Back in 2003, using publicly published data from Congressional Quarterly, we averaged a couple of different types of scores that they provided. We examined all votes going back to January 1, 1991. That gave us our original progressive cohort. Members are added to or subtracted from the progressive cohort as follows: Members of the House and Senate subsets of the cohort are removed as they leave that body of Congress for any reason. They can also be removed if their voting record becomes more conservative over time and their lifetime crucial vote score drops below 91.5%. Newly elected members are added if their crucial vote score is 91.5% or higher for their first year. Members of a body can be added to the progressive cohort after their first year if their lifetime crucial vote score moves above 91.5%.
Our current list includes 13 hard-core progressive United States Senators (13% of that body) and 33 hard-core progressive United States Representatives (about 8% of that body). The algorithm that we've used to come up with these progressive scores is as follows: We take ANY VOTE in which a majority of the progressives we've identified--so in the House say, if there were no absences, it would be 17 of 33--voted in opposition to a majority of the Republican caucus and have that vote qualify for the database. The same process is used in the Senate. So, non-ideological votes such as National Groundhog Day: 429-0 with 6 absences, do not qualify for the database. ANY vote in which a majority of progressives in the progressive cohort listed just below here votes against a majority of Republicans qualifies for the database and is included in the Overall % scores.
"The Progressive Position" by definition, is the position of the majority of the Progressives. The “Conservative Position” is the position of the majority of the Republicans. We’ve tested this algorithm in the real world and it works extremely well. In the case of members of Congress elected before November 1990, the “Progressive Lifetime Scores” include only votes cast in Congress since January 1, 1991 (1991-92 was the first full Congress where vote records were computerized). In the case of members of Congress elected on or after November 1990, the scores include all votes that have ever been cast while that member has been in Congress. The column labeled “Progressive ‘19-‘20 Scores” is for the current Congress and shows scores for votes since January 2019, which allows for an apples-to-apples comparison for the same time period of all current members of Congress. For example, the total number of qualifying votes according to this criteria in 2007 was 747 in the House and 269 in the Senate. By searching on an individual member of Congress, you will be able to see the specific roll call vote numbers of the votes that qualified for inclusion on Progressive Punch scores. The composite scores include ALL votes qualified by our algorithm, whether we've written the narrative vote descriptions that allow us to put them into categories or not (categories were discontinued about 2015). So the category scores can look different from the composite scores.
The votes used to calculate the scores in the Crucial Votes % columns are a subset of the overall votes that qualify according to the Progressive Punch algorithm described above. They show the impact that even a small number of Democrats have when they defect from the progressive position. Crucial Votes are really “where were you when we needed you” votes. The Crucial Votes % categories include all roll call votes where the margin between yes votes and no votes was narrow and could have been changed by a small group of Democrats voting differently. Narrow margins are defined as votes in which the winning side came out ahead by 20 votes or fewer in the House (so a shift of 10 votes from one side to the other would have changed the result) or by 6 votes or fewer in the Senate (so a shift of 3 votes from one side to the other would have changed the result).
Similarly, we treat absences differently depending on whether the vote was close. If the absence was on a vote where the overall margin was less than or equal to 20 votes in the House or six in the Senate then we count it as a “bad” vote, which gets counted against the member's score. However, if the margin was greater than 20 votes — say, 320–80 — then we remove the absence from the numerator and denominator, and it won't count at all. Conversely, if a member votes “Present” rather than “Yea” or “Nay,” we will always count that as a bad vote as long as the vote qualifies for the data base at all.
Prior to 2011, we had also included in the Crucial Votes % category any vote in which the progressive side was on the losing side, irrespective of how close the vote was. However when Republicans are in a majority in either the House or Senate — due to their strong ideological cohesion in their voting patterns — they're almost invariably on the winning side. That is to say they stick together and when they're in a majority that sticking together is enough for them to win most of the time. So we're actually using slightly different methodology for determining which votes fall in the Crucial Votes % category when the vote isn't a close vote. In situations where the Democrats are in control of a legislative body — as was the case with the US House of Representatives for 2019-2020 — we continue to classify as Crucial any vote in which the progressive side was on the losing side, in addition to all close votes as defined above. However, in situations where the Republicans control a legislative body — as is the case with the US Senate for 2019–2020 — we classify votes where at least three-quarters of the progressive cohort listed just below voted against the Republicans AND at the same time at least 10% of the Democratic Caucus voted with the Republicans as a Crucial Vote, along with still including all close votes as well. In other words “Crucial” votes are votes in which there was strong progressive cohesion and at the same time a significant defection on the part of more conservative Democrats to the Republicans. Summing all this up, “where were you when we needed you” votes would probably be a better title for this column than Crucial Votes % but the word crucial fits much more easily at the top of a column. We have made this calculation retroactive to all previous periods where the Republicans controlled the Senate or the House in the past; this change affected Members' Lifetime Crucial Votes % scores very minimally.
Votes that otherwise conform to the algorithms above have been eliminated if they're pro-forma votes to either adjourn the House or approve the House journal, unless the vote qualified as both a Crucial vote and an overall vote, in which case it was retained in both categories.
There is no surefire objective way to compute how progressive, or for that matter how conservative, a member of Congress is. A lot of thought went into coming up with this methodology. That doesn't mean it can't be critiqued. What we have done is to try to take human beings out of the equation as much as possible. In other words, the percentages calculated on this site do not necessarily correlate with the individual political positions of Joshua Grossman, the primary author of this website. There are some criticisms that could be levied against our methodology. One is that it treats every vote equally, when they're obviously not all equally important. Another is that lonely principled stands, that might be viewed by some as progressive, such as Barbara Lee's sole vote against war in Afghanistan, do not qualify for the database, because not enough Progressives rallied around her flag (no pun intended). One other thing that no voting index can measure is intensity of support/leadership.