On the Use of Two-Way Fixed Effects Regression Models for Causal Inference with Panel Data
Open accessAbstract
Abstract The two-way linear fixed effects regression ( 2FE ) has become a default method for estimating causal effects from panel data. Many applied researchers use the 2FE estimator to adjust for unobserved unit-specific and time-specific confounders at the same time. Unfortunately, we demonstrate that the ability of the 2FE model to simultaneously adjust for these two types of unobserved confounders critically relies upon the assumption of linear additive effects. Another common justification for the use of the 2FE estimator is based on its equivalence to the difference-in-differences estimator under the simplest setting with two groups and two time periods. We show that this equivalence does not hold under more general settings commonly encountered in applied research. Instead, we prove that the multi-period difference-in-differences estimator is equivalent to the weighted 2FE estimator with some observations having negative weights. These analytical results imply that in contrast to the popular belief, the 2FE estimator does not represent a design-based, nonparametric estimation strategy for causal inference. Instead, its validity fundamentally rests on the modeling assumptions.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
0 of 3 cited sources checked · 0 supported · 0 flagged
- Findingreasoning
The multi-period difference-in-differences estimator is equivalent to the weighted FE estimator with some observations having negative weights.
“Instead, we prove that the multi-period difference-in-differences estimator is equivalent to the weighted FE estimator with some observations having negative weights.”
- Also stated in: § . The Multi-period Difference-in-Differences Estimator · boosted
“Instead, the following theorem establishes that the general multi-period DiD estimator given in Equation ( ) is equivalent to a weighted two-way fixed effects regression estimator.”
- Also stated in: § T (Difference-in-Differences · plain
“Second, although the multi-period DiD estimator can be shown to be equivalent to the weighted FE estimator, some control observations will have negative regression weights.”
Quote in the paperSubstantial claim - Also stated in: § . The Multi-period Difference-in-Differences Estimator · boosted
- Findingreasoning
The two-way matching estimator that eliminates mismatches within-unit and within-time dimension can be written as a weighted FE estimator.
“The next proposition establishes that this two-way matching estimator, which eliminates mismatches within-unit and within-time dimension, can be written as a weighted FE estimator.”
Quote in the paperSubstantial claim- The authors state this as an assumption: The treatment varies within each unit as well as within each time period.
- Findingreasoning
Under the difference-in-differences design, all mismatches can be eliminated.
“Although it is generally impossible to eliminate all mismatches, in this section we show that we can do so under the difference-in-differences (DiD) design.”
- Also stated in: § . The Multi-period Difference-in-Differences Estimator · plain
“It is worth noting that all three sets only contain control observations, thereby eliminating all mismatches.”
Quote in the paperSubstantial claim- The authors state this as an assumption: The parallel trend assumption holds for the DiD design.
- Also stated in: § . The Multi-period Difference-in-Differences Estimator · plain
- Findingcitation
The two-way fixed effects estimator is equivalent to a two-way matching estimator.
“The two-way fixed effects estimator defined in Equation ( ) is equivalent to the following matching estimator,”
- Also stated in: § Adjustment for Unobserved Confounders · plain
“The proposition shows that the estimated counterfactual outcome of a given observation, that is,”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Imai and Kim
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Also stated in: § Adjustment for Unobserved Confounders · plain
- Findingreasoning
Even the weighted FE estimator, which has fewer mismatches than the standard FE estimator, suffers from some mismatches.
“Therefore, even the weighted FE estimator, which has fewer mismatches than the standard FE estimator, suffers from some mismatches.”
Quote in the paperSubstantial claim
- Findingreasoning
The weighted FE estimator has an adjustment set whose observations belong to neither the same unit nor the same time period as the observation being matched with.
“The estimator also has an adjustment set whose observations belong to neither the same unit nor the same time period as the observation being matched with.”
Quote in the paperSubstantial claim - Interpretationreasoning
The equivalence between the FE estimator and the difference-in-differences estimator does not hold under more general settings commonly encountered in applied research.
“We show that this equivalence does not hold under more general settings commonly encountered in applied research.”
- Also stated in: § Introduction · plain
“However, we show that this equivalence does not hold under more general settings frequently encountered in applied research.”
Quote in the paperSubstantial claim - Also stated in: § Introduction · plain
- Interpretationnone given
The two-way FE model's ability to simultaneously adjust for unit-specific and time-specific unobserved confounders critically relies upon the assumption of linear additive effects.
“Unfortunately, we demonstrate that the ability of the FE model to simultaneously adjust for these two types of unobserved confounders critically relies upon the assumption of linear additive effects.”
- Also stated in: § Introduction · boosted
“Unfortunately, we show that the FE's ability to simultaneously adjust for the two types of unobserved confounders critically hinges upon the assumption of linear additive effects.”
Quote in the paperSubstantial claim - Also stated in: § Introduction · boosted
- Premisecitation
The FE estimator is equivalent to the difference-in-differences estimator under the simplest setting with two groups and two time periods.
“Another common justification is based on the fact that the FE estimator is equivalent to the difference-in-differences estimator under the simplest setting with two groups and two time periods (e.g., Bertrand, Duflo, and Mullainathan ; Angrist and Pischke ).”
- Also stated in: § . The Multi-period Difference-in-Differences Estimator · boosted
“It is well known that the standard nonparametric DiD estimator is numerically equivalent to the FE estimator in the simplest setting, in which there are only two time periods and the treatment is administered only to one group of units in the second time period.”
Quote in the paperSubstantial claim0 of 2 sources checked- cites Bertrand, Duflo, and Mullainathan
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- cites Angrist and Pischke
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Also stated in: § . The Multi-period Difference-in-Differences Estimator · boosted
- Interpretationreasoning
The justification of the FE estimator as the DiD estimator is not warranted unless the linearity assumption is imposed.
“Since the regression weights should generally positive, the results of this section shows that the justification of the FE estimator as the DiD estimator is not warranted unless the linearity assumption is imposed.”
Quote in the paperSubstantial claim - Interpretationreasoning
The FE estimator does not represent a design-based, nonparametric estimation strategy for causal inference.
“These analytical results imply that in contrast to the popular belief, the FE estimator does not represent a designbased, nonparametric estimation strategy for causal inference.”
- Also stated in: § Introduction · boosted
“All together, we show that in contrast to the popular belief, the FE estimator does not represent a design-based, nonparametric estimation strategy for causal inference.”
- Also stated in: § Concluding Remarks · plain
“We show that contrary to the common belief, the standard two-way fixed effects regression estimator does not represent a design-based, nonparametric causal estimator.”
Quote in the paperSubstantial claim - Also stated in: § Introduction · boosted
- Interpretationreasoning
Mismatches imply the partial comparison of observations with the same treatment status, which generally leads to an attenuation bias.
“Therefore, mismatches imply the (partial) comparison of observations with the same treatment status, which generally leads to an attenuation bias.”
Quote in the paperSubstantial claim - Interpretationreasoning
It is impossible to simultaneously and nonparametrically adjust for unit-specific and time-specific unobserved confounders under the two-way fixed effects framework.
“This implies that it is impossible to simultaneously and nonparametrically adjust for unit-specific and time-specific unobserved confounders under the two-way fixed effects framework.”
- Also stated in: § Concluding Remarks · plain
“It is impossible to simultaneously adjust for unobserved unit-specific and time-specific confounders.”
Quote in the paperSubstantial claim - Also stated in: § Concluding Remarks · plain
- Interpretationreasoning
The validity of the FE estimator fundamentally rests on the modeling assumptions.
“Instead, its validity fundamentally rests on the modeling assumptions.”
- Also stated in: § Introduction · plain
“Instead, its validity fundamentally rests on the modeling assumptions.”
Quote in the paperSubstantial claim - Also stated in: § Introduction · plain
- Interpretationreasoning
If the model in Equation (1) is assumed correct, then the FE estimator is consistent for τ under the multi-period DiD design.
“If the model in Equation ( ) is assumed to be correct, then the FE estimator is consistent for τ under the multi-period DiD design.”
Quote in the paperSubstantial claim- The authors state this as an assumption: The parallel trend assumption holds for the DiD design.
- Interpretationreasoning
The equivalence between the FE estimator and the DiD estimator critically hinges on the linearity assumption.
“This implies that the equivalence between the FE estimator and the DiD estimator critically hinges on the linearity assumption.”
Quote in the paperSubstantial claim
The paper relies on this without support.
The two-way FE regression is a default method for estimating causal effects from panel data among applied researchers.
“The two-way linear fixed effects regression ( FE) has become a default method for estimating causal effects from panel data.”
“Many social scientists use the two-way fixed effects ( FE) regression, or linear regression with unit and time fixed effects, as the default methodology for estimating causal effects from panel data.”
The paper relies on this without support.
Many applied researchers use the FE estimator to adjust for unobserved unit-specific and time-specific confounders at the same time.
“Many applied researchers use the FE estimator to adjust for unobserved unit-specific and time-specific confounders at the same time.”
“Applied researchers o en use the FE regression to adjust for unobserved unit-specific and time-specific confounders at the same time.”
The authors state this as an assumption.
The two types of unobserved confounders (unit-specific and time-specific) are additive and separable, with no interaction between them.
“Thus, although the interaction between these two types of unobserved confounders is assumed to be absent, there is no functional-form restriction on h(•) and f (•).”
“In other words, since the treatment is binary, the model makes no restriction other than the additivity and separability of the two types of unobserved confounders.”
The authors state this as an assumption.
The parallel trend assumption holds for the DiD design.
“To establish the relations between the FE and DiD estimators, we begin by considering the following parallel trend assumption, A (Parallel Trend).”
The authors state this as an assumption.
The treatment varies within each unit as well as within each time period.
“Assume that the treatment varies within each unit as well as within each time period, that is, 0 < T t =1 X i t < T for each i and 0 < N i =1 X i t < N for each t.”
The authors state this as an assumption.
There is at least one treated and control unit, and at least one unit with D_it = 1.
“Assume that there is at least one treated and control unit, that is, 0 < N i =1 T t =1 X i t < NT , and that there is at least one unit with D i t = 1, that is, 0 < N i =1 T t =1 D i t .”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
This approach yields an interpretable causal effect under generalizations of the parallel trends assumption and an additional no carryover assumption that imposes that potential outcomes depend only on current treatment status and not on the full treatment history.
“This approach yields an interpretable causal effect under generalizations of the parallel trends assumption and and an additional 'no carryover' assumption that imposes that the potential outcomes depend only on current treatment status and not on the full treatment history.”
From What's trending in difference-in-differences? A synthesis of the recent econometrics literature (2023) · cites this paper as Imai and Kim (2021)
Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.
Experimental — accuracy not yet measuredde Chaisemartin and D'Haultfœuille (2020) and Imai and Kim (2021) consider settings where units are treated at different times, but do not necessarily require that treatment is an absorbing state.
“de Chaisemartin and D'Haultfœuille (2020) and Imai and Kim (2021) consider settings where units are treated at different times, but do not necessarily require that treatment is an absorbing state.”
From What's trending in difference-in-differences? A synthesis of the recent econometrics literature (2023) · cites this paper as Imai and Kim (2021)
Their estimators compare changes in outcomes for units whose treatment status changed to other units whose treatment status remained constant over the same periods.
“Their estimators intuitively compare changes in outcomes for units whose treatment status changed to other units whose treatment status remained constant over the same periods.”
From What's trending in difference-in-differences? A synthesis of the recent econometrics literature (2023) · cites this paper as Imai and Kim (2021)
Cites
How Much Should We Trust Differences-in-Differences Estimates?
Bertrand, M., Duflo, E., & Mullainathan, S. (2004) · The Quarterly Journal of Economics · doi:10.1162/003355304772839588 · Cited by 10,641 · cited 2× · see citing works
“Another common justification is based on the fact that the 2FE estimator is equivalent to the difference-in-differences estimator under the simplest setting with two groups and two time periods (e.g., Bertrand, Duflo, and Mullainathan 2004; Angrist and Pischke 2009).”
MethodMostly Harmless Econometrics: An Empiricist's Companion
Angrist, J. D., & Pischke, J.-S. (2009) · doi:10.1515/9781400829828 · Cited by 9,467 · cited 2× · see citing works
“Our result is also related to that of Arkhangelsky et al. (2018) who shows that the weighted two-way fixed effects estimator can improve the synthetic control method of Abadie et al.”
MethodResultAbadie, A., Diamond, A., & Hainmueller, J. (2010) · Journal of the American Statistical Association · doi:10.1198/jasa.2009.ap08746 · Cited by 5,513 · cited 1× · see citing works
“Recently, a number of researchers have extended the synthetical control method of Abadie, Diamond, and Hainmueller (2010) to more general settings (e.g., Xu 2017; Ben-Michael, Feller, and Rothstein 2019).”
BackgroundTwo-Way Fixed Effects Estimators with Heterogeneous Treatment Effects
de Chaisemartin, C., & D’Haultfœuille, X. (2020) · American Economic Review · doi:10.1257/aer.20181169 · Cited by 5,031 · see citing works
Solon, G., Haider, S. J., & Wooldridge, J. M. (2015) · Journal of Human Resources · doi:10.3368/jhr.50.2.301 · Cited by 1,367 · see citing works
Generalized Synthetic Control Method: Causal Inference with Interactive Fixed Effects Models
Xu, Y. (2017) · Political Analysis · doi:10.1017/pan.2016.2 · Cited by 938 · cited 1× · see citing works
“Recently, a number of researchers have extended the synthetical control method of Abadie, Diamond, and Hainmueller (2010) to more general settings (e.g., Xu 2017; Ben-Michael, Feller, and Rothstein 2019).”
BackgroundDesign-based analysis in Difference-In-Differences settings with staggered adoption
Athey, S., & Imbens, G. W. (2022) · Journal of Econometrics · doi:10.1016/j.jeconom.2020.10.012 · Cited by 917 · cited 2× · see citing works
“In addition, a number of scholars have recently considered causal interpretations of the standard 2FE estimator (see, e.g., Borusyak and Jaravel 2017; Abraham and Sun 2018; Athey and Imbens 2018; Chaisemartin and D’Haultfœuille 2018; Goodman-Bacon 2018).”
BackgroundDifference-in-Differences with Variation in Treatment Timing
Goodman-Bacon, A. (2018) · doi:10.3386/w25018 · Cited by 757 · see citing works
Revisiting Event Study Designs
Borusyak, K., & Jaravel, X. (2016) · SSRN Electronic Journal · doi:10.2139/ssrn.2826228 · Cited by 463 · see citing works
When Should We Use Unit Fixed Effects Regression Models for Causal Inference with Longitudinal Data?
Imai, K., & Kim, I. (2019) · American Journal of Political Science · doi:10.1111/ajps.12417 · Cited by 371 · cited 5× · see citing works
“Indeed, by extending the matching framework of Imai and Kim (2019), we show that the simultaneous adjustment for the two types of unobserved confounders cannot be done nonparametrically under the 2FE framework.”
MethodInfluentialSelf-citationKahn-Lang, A., & Lang, K. (2020) · Journal of Business & Economic Statistics · doi:10.1080/07350015.2018.1546591 · Cited by 367 · cited 2× · see citing works
“We emphasize that this assumption may not be credible in some settings (see, e.g., Bilinski and Hatfield 2018; Kahn-Lang and Lang 2019; Rambachan and Roth 2019).”
BackgroundTwo-way Fixed Effects Estimators with Heterogeneous Treatment Effects
de Chaisemartin, C., & D'Haultfoeuille, X. (2019) · doi:10.3386/w25904 · Cited by 350 · see citing works
Does Regression Produce Representative Estimates of Causal Effects?
Aronow, P. M., & Samii, C. (2015) · American Journal of Political Science · doi:10.1111/ajps.12185 · Cited by 190 · cited 1× · see citing works
“…Kim 2019) to the 2FE estimator in order to understand the causal interpretation of these widely used estimators within the nonparametric framework (see, e.g., Humphreys 2009; Aronow and Samii 2015; Solon, Haider, and Wooldridge 2015, for related work on causal inference with crosssectional data).”
MethodEstimating Dynamic Treatment Effects in Event Studies With Heterogeneous Treatment Effects
Sun, L., & Abraham, S. (2018) · SSRN Electronic Journal · doi:10.2139/ssrn.3158747 · Cited by 118 · cited 1× · see citing works
“In addition, a number of scholars have recently considered causal interpretations of the standard 2FE estimator (see, e.g., Borusyak and Jaravel 2017; Abraham and Sun 2018; Athey and Imbens 2018; Chaisemartin and D’Haultfœuille 2018; Goodman-Bacon 2018).”
BackgroundNothing to See Here? A non-inferiority approach to parallel trends
Bilinski, A., & Hatfield, L. A. (2018) · arXiv · doi:10.48550/arxiv.1805.03273 · Cited by 104
(2018) · arXiv (Cornell University) · Cited by 16 · see citing works
Synthetic Controls and Weighted Event Studies with Staggered AdoptionPreprint
Ben‐Michael, E., Feller, A., & Rothstein, J. (2019) · arXiv (Cornell University) · Cited by 12 · see citing works
On the Use of Linear Fixed Eects Regression Models for Causal Inference
Imai, K., & Kim, I. S. L. (2012) · Cited by 6 · see citing works
Self-citationSynthetic Controls with Staggered Adoption
Ben-Michael, E., Feller, A., & Rothstein, J. (2019) · arXiv · doi:10.48550/arxiv.1912.03290
Flagged first, then by citations
1 of the 20 references couldn’t be matched to a known paper.
Cited by
Sorted by citations
Difference-in-differences with variation in treatment timing
Goodman-Bacon, A. (2021) · Journal of Econometrics · doi:10.1016/j.jeconom.2021.03.014 · Cited by 8,046 · see citing works
How Much Should We Trust Staggered Difference-in-Differences Estimates?
Baker, A. C., Larcker, D. F., & Wang, C. C. Y. (2022) · Journal of Financial Economics · doi:10.1016/j.jfineco.2022.01.004 · Cited by 3,149 · see citing works
What's trending in difference-in-differences? A synthesis of the recent econometrics literature
Roth, J., Sant’Anna, P. H. C., Bilinski, A., & Poe, J. (2023) · Journal of Econometrics · doi:10.1016/j.jeconom.2023.03.008 · Cited by 2,027 · see citing works
How Much Should We Trust Staggered Difference-In-Differences Estimates?Preprint
Baker, A., Larcker, D. F., & Wang, C. C. Y. (2021) · SSRN Electronic Journal · doi:10.2139/ssrn.3794018 · Cited by 670 · see citing works
“…where a treatment occurs across units at different points in time, 5The body of work examining these issues include: Athey and Imbens (2018), Borusyak and Jaravel (2017),Callaway and Sant’Anna (2020), Goodman-Bacon (2019), Imai and Kim (2020), Strezhnev (2018), and Sun and Abraham (2020).”
BackgroundTwo-way fixed effects and differences-in-differences with heterogeneous treatment effects: a survey
de Chaisemartin, C., & D’Haultfœuille, X. (2023) · The Econometrics Journal · doi:10.1093/ectj/utac017 · Cited by 522 · see citing works
Liu, L., Wang, Y., & Xu, Y. (2024) · American Journal of Political Science · doi:10.1111/ajps.12723 · Cited by 390 · see citing works
Two-way Fixed Effects Estimators with Heterogeneous Treatment Effects
de Chaisemartin, C., & D'Haultfoeuille, X. (2019) · doi:10.3386/w25904 · Cited by 350 · see citing works
Matching Methods for Causal Inference with Time‐Series Cross‐Sectional Data
Imai, K., Kim, I. S., & Wang, E. H. (2023) · American Journal of Political Science · doi:10.1111/ajps.12685 · Cited by 349 · see citing works
Self-citationTwo-stage differences in differences
Gardner, J. (2022) · arXiv · doi:10.48550/arxiv.2207.05943 · Cited by 333 · see citing works
Greenwashing behavior and firm value – From the perspective of board characteristics
Chen, P., & Dagestani, A. A. (2023) · Corporate Social Responsibility and Environmental Management · doi:10.1002/csr.2488 · Cited by 271 · see citing works
Can Technology Solve the Principal-Agent Problem? Evidence from China’s War on Air Pollution
Greenstone, M., He, G., Jia, R., & Liu, T. (2022) · American Economic Review Insights · doi:10.1257/aeri.20200373 · Cited by 275 · see citing works
Systematic Evaluation of State Policy Interventions Targeting the US Opioid Epidemic, 2007-2018
Lee, B., Zhao, W., Yang, K.-C., Ahn, Y., & Perry, B. L. (2021) · JAMA Network Open · doi:10.1001/jamanetworkopen.2020.36687 · Cited by 186 · see citing works
Sustainability or Greenwashing: Evidence from the Asset Market for Industrial Pollution
Duchin, R., Gao, J., & Xu, Q. (2024) · The Journal of Finance · doi:10.1111/jofi.13412 · Cited by 249 · see citing works
Education and Political Participation
Willeck, C., & Mendelberg, T. (2021) · Annual Review of Political Science · doi:10.1146/annurev-polisci-051120-014235 · Cited by 159 · see citing works
Synthetic controls with staggered adoption.
Ben-Michael, E., Feller, A., & Rothstein, J. (2022) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/rssb.12448 · Cited by 120 · see citing works
The Effect of Job Loss and Unemployment Insurance on Crime in Brazil
Britto, D. G. C., Pinotti, P., & Sampaio, B. R. (2022) · Econometrica · doi:10.3982/ecta18984 · Cited by 110 · see citing works
Nothing to See Here? A Non‐Inferiority Approach to Parallel Trends
Bilinski, A., & Hatfield, L. A. (2026) · Statistics in Medicine · doi:10.1002/sim.70296 · Cited by 137 · see citing works
Dong, C., Zhou, R., & Li, J. (2020) · Applied Energy · doi:10.1016/j.apenergy.2020.116007 · Cited by 103 · see citing works
Do Commodity Price Shocks Cause Armed Conflict? A Meta-Analysis of Natural Experiments
Blair, G., Christensen, D., & Rudkin, A. F. (2021) · American Political Science Review · doi:10.1017/s0003055420000957 · Cited by 103 · see citing works
Sustainability or Greenwashing: Evidence from the Asset Market for Industrial PollutionPreprint
Duchin, R., Gao, J., & Xu, Q. (2022) · SSRN Electronic Journal · doi:10.2139/ssrn.4095885 · Cited by 98 · see citing works
Scott, R. (2022) · Electoral Studies · doi:10.1016/j.electstud.2022.102471 · Cited by 97 · see citing works
Neighborhood Externalities and Coordinated Rebuilding after Wildfires: Evidence from CaliforniaPreprint
Issler, P., Stanton, R., Vergara‐Alert, C., & Wallace, N. E. (2020) · SSRN Electronic Journal · doi:10.2139/ssrn.3511843 · Cited by 85 · see citing works
Causal Inference With Observational Data and Unobserved Confounding Variables
Byrnes, J. E. K., & Dee, L. E. (2025) · Ecology Letters · doi:10.1111/ele.70023 · Cited by 90 · see citing works
“Solutions are evolving rapidly, including for heterogeneous effects (see Roth et al. 2023), nonlinearities (Imai and Kim 2021), and continuous causal variables (Callaway, Goodman-Bacon, and Sant'Anna 2024).”
Yoder, J. (2020) · American Political Science Review · doi:10.1017/s0003055420000556 · Cited by 81 · see citing works
“The estimator, proposed in Imai and Kim (2020), is equivalent to a weighted version of the standard two-way fixed effects regression.”
MethodThe Correlates of State Policy and the Structure of State Panel Data
Grossmann, M., Jordan, M. P., & McCrain, J. (2021) · State Politics & Policy Quarterly · doi:10.1017/spq.2021.17 · Cited by 69 · see citing works
Two-Way Fixed Effects and Differences-in-Differences with Heterogeneous Treatment Effects: A Survey
de Chaisemartin, C., & D'Haultfoeuille, X. (2022) · National Bureau of Economic Research · doi:10.3386/w29734 · Cited by 70 · see citing works
Two-Way Fixed Effects and Differences-in-Differences with Heterogeneous Treatment Effects: A SurveyPreprint
de Chaisemartin, C., & D’Haultfœuille, X. (2022) · SSRN Electronic Journal · doi:10.2139/ssrn.4034184 · Cited by 71 · see citing works
Chen, P., Dagestani, A. A., & Kim, S. (2023) · Technology Analysis and Strategic Management · doi:10.1080/09537325.2023.2196585 · Cited by 70 · see citing works
“However, these two unobserved confounders rely heavily on the assumption of linear additive e ff ects (Imai and Kim 2021), which needs further testing.”
Xiao, D., Yu, F., & Guo, C. (2022) · Energy Economics · doi:10.1016/j.eneco.2023.106770 · Cited by 70 · see citing works
Did border closures slow SARS-CoV-2?
Shiraef, M. A., Friesen, P., Feddern, L., Weiss, M. A., COBAP Team, Janabi, H. A., Beling, E., Falcone, J., Feddern, L., Hirst, C., et al. (2022) · Scientific Reports · doi:10.1038/s41598-022-05482-7 · Cited by 60 · see citing works
Job Displacement, Unemployment Benefits and Domestic Violence
Bhalotra, S. R., Britto, D. G. C., Pinotti, P., & Sampaio, B. R. (2021) · The Review of Economic Studies · doi:10.1093/restud/rdaf004 · Cited by 67 · see citing works
Panel Experiments and Dynamic Causal Effects: A Finite Population Perspective
Bojinov, I., Rambachan, A., & Shephard, N. (2021) · Quantitative Economics · doi:10.3982/qe1744 · Cited by 60 · see citing works
“For example, Sobel (2006), Athey and Imbens (2018), and Imai and Kim (2019a) study β̂TWFE in the causal literature on panel data models.”
BackgroundCirone, A. E., & Pepinsky, T. B. (2021) · Annual Review of Political Science · doi:10.1146/annurev-polisci-051120-104325 · Cited by 67 · see citing works
The constrained politics of local public investment under cooperative federalism
Bremer, B., Carlo, D. D., & Wansleben, L. (2022) · Socio-Economic Review · doi:10.1093/ser/mwac026 · Cited by 60 · see citing works
On the causality and plausibility of treatment effects in operations management research
Mithas, S., Chen, Y., Lin, Y., & Silveira, A. D. O. (2022) · Production and Operations Management · doi:10.1111/poms.13863 · Cited by 57 · see citing works
“Recent work provides new insights on the assumptions of unit FEs (e.g., Imai & Kim, 2019) and two-way FEs models (e.g., Imai & Kim, 2021) when one focuses attention on binary treatments.”
BackgroundCarnegie, A. S., & Carson, A. (2020) · Cambridge University Press eBooks · doi:10.1017/9781108778114 · Cited by 56 · see citing works
Economic freedom and foreign direct investment: Are they related?
Tağ, M. N., & Deği̇rmen, S. (2022) · Economic Analysis and Policy · doi:10.1016/j.eap.2021.12.020 · Cited by 53 · see citing works
Two-Way Fixed Effects and Differences-in-Differences with Heterogeneous Treatment Effects: A SurveyPreprint
de Chaisemartin, C., & D’Haultfœuille, X. (2022) · SSRN Electronic Journal · doi:10.2139/ssrn.4021785 · Cited by 50 · see citing works
Causal models for longitudinal and panel data: a survey
Arkhangelsky, D., & Imbens, G. (2024) · The Econometrics Journal · doi:10.1093/ectj/utae014 · Cited by 50 · see citing works
Zor, S. (2023) · Journal of Cleaner Production · doi:10.1016/j.jclepro.2023.139815 · Cited by 49 · see citing works
Trauma and Turnout: The Political Consequences of Traumatic Events
Marsh, W. Z. C. (2022) · American Political Science Review · doi:10.1017/s0003055422001010 · Cited by 46 · see citing works
Two-Way Fixed Effects and Differences-in-Differences with Heterogeneous Treatment Effects: A SurveyPreprint
de Chaisemartin, C., & D’Haultfœuille, X. (2021) · SSRN Electronic Journal · doi:10.2139/ssrn.3980758 · Cited by 45 · see citing works
“The multi-period DID estimator in Imai and Kim (2021) is related to the DID M estimator.”
MethodDennison, J., & Kustov, A. (2023) · Public Opinion Quarterly · doi:10.1093/poq/nfad052 · Cited by 45 · see citing works
“Born for a Storm”: Hard-Right Social Media and Civil Unrest
Karell, D., Linke, A. M., Holland, E. C., & Hendrickson, E. (2023) · American Sociological Review · doi:10.1177/00031224231156190 · Cited by 48 · see citing works
Behavioural nudges for water conservation in unequal settings: Experimental evidence from Cape Town
Brick, K., Martino, S. D., & Visser, M. (2023) · Journal of Environmental Economics and Management · doi:10.1016/j.jeem.2023.102852 · Cited by 44 · see citing works
The Real Consequences of Symbolic Politics: Breaking the Soviet Past in Ukraine
Rozenas, A., & Vlasenko, A. (2021) · The Journal of Politics · doi:10.1086/718210 · Cited by 44 · see citing works
Interaction effects of market-based and incentive-driven low-carbon policies on carbon emissions
Jia, Z., & Wen, S. (2024) · Energy Economics · doi:10.1016/j.eneco.2024.107776 · Cited by 43 · see citing works
The Supply-Equity Trade-Off: The Effect of Spatial Representation on the Local Housing Supply
Hankinson, M., & Magazinnik, A. (2023) · The Journal of Politics · doi:10.1086/723818 · Cited by 41 · see citing works
Turunen, J., Karhula, K., Ropponen, A., Koskinen, A., Hakola, T., Puttonen, S., Hämäläinen, K., Pehkonen, J., & Härmä, M. (2020) · International Journal of Nursing Studies · doi:10.1016/j.ijnurstu.2020.103716 · Cited by 39 · see citing works
Ambient air pollution associated with lower academic achievement among US children
Lu, W., Hackman, D. A., & Schwartz, J. D. (2021) · Environmental Epidemiology · doi:10.1097/ee9.0000000000000174 · Cited by 1 · see citing works
Energy security and the shift to renewable resources: The case of Russia-Ukraine war
Nguyen, H. H., Nguyen, P. V., & Ngo, V. M. (2024) · The Extractive Industries and Society · doi:10.1016/j.exis.2024.101442 · Cited by 38 · see citing works
The effects of new energy vehicle subsidies on air quality: Evidence from China
Li, P., & Zhang, Z. (2023) · Energy Economics · doi:10.1016/j.eneco.2023.106624 · Cited by 39 · see citing works
Access to outpatient services in rural communities changes after hospital closure
Miller, K. E. M., Miller, K., Knocke, K., et al. (2021) · Health Services Research · doi:10.1111/1475-6773.13694 · Cited by 37 · see citing works
Making housing affordable? The local effects of relaxing land-use regulation
Büchler, S., & Lutz, E. C. (2024) · Journal of Urban Economics · doi:10.1016/j.jue.2024.103689 · Cited by 35 · see citing works
A panel data analysis of policy effectiveness for renewable energy expansion on Caribbean islands
Kersey, J., Blechinger, P., & Shirley, R. G. (2021) · Energy Policy · doi:10.1016/j.enpol.2021.112340 · Cited by 41 · see citing works
“A key constraint of the present work is that the regression specifications are limited in their ability to be causally interpreted in applied research (Imai and Kim, 2020).”
BackgroundThe impact of climate policy uncertainty on urban climate risk: Evidence from 274 cities in China
Zhang, H., Liu, J., Guo, Q., Zhang, X., & Hu, X. (2024) · Sustainable Cities and Society · doi:10.1016/j.scs.2024.106097 · Cited by 35 · see citing works
Difference-in-Differences Designs: A Practitioner’s Guide
Baker, A., Callaway, B., Cunningham, S., Goodman-Bacon, A., & Sant’Anna, P. H. C. (2026) · Journal of Economic Literature · doi:10.1257/jel.20251650 · Cited by 105 · see citing works
He, Y., Wen, C., Fang, X., & Sun, X. M. (2024) · Landscape Ecology · doi:10.1007/s10980-024-01926-9 · Cited by 35 · see citing works
“Scholars always utilize this model to adjust for unobserved inherent individual or temporal attributes that have confounding effects (Imai and Kim 2021).”
Transparency in Structural Research
Andrews, I., Gentzkow, M. A., & Shapiro, J. M. (2020) · Journal of Business and Economic Statistics · doi:10.1080/07350015.2020.1796395 · Cited by 32 · see citing works
Heinzel, M., & Koenig‐Archibugi, M. (2023) · The Review of International Organizations · doi:10.1007/s11558-023-09510-7 · Cited by 33 · see citing works
“A2-A5, Table A3) (Blackwell & Glynn, 2018; De Chaisemartin & d’Haultfoeuille, 2020; Goodman-Bacon, 2021; Imai & Kim, 2021; Liu et al., 2022).”
Two-Way Fixed Effects and Differences-in-Differences with Heterogeneous Treatment Effects: A Survey
de Chaisemartin, C., & D'Haultfoeuille, X. (2022) · doi:10.3386/w29691 · Cited by 31 · see citing works
Fixed Effects and Causal InferencePreprint
Millimet, D. L., & Bellemare, M. F. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4467963 · Cited by 32 · see citing works
Rare and highly destructive wildfires drive human migration in the U.S.
McConnell, K., Fussell, E., DeWaard, J., Whitaker, S. D., Curtis, K. J., Denis, L. A. S., Balch, J. K., & Price, K. (2024) · Nature Communications · doi:10.1038/s41467-024-50630-4 · Cited by 31 · see citing works
Public-Private Governance Initiatives and Corporate Responses to Stakeholder Complaints
Thrall, C. (2021) · International Organization · doi:10.1017/s0020818321000199 · Cited by 30 · see citing works
“Imai and Kim 2020.”
Liu, X., Liu, Z., Zhu, Z., et al. (2024) · Building and Environment · doi:10.1016/j.buildenv.2024.111618 · Cited by 29 · see citing works
Egami, N., & Yamauchi, S. (2022) · Political Analysis · doi:10.1017/pan.2022.8 · Cited by 38 · see citing works
“Examples include Dube, Dube, and García-Ponce (2013), Truex (2014), Earle and Gehlbach (2015), Hall (2016), and Larreguy and Marshall (2017). Another is to stick with the two-time-period DID and limit the use of additional pre-treatment periods only to the assessment of pre-treatment trends.”
Stokes, J., Bower, P., Guthrie, B., et al. (2022) · The Lancet Regional Health - Europe · doi:10.1016/j.lanepe.2022.100436 · Cited by 29 · see citing works
Testing the backlash argument: voter responses to (pro-)immigration reforms
Kustov, A. (2022) · Journal of European Public Policy · doi:10.1080/13501763.2022.2061034 · Cited by 28 · see citing works
“…however, has been recently challenged by some scholars who point out to the unclear counterfactual interpretation of twoway FEs and the impossibility of simultaneously adjusting for unit-specific and time-specific confounders in a nonparametric way (Kropko & Kubinec, 2020; Imai & Kim, 2021).”
BackgroundDong, H., Tao, M., Wang, J., & Baiocchi, G. (2024) · Sustainable Development · doi:10.1002/sd.3063 · Cited by 31 · see citing works
Lu, W., Levin, R., & Schwartz, J. (2021) · BMC Public Health · doi:10.1186/s12889-021-12474-1 · Cited by 26 · see citing works
Unexpected events during survey design and trust in the police: a systematic review
Nägel, C., & Nivette, A. E. (2022) · Journal of Experimental Criminology · doi:10.1007/s11292-022-09508-y · Cited by 29 · see citing works
“We recommend to adjust for unobserved unit-specific and time-specific confounders using a two-way-fixed-effects modeling strategy even though this does not represent a design-based estimation strategy (Imai and Kim 2021) that can yield similar confidence in the results as the straightforward UESD.”
Uncovering Latent Biases in Text: Method and Application to Peer Review
Manzoor, E., & Shah, N. B. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i6.16608 · Cited by 28 · see citing works
“(5) is typically estimated using a “two-way fixed-effects” regression (Imai and Kim, 2020) on peer reviews from the years tSB and tDB:”
MethodRural household energy transition in China: Trends and challenges
Zhang, X., Barrington-Leigh, C. P., & Robinson, B. E. (2024) · Journal of Cleaner Production · doi:10.1016/j.jclepro.2024.141871 · Cited by 25 · see citing works
Extreme weather events do not increase political parties' environmental attention
Wappenhans, T., Valentim, A., Klüver, H., et al. (2024) · Nature Climate Change · doi:10.1038/s41558-024-02024-z · Cited by 25 · see citing works
“25, allowing for both staggered and repeated treatment exposure.”
Does Transparency Inhibit Political Compromise?
Harden, J. J., & Kirkland, J. H. (2020) · American Journal of Political Science · doi:10.1111/ajps.12564 · Cited by 25 · see citing works
How Exile Shapes Online Opposition: Evidence from Venezuela
Esberg, J., & Siegel, A. A. (2022) · American Political Science Review · doi:10.1017/s0003055422001290 · Cited by 30 · see citing works
“…has also raised concerns about the causal interpretation of two-way fixed effects difference-in-differences designs with staggered treat-ment timing when effects are heterogenous across time (which is often likely to be the case; Brantly and Sant ’ Anna 2021; Goodman-Bacon 2021; Imai and Kim 2020).”
Secrets in Global Governance : Disclosure Dilemmas and the Challenge of International Cooperation
Carnegie, A. S., & Carson, A. (2020) · Cited by 24 · see citing works
Does Abortion Liberalisation Accelerate Fertility Decline? A Worldwide Time-Series Analysis
Fernández, J. J., & Juif, D. (2023) · European Journal of Population / Revue européenne de Démographie · doi:10.1007/s10680-023-09687-y · Cited by 26 · see citing works
“In the TWFE model, changes in abortion policy index 1 are positively related to increases in fertility rates (model 1).”
MethodInfluentialThe effect of digital transformation: Boosting productivity in the restaurant industry
Lee, W., Jang, S. (., & Kim, H. (2024) · International Journal of Hospitality Management · doi:10.1016/j.ijhm.2024.103896 · Cited by 24 · see citing works
Causal graphical views of fixed effects and random effects models
Kim, Y., & Steiner, P. M. (2020) · British Journal of Mathematical and Statistical Psychology · doi:10.1111/bmsp.12217 · Cited by 26 · see citing works
Effects of state opioid prescribing cap laws on opioid prescribing after surgery
Schmid, I. T., Stuart, E. A., McCourt, A. D., Tormohlen, K. N., Stone, E. M., Davis, C. S., Bicket, M. C., & McGinty, E. E. (2022) · Health Services Research · doi:10.1111/1475-6773.14023 · Cited by 24 · see citing works
Synthetic Controls with Staggered Adoption
Ben‐Michael, E., Feller, A., & Rothstein, J. M. (2021) · National Bureau of Economic Research · doi:10.3386/w28886 · Cited by 24 · see citing works
Perceived group discrimination and the integration paradox in stigmatized neighborhoods
Esaiasson, P., Lajevardi, N., & Sohlberg, J. (2023) · Politics Groups and Identities · doi:10.1080/21565503.2023.2227438 · Cited by 23 · see citing works
“Two-way fi xed e ff ects regressions are sometimes problematic because of the assumption of linear additive e ff ects (Imai and Kim 2021).”
Ren, Y., Tian, Y., & Zhang, C. (2021) · Environment Development and Sustainability · doi:10.1007/s10668-021-01971-3 · Cited by 23 · see citing works
Tiscornia, L. (2023) · Journal of Peace Research · doi:10.1177/00223433221128846 · Cited by 23 · see citing works
“Even though the inclusion of fixed effects is a common strategy to capture unobserved time-invariant confounders, recent research suggests that this strategy might be insufficient to identify causal effects in dynamic settings (Imai & Kim, 2021).”
BackgroundElectric vehicle adoption and local PM2.5 reduction: Evidence from China
Wu, Y., Гу, Ф., Ji, Y., Ma, S.-C., & Guo, J. (2023) · Journal of Cleaner Production · doi:10.1016/j.jclepro.2023.136508 · Cited by 24 · see citing works
Air pollution and internal migration: evidence from an Iranian household survey
Farzanegan, M. R., Gholipour, H. F., & Javadian, M. (2022) · Empirical Economics · doi:10.1007/s00181-022-02253-1 · Cited by 22 · see citing works
Morales, L. F., Bonilla‐Mejía, L., Pulido, J. D., Flórez, L. A., Hermida, D., Mahecha, K. L. P., & Lasso-Valderrama, F. J. (2022) · Canadian Journal of Economics/Revue canadienne d économique · doi:10.1111/caje.12549 · Cited by 22 · see citing works
Inviting oversight: Effects of forest certification on deforestation in the Brazilian Amazon
Rana, P., & Sills, E. O. (2023) · World Development · doi:10.1016/j.worlddev.2023.106418 · Cited by 21 · see citing works
Pipeline or Pipedream: Gender Balance Legislation’s Effect on Women’s Presence in State Government
McQueen, S. (2021) · State Politics & Policy Quarterly · doi:10.1017/spq.2020.8 · Cited by 20 · see citing works
Exploring EU’s Regional Potential in Low-Carbon Technologies
Bergamini, E., & Zachmann, G. (2020) · Sustainability · doi:10.3390/su13010032 · Cited by 20 · see citing works
The short and long-run effects of international environmental agreements on trade
Ederington, J., Paraschiv, M., & Zanardi, M. (2022) · Journal of International Economics · doi:10.1016/j.jinteco.2022.103685 · Cited by 19 · see citing works
Politicians’ Private Sector Jobs and Parliamentary Behavior
Weschle, S. (2022) · American Journal of Political Science · doi:10.1111/ajps.12721 · Cited by 19 · see citing works
Sun, Y., Zhang, J., Han, J., & Zhang, Q. (2023) · Business Ethics the Environment & Responsibility · doi:10.1111/beer.12564 · Cited by 19 · see citing works
Accrual and real earnings management by rumored takeover targets
Davis, F. J., & Khadivar, H. (2024) · International Review of Financial Analysis · doi:10.1016/j.irfa.2024.103105 · Cited by 19 · see citing works
On Inference When Using State Corporate Laws for IdentificationPreprint
Spamann, H. (2019) · SSRN Electronic Journal · doi:10.2139/ssrn.3499101 · Cited by 17 · see citing works
Voice Analytics of Online InfluencersPreprint
Hwang, S., Liu, X., & Srinivasan, K. (2021) · SSRN Electronic Journal · doi:10.2139/ssrn.3773825 · Cited by 17 · see citing works
Causal Models for Longitudinal and Panel Data: A Survey
Arkhangelsky, D., & Imbens, G. W. (2023) · National Bureau of Economic Research · doi:10.3386/w31942 · Cited by 20 · see citing works
Newman, A. L., & Zhang, Q. (2023) · Review of International Political Economy · doi:10.1080/09692290.2023.2267051 · Cited by 20 · see citing works
“The TWFE model, with two groups and two periods (pre-and during sanctions), is equivalent to difference-in-differences (DiD) estimators (de Chaisemartin & D’Haultfoeuille, 2022; Imai & Kim, 2021).”
MethodZhang, L., & Wen, S. (2024) · Resources Policy · doi:10.1016/j.resourpol.2024.105091 · Cited by 18 · see citing works
- Authors
- Imai, K., & Kim, I. S.
- Affiliations
- Harvard University (US)
- Massachusetts Institute of Technology (US)
- Corresponding author
- Kim, In Song
- Venue
- Political Analysis
- Publisher
- Cambridge University Press (CUP)
- Published
- 2020-11-12
- Volume
- 29
- Issue
- 3
- Pages
- 405-415
- ISSN
- 1047-1987 (ISSN-L)
- 1476-4989
- OpenAlex
- W3098174737
- Semantic Scholar
- 660ac5f400de68ad64a9f31370bc59a30ec98d22
- Cited by
- 623
- References
- 20
- Citations by year
- 2026: 105 · 2025: 142 · 2024: 126 · 2023: 92 · 2022: 68 · 2021: 54 · 2020: 30 · 2019: 3 · 2018: 2 · 2017: 1
- FWCI
- 42.96
- Type
- journal article
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- Political Analysis (publishedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by-nc-sa)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
- Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Spatial and Panel Data Analysis
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Keywords
- causal inference, panel data, difference-in-differences, unmeasured confounding, negative weights, design-based estimation, model assumptions
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-11.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-11.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
