Technology for Voter Registration

The process of registering voters and producing voter lists is one of the most important and time-consuming activities carried out by election management bodies (EMBs). Several types of technology can be used for voter registration application, including:

• database management software, for storing and manipulating data

• geographic information systems (GISs), for assigning voters to geographic locations and for ensuring quality and integrity of voter data

• imaging technologies that can be used for data entry as well as for identification needs, data matching and data recognition

• telecommunications technologies to support networked computer systems and other communication needs

• data storage media, such as disks, CD-ROMS and magnetic tapes

The first step in using technology for voter registration is to determine the requirements of the country voter registration such as:

• centralised versus decentralised

• permanent voter register with ongoing registration versus periodic registration

• the need or not to produce voter cards

• the need or not to use a voter card and a national identity card for voting

This analysis helps determine which available technologies are more appropriate to implement a specific kind of voter register.

Purposes of Voter Registration

When designing an automated voter registration system it is useful to define the purposes of the voter registration process. These would typically include:

• pre-identification of eligible voters (thereby removing the need to prove the identity of voters at the polling place during voting)

• pre-identification of eligibility for candidature (thereby removing the need to prove eligibility for candidature during the nomination period)

• pre-identification of eligibility of members of political parties (using voter registration as means of verifying the qualifications of a political party for registration)

• integrity control (using the voter registration process as a means of preventing multiple voting and impersonation)

• Production of identity documents (such as voter registration cards)

• planning for polling (using the voter register as a means of determining where polling places should be located and how many staff and how much material are needed at each polling place)

• producing lists of eligible voters for use in polling (using the voter register as proof of a person's right to vote)

• identifying names of people who have or have not voted in an election (using records generated during polling based on the voter register, particularly to identify instances of multiple voting and, in compulsory voting systems, instances of failure to vote)

• voter education (using the voter register as a resource to contact electors and to understand the demographics of the voting population)

• providing voter register data to political parties and candidates (as required by law for campaigning and transparency purposes)

• publicly publishing voter register lists (for transparency purposes)

• non-electoral purposes (such as jury lists and sharing data with other government agencies for law enforcement and protection of the public revenue purposes)

These various purposes for which voter registers can be used tend to dictate the form in which voter register data is stored and manipulated. In order to produce this wide range of voter register products, the data contained in voter registers needs to be electronically captured and stored in electronic form, usually in a database system.

A voter register database can be used by an election management body (EMB) to assist various aspects of the electoral process, such as election planning and redistricting.

Voter registration data can be aggregated to various geographic levels to assist with determining where polling places should be located and what resources (both staff and material) should be devoted to each polling place. It may be feasible for an EMB to link its voter registration data with its polling place management database to automatically calculate quantities of material and the number of staff needed. This can be done through assigning “catchment areas” to polling places and calculating the number of voters registered within each polling place's catchment area.

Voter registration data can also be aggregated to various geographic levels to assist with redrawing electoral boundaries or redistricting. Where a redistricting regime requires expected future voter numbers to be taken into account, a voter registration database can be used to plot voter mobility and to determine growth trends over time. However, while voter register data may show trends in growth or decline, care should be taken when using this data for redistricting purposes to include other factors, such as plans for future housing developments or plans to demolish housing.

Voter Registration and the Electoral System

The applicable electoral system and its enabling legislation in each jurisdiction will largely determine what data has to be included in a voter register and what the voter register will be used for.

Voter registration qualifications and disqualifications are generally determined by legislation. The method of applying for registration usually includes the steps an applicant must take to prove eligibility for registration. Such steps can include completing a form with personal data, a declaration of eligibility, proof of eligibility by, for example, providing representatives of the electoral management body (EMB) with relevant identity documents such as birth certificates, passports, immigration documents or a declaration by recognized witnesses if no proof of identity can be provided for a justified reason.

Usually, the electoral legislation determines how an electoral district is defined. This may require that a voter register include voter geographic data to ensure that the voter is registered to vote in the electoral district correspondent to the voter residence address.

Again, these requirements indicate the desirability of electronically capturing and manipulating voter register data.

The applicable electoral legislation is also expected to indicate how the voter register is to be used in the voting process. In most cases, the voter register is used in voting places and EMB offices to determine a person’s right to vote and therefore has to be suitable for voting purposes. This means that copies of relevant information stored in the voter register are available, for instance, in printed hardcopy list formatted in a logical order, so that voter information can be found easily, or in an electronic list accessed by computer. In either case, there is a need to electronically capture and manipulate voter register data.

A requirement to record the fact that a registered voter has voted also impacts on the design of the voter register data base and on how it is used.

The voter data to be electronically captured and manipulated is also determined by additional requirements such as the need to compile master voters lists for the purpose of determining instances of multiple voting or, in the case of compulsory voting systems, instances of failure to vote.

Impact of Voter Registration Methods

The voter registration methods set down in electoral legislation, guidelines or policies will also impact on how technology needs to be used in the voter registration process.

Voter registration can be active or passive. That is, a person may be required to actively apply for voter registration, or a person may be automatically (or passively) registered through participation in another process, such as holding a drivers licence or being included on a national population register.

Where voter registration is active, a process is needed to convert an application into a voter register record. While this process can be conducted using an entirely manual paper-based system, the various outputs required of modern voter registers generally dictate that an application for registration be converted to an electronic form.

Where voter registration is passive, data is generally taken electronically from another source, such as a drivers licence authority or a national population register. This process is generally not practicable without the ability to use electronic data transfer.

Voter identification requirements

Electoral legislation will generally set down what, if any, proof of identity needs to be submitted by a person applying for voter registration. Where proof of identity is required, a system must exist to record the fact that proof has been provided. While a manual paper-based system could be used in this case, an electronic database system would generally be much more efficient.

Where voter identification details need to be recorded and used at a later date for identity purposes, technology can be used to record and store those details. For example, signatures can be scanned and electronically recorded. Photographs can be taken and stored in digitised form. Where digitised bio-identity data is stored, software can be used to determine whether a person has registered more than once.

Types of voter registers

The type of voter register specified in electoral legislation will also impact on the use of technology for voter registration.

A continuous register is a register that is constantly kept up-to-date by amending and adding voter records whenever necessary. An electronic database can be used to easily update records and add new records, as well as keep track of amended and deleted records.

Periodic registers are another option. In these cases, a register could be established for a single electoral event or for any electoral events occurring within a defined period. Periodic registers generally require voters to register anew and previous registrations are not taken into account. While this type of register can be recorded using manual paper-based systems, as there is not as great a need to allow for amendments and updates as with continuous registers, many advantages can still be gained by electronically capturing and storing registration data.

In some jurisdictions no voter register is produced. In this case, voters apply for registration and are given a registration card as proof of registration but no consolidated register is compiled. Where no voter register is produced, the benefits that flow from holding an electronic register are not achievable, and the need to electronically capture data is not present. Therefore, the only benefit technology can give to this process is to assist with the process of issuing voter registration cards. Of all the types of voter registration processes, this is the one least amenable to using technology.

Some jurisdictions do not have a separate voter register, but make use of other government registers, such as a national civil register. In these cases, electronic storage and manipulation is virtually essential in order to manipulate data and produce products that can be used for electoral purposes.

Voter Identification Methods

Technology can be used to assist voter identification processes undertaken by election management bodies (EMBs) during voter registration.

The level of required proof of identity varies considerably between jurisdictions. Some jurisdictions simply require voters to sign a declaration stating their eligibility when registering to vote. Others require voters to provide documentary proof of identity. Some jurisdictions record evidence of identity such as fingerprints or portrait photographs for use during the voting process.

The following topics examine issues associated with voter identification during voter registration:

• identity checks for fraud control

• electronic proof of identity

• technology to verify and record identities of voters

EMBs may issue identity documents or proof of registration documents to registered voters. Technology can be used to automate this process.

Voter register databases can be used to generate the data to be printed on an electoral identity document. Where photographs, signatures or finger/thumb prints have been digitised and stored in a database, they can also be printed on identity documents generated by the voter register database.

Identity documents containing photographs, signatures or finger/thumb prints can also be generated using specialised systems designed to produce identity cards while the subject is present. In these cases, textual information is printed on hardcopy (using data either provided on the spot or data extracted from a database). The voter usually signs this hardcopy record, and/or makes a fingerprint or thumbprint. The operator places the hardcopy printout, including the signature and/or finger/thumb print in the device, and takes a photograph of the person. The device then prints an identity card including a copy of the printed data, the signature and/or finger/thumb print and the photograph. The card is usually laminated and can include tamper-evident security devices such as holograms or embedded print to make it difficult to forge or alter the card.

Identity cards can also be produced that do more than simply show images of text and pictures. “Smart cards” incorporating magnetic strips or data chips can also store electronic data about the person who is the subject of the card. This data can include bio-identification data, so that smart cards can be used with smart card readers and bio-identification readers (such as finger print scanners) to automatically verify a person's identity.

Smart cards can be “read only” cards that simply contain information about the subject. Other types of smart cards can be “read-write” cards, which have the information contained on the card updated as the cards are used. For example, a read-write card used to verify a person's right to vote could, once used, be recorded as having been used for that election, so that it could not be used for voting in that election again.

Where smart cards are used in polling places, they could be used to replace current methods of recording that a person has voted. Where a voter uses a smart card at a polling place to verify his or her right to vote, the smart card reader could at the same time record that that person had voted and transmit that data to a central database during or after polling.

The provision of smart cards to voters and smart card readers to polling places is expensive, so an EMB  needs to determine if the advantages are worth the expense.

Smart cards incorporating an electronic identity could also be used for voter registration or voting by computer over the Internet or at a computer kiosk, provided the computer was equipped with a smart card reader.

Computer software can be used to perform a range of tasks that can assist an EMB in reducing instances of fraudulent registration or voting and to identify and delete instances of duplicated voter registration records.

There are various types of techniques and technologies to verify and record identities of voters, namely:

• matching techniques and comparison routines

• signature recognition techniques

• digitized photographs

• bio-identification systems

• identity documents produced by election management bodies

• electronic proof of identity

Proving identity during an electronic transaction emerged as an identity  issue in the 1990s and the 2000s. Particularly with moves towards electronic voter registration and voting over the Internet, the issue of proving that the person sending an electronic transaction is the person named in the transaction has become important. Various strategies such as the provision of electronic signatures, public key encryption and the use of Personal Identity Numbers (PINs) are being developed. This field is still in its infancy, however, it is evolving rapidly and improvements can be expected in the 2000s.

Matching techniques and comparison routines

Software can perform various comparison routines to determine whether a person applying for registration is already registered, perhaps at another address. Electronic searches can be programmed into voter registration databases to identify whether a person applying for registration is already on the register. Since people do not always use the exact same name every time they complete a form, software comparison routines need to make use of “fuzzy matching” techniques to help identify possible duplicates. Using fields other than names can also assist, particularly date of birth records. Even with dates of birth, people can show different dates on forms over time, and “fuzzy matching” may also be needed when comparing birth dates.

“Fuzzy matching” involves programming a computer system to search for matching data by applying various criteria. For example, a possible match could be identified where a name differs by only one letter. Possible matches could also be identified where first names and surnames matched but middle names did not match or were absent in one record but not in another.

Names can also be matched using aliases. A system can be set up that compares different variations on the same name, which could be used by a person at different times. For example, aliases for John could include Jon, Jonathon or Johnstone; aliases for Catherine could include Katherine, Caterine, Katerine, Kate, Cate and so on.

Matching techniques also need to take account of people changing names. In some societies, a woman typically changes her surname when she marries. People also change their names by deed poll. Where this occurs, matching by surname will generally not find a match with an earlier record using the old name. In these cases, matches can be found using first name or given name fields and dates of birth. To assist with this process, it is desirable to ask voters on their registration forms whether they have changed their names since their last registration, and to ask for their previous registered address. This data, if supplied, can greatly assist in finding out of date entries.

Fuzzy matching programs often produce several possible matches with a current record, some or all of which may not be actual matches with the person concerned. A human operator should be used at this point to determine whether these are actual matches and to alter the relevant records as necessary.

Signature recognition techniques

Scanning technology can be used to electronically capture handwritten signatures. These digitised images of signatures can be made available to EMB staff over a computer network, so that they can perform visual comparisons of digitised signatures with signatures provided on later documents.

Digitised signatures can also be automatically analysed and compared using appropriate software. This software compares patterns in digitised signatures and can flag possible mismatches for a human operator to investigate.

Digitized photographs

Digitized photographs of people's faces can be used as a method of determining whether one individual has registered more than once. Computer software can be used to compare different photographs to determine whether the same person appears in more than one photograph. This technique could be used for voter registration purposes to determine whether a person had fraudulently registered more than once, if all registered voters have their photographs taken and digitised.

Bio-identification systems

Bio-identification systems can be divided into two main types: visual and electronic.

Visual bio-identification methods include use of photographs, signatures and/or finger or thumb prints on identity cards. These are relatively low cost to implement and administer. Many off-the-shelf identity card issuing systems incorporate inclusion of photographs and signatures. At polling places, polling staff are expected to compare the photographs, signatures and/or finger or thumb prints on the identity cards with the voters.

This process has its weaknesses, as signature and finger print comparison is a highly skilled process that polling staff can not be expected to master, and a person's appearance can often change significantly from that contained in an identity photograph, particularly if the photograph is not updated regularly. Nevertheless, this kind of identity system will probably suffice in most cases where the risk of voting fraud is not unacceptably high.

If visual bio-identification systems are not considered secure enough, electronic bio-identification systems can be used. Electronic bio-identification systems can include digitised voice, hand print, finger/thumb print, or retinal (eye) images. Using these systems, digital records of a person's voice or physical features are stored on disk or smartcard and compared to the actual features of the person using some form of electronic reader. Access is only provided where there is an exact match.

Electronic bio-identification systems are expensive to acquire and administer, and may not be practical for use with the general voting population (although this may change as the relevant technology becomes more widespread and cheaper). However, they are a feasible option for providing security at sensitive locations, such as work places and counting centres.

Voter Registration Processing

The most effective way to use technology for voter registration purposes is to capture voter data and geographic data electronically and store and manipulate the data using a database system.

The mechanics of how to input, store and manipulate voter register data can be a complex process.

The following topics examine various aspects of processing voter registration data:

• The Voter Register Processing Environment

• Database Design

• Processing Functions

Environment

The type of computer system used to process a voter register, and the complexity of it, will depend on:

• the size of the voter population (or number of records) to be stored on the system

• the number of different kinds of data (or data fields) to be stored

• the size of the geographic base to be used in the system
  

• the number of expected data transactions

• whether the system is to be used for a one time event, or whether it will be used for an ongoing or continuous voter register

• how data is going to be input

• whether the system will store out-of-date records as well as current records (to keep voters' registration history or for auditing purposes, for example)

• how many users will access the system and how often

• whether the system will be used in one office, or whether it will be used in several different geographically dispersed offices

• whether users will expect to access data on line, and whether they will expect data to be up-to-date

• what products will be output from the system
  

• what hardware the system will be used on

• whether the system will be run on stand-alone computers or over a network

• how data will be sorted and manipulated

A system designed to cater to a relatively small number of voters, to be used in a small central office and that does not have complex input and output requirements will be relatively straightforward. However, where large numbers of voters are to be recorded, and/or where large numbers of users will access the system and/or where complex input and output requirements exist, the voter registration database will need to be more carefully designed and managed to maximise its effectiveness.

Batch or on-line processing

A key decision to make regarding a voter register processing environment is whether to use batch or on-line processing for data input. With batch processing, data is input in "batches" of many records (each record representing a transaction related to one voter) and stored in a temporary data file. At a regular interval (often overnight when the computer system is not in heavy demand), the batched data is uploaded to the main data file, so that new records are added, changed records are updated and old records are deleted or archived in one process.

Batch processing is useful where the available computers are not powerful or are not joined in a network that lends itself to on-line processing of data. With batch processing, any loss of system performance resulting from data updating normally takes place overnight, when system demand is low or non-existent. With some database systems, users need to log out of the database before any updates can be processed, making overnight batch updates a useful method that avoids restricting officers' productivity. Batch processing is also useful where a system is dispersed over different physical networks, so that separate versions of the database need to be updated. In these cases updating each database once a day using a batch update is preferable to updating each database every time a single record is updated.

On-line processing is feasible where powerful computers are used and users share a network capable of allowing on-line updates. In this case, data is entered into the live database rather than a temporary batch file. As soon as an on-line record is updated, it becomes available to other users of the system. On-line processing is more convenient for users as data is kept constantly up–to-date and they do not have to wait for batch updates to be run overnight. However, on-line processing is more difficult to organise as it requires a complex (and usually expensive) network system to make it feasible, particularly where users are geographically dispersed. Special care also has to be taken with database design to ensure that different users are not able to update the same record at the same time, leading to the possibility of errors.

From a usability point of view, on-line processing gives users more current information than batch processing. However, batch processing can be used on less expensive systems and is generally less complex to design and manage than on-line processing. In many cases, the lack of current data using batch systems will not be a significant problem.

Distributed or centralised systems

A complex database like a voter register can be used in essentially two ways: as a distributed system or a centralised system.

A distributed system can be utilized where users of the system are spread over more than one network. In most cases, this occurs where users are geographically separated from other users, such as regional offices in different cities. In a distributed system, each component of the system maintains its own copy of the database and (usually) a subset of the data. For example, in a jurisdiction with six different regions, each of the six regions could maintain data for voters registered only in that region. Whenever there is a need to coordinate data between the distributed regions, this could occur by way of batch updates (see above).

In a centralised system, all the data is kept on one centralised database, which is accessed over a network connecting all the regional offices (if any). Taking the above example, in a jurisdiction with six different regions and a centralised system, users in any region could access data held on any of the regions. Any updates having a cross-regional effect (such as a voter moving from one region to another, leading to a new record in one region and a deleted record in another region) could take immediate effect. In a distributed system, such cross-regional affects would have to wait for a batch update to take effect.

The main advantages of a distributed system include lower costs stemming from less need for high-capacity, cross-region networks, and improved system performance resulting from smaller file sizes, as the amount of data dealt with in a distributed system is less than in a centralised system.

A centralised system, while it is more expensive since it needs greater network capacity and larger file sizes, has the advantages of providing access to all data to all users and of allowing for automatic updating of records across regions.

Database Design

The process of designing databases is a technical skill. Databases are best designed by trained computer professionals.

The design of a voter register database will depend on the complexity of the requirements of the users and of the specified inputs and outputs required of the system.

Factors that will influence the design of a voter register database might include:

• the size of the voter population (or number of records to be stored on the system)

• the number of different kinds of data (or data fields) to be stored

• the size and complexity of the geographic base to be used in the system

• the number of expected data transactions

• whether the system is to be used for a one time event, or whether it will be used for an ongoing or continuous voter register

• how data is going to be input

• whether the system will store out-of-date records as well as current records (so that voters' registration history can be kept and for auditing purposes, for example)

• how many users will access the system, and how often

• whether the system will be used in one office, or whether it will be used in several different geographically dispersed offices

• whether users will expect to access data on line, and whether they will expect data to be up-to-date

• what products will be output from the system

• what hardware the system will be used on

• whether the system will be run on stand-alone computers, or over a network

• how data will be sorted and manipulated

As voter register databases typically contain many thousands of records and may be used by large numbers of users across one or more networks, they need to be carefully designed to ensure they perform satisfactorily. Many strategies can be adopted to improve the performance and usability of a voter register database. Some strategies to consider include:

• choosing an appropriate software platform—off-the-shelf database systems that are supplied with office automation suites are generally not suited to complex databases containing large amounts of data; specialist database software intended for large databases is preferable

• ensure that the hardware used has sufficient processing power and network speed to maximise the performance of the database

• map out the “architecture” of the database before programming commences (that is, plan how all the different elements of the database will interact before programming starts; this will make for a better finished product)

• use a relational or object-oriented database design to maximise the flexibility of the database

• avoid storage of repeated data by using linked tables and relationships between data (for example, rather than store a complete address for each person in the register, store only a linking address code next to each person, linked to a complete address in a separate address file)

• use computer code that follows industry conventions, so that a person other than the original programmer can correct, modify, update or audit the database.

• document the database, both for users and for programmers, providing user manuals as appropriate and including descriptions of code and fields in the database design to aid programmers

• use standard naming conventions for all the objects in a database (such as tables, queries, reports, forms, code modules, data fields, controls used in forms, and so on)

• use programming code that is of modular design (that is, code that is made up of discreet modules that can be separately tested and evaluated)

• use indexes and unique identifiers (such as a “primary key” to provide a unique identifier for each record) to speed searches and to enable linking between data tables

• use a consistent user interface across the system's range of forms and reports, to increase usability

Ad hoc enquiries

From time to time managers and other users of voter register data may wish to extract data from the voter register database that has not been anticipated by the designers of standard output reports. One way of catering for ad hoc requests is to provide for an on-line reporting facility that permits users to specify a range of variables, such as start dates and finish dates.

Where a user requests data that is not catered for by any standard reports or by a flexible on-line reporting facility, usually the only option left (short of refusing to supply the information) is to ask a programmer to extract the information from the database. This can be an expensive option, particularly where contract programmers are involved, so users should be aware of the cost involved. Where special requests are specially programmed in this manner, it may be desirable to include the request as part of the standard set of reports so that, if the information is requested again, the programmers will not need to be called again.

Data sorting

Usually, whenever data is being used for a report or an output product such as a printed voter register, it needs to be sorted according to logical criteria. Some ways of arranging data include sorting by:

• surname (alphabetically, with a sub-sort on given names, to facilitate finding names on lists)

• *registration number (where it has a logical use, such as where the registration number is used to identify a voter)

• electoral district (usually as a preliminary sort before sorting according to a further criteria, such as surname)

• street number, street name and/or locality (for use in door-to-door canvasses, for example)

• polling division/voting station (for use during polling)

• criteria requested by political players (where they have a right to obtain data)

• data requested for non-electoral purposes (where other agencies have rights to obtain data, such as jury lists supplied to courts, sorted by jury district)

Consideration for likely sorting needs should be taken into account when the voter register database is designed. Some database programs provide for “indexing” of certain fields that might be used for sorting or searching. Indexing fields helps to speed up database performance by recording sort information as the data is stored, thereby reducing the processing time involved when a sort or search is requested.

It may be desirable to include special “sort” fields in a database table. For example, there are various conventions associated with sorting surnames in some cultures. For example, the Scottish “Mac” and “Mc” prefixes are often printed in a sorted list as if they all started with “Mac” to assist users to find names when they are unsure of the spelling. (Therefore, using a special sort field, the following names would be listed as McPhail, Macphee, McPherson, Macphillamy, whereas without a special sort field, a program would automatically sort these names as Macphee, Macphillamy, McPhail, McPherson.)

Random sampling facility

It may be desirable to include a random sampling facility in a voter register database where random samples might be a feature of output products. Database software will usually have a random sample feature that can generate the necessary samples. Where a random sample is required by legislation or other guidelines, an election management body may wish to ensure that the functions can be satisfied by the software before it is purchased.

Some occasions where a random sample facility might be used include:

• jury lists

• selection of poll workers

• research purposes

• verification of party membership lists

Functions

The processes of adding, amending or deleting data in a voter register will likely need to be carefully controlled. Safeguards can be enacted to ensure that only authorised users can gain access to perform these tasks, and adequate audit trails can be designed to verify that all additions, changes and deletions are properly made. This is essential to preserve the integrity of the voter register and to guard against fraud.

Data Creation and Maintenance

There are several different elements involved in processing data in an election management body's (EMB's) voter register database.

The methods used to add data to a voter register database depend on the type of data collected (such name, address, birth date, photo and other biometric data), required documentation and the different ways in which voter data can be received meaning personally, by telephone or by e-mail.  

On the other hand, data amendment can be required for a number of reasons, including:

• where the voter has notified the need for an amendment (such as a change of address notification)

• where an external agency has notified the need for an amendment

• where the EMB has identified an error in data input

• where an event has occurred requiring data to be updated (such as a change to electoral district boundaries)

The methods used to amend data in a voter register database will also depend on the different ways in which the amendments can be notified.

Data may need to be deleted when:

• new data is received that supersedes old data

• notice is received that a voter is deceased or has left the jurisdiction

• notice is received that a record was created through fraud or error

For transparency and audit purposes, it is usually appropriate to keep a record of data deletions. In many cases, rather than delete data, it is appropriate to move out-of-date data to an archive file where it can be accessed if needed.

In particular, where changes are made to a voter's records over time such as changes to names or current addresses, it may be desirable (depending on local operational and legislative requirements) to keep a history of such changes.

Data validation

Various data validation routines can be applied to data to assist in identifying errors or fraud. Some validation routines that can be used include:

• requiring address data to conform to a standard geographic base

• performing automatic qualifications tests (such as ensuring the voter's date of birth is within the permitted age range or checking that a person's declared citizenship details satisfy the voter eligibility criteria

• checking whether spelling of names conforms to a standard list (thereby identifying possible spelling errors)

• checking whether the number of voters registered at a particular address exceeds the number of persons capable of living at that address

• where signatures, photographs or finger/thumb prints are digitised, electronically checking whether any of these identifying features submitted by a voter matches any other records on the database

It is important that a voter register should only include one current record for each eligible voter, to ensure the register facilitates the democratic principle of one person, one vote.

Software can perform various comparison routines and electronic searches to determine whether a person applying for registration is already registered, perhaps to another address. As people do not always use the exact same name every time they complete a form, software comparison routines need to make use of “fuzzy matching” techniques to help identify possible duplicates. Using fields other than names can also assist, particularly date of birth records. Even with dates of birth, people can show different dates on forms over time, and “fuzzy matching” may also be needed when comparing dates of birth.

“Fuzzy matching” involves programming a computer system to search for matching data by applying various criteria. For example, a possible match could be identified where a name is the same except that one letter might be different. Possible matches could also be identified where first names and surnames match but middle names do not match or are absent in one record but not another.

Names can also be matched using “aliases”. A system can be set up that compares different variations on the same name that can be used by a person at different times. For example, aliases for John could include Jon, Jonathon or Johnstone; aliases for Kate could include Katherine, Catherine, Caterine, Katerine, Cate and so on.

Matching techniques also need to take account of people changing names. In some societies a woman typically changes her surname when she marries. People also change their names by deed poll. Where this occurs, matching by surname will generally not find a match with an earlier record using the old name. In these cases matches can be found using first name or given name fields and dates of birth.

To assist with the data duplication process, it is desirable to ask voters on their registration forms whether they have changed their names since their last registration, and to ask for their previous registered address. This data, if supplied, can greatly assist in finding out of date entries.

Fuzzy matching programs often produce several possible matches with a current record, some or all of which may not be actual matches with the person concerned. A human operator at this point can determine whether any possible matches are actual matches and alter the relevant records as necessary.

Reports

A voter register database should be capable of producing a range of statistical collations of data. This information could be used for a wide range of purposes, including performance management and statutory reporting. Statistical collations are usually produced as reports, which can be printed or made available on-line.

Some examples of statistical reports that could be produced by a voter register include:

• total numbers of voters in the whole jurisdiction, with subtotals for other geographic areas, such as states, electoral districts and sub districts

• total numbers of voters in various voter categories

• total number of transactions processed in the database, such as additions, deletions, alterations, transfers into electoral districts, transfers out of electoral districts and transfers within electoral districts

• total numbers of voters registered for small geographic areas, such as census tracts for redistricting purposes

• summaries of data from the geographic base, such as types of dwellings, numbers of voters registered at dwellings and numbers of dwellings with no voters registered

Production of some reports can be a considerable load on a database system, particularly where calculations on large data files are undertaken. Such system overload can be prevented or minimized by running complex reports as overnight “batch” requests, that run during system "down times" to avoid overloading the system during peak usage times. Another method is to run routine report requests on a regular basis, often as overnight batch requests, and make those available to users, rather than make on-line requests an option.

Where system resources permit, however, it may be desirable to allow users to have on-line access to reports on demand. This empowers users to receive the specific data they want, when they want it, and can be useful in circumstances where up-to-date information is needed quickly (such as during an election period). On-line access to reports can include user-specified variables, such as start dates and finish dates that allow users to extract customized data.

Targeting groups for special attention

A voter register database can be used for targeting groups for special attention. These might include:

• specific addresses or types of addresses (such as vacant dwellings and institutions, e.g. universities, hospitals, nursing homes, military barracks and flats/apartments)

• societal groups (such as people from specific ethnic backgrounds or people living in remote areas or disadvantaged areas)

Voter Register Data

Voter registration systems reliability and integrity can be maximized by using electronic capture, storage and manipulation of data.

Voter registration system inputs may include:

• paper forms completed by voters

• electronic forms completed by voters

• electronic data provided by other agencies

• hardcopy information provided by other agencies

• verbal advice provided by telephone or in person

• information obtained from field workers in hardcopy or electronic form

• information derived from returned mail

Capturing Data

Before taking a decision on  a data capture method, an election management body (EMB) needs to  define what data needs to be captured. This data depends largely on the legislative requirements applicable to the voter register, but it can also include requirements specified by administrative and operational convenience.

This means that data only needs to be captured if there is a legislative, administrative and/or operational reason for doing so. It may not be necessary to capture all of the data included by a voter on an applicaton form because some of these fields may be used by EMB staff to determine eligibility to register, but there may be no need to record the data in those fields the register database.

Some voter registration systems for countries with a large percentage of an immigration population, for example, might record on voter registration forms (among other things) the voter place of birth as well as citizenship information for persons born outside the country. This information is used by staff to determine if a person is eligible to register, but once eligibility has been determined there might be no need to store all the related data on the voter’s register for future reference. Another database may be created to archive the digitized images of the original form so that, later on, if necessary, the information can be accessed and extracted.

Data input requirements can be designed taking in consideration the output requirements, which could include, for example:

• data mandated by legislation on voter registration or required by administrative and operational needs

• output fields that could be useful for sorting data or for selecting subsets of data

• special categories that may apply to voter registrations

• needed auditing/tracking fields(such as date and time of data entry or amendment, name of data entry operator, records of previous entries related to each voter)

• output fields needed on the complete range of products, such as assignement of voter to voters stations,  statistical and other reports to be derived from the voter register.

Some output fields such as the voter’s electoral district can be calculated by the computer software from other input fields like the address fields and do not need to be data entered. .

A typical list of fields captured at the input stage could include:

• name (which can be subdivided into more precise fields, such as first name, middle names, surname and family name as applicable)

• address (which can be subdivided into more precise fields, such as apartment number, street number, street name, locality, district, province, state, postal  or zip code and country)

• date of birth

• gender

• former name (if the person's name has changed, for example by marriage or deed poll)

• former address (so that an earlier registration can be cancelled/updated)

• place and country of birth

• identity number(s) (as applicable to the particular jurisdiction, such as ID card number, passport number or  social security number)

• citizenship details (for example, if proof of citizenship is required for registration)

• postal address (for those whose postal address is different from their residential address)

• special voter category indicator (for example, a code to indicate whether a voter belongs to a special category of voters, such as a voter whose address is to be suppressed from the public voter register or a voter who currently resides outside the home country)

Capturing hardcopy data

There are two ways to convert hardcopy data into electronic data. The first is to use data entry operators to type or key the data into a computer system. The second is to use optical scanning hardware and intelligent character recognition (ICR) software to convert images into electronic text. Both methods have advantages and disadvantages.

Manual data entry may be preferable to scanning because human operators are generally better able to interpret handwriting than ICR software. Human operators are also able to make decisions about voter eligibility that may not be able to be automated, such as deciding whether a signature looks acceptable or whether the information provided is sufficient. However, manual data entry can be a tedious, monotonous task and lack of concentration can lead to mistakes.

Several measures can be taken to increase the accuracy of manual data entry. A common method is to require data to be entered by one person and then verified by a second person. This verification process can take the form of keying all data twice, keeping both electronic copies separate. The two copies are then electronically compared. If they are both the same, the record is accepted. If they are different, a supervisor can check the record against the original to ensure the record is correctly keyed. Another verification method is to have a second person check the data keyed by the first person against the original form.

Another measure is to design the input screen used for data entry to maximize the accuracy rate. For example, a data entry screen should follow the same logical order as the form being keyed, with design elements used to force the operator's eyes to follow a logical path.

Software can also be programmed to perform logic tests as data is entered to minimize errors. For example, “input masks” can be used, so that only numbers within a specified range can be added in a field where a number is required and only valid dates can be added in date fields. Software can force data entry operators to add valid data to every field, so that fields cannot be skipped or left blank by accident. Where data in a field must conform to a particular standard, such as a defined list of variables, software can reject any entry that does not conform to the standard. Better still, where data in a field must conform to a defined list of variables, the system can offer only those variables, often in a “drop down box” or a “list box”. For example, a gender field could allow the operator to only select “male” or “female” as options.

Where data in the voter component of a voter register database is linked to another part of the database, such as the address component of the database, software can force data entry operators to select only a valid address from the address database. Any address given by a voter that does not conform to an address in the address database is rejected by the system, thereby forcing the operator or a supervisor to investigate the legitimacy of the claimed address.

In some cases, the address given may be an unofficial variation of an official address. In others, the address may be fraudulent. If the address turns out to be legitimate but it is not contained in the address database, a separate process should be undertaken to update the address database before the voter registration can be processed, thereby preserving the integrity of the address database.

Data entry using optical scanning hardware and ICR software to convert images into electronic text may be preferable to manual data entry where large quantities of data have to be processed and the process of manual data entry is not likely to add enough value to the process to make it worthwhile.

The biggest drawback with using ICR for data capture is the level of accuracy achieved. Since hardcopy voter registration forms tend to be handwritten, the varying qualities of handwriting can make it difficult for ICR systems to accurately convert handwriting into text, particularly where names are being interpreted, as they do not give ICR software predictable grammatical patterns to follow. However, the accuracy of ICR software is continually improving, and error rates of modern ICR software are much lower than those achieved a few years ago..

ICR software can be effective if the accuracy of the data capture is checked by a human operator against the original form, in much the same way as data is verified using manual operators. This process can be streamlined by software that captures both interpreted text and a picture image of the original form and displays them side by side on screen for operators to check. This method removes the need to refer back to the original forms and means the checking process can be undertaken relatively quickly by a trained and experienced operator.

ICR software is well suited to capturing typed text and can be “taught” to understand various typed or printed fonts with high degrees of accuracy.

Capturing electronic data

Compared with capturing paper/hardcopy data, capturing electronic data is a relatively straightforward process. However, difficulties may arise if the data provided is not formatted in the same way as the data tables into which the data is to be included.

For example, an external agency may provide an EMB with a list of persons who are to be included on the voter register. The voter register will be set up so that data will be included in several defined fields, with each field referring to a particular type of data, such as a surname field. If the imported data does not contain information formatted in the same field structure, the data will have to be converted to fit into the desired structure.

For instance, an EMB may split voters' addresses into separate fields, such as apartment number, street number, street name, locality, state or district, post or zip code. Address data from the external agency might be provided in a “free field” format, that is, the entire address might be typed in one field, with no breakdown of the address into its component parts. In this case, some means must be devised to convert the imported data into the desired format. Unfortunately, this often can only be achieved by considerable manual intervention, making the electronic data exchange a more complicated exercise than it might be expected.

The solution to importing electronic data into a voter registration database is to coordinate data field structures with the agency supplying the data so as to ensure consistency. The best way to do this is to develop an agreed set of data structure standards that can be used across a range of agencies with similar data. Several such standards exist in several countries..

Problems of data structure standards should not arise where EMBs collect electronic voter registration data directly from voters. For example, electronic registration forms provided on the Internet or at computer kiosks can be structured to fit directly into the correct database structure, if practicable.

Electronic voter register data can also be collected by supplying field workers with portable data entry devices. Data can be downloaded from these devices by using removable disks, by connecting the devices directly to a computer or by downloading data over the Internet.

As with manual entry of data, software logic tests can be applied to data captured electronically to identify any possible data errors. For example, any data containing letters in fields that should only contain numbers can be flagged, and operators can investigate the problem and, hopefully, correct it, going back to the source if necessary. Similarly, any addresses submitted electronically that do not conform to the standard address database can be investigated and corrections made as needed.

Capturing data by telephone

In some cases it may be possible to allow voters to register or update their voter register details automatically by telephone, but the opportunities for this type of transaction are rare.

However, it may be feasible (electoral legislation permitting) to accept changes to the voter register by telephone. In these cases, the verbal message must be translated into an electronic form in order for it to update the electronic voter register. This could take the form of a handwritten or typed form completed by the operator, which is then keyed or scanned into the computer system. The advantage of this approach is that it leaves a paper audit trail that can be used to verify the legitimacy of changes to the register.

Alternatively, the operator taking the telephone call could update the register on screen. This has the advantage of saving time by eliminating the step of producing a hard copy record. In this case, the database should record that the change was reported by telephone, so as to leave an audit trail for the change.

Capturing information from returned mail

Where information derived from the voter register is used to address mail to voters that mail may be returned with annotations that may be useful for updating the register. For example, mail that is returned "not known at this address" can be used to annotate the voter register and (dependent on local legislation) either serves to remove the person from the register or to trigger action to investigate the person's right to remain registered.

In other cases, returned mail may indicate corrections to spelling of names or to addresses. This information could also be used to correct the register.

Depending on the type of annotation made on returned mail, the processing of capturing the data on the annotations can be automated to varying degrees. If outgoing address labels include an identifying bar code or identity number or code, that identifier could be used to simplify the data capture of any annotations on returned mail. If annotations fall into defined categories, then data capture of such information can be automated to a high degree. For example, mail containing identifying bar codes that state the voter no longer lives at the registered address could be separately categorised and run through a bar code reader, so as to record the relevant data in the voter register database.

Where annotations show corrections to names or addresses, an operator would be required to key the changes into the electronic register. This process could be speeded up by using bar codes or identity numbers to quickly bring up the voter's record for correction.

Functionality of data entry systems

Data entry systems used for inputting voter register data, from either hard copy or electronic sources, can be designed to perform a range of functions that will add value to the process.

A voter register is typically a continually changing entity, particularly where a continuous register is used. Even where a periodic register is used, changes must be made. A data entry system should permit addition of new records, amendment of existing records and deletion of records. A voter register can also (ideally) be designed to track changes to its records over time, so that a voter's registration history is available.

Voter registers can also be designed to accept data impacting on a voter's record from a variety of different sources, where practicable. For example, a voter's original record may derive from an application form completed by the voter. That voter's address details may be updated at a later date by data provided electronically by another government agency.

Voter Identification

Voter identification is required during two phases of the electoral process: first for registration to establish the right to vote, and afterwards, at voting time, to allow a citizen to exercise their right to vote by verifying that the person satisfies all the requirements needed to vote.

In most countries, the process of voter authentication and verification of identity is done manually, but some countries have implemented and others are experimenting with an automated or at least semi-automated method to verify the identity of voters and their right to vote. This also implies the existence of an electronic voter register.  Actually, the technologies used for voter identification at election time depend on the technologies used to establish the voter register.

Most voter identification technologies need to use digitalized voter information and may include the use of:

• smart cards that record a person’s personal information and even biometric data

• database management systems that store and manage digitalized data

• biometric information, such as finger print identification

Smart cards

Smart card technology permits the storage of digital information that can be updated and accessed with an inexpensive reading device that may or may not be linked to a computer network. The smartcard, itself, is a plastic card that resembles a credit card and contains a small chip, which includes memory and sometimes a microprocessor. Gold contacts connect the smart card to the reading device. Since it can store more data than a magnetic strip, a smart card can keep the voter’s relevant data, including biometric data, and can also store non permanent data, such as the polling station where the voter is supposed to vote, for instance. Encryption techniques secure the data, and the tiny processor, if it is there, allows the smart card to be programmed for different applications.

Database management systems

This technology enables the recording, storage and management of required voter data.

Finger printing recognition

While this technology is not new, the electronic methods of recording and recognizing an individual finger print advanced substantially during the last decade of the 20th century. Today, identification can be achieved in a few seconds with reasonable accuracy. As a result, the use of automated fingerprint identification systems (AFIS) that record, store, search, match and identify finger prints is rapidly expanding. AFIS can be integrated in a suite of applications that work together to provide a comprehensive fingerprint and palm print identification solutions to accommodate the needs of voter registration systems as well as voter authentication and voter identification systems used at voting time.

Geographic Information

Data held on a voter register can essentially be divided into two kinds: data related to persons and data related to geography. The geographic base of a voter register is a key element of the register.

The geographic base of a voter register can be used for a wide variety of functions. These include:

• recording the address of a voter on the register

• determining and recording the boundaries of electoral districts

• determining and recording which electoral districts voters are assigned to

• recording lists of habitable and non-habitable addresses (to ensure that voters do not fraudulently enrol for non-existent or non-habitable addresses)

• identifying cases where too many people are enrolled at an address (in order to identify fraudulent or out-of-date entries on the register)

• enabling targeted reviews of the register (by identifying addresses where no persons are registered or addresses with high turnover)

• facilitating reviews of the register (by producing outputs such as lists of names and addresses in street order to facilitate visits by review officers, or the mailing of review letters)

• producing lists of addresses in a variety of electronic or printed form for non-electoral purposes (such as mailing lists for other organisations or for verification of geographic bases held by other agencies)

• There are several different ways in which geographic data on a voter register can be stored depending on the type of voter register.

Name based voter registers

A voter register does not need a separate geographic base (consisting of files containing geographic data only) in order to be functional. A geographic base for a voter register could simply consist of the sum of addresses recorded for each voter on the register. This kind of register can be called a “name based” register, where addresses are contained as fields in records devoted primarily to individual persons (that is, one record per person).

A name based register, however, is an inefficient way to store geographic data. As most addresses usually contain more than one person, a database that stores address data under individual persons' names will contain many repetitions of addresses. Where many thousands or millions of names are stored, this practice will dramatically increase the size of data files and decrease the performance of computer systems accordingly.

A name based register also deprives an election management body (EMB) of the functionality of a combined list of addresses. Such a list can be used to identify anomalies, such as incorrect addresses and fraudulent registrations. It is also difficult to use a name based register to perform other electoral tasks based on addresses, such as redrawing of electoral district boundaries and reviews of the voter register.

Address based voter register

An “address based” voter register generally records addresses in a separate database file (or files) from the list of voters' names. A unique code is assigned to each unique address, and this code is used to link the list of addresses with the list of names on the register.

This approach has several advantages. It dramatically reduces the size of the files contained in the register, as each address only has to be listed once. (With a name based register, each address is repeated for as many people are registered at that address.) Assuming the address register is accurate, it should ensure that voters are registered for valid, habitable addresses. It also facilitates a range of other electoral tasks based on addresses, such as redrawing of electoral district boundaries and reviews of the voter register.

Geographic Information Systems (GISs)

The simplest form of geographic base is a relatively simple text file showing a list of addresses. This could take the form of a list contained in a name based register (see above) or a more useful address based register.

A Geographic Information System (GIS) is a more complex software product that combines text files of addresses with other geographic and spatial data such as maps, aerial photographs, building plans and “spot on the Earth” coordinates.

Developed originally in the early 1960s, GISs are computer-based tools for mapping and analysing data that contain a spatial component. GISs allow users to integrate information, create maps, visualise scenarios and solve complex problems that have a geographic element to them. GIS technology has been applied to a wide variety of enterprises, including environmental and natural resource management, infrastructure planning for transportation and utilities, governmental services and resource allocation, and military and defence planning.

GISs have several different applications for electoral purposes, particularly for drawing or redistricting electoral boundaries.

Geographic data collection and administration

A geographic base is only as good as the data it contains. The data used by an EMB will depend in large part on the data available in that jurisdiction. Some countries will have well resourced government or commercial organisations responsible for collecting and computerising geographic data. EMBs may be able to make use of these resources, if they are available. If these resources are not available, an EMB may have to develop its own collection of geographic data.

An EMB can develop a rudimentary geographic base by extracting addresses from its voter register as submitted by the voters themselves. Depending on the completeness of the voter register, this may be a good start. However, a geographic base derived solely from addresses submitted by electors will have several drawbacks. It will not be complete, as it would be unlikely that persons would be registered for every habitable address, even in countries where registration is compulsory. It is likely to contain errors, as voters often do not use the correct or “official” address of their residence when completing their registration forms. (For example, it is common for people to list unofficial place names as their local address, with the result that voters can give varying unofficial addresses for the same official address.) Perhaps most importantly, it will not be possible to use such a geographic base to determine whether an address shown on a voter registration is fraudulent, since there is no way to independently assess whether an address is accurate.

In some countries it will be possible to obtain or buy a comprehensive geographic base, or a series of them, from other government or private organisations. Local government authorities are often good sources of geographic data. These data sources can be combined into one geographic base for use for electoral purposes. Combining geographic data from a range of organisations can be a complex process. Using defined standards of geographic data is one way to reduce the complexities, but even then differences in use of names of addresses and variation in address numbering schemes can make combining of geographic data a time-consuming process.

Software can be helpful in this regard. Programs can be obtained or developed that will use matching techniques to search for duplicates and identify anomalies.

It may be necessary to conduct fieldwork to compile an accurate geographic base. Staff can be employed to travel around a jurisdiction to ascertain whether addresses listed from the geographic database are accurate, and to identify any missing addresses. The results from the fieldwork can then be entered into the geographic database.

After a geographic base has been established, it will need ongoing maintenance and administration. The data held in a geographic base will need to be updated regularly, as new dwellings are built and old ones demolished. Addresses also change when houses, flats or apartments are subdivided or added to. An EMB will need to keep abreast of changes to addresses by receiving details of changes from responsible government authorities, where possible.

In some cases, the first notice an EMB will receive of a new address might be when a voter attempts to register for the address. In these cases, the EMB will need to investigate the accuracy of the claimed address and attempt to verify it with the appropriate authority.

Geo coding and geo referencing

Every “spot on the Earth” can be defined in terms of geographic code, or “geo code” or “geo reference”. The most common such geo code is latitude and longitude. Several standard geo codes exist. An EMB can consult its appropriate government mapping authority to determine the standard geo code used in its jurisdiction.

Every address in a geographic base can be assigned a geo code. This is a useful practice as it can be used to facilitate data matching of addresses with other organisations that use the same geo code. It can also be used to plot addresses on maps and to allocate voters to electoral districts by calculating the geo codes encompassed within each district.

Geographic Positioning Systems (GPSs)

Geographic Positioning Systems are devices that can plot the precise location of the user, using satellite transmissions to define a “spot on the Earth”. For an EMB's purposes, GPSs can be used by field workers to determine the spot on the Earth of addresses for the purpose of geo coding addresses in a geographic base.

Special address categories

A geographic base used by an EMB needs to be able to allow for special address or voter categories.

Non-habitable or invalid addresses should be identified in a geographic base, to ensure that voters do not fraudulently register for such addresses and to ensure that the geographic base is complete and without any unexplained gaps. Such addresses could include parks, cemeteries and golf courses. Shops, factories and industrial areas may not generally be habitable addresses, but EMBs should be aware that such addresses may still be habitable, since owners, employees or caretakers may reside in them.

While most habitations will be houses, apartments and flats containing relatively small numbers of people, some habitations may legitimately show many people entitled to be registered at the one address. These include institutions, hospitals, universities, military bases, dormitories and prisons. EMBs will need to be aware of these types of habitations and take account of them when using software to identify addresses that appear to have too many voters registered at them.

Geographic bases will also need to account for vacant dwellings, including dwellings where no one resides and dwellings that are vacant in the sense that no persons are registered at them. For the purposes of reviewing the accuracy of the voter register, vacant addresses can be targeted to ascertain whether anyone should be registered at those addresses. It is generally desirable to include vacant addresses in the geographic base so that they can be identified as valid addresses should a person apply to be registered at that address in the future.

Special voter categories

A comprehensive voter register will be able to cope with voters who do not have addresses that fit neatly in a geographic base. Depending on the applicable electoral legislation, voters may be able to register without having a permanent address. These categories of voters might include homeless people, itinerant people who move frequently and have no permanent resident, and those people living outside the country or district who retain an entitlement to be registered. These categories of people will need to be assigned a notional geographic attachment with a particular constituency, to enable their eligibility to vote for a particular district to be ascertained.

Linking addresses with electoral districts

A geographic base will be capable of linking voters' addresses with electoral districts. This will enable the production of lists showing eligibility to vote for particular districts as well as facilitate the redistricting process.

When electoral boundaries are altered by redistricting, a process needs to be in place to update the register to reflect the new boundaries.

Addresses can be linked with electoral districts by assigning each address with a “spot on the Earth” indicator such as geo code, and defining each electoral district in terms of the “spots on the Earth” or geo codes included within it.

Where a unique geo code is not available for each habitable address, addresses can be linked to electoral districts by assigning each habitation to a small geographic unit, such as a census tract or a municipal authority district. Electoral districts can then be defined in terms of the smaller geographic units they contain in order to allocate habitations to them.

When taking the approach of using small geographic units such as census tracts to assign habitations to electoral districts, care needs to be taken where an electoral district boundary divides the geographic unit. In these cases, an EMB will need to determine on which side of the boundary each habitation is located in order to decide which electoral district should be assigned to the habitation. In some cases, this can be done using maps; in other cases a field inspection might be necessary.