# sdid-stata **Repository Path**: aieconlab/sdid-stata ## Basic Information - **Project Name**: sdid-stata - **Description**: Synthetic Difference in Differences for Stata - **Primary Language**: Unknown - **License**: GPL-3.0 - **Default Branch**: main - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 0 - **Forks**: 3 - **Created**: 2026-02-10 - **Last Updated**: 2026-02-10 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README # sdid -- Synthetic Difference-in-Differences for Stata This Stata package implements the synthetic difference-in-differences estimation procedure, along with a range of inference and graphing procedures, following Arkhangelsky et al., (2021). Arkhangelsky et al. provide a code implementation in R, with accompanying materials here: [synthdid](https://synth-inference.github.io/synthdid/). Here we provide a native Stata implementation, principally written in Mata. This package extends the funcionality of the original R package, allowing very simply for estimation in contexts with staggered adoption over multiple treatment periods (as well as in a single adoption period as in the original code). To install directly into Stata: ```s ssc install sdid ``` ## Inputs + Y: Outcome variable (numeric) + S: Unit variable (numeric or string) + T: Time variable (numeric) + D: Dummy of treatement, equal to 1 if units are treated, and otherwise 0 (numeric) ## Syntax ```s sdid Y S T D, vce(method) seed(#) reps(#) covariates(varlist [, method]) graph g1_opt(string) g2_opt(string) unstandardized graph_export([stub] , type) ``` + vce(): bootstrap, jackknife and placebo standard error. + seed(): seed define for pseudo-random numbers. + reps(): repetitions for bootstrap and placebo se. + covariates( varlist [, method]): covariates included to adjust Y. A varlist of covariates should be included, and optionally an option for the method used to adjust. This can be "optimized" in which case it follows the method proposed by Arkhangelsky et al., or "projected", in which case it follows the procedure proposed by Kranz, 2021 (xsynth in R). Where method is not specified, optimized is used as default. Kranz has shown that the projected method is preferable in a number of circumstances. In this implementation, the projected method is often considerably faster. + graph: If this option is specified, graphs will be displayed showing unit and time weights as well as outcome trends as per figure 1 from Arkhangelsky et al. + g1_opt() and g2_opt(): option to modify the appearance of the aforementioned graphs. g1 refers to the unit-specific weight graph, while g2 refers to the outcome trend groaphs. Options requested have to follow the syntax of Stata's twoway_options. + unstandardized: If controls are included and the "optimized" method is specified, controls will be standardized as Z-scores prior to finding optimal weights. This avoids problems with optimization when control variables have very high dispersion. If unstandardized is specified, controls will simply be entered in their original units. This option should be used with care. + graph_export( [stub], type): Generated graphs can be saved to the disk using this option. Graphs will be saved as weightsYYYY and trendsYYYY for each of the unit-specific weights and outcome trends respectively, where YYYY refers to each treatment adoption period. Two graphs will be generated for each treatment adoption period. If this option is specified, type must be specified, which refers to a valid Stata graph type (eg ".eps", ".pdf", and so forth). Optionally, a stub can be specified, in which case this will be prepended to exported graph names. ### References Dmitry Arkhangelsky, Susan Athey, David A. Hirshberg, Guido W. Imbens, and Stefan Wager. Synthetic Difference in Differences, American Economic Review, December 2021.   ![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/Lianxh_装饰黄线.png) > [主页](https://www.lianxh.cn/news/46917f1076104.html) || [视频](http://lianxh.duanshu.com) || [推文](https://www.lianxh.cn/news/d4d5cd7220bc7.html) || [知乎](https://www.zhihu.com/people/arlionn/) || [Bilibili 站](https://space.bilibili.com/546535876) --- ## 1. 计量专题   > [最新上线](https://gitee.com/arlionn/collections/278581) | [公开课](https://gitee.com/arlionn/collections/278579) | [小专题](https://gitee.com/arlionn/collections/1257)   > 推文集合：[PDF 合集](https://file.lianxh.cn/KC/lianxh_Blogs.pdf) - Stata：[教程](https://www.lianxh.cn/blogs/17.html) | [资料](https://www.lianxh.cn/blogs/35.html) | [新命令](https://www.lianxh.cn/blogs/43.html) | [结果输出](https://www.lianxh.cn/blogs/22.html) | [绘图](https://www.lianxh.cn/blogs/24.html) | [数据处理](https://www.lianxh.cn/blogs/25.html) | [程序](https://www.lianxh.cn/blogs/26.html) - [回归分析](https://www.lianxh.cn/blogs/32.html) | [面板数据](https://www.lianxh.cn/blogs/20.html) | [交乘项-调节](https://www.lianxh.cn/blogs/21.html) | [IV-GMM](https://www.lianxh.cn/blogs/38.html) | [Logit](https://www.lianxh.cn/blogs/27.html) | [空间计量](https://www.lianxh.cn/blogs/29.html) - [因果推断](https://www.lianxh.cn/blogs/19.html) | [DID](https://www.lianxh.cn/blogs/39.html) | [RDD](https://www.lianxh.cn/blogs/40.html) | [PSM](https://www.lianxh.cn/blogs/41.html) | [合成控制](https://www.lianxh.cn/blogs/42.html) | [文本分析](https://www.lianxh.cn/blogs/36.html) | [机器学习](https://www.lianxh.cn/blogs/47.html) - [Markdown](https://www.lianxh.cn/blogs/30.html) | [工具软件](https://www.lianxh.cn/blogs/23.html) | [其它](https://www.lianxh.cn/blogs/33.html) > 热门推文 - [稳健性检验！稳健性检验！](https://www.lianxh.cn/news/32ae13ec789a1.html) - [安慰剂检验！安慰剂检验！](https://www.lianxh.cn/news/cf8bd8363200d.html) - [Stata: 面板数据模型一文读懂](https://www.lianxh.cn/news/bf27906144b4e.html) - [Stata：聚类调整后的标准误-Cluster-SE](https://www.lianxh.cn/news/a7a8e613b2699.html) - [Stata：DID入门教程](https://www.lianxh.cn/news/3849f237b6d36.html) - [DID偏误问题：两时期DID的双重稳健估计量(上)-drdid](https://www.lianxh.cn/news/e6ef033e13c3e.html) - [DID偏误问题：多时期DID的双重稳健估计量(下)-csdid](https://www.lianxh.cn/news/762e878e7063b.html) - [Stata+R：一文读懂精确断点回归-RDD](https://www.lianxh.cn/news/96fb6b7e847e1.html) - [内生性！内生性！解决方法大集合](https://www.lianxh.cn/news/224e2b4e170e4.html) - [连享会：论文重现网站大全](https://www.lianxh.cn/news/e87e5976686d5.html)   > ### ${\color{blue}{\blacktriangleright}}$ [课程主页](https://www.lianxh.cn/news/46917f1076104.html)   [![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/MayunKC-001.png)](https://www.lianxh.cn/news/46917f1076104.html) [![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/MayunKC-002.png)](https://www.lianxh.cn/news/46917f1076104.html) [![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/MayunKC-003.png)](https://www.lianxh.cn/news/46917f1076104.html) [![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/MayunKC-004.png)](https://www.lianxh.cn/news/46917f1076104.html) [![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/MayunKC-005.png)](https://www.lianxh.cn/news/46917f1076104.html) [![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/MayunKC-006.png)](https://www.lianxh.cn/news/46917f1076104.html) > ### ${\color{red}{\blacktriangleright}}$ [计量专题](https://gitee.com/arlionn/Course) ![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/Lianxh_装饰黄线.png)   ## 2. 公开课/专题课 ### 公开课 - 侯丹丹，Stata软件及计量基础，6 次课 x 2h，[去听课](https://lianxh.duanshu.com/#/brief/course/49c3dbf11982456db9be1b5f595285dc)，[课件](https://gitee.com/arlionn/stataopen) - 连玉君，直击面板数据模型，1.4h，[去听课](http://lianxh-pc.duanshu.com/course/detail/7d1d3266e07d424dbeb3926170835b38)，[B站视频](https://www.bilibili.com/video/BV1oU4y187qY)，[课件](https://gitee.com/arlionn/PanelData) - 连玉君，Stata 33 讲，15min/讲. [去听课](http://lianxh-pc.duanshu.com/course/detail/b22b17ee02c24015ae759478697df2a0)，[B站视频](https://space.bilibili.com/546535876/channel/detail?cid=160748)，[课件](https://gitee.com/arlionn/stata101) - 连玉君，论文复现：why？what？How？，2h，[去听课](https://www.lianxh.cn/news/3544e941a02d8.html)，[B站视频](https://www.bilibili.com/video/BV1D54y1n7Fa)，[课件](https://gitee.com/arlionn/rep) - 龙志能，Stata小白的取经之路，1.5h，[去听课](https://lianxh.duanshu.com/#/brief/course/137d1b7c7c0045e682d3cf0cb2711530)，[课件](https://gitee.com/arlionn/StataBin) - 涂冰倩，微观数据库清理经验分享，2h，[去听课](https://gitee.com/arlionn/dataclean/blob/master/%E5%85%AC%E5%BC%80%E8%AF%BE%EF%BC%9A%E6%B6%82%E5%86%B0%E5%80%A9-%E5%BE%AE%E8%A7%82%E6%95%B0%E6%8D%AE%E5%BA%93%E6%B8%85%E7%90%86%E7%BB%8F%E9%AA%8C%E5%88%86%E4%BA%AB.md#https://gitee.com/link?target=https%3A%2F%2Fwww.lianxh.cn%2Fblogs%2F16.html)，[课件](https://gitee.com/arlionn/dataclean) - 游万海，实证分析中的数据可视化，2h，[去听课](https://lianxh.duanshu.com/#/brief/course/da1a75bc3acc4e238f489af3367efa26)，[课件](https://gitee.com/arlionn/Rplot) - 杨海生，如何玩转空间计量？，1.5h，[去听课](https://lianxh.duanshu.com/#/brief/course/675437c0c52f4e48947531286cef1e87)，[课件](https://gitee.com/arlionn/SP/blob/master/%E5%85%AC%E5%BC%80%E8%AF%BE-%E6%9D%A8%E6%B5%B7%E7%94%9F-%E7%8E%A9%E8%BD%AC%E7%A9%BA%E9%97%B4%E8%AE%A1%E9%87%8F.md) - 范巧，空间计量经济学的基本框架及分析范式，[去听课](https://lianxh.duanshu.com/#/brief/course/6b294c08454042ce860e43363bbeed5b)，[课件](https://gitee.com/arlionn/SP/blob/master/%E5%85%AC%E5%BC%80%E8%AF%BE-%E8%8C%83%E5%B7%A7-%E7%A9%BA%E9%97%B4%E8%AE%A1%E9%87%8F%E7%BB%8F%E6%B5%8E%E5%AD%A6%E5%9F%BA%E6%9C%AC%E6%A1%86%E6%9E%B6%7C%E5%88%86%E6%9E%90%E8%8C%83%E5%BC%8F.md) - 范巧，空间计量建模与高质量论文撰写，[去听课](https://lianxh.duanshu.com/#/brief/course/35252658d4a046a0b0b74ceea1f8e0ee)，[课件](https://gitee.com/arlionn/SP/blob/master/%E5%85%AC%E5%BC%80%E8%AF%BE-%E8%8C%83%E5%B7%A7-%E7%A9%BA%E9%97%B4%E8%AE%A1%E9%87%8F%E7%BB%8F%E6%B5%8E%E5%AD%A6%E5%9F%BA%E6%9C%AC%E6%A1%86%E6%9E%B6%7C%E5%88%86%E6%9E%90%E8%8C%83%E5%BC%8F.md) - 龙志能，如何高效运用搜索解决问题？，1h，[去听课](https://lianxh.duanshu.com/#/brief/course/ac14b768d2314d43a8a805205a45d3e3) - 王昆仑，双重差分方法的新进展-交错型DID，2h，[去听课](https://lianxh.duanshu.com/#/brief/course/9f83a4513f7f409b96e40a1bd0c79379) - [连享会码云：100多个精选计量项目](https://www.lianxh.cn/news/944a69d75cec9.html) | [新浪视频](https://weibo.com/tv/show/1034:4479228373303338) - [五分钟 Markdown / Markdown 幻灯片](https://gitee.com/arlionn/md) | [新浪视频](https://weibo.com/tv/show/1034:4484204327796746) - [连老师给你的-听课建议](https://www.lianxh.cn/news/69706e871c9ad.html) ### 专题小课 - 连玉君，我的特斯拉-实证研究设计，2.2h，[去听课](https://lianxh.duanshu.com/#/course/5ae82756cc1b478c872a63cbca4f0a5e)，[课件](https://gitee.com/arlionn/Live/tree/master/%E6%88%91%E7%9A%84%E7%89%B9%E6%96%AF%E6%8B%89-%E5%AE%9E%E8%AF%81%E7%A0%94%E7%A9%B6%E8%AE%BE%E8%AE%A1-%E8%BF%9E%E7%8E%89%E5%90%9B) - 连玉君，我的甲壳虫-论文精讲与重现，6h，[去听课](https://lianxh.duanshu.com/#/brief/course/c3f79a0395a84d2f868d3502c348eafc)，[课件](https://gitee.com/arlionn/paper101) - 连玉君，动态面板数据模型，2.2h，[去听课](https://efves.duanshu.com/#/brief/course/3c3ac06108594577a6e3112323d93f3e)，[课件](https://gitee.com/arlionn/Live/tree/master/%E8%BF%9E%E7%8E%89%E5%90%9B-%E5%8A%A8%E6%80%81%E9%9D%A2%E6%9D%BF%E6%A8%A1%E5%9E%8B) - 游万海，Stata 数据清洗之实战操作 (第一季)，2h，[去听课](https://lianxh.duanshu.com/#/brief/course/c5193f0e6e414a7e889a8ff9aeb4aaef)，[主页](https://gitee.com/arlionn/dataclean) - 游万海，Stata 数据清洗之实战操作 (第二季)，2h，[去听课](https://lianxh.duanshu.com/#/brief/course/23924488072b4e458ec3bb0a830b187f)，[主页](https://gitee.com/arlionn/dataclean) - 游万海，分位数回归，2h，[去听课](https://lianxh.duanshu.com/#/brief/course/f0bfb3102ada48969966c92123a7ebf0)   ## 3. 资源分享 连享会主要通过三个渠道分享实证分析经验。 > **主页**：[www.lianxh.cn](https://www.lianxh.cn) - 目前已分享 1000 余篇推文，包括 Stata/R/Python 等软件使用经验，[各类](https://www.lianxh.cn/blogs.html) 计量模型，内生性、DID、RDD 专题等。 - **人手一份：Stata 101**，[PDF下载](https://file.lianxh.cn/KC/Slides/lianxh_Stata101.pdf), [HTML下载](https://file.lianxh.cn/KC/Slides/lianxh_Stata101.html) | [Stata Journal](https://www.lianxh.cn/news/12ffe67d8d8fb.html) > **b 站**： - 亦可在 b 站搜索关键词 **连享会** - 内容：浏览量超过 10 w 的公开课视频，包括 [Stata 33 讲](https://space.bilibili.com/546535876/channel/detail?cid=160748)、[直击面板数据模型](https://www.bilibili.com/video/BV1oU4y187qY)、[论文复现：why？what？How？](https://www.bilibili.com/video/BV1D54y1n7Fa) 等。 > **码云仓库**： - 连享会收集整理的代码仓库，分为 30 多个 [专题](https://gitee.com/arlionn/collections)，500+ 个 [仓库](https://gitee.com/arlionn/projects)，涉及论文复现代码、RDD，DID，Panel，SCM，Lasso 等最新方法实现代码等。 - 用法：申请一个码云账号，然后点击仓库右上角的 **[Fork](https://gitee.com/arlionn/list#)** 按钮，将喜欢的仓库克隆到你的账号下，随后点击仓库名称右边的 $\curvearrowright$ 按钮强制同步。 ### Data - [CSMAR-国泰安](http://www.gtarsc.com/#/datacenter/singletable) | [Wind-万德](https://www.wind.com.cn/Default.html) | [Resset-锐思](http://www.resset.cn/databases) - [常用数据库](https://www.lianxh.cn/news/0b65fd5165c2c.html) | [人文社科开放数据库](https://www.lianxh.cn/news/6f06c914acde8.html) | [知乎-Data](https://www.zhihu.com/question/20179699/answer/681756635) - [徐现祥教授-IRE-官员交流、方言等](https://www.lianxh.cn/news/8c9f81a5f19ee.html) ### Papers - 学术论文复现 - [论文重现网站大全](https://www.lianxh.cn/news/e87e5976686d5.html) - Stata论文重现：[Harvard][harvd] | [Top 期刊论文](https://ejd.econ.mathematik.uni-ulm.de/) | [JFE][jfe] | [github][git1] - 学者主页：[Angrist][Ang1] | [Acemoglu][acem] | [Levine][ross] | [Duflo][Duflo] | [Imbens](https://scholar.harvard.edu/imbens/software) | [Chetty](http://www.rajchetty.com/) [harvd]:https://dataverse.harvard.edu/dataverse [jfe]:http://jfe.rochester.edu/data.htm [Ang1]:http://economics.mit.edu/faculty/angrist/data1/data [acem]:http://economics.mit.edu/faculty/acemoglu/data [ross]:http://faculty.haas.berkeley.edu/ross_levine/papers.htm [duflo]:http://economics.mit.edu/faculty/eduflo/papers [git1]:https://github.com/search?utf8=%E2%9C%93&q=stata&type= [yahoogit]:https://search.yahoo.com/search;_ylt=AwrBT8di2LBZqyEAuG9XNyoA;_ylc=X1MDMjc2NjY3OQRfcgMyBGZyA3lmcC10LTQ3MwRncHJpZAMEbl9yc2x0AzAEbl9zdWdnAzAEb3JpZ2luA3NlYXJjaC55YWhvby5jb20EcG9zAzAEcHFzdHIDBHBxc3RybAMwBHFzdHJsAzE0BHF1ZXJ5A3N0YXRhJTIwZ2l0aHViBHRfc3RtcAMxNTA0NzYxODcz?p=stata+github&fr2=sb-top&fr=yfp-t-473&fp=1   ## 4. 关于我们 - **Stata连享会** 由中山大学连玉君老师团队创办，定期分享实证分析经验。[直播间](http://lianxh.duanshu.com) 有很多视频课程，可以随时观看。 - [连享会-主页](https://www.lianxh.cn) 和 [知乎专栏](https://www.zhihu.com/people/arlionn/)，500+ 推文，实证分析不再抓狂；[Bilibili 站](https://space.bilibili.com/546535876) 有视频大餐。 >New！ **`lianxh` 命令发布了！** - 随时搜索连享会推文、Stata 资源，安装命令如下：   `. ssc install lianxh` - 使用详情参见帮助文件 (有惊喜)：   `. help lianxh` > 连享会小程序：扫一扫，看推文，看视频…… [![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/连享会小程序二维码180.png)](https://www.lianxh.cn/news/46917f1076104.html) > 扫码加入连享会微信群，提问交流更方便 ![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/连享会-学习交流微信群001-150.jpg) ![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/Lianxh_装饰黄线.png) ![](https://fig-lianxh.oss-cn-shenzhen.aliyuncs.com/连享会-草料主页-一码平川600.png)