A quantitative measure of conflict in the Washington State Democratic coalition

In the articles listed below I presented combined scorecards for Washington State legislators, based on five component scorecards: labor, conservation, racial justice, choice, and conservative.

In this article I examine the degree to which the various component scorecards agree and disagree. To compare two scorecards, I take the correlation coefficient of  the legislators’ scores on the respective scorecards.  Intuitively, a correlation coefficient of 1.0 means that the two scorecards track each other exactly and the two scorecards make the same ranking. A correlation coefficient of 0.0 would suggest that there is no predictable relation between a legislator’s score on one scorecard and his or her score on the other scorecard. A correlation coefficient of -1.0 would mean that the two scorecards make opposite evaluations.

Technical notes: I used the following query to calculate the correlation coefficient: http://stackoverflow.com/questions/2457645/mysql-math-is-it-possible-to-calculate-a-correlation-in-a-query. For Choice, I restricted the calculation to the 32 senators and 79 representatives for which the Choice scorecard had sufficient data to assign a score.

State Senate Results

(correlation) Conservation Labor Racial Choice Conservative
Conservation 1.0 0.93 0.92 0.89 -0.96
Labor 0.93 1.0 0.87 0.89 -0.98
Racial 0.92 0.87 1.0 0.85 -0.92
Choice 0.89 0.89 0.85 1.0 -0.93
Conservative -0.96 -0.98 -0.92 -0.93 1.0

Interpretation

The lowest correlations for the Senate Democrats is between Choice and Racial (0.85),  Racial and Labor (0.87),  and Choice and Conservation (0.89). The data suggests that the greatest conflict will be between Choice and the other coalition members. And indeed I think I see some evidence of that, since some of the centrist Democrats (including Rodney Tom and members of the Road Kill caucus) are pro-choice.

But a correlation of 0.85 is high, so the conflict would seem not to be severe.

The correlation of -0.98 between Labor and Conservative suggests that conservatives and union members have almost opposite policies.

Some senators are progressive on social issues but conservative on economic issues, and this is reflected in the data.


State House

(correlation) Conservation Labor Racial Choice Conservative
Conservation 1.0 0.93 0.92 0.97 -0.92
Labor 0.93 1.0 0.88 0.95 -0.94
Racial 0.92 0.88 1.0 0.93 -0.88
Choice 0.97 0.95 0.93 1.0 -0.93
Conservative -0.92 -0.94 -0.88 -0.93 1.0

Interpretation

The lowest correlations within the House Democratic coalition are between Racial and Labor (0.88) and Racial and Conservation (0.92).  This suggests that these coalitions would have the most pair-wise disagreement about policies.  But 0.88 is a rather high correlation, so this suggests there’s not much conflict within the Democratic coalition in the state House.

The conflicts in the coalition seem sharper in the Senate than in the House.

The difference between Democrats and Republicans seem less severe in the House than in the Senate. The correlation coefficients for Conservatives relative to the Democratic coalition scorecards in the Senate are -0.96, -0.98, -0.92, -0.93. In the House, the coefficients are -0.92, -0.94, -0.88, and -0.93:  further from -1.0.

Leave a Reply