(Originally published to the My Old Kentucky Podcast Newsletter)

Ideology, for the most part, is subjective. Despite that, there are a lot of legislators who might consider themselves the “most progressive” or “most conservative.” Several years ago, I did some work to try to determine quantitatively where legislators fall on the left-right scale, by using an algorithm named W-NOMINATE.

W-NOMINATE was developed in the 1980s by a pair of political scientists who combined, for the first time, roll call votes with statistical software. They were able to identify clearly that two dimensions exist within American political behavior, and their ability to suss when and why Southern Democrats, Northern Democrats, and Republicans all voted differently managed to win them several awards as well as the respect of academics of all stripes, who were beginning to bring a level of quantitative rigor to social studies.

Nowadays, the algorithm is available to anybody using just about any open source software. As long as, importantly, that person also has access to a dataset of roll call votes. Here in Kentucky, that list did not exist until 2016, when I started creating it. If you’ve ever been to kypoliticaldata.com, you know that there is information about votes in the legislature. What you might not know is that in order to get at that data, I wrote multiple web and pdf scrapers that collect that data nightly from the LRC website. Scraping that data allows me to create the data needed to run the W-NOMINATE algorithm over Kentucky’s legislative sessions. (As an aside–back in 2018, I worked directly with the LRC to try to have this data released by them directly. They did make it available – for one day – before it was taken down. The LRC Director was relieved of his duties a few months after that, and since then, the agency has become significantly more partisan than it was when Democrats had some control).

The first time I ran W-NOMINATE over the legislature and released the results via social media, it caused a pretty fun uproar. In the early days of MOKP, we had to work pretty hard to get legislators to pay attention to us – but not on the day I released a list showing how liberal or conservative they were. Lots of legislators reached out and asked a lot of questions about how I put the list together. I am not sure all of them were very happy.

Nevertheless, putting together the W-NOMINATE scores is a useful tool for determining how the legislature operates, and which legislators are most effectively voting in ways that represent their districts. Here are the 2025 results:

In the House, we can see a significant spread in the Democratic caucus, with progressive urban legislators like George Brown, Lindsey Burke, Sarah Stalker, Pamela Stevenson, Mary Lou Marzian, and Anne Donworth showing as the most progressive. Ashley Tackett Laferty is the most conservative Democrat, which makes sense as she is the only Democrat from Appalachia. Democrats from swing districts like Rachel Roarx and Adam Moore also tend to be more conservative.

On the Republican side, the clear story is that even though the caucus is very large, a significant group exhibits very similar ideology. There is a very clear caucus on the far right (we often call them the “Liberty” caucus), which consistently includes Savannah Maddox, Steven Doan, and Felicia Rabourn. Three other legislators (T.J. Roberts, Josh Calloway, and Marianne Proctor) join the far right caucus on a regular basis. Several Republicans form a group on the left side of the GOP caucus, including Scott Lewis (Owensboro), Kevin Jackson (Bowling Green), Kim King (NKY), and Denanna Frazier Gordon (Richmond). Most of the Republicans on the left side of the caucus come from smaller cities or Northern Kentucky. 

On the Senate side, the story is pretty clear: the Republicans are very much in line with one another, with potentially a small exception in Jimmy Higdon. On the Democratic side, Robin Webb is actually to the right of average (similar to Ashley Tackett Laferty, she is the only Democratic Senator from Appalachia). Besides her, the only Democrat even slightly out of line with average is David Yates, whose district is the least safe of any of the Democrats from Louisville or Lexington (his district is still quite safe).

There is plenty we can learn from these analyses, but the biggest takeaway from me is that the leadership of the GOP directs the legislative sessions in such a way that they can exhibit enough discipline to form an impregnable legislative force. That’s not really news, but any hope that the Democrats could peel away more moderate legislators in order to have more control over events in Frankfort (such has been done in recent years in Ohio and Alaska) seems silly after looking at W-NOMINATE scores. The way to control Frankfort is to win elections.

Our legislature is not a transparent one, and as Republicans become more and more entrenched, they become less and less likely to let the public know about how our government operates. Pulling apart the archaic and obtuse data that they let us have is one way I try to bring some sunlight to how Frankfort operates. I hope this is useful to you! I am always happy to answer questions about these analyses or hear observations that I’ve missed. Let me know if you have any.