The U.S. Census Bureau released “Voting and Registration in the Election of November 2008” on July 20. The Census Bureau’s press release is here. The press release contains a link to the report itself, and the previous reports. The Census study depends on survey data, not election returns or voter registration data.
In 2008, the Census reports that 206,072,000 people could have registered and voted. However, of those, only 146,311,000 were registered. And of those who were registered, only 131,144,000 voted. So, 89.6% of the registered voters cast a ballot. But because only 71.0% were registered, the final turnout (voters divided by the number of people who could have registered and voted) was only 63.6%.
By contrast, in 2004, there were 197,005,000 people who could have registered and voted. 142,070,000 were registered. 125,736,000 voted. So, the November 2004 turnout was 63.8%, slightly higher than the 2008 turnout.
I would like to see the same data on the 1992 and 1996 elections. My figures on 1996, taken from Curtis Gans, considered to be the expert on election demographics at the time, reported that less than half of possible voters actually voted in the November general election that year. Gans stated that “One hundred million (100,000,000), over 50%, of 196,000,000 age eligible citizens did not vote in 1996.” The voter “final turnout” was about “49.9%” in 1996. I wonder how many registered voters there were for that election. The number of registerec voters must have increased a lot due to motor voter and registration drives since the 2000 election.
I will go to the report with great interes to see what the Census Bureau report has to say about the 1992 and 1996 elections when Ross Perot was on the ballot. Thanks for the source.
The census reports are a gold mine of info on voting.
I would assume that they are the primary source on the number age eligible voters at any given time in any geographical area.
The SOS of the various states would be the primary source for the number of registered voters and the actual votes cast in any electoral area.
Voter registration (millions) and voter turnout (millions) both decreased (millions) a lot from 1992 to 1996, when in 1996 there was no independent or third party candidate in the Presidential debates. Perot was blocked from buying 30 minute media programs and was not allowed to be in any of the Presidential debates as he had been in 1992. As Perot would say, the people did not vote because they were unaware of non D and R candidates and “had no dog in the hunt.”
Thanks again for a very productive link to election data.
[a] and this is the purview of the Commerce Department rather than the [evil, political, anti alternative politics] Federal Election Commission BECAUSE ……….
[b] sadder still, P2008 COULD have been such a year for non Dems and non GOP! “Saddest comment of lips or pen, —– what could have been, what could have been”
Lots of the NON-voters are, of course, barely functional illiterates — due to the rotted Pubbbblik Skooools.
What percentage of the actual votes were OVER votes ???
Lots of under votes due to having unacceptable party hack choices.
A few non-voters are not the sharpst tool in the shed. That is true.
Most are reasonably intelligent people who wisely know that a D or R vote is a non-choice, a wasted vote, a waste of time.
What is the meaning of “Demo Rep?” Are you a supporter of the bipartisan two-party monopoly? It seems to me the pot may be calling the kettle black.
# 5
Democratic Republic
Equal nominating petitions — NO party hack caucuses, primaries and conventions.
P.R. legislative bodies.
A.V. executive / judicial offices.
— pending MAJOR public education about head to head math.
The Census Bureau bases their voting reports on the Current Population Survey, which is primarily intended for measuring (un)employment.
Until 1994, the Current Population Survey did not include citizenship information, and thus the Census Bureau reported percentage voting relative to the civilian voting age population.
In their report on the 1996 election, most of the report was based on the civilian population, but they did include some information on citizenship.
Since then, the voting age population has since been reduced to a single column in a table, with the citizen voting age population broken out into details such as percentage registered, voting, etc.
It would appear that the Gans claim for 1996 is based on the voting age population, plus he reports 2.3 million more persons than the Census Bureau.
The CPS asks persons being surveyed whether they had voted at the previous month’s election. Historically, there have been about 9-11% more persons reporting have voted, than votes that are counted.
In the 2000 and 2004 reports this was more like 3 to 4%, and in 2008, there may have been slightly more votes cast than persons reporting they may have voted.
For the slight declince in voting percentage from 2004 to 2008 to be accurate, voting would have had to only have increased by 5.5 million voters, rather than the 9.4 million additional votes actually recorded.
The inclusion of citizenship information may have reduced over-reporting of voting. If an interviewee has reported not being a citizen, he may be less likely to say that he has voted; or if he does, his response might be discounted.
The survey has switch to more phone surveys, and persons may be more willing to admit that they did not vote over the phone. And there has been a general trend to giving reasons such as disinterest, as the reason for not voting, rather than reasons that may have been rationalization. So there may be an overall trend to being more honest, with more non-voters actually admitting that they hadn’t voted and that they weren’t interested in the election.