Skip Navigation

ICES Journal of Marine Science: Journal du Conseil 2006 63(9):1759-1764; doi:10.1016/j.icesjms.2006.06.012
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Esmaeili, A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Esmaeili, A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© 2006 International Council for the Exploration of the Sea

Technical efficiency analysis for the Iranian fishery in the Persian Gulf

Abdoulkarim Esmaeili*

Department of Agricultural Economics, College of Agriculture Shiraz University, Shiraz, Iran

*tel: +98 917 1612327; fax: +98 711 2273517. e-mail: esmaeili{at}shirazu.ac.ir.

There are many fish landing areas in southern Iran, distributed all along the coastline, and despite gradually increasing effort, the total catch has fluctuated in recent years. This study examines the technical efficiency of the fishing industry, and identifies factors that could be causing inefficiency. A stochastic production function among fishery vessels is estimated. Technical efficiency measures the ability of firms to maximize output using a given set of inputs and technologies. The results indicate that technical efficiency in the fishery is relatively low, and that wooden vessels of medium size are more efficient than small fibreglass vessels. Both skippers' socio-economic drivers and vessel instrumentation have a significant impact on efficiency. Ownership of two-way radio and ownership of GPS are important considerations that influence fishing efficiency, and the skipper's level of education and experience are qualities that also affect it. Owner-operated vessels and younger skippers are more efficient than others. Understanding these constraints may contribute to increasing the efficiency of the Iranian fishery in the Persian Gulf.

Keywords: fisheries, Iran, stochastic production frontier, technical efficiency

Received 4 January 2006; accepted 30 June 2006.


   Introduction
Top
 Introduction
Methods
 Results
 Discussion
 References
 
The Persian Gulf is a body of water 950 km long that separates Iran from Iraq, Saudi Arabia, Bahrain, Kuwait, Qatar, and the United Arab Emirates. The Gulf is one of the world's strategic areas owing to its importance in global oil transportation. Fisheries in the Persian Gulf also play an important role in the economics of the countries that abut it. Three coastal provinces of Iran have important fisheries on their side of the Gulf: Khozestan in the northwest, Hormozgan in the northeast, and Boushehr in the central Persian Gulf. Hormozgan is the largest province, and some 54% of the 198 987 t of fish landed in 2003 on the Iranian side of the Gulf came from this province. To make that total catch, there were 25 969 fishers and 3508 fishing vessels. Fish is an important component of the local economy in the region, and the main problems facing the fisheries currently are the uncertain availability of fish, and its corollary, overfishing. Because of the declining fish landings and the ever increasing number of fishing vessels, the per capita catch has decreased in recent years (Figures 1 and 2).


Figure 1
View larger version (13K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1 Relationship between total fish catch and the number of vessels operating in Hormozgan Province, northern Persian Gulf, 1991–2003 (data courtesy of the Iranian Fisheries Company).

 


Figure 2
View larger version (8K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 2 Trends in catch per vessel operating in Iran's Hormozgan Province, northern Persian Gulf, 1991–2003 (data courtesy of the Iranian Fisheries Company).

 
Catch per vessel is declining year on year, despite effort increasing; the problem is that too many vessels are pursuing too small a fish resource. Poor management, inefficiency, high inflation, and a lack of regulation have caused some fish stocks to decline noticeably over the past few years.

Although much research has been conducted to investigate the efficiency of the world's fisheries (Kirkley et al., 1995; Sharma and Leung, 1999; Pascoe et al., 2001; Fousekis and Konaris, 2003; Herrero and Pascoe, 2003; Kompas et al., 2003; Garcia del Hoyo et al., 2004; Herrero, 2005; Tingley et al., 2005), only a few have looked at the situation in Iran (Yazdani and Esmaeili, 1995). The first study of technical efficiency of fisheries using a stochastic frontier function was that of Kirkley et al. (1995). Such an approach was also employed by Sharma and Leung (1999) to investigate the effect of vessel characteristics and targeted species on vessel technical efficiency in the Hawaiian longline fishery. Data envelopment analysis (DEA) is a recently developed deterministic approach for measuring efficiency in fisheries (Felthoven, 2002; Pascoe and Herrero, 2004).

The main purpose of the current study is to examine the technical efficiency of the Iranian Persian Gulf fishing industry, and to identify factors that may be causing inefficiency.


   Methods
Top
 Introduction
 Methods
Results
 Discussion
 References
 
Efficiency analysis
Technical efficiency measures the ability of firms to maximize output using a given set of inputs. Farrell (1957) introduced a methodology for measuring efficiency nearly five decades ago, and his methodology is still undergoing refinement and improvement. There are two approaches to estimating technical efficiency, parametric and non-parametric. The stochastic production frontier (SPF) developed by Aigner et al. (1977) and Meeusen and van den Broeck (1977) is a parametric approach. Data envelopment analysis (DEA), developed by Charnes et al. (1978), is a non-parametric approach. SPF uses a parametric function, whereas DEA is based on a linear programming technique. The production frontier in DEA is deterministic, so any deviations from it are related to inefficiency. In an SPF, the production frontier function is sensitive to random shocks by including a random error term to the production frontier. As a consequence, only deviations caused by controllable decisions can be attributed to inefficiency. The model used in this paper is based on the Battese and Coelli (1992, 1993, 1995) approach, which uses a stochastic frontier model to estimate efficiency:


Formula 1

(1)
where yi is the output of the i(th) vessel, Xi a vector of production inputs, ß a vector of parameters, and {nu}i refers to independent identically distributed random variables Formula that measure errors and exogenous shocks beyond the control of the manager (such as bad weather). Parameter ui is assumed to be truncated normally with variance Formula , and the mean Formula is represented as a linear combination of the inefficiency variable.

The inefficiency determinants function following the general form


Formula 2

(2)
where Zi is a vector of factors affecting the efficiency level, {delta} is a vector of parameters, and wi is the error term. Following Battese and Corra (1977) and Battese and Coelli (1993), variance terms are parameterized by replacing Formula and Formula with Formula and Formula . The technical efficiency of the i(th) firm can then be defined as


Formula

where E is the expectations operator and a technical efficiency measure by conditional expectation. The expected maximum value of Yi is conditional on ui = 0.

The frontier efficiency model (Equation (1)) and inefficiency model (Equation (2)) can be estimated together by maximum likelihood. The particular frontier software used is FRONTIER 4.1, developed by Coelli (1996), which uses a three-step estimation method to obtain final estimates of maximum likelihood. First, unbiased estimates of the ß parameters are obtained via OLS (ordinary least squares). A two-phase grid search of {gamma} is conducted in the second step, with ß set to the OLS estimates and other parameters set to zero. The third step involves an iterative procedure to obtain the estimated maximum likelihood.

Data and the fishery
Iran has a coastline of more than 1800 km along the Persian Gulf and Oman Sea, and 900 km in the Caspian Sea. Fishing in southern Iran is carried out mainly by artisanal boats and fishers, and to try to conserve the marine environment and to protect the fisheries resources of the country, large modern fishing vessels are, by decree, not allowed to operate in Iran's waters. The waters of southern Iran are characterized by a diversity of fish species. Demersal and pelagic species are differentially available during different times of the year.

According to a recent Iranian law, the Iranian Fisheries Company is responsible for managing and developing exploitation of the country's fisheries resources. Control of the fishery is through a comprehensive licencing policy, supported by legislation. The Iranian Fisheries Company collects catch and effort data by sampling, and it also collects catch and effort data from trawlers, purse-seiners, and longliners operating in the Oman Sea, using a logbook system. This study focuses on the Hormozgan Province in the northern Persian Gulf. Fishing vessels there consist of small fibreglass and medium-sized wooden vessels, which make short trips close to the coast, not far from their home base. In 2003, the major species caught were longtail tuna (Thunnus tonggol, 53.2%), Spanish mackerel (Scomberomorus comorus, 11%), kawakawa (Euthynnus affinis, 11.2%), yellowfin tuna (Thunnus albacares, 13.8%), and other species (10.8%). Gear and season are important factors in the catches of the different species. Because for 8 months of the year (autumn, winter, plus 2 months of spring), the gear deployed by small- and medium-sized vessels is similar (gillnets), the catch by species is similar among vessels.

The data for the study were derived from a field study of the fishery in Hormozgan Province. A sample of 142 fishing vessels was selected using stratified random sampling, and the vessels were classified into two groups of wooden (89 vessels) and fibreglass (53) construction. Information about the vessels and socio-economic data on the fishers were collected through face-to-face interviews with skippers during 2003.


   Results
Top
 Introduction
 Methods
 Results
Discussion
 References
 
Annual catches and vessel characteristics are listed in Table 1 along with information on the socio-economic background of the skippers.


View this table:
[in this window]
[in a new window]

  		Table 1 Vessel characteristics and socio-economic background of skippers in the study fleets of Hormozgan Province.

 
The fibreglass vessels are relatively small, with an average horsepower of 41.7, and 2.5 crew members each. By contrast, wooden vessels are larger, with an average horsepower of 85.7, and carrying on average 6.5 crew members. Although catches by wooden vessels are larger than those of fibreglass vessels, the operating costs of the former are higher. Skippers of wooden vessels are better educated, but fibreglass vessel ownership is more often private than company. Other socio-economic characteristics of the skippers of the two types of vessels are relatively similar (Table 1). Wooden vessels operate in a region that has better infrastructure and facilities. In the sample used here, 90% of wooden vessels operated out of an established port facility, and 75% used two-way radio.

For the analysis of efficiency and assuming a Cobb–Douglas functional form, the production function mentioned in Equation (1) can be written as


Formula 3

(3)
where i represents the i(th) vessel (i = 1, 2, ..., 142), yi the catch of the i(th) vessel, and Xj,i represents the input j required by the i(th) vessel. The dependent variables are defined as follows: X1i, the number of crew used in the i(th) vessel; X2i, the days operating by the i(th) vessel annually; X3i, the engine horsepower of the i(th) vessel; and X4i, the gear used by the i(th) vessel. The number of crew, the number of days operated, and the gear are directly related to fishing effort, whereas a greater engine horsepower allows a vessel to steam at higher speed and for longer distances from the home base (Fousekis and Kolonaris, 2003).

In order to determine differences in technical efficiency across vessels, six variables are used in the model. The inefficiency model used is:


Formula 4

(4)
where ui is the inefficiency of the i(th) vessel. Here Z1i denotes use of a two-way radio (Z1i = 1 if the vessel has a two-way radio, otherwise Z1i = 0), Z2i represents the level of education of the skipper (Z2i = 1 if skippers are well educated, otherwise Z2i = 0), Z3i denotes a skipper's experience in years, Z4i is the skipper's age, also in years, Z5i denotes use of GPS (Z5i = 1 if the vessel has GPS, Z5i = 0 if not), Z6i refers to whether or not the vessel is owner-operated (Z6i = 1 if the owner is its skipper, otherwise Z6i = 0), and Z7i represents financial security (Z7i = 1 if the owner has a loan on the vessel, otherwise Z7i = 0).

Skippers' skills and vessel characteristics are important factors in how a vessel performs. The skill of a skipper is a complex concept that has a profound influence on fisheries management models and the productivity of a fishery (Kirkley et al., 1995). Therefore, education, experience, and age are used as skipper-specific variables. Financial security and whether or not the vessel is owner-operated are also factors that closely relate to management and fishery performance. In addition, possession of a two-way radio and possession of a GPS are important predictors of how a vessel will perform.

The stochastic frontier and inefficiency models are estimated in a single stage by the econometric package FRONTIER 4.1 (Coelli, 1996).

The frontier model is used to investigate technical efficiency. Estimates of the coefficients of the frontier and inefficiency models are presented in Table 2. In both models, the coefficients estimated for most parameters have the anticipated impacts on production and efficiency. In the frontier model, the coefficients of labour, number of fishing days, and engine horsepower are significant and have the anticipated positive signs, implying that any increase in each variable would cause higher production. However, the coefficient for gear used was not according to expectation, and was not significant. This means that the amount of gear used by the fishing vessels is already greater than the optimum. The functional form used in the efficiency model was that of Cobb–Douglas, so the coefficients are elasticity. The elasticity for all input parameters is <1, meaning that a 1% increase in the input of each coefficient would cause a <1% increase in the fish catch. The output elasticities for labour, number of fishing days, and vessel horsepower were calculated as 0.42%, 0.52%, and 0.51%, respectively. This means that a 10% increase in labour, number of fishing days, and horsepower would lead to increases in the fish catch by 4.2%, 5.2%, and 5.1%, respectively. The elasticities estimated in the model also indicate increasing returns to scale. The return to scale for this fishery is calculated as 1.42, less than the 1.87 calculated by Sharma and Leung (1999), and substantially less than the 2.65 calculated by Garcia del Hoyo et al. (2004), but greater than the 1.26 deduced by Fousekis and Konaris (2003). A trans-log functional form was used to represent cross-elasticity, but based on the likelihood ratio test, the Cobb–Douglas specification was preferred (log-likelihood = 9.92, i.e. less than the critical value of 14.85). Likelihood ratio tests were also used to test the null hypothesis involving restriction on the variance parameters ({gamma}) in the stochastic production frontier model, and for the coefficients ({delta}k) in the inefficiency model. The null hypothesis that the technical inefficiency effect is absent, {gamma} = {delta}0 = {delta}1 = {delta}2 = {delta}3 = {delta}4, is rejected (log-likelihood = 24.72, greater than the critical value of 16.27). This means that there is technical inefficiency in the fishing industry. Further, the null hypothesis that vessel and skipper specifications do not influence the technical inefficiency, {delta}1 = {delta}2 = {delta}3 = {delta}4 = {delta}5 = {delta}6, is also rejected (log-likelihood = 22.84, greater than the critical value of 13.4). Finally, the null hypothesis that the technical efficiency effect is non-stochastic, {gamma} = 0, is also rejected (Table 3).


View this table:
[in this window]
[in a new window]

  		Table 2 Parameters estimated for the frontier and inefficiency models.

 


View this table:
[in this window]
[in a new window]

  		Table 3 Likelihood ratio test of the hypotheses for parameters of the stochastic production frontier and technical inefficiency.

 
The coefficients of the inefficiency model have the signs expected. A positive sign on the parameters in the inefficiency model implies negative effects on technical efficiency, and vice versa. The coefficients for two-way radio, GPS, owner-operation, skippers' experience, and skippers' level of education are significant, and have a negative sign, implying that these parameters have a positive effect on efficiency in the fishery. The coefficients for skippers' age and financial security are significant and have a positive sign, meaning that they result in lesser efficiency. The value of {gamma} suggests that the variance in production inefficiency effects accounts for 57% of the total variance in production. Although this value is not particularly large, it is still higher than that found by Fousekis and Konaris (2003). Values of {gamma} lie between zero and one, zero indicating that all deviations from the frontier are attributable to noise, and a value of one indicating that all deviations are attributable to technical inefficiency. The values estimated for {sigma}2, Formula , and Formula were 0.44, 0.19, and 0.25, respectively, implying that variance in the specific error term is greater than variance in the stochastic error term, and that a one-sided inefficiency random component dominates the measurement error and other random disturbances.

The overall technical efficiency calculated from the model, for the entire sample, is 78%. Comparison of the two vessel groups shows that the mean technical efficiencies for the wooden and fibreglass vessels were 85% and 60%, respectively. Statistical comparison reveals that this difference is significant (Table 4).


View this table:
[in this window]
[in a new window]

  		Table 4 Statistical comparison between the efficiency of fibreglass and wooden vessels.

 

   Discussion
Top
 Introduction
 Methods
 Results
 Discussion
References
 
Here, a frontier production model was used to assess the technical efficiency of fishing vessels operating in Iranian waters of the Persian Gulf. Both trans-log and Cobb–Douglas frontier production models are estimated using maximum likelihood estimation, but based on statistical criteria, the Cobb–Douglas form was preferred.

Although the coefficient for amount of gear operated was not significant, all other coefficients in the frontier model were significant, and had the expected signs. The output elasticity for labour, number of fishing days, and vessel horsepower is positive and <1, whereas the output elasticity of the amount of gear used is <0. An explanation for this is that vessels use more gear than is optimal, i.e. they operate in the third part of the production function. Fousekis and Konaris (2003) also found a negative relation between gear use and technical efficiency in Greece. Moreover, Kompas et al. (2003) concluded that gear length is negatively related to efficiency in Australia's prawn fishery. Labour and gear are two important variable inputs (and costs) in the fishing industry. The difference between output elasticity for labour and gear may relate to the wage payment regime in Iranian fisheries. As crew members receive a share of the fishing income, there is little or no attempt to increase the number of crew operating the fishing vessels.

The mean technical efficiency for the vessels in the sample is relatively low, implying that there is potential to increase productivity. The result also shows that, on the whole, larger wooden vessels are more efficient than smaller fibreglass vessels in the region, a finding similar to those of Sharma and Leung (1999) and Tingley et al. (2005) but contrary to that of Fousekis and Konaris (2003), who analysed trammel netters in Greece. Average vessel size in the Fousekis and Konaris (2003) study was 38% greater than in this study. However, although the technical efficiency of wooden vessels was greater than that of smaller fibreglass vessels, it would be erroneous to conclude simplistically that wooden vessels are far better than fibreglass vessels in this fishery. The reason is that the opportunity cost of investment (capital) was not included in the model, and additionally the two types of vessels can complement each other in their ability to catch the different species.

In the inefficiency model, the estimated coefficients reveal that the possession of two-way radio and/or possession of GPS are important influences on fishing efficiency; both help skippers to operate more efficiently. Further, the skipper's level of education and his experience are two important crew qualities that affect efficiency, both attributes helping skippers to allocate their input factors more efficiently. Owner-operation and relative youth of skippers are other attributes, which can make them more efficient than others in the fishery. A possible explanation for this is that owner-operated vessels and younger skippers are more willing to change their fishing patterns in order to succeed, and therefore take more risks than those who do not own their vessels or who are older. Tingley et al. (2005) also found young skippers in their study to be more efficient than older ones.

The overall finding of this work is that, if improvement of technical efficiency in the Persian Gulf fishery is to be an aim of the Iranian government, support in providing better electronic instrumentation, further education, and extension services would likely achieve the objective. The results of this research would then have helped to increase fishing efficiency in the region.


   Acknowledgements
 
I gratefully acknowledge the suggestions on an earlier draft of two anonymous reviewers, and the grammatical improvements by the Journal's editor.


   References
Top
 Introduction
 Methods
 Results
 Discussion
 References
 

    Aigner D.J., Lovell C.A., Schmidt P. (1977) Formulation and estimation of stochastic frontier production function models. Journal of Econometrics 6:21–37.[CrossRef][Web of Science]

    Battese G. and Coelli T. (1992) Frontier production functions, technical inefficiency and panel data: with application to paddy farmers in India. Journal of Productivity Analysis 3:153–169.[CrossRef]

    Battese G. and Coelli T. (1993) A stochastic frontier production function incorporating a model for technical inefficiency affects. Working Papers in Econometrics and Applied Statistics, 69. Department of Econometrics, University of New England, Armidale.

    Battese G. and Coelli T. (1995) A model for technical inefficiency effects in a stochastic frontier production function for panel data. Empirical Economics 20:325–332.[CrossRef]

    Battese G. and Corra G.S. (1977) Estimation of a production frontier model: with application to the pastoral zone of eastern Australian. Australian Journal of Agriculture and Resource Economics 21:169–179.

    Charnes A., Cooper W.W., Rhodes E. (1978) Measuring the efficiency of decision making units. European Journal of Operational Research 2:429–444.[CrossRef][Web of Science]

    Coelli T. (1996) A guide to Frontier, version 4.1. A computer program for frontier production function. CEPA Working Paper 96/07. Department of Econometrics, University of New England, Armidale.

    Farrell M.J. (1957) The measurement of production efficiency. Journal of the Royal Statistical Society 120:253–290.

    Felthoven R. (2002) Effects of the American Fisheries Act on capacity, utilization and technical efficiency. Marine Resource Economics 17:181–205.

    Fousekis P. and Konaris S. (2003) Technical efficiency determination for fisheries: a study of trammel netters in Greece. Fisheries Research 63:85–95.[Web of Science]

    Garcia del Hoyo J.J., Castilla Espino D., Jiménez Toribio J. (2004) Determination of technical efficiency of fisheries by stochastic frontier models: a case on the Gulf of Cádiz (Spain). ICES Journal of Marine Science 61:416–421.[Abstract/Free Full Text]

    Herrero I. (2005) Different approaches to efficiency analysis. An application to the Spanish trawl fleet operating in Moroccan waters. European Journal of Operational Research 167:257–271.[CrossRef][Web of Science]

    Herrero I. and Pascoe S. (2003) Value versus volume in the catch of the Spanish South-Atlantic trawl fishery. Journal of Agricultural Economics 54:325–341.[Web of Science]

    Kirkley J., Squires D., Strand I. (1995) Assessing technical efficiency in commercial fisheries: the mid-Atlantic sea scallop fishery. American Journal of Agricultural Economics 77:686–697.[CrossRef][Web of Science]

    Kodde D.A. and Palm F.C. (1986) Wald criteria for jointly testing equality and inequality restriction. Econometrica 54:1243–1248.[CrossRef][Web of Science]

    Kompas T., Che T. N., Grafton R. Q. (2003) Technical efficiency effects of input controls: evidence from Australia's banana prawn fishery. Working Paper No. 03-3. Australian National University.

    Meeusen W. and van den Broeck J. (1977) Efficiency estimation from Cobb–Douglas production function with compared error. International Economics Review 18:435–450.[CrossRef]

    Pascoe S., Andersen J., Wilde J.D. (2001) The impact of management regulation on technical efficiency of vessels in the Dutch beam trawler fishery. European Review of Agricultural Economics 28:187–206.[Abstract]

    Pascoe S. and Herrero I. (2004) Estimation of a composite fish stock index using data envelopment analysis. Fisheries Research 69:91–105.[CrossRef][Web of Science]

    Sharma K. and Leung P. (1999) Technical efficiency of the long-line fishery in Hawaii: an application of a stochastic production frontier. Marine Resource Economics 13:259–274.

    Tingley D., Pascoe S., Coglan L. (2005) Factors affecting technical efficiency in fisheries: stochastic production frontier versus data envelopment analysis approaches. Fisheries Research 73:363–376.[CrossRef][Web of Science]

    Yazdani S. and Esmaeili A. (1995) Investigating the economic efficiency of fishery industry in Bandar Lengeh. Iranian Journal of Agricultural Science 26:41–47.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Esmaeili, A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Esmaeili, A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?