Last week, Bloomberg reporter Steven Dennis tweeted  an idea for a “fun data project:” “For every senator give them a deficit score. / What is the net deficit as scored by CBO/JCT of all the *laws* they voted for, adjusted for inflation.”
As it turns out, Spending Tracker essentially does this—with a few important differences. First, the Spending Tracker doesn’t adjust for inflation—we simply take CBO’s numbers at face value. Whether or not you want to consider numbers in inflation-adjusted terms depends in part on what question you want to answer.
On the one hand, inflation-adjusting helps make more accurate comparisons across time—it’s necessary to contrast the relative “fiscal conservativeness” of someone who served in office decades ago with someone serving today. For this reason, we aim to inflation-adjust CBO’s figures in the future as our database grows. On the other hand, the national debt itself is not inflation-adjusted—a dollar of debt ultimately has to be paid back regardless of when it was accrued. Thus, looking at raw figures may be more appropriate for assessing who is most responsible for the existing national debt.
Regardless, the bigger difference between Spending Tracker and what Dennis describes is that we track spending, not deficits. In other words, the site currently doesn’t include revenue implications of legislation. We do, however, crunch these numbers in spreadsheets alongside Spending Tracker, and while the site doesn’t track deficits per se, the two are highly correlated.
For the current session in the Senate, deficit scores correlate 96.8 percent with the spending score displayed on Spending Tracker. In the House, that correlation is 89.8 percent. That’s because spending is by and large the primary driver of the national debt. Relatedly, it’s also because Congress tends to vote overwhelmingly more on bills that impact outlays rather than revenues. Of the bills signed by the president that were judged by CBO to have either a spending or revenue impact, 98 percent had an effect on outlays while only 41 percent affected expected revenue.
In a follow-up tweet , Dennis offered a prediction for how he thought ranking members by deficit scores would shake out: “Pretty sure zero U.S. senators have a negative net deficit score (aka a surplus). And no, bills that don’t become law do not count. Yoda rule: Do or not do.”
Looking closer at the data, Dennis’s conjecture is pretty much true—with an important caveat.
The Senator who’s voted to add the least to the deficit this session is Rand Paul (R-Ky.). Paul has voted so far to add about $494 billion to the national debt; close behind him is Sen. Mike Lee (R-Ut.) with $502 billion.
By comparison, three Senators—Jack Reed (D-Ri.), Brian Schatz (D-Hi.), and Sheldon Whitehouse (D-Ri.)—have voted to add almost $3.3 trillion to the national debt this session of Congress alone. $2.3 trillion comes from spending increases and another $957 billion from revenue decreases. Naturally, a big driver of deficit and spending scores this session are votes on the various pieces of coronavirus relief legislation; the CARES Act  alone increased spending by more than $1.3 trillion while also cutting revenue by another $445 billion.
The tables below show the Members who voted to add the most and the least to the national debt during the current session of Congress. You can compare these deficit rankings to the spending-only numbers here .
Voted for the Least New Debt, 116th Congress
Voted for the Most New Debt, 116th Congress
|Tie between 22 D’s||$3,270,989,260,000|
|16 R’s and 1 Ind.|
As for the important caveat mentioned earlier: in addition to counting bills that are voted on via a roll call, Spending Tracker also incorporates bills that are approved by the entire Congress (voice vote in the House; unanimous consent in the Senate), by adding their scores to all Members equally.
Normally, only legislation with small budgetary implications are approved unanimously, because they are typically less controversial. If a bill proposes to spend a lot or make radical tax changes, it almost certainly will not have the level of support necessary to avoid a roll call being taken.
This tendency has been upended a bit in the era of coronavirus, with portions of significant bills being passed by voice in the House or unanimous consent in the Senate. For example, in the House, the CARES Act was approved by voice, over the now-infamous objections of Rep. Thomas Massie (R-Ky.). In the Senate, the CARES Act was voted on by individual members, but H.R. 266 , the “plus up” for the Paycheck Protection Program, was not.
That bill carried an estimated cost from CBO of about $483 billion, which was added to every Senator’s score. But if we exclude this bill and others passed in the Senate by unanimous consent, we find that Sen. Rand Paul has actually voted to cut the deficit by about $2.2 billion. He is the only Senator to do so during this session of Congress—though Sen. Mike Lee is not far behind, having voting for a net debt increase of only $5.9 billion. In the House, one Member this session, Rep. Ken Buck (R-Co.), has voted for a net deficit cut (-$1.4 billion).
Of course, many members, if given the chance, would vote to reduce the national debt by far more (either through reducing spending or increasing revenues), but they either (1) aren’t given the chance to vote on such legislation, or (2) those bills aren’t enacted when representatives or senators do get to vote on them.
In a future post, I’ll explore further what these figures might look like if we examined all votes taken—not just bills that were signed into law—as well as how member records, both with respect to spending and deficits, have changed over time.
- “tweeted”: https://twitter.com/StevenTDennis/status/1285725267269423105
- “tweet”: https://twitter.com/StevenTDennis/status/1285727130303762435
- “CARES Act”: http://spendingtracker.org/bills/hr748-116
- “here”: https://spendingtracker.org/rankings?chamber=senate&year=10&baseline=with_baseline&congress=116&status=enacted&ajaxed=true&page=all
- “H.R. 266”: https://spendingtracker.org/bills/hr266-116