Bonin v. Sabine River Authority: Causation as a Threshold Requirement for Temporary Flooding Takings Claims (and No “Resurrection” of Stricken Expert Evidence)

1. Introduction

Parties. More than 700 downriver landowners (“Plaintiffs”) sued the Sabine River Authority of Texas (“SRA-T”) and the Sabine River Authority, State of Louisiana (“SRA-L”) (together, “the Authorities”), the co-operators of the Toledo Bend Dam and Reservoir on the Sabine River.

Factual setting. During a record storm in March 2016, the Toledo Bend Reservoir received extreme inflows described as exceeding a “500-year flood.” To prevent “overtopping” and potential dam failure, the Authorities opened nine spillways, releasing large flows into the Sabine River for weeks. Downriver flooding damaged Plaintiffs’ properties.

Claims and posture. Plaintiffs brought a Fifth Amendment Takings Clause claim via 42 U.S.C. § 1983, seeking compensation for flood damage. After earlier sovereign-immunity litigation (including Bonin v. Sabine River Auth., 65 F.4th 249 (5th Cir. 2023)), the district court granted summary judgment for Defendants. On appeal, the Fifth Circuit affirmed.

Key issues on appeal. (1) Whether the district court properly struck Plaintiffs’ untimely expert affidavits (and refused reconsideration); and (2) Whether Plaintiffs produced sufficient evidence—especially on causation—to survive summary judgment on a temporary-flooding takings theory.

2. Summary of the Opinion

The Fifth Circuit affirmed on two principal grounds:

  1. Expert affidavits properly excluded. Plaintiffs’ expert affidavits (Purcell and Knack) were untimely under the scheduling order, had been struck by a magistrate judge, and were later re-filed without permission. The district court acted within its discretion in striking them again; allowing them would severely prejudice Defendants by reopening expert work late in a long-running case.
  2. Summary judgment affirmed for lack of causation evidence. Applying guidance from Arkansas Game & Fish Comm'n v. United States, 568 U.S. 23 (2012), the court held that causation is a necessary prerequisite for government-induced temporary flooding takings claims. The Fifth Circuit adopted the Federal Circuit’s causation inquiry for dam cases: whether operation of the dam caused additional flooding above what would have occurred from the storm absent the dam. Plaintiffs’ surviving expert evidence was too conclusory and too dependent on excluded material to create a genuine fact dispute.

    3. Analysis

    3.1. Precedents Cited

    A. Expert exclusion and discovery-sanction framework

    • In re Complaint of C.F. Bean L.L.C., 841 F.3d 365 (5th Cir. 2016): supplied the standard of review (abuse of discretion) and reiterated the four-factor test for excluding untimely expert evidence as a discovery sanction (drawing from Geiserman).
    • Sierra Club, Lone Star Chapter v. Cedar Point Oil Co., 73 F.3d 546 (5th Cir. 1996), quoting O'Malley v. U.S. Fid. and Guar. Co., 776 F.2d 494 (5th Cir. 1985): underscored the deference appellate courts give trial courts on evidentiary/expert-admissibility management.
    • Geiserman v. MacDonald, 893 F.2d 787 (5th Cir. 1990): provided the four-factor test applied to late-disclosed expert testimony: (1) explanation; (2) importance; (3) prejudice; (4) continuance availability.
    • Fed. R. Civ. P. 72(a): used to emphasize that a party cannot assign error to a magistrate judge’s nondispositive order without timely objection—important here because Plaintiffs did not object to the order striking the affidavits and later attempted to re-file them.
    • F.D.I.C. v. Conner, 20 F.3d 1376 (5th Cir. 1994), quoting Hastings v. North E. Indep. Sch. Dist., 615 F.2d 628 (5th Cir. 1980): recognized that appellate review of sanctions can “necessarily include” review of the underlying discovery order—while the panel noted potential waiver, it affirmed without resolving waiver definitively.
    • United States v. Olguin, 643 F.3d 384 (5th Cir. 2011): applied to reject an inadequately supported argument that SRA-L could not benefit from striking affidavits because it had not answered when the magistrate first struck them.

    B. Summary judgment standards

    • Rogers v. Bromac Title Serv., L.L.C., 755 F.3d 347 (5th Cir. 2014): confirmed de novo review of summary judgment.
    • Fed. R. Civ. P. 56(a): articulated the governing “no genuine dispute of material fact” standard.

    C. Takings doctrine and temporary flooding—causation, foreseeability, and the “no-dam” baseline

    • Arkansas Game & Fish Comm'n v. United States, 568 U.S. 23 (2012): the central Supreme Court decision rejecting a categorical bar against takings liability for temporary flooding and supplying situation-specific considerations. The Fifth Circuit treated Arkansas Game as also illustrating that lack of causation is independently dispositive—citing its discussion of Sanguinetti.
    • Armstrong v. United States, 364 U.S. 40 (1960): provided the classic fairness rationale for the Takings Clause (public burdens should be borne by the public as a whole).
    • Sanguinetti v. United States, 264 U.S. 146 (1924): used in Arkansas Game (and here) to emphasize “foreseeability and causation”; the landowner’s failure to prove a causal connection defeated the claim.
    • Orr v. United States, 166 Fed. Cl. 1 (2023): cited for a five-factor formulation that includes causation and, importantly, for its holding that failure to prove causation prevents success under Arkansas Game.
    • St. Bernard Par. Gov't v. United States, 887 F.3d 1354 (Fed. Cir. 2018): supplied the operative causation baseline in dam/flooding contexts: whether government action caused flooding beyond what would have occurred anyway; if not, no taking.
    • United States v. Sponenbarger, 308 U.S. 256 (1939): reinforced the principle that government is not liable for a taking when flooding occurs “despite—not because of—the Government’s best efforts.”

    D. Sufficiency of expert proof at summary judgment

    • Lawrence v. Fed. Home Loan Mortg. Corp., 808 F.3d 670 (5th Cir. 2015), quoting Likens v. Hartford Life & Accident Ins. Co., 688 F.3d 197 (5th Cir. 2012): used to reject summary-judgment “evidence” built on speculation or unsubstantiated assertions.
    • Michaels v. Avitech, Inc., 202 F.3d 746 (5th Cir. 2000), quoting Anderson v. Liberty Lobby, Inc., 477 U.S. 242 (1986): permitted the court, even without a full appellate Daubert analysis, to recognize that expert evidence lacking “rational probative value” cannot create a genuine dispute; also warned that multiple equally plausible inferences can render proof not “significantly probative.”
    • Orthopedic & Sports Injury Clinic v. Wang Lab., Inc., 922 F.2d 220 (5th Cir. 1991): supplied the “level of conclusoriness” floor for affidavits used to defeat summary judgment.

    3.2. Legal Reasoning

    A. The “resurrection” problem: once struck, expert evidence does not come back by self-help

    The panel treated the procedural sequence as decisive. Plaintiffs’ affidavits were struck as untimely; Plaintiffs did not object; almost two years later they attached the same affidavits to a summary-judgment response without requesting leave or a scheduling-order modification. The Fifth Circuit viewed this as violating both the original scheduling order and the unobjected-to magistrate order.

    Applying the Geiserman v. MacDonald factors through In re Complaint of C.F. Bean L.L.C., the court emphasized:

    • Explanation: Plaintiffs offered no acceptable explanation for waiting and then re-submitting without permission.
    • Prejudice and case management: Allowing the affidavits “at the final hour” would force Defendants to redo expert work and briefing in an already years-old case—“severe prejudice.”
    • Continuance: While a continuance might theoretically cure prejudice, granting one would prolong the case further—undercutting orderly docket control.

    The opinion also signals an institutional concern: parties cannot sidestep Rule 72(a) (and waiver risks) by re-filing previously stricken material in hopes of a later, more favorable ruling.

    B. The substantive pivot: causation as a threshold element in temporary-flooding takings claims

    The Fifth Circuit characterized its principal doctrinal move as the first time it applied Arkansas Game & Fish Comm'n v. United States guidance. Although Arkansas Game listed several considerations (duration, intent/foreseeability, investment-backed expectations, severity), the Fifth Circuit read the decision’s discussion of Sanguinetti v. United States as establishing that causation is a prerequisite—and that failure to prove it is “more than sufficient to dispose” of the claim.

    To operationalize causation in the dam setting, the court adopted the Federal Circuit’s baseline inquiry (as framed in Orr v. United States and St. Bernard Par. Gov't v. United States): whether operation of the dam caused additional flooding above what the storm would have produced absent the dam. This is effectively a “but-for plus increment” test—requiring a counterfactual comparison to the no-dam scenario.

    C. Why Plaintiffs’ proof failed: dependence on excluded modeling and a missing bridge from history to the 2016 event

    On summary judgment, Defendants offered expert evidence (Dr. Simons) that the reservoir stored substantial stormwater and mitigated peak flows; Defendants also introduced a FERC review indicating significant downstream tributary contributions.

    Plaintiffs’ admissible proof centered on the “Rimkus Report” and limited deposition excerpts. The court concluded the report could not create a genuine fact dispute on causation for two related reasons:

    1. Excluded “no-dam” flow figure (75,000 cfs) was the keystone. The report’s chain of reasoning depended on a 75,000 cfs no-dam estimate originating in a master’s thesis. Once the portion of the report “parroting” that thesis was struck, the court found the remaining conclusion—that flooding was “more extensive” with the dam—became unsubstantiated.
    2. Historical pre-dam peak discharges did not establish the counterfactual flooding level for March 2016. Plaintiffs cited historic station data (e.g., 90,600 cfs in 1945 and 61,000 cfs in 1957 at Logansport; 52,000–53,000 cfs downstream in 1957) as a range for a no-dam scenario. But the court held Plaintiffs did not connect those numbers to predicted 2016 inundation levels, nor rule out competing inferences (differences in storm location, duration, watershed conditions). Under Michaels v. Avitech, Inc. and Anderson v. Liberty Lobby, Inc., the evidence lacked sufficient probative value and left Plaintiffs’ inference as only one among many equally plausible ones.

    The panel acknowledged the inherent difficulty of proving a counterfactual (“no-dam scenario”) yet insisted that difficulty does not lower the evidentiary floor at summary judgment: a takings claimant must still produce nonconclusory, methodologically supported proof that government action caused incremental flooding.

    D. Necessity doctrine left undecided on appeal

    The district court also relied on the necessity doctrine (emergency response) as an alternative ground. The Fifth Circuit affirmed without reaching it because causation failure was dispositive. As a result, the opinion’s enduring doctrinal contribution lies in its causation holding and its strict approach to post-sanction expert practice, not in elaborating the necessity defense.

    3.3. Impact

    • Fifth Circuit alignment with Federal Circuit flooding-causation doctrine. By adopting the “additional flooding above no-dam conditions” inquiry, the Fifth Circuit provides a clearer, more defendant-protective threshold screen in dam-operation flood cases. Plaintiffs must present credible hydrologic/hydraulic counterfactuals rather than general assertions that releases worsened flooding.
    • Reframing Arkansas Game in this circuit. The court treats causation as analytically prior to the Arkansas Game factors. Practically, litigants should expect courts to demand causation proof before engaging the more holistic, multi-factor takings inquiry.
    • Litigation-management signal on expert evidence. The opinion discourages “self-help” reintroduction of excluded expert materials and underscores the necessity of timely objections (Rule 72(a)) and timely motions to modify scheduling orders.
    • Strategic implications for plaintiffs and defendants. Plaintiffs will likely need robust modeling, transparent assumptions, and sensitivity analyses to show incremental causation (and to survive Daubert and summary judgment). Defendants, conversely, can focus on (i) storage/mitigation evidence, (ii) downstream tributary contributions, and (iii) challenges to counterfactual models as speculative or improperly sourced.

    4. Complex Concepts Simplified

    Takings Clause (Fifth Amendment)
    The Constitution requires the government to pay “just compensation” when it takes private property for public use. A “taking” can be physical (e.g., repeated flooding) or regulatory, but the claimant must prove the government’s action actually caused the invasion of the property interest.
    Temporary flooding takings
    Under Arkansas Game & Fish Comm'n v. United States, temporary flooding is not categorically exempt from takings liability. But courts examine the facts closely—especially causation, duration, foreseeability, severity, and expectations.
    Causation in dam-release cases (“no-dam” counterfactual)
    The question is not merely whether the dam released water. The question is whether operating the dam caused more flooding than would have occurred from the same storm if the dam did not exist. That requires credible evidence of the counterfactual baseline.
    Summary judgment
    A case can be decided without trial if the nonmoving party lacks evidence to create a real dispute over facts that matter to the legal claim. Conclusory expert assertions and unsupported inference chains do not suffice.
    Striking expert evidence
    Courts can exclude late or improper expert submissions as a sanction for violating scheduling and discovery orders. If a magistrate judge strikes evidence and no timely objection is filed under Rule 72(a), parties generally cannot later treat the issue as preserved by simply re-filing the same material.
    Discharge / flow (cfs)
    River “discharge” is the volume of water moving past a point per unit time, commonly measured in cubic feet per second (cfs). Discharge estimates are often inputs to modeling inundation (flood extent and depth).

    5. Conclusion

    Bonin v. Sabine River Authority makes two practical contributions to Fifth Circuit law. First, it enforces strict procedural discipline: expert affidavits struck as untimely cannot be reintroduced without permission, and doing so late in litigation can be excluded as highly prejudicial. Second—and more significantly for takings doctrine—it treats causation as a threshold requirement in temporary flooding takings claims and adopts the Federal Circuit’s “additional flooding above the no-dam scenario” inquiry for dam operations. The result is a clear roadmap: without rigorous, admissible proof that government action caused incremental flooding, courts need not reach the more nuanced Arkansas Game factors, and summary judgment will follow.