Loading pattern matches
VA's Geographic Distribution of VA Expenditures (GDX) and USAspending.gov show different amounts for the same state, fiscal year and spending category, by more than the tolerance band set by all states that year. GDX reports expenditures; USAspending reports obligations.
Aggregated from the matches below. Severity bands are calibrated per pattern; counts and dollar totals reflect this snapshot only.
Glassbox VA surfaces correlations and observations, not findings of misconduct. Severity is the size of the match (dollars, growth ratio, concentration or firm count), not the likelihood of a problem; evidence strength (0–1) shows how well the data supports the match. Both are a starting point for review, not a verdict.
The match meets the pattern threshold, in the lower size band.
A mid-sized match. Pull the source records and read the listed benign explanations first.
The largest size band for this pattern. Larger matches are listed first; size is not evidence of a problem.
Severity is computed from the detection signal · gdx.reconciliation (scripts/va/gdx.ts): state × fiscal year × category for Compensation & Pension, E&VRE, Insurance and Construction only (Medical Care, GOE, Loan Guaranty and Memorial Benefits are not comparable) · ratio = USAspending obligations of the crosswalked assistance listings / PSC Y contracts ÷ GDX expenditure · band = Tukey fences exp(Q1 − 1.5·IQR, Q3 + 1.5·IQR) over ln(ratio) of all compared states that year (≥ 10 states) · variance when outside the band AND |ratio − median| ≥ 5 points AND |gap| ≥ $1M · GDX < $1M not compared · years compared only when all 12 months of USAspending are loaded · signal = |gap| $
GDX reports $3.9M of Insurance & Indemnities expenditures for Alaska in FY2014; USAspending shows $2.0M of obligations under the same category definitions (50.8% of GDX). The ratio sits below the tolerance band set by 52 compared states that year (61.7% to 111.2%; median 83.0%). GDX figures are expenditures and USAspending figures are obligations; the difference is a variance to review, not a finding.
Why this may be benign · The sources place dollars differently: GDX follows the beneficiary's residence or the project site; USAspending follows the reported place of performance or the recipient's address.
Each count is a state, category and fiscal year where USAspending obligations sit outside the tolerance band set by all states that year (GDX expenditure vs USAspending obligations, compared categories only). Dark states have none. A variance is a question to ask about two public datasets, not a finding. Source: VA NCVAS GDX · USAspending.gov (VA, agency 036).
Select a state to filter the list below.
GDX reports $3.1M of Insurance & Indemnities expenditures for Puerto Rico in FY2019; USAspending shows $1.8M of obligations under the same category definitions (57.6% of GDX). The ratio sits below the tolerance band set by 52 compared states that year (60.6% to 109.0%; median 82.1%). GDX figures are expenditures and USAspending figures are obligations; the difference is a variance to review, not a finding.
Why this may be benign · The sources place dollars differently: GDX follows the beneficiary's residence or the project site; USAspending follows the reported place of performance or the recipient's address.
GDX reports $14.1M of Education & Vocational Rehabilitation/Employment expenditures for Vermont in FY2022; USAspending shows $15.2M of obligations under the same category definitions (108.2% of GDX). The ratio sits above the tolerance band set by 53 compared states that year (97.6% to 108.0%; median 102.8%). GDX figures are expenditures and USAspending figures are obligations; the difference is a variance to review, not a finding.
Why this may be benign · For FY2012–FY2023 VA states that GDX Compensation & Pension and E&VRE (and FY2023 Insurance) were taken from USAspending.gov, so agreement is expected; a variance points to a later revision or a state-assignment difference.