In a post election blog I explained how a 20% increase in voter turnout blurred the final outcome and made a closer election. That was true. I blamed the increased voter turnout on 527 spending. That was probably wrong. So you get a semi-recant.
The final data on the election won’t be out for another month. The first reliable total of the San Francisco vote, including absentees and provisionals, showed a 20% increase in the total turnout over the 2000 presidential race. That is the same increase in turnout as in the national race. Since not one penny was spent in San Francisco, nor one finger raised to get out the vote, that suggests voter turnout efforts and the money spent by 527s was irrelevant. San Francisco/no effort vs. nationwide/big effort and money = same results.
Now we old codgers get to put in a word. Something few living people know about.
First you should know that the turnout for an election is about 30% of registered voters for a special election, 40% for a primary, 50% for a general election and 60% for a presidential election. Right off the bat that should tell you that voter interest is the only variable that affects voter turnout.
In 1963 I started a group in San Francisco called Research in Politics. It lasted five years and did a great deal of research. I started it after working a few years as an advance man, legislative aide and campaign manager.
I wanted to know how effective get-out-the–vote efforts were. I enlisted both political parties in a cooperative research endeavor. We focused on two special elections on the premise that special elections have the lowest turnout and would be most susceptible to door-to-door voter turnout efforts.
I designed a 25 precinct Latin Square test for a San Francisco Congressional special election in February 1964 for Democrat Phil Burton and the same design for a special election in March in Modesto for a Republican.
The 25 precincts varied from five with absolutely no voter turnout effort by either party to five precincts with full effort at turnout by both parties. There were two sets of five precincts with heavy effort by only one party and the last five with light effort by both parties.
Result: even with intense efforts (I personally worked hard knocking on doors in five precincts with four friends), absolutely no effect on voter turnout. None, nada. The data was overwhelming and statistically significant.
I reported the results to campaign managers all over the U.S. in a variety of training sessions. Everyone continued and still continues to do voter turnout for one reason: it's something to keep volunteers busy and make them feel useful.
I personally shouldn't have been surprised that the 527 voter turnout effort was probably insignificant.