Residual Analysis Discussions External Thesis Examiner

39 Similar to the findings in the plywood model, the VECM confirms the results of Granger-causality test as well, which emphasizes both the domestic and the world price have an influence to each other. Table 9 Error Correction Model for Sawn wood Variable Parameter Variable Parameter Δdst Δrwst ectt-1 -0.072 ectt-1 0.037 Δdst-1 0.277 Δdst-1 0.294 Δrwst-1 -0.070 Δrwst-1 0.119 Long run dst-1 Intercept 3.509 rwst-1 0.350 Trendt-1 0.004

5.7 Residual Analysis

Residual analysis is necessary with regard to check the presence of autocorrelation and heteroscedasticity as well as the assumption of normality in the model. As shown in the table below, based on the portmanteau test, the null hypothesis is not rejected for logwood and sawn wood model which implies there is no problem with autocorrelation. However, based on the LM test, the result indicates the contradiction in the logwood model, which finds an indication of autocorrelation. Furthermore, for plywood model, both portmanteau test and LM test suggest the problem of autocorrelation. Even though the sawn wood model does not contain the problem of autocorrelation, but based on ARCH-LM test, it is indicated the problem of heteroscedasticity in the sawn wood model. In addition, all of the models do not contain the problem of normality as the null hypothesis of non-normality is rejected at any level of significance. 40 Table 10 The Results of Residual Analysis Diagnostic Test p-value Logwood Plywood Sawnwood Portmanteau test for autocorrelation 0.1276 0.0173 0.8135 LM test for autocorrelation 0.0000 0.0010 0.5277 Test for non-normality 0.0000 0.0000 0.0000 ARCH-LM test U 1 : 0.9667 U 2 : 0.8496 U 1 : 1.0000 U 2 : 0.9999 U 1 : 0.0411 U 2 : 0.9766

5.8 Discussions

According to both two step Engel-Granger test and Johansen technique, it can show the presence of co-integration relationship between world price and domestic price for all wood products i.e. logwood, plywood, and sawn wood. Co- integration relationship between wood prices in the world market and in the domestic market implies that there is a long run relationship between them. Therefore, it can be concluded that domestic market of primary wood products of Indonesia is integrated with the world market. These are not surprising findings since Indonesia is one of the biggest producer as well as exporter of tropical wood products in the world, particularly for plywood and sawn wood. Even though Indonesia has applied an export ban for logwood since 1987, but based on the result of co-integration test, the restriction on logwood export might not prevent the integration of logwood market of Indonesia with the world market in the long run. This proved that physical flow of the good is not always a necessary condition to the presence of market integration. If the two markets are a part of one marketing system, then the integration of markets can occur without the occurrence of physical trade flow as happened in the Indonesian logwood market. Although domestic logwood market is integrated with the world market implying a long-run relationship, but based on the results of error correction model, it is not indicated any response in the short-run between domestic and world prices of logwood. As mentioned before, this might be due to the export ban of logwood, thus the changes on world prices will not be transmitted to 41 domestic prices and so does the other way around. Before the implementation of logwood export ban in 1987, Indonesia had the highest market share of tropical logwood in the world market which was responsible for around 40 percent of tropical logwood supply in the world market. By imposing this export ban policy, Indonesian government tries to encourage the development of domestic wood processing sector. Therefore, logwood as the main raw material in the wood industry is heavily consumed for domestic demand. Besides being the biggest tropical logwood producer, Indonesia is the biggest tropical logwood consumer as well ITTO, 2010. However, suspicion of illegal logging might occur against this policy. This will be discussed later in the next chapter. Based on sample split chow-test, it is indicated the structural break in the last quarter of 2009. This structural change might be related to the amendment of logwood export ban policy in 2009. Due to the financial crisis particularly in Europe and USA, which started in 2008, the Indonesian government decided to amand the logwood export ban in order to compensate the income loses from export of wood products. According to this amendment, the export of logwood is not fully banned, but is allowed for several types of logwood. As mentioned before that after the last quarter of 2009, the speed of adjustment of both domestic and world logwood price becomes higher. This finding is plausible since the domestic market of logwood is not as restrictive as before the amendment. According to both Granger-causality test and the result of ECM, it is found that both the domestic price and world price of plywood and sawn wood have an influence to each other. These results might confirm the position of Indonesia as one of the biggest exporter in the world market. However, based on the ECM of sawn wood, that the speed of adjustment of the domestic price is higher than that of the world market, implies the role of the other exporting countries of sawn wood, such as Malaysia and Brazil as the biggest exporting countries. Meanwhile, in the logwood market, that the world price will adjust the disequilibrium faster than will the domestic price also indicates the importance of Indonesian position in the world market, though Indonesia is no longer one of the biggest exporter of logwood due to the export ban policy. Though Indonesia faces declining forest area, but generally Indonesia has a comparative advantage of forest resource 42 abundance compared to the other tropical countries. This situation also occurs in the plywood industry, in which the world price will adjust faster than the domestic price to achieve the equilibrium condition. Since this study only employs standard error correction model to analyze the price transmission between the world and the domestic wood products, there are several drawbacks due to the presence of structural break as resulted in the stability analysis. Based on the results of chow test, generally, the parameters in the model are seen to be not stable. In addition to that, the presence of autocorrelation and heteroscedasticity in the model might also danger the validity of the error correction term. If the residuals are correlated each other autocorrelation, the standard deviation might be underestimate and thus could produce the overestimate the value of statistics. On contrary, if the variance of the residuals is not constant, the standard deviation might be overestimate and thus the value of statistics could be underestimated. The presence of heteroscedasticity might be the reason why the t-statistic in the long run equation of sawn wood model is too small. According to these drawbacks, thus it is recommended for the further research to improve the model of price transmission which considers more precisely the presence of structural breaks using the regime dependent model and so forth. Since the logwood as the main raw material for the processed wood products i.e. sawn wood and plywood, it might be useful to introduce panel data VECM to address this issue. 43 6 THE IMPLICATION OF WOOD PRICE CHANGES ON DEFORESTATION IN INDONESIA This chapter tries to answer the question “how is the implication of wood price changes on deforestation in Indonesia” as the main objective of this study using regression analysis. First, it is started with the estimation of potential deforestation in Indonesia, and then it continues with the explanation on the result of regression analysis, which emphasizes the variables that have statistically significant effect on deforestation. Finally, the last part is the conclusion, which summarizes the interpretation of the result of regression analysis, including the benefits and drawbacks of the model, and the policy implication as well.

6.1 Potential Deforestation in Indonesia