Where did federal contract competition actually decline — and who drove it?

Overview

Five years of spending, one question about competition.

Competition is one of the clearest signals of health in federal contracting: when a larger share of dollars is competed, the government generally gets better pricing and lower risk. Headlines periodically claim competition is "falling," but the topline is noisy — inflation inflates raw dollars, and a handful of large buyers can move the government-wide number on their own.

I analyzed $3.58 trillion in federal contract obligations from FY2021 through FY2025 to answer a deliberately practical question: is competition really declining, and if so, where — at a level specific enough to act on.

The Challenge

The topline hides more than it reveals.

Three problems make the raw numbers untrustworthy for decision-making:

• Inflation. Nominal growth overstates real change; a rising dollar figure can mask flat or declining activity.
• Aggregation. A single large department — or one service branch within it — can drive a government-wide shift, so the headline tells you that something moved, not who moved it.
• Data quality. Government contract data carries nulls, revisions, and inconsistent coding that quietly distort naive queries.

Questions

What the analysis had to answer.

• After adjusting for inflation, did federal contract spending actually grow — and by how much?
• Did the competed share of obligations decline, and in which fiscal year?
• Which department and which sub-agency accounts for most of the decline?
• Are there timing patterns (e.g., end-of-fiscal-year concentration) that signal risk?

Approach

A reproducible pipeline, not a one-off query.

I built the analysis as an engineered pipeline so every number is traceable and re-runnable:

• Ingested contract obligations directly from government APIs into a local analytical store using DuckDB.
• Modeled the data into clean analytical tables, applying data-quality validation to handle nulls, revisions, and inconsistent agency coding.
• Adjusted nominal dollars into inflation-adjusted terms to measure real change.
• Computed competed-share metrics government-wide, then decomposed the change by department and sub-agency.
• Wrapped the pipeline in automated tests and GitHub Actions so results regenerate identically and regressions surface immediately.

Analysis

Separating real change from noise.

Adjusting for inflation reframed the entire story. Nominal obligations rose sharply, but the real increase was modest — most of the "growth" was the dollar losing value, not activity expanding.

With real growth established, I turned to competition. The competed share of obligations declined, with the sharpest movement landing in FY2025 — then decomposed that decline to find its source.

Key Findings

The decline is real, small in aggregate, and highly concentrated.

Business Implications & Recommendations

Where oversight and strategy should focus.

The decline in competition is not a diffuse, government-wide trend to be addressed with broad policy. It is concentrated enough to act on precisely:

• Target oversight at the Navy. Because one sub-agency explains roughly 70% of the government-wide decline, competition-improvement efforts there would move the national number more than any broad initiative.
• Judge programs in real terms. Inflation-adjusted analysis should be the default; nominal growth overstates activity and can justify budgets that real trends don't support.
• Watch year-end concentration. Rising September concentration is a known risk marker for rushed, less-competed awards and deserves monitoring.
• Institutionalize the pipeline. Because the analysis is reproducible and tested, it can be re-run each fiscal year to track whether interventions actually change the trend.

Tools & Methods

How it was built.

Reproducibility: the analysis runs as an automated, tested pipeline so every figure regenerates from source. Project repository: github.com/Akemp787/federal-spending-efficiency-analysis.