DW-NOMINATE Scores With Bootstrapped Standard Errors

Royce Carroll, Jeff Lewis, James Lo, Nolan McCarty, Keith Poole, and Howard Rosenthal

Updated 17 February 2013

This updated release of the DW-NOMINATE scoresfor the 1^st to the 112^th Congresses(1789 - 2012) contains parametric bootstrapped standard errors.For an explanationof the basic theory of the parametric bootstrap see:

Political Analysis

Jeffrey B. Lewis

Keith T. Poole

Political Analysis

This research was made possible by NSF Grant 0611880 to Jeffrey B. Lewis, Keith T. Poole, and Howard Rosenthal.This work was also supported in part by the Rice Terascale Clusterfunded by NSF under Grant EIA-0216467, and a partnership between RiceUniversity, Intel, and HP. We thank the National Science Foundation,Rice University, and the San Diego Supercomputer Center for theirsupport.The files below contain DW-NOMINATE scoresfor the 1^st to the 112^th Congresses(1789 - 2012). For an explanation of how the data is organized and formattedread this text file.

As Poole and Rosenthal explain in Ideology & Congress (the 2^nd edition ofCongress: A Political-Economic History of Roll Call Voting), the first dimension can be interpreted in most periods as government intervention in the economy or liberal-conservative in themodern era. The 2^nd dimension picks up the conflict between North and South on Slavery before the Civil War and from the late 1930s through the mid-1970s, civil rights for African-Americans. After 1980 there is considerable evidence that the South realigns and the 2^nd dimension is no longer important. See our discussion ofthis period in our monograph:Income Redistributionand the Realignment of American Politics (joint withNolan McCarty, 1997, AEI Press). Further discussion can also be foundinSpatial Models of ParliamentaryVoting by Keith Poole and inPolarized America(joint with Nolan McCarty).

In the House scaling, the second dimension weight is 0.4088 and the Betaparameter (proportional to 1/ where is the standard deviation of theerror) is 7.355. The correct classification was 87.60 percent with anAPRE of 0.6308 and a geometric mean probability of 0.758. The correspondingvalues for the Senate scalingare .5513 and 9.0609, for the second dimension weight and Beta, respectively.The correct classification was 86.00 percent with an APRE of 0.5866 and ageometric mean probability of 0.742.

In order to calculate distances from the DW-NOMINATE scores you must multiply thesecond dimension by the weight parameter. To calculate the choiceprobabilities you must apply both the second dimension weight andthe Beta parameter.Use the Yea and Nay outcome coordinateswith considerable caution because, as we explain inCongress: A Political Economic History of Roll Call Voting, theyare poorly identified. However, the cutting line is identified andcan be used safely.

If you have used previous releases of the DW-NOMINATE scores pleasego to theCurrent vs. Past Releases ofDW-NOMINATE Scores page. The scores computed for the 1^st tothe 112^th (through 2010) are essentially identical to the previous release (i.e.,the corresponding scores for Congresses 1 to 111). However, when a new Congress is added to the dataset this will slightlychange the scores for more recent members because their scores areestimated using theirentire voting history. This will also slightlychange the overall means of the dimensions. Finally, the past few Congressesare nearly unidimensional with correct classifications of 90 percent or better.Consequently, the overall fit of the DW-NOMINATE estimation has increased asrecent Congresses have been added to the dataset. Also note that the2^nd dimension weight can vary somewhat from scaling to scalingbecause the 2^nd dimension coordinates can be contracted/expandedslightly. The weight will compensate for this. TheCurrent vs. Past Releases ofDW-NOMINATE Scores page shows the corrections necessary to map the pastscalings into the current scalings along with STATA files containing the currentand past releases.

Please note that the House files now contain scores for most Presidents.For Presidents prior to Eisenhower these are based on roll calls correspondingto Presidential requests. These roll calls were compiled by an NSF projectheaded by Elaine Swift (Study No. 3371, Database of CongressionalHistorical Statistics, 1789-1989). Many of these scores are based upon a small numberof roll callsso use them with caution!

If you have questions or need help with these files please send us e-mail atjblewis@ucla.edu (Jeff Lewis) orktpoole@uga.edu (Keith Poole).

To see a QuickTime Movie of the DW-NOMINATE scalings download this movie file:

Two Dimensional Movie of the 1^st to 111^th Houses and Senates (through 2009, high resolution, 458 Megabytes!!)

The format of the legislator files is:

 1.  Congress Number 2.  ICPSR ID Number:  5 digit code assigned by the ICPSR as                        corrected by Howard Rosenthal and myself. 3.State Code:  2 digit ICPSR State Code. 4.  Congressional District Number (0 if Senate) 5.  State Name 6.  Party Code:  100 = Dem., 200 = Repub. (SeePARTY3.DAT for a full set of codes of minor and historical parties) 7.  Name 8.  1st Dimension Coordinate 9.  2nd Dimension Coordinate10.  1st Dimension Bootstrapped Standard Error11.  2nd Dimension Bootstrapped Standard Error12.  Correlation Between 1st and 2nd Dimension Bootstrapped Estimates over the 1000 trials (for computing the ellipsoid of estimated points)13.  Log-Likelihood14.  Number of Votes15.  Number of Classification Errors16.  Geometric Mean ProbabilityThe format of the roll call files is: 1.  Congress Number 2.  Roll Call Number 3.  "H" if House, "S" if Senate 4.  Number of Yeas 5.  Number of Nays 6.  Month of Roll Call 7.  Day of Roll Call 8.  Year of Roll Call 9.  Number Correctly Classified10.  Predicted Yea/Actual Yea11.  Predicted Yea/Actual Nay12.  Predicted Nay/Actual Yea13.  Predicted Nay/Actual Nay14.  Proportion Correctly Classified (#9 divided by #4 + #5)15.  Proportional Reduction in Error (PRE) -- (Min. on RC - Error)/Min. on RC16.  Geometric Mean Probability17.  Spread on 1st Dimension    -- if the roll call was not scaled, there18.  Midpoint on 1st Dimension  -- are 0.000's in all four fields 19.  Spread on 2nd Dimension    --20.  Midpoint on 2nd Dimension  --