Changing Election Methods: How Does Mandated Vote-By-Mail Affect Individual Registrants?

Introduction

What is the impact of vote by mail on individual registrant turnout? That is a question election administrators and policymakers alike are grappling with across the country as more and more states consider moving to some form of vote-by-mail system. For now, Oregon remains alone in conducting all elections by mail. In Washington all but one of its counties vote by mail.¹ Montana allows vote by mail for municipal elections,² and Idaho³ and New Jersey⁴ have taken steps in that direction. California and Colorado already permit permanent no excuse absentee ballots, and twenty-nine states have provisions for no excuse absentee voting.⁵

Turnout rates from recent elections show that voters are taking advantage of laws that provide options for how they cast their ballot. For example, in November 2008 more than a third of the electorate voted early, up from 15% in 2000 and 22.5% in 2004 (McDonald 2008). In closely contested Colorado, Nevada, and New Mexico, half of all voters cast ballots before Election Day, and in Florida the early vote was 40%.⁶ Among early voting options, momentum for vote by mail in particular has been strong—California's May 2009 election, in which 62.19% of voters used this method, is one striking example.⁷ However, it is important to recognize that in all these instances voting by mail is optional. This means that the voter behavior captured in these turnout statistics is based on self-selected behavior. In other words, 62.19% of voters in California chose to cast their ballot by mail. What would happen if voting by mail became compulsory? How would changing the method of voting affect the behavior of the other 37.81% who did not choose to vote by mail? The answer is that we do not know, save for the oft-cited example of Oregon, a state without the challenges posed by population demographics, high density, or language diversity. That is the question we sought to answer in this study: what happens to individual registrants when their mode of voting is changed to require mail-only balloting?

We conducted the most comprehensive research to date on the impact of vote-by-mail systems on individual voter turnout. As Kousser and Mullin (2007) point out, an ideal research design would randomly assign a group of voters to vote by mail. Arguably, such a design occurs in California where any of 58 county election officials can assign voters to mandatory mail precincts in any election—within precinct size limitations that are addressed in the next section—effectively creating natural variation in the use of mail balloting. We collected individual-level data from five counties in California⁸ and then studied all mandatory mail voters in those counties. We examined the behavior of 97,381 individual voters across four elections from 2006 through 2008 to ascertain whether vote-by-mail increases individual voter participation. Our hypothesis is that being required to change one's mode of voting, what we refer to as being a “mandatory mail” voter, increases turnout. We test our hypothesis through a design that holds the electoral context constant and naturally assigns voters to the treatment, in-person or mandatory mail precinct, thereby allowing us to assess the impact of being required to vote by mail on turnout.

As mentioned, our individual panel data of mandatory mail voters spanned five California counties. Panel data is the best way to examine the individual behavior of voters, but such studies are rare due to the cost of acquiring validated voter records from local elections offices (Berinsky 2005:479).⁹ Using weighted least squares logit models for analysis of the validated records of all mandatory mail voters in the study counties, we estimated the impact of mandatory mail balloting across voter stratifications while removing the self-selection bias inherent to absentee voters. Individual level work using panel data has not been conducted since scholars (Berinsky et al. 2001) did so using Oregon data from 1994–1996, where the sample was 811 voters.

Our findings indicate that when a mandatory vote-by-mail system is implemented, the estimated odds of an individual voter voting decreases by 13.2%. However, election officials can play a role in ameliorating that effect in the number of communications they send to voters; we find four mailings critical to overcoming negative effects.

This article briefly recaps the recent research on vote by mail to date. We then turn to a discussion of our data and methods. We then discuss findings and implications.

Previous Research

Proponents of alternative voting methods argue that citizens will be more likely to participate in elections when they have more flexibility in casting their ballot. Opponents argue that broadening election parameters opens the door to fraud and may limit civic engagement (Thompson 2004). On the threat to civic engagement, Richey (2005) reports that mail ballot voters engage in more political discussion than in-person voters. With respect to turnout impacts specifically, scholars find both negative and positive effects from vote by mail. Recent work (Kousser and Mullin 2007, Richey 2008, Southwell 2009) indicates that vote by mail has some negative effect (2–3%) on turnout in statewide general elections but a stronger positive effect in local (8%) and special elections, and a large (10%) positive effect in national elections—but those turnout gains are short-lived (Giammo and Brox 2010). Other studies (Barreto et al. 2006, Gronke and Miller 2007, Hanmer and Traugott 2004, Karp and Banducci 2000, Southwell and Burchett 2000) find insignificant effects from vote by mail on turnout and on the composition of the electorate. Berinsky, Burns, and Traugott (2001) find no effect on the partisan composition of the electorate but do find vote-by-mail systems advantage resource-rich (i.e., older, educated, interested) members of the electorate.

In California, where legislation (e.g., AB1654) to expand mail ballot elections failed to garner sufficient support in the legislature, the opposition argued that mail ballot elections would decrease turnout, especially among low-income and minority voters; however, opponents acknowledged during testimony that “there is no definitive research on the impact of mail-only elections on minority voter turnout.”¹⁰

Kousser and Mullin (2007: 429) point out that much of the scholarship in the vote-by-mail subfield “cannot be used” in attempting to understand the impact on voters of the move to mail-only elections because it focuses on people who have chosen to vote absentee or by mail and are therefore a high propensity voter group. These researchers bring new methodological rigor to the field but acknowledge “it is individual factors that matter the most in explaining turnout,” (430) and their design does not account for individuals over time. Furthermore, their study did not follow the same voters across elections; even within the confines of one election, the treatment is observed on different voters thus resulting in a turnout comparison of different groups of voters.¹¹

Research Setting

While the body of research summarized here provides important information about the effect of various voting methods on aggregate turnout, it does not seek to, nor can it, address the impact that instituting compulsory vote-by-mail systems will have on individual voter behavior. Our research advances knowledge in this subfield through the use of a natural quasi-experiment that randomly assigns individuals to different election rules—an ideal design because it holds the election context constant. With panel data we can assess the impact of changing election systems on individuals because the mode of voting is changed. County voting records enable us to track the same voter across four elections when election officials in at least one of those four elections altered the mode of voting.

This is possible because California law¹² allows local election officials (LEOs) to designate any precinct with fewer than 250 voters a mandatory mail precinct.¹³ This precinct threshold gives the LEO flexibility in managing the myriad of elections with changing geographic boundaries that can occur in any given year in California and that determine what ballot a registrant must receive. Every election contains either a single item or multiple items that each registered voter is asked to make a choice on—these issues determine ballot type. Each item may be something that is voted on by everyone in the county or it may be something that pertains only to a specific area or district. Each district has legal boundaries that sometimes split down the middle of a street or between next-door-neighbors. Prior to an election the LEO in a county reviews what districts and precincts are affected by ballot type.

Every residence is assigned a “home precinct” number that is determined by all the voting districts the property resides in. Once the LEO has determined all of the home precincts that are within the affected districts, they “consolidate” all precincts that will vote on the same issues, and assign a “consolidated precinct” number. Factors that are considered when consolidating precincts include: 1) the number of registered voters at the time of consolidation, 2) the legal limitations for maximum size (number of voters) of a polling place, 3) any special legal considerations (like language requirements stipulated by federal law), 4) distance the voter must travel, 5) geographic accessibility (i.e., distance might be short but could involve crossing busy streets or passing other polling places, or could involve difficult mountain roads), 6) accessibility to the elderly and voters with disabilities, 7) history of consolidations from past elections (to maintain consistency of polling places whenever possible), and more recently 8) the cost of conducting an election (e.g., special elections where state reimbursement is not available). LEOs generally try to send the same voters to the same polling places in every election to maintain consistency in polling places as long as the ballot style is the same (i.e., ballots have the same candidates and measures). However, this is not always possible because of ballot variation; in those instances when it is not, a voter who voted in-person in a prior election may be assigned to vote by mail in the next election.¹⁴ Thus, the mode change is imposed by the LEO and is not chosen by the voter.

What would happen if all voters who did not initially choose to use the vote-by-mail option were then required to vote by mail? The literature suggests that convenience can reduce the cost of participation by making it easier for voters, thereby increasing the likelihood that they will cast a ballot. Voting by mail, early voting, and vote centers are all considered “convenience voting” mechanisms. Thus a mode change from in-person polling place voting to voting by mail could be expected to increase turnout. In addition, the repeated nature of the mode change experience over time, necessarily involving more contact with the election system in the form of mailings that mail-ballot voters receive from election administrators, might also increase the likelihood of casting a ballot. Direct mail has been shown to increase turnout by up to three percent (Gerber et al. 2008). Mail also serves the important functions of making voters aware and educating them about process, factors demonstrated to increase turnout (Wolfinger and Rosenstone 1980, Wolfinger et al. 2005).

Scholars have noted how the assignment of voters to mail-only precincts in California provides a unique research opportunity (Arceneaux et al. 2009, Kousser and Mullin 2007, Meredith and Malhotra 2008). In addition, such assignment does not sacrifice voter diversity for two reasons: first, because individuals do not self-select into precinct type; and second, though like persons may flock together in small residential locations, small all-mail precincts exist as a function of multi-jurisdiction ballot needs and are frequently nested in larger metropolitan areas. Aistrup (2004) noted that urban jurisdictions are generally more diverse than rural ones. Furthermore, due to changing ballot requirements, LEOs may redraw precinct boundaries for every election, so voting groups may not remain static or homogeneous and likely fluctuate.

Data and Methods

We obtained data from county election offices in five California counties. Demographic data shown in Table 1 indicates that the study counties are highly diverse.

The data for this project are the actual records of registered voters in the study counties.¹⁵Table 2 shows the number of voter records received per county and the number of records utilized in our analysis. Of these registered voters, we identified 110,459 as mandatory mail voters who were used in our analysis.

A common perception about vote-by-mail precincts is that they are located in rural areas. However, according to the U.S. Census Bureau definition, only Marin and Fresno counties are rural;¹⁶ the other three counties are urban. Alameda has a density of 2,069 persons per square mile, followed by San Mateo at 1,633 and Santa Clara at 1,401.¹⁷ Fully 75% of mandatory mail ballot precincts in this study were urban metropolitan area precincts in 2008. We coded the geographic density variable a “1” for urban or a “0” for rural based on the voter's address (specifically their city and zip code) and its corresponding location on Census density maps.

Furthermore, the number of mandatory mail precincts in a particular county was not skewed in a given election year but fairly evenly distributed across years. Table 3 shows the number of mail ballot precincts by county in a given election year. Overall, we have more than 700 mail ballot precincts per election year.

We tracked all registered voters assigned to a mandatory mail precinct at any time during the study period to ascertain whether or not they voted by mail in any of the four elections under examination: the June 2006 direct primary, November 2006 gubernatorial/midterm, June 2008 direct primary, and November 2008 presidential election. Across years these elections had large fluctuations in turnout levels. However, the variations are not due to anything distinct about the elections, but rather mirror the general pattern of low turnout in primaries and higher turnout in general elections, as well as the national pattern of low midterm and high presidential turnout. For example, in June 2006 turnout was 33.63%, compared with the November midterm turnout of 56.19% of registered voters; in 2008, the June direct primary had 28.22% turnout, compared with the November presidential election of that year when turnout of registered voters was 79.42%.¹⁸ We expect our results will follow this pattern.

In addition to election, voting method, and frequency of voting, the voter files contain age and party registration information. Other data necessary to the project are not contained in the validated vote records, so we appended additional data to every individual record based on the appropriate information (e.g., Assembly District based on address¹⁹) that applied to the individual voter.²⁰

Race and ethnicity are also not available in county voter files. We appended ethnicity to the individual voter records using a process developed by the U.S. Census (Word and Perkins 1996) and used by scholars (see for example Atkeson et al. 2009) for estimating values when the respondent does not supply data on a census form. Ethnicity is inferred through surname matching (approximately 20 percent of the Spanish surnamed population in the United States is concentrated in a dozen names, and 95 percent of individuals with the 12 most frequently occurring Spanish surnames self-identify as Hispanic). A similar method is used for matching Asian surnames (Lauderdale and Kestenbaum 2000), using the top 10 Chinese and top 10 Japanese surnames. Surname matching has been shown (Bullock et al. 2007) to underestimate Hispanic voters, and we assume the same effects could exist for Asian voters, so we weighted the ethnicity data based on the proportion of those populations per county.²¹

Political control variables include an indicator for the number of mailings a voter received from county election officials and measures for political competition. “Mailings” is a count of mailings sent out by each county to all voters (typically a Sample Ballot, the State Voter Guide, and the Official Ballot) and any additional pieces sent to mandatory mail voters, including reminder letters and postcards. Fresno and San Mateo counties send five mailers to registrants required to vote by mail. Generally, in addition to the three items listed above, voters get a letter notifying them of the change in mode of voting, then at a subsequent date they get a reminder letter which also lists the two closest polling places, should they opt to return their ballots in person. Marin and Santa Clara counties generally send out one notification letter, combining the mode change information and the voter's closest precincts, with the Sample Ballot, State Voter Guide, and Official Ballot. Alameda County incorporates mandatory mail balloting requirements into the standard three mailings (e.g., a page is added or a notification is printed on the page containing vote-by-mail information).

We hypothesize that these mailings will increase turnout because of both the repeated nature of the communications, as well as their official source. There is some basis for such a hypothesis. Kousser and Mullin (2007) presented anecdotal evidence from California that demonstrated the importance of a public information campaign when a new voting system is implemented. Table 4 shows the number of mailings sent to mandatory mail voters by each election official in the study counties. In three of the five counties, mandatory mail voters received more communications than absentee voters. We do not know the reason for the differing number of communications.

We created two variables to measure political competition: a variable based on the margin of victory per Assembly District expressed as the percentage difference between the two major party candidates or “1.00” where there was no competition, and a second variable reflecting the total amount of money raised in an Assembly race by all candidates in both the primary and general elections. Finally, we created an interactive variable to account for the number of elections covered. The interaction is the treatment over time: TRT × Time.

A logit model for repeated measures was fit to the voter data. The response modeled is the probability of a registrant voting, denoted as π(x). The model for the i^th treatment group (mandatory mail versus not mandatory mail) and the g^th election (November 2008, June 2008, November 2006 versus June 2006) is

where β′x are any additional covariates being controlled for in the model, γ_ig represents the interaction between the i^th treatment and the g^th election, and ν_g represents the number of years between the g^th election and the first election. The research design for this project allows us to test whether voter turnout is affected by the switch from in-person precinct voting to mandatory vote by mail and vice versa. We also investigate whether the effect of voting by mail varies across election type: primary, general, mid-term, and presidential. Furthermore, the model allows us to not only obtain predicted probabilities and their asymptotic standard errors for each treatment-election combination, but also the estimated odds of voter turnout for various treatments, elections, and values of the covariates incorporated into the model. Yang, Goldstein, and Heath (2000) worked with a similar model using panel data from constituencies to model voting over three elections.

Results and Findings

Individual voting history records were collected for registered voters from Alameda, Fresno, Marin, San Mateo, and Santa Clara counties. All voters who were permanent absentee voters were removed from the data set. We also removed UOCAVA and military voters from the sample.²² Of the remaining voters, only those who were required to vote by mail at any time over the four elections were kept for the analysis. These remaining voters were monitored during June 2006, November 2006, June 2008, and November 2008. The dependent variable of interest was whether or not the mail voters returned their ballot by mail—if the voters returned their ballot in person at a polling place, they were dropped from the sample.²³ Independent variables included those discussed in the previous section. With these panel data there is no attempt at creating a randomized sample as all non-permanent absentee mail voters had the opportunity to be included in the study.²⁴

The results are presented in Table 5. Only three counties, Alameda, San Mateo, and Santa Clara, are included due to low cell counts for Fresno and Marin counties. (See Appendix B, Tables A–E for explanation.) The metric for interpreting these effects with the model used is the odds ratio, which is calculated as follows. Let π₁ be the probability that a mandatory mail registrant actually votes and 1-π₁ be the probability that they do not vote. Similarly, let π₂ be the probability that a polling place registrant actually votes and 1-π₂ is the probability that they do not vote. The odds ratio of a mandatory mail registrant voting versus a polling place registrant voting is (π₁/1-π₁)/(π₂/1-π₂). All odds are reported at the 95% confidence level.

The estimated odds of a mandatory mail registrant voting, given a fixed point of time (i.e., year equals any fixed value), election characteristics, and voter demographics, is 0.8680²⁵ times the odds of a registrant who is not required to vote by mail voting in an election, equating to a 13.2% decrease in the odds of voting (1.0–.8680). For each additional year ceteris paribus the estimated odds of a mandatory mail registrant voting is 1.005²⁶ times the odds of that same registrant voting the previous year. Since 1 falls in the 95% confidence interval, the effect of voting by mail on voter turnout over time (for each additional year) is deemed to be insignificant. Of greater interest is the estimated probability of this effect over time when taking into consideration a specific election's effect on voter turnout. Moving from primary to general election in a given year increases the probability of turnout. The model estimates that a mandatory mail registrant was 5.60% more likely to vote in November 2006 compared to June 2006,²⁷ and 34% more likely to vote in the November 2008 versus June 2008. Not surprisingly, turnout declines between the general and primary elections across years; we estimate that decrease was 6.26% from November 2006 to June 2008. These estimated probabilities are mainly due to the impact of general elections on turnout, and are not a function of repeated exposure to voting by mail.

We now turn to the impact of demographics. To interpret the estimates in Table 5, one has to fix the level of treatment and type of election. We have set these to be a mail-only precinct and voting in the same election. These results are summarized in Table 6.

The model predicts that the odds of an urban registrant voting is 0.4992 times the odds of a rural registrant voting. Under such conditions, the odds range is 0.4720 to 0.5280, a 47.2% to 52.8% decrease in the odds of voting for urban registrants. Following the same procedure, Hispanic voters have a 19.8% to 34.1% lower odds of voting and Asian voters have a 23.8% to 36.3% lower odds. Finally, Democrats show a 2.89% to 9.19% increase in the odds of voting compared to Republicans. This is a somewhat anomalous finding in that Republican usage of absentee voting is generally higher than Democratic usage. However, it is possible that unusually good Democratic years can enhance turnout vis-à-vis Republicans. Arguably this was the case in 2006 when Democrats gained the majority in the House of Representatives for the first time in 12 years, and in 2008 when Democrats “were building on furious registration gains” in California.²⁸ When we calculate the impact on turnout, we find that this equates to an increase in Democratic voter turnout of 0.18% to 1.45% versus Republican voter turnout in mandatory mail precincts.²⁹ See Table 7.

Finally, we turn to the effects of the political environment. Communications from the Registrar of Voters are always written. As counties differ in the content and timing of their mailings, we can only generalize based on the number of mailings that a voter received. Ceteris paribus, each additional communication improved the odds of voting by 3.94%. Perhaps more important, communications could overcome the negative effects of forced vote by mail. A voter in a mandatory vote-by-mail precinct who received five communications was 4.05% more likely to vote than a precinct place voter who received no mailings.³⁰ Finally, and not surprisingly, the level of political competition has a substantial impact on turnout.

Discussion

We sought to examine whether being assigned to vote by mail increased a registrant's likelihood of voting. Focusing exclusively on registrants who had their voting method changed not by choice but by the county registrar and using turnout data at the individual level, we demonstrated that this was not the case. The required change to this form of voting decreased the likelihood an individual registrant would vote, and this was true even after the first time a voter was assigned to a mandatory mail precinct as repeated exposure had no significant positive effect. However, that said, election officials do have a role to play in mitigating and possibly reversing the negative effect switching to mail-only systems has through repeated communication with registrants. Our results show that increasing the number of communications a registrant receives from election administrators about a prospective voting method change has a positive payoff by enhancing the likelihood of turnout. Also, given the language diversity in California,³¹ the requirements of Section 203 of the Voting Rights Act in four of the study counties,³² and the strong negative impact of mandatory mail on Hispanic and Asian voters, it is possible that multilingual mailings might have an impact with these populations.

The larger electoral context was found to play an important role in establishing incentives and disincentives to vote. The implication is that administrators are unlikely to turn away significant proportions of voters in statewide or national elections because of a change from in-person polling place to all-mail balloting; likewise, neither is the change in the way a registrant casts a ballot likely to overcome the electoral context and increase turnout. Put simply, mandatory mail did not have a larger effect on the voter than did the type of election. Finally, any partisan advantage we found was less than a percent and a half, whereas big shifts in campaign activity can change the partisan complexion of the electorate by twice as much (McGhee and Sides 2010).

We need to highlight some potential caveats. First, focusing on northern California may limit the generalizability of these findings. Nevertheless, testing the effects of shifting registrants from polling place to vote-by-mail precincts requires a focus on regions containing significant numbers of such precincts. The fairly broad distribution of these types of precincts across the study counties makes this environment well suited to our purposes. Second, we have little to say about the impact on African Americans since we are not able to identify black registrants.

The negative effect that the implementation of compulsory vote-by-mail systems has on an individual registrant's likelihood to vote is not overcome with time, experience, or convenience but rather with continued contact and information provided by the LEO. The potential advantages of this increase in frequency and content of communication are profound.

Appendix A

The variable Communications lists the number of mailings sent to voters by the election official in each county. This includes the basic information mailed to all voters: 1) Sample Ballot, 2) the State Election Guide, and 3) the Official Ballot. These three items are referred to as “Standard” below. Some counties also sent out letters and postcards reminding voters to return their mail ballots. A county-by-county summary follows:

Alameda:

• - Polling place voters and mandatory mail voters get three (3) Standard items.

Fresno:

• - Polling place voters gets four (4)—including Standard and a “pre-notice” flyer with information on voting by mail.

• - Mandatory mail voters get five (5)—including the Standard plus a “pre-notice” flyer as well as a letter listing their closest polling places.

Marin:

• - Polling place voters gets three (3) Standard items only.

• - Mandatory mail voters get four (4)—including Standard and a letter listing their closest polling places.

San Mateo:

• - Polling place voters gets three (3) Standard items only.

• - Mandatory mail voters get five (5)—including Standard and a “pre-notice” flyer with information on voting by mail and a letter listing their two closest polling places.

Santa Clara:

• - Polling place voters gets four (4)—including Standard and a “pre-notice,” a reminder of an upcoming election, information on when to expect the ballot, and how to return it.

• - Mandatory mail voters get four (4)—including Standard and a letter listing their two closest polling places.

Appendix B

Tables A through E show the number of voters that fell into each treatment-election combination. The numbers come from the actual panel data for each county before the logistic regression models were run. These results include voters with missing and inaccurate age data in the data sets; such values caused problems in the metafile merges so these voters were ultimately excluded from the model. Because these cases amounted to a small number (<5%) of the cases in the combined dataset and were randomly excluded, we do not sacrifice analytic capability.

Looking at Table 5c, the odds of a mail ballot registrant voting in June 2008 is

times the odds of a polling place registrant voting in June 2008, where ∞ is an undefined value for the odds ratio. The odds ratio for Marin County in 2009 would have the same exact value. Even small “cell” counts may eventually lead to faulty interpretations over the entire analysis. Looking at Table 5b, the odds of a mail ballot registrant in Fresno County voting in November 2006 is

times the odds of a polling place registrant voting in November 2006. Because of these zero values in Tables 5b and 5c, Fresno and Marin counties are excluded from the logistic regression model.

Table A. Count Data for Alameda County

Precinct	Vote	June 2006	November 06	June 2008	November 08
Mandatory Mail	Yes	645	1474	788	3684
	No	1128	945	2109	165
Polling Place	Yes	11841	19397	9832	31851
	No	23776	15574	24661	1690

Table B. Count Data for Fresno County

Precinct	Vote	June 2006	November 06	June 2008	November 08
Mandatory Mail	Yes	1033	2199	1305	6474
	No	353	3	0	2357
Polling Place	Yes	92	116	184	144
	No	9028	8188	9017	1531

Table C. Count Data for Marin County

Precinct	Vote	June 2006	November 06	June 2008	November 08
Mandatory Mail	Yes	56	185	754	983
	No	219	350	1307	610
Polling Place	Yes	121	73	0	0
	No	2176	1964	511	979

Table D. Count Data for San Mateo County

Precinct	Vote	June 2006	November 06	June 2008	November 08
Mandatory Mail	Yes	577	963	4375	11109
	No	797	478	7561	2659
Polling Place	Yes	2451	3551	217	83
	No	10238	9071	1910	212

Table E. Count Data for Santa Clara County

Precinct	Vote	June 2006	November 06	June 2008	November 08
Mandatory Mail	Yes	519	1798	4778	20542
	No	4872	7870	29366	13237
Regular	Yes	7653	12482	1294	7263
	No	32884	23778	10490	4886

¹ “KingCo's all-mail voting gets a thumbs-up.”<http://blogs.secstate.wa.gov/FromOurCorner/index.php/2009/02/kingcos-all-mail-voting-gets-a-thumbs-up/>. Retrieved June 24, 2009.

² O'Connell, Sue. 2007. “HJR 46: Revision of Election Laws. Your Ballot's in the Mail: An Overview of Vote-by-Mail Issues.” <http://leg.mt.gov/content/committees/interim/2007_2008/st_admin_vet_affairs/assigned_studies/hj46overview11202007.pdf>. Retrieved June 24, 2009.

³ “Initiative would let Idaho voters sign up for permanent absentee ballots.” The Spokesman-Review. <http://www.spokesman.com/blogs/boise/2009/jun/18/initiative-would-let-idaho-voters-sign-permanent-absentee-ballots/>. Retrieved June 24, 2009.

⁴ Paul Gronke, June 19, 2009, “Changes in no-excuse absentee coming in NJ?”<http://electionupdates.caltech.edu/>. Retrieved June 24, 2009.

⁵ The Canvass: States and Election Reform. <http://www.ncsl.org/default.aspx?tabid=16604>. Retrieved July 14, 2010.

⁶ “Early voting starts today in some states.” <http://www.usatoday.com/news/politics/election2008/2008-09-21-early-voting_n.htm>. Retrieved January 31, 2009.

⁷ Voter Participation Statistics by County. <http://www.sos.ca.gov/elections/sov/2009-special/3-voter-participation-stats-ss09.pdf>. Retrieved June 29, 2009.

⁸ We selected counties in Northern California because of geographic proximity to the researchers, project resource limitations, and access to diverse voter populations. Certainly, researchers would do well to study more than the five counties we covered.

⁹ We are referring to records containing the voter's voting history as maintained in the jurisdiction of registration—in the case of California, the registering jurisdiction is the county. Per the Help America Vote Act (HAVA), voter registration must be verified. To do this, files go from the county office to the Secretary of State's (SOS) office to be matched against the California Department of Motor Vehicle files to verify that the voter's driver's license matches in order to validate the registration of the voter. Local election officials (LEOs) send new registrant files to the SOS every day during an election season. The SOS matches the data and sends back the file. Those people whose registration data does not match the data in the SOS databases get put in the “pending” file and don't become active until their information is verified via a mailing to them. Only voters whose registration has been validated in this way are eligible to vote.

¹⁰ California Assembly Bill 1654, page 7, from <http://www.leginfo.ca.gov/pub/07-08/bill/asm/ab_1651-1700/ab_1654_cfa_20070416_111314_asm_comm.html>. Retrieved Oct. 9, 2007.

¹¹ Mean turnout values of mail precincts (the treatment) are compared with mean turnout of matched in-person precincts (non-treatment). Matching mail and in-person precincts based on covariates might reduce some voter differences; however, the voters are still not the same voters.

¹² California Elections Code Section 3005:

Whenever, on the 88th day before the election, there are 250 or less persons registered to vote in any precinct, the elections official may furnish each voter with a vote by mail ballot along with a statement that there will be no polling place for the election. The elections official shall also notify each voter of the location of the two nearest polling places in the event the voter chooses to return the ballot on election day. The voter shall not be required to file an application for the vote by mail ballot and the ballot shall be sent as soon as the ballots are available. No precinct shall be divided in order to conform to this section.

History of Cal Elec Code § 3005: Added Stats 1994 ch 920 § 2 (SB 1547). Amended Stats 2007 ch 508 § 17 (AB 1243), effective Jan. 1, 2008. [2007 Amendment: Amended the first paragraph by substituting (1) “a vote by mail” for “an absentee” in first sentence; and (2) “vote by mail” for “absentee” in the last sentence.] Historical Derivation: (a) Former § 1005, as added Stats 1976 ch 1275 § 18, amended Stats 1978 ch 836 § 1, Stats 1992 ch 527 § 1. (b) Former § 14620.5, as added Stats 1965 ch 2004 § 1, amended Stats 1967 ch 186 § 1, Stats 1970 ch 379 § 3. Source: Deering's California Code Annotated.

¹³ The maximum allowed by law is 1,000 registrants (Section 12223 of the California Elections Code).

¹⁴ Registrants assigned to mandatory mail precincts may still vote in person if they wish in some jurisdictions, and they may return their ballots to various local government offices. Such information is provided in mail correspondence sent by the LEO.

¹⁵ It is worth noting that the validated voter data files received from county election officials were not uniform or consistent. There is no uniform standard, either across the nation or within California, for gathering, storing, updating, or validating this information. While HAVA legislation required funded states to update their voter registration databases, that mandate does not extend to validated voter records or post-election files. This is an important distinction for both researchers and policymakers as it poses a challenge for data analysis, cross-jurisdiction comparisons, and other assessments.

¹⁶ U.S. Census defines rural as having a population density less than 1,000 people per square mile. Source: <http://www.census.gov/geo/www/ua/ua_2k.html. Retrieved Jan. 6, 2009.

¹⁷ Source: RAND California: Wed. Oct. 21 09:26:12 2009.

¹⁸ The 2008 presidential election did have historic levels of voter turnout across the nation. However, the increase must be compared to the previous presidential election results, not the 2006 midterm. In California in the 2004 presidential election, 76% of registered voters cast a ballot. All voter turnout statistics are from the California Secretary of State.

¹⁹ Assembly District maps. Source: The California Voter Foundation. <http://www.calvoter.org>. Retrieved Jan. 6, 200c9.

²⁰ Education and income are variables not in the registrar's database. Obtaining a registrant's information would require costly surveying. Controlling for these factors has not been done at the individual level with data of this type. Other scholars (Barreto et al. 2004) who conducted similar individual-record analyses with large voter databases resorted to aggregate measures (e.g., percent of residents with a college education and median household income by zip code) in order to control for these effects. We chose not to do this as such results would suffer from problems of ecological inference, masking crucial individual effects in small VBM precincts.

²¹ Kousser and Mullin (2007) calculate racial (including Hispanic) composition as a percentage of a precinct's voting age population (VAP). While VAP is a more restrictive measure than proportion of population, the American Community Survey (ACS) of the Census does not combine age and ethnic classifications in data; only disaggregated data is available. Thus, we recognize that our weighting is somewhat inaccurate as it includes 0 to 17 year-olds who would not be part of VAP.

²² The Uniformed and Overseas Citizens Absentee Voting Act (UOCAVA) of 1986 stipulates that these citizens are allowed to vote absentee. As such, they would not have the option to vote in person at a polling place, nor be assigned to a mandatory mail ballot precinct. Therefore, these voters would not qualify according to the parameters of our study.

²³ The reason for this is that the “treatment” is voting by mail (i.e., returning the ballot by mail not in-person). However, we did run the model retaining those mandatory mail voters who delivered their ballot to a polling place; results of that model indicate that this treatment effect is not significantly different from the treatment effect reported in Table 6 (Z=0.4898, p-value=0.6243).

²⁴ “Opportunity” means that we strove to include every possible case. However, some cases dropped out of the model due to missing values or corrupt data. This explains the difference in our starting sample of 110,459 versus our ending sample of 97,381. Arguably, excluded cases are random.

²⁵ Exp(-0.1416)=0.8680.

²⁶ Exp(0.0047±1.96×0.0112)=0.9829 to 1.0270.

²⁷ This can be computed by taking the difference of: exp (-3.3607 - 0.1416+1.1275+0.0047×5/12)/(1+exp(-3.3607 - 0.1416+1.1275+0.0047×5/12)) - exp(-3.3607 - 0.1416)/(1+exp(-3.3607 - 0.1416).

²⁸ Decker, Cathleen. 2008. “California gives GOP the blues.” Los Angeles Times. <http://articles.latimes.com/2008/nov/06/local/me-calif6>. Retrieved Jan. 17, 2011.

²⁹ Predicted probabilities calculated from a logistic regression model:

where all of the parameters are defined in equation (1).

³⁰ The estimated odds of a mandatory mail registrant who received a communication voting is 0.9021 times the odds of a polling place voter who received no communications voting. Under these fixed conditions, the odds of a mandatory mail registrant who received a communication voting is 0.8640 to 0.9419 times the odds of a polling place registrant who received no communications voting. Looking at the general form for this estimated odds ratio, exp(-0.1416+0.0386x), where x represents the number of communications, the estimated odds are first greater than 1 when the number of communications sent out is four. In other words, for a given county, election characteristics, and voter demographics, the estimated odds of a mandatory mail registrant who received four communications voting is 1.0129 times the odds of a polling place voter who received no communications voting. Under these fixed conditions, the odds of a mandatory mail registrant who received four communications voting is 0.9736 to 1.0537 times the odds of a polling place voter who received no communications voting. Since 1 falls inside of this confidence interval, both types of voters are equally likely to vote. Five communications ensures that both limits will be greater than 1.

³¹ 43% of the Latino population in the state is limited-English proficient (LEP), and 39% of the Asian population is LEP (Avila et al. 2006).

³² There are 25 counties in California subject to Section 203 that require election materials in a language other than English; of those, four are study counties, including: Alameda (Chinese, Spanish), Fresno (Spanish), San Mateo (Chinese, Spanish), and Santa Clara (Spanish, Chinese, Filipino, Vietnamese) (Avila et al. 2006).