The image “” cannot be displayed, because it contains errors.


The nation's clearinghouse for election audit information!


Why Computerized Voting Machines Cause Long Lines and Disenfranchise Voters PDF  | Print |  Email
By William A. Edelstein, Ph.D.   
July 08, 2007

Why Paper Ballot-Optical Scan Eliminates Lines

Technical details with calculations available for download here

New York counties will soon choose computer touchscreen/pushbutton DRE (direct recording electronic) (DRE) voting machines or paper ballot-optical scanner systems (PBOS) to replace lever voting machines. How many new voting machines will be needed? The answer to this question is critical to ensure that voting will go smoothly and that costs will be within reason.
Long lines have occurred during elections using DREs in many states,1-9 causing some voters to give up and go home, effectively disenfranchising them. It is prohibitively expensive to buy a large number of DREs, which makes it likely that a substantial number of voters using DREs will end up in long lines. In contrast, PBOS uses simple marking booths whose numbers can be inexpensively increased to eliminate lines and long waits.

I believe that inevitable long voter waits—and consequent disenfranchisement of those who cannot wait—is a fundamental and disqualifying problem for DREs.


A New York City Board of Elections report10 suggested that one DRE could replace each of our present lever machines and serve 277 voters in a day. They posit that each voter using a DRE with voter verified paper trail takes 3.25 minutes. They then divide a 15 hour election day (900 minutes) by 3.25 minutes and get 277 voters.


If we accept their figure of 3.25 minutes to vote, is it really possible for 277 voters to finish in a 15 hour day? No, it is not. 277 voters at 3.25 minutes each would only work if everybody arrived at precise 3.25 minute intervals. On Election Day there are busy periods, such as early morning, lunch and dinner, when people come at a higher rate than average, and other slack periods. At all times, people come randomly. Sometimes many more than average will come, sometimes many less.


Here is what the NYC report says:

On Election Day, there are “peaks and valleys” of usage by voters depending upon the time of day, the weather, traffic and other variables outside of the control of election staff. Thus there will always be times when voters are waiting, but on the whole, there should be some insurance that waits will not be over long durations throughout the day and that on the whole, voting can be accomplished expeditiously.10
Unfortunately, these are unsupported assumptions, contradicted by experience with DREs in real elections and the mathematics of queuing theory that governs the flow of voters. I have applied this mathematical approach to simulate elections in districts with 1, 2 or 4 DREs, allotting 1 DRE to each 277 voters as suggested by the NYC report. The result is that in more than 80% of these pollsites there will be people waiting over an hour to vote. There will be many polling places in which voters are kept waiting even longer.11


This is not just math—long lines with DREs have been endemic to that technology. This occurred in the 2004 general election in Florida and Mississippi.1, 2 In Ohio long lines caused voters to give up and leave without voting.4 There were more long lines in the recent 2006 primary in Cuyahoga County, OH.12 There were long lines—along with other DRE issues—in the September 2006 problematic primary in Maryland.3 Insufficient DREs and DRE malfunctions caused more long lines and voter frustration in Tennessee, Utah, Ohio, California, Pennsylvania, Florida and other states in the November, 2006 general election (e.g. refs 5-9).


Is there a cure for this problem? Yes, buy more machines—many more machines. But we can’t afford that.


The picture for paper ballot optical scan is very different. It is easy and inexpensive to buy marking booths for PBOS. Marking booths represent the same potential bottleneck for voter flow as do the DREs. The difference is that the marking booths each cost about $150, and a DRE costs more than $8,000 in New York.


Lee, MA uses PBOS and had 3200 voters in the 2004 general election served by a single scanner.13 They had 35 marking booths for a 13-hour Election Day, i.e. 91 voters per marking booth.


The Lee town clerk Susan Scarpa said there were no lines to use the marking booths. They previously had 8 lever machines with, as described by Ms. Scarpa, “long, long lines.” So eight lever machines were replaced by 35 marking booths, and the lines disappeared.13


I spoke to the Londonderry, NH town clerk about their PBOS experience.14 In the 2004 general election they had 12,000 voters served by two scanners in 13 hours. They had 100 marking booths and no lines. So that is 120 voters per booth. The town clerk told me that during slack periods there were many empty booths, but they were 90% occupied during busy times.
These two PBOS examples really work. They are equivalent to 105 and 140 voters/per marking booth in a 15 hour day. These figures are less than 1/3 the number of voters per DRE suggested by the NYC Board of Elections, or by a more recent proposal by the New York State BOE,15 when a possible high voter turnout is taken into account. Lee, MA and Londonderry, NH do not have lines at their marking booths. Ohio, Maryland and other states have had long lines at their DREs.


Voter waits will be increased by DRE outages (roughly 10% in recent elections) or voters with special needs taking an average of 30 minutes each to use machines with disability aids, as indicated by tests conducted by the NY State Board of Elections.16


This brings us to cost. Are we prepared to buy 3 DREs for each lever machine we now own? As a taxpayer, I hope not. But that is what we would need to do to make the voting process run efficiently with DREs.


DREs and ballot marking booths represent the same bottleneck for voting. Buying more booths is effective and inexpensive. Nobody is going to be willing to buy enough DREs. Instead, they will buy an insufficient number, and long lines and voter disenfranchisement will follow.



1 D. Karp, T. Zucco, S. Nohlgren, and L. LaPeter. "Biggest voting gripe: long lines."  2004  [cited 2006 10/15/2006].
2 "Machine Problems Cause Long-Lines, Vote Dissatisfaction and Ultimately a New Election In Hinds County, Mississippi."  2004  [cited 2006 10/15/06].
3 I. Urbina. "Officials Wary of Electronic Voting" New York Times  2006  [cited 10.4.06].
4 M. Riess, T. Wang, R. Randhava, B. Burt, M. Schaffer, and S.T. Steigleder, "Voting in 2006: Have We Solved the Problems of 2004?" The Century Foundation, Accessed 10/15/06.
5.    B. Schrade, A. Paine, R. Loos, and L.A. O'Neal. "Metro wants more voting machines"  2006  [cited 2006 11/9/06]; Available from:
6.    C. Jones. "Voters head to the polls, despite some glitches"  2006  [cited 11/9/2006].
7.    "Judge orders Cuyahoga County to keep polls open late"  2006  [cited 2006 11/9/06].
8.    G. Kane. "Polling woes in S. J. [San Joaquin, CA)"  2006  [cited 11/9/06].
9.    "Rage against the machines"  2006  [cited 2006 11/9/06].
10.    "An Analysis of the Number of Voters per Voting Machine"  2006 8/21/06 [cited 2006 11/26/06].
11.    W.A. Edelstein. "New Voting Systems for New York--Long Lines and High Cost"  2006  [cited 2006 2006.12.09].
12.    "DRE Analysis for May 2006 Primary--Cuyahoga County, OH" 2006  [cited 2006 2006.08.17].
13.    R. Millman. "Do the Math"  2006  [cited 2006 8.19.06].
14.    M. Seymour, 2004 PBOS Voting in Londonderry, NH, W. Edelstein, Editor. 2006.
15.    Proposed Regulations--Minimum Number of Voting Machines, N.Y.S.Board.of. Elections, Editor. 2007.
16.    A.E. Svizzero, "Authorization Testing for HAVA Plan B Ballot Marking Devices" New York State Board of Elections.
Comment on This Article
You must login to leave comments...
Other Visitors Comments
You must login to see comments...
< Prev   Next >