Dollars Just for DC
Do significant federal campaign donors also contribute to NYC’s local races?
Overwhelmingly, no: only 13% also donated to any city race, and among donors who work at large employers, just 8% did.
Local politics shapes the buildings we live and work in, the streets and spaces between them, the schools that teach our kids, and the police and fire departments we depend on.
But federal politics can absorb all our political attention. It is high stakes and high profile. It’s also diffuse, and it can hook us into following the show on our screens as a hobby.
How wide is the gap between federal and local political involvement?
On this new page, Dollars Just for DC, I dig into that gap by looking at New York City residents who made federal campaign contributions over the $200 reporting thresholdi and exploring whether they made any local New York City campaign contribution in the following year.
Only 13% of the 130k New York City residents who donated more than $200 to a federal campaign in 2024 donated to any city campaign in 2025. Among donors who work at large employers?, the city campaign contribution rate was only 8%.
City campaign contribution rates among federal donors fell from the prior 2020 & 2021 election cycles: from 16% to 13% citywide and from 11% to 8% among donors who work at large employers.
| Election window | Federal donor group | Federal donors ($200+) | Gave to any city campaign | Rate |
|---|---|---|---|---|
| 2024 federal → 2025 city Latest city election | All NYC residents | 130,002 | 17,048 | 13% |
| Donors who work at large employers | 12,146 | 957 | 8% | |
| 2020 federal → 2021 city Prior city cycle | All NYC residents | 184,640 | 30,156 | 16% |
| Donors who work at large employers | 19,166 | 2,118 | 11% |
The analysis below shows how the rate varies across industries and firms, how federal pay-to-play rules appear to reduce local giving at affected firms, and that there was more room to donate locally, with about $19M in unused public matching-funds capacity for viable candidates in close 2025 races.
I’m looking at campaign contributions as an imperfect but concrete proxy for involvement in political campaigns. I’ve found that getting involved and contributing to local campaigns has strengthened my own connection to how our city is run and inspired me to share that enthusiasm with others.
As far as I could tell, there has been no prior research that looked at how often federal campaign donors contributed at the local level too. I hope this work to understand that gap spurs more exploration of how local political engagement varies across the city, and as a starting place, the methodology at the bottom has the public repo with all the code used for this analysis.
The goal is broader than donations. I hope this analysis makes the local civic gap more visible and inspires more involvement in local politics, so we can keep building a magnificent city worthy of the mighty woman with a torch in our harbor.
Key findings on the federal vs. local political contribution gap in New York City
- The federal vs. local contribution gap is large. Only 13% of the 130,002 New Yorkers who gave at least $200 to a federal campaign in 2024 gave to any city political campaign (primary or general) in 2025.
- The gap got larger in the most recent election cycle. The city campaign contribution rate was higher at 16% in the 2020 federal to 2021 city cycle.
- The gap is larger among donors who work at large employers. NYC donors in this group show an even lower local-giving rate: only 8% of these 2024 federal donors gave to a 2025 city campaign.
- Some industries are much more locally involved. The federal donor contribution rate to local campaigns was 34% in commercial brokerage firms, 18% in Big Law, and 12% in Software / Internet firms, compared to the large-employer average of 8%. Read more →
- Compliance restrictions appear to cut local giving roughly in half at affected firms. Employees at firms with direct pay-to-play restrictions give locally at 5%, vs. 9% at unaffected firms. Employees at such firms may be prohibited from donating to local officials such as the city comptroller (who oversees the pension fund and its allocations to investment managers), and compliance policies can more broadly restrict local giving. Read more →
- There’s more room to give. Across city council, borough, and city-wide primary and general election races in 2025, 28 campaigns in close races (initial vote margin within 15 percentage points) remained below their public matching cap, with about $19M of unused capacity collectively. Read more →
How does the pattern vary by industry?
Most large sectors sit in the low-to-mid single digits. Two exceptions stand out among the sizeable donor pools:
- Big Law firms — 18%
- Software / Internet — 12%
One smaller sector runs well above the citywide rate:
- Commercial Brokerage — 34%
Outside those exceptions and a few others in smaller sectors, sector-level participation runs in the low-to-mid single digits.
How firms are grouped into sectors. Each firm’s GICS industry classification is pulled from SEC EDGAR where available, then a hand-tuned custom subsector is assigned (e.g. “Big Law,” “Investment Bank,” “Software / Internet”) so the buckets match how readers think about the NYC employer base rather than the broader GICS sector hierarchy.
One-firm sectors to read carefully. The Healthcare / Pharma row is effectively one firm: Pfizer accounts for nearly all of the donors and the 1% rate. Read that row as a single-firm finding, not a broad-sector finding.
See the methodology appendix for firm-by-firm sector assignments.
How much of the gap do compliance restrictions explain?
Significant federal donors at compliance-restricted firms contribute to local campaigns at roughly half the rate of unrestricted firms. In the 2024 federal → 2025 city window, significant federal donors who worked at firms without pay-to-play? restrictions had 9% donate to a city campaign, while for significant federal donors at restricted firms, it was only 5%.
The gap is large enough that the regulatory difference is the most plausible driver, even if other differences between donors at these firms also matter. Even so, at unrestricted firms, almost 90% of significant federal donors still don’t contribute to any local campaign.
How the rules work, in detail
The rules creating that friction are well-defined. SEC Rule 206(4)-5 (investment advisers) and MSRB Rule G-37 (municipal-securities dealers and municipal advisors) impose a two-year ban on compensated government work after certain employee contributions to officials with influence over adviser, dealer, or municipal-advisor selection. The NYC Comptroller — the sole fiduciary of the city pension systems — and the Mayor, who appoints the pension-board trustees, are commonly treated as covered.
But the legal trigger is narrower than firm practice. The statutes cover firms, certain covered executives, solicitors, and municipal-finance professionals. In practice, large firms typically apply pre-clearance, low internal thresholds, and conservative interpretations to every employee, not just those legally covered. Some maintain a practical culture of avoiding state and city giving altogether.
For bucket definitions, regulatory citations, examples of published firm policies, SEC enforcement history, and a fuller justification of the Partial bucket, see the pay-to-play background section in the appendix.
How does the pattern vary across individual firms?
Most large employers in the analysis sit in the low-to-mid single digits. A handful run higher: Sullivan & Cromwell (34%), Davis Polk (20%), Paul Weiss (17%), Bloomberg (16%), Greenberg Traurig (16%) — and, further down the full firm list, Newmark at 48%. The higher rates are concentrated in Big Law, commercial brokerage, and a small number of media and tech firms.
Which firms are in the table. The table covers about 150 large employers with at least 10 NYC-resident employees who made a federal contribution above the $200 reporting threshold in 2024. The remaining employers in the source list have fewer than 10 such donors and were excluded to keep statistically thin pools out of the chart. The methodology appendix explains how the employer group was defined.
How does the pattern vary by industry?
Same picture as 2024/2025 — most large sectors sit in the low-to-mid single digits. Two exceptions stand out among the sizeable donor pools:
- Big Law firms — 22%
- Software / Internet — 15%
Three smaller sectors (under ~200 NYC-resident federal donors) also exceed the citywide rate:
- Commercial Brokerage — 32%
- Sports / Entertainment — 20%
- Real Estate Developers — 18%
Outside those exceptions and a few others in smaller sectors, sector-level participation runs in the low-to-mid single digits.
How much of the gap do compliance restrictions explain?
Significant federal donors at compliance-restricted firms contribute to local campaigns at roughly half the rate of unrestricted firms. In the 2020 federal → 2021 city window, significant federal donors who worked at firms without pay-to-play? restrictions had 12% donate to a city campaign, while for significant federal donors at restricted firms, it was only 8%.
The gap is large enough that the regulatory difference is the most plausible driver, even if other differences between donors at these firms also matter. Even so, at unrestricted firms, almost 90% of significant federal donors still don’t contribute to any local campaign.
How the rules work, in detail
The rules creating that friction are well-defined. SEC Rule 206(4)-5 (investment advisers) and MSRB Rule G-37 (municipal-securities dealers and municipal advisors) impose a two-year ban on compensated government work after certain employee contributions to officials with influence over adviser, dealer, or municipal-advisor selection. The NYC Comptroller — the sole fiduciary of the city pension systems — and the Mayor, who appoints the pension-board trustees, are commonly treated as covered.
But the legal trigger is narrower than firm practice. The statutes cover firms, certain covered executives, solicitors, and municipal-finance professionals. In practice, large firms typically apply pre-clearance, low internal thresholds, and conservative interpretations to every employee, not just those legally covered. Some maintain a practical culture of avoiding state and city giving altogether.
For bucket definitions, regulatory citations, examples of published firm policies, SEC enforcement history, and a fuller justification of the Partial bucket, see the pay-to-play background section in the appendix.
How does the pattern vary across individual firms?
Same shape as 2024/2025: a handful of higher-engagement firms, the long tail clustered near zero.
Was there room left to give?
Yes. In the 2025 city elections, there was about $19M left in viable campaigns in competitive races before they hit their public matching caps. The $19M came from 28 below-cap campaigns (21 in primaries, 7 in general elections) and represents about 25% of the public matching capacity at those viable campaigns.
We defined viable campaigns as those that either raised at least 50% of their public-match cap or cleared 1/(n+1) of the initial vote in their contest (17% in a five-candidate race, 33% head-to-head). A race counts as competitive when at least two viable campaigns ran in it and the top two finishers ended within 15 percentage points of each other — tight enough that more donations to a runner-up could have plausibly mattered. Both screens are imperfect but useful proxies for “was there more room for donations that could have made a difference?” More detail in the appendix →
Capacity left
Breakdown by level
| Race level | Competitive races (default) | Less competitive races |
|---|---|---|
| Mayoral | 2 contests7 campaigns below cap4 primary; 3 general$15M remaining | -- |
Show race examplesCompetitive races with roomCandidate Unused Cap used Vote MayorPrimary
Mamdani, Zohran Kwon $0 100% 44% initial Cuomo, Andrew M $3M 60% 36% initial Lander, Brad $632K 91% 11% initial Stringer, Scott M $3M 63% 2% initial Myrie, Zellnor $4M 50% 1% initial MayorGeneral
Mamdani, Zohran Kwon $955K 86% 51% general Cuomo, Andrew M $3M 54% 41% general Sliwa, Curtis A $2M 76% 7% general | ||
| Citywide and boroughwide | 2 contests4 campaigns below cap4 primary; 0 general$2.5M remaining | 10 contests10 campaigns below cap3 primary; 7 general$20M remaining |
Show race examplesCompetitive races with roomCandidate Unused Cap used Vote ComptrollerPrimary
Levine, Markwon $504K 89% 48% initial Brannan, Justin $1M 69% 34% initial BP MNPrimary
Hoylman-Sigal, Bradwon $259K 84% 49% initial Powers, Keith $415K 74% 38% initial | ||
| City Council | 10 contests17 campaigns below cap13 primary; 4 general$782K remaining | 42 contests52 campaigns below cap24 primary; 28 general$3.8M remaining |
Show race examplesCompetitive races with roomCandidate Unused Cap used Vote CD 13Primary
Aldebol, Shirleywon $32K 84% 30% initial Torres, Jacqueline J $133K 34% 21% initial Diaz, David A $149K 26% 19% initial CD 21Primary
Thomas-Henry, Shanelwon $8K 96% 31% initial Montoya, Erycka $0 100% 26% initial Henriquez, Yanna M $94K 53% 23% initial Aiken, David $91K 55% 20% initial CD 28Primary
Hankerson, Tywon $0 100% 35% initial Singh, Japneet $0 100% 25% initial LeGrand, Latoya $96K 52% 17% initial Hitlall, Romeo $0 100% 9% initial CD 30General
Wong, Phil Wwon $20K 90% 54% general Vaichunas, Alicia B $0 100% 44% general CD 13General
Aldebol, Shirleywon $15K 92% 52% general Marmorato, Kristy $1K 99% 46% general CD 19General
Paladino, Vickiewon $10K 95% 57% general Chou, Benjamin $0 100% 43% general Show 4 more racesCandidate Unused Cap used Vote CD 08Primary
Encarnacion, Elsiewon $0 100% 29% initial Lopez, Wilfredo $20K 90% 18% initial Alayeto, Clarisa $39K 81% 15% initial CD 30Primary
Wong, Phil Wwon $25K 87% 36% initial Pogozelski, Paul J $8K 96% 32% initial Smyth, Dermot $0 100% 31% initial CD 17Primary
Sanchez, Justinwon $0 100% 40% initial Ortiz, Antirson R $31K 84% 25% initial Santana, Elvis $0 100% 16% initial CD 04Primary
Maloney, Virginiawon $0 100% 27% initial Aronson, Vanessa T $0 100% 26% initial Wetzler, Benjamin D $1K 99% 15% initial Bondy, Faith $0 100% 8% initial | ||
| Total | 14 contests28 campaigns below cap21 primary; 7 general$19M remaining | 52 contests62 campaigns below cap27 primary; 35 general$24M remaining |
About me & my motivation
I’m Ramon Gonzalez, and this is an independent civic data project. It grew out of my involvement with Abundance New York and Maximum New York, but is not affiliated with either organization, any campaign or candidate, or Devoted (where I work).
I’m interested in helping more New Yorkers get involved in local politics and shape the communities they live in. This project looks at one lens on that broader question: the gap between involvement in federal politics and local NYC politics, proxied through political contributions because donations are visible in comparable public records.
For readers who want to inspect or adapt the work, the public repository at github.com/ramongonzalez12/dollars-just-for-dc includes the code, methodology, committed aggregate data, and instructions for pulling the underlying public data: FEC bulk individual-contribution files, NYC Campaign Finance Board contribution records, NYC Board of Elections vote results, and the public employer directory used to define the large-employer group.
Other work
I lead population health at Devoted and write at Spheres of Possibility.
I run Art of the Near Future World, a community project that solicits, reads, and celebrates short fiction imagining the world a few months to a few years from now. The near future comes quickly and we meet often: our next event is June 8th, and you can read past winning stories from April 2026, October 2025, and April 2025.
I live in New York City, and I’m a member of Abundance New York to help build a more vibrant and magnificent city.
Corrections or questions welcome: DM @ramonjgonzalez7 or message me on Substack at Spheres of Possibility.
Additional checks, methods, and background
A short methodology overview, then the detailed disclosures: matching sensitivity, full methodology and source files, compliance regime classification, the campaign-capacity test, and prior-research comparison.
A1. How the analysis works
The analysis follows six steps. The detailed sections in this appendix unpack each step, with sensitivity tables where the choice affects the result.
- Identify significant federal donors. Federal campaign-finance law (52 U.S.C. §30104) itemizes individual contributions of $200 or more. That threshold defines the only publicly identifiable universe of federal donors, so the analysis is restricted to itemized donors — people who gave $200+ to a single federal committee in a cycle.
- Restrict to NYC residents. Using FEC-reported addresses, the analysis keeps only donors whose address ZIPs fall in the five boroughs.
- Match to NYC city campaign records. The NYC Campaign Finance Board publishes every city contribution down to $5 — no reporting threshold. Each federal donor is matched to the city file using full first name + last name + ZIP5. Section A2 details what this key can miss and how the miss rate is bounded.
- Calculate the rate. For each cohort, the analysis calculates the share of federal donors who also appear in the city file with any positive city contribution in the following year’s city cycle.
- Break out by employer. For Partnership for NYC member firms, each federal donor’s self-reported FEC employer string is mapped to a Partnership firm using a token shortlist plus a Claude Haiku classifier, with a small alias layer for short names (EY, AMEX, NBC Universal, Starbucks).
- Test for capacity constraints. For competitive 2025 races, the analysis calculates how much public matching-funds capacity remained unused at viable below-cap campaigns, to test whether the absence of room to give explains the gap.
A2. Methodology and matching sensitivity
Federal and city campaign-finance records do not share a donor ID. The analysis turns on two methodological choices: which federal donors enter the cohort, and how we recognize the same person in the city file.
Who counts as a federal donor?
Itemized FEC individual contributions of $200 or more from a NYC-borough ZIP, drawn from the bulk individual-contribution files for the relevant federal cycle. The Partnership for NYC member-firm cohort is the subset of those donors whose self-reported FEC employer string normalizes to a Partnership member company.
- Employer normalization uses a token shortlist plus a cached Claude Haiku classifier, with a small exact-alias layer for short names like EY, AMEX, NBC Universal, and Starbucks.
- Classification noise is small and balanced across firms; it doesn’t move the aggregate rate. Full normalization mechanics live in the “How we did this” fold below.
How do we identify the same person across federal and city files?
A federal donor counts as a city donor when the same full first name + last name + ZIP5 appears in the NYC Campaign Finance Board contribution file for the paired city cycle. Middle names, suffixes (Jr., Sr., III), and honorifics are dropped, so “James R. Smith, Jr.” and “James Smith” collapse to one person. It is a clean, reasonable person key, forgiving on title variation but strict on first-name agreement. Any positive NYC contribution counts.
- What the headline key can miss. Nicknames (“Michael” vs “Mike”), marriage-changed surnames, or a donor who moved ZIPs between cycles will split one person into two and miss the match.
- An alternative match to bound that miss. Last name + ZIP5 alone, restricted to last names that appear in two or fewer ZIPs across the NYC CFB universe. In that rare-last bucket, two unrelated people sharing the same last name in the same ZIP is unlikely.
- What the alternative recovers. The alternative finds 18% more matches than the headline rule, raising the citywide rate from 13% to about 15%. The lift is stable across both 2024/25 and 2020/21 election pairs.
- What it conflates. Those extra matches mix two things we can’t cleanly tell apart: real same-person misses the headline rule didn’t catch, and household members with the same rare last name who gave separately (typically spouses).
- How to read it. 15% is the ceiling that assumes every extra match is the same person. The true correction is somewhere between 13% and 15%, probably closer to 13%.
Full source-file inventory, FEC and CFB filters, campaign-capacity joins, and AI disclosure are in the “How we did this” fold immediately below.
A3. Source files, full methodology, and disclosure
Open the full methodologySource files, matching rules, campaign totals, vote results, and AI disclosure
Two measurement layers. The main analysis is a donor-crossover measure: among NYC-resident itemized federal donors, who also gave to any NYC city campaign in the paired local cycle? The local-giving-room appendix is a campaign-capacity measure: among 2025 campaigns, which viable campaigns in competitive races still had unused public matching-funds capacity? The method below keeps those two layers separate.
1. Source files and vintages
- Firm universe. 325 Partnership for New York City member companies, captured from the public member directory on 2026-04-22. 300 have at least one NYC-resident federal donor in the analysis window.
- Federal donor records. FEC bulk individual-contribution files for the 2018, 2020, 2022, and 2024 federal cycles, with OpenFEC per-firm pulls retained as a cross-check. These files capture itemized contributions from individuals (filtered to
ENTITY_TP == 'IND') to all federal recipients: candidate campaigns, national and other party committees, leadership PACs, and super PACs. PAC and corporate-treasury giving is excluded. FEC pulls were completed 2026-04-26. - City donor records. NYC CFB Campaign Contributions full-cycle pulls for 2021 and 2025. The 2025 pull was completed 2026-04-27, so later filings, amendments, and processing updates can change the 2025 total.
- Campaign-capacity records. CFB finance-summary totals were captured from the page stamped 2026-05-01 14:48 ET. The public-funds and BOE results merge was built 2026-05-03.
2. Employer normalization
- Source field. The FEC
contributor_employerfield is self-reported and inconsistent. - Classifier path. Roughly 45,000 raw employer strings are normalized to Partnership canonical names with a token shortlist, a cached Claude Haiku classifier, a tail-classification pass, and a small exact alias table for short names such as EY, AMEX, NBC Universal, and Starbucks.
- Alias rule. The alias layer is exact-match only after light normalization; it is not fuzzy substring matching.
3. Donor-crossover layer
- Federal filter. Individual, real-money rows only; NY state; NYC-borough ZIPs; employer mapped to the Partnership list when computing the large-employer group. The FEC file is itemized-only, so donors below the federal disclosure threshold are outside the visible universe.
- Local donor file. NYC Open Data Campaign Contributions, Socrata endpoint
rjkp-yttg. The headline cycle pairs use full-cycle 2025 CFB records for the 2024 federal cohort and full-cycle 2021 CFB records for the 2020 federal cohort. - Local filter. Positive individual contribution rows, excluding refunds and non-individual contributor codes. This file answers the headline question: whether the same person gave to any NYC city campaign.
- Person key. A federal donor counts as local when the same full first name, last name, and ZIP5 appears in the CFB donor file. Middle names, suffixes, and honorifics are dropped, so “James R. Smith” and “James Smith” collapse to one person in the same ZIP.
- Firm rule. The headline Partnership statistic uses employer only on the federal side, to define the cohort. It does not require the city-side employer field to match.
4. Campaign-capacity layer
- Campaign totals. Source: the NYC CFB 2025 Campaign Finance Summary page, parsed by
scripts/analyze_local_campaign_capacity.py. Official candidate-level private funds, public funds, spending, estimated balance, and outstanding loans are the totals used when the article discusses how much campaigns raised. - Row-level gifts. Direct candidate-gift diagnostics are deduped by recipient, schedule, and CFB reference number. Row-level contribution sums are useful for gift-size diagnostics, but official CFB summary totals are the campaign-total source of record.
- Public funds. Source: CFB Campaign Public Funds Payments, Socrata endpoint
u69g-mvrb. The phase-specific fieldsprimarypay,generalpay, andrunoffpayare used where present. Early-payment rows are kept separately and used only when no phase-split row exists. - Vote results. Source: official NYC BOE 2025 election results from
vote.nyc. BOE RCV primary HTML pages provide first-round, final-round, and winner fields; BOE general-election recap CSVs provide general-election votes, percentages, and winners. - Capacity join. The public-cap appendix joins CFB candidate totals, CFB public-funds payments, CFB row-level gift diagnostics, and BOE vote results by normalized candidate name and office. Exact candidate-name matches are used first; an unambiguous first-initial plus last-name match is used only when it points to one candidate in the same office.
- Competitive-race screen. A race phase is counted as competitive only when (a) at least two campaigns in that same primary or general contest pass the viability screen, AND (b) the top two vote-getters finished within 15 percentage points of each other. A campaign is treated as viable if it received at least half of the phase-specific public-funds cap or cleared
1 / (n + 1)of the initial vote in its BOE contest. The 15-point margin gate keeps fundraising blowouts out of the “competitive” list: a runner-up who qualified for full matching funds but lost 85 to 8 (e.g., the CD 35 primary 2025) clears viability on fundraising alone but fails the margin gate.
5. Disclosure thresholds
- Federal threshold. The FEC threshold defines the federal donor universe. Smaller federal donors may exist, but they are not disclosed by name and cannot be matched person by person.
- City records. NYC CFB records capture very small city gifts from participating campaigns. In the CFB contribution file used here, 68% of city contribution records are under $200, 29% are under $50, and the smallest recorded contribution is $5.
- Interpretation. A missing city match is unlikely to be a hidden sub-$200 city gift. Even a small city contribution would normally appear in the public CFB records.
6. Reproducibility and AI use
- Reproducibility. Pipeline code, committed aggregate inputs, source URLs, and run instructions are at github.com/ramongonzalez12/dollars-just-for-dc. Raw FEC, CFB, and BOE pulls are regenerable and are kept out of git when they are large.
- Disclosure: heavy use of Codex and Claude. AI is used in two different ways here.
- Classification: Claude Haiku is the production employer classifier that maps free-text employer strings to Partnership canonical firms, and Claude also helps assign firm subsectors for the industry rollup. Those classification decisions are cached or committed for audit.
- Coding and presentation: Codex and Claude helped write, refactor, and debug the Python pipeline, add tests, draft prose, and refine the page design. The headline rates, campaign totals, public-cap calculations, vote-result joins, regulatory citations, and matching-rule decompositions are computed by deterministic Python code in this repo and checked against the underlying public records where practical.
A4. Compliance regime classification
These rules do not make every local donation illegal. They create a severe business penalty: after certain employee contributions to covered officials, a firm can be barred from compensated government work for two years.
- The legal setup
- SEC Rule 206(4)-5 covers investment advisers. MSRB Rule G-37 covers municipal-securities dealers and municipal advisors. Both regimes can impose a two-year business ban after certain contributions to officials who can influence adviser, dealer, or municipal-advisor selection.
- Who gets swept in
- The legal trigger is a contribution by the firm itself, a covered executive, a solicitor, a municipal finance professional, a municipal advisor professional, or a controlled PAC. Many firms apply pre-clearance to a broader employee population than the rule itself reaches.
- Why NYC offices matter
- The Comptroller is the sole fiduciary of the five NYC pension systems (NYCERS, TRS, Police, Fire, BERS), and the Mayor appoints trustees to several of those boards under the NYC Charter. Either office can therefore influence the selection of advisers, dealers, and municipal advisors handling pension assets, which is the activity Rule 206(4)-5 and MSRB G-37 are designed to police.
- What employees see
- For many employees, the practical barrier is not the statute itself but the compliance layer built around it: all-employee pre-clearance, low internal thresholds, conservative interpretations, or a culture that avoids state and city donations entirely.
- What the data show
- In the latest cycle, employees at firms without direct pay-to-play restrictions gave locally at nearly twice the rate of employees at directly restricted firms, 9% versus 5%. That’s roughly half the gap among restricted firms — but the unrestricted-firm rate is still only 9%.
Show legal backup and classification detailRules, firm practice, and implications
The notes below document the rule text, firm-practice assumptions, sector classifications, and empirical literature behind the plain-English summary above. Open an item for the relevance note.
-
The federal “pay-to-play” rule for investment advisers. After a “covered associate” (defined at §275.206(4)-5(f)(2) as a general partner, executive officer, or any employee soliciting government business, plus their direct supervisors) contributes to an official of a government entity who can influence the entity’s selection of investment advisers, the firm is barred from receiving compensation for advisory services to that entity for two years. The rule includes a de minimis exception: contributions of $350 per election to a candidate the covered associate is entitled to vote for, and $150 per election to other candidates, do not trigger the ban. The narrowness of those carve-outs and the severity of the two-year ban are the practical reasons many advisers extend pre-clearance beyond the rule’s covered-associate population.
-
The MSRB analog to Rule 206(4)-5, covering broker-dealers in municipal securities and (since 2016) municipal advisors. After a “municipal finance professional” or “municipal advisor professional” contributes to an issuer official with influence over the selection of dealers or advisors, the firm is barred from municipal-securities business or municipal-advisory business with that issuer for two years. FINRA Rule 2030 extends a parallel two-year ban to broker-dealers acting as placement agents for investment advisers.
-
Why NYC officials are coveredThe NYC Comptroller is the sole fiduciary of the five NYC pension systems (NYCERS, TRS, Police, Fire, BERS), and the Mayor appoints trustees on several of those boards under the NYC Charter. Either office can therefore influence the selection of investment advisers, broker-dealers, and municipal advisors handling pension assets — the activity Rule 206(4)-5 and MSRB G-37 are designed to police. Whether other NYC offices (Council, Borough Presidents) are reached by the rules is fact-specific and turns on each office’s actual influence over pension-system, dealer, or advisor selection.
-
Firm policies often go beyond the ruleMany large registered investment advisers, broker-dealers, and municipal advisors implement pay-to-play policies broader than the rules formally require — commonly applying pre-clearance to all employees rather than only the covered-associate population, and in some cases prohibiting state and local contributions entirely. Federal-level contributions are generally outside the scope of these policies because Rule 206(4)-5 and MSRB G-37 are triggered only by contributions to officials of government entities that hire advisers or dealers. Firm-by-firm policies vary and change over time; readers can pull any registered firm’s current Form ADV Part 2A Item 17 disclosure from IAPD to see exactly how a given firm implements the rules. One published example: AllianceBernstein’s SEC-filed Code of Ethics requires “all employees” to pre-clear personal political contributions with the Compliance Department, including contributions made by spouses and dependent children — meaningfully broader than Rule 206(4)-5’s covered-associate scope.
-
Why certain sectors are classified as “partial exposure”The Partial exposure bucket covers sectors where the firm itself is not a registered investment adviser but a meaningful share have practice-area or subsidiary exposure that triggers equivalent restrictions:
- Law firms. Muni-bond counsel groups can fall within the statutory definition of municipal advisor under Dodd-Frank Section 975 (amending Securities Exchange Act §15B), bringing them within MSRB Rule G-37’s scope.
- Audit and consulting firms. Public-sector engagements with pension systems and issuer entities can create direct or contractual pay-to-play exposure for the practice groups handling those clients.
- Insurance holding companies. Many insurance holding companies own registered-adviser subsidiaries; firm-wide political-contribution policies often extend across the corporate group rather than only to the regulated subsidiary.
- Broker-dealers and adviser-affiliated platforms. Broker-dealers acting as placement agents for advisers, and any firm holding a broker-dealer or investment-adviser registration, fall within the reach of MSRB G-37 and FINRA Rule 2030.
-
SEC and MSRB pay-to-play enforcement historyTwo anchor cases: TL Ventures (2014) was the first Rule 206(4)-5 enforcement action; the SEC ordered $256,697 in disgorgement, $3,197 in prejudgment interest, and a $35,000 civil penalty after a covered associate contributed to candidates for Mayor of Philadelphia and Governor of Pennsylvania. Earlier, the 2012 Goldman Sachs / MSRB G-37 settlement resolved charges that a Goldman vice president made undisclosed in-kind contributions to Massachusetts state treasurer (and 2010 gubernatorial candidate) Timothy Cahill while Goldman was engaged in muni-securities business with the state; the firm paid roughly $14–16 million across SEC, MSRB, and Massachusetts components. Smaller Rule 206(4)-5 actions have followed in subsequent years.
-
The direct implication for this analysisThe compliance section above splits Partnership member firms into three buckets. Restricted: Investment Bank, Bank, Asset Manager, PE, Hedge Fund, Broker-Dealer (direct 206(4)-5 / G-37 exposure). Partial: Big Law, Big Four, Consulting, Insurance, Crypto (practice-area exposure and voluntary firm policies). Unrestricted: Software, Real Estate, Consumer, Media, Utilities, Telecom (no federal pay-to-play regime). In the latest cycle, the unrestricted bucket was nearly twice the restricted bucket: 9% versus 5%. The compliance gap is associated with regulatory exposure, not necessarily caused by it — industry composition, donor-pool size, and other firm-level factors plausibly contribute. Either way, the gap is not enough to explain the whole federal-vs-local gap, since the unrestricted-firm rate is still only 9%.
A5. Campaign-capacity test
The body section above shows the headline answer and the by-level breakdown. This appendix subsection covers how the test was constructed, what it does and does not test, and the sensitivity of the result to looser screens.
What counts here. The default screen has three steps:
-
Viable campaign, imperfectly measured.
A rough proxy — not a judgment about every campaign. A campaign must have drawn public matching funds and passed either a fundraising-strength or vote-share screen.
- Fundraising strength: it had already received at least half of the public-funds cap for that primary or general contest.
- or
- Vote-share strength: it cleared
1 / (n + 1)of the initial vote in its BOE contest. For example, that means 17% in a five-candidate race, or 33% in a two-candidate race.
-
Competitive race.
A race counts as competitive when both of the following hold:
- At least two viable campaigns in the same primary or general contest passed one of the screens above.
- and
- Within 15 percentage points at the ballot box: the top two finishers ended within 15 percentage points of each other — tight enough that more donations to a runner-up could have plausibly mattered.
- Unused capacity ? The dollars below count remaining public matching-funds capacity for viable campaigns in competitive races, not the exact private dollars needed to unlock the match.
Method details, sensitivity, and source notes
The default view is intentionally conservative: it drops uncontested or weakly contested races, zero-payment phases, and campaigns that did not show either fundraising strength or initial-vote strength. If less competitive races are included too, the estimate rises to $19M in the primary and $24M in the general.
The next regular Council cycle is 2029, not 2027. The Charter schedule puts all 51 Council seats back on the regular four-year cycle, and the current Council composition table marks 36 districts as term-limited in 2029.
The cap test uses CFB public-funds payment fields split by primary and general election phase. When the payment file lists a separate early-payment row, the pipeline uses the phase-specific amount when present and uses early payment only when no phase split exists. The phase-payment file and the CFB finance-summary page do not always reconcile exactly candidate by candidate, so cap usage is based on the phase-payment file; the reconciliation difference is retained in the processed CSV.
A campaign below the public-funds cap may still face eligibility, timing, spending-limit, contribution-limit, or strategic constraints. The cap test answers only whether the public-funds cap itself was binding. The default room screen excludes candidate-phases with zero phase-specific public funds (because zero-payment phases are not strong evidence of usable matching-cap room) and excludes races where fewer than two campaigns in the same primary or general contest pass the viability screen. The 1 / (n + 1) screen is post-hoc and uses matched CFB/BOE candidate result rows grouped by BOE contest. The 50% public-cap rule uses phase-specific public funds received as a share of the phase cap.
A6. Comparison with prior research
No prior work on this question. We didn’t find a published study that measures, for any city or geography, what share of federal donors also give locally. Cross-level studies in the existing literature tend to look at aggregate dollar flows, or to study one level (federal or local) in isolation. The closest published analogues — donor-geography work, pooled cross-level donor databases, and pay-to-play studies on narrower regulated populations — answer different questions.
Closest related study: THE CITY’s Mamdani-by-employer analysis. On April 24, 2026, THE CITY reported that 1,971 Partnership-firm employees gave about $242,000 to Zohran Mamdani’s 2025 mayoral campaign. THE CITY classifies city-form employer strings; this project classifies federal employer strings. That makes their work a useful cross-check on our employer classifier: if we point the same classifier at the city-side employer field and run their single-candidate slice, do we land in the same place?
Yes, very close — within 3 donors out of 1,971, and within 1% on dollars.
| Check | Donors | Dollars | What it shows |
|---|---|---|---|
| THE CITY reported | 1,971 | ~$242,000 | Published benchmark. |
| Our classifier on city employer field | 1,968 | $244,591 | Reproduces 99.8% of donors. |
The Google donor set illustrates the remaining pattern. Our repo finds 364 positive Google contribution rows that collapse to 271 donor keys and $56,058; THE CITY reports 323 donors and $55,727. The dollars are essentially identical; the donor count falls between our person-key count and row count, which points toward de-duplication or employer-string treatment differences rather than missing money. The match is tight enough that the same classifier, applied to the federal employer field for this page’s headline analysis, can be trusted the same way.
2025 data lag. This dataset was pulled on 2026-04-27, when the 2025 NYC cycle was still being filed and processed. Later filings, amendments, and CFB processing updates can raise the 2025 totals, so the 2024 → 2025 comparison should be read as the public-record picture as of that date.
Closest published analoguesBonica (DIME); Heerwig & Zhang; Briffault; Brennan Center / CFI; NYC CFB; Boatright; Fulcrum / OpenSecrets; CFI / NIMSP; Hill & Huber
-
Introduces the DIME database, which pools federal, state, and (some) local individual donor records from 1979 onward. This is the data infrastructure that makes a cross-level overlap analysis technically possible. Bonica and collaborators establish that a "common pool" of donors gives across levels, but they do not publish the cross-level overlap rate for any specific city, employer, or residence-restricted subpopulation.
-
Documents the "constituent vs. contributor" gap: non-constituent (often out-of-state) money dominates itemized giving at every level of government. The closest frame to this finding, but it measures geographic mismatch at the aggregate level, not individual-level federal/local donor overlap.
-
The closest work on the corporate-employee angle: studies federal congressional giving by corporate executives from 1980 to 2014. Federal-only; does not ask whether those same executives also give to state or municipal races.
-
Census-block analysis of NYC City Council vs. State Assembly small donors. Finds City Council neighborhoods more racially and economically representative than Assembly ones. Characterizes the NYC local donor pool in detail but does not cross-reference FEC records or ask about federal donors' local participation.
-
Official NYC Campaign Finance Board characterizations of the local donor pool: small-donor share, geographic distribution, matching-fund effects. Treat the NYC donor universe as self-contained; never compared against the FEC population.
-
Four-city Massachusetts study of municipal donors. Finds local donors are unrepresentative and network-sorted. Does not compare against the federal donor population or measure the same individuals across levels.
-
Finds that roughly 77% of federal campaign funds come from 5% of ZIP codes; in-district share of federal giving has collapsed. Directionally supports the "localism gap" framing but measures the flow of money into federal races, not whether federal donors also give locally.
-
Notes that state races are 87 to 90% in-state funded while federal races are increasingly out-of-state. Treats the federal and state donor pools as separate aggregates, without linking individuals across levels.
-
Survey-based study of donor ideology, demographics, and motivations. Does not address participation across levels of government.
Three design choices make this finding unusual in the existing literature: restricting the universe to NYC residents (most cross-level studies use statewide or national frames), restricting to employees of a defined set of companies (Heerwig and Zhang study executives but federal-only), and measuring the non-participation rate rather than the aggregate dollar flow.