{"title":"采用双重/偏差机器学习的连续差分法","authors":"Lucas Zhang","doi":"arxiv-2408.10509","DOIUrl":null,"url":null,"abstract":"This paper extends difference-in-differences to settings involving continuous\ntreatments. Specifically, the average treatment effect on the treated (ATT) at\nany level of continuous treatment intensity is identified using a conditional\nparallel trends assumption. In this framework, estimating the ATTs requires\nfirst estimating infinite-dimensional nuisance parameters, especially the\nconditional density of the continuous treatment, which can introduce\nsignificant biases. To address this challenge, estimators for the causal\nparameters are proposed under the double/debiased machine learning framework.\nWe show that these estimators are asymptotically normal and provide consistent\nvariance estimators. To illustrate the effectiveness of our methods, we\nre-examine the study by Acemoglu and Finkelstein (2008), which assessed the\neffects of the 1983 Medicare Prospective Payment System (PPS) reform. By\nreinterpreting their research design using a difference-in-differences approach\nwith continuous treatment, we nonparametrically estimate the treatment effects\nof the 1983 PPS reform, thereby providing a more detailed understanding of its\nimpact.","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"23 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Continuous difference-in-differences with double/debiased machine learning\",\"authors\":\"Lucas Zhang\",\"doi\":\"arxiv-2408.10509\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This paper extends difference-in-differences to settings involving continuous\\ntreatments. Specifically, the average treatment effect on the treated (ATT) at\\nany level of continuous treatment intensity is identified using a conditional\\nparallel trends assumption. In this framework, estimating the ATTs requires\\nfirst estimating infinite-dimensional nuisance parameters, especially the\\nconditional density of the continuous treatment, which can introduce\\nsignificant biases. To address this challenge, estimators for the causal\\nparameters are proposed under the double/debiased machine learning framework.\\nWe show that these estimators are asymptotically normal and provide consistent\\nvariance estimators. To illustrate the effectiveness of our methods, we\\nre-examine the study by Acemoglu and Finkelstein (2008), which assessed the\\neffects of the 1983 Medicare Prospective Payment System (PPS) reform. By\\nreinterpreting their research design using a difference-in-differences approach\\nwith continuous treatment, we nonparametrically estimate the treatment effects\\nof the 1983 PPS reform, thereby providing a more detailed understanding of its\\nimpact.\",\"PeriodicalId\":501293,\"journal\":{\"name\":\"arXiv - ECON - Econometrics\",\"volume\":\"23 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-08-20\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - ECON - Econometrics\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2408.10509\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - ECON - Econometrics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2408.10509","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Continuous difference-in-differences with double/debiased machine learning
This paper extends difference-in-differences to settings involving continuous
treatments. Specifically, the average treatment effect on the treated (ATT) at
any level of continuous treatment intensity is identified using a conditional
parallel trends assumption. In this framework, estimating the ATTs requires
first estimating infinite-dimensional nuisance parameters, especially the
conditional density of the continuous treatment, which can introduce
significant biases. To address this challenge, estimators for the causal
parameters are proposed under the double/debiased machine learning framework.
We show that these estimators are asymptotically normal and provide consistent
variance estimators. To illustrate the effectiveness of our methods, we
re-examine the study by Acemoglu and Finkelstein (2008), which assessed the
effects of the 1983 Medicare Prospective Payment System (PPS) reform. By
reinterpreting their research design using a difference-in-differences approach
with continuous treatment, we nonparametrically estimate the treatment effects
of the 1983 PPS reform, thereby providing a more detailed understanding of its
impact.