Secondly, we have a "good" institutional environment. 192 1 year, the Australian government set up a tariff commission (later industrial mutual aid commission, now production commission) to make suggestions on the implementation of tariffs and quotas. For a long time, it took the route of trade protection, but surprisingly, when Alf Rattigan was chairman at the end of 1960s, he realized that tariffs had harmed some people, and he hoped to find a quantitative model to identify the victims and quantify their losses. He suspects that if all the losses are fully recognized, it may shake the traditional trade protection policy.
Rattigan realized the importance of economic modeling, so he persuaded the government to invest in the university to establish a modeling project in 1969, but failed. One of the reasons is that the model design of scholars is far from the actual needs of decision makers. But Rattigan didn't stop. He established a new project in 1975, which was later called IMPACT. By setting up a public government agency (instead of still in the university) and letting scholars come to this agency, researchers can be required to focus on specific policy analysis at any time. At the same time, the largest chemists have academic freedom to accept comments on the model, which ensures the openness and transparency of the project. IMPACT project is open, and one of the important aspects is to provide training courses, encourage more people to apply CGE model, and constantly accept opinions to achieve improvement. This tradition continues to this day, and GTAP actively promotes training programs, which is also an important reason for its success in the world.
Finally, we have a "good" model. ORANI, the main model of IMPACT project development, was officially released and used in 1977. What are the characteristics of Aranyi model? First of all, scholars' research on the problem is mainly theoretical, and only need to examine the relevant relationship clearly. The government needs details when formulating policies, for example, it needs to know the impact of a policy change on the production of an automobile part, not the impact of the whole industry or economy; For example, policy makers are concerned about the regional economy, such as the influence of a certain district in Melbourne rather than the whole Melbourne or the whole country. Another is that decision makers pay attention to actual data, rather than qualitative or approximate data expressed by scholars (enough to explain the problem). Considering the influence of transportation cost on the marginal price of products, if the CIF price of an imported product is 1 USD and the tariff is 50%, how much does the price actually increase for consumers? This requires consideration of domestic transportation costs and other issues. The actual increase may be only 16%, not 50%. Therefore, decision makers pay attention to actual data, while ORANI model pays attention to actual data and details from the beginning.
At the same time, there are many inventions in the model structure, including Armington hypothesis, that is, the same goods, imported and domestic, are set as different goods because of different prices; Pay attention to the variables of technological change; The calculation method of the model is very simple, and the influence of policy change can be seen by differentiating the initial solution, so that the large model can be solved; And the invention of model closure-telling us which variables are endogenous and which are exogenous. It is very important to close the model, because generally speaking, the number of variables in the model is more than the number of equations, so it is impossible (and unnecessary) to solve it completely.
The following mainly talks about the calculation method and the problem of model closure. ORANI model is a more complex version of the algorithm used by Johansen in 1960. Johansen linearizes all the equations of the model, making the model a linearized system (variables become percentages). Although this method has been able to deal with the large-scale system model of 1960s, it has been questioned by many economists because of its linear error. ORANI followed this method and extended it, that is, using multi-step Johansen algorithm, which can eliminate the error. This simple and correct method can help us solve thousands of complex equations.
The closed model means that exogenous variables can be selected flexibly, which is also the characteristic of ORANI model, thus making the model widely used. For example, in the analysis of ORANI policy effect, short-term capital stock can be regarded as exogenous variable, while return on capital can be regarded as endogenous variable; In the long run, the rate of return is an exogenous variable, while the capital stock is an endogenous variable. For another example, in the short term, the real wage rate is exogenous and employment is endogenous; In the long run, employment is exogenous and the real wage rate is endogenous; Wait a minute.
Recently, closures have been introduced into dynamic models. There are four kinds of basic closures: (1) Historical closures, that is, many variables such as consumption, investment, government expenditure, export, import, employment and capital stock can be introduced as exogenous shocks by selecting exogenous variables. Through calculation, we can make detailed estimates of technological changes and preference changes, and update the input-output table data. (2) Decomposed and closed, that is, technology and preference can be regarded as exogenous, and historical activities will impact technology and preference. This closure can analyze the role of industry output growth and the influence of changes in natural exogenous variables such as technology and preference on endogenous variables. (3) Prediction is closed, that is, we regard our predictable variables as exogenous, and then analyze their influence on other unpredictable variables (endogenous variables). (4) Policy closure takes policy as an exogenous variable and analyzes how policy changes will affect the economy.