Motor Voter Registration Programs by nyut545e2


									Motor Voter
                                Motor Voter

Robert S. Montjoy
Director, MPA Program
1224 Haley Center
Auburn University, Alabama 36849

Managed and Edited by:
William C. Kimberling
National Clearinghouse on Election Administration

Published by:
National Clearinghouse on Election Administration
Federal Election Commission
Washington, D.C. 20463

September 1992
Introduction by the

   This report is another in the series on Inno-        We welcome you comments on these reports as
vations in Election Administration being pub-        well as any suggestions you may have for addi-
lished by the FEC's National Clearinghouse on        tional topics. You may mail these to us at:
Election Administration.
   The purpose of this series is to acquaint State
and local election officials with innovative elec-           The National Clearinghouse on
tion procedures and technologies that have been                 Election Administration
successfully implemented by their colleagues                 Federal Election Commission
around the country.                                          999 E. Street, N.W.
                                                             Washington, D.C. 20463
  Our reports on these innovations do not
necessarily constitute an endorsement by             or else call us
the Federal Election Commission either of                   toll free on 800/424-9530 or
the procedures described or of the vendors                  direct on 202/219-3670.
or suppliers that might be listed within the
report. Moreover, the views and opinions
expressed in these reports are those of the
author and are not necessarily shared by
the Federal Election Commission or any
division thereof.
                                                       Motor Voter

Introduction                                           National Clearinghouse of Elections
                                                       Administration's Technical Report Series. All
   "Motor Voter" is the popular name for a system      population figures and estimates of voter regis-
in which people can apply for voter registration in    tration are taken from the U.S. Census Bureau,
the same offices where they obtain driver's li-        Current Population Reports Series P-20. Infor-
censes or, in a few cases, where they register         mation on licensed drivers and driver licensing
vehicles. About half of the states now use some        requirements came from Federal Highway Ad-
version of this system and legislation is pending      ministration Driver License series.
in several others.
   This report is designed for the benefit of juris-   Motor Voter Systems
dictions that are considering adoption or modifi-
                                                          This report details many of the variations in
cation. For this reason it draws upon examples
                                                       motor voter systems currently being used. It
from many jurisdictions to generalize about the
                                                       begins, however, with a basic model having a few
effects of different program designs. It does not
                                                       common steps. (1) People waiting in an office to
attempt to describe the complete or current op-        get driver's licenses learn that they may also
eration of any single jurisdiction.                    apply for voter registration or change their reg-
  The implementation of a motor voter system           istration information there. (2) Those who wish
requires the interaction of officials from two         to do so complete the necessary forms, with or
agencies, one having responsibility for voters and     without assistance from the staff in the office. (3)
the other having responsibility for motor vehicle      The completed form is sent, directly or indirectly,
                                                       to the registration office of the locale where the
operators, who have no reason to be familiar with
                                                       applicant resides. (4) Registration officials act on
each other's basic procedures. The report at-
                                                       the application more or less as they would on any
tempts to address both audiences and so includes
                                                       other application.
some material that will seem basic and unneces-
sary to each.                                             Motor voter systems are often classified in
                                                       three categories - passive, active, and com-
   The primary data sources used in this report        bined - based upon variations of the first two
are written surveys, telephone interviews, and         steps described above. Passive systems place
site visits with voter registration and driver li-     voter registration forms in the driver's license
censing personnel in the states and a sample of        office and rely on the voter's initiative to complete
local jurisdictions that use motor voter. Informa-     them. The forms are separate from the driver's
tion on registration procedures and administra-        license application and can be the same form that
tive structure in all of the states came from the      is used in other registration systems within that
jurisdiction. Active systems also use separate          to the regular voter registration authorities, usu-
registration forms but provide a proactive means        ally an office in the county or township where the
of notifying the voter. This may be a question          prospective voter resides, and there they are
printed on the driver's license application or a        processed just as other applications are, with a
requirement that a clerk ask the applicant if he or     few procedural adjustments in some cases.
she wishes to register.
   Combined systems place the driver's license             Second, motor voter does not require an inte-
and voter registration applications on the same         gration or even an interface of computer systems
form, often sharing information to reduce redun-        between the motor vehicle and election offices.
dancy. Notification of the voter is automatic; he       While such interaction is technically possible,
or she encounters the voter registration opportu-       only one jurisdiction studied for this report has
nity in the process of completing the form.             attempted it, the State of Washington. Of course,
                                                        the desirability of a computer link between driver
   Several jurisdictions have adopted acomputer-        licensing and voter registration systems may
assisted alternative to the combined form. A            depend, in part, upon unique requirements of
clerk collects the information necessary for a          each state's election system. It is not, however, an
driver's license application and enters it into a       inherent part of a motor voter system.
computer. The computer then prints out a voter
registration application for the applicant to re-          The ways in which a motor voter system can be
view, modify as necessary, and sign. This system        implemented depend in large part upon the legal
is like the combined system in that the applicant       and organizational environment in which it is
does not have to write the same information on          placed. Motor voter is a marriage of two distinct
two different forms. It is like the active system in    systems: the driver's license system and the voter
that the clerk typically asks the voter if he or she    registration system. Each has its own personnel,
wishes to register. This system is sufficiently         procedures, legal constraints, organizational cul-
distinct to merit separate discussion in this report.   ture, habits, and all the other characteristics
    Thus we have four basic systems:                    which build up over time and which make each
•   passive                                             organization different. Joining certain opera-
•   active                                              tions of the two will require at least one organi-
•   combined                                            zation to change procedures. The amount of
•    computer-assisted.                                 change necessary, the willingness of the person-
                                                        nel from different agencies to cooperate, and the
   Each system has variations and each distrib-         amount of outside resources or pressure available
utes the tasks in the registration process differ-      to facilitate the transition will all affect the suc-
ently among the applicants, motor vehicle agency        cess of the system.
staff, state election officials, and local election
   That, essentially, is what motor voter is. Be-       Voter Registration Systems
cause of the debate that has surrounded propos-            The task of voter registration is to maintain a
als for these systems in many jurisdictions, it is      current list of persons who are eligible to vote in
important to recognize what motor voter is not.         a specific jurisdiction or electoral district by reason
   First, and most importantly, no system pro-          of citizenship, residence, age, and the absence of
vides for an automatic registration at the driver's     any disqualifying conditions. In the United States
license bureau and no system allows people who          the primary responsibility for initiating registra-
could not otherwise vote to do so on the basis of a     tion and keeping it current lies with the voter, but
driver's license alone. The applications are sent       registration officials use a number of means to
ease the burden of registration and to collect           ent parties. Some registration systems are de-
information for voter file maintenance. Motor            centralized to townships and/or municipalities,
voter serves both functions of voter registration        as is the case in some New England and
and voter file maintenance.                              midwestern states. These offices can be very
                                                         independent of each other, even to the point of
   The organization and procedures for voter             designing their own registration forms. The
registration vary from state to state. They include      authority that state election officials can wield
a few common elements. (1) A voter obtains and           over the registration process varies widely, but it
completes an application. (2) The voter delivers         is generally less than that of their counterparts in
or sends the application to an election official or      the state driver's license systems.
board. (3) The board reviews the application for
completeness and conformity to state law. (4)               The degree of decentralization in a voter regis-
The board registers the names of successful ap-          tration system is important because motor voter
plicants and notifies those who were rejected. (5)       systems require coordination among registrars
The board maintains the registration records and         as well as between registrars and driver's license
provides them to other officials as needed. The          personnel. Coordination is necessary in the de-
most common usage is for checking voter eligibility      sign of forms and in the transfer of applications to
at the polling places on election day. (6) Between       the appropriate jurisdictions. Both of these steps
elections registrars try to keep their files current     are discussed later in this report. Coordination is
by recording changes of name or address and by           also important, especially important, in the plan-
removing the names of persons who have died,             ning of the many features of the program, includ-
moved away, or otherwise becoe ineligible to vote        ing legislation, administrative procedures, train-
in that jurisdiction. Motor voter is directly in-        ing, monitoring, and evaluation. Thus, it is
volved in elements 1,2, and 6.                           essential to recognize who the players are in any
                                                         state system and to ensure that their views and
   Jurisdictional boundaries are much more im-           existing procedures and constraints are taken
portant in voter registration systems than they          into account.
are in driver's license systems. A driver's license
is valid for the entire state and, temporarily, in all
other states. Voter registration is valid only in        Driver Licensing Systems
the electoral district where the voter resides. The      Except in the District of Columbia, the licensing
reason, of course, is that representation in the         of drivers is a state function, although various
United States is based upon geographic districts.        local governments and even private agencies may
Where one lives determines where one votes.              also be involved in the process. The basic steps
   Because of its historical association with place,     appear to be common to all systems. (1) New
voter registration has almost always been a              applicants go to an office to apply for a driver's
function of local government. (Alaska is an ex-          license. Some applicants may have to visit two
ception, using a statewide system of four regional       offices, depending upon the state and whether
offices and 3,000 appointed registrars.) The most        they require a driving test. (2) They are tested.
common form of organization is based on the              (3) Each applicant takes her or his test results
county. The registration function may be housed          and completed application form to a member of
in the office of the county clerk, in a separate         the office staff, who reviews it and, usually, en-
election office, or even in a separate voter regis-      ters the relevant information into a computer on
tration office. Many of these offices are headed by      the spot. The computer is likely to be a local
elected officials. In other cases the registration       terminal for a statewide system. (4) A staff
officials are appointed, usually through some            member photographs the applicant. (5) The
procedure to guarantee representation of differ-         photograph will be laminated to a card which
contains the applicant's name and other personal         reported to have copied Michigan's system with-
information. If this step is not completed in the        out any adjustments for the different authority
office, the applicant receives a temporary card          structures and to have suffered a number of
and the regular card will be mailed later. (6) The       coordination problems as a result.
applicant receives the card and leaves the office.

   The driver licensing system is run by the state
and designed to be accessible statewide. This fact
                                                         Points of Comparison between
has several consequences for motor voter. Be-            Voter Registration and
cause the system is statewide, changes in official       Driver Licensing Systems
procedure will normally have to be statewide,
especially any changes involving the central com-          As the preceding discussion suggests, voter
puter. Thus, the scope of any implementation             registration and driver licensing systems are
effort will be large. On the other hand, there will      similar in some ways and different in others.
be a central state office and a hierarchy which can      Since motor voter requires a joining of these two
set policy and facilitate implementation in local        systems, a systematic comparison on key points
offices. In fact, the state agency that licenses         may be helpful.
drivers will normally have more authority over
its field offices than the state's chief election or     • Function. The systems perform similar func-
registration officer will have over local registrars.    tions. Each establishes applicants' eligibility to
                                                         perform officially sanctioned acts (voting or driv-
   Another, and equally important, implication           ing). Each collects and stores data on individuals
of the system is that the jurisdiction of the            in order to certify that eligibility as necessary. It
driver's license offices is statewide. Applicants        is this similarity that creates the prospect of
are not limited to offices in their city, county, or     using part of the driver licensing system (infor-
township of residence. There are cities, town-           mation collection) in the voter registration process.
ships, and even counties that have no full-time          And because they perform these functions for
driver's license office. As a result, the jurisdic-      large numbers of people, the two systems share
tion of most driver's license offices will not           another feature: the effect of anything that
be the same as the jurisdiction of voter reg-            changes the routine processing of individual cases
istration offices. This situation complicates            is multiplied many times over. Small changes
the transfer of applications between offices.            can have big consequences.
Different states solve this problem in different
ways. The major approaches are discussed under           • Jurisdiction. As noted above, the systems
the heading "Transfer Responsibilities" below.           typically differ with regard to jurisdiction. Driver
                                                         licensing is performed by a state agency with
   Statewide organization can be most facilita-          district offices or, as in the case of Ohio, by private
tive of motor voter implementation in those few          contractors working for a state agency. The
states where the chief election officer is in charge     organization is hierarchical; a state official is in
of driver's licenses as well. In fact, the first motor   charge. Voter registration is usually a county or
voter system was initiated by the Secretary of           township function. Local registrars are often
State in Michigan where this is the case. Having         elected, and the authority of state election offic-
a common authority over both functions solves a          ers varies considerably from state to state.
lot of coordination problems. The most important
point for most states, however, is that they will        • Identification of Applicants. Every state
lack that common authority and must provide              requires some positive identification before the
other means of coordination. At least one state is       applicant can receive her or his first license to
drive in that jurisdiction. (FHWA 1992) In con-         portant point is that there is a large difference
trast, only fifteen states require positive identifi-   between the rates. Approximately one fourth of
cation by all applicants for voter registration.        the age-eligible population is licensed to drive but
(Kimberling 1990)                                       not registered to vote. This is the target for motor
                                                        voter programs.
• Information Requirements. The ability to
share information is an important consideration            Second, the patterns are different among the
in combined-form and computer-assisted systems.         three groups. Except for an anomalous bump in
All driver licensing and voter registration agen-       1974, the driver's license line follows a steady
cies in the United States require the name and          path of gradual increase. Registration and voting
                                                        rates, on the other hand, traced similar patterns
the age or birth date of the applicant. All require
                                                        of decline during the time frame covered by this
an address, as well, but driver licensing agencies
                                                        chart. The largest drop came between 1968 and
will typically accept a mailing address while
                                                        1974. The patterns oscillated around a nearly
voter registration agencies must have a resi-
                                                        constant rate from 1974 through 1988 and then
dence. The following information items are re-
                                                        dropped slightly in 1990. Both rates reflect the
quested on voter registration applications in at
                                                        four-year cycle of participation associated with
least twenty jurisdictions (states or the District of
                                                        presidential elections, but the amplitude is greater
Columbia) and are not normally collected by
                                                        for voting rates than for registration rates. This
driver licensing systems: place of birth, political     difference reflects the facts that voting requires a
party, citizenship status, and place of prior regis-    positive act at each election but names stay on
tration. (Kimberling 1990, FHWA 1992)                   registration lists until they are purged. Never-
                                                        theless, registration rates are affected by the
• Renewals. Unlike voter registrations, driver's
                                                        appeal of particular elections. This fact compli-
licenses are issued for specified terms, which
                                                        cates any attempt to measure the effects of differ-
vary with the type of license and individual ages
                                                        ent registration systems. One has to take into
and driving records. Forty-three states and the
                                                        account the effects of different elections.
District of Columbia have maximum terms of
four years or less. The remaining seven states             Driver licensing and voter registration rates
have maximum terms of five years. At a mini-            vary by age group, as well. Figure 2 shows that
mum, the renewal process requires the driver to         both rates start low and rise steadily with in-
review the information on the license and attest        creasing age. Yet the driver licensing rate starts
to its correctness by signature. All but eleven         higher and peaks at an earlier age. Among the 18-
states mail renewal notices to drivers. In order to     19 age group 80 percent have driver's licenses but
spread the workload, most states schedule re-           only 30 percent are registered to vote. The driver's
newals on or around the applicant's birthday. A         license rate surpasses 90 percent at the 20- 24
few use the issuance date. (FHWA 1992)                  age, remains above that mark until the 55-59 age
                                                        group, and then declines to 67 percent for the group
• Populations Served. Figure 1 displays the             above age 70. The voter registration rate is only
relative size of three groups - persons licensed to     43 percent for the 20-24 age group, but it contin-
drive, persons registered to vote, and persons          ues upward to 78 percent for the 65-69 age group
voting - all as percentages of the voting age           and declines only slightly to 75 percent for those
population. This graph illustrates two important        over age 70. There are several possible expla-
points. First, more people have driver's licenses       nations for this pattern, including the different
than are registered or vote. In 1990, almost 90         experiences of the generations as they reached
percent of the voting age population had driver's       driving and voting ages at different times in our
licenses, but only 62 percent were registered to        history. To the extent that this pattern is stable,
vote. Both figures are approximations. The im-          however, it represents an especially large target
for motor voter among young people. And the            observers have expressed concern over the cost of
table also suggests that, once registered, people      reprogramming to interface the separate com-
tend to stay registered.                               puters handling driver's licenses and voter regis-
                                                       tration files. That has not been an issue in the
                                                       states which have motor voter systems because
Other Registration Systems                             only two of them reported any conversion or any
   In addition to motor voter, states use several      interface between the systems. In every system
other methods to facilitate voter registration and     there is a manual transfer of a physical document.
file maintenance. These include: deputy regis-         The potential for computer interaction has not
trars, agency registration, mail registration,         been fully developed, and a number of systems
election day registration, and statewide computer      are working quite well without it.
registration files. Do these affect the implemen-
tation of motor voter? Not directly, it appears,
except when the law for deputy registrars is           Elements of Motor Voter Systems
applied to motor vehicle personnel. That issue is
covered under "Legal Requirements" below.                 No two states have identical motor voter sys-
                                                       tems. Indeed, there is some variation of practice
   There are some indirect effects, however. The       even within states. The following section ana-
most common of these is the familiarization of the     lyzes motor voter systems according to seven
public. Where they have become comfortable             principal elements: methods of notifying pro-
with one method of voter outreach, they are likely     spective applicants, application forms, roles of
to accept another. For example, one official pointed   staff in driver's license offices, legal requirements
out that the debate over mail registration in his      affecting the role of staff, responsibilities for
state had been heavily concerned with the issue        transferring completed applications from the
of potential fraud or multiple registration. This      driver's license office to the voter registration
issue did not arise later in the discussion of motor   office, special activities undertaken by registrars,
voter. He attributed that fact to the general          and renewals. Each element has several possible
satisfaction with the way mail registration had        variations, and the way in which these variations
worked out. Motor voter is the more secure of the      are combined defines a motor voter system.
two systems because, in most cases, people apply
in person before a government employee.
   An official in Minnesota pointed out that elec-     Notification
tion day registration reduced the pressure for           Because applicants will not necessarily have
rapid transfer of applications from the driver's       come to the driver's license office knowing that
license agencies to the voter registration agencies.   they can also register to vote, each motor voter
The potential for a person to apply at a motor         system has some means of informing them. There
vehicle office just before the announced deadline      are three methods in use: passive, active, and
for registration but too late for the application to   automatic.
reach the voter registration office is an issue of     • Passive. The passive method relies on signs
concern in some states. That is not a problem in       posted in the office to inform voters that they may
Minnesota because the person would be allowed          register to vote. It is always used in conjunction
to vote in any case, due to election day registra-     with separate application forms, rather than
tion at the polls.                                     combined forms. The New Mexico system offers
   Perhaps most interesting is the lack of any         an example. Signs are posted in each district
interaction with existing state registration files     office of the Department of Motor Vehicles. The
in those states that have them. A number of            applicant must request a form from an official in
the office. (See Exhibit 1.) At least in the              in the nation, fits this category and serves as a
jurisdictions surveyed, the applicant and the of-         useful illustration. The Secretary of State has
ficial complete the form on the spot. That is, the        responsibility for the issuing of driver's licenses.
official types in information provided by the ap-         Staff in each of the 179 Secretary of State Branch
plicant, and the applicant checks the information         Offices are supposed to offer voter registration as
and signs the form. The motor vehicle office              part of each transaction. Persons who wish to
takes responsibility for sending the completed            register receive the application pictured in Ex-
form to the appropriate county clerk.                     hibit 2. They complete it in the presence of a staff
                                                          member, who also signs the application, thereby
   The advantage of the passive system is ease of         satisfying Michigan's requirement for registra-
initial implementation. Each office continues to          tion in person.
use its own forms and most of its former proce-
dures. The chief disadvantage of the system is              The obvious advantage of an active system
that it relies for its effectiveness upon the public      over a passive one is that, if it is properly imple-
to read, understand, and act upon the signs.              mented, it insures that each client in the office
Then, to the extent that the public does attempt          recognizes the opportunity to register. And the
to use the system, it will impose some demands            person who notifies the applicant also provides
upon staff in the driver's license office. If the staff
                                                          the application form and any directions needed.
types the form for the applicant, or even checks it
for accuracy, there is an obvious investment of               Michigan adds another step by including a
time. Conversely, without such checking the               voter registration application in the driver's li-
system loses an important quality control. More-          cense renewal notice, which is mailed to each
over, in a busy office if the voter has to obtain the     resident forty-five days before her or his license
form, get out of line to complete it, and then stand      expires. The recipient must still appear in person
in line to have it checked, much of the convenience       to make a change in voter registration, but the
of motor voter registration will be lost. Finally,        application serves as an additional reminder and,
every motor voter system, including the passive
                                                          if the applicant brings in a completed application,
system, creates some expectation on the part of
                                                          it is simply affixed to the form shown in Exhibit
the applicant that the public official working in
                                                          2 so that the applicant does not have to repeat the
the office will be able and willing to answer
                                                          same information.
questions about voting. Systems vary in the way
in which staff are instructed to respond to ques-
                                                             The principal disadvantage of this system is
tions. (This issue is covered more fully under
                                                          that it relies upon personnel in the driver's li-
"Role of Staff' below.) A passive system imposes
                                                          cense offices to make it work. Voter registration
some demands on office staff. The chief savings
                                                          officials in several of the states using such sys-
seem to come, not in processing time per voter,
but in start-up costs and in smaller numbers of           tems complain that the staff do not always ask.
applicants who take advantage of this system              Reasons suggested include lack of training, indif-
compared with other systems.                              ference or even hostility to the goals of the pro-
                                                          gram, and the extra work required. Assisting the
• Active. In active notification systems the              applicant and signing the form takes a little time,
driver's license offices take a proactive role by         but a little time per client in a busy office can add
asking their clients, individually, whether they          to a lot. When there is a line waiting, the clerk
wish to register to vote or change their registra-        must take time from the primary function of the
tion. Those who respond affirmatively are given           office (driver's licenses) to serve the primary
a separate voter registration form to complete.           function of another office (voter registration).
Michigan's system, the first motor voter program          Finally, a clerk who is unsure of what to do and is
anxious about the consequences of an error can           An obvious advantage of the automatic system
avoid the problem by simply not asking the ques-      is that it does not depend upon office personnel to
tion.                                                 notify the voter. Indeed, when properly imple-
                                                      mented, it forces the applicant to make a written
   States have used several techniques to combat      response to the question. On the other hand, the
the problem of staff non-compliance. One is           system does require redesigning and printing of
training, making sure that the staff know what is     driver's license applications. More importantly,
to be done and how to do it. Where there is a high    it, too, is not absolutely foolproof. Some people
turnover or transfer rate among driver's license      have checked "yes" and assumed that they would
staff, training must be repeated often. Obviously,    be registered without further action on their part.
the degree of support exhibited by the state driver's This problem can be avoided by having the office
license agency is very important. Oversight is        staff check the response to the voter registration
vital in this regard. Michigan has an ideal.situ-     question and advise the applicants how to proceed.
ation in that the author and chief proponent of the   But relying on staff to do so undermines some of
program has line authority over the personnel         the advantage of the automatic system over the
who issue driver's licenses. His office makes clear   active system.
the policy that this is an important program for
agency personnel. And it monitors compliance by          Combined forms are also automatic and, in this
comparing the number of address changes for           case, the request for voter information is on the
                                                      same form. (See Figure 2.) Thus it seems even
driver's licenses and voter registration received
                                                      more unlikely that the prospective voters will
from each office. If the system is working prop-
                                                      either miss the notice or mistakenly believe that
erly, these figures should be close. If they are not,
                                                      they have registered simply by checking a box.
the state office makes an inquiry.
    Another technique is the use of deputy regis-
trars in the driver's license offices. The local      Application Forms
registrar deputizes at least one clerk in each           There are three types of motor voter application
office. This step is often taken as a requirement     forms: separate, combined, and computer as-
of state law (see "Legal Requirements," below),       sisted. The first two of these are completed
but it also establishes a direct link between the     manually. One is physically separate from the
registrar and the implementing personnel in the       driver's license application and the other is
driver's license office. Moreover, people who         combined on the same form with the driver's
volunteer for the position will usually be favor-     licens application. With the third type most of
able to the goals of the program. A disadvantage      the voter information is printed by a computer on
of this approach is that, if only one or two people   a form or transferred directly in machine readable
are deputized in an office, the program may falter    form. These types have some important simi-
if these people are absent or even very busy.         larities and differences which affect the way in
                                                      which motor voter programs work.
• Automatic. Automatic systems place a ques-
tion on the driver's license application to notify      All motor voter systems, including those that
the applicant that he or she may register to vote,    are computer assisted, use manual forms to record
using a separate form. The Arizona Amended            the necessary information and the voter's signa-
Motor Voter Form (Exhibit 3) illustrates the          ture. The information requirements vary as a
driver's license application in such a system.        function of state law. The National Clearinghouse
Those who respond affirmatively proceed to ap-        on Election Administration has published a
ply for registration much as they would in other      summary of these requirements for each state
systems using separate forms.                         and the District of Columbia. (Kimberling, 1990)
In every system studied for this report the voter        place.) The County Clerk's office organizes the
registration office receives a form with an original     cards within separate binders for the different
signature. This point is very important for law          precincts to which they must be delivered. New
enforcement officials investigating voter fraud          cards must be placed in their proper alphabetical
cases. The ways in which the forms capture the           order within the binders so that poll workers can
necessary information and original signature vary        locate them quickly during the rush of business
from state to state. The primary considerations          on election day. Thus, Cook County has a sub-
involved in the design of forms seem to be compat-       stantial sunk cost in its present system. Cook
ibility with the existing forms used by registra-        County and the other election jurisdictions of
tion offices or driver licensing offices and with the    Illinois may have to deal with this issue soon
system for transferring forms (individual mail,          since they do not share a standard registration
bulk mail, etc).                                         form and the legislature has passed a motor voter
                                                         bill (which the governor had not acted on at the
   A special problem arises for a state that does        time of this report).
not already have a standard registration form in
use throughout the state at the time that the               Note that the Illinois example illustrates a
motor voter system is adopted. The problem               combination of two distinct conditions that affect
stems from the inconsistency of jurisdictions be-        the implementation of motor voter programs.
tween voter registration and driver licensing of-        One is the lack of uniformity in the size and shape
fices. A driver's license office can typically serve     of voter registration applications among local
anyone in the state. People may visit a driver           jurisdictions. The other is the legal requirement
licensing office near their place of employment or       to have documents containing the voters' original
where they go to school, for example, in a different     signatures at the polling places. A state can have
county or township from their voting residence.          one problem without the other. Like most prob-
It is not practical for the driver's license office to   lems these are easier to solve in isolation than in
maintain voter registration forms for every county       combination.
in the state, so some voters will use a registration
                                                            If nonuniformity is the only issue, then the
form that is different from the one used in their
                                                         problem is one of integrating motor voter forms
                                                         with the existing physical storage and retrieval
   The effects of the problem become apparent            systems of the local jurisdictions. The magnitude
when the application reaches the voter's home            of this task will vary with the size of the juris-
jurisdiction. The issue is usually not one of            diction. Clearly, the sunk cost in filing cabinets
eligibility since most applications will contain the     and the importance of strict procedures will be
essential information in one place or another. It        much greater in a jurisdiction having hundreds of
is, rather, one of fitting an outsized form into the     thousands of registrants than in one which has a
file system. The jurisdiction's application form         few thousand.
was designed to fit the storage system, or vice
                                                            Local procedures also affect the task. Many
versa. The form from another county does not fit,
                                                         jurisdictions that do not require signature verifi-
but it has to be kept because it contains the voter's    cation at the polling place routinely transfer the
original signature.                                      information from all application cards to some
  Cook County, Illinois, provides an example. It         other medium. Most large and many small ju-
has over 1,000,000 registered voters outside of          risdictions use computers to store, sort, retrieve,
the City of Chicago. The cards containing the            and print the voter registration information as
voters' signatures must be delivered to the polls        needed. If necessary, a field can be added to
on election day. (Illinois is one of eighteen states     computer records to indicate that the original
that require signature verification at the polling       application came through the motor voter pro-

gram and is stored in a different file from the        tures to the polling place for signature verifica-
regular applications. (Such a field would also be      tion on election day. Nancy Van Meter, Director
useful in measuring the impact of the motor voter      of the Ashland County Board of Elections, de-
program. See "Motor Voter Measurements," be-           scribed the way in which they make the system
low.) Many offices without computers still rou-        work in her county. The nearest driver's license
tinely type registration information onto new          office is the Ashland Auto Club (AAA), a private
cards, which they use in their normal operations
while storing the original for safekeeping. In         organization that works on contract with the
these cases the impact of a differently sized or       state of Ohio. Her office supplies the auto club
shaped form associated with a motor voter pro-         with Ashland County voter registration forms
gram is not great.                                     (Exhibit 5), which the club then provides to county
                                                       residents who need them. These applications fit
   If, on the other hand, the regular voter regis-     the county's binders and require no special pro-
tration forms are uniform in all local jurisdictions   cessing when they reach the election board. On
of a state, it is possible to simply adopt a motor     the other hand, the auto club and similar organi-
voter application of the same size and shape.          zations around the state give the state form
Such applications will fit the existing files of all   (Exhibit 4) to motor voter applicants who reside
jurisdictions and can be sent to the polling places
in the same manner as regular applications where       in other counties. When applications using the
signature verification is required. This was the       state forms come in to the Board of Elections, the
approach taken by Nevada, which standardized           staff must copy the information to an Ashland
its voter affidavits before adopting motor voter.      County form and then cut out and paste the
                                                       original signature to that form. This is a tedious
   The two considerations discussed here -             process, but it allows the Board to accommodate
nonuniformity of local registration forms and the      the motor voter forms within its existing system.
requirement for signature verification at the polls    The fact that most motor voter applications use
- constrain, but do not necessarily determine, the     the county form greatly reduces the burden.
type of form chosen for a motor voter program.
Other issues are involved and each type has               The State of Michigan also uses separate forms
advantages and disadvantages, as described be-         for motor voter applications and has encountered
low.                                                   the same compatibility problem. Its solution was
                                                       to print the motor voter application on one side of
• Separate forms. In most motor voter sys-
                                                       heavy stock paper and to put adhesive and a peel-
tems the voter registration application and the
driver license application are printed on separate     off backing on the other side (see Exhibit 2). The
forms. The registration application may be the         card is perforated along the dotted lines so that it
same one that is used in the voter registration        can be easily divided into sections. When the
office or it may be a different form designed for      backing is removed, the sections can be stuck onto
statewide use, especially where the local jurisdic-    appropriately sized forms at each local voter
tions do not use a standard form. Exhibit 4            registration office. This approach makes trans-
displays the Ohio form, which is used for both         ferring the information and the original signa-
motor voter and mail registration. Note that the       ture easy, although it does increase the cost of the
forms are self-mailing. Applicants may mail            forms. The Secretary of State's office recently
them to the Secretary of State for further distri-     paid $45,000 for 850,000 forms, a price of just over
bution, take them to their local board of elections,   five cents per copy.
or leave them at the driver's license office to be
picked up by the local board of elections.               A number of other jurisdictions use separate
                                                       forms as well. This type of form has several
  Like Illinois, Ohio requires election officials to   advantages and disadvantages. It appears to be
send cards containing the voters' original signa-      the least expensive option in terms of printing or

start-up costs. It does not require a change in the     On the other hand, there are ways to deal with
existing driver's license applications, although        this problem and some of the most successful
some jurisdictions add a question to the driver's       systems in the country use separate forms. To
license form asking whether the applicant wishes        preview a conclusion of this report: the success of
to register to vote. (This point is discussed under     motor voter programs using separate forms is
"Notification," below.) Election officials are free     more dependent upon the attitude and effort of
to adopt an existing application or to design a new     implementing personnel than are programs us-
one.                                                    ing other (combined or computer-assisted) forms.
                                                        • Combined forms, as the name implies, put
   There are disadvantages, as well. Separate           both the driver's license and voter registration
forms require more work on the part of the voter,       applications on a single document. There are two
who must manually fill in much of the same              types of combined forms - tearoff and duplicate.
information already provided on the-driver li-
cense application. In most cases this is not a              Tearoff forms. This type of form consists of two
critical problem, as the success of some systems        applications printed on the same sheet or card,
using separate forms will attest. Nevertheless,         which is perforated for easy separation. Iowa
a few officials in driver licensing offices suggested   uses a tearoff form, pictured in Exhibit 6. In this
that they lose some potential applicants as a           case the voter application is attached to the writ-
result. The scenario they present is a large, busy      ten test for a driver license. The voter registration
office in which the applicant has already been          section is the top portion of the form. The tearoff
waiting for some time. This is, of course, a            combined form is one step removed from the
subjective evaluation that addresses the reaction       separate form in that the driver's license and
of a relatively small number of potential appli-        voter registration applications are attached. This
cants. The available evidence indicates a generally     step is important because it guarantees that the
positive reaction on the part of potential voters to    voter registration process will be integrated with
a well implemented motor voter program using            the routines of the driver licensing office. Notifi-
separate forms. (See "Motor Voter Measure-              cation of the prospective applicant is automatic.
ments," below.)                                         On the other hand, the applicant must still com-
                                                        plete all necessary information on each form. In
   A second problem with separate forms is that         this respect the tearoff combined form is like the
they are not well integrated into the organiza-         separate form. (Minnesota is another state with
tional routines of the driver licensing office. This    experience using a tearoff form; this year they
is the flip side of the ease-of-implementation          have changed to a carbon duplicate form.) Com-
advantage. As will be discussed in the following        bining applications, either as tearoff or dupli-
section, separate forms always require some in-         cates, adds an additional constraint to the design
dependent means of notifying the prospective            of the forms. They must meet the needs of the
applicant of the opportunity to register. Some          driver licensing organization as well. This point
election officials interviewed for this project com-    will be explored more fully after the presentation
plained that the driver's license personnel did not     of the duplicate combined forms below.
always ask clients if they wished to register. This
possibility is understandable, especially in large,        Duplicate forms. This type of form goes one
busy offices.      Clerks must take time from a         step further than the tearoff form. It is designed
mission for which they were hired and trained in        to copy information from one application to an-
order to perform a job they consider to be the work     other so that the voter does not have to write the
of another agency. And there is clear evidence of       same information twice. Two jurisdictions use
wholesale non-implementation in some cases.             duplicate combined forms, Colorado and Wash-
(See "Motor Voter Measurements" in this report.)        ington, D.C.

   The Washington, D.C., form uses pressure sen-         certificate or naturalization papers to establish
sitive paper to make a copy of part of the com-          age, that fact could trigger an inquiry into citizen-
bined application. Exhibit 7 displays the top            ship status for voting purposes.
sheet of the combined form. This is the applica-
                                                            Finally, the bottom sheet of the combined form
tion for a motor vehicle operator's permit. It
                                                         is printed on heavy stock paper so that it can be
overlays the second sheet, which is the voter
                                                         used as a file copy. The District of Columbia does
registration application. Because the first sheet
                                                         not require signature verification at the polling
is shorter, the lower sections of the second sheet
                                                         place, so sending the forms there is not an issue.
(Sections 7 through 9) are visible to the applicant.
                                                         This example does show, however, thata combined
Exhibit 8 illustrates. The first sheet is pressure
                                                         form can be designed to meet physical filing and
sensitive so that the needed information from the
                                                         retrieval needs of a jurisdiction.
top form copies onto the second. Exhibit 9 shows
the entire voter application form. When it is               Colorado did require signature verification at
completed, Sections 1 through 3 and Section 6            the time that it adopted a combined-form motor
will contain information duplicated from the             voter system. Its form is smaller and does not
driver's license application. Sections 7 through 9       produce a heavy-stock copy (Exhibit 10). The
will contain original information, including an          solution for the first year of operation was to
original signature. When the forms are sepa-             paste the motor voter application onto the regu-
rated, the Bureau of Motor Vehicles receives top         lar application, which was of the right size and
sheet (Exhibit 7) and the Board of Elections             shape for the binders used in the polling places.
receives the bottom sheet (Exhibit 9).                   Shortly thereafter Colorado dropped the re-
                                                         quirement for original signatures at the polling
   Note that the information captured in Sections        place.
1,2, and 3 of this form goes to both agencies. The
only additional information required for voter              The advantages of the combined form are
registration is the individual's party, last voter       several. First, the voter registration process is
registration address, signature, date, and daytime       completely integrated into the operating routines
telephone number. Three features of this form            of the office and notification of the applicant is
deserve special attention here. First, note that         automatic. There is no way for an applicant to
Section 2 asks for "Current Residence (Street            complete a motor vehicle transaction without
Address)." This is an important feature from the         learning that he or she may also apply for voter
point of view of election officials who must assign      registration. Second, the combined form does not
voters to districts based upon the geographic            require any redundant information on the part of
location of their residences. Driver licensing           the voter. Finally, the combined form may pro-
agencies will often accept a commercial address          duce some useful information that is not normally
or a post office box, so it is important that the form   captured by voter registration applications.
is specific on this issue.                                  The disadvantages of the combined form are
   Second, the form asks for a positive declaration      the amount of coordination needed to initiate the
that the individual meets the qualifications for         program and the continuing costs of printing.
voting. One does not have to be a U.S. citizen, for      Driver's license and election officials must together
example, to get a driver license. TheD.C. Director       design a form that meets the needs of both
of Elections, Emmett Fremaux, points out that he         agencies. This step will take time and a willingness
gets an additional check on citizenship from the         to cooperate on both parts. The driver licensing
information contained in the box labeled "FOR            agency may have substantial sunk costs in its
BMVS USE ONLY" in the upper right corner of              own forms and the operating procedures built
the form. If the applicant did not use a U.S. birth      around them. Change will require some adjust-

ment on their part. Costs can be offset and              individual form, more comparable to the District
resistance can be overcome, but they must be             cost of $.06 per copy than to the $.18 per com-
recognized as important factors in motor voter           pleted application. The D.C. form is larger, more
implementation. Of course, change can bring              complicated, and pressure sensitive, hence the
benefits as well as costs. The head of the driver        greater cost. Yet since the design of the forms
licensing agency in the District was initially           affects numerous other costs in the program (per-
concerned that motor voter would increase the            sonnel, transportation, storage, etc.), these costs
cost of forms and slow down operations in the            do not represent net efficiencies of one program
agency. Yet in designing the new motor voter             over another. For example, Colorado added
form the Bureau of Motor Vehicles was able to            personnel to its motor vehicle offices and D.C. did
combine three of its existing forms into the new         not. Different programs absorb costs in different
application, thereby streamlining its own opera-         ways. Still, these figures are useful benchmarks
tions. There were delays during the initial              for other jurisdictions considering motor voter
implementation of the program (ten to fifteen            programs.
minutes longer waits than usual), but the process
                                                         • Computer assisted applications. Three
is now faster than before.
                                                         states—Montana, Oregon, and Washington—
   Duplicate combined forms do cost more. At six         have computer assisted application procedures.
cents per copy the individual forms are not much         All of them capture an applicant's driver license
more expensive than Michigan's peel-off form.            information in a local computer, either from a
However, Michigan's separate forms are used              manually completed application form or by the
only for motor voter transactions while the D.C.         applicant telling it to an official at the computer.
combined forms are used for voter registration,          For renewals this information should already be
operator permits, learner permits, identification        in the computer. The Oregon and Montana sys-
cards, organ donor designations, and address             tems use the computer to print out most of the
changes on the above. Thus the increased cost of         needed information on the voter registration card,
the form is multiplied by the greater number of          which is attached to the driver's license "camera
transactions in which it is used. And as Leroy           card." Washington uses a separate form to cap-
Bennett of the D.C. driver's license office pointed      ture an original signature for the registrars' files,
out, forms with a lot of white space are sometimes       but has electronic transfer of information be-
wasted by people who are looking to write notes or       tween driver's license and voter registration
give their children something to draw on while           systems.
they wait. Emmett Fremaux, Executive Director
                                                            Exhibit 11 displays the Oregon form. This is
ofthe D.C. Board ofElections and Ethics, estimates
                                                         the camera card, a computer generated card that
the increased expense attributable to motor voter        is used with the applicant's picture to make up
as three cents per form. A total usage rate of           the actual driver's license. The personal informa-
180,000 forms per year for 30,000 motor voter            tion that will appear on the driver's license is
applications produces a cost attributable to motor       printed from the computer onto the middle sec-
voter of $.18 per completed application. This            tion of the card. The applicant's name, address,
price is, for his jurisdiction, considerably less        date of birth, mother's maiden name, and place of
than that of other forms of voter outreach.              birth are printed on the voter registration appli-
(Fremaux, 1991)                                          cation. If the applicant does not wish to register
   Colorado recently switched from pressure              or change registration, the official tears off the
sensitive forms to carbon paper in order to reduce       last section and places it in a designated container
costs. A batch of 200,000 new forms cost $5,000,         at the work station. Those who receive the
for a price of $.025 per copy. This is the price of an   registration card review it, add information as

necessary, and sign it. In this way the Oregon            concerned are working through the problems in
system uses the computer to avoid requiring               an orderly fashion.
redundant information from the voter whereas
combined forms use the pressure sensitive paper              All of the computer assisted systems, like the
or carbon copies to accomplish the same purpose.          combined systems, save the voter the trouble
                                                          filling in the same information twice. This ad-
    The state of Washington uses a different pro-         vantage has a downside. Some voters do not take
cedure. The driver's license applicant who wishes         the trouble to enter additional information even
to register to vote does not receive a printout.          when it is needed. This problem seems to occur
Instead, the driver licensing official simply keys        most often with addresses. Some people put post
a "yes" button to a response on a computer screen,        office boxes or commercial addresses on driver's
thereby flagging the record for copying and               license applications and they fail to fill in the
transfer to the Secretary of State's office. The          blank on the voter registration card that asks for
applicant receives a separate form (Exhibit 12).          residence address. Registration officials in
The applicant completes the form by printing              Washington and Oregon pointed out this problem,
her or his name in block 6 and signing in the two         but none characterized it as being a major prob-
spaces provided. The applicant is supposed to             lem.
fill in any other information that is not included          As is also true with the combined form, the
in the driver's license record. These forms and           start-up costs are greater than with separate
the electronically coded data on the applicants           forms. In this case the driver's license computers
go to the Secretary of State's office, where a            have to be reprogrammed. Oregon estimated a
specially designated motor voter staff match them         cost of $4900 for this step. In the Oregon system,
and sort them by county for transfer to the ap-           the voter application form had to be compatible
propriate local election officials. The voter in-         with the driver's license form, but the marginal
formation is currently sent on printouts, although        cost of the addition was small, an estimated
the state may develop procedures to transfer it           $2143 in the first year of operation. The Wash-
in machine readable form to the counties.                 ington and Montana forms are separate. This
                                                          choice eases the design constraints and also means
   This is a complex system which has experi-             that forms are not wasted when driver's license
enced some start-up problems, as one might ex-            applicants do not need or wish to register. On the
pect. The program began in January of 1992.               other hand, the applicant does not see and verify
Sometimes signature cards do not match the                the voter registration information.
names on the printout, or there are multiple
matches because people have the same name and
the staff cannot immediately determine which              Legal Requirements
signature goes with which record. The Secretary              Two common legal requirements affect the
of State's staff tries to reconcile these difficulties.   way in which motor voter operates: deputy reg-
If it cannot, it sends a report to the counties for       istrars and registration in person. Some states'
their help. This example illustrates an important         election laws require that all applications for
point for jurisdictions that are considering adopt-       voter registration be administered by registrars
ing or changing motor voter systems. Any new              or deputy registrars, and some of these states
program is likely to encounter a shakedown pe-            have carried this rule over into their motor voter
riod. What appears most impressive from site              systems by providing that driver licensing staff
visits to the states of Washington and Oregon is          be deputized. A problem arises where
that they reduced the number of such difficulties         deputization is done by a local registrar or county
through careful planning and that the officials           clerk and the jurisdiction of the deputy is, by

law, the same as that of the person deputizing.        office, many people will assume that the staff are
Because the jurisdiction of the driver's license       both able and obligated to answer all manner of
office is normally statewide, this requirement         questions about elections. Agency officials will
can limit the effect of motor voter and can cause      have to consider what information the staff should
much confusion as people try to register in offices    provide and how it should be provided. Too little
outside their voting jurisdiction. A few states        information can result in improper registrations;
have dealt with this problem by providing in the       too much takes time away from other duties and
enabling legislation that any employee of the          increases the chances of conflict with information
motor vehicle or driver's license department has       disseminated by regular election officials such as
authority to take registration statewide. If the       the local registrar. Oregon provides staff with a
statute is silent on this issue, existing voter        pamphlet to give to applicants. The pamphlet
registration requirements will generally govern.       answers some questions and tells applicants to
                                                       direct all other questions to their county clerk's
   Many states require that all applicants, except     office. This system seems to accomplish its pur-
for the disabled and those temporarily out of the      pose, although it produces some frustration within
jurisdiction, register in person. This requirement     the Department of Motor Vehicles, which strives
has actually been a stimulus for motor voter,          to imbue staff with a strong service orientation
which provides for registration in person before       yet must tell them to limit their responses on
an employee of the state. Registration in person,      election issues.
coupled with a very decentralized registration
system, was a reason for the creation of the first     • Assistance.        Staff may be called upon to
motor voter system, according to Michigan Secre-       provide assistance in completing forms for handi-
tary of State Richard H. Austin, who pioneered         capped or illiterate applicants. In addition, some
the system. Michigan has 269 cities and 1242           registration offices have complained about illeg-
townships. Many people had difficulty discover-        ible handwriting and have requested motor voter
ing where they could go to register, and the fact      staff to check the forms submitted at their offices.
that some of the small offices kept irregular hours    Registrars want the problem corrected at the
compounded the problem. The motor voter sys-           source, whenever possible, because it is easier to
tem allowed a citizen to apply in any driver's         ask questions of the applicant standing in the
license office in the state during regular business    office than it is to telephone or mail a question
hours and, in many offices, on Saturday as well.       after the fact, assuming that the registrars can
                                                       even identify the applicant from an illegible form.
                                                       This problem is not unique to motor voter; it can
Role of the Staff in                                   occur with deputy registrars and mail registra-
Driver's License Office                                tion as well. The principal issue here is time. In
  The preceding discussion has frequently al-          some jurisdictions the staff provide assistance to
luded to the staff in the driver's license office.     all applicants by typing or printing the applica-
This section summarizes the various duties which       tion. Systems which use the computer-assisted,
should be considered in the design of a motor          tear-off combined form are best for minimizing
                                                       this problem. Staff make the required computer
voter system.
                                                       entries for the driver's license system while the
• Notification.      The active system of voter        applicant is in the office and then print out both
notification depends for its success upon the staff    the driver's license and the voter registration
to inform voters of their opportunity to register.     information for the applicant to check before
Other systems are less dependent.                      signing.
• Questions. Upon learning that they may               • Collecting forms. In most systems appli-
apply for voter registration in the driver's license   cants submit the completed voter registration

forms to the driver licensing staff, who collect        codes on the cards, and sends the cards and the
them for transfer to the appropriate registration       printouts to the counties for entry into their
officials. (Some systems give the applicant the         system.
option of mailing the form directly.) The various
means of transferring forms are discussed in the           Finally, there are various combinations. In
following section.                                     Michigan, where registration records are main-
                                                       tained by 1511 separate jurisdictions, driver's
                                                       license offices forward applications and voter
Transfer Responsibilities.
                                                       updates once a month. They send applications for
   Who transfers the completed applications from       townships or cities within the county where the
the driver's license office to the appropriate voter   driver's license office is located directly to the
registration office? In some jurisdictions, the        appropriate clerk. For applicants living in neigh-
voter takes responsibility for mailing or deliver-     boring counties, they send them to the county
ing the completed form, but usually the driver's       clerk, who redistributes them to the appropriate
license office initiates the transfer. The task is     township or city clerk. They send all other
more complicated than it might seem at first           applications to the Secretary of State's Elections
glance because the jurisdiction of the driver's        Division in Lansing for redistribution. As the
license office will rarely coincide with that of a     registration deadline for an election approaches,
voter registration office. So applications must be     the Branch Offices transfer the forms daily in-
sorted by voting jurisdiction, and in many cases       stead of monthly.
the applicant will not know the appropriate des-          As the Michigan example illustrates, the most
tination.                                              suitable mechanism will depend upon the cir-
    The most common procedure is for the driver's      cumstances of the state in question. The way in
license office to send all completed applications to   which the state arranges its transfer procedures
the nearest voter registration office, which then      will affect two related issues: safeguard proce-
forwards applications to other voter registration      dures and the close of registration before elec-
offices as necessary. Oregon is an example. In         tions.
some cases the driver's license office will do an      • Safeguard procedures. When an applica-
initial sort and send applications to several nearby   tion is handled by two or three different organiza-
voter registration offices with which the staff are    tions, tracking is both more difficult and more
familiar. The registration offices spend the time      important than in a single office. What happens
to look up obscure addresses and forward appli-        if a file is lost between the motor voter office and
cations to other offices as necessary. In some         the registration office? There are several possible
states the driver's license offices send the appli-    ways to discover the problem or to mitigate its
cations to the chief election officer of the state,    effect on the voter.
who then redistributes them to the appropriate
                                                          The North Carolina system appears to be the
local registration offices. North Carolina follows     most thorough. Every month the Department of
this pattern.                                          Motor Vehicles provides the State Board of Elec-
   The Washington system transfers the signed          tions with a computer-printed Voter Transaction
voter registration cards by mail and the voter         List sorted by the client's county of residence.
information in machine readable form (tapes,           This list contains the name, address, date of
etc.) to the Department of Licensing, which then       birth, and date of the transaction for each person
transfers from its computer to the one used by the     who conducted a transaction in any driver's license
Secretary of State. The motor voter staff within       office of the state. The State Board of Elections
the Secretary of State's office matches the signa-     sends each county its portion of the list with the
ture cards with the computer data, writes county       following directions:

     You are to compare the names listed               The solution adopted by most of these states is to
  thereon for accuracy and completeness.               provide a receipt at the motor voter office and to
  Should there be a name shown for which               allow people to vote upon presentation of that
  your office has no voter registration applica-       receipt at the polling place.
  tion or change of address record your board
  is directed to contact these people and pro-            Other states use the date that the application
  vide an opportunity for them to be properly          is received in the voter registration office for
  registered to vote or make whatever change           determining whether the deadline has been met.
  they intended.                                       Here the problem is to insure an expeditious
                                                       transfer of forms between the motor voter offices
   This is obviously a strong system, and most         and the voter registration offices. In Oregon, for
states do less. In several jurisdictions the motor     example, the county clerks take responsibility for
voter office sends a separate notice of the number     picking up forms at the Bureau of Motor Vehicle
of forms in each shipment to a voter registration      offices on the last day of registration.
office. The voter registration office may repeat
that process for any forms that it forwards to
other voter registration offices. This step alerts     Special Activities in Voter
the receiving office if any forms fail to arrive.      Registration Offices
Most voter registration offices complete the loop         Once the application arrives in the voter regis-
by sending out voter identification cards or other     tration office, it can be treated like any other
notices to successful applicants and rejection         application that is received from outside the of-
letters to any who were unsuccessful. Thus, an         fice (through deputy registrars, etc).
alert applicant might notice if he or she failed to
receive a card within a reasonable period of time.     • Staff will need to determine whether the new
Finally, most systems give the applicants a receipt    registration duplicates an existing one. Some
at the motor voter office so that they can establish   voters will not remember whether they are regis-
the fact that they attempted to register.              tered or whether their registration is current and
                                                       they may inadvertently re-register. This step is
• Close of registration. The concern with the          not difficult since the duplication will normally be
close of registration is that voters may apply at      revealed in the process of filing the registration
the motor voter office before the deadline but         card.
their files will not reach the registration office
before the deadline. The problem is not as great       • They will need to check addresses to determine
as it might seem because the motor voter system        whether the application has been sent to the
tends to smooth out the flow of applications           proper jurisdiction. This is not an onerous step
throughout the year, thereby avoiding the rush         because they will be familiar with most address-
just before elections. (See discussion under "Ef-      ees in their jurisdiction and will have to check in
fects," below.) Nevertheless, some individuals         any case to make a precinct assignment.
will still register at the last possible moment and
their rights to vote must be protected.                • In most jurisdictions, the staff will also notify
                                                       the voter of acceptance and precinct assignment
   Some states use the date that the application is    or provide the reasons for rejection along with
received in a motor voter office for determining       instructions for correcting the problem.
whether the deadline has been met. The problem
for these states is that valid applications may not    • In some jurisdictions the staff will have to
arrive at the registration office in time for offi-    prepare a registration card for the voter because
cials there to put the names on the list of regis-     the application is not in a form for filing. This can
tered voters which is sent to the polling place.       be an issue where combined forms are used and

should be addressed in negotiations between voter
registration and driver's license offices. The D.C.
                                                        Measuring Motor Voter
system, for example, puts the registration appli-          Does motor voter work? A short answer is that
cation on the bottom sheet of a multi-page, pres-       we cannot know yet because the existing systems
sure sensitive form. The bottom sheet is printed        are quite new, because so many factors can affect
on heavier stock than the others so that it can be      the result, and because most jurisdictions do not
used as a file card. In some cases the applications     have data that would allow them to track it
are not uniform statewide, so no single motor           through time or to compare results across sys-
voter form would fit all local needs. But the form      tems. Nevertheless, policy makers are currently
that contains the original signature must still be      deciding on new systems, changing the design of
maintained as part of the record. Michigan devel-       old ones, and looking carefully at funding levels
oped the peel-off system so that the signature can      for both. Events will not wait for research. This
be affixed to ocally designed registration cards.       section describes the available evidence in order
Some jurisdictions attach the entire motor voter        to suggest some tentative conclusions and to
form to another card.                                   promote a discussion within the election commu-
                                                        nity of the type of information it needs to have
                                                           What do we want motor voter to do? We have
    The preceding list of elements describes the        seen that there are a variety of systems. There
motor voter registration process associated with        are also a number of possible goals: showing the
initial driver licensing. The procedure may or          public that the government is trying to do some-
may not be the same when people renew driver's          thing about participation rates, making registra-
licenses. This issue is important because it af-        tion more convenient for people who would regis-
fects the number of opportunities that motor            ter even without the system, providing a registra-
voter has to capture new voter registrants. If it is    tion opportunity for many people who would not
limited to initial licenses, the opportunities will     otherwise register, improving the accuracy and
be severely constrained. As Figure 2 illustrated,       timeliness of voter files, increasing registration
half of the population obtains driver licenses          rates, and increasing voter turnout. Information
before the age of 18. The Iowa combined system          needs vary with goals.
is limited in this way because the voter registration
application is printed on the written test that            Before a program can begin to accomplish any
people take before obtaining initial driver's li-       of its goals it must be put into place. This
censes in that state. The license renewal process       seemingly trite statement directs our attention to
is paperless, so there is no automatic opportunity      what may be the most important factor in motor
to capture new registrants on the combined form.        voter success, implementation. A large body of
Staff do have postcards for mail registration that      literature in political science and public admin-
they are to make available for renewals. Oregon,        istration amply documents the fact that simply
on the other hand, mails voter registration ap-         passing a law does not insure that the desired
plications along with driver's license renewal          activity will take place. The work is done, or left
notices. This solution is possible there because        undone, by people in the field who must allocate
Oregon also has mail registration. Michigan             scarce resources among competing demands and
sends a voter registration card with the renewal        solve the myriad practical problems unanticipated
notice, and has recently changed its law so that        by the original mandate. Scholars differentiate
the Secretary of State can receive registration         between implementation, referring to what those
applications in the mail and forward them to the        charged with administering a program actually
appropriate local jurisdictions.                        do, and impact, referring to the result of those

activities. A program can be implemented with-         ways been strong support at the head of the
out having the desired impact, but it cannot have      agency where the transactions took place. The
an impact without being implemented.                   North Carolina experience was different. There
                                                       the driver's license agency was the recipient of a
                                                       mandate, not the initiator. When the program
Michigan and North Carolina                            began, the legislature and the governor who had
                                                       passed and signed the bill were in office and of the
   Two examples illustrate this point. Among the       same party, but the governorship changed party
four earliest states to adopt motor voter, Michi-      the following year. Priorities changed and with-
gan (1975) and North Carolina (1984) are the only      out active support the program languished. More
two with transaction data from the beginnings of       recently an agreement between the State Board
their programs to the present. "Transactions"          of Elections and the Driver License Section of the
simply refers to the number of motor voter forms       Division of Motor Vehicles has produced a much
filled out in driver licensing offices and forwarded   more vigorous implementation, as the steep rise
to registration authorities. The prescribed ac-        in transaction rates attests. Given this history, it
tivities in the Michigan and North Carolina pro-       is reasonable to attribute the different transac-
grams are similar. In both states the driver's         tion patterns to implementation differences rather
license and voter registration applications are        than design (impact) differences.
printed on separate forms. Both relied on driver
licensing staff to notify applicants of their op-         These two cases support several generaliza-
portunity to register until 1989, when North           tions. First, the Michigan experience shows that
Carolina put a voter registration question on the      an active motor voter program can produce re-
driver's license form.                                 sults, at least in the form of transactions at the
                                                       driver's license offices. Second, the irregular
   Yet their experiences with motor voter could
                                                       pattern traced by Michigan's transactions can
hardly be more divergent. Figure 3 displays the
raw transaction rates from these two states from       serve as a useful benchmark for motor voter
the beginning of each program through 1991. The        programs of the same design and perhaps for
heights of the two lines are quite different, par-     those of other designs as well. The sharp increase
tially because these are raw figures and Michigan's    during the first two years suggests an organiza-
population is larger than North Carolina's. The        tional learning curve as the various branch of-
important characteristic here is the shape of the      fices assimilated the new procedure. The irregu-
lines. Note that Michigan started the program in       lar pattern afterwards suggests that even well
1975 with 128,633 transactions. It grew rapidly        implemented programs will experience a good bit
to 661,736 in 1976 and varied from year to year        of yearly variation. Yet there is also a pattern to
thereafter around an overall pattern of slow           the variation. After the second year of the pro-
growth.     North Carolina began with 60,507           gram every even-numbered year brought an in-
transactions in 1984 and then dropped to a few         crease and every odd-numbered year brought a
thousand per year until 1989, when the number          decline. This pattern obviously fits the election
increased again.                                       cycle even to the point of the highest rates falling
                                                       on presidential years. It suggests that motor
   What happened? The difference appears to lie        voter programs, like regular registration pro-
in implementation. Richard Austin, Michigan's          grams, are sensitive to the biennial election cycle,
Secretary of State, pioneered the motor voter          though perhaps not to the same degree.
concept. He has been in office throughout the
history of the program and Michigan is one of the        The number of transactions is not an ideal
few states in which the Secretary of State is in       measure of implementation. It reflects both the
charge of driver licensing. Thus, there has al-        activity of the office staff (implementation) and

the response of the clients in the office (impact).      average of seven percentage points higher from
Yet it is a more direct measure than is a state's        1978 through 1990. Since the previous difference
registration rate (registrants as a percentage of        (1970-72) was just under one percentage point,
voting aged population), which reflects the impact       the net increase was six percentage points. Sec-
of motor voter plus all other sources of registra-       ond, the Michigan line is more stable and less
tion. The two figures can best be used in tandem,        sensitive to the political cycle than is the US rate.
each helping in the interpretation of the other.         Although motor voter transaction rates do follow
   An increase in transactions will not necessarily      the election cycle to some extent, they seem to
result in an increase in registration. Consider the      have a dampening effect on the swing of total
hypothetical case in which all the transactions          registration rates.
were changes, duplicates, ineligibles, etc., or,            North Carolina initially follows the US pattern
more likely, if the motor voter program simply           at a lower level, except for a deeper drop in 1978.
were to provide a more convenient mode for               Then in 1984 it rises more steeply than the US
people who would have registered by other means
                                                         rate and maintains a more stable pattern there-
anyway. An increase in transactions followed
                                                         after, declining as does the US rate but without
closely by an increase in registration rates would
provide much stronger support for the conclusion         the swings associated with presidential elections.
that the program had had an impact on regis-             Since 1984 was the year in which North Carolina
tration rates. On the other hand, if we should find      began its motor voter program, we might easily
an increase in registration rates following no or        attribute the change in registration rates to it but
few transactions in driver's license offices, we         for our knowledge of transaction rates. If all of
have to conclude that the change came from some          North Carolina's motor voter transactions in 1984
source other than motor voter.                           had been new registrants, they could have ac-
                                                         counted for a maximum increase of one percent-
   Figure 4 displays registration rates for the          age point. Although motor voter may have con-
Michigan, North Carolina, and the United States          tributed to the rise in 1984, it does not appear to
for the period of 1972 through 1990. That is the         be the primary cause of the change in North
time frame for which Bureau of Census estimates          Carolina's pattern that started then.
of state registration rates are available. We have
already seen that registration rates are sensitive         The experiences of these two states suggest
to the national political cycle, so it is necessary to   three points. First, motor voter works when it is
use the US rate as a benchmark. The U.S. rate is         implemented. Second, implementation is not
shown in Figure 4 as the plain line in the middle.       automatic. Third, we need some independent
Note the nationwide drop between 1972 and 1974,          indicator that a program has been implemented
the pattern of rise and fall associated with the         before we try to judge its impact. Yet these are
presidential election cycle, and the downturn in         but two examples. The following section at-
recent years.                                            tempts to apply these conclusions to other juris-
                                                         dictions, recognizing that data is more limited in
   Michigan is represented by the line with the
                                                         these cases.
plus marks, the highest line in the figure. It
parallels the U.S. line exactly from 1972 to 1974
but rises much more rapidly between 1974 and
1976 than does the U.S. line. This departure
                                                         Comparing Motor Voter Systems
coincides with the state's institution of its motor         One motive for studying more jurisdictions is
voter program in 1975. Afterwards, the Michigan          to determine whether motor voter results vary
line generally parallels the US line again with          depending upon the design of the system in use.
two differences. First, it is at a higher level, an      This is an important question because some

systems are more difficult to implement than           sources, made available by Multnomah County
others. Are they worth the trouble?                    (Portland) revealed some interesting results.
                                                       Comparing the period of October through Febru-
• Passive systems. We have little information          ary in 1991-92 with the same period in 1989-90,
on passive systems. State and local election           both off-years in the national election cycle, shows
officials in jurisdictions having passive systems      a 62 percent increase in registration activity after
reported mixed results with implementation. No         the introduction of motor voter. Of course, the
jurisdictions reported a large increase in regis-      Portland area's population grew rapidly during
trations as a result of a passive system. On the       this time period, so the change might not be due
other hand, passive systems generally do not           to motor voter. However, Figure 5 shows that the
have systematic monitoring systems, so it is dif-      number of weekly registrations increased mark-
ficult to tell what the result is. A survey by a       edly when motor voter started in October, 1991,
public interest group in Washington found a very       and maintained a pattern of irregular growth
low rate of implementation in the passive system       thereafter. It would appear that motor voter
used by that state before 1992. Such measures          touched off a significant increase in registration
can be very useful, as would random telephone          activity, possibly by making registration conve-
surveys of people who have recently received           nient for the newly arriving population. It will be
driver's licenses. Without some standard indica-       interesting to see whether the pattern stabilizes
tor of implementation across several states, how-      in the future.
ever, it would be dangerous to attempt to assess
the impact of the passive system.                        The State of Washington began its motor voter
                                                       program in January of 1992. In King County
   Maryland recently reported an interesting in-       (Seattle) the number of new registrations averaged
novation to its passive system. Prince Georges         during the first five months of this year was 95
and Montgomery Counties have opened voter              percent higher than the average for the compa-
registration counters in driver licensing offices,     rable period in 1988. It was 5 to 13 times higher
where there is plenty of space available. Staff will   than the average in any of the intervening years.
rotate from the local voter registration offices.
Other counties are expected to follow suit. The           The examples of Portland and Seattle raise two
two sites are now averaging around 200-300 reg-        interesting questions which can only be answered
istrations per month.                                  with time. Could the dramatic increase in reg-
                                                       istration associated with motor voter be a tem-
• Computer-assisted systems. At the other              porary effect related to the novelty of the program
extreme of complexity is the computer-assisted         and, perhaps, to any publicity surrounding its
system. The three examples now in place are too        introduction? To what extent, if any, are the
new for comprehensive before-and-after com-            increases observed in months of normally low
parisons. Nevertheless, the available data are         registration displacing activity from the normal
instructive. Montana began its program on Octo-        peak months near elections rather than increasing
ber 1,1991. In the first three months of operation     the total registration rate? Both effects identified
the system generated 3531 motor voter transac-         in the second question are desirable, but it would
tions. This figure represents 6 percent of the         be useful to sort them out.
driver license transactions.
                                                         At this point in our comparison of motor voter
  Oregon's computer-assisted system began              systems, we have no measures of activity for
around the same time, on September 30, 1991.           passive systems and selective evidence of dramatic
Site visits to three counties in December revealed     increases for computer-assisted systems. Fortu-
that the program was being implemented. An             nately, there is more data available for the other
analysis of weekly registration figures, from all      two types of systems.

• Active and combined-form systems. A                   maturity of the program and the strong adminis-
number of states using other systems have re-           trative support behind it. If so, its rate could be
ported transaction data sufficient to support some      a benchmark to which other programs could as-
rough comparisons. The most useful data come            pire. On the other hand, this rate may reflect
from states with active or combined-form sys-           conditions not shared with all other states. One
tems. None of the computer-assisted systems has         possibility is the fact that local registration is a
been in place long enough to generate even one          function of the township, not the county. As a
year's worth of data.                                   result, people moving within the local areas in
                                                        Michigan, and in other states with township
   In order to make comparisons across jurisdic-
                                                        registration systems, will change registration
tions we have to somehow standardize the
                                                        offices more often than in county-based systems.
transaction rates. Dividing the number of
                                                        One of Richard Austin's reasons for starting the
transactions by the voting aged population of the
                                                        motor voter program in Michigan was that many
state helps to control for the differences in size.
                                                        voters complained of not knowing where to go to
Seven states and the District of Columbia have
                                                        change their registration. Driver's license offices
reported transaction data for at least one year
                                                        are a more easily identifiable and convenient
between 1986 and 1991. We know that transac-
tion rates vary from year to year within the same
state, so we should only attach significance to            This reasoning suggests that Michigan's un-
large differences. It turns out that the jurisdic-      usually high transaction rate may come more
tions cluster in three groups. Michigan stands by       from an unusually high rate of address changes
itself with a rate that varied between 10 and 12        rather than from new registrations. Address
percent of voting aged population. Colorado,            changes comprised 60 percent of total motor voter
Nevada, and DC form another group generally in          transactions in Michigan from January through
the 6 to 9 percent range, with a few exceptions to      November of 1991. In Colorado, by contrast,
be noted below. Iowa, Maine, Minnesota, and             changes averaged 23 percent of transactions in
North Carolina fall with a range of 1 to 3 percent,     1987-88. A 1991 report from the District of
although we will see exceptions here, too.              Columbia put address changes at 21 percent of
   These clusters mean very little in themselves        transactions. Emmett Fremaux, Executive Di-
unless we can find some reasonable explanations         rector of the DC Board of Elections and Ethics has
for the differences. We can start by asking whether     suggested that the percentage of registration
the choice of motor voter systems (active or            changes among total transactions will naturally
combined) makes a difference. The answer ap-            increase with the age of the program. There is
pears to be that it does not, at least not by itself.   some evidence of this trend in the Michigan data.
States using active, separate-form systems fall in      The percentage of changes increased from a low of
all three groups - high (Michigan), medium (Ne-         51 percent in the first year of the program, 1975,
vada), and low (Maine and North Carolina). Ju-          to 66 percent in 1984. Yet even the low of 51
risdictions using combined-form systems fall into       percent in Michigan was higher than the percent-
two categories - medium (Colorado and DC) and           ages from Colorado and DC. It seems likely that
low (Minnesota). The remaining state in the low         both factors are at work here - program matura-
group, Iowa, uses a combined form for original          tion and the township registration system.
licenses and mail-in postcards available at the           This result leads to another. Without the extra
driver license offices for renewals.                    address changes, Michigan's transaction rate
  Perhaps we can find a better explanation by           would fall well within the same range as those of
looking at the groups of states. The singularly         Colorado, DC, and Nevada. Perhaps 6 to 9 per-
high transaction rate of Michigan may reflect the       cent ofvoting aged population is a good benchmark

for motor voter programs of this type. Site visits     tration forms. It is clear that Iowa's system is not
at two of these jurisdictions, Colorado and DC,        directly comparable to those of the other states in
indicated that those programs were being imple-        this group.
mented as they were designed to be. The Colo-             Minnesota's motor voter program has not had
rado transaction rate did drop from 6 to 4 percent     the restricted reach of Iowa's. The data presented
in 1989, the last year for which complete data         here are for Minnesota's combined, tear-off form,
were available from that state. In that same year      which was available for original applications and
Colorado changed from a 4-year cycle to a 5-year       renewals until 1992. (The state has just replaced
cycle for driver's license renewals. Nevada, on        that application with a combined-duplicate form.)
the other hand, increased from 8 percent to 11         The explanation for the low level of activity in
percent between 1990 and 1991, following the           Minnesota's motor voter program probably lies in
addition of vehicle registration offices to the pro-   its even greater availability of alternatives. The
gram.                                                  state has both mail registration and election day
                                                       registration. Moreover, the target for motor voter
   What about the remaining group of states,           may also be smaller in Minnesota than elsewhere.
those for which all observations were well below       Although nationally the percentage of the popu-
the 6 percent mark? Two of them, Maine and             lation licensed to drive greatly exceeds the per-
North Carolina appear to be temporary residents        centage registered to vote, the difference is not
of this category. Maine started its program in         evenly distributed across the states. In 1990 it
1990 and North Carolina began implementing its         ranged from a high of 40 percentage points in
program again in 1989. Both rates were moving          Delaware to -7 percentage points in Minnesota,
up between those starting points and 1991, and         which was the only state to have more registrants
both were at 3 percent in 1991. Only time will tell    than licensees.
whether these trends continue. The other two
states require more explanation.                          The hypothesis that convenient alternatives
                                                       for registration depress the number of motor
   Iowa uses a combined, tear-off form that is         voter transactions in Iowa and Minnesota is the
attached to the written test taken by first-time       flip side of the argument that Michigan's rate is
applicants. Driver's license renewal does not          high partially because of its township-based
require a form. Driver licensing staff do have post    registration system. Both rest on the fundamental
cards for mail registration and they are supposed      assumption that people will respond to greater
to ask clients whether they wish to register or        ease or less burden of action. This observation
change registration. Yet neither process seems to      leads to the question of whether motor voter
generate a large number of applications. In 1990,      systems are really bringing new registrants into
a total of 5532 individuals applied on the combined    the system or simply making participation easier
form, and the Department of Transportation re-         for people who would have done so anyway.
ported giving out 15,509 postcards that year.
Both figures declined significantly in 1991. Most         In one sense the answer does not matter.
residents get their original license before the age    Transaction rates may be used directly as a
of 18, and the state has had a loss of population,     measure of public acceptance. A high transaction
indicating that few people would encounter the         rate means that a large portion of the public
combined form because of moving into the state.        prefers this method of conducting its voter reg-
Why the postcards do not generate more activity        istration business to the other options available,
is less clear. One reason may be the availability      including the option of doing nothing. Interviews
of alternatives. Mail applications are printed in      with driver's license staff in a number of states
telephone books and on tax forms. Candidates           turned up unsolicited comments in praise of the
and political groups may also deliver mail regis-      program as a public service. Although the pro-

gram added something to their work loads, they          tional average, and they increased most in those
liked the positive feedback they got from the           jurisdictions that had the program in place the
public. So transactions have some value irre-           longest. Michigan's second place may seem at
spective of the number of new registrations or          odds with its very high transaction rate, but we
address corrections they produce.                       must remember that an unusually high percent-
• Transaction and registration rates. An-               age of that rate was made up by address changes.
other goal of motor voter programs is to increase       Actually, we should not put too much confidence
registration. As noted above, transaction rates         in the numbers for any one state because of the
include address and name changes, duplications,         many other factors that can affect registration.
and ineligible applications, as well as registra-       Still, the consistent results from the available
tions. Even motor voter registrations may not           data provide support for the conclusion that higher
actually increase the jurisdiction's total registra-    transaction rates are followed by noticeable in-
tion rate. Individuals who would have registered        creases in voter registration.
in any case may simply find motor voter to be
                                                        . This is not the same thing as saying that motor
more convenient. So we must examine registra-
                                                        voter will increase relative registration rates in
tion rates over time to determine whether there
has been any change.                                    all cases. Witness the examples of Iowa, Minne-
                                                        sota, and North Carolina until 1989. Program
   Table 1 shows the relative change in registra-       design and implementation, the size of the target,
tion rates for four states following the implemen-      and the availability of other attractive options
tation of motor voter programs. These are all the       may all make a difference. The latter two charac-
states which were shown in the preceding section        teristics may not be critical since a small motor
to have had significant transaction rates by 1990.      voter target and convenient alternatives for reg-
The question here is whether these transactions         istration generally are found in jurisdictions that
had any apparent effect on registration rates.          have high registration rates anyway.
The second column, headed "Election Dates,"
gives the time frame for each comparison. It
consists of the last national election before the
implementation of motor voter and the election
that followed four years later, so we are compar-          Policy makers and administrators are usually
ing presidential with presidential and off-year         concerned about the budgetary impact of pro-
with off-year. "Years of Implementation" refers         grams. How much new money will it cost? Motor
to the number of years in which motor voter was         voter programs vary from no apparent budgetary
implemented between the elections. On the aver-         impact to over $100,000 per year in some juris-
age, only a fourth of each state's prior licensees      dictions. These figures can be misleading because
would encounter motor voter in a year, but if they      some costs, especially personnel time, are simply
registered, their registration would normally stay      combined with other activities. In many cases the
on the books for at least four years. So the effects    costs are too small to justify separate tracking.
should build over time. The net change is the           Because motor voter programs use facilities and
difference in registration rates between the two        services of agencies that exist for other purposes,
dates, controlling for the change in the U.S.           the assignment of costs to the program can be
registration rate. That is, if a state's rate went up   tricky. The hiring of new personnel is an obvious
by 5 percentage points and the national rate went       impact. But in some cases, at least, the program
up by 2 percentage points, the net change would         seems to have resulted in increased productivity
be 3 percentage points.                                 of existing personnel. The following account
  The results are about what one would expect.          breaks down the major ways in which motor voter
Registration rates did go up, relative to the na-       programs use resources and discusses the avail-

able information about the budgetary and organi-        where the driver licensing staff records applicant
zational impact of each.                                information directly onto a computer, jurisdic-
                                                        tions using active systems may add a question to
• Start-up costs. Some new programs will have
                                                        the computer screen. It is both a prompt for staff
identifiable start-up costs, one-time expenditures
                                                        to ask about voter registration and a method of
of money or time to begin implementation. These
                                                        recording applicant responses. Oregon, which
fall in three categories: training, the design of
                                                        uses a computer-assisted system, estimated pro-
new forms, and computer programming.
                                                        gramming start-up costs at $4900. The State of
   Passive systems and most of the active systems       Washington, which has electronic transfer of
did not report any start-up costs. They did not         registration information, recorded start-up com-
need to design new forms or reprogram comput-           puter costs of $17000.
ers. Any training or notification about the new
program was conducted as part of normal opera-            Washington's total start-up costs for the De-
tions. As noted above, some election officials have     partment of Licensing, not including the time of
complained about the lack of training in a few          implementing staff and administrators, was
jurisdictions and have expressed the desire to          $25,426.50. This state has the most complete
conduct it themselves. This might be a cost-            computer-integrated system adopted so far and is
effective means of communicating registrars' goals      the only state to have prepared a training video.
and concerns to driver licensing personnel. Nor-        So this figure does not represent average costs. It
mally, any procedural training will have to be          covers the most extensive start-up program en-
conducted by or carefully coordinated with the          countered for this report.
parent agency, which is the only one with author-
ity to set priorities and modify activities of agency   • Supplies and Equipment. The most obvi-
personnel.                                              ous supply requirement is that of the application
                                                        forms, which vary widely in cost. Many juris-
   Training methods and costs vary among the            dictions do not report form costs separately. They
remaining jurisdictions. Several have produced          are quite low and are combined with other printing
pamphlets or instruction sheets for driver licens-      costs. Prices for the more complicated forms
ing personnel. These describe the routine pro-          range form 2 cents to 6 cents apiece. However,
cedures that the staff are to follow and tell what      printing costs associated with different designs
to do in the case of questions or problems. West        cannot be separated from other programmatic
Virginia has an excellent manual but has not            costs for two reasons.
broken out the development costs or printing
costs. The State of Washington has produced a              First, design determines usage and the num-
video presentation. Its estimated development,          ber of forms actually used can also vary widely.
production, and field training costs for the initial    Active systems and Washington's computer-as-
implementation of the program was $4784.                sisted system provide forms only for those who
                                                        say they need them. Combined forms, by defini-
   Many jurisdictions need to redesign forms.           tion, go to everyone who requests any type of
This step can be as simple as adding a question to      application offered on that form. So the additional
the driver's license application or as complicated      cost of the motor voter application must be mul-
as the integration of multiple applications. Most       tiplied by the total number of forms used, not just
form design is done in-house and not separately
                                                        the number of motor voter applications received.
costed. This was the case for the DC form, which
                                                        Wastage also varies with the use of the form.
appears to be the most comprehensive redesign
                                                        Those placed out in the office are more likely to be
yet accomplished.
                                                        wasted than those given out at the counter. Mail-
   Computer programming is obviously required           in forms, such as those that accompany renewal
for the computer-assisted systems. In addition,         notices in some jurisdictions, can use a lot of

forms per application received. One official com-         Several jurisdictions have estimated the time
mented that her jurisdiction had distributed           required for each motor voter transaction in order
enough postcards to cover the voting aged popu-        to figure total personnel requirements. Oregon
lation several times over.                             has made the useful distinction between the time
   Second, form design also affects the amount of      it takes to determine whether a motor voter
staff time required to operate the program. The        transaction is needed and the time required to
DC form increased printing costs but lowered           actually conduct a motor voter transaction. The
transaction time, for example. Estimates of            former estimate, 15 seconds, must be applied to
transaction times are provided below.                  all applications in the agency. The latter esti-
                                                       mate, 30 seconds, applies only to motor voter
    Postage is another expense for some jurisdic-      transactions. Using these numbers, plus inci-
 tions, and the cost will vary depending upon          dentals and some executive time, Oregon esti-
usage. Minnesota spent $900 for postage in a           mated the personnel cost to the DMV of $94,796
year. Montana estimates $25 per month for              in the first biennium and $ 114,918 in the following
postage. The greatest potential postage expense        one. This is just over $50,000 per year. DMV staff
that is unique to motor voter programs is the cost     now think this estimate may be a little low.
of mailing applications from the driver licensing
offices to the voter registration offices. We do not      Other jurisdictions have different estimates
have a postage cost from a high-volume state that      depending upon the requirements of their sys-
used the mail to transfer applications. Yet post-      tems. Several fall within the 2-3 minute range for
age costs should be fairly easy to estimate by         normal transactions. Those that require or offer
jurisdictions considering motor voter programs.        any unusual activity will take longer. In Colo-
                                                       rado, the driver licensing staff must verbally
   An equipment expense could be incurred by
local voter registration offices if the size and       administer an oath to each applicant. In North
shape of the form produced, coupled with require-      Carolina, the staff members generally type the
ments for access to original signatures, necessi-      necessary information onto the registration ap-
tates the acquisition of new filing cabinets or        plication. The best guidance for any jurisdiction
binders. Some states, such as Michigan, have           considering adoption or modification of a motor
chosen to use more elaborate forms in order to         voter system is to run simulations and field tests.
avoid this difficulty.                                 It is very difficult to generalize from one system
                                                       to another.
• Personnel. People are at once the most ex-
pensive and the least expensive part of a motor           There may be changes in the work load of
voter program. They are the most expensive part        election personnel, too. One would expect their
because multiplying even a small amount of time        work to increase as the number of transactions
devoted to motor activities by the number of           increases. That would be true of any program.
people involved will produce a large figure. The       Particular types of change may be associated
State of Washington's Department of Licensing          primarily with motor voter, however. Some states
estimated the total staff time spent on their          collect the applications in a central office and
program at $209,672 per year. Yet people are the       then send them to the appropriate local regis-
least expensive item in many budgets because no        trars. This function requires staff time. Effects
personnel are added. Licensing personnel in            at the local registration offices will probably de-
Washington are doing motor voter with the same         pend upon the design of the forms and the care
staff as before. The need for additional personnel     with which they are completed. If registrars have
is partially a function of system design and par-      to paste forms onto standard sized cards or do
tially a function of how busy the existing person-     extensive follow-up for incomplete or illegible
nel are with their normal duties.                      applications, then motor voter applications will

be more difficult to process than others. So time        Election officials have been almost universal in
saved by system designers or at driver licensing      praise of this feature of the program. One did
staff may have to be paid back in voter registra-     comment that frequent address changes between
tion offices. Conversely, several registrars com-     elections created more work than was necessary,
mented that motor voter applications were easier      but even this complaint reflects the registrars'
to process than those produced by other outreach      greater ability to keep registration files current.
programs because they were typed or computer-         In fact, most responses from local registrars fo-
printed.                                              cused on the benefits of increased efficiency and
                                                      better file maintenance rather than increases in
   Work load changes do not have to be negative       registration rates. One official in Colorado even
for either driver licensing or voter registration     reported cutting back on permanent positions.
offices. The change of forms in the District of
Columbia seems to have brought an absolute
increase in staff productivity. Because of the        Design and Implementation
consolidation of applications they are able to        Strategies
handle their total work load more efficiently than
before. Even changes that impose real increases          Does program design matter? There are ex-
in transaction time may not result in comparable      amples of apparently successful programs in three
increases in costs. The actual impact depends         of the four design categories (active, combined-
upon how much down time there is between              form, and computer-assisted). Within these three
applications. Some offices, especially those in       categories, at least, implementation appears to
large urban areas, work at or near capacity all of    be the most important variable that can be affected
the time. Less active offices often have an inter-    by program sponsors. Yet it turns out that
mittent flow of clients. A small increment of time    implementation is related to design in a para-
per client will have a much greater impact on the     doxical way. The systems that require the most
former than on the latter.                            change in existing forms and procedures would
                                                      appear to be the most difficult to implement. Yet
   In voter registration offices motor voter tends    no official in combined-form or computer-assisted
to increase efficiency by smoothing the transac-      jurisdictions mentioned non-implementation or
tion rate throughout the year. Normal registra-       partial implementation as a problem. At least
tion activity is heavily concentrated around elec-    one official in 9 of 15 states with statutes man-
tions. Motor voter activity tends to be more          dating passive or active programs mentioned this
evenly dispersed throughout the year, although it     problem.
does follow the election calendar to a limited           This result makes sense in light of implemen-
degree in some jurisdictions. A possible explana-     tation theory and the design of motor voter pro-
tion for this pattern is that people interact with    grams. Combining forms or modifying computer
driver licensing offices for reasons totally inde-    programs commits the driver licensing and voter
pendent of the election schedule, but the publicity   registration organizations to a joint course of
surrounding elections makes some difference in        action. There may be problems, as both Oregon
the number of them who take advantage of their        and Washington have experienced to some extent
motor voter opportunities. Data from the District     during their start-up phases, but so far the agen-
of Columbia shows a much more even rate, month        cies involved have been able to solve them or live
by month, for motor voter registrations than for      with them. Where the organizational forms and
other registrations. Offices can use their staff      procedures are not so tightly integrated and other
much more efficiently and can avoid hiring tem-       commitments take priority, non- implementation
porary help if their work loads are steady.           or reduced implementation is a possibility.

   A number of jurisdictions have taken steps to           Field-testing is also a useful device. New York
improve the design of their programs and to             state ran a pilot program of mailing voter regis-
facilitate implementation before starting full-         tration request cards with motor vehicle regis-
scale operations. In Oregon, for example, the           tration renewal applications from August 1987
Secretary of State's office, the county clerks' asso-   through September 1988. That program was not
ciation , and the Division of Motor Vehicles worked     continued. New York does include Department of
cooperatively to refine the original proposal as it     Motor Vehicle offices in its agency-based regis-
moved through the legislative process. DMV              tration program. Maine also ran a pilot program
staff ran simulations to estimate the staff time        tied to the registration of motor vehicles, begin-
required per transaction as a result of motor           ning in April of 1990. It moved to full imple-
voter, and they contacted other states to compare       mentation in August of 1990.
figures. The county clerks agreed to take respon-          At least four states employ mechanisms to
sibility for transferring the completed forms from      audit their systems. Michigan and West Virginia
the driver's license offices to the appropriate         compare motor voter transactions with other
registration offices. Several participants com-         transactions coming from each branch office.
mented that the level of cooperation made the           North Carolina sends to each county a voter
difference in the passage and implementation of         transaction list from the DMV each month. The
motor voter during a period of severe financial         State Board of Elections directs the county boards
constraint.                                             to contact each listed person for whom they have
                                                        not received a registration or change of address
   In the state of Washington the Secretary of          record. New York has sent inspectors to agency
State created a special office to implement the         offices to insure that the staff was asking clients
transfer of signature forms and application infor-      about registration.
mation from the driver's license offices to the
counties. Meanwhile, the Department of Licens-             The District of Columbia has the most thorough
ing involved its district offices in the design of      system for tracking the implementation and im-
                                                        pact of its motor voter program. It has added a
detailed procedures, such as deciding at what
                                                        field on the computer record for each voter to
point to ask the applicants if they wished to
                                                        designate the source of registration. This step
register. Their participation improved the ulti-        permits the development of statistics on the
mate design and gave them an opportunity to feel        subsequent turnout of motor voter registrants
some ownership of the program, according to             compared to other registrants.
John Specht, the DMV official charged with
implementation. DMV also produced a videotape
to train its personnel in the new procedure.            Legislation
   Training is an important part of implementa-           Although a few programs are based upon ex-
tion. Several respondents who felt that driver          ecutive order, most are based upon legislation. A
licensing personnel were not fully implementing         properly drafted law can resolve many problems
the program in their states pointed to a lack of        that would otherwise impede implementation or
training. It may be the case that the jurisdictions     limit impact. The specific language of bills will
with the less complicated programs (passive and         have to vary depending upon the type of system
active) are also the ones less likely to offer com-     desired and the existing elections law. There are,
prehensive training in motor voter administra-          however, a number of elements that should be
tion. Yet the need for training may be greatest         considered. 100% VOTE/Human SERVE has
where the desired action is not automatically           published a recommended checklist for model
built into the operating routine.                       legislation in its News on Agency-Based Voter

Registration of February 15, 1992. That list              Other jurisdictions will make different choices
provided the idea for this one, which covers the       and provide different levels of detail in their
same items and suggests a few more. The pur-           statutes. Below is a list of issues for consideration
pose here is to point out certain issues that should   in the drafting of statutes.
be resolved and not to recommend a particular          • Target. Passive systems do not need to specify
resolution.                                            a target because the forms are available to any-
   The first issue to address is how much detail to    one in the office. Other systems specify who is to
put in the legislation. The answer will vary from      be offered the opportunity to register by listing
state to state. Some jurisdictions traditionally       the types of applications that are to be included in
leave matters to administrative rule- making           the motor voter program. These may include:
that others put in legislation. The Minnesota          initial driver's licenses, renewals, duplicates,
statute is quite brief, but it contains a number of    name or address changes, non-driver identifica-
essential elements.                                    tion cards, organ donor designations and vehicle
                                                       registrations. Statutes typically begin this with
   The department of public safety shall change its    "All applicants for..." Then they list the licenses
applications {agency required to take action) for      or other transactions to be included. They con-
an original, duplicate, or change ofaddress driver's   clude with the action to be taken such as asking
license or identification card (type of driver li-     a specific question or simply providing an oppor-
censing actions to be included in motor voter} so      tunity to register.
that the forms may also serve as voter registration
cards, {type of form - combined) The forms must        • Notification.     If the goal is to have a staff
contain spaces for the information required in         person ask each applicant whether he or she
section 201.071, subdivision 1 {specific informa-      wishes to register, that requirement must be in
tional requirements for new form to meet) and          the statute. Failure to mention notification or
applicable rules of the secretary of state, {rule      language requiring only that registration forms
making authority over registration applies here,       be "available" may be interpreted as allowing a
too) Applicants for driver's licenses or identifi-     passive system. On the other hand, it is not
cation cards {target of the program) must be asked     necessary to specify the point in the process at
if they want to register to vote at the same time.     which voter registration will be offered.
(method of notification) A copy ofeach application     • Form. It is necessary to prescribe the type of
containing a completed voter registration must be      application form and establish responsibility for
sent to the county auditor of the county in which      designing and providing it. It is neither practical
the voter maintains residence or to the secretary of   nor necessary to describe the form in great detail.
state {establishes transfer responsibilities and       Doing so limits the ability of administrators to
options) as soon as possible, {provides guideline      design the most efficient form for the current
for timeliness of transfer without setting specific    system or to adjust to changing technologies. The
standard)                                              most important distinction is between separate
   The statute goes on to make relevant computer       forms, on the one hand, and combined-form or
records available to the secretary of state and for    computer-assisted applications on the other.
use in the statewide voter registration system         Without a specific mandate the driver licensing
being developed. Elsewhere the secretary of state      agency will generally not incur the expense and
is given broad rule-making authority over the          disruption of redesigning an essential component
statewide registration system, including the           of its system.
prescribing of procedures for the transfer of mo-        Expressing the desired policy in a statute re-
tor voter forms from the licensing offices to the      quires considerable care. For example, a man-
secretary of state or to the county registrars.        date for a single form would permit tear-off

forms (Iowa and, formerly, Minnesota) or dupli-        • Transfer responsibility. Legislation should
cate forms (DC and Colorado). It would include         fix responsibility for transferring forms. Most
the computer-assisted systems of Oregon and            legislation also addresses the issue of timeliness.
Montana, but not include the state of                  Some specify a weekly transfer.
Washington's system as it is now configured. A
mandate for a motor voter registration system          • Record keeping and reporting. Since the
that requires only an additional signature would       record keeping systems of the implementing
allow either combined-duplicate forms or com-          agencies will not have been designed to monitor
                                                       and evaluate a motor voter system, drafters should
puter-assisted applications, but it would also
                                                       consider whether it is necessary to mandate record
necessitate putting all of the necessary informa-
                                                       keeping and reporting in the statute. The discus-
tion for voter registration, e.g., residence address
                                                       sion of motor voter measurement in this report
or affirmation of eligibility, on the basic applica-
                                                       suggests the use of some common statistics. The
tion. Involving the implementing agencies in
                                                       number of motor voter transactions compared to
the drafting process helps the legislature craft a
                                                       total transactions in individual driver licensing
statute that provides the necessary stimulus and
                                                       offices and in the system as a whole provide a
authority while avoiding unnecessary restric-
                                                       measure of implementation and initial impact.
tions.                                                 (Low transactions in a few offices might reflect
• Assistance and other staff duties. Statutes          implementation problems while low transactions
generally authorize driver licensing staff to ac-      in the whole system might reflect design prob-
cept applications and describe any particular          lems.) As noted in the case of Michigan, it would
actions that they are to take. Drafters should         be useful to break down motor voter transactions
consider the issue of deputization here. Ap-           into new registrations and address or name
proaches range from permitting deputization, to        changes. Registration offices could record num-
requiring deputization of at least one employee in     ber of registrations received by source if the forms
each office, to exempting driver licensing person-     permit such identification, such as motor voter vs
nel from deputization requirements by directly         mail vs deputy registrars. Record keeping is
granting them authority to accept applications.        easiest where files are stored on computer. Fields
Whether assistance should be addressed in the          can be added to indicate source and type of regis-
statute is an issue. Some laws are silent on this      tration transaction. In jurisdictions that keep
point, leaving discretion to the implementing          voting histories, turnout rates of motor voter
agency. Others, such as the new Texas law,             registrants can be compared with those of people
require nonpartisan assistance upon request. Still     who registered through other systems. The Dis-
others mandate a variety of activities including       trict of Columbia has already undertaken such
administering oaths, witnessing signatures,            studies. In order to evaluate the whole system a
stamping applications, and providing receipts.         central office will have to aggregate statistics for
In some cases these duties are assigned by refer-      the whole jurisdiction. Thus it is useful to desig-
ence. Applying an existing deputization law to         nate reporting and monitoring requirements.
licensing employees brings into the motor voter
process all of the existing requirements for the       • Organizational responsibilities and re-
acceptance of applications by deputy registrars.       lationships.     Because independent organiza-
The amount of staff interaction involved deter-        tions are involved, it is usually necessary to
mines the transaction times described earlier in       designate which agency is responsible for each
this report, and depending upon the design of the      element of the system. This is usually accom-
system, may also determine the quality of appli-       plished in the language setting forth the require-
cations received by registrars. It is important        ment as the Minnesota example shows. It tasks
that system designers weigh carefully the costs        the department of public safety with changing its
and benefits of each requirement.                      application and requires that the voter registra-

tion portion comply with rules promulgated by          volved in implementation, if it is possible to do so.
the secretary of state. Statutes frequently desig-     Learn how any proposed changes will affect their
nate a lead agency and mandate consultation            internal procedures. It may be necessary to
with other agencies.                                   create or designate the organizational structure
                                                       for rule making, monitoring, and problem reso-
• Training. No statute reviewed for this report        lution in the motor voter program.
mandated a training program for licensing per-
sonnel. Yet given the frequency with which             • Review existing election law. Issues of
training was mentioned as a potential remedy for       particular importance to the design of motor
implementation problems, it is worth consider-         voter programs are requirements for: deputization
ing. Putting a training requirement in the stat-       of driver license staff, the administration of oaths
ute also increases the probability that the associ-    to applicants, witnessing of applicants' signa-
ated costs will be recognized and provided for.        tures, the delivery of original signatures to the
                                                       polling places on election day, and polling place
• Adjustments to other state law. This is a            procedures for dealing with anyone whose regis-
catch-all category for provisions necessary to ad-     tration application is not processed in time (e.g.
dress any gaps between the motor voter program         whether a receipt issued at a driver licensing
and existing state law.                                office would be accepted). Some jurisdictions
                                                       have changed existing law or exempted the motor
                                                       voter program in order to streamline procedures.
A Checklist for Program Design                         Others have designed their motor voter programs
and Implementation                                     to more easily fit existing requirements.
   The preceding section focused on motor voter        • Review different motor voter designs.
legislation. The following list subsumes those         Consider (a) how each would fit the proposed
recommendations under a checklist of items to          target for the program, the existing organization
consider in the total design and implementation        of the driver licensing and voter registration
of motor voter programs. Not all of them will be       functions in the jurisdiction, and the relevant
necessary in every case and some of them may not       election laws and/or (b) how any of these might be
be politically or practically feasible in particular   changed to accommodate a desired motor voter
cases. There may also be unintended omissions.         program. The descriptions of motor voter ele-
The purpose is to provide a summary statement          ments in this report provide a starting place and
that can be modified by others based upon their        may help in narrowing the search. Contact with
experiences.                                           officials who have experience implementing the
                                                       type of program being considered is especially
• Determine the target for the proposed sys-           important. Such contact should not be limited to
tem. Depending upon the jurisdiction, different        officials representing one agency but should be
procedures may be needed to reach initial licens-      extended to experienced representatives of all
ees and persons applying for renewals, duplicates      offices involved in the implementation. The
or changes. There may be additional opportuni-         experiences of local offices (county or township
ties related to non-driver ID cards, organ donor       offices for registrars, district or mobile offices for
designations, and vehicle registrations.               driver's license personnel) tend to differ depend-
                                                       ing upon the size of the jurisdiction.
• Review the organization of the driver li-
censing and voter registration functions in            • Estimate costs and benefits of the designs
he state. Identify the key officials and their         still under consideration. Suggestions are
organizational relationships. Create a program         contained in the section on motor voter measure-
design team representing the parties to be in-         ment above. Two points require emphasis. First,

there is rarely a one-to-one correspondence be-       likes the program. Clearly, many people register
tween the results in one jurisdiction and results     to vote or change registration information through
of the same program in another. Projections must      motor voter. Convenience to voters and improved
take local conditions into account. Second, both      voter file maintenance are valuable and attain-
immediate and long-range consequences are im-         able goals, irrespective of other effects. In addi-
portant. The careful integration of organiza-         tion, at least some of the people who register in
tional routines between the regular driver li-        this way probably would not have registered
censing program and the motor voter program           through other means. There has been a positive
will usually increase start-up costs but improve      change in the registration rate, relative to the
program implementation thereafter.                    national registration rate, in all four jurisdictions
                                                      where a high transaction rate has been docu-
• Choose a design and prepare legislation.
                                                      mented. However, any conclusion regarding the
In practice a legislative proposal will often pre-
                                                      effects of motor voter on participation rates must
cede the steps described above, not follow them.
                                                      be regarded as preliminary. There is not yet
Yet, ideally, a jurisdiction would have done a good
                                                      enough data to sort out the effects of motor voter
bit of information gathering and decision making
                                                      from the other factors that can influence regis-
before finalizing the legislative mandate.
                                                      tration and turnout.
• Pass legislation.
                                                         Passage of a motor voter statute will not have
• Prepare a detailed implementation plan.             the same effect in every jurisdiction. The most
This is the point at which agencies work out          important determinants of motor voter impact
specific procedures and establish responsibilities    seem to be program design, the degree of imple-
and timetables.                                       mentation, the size of the target, and the relative
• Develop materials and software as re-               attractiveness of available alternatives.
quired.                                                  Design and implementation are related. Pro-
• Test the program and adjust as necessary.           grams using separate forms (passive and active)
It is usually necessary to test the assumptions       are the easiest to implement initially because
built into a program on an audience that is not       they have the lowest start-up costs. Yet all of the
familiar with its history or purpose. There are a     reports of incomplete implementation after the
variety of ways to pre-test programs ranging from     intial start-up come from jurisdictions of this
simulations with a few volunteers to full-blown       type. The reason seems to be that the other two
pilot programs such as those employed by several      types (combined-form and computer-assisted)
of the jurisdictions studied for this report.         require permanent change in the operating rou-
                                                      tines of driver licensing offices. The transition
• Train personnel.                                    will almost certainly encounter organizational
• Launch program. Publicity can help notify           and technical difficulties, but once it is accom-
the public and impress upon implementing per-         plished motor voter becomes a more or less auto-
sonnel that this is an important project.             matic part of the office procedures.

• Monitor implementation and impact.                     Where an active program is implemented, as in
                                                      the case of Michigan, the impact on registration
                                                      rates appears to be as strong as for combined-
                                                      form programs. The computer-assisted programs
Conclusion                                            are too new to have comparable impact data.
  Motor voter can make a difference. The high         Jurisidictions having convenient registration al-
numbers of transactions in some jurisdictions         ternatives and high registration rates before motor
indicate that a significant portion of the public     voter would seem less likely to experience a

dramatic change after adopting a program. Data
now coming in from new programs in states with
already high registration rates (Maine, for ex-
ample) may cast some doubt on this generaliza-
tion, however. It would be useful to learn whether
the transaction rates continue to increase and
whether they result in increased registration.
   Different program designs incur different costs
and they distribute those costs differently, both
among implementing agencies and between start-
up and continuing implementation. Failures to
provide adequate resources or to distribute them
properly have been cited as explanations for
implementation problems. Yet there are examples
of improved productivity as a result of motor
voter and evidence that it can be very cost-effective
relative to other forms of registration outreach.
The District of Columbia has excellent compara-
tive data on this last point.
   Thus it appears that both costs and benefits
will vary depending upon the type of program and
characteristics of the jurisdiction implementing
it. This report has attempted to describe the basic
program types as well as some important juris-
dictional characteristics of programs now in
existance and to suggest some generalizations
that policy makers might use in designing new
programs or modifying old ones. The knowledge
base in this field is changing rapidly as jurisdic-
tions experiment with new programs and more
information accumlates from existing programs.
The election community should continue to follow
these developments and to build a base of com-
parable data from which future policy makers
and election administrators can draw.

Appendix 1


Exhibit 1
Exhibit 2
Exhibit 3
Exhibit 4
Exhibit 5
Exhibit 6
Exhibit 7
Exhibit 8
Exhibit 9
Exhibit 10
Exhibit 11
Exhibit 12
Appendix 2


Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

                         FROM 3/4/91 TO 2/3/92
Appendix 3


Table 1


                                           YEARS OF
     Colorado               84-88              4             7.8%
     Michigan               74-78              4             6.6%
     Nevada                 86-90              2             4.8%
     DC                     86-90              1             2.7%

                             For information about other

    Innovations in Election Administration

   National Clearinghouse on Election Administration
   Federal Election Commission
   999 E. Street, N.W.
   Washington, D.C. 20463

   Toll Free 800/424-9530
   Direct    202/219-3670
   FAX       202/219-3880

    WASHINGTON, DC 20463                                   Bulk Rate Mail
                                                      Postage and Fees Paid
      OFFICIAL BUSINESS                             Federal Election Commission
       Penalty for Private                             Permit Number G-31

To top