DW-NOMINATE took a bit more work for me to understand than some other data sources, even with a bit of help from one of its creators. But I didn’t want to let commenter Tarl down, so I didn’t give up. DW-NOMINATE uses two dimensions to estimate the distance between legislators. The first one is liberalism vs conservatism, with positive numbers indicating conservative, the latter reflects regional divides with positive numbers being associated with the south/rural areas. Richard Nixon served in the 80th and 81st Congresses in the House of Representatives. His DIM1 is 0.18 and his DIM2 is -0.41, while the average for Republicans of that period is about 0.3 and -0.37, respectively. I’m rather new to using Open Office Calc/Excel so I’m not terribly confident in my work. I’ve uploaded my data file, which should contain the formula I used near the top of the sheet and to the right of the main data. If those who understand that sort of thing verify that I did it correctly, I’ll update this post to say how Nixon compared to other Republican senators that he worked alongside.
UPDATE 3: In the absence of outside verification, I used the same method for Nixon’s time in the Senate. There he had scores of 0.13 and -0.57, for Dim1 and Dim2 respectively, while his Republican peers in the Senate had about 0.26 and -0.34 on average for the 82nd Congress he served in. These scores should not be taken to indicate that he shifted in his ideology between bodies of Congress as scores in different houses are not comparable. I’m not going to upload my file for the senate because I don’t have room at my tripod site. The standard deviations for his GOP peers in the house were 0.16 and 0.32, making Nixon about 0.73 and 0.12 standard deviations below average for those dimensions. The standard deviations for the senate were about 0.19 and 0.42, making him about 0.68 and 0.55 standard deviations below average there.
UPDATE 2: My excel file (with formula and all) is now here. Its shaggy dog story is below the fold. Some months ago my hard-drive failed, taking MS Office with it. As a replacement, I had Open Office, which is very slow and annoying. When I initially did this analysis I saved it in the open document format (.ods), but then when I tested out downloading it I was informed that the file was corrupted and the “repair” removed all the data. I had the option of saving it in excel format, but that quadrupled the size and left hardly any space at my tripod site. Figuring that Open Office must not be optimized for Excel, I went to the pirate bay, failed to make any progress for a while, then copy-pasted some new trackers and in a short amount of time had MS Office back. Resaving the excel file didn’t shrink it, nor did copy-pasting from the ods file into excel. Figuring there must be formatting taking up space that was being copied over, I decided to redownload the data to open in excel and then do the analysis again. In the download window I then noticed that the original file was roughly the size of the bloated one I was trying to avoid. I guess sometimes you just have to accept things are as good as they get.