Follow the Money

Think of it like a seesaw. Each bill is weighed: does its text name more everyday beneficiaries (veterans, families, workers, small businesses) or more institutional ones (federal agencies, regulators, administrative offices)? The average of all those weights per congress becomes one dot on the chart.

Zero = the average across all 7 congresses. A dot above zero doesn't mean that congress was "good" — it means its bills named people-type beneficiaries more often than the historical norm. Below zero means institutional beneficiaries showed up more. This removes the model's built-in lean so you see real shifts, not artifacts.

Solid line = everything introduced. This includes all bills, even the thousands that never received a vote. Dashed green line = only the bills signed into law. When these two lines diverge, it tells you something: Congress may introduce a lot of people-focused bills, but the laws that actually pass may lean a different direction — or vice versa.

D and R lines break the same data by sponsor party, so you can see if the two parties propose legislation with different beneficiary profiles.

Green bars at the bottom show the law passage rate — what percentage of introduced bills actually became law. Most congresses enact only 2–4% of introduced legislation. Hover or tap any bar or dot for exact numbers.

Filter by presidency to highlight which congresses overlapped with a given administration. Each president spans at least two congresses. The filter highlights the relevant sessions so you can compare across administrations.

Beneficiaries below the chart list the specific entities the model found. Click any dot to see that congress's top beneficiaries.

Dollar-weighted beneficiaries go further — for bills that mention specific dollar amounts near an identified beneficiary, the model associates that amount with the entity. This shows not just who legislation names, but how much money is referenced near each group. Coverage varies by congress (strongest for the 113th–118th where bill summaries are most detailed).

Note: These scores are generated by an automated model analyzing bill text. A bill that creates a new federal program may score toward “government” even if its ultimate purpose is to help people. Dollar amounts reflect what bill summaries reference, not final enacted appropriations. The scores are one lens for exploration, not a definitive judgment.