Andrés Aradillas Fernández, José Luis Montiel Olea, Chen Qiu, Jörg Stoye, Serdil Tinda
{"title":"部分识别的稳健贝叶斯治疗选择","authors":"Andrés Aradillas Fernández, José Luis Montiel Olea, Chen Qiu, Jörg Stoye, Serdil Tinda","doi":"arxiv-2408.11621","DOIUrl":null,"url":null,"abstract":"We study a class of binary treatment choice problems with partial\nidentification, through the lens of robust (multiple prior) Bayesian analysis.\nWe use a convenient set of prior distributions to derive ex-ante and ex-post\nrobust Bayes decision rules, both for decision makers who can randomize and for\ndecision makers who cannot. Our main messages are as follows: First, ex-ante and ex-post robust Bayes\ndecision rules do not tend to agree in general, whether or not randomized rules\nare allowed. Second, randomized treatment assignment for some data realizations\ncan be optimal in both ex-ante and, perhaps more surprisingly, ex-post\nproblems. Therefore, it is usually with loss of generality to exclude\nrandomized rules from consideration, even when regret is evaluated ex-post. We apply our results to a stylized problem where a policy maker uses\nexperimental data to choose whether to implement a new policy in a population\nof interest, but is concerned about the external validity of the experiment at\nhand (Stoye, 2012); and to the aggregation of data generated by multiple\nrandomized control trials in different sites to make a policy choice in a\npopulation for which no experimental data are available (Manski, 2020; Ishihara\nand Kitagawa, 2021).","PeriodicalId":501293,"journal":{"name":"arXiv - ECON - Econometrics","volume":"1 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Robust Bayes Treatment Choice with Partial Identification\",\"authors\":\"Andrés Aradillas Fernández, José Luis Montiel Olea, Chen Qiu, Jörg Stoye, Serdil Tinda\",\"doi\":\"arxiv-2408.11621\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"We study a class of binary treatment choice problems with partial\\nidentification, through the lens of robust (multiple prior) Bayesian analysis.\\nWe use a convenient set of prior distributions to derive ex-ante and ex-post\\nrobust Bayes decision rules, both for decision makers who can randomize and for\\ndecision makers who cannot. Our main messages are as follows: First, ex-ante and ex-post robust Bayes\\ndecision rules do not tend to agree in general, whether or not randomized rules\\nare allowed. Second, randomized treatment assignment for some data realizations\\ncan be optimal in both ex-ante and, perhaps more surprisingly, ex-post\\nproblems. Therefore, it is usually with loss of generality to exclude\\nrandomized rules from consideration, even when regret is evaluated ex-post. We apply our results to a stylized problem where a policy maker uses\\nexperimental data to choose whether to implement a new policy in a population\\nof interest, but is concerned about the external validity of the experiment at\\nhand (Stoye, 2012); and to the aggregation of data generated by multiple\\nrandomized control trials in different sites to make a policy choice in a\\npopulation for which no experimental data are available (Manski, 2020; Ishihara\\nand Kitagawa, 2021).\",\"PeriodicalId\":501293,\"journal\":{\"name\":\"arXiv - ECON - Econometrics\",\"volume\":\"1 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-08-21\",\"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-2408.11621\",\"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-2408.11621","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Robust Bayes Treatment Choice with Partial Identification
We study a class of binary treatment choice problems with partial
identification, through the lens of robust (multiple prior) Bayesian analysis.
We use a convenient set of prior distributions to derive ex-ante and ex-post
robust Bayes decision rules, both for decision makers who can randomize and for
decision makers who cannot. Our main messages are as follows: First, ex-ante and ex-post robust Bayes
decision rules do not tend to agree in general, whether or not randomized rules
are allowed. Second, randomized treatment assignment for some data realizations
can be optimal in both ex-ante and, perhaps more surprisingly, ex-post
problems. Therefore, it is usually with loss of generality to exclude
randomized rules from consideration, even when regret is evaluated ex-post. We apply our results to a stylized problem where a policy maker uses
experimental data to choose whether to implement a new policy in a population
of interest, but is concerned about the external validity of the experiment at
hand (Stoye, 2012); and to the aggregation of data generated by multiple
randomized control trials in different sites to make a policy choice in a
population for which no experimental data are available (Manski, 2020; Ishihara
and Kitagawa, 2021).