Home » Analysis » What’s The Administrative Cost of Voter I.D. in Texas?

What’s The Administrative Cost of Voter I.D. in Texas?

[Post corrected to reflect that DPS didn’t get funding through a contingent rider in 2011 for improvements in drivers’ license services].

Yesterday, I got an interesting email from a reader asking a slight variation on a question I’ve looked at before. Previously, I’ve examined how expensive it is to acquire an election I.D. for an individual voter, but this reader asked, “Would you have any idea how much it will cost the State of Texas to issue the “free” ID cards and where that money is coming from?  Is DPS covering the cost and if not, which agency will cover it?”

After I took a stab at answering the question, the reader asked in passing if I had ever blogged about the issue, and I thought, “no, but why not?”

First, the question is an excellent one, and not easily answered based on the limited budget information provided by the State of Texas. The most accurate answer would be for me to say, “I don’t know” and leave it at that, but there are some clues that allow us to guess what Texas taxpayers spend to implement the provisions of the 2011 Texas picture I.D. law.

Digging through the DPS budget is a bit numbing, so let me summarize what I found, before you have to go read through my long post.

It looks like the cost of implementing the issuance of “free” election identification certificates might have been included in a 2011 contingent appropriations rider for improvements in DPS computer systems, database management, and driver license systems, and that the cost was part of a larger $64 million package. My guess is that voter i.d. didn’t account for the whole $64 million (although that would certainly be eye-opening).


What complicates things is that the contingent rider never went into effect – the associated public safety bill (S.B. 9, 2011 Tex. Leg., R.S.) never passed, and the money that DPS might have hoped for didn’t materialize, at least not in that one-time rider.


My guess is that the portion of the cost to the taxpayers attributable to the free I.D. program in Texas is somewhere in the neighborhood of $14-20 million, based on cost estimates from other states that adopted picture i.d. requirements and that are a little less shy about discussing their budgets.

(Although I’m not consistent about this myself, I think it’s more accurate to call the Texas law a “picture I.D. law” than a “voter I.D. law.” Texas already had voter I.D. law for over a hundred years, and it worked perfectly well for that entire time, it just didn’t involve a restrictive list of a few select types of picture I.D. in order to qualify to cast a ballot).


Anyway, proponents of new, punitive picture I.D. laws tend to be quite shy about discussing the price tag associated with providing free I.D.s – that’s been the pattern not just in Texas, but also in other states such as Pennsylvania, Missouri, Georgia, and elsewhere).


But this budgetary reticence is quite pronounced in Texas, and might be worse here than in other jurisdictions, in part because the Texas Legislature has proven to be very allergic to the idea of spending any state money whatsoever on social programs or services.

In 2009, the proponents of an earlier version of the strict photo-ID requirements for voting were unsuccessful in getting that law passed (H.B. 125, 81st Leg. R.S. (2009)). The reason for that difficulty was that the Lieutenant Governor (David Dewhurst) feared the legal and political consequences that would result from a passage of the law. But the specific mechanism by which the law was killed was with the addition of a fiscal note.


The Texas Legislature is famously hostile to the creation of any program, benefit, or service that costs money to implement. Therefore, a quick and easy way to derail legislation is to have the Texas Comptroller or some other state agency attach a fiscal note estimating the cost of implementation of the law.

Like the 2011 bill that ultimately passed and became law (S.B. 14, 82nd Leg., R.S. (2011)), the proposed 2009 law required the Texas Department of Public Safety to issue photo i.d at no cost upon request from voters who qualified for them. DPS (allegedly bowing to political pressure from the Office of the Governor) declared that any expenses associated with the issuance of I.D.s could be absorbed seamlessly into the existing agency funds.
The Secretary of State’s office, meanwhile, indicated that statewide voter education relating to the 2009 law would require a one-time expenditure of $2 million. Although this is a minuscule amount in comparison to the total state budget, it was more than enough of a fiscal note to torpedo the bill.
In 2011, the bill’s supporters were better organized, and applied political pressure more effectively to avoid the pitfalls that had doomed the 2009 bill – the new law was declared to be completely cost-neutral to the state. In other words, DPS and the Secretary of State both declared that the adoption of more restrictive photo I.D.s would have absolutely no impact on the state budget.
Needless to say, most people thought that the lack of a fiscal note was crazy. How could an entirely new photo I.D. program be launched and implemented statewide at zero expense? Nevertheless, the photo I.D. program officially costs nothing to the State. (And to be fair, this pathological need for public reassurance that programs which will obviously cost money must be enacted at no-cost is a not-uncommon event in Texas lawmaking, and is certainly not peculiar just to election bills. It’s part of a larger endemic structural problem with the state budget process, and with our state’s simultaneously hilarious and tragic inability to make use of deficit financing in any sort of mature or intelligent way).
No separate line-item accounting has been done for free I.D.s, and DPS has issued so few of the I.D.s that the variable costs are probably quite low. But that still leaves the fixed costs that are more-or-less unaffected by the level of demand for the free I.D.s (equipment and computer programming costs, employee training, advertising, etc.)
To get some sense of costs, I looked at the reported estimates for the fixed costs associated with implementation of changes in voter I.D. requirements in other states.
State officials estimated that the costs for implementation of picture I.D. laws were over $16 million for Missouri and around $11 million for Pennsylvania, but there was considerable uncertainty about these numbers. As with Texas, both of the other states tended to low-ball the estimated costs in response to legislative inquiries, (for instance, proponents of the law in Pennsylvania estimated a budget cost of just $1 million, paid for with spare Federal grant money for implementation of the Help America Vote Act of 2002).
There are a lot of variables here that make it hard to treat costs in other states as being directly comparable to the costs in Texas. In Texas, administrative costs associated with statewide laws are often shifted “off the books” to county budgets and paid for with county property tax money. Also, it’s fair to point out that in terms of population, Texas is significantly larger than either Missouri or Pennsylvania, and that the Texas picture I.D. law was not directly comparable to the generally more forgiving laws adopted in other states.
My guess is that $16 million is probably at least in the same order of magnitude as the costs experienced by the state (i.e. that the taxpayer’s burden of voter I.D. in Texas is somewhere in the tens of millions of dollars, with a wide margin of uncertainty due to the lack of any public audit). Some of the variables likely make the program cheaper in Texas (i.e., the fact that so few people who lack adequate I.D. have bothered to jump through the hoops to get picture I.D.s for voting in Texas), and some variables make the Texas program more expensive (the physical size of the state, the number of DPS offices, the significantly larger population, and the concomitant increase in fixed labor and equipment costs).
The DPS budget request for 2014-2015 is online (warning – the .pdf is 1,038 pages long), and the parts that apply to this question are found in Section 3.A.71 (rolled into driver license services, which is the division of DPS that issues the free election I.D.s).
As a general overview, the actual expenses of the division (before implementation of the voter I.D. law, which began in very late FY2012) for FY2011 (fiscal year 2011, from Sept. 1, 2010 to Aug. 30, 2011) had been just shy of $19 million, the estimated expenses for FY2012 were just shy of $22 million, and the budget for 2013 was $25.6 million. Despite the one-year $3.6 million bump in the budget compared to the previous biennial funding request (which had been made when the 2011 picture I.D. law was still just a twinkle in Rep. Debbie Riddle’s eye), the 2014-2015 requests were for a modest increase (a smidge less than $23 million for 2014, and $22 million for 2015).
But … this does not tell the whole story. Article IX, Section 18.09 of the 2011 general appropriations act (H.B. 1 (2011)  included a contingent rider for around $64 million (broken into $27.7 million in estimated FY2012 expenses, and $36.3 million in the FY2013 budget) to cover “improvements in the driver license system” for 2012-2013. The related rider in Article IX, Section 18.07 of the same budget included provisions for an increase of 112 FTE (full-time equivalent) staff positions in FY2012 and 361 FTE positions budgeted for FY2013 to implement improvements of the driver license system.
But DPS would only get the money if another bill (describing various improvements in DPS drivers’ license services) passed. The other bill (S.B. 9, 82nd Leg. R.S. (2011)) didn’t pass, and so the contingent rider failed. But was that rider at least possibly part of the funding for voter I.D.?
Digging further into the budget, one finds that the Drivers’ License Division also budgeted about $7 for capital improvements in FY2013, having estimated that it would spend roughly that amount in FY2012.
But there’s only so much that one can read into all this – DPS had a number of budget concerns associated with the driver license system – most of these concerns had nothing to do with the issuance of free election I.D.s. For one thing, DPS was faced with a number of complaints regarding poor customer service for such things as driving tests, motorcycle safety, online license renewal, and records management, and years of budget cuts had drastically reduced staffing in the Drivers’ License Division.
Further complicating matters is that DPS separated out some of its requests for improvements in communications and IT infrastructure (directly impacting issues like identification records and inter-agency database operations with the Voter Registration section of the Secretary of State) into line items for capital improvements to its central headquarters division. As a result, it’s likely that pointing to any one line item and declaring “that’s how much the free I.D.s cost” is risky, since the actual expenditures for that system could be subsumed under multiple funding sources relating to DPS field office improvements, computer systems, field office staff, customer service improvements, and so on.
In other words, the DPS budget probably folds the cost of free I.D. into at least two or more funding sources, and that program probably shares resources with at least two other strategic funding objectives (namely, general improvements to the driver license system, and general improvements to DPS computer systems and database management).
On March 31st of this year, the plaintiffs in Veasey v. Perry asked how much the election identification certificate program cost – as part of a discovery request ahead of the voter I.D. trial. (See item 6 on page 8 of the discovery request). As far as I can tell, the State of Texas never answered this question (presumably arguing that no record exists of the cost of the program) – if the answer is in the court records, I’d appreciate a cite.
My guess is that the fixed “start-up” cost of election identification certificates in Texas was expected to be some fraction of the $36 million appropriations rider for FY2013 for “improvements to the driver license system.” When that source became unavailable, I suspect there may have been some “robbing Peter to pay Paul” shifting of funds from capital improvements and appropriations for database and inter-agency communication to cover the expense.

Aside from its historical allergic reaction to spending money, there’s probably another reason why Texas is so coy about the start-up costs. Given the miniscule number of people who have applied for election identification certificates, the per-unit cost is huge. If you were at DPS, would you want to admit that you had spent millions in order to get laminated cards made for half a dozen people?



  1. N. E. Longoria says:

    Great article! Thank you. I can only hope the news media and the public will insist our legislators determine the actual cost of this patently unconstitutional requirement.

  2. […] Texas Election Law Blog makes a valiant effort to calculate the administrative cost of voter ID in […]

  3. […] Texas Election Law Blog makes a valiant effort to calculate the administrative cost of voter ID in […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: