Occasional thoughts on diverse subjects

The Red and the Blue

By Dave Menendez
Sunday, January 27, 2008, at 5:27 PM

Summary: Over the last seven years, Americans have associated red with Republicans and blue with Democrats. Why? One possibility is a formula allegedly used by TV news for assigning the colors in presidential elections—but it assigns blue to Republicans and red to Democrats in the 2008 elections.

As I understand it, the use of “Red State” and “Blue State” to refer to U.S. states voting Republican and Democratic, respectively, comes from the 2000 presidential elections. The electoral maps used by the TV networks that year all colored the states that voted for Gore blue, and those that voted for Bush red. For some reason—I’m guessing the length of time it took for the election to resolve and the perceived depth of difference between Bush and Gore supporters—this color association stuck with us.

This particular color association strikes some as odd, because red is the traditional color of left-wing parties, and blue is associated with the conservatives. (Possibly a reference to blue-bloods? Or just because it’s common to oppose red and blue?) In Canada, for example, the Liberals are red and the Tories are blue.

So why are the Democrats blue and the Republicans blue? Well, because the TV networks all used that color scheme in 2000. And why did they do that? Back in November 2004, one of Kevin Drum’s commenters suggested that the networks alternate the color of the incumbent party.

This table shows what colors the formula assigns for each election since 1976:

Year Incumbent Party Challenging Party
1976 FordRBlue CarterDRed
1980 CarterDRed ReaganRBlue
1984 ReaganRBlue MondaleDRed
1988 BushRRed DukakisDBlue
1992 BushRBlue ClintonDRed
1996 ClintonDRed DoleRBlue
2000 GoreDBlue Bush IIRRed
2004 Bush IIRRed KerryDBlue
2008 TBDRBlue TBDDRed

That makes the Democrats blue in 1988, 2000, and 2004, and red in 1976, 1980, 1984, 1992, 1996, and 2008.

This raises two questions. First, have the networks actually followed this formula? This is actually a two-part question. First, did the networks actually assign the candidates the colors given in the tables in their election coverage? I don’t have any convenient way of checking this. I vaguely recall the states Clinton won being colored red, but that’s far from enough. Drum says the table “fits all the available evidence”, but I’m not sure what he means by that.

Second, even if the networks did use these colors in those years, did they do so because of the formula? The key here is the 1988 election. Aside from that, every election from 1976 to 1996 used red for the Democrats and blue for the Republicans. It seems unlikely that the networks would all switch the parties’ colors for one election and then switch back by coincidence.

The only references I’ve found to this formula are Drum’s post and a few others linking to his post. The discussion in Wikipedia does not mention a formula, and suggests that the networks have not been consistent with each other about assigning colors, or even consistent about using red and blue.

The second question is whether the networks will continue using the formula this year (assuming that they have used the formula in the past). As the table shows, it assigns the Democrats red and the Republicans blue in 2008. Would the networks really use that association after seven years of the public associating red with Republicans and blue with Democrats? It’s so ingrained that one major conservative web site is simply named “Red State”.

Of course, if the networks do use red for Democrats and blue for Republicans this year, it will be strong evidence that the formula is real and that the networks do use it.