[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-116839-en":3,"doc-seo-116839-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},116839,1099513958607,"Jiven","https://ap-avatar.wpscdn.com/avatar/100002390cf8733938c?x-image-process=image/resize,m_fixed,w_180,h_180&k=1778829742770036399",8,"Research & Report","ddml - Double/Debiased Machine Learning in Stata","The ddml package introduces Double/Debiased Machine Learning (DDML) in Stata for estimating causal parameters across five econometric models. It enables flexible identification of causal effects of endogenous variables when functional form and/or the set of exogenous variables is unknown. The approach leverages Neyman orthogonality and cross-fitting, supports integration with existing supervised machine learning tools in Stata, and recommends pairing DDML with stacking to combine multiple learners into a final predictor. Monte Carlo evidence supports the recommendation.","arXiv :2301 .09397v1 [ econ .EM] 23 Jan 2023  \n\n| The Stata Journal (yyyy) vv, Number ii\u003Cbr>ddml: Double/debiased machine learning\u003Cbr>in Stata |  |  |\n| --- | --- | --- |\n| Achim Ahrens ETH Z􀁿urich\u003Cbr>achim.ahrens@gess.ethz.ch | Christian B. Hansen University of Chicago [christian.hansen@chicagobooth.edu](christian.hansen@chicagobooth.edu) |  |\n| Mark E. Scha􀀋er Heriot-Watt University Edinburgh, United Kingdom m.e.scha􀀋er@hw.ac.uk |  | Thomas Wiemann University of Chicago [wiemann@uchicago.edu](wiemann@uchicago.edu) |\n| Abstract. We introduce the package ddml for Double/Debiased Machine Learning (DDML) in Stata. Estimators of causal parameters for 􀀌ve di􀀋erent econometric models are supported, allowing for 􀀍exible estimation of causal e􀀋ects of endogenous variables in settings with unknown functional forms and/or many exogenous variables. ddml is compatible with many existing supervised machine learning programs in Stata. We recommend using DDML in combination with stacking estimation which combines multiple machine learners into a 􀀌nal predictor. We provide Monte Carlo evidence to support our recommendation.\u003Cbr>Keywords: st0001, causal inference, machine learning, doubly-robust estimation\u003Cbr>1 Introduction\u003Cbr>Identi􀀌cation of causal e􀀋ects frequently relies on an unconfoundedness assumption, requiring that treatment or instrument assignment is su􀀎ciently random given observed control covariates. Estimation of causal e􀀋ects in these settings then involves conditioning on the controls. Unfortunately, estimators of causal e􀀋ects that are insu􀀎ciently 􀀍exible to capture the e􀀋ect of confounds generally do not produce consistent estimates of causal e􀀋ects even when unconfoundedness holds. For example, Blandhol et al. (2022) highlight that TSLS estimands obtained after controlling linearly for confounds do not generally correspond to weakly causal e􀀋ects even when instruments are valid conditional on controls. Even in the ideal scenario where theory provides a small number of relevant controls, theory rarely speci􀀌es the exact nature of confounding. Thus, applied empirical researchers wishing to exploit unconfoundedness assumptions to learn causal e􀀋ects face a nonparametric estimation problem.\u003Cbr>Traditional nonparametric estimators su􀀋er greatly under the curse of dimensionality and are quickly impractical in the frequently encountered setting with multiple observed covariates.1 These di􀀎culties leave traditional nonparametric estimators essentially inapplicable in the presence of increasingly large and complex data sets, e.g.\u003Cbr>\u003Cbr>1. For example, the number of coe􀀎cients in polynomial series regression with interaction terms increases exponentially in the number of covariates.\u003Cbr>© yyyy StataCorp LP st0001 |  |  |\n\n2 ddml  \ntextual confounders as in Roberts et al. (2020) or digital trace data (Hangartner et al. 2021) . Tools from supervised machine learning have been put forward as alternative estimators. These approaches are often more robust to the curse of dimensionality via the exploitation of regularization assumptions. A prominent example of a machine learning-based causal e􀀋ects estimator is Post-Double Selection Lasso (PDS-Lasso) of Belloni et al. (2014), which 􀀌ts auxiliary lasso regressions of the outcome and treatment(s), respectively, against a menu of transformed controls. Under an approximate sparsity assumption, which posits that the DGP can be approximated well by a relatively small number of terms included in the menu, this approach allows for precise treatment e􀀋ect estimation. The lasso can also be used for approximating optimal instruments (Belloni et al. 2012) . Lasso-based approaches for estimation of causal e􀀋ectshave become a popular strategy in applied econometrics (e.g. Gilchrist and Sands 2016; Dhar et al. 2022), partially facilitated by the availability of software programs in Stata (pdslasso, Ahrens et al. 2018; StataCorp 2019) and R (hdm, Chernozhukov et al. 2016) .  \nAlthough approximate sparsity","cbCaim3266GjyFKv","https://ap.wps.com/l/cbCaim3266GjyFKv","pdf",692546,1,50,"English","en",105,"# Abstract\n# Introduction\n## Causal effect identification and unconfoundedness\n## Limitations of traditional nonparametric methods\n## Machine learning approaches and regularization\n## Post-Double Selection Lasso and extensions\n## Double/Debiased Machine Learning (DDML)\n## Stacking as a practical learner-combination strategy","[{\"question\":\"What does the ddml package provide in Stata?\",\"answer\":\"It provides tools for implementing Double/Debiased Machine Learning (DDML) in Stata, supporting causal parameter estimation across multiple econometric models.\"},{\"question\":\"Why are unconfoundedness assumptions important for causal effect estimation?\",\"answer\":\"They require treatment or instrument assignment to be sufficiently random given observed covariates, which justifies conditioning on controls to learn causal effects.\"},{\"question\":\"How does DDML improve causal inference when the functional form is unknown?\",\"answer\":\"DDML uses Neyman orthogonality and cross-fitting to establish asymptotic normality under relatively mild convergence-rate conditions for the underlying nuisance estimators.\"}]","ddml - Double/Debiased Machine Learning in Stata | PDF",1785672019,126,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"ddml-doubledebiased-machine-learning-in-stata","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/ddml-doubledebiased-machine-learning-in-stata/116839/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-02",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What does the ddml package provide in Stata?","Question",{"text":75,"@type":76},"It provides tools for implementing Double/Debiased Machine Learning (DDML) in Stata, supporting causal parameter estimation across multiple econometric models.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why are unconfoundedness assumptions important for causal effect estimation?",{"text":80,"@type":76},"They require treatment or instrument assignment to be sufficiently random given observed covariates, which justifies conditioning on controls to learn causal effects.",{"name":82,"@type":73,"acceptedAnswer":83},"How does DDML improve causal inference when the functional form is unknown?",{"text":84,"@type":76},"DDML uses Neyman orthogonality and cross-fitting to establish asymptotic normality under relatively mild convergence-rate conditions for the underlying nuisance estimators.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,114,119,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":21,"slug":113},6,"Technology","technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]