Methodology

This site makes an argument, and it's good to be transparent about it. Every number here is reproducible from public data with the scripts in the repository. Data generated 2026-07-24.

Sources

• FEC bulk data: committee-to-candidate contributions, committee master, and candidate master files for the 2024 and 2026 cycles. Public domain.
• unitedstates/congress-legislators: the roster of sitting members and the FEC-ID crosswalk that joins members to their campaign committees. Public domain (CC0).
• Endorsements, scraped from the Internet because no dataset exists. The 12 organizations on the endorsements page were scanned one at a time from published lists and reporting, and stored as 497 entries in the repository, each carrying the URL it came from. Nothing is counted without a citation, and every organization has a written note saying how complete its list is. This is the least reliable data on the site and the notes say so individually.

What counts as money here

Only direct contributions from a committee to a member's campaign committee (FEC transaction types 24K and 24Z). That is $755M across 4,566 committees and 537 members.

Of that, $106M is excluded from every share this site reports: leadership PACs, joint fundraising committees, transfers between members' own campaign committees, and party committees. That money is the political system moving money within itself, not an outside interest funding it, and counting it would answer a different question than the one this site asks. It also fluctuates with who holds which seat from cycle to cycle, which makes it noise against the thing being measured. Every reported share is therefore against the remaining $649M of outside-interest money.

The honest caveat: that money came from somewhere. Industry PACs give to a member's leadership PAC, which then gives to colleagues, so excluding it hides a channel through which industry money reaches members indirectly rather than eliminating that influence. Tracing it properly would mean following committee-to-committee transfers and apportioning a leadership PAC's receipts across its disbursements. That would be interesting to see, but not really important to what we're trying to do here.

Addressing the whale in the room: we do not pull in spending information from Super PACs and the large amounts of dark money circulating in the system. This is a conscious choice because the scale of that money is beside the point. If a bank writes you a check, we can reasonably deduce that you're favorable to their interests. Beyond that, the $5M they spent against your opponent only tells us how favorable.

What "membership" means

A member belongs to a sector if any PAC in that sector gave them any amount. That is a deliberately loose definition, and it produces a genuine finding: on this definition nearly every member belongs to nearly every major business sector. It also makes the map useless as a picture, because if everyone belongs to everything, position cannot distinguish anyone.

So the map defaults to weighting each tie by dollars, and you can switch it to the binary definition to see the collapse for yourself. Both views are drawn from the same underlying data; the pipeline records the dollar amount on every member-to-PAC tie regardless of which view you pick.

How PACs are grouped into sectors

Sector names and boundaries follow the 13-sector taxonomy used by OpenSecrets, plus separate buckets for party committees and leadership PACs. Committees are assigned in this order:

1. Hand review. The 100 largest committees by dollars were each reviewed individually. Where review disagreed with the automatic rules, the human call wins and is recorded in the repository with a written rationale.
2. Structure. Party committees, leadership PACs, joint fundraising committees and candidate-to-candidate transfers are classified by their FEC type and designation, not by name. These are internal political ties rather than industry ties.
3. Name rules. Everything else is matched on committee name and connected organization.

Rule-based matching leaves 22.0% of dollars unclassified, shown throughout as "Other": 1,861 mostly small committees, averaging roughly $77k each. That residual is displayed rather than hidden, and it is large enough that you should treat individual sector totals as approximate.

The unclassified bucket is counted in every total but is given no circle on the map. It is real money but we do not know enough about those dollars to judge how it should fit into our chart. We ignore it because it doesn't inform political alignment.

When one PAC is the size of a sector

A single PAC giving more than $5M is listed as its own segment rather than folded into a sector. Two currently clear that bar: AIPAC at $7.9M and the National Association of Realtors at $5.7M. At that size a single organization rivals an entire category. AIPAC alone is roughly 90% of the Lawyers & Lobbyists sector and nearly twice all party committees combined. The intent here is to make the chart as informative as possible while limiting noise.

The rule is a threshold, not a list. It applies to whichever committees clear it in the data, so no organization is singled out by name, and the next largest PAC is around $4.3M. The boundary is not cutting through a cluster. Segments keep the kind of the sector they came from, so business-versus-labor totals are unchanged by the split; only the named lines separate. Note also that concentration like this is not unique: WinRed is 40% of all party committee money and the American Association for Justice is 38% of Lawyers & Lobbyists, which is why the table reports how many PACs sit behind every figure.

A trap worth naming

It is tempting to simplify this kind of analysis by keeping only the largest PACs. Doing so inverts the result. Labor money is concentrated in about twenty large union PACs, while business money is fragmented across thousands of company and trade-association PACs. Inside the top 100 committees alone, labor appears to rival finance as the biggest force in Congress. Measured across all 4,566 contributing committees, business money outweighs labor money by more than five to one.

Sector totals on this site are therefore always computed over every contributing committee. The top 100 are named and browsable because they are the ones we could verify individually, not because they are the whole picture.

The analysis built on top of the data

Most of this site is arithmetic on contributions. One page is not, and it carries its own assumptions.

Ideology splits the belief-driven money into 16 causes by matching PAC names, and deliberately keeps opposed causes apart. A single netted "guns" or "abortion" figure would describe neither side. Name matching is weaker here than for sectors, because a cause is less often stated in a committee's name than an industry is, so treat the smaller cause totals as approximate.

What this does not show

• Not causation. Contribution data shows that money flowed from an interest to a member. It does not show that the member voted a particular way because of it, and this site never claims otherwise. Contributions also follow power. Committee chairs and leadership attract money because of the seats they already hold.
• Affiliation. The contention here is that donors tend to give to candidates they believe will align with their interests, and that members of Congress are unlikely to take money from a PAC strongly opposed to their own ideology. An admitted limitation is that some sectors spread money far and wide, purchasing access if not influence. Still, there is signal in the noise.
• Not all the money. No independent expenditures, no dark money, no state-level spending, no lobbying spending, no bundled individual contributions.
• Endorsement counts are floors, not measurements. The 336 members shown with a documented endorsement are the ones we could find a citation for. No union publishes a national congressional endorsement list, so the labor numbers in particular are certainly low. Read those bars as "at least this many", and read a small bar as possibly meaning the organization endorses quietly rather than rarely.
• Not a scorecard. A large number next to a member's name is not an accusation. Incumbency, committee assignment and seat competitiveness all drive these totals.

Corrections

Sector assignments are judgment calls and some are arguable. Every one of them lives in version control with a stated reason, so a disagreement can be raised against a specific committee and a specific rationale rather than against the site as a whole.