Social:Election audits

Election auditing refers to any review conducted after polls close for the purpose of determining whether the votes were counted accurately (a results audit) or whether proper procedures were followed (a process audit), or both. Both results and process audits can be performed between elections for purposes of quality management, but if results audits are to be used to protect the official election results from undetected fraud and error, they must be completed before election results are declared final.

Recounts may be considered to be a specific type of audit, but not all audits are recounts. The Verified Voting Foundation explains the difference between audits and recounts: Post-election audits are performed to “routinely check voting system performance…not to challenge to the results, regardless of how close margins of victory appear to be", while "recounts repeat ballot counting (and are performed only) in special circumstances, such as when preliminary results show a close margin of victory. Post-election audits that detect errors can lead to a full recount.” In the US, recount laws vary by state, but typically require recounting 100% of the votes, while audits may use samples. Recounts incorporate elements of both results and process audits.

The need for verification of election results

In jurisdictions that tabulate election results exclusively with manual counts from paper ballots, or 'hand counts', officials rarely rely on a single person to view and count the votes. Instead, valid hand-counting methods incorporate redundancy, so that more than one person views and interprets each vote and more than one person confirms the accuracy of each tabulation. In this way, the manual count incorporates a confirmation step, and a separate audit may not be considered necessary.

However, when votes are read and tabulated electronically, confirmation of the results' accuracy must become a separate process.

Within and outside elections, use of computers for decision support comes with certain IT risks. Election-Day electronic miscounts can be caused by unintentional human error, such as incorrectly setting up the computers to read the unique ballot in each election and undetected malfunction, such as overheating or loss of calibration. Malicious intervention can be accomplished by either corrupt insiders or external hackers who accessed the software before Election Day.

Computer-related risks specific to elections include local officials’ inability to draw upon the level of IT expertise available to managers of commercial decision-support computer systems and the intermittent nature of elections, which requires reliance on a large temporary workforce to manage and operate the computers.

To reduce the risk of flawed Election-Day output, election managers like other computer-dependent managers rely on testing and ongoing IT security. In the field of elections management, these measures take the form of federal certification of the electronic elections system designs; security measures in the local election officials’ workplaces; and pre-election testing.

A third risk-reduction measure is performed after the computer has produced its output: Routinely checking the computers' output for accuracy, or auditing. Outside elections,auditing practices in the private sector and in other government applications are routine and well developed. In the practice of elections administration, however, the Pew Charitable Trusts stated in 2016, “Although postelection audits are recognized as a best practice to ensure that voting equipment is functioning properly, that proper procedures are being followed, and that the overall election system is reliable, the practice of auditing is still in its relative infancy. Therefore, a consensus has not arisen about what constitutes the necessary elements of an auditing program.”

Routine results audits also support voter confidence by improving election officials' ability to respond effectively to allegations of fraud or error.

Audit challenges unique to election results

Confirming that votes were credited to the correct candidates’ totals might seem to be a relatively uncomplicated task, but election managers face several audit challenges not present for managers of other decision-support IT applications. Primarily, ballot privacy prevents election officials from associating individual voters with individual ballots. This makes it impossible for election officials to use some standard audit practices such as those banks use to confirm that ATMs credited deposits to the correct account.

Another challenge is the need for a prompt and irrevocable decision. Election results need to be confirmed promptly, before officials are sworn into office. In many commercial uses of information technology, managers can reverse computer errors even when detected long after the event. However, once elected officials are sworn into office, they begin to make decisions such as voting on legislation or signing contracts on behalf of the government. Even if the official were to be removed because a computer error was discovered to have put that official in office, it would not be possible to reverse all the consequences of the error.

The intermittent nature of elections is another challenge. In most jurisdictions, elections take place no more than four times a year. This prevents development of a full-time, practiced workforce, for either the elections or the audits. Turning election audits over to an independent, disinterested professional accounting firm is another option not available to election officials. Because election results affect everyone, including the election officials themselves, truly disinterested auditors do not exist. Therefore, audit transparency is required to provide credibility.

Attributes of a good election results audit

No governing body or professional association has yet adopted a definitive set of best practices for election audits. However, in 2007 a group of election-integrity organizations, including the Verified Voting Foundation, Common Cause, and the Brennan Center for Justice at NYU School of Law collaborated with the American Statistical Association to produce a set of recommended best practices for post-election results audits:

• Transparency: The public must be allowed to observe, verify, and point out procedural mistakes in all phases of the audit. This requires that audit procedures and standards be adopted, in written form, and made available before the election.
• Independence: While the actual work of post-election audits may be best performed by the officials who conduct the elections, the authority and regulation of post-election audits should be independent of officials who conduct the elections.
• Paper Records: Vote-counting in the audit should be performed with hand-to-eye counts of voter-marked, voter-verified paper ballots.
• Ballot Accounting (chain of custody, or internal control): The records used in the audit must be verified to be true and complete records of the election.
• Confirmation of the correct winners: The audits must reach statistical confidence that the computer-tabulated results identified the correct winners.
• Addressing Discrepancies: When discrepancies are found, investigation is conducted to determine cause of the discrepancies.
• Comprehensive: The ballot-sample selection process includes all jurisdictions and all ballot types (e.g., absentee, mail-in and accepted provisional ballots).
• Additional Targeted Samples: The audit includes a limited non-random sample selected on the basis of factors useful for building voter confidence or improving election management, such as Election-Day problems or preliminary results that deviate significantly from historical voting patterns.
• Binding on Official Results: Post-election audits must be completed before election results are declared official and final, and must either verify or correct the outcome.

Current practices in election results auditing in the United States

Computerization of elections occurred rapidly in the United States following the presidential election of 2000, in which imprecise vote-counting practices played a controversial role, and the subsequent adoption of the Help America Vote Act (HAVA) of 2002. The rapid switch to computerized vote tabulation forced election officials to abandon many pre-automation practices that had been used to verify vote totals, such as the redundancy included in valid hand-counting procedures.

According to information from state profiles provided by Verified Voting, as of early 2018, only Arizona, Colorado, Minnesota, New Mexico, the District of Columbia, and (in legislation that will not be fully implemented until 2020) Rhode Island require local election officials to perform audits that:

1. are completed promptly, before official results can be certified;
2. expand the audit to a full recount whenever the audit detects serious discrepancies in the original sample; and
3. are binding on the official results.

If followed by local election officials, such requirements create an opportunity to detect and correct any outcome-altering miscounts (whether caused by accident or fraud) that affected the preliminary Election-Night counts.

An additional 19 states prescribe audits that check a flat percentage, typically between 1 and 3 percent, of voting machines or precincts. These audits can detect problems in the individual voting machines selected for audit, but cannot confirm the correct results except in races with very large margins between the winner and losers.

Finally, 26 states finalize official election results without verifying computer-tabulated vote totals. These states either have no audit requirements, or in four states (Florida, Vermont, Virginia, and Wisconsin) allow audits to be delayed until after winners are certified.

Other ways to group the states include that 20 states audit by means of hand counts; others use machines or a mix. Ten states audit all races on the ballot; others generally audit the top race and 1-4 others. In 14 states the audit results are used to revise official winners when there is a discrepancy. Another 10 may revise official results, depending on local judgment, and in 9 states the audit creates a report without changing official results, including large states like California, Florida and Illinois. Only two states use hand counts to audit all races, and use the results to revise winners; both are small, Alaska and West Virginia.

Election audit practices, by state

Sortable list of US states' election audits, initially sorted by population

State	Method of Auditing Ballots	Do Audits Revise Election Outcomes?	Number of Races Audited	Level of Discrepancy to Trigger Action	Action if Discrepancy	State Gets Report?	Which Units Are Sampled?	Sample Size	Can Public See Ballot Marks?	What's Done if Some Ballots Are Missing?	Law	Rules	2017 Population, Millions
California	Hand count of precinct ballots	No	All	No rules	Report	Yes	precincts or machines	1%	No		Cal. Elec. Code §336.5, §15360 (West 2015)		40
Texas	None on paperless machines, which most counties use, hand count otherwise	Maybe	6-Sec. of State chooses 3 races + 3 ballot items	Any	Determine cause	Yes	precincts	at least 1% or 3, not paperless machines	No		Tex. Election Code Ann. §127.201 (Vernon 2015)	Election Advisory No. 2012-03	28
Florida	Count with machines, independent of election, or by hand	No	all by machine, or 1 by hand, randomly chosen by each county	No rules	Report	Yes	precincts	20%-100% by machine or 2% by hand	No		Fla. Stat. Ann. §101.591		21
New York	Count with machines different from voting system or by hand	Yes	All	Any	Expand sample in stages to full recount. Unclear if just the affected office	Yes	machines	3%	No		N.Y. Election Law § 9-211 (McKinney 2015)	9 N.Y. Comp. Rules & Regs. 6210.18	20
Pennsylvania	Count with machines different from voting system or by hand: the ballot images created by the election machines	Maybe	All	No rules	No rules	No	ballot images from paperless machines	lesser of 2% or 2,000 ballot images	No		Pa. Cons. Stat. tit. 25 §3031.17		13
Illinois	Count with machines the optical ballots, which are common, others by hand or different machine from election	No	All	Any	Report	Yes	precincts + machines	5%	No		Il. Rev. Stat. ch. 10 §5/24A-15, 10 §5/24C-15		13
Ohio	Hand count	Yes	3-President, random state-wide, random county-wide, in general elections in even-numbered years	0.5% (0.2% if margin under 1%)	Expand sample. Sec. of State may order full recount of county	Yes	machines, precincts or polling places	at least 5% of votes	No		Secretary of State Directive 2014-36,[3] 2015 Election Official Manual	settlement agreement in League of Women Voters, et al. v. Brunner, Case No. 3:05-CV-7309, US N.District of Ohio	12
Georgia	None		0										10
North Carolina	Hand count	Yes	1-President or state-wide ballot item	"Significant"	Expand sample to whole county, all races	No	precincts	risk-limiting	No	No audit	N.C. Gen. Stat. Ann. §163-182.1		10
Michigan	None		0										10
New Jersey	None		0								N.J. Stat. Ann. §19:61-9		9
Virginia	Hand count	No	Some	Any	Analyze	Yes	localities	risk-limiting	No		Code of VA 24.2-671.1		8
Washington	If parties agree or county auditor requires, then hand count sample of mail ballots, which are the vast majority	Maybe	1	Any	investigate + resolve	Yes	precincts or batches of mailed ballots	up to 3 precincts or 6 batches	No		Wash. Rev. Code Ann. §29A.60.185, §29A.60.170		7
Arizona	Hand count	Yes	5-President, random federal, random state-wide, random legislative, ballot measure	Statistical committee chooses cutoffs	Expand sample, up to whole county, for affected office	Yes	precincts	at least 2% or 2	No		Ariz. Rev. Stat. §16-602		7
Massachusetts	Hand count	Maybe	6-President, state and national Senator and Representative, random ballot question	Cast doubt on outcome	Sec. of State may expand audit	Yes	precincts	3%	Yes		Mass. Gen. Law Ann. ch. 54 § 109A		7
Tennessee	Count with same method as election day, different optical scanner	No	1-President or Governor in general	1%	Expand sample to 3% of precincts. Then no action, but audit is evidence in court	No	precincts	1 to 5	No		Tenn. Code Ann. § 2-20-103		7
Indiana	If requested by party chair, hand count or machine different from voting system	Yes	All	No rules	Correct errors	No	precincts	up to 5% or 5	No		Indiana Code §3-12-5-14, §3-11-13-39		7
Missouri	Hand count	Maybe	5-state-wide candidate + ballot issue, legislator, judicial, county	0.50%	Investigate and resolve	Yes	precincts	5%	No		15 Mo. Code of State Regs. §30-10.110		6
Maryland	Independent machines re-count ballot images from election machines	Yes	All	Any	Investigate and resolve	Yes	All	100%	No			Code of Md. Regs. §33.08.05	6
Wisconsin	Hand count	No	4-President or Governor, and 3 random state contests, only in general	Any	If no explanation, get manufacturer to investigate	Yes	wards or other districts reporting results	100 per state	No		Wis. Stat. Ann. §7.08(6)	Wisconsin Elections Commission Voting Equipment Audits	6
Colorado	Hand count	Maybe	In each county: 1 contest in general election, 1 per party in primary, non-randomly chosen by Sec. of State	Statistical cutoffs	Investigate + decide	Yes	ballots	risk-limiting. 5% of machines, in sampled machines, 20% if machine has 1-500 ballots, otherwise lesser of 5% or 500	No		Colo. Rev. Stat. §1-7-515	Colo. Sec. of State Election Rule 25	6
Minnesota	Hand count	Yes	2-3-Governor + federal	0.5% or 2 votes	Expand samples for Governor + federal up to whole county, and if counties with 10% of state's ballots discover problems in a race, hand-count affected race statewide	Yes	precincts	3% or at least 2-4	No		Minn. Stat. Ann. §206.89		6
South Carolina	None		0										5
Alabama	None		0										5
Louisiana	None		0										5
Kentucky	None on paperless machines, which most counties use, hand count otherwise	Yes	All	No rules	Correct errors	Yes	precincts	3% to 5%	No		Ky. Rev. Stat. §117.383		4
Oregon	Hand count	Yes	2-3-heaviest vote in each county, statewide office + ballot measure	0.50%	Expand sample to whole county for that vote tally system, all races	Yes	precincts with 150 or more ballots	3% to 10% depending on margin of victory	No		Or. Rev. Stat. §254.529		4
Oklahoma	None		0										4
Connecticut	Count with machines or by hand	Maybe	3-President or Governor and 2 other races; 20% of races in primaries and municipal ballots	0.50%	Investigate and if needed recount county with mix of hand and machine counts	Yes	voting districts	5%	No		Conn. Gen. Stat. §9-320f		4
Iowa	Hand count	No	1-President or Governor	No rules	Report	Yes	precincts	?	No		HB 516, enacted in 2017		3
Utah	Hand count	No	All	Any	Record reasons	Yes	machines	1%	No		Election Policy Directive from the Office of the Lieutenant Governor		3
Arkansas	None		0										3
Nevada	Count with machines or by hand	Maybe	All	No rules	No rules	Yes	machines	2% or 3%	No		Nev. Admin. Code 293.255, Nev. Rev. Stat. §293.247		3
Mississippi	None		0										3
Kansas	None		0										3
New Mexico	Hand count	Yes	3-4: federal, governor + closest statewide race	error over 90% of the margin of victory	Expand sample, up to whole state for affected office	Yes	precincts	none in race with over 15% margin of victory, otherwise 4-165 in state	No		N.M. Stat. Ann. §1-14-13.2 et seq.		2
Nebraska	If requested by Sec of State, method unclear	Maybe	3	Unclear	Report	Yes	precincts	2%	Unclear				2
West Virginia	Hand count	Yes	All	1% or changed outcome	Expand sample to whole county, all races	No	precincts	3%	No		W. Va. Code, §3-4A-28	2016 Best Practices Guide for Canvass and Recount	2
Idaho	None		0								Idaho Code §34-2313		2
Hawaii	Hand count	Maybe	3-state-wide, county-wide, district-wide	Any	Expand audit until staff are satisfied with results	Yes	precincts	10%	No		Hawaii Rev. Stat. §16-42, Haw. Admin. Rules § 3-172-102		1
New Hampshire	None		0										1
Maine	None		0										1
Rhode Island	Hand count	Yes	Some	Statistical cutoffs	Expand sample to full recount	Rules not yet published	Not yet published	risk-limiting	Not yet published		R.I. Stat. Ann. §17-19-37.4		1
Montana	Hand count	Yes	4-random federal, state-wide, legislative, ballot measure	0.5% or 5 votes	Expand sample but not full recount except in small counties with up to 6 precincts. Check machines with errors	Yes	precincts	at least 5% or 1	No		Mont. Code Ann. §13-17-503		1
Delaware	None		0										1
South Dakota	None		0										1
North Dakota	None		0										1
Alaska	Hand count	Yes	All	1%	Expand sample to whole district (1/40th of state), all races	Yes	large precincts	1 per House district, with at least 5% of district ballots, so about 600 registered voters	No		Alaska Stat. §15.15.430		1
District of Columbia	Hand count	Yes	4-random District-wide, 2-ward-wide, 1 any type	0.25% or 20% of margin of victory	Expand sample, up to whole county, for affected office	Yes	precincts + ballots	5%	Yes		D.C. Code Ann. §1-1001.09a		1
Vermont	Each town counts with same method as election day, optical scan or hand count. Sec.of State randomly picks 6 towns to scan & machine-count all contests	No	Unknown by county. All by Sec.of State	Any	Investigate	Yes	polling places	?	No		17 Vt. Stat. Ann. §2493, §2581 - §2588		1
Wyoming	None		0										1

Colorado elections in 2017, sample sizes needed for risk-limiting audits

Risk-limiting audits

Risk-limiting audits are required in Colorado, North Carolina, Rhode Island and Virginia. These choose random representative samples big enough to be sure results are accurate, up to an acceptable level of risk, such as 9%. For each audited race, if the original computer count identified the wrong winner, there is a 9% chance in Colorado that the audit will miss it, and the wrong winner will take office. A lower risk limit would let fewer errors through, but would require larger sample sizes. Close races also require larger sample sizes. Colorado audits only a few races, and audits none of the closest races, which need the biggest samples. Auditing all races would require several counties to hand-count thousands of ballots each.

If the samples do not confirm the initial results, more rounds of sampling may be done, but if it appears the initial results are wrong, risk-limiting audits require a 100% hand count to change the result, even if that involves hand-counting hundreds of thousands of ballots.

Colorado notes that they have to be extremely careful to keep ballots in order, or they have to number them, to be sure of comparing the hand counts with the machine records of those exact ballots. Also they have to re-tally all the machine records, independently of the election software, to be sure the machine records they audit generate all the same winners reported by the election software. Colorado says it has a reliable system to re-tally the records, but it is not yet publicly documented.

In 2010, the American Statistical Association endorsed risk-limiting audits, to verify election outcomes. With use of statistical sampling to eliminate the need to count all the ballots, this method enables efficient, valid confirmation of the outcome (the winning candidates). In 2011, the federal Election Assistance Commission initiated grants for pilot projects to test and demonstrate the method in actual elections. In 2014, the Presidential Commission on Election Administration recommended the method for use in all jurisdictions following all elections, to reduce the risk of having election outcomes determined by undetected computer error or fraud. In 2017, Colorado became the first state to implement risk-limiting audits statewide as a routine practice during the post-election process of certifying election results.

Ballot scans for 100% audits

Humboldt County, CA, Clear Ballot, and TrueBallot scan all ballots with a commercial scanner, so extensive audits can be done on the scans without damaging paper ballots, without hand-counting, by multiple groups, independently of the election software. Clear Ballot is certified by the US Election Assistance Commission for voting, and they also have an auditing system, ClearAudit. TrueBallot does not currently serve government elections, just private groups. Both use proprietary software, so if it were hacked at the vendor or locally to create false images of the ballots or false counts, then local officials and the public could not check it.

The most open audit system is the Elections Transparency Project, as used in Humboldt County, California, a medium sized county with 58,000 ballots in the 2016 general election. They use a commercial scanner to scan all ballots and put them in a digitally signed file, so true copies of the file can be reliably identified. Ballots with identifying marks are first hand-copied by election staff, to anonymize them, since they are valid under California law. The scanner prints a number on each ballot before scanning it, so the scan can be checked against the same physical ballot later if needed.

The project has written open source software to read the files, and they check all races on all ballots to see if official counts are correct. The first time they scanned and checked, in 2008, they found 200 missing ballots, showing the value of complete checks. Other jurisdictions can similarly scan ballots and use Humboldt's software or their own to audit all races. They need to ensure their in-house counting software is secure and independent of any bugs or hacks in the election software, just as users of risk-limiting audits do.

Reading these scans, with software independent of the election system, is the only practical way to audit a large number of close races, without large hand counts. If scanned on election night, scans are also the only practical way to bypass problems of physical security, since scans are made and digitally signed before ballots are stored, while other audit methods are too slow to do election night. Scanners can process thousands of ballots per hour, and several scanners can operate simultaneously in larger jurisdictions. Once ballots are scanned with a digital signature, they can be analyzed at any convenient time.

Humboldt has gone beyond this internal approach, and decided to release the digitally signed files of ballot images to the public, so others can use their own software, independent of the election system and officials, to check all races on all ballots. Public release lets losing candidates who mistrust the election officials' security measures do their own checks. Humboldt County believes the scans give their citizens high confidence in the county election results.

The "Brakey method" is a variant of the Humboldt scans, with a public release of electronic ballot images already created by most current US election systems.. The Brakey Method requires that the ballot images be connected to their physical ballots through a unique identifier that appears on both digital image and physical ballot. Because ballot images can be hacked if the election software itself is hacked, either at the vendor or elsewhere, ballot images do not provide an independent check unless they can be compared with physical ballots.

Pennsylvania uses these ballot images for its audits, though it does not release them to the public for the public to audit.

Samples used for election audits

Other variations

Hand and machine counts. Current audits in most states involve counting paper ballots by hand, but some states re-use the same machines used in the election. Three states use different machines, to provide some independent check.

Sample sizes. When states audit, they usually pick a random sample of 1% to 10% of precincts to recount by hand or by machine. If a precinct has more than one machine, and they keep ballots and totals separate by machine, they can draw a sample of machines rather than precincts and count all ballots processed by that machine. These samples can identify systematic errors widely present in the election. They have only a small chance of catching a hack or bug which was limited to a few precincts or machines, even though that could change the result in close races.

Number of races. In the random sample, most states audit only a few races, so they can only find problems in those races.

Physical security. Auditing is done several days after the election, so paper ballots and computer files need to be kept securely. North Carolina specifies that no audit is done if ballots are missing or damaged.

Physical security has its own large challenges. No US state has adequate laws on physical security of the ballots. Security recommendations for elections include: starting audits as soon as possible after the election, preventing access by anyone alone, having risks identified by people other than those who design or manage the storage, using background checks and tamper-evident seals. However seals on plastic surfaces can typically be removed and reapplied without damage.

Experienced testers can usually bypass all physical security systems. Security equipment is vulnerable before and after delivery. Insider threats and the difficulty of following all security procedures are usually under-appreciated, and most organizations do not want to learn their vulnerabilities.

External links

• Election Verification Network
• Verified Voting
• Principles and Best Practices for Election Auditing