Causal Effects Model 的估計

==Causal Effects 估計範例==

in Angrist and Pischke (2017, JEP, p130-132),主要提出觀念的是 Dale and Kruger (2002, QJE)

他們想估計美國唸私立大學和唸公立大學的差異,用較明確的因果關係為主的估計法。

Casusal Effect: 在申請入學時,同時接到私大和州大入學許可,但最後後選擇唸私大、或州大的學生為樣本:

Yi: 全部的樣本的畢業生所得,可觀察的,其中又分為兩類

Y1i : 第 i 個樣本的「受私校教育後」所得

Y0i : 第 i 個樣本的「受州校教育後」所得

以上兩個,令

Pi: =1 if 第 i 個樣本唸私立,=0 otherwise 唸州立大學

合理的假設是,每個人在受大學教育之前,原本就有一定的能力

Y10 : 第 i 個樣本的原來能力
重點:美國私立大學教學效果,是否來自教學,還是學生的本質。

因為好學生集中去唸名私校,所以畢業後收入高,不見得是私校的努力。這個稱為 selection bias 樣本選擇偏誤。

私校的教學效果 (用大概畢業20年後的 earning 來衡量) 之差異為:

Y1i – Y0i

若教學有效的話,然後差異的平均是 β

H0 : E(Y1i – Y0i) = β>0

假設E(Y0i) = α, 即

Y0i =α + ηi

α 為學生原來的潛力, ηi是誤差,或個別差異,這個個別差異會和選私校有關係,例如家庭背景、爸媽是否畢業於私立..。

Caussal-Effect model

Yi = α+βPi+ηi

Pi 和 ηi 是(統計上)不獨立的,也就是無法滿足迴歸上原來的獨立性要求。

這個 causal-effect model 的想法創新就在此,他們提出比較不嚴格的「條件獨立性假設」(conditional independence assumption),

E(ηi|Pi,Xi) = E(ηi|Xi)

所以要找其它的可能影響畢業後所得能力的變數 X (例如 SAT 的分數…),又稱為控制變數 control variable,來加入估計,觀念上是

E(ηi|Pi,Xi) = E(ηi|Xi) = γXi

所以, causal-effect model 最後就變成

Yi=α+βPi+γXi+ηi

此法可建構出 unbiased 和 consistent 的 β 估計,而且它有明確的意義:唸私校和唸公校的「效果差異」平均值。
==ref==

Angrist, Joshua D., and Jörn-Steffen Pischke. “Undergraduate econometrics instruction: through our classes, darkly." Journal of Economic Perspectives, 31.2 (2017): 125-44.

Dale, Stacy Berg, and Alan B. Krueger. “Estimating the payoff to attending a more selective college: An application of selection on observables and unobservables." The Quarterly Journal of Economics, 117.4 (2002): 1491-1527.

New directions for modelling strategic behavior: Game-theoretic models of communication, coordination, and cooperation in economic relationships

Crawford, Vincent P. “New directions for modelling strategic behavior: Game-theoretic models of communication, coordination, and cooperation in economic relationships." Journal of Economic Perspectives 30.4 (2016): 131-50.

URL:http://pubs.aeaweb.org/doi/pdfplus/10.1257/jep.30.4.131

==original Abstract==

In this paper, I discuss the state of progress in applications of game theory in economics and try to identify possible future developments that are likely to yield further progress. To keep the topic manageable, I focus on a canonical economic problem that is inherently game-theoretic, that of fostering efficient coordination and cooperation in relationships, with particular attention to the role of communication. I begin with an overview of noncooperative game theory’s principal model of behavior, Nash equilibrium. I next discuss the alternative “thinking" and “learning" rationales for how real-world actors might reach equilibrium decisions. I then review how Nash equilibrium has been used to model coordination, communication, and cooperation in relationships, and discuss possible developments

GIMS, an open source market software

S. Palan, GIMS-Software for asset market experiments. J. Behav. Exp. Finance 5, 1–14, (2015). Medline doi:10.1016/j.jbef.2015.02.001 (可免費閱讀)

==by YNY==

GIMS 是架在 z-Tree 上,專門用來跑財務資產市場 (又稱 double auction asset market) 的實驗平台軟體,採開放源碼 (open source) 授權。

此文亦介紹、比較了其它相關的財務市場實驗平台軟體,參見文中的 Table 1。

  • EconPort MarketLink(based on Java and experiments can be run over the internet)
  • Flex-E-Markets(not open source)
  • GIMS(based on z-Tree)
  • jMarkets (open-source software based on Java)
  • Rotman Interactive Trader (非免費)
  • SoPHIE Labs (非免費)

Continuous double auction market interface

新經濟學範式告別曼昆?

yinung: 終於有新式經濟學教科書問世了… 用三個觀點重新描述「數據時代」的經濟學。

原文網址:https://kknews.cc/finance/nnxg8.html

…有一本全新的《經濟學》面世,曼昆《經濟學原理》突然面臨強有力的挑戰。這就是阿西莫格魯、萊布森與李斯特合著的《經濟學》。這三人都是當下紅得發紫的經濟學家,也都已在多個領域做出突出貢獻,被認為是未來數年諾貝爾經濟學獎理所當然的候選人…

三人版《經濟學》比曼昆的《經濟學原理》還薄了許多。仔細一讀,《經濟學》確實增添了大量三人自身的研究成果,但同時也刪減了大量經典內容。也就是說,三人都認為那些內容不再是探索未來經濟學所必須掌握的知識。刪去教科書里的經典內容,比補充內容需要更多的勇氣。

他們三人沒有拋出曼昆式的「十大原理」,但也開宗明義地提出了「三個渾然一體的主題」,分別是優化(optimization)、均衡(equilibrium)和經驗主義(empiricism)

 

 

微观经济与宏观经济实验原理.方法与案例

作者: 刘贞 / 施於人
出版社: 机械工业出版社
出版年: 2012-9
页数: 186
定价: 26.00元
ISBN: 9787111389781

Saving Face and Group Identity

==notes by yinung==
很有趣的保留面子 (saving face) 的實驗?
Date: 2015
By: Tor Eriksson (Department of economics – University of Aarhus)
Lei Mao (GATE Lyon Saint-Étienne – Groupe d’analyse et de théorie économique – ENS Lyon – École normale supérieure – Lyon – UL2 – Université Lumière – Lyon 2 – UCBL – Université Claude Bernard Lyon 1 – Université Jean Monnet – Saint-Etienne – PRES Université de Lyon – CNRS, Central University of Finance and Economics)
Marie Claire Villeval (GATE Lyon Saint-Étienne – Groupe d’analyse et de théorie économique – ENS Lyon – École normale supérieure – Lyon – UL2 – Université Lumière – Lyon 2 – UCBL – Université Claude Bernard Lyon 1 – Université Jean Monnet – Saint-Etienne – PRES Université de Lyon – CNRS)
URL: http://d.repec.org/n?u=RePEc:hal:wpaper:halshs-01161750&r=net
Are people willing to sacrifice resources to save one’s and others’ face? In a laboratory experiment, we study whether individuals forego resources to avoid the public exposure of the least performer in their group. We show that a majority of individuals are willing to pay to preserve not only their self-but also other group members’ image. This behavior is frequent even in the absence of group identity. When group identity is more salient, individuals help regardless of whether the least performer is an in-group or an out-group. This suggests that saving others’ face is a strong social norm.

Experiments in EJAE

The advantage of resource queues over spot resource markets: decision coordination in experiments under resource uncertainty

Fonseca, Miguel A., Alexander Pfaff, and Daniel Osgood. “The advantage of resource queues over spot resource markets: decision coordination in experiments under resource uncertainty." American Journal of Agricultural Economics 94.5 (2012): 1136-1153. [oxfordjournals.org]; [Open Access PDF]

==Notes by yinung==

本文用實驗法探討灌溉水源 (以 water queues 的形式) 的事前分配 (用 bidding) vs 事後水源拍賣,對生產決策 (種高/低水量需求作物) 之影響。以2007-08 英國 Exeter 大學學生和巴西農業區半工半讀的學生或公務員為受試者,實驗結果證明,事前的 water queues 制度,相對於事後的 spot market 分配水源的制度,有助於提昇效率。
其原因主要是在,決策者面對 environmental and strategic 兩種不確定之下, 可以利用先知道對手對環境不確定之選擇,用來當做生產決策協調之依據。
[引文] The advantage of resource queues over spot markets concerns two types of uncertainty, environmental and strategic.
(較早寫的)
此篇用經濟實驗法, 研究農民在灌溉用水不確定下之決策; 投入種子、肥料、整地的成本, 但利潤取決於不確定供水。在環境不確定 (雨量多寡) 前,先以 bidding (water queues) 決定用水優先序,可做為投入種植決策之協調機制,使得在事前 water queues 制度下, 比雨量確定後才拍賣決定水權制度更有效率。

==abstract==

Farmers have to make key decisions, such as which crops to plant or whether to prepare the soil, before knowing how much water they will get. They face losses if they make costly decisions but do not receive water, and they may forego profits if they receive water without being prepared. We consider the coordination of farmers’ decisions, such as which crops to plant or whether to prepare the soil when farmers must divide an uncertain water supply. We compare ex-ante queues (pre-decision) to
an ex-post spot market (post-decision & post-rain) in experiments in rural Brazil and a university in England. Queues have greater coordination success than does the spot market.

==References==

Lefebvre, M., L. Gangadharan, and S. Thoyer. 2012. “Do Security-Differentiated Water Rights Improve the Performance of Markets?” American Journal of Agricultural Economics. doi:10.1093/ajae/aas060.
本文回顧到這篇:
The most relevant experiment of which we are aware is that of Lefebvre et al. (2012), which studies profits with security-differentiated water rights (mimicking the essence of a queue) versus non-differentiated, equal share rights. The differentiated rights are seen to have profit and risk-management benefits relative to the non-differentiated rights.

Consumer Price Search and Platform Design in Internet Commerce

==noted by yinung==
這個網站搜尋、比價的主題很久了, 但仍是有趣, 也是有人用 law of one price 的觀念來看, 但裡頭的實驗怎麼做, 再了解。
Date: 2014-08
By: Michael Dinerstein
Liran Einav
Jonathan Levin
Neel Sundaresan
URL: http://d.repec.org/n?u=RePEc:nbr:nberwo:20415&r=net
Search frictions can explain why the “law of one price" fails in retail markets and why even firms selling commodity products have pricing power. In online commerce, physical search costs are low, yet price dispersion is common. We use browsing data from eBay to estimate a model of consumer search and price competition when retailers offer homogeneous goods. We find that retail margins are on the order of 10%, and use the model to analyze the design of search rankings. Our model explains most of the effects of a major re-design of eBay’s product search, and allows us to identify conditions where narrowing consumer choice sets can be pro-competitive. Finally, we examine a subsequent A/B experiment run by eBay that illustrates the greater difficulties in designing search algorithms for differentiated products, where price is only one of the relevant product attributes.
JEL: D12 D22 D83 L13 L86

Medical Insurance and Free Choice of Physician Shape Patient Overtreatment. A Laboratory Experiment

==Noted by yinung==

終於看到有健康與醫療的實驗研究了

Date: 2014-09-30
By: Steffen Huck (Wissenschaftszentrum Berlin für Sozialforschung (WZB))
Gabriele Lünser (University College London – Centre for Economic Learning and Social Evolution (ELSE))
Florian Spitzer (Department of Economics, Vienna Center for Experimental Economics (VCEE), University of Vienna)
Jean-Robert Tyran (Department of Economics, Copenhagen University)
URL: http://d.repec.org/n?u=RePEc:kud:kuiedp:1419&r=net
In a laboratory experiment designed to capture key aspects of the interaction between physicians and patients in a stylized way, we study the effects of medical insurance and competition in the guise of free choice of physician. Medical treatment is an example of a credence good: only the physician (but not the patient) knows the appropriate treatment, and even after consulting, the patient is not sure whether he got proper treatment or got an unnecessary treatment, i.e. was overtreated. We find that with insurance, moral hazard looms on both sides of the market: patients consult more often and physicians overtreat more often than in the baseline condition. Competition decreases overtreatment compared to the baseline and patients therefore consult more often. When the two institutions are combined, competition is found to partially offset the adverse effects of insurance: most patients seek treatment, but overtreatment is moderated.
Keywords: Credence good, Patient, Physician, Overtreatment, Competition, Insurance, Moral hazard
JEL: C91 I11 I13