How to Read an FEC Independent Expenditure Filing Without Getting Fooled

If you've ever clicked through an "[INDEPENDENT EXPENDITURE REPORT]" on the FEC website, you've seen the schedule E rows: dozens of entries tagged SUPPORT or OPPOSE, each one naming a candidate, a dollar amount, a purpose, and a date. The format looks simple. It isn't.

This post is a field guide to the underlying data — the same rows that populate LobbyVault's independent expenditure pages — and the three analytical mistakes that turn otherwise good campaign-finance reporting into misleading numbers.

The Key Fields

The FEC's bulk independent expenditure file (independent_expenditure.csv) has a wide schema, but the columns that matter for most reporting are:

The sup_opp column is the single most important field and the most commonly misread.

Mistake 1: Treating "Oppose" as Hostile to the Candidate

An O (oppose) expenditure against Candidate X often helps Candidate X's opponent. When you aggregate by the cand_id field and call that number "spent on Candidate X," you can badly mislead readers.

The correct aggregation for "money that helped Candidate X" is:

SUM(exp_amo) WHERE cand_id = X AND sup_opp = 'S'
+ SUM(exp_amo) WHERE cand_id = Y AND sup_opp = 'O'   -- for each opponent Y

LobbyVault's candidate pages try to surface both views separately: money for (supportive IE + opposition IE against opponents) and money against (opposition IE against the candidate + supportive IE for opponents). The raw FEC export only gives you one side; the flip has to be reconstructed.

Mistake 2: Summing Amendments Without Deduplication

FEC filings are amended frequently. When a committee files an amendment, it does not update the original rows — it files a new report, and the bulk data keeps both. If you simply SUM(exp_amo) across all filings you will double-count every amended row.

The standard defensive move is to partition by file_num and keep only the latest filing per transaction_id. A more conservative approach is to join against the form_type column and exclude rows from superseded reports.

If the numbers you're computing look too high compared to press coverage of a race, duplication from amendments is the first thing to check.

Mistake 3: Trusting Free-Text pur Fields

The pur (purpose) column is free-text. Committees write things like "media", "television", "TV", "MEDIA BUY", "ADVERTISING - CABLE", and "consulting" interchangeably. Building a "spending by category" chart by grouping on the raw pur string produces a long tail of near-duplicates.

A workable normalization is:

• Uppercase everything.
• Collapse any value containing MEDIA, TV, CABLE, BROADCAST, RADIO into MEDIA.
• Collapse MAIL, PRINT, POSTCARD into DIRECT MAIL.
• Collapse DIGITAL, ONLINE, WEB, SOCIAL into DIGITAL.
• Bucket CONSULTING, STRATEGY, RESEARCH into CONSULTING.
• Everything else → OTHER.

This is crude but survives contact with real FEC data.

Timing: When Do These Rows Appear?

Most independent expenditures are disclosed via 24-hour notices (filed within 24 hours of the expenditure) during the 20 days before an election, and 48-hour notices before that. Outside the pre-election window, they appear in quarterly reports.

This means the exp_dat field and the filing date can drift. If you're producing a time-series chart of IE activity, use exp_dat (the dissemination date) for the story-telling axis but sanity-check against file_num's filing date when looking for reporting anomalies.

Putting It Together

A defensible "top IE spenders of the cycle" query looks roughly like:

SELECT cmte_id,
       SUM(CASE WHEN sup_opp='S' THEN exp_amo ELSE 0 END) AS supporting,
       SUM(CASE WHEN sup_opp='O' THEN exp_amo ELSE 0 END) AS opposing,
       SUM(exp_amo) AS total
FROM fec_independent_expenditure
WHERE exp_dat BETWEEN '2026-01-01' AND '2026-12-31'
  AND is_superseded = 0             -- amendment handling
GROUP BY cmte_id
ORDER BY total DESC
LIMIT 100;

Most of LobbyVault's committee pages run a version of this query, then layer on the support/oppose breakdown per targeted candidate.

The Bottom Line

The FEC publishes enough data to answer almost any question about independent-expenditure spending — but the schema rewards readers who understand the distinction between cmte_id, cand_id, and the sup_opp flag, and who deduplicate amendments. Most of the headline numbers you see in campaign-finance coverage are a rough approximation; the interesting story is usually one layer deeper in the raw rows.