Published September 22, 2003, in The National Review Online

Voting Data
Sometimes mistakes aren't mistakes.

By John R. Lott Jr.

The Ninth Circuit Court of Appeals is surely one of the most-liberal and most-reversed courts in the U.S., but that only explains part of the court's recent decision to delay California's recall election. Both the California attorney general and the secretary of state played an important role by leaving completely uncontested the ACLU's claims on how punchcard machines will result in votes being uncounted. This still should not have been enough and today's eleven-judge panel from the same Ninth Circuit seems likely to fix the mistake. Yet, these state officials' concessions and weak defense of the recall gave the liberal judges just the excuse they were looking for.

Like many lawyers who concentrate on arguing the law and avoiding numbers, the California Democratic attorney general's office made vague general, though plausible, claims that delaying the election will harm the state. And such a delay will surely produce uncertainty and continued confusion. This left the judges to choose between a claimed unconstitutional harm in treating different voters differently based upon which county that they live in and their guaranteed right under the California constitution to quickly recall a politician before more harm is done. The lawyers supporting the recall could have and should have challenged that tradeoff.

There are a lot of reasons for voters not to vote in all races in an election. During the 2000 presidential race in California, not trying to account for any other factors, the average county using punchcard machines had 1.36 percent of voters using punchcard machines did not have a recorded a vote for president. The average county without punchcards did have a lower rate of 1.04 percent.

Assuming that nothing else mattered, simply replacing the punchcard machines for the six counties that are still planning on using punchcard machines for the recall election implies about 15,000 fewer non-voted ballots. A little over one-tenth of one percent of the votes in the 2000 presidential election would have been affected.

Yet, other factors, such as the age of the population, income, education, and demographics, do matter in explaining non-voted ballots. What voting machine is used explains little of whether a voter is shown as voting for a particular candidate. During the three presidential elections in 1992, 1996, and 2000, the type of voting machine used explains a mere 2.5 percent of the difference in this non-voted ballot rate across counties. African Americans didn't have a statistically higher rate of non-voted ballots for president in counties with punchcards or with optical readers or electronic touchscreen voting machines.

But the most problematic punchcard machines are the DataVote ones and, fortunately, none of the six counties that are planning to use punchcards will use that type of machine.

Using this additional information, how many extra blank ballots might there erroneously be due to the punchcard machine still in existence? Out of every 1,000 voters there will be fewer than one additional non-voted ballot. Even if as many people vote as voted during the 2000 presidential election, there will be about 10,500 more non-voted ballots. In other words, fewer than one-tenth of one percent of the votes will be affected.

Punchcards (and particularly DataVote machines) have potential problems with the infamous hanging chad, and touchscreen systems also have the benefit of preventing voters from accidentally voting for more than one candidate in a race. But touch-screen systems have their own risks. The machines can be prone to tampering and vote rigging and the very lack of a paper trail means that officials might never be able to discover if votes had been correctly recorded. Optical vote readers can also misread votes.

Not all decisions to abstain from voting for a candidate are a mistake. Nor are all punchcard machines the same. The "bias" introduced by different voting machines is not as great as the ACLU makes us believe. But all this confusion could have been avoided if the California government attorneys had done their job.

óJohn Lott, a resident scholar at the American Enterprise Institute, is the author of The Bias Against Guns .

Home (description of book, downloadable data sets, and discussions of previous controversies)

Academic papers:

Social Science Research Network

Book Reviews:

For a list of book reviews on The Bias Against Guns, click here.

List of my Op-eds

Posts by topic

Appalachian law school attack

Baghdad murder rate

Arming Pilots

Fraudulent website pretending to be run by me

Ayres and Donohue

Stanford Law Review

Mother Jones article

Vin Suprynowicz quote


Craig Newmark

Eric Rasmusen

William Sjostrom

Dr. T's

Interview with National Review Online

Lyonette Louis-Jacques's page on Firearms Regulation Worldwide

The End of Myth: An Interview with Dr. John Lott

Cold Comfort, Economist John Lott discusses the benefits of guns--and the hazards of pointing them out.

An interview with John R. Lott, Jr. author of More Guns, Less Crime: Understanding Crime and Gun Control Laws

Some data not found at

Updated Media Analysis of Appalachian Law School Attack

Since the first news search was done additional news stories have been added to Nexis:

There are thus now 218 unique stories, and a total of 294 stories counting duplicates (the stories in yellow were duplicates): Excel file for general overview and specific stories. Explicit mentions of defensive gun use increase from 2 to 3 now.

Journal of Legal Studies paper on spoiled ballots during the 2000 Presidential Election

Data set from USA Today, STATA 7.0 data set

"Do" File for some of the basic regressions from the paper