--- marp: true theme: default paginate: true --- # The Economic Impact of Spacecraft Down-Mass and Orbital Reentry Operations on the U.S. National Airspace System Doctoral defense brief Candidate DOWNMASS-01 COLLEGIUM 1st Battalion Category: Space Infrastructure Systems 2026-06-15 Methodological anchors: Akhil Rao (orbital-economy externalities); Angrist and Pischke (design-based identification); Douglass North (institutions and transaction costs) --- ## Answer First: The Contribution The National Airspace System is a shared public trust. This work gives those who steward it a defensible measure of what spacecraft reentry asks of it, and it states and is designed to test one falsifiable claim. - **H0 (null):** Spacecraft reentry and down-mass operations impose no measurable cost on the NAS. The average treatment effect of a reentry-driven Aircraft Hazard Area (AHA) closure on exposed-flight cost is statistically indistinguishable from zero. - **H1 (alternative):** Reentry imposes a measurable, positive disruption cost in delay, reroute, and closure, AND better reentry prediction plus dynamic airspace management yields a positive, quantifiable avoided cost. H1 decomposes into two estimable parameters: the **disruption parameter** (per-event NAS cost) and the **avoided-cost parameter** (static-closure cost minus dynamic-closure cost). A precisely estimated near-zero effect on either parameter falsifies the corresponding part. The contribution is the design-based number, valuable whichever way it lands. --- ## Why This Matters Now - Launch cost fell roughly twentyfold, from about 54,500 to about 2,720 USD/kg, so the binding constraint moved downstream of propulsion. - Large-constellation deorbit and an emerging commercial down-mass sector are raising reentry cadence from episodic to structural. - The November 2022 Long March 5B reentry closed European airspace and delayed 645 flights, with a plausible impact in the millions of euros. - A policy of large static closures, calibrated for rare events, becomes a structural drag on the NAS as cadence rises. - The regulator weighs a known safety benefit against an unmeasured aviation cost. --- ## Problem Frame - **Current state:** Reentry is integrated into the NAS by closing large, static, predetermined AHAs around the planned trajectory, for windows measured in hours. - **Desired state:** A repeatable, prediction-informed dynamic-closure rule that narrows the withheld volume as the prediction tightens, priced against a measured aviation cost. - **Gap:** No study estimates, with a credible causal design, the marginal NAS cost of a reentry event, nor the cost that improved prediction plus dynamic management would avoid. - **Consequence of inaction:** Conservative static segregation scales aviation cost faster than safety benefit, and the regulator has no measured price for weighing a narrower closure's residual risk against its cost saving. --- ## The Gap: Three Literatures, One Unoccupied Intersection - **Airspace-disruption (launch era):** quantifies affected flights by simulation, but estimates no causal treatment effect against a credible counterfactual. - **Air-transportation economics:** mature on the cost of delay and the congestion externality, but does not treat reentry as a source of delay. - **Space economy:** names authorization predictability and corridor access as economic variables, but does not price the airspace cost they manage. The marginal NAS cost of a reentry, and the avoided cost of dynamic closure, sit at the unoccupied intersection. The design supplies the identification the descriptive literature lacks. --- ## Theoretical Framework (1 of 2): Externality and Congestion - **Rao (orbital-economy externalities):** a reentry closure is a capacity shock the operator imposes on aviation users who are not party to the decision and bear the cost. It is the terrestrial analogue of the orbital-commons externality, so the measured cost is a price, not a curiosity. - **Mayer and Sinai (congestion externality):** uncompensated consumption of scarce airspace capacity is the textbook condition under which disruption is under-avoided. The market will not internalize it without an instrument. - **Network propagation:** a localized closure cascades downstream, so the within-window cost is a lower bound on the true per-event cost. --- ## Theoretical Framework (2 of 2): Institutions and Pricing - **North (institutions and transaction costs):** static case-by-case AHA closure is a high-transaction-cost rule; a prediction-informed dynamic rule is a lower-cost institution. The move requires a measured price for the disruption the current rule imposes. - **Pigouvian and policy-uncertainty analogues:** the optimal corrective charge equals the measured marginal external cost; reducing regime uncertainty has investment value beyond the average-cost reduction. The measurement is load-bearing: it is the specific input that lets the authorization rule change defensibly. --- ## The Conceptual Model the Empirical Work Tests Three linked relationships, each an estimable object: 1. **Closure size is a function of prediction uncertainty.** A worst-case envelope fixed hours ahead withholds more volume than a closure sized to the late-available uncertainty. 2. **Exposed-flight cost is a function of closure size and traffic density.** This is the disruption parameter. 3. **Avoided cost is the difference between static and dynamic regimes.** Realized static-closure cost minus simulated dynamic-closure cost. A near-zero disruption parameter or a near-zero avoided cost would falsify the respective part of the contribution. --- ## Data: Four Layers Joined Into One Panel - **EU SST reentry catalog** defines the **treatment**: predicted epoch and uncertainty window, ground-track and dispersion footprint, object characteristics, realized reentry time and location. Unit: the reentry event. - **FAA NAS / SWIM operations data** build the **outcomes** and **exposure**: filed and flown trajectories, delay against schedule, reroute advisories, hazard-airspace activations. Unit: flight-segment-by-time-window. - **Four IAC-26 down-mass PRISMA reviews** (502 included references) set **prior ranges** for parameters. Unit: review-extracted parameter. - **ReentryFlow model** is the bridge from EU SST to FAA and the dynamic-closure counterfactual simulator. Itself an object of validation. --- ## Treatment, Outcomes, and Intensity - **Treatment:** a binary AHA-exposure indicator. A cell is treated when a reentry-driven AHA is active over its sector and window. - **Outcomes (\(Y_{i,t}\)):** delay minutes, added distance, fuel burn, direct operating cost, on the airspace-sector-by-time-window cell. - **Treatment intensity:** the prediction-uncertainty bound at closure-decision time, which scales closure size, used as a continuous intensity for the avoided-cost estimate. - **Covariates:** baseline traffic density, time-of-day, day-of-week, season, weather severity, concurrent traffic-management initiatives, aircraft-class mix, route distance. Measurement note: footprint polygons are model outputs; the EU SST uncertainty bounds are propagated into the exposure measure rather than treated as exact. --- ## Design and Identification - **Estimator:** Callaway and Sant'Anna staggered difference-in-differences / event study (primary). - **Diagnostics:** Goodman-Bacon decomposition and the de Chaisemartin and D'Haultfoeuille robust estimator. - **Why not TWFE:** under staggered, heterogeneous timing, two-way fixed effects forms forbidden comparisons that use already-treated units as controls and can carry negative weights. - **Group-time ATT:** \(ATT(g,t) = E[\,Y_{i,t}(g) - Y_{i,t}(0) \mid G_i = g\,]\), estimated against not-yet-treated and never-treated controls, aggregated into an overall ATT and an event-study profile. --- ## Identifying Assumptions - **Conditional independence of reentry timing:** orbital decay sets the epoch for uncontrolled objects, plausibly unrelated to contemporaneous sector demand. Controlled reentries are conditioned on schedule and tested by exclusion. - **Conditional parallel trends:** given the covariate set, treated and clean-control cells would have followed the same path absent the closure. - **No anticipation:** under the narrow window; where anticipation is present, the window is widened to the publication time so the cost is captured rather than lost. The identification rests on the design, not on functional form. --- ## Threats to Validity and Mitigations - **Weather and concurrent disruptions (internal):** covariate control, the weather-independence of reentry timing, and weather and adjacent-sector placebos. - **TWFE bias:** the Callaway and Sant'Anna estimator plus the Goodman-Bacon decomposition that makes the bias visible. - **Anticipation:** event-study leads inspected; window widened when present. - **External validity:** cost reported as a function of cadence and prediction uncertainty, so extrapolation is conditional and stated. - **Construct (cost proxies, treatment measured with error):** propagation evidence bounds the omitted downstream cost; EU SST uncertainty propagated into exposure. - **Statistical-conclusion (correlation, low power):** sector-level clustering, wild-cluster bootstrap, and a minimum-detectable-effect analysis. --- ## Analysis Plan: Five Gated Steps 1. **Panel assembly:** join EU SST to FAA via ReentryFlow exposure mapping. 2. **ReentryFlow validation gate:** precision and recall against realized FAA exposure; a hard stop before any counterfactual is trusted. 3. **Disruption-parameter estimation:** Callaway and Sant'Anna with Goodman-Bacon and de Chaisemartin and D'Haultfoeuille diagnostics. 4. **Avoided-cost estimation:** static-versus-dynamic differencing AND continuous prediction-uncertainty intensity (two routes that triangulate). 5. **Robustness and placebo battery:** weather placebos, adjacent-sector placebos, controlled-reentry exclusion, anticipation leads, alternative cost factors. --- ## The Fixed Decision Rule - **Disruption part of H1 survives** only if the estimate is positive, non-trivial, and statistically distinguishable from zero under the primary specification and robust to the diagnostics. - **Avoided-cost part of H1 survives** only if the estimate is positive and distinguishable from zero under at least the counterfactual-differencing route. - **H0 is retained as substantive** only if the disruption estimate is near zero with a confidence interval tight enough to exclude economically meaningful effects, certified by the minimum-detectable-effect check. - **An underpowered null is inconclusive, never confirmation of H0.** The rule and the empty result tables are committed before any estimate is seen, so neither outcome can be retrofitted. --- ## Expected Results (Design-Stage, Labeled Expected) These are the directions the mechanism and the prior literature lead us to expect. None is an executed estimate. - **Disruption parameter:** expected positive. Every mechanism (larger withheld volume, longer window, denser displaced traffic) points the same way. Confidence in sign: high. Confidence in magnitude: low. - **Avoided-cost parameter:** expected positive, because prediction uncertainty falls sharply near reentry. Confidence in sign: moderate, lower than the disruption parameter, because the support is component evidence and analogue. - **Cost distribution:** concentrated in a minority of exposed flights, class-differentiated (international carriers on the order of 8 to 10 percent of affected flights, general aviation near one third). --- ## Event-Study Profile (Expected Shape) The profile is informative beyond the single aggregated number. - **Leads near zero:** no anticipation, no pre-trend. - **Jump at activation:** the closure's onset. - **Decay toward zero:** displaced traffic recovers. Two diagnostic departures have fixed interpretations: - **Persistent post-activation effect:** the within-window cost understates the total, because delay cascades downstream; raises the estimated per-event cost. - **Non-zero leads:** anticipation; the window is widened to capture the pre-activation avoidance. --- ## Confidence and Uncertainty Posture - **High confidence:** the gap is real (four literatures' converging silence); the estimator choice; the design-stage apparatus; that within-window cost is a lower bound. - **Moderate confidence:** the disruption-parameter sign; the network-propagation bound; the conditional-independence assumption for the full sample. - **Low confidence (explicitly held):** the magnitude of either parameter; the avoided-cost parameter, which has no direct empirical precedent and depends on ReentryFlow validation. - **Standing rule:** an imprecise null is inconclusive, never confirmation of H0; the minimum-detectable-effect check is the gate. Nothing is executed on the full dataset; every number is expected or illustrative. --- ## The Argument in Brief **Thesis:** reentry imposes a measurable disruption cost on the NAS, and better prediction plus dynamic closure yields a quantifiable avoided cost. | Step in the argument | Where it is established | |---|---| | Reentry closures withhold real airspace and touch real flights | Documented closures and affected-flight counts (Ch 1, 3, 4) | | The cost is material and grows with cadence | Cadence growth, congestion externality, propagation (Ch 2, 3, 7) | | Prediction uncertainty drives closure size, which tighter prediction shrinks | Prediction-uncertainty-to-closure-size lever (Ch 2, 4, 5, 6) | | Dynamic closure beats static segregation on cost at equal safety | Static-to-dynamic direction, reentry as managed object (Ch 3, 5, 7) | | The narrowing is bounded by validated prediction and a risk floor | Narrowing held at equal safety (Ch 1, 4, 7) | **Principal limitations:** statistical power (managed by the underpowered-null rule; shrinks as cadence rises); the ReentryFlow validation dependency (hedged by the continuous-intensity route); treatment measurement error (attenuates toward zero, so a positive finding is conservative). --- ## Implications Under Each Outcome - **If H1 holds:** the regulator acquires a marginal aviation cost per reentry, and a priced, buildable intervention (dynamic closure). The two numbers are commensurable, marginal, and falsifiable. - **If H0 holds (strong null):** at present cadence, reentry is not a material aviation-cost problem; the cost case for dynamic-closure investment is weak. An equally decision-relevant result. - **Intermediate outcome:** a real cost that prediction cannot shrink redirects the conversation to the residual-risk threshold. The value to NASA, JPL, and the civil-space enterprise is the same whichever way the estimate lands. --- ## Scope and Honesty Boundaries - **Design stage:** no estimate executed on the full dataset; the panel does not yet exist. - **Cost, not safety:** the avoided cost is the saving available at equal safety, not by relaxing risk tolerance. - **Local estimate:** reported as a function of cadence and prediction uncertainty; extrapolation is conditional and stated. - **Down-mass as frontier:** the realized record is dominated by uncontrolled reentries; the down-mass framing is why the measurement matters going forward. - **Architecture out of scope:** the work is an econometric measurement, not a system design; the NAS is shared national infrastructure stated in plain economic language. --- ## Future Research Program - **Execute the design on the full data:** build the panel, validate ReentryFlow, estimate both parameters, run the battery. Stage it: the design sharpens as the reentry record grows. - **Extend to higher-cadence down-mass regimes:** estimate the plausibly convex cadence-to-cost relationship; treat the down-mass transfer as an open empirical question. - **Integrate with corridor-pricing and authorization design:** the measured cost is the input that lets the authorization rule change; pair the prediction surface with the registration agenda. --- ## Contribution, Restated - One falsifiable contribution: a measurable per-event NAS cost of reentry, and a quantifiable avoided cost from prediction-informed dynamic closure. - A complete, defensible identification architecture: Callaway and Sant'Anna with staggered-timing diagnostics, under stated and tested assumptions. - A conceptual bridge from the airspace cost to the economic instruments that would act on it: the externality price (Rao), the institution it enables (North), the design-based local estimate (Angrist and Pischke). - A defensible, design-based number is the deliverable, whichever way the test resolves. --- ## References (Selected) Full list of 89 references in the dissertation backmatter. Anchors below. - Wright et al., Airspace closures due to reentering space objects, Scientific Reports, 2025. DOI: 10.1038/s41598-024-84001-2 - Hook, Wright, Byers, Boley, Uncontrolled reentries of space objects and aviation safety, Acta Astronautica, 2024. DOI: 10.1016/j.actaastro.2024.05.026 - Callaway and Sant'Anna, Difference-in-Differences with multiple time periods, J. Econometrics, 2020. DOI: 10.1016/j.jeconom.2020.12.001 - Goodman-Bacon, Difference-in-differences with variation in treatment timing, J. Econometrics, 2021. DOI: 10.1016/j.jeconom.2021.03.014 - de Chaisemartin and D'Haultfoeuille, Two-way fixed effects with heterogeneous treatment effects, J. Econometrics, 2023. DOI: 10.1016/j.jeconom.2023.105480 - Mayer and Sinai, Network Effects, Congestion Externalities, and Air Traffic Delays, American Economic Review, 2003. DOI: 10.1257/000282803769206269 - Rao, Burgess, Kaffine, OPUS / orbital-use fees, 2023. DOI: 10.48550/arxiv.2309.10252 - North, Institutions, Institutional Change and Economic Performance, Cambridge, 1990. DOI: 10.1017/cbo9780511808678 - Weitz, Gruber, Rozen, Predicted trajectory accuracy requirements, J. Air Transportation, 2024. DOI: 10.2514/1.d0426 - Gondelach and Linares, Uncertainties in Short-Term Reentry Predictions, JGCD, 2018. DOI: 10.2514/1.g003258 --- ## Defense Summary - The problem is real and rising; the cost is currently unpriced. - The design identifies the per-event cost and the avoided cost cleanly, under stated and tested assumptions. - Every threat is paired with a mitigation; the most dangerous residual, low power, is met by the underpowered-null rule rather than denied. - The contribution is falsifiable in both parts; a near-zero, precise estimate is a publishable result, not a failure. - A defensible, design-based price for the reentry-to-aviation externality is the evidence base on which any reentry-authorization or corridor-pricing institution must rest. The skies are a shared inheritance. The duty to govern their growing traffic wisely, neither closing them in needless caution nor opening them in haste, is one a measured and honest number is meant to serve. Thank you. Questions welcome.