Abstract
The countries’ factor endowments and use of advanced technology are important parameters for decreasing costs relating to agric ultural production. As regards competitiveness in cotton markets, reducing cost of production is no less a requisite than attaining a high level of productivity. Mechanization of harvest is of immense importance for decreasing cost of cotton production. Manual harvesting is the most common harvesting method in Turkey. However, harvest mechanization is becoming more commonplace, owing to higher costs. This study performs an ec onomic analysis of cotton production on basis of farms utilizing various methods of harvesting in Turkey. Moreover, in the study, the factors affecting producers’ adoption of harvest mechanization were investigated. There are two different methods adopted for harvest mechanizati on, the first one being through use of rented harvesting machines and the second one through using proprietary equipment. According to research, the mean of cotton areas harvested using manual methods, rented equipment and proprietary equipment are 7.44 ha, 14.12 ha and 44.55 ha, respectively. The corresponding unginned cotton yields are 3717.37 kg ha-1, 3444.05 kg ha-1, and 3426.76 kg ha-1, respectively. The average total cost is 3855.98 US$ ha-1 for farms harvesting manually, while 3412.38 US$ ha-1 for farms opting to rent a harvesting machine and 3610.63 US$ ha-1 for farms utilizing their own harvesting machines. The gross margins are 159.69 US$ ha-1, 306.54 US$ ha-1 and 374.75 US$ ha-1, respectively, with net profits being -1155.69 US$ ha-1, -970.55 US$ ha-1 and -1226.98 US$ ha-1, respectively. The net profits of farms utilizing rented equipment become positive with government subsidies. Cotton farmers are able to continue producing only with government support. Renting a harvesting machine is the most suitable method for harvesting, vis-à-vis owning the equipment. Cotton harvesting using rented equipment is becoming commonplac e. The size of the cotton field, the experience of the farmer and age of farmer are important variables affecting farmers’ decision to adopt m echanized harvesting. The size of cotton area and farmer’s age have a positive effect on farmers’ preference for harvesting using rented equipment, r ather than harvesting manually.
Key words: Cost analysis, harvest methods, net profit analysis, gross margin, multinomial logit. www.world-food.net
Introduction
Cotton (Gossypium hirsitum L.) is an important crop product in the world economy and trade. It is a main source of income for many developing countries in Asia and West and Central Africa18. The largest producer of cotton in the world is China, followed by India, the US and Uzbekistan. Turkey ranks in seventh place in the world cotton production. China consumes the highest amount of cotton, followed by India, Pakistan and Turkey, in this order. In global importation of cotton, China ranks first and Turkey second whereas in exportation of cotton the US and India rank in first place 19.
The international cotton market has become a global market and is influenced by globalization. Every producer, consumer, exporter or importer country asserts influence on the market. It is considered that cotton markets operate freely and comparative advantages have prevalence. However, internal interventions and support policies implemented by producer countries have their repercussions on the markets. Low prices in international cotton markets complicate competition.
Turkey is an important importer of cotton. In 2006, the share of the USA in Turkey’s total cotton imports is approximately 56%, and Greece’s share is approximately at 22% level. Turkmenistan, Syria and Uzbekistan follow 17. Subsidies implemented by the US and the EU on cotton have significantly reduced prices in foreign markets.
The most important factors for competing in cotton markets are the use of advanced technology, high productivity, low cost and government subsidies. Monitoring cost is of ultimate importance for setting cotton policies, also. In the USA, cost analyses are carried out on cotton and costs are monitored 20. The EU also carries out some studies on cotton production costs 11. In EU member states (Greece and Spain) and in the USA, cotton harvest is mechanized. Harvest mechanization is of most importance for reducing costs.
In recent years, cotton producers in Turkey started facing difficulties in competing against imported cotton in domestic markets. While cotton production fields diminish particularly in Aegean and Mediterranean regions of Turkey, they expand in the Southeast Anatolia region 16. In the Aegean region, cotton producers are abandoning cotton production; the continuation of cotton production depends mostly on reducing production costs. The most significant item of cost in cotton production in Turkey is the expenses related to temporary workers. The most labor-intensive phase of production is the harvest. Harvest mechanization is an important factor in reducing costs of production. This study performs a comparative economic analysis of enterprises utilizing harvesting machines as opposed to those harvesting manually to underline the importance of harvest mechanization as regards the cost. Moreover, the study analyses factors of affecting producers’ adoption of harvest mechanization.
Material and Methods
This study covers the Aegean provinces of Izmir, Aydin, Denizli, Manisa and Mugla. These provinces account for 99.9% of total cotton production in the Aegean Region and 21.6% of total production in Turkey. TARIS (Figs, Raisins, Cotton and Oil Seeds Agricultural Sales Cooperatives Unions) members constitute research population. TARIS currently has 44 cooperative units in the Aegean Region, 19 of which have been included in this study. Research population consists of 14432 producers 15. Using the formula below 14, a sample size has been determined at 90% significance and with a 5% error margin. 266 farmers have been interviewed. ) 1( ) 1 () 1(2p p Np Npnpx ) 1( ) 1 () 1(2p p Np Npnpx
Interviewed farmers were specialized in cotton production. The data obtained from producers relate to the production year 2007. In the region, cotton seeds are sown in April and harvesting begins in September and continues until November. Activities relating to marketing and sale of the produce are completed by the end of January. Farmers market the harvested cotton in unginned form.
The costs of cotton production were classified into variable costs and fixed costs. The variable costs include those relating to seed, fertilizer, agricultural chemicals, fuel, tools and machinery repair and maintenance, temporary workers, harvesting machine rent, water etc. items. Variable costs are calculated over farm-gate prices. The interest accruing on variable costs are included in the costs. When calculating the interest rate, the reference has been the rate which Agricultural Bank of Republic of Turkey applies to agricultural loans (i.e. 13%). Variable costs spread homogenously along the production period, which is six months for cotton. Therefore, the interest rate was applied to be 3.25% of variable costs.
Fixed costs include depreciation, interest on fixed capital assets, land rent, wages of permanent labor and family labor, interest, insurance and property tax. Depreciation has been calculated on fixed capital assets of studied farms. The figure corresponding to the value of capital assets less salvage value was divided to the economic utility life to calculate depreciation 2, 8, 10. Fixed capital assets make up a substantial portion of a farm’s total assets. Therefore, it is necessary to also calculate interest cost on fixed capital assets. The mean investment amount has been calculated for fixed capital assets.
The interest rate applied to fixed capital is the real interest rate 1, 10. The real interest rate is calculated as 5%, considering the net interest applied to savings deposits and the overall inflation rate 10. The overhead is calculated as 3% of total variable costs 10.
The findings of the study have been presented by making a distinction between total costs, variable costs and fixed costs. Gross production value has been calculated in reference to farm- gate selling price of unginned cotton and cotton production levels. Gross margin has been calculated by subtracting variable costs from total gross production value. Net profit has been calculated by subtracting total costs from total production value.
Farms utilize three different forms of harvesting, being manual harvesting, harvesting using rented machinery and harvesting using proprietary machinery. The study results have been compared based on harvesting methods: 53.4% of cotton farms utilize the manual harvesting method (Group MH), 39.5% harvest using rented machinery (Group RM) and 7.1% own their harvesting machines and harvest using their own means (Group OM). Farms that harvest using harvesting machines time to time employed temporary workers during the harvest. Conformance of variables to the normal distribution was determined using the Kolmogorov–Smirnov test. When comparing group averages derived from farm averages, the one-way ANOVA has been applied to variables conforming to the normal distribution.
Since there are three different methods of harvesting utilized during cotton production, the multinomial (polytomous) logit model has been used for evaluating factors that affect producers’ choice of harvesting method. The harvesting methods preferred by farmers were taken as the dependent variable.
In multinomial logistic regression, for dependent variables in M number of categories, M-1 number of equations is needed, which defines the relationship between dependent and independent variables and which compares each category with the reference category. Other than the reference category, the equation for each category of dependent variables is as follows 3, 5, 13.: where the subscript k refers, as usual, to specific independent variables X and the subscript h refers to specific values of the dependent variable Y for the reference category, The probability that Y is equal to any value h other than the excluded value h0 is and for the excluded category h0=M or 0,
The dependent variable (Y) is farms’ harvesting method. Manually harvesting farms = 0 (reference category), producers harvesting using rented machinery = 1, producers harvesting using own machinery = 2. Accordingly, a model of 3 categories has been created. The model has two logit functions. In the first function Y= 1 (harvesting using rented machinery) against Y = 0 (manual harvesting) and in the other, Y = 2 (harvesting using own machinery) against Y = 0 (manual harvesting). The group where Y= 0 is designated as the reference group.
The study incorporates eight independent variables (X) into the model. These variables comprise variables relating to the farms 012 ( , ,..., ) 1kgXX X 11 2 211 2 2(...)12 1( ... )1(,,...,)1hh h hkkhh h hkkabXbX bXk MabXbX bXhePY h X X Xe _ ¦ 1, 2,..., 1 hM , (0.2) 11 2 2012 1(...) 11( , ,..., )1hh h hkkk MabXbX bXhPY h X X Xe _ ¦1, 2, ... , 1 hM . ( 0.3) 11 2 2( ..... )12 ( , ,...., )hh h hkk abXbX bXhk gXX X e 1, 2,....., 1, hM (0.1) Journal of Food, Agriculture & Environment, Vol.7 (2), April 2009 389 and those relating to the producers. Mean and standard deviation values relating to explanatory variables are provided in Table 1. Variables relating to the farms are the total farms area, cotton area, number of cotton parcels, average cotton parcel size and the farm population. The variables relating to the farmer are the farmer’s age, education level and cotton production experience.
The hypothesis that harvest mechanization would have a reducing effect on the production cost has been tested in this study. Accordingly, the costs, gross margin and net profit per hectare were compared on basis of harvesting methods.
Moreover, it is considered that farmers’ preference of mechanized harvesting is positively affected by farm size and cotton area. It is estimated that number of parcels has a negative effect whereas the parcel size has a positive effect in opting for mechanized harvesting. On the other hand, a farmer’s cotton production experience and education level are also thought to have a positive effect on adopting harvest mechanization.
Results
Economic analysis results: The average farm size was 20.65 ha, while average cotton production area was 12.73 ha. The average farm sizes and cotton production areas are shown in Table 1 by harvesting method. Farms owning their harvesting machines have the largest cotton production area with 44.55 ha.
The average production costs of cotton in farms harvesting manually, rental machinery and their own harvesting machines were 3855.98 US$ ha-1, 3412.38 US$ ha-1 and 3610.63 US$ ha-1, respectively (Table 2). According to one-way ANOVA, performed over average cost of farm per hectare by calculating group averages, the difference between the groups was significant at 0.05 probability level (F = 13,903, p<0.05). According to Duncan test results, the group averages are particularly diverse for manually harvesting farms and those using rental machinery, while averages for farms using proprietary machinery are close to both groups.
The variable and fixed production costs are 65.89 and 34.11%, respectively, for farms harvesting manually, 62.57 and 37.43%, respectively, for those harvesting using rented machinery and 56.83 and 43.17%, respectively, for farms that have their own harvesting machinery. The greatest portion of variable costs incurred by manually harvesting farms consists of labor costs with approximately 45%, of labor cost 67.42% consists of labor needed during the harvest. In farms harvesting using rental machinery, the renting cost has the largest share among variable cost items with 19.58%. The entire cost of machine rental incurred by these farms relates to rentals for harvesting purposes. For farms harvesting using own machinery, the variable cost item having the largest share was fuel costs with 29.01%. The harvest is an important factor in the fuel costs being high. The fuel consumed during harvesting takes up 24.29% share.
Input utilization was more intense in farms that mechanize harvesting as compared to manually harvesting producers. Other than diesel fuel, fertilizers and chemicals were consumed at higher levels. Moreover, the highest irrigation costs were incurred by farms that mechanize harvesting. Average variable costs were highest in manually harvesting farms while lowest in farms harvesting using own machinery. Particularly, the difference in the method of harvesting was the most important cause of the foregoing observation. Table 3 itemizes farms’ harvesting costs.
The most important fixed cost item is the land rent. Farms having their own harvesting machines have the largest fixed costs. The most important reasons for that are the depreciation costs and interest costs on the investment. In this group of farms, cost of depreciation and interest on investment constitute 32.23% of average fixed costs, causing an increase in average total costs. The cost of family labor is high in manually harvesting farms. These are small farms which intensely rely on family members’ labor.
Cotton yield is among important factors that affect gross margin and net profit. The average unginned cotton yield in farms harvesting manually, rental machinery and their own harvesting machines were approximately 3718.37, 3445.05 and 3426.76 kg ha-1, respectively. Means are not significantly different at the 0.05 level of probability according to one-way ANOVA (F = 0.716, P>0.05).
Average unginned cotton selling prices are 0.7262 US$ kg-1 for farms harvesting manually, while 0.7090 US$ kg-1 for farms opting to rent a harvesting machine and 0.6956 US$ kg-1 for farms utilizing their own harvesting machines.
Average gross margin obtained from cotton in farms which harvest using own machinery was 374.75 US$ ha-1 and highest. When subsidies for cotton are included, this figure climbs to 1411.59 US$ ha-1. In contrast, manually harvesting farms have the Variable Unit Manually harvesting farms (Group MH) (Y= 0) (n =142) Farms harvesting using rented machinery (Group RM) (Y= 1) (n = 105) Farms harvesting using own machinery (OM) (Y=2) (n=19) General total (n = 266) Mean Standard deviation Mean Standard deviation Mean Standard deviation Mean Standard deviation Farm size (FS) ha 15.46 14.39 20.41 16.54 60.71 55.25 20.65 23.54 Cotton production area (CPA) ha 7.44 8.59 14.12 11.99 44.55 31.85 12.73 15.90 Number of parcel in cotton production area (CPN) Number 3.37 3.12 5.19 4.65 10.00 10.00 4.56 4.84 Average cotton parcel size (APS) ha 2.51 2.28 3.10 2.20 7.55 12.97 3.10 4.21 Family size (FAS) Number 4.54 1.88 4.82 1.66 5.68 2.56 4.73 1.87 Age of farmers (AGE) Years 47.94 10.22 47.69 9.45 45.63 11.01 47.67 9.96 Education level of farmers (EL) Years 6.91 2.82 7.34 2.92 8.00 3.33 7.16 2.90 Cotton production experience of farmers (EXP) Years 28.89 9.61 26.85 9.97 29.26 13.19 28.11 10.05 Table 1. Sample means and standard deviation of variables according to harvest methods. lowest gross margin figures (159.65 US$ ha-1) (1274.64 US$ ha-1 including subsidies) (Table 4). According to one-way ANOVA results, the means are significantly different at the 0.05 level of probability (F = 3.180, p<0.05).
In all groups, net profit was negative. The group sustaining the least amount of loss in terms of net profit is the group of farms that harvest using own machinery. When government subsidies are included, the net profit figure becomes positive for this group (70.93 US$ ha-1) (Table 5). Results of multinomial logit model: The independent variables included in the multinomial logit model are provided in Table 1.
Results of the model estimation are shown in Tables 6. The likelihood ratio test statistic indicates that the model is statistically significant. The model’s chi-square probability is below the 0.05 significance level. The results obtained from the model support existence of a relationship between independent variables and dependent variables.
Multinomial logit results yielded separately two equations (Table 6). The first equation demonstrates the distinctiveness of independent variables between the group harvesting using rental machinery (Group RM) and the manually harvesting group (Group MH). The second equation demonstrates the distinctiveness of independent variables between the group harvesting using own machinery (Group OM) and the manually harvesting group (Group MH).
According to the first equation’s estimations, the farm size (FS) (p<0.05), cotton production area (CPA) (p<0.05), age of farmers (AGE) (p<0.10) and cotton production experience of farmers (EXP) (p<0.10) significantly distinguish manually harvesting group from the group harvesting using rental machinery. The farmer size and experience have a negative effect, whereas cotton production area and farmer’s age have a positive effect.
Each unit of increase in the farm size reduces the odds that a farmer would harvest using rented machinery by 3.2% (0.968-1). Each unit of increase in farmer’s experience reduces the odds that a farmer would harvest using rented machinery by 4.7% (0.953-1). Each unit of increase in farmer’s age increases the odds that a farmer would harvest using rental machinery by 4.8% (1.048-1). Each unit of increase in the cotton production area increases the odds that a farmer would harvest using rental machinery by 10.9% (1.109-1). Table 2. Cotton production costs per planted hectare (US$ ha-1), excluding government payments. Item Group MH Group RM Group OM General total Total cotton area (ha) 1055.95 1482.90 846.40 3385.25 Number of farms (n) 142 105 19 266 Variable costs x Seed 105.21 91.68 93.35 96.32 x Fertilizer 296.16 313.27 315.01 308.37 x Chemicals 70.14 106.86 162.37 109.28 x Fuel 435.41 389.49 582.80 452.15 x Repair 131.10 97.01 112.28 111.46 x Labor 1142.42 342.29 342.54 591.93 x Machinery rent 0.00 418.08 0.00 183.14 x Irrigation 276.45 304.82 332.02 302.77 x Product insurance 1.26 0.98 5.29 2.15 x Others 2.51 3.59 0.00 2.36 x Interest on variables costs 79.97 67.22 63.24 70.20 Total variable costs 2540.63 2135.29 2008.90 2230.13 Fixed costs x Depreciation 215.46 208.78 415.01 262.43 x Opportunity interest cost of the financial capital 78.51 76.62 149.32 95.39 x Opportunity cost of land (rent) 761.86 819.78 840.30 806.84 x Land tax 1.86 0.92 4.39 2.08 x Opportunity cost of unpaid labor (for family) 174.03 94.15 122.10 126.05 x Keeper fee 7.42 12.78 10.35 10.50 x General farm overhead 76.21 64.06 60.26 66.90 Total fixed cost 1315.35 1277.09 1601.73 1370.19 Total costs 3855.98 3412.38 3610.63 3600.32 Table 3. Cotton harvest costs by machine and by hand (variables costs) per hectare (US$ ha-1). Cost item Group MH Group RM Group OM General tota l Temporary labor 770.20 38.27 30.97 264.75 Machinery rent 0.00 418.08 0.00 183.13 Fuel 0.00 0.00 141.54 35.39 Total 770.20 456.35 172.51 483.
According to the second equation’s estimations, the cotton production area (CPA) (p<0.05), farmer age (AGE) (p<0.10) family size (FAS) (p<0.10) and farmer experience (EXP) (p<0.10) meaningfully distinguish manually harvesting group from the group harvesting using own machinery. Farms owning harvesting machines are large-scale farms, which have an average cotton field size of 44.55 ha. The cotton production area, the family size and farmer’s experience have a positive effect while farmer’s age has a negative effect. As the cotton production area in farms increase, the odds that harvest would be mechanized using rented machinery increases. Each unit of increase in the cotton production area increases the odds by 23.2% that a farmer would be in the group that harvests using own machinery (1.232-1). Each unit of increase in the family population increases the odds by 29.1% that a farmer would be in the group that harvests using own machinery (1.291-1). Each unit of increase in farmer experience increases the odds by 15.3% that a farmer would be in the group that harvests using own machinery (1.153-1).
Each unit of increase in farmer age reduces the odds by 24.1% that a farmer would be in the group that harvests using own machinery (0.859-1).
Discussion
Cotton production per hectare varies by farms, on regional scale and even between countries. In this study, mean cotton yield is calculated as 3525.3 kg ha-1. For year 2007, the overall yield in Turkey is estimated to be 4290 kg ha-1, whereas for the Aegean Region it is estimated at 3570 kg ha-1. In Turkey 16, the average cotton fiber yield is 1650 kg ha-1. In year 2007, the cotton yield fell in the Aegean Region due to drought. The mean unginned cotton yield of the studied farms is close to that of the Aegean Region. Although it varies from region to region 11, the yield in Greece, an EU member state, is approximately 3200–3500 kg ha-1. The figure for Spain11 is approximately 3700 kg ha -1. In the USA, however, the fiber cotton yield is 817 lb acre-1 (915.7 kg ha-1) and seed yield 1315 lb acre-1 (1490 kg ha-1) (year 2005 figures) 20. In China, a prominent cotton producing country, seed cotton yield is 3467 kg ha-1, cotton fiber yield is 1058 kg ha-1 whereas in India seed cotton yield is 751 kg ha-1 and cotton fiber yield is 312 kg ha-1 18. In this study the average total costs were found to be lowest in farms harvesting using rental harvesting machinery. In these farms, the cost of labor, depreciation and interest on fixed capital assets are lower. The highest average total cost is incurred by manually harvesting farms (3855.90 US$ ha-1), due mostly to manual labor employed during the harvest. In farms that have their own harvesting machines the cost of depreciation and interest on fixed capital assets increase (average total cost 3610.63 US$ ha-1).
In this study, the total overhead is 3600.32 US$ ha-1. In the USA, the total expense figure for cotton production in year 2005 was 557.24 US$ acre-1 (1377 US$ ha-1) 20. In the Macedonia- Thraki region of EU member state Greece, the average total cost incurred for cotton production is 2697 € ha-1, while in Thessallia/Sterea Ellas region it is 2455 € ha-1 as compared to 3070 € ha-1 of Spain 11.
In this study, the gross margin and net profit are 277.76 US$ ha-1 and -1092.43 US$ ha-1, respectively. With subsidies, the gross margin and net profit become 1341.01 US$ ha-1 and -29.12 US$ ha-1, respectively. According to a study by LMC, in 2005 the gross margin and net profit in Greece’s Macedonia- Thraki region were 1470.1 € ha-1 and -57.8 € ha-1, respectively, and in Ipiros-Peloponi/ Thessallia/Sterea Ellas region, 2030.2 € ha-1 and 684.8 € ha-1, respectively. In Spain, the figures were 2058.1 € ha-1 and 720.3 € ha-1 (unginned cotton prices include subsidies per pre-2006 regime)11 (the exchange rate on 30.12.2005, €1 = US$ 1.1842 ) 4. In the USA, the gross margin figure for year 2005 was 107.43 US$ acre-1 (265.5 US$ ha-1) and net profit -87.54 US$ acre-1 (-21.63 US$ ha-1).
Conclusions
Cotton production is not profitable in the Aegean Region of Turkey. Cotton subsidies allow farmers to continue producing cotton. In some years, the net profit from cotton turns out negative even with subsidies. Farmers continuing to produce cotton thought Item Group MH Group RM Group OM General total Support premium* (US$ ha-1) 996.52 923.01 918.37 944.78 Direct income support (DIS) (US$ ha-1) 53.85 53.85 53.85 53.85 Diesel support (US$ ha-1) 41.54 41.54 41.54 41.54 Chemical fertilizer support (US$ ha-1) 23.08 23.08 23.08 23.08 Total support (US$ ha-1) 1114.99 1041.48 1036.84 1063.25 Table 5. Government payments for cotton. Source: MARA 12. General Directorate of Agricultural Production and Development, http://www.tugem.gov.tr/tugemweb/destekler.html (in Turkish). * The premium (in use certified seed) was 0.268 US$ kg-1. Premium per hectare derived from premium per kg related to yield. Table 4. Cotton production returns, gross margin and net profit per planted hectare, excluding government payments. Item Group MH Group RM Group OM General tota lCotton production (kg ha-1) 3718.37 3444.05 3426.76 3525.29 Average sale price for cotton (from farms) (US$ kg-1) 0.7262 0.7090 0.6956 0.7114 Gross production value (US$ ha-1) 2700.28 2441.83 2383.65 2507.89 Variable costs (US$ ha-1) 2540.63 2135.29 2008.9 2230.13 Fixed costs (US$ ha-1) 1315.35 1277.09 1601.73 1370.19 Total costs (US$ ha-1) 3855.98 3412.38 3610.63 3600.32 Gross margin(US$ ha-1) 159.65 306.54 374.75 277.76 Net profit (US$ ha-1) -1155.69 -970.55 -1226.98 -1092.43 odel fitting criteria Likelihood ratio tests -2 Log likelihood Chi-Square df Sig. Intercept only 473.745 Final 376.595 97.150 16 0.000 Cox Snell 0.306 Nagelkerke 0.368 McFadden 0.205 Effect Model fitting criteria Likelihood ratio tests -2 log likelihood of reduced model Chi-Square df Sig. Intercept 382.914 6.319 2 0.042* FS 380.810 4.216 2 0.122 CPA 396.578 19.983 2 0.000* CPN 378.401 1.807 2 0.405 APS 377.231 0.637 2 0.727 FAS 380.022 3.427 2 0.180 AGE 387.391 10.796 2 0.005* EL 376.859 0.264 2 0.876 EXP 386.845 10.251 2 0.006* Parameter estimates Harvest method B Std. error Wald df Sig. Exp(B) Intercept -2.475 1.018 5.910 1 0.015* FS -0.033 0.017 3.920 1 0.048* 0.968 CPA 0.104 0.037 7.666 1 0.006* 1.109 CPN 0.035 0.073 0.231 1 0.631 1.036 APS -0.002 0.113 0.000 1 0.984 0.998 FAS 0.113 0.082 1.878 1 0.171 1.119 AGE 0.047 0.024 3.734 1 0.053** 1.048 EL 0.020 0.053 0.139 1 0.709 1.020 Group RM (1.00) vs Group MH (0.00) EXP -0.048 0.025 3.669 1 0.055** 0.953 Intercept -2.763 2.275 1.475 1 0.225 FS -0.033 0.035 0.911 1 0.340 0.968 CPA 0.208 0.053 15.726 1 0.000* 1.232 CPN -0.064 0.100 0.415 1 0.519 0.938 APS -0.124 0.171 0.521 1 0.470 0.884 FAS 0.255 0.154 2.751 1 0.097** 1.291 AGE -0.152 0.080 3.625 1 0.057** 0.859 EL 0.055 0.124 0.199 1 0.656 1.057 Group OM (2.00) vs Group MH (0.00) EXP 0.143 0.079 3.274 1 0.070** 1.153 Table 6. Multinomial logistic regression results. The reference category is: Group MH (0.00). *) Significant at p<0.05, **)Significant at p<0.10 that cotton as the production technique they know best and state that investments they make on their tools and equipment is suitable for cotton. Moreover, the availability of selling guarantee in cotton and it being possible to store the produce influences cotton farmers’ decision to continue producing it. It was also observed that farmers in the studied region were investigating alternative products. These findings confirm the results of earlier studies 9, 21.. In certain provinces of the Aegean Region, farmers incline toward crops like tomato, which has a high gross margin (a study by Engindeniz 6 in 2007 found the net profit from growing tomatoes to be 1794 US$ ha-1), industrial tomato and corn.
In the USA and EU member states, cotton harvest is entirely mechanized, whereas in Turkey manual harvesting is prevalent. With the augmenting of unginned cotton yield and with greater penetration of mechanized harvesting, it shall be possible to elevate profitability. However, optimal field size levels are yet to be attained for using proprietary harvesting machines. Using rental machinery is seen to be the most suitable method of harvesting. The number of farmers opting to rent machinery for harvesting is increasing. The farm size and cotton production experience have a negative effect while cotton field size and farmer’s age have a positive effect on the odds that a farmer would be in the group of farmers harvesting using rental machinery. The cotton field size, family population and production experience have a positive effect, while farmer’s age has a negative effect on the odds that a farmer would be in the group of farmers that own and harvest using own harvesting machines (harvesting machine owners). It may be recommended that farmers should jointly acquire machinery and create machinery pools so as to reduce costs. It may also be recommended that cooperatives should acquire harvesting machines to promote harvest mechanization.
Acknowledgement
This study was conducted under the research project “Calculating Cotton Production Costs and Gross Margin and Factors Affecting Costs in the Aegean Region for year 2007”, undertaken by TARIS (Figs, Raisins, Cotton and Oil Seeds Agricultural Sales Cooperatives Unions, the Cotton and Oil Seeds Union) and Ege University, Agricultural Faculty, Department of Agricultural Economics. We extend our gratitude to Cotton and Oil Seeds Union Board for their support. We also thank TARIS R&D department staff for their assistance during our field studies and farmers who cordially responded to our questions.
References
1AAEA 2000. Commodity Costs and Returns Estimation Handbook. A Report of the AAEA Task Force on Commodity Costs and Returns, February 1, 2000 Ames, Iowa. 2AERI 2001. Input Use and Production Costs for Important Crops in Some Regions of Turkey. Agricultural Economics Research Institute, Ankara,Turkey, Research Report, Publication No. 64, 248 pp. (In Turkish). 3Borooah, V. K. 2002. Logit and Probit, Ordered and Multinomial Models, Series: Quantitative Applications in the Social Sciences. Sage Publications (CA), Paper 138. 4CBRT (Central Bank of The Republic of Turkey) 2008. Electronic Data Delivery System (http://evds.tcmb.gov.tr). 5Christensen, R. 1997. Log-Linear Models and Logistic Regression. 2nd edn. Springer-Verlag, New York. 6Engindeniz, S. 2007. Economic analysis of processing tomato growing: The case study of Torbali, West Turkey. Spanish Journal of Agricultural Research 5(1):7-15. 7Inan, H. 1994. Agricultural Economics. Trakya University, Tekirdag Agricultural Faculty, 259 pp. (in Turkish). 8Inan, H. 1999. Farm Management and Planning. Tekirdag. 9Isin, F. 2000. A Study on Reflection of Subsidy Policies Implemented on Producers in a Selected Region within the Province of Izmir in Recent Years; Producers’ Opinions. Ege Univ. Agricultural Faculty, Department of Agricultural Economics, Izmir (in Turkish). 10Kiral, K., Kasnakoglu, H., Tatlidil, F., Fidan, H. and Gündogmus, E. 1999. A Guide of Cost Calculation Methodology and Database for Agricultural Products. AERI, Ankara (in Turkish). 11LMC International 2007. Study on the Cotton Sector in the European Union. Final Report, SFSA607. 12MARA (Ministry of Agriculture and Rural Affairs) 2008. General Directorate of Agricultural Production and Development, http:// www.tugem.gov.tr/tugemweb/destekler.html (in Turkish). 13Menard, S. W. 2002. Applied Logistic Regression Analysis. 2nd edn. Quantitative Applications in the Social Sciences. Sage Publications, CA, No.106. 14Newbold, P. 1995. Statistics for Business and Economics. Prentice Hall Int., USA. 15TARIS (Figs, Raisins, Cotton and Oil Seeds Agricultural Sales Cooperatives Unions) 2008. TARIS Records. 16TURKSTAT 2008. Turkish Statistical Institute (www.turkstat.gov.tr). 17UN Comtrade 2008. http://comtrade.un.org. 18UNCTAD 2008. Infocomm, Information on Cotton (http:// www.unctad.org/infocomm). 19USDA 2008a. FAS (Foreign Agricultural Service), Circular Series (www.fas.usda.gov). 20USDA 2008b. ERS, United States Department of Agriculture, Economic Research Service, http://www.ers.usda.gov/Data/CostsAndReturns/ methods.htm. 21Uzmay, A. 2003. A Study on the Results of Price and Income Policies Implemented in Agriculture in Turkey after 1980. Doctoral thesis, Ege Univ., Faculty of Agriculture, Department of Agricultural Economics, Izmir, 365 p. (in Turkish).
