Watching Steve Kornacki gestating in front of the Big Board as Georgia runoff results came in, I realized that as informative and entertaining as Steve is, the scoreboard-style-winner-past-the-post method of analyzing elections is still limited, the limitations reflected on how vote totals and winners are displayed.
Because designers and coders can only work with three primary colors when creating the interactive displays playing such a central role in Kornacki’s presentations, the end results over-emphasize the horse-race aspects of elections, and diminish the role election analysis could play in promoting voting and civic participation. How could something as seemingly trivial as color schemes have such a dramatic effect?
First, let’s look at how results are displayed. Most races are a contest between the two major parties, represented by primary colors Red (GOP) and Blue (Dems). The Big Board will sometimes represent a third-party candidate receiving significant votes in the remaining (RYB) primary color, Yellow. When Kornacki shows a statewide map by county- the smallest data set- the color for each county represents the party receiving the plurality of votes.
Here's the problem- voting patterns are increasingly split along urban vs. rural lines, and most here on DK know how that breaks down. There are, of course, many more rural counties in any one state than urban ones, especially in the South where counties were created to increase rural (read conservative) political strength. However, the population (and voting levels) of urban counties is many states is multiples higher than the more numerous rural counties.
As a result, Kornacki & his Big Board can be confusing to watch and listen to. Statewide totals may show a Dem candidate with a solid lead, but the Board will show a sea of Red counties with specks of Blue here-&-there to represent Dems. Because only two colors are used the display gives a distorted picture of relative voting strength. Sure, you know there are more voters in the urban counties, but you’d have to have more than a cursory knowledge of a state’s demographics to really understand to what extent votes from a particular Blue county will offset votes from the preponderance of Red counties.
To compensate, Kornacki basically does a lot of talkin’ and explain’. It’s even a bit amusing when Steve talks about a candidate having to “make up votes in Such-and-Such County” as vote totals come in. I’m thinking- people have already voted, so what can a candidate or their campaign do to “make up the votes”? I won’t offer any opinions here on Kornacki’s analysis, but in the end the presentations are more entertaining than informative.
The way to make the Big Board more in-depth and informative would be to display not just the plurality of votes, but also the voting strength of each county, i.e. the number of votes cast for each party. Here, as they say, lies the rub- designers only have three primary colors to work with. The different variables are plurality and number of votes for each party. Plurality can be represented by distinct colors, but simultaneously displaying the number of votes presents a design challenge.
Here's a possible work-around that could provide more visual info without overloading the viewer:
As now done, the plurality of one party over another is represented by the colors Red and Blue. I don’t know why the Big Board map doesn’t represent the percentage of votes, as this is easily done on web or print by adjusting the color density (%). A 50-50% vote split would be represented as white, with intensity of Red or Blue dependent on percentage of total votes won by either party. An alternative would be to simply put Red and Blue on opposite ends of a spectrum; in this scenario, the 50-50% split is a solid purple. It’s above my pay grade why Kornacki & Co. can’t do either of these, but they should consider if possible.
Number of votes would be trickier, because you’d need to combine two elements into a system of single-color tones. The parameter to be measured is the difference in votes between the top two candidates. You could create 5-6 different hues of Yellow representing the number of votes between the two candidates, such as setting a 0-10, 20K difference as pale Yellow, 150K+ in bold, etc.
When you superimpose the two-color schemes together- Red/Blue and shades of Yellow- the result is more like a “heat map”, with different ranges of color adding more visual depth than the standard Red-Blue scheme. Using the “white center” method, a populous urban county where the Dem candidate had a healthy percentage lead and large vote total difference would appear as a thick money-green. A small rural county where the GOP candidate with a strong percentage but a vote difference of only a few thousand votes would show as Red with an orange glow. Using the Red/Blue method would present a different display, the only difference is that a 50-50% split would appear brown as opposed to solid Yellow.
Neither method would be ideal and would take some adjustment on the part of both map creators and viewers, as by now we’re all used to the Red/GOP-Blue/Dem motif. For example- when using the white-center scheme, most urban counties- populous and heavily Dem-would display as some shade of green, with populous GOP counties showing shades of orange. Because vote percentages won by either party are never 100%, the Red/Blue spectrum would put a brown hue over most counties, likely not something appealing to most viewers. Another problem would be the difficulty of displaying votes of 3rd party candidates; adding more color variations could also prove not-as-useful for primary contests with three-or-more candidates.
While a new color scheme for Kornacki’s Big Board (and similar vote results maps) would have flaws and take some getting used to, it could add a level of subtlety illuminating the nuances behind election results. Discussion would shift away from focus on vote percentages (instead of vote totals), which currently obscures understanding of how votes from a handful of urban counties can offset votes from a plethora of rural ones. It could also stimulate more discussion on the urban-rural split in American society, something that will have a huge impact on politics and social dynamics as the demographic composition of the country changes.