人造游戏尚难以用模型以概之,何况真实世界?
经济增长即是总产出增长
Cobb-Douglas 生产函数:
\[Y = A K^{\alpha} L^{\beta} \]产出 \(Y\) (total production)和什么相关?
- 可累加投入(规模)
- 劳力投入(Labor) \(L\)
- 资本投入(Capital) \(K\)
- 固定因子
- 技术、管理等因素:全生产要素(Total Factor Productivity, TFP)或者知识(Knowledge) \(A\)
生产函数和行业相关,不同的行业对投入的敏感程度不一样,\(\alpha\) 和 \(\beta\) 表示敏感度,又称产出弹性。
衡量增长有俩个指标,一个是大家总共的钱袋子 \(Y\) ,另一个是每个人的钱袋子 \(y=\frac{Y}{AL}\) (output per effective unit of labour, 其中 \(AL\) 称作有效劳力 effective labour)。一般从宏观国家层面,经济增长指前者 \(Y\) 的增长。
经济发展的第一推动力
经济的长期增长与什么有关?新古典模型中有外生增长(Exogenous)和内生增长(Endogenous)模型两种思路:
- Exogenous Growth Model
- Solow-Swan Growth Model
- Harrod-Domar Growth Model
- Endogenous Growth Model
- AK Model
- Romer Model
- Aghion-Howitt Model
- Grossman-Helpman Model
外生模型即生产函数中某些因素不受模型所影响,比如恒定的变化率,像从外界无形的手推动经济不断往前,而内生模型则将经济的长期增长看作模型内部变量作用的结果。
外生增长 Solow-Swan Growth Model
Solow-Swan Growth Model[1]: 又称 Exogenous Growth Model(外生增长模型),描述了经济是如何增长的。
而在 Solow-Swan Model 里,假设规模报酬(Returns to Scale, RS)为恒定规模报酬(Constant Returns to Scale),即按相同比例扩大劳力和资本规模,产出同比例扩大。即 \(\alpha + \beta = 1\)。
产出要不断扩大,要么提高技术 \(A\) ,要么增大规模 \(K\) 和 \(L\) 。其三者是随时间的函数,技术和劳力的增长按指数表示
\[\begin{align} A(t) &= A(0)e^{gt}\\ L(t) &= L(0)e^{nt} \end{align} \]使用指数表示增长蕴涵假设有2:(1)视野极其宏观;(2)增长在自由环境中成指数关系(比如技术按指数可以用技术大爆炸解释,人口按指数可以用种群远远小于环境承载力时的种群增长关系解释)。
很显然,这里对于 A 和 L 的增长假设是粗糙的,且不是模型关注的重心,而做为模型的基本假设。此之谓“外生”,更确切地说是对 A 外生 —— 将 A 放在模型之外不纳入考虑。
而资本的关系可以用 投资 - 损耗表示,资本累积方程:
\[\dot{K}(t) = I(t) - \delta K(t) \]投资看作和产出相关,则:
\[\dot{K}(t) = sY(t) - \delta K(t) \]该模型的特点是,\(y\) 存在稳态解,也就是每个人富裕程度最终不变,而 \(A\) 和 \(L\) 的自然增长推动着总量 \(Y\) 不断增长。
资本累积方程蕴含资本扩大的内在趋势,即产出 \(Y\) 会用以扩大资本,从而进一步扩大产出 \(Y\) 。换句话来说,资本的来源本质是产出(废话嘛,设备不是钱买的)。Solow-Swan 中的稳态实际上来自于资本损耗 \(\delta\),这被解释为生产资料老化维修等问题(维护成本)。
Victoria 3 的经济模型
@Fakeseoi_into_osoto 在 Reddit [2]上说 Victoria 3 的经济模型更类似 Solow-Swan Growth Model。理由是无论怎么经营,最终总会达到一个不能再提高的值[3],即 Solow-Swan 中的稳态,此时只能追求通过应用新技术即提高 TFP 来进一步扩大收益。
"Unless you've broken the game by being so goddamn good, there's a solid chance that you reach a point of economic stagnation where further development in your economy is really difficult. This is a sign you've reached equilibrium-- the cost to expanding further is too taxing since your replacement rate of capital is too high. It simply requires too many engines to replace... until you unlock electric engines." —— Fakeseoi_into_osoto
另外 @Fakeseoi_into_osoto 认为游戏中“使用”的概念和资本损耗相近。盲目地扩大资本并不能带来经济的提升。
"In this game we can roughly assume that the depreciation of the machines is in the form of being "used up" in processes, like when mining takes a certain number of engines. But as anyone who has played the game multiple times will know, engines are fairly unprofitable for most of the game and it serves as a cap on industrializing your economy." —— Fakeseoi_into_osoto
这似乎比较勉强,这实际是描述边际递减效应,或者资本和劳动的互补性。这被蕴含在了 Cobb-Douglas 的产出弹性 \(\alpha < 1\) 中。
一位已经消耗的用户反驳了这个观点。他认为游戏中对于新技术的使用添加了限制,即当规模较小时,新技术带来的产出小于投入研发新技术的成本,而 Solow-Swan 模型 \(A\) 的增长表明只要有新技术就会立刻纳入使用。这实际是对 Solow-Swan 模型外生的本质进行驳斥。
"One of the major limiting factors in the game to adopting new production methods seems to me to be that your economy may not be harnessing the throughput bonuses enough to supply firms to use them. This is similar to the sort of arguments you'd get from economists like Adam Smith, Karl Marx, or the contemporary complexity economists (namely W. Brian Arthur). In the Solow model, there wouldn't be cases where production can't just adopt the new technology immediately (whether or not Solow believed that to be true in the real world is another case)." —— Unknown
他随后进一步举了大国小国间自由贸易的例子论证技术需要在规模效应(throughput bonuses)下才能发挥作用。
"If you're already an industrial powerhouse, free trade tends to work out pretty well since you'll be able to harness your throughput bonuses to export high value added goods. If you're a less developed country though, it seems better in my experience to be more protectionist to develop your strategic industries without them having to compete with the industries in the Great Powers." —— Unknown
TODO ...