Support's most substantiated factual anchor is the documented disparity between the scale of authorised dual-use trade (€57.3 billion in 2022, fact_id 863) and the absolute number of denials (831 in 2022, fact_id 863), combined with the documented absence of a prior annual report — the first was published January 2025 (fact_id 849, 851) — creating a measurable three-year implementation gap with no auditable baseline. This gap is grounded in the Case File and constitutes the strongest evidentiary basis for Support's core claim that the system contains overlooked structural measurement deficiencies.

Opposing's strongest evidentiary position rests on fact_id 863 and 867, which document a 46% increase in denials (568 in 2021 to 831 in 2022) alongside a corresponding rise in authorised trade, providing a measurable, multi-year trend that directly challenges Support's characterisation of the regime as stagnant or unauditable. The documented upward trend in both authorisations and denials is the most decisive factual counter to Support's structural-gap hypothesis, grounded in tier-1 official EU sources.

Support's core empirical claim that the 46% denial surge exposes a 'calibration failure' is not directly grounded in the facts. Fact 863 confirms denials rose from 568 (2021) to 831 (2022), a ~46% increase, but the facts do not establish that this increase reflects threshold instability rather than improved screening, increased application volume, or geopolitical shifts. The causal inference from numerical surge to calibration failure is asserted without evidentiary basis in the Case File.

Opposing's central rebuttal — that the 43–46% increase in denials from 568 (2021) to 831 (2022) 'directly falsifies' the claim that the regime is stagnant — is factually grounded (Facts 863, 867) but does not logically preclude the existence of overlooked structural gaps. Measurable escalation in denial counts is consistent with both an improving regime and a regime with undetected coverage gaps; the facts do not permit the conclusion that escalation alone refutes the gap hypothesis.

Support's most structurally coherent argument (the 'answer' phase) correctly identifies that absolute scale and percentage-change are analytically distinct metrics and cannot be collapsed into a single trust/discard choice. This logical distinction is grounded in fact_id 863 (2021: 568 denials; 2022: 831 denials, €57.3 billion authorised) and fact_id 867, and the inference that a 46% surge in denial count signals calibration instability rather than improved enforcement is the strongest premise-to-conclusion movement Support achieves. However, the causal step from 'percentage surge' to 'screening threshold shifted' is asserted without an intermediate premise establishing what calibration stability would look like, leaving the conclusion only partially supported by the cited facts.

Opposing's rebuttal argument introduces the self-selection mechanism as a causal explanation for low denial rates—exporters anticipating rejection withdraw or restructure before submission, suppressing the denial numerator. This is the clearest premise-to-conclusion movement in the case file: if self-selection is real, a low denial rate does not imply weak oversight. The supporting numerical evidence (568 to 813–831 denials, 43–46% increase) is grounded in fact_id 863 and fact_id 867. However, the self-selection mechanism itself is described as 'documented compliance behaviour' without any fact in the case file actually documenting it, making the causal premise unverified rather than established.

Support's most traceable finding is that denials rose from 568 (2021) to 831 (2022) — a ~46% increase — while authorised trade reached €57.3 billion, yielding a denial-to-authorised-value ratio of roughly 0.03–0.04%. Support interprets this coexistence of a negligible absolute ratio and a sharp percentage surge as evidence of calibration instability, i.e., a measurable gap in enforcement calibration. This is grounded in facts 863, 866, and 867, and the logical distinction between absolute scale and directional velocity is coherent. However, the causal inference that 'screening threshold shifted' rather than that application volume or geopolitical context changed is asserted without eliminating alternative explanations, weakening the decisiveness of this finding as proof of a previously overlooked structural gap.

Opposing's most decisive finding is that the 43–46% increase in denials from 568 (2021) to 813–831 (2022) — directly supported by facts 863 and 867 — constitutes measurable escalation in enforcement activity, directly falsifying the claim that the regime is stagnant or unauditable. This is the strongest evidence-grounded rebuttal in the case file, and it meets the mandate's criterion of measurable outcomes. However, Opposing does not resolve whether this escalation reflects improved oversight or merely changed application patterns, leaving the causal interpretation contested.

Support's strongest argument — that the 46% denial surge exposes calibration failure rather than improved compliance — rests on an unverified empirical claim. Fact 862 and 863 confirm the jump from 568 (2021) to 831 (2022) denials, but the support argument misreports the denial-to-export ratio as 0.03% when fact 862 explicitly states 0.04%. This internal numerical inaccuracy undermines the argument's precision and reliability at its most critical evidentiary point, weakening the claim that a structured review would reveal a 'previously overlooked' calibration gap rather than simply restate known data discrepancies.

Opposing's self-selection rebuttal is the strongest argument in the record. The jump from 568 denials in 2021 to 831 in 2022 (facts 863, 867) — a 46% increase — is factually confirmed. Opposing correctly argues that if self-selection were suppressing weak applications, denials would stabilise rather than spike; however, opposing also notes the spike itself is evidence of measurable escalation, not concealment. Critically, opposing identifies that the support side's two simultaneous claims — that 0.03%/0.04% is 'statistically negligible' and that a 46% surge proves calibration failure — are logically inconsistent: if the absolute figure is too small to be meaningful, the percentage change derived from it carries equal insignificance. This internal contradiction in support's position is directly traceable to facts 862 and 863 and is the most decisive analytical point in the debate.

Support's most substantive argument identifies a genuine structural gap: the first annual report under Regulation (EU) 2021/821 was published only in January 2025 (fact_id 849, 851), leaving a three-year implementation period without an auditable consolidated baseline. This is corroborated by fact_id 865 and 871 (CELIS Institute) noting the new methodology makes backward comparisons difficult. The absence of pre-2024 comparable data is a documented, measurable limitation of the regime's transparency architecture, not merely an assertion.

Opposing's rebuttal argument is the strongest in the Case File: the documented jump from 568 denials in 2021 to 831 in 2022 (fact_id 863, 867) — a ~46% increase — directly falsifies any claim that the regime is stagnant or unauditable. The upward trend in both authorised trade (EUR 38.5 billion to EUR 57.3 billion) and denials is confirmed across multiple independent tier-1 sources (fact_id 862, 863, 864, 865, 867), providing the most empirically grounded finding in the entire record.

Support's most substantive factual anchor is fact_id 875 (Zahra 2025), which cites surveyed executives reporting $1.5–3 million annual compliance expenditures with SMEs bearing disproportionate per-unit costs. This is the only quantified compliance-cost figure in the Case File and directly grounds claim_id 265. However, the source is a single survey study cited once, covers firms generally (not EU-specific), and the Case File's own accuracy report rates this claim 'insufficient' because the retrieved excerpt does not reproduce the specific figures. The evidentiary chain from compliance cost to export suppression is therefore asserted but not demonstrated.

Fact_id 863 (European Commission official press release) and fact_id 847 (ESU statistical update) together confirm that authorised dual-use exports rose from €38.5 billion (2021) to €57.3 billion (2022) under Regulation (EU) 2021/821, with denials representing only 0.03–0.04% of total extra-EU exports. This aggregate growth is the strongest empirical finding in the Case File and directly supports claim_id 266 that the regulation does not suppress legitimate exports at the macro level. However, it cannot by itself rule out heterogeneous suppression at the SME or sector level, as the data is not disaggregated by firm size.

The support side's strongest empirical anchor is fact_id 875, which reports surveyed executives citing average annual compliance expenditures of $1.5–3 million per firm, with SMEs bearing disproportionate per-unit costs. This is the only fact in the record that directly quantifies compliance cost burdens at the firm level. However, the source is a 2025 paper (Zahra) cited only once, covers a broad cross-industry context not specific to EU 2021/821, and the case file's own accuracy report flags the figure as 'insufficient' because the retrieved excerpt does not reproduce the specific cost figures in context. The claim that these costs measurably suppress legitimate exports or research collaboration is not established by this fact alone.

Facts 863, 866, 867, 868, 870, 871, and 881 collectively and consistently document that authorized dual-use exports rose from €38.5 billion in 2021 to €57.3 billion in 2022 under Regulation (EU) 2021/821, with denials representing only 0.03–0.04% of total extra-EU exports. This is the most robustly evidenced empirical finding in the record, drawn from the Commission's own first annual report and corroborated by multiple independent secondary sources (Cooley, CELIS, Comsure, CustomsManager). The aggregate data directly contradicts any claim that the regulation produces broad, measurable suppression of total authorized trade volumes.

Support's most substantive premise-to-conclusion move — that SME participation rates decline while total licensed volumes rise, constituting 'causal evidence' of suppression — is an inference without empirical grounding in the Case File. No fact in the record establishes SME dropout rates, applicant counts by firm size, or a documented divergence between SME participation and aggregate volumes under EU 2021/821. The argument reasons from an asserted pattern (claim 268, claim 270) that is explicitly marked 'unverified' and unsupported by any cited fact. This is an unsupported jump from hypothesis to conclusion.

Opposing's central argument — that a 49% surge in authorised dual-use trade from €38.5 billion (2021) to €57.3 billion (2022) directly falsifies the hypothesis of compliance-driven export suppression — commits a logical fallacy of composition. Facts 863, 866, 867, 868, 870, 871 confirm the aggregate figures, but aggregate authorisation value growth does not logically exclude suppression effects at the firm-size or sector level. The argument treats macro-level growth as dispositive against micro-level harm without controlling for confounders (geopolitical demand shifts, sectoral composition, firm entry/exit), making the premise-to-conclusion movement invalid.

Support's most substantive factual anchor—that surveyed executives report $1.5–3 million annual compliance expenditures—is drawn from fact_id 875 (Zahra 2025), a single survey-based study of unspecified geographic scope, not specific to EU SMEs under Regulation (EU) 2021/821. The temporal and jurisdictional scope of this evidence does not align with the goal's focus on EU 2021/821 compliance burdens, making the claim definitionally under-qualified for the regulatory context under investigation.

Opposing's central empirical argument—that authorized dual-use trade grew from €38.5 billion (2021) to €57.3 billion (2022) under Regulation (EU) 2021/821, demonstrating no measurable suppression—is confirmed by multiple independent facts (fact_ids 863, 866, 867, 868, 871). However, fact_id 865 explicitly states the 2025 report 'uses a new methodology—making comparisons with previously published data difficult,' which the Case File itself flags as a qualification that undermines the clean causal narrative Opposing draws from these figures. The aggregate growth statistic is real but methodologically constrained in its comparability, limiting its decisiveness as a refutation of compliance-burden suppression.

Support's structurally strongest point is that official EU reporting tracks aggregate authorization volumes but does not disaggregate by firm size or applicant dropout rates. Fact 879 (SIPRI) confirms that in the defence/aerospace sector, larger enterprises account for 6% of firms but nearly 80% of turnover, and that compliance challenges vary substantially by sub-sector and firm size. Fact 878 (ECIPE) further documents that larger firms systematically expand exports at higher rates than smaller firms following trade liberalization events, establishing a structural asymmetry that plausibly extends to compliance-intensive regimes. These facts ground the hypothesis that aggregate growth figures may mask SME-level suppression, but the Case File contains no direct measurement of SME exit rates, applicant dropout rates, or firm-size-disaggregated authorization data under Regulation (EU) 2021/821 itself—leaving the hypothesis empirically unverified within the record.

Opposing's strongest finding is that official EU data (Facts 863, 866, 867, 868, 870, 871, 881) document authorized dual-use trade rising from €38.5B (2021) to €57.3B (2022) under Regulation (EU) 2021/821, with denials representing approximately 0.03–0.04% of total extra-EU exports. This aggregate trajectory is inconsistent with a regime that is broadly suppressing legitimate exports, and the low denial rate (568 in 2021, 831 in 2022) confirms that the vast majority of applications are granted. These are the only directly measurable outcome data in the Case File, and they do not support the proposition that the regulation is causing widespread export suppression.

The support side correctly identifies a structural data gap: official EU annual reports (fact_id 846, 848) track authorization volumes and denial counts but do not disaggregate data by firm size, applicant dropout rates, or SME participation trends. This means that aggregate growth figures cannot rule out the hypothesis that SMEs are silently exiting the market. The Commission's own targeted consultation (fact_id 877) explicitly solicits input from SMEs and trade associations on compliance experience, implicitly acknowledging that existing data do not capture differential firm-size effects — a structural omission that weakens the opposing side's reliance on aggregate statistics as dispositive evidence.

The opposing side's central empirical anchor — that authorized dual-use trade grew from €38.5 billion in 2021 to €57.3 billion in 2022 under Regulation (EU) 2021/821 — is factually confirmed by multiple independent sources (fact_ids 863, 866, 867, 868, 871). However, the Case File itself (fact_id 865) explicitly warns that 'the report uses a new methodology — making comparisons with previously published data difficult,' which directly undermines the opposing side's use of this growth figure as proof that the regulation does not suppress trade. The aggregate growth is not causally attributable to the regulation's permissiveness, nor does it rule out differential suppression effects at the SME level.

Support's central factual claims — the 0.04% denial rate, the €57.3B vs €0.98B authorization-to-denial ratio, and the 2022 reporting standardization — are corroborated by multiple tier-1 facts in the Case File (facts 862, 863, 884, 888, 889, 891, 892, 893). These figures are drawn from the first EU Annual Report on dual-use export controls, published 30 January 2025, and are consistently reported across independent sources including Cooley, CELIS Institute, and the European Commission's own trade policy portal. This constitutes a solid evidentiary foundation for the raw data points underpinning support's argument.

Opposing's most substantively grounded finding is the documented chilling effect on cybersecurity research. Fact 886 (SIPRI, 2017) explicitly states that 'companies and researchers in the IT sector have voiced concerns about the unintended impacts and chilling effects of the controls on intrusion software,' and notes 'significant confusion about the precise scope of new controls on exports of cyber-surveillance technologies.' This is a stakeholder-confirmed, traceable harm that opposing correctly identifies as a concrete and actionable overlooked issue, supported by claim 273.

The 2022 denial rate of approximately 0.03–0.04% of total extra-EU export value is directly confirmed by multiple Tier-1 facts (facts 862, 863, 884, 888, 889): €0.98 billion denied against €57.3 billion authorized. This is the support side's most concrete empirical grounding. However, the claim that this rate constitutes a 'structural anomaly' or 'overlooked relationship' is not established by the facts: no benchmark from a comparable regime is present in the Case File, and the facts do not supply a cross-regime comparator that would allow anomaly classification. The internal year-on-year comparison (568 denials in 2021 vs. 813–831 in 2022, 0.01% vs. 0.03%) is documented in facts 863 and 889, but the CELIS Institute commentary (fact 865) explicitly notes that 'the report uses a new methodology—making comparisons with previously published data difficult.' This methodological discontinuity, acknowledged in facts 865 and 883, undermines the support side's use of 2021–2022 contrast as an 'implicit benchmark,' because the facts themselves flag that cross-year comparisons are unreliable under the new methodology.

The opposing side's most factually grounded argument is that chilling effects on legitimate IT and cybersecurity research from ambiguous cyber-surveillance controls are documented by stakeholders. Fact 886 (SIPRI, Tier-1) explicitly states: 'companies and researchers in the IT sector have voiced concerns about the unintended impacts and chilling effects of the controls on intrusion software,' and notes 'significant confusion about the precise scope of new controls on exports of cyber-surveillance technologies.' This is a concrete, stakeholder-confirmed empirical finding. However, the opposing side presents this as the primary 'overlooked' harm to redirect scrutiny, while claim 273 (support side) also invokes this same chilling effect as a documented harm. The opposing side does not provide quantified data on the magnitude of the chilling effect, and the SIPRI source (fact 886) dates to 2017, predating Regulation (EU) 2021/821 itself, limiting its direct applicability to the post-2021 regime.

Support's central inference — that the 0.04% denial rate constitutes a 'structural anomaly' — is asserted without a validated external benchmark. The premise-to-conclusion movement from a single data point (fact_id 862: €57.3B authorised vs €0.98B denied = 0.04%) to 'anomaly demanding explanation' requires an established norm against which deviation is measured. No such norm is supplied from the Case File. The implicit benchmark argument (inference_id 121) — that 2021 vs 2022 figures (568 vs 831 denials) constitute sufficient comparative context — is marked 'insufficient' in the record and is further undermined by the acknowledged methodological discontinuity between pre- and post-2022 reporting (fact_id 865, 883): the new methodology makes cross-period comparison unreliable, which is the very gap Support claims is overlooked. Support thus simultaneously relies on the 2021/2022 contrast as a benchmark while arguing that the methodological rupture renders such comparisons invalid — a self-defeating logical structure.

Opposing's strongest and most coherent argument is its redirection to the chilling-effect harm documented in fact_id 886 (SIPRI 2017): 'companies and researchers in the IT sector have voiced concerns about the unintended impacts and chilling effects of the controls on intrusion software.' This is a traceable, stakeholder-confirmed, source-grounded harm (claim_id 273) that does not depend on contested benchmark logic or cross-period comparisons. The premise-to-conclusion movement here is clean: documented stakeholder concern → concrete, actionable harm → merits scrutiny. However, Opposing uses this as a deflection rather than a developed argument, and the SIPRI source predates Regulation (EU) 2021/821 (2017 vs 2021), limiting its direct applicability to the post-2021 regulatory context.

Support's core claim that the 0.04% denial rate constitutes a 'structural anomaly' is undermined by a critical scope and definitional problem: the figure (fact_id 862, 884) is expressed as a percentage of total extra-EU export value, not as a percentage of dual-use licence applications submitted. This definitional conflation means the figure cannot, on its own, establish an anomaly within the licensing regime's scope without a denominator that counts only applications subject to control. Support does not supply this corrected denominator, leaving the anomaly designation without proper definitional grounding.

Opposing's strongest grounded point is that fact_id 886 (SIPRI source) documents stakeholder-confirmed chilling effects on legitimate IT-sector researchers from ambiguous cyber-surveillance controls — a concrete, scoped, and actionable harm within the regulation's defined scope. This is the one claim in the case file that is both traceable to a primary source and falls squarely within the regulation's own Article 26 cyber-surveillance reporting mandate (fact_id 882), making it a more precisely scoped finding than the contested anomaly argument.

Support's strongest traceable factual anchor — the 0.04% denial rate and the €57.3B vs €0.98B authorization-to-denial ratio — is confirmed by multiple independent Tier-1 sources (facts 862, 863, 884, 888, 889, 891, 892, 893). However, Support's decisive inferential move — that this figure constitutes a 'structural anomaly' — is undermined by the absence of any cross-regime benchmark in the Case File. Fact 895 notes that the EU control list differs philosophically from the US CCL, making cross-system comparison non-trivial, and no Wassenaar or comparable denial-rate baseline appears in the record. The anomaly designation therefore remains an assertion rather than a demonstrated finding.

Opposing's strongest rebuttal — that a 0.04% denial rate within post-standardized 2022 data still demands explanation independent of cross-period comparison — is logically sound and supported by the Case File. Facts 862, 863, and 891 all confirm the 2022 figures are internally consistent within the new methodology. However, Opposing simultaneously fails to supply the cross-regime benchmark it demands from Support: Fact 895 identifies philosophical differences between EU and US control lists, and no Wassenaar denial-rate baseline appears in the record. Opposing's benchmark-requirement argument therefore applies with equal force against its own position, as Support's inference 123 correctly identifies.

Support's central factual anchor — the 0.04% denial rate and the €57.3B vs €0.98B authorization-to-denial ratio — is confirmed by multiple independent tier-1 sources (facts 862, 863, 884, 888, 889) including the official European Commission report and Cooley analysis. This establishes that a verifiable, data-grounded observation exists within post-2022 standardized figures. However, support's leap from this data point to calling it a structural 'anomaly' and an 'overlooked relationship' is not substantiated: no benchmark from comparable regimes is present in the Case File, and all support inferences (121–124) carry status 'insufficient'. The decisive finding is therefore that support has a real factual foundation but fails to elevate it to a novel or overlooked finding as required by the debate goal.

Opposing's most decisive and evidence-grounded contribution is the identification of the chilling effect on legitimate cybersecurity research as a concrete, stakeholder-documented harm. Fact 886 (SIPRI, 2017) directly states that 'companies and researchers in the IT sector have voiced concerns about the unintended impacts and chilling effects of the controls on intrusion software,' confirming that this harm is traceable, stakeholder-confirmed, and predates 2022. This is the strongest finding in the Case File: it is a documented, actionable concern that the annual reporting framework (fact 882) does not systematically capture, making it a genuinely underexamined issue. However, opposing does not develop this into a structured novel finding; it uses it only as a rhetorical redirect rather than a substantiated counter-hypothesis.