2024.8.19 詳細(xì)議程
會(huì)議主持：李俊青-教授 南開大學(xué)
2024.8.19 上午安排如何解決能源消費(fèi)所帶來的環(huán)境污染對(duì)公眾健康的負(fù)面影響，是實(shí)現(xiàn)健康中國(guó)目標(biāo)的重要挑戰(zhàn)，而清潔能源發(fā)展則為此提供了一個(gè)可行的治理路徑。本文以西氣東輸二線工程的投產(chǎn)運(yùn)營(yíng)作為準(zhǔn)自然實(shí)驗(yàn)，利用2006~2015年的中國(guó)健康與營(yíng)養(yǎng)調(diào)查數(shù)據(jù)（CHNS），實(shí)證考察了清潔能源發(fā)展如何影響公眾健康。研究發(fā)現(xiàn)：西氣東輸工程產(chǎn)生了健康效應(yīng)，在通過多種穩(wěn)健性檢驗(yàn)后，仍能顯著提升沿線地區(qū)公眾健康水平。但這一效應(yīng)主要體現(xiàn)在城市居民和老年人群體，且家庭用能結(jié)構(gòu)改善、企業(yè)污染減排、城市環(huán)境質(zhì)量提升是主要作用渠道。進(jìn)一步分析表明，“煤改氣”政策有助于增強(qiáng)工程的健康效應(yīng)。福利分析顯示，該工程降低了個(gè)人及地區(qū)醫(yī)療支出，并促進(jìn)了當(dāng)?shù)鼐蜆I(yè)。本文的發(fā)現(xiàn)對(duì)如何深化能源供給側(cè)結(jié)構(gòu)性改革，助力實(shí)施健康中國(guó)戰(zhàn)略，提供了鮮明的政策啟示。
關(guān)鍵詞：西氣東輸工程 清潔能源發(fā)展 健康效應(yīng)
This talk will demonstrate how to produce informative, robust, and complex graphs using reproducible official and community-contributed routines in Stata. We will also discuss commonly used programming tools and tips for creating more engaging graphs.
To make the conventional synthetic control method more flexible to estimate the average treatment effect (ATE), this article proposes a quasi- synthetic control method for nonlinear models under the index model framework with possible high-dimensional covariates, together with a suggestion of using the minimum average variance estimation (MAVE) method to estimate parameters and the LASSO type procedure to choose high-dimensional covariates. We derive the asymptotic distribution of the proposed ATE estimators for both finite and diverging dimensions of covariates. A properly designed Bootstrap method is proposed to obtain confidence intervals and its theoretical justification is provided. When the dimension of covariates is greater than the sample size, we suggest using the robust version of sure independence screening procedure based on the distance correlation to first reduce the dimensionality and then apply the MAVE approach to estimate parameters. Finally, Monte Carlo simulation studies are conducted to examine the finite sample performance of our proposed estimators and Bootstrap procedure. In addition, an empirical application to reanalyzing data from the National Supported Work Demonstration demonstrates the practical usefulness of our proposed method.
會(huì)議主持：王群勇-教授 南開大學(xué)
2024.8.19 下午安排本文重點(diǎn)討論新質(zhì)生產(chǎn)力研究現(xiàn)狀以及測(cè)度、評(píng)估過程中應(yīng)該注意的問題。并沒有進(jìn)行新質(zhì)生產(chǎn)力發(fā)展指數(shù)的實(shí)質(zhì)性計(jì)算。建議國(guó)家統(tǒng)計(jì)局把“新質(zhì)生產(chǎn)力發(fā)展指數(shù)”作為產(chǎn)品定期生產(chǎn)、發(fā)布。而不是由民間自行發(fā)布。
Average causal response function (ACRF) is a useful tool to assess treatment effect with dose functions, especially when the treatment is endogenous. This paper presents the identification and estimation of an ACRF with sample selection and a high dimensional controls. We derive the Ney- man orthogonal moments with multiple nuisance parameters and utilize double machine learning method and typical nonparametric techniques to estimate the proposed estimators. Asymptotics for proposed estimators are derived and Monte Carlo simulations demonstrate their good finite sample properties. Our identification and estimation results could be readily extended to the case with more complex sample selection mechanisms. We apply the proposed method to US Job Corps data to evaluate the heterogeneous effect of residential components, which yields new insights for policy makers.
作為最重要的準(zhǔn)實(shí)驗(yàn)因果推斷方法之一，斷點(diǎn)回歸設(shè)計(jì)有兩大分析框架，二者無論在前提假定、帶寬選擇還是推斷方法上均有相當(dāng)差異。其中，基于連續(xù)性的框架假定潛在結(jié)果的條件期望連續(xù)，在實(shí)證研究中廣泛應(yīng)用。局部隨機(jī)化的框架則為后起之秀，該框架假定在斷點(diǎn)附近的小窗口，驅(qū)動(dòng)變量可視為隨機(jī)分配。本講座將介紹這兩大框架的原理與技術(shù)，包括識(shí)別、估計(jì)、推斷，并通過蒙特卡羅模擬與Stata案例比較二者的差異，以及應(yīng)用前景。
The interaction effect in endogenous probit model with an interaction term is consistently estimated in Zhou and Li (2021). However, the estimation and test are time-consuming when the sample size is large. In this article, a new Stata command, eivprobit, is developed to implement Zhou-Li’s method with much less time. Besides, the marginal effects of the two interacted regressors and the quadratic effect of a regressor with a squared term can also be estimated by the command. The eivprobit estimation is based on the control function approach and the standard errors of the estimated effects are obtained by nonparametric bootstrapping. Moreover, the finite sample Monto Carlo simulation shows that the estimator of the interaction effect behaves well and better than the usual methods such as Ai and Norton (2003)’s estimator ignoring endogeneity or the coefficient estimator of the interaction term in IV-probit estimation.
Mediation models with censored outcomes play a crucial role in social and medical sciences. However, the inherent censoring characteristics of the data often lead existing models to rely on assumptions of linearity, homogeneity, and normality for estimation. Unfortunately, these assumptions may not align with the complexities of real-world problems, limiting the persuasiveness of causal analyses. In this study, we investigate causal mediation analysis within a counterfactual framework by framing it as a neural style transfer problem commonly encountered in image processing. Acknowledging the impressive capabilities of generative adversarial networks (GANs) in handling neural style transfer, we propose a novel GAN-based model named generative adversarial censored mediation network to address mediation issues under our concern. Our model employs rectified linear unit (ReLU) activation function and designs a particular multi-channel network structure to implement the censored outcome mechanism while accommodating multiple mediators. To guide our model in accurately learning the underlying data patterns, we also develop a novel min-max optimization problem. Leveraging the strengths of GANs, our model fundamentally relaxes the stringent assumptions present in traditional models, resulting in more precise estimations of mediation effects and promising inference outcomes, especially in the context of intricate data patterns. Through unique insights and techniques, this study illustrates how generative learning methods can serve as an effective and robust approach for diverse causal mediation problems. We substantiate our claims with numerical results obtained from synthetic and realistic datasets, showcasing the superior performance of our method.
2024.8.20 詳細(xì)議程
會(huì)議主持：李寶偉-副教授 南開大學(xué)
2024.8.20 上午安排You can use treatment-effects estimators to draw causal inferences from observational data. You can use lasso when you want to control for many potential covariates. With standard treatment-effects models, there is an intrinsic conflict between two required assumptions. The conditional independence assumption is likely to be satisfied with many variables in the model, while the overlap assumption is likely to be satisfied with fewer variables in the model. This presentation shows how to overcome this conflict by using Stata's telasso command. telasso estimates the average treatment effects with high-dimensional controls while using lasso for model selection. This estimator is robust to model-selection mistakes. Moreover, it is doubly robust, so only one of the outcome or treatment model needs to be correctly specified.
過去三十多年間，項(xiàng)目評(píng)估計(jì)量經(jīng)濟(jì)學(xué)經(jīng)歷了長(zhǎng)足的發(fā)展，同時(shí)得益于計(jì)量分析軟件技術(shù)的不斷進(jìn)步和數(shù)據(jù)資料的日益豐富，經(jīng)濟(jì)學(xué)實(shí)證研究范式乃至整個(gè)經(jīng)濟(jì)學(xué)的研究范式發(fā)生了巨大轉(zhuǎn)變，深刻影響了經(jīng)濟(jì)學(xué)的教學(xué)和研究。過去五年間，圍繞DID, IV, RD等主流項(xiàng)目評(píng)估計(jì)量經(jīng)濟(jì)學(xué)方法，又出現(xiàn)了不少新的計(jì)量理論進(jìn)展，一方面對(duì)原有的理論方法和應(yīng)用實(shí)踐做了修補(bǔ)和完善，另一方面也推動(dòng)了項(xiàng)目評(píng)估計(jì)量經(jīng)濟(jì)學(xué)方法進(jìn)一步向前發(fā)展。我將對(duì)這些進(jìn)展做一個(gè)概括性的介紹。
同行效應(yīng)（或鄰居效應(yīng)）模型是研究個(gè)體之間相互影響的重要模型，其設(shè)定與空間計(jì)量模型相類似。但在空間計(jì)量模型中，鄰接矩陣往往被視作外生的。如果鄰接矩陣不是地理網(wǎng)絡(luò)而是社會(huì)或經(jīng)濟(jì)網(wǎng)絡(luò)，那么外生性假定是不合理的。本文提出了異質(zhì)性同行效應(yīng)模型和內(nèi)生性同行效應(yīng)模型的Stata估計(jì)指令，snreghnet和snregenet。snreghnet可以考察模型的行異質(zhì)性和列異質(zhì)性。snregenet計(jì)算模型的兩階段工具變量估計(jì)，并采用野蠻自舉計(jì)算標(biāo)準(zhǔn)誤差。
會(huì)議主持：顏冠鵬-青年講師 山東財(cái)經(jīng)大學(xué)
2024.8.20 下午安排Li et al. (2024) 擴(kuò)展了用于估計(jì)和推斷交互固定效應(yīng)面板模型處理效應(yīng)的因子化方法 (the factor-based approach)。本演講介紹了Stata新命令——xtteifeci，該命令可逐期生成處理效應(yīng)的置信區(qū)間和p值，且支持多種模型設(shè)定，包括模型中包含協(xié)變量和/或非平穩(wěn)趨勢(shì)等。最后，以經(jīng)典案例詳細(xì)介紹該命令的具體操作。
隨著信息技術(shù)的飛速發(fā)展和全球經(jīng)濟(jì)一體化的深入，數(shù)字經(jīng)濟(jì)已經(jīng)成為推動(dòng)全球經(jīng)濟(jì)增長(zhǎng)的重要引擎。然而，由于不同地區(qū)的資源稟賦、經(jīng)濟(jì)基礎(chǔ)、政策支持等因素存在差異，數(shù)字經(jīng)濟(jì)的發(fā)展水平在不同地區(qū)呈現(xiàn)出明顯的空間異質(zhì)性。因此，對(duì)數(shù)字經(jīng)濟(jì)發(fā)展水平的空間異質(zhì)性進(jìn)行深入分析，具有重要的必要性和現(xiàn)實(shí)意義。這不僅有助于我們?nèi)媪私鈹?shù)字經(jīng)濟(jì)在全國(guó)乃至全球范圍內(nèi)的分布狀況和發(fā)展趨勢(shì)，還有助于揭示數(shù)字經(jīng)濟(jì)與區(qū)域經(jīng)濟(jì)發(fā)展之間的互動(dòng)關(guān)系、識(shí)別數(shù)字經(jīng)濟(jì)在區(qū)域間的發(fā)展差距和潛在風(fēng)險(xiǎn)，以及推動(dòng)經(jīng)濟(jì)的可持續(xù)發(fā)展。
隨著時(shí)代的發(fā)展與技術(shù)的進(jìn)步，一般性的統(tǒng)計(jì)數(shù)據(jù)得到了廣泛的應(yīng)用。與此同時(shí)，以文本形式存在的非結(jié)構(gòu)化數(shù)據(jù)也正在逐漸成為經(jīng)管實(shí)證領(lǐng)域的中堅(jiān)力量。值得注意的是，廣大研究者在進(jìn)行文本分析時(shí)通常會(huì)優(yōu)先考慮使用Python等工具。但從Stata到Python的工具遷移往往伴隨不小的學(xué)習(xí)成本。在這種情況下我們不禁會(huì)想，是否可以使用Stata做文本分析內(nèi)容？本主題旨在介紹經(jīng)管領(lǐng)域主流的文本分析方法，并探討使用Stata進(jìn)行這些文本分析的可能性與局限性。
當(dāng)一個(gè)單位的處置也會(huì)影響其他單位的結(jié)果時(shí)的干擾情況下，傳統(tǒng)的因果推斷的SUTVA假定被違反，當(dāng)干擾起作用時(shí)，政策評(píng)估主要依賴于在集群干擾和二元處理下的隨機(jī)實(shí)驗(yàn)的假設(shè)。相反，我們考慮在連續(xù)治療和網(wǎng)絡(luò)干擾下的非實(shí)驗(yàn)處置。具體來說，我們通過將網(wǎng)絡(luò)處置的暴露程度定義為通過物理、社會(huì)或經(jīng)濟(jì)互動(dòng)連接的單位所接受的處理的加權(quán)平均值來定義溢出效應(yīng)。在Forastiere等人(2021)的基礎(chǔ)上，我們提供了一個(gè)基于廣義傾向得分的估計(jì)量來估計(jì)連續(xù)處置的直接和溢出效應(yīng)。我們的估計(jì)量還允許考慮以不同強(qiáng)度為特征的非對(duì)稱網(wǎng)絡(luò)連接。本演講介紹了一個(gè)Stata新命令，該命令結(jié)合Mathematica的優(yōu)點(diǎn)，采用線性回歸和機(jī)器學(xué)習(xí)的方法估計(jì)個(gè)體傾向得分和鄰域傾向得分，支持多種模型設(shè)定，采用廣義線性模型估計(jì)結(jié)果模型和劑量反應(yīng)函數(shù)（ADRF），并采用自舉法計(jì)算標(biāo)準(zhǔn)誤差。最后，以一個(gè)實(shí)際的案例詳細(xì)介紹該命令的具體操作。
Stata官方命令merge和joinby可以進(jìn)行1:1、m:1、1:m和m:m的數(shù)據(jù)橫向合并。但這兩個(gè)命令需要將使用數(shù)據(jù)集保存到硬盤中，不僅增加了時(shí)間成本并可能產(chǎn)生大量中間文件，影響效率。Stata的數(shù)據(jù)框（frame）功能允許用戶在內(nèi)存中同時(shí)操作多個(gè)數(shù)據(jù)集，無需保存數(shù)據(jù)到硬盤，但其frlink和frget命令僅支持1:1和m:1的合并類型，不支持1:m和m:m。通常的解決方法是將使用數(shù)據(jù)框的數(shù)據(jù)保存到硬盤，再借助merge和joinby命令將數(shù)據(jù)合并到主數(shù)據(jù)框中。顯然，當(dāng)需要合并的數(shù)據(jù)量較大或涉及的數(shù)據(jù)集較多時(shí)，這種方式的效率很低。本演講介紹一個(gè)新命令——framerge，以解決上述問題。framerge命令不僅支持多種數(shù)據(jù)合并關(guān)系（包括1:1, m:1, 1:m和m:m），還能夠在內(nèi)存中直接操作數(shù)據(jù)，無需讀寫硬盤，從而提高了數(shù)據(jù)處理的效率。最后以案例詳細(xì)介紹該命令的具體操作。傾向得分，支持多種模型設(shè)定，采用廣義線性模型估計(jì)結(jié)果模型和劑量反應(yīng)函數(shù)（ADRF），并采用自舉法計(jì)算標(biāo)準(zhǔn)誤差。最后，以一個(gè)實(shí)際的案例詳細(xì)介紹該命令的具體操作。
會(huì)議官網(wǎng)鏈接
