Skip to content

Hypotheses

Testable predictions for Sourcing under Sanctions: Judicial Urgency and Pharmaceutical Procurement Costs. Each hypothesis lives in its own file and is referenced project-internally by H:<slug>.

Hypotheses are organized into four thematic clusters aligned with the canonical manuscript v10-causal-mechanism/manuscript/paper/main.tex.

How to read a hypothesis page

Each page follows the same structure — the design → what's known → what's left:

  • Title + lede: the H:<slug> title, then a plain-words statement of the claim.
  • Evidence-strength callout: an at-a-glance verdict (Very strong / Strong / Moderate / Weak / Not yet tested) plus a one-line status.
  • Theory: the theoretical motivation, with references.
  • Prediction: the precise directional claim.
  • Competing prediction: what a non-sourcing explanation predicts instead — the alternative the test has to rule out (typically an incumbent same-firm markup).
  • Setting evidence: institutional and descriptive grounding from the BEC-SP record and the paper.md setting.
  • Empirical test: the concrete specification — outcome, variation, fixed effects.
  • Data requirements and limitations: datasets needed and threats to identification.
  • Evidence: a table of the project's analyses bearing on the hypothesis — Analysis (AN-ID, linked) · Bearing (Supports / Against / Mixed / Pending) · Key takeaway.
  • Open tests: forward-looking only — analyses not yet run.

The scorecard below rolls up the current status of every hypothesis against the analyses (AN-NNN) that bear on it.

Cluster A: The urgent-procurement margin (motivation)

Maps to paper §5.1 and to docs/results.md. Establishes the procurement environment — urgency raises costs and weakens competition — before the design moves inside urgent procurement.

Maps to paper §5.2–§5.3 and §4 (Empirical Framework). The selection-bounded under-the-gun contrast and the within firm-buyer-item pricing test that separates same-firm pricing from supplier-set sourcing.

  • H:utg-gap-selection-bounded — Court-mandated urgent purchases cost more than the closest administrative urgent comparison, and the gap survives selection bounding.
  • H:no-broad-same-firm-markup — In deep repeated urgent markets, the sanction-related cost margin does not appear as a broad same-firm markup.
  • H:thin-market-supplier-leverage — The deep-market null is not universal: a residual within-firm gap persists in the earlier period (the quantity dimension reflects scale).

Cluster C: The sourcing mechanism

Maps to paper §5.4 and to docs/results.md. Where the procurement-cost margin operates: lost scale and supplier-set reallocation.

  • H:lost-scale — Court-mandated buying gives up scale; administrative urgent orders are larger and capture bulk discounts.
  • H:supplier-set-reallocation — Legal urgency reallocates the winning supplier set.

Cluster D: Identification and robustness

Maps to paper §6 (Falsification and Robustness) and to docs/robustness.md + docs/advanced.md.

  • H:placebo-and-dynamics — The urgent-procurement pattern is specific to litigated items; dynamic evidence is diagnostic, not the primary design.

Scorecard

The theory-driven prediction, the intuition behind it, the analyses that bear on it, and the current verdict. Click a hypothesis for the full design and detailed Evidence table; click an AN-NNN for the backing analysis. Status runs Not yet tested → Not confirmed / Mixed / Moderate → Partial (strongly supported). Confirmed requires independent cross-data replication; no single-jurisdiction result here reaches it, so the ceiling is Partial (strongly supported).

# Prediction Intuition Evidence Status
H1 Urgent procurement → higher prices and fewer bidders than ordinary Compressed urgent sourcing cannot wait for aggregation or regular competition AN-001 negotiated price +5.4%, reference price +2.7%, bidders −5.4%, tender success +2.1pp (item+year+PBU FE, PBU-clustered). All four outcomes significant and consistent in direction; establishes the motivational urgent-procurement margin, not the sanction-exposure design Partial (strongly supported)
H2 Litigated urgent > administrative urgent, gap survives selection bounding One-sided sanctions raise the cost of court-mandated buying relative to the closest feasible urgent comparison AN-002 naive 29.5%, Lee bounds 15.9%–21.1% after trimming the selected administrative group; AN-007 wild-cluster \(p=0.0080\) (preferred), 0.0390 (item-by-year-month) Partial (strongly supported)
H3 No broad same-firm markup in deep repeated urgent markets Conditional on the same firm, buyer, and item, price is indistinguishable across urgent regimes AN-003 within firm-buyer-item coefficient 0.035 (SE 0.041); near zero in above-median quantity (−0.005) and SUS-formulary (−0.001) subsamples Partial (strongly supported)
H4 The deep-market null is not universal: a within-firm gap persists in thin/early cells Thin or early urgent pools could give incumbents room to price AN-003 below-median quantity +0.066***, earlier period +0.117***, non-formulary +0.101; AN-004. Disambiguation: quantity axis is scale (within-FBI log-qty −0.259, 61); earlier-period gap survives a quantity control (+0.168, \(p=0.007\), 62) but is administrative-dearer and time-declining Partial (strongly supported)
H5 Court-mandated buying gives up scale Fragmented urgent buying forgoes bulk discounts AN-005 quantity/scale component −32.8% of the gap, administrative orders 3.3× larger, bulk-discount elasticity −0.329; AN-009 aggregation diagnostic (mixed) Partial (strongly supported)
H6 Legal urgency reallocates the winning supplier set Compressed timelines and fragmented demand change which supplier wins AN-006 among 2,134 item-buyer pairs, modal winner differs in 70.2%, mean winner-set Jaccard similarity 0.268, 48.5% with no winner overlap Partial (strongly supported)
H7 Pattern is litigation-specific; dynamics are diagnostic A generic platform/time-trend story would reproduce the pattern off the litigation margin; dynamic trends are sensitive AN-008 never-litigated placebo −0.020 (SE 0.032), null; AN-013 matched buyer-class placebo −0.046 (SE 0.053), null; AN-010 BJS event study diagnostic, Honest-DiD does not survive at the observed maximum pre-period scale Partial (strongly supported)