The Argument Against Voter ID That You Won’t See in the Media

December 12, 2011 at 10:54 pm

When it comes to voter ID programs, what’s the typical mainstream media article look like?

One side says that the programs prevent voter fraud, the other says that they are “designed to stifle turnout among students, poor people and minorities, who are more likely to vote for Democrats.”

The media reports what both sides say.  What you rarely see the media doing is any type of objective cost/benefit analysis.

Think of this as the perspective you might want to take if you were a business and you were looking at investing your money in a new voter ID program.

You’d want to see what the returns or benefits look like in comparison to the costs.

Basically, is it worth it?

What’s a simple analysis look like?

To start, you want to have reasonable estimates that everyone can agree upon for at least 3 things:

1) Initial costs
2) Current situation
3) Benefit of a new solution

You can also add in ongoing costs. For the purpose of this simple estimate, I’m going to assume ongoing costs are roughly similar between any new solution and current voter ID measures.


I will also state, for the record, that this is a quick and dirty estimate. The idea is not to come up with perfect numbers but to get a sense for whether we’re more or less likely to realize some significant improvements from a new voter ID system.

If you are doing this cost benefit analysis for or with a particular person or audience, check along the way to make sure that you can both agree that the numbers are reasonable estimates.

Another quick reminder: Focus the discussion on the analysis and keep personal politics out of it. The idea here is to:

Work together to analyze a proposed solution using numbers everyone can agree upon.

This sounds simple, but this can be the hardest part of the discussion.

Onto the analysis.

Initial costs

For initial costs, there are a number of programs already in place or under study so it was pretty easy to find some numbers.

Some of the costs include:

  1. Training for workers on the new process
  2. Documentation
  3. A communication process and implementation plan for the new process; how you are going to make it happen
  4. Public education and outreach programs to explain the changes
  5. Cost of issuing free IDs to satisfy Constitutional requirements
  6. Updates to forms and systems

The analysis for the State of Missouri was between $6-8 million per year. Wisconsin estimates $2.7 million a year to provide free IDs plus $2 million for the first 2 years to implement.

In Indiana, the BMV has issued 771,017 free photo IDs at a total cost of just over $10 million from 2007-2010. In Georgia, the costs to implement the new photo ID system was stated as $1.6 million from 2006-2008. An estimate for North Carolina is $18-25 million over the course of 3 years.

Estimates range from roughly $1 million per year to upwards of $8 million for the first couple years. After the initial couple years, costs drop significantly as changes have been implemented and the public and election workers understand the new requirements.

Current situation

What’s the current situation look like? How much voter fraud is occurring?

The New York Times reported that from 2002 to 2005, exactly 55 people were convicted of voter fraud.

The Republican National Lawyers Association (RNLA) compiled stats by state on voting convictions from 2000-2010. During the 10 year period, 21 states had only 1 or 2 convictions for voting irregularities.

So as not to be accused of bias, let’s look at the RNLA statistics for the State of Missouri. From 2000 to 2010, 17 convictions for voting irregularities were observed. This is roughly 1.7 voting irregularities per year.

Now let’s assume that we only catch 5% of those people who voted illegally so the 1.7 number is 5% of the total. We likely catch a greater percentage but, for the sake of argument, let’s assume we catch very few people who vote illegally. If we only catch 5%, then there would be roughly 34 cases of illegal voting in Missouri each year.

The total number of voters in Missouri in 2008 was 4.2 million. Let’s say only 30% of these voters vote in an average election for an average voting number of  1.26 million.

The percentage of illegal votes under the current system = 34 / 1.26 million x 100 % = .0027%.

In other words, 500 times less than 1% of the vote in the State of Missouri. This is likely less than the margin of error in the voting system itself. In other words, you’ll likely see more clerical errors than illegal votes.

Benefit of a new solution

Let’s assume the new system works well and that it improves upon the old by limiting 80% of the current illegal votes.

This would mean that the State of Missouri would reduce the number of illegal votes from 34 to 7 (6.8 rounded up) illegal votes per year.

This would reduce the percentage of illegal votes under the new system to 7 / 1.26 million x 100% = .00056% per year.

Summary of Results

Here’s the results of our simple cost/benefit analysis:


$6-8 million for first 3 years Reduced illegal voters from 34 per year to 7
Percentage of illegal votes reduced from .0027% to .00056%

For the first 3 years of the program, the cost is roughly $220,000 per illegal vote prevented.

No election I have ever heard of has been within .0027%.

If it were, the laws usually state that you have to have a recount or some type of runoff election.

When you look at the numbers, I don’t see the value of changing the voter ID laws. Even if we use conservative numbers and assume that we hardly catch any illegal voters or that the costs for implementation are lower.

The current system works well, no elections are being influenced by illegal voting, and the changes come with a significant cost.

Post Mortem

The real question is, why don’t we see more of this type of analysis in the media?

Its not hard to do, it would help people make better decisions, and full-time reporters, with the ability to do better research, could likely find even more accurate numbers than those I’ve used and validate the analysis further.

I believe that this is because the media, by and large, has adopted a certain view of the news. This view is that “fair and balanced” equals presenting both sides of the story equally without any analysis. Equal media time and coverage as opposed to fair analysis.

This means that if one side said the world was round and the other said the world was flat, we would see equal coverage of the situation. The only thing that you can’t do under these rules, is give less time to one side.

In this case, articles end up looking like this one in the Atlantic or this one in USA Today.

Most report simply quote what each side says.

Any type of analysis runs the risk of disagreeing with one side or another and therefore would be unfair.

This view of the media says that to be fair, we can only print what each side says.

A working media would question the conclusions and help the public understand the methods used and how results were verified. It would question the claims made by both sides. It would perform analysis and ensure that the analysis itself was fair.

If the methods were deemed proper and reviewed appropriately, then conclusions would be trustworthy and non-partisan.

Today’s media disagrees. Fair is printing what both sides say without any analysis of the claims. Any critical analysis which favors one conclusion over another is considered bias.

Is it any wonder that so many people are confused when the media, which should be responsible for vetting claims, merely transcribes what each side says?

Cross posted at: Daily Kos.