pageok
pageok
pageok
There (Probably) Was No Great Increase in Voter Turnout this Year:

Many analysts predicted that there would be a massive increase in turnout in this year's election. In actual fact, that doesn't seem to have occurred. With 97% of precincts reporting, CNN indicates that there were just under 120 million votes cast in the presidential election (63.4 million for Obama, and just over 56 million for McCain). That is pretty similar to the 121.4 million cast in 2004. Obviously, as the remaining precincts report in, the 2008 total will climb, perhaps reaching 125 million or somewhat more. That increase, however, will not be significantly greater than what one could expect based on population growth. The US population increased from an estimated 293 million in 2004 to approximately 305 million today. This 4% increase in population is almost exactly the same as the 4% increase in vote totals between the 2004 election and 2008 (assuming that this year's popular vote total ends up around 125-127 million votes or so). It's possible that the remaining precincts that have yet to report in are very densely populated, containing much more than their proportional 3% of the population. Still, I would be surprised if the final vote count rose significantly above 127 million or so (a 5-6% increase over the currently tabulated vote).

We also have not had the major anticipated increase in relative turnout by younger voters. Exit polls show that voters between the ages of 18 and 29 were 17% of the electorate in 2004, and 18% this year, a statistically insignificant difference. It is, however, worth noting that black turnout increased from 11% of the electorate in 2004 (almost exactly equal to the African-American percentage of the population) to 13% this year. Quite understandably, black voters were enthusiastic about supporting the first African-American candidate with a real chance to win the presidency.

I am not a big fan of increased turnout, which I think is an overrated objective. Increasing voter turnout tends to increase the percentage of voters who are likely to be highly ignorant, an already very serious problem in our politics. As a general rule, turning out to vote is correlated with education and interest in politics, which in turn are highly correlated with one's political knowledge, as I discuss in this article. An increase in turnout is likely to attract new voters with lower average knowledge levels than the previous electorate. In this 2006 post on turnout, I emphasized that there is no evidence showing that increased turnout improves the quality of government in any way. It worries me that so many people are concerned about increasing turnout and relatively few worry about the fact that most of the voters have little understanding of the issues and candidates they are voting on.

That said, I think the lower than expected turnout is a good sign for Obama and the Democrats. It suggests that their victory was not primarily driven by first-time and "sporadic" voters who are unlikely to return to the polls in future, less emotionally charged elections.

UPDATE: Some commenters note that other sources predict turnout to be around 130 million, rather than the 126-27 million I estimate based on the CNN numbers; it was probably a mistake to rely on the (still incomplete) CNN figures. However, the larger estimate of 130 million still represents only about a 6.3% increase in vote totals over the 122.3 million the same source indicates were cast in 2004. That is still only a slight net increase above the 4% population growth over the last 4 years.

UPDATE #2: In response to commenters who worry about the distinction between total population and those eligible to vote, it's worth noting that the total "voting eligible population" (US citizens old enough to vote minus those ineligible because they are convicted felons or for other reasons), increased from about 203.5 million in 2004 to 213 million this year. This is a roughly 4.7% increase, very similar to the 4% increase in the general population during the same time period. If anything, using the VEP figure rather than the general population figure makes this year's turnout increase seem even smaller.

UPDATE #3: Some experts are predicting final vote totals of about 134 to 136.6 million, which would be a much larger increase over the 122 million of 2004. That would be a roughly 10-11% increase in total vote, still significant even after factoring in the 4.7% increase in VEP noted above. Michael McDonald, the George Mason University scholar quoted in the above article as predicting the highest turnout number for this year estimates a 64.1% total turnout rate for 2008. However, it's worth noting that his calculations show a 60% rate for 2004. An increase from 60.1% to just over 64% is statistically notable, but hardly earthshattering.

Related Posts (on one page):

  1. Additional Evidence on Voter Turnout:
  2. There (Probably) Was No Great Increase in Voter Turnout this Year:
Cityduck (mail):
I'd like to have a one line response: Excellent post.

But, I disagree with the "masses are asses" philosophy that undergirds your opposition to efforts to get more people involved in our democratic processes.

So I'm forced to clarify that I appreciate this insight, which I haven't seen voiced elsewhere, that: the lower than expected turnout is a good sign for Obama and the Democrats because their victory was not primarily driven by first-time and "sporadic" voters.

Somewhere a political science Prof has started an article.
11.5.2008 3:33pm
Ilya Somin:
I disagree with the "masses are asses" philosophy that undergirds your opposition to efforts to get more people involved in our democratic processes.

The masses are not "asses." They are ignorant about politics for perfectly rational and understandable reasons, as I have pointed out time and again on this blog. The rationality of their ignorance, however, does not make it any less dangerous.
11.5.2008 3:37pm
Danny J. Boggs (mail):
There are websites that give "Voting-age population" and, even better, "Voting-eligible population" (excluding aliens, felons, etc). If you take the 2004 total vote and inflate, state by state, for the percentage increase in VEP, you get a projected vote of almost exaclty 130 Million. After all of the absentees, late mail-ins, provisionals, etc, we may get close to thart, but certainly not well over. That is the number to watch for.
11.5.2008 3:39pm
Nate in Alice:
I don't understand how this post gels with everyone else saying, well, exactly the opposite:

Washington Montly Predicts Highest Turn-Out Since Women won the Right to Vote

(Adding totals on CNN is, Ms. Somin, not a very competent way of going about answering this question. And you're probably going to have egg on your face when all the votes have been counted.)
11.5.2008 3:41pm
HoyaBlue:
Based on other sources I've read, this is most likely just plain wrong.

http://www.politico.com/news/stories/1108/15306.html

Summary: Turnout was probably over 130 million, and as a percentage, that might be the greatest since women were granted suffrage.

Time will tell.
11.5.2008 3:47pm
Fraggle Rock (mail):
Ilya is a man, Mr. Smug.
11.5.2008 3:49pm
Light Hearted (mail):
It would follow from you analysis, Ilya, (an analysis which I largely agree with) that off year elections, since they involve less voters and therefore a higher percentage of interested and knowledgeable voters, tend to suffer less from problems associated with rational ignorance/irrationality. Is there any empirical data to support that?
11.5.2008 3:54pm
wm13:
The Washington Monthly blog entry is citing Politico as its only authority, and Politico is citing no authority. I don't see any evidence that there were 130 million voters.
11.5.2008 4:01pm
steve lubet (mail):
Wouldn't increased voter turnout usually be lower than population increase, given that population increase is mostly due to babies and immigrants, neither of whom can vote? Stated differently, wouldnt' voter turnout normally rise at a lower rate than population?

Just wondering . . .
11.5.2008 4:02pm
The Unbeliever:
We also have not had the major anticipated increase in relative turnout by younger voters. Exit polls show that voters between the ages of 18 and 29 were 17% of the electorate in 2004, and 18% this year, a statistically insignificant difference.

Once again people, repeat after me: NO, the "surge" in youth voting will not "make the difference" this year. Or next time. Or the time after that, and will continue to fail to do so in the forseeable future. Just like it has failed to do so even though MTV and "Vote Or Die" keep pretending they can rock elections.

Nate and HoyaBlue, I keep seeing that Politico article bandied around, but actually reading the article I don't see any sources to support the claim (the Washington Monthly article is just a reference to the Politico article). Where did Barr come up with his 130 million number?
11.5.2008 4:03pm
TomH (mail):
Also, a 4% increase in population from 2004 to 2008 is not exactly relevant. That just means we have a few million more toddlers in the country rather than voters. I may be wrong, but I cannot imagine that the number of persons granted the franchise since 2004 can be too significant.

The relevant increase is in voting population or persons born before 1990 plus the new citizens since that time.

And to make things more confusing, does that increase in "U.S. population only count citizens? or does it include permanent residents and other legal visitors.
11.5.2008 4:04pm
Ilya Somin:
Wouldn't increased voter turnout usually be lower than population increase, given that population increase is mostly due to babies and immigrants, neither of whom can vote? Stated differently, wouldnt' voter turnout normally rise at a lower rate than population?

Just wondering . . .


You would expect a roughly proportionate increase in newly eligible voters, unless birth rates have been rising recently (which they haven't).
11.5.2008 4:05pm
steve lubet (mail):

You would expect a roughly proportionate increase in newly eligible voters, unless birth rates have been rising recently (which they haven't).


Except that 18-22 year olds do not tend to vote at the same rate as older people (and I believe that immigration has been rising). Thus, it still seems that voting rate would normally lag population increase -- unless, that is, the voting rate went up.
11.5.2008 4:09pm
Justin (mail):
Ilya,

I think you're neglecting three things:

1) The AP is reporting 136.6 million, which matches the 64.1% turnout rate

1) 64% is still the highest turnout in many decades. You may be setting the bar too high when pooh-poohin the numbers.



2) In the internet age, where the election is segmented and decided in fewer and fewer states, it's a lot harder to get the 90% of the country who are ignored as in non-battleground states out to the polls.

3) The 130-140 million area was what most people who projected a high turnout were predicting.
11.5.2008 4:09pm
ys:

TomH (mail):
Also, a 4% increase in population from 2004 to 2008 is not exactly relevant. That just means we have a few million more toddlers in the country rather than voters. I may be wrong, but I cannot imagine that the number of persons granted the franchise since 2004 can be too significant.

No need to use imagination. All you need to do is assume that the population distributuion by age and other relevant parameters has not significantly changed in 4 years, a reasonable assumption, then you can apply 4% to the relevant metrics.

Best of luck
11.5.2008 4:10pm
JohnK (mail):
"The masses are not "asses." They are ignorant about politics for perfectly rational and understandable reasons, as I have pointed out time and again on this blog. The rationality of their ignorance, however, does not make it any less dangerous."

I don't know about that Illya. Having spent some time in the Balkins and the middle east I can tell you people in those places are very knowledgeable and engaged in politics. Most of them would be the envy of most civics teachers. They also have a bad habbit of wanting to kill each other over politics.

I know there are disadvantags to our disinterested population. But I find the disadvantages of an engaged population to be a whole lot bigger.
11.5.2008 4:12pm
Mahan Atma (mail):
Much of the 4% increase in population was due to immigration, most of whom probably haven't attained citizenship yet.
11.5.2008 4:13pm
Anderson (mail):
Steve Benen:

But a 64% turnout rate is a very big deal. According to this Wiki entry, turnout four years ago was about 56%, and that was considered a pretty good year. More notably, turnout was 63% in the Kennedy-Nixon race in 1960, and that was the high watermark of the modern political era.

Please, enough of the Lindgrenesque weirdness. I suppose next Prof. Somin will say that there wasn't really that much turnout of *real* voters.
11.5.2008 4:17pm
E:
Perhaps there's higher turnout among first-time voters, and lower turnout among republicans?
11.5.2008 4:17pm
Bob from Ohio (mail):
No need for the update. You are actually using numbers and making an educated guess.

The "biggest turnout ever" side are using one article quoting a second article quoting an estimate.

There are 119,000,000 plus votes actually counted to date. Where 11 milion votes comes from, they don't say. They just go back to the original estimate.

The pro-record side wants to prove the point that Obama is the most exciting candidate ever!
11.5.2008 4:33pm
John M. Perkins (mail):
In Georgia, the turnout climbed 0.6% to 66.9% of the registered voters, but voter registration increased so much that "almost 567,000, or 17.3 percent more Georgians cast ballots Tuesday than did in 2004, according to data from the Secretary of State's office.


AJC
11.5.2008 4:40pm
wfjag:
Careful Ilya. You're bordering on blasphemy by calling into question the miracle of the masses. Next you'll likely question The One's ability to feed the masses -- or give them tax "refunds" and all the other promised benefits of Change and Hope -- costing $1 Trillion with only $100 Billion of increased tax revenues. That's the modern version of the loaves and fishes story. So repent and believe. "The Truth" is more important than the facts.
11.5.2008 4:46pm
ECJ:
From the NYT (doesn't do a good job of explaining where the numbers came from though):

Michael McDonald, a professor and voting expert at George Mason University, estimated the popular vote total would reach 133.3 million, after hundreds of thousands of absentee and provisional ballots that are still outstanding are finally counted, eclipsing the roughly 123 million voters who turned out in 2004. Mr. McDonald said there may be nearly a half million outstanding ballots in Georgia alone, and thousands more in other states — including the one with the largest turnout, California — that are still being counted.

Based on early figures, Mr. McDonald projected about 62.5 percent of all eligible voters cast ballots, just shy of the 62.6 percent figure that was recorded in the 1964 election. But that figure could very well rise as the remaining votes are finally counted.

"That's just my sense in how this has gone in the past," Mr. McDonald said.
11.5.2008 4:46pm
Frank M Howland (mail):
I think the increase in black turnout is the biggest story here. A 2 percentage point rise in the black vote divided by the overall vote from 11% to 13% would mean something like a 25% increase in the number of black voters when you factor in the increase in the overall number of voters. That's a lot. (Caveat: I am going with current rough numbers--I'm unsure where the 11% and 13% come from.)
11.5.2008 5:00pm
Stephen C. Carlson (www):
Increased voter turnout (of actual voters) is good in that it increases the actual margins of victory and consequently makes voter suppression and voter fraud harder to get away with.
11.5.2008 5:18pm
Suzy (mail):
Increased turnout is good because it's our precious civic duty to vote. Yes, some will be uninformed, but I don't prize the reasoning of the relatively more-informed so much that I'm concerned about opening the gates as far as possible. To me this has nothing to do with which candidate wins or won this time; it's simply Good to Vote because it's the right thing to do.
11.5.2008 5:24pm
Linc Wolverton (mail):
You stated: Exit polls show that voters between the ages of 18 and 29 were 17% of the electorate in 2004, and 18% this year, a statistically insignificant difference.

Technically speaking, the increase from 17 to 18% is statistically significant, because both samples are 100% of the population. The increase may be considered insignificant but not statistically insignificant.
11.5.2008 5:33pm
William Ayers:

It worries me that so many people are concerned about increasing turnout and relatively few worry about the fact that most of the voters have little understanding of the issues and candidates they are voting on.


From the point of view of ACORN and their allies, this is not a bug, it's a feature. Ignorant people are easier to lead.
11.5.2008 7:00pm
PubliusFL:
"That said, I think the lower than expected turnout is a good sign for Obama and the Democrats. It suggests that their victory was not primarily driven by first-time and 'sporadic' voters who are unlikely to return to the polls in future, less emotionally charged elections."

Unless, as E suggested above, Obama's victory was driven by first-time and sporadic voters but the effect of this on overall turnout was offset by poor turnout among Republican voters.
11.5.2008 7:10pm
PubliusFL:
Stephen C. Carlson: "Increased voter turnout (of actual voters) is good in that it increases the actual margins of victory and consequently makes voter suppression and voter fraud harder to get away with."

As Prof. Somin observed, as voter turnout increases the additional voters turning out are increasingly ignorant. So increased voter turnout would only significantly increase the margin of victory if the side supported by relatively more ignorant voters wins.
11.5.2008 7:18pm
PubliusFL:
Linc Wolverton: "Technically speaking, the increase from 17 to 18% is statistically significant, because both samples are 100% of the population. The increase may be considered insignificant but not statistically insignificant."

No, the claim was based on exit polls. Exit polls do not poll everyone who voted, so the sample is not 100% of the population.
11.5.2008 7:21pm
lzhou10 (mail):
I have to disagree with your argument on voter turnout. If the issue is ignorance of the masses, the solution is not to discourage their right to vote; it's to increase efforts to get people politically involved and aware. I realize that this is easier said than done.

Wanting only a certain portion of the population that you deem adequately educated to vote seems elitist and hardly democratic. There are always bound to be voters unaware of the issues, but that comes with the territory of democracy. High voter turnout may not necessarily be correlated with quality of government, but at least we are living up to our principals.

Also: Possibly Highest Turnout since 1968
11.5.2008 8:23pm
josil (mail):
Suppose only university faculties could vote and let's assume that this segment of the population is more politically engaged than the "masses" (of which they tend to speak pejoratively). Now, would the resulting govermnment be an improvement on what currently obtains? Would it be Marxist? Would the national government perform like, say, an educational institution?
11.6.2008 12:37am