4. DISEÑO E IMPLEMENTACION DE LA APLICACIÓN
4.3. Análisis del problema
The unrestricted error correction modelling (UECM) procedure is used in this section to estimate short run and long run relationships between the dependent and independent variable in the models of Thailand’s demand for Australia’s exports of three SITC categories of dairy products, i.e. 022 milk dry (milk powder), 023 butter and 024 cheese and curd. The models are estimated with difference and level terms of the variables and in log-log (double log) functional form. Log-log form is used for the ease of estimation and derivation of long run elasticities. In Tables 6.2, 6.3, and 6.4, the estimated demand functions are presented together with the relevant diagnostic statistics. As the diagnostic statistics show, the estimated models are statistically well performed. Hence, the presence of problems in relation to serial correlation, functional form misspecification, non-normality and heteroscedasticity can be rejected.
6.7.1.1 Estimated Export Demand Function for Milk Dry
The estimated export demand function for milk dry (SITC 022) is presented in Table 6.2. The results of the diagnostic tests are satisfactory although the value of the adjusted R2 is not very high. The short run coefficient for the difference term of the relative price of exports of milk dry (ΔLRP22) has the expected negative sign, which is significant. This indicates that in the short run, the quantity of Australia’s milk dry exports demanded in Thailand falls when Australia’s export price relative to that of competing countries increases. As shown by the positive but insignificant coefficient for ΔLGDPT, in the short run, the quantity of Australia’s milk dry exports demanded in Thailand is not responsive to Thailand’s real national income.
The long run coefficient for relative price variable (LRP22t-1)is negative as expected and significant. This indicates that in the long run, the quantity of Australia’s milk dry exports demanded in Thailand declines when Australia’s export price relative to that of competing countries’ price increases. The long run relative price elasticity of demand of –2.76 indicates that a 1 per cent increase in Australia’s export price relative to competitors’ export prices, ceteris paribus, will result in a 2.76 decrease in the Australia’s milk dry exports demanded in Thailand. The less than infinite price elasticity may indicate that Australia has some market power in Thailand in relation to
the exports of its milk dry products. This may be in contrast to “small country assumption, an aspect that warrants further examination in future studies.
The long run coefficient for Thailand’s real national income variable (LGDPTt-1)is positive as expected but the coefficient is not significant. This indicates that in the long run, the quantity of Australia’s milk dry exports demanded in Thailand does not change significantly in response to changes in Thailand’s real national income. The coefficient for the dummy variable for import tariffs of Thailand (DTT) has the expected negative sign but it is not significant. Thus, in high tariff years in Thailand, the demand for Australia’s milk dry exports do not seem to be significantly different from that in low tariff years in Thailand. However, the coefficient for the dummy variable for the crisis (DAC) has the positive sign contrary to the expectation and the coefficient is not significant. Hence, it appears that Thailand’s demand for Australia’s milk dry exports during the financial crisis was not significantly different from that in non-crisis years.
Table 6.2: Estimated UECM of Thailand’s Demand Function for Australia’s Milk Dry Exports (SITC: 022)
Dependent Variable = ΔLRX22
Regressor Coefficient t-ratio Long Run Elasticity t-ratio Constant 18.59 2.38* - - ΔLRP22 -1.54 -1.81* - - ΔLGDPT 4.48 0.96 - - LRX22 (t-1) -0.94 -4.58*** - - LRP22(t-1) -2.59 -2.42** -2.75 -2.67*** LGDPT(t-1) 1.09 1.61 - - DTT -0.72 -1.13 - - DAC 0.80 1.42 - - R2 = 0.59 Adj. R2 = 0.41 DW = 1.64 LMS F (1,16) = 1.88 RESET F (1,16) = 2.60 JBN, χ2 (2) = 0.05 HSC F (1,23) = 0.02
*** significant at the 1% level, ** significant at the 5% level, * significant at the 10% level. LMS = Lagrange multiplier test for serial correlation,
RESET = Ramsey’s test functional form using the square of the fitted values, JBN = Jarques-Bera test for the normality of residuals (based on χ2 distribution), HSC = Heteroscedasticity test based on the regression of squared residuals on squared fitted values.
6.7.1.2 Estimated Export Demand Function for Butter
The estimated export demand function for butter (SITC 023) is presented in Table 6.3. The results of the diagnostic tests are satisfactory with a reasonable value of the adjusted R2. The short run coefficient for the difference term of the relative price of exports of butter (ΔLRP23) has the expected negative sign, which is significant. This indicates that in the short run, the quantity of Australia’s butter exports demanded in Thailand falls when Australia’s export price relative to that of competing countries increases. As shown by the negative but insignificant coefficient for ΔLGDPT, in the short run, the quantity of Australia’s butter exports demanded in Thailand is not responsive to Thailand’s real national income. The coefficient for Thai baht/AU$ exchange rate (ΔLEXR) is negative as expected and significant. This indicates that in the short run, the quantity of Australia’s butter exports demanded in Thailand declines when the Thai baht depreciates against the Australian dollar (or, when AU$ appreciates against the bath).
The long run coefficient for relative price variable (LRP23t-1)is negative as expected and significant. This indicates that in the long run, the quantity of Australia’s butter exports demanded in Thailand decreases when Australia’s export price relative to that of competing countries’ price increases. The long run relative price elasticity of demand of –1.13 indicates that a 1 per cent increase in Australia’s export price relative to competitors’ export prices, ceteris paribus, will result in a 1.13 decrease in the Australia’s butter exports demanded in Thailand. The less than infinite price elasticity may indicate that Australia has some market power in Thailand in relation to the exports of its butter. This may also be in contrast to “small country assumption, an aspect that can be investigated further in future studies.
The long run coefficient for Thailand’s real national income variable (LGDPTt-1)is positive as expected but the coefficient is not significant. This indicates that in the long run, the quantity of Australia’s butter exports demanded in Thailand does not change significantly in response to changes in Thailand’s real national income. The long run coefficient for relative exchange rate variable (LEXRt-1) is negative as expected and significant. This indicates that in the long run, the quantity of Australia’s butter exports demanded in Thailand declines when the Thai baht depreciates against
the Australian dollar (or, when AU$ appreciates against the bath). The long run exchange rate elasticity of demand of –6.34 indicates that a 1 per cent depreciation of Thai bath against the AU$, ceteris paribus, will result in a 6.34 decrease in the Australia’s butter exports demanded in Thailand.
The coefficient for the dummy variable for import tariffs of Thailand (DTT) has the unexpected positive sign but it is not significant. Thus, in high tariff years in Thailand, the demand for Australia’s butter exports do not seem to be significantly different from that in low tariff years Thailand. The coefficient for the dummy variable for the crisis (DAC) also has the positive sign contrary to the expectation and the coefficient is not significant. Hence, it appears that Thailand’s demand for Australia’s butter exports during the financial crisis was not significantly different from that in non- crisis years.
Table 6.3: Estimated UECM of Thailand’s Demand Function for Australia’s Butter Exports (SITC: 023)
Dependent Variable = DLRX23
Regressor Coefficient t-ratio Long Run Elasticity t-ratio Constant 26.36 2.02* - - ΔLRP23 -0.68 -2.24** - - LGDPT -0.11 -0.03 - - ΔLEXR -5.73 -2.15** LRX23 (t-1) -0.92 -3.77*** - - LRP23(t-1) -1.04 -2.48** -1.13 -2.33** LGDPT(t-1) 1.18 1.36 1.28 1.49 LEXR(t-1) -5.85 -2.13** -6.34 -3.25*** DTT 0.80 1.26 - - DAC 1.15 1.60 - - R2 = 0.70 Adj. R2 = 0.53 DW = 2.17 LMF F (1,14) = 0.45 RESET F (1,14) = 8.29 JBN, χ2 (2) = 0.78 HSC F (1,23) = 0.24
*** significant at the 1% level, ** significant at the 5% level, * significant at the 10% level. LMS = Lagrange multiplier test for serial correlation,
RESET = Ramsey’s test for functional form using the square of the fitted values, JBN = Jarques-Bera test for the normality of residuals (based on χ2 distribution),
HSC = Heteroscedasticity test based on the regression of squared residuals on squared fitted values.
The degrees of freedom are given in parentheses.
6.7.3 Estimated Export Demand Function for Cheese and Curd
The estimated export demand function for cheese and curd (SITC 024) is presented in Table 6.4. The results of the diagnostic tests are satisfactory although the value of the adjusted R2 is fairly low. This estimated function is relatively poor in that only two of the significant coefficients being derived. The low explanatory power is a result of the exclusion of the exchange rate variable and the dummy variables for Thailand’s tariffs and the Asian currency crisis. An estimation with these variables resulted in none of the coefficients being significant.
The short run coefficients for the difference terms of the relative price of exports of cheese and curd (ΔLRP24) and Thailand’s real national income (ΔLGDPT) have the expected signs, but they are not significant. This indicates that in the short run, the quantity of Australia’s cheese and curd exports demanded in Thailand is not responsive either to the relative price of exports or to Thailand’s real national income. The long run coefficient for Thailand’s real national income variable (LGDPTt-1)is positive as expected and the coefficient is significant. This indicates that in the long run, the quantity of Australia’s cheese and curd exports demanded in Thailand changes significantly in response to changes in Thailand’s real national income. The long run income elasticity of demand of 1.84 indicates that a 1 per cent increase in Thailand’s real GDP, ceteris paribus, will result in a 1.84 increase in Australia’s cheese and curd exports demanded in Thailand. This further suggests that Australian cheese and curd is regarded as something of a “luxury” in Thailand.
Table 6.4: Estimated UECM of Thailand’s Demand Function for Australia’s Cheese and Curd Exports (SITC: 024)
Dependent Variable = DLRX24
Regressor Coefficient t-ratio Long Run Elasticity t-ratio Constant -1.35 -0.28 - - ΔLRP24(t-1) -1.02 -1.35 - - ΔLGDPT 0.59 0.30 - - LRX24 (t-1) -0.56 -2.73** - - LRP22(t-2) 0.12 0.14 - - LGDPT(t-1) 1.04 2.34** 1.84 5.17 *** R2 = 0.41 Adj. R2 = 0.23 DW = 1.87 LMF F (1,15) = 0.05 RESET F (1,15) = 1.08 JBN, χ2 (2) = 1.10 HSC F (1,20) = 1.29
*** significant at the 1% level, ** significant at the 5% level, * significant at the 10% level. LMS = Lagrange multiplier test for serial correlation,
RESET = Ramsey’s test using the square of the fitted values,
JBN = Jarques-Bera test for the normality of residuals (based on χ2 distribution),
HSC = Heteroscedasticity test based on the regression of squared residuals on squared fitted values.
6.8 Conclusion
Econometric models of Thailand’s demand for Australia’s dairy product exports, specifically for milk dry, butter, and cheese and curd exports, were developed and estimated in this chapter. The unrestricted error correction modelling (UECM) procedure was employed to estimate the models of export demand.
The findings indicate that in the short run, the quantity of Australia’s milk dry exports demanded in Thailand falls when Australia’s export price relative to that of competing countries increases, while it is not responsive to Thailand’s real national income. In the long run, the quantity of Australia’s milk dry exports demanded in Thailand declines when Australia’s export price relative to that of competing countries’ price increases. The estimated long run relative price elasticity of export demand for milk dry is -2.76. The less than infinite price elasticity may indicate that Australia has some market power in Thailand in relation to the exports of its milk dry products, an aspect worthy of further examination in future studies. In the long run, the quantity of Australia’s milk dry exports demanded in Thailand does not change significantly in response to changes in Thailand’s real national income.
In the short run, the quantity of Australia’s butter exports demanded in Thailand falls when Australia’s export price relative to that of competing countries increases, but it is not responsive to Thailand’s real national income. The quantity of Australia’s butter exports demanded in Thailand declines when the Thai baht depreciates against the Australian dollar (or, when AU$ appreciates against the baht). In the long run, the quantity of Australia’s butter exports demanded in Thailand decreases when Australia’s export price relative to that of competing countries’ price increases. The estimated long run relative price elasticity of demand is –1.13. The less than infinite price elasticity may indicate that Australia has some market power in Thailand in relation to the exports of its butter. This aspect that can also be investigated further in future studies. In the long run, the quantity of Australia’s butter exports demanded in Thailand does not change significantly in response to changes in Thailand’s real national income. The quantity of Australia’s butter exports demanded in Thailand declines when the Thai baht depreciates against the Australian dollar (or, when AU$
appreciates against the bath). The estimated long run exchange rate elasticity of demand is –6.34.
In the short run, the quantity of Australia’s cheese and curd exports demanded in Thailand is not responsive either to the relative price of exports or to Thailand’s real national income. In the long run the quantity of Australia’s cheese and curd exports demanded in Thailand changes significantly in response to changes in Thailand’s real national income. The estimated long run income elasticity of demand is 1.84.
The results also indicate that in high tariff years in Thailand as well as in the Asian financial crisis years, Thailand’s demand for Australia’s exports of any of the dairy products considered is not significantly different from that in other years.
The findings of this chapter can be used to explain the relationship between various macro level determinants (relative price, real income, exchange rate) and Thailand’s demand for Australia’s dairy product exports. In chapter 7, the findings from interviews with export managers in several Australian companies that export dairy products to Thailand will be presented in order to explore micro (firm or company) level information on the experiences and problems faced by the firms (companies) in the Thai market. This information will be used only as a supplement to the analysis in Chapter 6.