For real election obsessives, I've posted an Excel spreadsheet with complete primary results so far:
Not surprisingly, it is a very large file and may take a while to download. I will be posting maps from the data after the final primaries, in Montana and South Dakota on June 3. Disclaimers below:
NOTE: The purpose of this spreadsheet is to show geographical patterns in support for specific presidential candidates. For that reason, all popular vote data is included, and primary results are included for states that hold both primaries and caucuses -- even when those primaries are not used to award national convention delegates. This inclusion should not be interpreted as an argument for using these primaries as the means for determining a party's presidential nominee. On the Democratic side this chart includes results from the Florida and Michigan primaries, but those early primaries were not sanctioned by the national party (and Barack Obama was not on the ballot in Michigan) and thus did not count toward the awarding of delegates. This chart includes only primary results from Texas, but delegates from that state were awarded in accordance with the results of both the primary (which Hillary Clinton won) and caucuses held on the same day (which Obama won). This chart includes results from nonbinding primaries in Nebraska and Washington, but delegates from those states were awarded in accordance with the results of earlier caucuses. (In both states, Obama won both the caucuses and primaries, but won the latter by smaller percentage-point margins.) Finally, popular votes from the Iowa and Maine Democratic caucuses are estimates based on each county's awarding of delegates to statewide Democratic conventions.
SECOND NOTE: You may have noticed that I flagged “Marengo County, Alabama” by putting “NA” in the column that compares votes cast in the Democratic and Republican primaries. That’s because I am certain that the 6,175 votes recorded for Mike Huckabee is impossibly high, given the overwhelming Democratic bent of the county. But the state of Alabama has not corrected this and probably never will. There are always errors like this (someone probably put an extra digit in a handwritten sheet given to the state) that aren’t corrected unless they would significantly change the outcome of the race, and every state has a different level of carefulness in these matters. I will try to catch anything unusual like this, but it's inevitable that there will be (hopefully very slight) errors at the county level.