Skip to content
Benchmarking

VRscores versus donations-based and population benchmarks

This page summarizes comparisons between VRscores (workforce two‑party partisan shares based on voter registrations) and donation‑based employer measures (FEC‑reported donors), alongside population benchmarks from the Current Employment Statistics (CES), the Bureau of Labor Statistics (BLS), and the American Community Survey (ACS). We highlight three themes: (1) VRscores has greater coverage and scale, (2) VRscores has better workforce representativeness, and (3) firm‑level alignment between VRscores and donation‑based measures of workforce partisanship is only moderately correlated. The visuals mirror the analyses in the VRscores research paper, using the same definitions and thresholds.

Comparison of FEC and VRscores. We compare VRscores (workforce partisanship from linking L2 voter registrations to Revelio Labs workers within MSAs) with FEC donors (individual contribution records grouped to canonical employer names). Totals are aggregated to 2024 for VRscores and to the 2023–2024 cycle for FEC, after applying a consistent employer threshold (≥5 matched workers or donors). Employer name grouping for FEC follows the paper’s approach (lower‑casing, punctuation removal, standardized corporate suffixes, token blocking, TF–IDF clustering).

Data snapshot: These comparisons use the 2024 VRscores panel and the 2023–2024 FEC donor cycle. All FEC comparisons require ≥5 donors per employer to mirror the VRscores public‑dataset threshold. FEC bulk data are based on the indiv24.zip file retrieved from fec.gov/data/browse-data.

Coverage & scale

Comparison of FEC and VRscores

We contrast VRscores (workforce partisanship from linking L2 voter registrations to Revelio Labs workers within MSAs) with FEC donors (individual contribution records grouped to canonical employer names). Totals are aggregated to 2024 for VRscores and to the 2023–2024 cycle for FEC, after applying a consistent employer threshold (≥5 matched workers or ≥5 donors). Employer name grouping for FEC follows the paper’s approach (lower‑casing, punctuation removal, standardized corporate suffixes, token blocking, TF–IDF clustering).

Total unique workers (millions)
24.5 M
VRscores
1.3 M
FEC
Total unique organizations (millions)
0.5 M
VRscores
0.1 M
FEC
Average workers per organization
45.9
VRscores
25.5
FEC

Employers by workforce / donor counts (2024 snapshot)

Each bar shows how many employers fall within the indicated worker or donor-count bins. VRscores requires at least 5 matched workers per employer; the FEC comparison applies the same donor threshold.

0110100100010,000100,0005–199800–9991,600–1,7992,400–2,5993,200–3,3994,000–4,1994,800–4,9995,600–5,7996,400–6,5997,200–7,3998,000–8,1998,800–8,9999,600–9,79910,200+Employers (logₑ(count + 1))Matched employees / donors per employer (capped at 10,000)
VRscores employersFEC donor employers|Default is log scale. Switch to linear to see absolute differences.

Workforce representativeness

Partisan representativeness

Comparison of Democratic, Independent/Other, and Republican shares. VRscores is shown as registered two-party and imputed two-party variants; donors reflect party of contribution. VRscores tracks the population benchmark more closely, whereas donor series classify contributions to both parties as Independent/Other.

0%10%20%30%40%50%60%70%Share of workers / donors (%)DemocratIndependent / OtherRepublican
VRscores (Registered)VRscores (Imputed)FEC Donors2022 CES registered voters benchmark

Job zone coverage

Higher job zones require more formal education and training. Donors skew heavily toward Zone 5 roles relative to the VRscores workforce and to national employment benchmarks.

Zone 5: Extensive Prep
15.2%
31%
0%
25%
50%
Zone 4: Considerable Prep
38%
43.6%
0%
25%
50%
Zone 3: Medium Prep
19.1%
13.8%
0%
25%
50%
Zone 2: Some Prep
24.8%
9.8%
0%
25%
50%
Zone 1: Little/No Prep
2.8%
1.8%
0%
25%
50%
All VRscores WorkersVRscores Donors OnlyBLS May 2024 employment benchmark

Seniority level coverage

Donors are much more concentrated in senior leadership roles, while VRscores captures many more entry‑ and junior‑level employees within each firm. National benchmarks are unavailable for seniority because the classification comes from Revelio Labs’s taxonomy rather than a federal series.

Executive Level
4.3%
13.4%
0%
20%
40%
Director Level
12%
23.5%
0%
20%
40%
Manager Level
10.7%
13.4%
0%
20%
40%
Associate Level
10.7%
10.8%
0%
20%
40%
Junior Level
25.1%
21.1%
0%
20%
40%
Entry Level
37.2%
17.9%
0%
20%
40%
All VRscores WorkersVRscores Donors Only

Industry coverage

Relative to the ACS national employment benchmark, VRscores more closely reflects the population’s industry mix, while donors are disproportionately concentrated in Finance, Information, and Professional and Business Services. Bars show sector shares for all matched VRscores workers and for the donor subset; dashed ticks denote the ACS 2024 industry composition.

Natural Resources and Mining
0.8%
0.9%
Construction
2.2%
2%
Manufacturing
13.4%
12.5%
Wholesale Trade
2.5%
1.9%
Retail Trade
7.2%
3.5%
Transportation, Warehousing, and Utilities
3.7%
3.9%
Information
8.2%
9.6%
Financial Activities
12.3%
14.1%
Professional and Business Services
13.5%
16.4%
Education and Health Services
19.1%
19.8%
Leisure and Hospitality
5.1%
2.7%
Other Services (except Public Administration)
3.5%
4.4%
Public Administration
8.4%
8.4%
All VRscores WorkersVRscores Donors Only|ACS 2024 industry benchmark

Employer-level correlation

VRscores vs FEC donation-based measure

VRscores and FEC donor shares are only moderately correlated (r ≈ 0.51). One in five matched employers have differing majority party categorizations between the two measures. Employers shown are those we could match by name between VRscores and FEC employer records. FEC bulk data: indiv24.zip.

0%0%25%25%50%50%75%75%100%100%FEC donor Republican share (two-party)VRscores workforce Republican share (two-party)

Circles scale with matched VRscores workers plus FEC donors. Employers are matched by name across VRscores and FEC employer records.

Consistent (majority match)Inconsistent (majority mismatch)

DIPI comparison

VRscores vs. Donation-Based Indicator of Political Ideology (DIPI; 10-year)

Public firms matched by GVKEY across datasets (VRscores 2024 vs. DIPI 2022). Grey markers indicate majority‑party agreement across the two measures; blue markers highlight disagreements. Panel buttons update the DIPI measure only; VRscores workforce measures remain the same across panels. Read A Donation-Based Indicator of Political Ideology (DIPI) for methodological details on the donations-derived ideology scores used here.

DIPI measure
0%0%25%25%50%50%75%75%100%100%% Republicans in Employee Donations (DIPI, 10-Year)% Republicans in VRscores (Imputed Two-Party Share)
n = 1,752r = 0.51Consistent: 72.1%Consistent (majority match)Inconsistent (majority mismatch)