RALPH BECKER MARGARET PLANE Mayor CITY ATTORNEY

CCT [ 3 20·14

ciTY couNCIL TRANSMITTAL Mayor ....

SCANNED TO: ;/)y Date Recei .ANNED BY: Date sent to Cou · TE: /0 -.3 , / c, '

TO: Salt Lake City Council DATE: October 2, 2014 Charlie Luke, Chair

FROM: Margaret Plane, Salt Lake City Attorney Cindi Mansell, City Recorder CJ1;V

SUBJECT: 2015 Vote-by-Mail I Election Information

STAFF CONTACT: Cindi Mansell, City Recorder 801-535-6223

COUNCIL SPONSOR: n/a Council Request

DOCUMENT TYPE: Written Information Briefing and Draft Resolution outlining City Council intent to conduct Vote-by-Mail (VBM) elections for Salt Lake City in 2015. Additional information is included relative to 2014 VBM statistics and information, VBM process and procedures, and technology updates. The "Evaluating the Feasibility of Vote by Mail in " report to the Utah Lieutenant 's Office is also attached.

RECOMMENDATION: Consider Information I Adopt Resolution.

BUDGET IMPACT: Varies depending on type of election to be administered. While there are many expenses from a traditional election that do not apply to a vote by mail election, there are some added expenses, including that Utah law requires the City to pay postage on the ballot return envelope.

Salt Lake County estimates that the costs will be the same for either type of election; however, other cities that have conducted VBM elections indicate a minimum of 10% cost savings. West Jordan is comparable in size to Salt Lake City, and their citywide VBM resulted in a 12% cost savings in 2013. Accordingly, the City could see a potential 12% or greater cost savings with a citywide vote-by-mail election.

1 CITY ATTORNEY'S OFFICE 451 SOUTH STATE STREET, ROOM 505A WWW.SLCGOV.COM P.O. BOX 145478, SALT LAKE CITY, UTAH 84114-5478 TEL 801-535-7788 FAX 801-535-7640 Estimated costs and percentage of voter turnout for elections (with variables) are: • 2013: $1.50/voter and 19% voter turnout. 4 precinct election. • 2011: $2.17 /voter and 23% voter turnout. 3 precinct/Mayor/Citywide election. • 2009: $1.56/voter and 20% voter turnout. 4 precinct election; included citywide General Obligation Bond Election for a proposed Salt Lake City Public Safety Complex. The Primary Election was only in 2 precincts. • 2007: $4.06/voter and 46% voter turnout. 3 precinct/Mayor/Citywide election. This was the first year with the electronic voting machines and the charge from the counties was high. • 2005: $.63/voter 4 precinct election with previous votamatic punch machine system.

By comparison, the estimated costs and percentage of voter turnout for the Salt Lake Citywide Local Opinion Survey Process was: • 2013: $4.93/voter and 24% voter turnout. However, the Local Opinion Survey was very expensive because without access to the State Voter Database we had to hire a vendor to run it.

BACKGROUND/DISCUSSION: At the April 8, 2014 Council Work Session, there was discussion and a straw poll indicating interest in pursuing Vote-by-Mail elections for Salt Lake City in 2015 .

In 2014, Davis County had a Vote-by-Mail primary to increase voter turnout, and to address the high cost associated with replacing electronic voting machines as they begin to fail over time. About 26% of all registered voters cast ballots in Davis County's first­ ever vote-by-mail primary election held June 24, 2014. Although voter turnout for that primary tops the turnout for conventional primaries by as much as 10-12%, county election officials want to reach more voters for the November 4, 2014 general election. The Davis County Elections Office plans to produce a three-minute promotional and information video about the general election vote-by-mail process. The county election office hopes to circulate the video as ballots for military and overseas voters are mailed out, starting as early as the end of September. A survey seeking information about voter impressions of vote-by-mail will be included with the mailed ballots.

Cache County intends to implement its new mail voting system for the November 2014 election. Because ballots are mailed a month before Election Day, another benefit to voters is having four weeks to study the ballots before they are due. No longer will people be surprised by a ballot question after already having stepped into the voting booth. Another advantage to the new system will be savings to the county. Although the county will have to hire more vote counters for the election, the cost of training and employing poll workers should fall significantly. Cache County feels it will be critical to get the word out regarding the change, and one of the biggest concerns is that only one polling place- the Cache County Building in Logan- is scheduled to be open on election day as a fallback for those who did not mail in their ballots. In contrast, Davis County had seven polling stations open. Some residents have expressed fear that their votes will

2 no longer be secret, and others have questioned the likelihood of someone stealing ballots out of mailboxes. Although these are legitimate concerns, there are procedures in place to circumvent ballot snooping and voter fraud.

The State Legislature tasked the Lieutenant Governor's Office with conducting a study about the possibility of a statewide vote-by-mail election, with some vote centers in 2015. The September 17, 2014 report is attached.

STEPS NEEDED TO IMPLEMENT VBM: (These can be delineated further upon Council request). 1. City Council adopt a resolution recommending support for the City of Salt Lake to conduct a citywide "Vote by Mail" Election in 2015 and encouraging the Administration to take the steps necessary to ensure the use of VBM for the 2015 Municipal Primary and General Elections. 2. City staff to work with Salt Lake County Elections staff to prepare an agreement providing for the 2015 VBM Election. 3. City staff will study and recommend the number of Vote Centers they would like to have available citywide for the 2015 elections; Council input is solicited. 4. City Recorder prepares public engagement information such as pamphlets, mailers, postcards, website information, publications, press releases, newsletter inserts, etc. as to the Vote-by-Mail process, why, what to expect and when, how to return the ballot, and Vote Center and ballot drop-box locations. 5. The Salt Lake County Clerk's Office offers information on their website that would be made available on the City website as to how to apply for a vote-by­ mail ballot; that ballots are mailed approximately 30 days before an election (and ballots cannot be forwarded by the Post Office); what to do in case of moving or dealing with a lost or spoiled ballot. 6. Ballots do not have to be returned by mail; they may be returned to any Vote Center or the Office of the City Recorder or County Clerk. 7. A ballot affidavit must be signed by the voter to confirm identity. 8. If a voter wants to vote at a polling location rather than by mail, they can take the VBM ballot and return envelope to a vote center and upon surrender of the un­ voted, unsigned ballot and envelope, they will be allowed to vote on the available electronic machines. 9. All other procedural items remain the same, with a few revisions to the receipt, verification, and counting process: • Ballots that were delivered in the morning by the Post Office are placed in a locked cage until two employees can begin the signature verification process; • After signature verification, the ballots are sorted into boxes that need further review or issues resolved; • The stack of ballots with defects (i.e. defective signatures, two ballots in one envelope, or no ballot, blank, etc.) will be noted prior to the official count; • Ballots are separated from the envelopes. The stack of envelopes is zip tied and the count must match the count of ballots. At this time, ballots

3 are placed in either a "good" pile or a pile that needs to be "enhanced" or duplicated; • In the same room, each ballot is reviewed for clarity, enhancing the ones that need enhancement, and counting the number of ballots that will go to be counted; • All ballots that have not been tabulated will be locked in a secure room, waiting for tabulation; • Tabulation includes three individuals for verification - two employees scan the ballots, while the other seals and labels the box for completion; • Final signatures and verification of ballots. • The boxes will not be opened again unless there is a request for a recount.

ADDITIONAL TECHNOLOGY INFORMATION: In the 2012 election, Go VoteColorado.com enabled Coloradans to register to vote through its mobile device, making the process more accessible and suitable for today's technology-driven lifestyle. The successful initiative had more than 80,000 Coloradans added to the polls from September to October of 2012. More than 229,000 total submissions and updates to existing registrations were logged between August 31 and October 9, 2012. Approximately 20% of these submissions were via the mobile version of Go VoteColorado.com. The mobile-optimized online registration site offered another way for Coloradans to register without having to use a paper form. Paper registrations are more prone to error and require more time to process. As a result of the project's success, the Colorado team is reviewing what went well and improvements that can be made for future elections. In Utah, voter registration is controlled by state statute and the City could not pursue this independent of the state.

In an effort to help voters stay connected and informed, the City of Winnipeg launched the 2014 City of Winnipeg Election mobile app. While the app was not designed to replace more traditional methods of providing information, it allows people to receive instant access to the information they need. The free app is available for download on lOS (iPhone, iPod, iPad), Android and Blackberry devices in both official languages. The City feels the 2014 Election app is a great reference tool and addition to the election outreach and engagement practice. The app is an innovation platform to disseminate information out to voters and may help reduce potential perceived barriers to voting. The app mirrors the information already offered on the City's election website, but with the added convenience of being in a mobile friendly format. A similar app has also been implemented in cities across North America- including Vancouver, Victoria, Calgary and Edmonton.

Previously, information was provided that in both 2011 and 2013, Norway's Parliament permitted internet voting on a trial basis. Between the two elections, use of internet voting more than doubled. However, as of August 2014, Norway's Government Ministry in charge of internet voting has announced that future trials of the technology will be discontinued for the time being. "Due to the lack of broad political will to introduce internet voting", writes the Ministry, it has "decided not to continue expending public resources on continuing the pilots".

4 SUMMARY: Various studies and findings indicate that Vote-by-Mail can be a legitimate and valid way to administer an election. Although cost-savings are typically realized at first, it is uncertain as to whether VBM would increase voter turnout or reduce voting costs over time. Studies indicate skepticism for various reasons. If Vote-by-Mail is adopted, then it should be executed carefully to ensure the integrity of the voting process.

HB ATTY-#4 !236-v2

5 RESOLUTION NO. - --OF 2014

Resolution expressing support for vote by mail for Salt Lake City' s 2015 municipal elections

WHEREAS, the City Council desires to increase voter turnout at municipal elections and to provide voters easier access to the election process, and therefore supports the conduct of the 2015 municipal primary and general elections through a city-wide "Vote By Mail" process; and

NOW, THEREFORE, be it resolved by the City Council of Salt Lake City, Utah, as follows:

Section 1. The City Council hereby expresses support for conducting the City's 2015 primary and general municipal elections through a city-wide "Vote by Mail" process.

Section 2. The City Council encourages the administration to take the steps necessary to ensure the use of a city-wide "Vote by Mail" process for the 2015 municipal primary and general elections.

Section 3. This Resolution shall take effect immediately upon its adoption.

PASSED and APPROVED by the City Council of Salt Lake City, Utah, this day ofNovember, 2014.

SALT LAKE CITY COUNCIL

By ______CHAIRPERSON ATTEST:

City Recorder

APPROVED AS TO FORM:

Salt Lake City Attorney' s Office HB_ATTY-#4 1234-v l STATE OF UTAH O FF I CE OF TH E LI EUTENANT GOVER NOR

SPE NCER J. Cox LIEUTEN ANT G OVER N OR

September 17, 2014

Senator Margaret Dayton Senate Chair - Government Operations Interim Committee 350 North State, Suite 320 Salt Lake City, UT 84114

Representative Jack R. Draxler House Chair - Govemment Operations Interim Committee 350 North State, Suite 350 Salt Lake City, 84114

Dear Senator Dayton and Representative Draxler,

As required by Utah State Code §20A-5-41 0, the Lieutenant Governor's office submits the attached study on voting by mail for your review.

We anticipate supplementing the study with one or more addendums before the September 30, 2014 statutory deadline. In an effort to provide the best possible information our study team is still analyzing data from the 2014 June Primary Election. We will forward their analysis to you and the committee as soon as it becomes available.

Thank you for your continued work and interest in making Utah's elections processes the best they can be.

pencer J. Cox Lieutenant Governor

cc: Representative Steve Eliason Senator Curtis S. Bramble

UTAH STATE CAPITOL, SuiTE 220 · P.O. Box 142325 ·SALT L AKE CITY, UTAH 84114·2325 T EL: ( 801) 538- 1041 • FAx: ( 8 01 ) 538- 11 33 Evaluating the Feasibility of Vote By Mail in Utah

A Report to the Utah Lieutenant Governor's Office

Damon Cann Thad Hall Quin Monson Dept. of Political Science Dept. of Political Science Dept. of Political Science Utah State University University of Utah Brigham Young University

September 17, 2014 Executive Summary

• Vote-by-mail systems are becoming more common across the U.S., and when administered carefully are a legitimate method of conducting elections. Vote-by-mail systems carry advantages and disadvantages that policymakers should be aware of to ensure they select a system that meets their goals and is consistent with their values. • In terms of general election turnout, vote-by-mail systems do not show a general long-term increase in turnout. Gains in turnout, are more likely to be observed in low-salience elections such as special elections and municipal elections. • The cost ofVBM relative to traditional Election Day voting varies across contexts. Ballots must be printed for all registered voters; mailing and postage costs are added to that total. Some jurisdictions pay for the postage to return the ballot as well, while others require the voter to pay return postage for their ballot (or allow voters to drop off their ballots at specified locations). These costs can be substantial. Our survey of county officials substantiates the idea that cost savings will vary across contexts. • Documented cases of voter fraud are rare under existing VBM systems in Oregon and Washington. Using security measures like barcodes, security cameras, and signature verification is important for conducting a high­ integrity election. • Public opinion in Utah tends to favor maintaining a system with choices for voters including mailed absentee ballots, early voting, and traditional Election Day voting. However, VBM is popular in states that have already made the transition. • Utahns generally seem to recognize that VBM has some important advantages (convenience and ease of voting) and perceive it to be a system that tends to count votes correctly, but they also express some reservations about the possibility of fraud and the possibility that they could cast their ballot before late-breaking campaign information becomes available. • A survey of county clerks shows substantial variation from county to county on perceptions of whether VBM is easier or harder to administer and whether VBM saves money or is more expensive. • Like all elections, VBM can be done to a high standard of integrity or it can be done haphazardly. Our survey of counties shows that many counties would need upgrades in security to protect ballots from theft and disaster (e.g. fire, water). A rigorous and careful signature verification process is important, and clearer voter intent standards could be helpful. • Most counties have the space necessary to transition to a VBM system, but many would need additional technology to track, open, sort, and scan ballots.

1 Introduction: Vote-by-mail (VBM) election adm ini stration has grown in popularity, particularly over the last ten years. Currently three states (Oregon, Washington, and Colorado) and multiple local jurisdictions in the United States use all vote-by-mail systems. Many more use mailed-in ballots for absentee balloting. The purpose of this report is threefold: To provide information to Utah policymakers about existing research on the consequences of a vote-by-mail system, to provide Utah policymakers with information about citizens' perceptions of vote-by-mail systems, and to provide information about best practices for administering a vote-by-mail so policymakers have advice on how to make such a system work. We remain neutral on whether the state should make such a change and instead seek to provide the most pertinent information about advantages and disadvantages of such a system to policymakers to help them make an informed decision.

Utah already makes use of mail-in ballots for absentee voters in a variety of ways. A handful of local jurisdictions have already switched to VBM systems, including Davis County and Duchesne County. Cache County will conduct its first full VBM election in the November 2014 general election, and Sevier, Cache, and Weber counties have used VBM for particular bond elections in the recent past. At least 33 Utah municipalities, mostly small towns, have conducted vote-by-mail elections. Utah Counties Reporting Use A typical vote-by-mail system of an all VBM Election involves mailing ballots (and instructions Beaver for completing them) to all registered Davis voters about a month in advance of the Duchesne election. Lost or damaged ballots may be Garfield replaced, though caution must be Grand exercised in distributing replacement Sanjuan ballots without restraint. Voters receive Sevier a ballot and a pre-addressed envelope for Weber returning the ballot. The voter marks their ballot, places it in the return envelope, and signs in the appropriate place on the return envelope indicating that they are the one who cast the ballot. When ballots are returned, signature verification should be performed to minimize the possibility of election fraud. Ballots are opened and counted by election judges at a time and place where election observers can be present. Votes are counted as received before a particular elate (as in Oregon) or as long as they are postmarked by a particular date (typically Election Day, as in Washington). States requiring ballots to arrive by a particular date and time can then announce election outcomes typically on the same cl ay as the deadline. When the cutoff is designated by th e date of postmark rather than the elate of receipt, complete results may be delayed by several days.

Our report proceeds with some background information on vote by mail and its impacts on jurisdictions that adopt it. We also discuss a number of aspects of

2 VBM systems th at policymakers may want to be aware of. We then discuss public perceptions ofVBM sys tems both nationally and using unique survey data collected for this project. The final section provides technical details on processes and procedures necessary for a successful implementation.

Section 1: Background Information on Vote by mail

Voter Turnout and Vote by mail

The first reason many jurisdictions consider adopting a vote-by-mail (VBM) system is because they believe a VBM system will increase voter turnout. The notion that VBM systems increase turnout comes from the idea that voting on Election Day is "costly" in terms of time and effort. U.S. elections are on a Tuesday during the work-day and typically involve voting on dozens of races and referenda, which require information in order to make informed vote choices. It has been argued that VBM will increase turnout by lowering the costs of voting, since the ballot for every election comes to the voter ahead of time and can be completed in the comfort of one's own home.

Early research on Oregon's VBM elections showed a considerable increase in turnout, as large as 10 percentage points in some of the elections immediately after the switch. Subsequent research shows that this initial effect is attributed to a "novelty" effect. 1 After the novelty wore off (typically after 3 elections), most elections showed little or no increase in turnout. In special elections and municipal, VBM may have some effects on turnout because receiving a ballot in the mail reminds voters to vote in these less visible elections.

To supplement the data on primary and general elections compiled from Oregon, we analyzed the levels of voter turnout in local elections for municipalities in Washington for the three elections before and three elections after a municipality switched to a VBM system. The data appear in Figure 1. While the election immediately following the transition to a VBM system shows a jump of nearly 10% in voter turnout, the effect is substantially diminished by the second and third elections after the VBM reform was implemented.

Several Utah jurisdictions have begun using a VBM process. However, only a handful have bee n using VBM for a long enough period of time for us to evaluate the effect ofturnout on their elections after the novelty has worn off. Figure 2 shows the levels of turnout observed in Utah as a whole compared against three municipalities which have used VBM over at least four elections: Garden City, Parago nah, and Rockville. Consistent with previous research from outside of Utah, th e tre nd lin es show most of these cities experiencing gains in the first year in which elections are conducted by mail followed by decreases in the level of turnout.

1 Gronke & Mill er 2012, So uthwe ll 2009 3 Figure 1: Turnout in Washington Municipal Elections Pre- and Post- VBM Reform

,i --65- --

60

55 ~ /~

::.R ':.. 50 ~ / =0 '" c - Turnout 5 45 --- E-< 40 L

--

I :: L - ~Year-3 Year-2 Year-1 Year 1 Year 2 Year 3 Note: Year 1 refers to turnout in the first year after the adoption ofVBM with each subsequent year representing the subsequent municipal election years. Year -1, Year -2, and Year -3 a re the three years preceding the switch to VBM Elections.

Figu!:~~ £_!urnou~ in _!!_~_!! MuniciQ_al ElectiO!!~ p_1~~~ ~nd _PQSt- VB_M R_efQ_~_l)! _

90 ~------

80

70

60 - Utah Turnout - Ga rden City - Paragonah - Rockville

20

- - .------.- t:~-~-e-a~- --~ 2-~-:_Y_ea-r-- -1 ---. Year 1 Year 2 Year 3 Note: Year 1 refers to turnout in the first year after the adoption ofVBM with each subsequent year representing the subsequent municipal election years. Year -1 and Year -2 are the two years preceding the switch to VBM Elections.

4 Ba sed on patterns observed in other states where VBM has been implemented, one would expect that if Utah adopted VBM there would be an initial increase in turnout in general elections, but that the initial increase would be short­ lived and as the novelty of the VBM system wears off there will be little or no gain in general election turnout (though some gain may be observed in special elections and municipal elections).

VBM may decrease the effectiveness of get-out-the-vote (GOTV) campaigns. A field experiment in San Diego showed that, among medium and high propensity voters, VBM precincts were less responsive to GOTV campaigns than traditional precincts. VBM and other early voting mechanisms may also negatively impact turnout by reducing the visibility of Election Day, reducing the impact of ads, and lowering the level of media coverage due to the extended voting window. Increases in VBM turnout have also been linked to the number of reminders or informational pamphlets sent to voters.

Residual Votes and Ballot Roll-off

The adoption of electronic voting machines after the 2002 passage of the Help America Vote Act (HAVA) resulted in decreases in ballot errors because the machines can "check" the activities of voters in real time while preserving the secret nature of the ballot. For example, in a race where the voter should only vote for one of three candidates, an electronic voting machine can alert a voter who attempts to mark their electronic ballot for two candidates in the same race while paper ballots cannot prevent these errors. Voting for more candidates than allowed is known as an "over-vote" and results in a vote that is uncounted in the over-voted race.

In other instances, voters may have intended to vote in a race but accidentally skipped marking off their preferred candidate in the race. Voting machines can check to ensure that a voter intended to skip a race. Instances where voters fail to select a candidate in a down-ballot election where they voted in up­ ballot races are known as under-votes. The total of over-votes and under-votes is known as the residual vote for a particular race.

A concern when switching to VBM is that it will lead to more mistakes on the ballot that would not be allowed by an electronic voting machine. A 2011 study examined how VBM affected the residual vote rate in California jurisdictions where it is used, and found that VBM increased the residual vote rate in most counties.2 During the 1990-2010 time frame the residual vote rate dropped as voting technology became more accurate, but that by 2010 VBM had essentially negated any gains in the residual vote rate du e to better technology.

2 Al varez, R. Michael, Dus tin Beckett, a nd Cha rl es Stewa rt Ill. "Voting Technology, Vote-by-Mail, a nd Residual Votes in Cali fornia, 1990-2010". Politica l Research Quarterly, no. 3 (S ept 2013): 658-6 70. 5 Our own analysis of data on res idu al votes in Washington elec ti ons from 1996-2012 fo und less of an effect, with the increased residual vote rate being mo st prevalent in judicial elections.3

Rejected Votes

Another important metric rel ating to VBM is the percent of ballots that are received but are not counted. The reasons for rejection may range from missing or incorrect signatures to sending back the wrong ballot or sending it back too late. For most states, the rejection rate for mailed absentee ballots is under 2%. 4 For comparison, Utah rejected 1.1% of absentee ballots received in the 2012 election while Washington rejected 1%, Oregon 2.3%, and Colorado .9%.s Since 2010, Washington State has required all counties to fill out reconciliation reports after each election that account for how many ballots were rejected to ensure election integrity. From these reports we found the percentage of ballots that were rejected. The rejection rate range d from 1%-1.5% of all ballots received.

While the percent of ballots that would likely be rejected in a VBM system would remain around the sa me as current absentee ballot rejection rates, it will end up affecting a much larger number of Utah voters.

Timing of Adoption

Th e timing of adoption for VBM can affect how smooth th e transition is, what the effects are, and how it is received by citizens. This section will compare Utah to where Oregon, Washington, and Colorado were when they made the switch to VBM.

Oregon has th e nation's longest experience with a VBM system. In 1981, the Oregon state legislature allowed several counties to test VBM in local elections. By 1987 VBM was a permanent election system for most counties, and expanded to include special elections. During the 1990's Oregon used VBM for several statewide races and primaries. Th ey conducted th e first fed eral election by VBM in 1996 when they repl aced a senator in a special election. Finally in 1998 the voters decided to implement VBM for all elections.

Because their VBM transition was the fir st in the nation, their implementation ofVBM over time helped make it successful. Usin g it in local

3 Our a nalysis was based on a two-way fi xed-effects regression to control for differences between election years and counties. Fu ll details of the stati stical analysis are avail able from the a uthors upon request. 4 Burde n, Barry C. and Brian]. Gaines. "Adm in istrati on of Absentee Ballot Programs." Prepared for the Presid e ntial Commission on Election Administration, july 15 2013. 5 U. S. Election Assistance Commission, "2012 El ection Administration and Voting Survey". http: j jwww.eac.gov j assets/1/Page/990-050%2 0 EAC %2 OVo terSurvey_508Co m pi iant.pcl f

6 elections allowed the state to find the most cost effective way of conducting a VBM election, all owed trust in the VBM system to grow before implementing for state and federal elections, and has ultimately led to a state th at is proud of th ei r VBM experience.

Washington's transition to VBM has be en much more recent. Prior to 2005, precincts with fewer than 200 voters were allowed to use VBM in state and federal elections. One county, Ferry County, began using VBM for all elections in the 1990's. Influenced by the success ofVBM in Oregon, Washington counties saw very high absentee voter rates that continued to grow in the early 2000's. This amounted to essentially running a VBM and traditional election simultaneously, which was costly. In 2005, th e legislature passed a bill allowing counties to switch to VBM for all elections. Most counties made the switch immediately, and in 2011 the last county (Pierce County) made the switch to VBM. As opposed to Oregon where complete VBM began with its success in local elections, Washington adopted VBM in response to demand from its voters.

Table 1 shows the absentee voter rate in Washington in the 2003 local election and 2004 federal election by county. This shows the very high absen tee voter rates for most counties, which made for a smooth transition for both election officials and the voters.

Colorado's transition to vote by mail earlier this year was similarly preceded by widespread use of mail ed absentee ballots. Most counties in Colorado had over 50% of the electorate alrea dy voting by mail through th e absentee mechanism and many had absentee rates over 60%. For the reader's convenience, Ta bl e 2 shows the percentage of permanent vote-by-mail voters in Col orado counties just prior to their switch to a VBM system.

Utah's position has not advanced nearly to the same point of frequent absentee voting observed in Washington or Colorado at th e time they switched to a VBM system. The percentages of Utah voters casting ballots by mail (either absentee or other VBM in counties offering VBM in th e 2012 election) are depicted in Tabl e 3 below. The Utah absentee or VBM figures are much lower than those in Colorado and Washington at the time of their switch.

7 mail Absentee V before Vote b Mail

Adams 90.33% Asotin 54.62% 41.86% Benton 71.57% 57 .38% Chelan 89.56% 82.2 1% Clallam 100% 100% Clark 88.29% 70.63% Columbia 64.17% 58.65% Cowlitz 92.24% 66.95% Douglas 83 .29% 77.82% Ferry 100% 100% Franklin 39.75% 40.34% Garfield 46.47% 74.23% Grant 68.47% 63.87% Grays Harbor 66.61% 62.56% Island 78.11% 73.91 o/o Jefferson 75.8% 73.06% King 71.33% 62.83% Kitsap 86.5% 78.11 o/o Kittitas 54.31% 46.85% Klickitat 57.33% 42.7% Lewis 91.15% 98.83% Lincoln 24.32% 25.63% Mason 61.21% 58.99% Okanogan 29.01% 100% Pacific 58.58% 53.06% Pend Oreille 100% 100% Pierce 86.15% 80.3% SanJuan 72.26% 63.53% Skagit 59.52% 57.91 o/o Skamania 100% 100% Snohomish 74.2% 67.62% Spokane 71.41% 64.3 4% Stevens 57.85% 51.44% Thurston 84.6% 73.91% Wahldakum 77.58% 73 .84% Walla Walla 61.42% 65.53% Whatcom 71.98% 72.85% Whitman 34.9% 34.46% Yakima 82.67% 73.88% TOTAL 75.66% 68.45%

" Information obtained through the Washington Office of the Secretary of State found at http: //www.sos. wa.gov / electi ons /absentee stats.aspx?y=2003. 7 Counties could use full vote-by-mail in precin cts that had fewer than 200 registered voters. Some counti es drew their precincts to fit the criteria, thus allowing them to be fu ll VBM prior to the 2005 legislation. 8 Adams 62.3% La Plata 52.7% Alamosa 63.9% Lake 36.3% Arapahoe 67.9% Larimer 67.2% Archuleta 50.3% Las Animas 53.3% Baca 61.8% Lincoln 60.4% Bent 47.7% Logan 65.8% Boulder 66% Mesa 68.2% Broomfield 70.2% Mineral 67.2% Chaffee 72.6% Moffat 37% Cheyenne 49.6% Montezuma 48.8% Clear Creek 56.7% Montrose 69% Conejos 51.1% Morgan 63 .8% Costilla 39.9% Otero 53.9% Crowley 37.3% Ouray 63.7% Custer 48.4% Park 61.3 % Delta 70.6% Phillips 63.5% Denver 61.5% Pitkin 34.3% Dolores 54.6% Prowers 59.3% Douglas 71.2% Pueblo 56.1 o/o Eagle 57.5% Rio Blanco 57.6% El Paso 58% Rio Grande 51.6% Elbert 64.9% Routt 52% Fremont 54.6% Saguache 56.9% Garfield 62.9% SanJuan 29 .1 o/o Gilpin 49.1 o/o . San Miguel 52.4% Grand 57.2% Sedgwick 60.9% Gunnison 55.9% Summit 46.7% Hinsdale 46.6% Teller 54.8% Huerfano 57% Washington 59.1% Jacl{son 44.5% Weld 62.9% Jefferson 72% Yuma 56.4% Kiowa 54.4% Kit Carson 44.8%

8 Data was obtained through the Colorado Secretary of State's website. The data may be found at http: I / www.sos.state.co.us /pu bs /elections IV oterRegN umbers /2 014 /lanuary /Total PM IVReq uests.p Qf. Permanent absentee statistics were last updated February 3~'", 2014. Active voters are de fined as those who have voted in the last election, updated their information sin ce the least election, or recently registered to vote.

9 Table 3: Methods of Voting in Utah by County, 2012 General Election

Beaver 8.46% 40.15% 50.03% 1.36% Box Elder 16.54% 20.43% 58.22% 4.81% Cache 8.10% 28.3 7% 62.37% 1.17% Carbon 4.95% 36.07% 55.70% 3.28% Daggett 48.62% 50.59% 0.00% 0.7 9% Davis 13.99% 34.89% 50.40% 0.71% Duchesne 100.00% 0.00% 0.00% 0.00% Emery 23.58% 75 .89% 0.11% 0.42% Garfield 33.87% 14.89% 50.79% 0.45% Grand 22.54% 77.33% 0.06% 0.06% Iron 17.21% 20.30% 57.32% 5.17% Juab 7.64% 27.66% 64.62% 0.08% Kane 22.67% 24.59% 52.35% 0.38% Millard 40.17% 55.30% 1.08% 3.46% Morgan 10.17% 31.50% 56.08% 2.25% Piute 11.16% 23.75% 65 .09% 0.00% Rich 34.53% 65.47% 0.00% 0.00% Salt Lake 27.09% 20.50% 49.92% 2.48% SanJuan 20.52% 33.63% 45.78% 0.07% Sanpete 19.95% 14.18% 65.32% 0.55% Sevier 9.26% 22.78% 67.41 o/o 0.55% Summit 11.46% 33.87% 51.05% 3.62% Tooele 15.92% 27.44% 51.22% 5.42% Uintah 8.10% 28 .87% 60.77% 2.25% Utah 11.03% 19.36% 69.48% 0.12% Wasatch 11.50% 28.96% 56.71% 2.84% Washington 20.80% 24.79% 48.85% 5.57% Wayne 14.43% 16.98% 68.52% 0.08% Weber 8.73% 30.40% 58.64% 2.23%

9 Data was obtained through the 2012 Utah Vo ter File. 10 Includes offic ial VBM and absentee voters. 11 Includes in-person early voting. 12 Includes tradi tional polli ng place vo ters. 13 Includes a ll provisional vo ters regardless of whether th ey were early, traditional, or absentee. 10 Voter Identification vs. Signature Verification

Like Utah, many states have adopted voter identification laws over the last several years. These have been intended to ensure that the individual casting the ballot is indeed the registered voter. While relatively few incidents of voter impersonation have been documented, perhaps it is the possibility that some have gone undocumented that motivated these laws.

In a VBM system, voter verification is conducted through a process of signature verification instead of voter id en tification. Colorado (like many other jurisdictions) uses software that is designed to take signatures from voter registration or DMV documents and compare them with signatures on ballot envelopes. Such software typically allows the user to set the program's sensitivity to differences between signatures. Signatures with a certain level of difference are flagged by the software for further evaluation by election judges (who should be trained in signature verification techniques). The effectiveness of signature verification will depend on the quality of the signatures available on voter rolls and/or through the DMV. We anticipate that issues with signature verification are less likely to occur with mass fraud (as ballots are kept secure and mass forgery of signatures with some degree of accuracy poses a formidable challenge). If an issue arises, we suspect it will be with valid votes that are rejected because of signature differences rather than massive numbers of fraudulent ballots with forged signatures.

Partisan Consequences ofVBM

One common worry with any change in an electoral system is that the change will benefit one political party or demographic group at the expense of another. Scholarly studies suggest that VBM does not alter the composition of the electorate in any meaningful way. VBM primarily makes voting easier for people who would have voted anyways rather than mobilizing new voters. Studies in Oregon have shown that switching to VBM has not favored any particular party or demographic group. 14

Section 2: Public Opinion on Vote by mail

When evaluating any policy change, it is natural to wonder about public support for the new policy. Surveys conducted in 2008 and 2012 by the Caltech/MIT Voting Technology Project found that, in Utah, only 11.9% of respondents in 2008 and 20.9% in 2012, supported all VBM elections. Across all states, VBM as an election reform received majority support only in Oregon and Washington (the two states th at had implemented full VBM elections when the survey was fielded).

11 Southwe ll 2009 11 Following up on the results of a national survey with a modest sa mpl e size for Utah, we fielded questions on the October 2013 Utah Voter PolllS to gauge public opinion about possible reforms. Utahn's tend to approve of the traditional system, with an overwhelming 87% in favor of keeping "the current system with a choice of absentee voting by mail, early in-person voting, or voting on Election Day," while only 3% favor a system that would "use only mail ballots for all voters."

When asked to rate their agreement with various statements about vote by mail, voters acknowledge various strengths and weaknesses of VBM systems. Voters agree with arguments on both sides of the issue. We have fielded several original polls to assess public support for VBM initiatives. Majorities agree with the ideas that a mail ballot could be cast privately and that VBM elections would "be more convenient for me." However, majorities also agree that VBM "would encourage fraud," and that "a single day where all voters cast their ballots is a valuable part of our election system."

We wanted to determine whether experience with a VBM election influences voters' opinions on this issue. To that end, we've fielded a survey in Davis County to ask those who voted in the VBM primary election and compare their responses to the attitudes of individuals who had a high probability of voting in elections (as determined by a statistical forecasting model) and compare their responses. 16

Our first question was a replication of the Utah Voter Poll question we r·eferenced earlier. Our results (presented in Table 4) show that for both VBM voters and non-voters in Davis County, there is a strong preference for a wide range of options in the type of voting allowed. Perhaps most surprising are the minimal differences between voters and nonvoters in their preferences.

ts The Uta h Voter Poll is fi elded by Brigham Young University's Center for the Study of El ections and Democracy pe ri odically throughout the year and has a s trong reputation for accuracy. t 6 This s urvey continues to amass responses in the fi eld as of this w riting and the final re port w ill share res ults w ith a larger sample s ize a nd appropriate margins of error. The survey was fi eld ed on line by using a data warehouse to match information from the voter fil e to ema il addresses for the voters. Our match rate was approximately 50%. We then sampled nearly 25,000 email addresses, s plit approximately evenl y between voters and the non-voters with the hi ghest probability of voting as determined by a statistical model incorpora ting vote hi story. 12 Table 4: Preferences for Election System in Davis County by June 2014 p nmary. p artlctpatwn VBM Voters Non-Voters System with absentee 83% 83% mail, early in-person, and Election Day voting All VBM 8% 6% Election Day only 9% 12% n=217 for non-voters and n =304 for voters

In Table 5, we show a comparison between those attitudes of voters and high-probability non-voters in the june 2014 Davis County primary election. The survey reveals a reasonable amount of consistency across groups, though non­ voters differ from voters in a number of potentially important respects. just over 10% more non-voters express concerns about making mistakes on the paper mail-in ballot than express those concerns among voters. Relative to those who voted, about 11% fewer non-voters agree that voting by mail is more convenient for them, though the percentages are quite high. Concerns about fraud are higher among non­ voters than they were among voters. Interestingly, voters were much more likely to perceive VBM as a cost-saving move than non-voters.

Across both groups, though, concerns about mailing ballots too late, failure of the post office to deliver their ballot, and th e ability to vote privately in a VBM format are minimal. The most substantial concerns raised pertain to th e potential for fraud, th e potential for late-breaking information to make them want to change their votes, and the value of a single Election Day.

Finally, we asked voters whether it was a straightforward task to complete the ballot according to the instructions they received. The results were very clear that voters felt the instructions were clear and straightforward. Fully 73% of voters expressed that following the instructions was very easy and an additional 24% said the instructions were somewhat easy to follow. Only 2% said they were somewhat hard (none replied the directions were very hard) and the remainder indicated they couldn't remem ber how hard the in struction s were to complete.17

17 The sample si ze for this question was n=291 voters. 13 Table 5: Attitudes about Pros and Cons ofVBM Systems in Davis County by June 2014 p nmary . p articipatwn. Percent ofVBM Voters Percent of Non-Voters agreeing agreeing I am concerned I would 34.0% 32.7% return a mail ballot too late for it to be counted I worry about making 17.0% 28.57% mistakes when completing and returning a mail ba ll ot I could vote with a mail 90.7% 86.51% ballot in privacy People in my family could 21.59% 27.07% try to cast my ballot for me or influence my vote in a by mail election I am confident the Post 87.05% 79.81% Office would deliver my mail ballot to the county clerk's office Casting a ballot by mail 81.79% 70.51% would be more convenient for me An election with mail 52.34% 63.77% ballots would encourage fraud An election conducted 63.0% 46.48% entirely by mail would cost less than traditional Election Day polling place voting A single day where all 54.63% 56.88% voters cast their ballots is a valuable part of our election system. I worry that I would cast 51.98% 55.76% my ballot early and then hear important information that might change my vote. n for non-voters ranges from 213 to 218 depending on skipped responses. n for voters ranges from 301 to 30 4- depending on skipped responses.

14 Section 3: Administering Vote-by-Mail Elections

As a part of our analysis of all-mail voting, we talked to individuals in the Secretary of State's office in Colorado, Oregon, and Washington, as well as local election officials in Colorado and Washington, to learn about their experiences transitioning to vote by mail. We also examined the laws, regulations, and materials related to voting by mail in these three states.

Based on this review, we developed a survey of questions that were sent to county election officials in Utah. The questions focused on specific policy areas that we identified from our document reviews and discussions with election officials in other states. We also asked Utah election officials to speculate as to whether moving to vote by mail will make elections easier to administer and less costly to administer.

Attitudes toward Cost and Ease of Administration

We asked county officials two questions related to cost and administration of vote­ by-mail elections. First, we asked: "If your county were to switch to all vote-by-mail elections, do you think it would increase the cost of holding an election, decrease the cost, or would the cost remain about the same?" Of the 24 counties that responded: • 42% (10/24) said it would increase costs; • 33% (8/24·) said it would decrease costs; and • 25% (6/24) said costs would stay the same.

Even among the six jurisdictions that have conducted vote-by-mail elections, three of them think it would increase costs and three think it would decrease costs.

Second, we asked: "If your county were to switch to all vote-by-mail elections but were required to also operate a limited number of early voting locations as well, do you think it would make your elections easier to administer, harder to administer, or about the same to administer?" Again, 24 counties responded and: • 21% (5/24) said it would be easier to administer; • 50% (12/24) said it would be harder to administer; and • 29% (7 /24) said it would be the same to administer.

Even among the six jurisdictions that have conducted vote-by-mail elections, three of them think it would make elections easier to administer and three think it would be harder to administer.

15 Electoral lntegritv

Any transition to a new voting technology is ensuring that the electoral process has a high level of integrity. Ballots need to have strong levels of security over an extended period, as voted ballots will be returned and stored for over a month. All three states with all vote-by-mail elections have extensive laws, rules, and procedures related to ensuring that ballots are located in areas with appropriate security and access control. We asked counties about the measures that they take to secure their absentee ballots currently. Of the 25 counties that answered the security questions: • 9 (35%) counties have 24· hour video surveillance cameras in their elections area; • 3 (12%) log on paper entry and exit to the area where ballots are secured; • 6 (23%) log electronically entry and exit to the areas where ballots are secured; • 5 (19%) have a secured space with a two-person access requirement for accessing secured space for ballots; • 4 (15 %) have an alarm system; • Half of counties (13) have fire security for ballots (sprinklers or fire boxes); and • Only 7 counties (27%) have security against water damage for ballots (e .g., if the sprinklers went off).

Of th e seven security items noted above, six (6) counties utilized zero of them, and (4) counties only had one of them.

The other key issue related to the integrity of absentee ballots is the signature verification process. Properly doing a signature verification is actually quite difficult. In Washington State, for example, there is a legal requirement that all county election officials receive training by the Washington State Police to learn about th e proper ways for conducting signature verification. In a vote- by-mail environment, signature verification is the equivalent of checking an individual's photo identification. We asked counties about several issues related to signature verification and determined that: • 6 (23%) counties have signature verification software; • 16 co unti es report doing staff training related to signature verification; • 5 counties provide a checklist of what to look for when doing signature training; • No co unti es do any sort of training via law enforcement on signature verification; and • 7 counties report doing no trainings related to signature verification.

16 Ballot Duplication an d Voter Intent

When voters cast ballots on paper, the ballot can be damaged or the voter may mark the ballot in a way th at cannot be read by the ballot scanning eq uipment. Ballots that have marking problems can be enhanced - the marks on the ballot are darkened- or duplicated- the voter's choices are transferred onto a new ballot and the new ballot and old ballot are linked by a unique identifier. More ballots cast on paper mean more potential for the need for ballot enhancement and duplication, both of which are time consuming processes requiring two staff members. Of the 26 co unties that answered questions related to duplication and enhancement, 22 (85%) have a process for reviewing absentee ballots to see if they need to be enhanced and 24 (93%) have a process for reviewing absentee ballots to see if they need to be duplicated because of a defect. In 2012, most counties duplicated fewer than 60 ballots but one county duplicated 200, one 670, and Salt Lake County duplicated 7,000.

Duplicating a ballot correctly requires being able to interpret voter intent. Voter intent laws and regulations should provide clear guidance regarding what types of ballot markings constitute a vote. For example, if a voter fills in the bubble completely next to "Gary Herbert", another voter places an "X" in the bubble next to "Gary Herbert, another voter puts a check mark in the bubble next to "Gary Herbert", and yet another voter does not mark the box but writes in "Gary Herbert", have all of th ese individuals voted for "Gary Herbert"? A clear voter intent standard answers this qu estion.

We asked th e counties if they thought ~hat the voter intent law and regul ations in Utah were clear and, of the 23 counties that res ponded • 7 (29%) said current law is very clear and specific; • 13 (54%) said it is somewhat clear and specific; • 3 (13%) said it is not very clear or specific; and • 1 (4%) said it is not at all clear or specific.

When asked how they train workers to discern voter intent, 21 co unti es said that have two (2) poll workers examine th e same ballot and compare th e results, 11 counties provide exa mples of correct and incorrect markings, and 2 counties have a short course for workers on voter intent. Th ere were 19 counties out of 24 who said that they send voter ed uc ation materials to voters with their absentee ballots so that they know how to mark the ballot correctly.

Space and Technology

For an all vote-by-mail election, a coun ty needs to have a secured space for processing ballots, space th at ca n be secured for up to six-weeks, as ballots are returned, then counted, and canvased. Only 4 of 26 counties res ponded th at they did not think th ey had th e space to process ballots over multiple days.

17 Counties also will likely need technology tools to make the processing of ballots easier. Ballots need to be marked as returned, signature checked, sorted by precinct, and sliced open so that the ballot can be removed. All of these processes can be made easier with technology, and most jurisdictions do not have at least some of the equipment they may need. There were 26 counties who responded to these questions, and: • 17 counties (65%) have software and hardware for scanning mail barcodes that would be effective in all VBM environments; • 3 counties (12%) have an automated letter opener for opening absentee ballots; • No counties have electronic ballot/mail sorting equipment, to sort ballots by precinct; and • 17 counties (65%) have enough ballot scanning equipment to scan ballots effectively in all vote-by-mail environments.

Disability Services

Individuals with disabilities need to be served effectively, regardless of the voting technology used. With no regular regime of polling places, voters with disabilities have to be served in other ways. We asked counties about various methods of serving individuals with disabilities - methods that are used in other all vote-by­ mail states. We asked counties whether they currently offer these services to permanent absentee voters with disabilities and then whether they thought that they could offer these services, in an all vote-by-mail environment. Many jurisdictions do not provide many of these services and some accommodations would need to be made so that the needs of individuals with disabilities were appropnatelyI I1an dl e d . Currently Could Provide Provide Two Election Workers visit a voter's home 3 4 Multiple Satellite Voting Locations 6 6 Visit Community Centers 3 4 Visit Senior Centers 8 12

Primary Elections Primary elections in Utah can be complicated because non-Republican voters who are affiliated with another party can vote in any primary election for any other party. This can create complications for election officials. In add ition, unaffiliated voters need to be informed of their ability to affi liate prior to the primary election. In asking about primary elections, we found that, of 24 counties responding, 22 counties (92%) notify unaffili ated absentee voters of their need to affiliate to participate in primary elections. Fo r non-Republican absentee voters, in primary elections 18 counties (75%) send the voter only the primary ballot for the party 18 with which they are affiliated and 6 (25%) send the voter all of the primary ballots for which they are eligible to vote and let the voter return one of them.

The Experienced Counties: Lessons from VBM Five counties with all vote-by-mail election experience responded to questions about their experiences with VBM and we learned that: • All sent information about the candidates on the ballot, if applicable; • All sent information about how to mark ballots correctly; • 1 of 2 sent information about the ballot measures on the ballot; • 1 of 5 provided information about where the voter could drop off a ballot; and • 1 of 5 provided information about where the voter could vote in person.

When asked about interactions with their USPS representative, we found that all met with their USPS representatives to discuss envelopes and ensure they were the right size, and all informed the postal service to expect to discuss hi gh volumes of mail. Four counties met to discuss any possible delivery issues (e.g., to rural areas) and four discussed the process of securing returned ballots at USPS. Three counties discussed the process of securely transferring ballots from the USPS to the county.

Mailing Services: In-Hou se or Contracting All vote-by-mail elections require sending every registered voter a ballot. This can be done in-house or contracted to a mailing house. Of the 24 counties who answered questions about these services, 8 would want to assemble ballot packets (ballot, return envelope, etc.) in-house and 16 would want to contract with a mailing house for this work. Contracting with a mailing house requires some effort to ensure that there is quality control regarding how the mailings are done. When we asked counties how they would supervise contracted mailing: o 4 would have staff visit the mailing house during the election period; o 3 would want a third party audit of mailing houses; o 7 would want the Lt. Governor's office have a certification program for mailing houses; and o 4 gave another answer (do it themselves, communicate with mailing house, depends).

Vote Centers In Colorado, vote by mail is done in conjunction with vote centers, where voters can cast ballots in the 14 days prior to the election. Smal l counties have to have at least one vote center and larger counties have to have more. We asked several questions that pertain to the implementation of vote centers.

First, we would note that there are 6 jurisdictions with a large number of ballot styles; 3 counties have between 17 and 29 ballot styles and 3 have between 100 and 372. To do vote centers, these jurisdictions would need to have robust ballot on

19 demand technologies or electronic voting machines to accommodate the large number of ballot styles.

Second, there are 12 jurisdictions who only have a single early voting location and one has no early voting locations. This could have implications were jurisdictions to have vote centers in conjunction with all VBM.

Third, 52 % of all responding jurisdictions do not have "live" voter registration­ voter lists connected to the internet- for Election Day voting and 69% (18 counties) have live VR for early voting. This also has implications for i.mplementing vote centers.

Conclusions

Our findings in this report suggest that vote by mail can be a legitimate and valid way to administer an election. However, just as traditionally-administered polling-place elections can be run with higher standards and better practices, the manner in which VBM is implemented is important for maintaining the integrity of the voting process. ln the course of our report we note that various counties have had different experiences with VBM and absentee systems. The counties, from their sundry vantage points, also have different perceptions about the impacts VBM would have on their counties. We see wisdom in accounting for the variety of experiences and perceptions among counties in crafting policy. We encourage policy makers to carefully consider the pros and cons ofVBM systems before making changes in the manner in which they administer their elections, and to ensure that where VBM may be adopted it is done carefully to ensure the integrity of the election process.

20 Election-by-mail trend changing how Utahns vote and campaign I The Salt Lake Tribune Page 1 of2

~b.e inlt !nkt Wribunt

Election-by-mail trend changing how Utahns vote and campaign•

Change o A third ofUtah counties are conducting elections entirely by mail.

BY LEE DAVIDSON

THE SALT LAKE TRIBUNE

PUBLISHED: OCTOBER 3, 2014 07 :26AM UPDATED: OCTOBER 2, 2014 10:05PM

The Nov. 4 election is still a month away, but hundreds of thousands of Utah voters already are receiving their ballots by mail - or will v-i thin days. The trend is changing how candidates campaign, and how parties get out the vote.

About a third of Utah's counties- nine of 29 -this year are conducting elections almost entirely by mail, and either have sent their ballots already or will by law by Tuesday. The rest of the counties also allow individual voters the choice to request ballots by mail, or to cast them in person. Salt Lake County is among that group and sent out 126,500 ballots by mail this week as requested by 29 percent of the area's registered voters, said County Clerk Sherrie Swensen. Francisco Kjolseth 1 Tribune file photo Utah Democratic Party Chairman Peter Corroon said it makes a difference in campaigning.

"It means our candidates have to be mailing out materials earlier and reaching voters earlier because people now have the opportunity to take a month to vote." Instead of waging one big get-out-the-vote drive, he said parties and candidates now must do it in three waves: one for those who vote by mail, another for those who vote early in-person, and a final push on Election Day. Utah Republican Party Chairman James Evans says voting by mail brings advantages.

Positives o "One of the upsides is candidates are able to see who has already voted" by looking at constantly updated lists compiled by county clerk offices. "So [candidates] can be more targeted in their communication so they do not have to send mailers to people who have already voted." Evans said by-mail voting also allows residents at home, with a ballot in hand, to study candidates and issues "for a longer period of time instead of just walking into a ballot booth and making a decision."

Both parties have encouraged voting by mail because it tends to increase turnout. For example, Swensen said 103,000 of the 130,000 Salt Lake County voters who requested mail ballots two years ago returned them. All-voting-by-mail counties estimate it sometimes has doubled voter turnout in early experiments. That is welcome because Utah's voter turnout has become one of the nation's worst in recent years. In the last mid-term election in 2010, only 36.2 percent of voting-age Utahns cast ballots- fifth lowest in the nation, according to studies by George Mason University. This year, nine counties are voting almost entirely by mail- Beaver, Cache, Davis, Duchesne, Garfield, Grand, San Juan, Sanpete and Sevier, said Justin Lee, deputy director of elections for Lt. Gov. Spencer Cox. Most also offer some in-person voting centers on Election Day. Lee said that only Duchesne previously conducted a general election entirely by mail, but most of the others held a by-mail primary this year or conducted some city elections by mail. Cache County Clerk Jill Zollinger said her county conducted all-by-mail voting in three precincts in this year's primary election, and turnout in them was twice as high as in other areas. Her county is mailing out 49,000 ballots on Monday.

Urban pioneer o Brian McKenzie, election manager for Davis County, said its by-mail primary this year also had turnout that was much higher than in neighboring counties, \·lith participation up to so percent of registered voters in some areas. "For a primary election, that just doesn't happen."

http://www .sltrib.com/csp/cms/sites/sltrib/pages/printerfriendly .csp?id=5 8481150 10/3/2014 Election-by-mail trend changing how Utahns vote and campaign I The Salt Lake Tribune Page 2 of2

He said it also allowed Davis County, the only urban county in Utah so far to adopt the practice, to release results much faster. That is because it processed votes as they were received in the mail over a month's time - although they were not tabulated until all polling places closed at 8 on election night. "At 8:02 on election night, we released those initial results," which did not include in-person votes cast that day and some ballots arriving in subsequent days' mail. "It was much earlier than we were able to do" previously.

McKenzie said Davis County is mailing ballots to its 135,000 registered voters on Monday, along with contact information for candidates on the ballot "to empower the voters, if they want to, to study the issues." More voting by mail is not the only big change in elections this year. Four counties will also conduct an experiment in allowing Election-Day voter registration: Davis, Weber, Salt Lake and Kane. Even though Davis is conducting by-mail voting, it allows same­ day registration at a few voting centers open on Election Day, McKenzie said. Swensen said Salt Lake County allowed Election-Day registration in this year's primary for the first time. She said most who sought it were first-time young voters ages 18-25. But she was surprised that many people older than 6o also took advantage of same-day registration - she figures because they had moved and never reregistered.

Deadlines • Elsewhere, Lee said the statewide deadline for registering by mail is Monday. But in-person registration at county clerk offices continues through Oct. 27, and online registration continues through Oct. 28. Voting early in-person is also an option in those counties not voting all-by-mail. That begins in county clerk offices (technically casting absentee ballots in person) on Monday. Early voting at a variety of other locations around counties is available Oct. 21-31. A list of locations statewide for early voting (and on Election Day) is available at vote.utah.gov. That website also allows online-voter registration, and has a guide with information about candidates and ballot initiatives. Swensen said she may ask the Salt Lake County Council to consider all-by-mail voting in 2016. Her office has conducted all-by-mail voting by contract for city elections in Cottonwood Heights and West Jordan, and turnout in last year's municipal elections more than doubled. Salt Lake County will offer other cities that option next year. Holladay is one of those exploring making the switch. "I still get a lot of people who just like going to their local polling locations" on Election Day, or even for early voting, Swensen said. "So right now, I like offering all those options, and let the voter choose."

General election twists • Nine of 29 counties will vote primarily by mail: Beaver, Cache, Davis, Duchesne, Garfield, Grand, San Juan, Sanpete and Sevier. • Four counties will offer election-day registration: Salt Lake, Davis, Weber and Kane. • In-person voting in county clerk offices begins Monday. Other in-person early voting is available Oct. 21-31. • Voter registration by mail ends Monday (for forms sent that day). In-person registration runs through Oct. 27, and online registration runs through Oct. 28.

© Copyright 2014 The Salt Lake Tribune. All rights reserved. This material may not be published, broadcast. rewritten or redistributed.

http ://www.sltrib.com/csp/cms/sites/sltrib/pages/printerfriendly .csp?id=5 8481150 10/3/2014