficiencies are a proportion measure with 0 meaning perfect inefficiency and 1 a maximum efficiency. The idea is to model the treatment effect of household irrigation ecology on the efficiency measures of small scale rice producers. As in [15] we use the endogenous treatment effect regression specification to assess the impact of irrigation ecology on technical, allocative, and economic efficiencies. This model assumes a joint normal distribution between the errors of the selection equation (irrigation ecology) and the outcome equation (the measure of technical, allocative, and economic efficiencies). We specify the outcome model as follows:

[Eff.sub.i] = [X.sub.i]'[beta] + [delta]I[E.sub.i] + [[epsilon].sub.i], (7)

where the effect of irrigation ecology ([IE.sub.i]) on technical, allocative, and economic efficiencies ([Eff.sub.i]) is expressed. The impact of irrigation ecology on technical, allocative, and economic efficiencies is not captured by the [delta], because these households were not randomly assigned to participate in irrigation farming or otherwise but were personal choices of the participants to participate in irrigation farming or rain fed (case of self-selection; self-selection bias arises in any situation in which individuals select themselves into a group, causing a biased sample with nonprobability sampling). Hence, neglecting the potential endogeneity (the problem of endogeneity occurs when the independent variable is correlated with the error term in a regression model) of irrigation ecology will produce wrong estimates of the treatment model and will confound the effect of irrigation ecology on household technical, allocative, and economic efficiencies. Household irrigation ecology choice (treatment) is based on the household, individual, community, and farm characteristics Gi, and is modeled as

I[E*.sub.i] = [G'.sub.i][alpha] + [u.sub.i] (8)

[mathematical expression not reproducible] (9)

where [IE*.sub.i] represent irrigation ecology and [X.sub.i] and [G.sub.i] are covariates that are unrelated to the error terms. [beta] and [alpha] are the parameter estimates. The assumption is that [[epsilon].sub.i] and [u.sub.i] are jointly normally distributed with mean vector zero and variance covariance matrix Z given as

[mathematical expression not reproducible] (10)

The model can be estimated using the two-step approach or the maximum likelihood approach. This is therefore modeled simultaneously as irrigation ecology model as in (8) and the efficiency model as in (7). Consistent estimates of impact of irrigation ecology decision on their technical, allocative, and economic efficiencies are obtained by accounting for the endogenous participation. The determinants of the efficiencies are jointly determined. The maximum likelihood approach is used to analyse the model.

4. Results and Discussion

4.1. Effect of Irrigation Ecology on Some Factors of Production of Farm Households. According to Table 3, the mean age difference between irrigation farms and rain fed farms is about 0.34 and is not significant. There is also no significant difference in rice production experience between the irrigation farmers and the rain fed farmers. Irrigation farmers are richer than rain fed farmers and this is significant at 1%. There is no difference in the household size and also the available arable lands of the two groups.

According to Table 4 the mean difference in rice farm size of irrigation farmers and rain fed farmers is about 0.34 acres and it is not significant. The mean differences in fertilizer use, seed use, and labour used are 68.47kg, 23.13kg, and 3 persons, respectively, which are all significant. There is no difference in the prices of fertilizer, seed, and labour used. Yield is an important variable in assessing farm level performance and it is evidently clear that irrigation farmers have higher yields than their rain fed farmer colleagues with mean difference of 681.26kg per acre. Total output of irrigation farmers was far more than the output of rain fed farmers with a mean of about 1823.29kg of paddy rice. Output price is not significant indicating that irrigation farmers and their rain fed counterparts receive the same price for their paddy rice. This implies their farm revenues will also be higher with a significant mean difference of 1,101.34GHS. Cost of production of irrigation farmers is far more than that of the rain fed producers with mean difference of 144.34 GHS. Gross margins mean difference is 957.00 GHS, indicating that irrigation farmers earn more profit than their rain fed counterparts.

4.2. Irrigation and Rain Fed Production Ecologies Analysis. The mean technical efficiency for irrigated and rain fed farms is 0.84 and 0.83, respectively. These imply that irrigated farms can achieve their current outputs with about 16% reduction in inputs used. The mean allocative efficiency of the study is 71% and 61%, respectively, for irrigated and rain fed systems. However, the economic efficiency of farms is 61% and 51% for irrigated and rain fed farms. Farmers in irrigation are producing rice at a better cost minimizing level compared to that of rain fed systems. This implies irrigated and rain fed farms can achieve their current production levels with about 39% and 41% reduction in cost of production. Details are shown in Table 5.

4.3. Mean Differences of Irrigation and Rain Fed Farms. There is significant difference in the allocative and economic efficiency means of irrigated and rain fed farms. This shows that the irrigated farms have higher allocative and economic efficiencies than their rain fed colleagues hence the null hypothesis is rejected. However, with regard to technical efficiency, there are no significant differences in their means. The null hypothesis is sustained as shown in Table 6.

4.4. Impact of Irrigation Ecology on Technical Efficiency and Its Determinants. Results of the endogenous treatment effect model on impact of irrigation ecology on technical efficiency are presented in Table 7. The maximum likelihood estimation approach was used to estimate the impact of irrigation ecology on technical efficiency of rice farms. Thus, the results of the selection equation are given in the 4th and 5th columns of Table 7. The results of the outcome equation which represents the impact of contract participation on rice farms technical efficiency are presented in the 2nd and 3rd columns of Table 7.

From the results, the Wald test is highly significant indicating the goodness of fit of our endogenous treatment effect model. This implies there are endogeneity problems; hence the use of the endogenous treatment effect model is justified. The likelihood ratio test of independence of the selection and outcome equations indicates that we can reject the null hypothesis of no correlation between irrigation ecology and technical efficiency. This implies irrigation ecology is negatively correlated with technical efficiency. The estimated average treatment effect (ATE) of participating in irrigation production is 0.05 of the technical efficiency. The impact of irrigation ecology on technical efficiency is about 0.05. This implies farmers participating in irrigation farming are more efficient (0.05 more efficient) in their rice production than those in rain fed production. The estimated correlation between the treatment assignment errors and the outcome errors is (-0.44) indicating that the unobservables that increased technical efficiency also tend to occur with the unobservables that discourage choice of irrigation production. The negative sign indicates a positive bias, suggesting that farmers with above average technical efficiency have a higher probability of participating in irrigation production and will prefer to produce under irrigation production.

4.5. Impact of Irrigation Ecology on Allocative Efficiency. Results of the endogenous treatment effect model on impact of irrigation ecology on technical efficiency are presented in Table 8. The maximum likelihood estimation approach was used to estimate the impact of irrigation ecology on technical efficiency of rice farms. Thus, the results of the selection equation are given in the 4th and 5th columns of Table 8. The results of the outcome equation which represents the impact of irrigation production on rice farms technical efficiency are presented in the 2nd and 3rd columns of Table 8.

From the results, the Wald test is highly significant indicating the goodness of fit of our endogenous treatment effect model. This implies there are endogeneity problems; hence the use of the endogenous treatment effect model is justified. The likelihood ratio test of independence of the selection and outcome equations indicates that we can reject the null hypothesis of no correlation between irrigation ecology and allocative efficiency. This implies irrigation ecology is correlated with allocative efficiency. The estimated average treatment effect (ATE) of irrigation production is 0.33 of the allocative efficiency. The impact of irrigation ecology on allocative efficiency is about 0.33. This implies farmers participating in irrigation farming are more allocatively efficient (0.33 more efficient) in their rice production than those in rain fed production. The estimated correlation between the treatment assignment errors and the outcome errors is (-0.87) indicating that the unobservables that increased allocative efficiency also tend to occur with the unobservables that discourage the choice of irrigation production. The negative sign indicates a positive bias, suggesting that farmers with above average allocative efficiency have a higher probability of producing under irrigation.

4.6. Impact of Irrigation Ecology on Economic Efficiency. Results of the endogenous treatment effect model on impact of irrigation ecology on economic efficiency are presented in Table 9. The maximum likelihood estimation approach was used to estimate the impact of irrigation ecology on technical efficiency of rice farms. Thus, the results of the selection equation are given in the 4th and 5th columns of Table 9. The results of the outcome equation which represents the impact of irrigation ecology on rice farms economic efficiency are presented in the 2nd and 3rd columns of Table 9.

From the results, the Wald test is highly significant indicating the goodness of fit of our endogenous treatment effect model. This implies there are endogeneity problems; hence the use of the endogenous treatment effect model is justified. The likelihood ratio test of independence of the selection and outcome equations indicates that we can reject the null hypothesis of no correlation between irrigation ecology and allocative efficiency. This implies irrigation ecology is positively correlated with economic efficiency. The estimated average treatment effect (ATE) of participating in irrigation production is 0.23 of the economic efficiency. The impact of irrigation ecology on economic efficiency is about 0.23. This implies farmers participating in irrigation farming are more economically efficient (0.23 more efficient) in their rice production than those in rain fed production. The estimated correlation between the treatment assignment errors and the outcome errors is (-0.88) indicating that the unobservables that increased economic efficiency also tend to occur with the unobservables that discourage the choice of irrigation production. The negative sign indicates a positive bias, suggesting that farmers with above average allocative efficiency have a higher probability of producing under irrigation.

Irrigation ecology had significant impact on technical, allocative, and economic efficiencies of rice farms. This impact was more on allocative efficiency followed by economic efficiency and then technical efficiency; see Table 10.

5. Conclusion and Recommendation

5.1. Main Findings of Impact of Irrigation Ecology on Technical Efficiency. The impact of irrigation ecology on technical efficiency is about 0.05. This implies farmers participating in irrigation farming are more efficient (0.05 more efficient) in their rice production than those in rain fed production. The estimated correlation between the treatment assignment errors and the outcome errors is (-0.44) indicating that the unobservables that increased technical efficiency also tend to occur with the unobservables that discourage the choice of irrigation production. The negative sign indicates a positive bias, suggesting that farmers with above average technical efficiency have a high probability of participating in irrigation production and will prefer to produce under irrigation production.

5.2. Main Findings of Impact of Irrigation Ecology on Allocative Efficiency. The impact of irrigation ecology on allocative efficiency is about 0.33. This implies farmers participating in irrigation farming are more allocatively efficient (0.33 more efficient) in their rice production than those in rain fed production. The estimated correlation between the treatment assignment errors and the outcome errors is (-0.87) indicating that the unobservables that increased allocative efficiency also tend to occur with the unobservables that discourage irrigation production. The negative sign indicates a positive bias, suggesting that farmers with above average allocative efficiency have a higher probability of producing under irrigation.

5.3. Main Findings of Impact of Irrigation Ecology on Economic Efficiency. The impact of irrigation ecology on economic efficiency is about 0.23. This implies farmers participating in irrigation farming are more economically efficient (0.23 more efficient) in their rice production than those in rain fed production. The estimated correlation between the treatment assignment errors and the outcome errors is (-0.88) indicating that the unobservables that increased economic efficiency also tend to occur with the unobservables that discourage the choice of irrigation production. The negative sign indicates a positive bias, suggesting that farmers with above average allocative efficiency have a higher probability of producing under irrigation.

5.4. Recommendation. Irrigated farms have higher technical, allocative, and economic efficiencies than their rain fed counterparts; hence we recommend that farmers should be encouraged to participate more in irrigation rice production than in the rain fed production since they are more efficient in their resource allocation.

https://doi.org/ 10.1155/2018/5287138

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

This work was carried out with the support of West African Science Service Center on Climate Change and Adapted Land Use (WASCAL), Federal Ministry of Education and Research (BMBF), and Alliance for Green Revolution in Africa (AGRA) funded Quality Rice Development Project (QRDP).

References

[1] N. Key and D. Runsten, "Contract farming, smallholders, and rural development in Latin America: The organization of agroprocessing firms and the scale of outgrower production," World Development, vol. 27, no. 2, pp. 381-401,1999.

[2] E. M. A. Smaling and L. O. Fresco, "A decision-support model for monitoring nutrient balances under agricultural land use (NUTMON)," Geoderma, vol. 60, no. 1-4, pp. 235-256,1993.

[3] A. De Jager, D. Onduru, M. S. Van Wijk, J. Vlaming, and G. N. Gachini, "Assessing sustainability of low-external-input farm management systems with the nutrient monitoring approach: A case study in Kenya," Agricultural Systems, vol. 69, no. 1, pp. 99-118, 2001.

[4] A. Abdulai and W. Huffman, "Structural adjustment and economic efficiency of rice farmers in Northern Ghana," Economic Development and Cultural Change, vol. 48, no. 3, pp. 503-520, 2000.

[5] A. H. Seidu, D. B. Sarpong, and R. Al-Hassan, "Allocative efficiency, employment and rice production risk: An analysis of small holder paddy farms in the Upper East Region of Ghana," Ghana Journal of Development Studies, vol. 1, no. 2, pp. 142-163, 2004.

[6] S. A. Donkoh, S. Ayambila, and S. Abdulai, "Technical efficiency of rice production at the Tono irrigation scheme in northern Ghana," American Journal of Experimental Agriculture, vol. 3, no. 1, p. 25, 2013.

[7] D. Aigner, C. K. Lovell, and P. Schmidt, "Formulation and estimation of stochastic frontier production function models," Journal of Econometrics, vol. 6, no. 1, pp. 21-37,1977.

[8] W. Meeusen and J. van Den Broeck, "Efficiency estimation from Cobb-Douglas production functions with composed error," in Proceedings of the International Economic Review, pp. 435-444, 1977.

[9] G. E. Battese and T. J. Coelli, Frontier Production Functions, Technical Efficiency and Panel Data: With Application to Paddy Farmers in India, pp. 149-165, Springer, Netherlands, 1992.

[10] T. J. Coelli and S. Perelman, "Efficiency measurement. multipleoutput technologies and distance fucntions: with application to european railways," CREPP, vol. 96, pp. 05-31,1996.

[11] T. Kompas and T. N. Che, "Technology choice and efficiency on Australian dairy farms," Australian Journal of Agricultural and Resource Economics, vol. 50, no. 1, pp. 65-83, 2006.

[12] R. Magreta, A. K. Edriss, L. Mapemba, and S. Zingore, "Economic efficiency of rice production in smallholder irrigation schemes: A case of Nkhate irrigation scheme in Southern Malawi," in Proceedings of the 4th International Conference of the African Association of Agricultural Economists, 2013.

[13] Souleymane Ouedraogo, "Technical and economic efficiency of rice production in the Kou valley (Burkina Faso): Stochastic frontier approach," Asian Journal of Agriculture and Rural Development, vol. 5, no. 2, pp. 53-63, 2015.

[14] T. Coelli, D. P. Rao, and G. E. Battese, "Additional topics on data envelopment analysis," in An Introduction to Efficiency And Productivity Analysis, pp. 161-181, Springer US, 1998.

[15] W. H. Greene, 2002, NLOGIT: Version 3.0: Reference Guide. Econometric Software.

[16] J. J. Heckman, "A partial survey of recent research on the Labor supply of women," American Economic Review, vol. 68, no. 2, pp. 200-207,1978.

[17] J. J. Heckman, Statistical Models for Discrete Panel Data, Department of Economics and Graduate School of Business, University of Chicago, Chicago, IL, USA, 1979.

John Kanburi Bidzakin [ID], (1,2) Simon C. Fialor, (2) Dadson Awunyo-Vitor, (2) and Iddrisu Yahaya (1)

(1) Savanna Agricultural Research Institute (SARI), Council for Scientific and Industrial Research (CSIR), P.O. Box TL 52, Nyankpala, Tamale, Ghana

(2) Department of Agricultural Economics, Agribusiness & Extension, Kwame Nkrumah University of Science & Technology, Kumasi, Ghana

Correspondence should be addressed to John Kanburi Bidzakin; bidzakin2@gmail.com

Received 29 December 2017; Revised 23 April 2018; Accepted 17 May 2018; Published 12 June 2018

Academic Editor: Innocenzo Muzzalupo

Caption: Figure 1: Production frontier.

Table 1: Study regions, districts, and communities. REGIONS DISTRICTS COMMUNITIES Northern Tolon Golinga Region Kumbungu Gbuli, Vogu, Kushebo, Zangbali, Kprim, Yipelgu, Wuba, Dalung, Kumbungu Savelugu Tarikpaa, Duko, Dinga, Gbanga, Nakpanzo Upper East Bolga Metro Nyariga, Yorugu, Gowrie, Zaare, Vea, Region Yebongu, Aguusi Kassena Pungu, Kajelo, Yogbania, Chuchulga, Nankana East Kogwania, Korania, Wuru Volta Region South Tong Sogakope Table 2: Sample frame. Region Production ecology # farmers NR Irrigation 51 Rain Fed 99 UER Irrigation 44 Rain Fed 106 VR Irrigation 38 Rain Fed 12 Total 350 Table 3: Production ecology effect of means of farm household's characteristics. Farm household Irrigation Rain fed Characteristics (n=133) (n=217) Mean Mean Age of household head 44.71 44.36 Farmer experience in years 22.50 24.42 Wealth of farm household 8385.63 3298.29 Household Size 7 7 Total household arable land 5.43 5.41 Farm household Mean Z- Characteristics Difference statistic Age of household head 0.34 0.28 Farmer experience in years -1.92 -1.73 Wealth of farm household 5087.34*** 4.5 Household Size -0.24 -0.7 Total household arable land 0.02 0.0447 *** 1% significance level; ** 5% significance level; * 10% significance level. Table 4: Production situation under rain fed and irrigation production. Irrigation Rain fed (n=133) (n=217) Variables Mean Mean Rice farm size in acres 2.89 2.66 Fertilizer used in kg 283.54 215.08 Seed used in kg 87.32 64.19 Labour used (average 12 9 number of people who worked) Fertilizer price in GHS 1.73 1.40 Seed price in GHS 0.62 0.66 Labour price in GHS 38.96 41.21 Yield per acres 1511.41 830.15 Total output kg 3994.67 2171.38 Output price per kg in GHS 1.40 1.36 Total Revenue GHS 2249.68 1148.34 Total cost of production 572.43 428.09 GHS Gross margins GHS 1677.25 720.25 Mean Z-statistic Difference Variables Rice farm size in acres 0.23 0.91 Fertilizer used in kg 68.47 *** 2.63 Seed used in kg 23.13 ** 2.13 Labour used (average 3 ** 2.29 number of people who worked) Fertilizer price in GHS 0.32 0.48 Seed price in GHS -0.04 -0.92 Labour price in GHS -2.25 -0.47 Yield per acres 681.26 *** 5.7 Total output kg 1823.29 *** 5.06 Output price per kg in GHS 0.04 1.47 Total Revenue GHS 1101.34 *** 5.52 Total cost of production 144.34 *** 3.95 GHS Gross margins GHS 957.00 *** 5.4 *** 1% significance level; 5% significance level; * 10% significance level. Table 5: Average efficiency measures of the irrigation and rain fed ecologies. Efficiencies Irrigation (n=133) Mean Min Max STD SEM Technical 0.84 0.58 0.94 0.06 0.01 Allocative 0.71 0.10 1.09 0.29 0.02 Economic 0.61 0.07 1.00 0.32 0.03 Efficiencies Rain fed (n=217) Mean Min Max STD SEM Technical 0.83 0.58 0.99 0.07 0.00 Allocative 0.61 0.09 1.07 0.29 0.02 Economic 0.51 0.07 1.00 0.32 0.02 Table 6: Mean differences of irrigation and rain fed farms. Efficiencies Irrigation Rain fed Mean Z-statistic (n=133) (n=217) Difference Mean Mean Technical 0.84 0.83 0.01 1.5 Allocative 0.71 0.61 0.09 *** 2.97 Economic 0.61 0.51 0.09 *** 2.64 *** 1% significance level; ** 5% significance level; * 10% significance level. Table 7: Estimates of the endogenous treatment effect model of impact of irrigation ecology on technical efficiency. Technical Efficiency Variables Coef. Std. Err. Improved seed -0.027 0.017 Gender -0.006 0.019 Age 0.001 * 0.000 Educational level -0.002 * 0.001 Farmer experience -0.001 ** 0.000 Wealth of farm HH 0.000 0.000 Household Size 0.001 0.001 Total HH arable land 0.000 0.002 Rice farm size 0.004 0.003 FBO Membership -0.017 ** 0.008 Fertilizer Seed Labour 1. Irrigation ecology 0.048 ** 0.021 _cons 0.840 *** 0.033 /athrho -0.46942 ** 0.207234 /lnsigma -2.706 *** 0.061 rho -0.438 0.168 sigma 0.067 0.004 lambda -0.029 0.013 Log likelihood 261.944 Waldtest rf (15) 4.95 *** LR test of independent equations (1) 3.84** Irrigation Ecology Variables Coef. Std. Err. Improved seed Gender -0.17723 0.37695 Age -0.00236 0.00831 Educational level -0.02381 0.01787 Farmer experience -0.01053 0.00866 Wealth of farm HH 3.3E-05 *** 1.15E-05 Household Size -0.03129 0.02902 Total HH arable land 0.02266 0.03297 Rice farm size -0.09362 0.063848 FBO Membership -0.06097 0.17115 Fertilizer 0.00091 * 0.00054 Seed 0.0016 0.00101 Labour 0.02072 ** 0.01074 1. Irrigation ecology _cons /athrho /lnsigma rho sigma lambda Log likelihood Waldtest rf (15) LR test of independent equations (1) 3.84** *** 1% significance level; ** 5% significance level; * 10% significance level. Table 8: Estimates of the endogenous treatment effect model of impact of irrigation ecology on allocative efficiency. Allocative Efficiency Variables Coef. Std. Err. Improved seed -0.040 0.060 Gender 0.042 0.093 Age 0.001 0.002 Educational level 0.006 0.004 Farmer experience 0.002 0.002 Wealth of farm HH 0.000 0.000 Household Size -0.001 0.007 Total HH arable land 0.005 0.008 Rice farm size 0.035 ** 0.015 FBO Membership 0.030 0.041 ISFM3 Adoption 0.026 0.036 Contract Farming 0.254 *** 0.037 Fertilizer 0.000 0.000 Seed -0.00043 *** 0.000247 Labour -0.005 *** 0.002 1. Irrigation ecology 0.330 * 0.050 _cons 0.23829 0.1427 /athrho -1.34707 *** 0.289909 /lnsigma -1.143 *** 0.084 rho -0.873 0.069 sigma 0.319 0.027 lambda -0.278 0.044 Log likelihood -202.687 Wald test [chi square] (15) 209.61 *** LR test of independent equations [chi square] (1) 5.03 ** Irrigation Ecology Variables Coef. Std. Err. Improved seed Gender -0.18091 0.382307 Age -0.00393 0.007909 Educational level -0.01303 0.017019 Farmer experience -0.01424 * 0.008081 Wealth of farm HH 3.34E-05 *** 1.15E-05 Household Size -0.01883 0.027527 Total HH arable land 0.021044 0.030328 Rice farm size -0.08274 0.05944 FBO Membership -0.0174 0.16182 ISFM3 Adoption Contract Farming Fertilizer 0.00055 0.00047 Seed 0.00141 0.00098 Labour 0.016 ** 0.00787 1. Irrigation ecology 0.015587 0.007868 _cons 0.057237 0.531055 /athrho /lnsigma rho sigma lambda Log likelihood Wald test [chi square] (15) LR test of independent equations [chi square] (1) 5.03 ** *** 1% significance level; ** 5% significance level; * 10% significance level. integrated Soil Fertility Management technique Table 9: Estimates of the endogenous treatment effect model of impact of irrigation ecology on economic efficiency. Economic Efficiency Variables Coef. Std. Err. Improved seed -0.022 0.063 Gender 0.104 0.102 Age 0.000 0.002 Educational level 0.004 0.005 Farmer experience 0.003 0.002 Wealth of farm HH 0.000 0.000 Household Size -0.003 0.008 Total HH arable land 0.005 0.009 Rice farm size 0.069 *** 0.017 FBO Membership 0.018 0.044 ISFM Adoption 0.028 0.039 contract farming 0.235 *** 0.039 Fertilizer 2.9E4 *** 0.000 Seed -0.001 *** 0.000 Labour -0.007 ** 0.002 1. Irrigation 0.23 *** 0.034 ecology _cons 0.047 0.154 /athrho -1.3758 *** 0.232817 /lnsigma -1.051 *** 0.071 rho -0.880 0.053 sigma 0.350 0.025 lambda -0.308 0.039 Log likelihood -232.204 Wald test [chi squaare] (15) 219.82 *** LR test of independent equations [chi squaare] (1) 10.05 *** Irrigation EcolStd. Variables Coef. Err. Improved seed Gender -0.20412 0.37845 Age -0.00242 0.00771 Educational level -0.01636 0.01694 Farmer experience -0.0131 0.008 Wealth of farm HH 3.1E-05 *** 1.1E-05 Household Size -0.02967 0.02774 Total HH arable land 0.01808 0.03016 Rice farm size -0.0871 0.05974 FBO Membership -0.03796 0.1604 ISFM Adoption contract farming Fertilizer 0.00055 0.00046 Seed 0.00159 0.00098 Labour 0.01743 *** 0.00751 1. Irrigation ecology _cons 0.109085 0.521881 /athrho /lnsigma rho sigma lambda Log likelihood Wald test [chi squaare] (15) LR test of independent equations [chi squaare] (1) 10.05 *** *** 1% significance level; ** 5% significance level; * 10% significance level. Table 10: Summary of impact of irrigation ecology on technical, allocative, and economic efficiencies. Study Unit Efficiency Type Impact Significance level Irrigation Technical 0.05 5% level ecology Allocative 0.33 10% level Economic 0.23 1% level

Printer friendly Cite/link Email Feedback | |

Title Annotation: | Research Article |
---|---|

Author: | Bidzakin, John Kanburi; Fialor, Simon C.; Awunyo-Vitor, Dadson; Yahaya, Iddrisu |

Publication: | Advances in Agriculture |

Date: | Jan 1, 2018 |

Words: | 6257 |

Previous Article: | Response of Two Chickpea (Cicer arietinum L.) Varieties to Rates of Blended Fertilizer and Row Spacing at Tselemti District, Northern Ethiopia. |

Next Article: | Estimating Profit Efficiency of Artisanal Fishing in the Pru District of the Brong-Ahafo Region, Ghana. |

Topics: |