The substitutability among Japanese, Taiwanese and South Korean fronzen tuna

The substitutability among Japanese, Taiwanese and South Korean fronzen tuna By: Lih-Chyun Sun * Hsi-Chiang Liu** Li-Fen Lei * Wann-ChangShyu**** * Associate Professor. Department of Agricultural economics, National Taiwan University, No. 1, Sec. 4. Roosevelt R-d.. Taipei. Taiwan. R. 0. C. ** Professor. Institute of Oceanography. National Taiwan University- No, 1. Sec. 4, Roosevelt Rd.. Taipei. Taiwan. R. 0. C. *** Graduate Student. Institute of Oceanography, National Taiwan University, No. 1, Sec. 4. Roosevelt Rd., Taipei. Taiwan. R. 0. C. The substitutability among Japanese, Taiwanese and South Korean Frozen tuna Lih-Chvun Sun. Hsi-Chianc Liu. Li-Fen Lei-Wann-Chang Shyu I- Introduction Japan is the number 1 country in the world in terms of tuna harvest[1]. For example, it started from about 294,961 tons in 75, then reached its peak at 371,103 tons in 85. Although the harvest declined since 85, it bounced back for the recent years and reached 343,611 tons in 93 (Table 1 and Figure I). While Japan has the largest tuna harvest in the world, it relies on import for both fresh and frozen tuna to meet its domestic demand. The amount imported has increased from about 84,703 tons in 81 to around 273,399 tons in 93 (Table 2) of which the frozen tuna has increased from about 73,000 tons to about 184,000 tons. For domestically demanded frozen tuna, Thunnus albacares (i.e. Yellow Fin tuna) and Thunnus obsess (i.e. Big Eye tuna) are the two major categories. Other than domestic supply, Taiwan and South Korea are the two major suppliers among about fifty countries. Starting 90, Taiwan exceeded South Korea and became the largest foreign frozen tuna supplier to Japan. The Taiwanese tuna deep-sea long-line fisheries has been an important part of its fishery industry. Compared to Taiwanese deep-sea fishery, the Tuna fishery took about 50% of the value before 91. Recently, the ratio increased to 62.9% (92), (75.0) (93) and dropped a little bit to 68.3% in 94 (Table 3). As compared to the fishery industry as a whole, Taiwanese tuna fishery took more than 20% of the total value for the past several years (Table 4). In order to develop the Japanese tuna sashimi market, the ultra low temperature long-liners was promoted in the Taiwanese tuna fishery industry at 76, and the number of ultra low temperature long-liners fishing vessels had reached 334 at 94. Although the number of Taiwanese ultra low temperature long-liners fishing

vessels is still far less than Japanese (716 in 94), it exceeded South Korean since 90 (Table 5 and Figure 3). In order to keep the reasonable profit level for all Japanese, Taiwanese and South Korean tuna fishery industries, protect the welfare of the Japanese consumers, maintain the sustainable harvest rate and the stability of the Japanese tuna sashimi market, it is very important to understand the Japanese tuna sashimi market, especially the demand side of the market. Although the demand for tuna sashimi can be divided into "fresh'' and "frozen" tuna sashimi, it is the frozen tuna sashimi that had been imported the most. Among the imported frozen tuna, Big Eye and Yellow Fin tuna are the two major categories which are also the major harvest for both Taiwanese and South Korean ultra low temperature long-liners fishing vessels. In this study, we concentrate on the Japanese tuna sashimi demand. To be more specific, the Japanese demand for Japanese, Taiwanese and South Korean produced Yellow Fin and Big Eye Tuna. Our goal is to study the substitutability among Japanese, Taiwanese as well as South Korean produced Yellow Fin and Big Eye tuna. The rest of this study is arranged in the following way. The theoretical demand system model and its empirical specifications are described in the second section. The third section presents the data description and the estimation results. Some of the important findings are also summarized in the section. Finally, a brief conclusion is given in the last section. 2. Demand system specification The main purpose of this study is to understand the substitutability among Japanese, Taiwanese and South Korean produced frozen tuna in the market of Japan. In order to achieve this goal, a demand system that describes the Japanese demand behavior among Japanese, Taiwanese and South Korean produced frozen tuna has to be constructed and estimated. In the following section, an ordinary demand system model will be briefly discussed. 2.1 Ordinary Demand System Following Huang and Hahn (95) and Huang 93, 91), let q been mx1 column vector which represents the quantities demanded for a consumer, p an 1 X m rowvector of the corresponding prices, y = p.q the consumer expenditure which is the inner product of p and q. and u(q) the utility function which is assumed to be quasiconcave in q. The utility maximization under constraint can be set up by the a Lagrangean function with the multiplier A : Max L=u(q)+ (y-p.q) q. According to the first order condition (FOC) of the above maximization problem, it is shown that

ui(q)= pi and y=p.q, where u,(c/) is the marginal utility of the ith commodity, I=1, 2,, m, and is known as the marginal utility of income. The second order condition (SOC) for the maximization problem implies that the Hessian matrix H = ui(q)= 2 u / 2 qi u 2 qj is, symmetric and negative definite. The ordinary demand system is obtained by solving the FOC: With the unknown utility structure, an ordinary demand system model can be approached by applying The above equation can be rewritten by the expression of price and income elasticities : where is the price elasticity of by the commodity with respect to a price change of the j commodity, and is the expenditure elasticity showing the effect of the quantity in reponse to a change in per capita expenditure. To ensure theoretical consistency in applying the differential-form demand model, given wi=piqi / y the expenditure weight of the commodity and e*ij a compensated elasticity, the following parametric constraints should be applied (Hicks. 36):

Negativity: eii +wi i <0, i =1,2,m. Linkage condition: e*ij = eij + wj, i, j=1,2 m. The Engel aggregation states that the sum of the expenditure elasticities weighted by the expenditure shares of corresponding commodities equals 1. The homogeneity condition implies that a consumer has no money illusion, and thus a proportional change in both price and expenditure leaves quantity demanded unchanged. The symmetry condition is derived from the symmetry of the Slutsky income compensated substitution terms. The negativity condition which implies that an increase in price with utility held constant must cause demand for that good to fall. Finally, the linkage condition is an expression of the Slutsky equation in terms of elasticity. 2.2 Empirical Models Specification Based on the derivation from the previous section, a theoretically consistent empirical differential-form demand system could be specified as Where q i,p i and y' are the relative changes in quantity, price, and per capita expenditure which can be defined as the first-order difference of the natural logarithm form[2]. The parameters ci,, eij and i, are constant term, price and expenditure elasticities respectively. In order to ensure the internal consistency with the demand structure provided by the demand theory, the parametric constraints of Engel aggregation

and symmetry which described in the previous section should also be imposed into the estimation procedure[3] 3. Estimation Results According to Huang (94), the demand for tuna as a whole are not related to the demand for beef, pork, chicken as well as to other kinds of fresh fishes. Huang's results suggested that the demand for different kinds of tuna can be modeled separately without considering the demand for other kinds of food, namely, beef, pork, chicken and other kinds of fresh fishes in the Japanese market. 3.1 Description of the Data Set To avoid modeling the short term shocks from the Japanese Tuna sashimi market, e.g. seasonably, storage, and stocking etc.. annual data ranged from 80 to 93 are adopted in the empirical model. Since the tuna sashimi consumption data are not fully available, Japanese tuna harvest and tuna import data are used to replace the consumption data. The average tuna prices are generated as the total value of the annual harvest (import) divided by the total quantities of the annual harvest (import). Variables used in estimating the demand for Japanese, Taiwanese, South Korean, and other countries produced Yellow Fin and Big Eye tuna are: DFJYFQ: quantity of Japanese produced frozen Yellow Fin tuna in log-difference form (unit:: Kg.)[4]. DFTYFQ: quantity of Taiwanese produced frozen Yellow Fin tuna in log-difference form (unit: Kg.)[5] DFKYFQ: quantity of South Korean produced frozen Yellow Fin tuna in logdifference form (unit: Kg.). DFOYFQ: quantity of other countries produced frozen Yellow Fin tuna in logdifference form (unit: Kg.). DFJBEQ: quantity of Japanese produced frozen Big Eye tuna in log-difference form (unit: Kg.). DFTBEQ: quantity of Taiwanese produced frozen Big Eye tuna in log-difference form (unit: Kg.). DFKBEQ: quantity of South Korean produced frozen Big Eye tuna in logdifference form (unit: Kg.). DFOBEQ: quantity of other countries produced frozen Big Eye tuna in logdifference form (unit: Kg.).

DFJYFP: price of Japanese produced frozen Yellow Fin tuna in log-difference form (unit: Japanese Yen / Kg.). DFTYFP: price of Taiwanese produced frozen Yellow Fin tuna in log-difference form (unit: Japanese Yen / Kg.). DFKYFP: price of South Korean produced frozen Yellow Fin tuna in log-difference form (unit: Japanese Yen / Kg.). DFOYFP: price of other countries produced frozen Yellow Fin tuna in logdifference form (unit: Japanese Yen / Kg.). DFJBEP: price of Japanese produced frozen Big Eye tuna in log-difference form (unit: Japanese Yen / Kg.). DFTBEP: price of Taiwanese produced frozen Big Eye tuna in log-difference form (unit: Japanese Yen / Kg.). DFKBEP: price of South Korean produced frozen Big Eye tuna in log-difference form (unit: Japanese Yen / Kg.), DFOBEP: price of other countries produced frozen Big Eye tuna in log-difference form (unit: Japanese Yen / Kg.). DTV: total value of the Japanese produced and imported Yellow Fin and Big Eye tuna in log-difference form. [6] 3.2 Estimation Results The empirical model is specified by eight equations: d Qi = Ci + i,1 DFJYFP + i,2 DFTYFP + i,3 DFKYFP + i,4 DFOYFP + i,5 DFJBEP + i,6 DFTBEP + i,7 DFKBEP + i8j DFOBEP + i,9 DOTV, where d QI, represents DFJYFQ, DFTYFQ, DFKYFQ, DFOYFQ, DFJBEQ, DFTBEQ, DFKBEQ, and DFOBEQ for different value of i s. Due to the serious multi-colinearity among regressors, a theoretically more efficient estimator, i.e. a SUR (seemingly unrelated regression) model with restrictions could not be applied. Instead, an OLS estimator without any restriction is then adopted for each equation. Since the regressors are exactly the same for each of the eight equations, OLS and SIJR estimators will produced the same estimates for the parameters, though the OLS estimates are less efficient. Results of the estimation are presented in Table 6 for Yellow Fin tuna and in Table 7 for Big Eye tuna respectively. Some of the important OLS results can be summarized as following: 1. As the price of the Taiwanese produced Yellow Fin tuna increases by 1%, the demand for Japanese produced Yellow Fin will increase by 0.57%, which shows that Japanese and Taiwanese produced Yellow Fin tuna are substitutes, although the substitutability are not very strong.

2. As the price of the South Korean produced Yellow Fin tuna increases by 1%, the demand for Taiwanese produced Yellow Fin will decrease by 4.10%, which shows that they are sort of complements and the complementability are quite large. 3. As the price of the Japanese produced Yellow Fin tuna increases by 1%, the demand for South Korean produced Big Eye tuna will increase by 0.45%, which shows that they are somewhat substitutable. 4. As the price of the Taiwanese produced Yellow Fin tuna increase by 1%, the demand for South Korean produced Big Eye tuna will decrease by 1.04%, which denoted a complementability between these two kinds of tuna. 5. When total Japanese expenditure increase, the consumption of Japanese produced Yellow Fin tuna and South Korean produced Big Eye tuna will increase. However, the consumption of the Japanese produced Yellow Fin tuna is more elastic. 4. Conclusions Japan is the only significant market of tuna sashimi in the world. Although Japan has the largest tuna harvest in the world, it still rely on import to satisfy its domestic demand, especially for frozen Yellow Fin and Big Eye tuna. Among all the fifty, some Yellow Fin and Big Eye tuna producer countries, Taiwan and South Korea, are the two most important ones in terms of both value and quantities produced. Tuna fishery industry is a very important part of Taiwanese fishery industry. The value produced by the Taiwanese tuna industry took about 30% of the total value of the Taiwanese fishery and about 75% of the total value of the Taiwanese deepsea fishery industry in 93. Since frozen Yellow Fin and Big Eye tuna are the major harvest of the Taiwanese tuna fishery industry, and most of the harvest are exported to the Japanese sashimi market, it is important for the Taiwanese tuna fishery industry to understand the Japanese tuna sashimi market. Based on the concept of a demand system, this study studied the substitutability among the Japanese, Taiwanese and South Korean produced frozen Yellow Fin and Big Eye tuna sashimi. Empirical results suggested that the consumption of Japanese produced frozen Big Eye tuna sashimi is really not related at all to other kinds of tuna sashimi which produced by other countries. There is no substitutability even between the Japanese produced Big Eye and Yellow Fin tuna sashimi. As for the Yellow Fin tuna sashimi, there is an interesting result. Unlike the case of the Big Eye tuna sashimi, it was shown that when the price of the Taiwanese produced Yellow Fin tuna sashimi increases, some of its demand will shirt to the Japanese produced Yellow Fin tuna sashimi (with the cross elasticity equal 0.57). However, when the price of the Japanese produced Yellow Fin tuna sashimi changes, the demand won't shift to the Taiwanese produced Yellow Fin tuna sashimi at all. Due to the serious multi-colinearity among regressors, a theoretically more efficient estimator. SUR model, with restriction estimated by FGES, could not be applied. OES was then used to estimate the empirical models. Although the 01.S

would produce the same parameters estimates as the SUR did. It was less efficient and all the theoretically consistent restrictions could not be applied. A SUR model with restrictions which estimated by MEE shall be adopted in the future in order to acquire more efficient parameters estimates. References Association of Agriculture and Forestry Statistics. Japan (March 95),Annual Statistics on Marketing of Fishery Product. Taiwan Fisheries Bureau (June 95). Fishery Yearbook. Taiwan Area. 94. Department of Agriculture and Forestry, Provincial Government of Taiwan. R. 0. C. Hicks, J. R. (36), Value and Capital. Oxford. U. K.: Oxford University Press. Huang, H. W. (94). "The Demand of Tuna in Japan and the Demand of Taiwanese Frozen Tuna Exported to Japan." M. S. thesis. Institute of Oceanography. National Taiwan University, Taipei, Taiwan. R, 0. C. Huang. K. S. And W. F. Hahn (95), "U. S. Quarterly Demand for Meats, "U. S. Department of Agriculture- Economic Research Service. Technical Bulletin No, 1841. Huang, K. S. (t993), "A Complete System of U.S. Demand for Food." U. S. Department of Agriculture. Economic Research Service. Technical Bulletin No. 1821. Huang. K. S. (91), "U. S. Demand for Food: A Complete System of Quantity Effects on Prices" U. S. Department of Agriculture. Economic Research Service. Technical Bulletin No. 1795. Japan Tariff Association. Japan Exports & Imports Country by Commodity, various issues. Japan Fisheries Agency (95), Statistical Tables of Fishing Vessels.

TABLE 1: Japanese Tuna Harvest (75-93) Unit : Tons Ye ar 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 Blu e Fin 407 16 418 05 5 00 465 55 442 41 494 94 584 85 442 06 366 63 357 18 299 48 234 55 253 35 181 64 9 38 137 98 157 91 165 53 168 58 BIG Eye 1134 45 1147 75 1283 33 1276 66 1304 66 1231 68 1105 13 1317 72 1391 64 1314 25 1489 40 1578 06 1409 85 1358 56 1172 63 1221 48 1245 01 1435 60 1395 07 Yello w Fin 73 9 8574 4 8284 5 9806 6 9965 9 10 0! 1100 09 1I42I 9 1124 93 1153 89 1343 95 1182 57 1152 25 1022 65 1000 02 9825 7 1080 88 1227 53 1269 15 Data Source: Annual Statistics on Marketing of Fishery Products. 93, Association of Agriculture and Forestry Statistics. Japan. March 95. Albac ore Total 68861 2949 61 10707 3493 1 95 54027 3171 05 87675 3599 62 66822 3411 88 69677 3613 40 64082 3430 89 70043 3602 40 51512 3398 32 63624 3461 56 57820 3711 03 51136 3506 54 46945 3284 90 45151 3014 36 45273 2824 76 43315 2775 18 38213 2865 93 48709 3315 75 60331 3436 11

Table 2: Japanese Imported Tuna (81-93) Ye ar 8 1 8 2 8 3 8 4 8 5 8 6 8 7 8 8 8 9 9 0 9 1 9 2 9 3 Total Imported 84703.30 3 95608.11 3 114529.5 52 99646.17 4 132393.4 64 132804.7 30 177878.7 46 203347.1 85 200358,7 52 230030.3 61 231377.4 84 2370.1 06 373399.1 33 Imported from Taiwan Imported from South Korea 48113.3 73 46551.4 63 11842.84 2 14292.68 9 284,44 57682.3 21 15963.34 43234.9 4 6 453.62 46126.7 3 09 24710.77 55954.3 2 84 36804.98 66026.7 6 58 40290.86 62485.8 5 04 39971.69 52790.0 8 67 61240.25 55331.0 0 87 69718.61 52378.0 1 65 879)2.90 49333.5 5 14 12H83.1 39613.0 09 61 Data Source: Japan Exports & Imports Country by Commodity, Japan Tariff Association. Imported from other Countries 24748.08 8 34763.96 1 37562.79 1 40447.87 0 66813.13 2 52139.57 4 75047.00 2 100570.5 16 107596.9 87 113459.0 24 109280.8 08 99943,68 7 112602.9 63

Table 3: Taiwanese Deep-sea Fishery Vs. Tuna Fishery (80-94) Deep-sea Fishery Tuna Fishery Year Quantity (1000 tons) value (N.T.$ 100.000.000) Quantity (1000 tons) (2) Value (N.T.$ 100.000.000) (2)/(1) 80 350.4 134.1 1.6 66.5 49.6% 81 321.6 152.4 94.9 65.5 43.0% 82 316.7 163.4 148.4 86.6 53.0% 983 3.9 171.4 149.5 84.1 48.2% 84 369.4 7.1 164.6 97.7 49,6% 85 413.7 215.3 7.1 107.1 49.7% 86 463.4 234.7 215.5 1.2 50.8% 87 573.9 285.5 247.1 141.3 49.5% 88 699.3 3.9 292 150.3 47.0% 89 734.5 335.1 321.6 173.3 51.7% 90 767 352.5 405.6 2.6 54.6% 91 714.3 322.1 332.6 154.3 47.9% 92 737.6 346.3 428.6 217.7 62.9% 93 835 427.0 476.9 320.2 75.0%. 94 683.8 360.5 391.9 246.4 68.3% Data Source: Fishery yearbook. Taiwan Area. 94. Taiwan Fisheries Bureau- Department of Agriculture and Forestry. Provincial Government of Taiwan. R. 0. C- June 95-

Table 4: Taiwanese Fishery Industry Vs. Tuna Fishery (83-94) Year (1) Total Fishery Industry (N.T.$1.000) (2) Tuna Fishery (N.T.$ 1,000) (2)/(I) (%) 83 62.Ut2.254 7.712.710 12 84 64,376,358 8.616.306 13 85 66.892.998 9.142-535 14 86 75.380.054 10.051.779 13 87 85.954.762 11.304.899 13 88 88.135,707 12,807.739 15 89 89.110.347 15.287.562 17 00 S9.154.i64 14.875.414 17 91 83.518,127 10-831.710 13 92 83.715.433 17,307.526 21 93 93.175.224 28-560.137 31 94 89,201.376 21.206.994 24 Data Source: Fishery Yearbook. Taiwan Area, 94. Taiwan Fisheries Bureau. Department of Agriculture and Forestry, Provincial Government of Taiwan. R. 0. C., June 95. Table 5: Number of Ultra Low Temperature Long-liners Fishing Vessels (83-94) SOUTH Year JAPAN TAIWAN KOREA 80 943 2 72 81 964 208 72 82 854 185 69 83 84 770 761 169 157 71 63 85 773 156 75 86 771 167 81 87 770 189 103 88 759 9 106 89 764 5 169 90 758 203 237 91 737 201 318 92 723 185 313 93 722 174 313 94 716 174 334 Data Source: Statistical Tables of Fishing Vessels, Japan Fisheries Agency- 05

Table 6: OLS Results: Frozen Yellow Fin Tuna Consumption Dependent Variable DFJYFQ DFTYFQ DFKYFQ DFOYFQ CONSTANT -0.05 0. -0.24* 0.18 (0.11) (0.06) (0.03) (0.65) DFJYFP -0.98* 0.24 0.12-1.03 (0.16) (0.51) (0.28) (3.03) DFTYFP 0,57* 1.96* 0.15-0.39 (0.181) (0.58) (0.32) (3.45) DFKYFP -0.04-4.10* -2.91* 2.57 (0.32) (1.04) (0.57) (6.13) DFOYFP 0.07-0.03-0.31* 0.25 (0.04) (0.14) (0.08) (0.82) DFTBEP -1.22 3.14 1.34 1.45 (1.43) (1.39) (0.76) (8.2;1) DFKBEP 0.24 0.86 1.85-4.91 (0.46) (1.48) (0.81) (8.73) DFOBEP 0.56-0.65-1.71 1.89 (0.44) (1.43) (0.78) (8.46) D1V 1.60* -1.97 1.42 3.49 (0.40) (1.31) (0.72) (7.76) R 2 Adj. R 2 F-Value 0.99 0.94 20.72 0.97 0.80 5.97 0.98 0.90 12.49 0.49 1.83 0.209 1.Annual data. 80-93. 2.Number in parentheses are standard errors. 3.* indicates significant at 90% level.

Table 7: OLS Results: Frozen Big Eye Tuna Consumption Dependent Variable DFJBEQ DFTBEQ DFKBEQ DFOBEQ CONSTANT -0.02 (0.05) 0.04 (0.07) -0.16* (0.02) 0.05 (0.14) DFJYFP 0.18 (0.23) -0.46 (0.34) 0.45* (0.09) 0.05 (0.65) DFKYFP -0. (0.26) -1.24* (0.39) -1.04* (0.11) -1.20 (0.73) DFKYFP 0.30 (0.46) 0.46 (0.70) -0.61* (0.) 1.05 (1.31) DFOYFPP -0.02 (0.06) -0.12 (0.09) -0.23* (0.03) -0.12 (0.17) DFJBEP -0.47 (0.42) 0.23 (0.64) 0.07 (0.17) -0.00 (1.20) DFTBEP -0.71 (0.62) -0.82 (0.93) 0.71 (0.25) -0.59 (1.75) DFKBEP 0.09 (0.65) -0.06 (0.99) 0.79* (0.27) 0.94 (1.86) DFOBEP -0.30 (0.63) 1.04 (0.96) -0.84* (0.26) -1.01 (1.80) DTV 1.45 (0.58) 1.35 (0.88) 0.84* (0.24) 2.17 (1.65) R 2 Adj. R 2 F-Value 0.96 0.77 5.10 0.96 0.77 5.07 0.99 0.96 33.16 0.74-0.41 0.64 1.Annual data. 80-93. 2.Number m parentheses are standard errors, 3.* indicates significant al 90% level.

[1] Including Blue Fin Tuna (Thunnus thunnus). Big Eye tuna (Thunnus obesus), Yellow Fin tuna (Thunnus albacares), and Albacore (Thunnus alaunga). [2] for example [3] Since there is neither efficiency gain nor refunction in the number of parameters the number estimated, the negativity condition is not incorporated. Because the compensated demand is not an issue in this study, the linkage condition is also ignored. [4] All the quantities and prices which related to the Japanese produced tuna are from Annual Statistics in Marketing of Fishery Products, 93. Association of Agriculture and Forestry Statistics. Japan. March 95. [5] All the quantities and prices which related to imported tuna are compiled from Japan Exports & Imports Country by Country. Japan Tariff Association. [6] A model which replaces DTV by DTTEXP, the Japanese total final expenditure of the private sector, are also estimated. The result are similar to what will be provided in the next section.