From Lowest Price to Highest Public Value
  • Home
  • Research
  • Papers
  • Findings
  • Literature
  • Federal Data
  • Vendor Journey
  • International
  • Briefs
  • Decision Tool
  • Tracker
  • People
  • About

Research Papers

Five empirical papers on procurement source selection

This research program comprises five empirical papers constituting the most comprehensive analysis of source selection method effects in federal procurement to date. Together they analyze 654,307 competitively awarded contracts spanning fiscal years 2017-2023, drawn from the Omari et al. (2025) comprehensive FPDS dataset.

The papers follow a deliberate narrative arc. Paper 5 establishes that only one prior peer-reviewed study (Landale et al., 2017; N=124) has directly compared LPTA and tradeoff outcomes using U.S. federal contract-level data – a striking evidence gap given that procurement policy affects hundreds of billions of dollars annually. Paper 1 fills that gap with a difference-in-differences design exploiting NDAA Section 813 as a natural experiment, finding cost growth reductions of 13-21 percentage points after robustness corrections but no detectable first-stage compliance shift. Paper 2 asks when evaluation method matters most, finding that transaction cost economics correctly predicts the effect is amplified for complex, small, and cost-type contracts. Paper 4 reveals an uncomfortable paradox: 41% of ostensibly competitive awards receive only one bid, and tradeoff evaluation is descriptively associated with 3.5 times higher single-bid odds. Paper 3 zooms out to compare procurement entry barriers across twelve countries and the EU, finding an 87-fold cost differential that dwarfs any evaluation method effect.

Integrating thesis: The LPTA-vs-tradeoff debate may be a red herring. The real problem is structural barriers that constrain the competitive pool, making evaluation method largely irrelevant.


Paper 5

The Source Selection Evidence Gap: A Scoping Review and Research Agenda

Target: Journal of Public Procurement
Draft Complete

A scoping review following the PRISMA-ScR framework that synthesizes findings from 87 relevant works across procurement economics, public administration, auction theory, and defense acquisition published between 1990 and 2025. The central finding is stark: only one peer-reviewed empirical study has directly compared LPTA and tradeoff outcomes using U.S. federal contract-level data (Landale et al., 2017; N=124).

Key Finding: Procurement policy affecting hundreds of billions of dollars annually rests on an evidence base that would be considered wholly inadequate in any adjacent policy domain.
Paper 1

The Policy Shock That Didn't Shock: A Difference-in-Differences Analysis of NDAA Section 813

Target: Journal of Public Procurement
Draft Complete

Exploits Section 813 of the NDAA FY2017 as a natural experiment, comparing DoD service and IT contracts (treatment) with comparable civilian agency awards (control) across 654,307 competitively awarded contracts. Baseline estimates show a 30.35-percentage-point reduction in cost growth and 0.55 fewer modifications per award. Five robustness corrections -- trend-adjusted DiD, Callaway-Sant'Anna, Bacon decomposition, IPW, and temporal trimming -- yield a preferred range of 13 to 21 percentage points.

Key Finding: No evidence that DoD's source selection composition diverged from civilian agencies post-treatment -- DoD's LPTA rate (39.8%) remained comparable to civilian agencies (38.8%), suggesting the policy did not operate through its intended mechanism.
Paper 2

Transaction Costs as Moderators of Procurement Design: When Does Evaluation Method Matter?

Target: JPART
Draft Complete

Tests transaction cost economics predictions using interaction models within the Section 813 DiD framework. Estimates how the policy effect varies across four dimensions of transaction complexity: contract size, industry sector (NAICS classification), pricing structure, and market competition. Results strongly support TCE predictions across all four moderator dimensions.

Key Finding: The effect is amplified for complex transactions -- substantially larger for small contracts than large ones, concentrated in restricted professional and scientific service categories where asset specificity is highest, and larger for cost-reimbursement contracts than firm-fixed-price.
Paper 4

Single-Bid Awards in Federal IT Procurement: Prevalence, Predictors, and a Competition Paradox

Target: PPMR
Draft Complete

Analyzes the prevalence and consequences of single-bid awards -- procurements receiving only one offer despite being classified as "competitive" under federal acquisition regulations. Documents that 41% of ostensibly competitive awards received only one offer, with civilian agencies exhibiting substantially higher rates than DoD. Logistic regression reveals that tradeoff-evaluated contracts are 3.5 times more likely to receive a single bid than LPTA awards.

Key Finding: The 3.5x descriptive odds ratio should not be interpreted causally due to endogenous source selection, but the DiD analysis finds that restricting LPTA in DoD slightly increased single-bid probability (logit coefficient = 0.051, p < .001), suggesting more complex evaluation may paradoxically reduce competition. With agency-clustered standard errors, this effect becomes non-significant (p = 0.288).
Paper 3

Comparing Procurement Entry Barriers Across Twelve National Systems

Target: IJPA
Draft Complete

Uses vendor journey mapping -- a process-tracing methodology documenting every step, cost, and time requirement -- to compare procurement entry barriers across 13 national systems: the United States, South Korea, Singapore, the United Kingdom, Australia, Canada, New Zealand, the European Union, Chile, Estonia, Georgia, Denmark, and Brazil. Proposes a three-tier typology of procurement regimes: open-access, moderate-barrier, and high-barrier.

Key Finding: An 87-fold differential in qualification costs between the most restrictive system (United States: $6,800-$516,000) and the least restrictive (Chile: $425-$5,800), with Brazil's reverse auction model producing per-bid costs 160-270 times lower than the US. Digital infrastructure shows the strongest association with low barriers.
© From Lowest Price to Highest Public Value 2026