Tuesday, July 1, 2008

CAUCUS FACTOR

Original Link: http://www.lynettelong.com/my_weblog/2008/06/caucus-factor.html

On March 4, 2008, Texas held its Democratic Primary, affectionately called the Texas-Two Step. After the polls closed at 7 pm, primary voters could participate in a caucus. Sixty-five percent of the pledged delegates were awarded based on the primary results and the other 35% based on the caucus results. According to CNN, 2,867,454 votes were cast in the primary - 51% (1,458,814) for Senator Hillary Clinton and 47% (1,358,785) for Senator Barack Obama, and a smattering of votes (49,855) for John Edwards, Bill Richardson, Joe Biden, and Chris Dodd combined. In the Texas caucus 42,538 votes were reported. Obama got 56% (23,918) and Clinton got 44% (18,620) of these votes. Caucus voters were required to have voted in their precinct. Consequently, caucus voters were a statistical subset of primary voters, but they did not vote the same way.

Comparing the Texas primary to the Texas caucus has some unique challenges. During the primary other candidates besides Clinton and Obama were on the ballot. To equalize the percentages, the ballots cast for Edwards, Richardson, Biden, and Dodd were eliminated and Clinton’s and Obama’s percentages recomputed assuming the eliminated voters would split their votes in the same proportion as the rest of the electorate. The recomputed percentages are Clinton 52% and Obama 48%.

Clinton’s adjusted percentage of 52% of the primary vote is 8% points higher than her 44% in the caucus vote. Clinton moved from a four point win to a twelve point loss, a sixteen point shift. Obama gained these eight percentage points moving from 48% in the primary to 56% in the caucuses. This statistically significant sixteen point difference is “the caucus factor,” a major factor in the Democratic presidential primary. The existence of a caucus factor poses three important questions. What accounts for the dramatic difference between the two results? Was the “caucus factor” present in the other caucus states? What do these results imply regarding the validity of other caucus results?

No one doubts the accuracy of the almost three million Texas voters who selected Hillary Clinton as their choice for the Democratic nominee--- but then why wouldn’t the caucus voters, a sample of the primary voters, make the same selection. The simplest explanation is that caucus voters are a sample of primary voters but not a random sample of primary voters. Caucuses are held in the evening and take several hours. Senator Clintons core voting groups, (senior citizens, shift workers, women, and women with children) are less likely to attend caucuses. Senior citizens are less likely to go out at night, have difficulty driving in the dark, and go to bed earlier. Mothers with young children are too busy in the evenings to spend several hours at a caucus. And some people don’t go to caucuses because of the public nature of the declaration. Voting is a private event, only you and God know who you voted for. At a caucus a voter must publically declare for their candidate and resist influencing by the opponents supporters.

Another factor that contributed to the Texas caucus factor was the less regulated nature of the caucuses. Participants in the Texas caucus complained about the lack of validation of identity, undo influencing, and individuals signing several names on the caucus sheets. How much these violations occurred and how much these violations impacted the results and in which direction is impossible to determine.

Finally, Democratic insiders will say that success at a caucus depends on organization. Did the Obama Campaign just out organize the Clinton campaign which contributed to their success in the Texas Caucus?

On February 9, Washington State caucused to determine the distribution of the pledged delegates. Obama won 21,629 to 9,992 votes or 68% to 31% and received 53 of the 78 pledged delegates. Ten days later Washington State had a primary election in which no delegates were awarded yet 669,856 people chose to vote in this beauty contest. Obama won this contest by 51% (354,112 votes) to 46% (315,744 votes). How could Obama win by 37 points in a caucus and only 5 points in the primary, a 32-point difference? The caucus factor is at work again. Which result accurately reflects the will of the voters? Over twenty times the number of people participated in primary, should we trust the masses or the sanctioned election?

Further proof of the caucus factor can be shown by contrasting the difference between the results of the caucuses and the primaries collectively. Fourteen states had caucuses -- Senator Obama won thirteen of these caucuses. The probability of that occurring in hotly contested race where popular vote is a statistical tie, is a statistical impossibility if the caucuses were a true representation of the voters. In the states where primary elections were held, Senator Clinton has won 19 primaries and Senator Obama has won 17 primaries, reflecting the close nature of the Democratic Primary race.

The difference between the margin of victory during caucuses and primaries again illustrates the inaccurate nature of the caucuses. The average point spread in the 13 caucuses Obama won is 32 points. The average of the point spread in the 17 primaries Obama won, is 21 points. This eleven point difference is the caucus factor.

The results in Texas, Washington, the number of caucuses won by Obama, and difference between the average margin of victory in the caucuses and primaries, all point to the existence of the caucus factor. The variability of the results of the caucus returns also points to the unreliability of the caucus results. Senator Clinton won two primaries in demographically similar states, Ohio and Pennsylvania, by the same ten point margin. Yet when caucus results in similar states are compared, the disparity of the results is profound. Obama won Idaho by 62 points and the demographically similar Wyoming by 23 points. If the caucus results were reliable, the results of these two similar states would not be so disparate.

Voters who participated in the caucuses had more influence in the election than voters in primaries. In California, 4,677,788 votes were cast and 363 delegates were awarded. In New York, 232 delegates were pledged in a primary that had 1,748,833 votes cast. In Alaska a total of 405 persons “caucused” for 13 pledged delegates. A person who participated in the Alaska Caucus had 242 times more influence in the Democratic Primary than a voter in the New York primary and 415 times more influence than a voter in the California primary. With so few voters accounting for delegate selection in caucus states, results and delegate totals can be easily influenced and manipulated.

The caucus factor is a real statistical event which inflated Obama’s lead over Clinton and misrepresented the will of the people. If all the caucus states had primaries, Obama’s margin of victory in those caucus states would have been smaller and most likely Clinton would have won some of the caucus states. The Democratic Primary was extremely close and questions should be raised regarding the validity and reliability of the caucus results. Unfortunately, a win in a caucus state was given the same weight as a win in a state with a primary election, allowing Obama supporters to claim, “He’s won more states,” even though Clinton won more primaries. Even more disturbing is the fact that of the 14 Caucus States, eight are Dark Red states. The Democrats have let the Red States like Wyoming, Nebraska, Alaska, Idaho, Kansas, North Dakota, Colorado and Nevada select their Democratic Nominee. The will result in certain disaster in the fall.

No comments: