Philip D. Adams教授，澳大利亚科学院院士，澳大利亚Victoria大学CoPS中心前任主任。Adams教授1989年于澳大利亚墨尔本大学取得经济学博士学位，先后在CoPS中心担任高级讲师、副研究员以及研究员，并曾担任太平洋经合组织经济预测小组的澳大利亚代表，在多区域多部门经济模型的构建和应用，以及相应的经济分析和经济预测方面具有丰富的经验。在CGE模型方面，更是作为主要研究人员直接参与了诸如MONASH模型、MMRF动态模型等大型经济模型的构建和开发。
Adams教授在众多国际期刊上发表论文60余篇，涉及自由贸易、环境经济学、区域经济学、宏观经济分析和CGE模型构建等多个领域。
近些年来，Adams教授的研究方向主要集中于气候变化经济学以及气候政策的经济影响方面。他不仅出色地完成了多项将气候变化及气候政策引入经济模型的重大项目，也为澳大利亚政府提供了大量可靠的政策建议。在2004年到2011年间，Adams教授为推广温室气体政策方面的研究作出了巨大的贡献。Garnaut气候变化评述项目和联邦公库项目中关于气候变化和排放交易模型的开发、修改、调试和应用，也都离不开Adams教授的辛勤工作。
采访内容
Question.1
Q：How long have you been using CGE models? What is the most influential work you have done to the policy field?
问：请问您用CGE模型用了多长时间？您认为在政策领域里做过最有影响力的工作是什么？
P：Well, I've been using CGE models for a very very long time. I started at my PHD, which I finished at 1987, so I guess something that I do modeling for, fulltime, thirty years now. The most influential work I've done is probably the work undertaken for the Australia Government on climate change, and on policy design to reduce the greenhouse gases emissions. That work was done between 2009 and 2013, which was highly influential in the government of the day, designing a policy based on the emission trading scheme to reduce greenhouse gas emissions. That was probably the most impactful work that we did. The government was very keen to understand what the very detailed impacts of carbon tax and emission trading scheme would have on the Australian government. CGE model was ideal because of the level of the detail to answer the detailed questions that the government was asking.
答：我已经使用CGE模型非常长时间。从1987年完成博士学位到现在，我使用CGE模型加起来超过30年时间。我做过最有影响力的工作可能就是为澳大利亚政府做的关于气候变化的项目，以及减少温室气体排放的政策设计。这一项目是在2009年至2013年间完成，对当今的政府决策影响很大，是一项旨在减少温室气体排放的碳交易机制政策设计，这也可能是我们做过的最具影响的工作。澳大利亚政府对碳税以及碳交易机制对政府部门所产生的具体影响非常感兴趣，而CGE模型对政府提出的问题能够提供非常详尽的解答。
Question.2
Q：What are the advantages of CGE model, relative to the econometric model, or Input-Output model?
问：相对于计量经济模型以及投入产出模型，CGE模型的优势是什么？
P: Well first of all, relative to the input-output model, CGE models include the input-output as a special case, but CGE models are more general, like allowing for relative price effects, where is input-output model generally make no allowance for prices. Indeed if prices are presented in input-output model, generally prices are fairly passive.
Well, relative to the econometric models, that's a very good question. First, you have to understand that econometric model is generally based on time series or panel data that reflect the past. So past behavior course is contained in that historical record, then effectively project into the future or extrapolate from the past into the future. The CGE model is different. CGE models don’t pin on the data. They start off with the structure or a theory. The theory is well grounded in established principles, typically in the neoclassical, so it involves profit maximization, or cost minimization etc.. So econometrics is well known in a CGE model, well known not only by the modelers, but by other people. When looking at the effects of the policy, if the policy reflects past and changes etc., then econometrics would pick up that changes very well. But if it doesn’t, if it is a new policy of the economic changing, then a structural approach such as the CGE approach which is far better predicting future consequences of changes to the economy.
答：首先，相对于投入-产出模型，CGE模型包括了投入-产出分析。但CGE模型更为全面，例如模型中包含了相对价格效应，而投入-产出模型一般来说并没有引入价格。即时投入-产出模型出现了价格因素，一般来说也是相当被动的。
而相对于计量模型(CGE模型的优势)，这是一个非常好的问题。首先，你必须了解计量模型一般是基于时间序列及面板数据，这些数据仅反映过去。因此过去的行为过程是包含在历史数据中，并且有效的预测未来或是根据过去的历史数据映射未来。但是CGE模型不一样，CGE模型并不受数据限制。CGE模型从结构或理论出发。这些理论是建立在一些成熟或经典理论基础上，例如包括了利润最大化以及成本最小化理论等。因此在CGE模型中计量方法不仅被建模者还被其他人所熟知。当我们评价政策效应时，如果政策反映的是过去以及变化，计量模型可以很好的解释这些变化。但如果不是，如果这是一项反映经济变化的新政策，那么像CGE模型这样的结构性方法则会更好的预测对经济造成的影响。
Question.3
Q：you have a great deal of experience in teaching CGE and policy analysis. What is your advice to a beginner, and what makes a good modeler essentially?
问：您在CGE教学及政策分析方面有丰富的经验。您对初学者有什么建议？以及什么因素从根本上造就一个优秀的建模者？
P：Personal experience, sticking to the task. Learning CGE modeling is not easy. CGE model can be quite large, and can be quite complex. And it involves a large amount of data, and involves quite detailed solution techniques. And understanding results, which is a key to any modeling task, can only be learnt essentially through experience. So learning the basic is the key. Understanding the basic is the key. But just like with the language, you have to keep using the model. You have to keep focusing on the model, so that you could learn how the model works. It is not something that you can pick up just by reading a textbook or by listening to a lecture or by talking to someone else over a short term. It is something that you have to concentrate and you have to continually learn. Although I have been working on this for thirty years, I am still learning a lot about CGE model.
答：个人经验来说，坚持学习。学习CGE模型不是一件容易的事情。CGE可以是非常庞大的、非常复杂（的模型），它包含了大量的数据以及非常详细的解决方案技巧。对于结果的理解最根本的是需要从经验中学习的，这也是学习任何模型的关键。因此学习基本的内容是关键，理解模型也非常重要。但是正像语言一样，你必须要持续使用模型，持续关注模型，这样你才能知道模型如何运行。你不能仅依靠短时间内读教科书、听讲座或者是跟其他人聊天就想快速学习，必须要专注且持续学习。尽管我已经在这一领域工作三十多年，但我仍然还在坚持不懈学习CGE模型。
Question.4
Q：As a modeler, you know CGE model is a very complex model, how to promote CGE and communicate with policy makers?
问：作为建模者，您知道CGE模型非常复杂，那要怎样去推广CGE模型以及和政策决策者沟通？
P： The great danger in any… this is not just CGE modeling, but in any modeling study， is that the model is regarded as a black box. That has to be avoided as much as possible. How do you avoid a model being described as a black box? You avoid it by explaining how the model works in detail to a person or to a level where an economist who is not familiar with the model itself can understand the key mechanisms. To promote CGE modeling, to get the CGE modeling accepted, getting the story of CGE models telling across to the general publics, across to the policy makers, across to the media etc.. are the most important. That is why we put some emphasis on interpretation. Interpretation is a key. If you can’t explain the results from your modeling exercise to people who are not familiar with the actual model, then you are doing the disservice. People would naturally gonna be turned off, and naturally gonna be thinking this is just a black box. That has to be avoided.
答：对于CGE模型，或者说对于任何一个模型来说，非常危险的一件事情就是模型经常被认为是一个“黑箱”。我们要尽量避免这样的问题。如何避免模型被认为是“黑箱”呢？你如果可以向其他人或其他不熟悉这个模型的经济学家解释清楚模型运行的细节，帮助他们理解模型机制，这样就可以避免（上述问题）。为了推广CGE模型，让更多人接受，让CGE模型相关的背景知识被普通大众、政策决策者以及媒体等所熟知，我们要强调解释（的重要性）。模型解释非常重要。如果你不能将你在建模过程中所产生的结果向不熟悉模型本身的人解释清楚，那你所做的一切都徒劳。大众自然而然会对这个模型失去兴趣，也会自然而然地认为这个模型是“黑箱”。这是需要避免的问题。
Q: So it is important that understanding the model first, and then interpreting the model.
问：因此首要的是理解模型，接着是解释模型。
P: Yes. Interpret, and be able to convey the interpretation in a way that others can understand and in a way that potentially can influence policy makers’ decisions.
答：是的，要解释，并且要以能够向所有人解释清楚的方式，以及能够影响政策决策者决定的方式。
Question.5
Q：What do you think of the future of the CGE model's development? What areas do you think got potential, and truly able to add value to clients, no matter government or private sectors?
问：CGE模型已有将近六十年的历史。您认为CGE模型发展的前景是什么？您认为在什么领域可以为客户提供更多的价值，不管是针对政府或是私人部门的用户？
P: I think CGE modeling has a very rich future. Computer is getting faster. CGE modeling software are getting smarter. We use the GEMPACK software. Other people use GAMS. Both software now can solve reasonably large models at reasonably quick speed, much larger and much quicker than I could even 5 years ago. As the software gets better, so the CGE model will become more accessible.
What about the theoretical developments within the CGE model? Well, there are enormous amounts of developments that could be embodied within the CGE framework, both macroeconomic and microeconomic. I think we should be increasingly trying to incorporating to CGE models, with the insides that macroeconomicsmodelers have about the behavior and so forth. I think increasingly we should be incorporating into CGE models, with the insides that the micro-simulation people have about individual and household behavior. One of the key areas where I think we could expand the CGE modeling base is through linking. I am not talking hard linking, I am talking essentially soft linking, but linking to scientific and engineering, and other forms of partial equilibrium models, of a really great scope for the CGE model to expand already the vast detail into an even greater detail within specific markets. The work on the emission before, the emissions trading and carbon tax works were enhanced by our ability to link the CGE system with the electricity supply system, so we are able to introduce into our model very detailed electricity technology stories, which really enriched the CGE modeling.
答：我认为CGE模型具有非常远大的前景。我们的计算机运行得越来越快，CGE软件也越来越智能。我们中心（CoPS）使用GEMPACK软件，其他（建模者）使用GAMS。这些软件都能以相当快的速度解决相当大型的模型，比我之前甚至是五年前能做的还要快、还要大。正因为软件变得更好了，CGE模型也将会被更多人所使用。
关于CGE模型理论发展的问题，我认为有相当一部分理论发展可以体现在CGE模型框架中，包括微观、宏观经济理论。我认为我们应该不断地将（更多）的宏观建模者所理解的行为理论，以及将微观经济学中对个体以及家庭行为嵌入到CGE模型框架中。其中我认为CGE模型拓展中一个最清晰的路径就是“连接”。我并不是在讨论“硬连接”，而是“软连接”。我讨论的是通过与其他自然科学、工程学、以及其他形式的局部均衡模型的连接，在一些具体市场领域里，CGE模型的拓展将具有更广阔的空间。像我们此前做的关于温室气体排放的研究工作，正是由于我们将CGE系统以及电力供应系统进行连接，这样大大提升了我们项目的质量。通过这种方式我们将电力技术的背景知识引入我们的模型，也极大拓展了我们的知识领域，丰富了CGE模型内涵。
专访录音可供*载下**，量化经济平台注册会员请到量化经济平台（www.econmod.cn）的交流中心——行业资讯——数据/资源板块自行*载下**。尚未注册量化经济平台会员的，可先在网站注册，然后共享平台资源。