{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,5]],"date-time":"2026-06-05T09:12:20Z","timestamp":1780650740449,"version":"3.54.1"},"reference-count":52,"publisher":"Institute for Operations Research and the Management Sciences (INFORMS)","issue":"6","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Management Science"],"published-print":{"date-parts":[[2026,6]]},"abstract":"<jats:p>Switchback experiments, in which experimental units are assigned to control and treatment interchangeably over time, have gained increasing popularity in recent years due to their ability to mitigate interference across units. Despite their wide adoption in the technology sector, drawing reliable inferences from such experiments remains challenging due to temporal interference. In this paper, we address this challenge by examining a special causal structure commonly found in a wide range of applications. Specifically, we consider scenarios where there exist surrogate variables at each experimental period that can fully capture the potential temporal interference from the past. We outline the necessary assumptions for identification in this context and propose an unbiased estimator using the technique of importance sampling. Additionally, we derive an experimental design with a near-optimal worst-case guarantee and compare it theoretically and empirically to the inverse propensity score estimator. To facilitate inferences, we introduce both an exact inference procedure based on Fisher\u2019s randomization test and an asymptotic inference procedure based on the central limit theorem. We demonstrate by extensive numerical experiments that the proposed estimator, coupled with the experimental design, results in lower risk compared with the traditional inverse propensity score estimator.<\/jats:p>\n                  <jats:p>This paper was accepted by J. George Shanthikumar, data science.<\/jats:p>\n                  <jats:p>Funding: Research was funded by the MIT Data Science Lab.<\/jats:p>\n                  <jats:p>Supplemental Material: The online appendix and data files are available at https:\/\/doi.org\/10.1287\/mnsc.2023.03818 .<\/jats:p>","DOI":"10.1287\/mnsc.2023.03818","type":"journal-article","created":{"date-parts":[[2025,9,22]],"date-time":"2025-09-22T16:28:00Z","timestamp":1758558480000},"page":"4854-4870","source":"Crossref","is-referenced-by-count":0,"title":["Efficient Switchback Experiments with Surrogate Variables: Estimation and Experimental Design"],"prefix":"10.1287","volume":"72","author":[{"ORCID":"https:\/\/orcid.org\/0009-0008-4227-3572","authenticated-orcid":false,"given":"Hongyu","family":"Chen","sequence":"first","affiliation":[{"name":"Institute for Data, Systems, and Society, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139; and Laboratory for Information & Decision Systems, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4650-1519","authenticated-orcid":false,"given":"David","family":"Simchi-Levi","sequence":"additional","affiliation":[{"name":"Institute for Data, Systems, and Society, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139; and Laboratory for Information & Decision Systems, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139; and Department of Civil and Environmental Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139; and Operations Research Center, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"109","reference":[{"key":"B1","unstructured":"Abadie A, Zhao J (2021) Synthetic controls for experimental design. Preprint, submitted August 4, https:\/\/arxiv.org\/abs\/2108.02196."},{"key":"B2","doi-asserted-by":"publisher","DOI":"10.1287\/mnsc.2021.4096"},{"issue":"2","key":"B3","first-page":"119","volume":"31","author":"Armitage P","year":"1982","journal-title":"J. Roy. Statist. Soc. Ser. D (Statistician)"},{"key":"B4","doi-asserted-by":"crossref","unstructured":"Aronow PM, Samii C (2017) Estimating average causal effects under general interference, with application to a social network experiment.\n                      Ann. Appl. Statist.\n                      11(4):1912\u20131947.","DOI":"10.1214\/16-AOAS1005"},{"key":"B5","doi-asserted-by":"publisher","DOI":"10.1080\/01621459.2016.1241178"},{"key":"B6","doi-asserted-by":"crossref","unstructured":"Athey S, Chetty R, Imbens GW, Kang H (2019) The surrogate index: Combining short-term proxies to estimate long-term treatment effects more rapidly and precisely. NBER Working Paper No. 26463, National Bureau of Economic Research, Cambridge, MA.","DOI":"10.3386\/w26463"},{"key":"B7","doi-asserted-by":"crossref","unstructured":"Bakshy E, Eckles D, Bernstein MS (2014) Designing and deploying online field experiments. Chung C-W, ed.\n                      Proc. 23rd Internat. Conf. World Wide Web\n                      (Association for Computing Machinery, New York), 283\u2013292.","DOI":"10.1145\/2566486.2567967"},{"key":"B8","unstructured":"Basse G, Ding Y, Toulis P (2019) Minimax crossover designs. Preprint, submitted August 9, https:\/\/arxiv.org\/abs\/1908.03531."},{"key":"B9","doi-asserted-by":"publisher","DOI":"10.1111\/1467-985X.00153"},{"key":"B10","doi-asserted-by":"publisher","DOI":"10.1080\/01621459.2018.1527225"},{"key":"B11","doi-asserted-by":"publisher","DOI":"10.1287\/mnsc.2022.4583"},{"key":"B12","doi-asserted-by":"publisher","DOI":"10.1080\/00401706.1971.10488845"},{"key":"B13","unstructured":"Chamandy N (2016) Experimentation in a ridesharing marketplace. Accessed September 4, 2025, https:\/\/eng.lyft.com\/experimentation-in-a-ridesharing-marketplace-b39db027a66e."},{"key":"B14","doi-asserted-by":"crossref","unstructured":"Farias VF, Li AA, Peng T, Zheng AT (2022) Markovian interference in experiments. Preprint, submitted June 6, https:\/\/arxiv.org\/abs\/2206.02371.","DOI":"10.52202\/068431-0039"},{"key":"B15","doi-asserted-by":"publisher","DOI":"10.1287\/msom.2015.0561"},{"key":"B16","doi-asserted-by":"publisher","DOI":"10.1136\/bmj.1.3923.554-a"},{"key":"B17","doi-asserted-by":"publisher","DOI":"10.1111\/j.0006-341X.2002.00021.x"},{"key":"B18","doi-asserted-by":"publisher","DOI":"10.1287\/mnsc.2021.4058"},{"key":"B19","first-page":"15054","volume":"33","author":"Glynn PW","year":"2020","journal-title":"Adv. Neural Inform. Processing Systems"},{"key":"B20","doi-asserted-by":"publisher","DOI":"10.2307\/2998560"},{"key":"B21","first-page":"619","author":"Hedayat A","year":"1978","journal-title":"Ann. Statist."},{"key":"B22","doi-asserted-by":"publisher","DOI":"10.1093\/biomet\/91.4.929"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.1017\/S0266466608080699"},{"key":"B24","doi-asserted-by":"publisher","DOI":"10.1080\/01621459.1986.10478354"},{"key":"B25","doi-asserted-by":"publisher","DOI":"10.1080\/01621459.1952.10483446"},{"key":"B26","doi-asserted-by":"publisher","DOI":"10.1111\/rssb.12017"},{"key":"B27","unstructured":"Hu Y, Wager S (2022) Switchback experiments under geometric mixing. Preprint, submitted September 1, https:\/\/arxiv.org\/abs\/2209.00197."},{"key":"B28","doi-asserted-by":"publisher","DOI":"10.1080\/10618600.1996.10474715"},{"key":"B29","doi-asserted-by":"publisher","DOI":"10.1017\/CBO9781139025751"},{"key":"B30","doi-asserted-by":"publisher","DOI":"10.1287\/mnsc.2021.4247"},{"key":"B31","doi-asserted-by":"crossref","unstructured":"Kenward MG, Jones B (2014) Chapter 27: Crossover trials.\n                      Methods and Applications of Statistics in Clinical Trials: Concepts, Principles, Trials, and Design\n                      (John Wiley & Sons Inc., Hobooken, NJ).","DOI":"10.1002\/9781118445112.stat06798"},{"issue":"5","key":"B32","first-page":"74","volume":"95","author":"Kohavi R","year":"2017","journal-title":"Harvard Bus. Rev."},{"key":"B33","doi-asserted-by":"publisher","DOI":"10.1017\/9781108653985"},{"issue":"2009","key":"B34","first-page":"39","volume":"11","author":"Kohavi R","year":"2009","journal-title":"Data Mining Case Stud."},{"key":"B35","doi-asserted-by":"publisher","DOI":"10.1002\/sim.4780111415"},{"key":"B36","doi-asserted-by":"publisher","DOI":"10.1080\/01621459.2017.1295865"},{"key":"B37","doi-asserted-by":"crossref","unstructured":"Li H, Zhao G, Johari R, Weintraub GY (2022) Interference, bias, and variance in two-sided marketplace experimentation: Guidance for platforms.\n                      Proc. ACM Web Conf.\n                      (Association for Computing Machinery, New York), 182\u2013192.","DOI":"10.1145\/3485447.3512063"},{"key":"B38","doi-asserted-by":"crossref","unstructured":"Ni T, Bojinov I, Zhao J (2023) Design of panel experiments with spatial and temporal interference. Preprint, submitted June 4, https:\/\/dx.doi.org\/10.2139\/ssrn.4466598.","DOI":"10.2139\/ssrn.4466598"},{"key":"B39","doi-asserted-by":"publisher","DOI":"10.1093\/biomet\/75.1.81"},{"key":"B40","doi-asserted-by":"publisher","DOI":"10.1002\/sim.4780080407"},{"key":"B41","unstructured":"Puelz D, Basse G, Feller A, Toulis P (2019) A graph-theoretic approach to randomization tests of causal effects under general interference. Preprint, submitted October 24, https:\/\/arxiv.org\/abs\/1910.10862."},{"key":"B42","volume-title":"Cross-over Experiments: Design, Analysis and Application","volume":"135","author":"Ratkowsky D","year":"1992"},{"key":"B43","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-9876.2005.00510.x"},{"key":"B44","doi-asserted-by":"publisher","DOI":"10.1037\/h0037350"},{"issue":"371","key":"B45","first-page":"591","volume":"75","author":"Rubin DB","year":"1980","journal-title":"J. Amer. Statist. Assoc."},{"key":"B46","doi-asserted-by":"publisher","DOI":"10.1002\/(SICI)1097-0258(19981230)17:24<2849::AID-SIM955>3.0.CO;2-O"},{"key":"B47","unstructured":"Sussman DL, Airoldi EM (2017) Elements of estimation theory for causal effects in the presence of network interference. Preprint, submitted February 12, https:\/\/arxiv.org\/abs\/1702.03578."},{"key":"B48","volume-title":"Experimentation Works: The Surprising Power of Business Experiments","author":"Thomke SH","year":"2020"},{"key":"B49","doi-asserted-by":"crossref","unstructured":"Ugander J, Karrer B, Backstrom L, Kleinberg J (2013) Graph cluster randomization: Network exposure to multiple universes.\n                      Proc. 19th ACM SIGKDD Internat. Conf. Knowledge Discovery Data Mining\n                      (Association for Computing Machinery, New York), 329\u2013337.","DOI":"10.1145\/2487575.2487695"},{"key":"B50","doi-asserted-by":"publisher","DOI":"10.1287\/mnsc.2020.3844"},{"key":"B51","unstructured":"Xiong R, Chin A, Taylor S (2023) Bias-variance tradeoffs for designing simultaneous temporal experiments.\n                      Proc. KDD Workshop Causal Discovery Prediction Decision\n                      (PMLR, New York), 115\u2013131."},{"key":"B52","doi-asserted-by":"crossref","unstructured":"Xiong R, Athey S, Bayati M, Imbens G (2019) Optimal experimental design for staggered rollouts. Preprint, submitted November 9, https:\/\/arxiv.org\/abs\/1911.03764.","DOI":"10.2139\/ssrn.3483934"}],"container-title":["Management Science"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pubsonline.informs.org\/doi\/pdf\/10.1287\/mnsc.2023.03818","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,5]],"date-time":"2026-06-05T08:57:49Z","timestamp":1780649869000},"score":1,"resource":{"primary":{"URL":"https:\/\/pubsonline.informs.org\/doi\/10.1287\/mnsc.2023.03818"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6]]},"references-count":52,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2026,6]]}},"alternative-id":["10.1287\/mnsc.2023.03818"],"URL":"https:\/\/doi.org\/10.1287\/mnsc.2023.03818","relation":{},"ISSN":["0025-1909","1526-5501"],"issn-type":[{"value":"0025-1909","type":"print"},{"value":"1526-5501","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6]]}}}