Cybernetics and Macroeconomics: A Comment
Cybernetics and Macroeconomics: A Comment
复制标题
控制论和宏观经济学：评论
DOI：
10.1111/j.1465-7295.1976.tb00391.x
复制
发表时间：
1976
期刊：
影响因子：
1.8
通讯作者：
A. Leijonhufvud
中科院分区：
文献类型：
--
作者：
M. Aoki;A. Leijonhufvud
控制论和宏观经济学：评论青木正雄和阿克塞尔·莱琼霍夫夫加州大学洛杉矶分校我们与科克伦和格雷厄姆没有什么特别的争执。这不是那种评论。他们关注的问题是：为什么受系统理论影响的经济学成果没有得到回报？为什么早期的努力逐渐消失了？虽然我们认为他们对这件事的评论和推断是合理的，但他们的总和给我们的印象是不确定的，相当令人不满意的。当然，这些问题可能没有非常有趣或有教育意义的答案。因为，毕竟，研究努力的失败--即使是那些大张旗鼓的努力--也是很常见的。比成功更常见的是成功。没有什么可预期的，在跨学科的风险投资中，失败是意料之中的。还有一个单调乏味的实物--他们如何失败的教训，我们几乎不需要。我们知道这种情况会如何发生：当人们不耐烦地想要迅速获得回报时，他们就会失败，他们更倾向于将自己从传统的智力纪律中解放出来，而不是急于同时服从两个领域的双重约束，而且他们不太理解为什么他们想要合并的两个领域一开始就是分开的。诸若此类。或许，系统论中的经济学家除了可以借用简单的类比和各种各样的技巧之外，没有其他东西可以借鉴了吗？当然，科克伦-格雷厄姆的调查是建立在这样一个观点之上的。我们同意这一观点。然而，他们没有告诉我们太多关于错过的机会是什么。然而，在我们看来，需要传达一些关于这一点的概念，以便为这种缺乏成功的跨学科故事注入兴趣。试图笼统地阐述我们对潜在机会的看法是不可取的。如果自命不凡的哲学观足以推动企业的发展，那么它早就开始竞争了。为了避免这种事情，我们将讨论几个简单的插图，使深刻和可怕的深刻几乎是不可能的。其中之一是由Cochrane和Graham提供的。1.CG论文中的说明性框图是塔斯汀用来表示他选择(根据经济学家同事的建议)工作的系统的图表类型的一个简单例子。它描绘了一种迭代方案或例程，用于找到简单凯恩斯主义的两个联立方程的解。实质经济利益依附于这种表示模式，其程度(如果有的话)使人愿意将连续回合251经济调查卷的结果联系起来。XIV，1976年6月对经济系统中假设的一系列事件的这种迭代的经济调查。在所有这种迭代方案的类中，我们想要区分一个特定的子类，其成员对应于我们将称为反馈的东西。现代系统理论专门处理属于这个子类的系统。事实上，这被系统理论家和控制工程师认为是理所当然的，以至于该领域没有公认的术语来区分这个子类和其他也可以由迭代例程表示的“系统”。CG框图所示的简单凯恩斯主义模型不属于这个子类。塔斯汀从系统科学借用的形式主义和术语中的经济模型也不属于错误激活反馈系统的类别。我们认为，这就是早期试图利用系统科学为经济理论谋利的地方。当时年轻纪律的概念(如果这是这样的话)是错误的系统类型；既然如此，他们的…
CYBERNETICS AND MACROECONOMICS: A COMMENT MASANAO AOKI and AXEL LEIJONHUFVUD University of California, Los Angeles We have no particular quarrels to pick with Cochrane and Graham. This is not that sort of Comment. The questions they concentrate on are: Why has the system theory-influenced work in economics not paid off? Why have the early efforts petered out? While we find their comments and inferences on the matter reasonable, their sum total strikes us as inconclusive and rather dissatisfying. Of course, it might be that these questions do not have very interesting or instructive answers. For, after all, failure of research efforts-even of the much bally-ho'ed ones-are common. Far more common than are successes. There is nothing about the expectable, and in interdisciplinary ventures failure is to be expected. Yet another humdrum object-lesson in how they fail we hardly need. We know how that will run: They fail when undertaken by people who are impatient for quick pay-offs, more prone to free themselves of a traditional intellectual discipline than anxious to submit to the doubled constraints of two at once, and less than appreciative of the reasons why the two fields they want to cojoin are separate in the first place. And so on. Is there, perhaps, nothing more for economists in system theory than facile analogies and sundry techniques to borrow? The Cochrane-Graham survey is, of course, predicated on the view that there is. We share that view. They do not tell us much, however, about what the opportunities were that have been missed. Yet, some conception of this need be conveyed, it seems to us, to inject interest into this interdisciplinary lack of success story. It would be inadvisable to try to couch our view of what the potential opportunities are in terms of generalities. If pretentious philosophizing on and could suffice to get the enterprise under way, it would be off to the races long since. To avoid that sort of thing, we will discuss a couple of illustrations so simple as to render deep and awful profundity rather impossible. One of these has been provided by Cochrane and Graham. 1. The illustrative block-diagram in the CG paper is a simple example of the type of diagrams utilized by Tustin to represent the systems he chose (on the advice of economist colleagues) to work with. It depicts an iterative scheme or routine for finding a solution to the two simultaneous equations of the simple Keynesian Substantive economic interest attaches to this mode of representation to the degree (if any) that one is willing to associate the results of successive rounds 251 Economic Inquiry Vol. XIV, June 1976 ECONOMIC INQUIRY of this iteration with a hypothetical sequence of events in an economic system. In the class of all such iterative schemes, we want to distinguish a particular subclass the members of which correspond to what we will call feedback Modern system theory deals exclusively with systems belonging to this subclass. In fact, this is taken so much for granted by system theorists and control engineers that the field has no accepted term to distinguish this subclass from other'systems' that may also be represented by iterative routines. The simple Keynesian model which the CG block-diagram illustrates does not belong to this subclass. Nor did the economic models which Tustin'redressed'in formalisms and terminology borrowed from system science belong to the class of error-activated feedback systems. This, we think, is where the early attempts to draw on system science for the benefit of economic theory went wrong. The concept of that then young discipline were (if that is the word) to the wrong type of systems; that being so, their …