Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Development and operation: major party voter databases: refereed paper presented to the Australasian Political Studies Association Conference.



Download PDFDownload PDF

Development and Operation: Major Party Voter Databases

Mr Peter van Onselen

School of Politics and International Relations

University of New South Wales

Dr Wayne Errington

Australian National University

Refereed paper presented to the

Australasian Political Studies Association Conference

University of Tasmania, Hobart

29 September - 1 October 2003

Peter van Onselen, Wayne Errington: Electoral Databases

Page 1

Abstract

Modern political campaigning is becoming increasingly professionalised such that in Australia today the major parties use information databases to assist with their various political activities. This article outlines the design and operation of the electoral databases of the Coalition and Australian Labor Party (ALP). More emphasis is given to the operation of the Coalition’s database, Feedback, with details of the ALP’s database, Electrac, introduced as necessary. The article also analyses the increased significance of political databases in modern political campaigning, and the ways in which databases complement other features of contemporary campaigns, such as the focus on party leaders. Through analysis of the variation and similarities between the major parties’ databases, this little-known aspect of Australian politics will be shown to be one of the most important tools in modern campaigning. Finally, the ethical considerations of political intelligence gathering will be discussed. While Australians have grown used to the idea that governments keep all manner of information about citizens on file, a political party is a private organisation. With the use of electoral databases a comparatively recent development, some discussion about their use and potential abuse with respect to an individual’s right to privacy, and the use of personal information in partisan campaigns, is well overdue.

Introduction

The stability of Australia’s party system is due in large part to the willingness of the

major political parties to adopt the latest policies, presentation and technology from

around the world (Weller and Young 2000: 157). An essential element of the ongoing

professionalisation of political campaigning in Australia is the gathering and

management of information by political parties. While there is no substitute for a

politician with a sound knowledge of their electorate, the rapid growth in

information technology in recent decades has allowed parties to comprehensively

record and store an enormous and varied range of information about voters, seats

and issues. Electoral databases have become an increasingly important source of

information for Australia’s two major political parties. While less than a decade in

operation, the major parties’ respective databases have become integral to the

operation of individual Members’ offices, and to the ways in which the party

organisations target their campaign communication to swinging voters in marginal

seats. As well, the databases identify potential party supporters, who can be

contacted for financial or other kinds of support to the candidate.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 2

Electoral Databases: Not-So-Secret Weapons

Members of the public contacting the office of their local Member of Parliament

would expect their details and concerns to be taken down by office staff, and are

rarely surprised to be told that their name and address is in a computer in the MP’s

office. How many such citizens would realise, though, that the details of their phone

call to their MP then forms part of a sophisticated national database aimed solely at

partisan political advantage? Both of Australia’s major political parties have such a

database.

In interview, Senator Robert Ray was not willing to disclose the details of the ALP

database. However he did acknowledge its immense value to ALP campaigning,

particularly in marginal seats.

Yes, we operate a database on constituents but I’m not going to disclose what it does or how it functions. I can say it is an enormously valuable campaign tool, as I am sure Liberal Party persons would suggest theirs is, too (Ray, 2001).1

The unwillingness of the major political parties to publicly discuss their electoral

databases is understandable. Well over a decade after their introduction, it is time to

shed some light on this crucial development in modern Australian politics.

The introduction of the databases of both major parties has co-incided with a period

in Federal politics where a great deal of attention, by the parties and the media, has

begun to be paid to a relatively small number of swinging voters, to complement the

national campaign. The ALP's database, Electrack, was developed in the late 1980s,

as part of Labor's ongoing efforts through it’s national secretary, Bob Hogg, to ensure

that the ALP was capable of maintaining electoral advantage through exploiting

incumbency, something it had failed to do previously. Originally entitled Polfile, the

ALP's database was a first in Australian politics. A database of voters was a logical

1 Campaign directors from both major parties have been similarly guarded about the details of their databases. See The Australian Financial Review. 1 February 1993.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 3

element of more general moves to centralise ALP campaigning in the national

secretariat. After commercial disagreements over ownership of the technology, the

ALP switched to its current Electrac system after the 1993 election (The Age. 24

August 1996). Ownership of Electrac is compulsory for Caucus members.

The development of the Liberal Party's Feedback database was part of a national

review of the Coalition's 1990 election campaign. In that election, it was generally

recognised that the ALP had out-campaigned the Coalition in key marginal seats,

allowing Labor to win the election without a majority of the two-party preferred

vote.2 A Liberal Party delegation led by Michael Wooldridge had studied campaign

methods in Britain and the United States in 1988 (Australian Financial Review. 24

March 1993). In the USA, the Republican Party led their opposition Democrats in the

development of targeted campaigning.3 Technical difficulties, and the differences

between the two political systems, prevented the wholesale importation of American

campaign techniques (as well as the relevant software). This, along with the federal

structure of the Liberal Party, saw it fall well behind the ALP in the development of

direct mail techniques. It was becoming increasingly clear, however, that the practice

of identifying and targeting individual voters was becoming significantly easier with

the increasing power of information technology.

The design and operation of electoral databases is fairly simple. Access to

commercially available information, Australian Electoral Commission data, and the

telephone directory provides the raw material of names and addresses of

constituents. That is where the hard work begins. The purpose of the databases is to

provide parties with information about the policy and voting preferences of

individual voters, and to collate this information in ways useful to political

campaigning. Their effectiveness, however, hinges upon the diligence with which

2 Coalition: 50.1%, ALP 49.9% two-party preferred. Source: Australian Government and Politics 2002. 3 The American databases now allow automated telephone messages from candidates, with separate messages for party loyalists (encouraging them to vote) and swinging voters. With compulsory voting and therefore less emphasis on getting out the vote, Australians have not yet had this inconvenience inflicted upon them.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 4

individual offices enter data (a decentralised process with variable adherence on the

part of MPs), and the way in which the information stored in the database is utilised

by the party and individual candidates.

The Coalition's Feedback program is automatically updated monthly with

information from the AEC roll. This process takes account of boundary

redistribution’s and adds and removes constituents who enter and leave the

electorate (allowing welcome letters to new constituents). The electoral roll data

contains every elector's full name and address, telephone number, gender, date of

birth, and occupation (an optional entry on electoral enrolment forms). An important

feature of the database is the transfer of constituent details with the AEC updates.

This effectively allows Coalition members to often have some briefing notes on new

electors moving into their electorate. Such information is not only valuable in relation

to individual record keeping, but it also helps the parties track demographic changes.

Office staff can add to the basic electoral roll data in two ways. Feedback employs a

kind of shorthand known as tagging. Constituents are tagged based on information

gathered through contact with the electorate office, local newspaper coverage (letters

to the editor provide good information about issues of interest to particular voters),

doorknocking and telephone canvassing. Feedback provides specific tags for voting

information (to identify swinging voters, strong or weak party identification), issues

of concern, any history of party donation, ethnic identity, and alternative contact

details.

Feedback allows voters to be tagged according to their interest in over 300 issues.

These tags may mark a general interest in heath, or a particular interest in the cost of

private health insurance. Tags can also be created to account for local issues (about

problems with particular roads or proposed developments) or new issues as they

arise (such as the debate over stem cells). Feedback has over 150 generic forms of

tags. These tags are easily viewed by office staff as icons as soon as they open the

Peter van Onselen, Wayne Errington: Electoral Databases

Page 5

constituent's file on the database, and thus provide an instant picture of the person

with whom they are dealing. Hidden tags allow MPs to make notes about

constituents that are not accessible by the central database. Seamless operation of the

database adds both an aura of professionalism to the office, as well as a personal

touch to dealings with constituents.

The second way of adding to a constituent's profile is to retain detailed information

about any contact with the office. Contact with the office may come through a letter4

addressing a particular concern, or as ‘off the record’ as an anonymous telephone call

to complain about a particular action of the government. In both cases the political

staffer is likely capable of adding information onto the constituents file. Staff are

trained to log all written correspondence into Feedback. In the case of anonymous

callers unwilling to give their names, the use of caller ID telephone technology

(although not universally available), allows staffers to identify the number the

constituent is calling from, and if that number is a home line it can be cross checked

with the Feedback system.

A summary of the new contacts is added to the database, so that the frequency and

nature of contacts is tracked. These general tags build up a picture of individual

voters and their suitability for party communication. It is this data that interests the

party organisation, and can be used by individual MPs to tailor letters to small

groups of voters. This method is both cheaper and more effective than an electorate-wide mail-out, as discussed below. If a letter needs to be composed to the

constituent, or to another party on the constituent’s behalf, this is done from within

the database, so that the letter stays on the constituent’s electronic file. However,

these attachments are unable to be downloaded by the central computer. Thus, the

party organisation only receives a summary of the issue involved as contained in the

tags or other notes.

4 Email correspondence is significantly less likely to generate a response from an MP, since much of it is generated through mass lists from outside the electorate.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 6

Feedback also allows the compilation of a community database. This database is

linked to the main database by connecting the files on individual members to the

organisation listed in the community database. This allows identification of voters

based on issues of potential interest (Chamber of Commerce members), identification

of potential donors, as well as a list of school or community groups within the

electorate who can be sent government information packages about such events as

ANZAC Day, or relevant government health and community programs. In

particular, MPs can direct these organisations towards grants programs aimed at

community groups for purposes such as regional development, community and

communications infrastructure. However, privacy laws prevent community and

public organisations from handing out lists of members, making the compilation of

the community database difficult. For this reason political electorate staff are

expected to scour the local newspapers for constituent affiliations to community

organisations.

A well-organised office (it is hard to estimate what proportion of MPs’ offices fit this

description, probably less than half) incorporates the database into all stages of

constituent contact. As soon as the telephone rings, or a letter is received, the first

response of the staffer is to find the constituent’s details in the database. The basic

information (name, address, date of birth, other members of the household) is

supplied by the Australian Electoral Commission from the electoral roll5. If that

constituent has contacted the office in the past, or been canvassed by telephone (even

if when living in another electorate), the staffer can see some basic information about

the constituent. This includes such things as political affiliation or leaning, their

occupation, membership of community organisations, and the issues in which they

have shown interest in the past. The design of the major parties’ databases differs at

this point but the goal is the same - to gather as much information as possible about

5 This information is not completely reliable. MPs have been known to receive angry phone calls from constituents bemused to receive letters of congratulation on their 90th birthday when they are only in their fifties.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 7

voters in a form useable to both the individual member and the party organisation. A

well-integrated electorate office information management system will also include

data about weak or swinging booths in the electorate, and a breakdown of Census

Collection Districts. The latter allows demographic information to be gathered from

the census about particular areas within the electorate. As well, since the Census

Collection Districts provide the basis for the geographic breakdown of all

Commonwealth statistic data collection, such information as areas of high and low

unemployment can be tracked.

Feedback is designed to be operated in individual member electorates (members and

candidates must purchase their own copy of the software). The limiting factor is that

feedback allows for logged entries that do not throw back to the central system -

thereby protecting members interests by allowing them to freely enter info they may

not want their name or their offices name linked to inside the party, such as unkind

descriptions of troublesome constituents. With pre-selection contests often bitter

affairs, there is an incentive for MPs to mostly tag their constituents in the section

that does not feed back to the central system (in case they get challenged in

preselection for example : they don’t want this person to have any benefits.). This in

turn has consequences for the level and quality of information gathered by the

central office. Further, where pre-selections end bitterly - either because a sitting

member is defeated or the sitting members preferred candidate is defeated on his/her

retirement, Feedback information is often removed from the system. This removal of

information can either occur simply where information has been added in the sealed

section not available to the central party, or with risk of central party reprisal by

removing information prior to a central load down and update, thereby wiping

previous entries and replacing them with the newly vacated information. An

example of these latter approaches occurring was following the pre-selection battle

for the Liberal held seat of Wentworth. The sitting member Thompson was defeated

at pre-selection by former Woollahra Mayor and State organisational President Peter

King. Thompson's office immediately began de-logging information off both the

Peter van Onselen, Wayne Errington: Electoral Databases

Page 8

central and local sections of the Feedback information. King thereby not only had

little to no Feedback information for the 2001 election, but his office had to begin the

task of re-entering such information from scratch after the election (Subject A, 2002).

The decentralised nature of the databases is therefore a source of weakness. New

members of parliament, more familiar with information technology and likely to

have worked with their party’s database in the past, are more inclined to use

Feedback. Some do so without much enthusiasm however, believing that office

resources can be better utilised. In future, even with the majority of MPs’ offices

using their database systematically as described above, each party will have

significant gaps in their database. For example, where a senator provides a presence

for a party in an opposition marginal seat, the public is less likely to contact a

senator’s office than their local MHR, limiting the comprehensiveness of the party’s

picture of that seat.

Further, members in their final term are unlikely to be motivated to make

comprehensive entries in the database in their final term. Electoral Commission data

can become out of date very quickly. Constant updating of the database is essential

given that as many as 48.2% of electors in any one seat changed their address from

1996-2001 (Australian Bureau of Statistics, 2001). Thus, electorate-specific information

is lost if a constituent leaves the electorate (even though generic information about

each constituent will follow them to their new address. Thus, new members often

begin their careers with little information on their electorate database even if their

hand-over from the previous member is comprehensive. A new member in this

situation has a strong incentive to have their office comprehensively oriented

towards quickly building up their database.

After each election, the Feedback State Audit comprehensively reviews the efforts of

each office in providing information through the database. Offices which are below

average in their identification of issues and voting preferences are strongly

Peter van Onselen, Wayne Errington: Electoral Databases

Page 9

encouraged by the Government Members Secretariat (GMS) to improve their use of

the system. The decentralised nature of the system also adds to the potential for

technical problems. If an office's database is, through crash or theft, unavailable for

any length of time, there is little the central office can do to assist in the compilation

of data while the office is offline (Blaemire 2001).

Government and Opposition

While both parties' databases have the same goal - maximising the precision with

which candidates and leaders can communicate with swinging voters - their

databases are slightly different. In the case of the ALP, the database system is run as

a collective, however candidates need to access the system through an existing

Senator. The weakness of this system is that Senators rarely devote their time to

unwinnable seats, the unwinnable seat candidate is therefore largely prohibited from

enjoying the benefits of the system.

The overall structure through which the databases are organised differs between the

parties and also depends on whether or not a party enjoys the resources of

government. Databases undoubtedly afford advantages to incumbents, since while

opposition candidates may have access to their party’s database, pre-selection may

take place a matter of months before an election. In any event, constituents are more

likely to take their concerns about a specific problem (especially one that pertains to a

government department) to an MP rather than an opposition candidate. MPs are also

in a better position to ingrate their office staff with the database than are opposition

candidates who rely to an overwhelming degree on unpaid volunteers. Opposition

MPs are also largely precluded from significant database entry. Shadow Ministers

only receive one additional staffer on top of their electorate staff - therefore using

their electorate staff to assist with portfolio issues. Backbench opposition MPs are

expected to devote staff to neighbouring marginal seats, again detracting from their

capacity to devote time to the party database. Each of these examples illustrates the

amount of manpower available to effectively utilise the party database from

Peter van Onselen, Wayne Errington: Electoral Databases

Page 10

opposition. It is particularly important in opposition for the coordination between

individual candidates and the central office to run smoothly, since candidates

without the resources of an MPs office are heavily reliant on outside assistance to

campaign effectively.

Both parties open access to the databases to pre-selected candidates as well as

members. Candidates in both winnable and unwinnable seats are afforded access.

The inhibiting factor for each party is derivative of their ideological roots. In the case

of the Liberal Party, Feedback is purchased on an individual basis. Cost is incurred at

the point of sale, and again at the point of election. This process effectively lowers the

costs for unwinnable contesting candidates. The party does this with a view to

encouraging such candidates to purchase the system to improve voter identification

in such electorates. This view is held by the party for a number of reasons. Boundary

changes may increase the importance of the information within a particular area;

electors may move and the AEC information transfers a particular elector into a

marginal or safe seat of the party; demographic changes may see the seat in question

shift to marginal seat status; for the purpose of senate votes to assist in securing the

number three senate ticket position; as a means of encouraging the local party

organisational wing to continue to use feedback for either state or local government

elections, again as a means of boosting the parties presence in a particular area.

Nevertheless the start-up cost as well as the ongoing cost to receive updates are

significant (over $1000 for a copy of the software) and therefore an inhibiting factor

in itself.

In the case of both major parties, their Senate team plays an extensive role adding

value to the databases uses. In the case of the ALP, the Senators run the party’s

database with a regional focus, designed to mirror the party’s approach of spreading

its senate team across marginal seats during election campaigns. The Coalition

senators, whilst involved in the databases usage, are not maximised in their value.

‘Duty Senator’ is a term used by both major parties to describe lower-house

Peter van Onselen, Wayne Errington: Electoral Databases

Page 11

electorates that their senators are responsible for assisting (van Onselen 2001). Major

parties allocate their own seats, as well as seats held by the opposition (usually

marginal), to their state Senate teams to assist the local member or candidate and

party branches (Payne 2002). It is a natural consequence that in the small states,

where senator representation per MHR is higher, the senators play a more significant

role in the operation of the database. Western Australian Senators usually take

responsibility for one seat (Bishop 1999). In Tasmania, where there is no Coalition

lower-house representation, the five Liberal senators act as quasi-representatives of

the five House electorates. While in Victorian and NSW duty Senators are

responsible for a number of seats, and cannot be expected to do as thorough a job as

their interstate counterparts.

An MPs office staff can only assist in the compilation, running and expansion of the

database when operating out of their member’s office, although Senators usually

have feedback set up in their offices for the duty electorates they have been assigned.

They can therefore enter information and check data already on the system entered

by the local member. The central office-electorate office division described above also

limits the extent of help the Senator offers the local member, primarily through

limiting the extent of information sharing. This factor in part acts as a causal reason

the Coalition Senators out-source staff to lower-house members more frequently than

do ALP senators.

Factional disputes may also limit the level of information sharing between a senator

and their duty electorates. For example, Senator Marise Payne, a prominent Liberal

Party moderate, is the duty Senator for western Sydney, taking in the Liberal-held

seats of Lindsay and Parramatta, occupied respectively by Jackie Kelly and Ross

Cameron, members of the Liberal Party’s Right faction. The factional divide thus

limits the incentive for the MHRs and the Senator to fully share information through

the feedback system. The extent to which this disincentive is played out in practical

terms is difficult to quantify. For the central office, it is a constant struggle to co-

Peter van Onselen, Wayne Errington: Electoral Databases

Page 12

ordinate the sharing of information in this regard. It also displays the importance of

duty senators and candidates in un-winnable seats involving themselves in the

feedback process. Failure to do so can result in large sections of electorates at all

levels not gaining the coverage required to stimulate adequate correspondence to

maximise the party’s electoral prospects.

Senator John Tierney, in interview, raised the fact that his operating of Feedback out

of the Hunter Valley area he serves as duty senator for, has provided new candidates

with strong constituent information for doorknocking, direct mail-outs and general

canvassing of the issues that concern local residents. He pointed out that from when

he entered the parliament and moved his office to the Hunter area, the Liberal Party

have picked up the seats of Paterson, Dobell and Robertson, where previously the

Liberal Party was unrepresented in the Hunter area at state, federal or even

senatorial levels (Tierney 2002). It is impossible to quantify the level of effect

Feedback had in the securing of lower-house seats in the Hunter region for the

Liberal party. Nevertheless, it is interesting that the duty Coalition senator based in

the Hunter, operating as the only sitting member of parliament involved in the

campaign, has identified Feedback as a vital cog in the campaign.

Feedback became fully operational just in time for the 1996 campaign. The Coalition

has therefore enjoyed the use of government resources in the compilation of its

database, and in turning the database to political advantage. The government, by

definition, has a numerical advantage in that it occupies more seats (and usually

more marginal seats) than the opposition, providing additional staff (including

ministerial staff) and resources to enhance the database. The government can also

make use of ministerial staff to contribute towards maintenance of the database. The

Coalition has also had, both in government and opposition, more Senators than the

ALP in recent decades, adding to the benefits of incumbency.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 13

Usage Across Tiers of Government

Both Feedback and Electrac are open to both state and federal members to use.

Theoretically, given that both major parties operate at all three levels of government,

including some local councils, party databases offer assistance to all three levels. In

practise resource limitations prevent local councillors from greatly utilising the

database systems. Such representatives are less ordered in their logging of

information. Most councillors do not even have access to the software, much less

understand how to use it effectively. In the limited instances where local councillors

use party databases they do so more as a way of gaining background information on

constituents. Occasionally they will log information back into the system. Such

actions require councillors to either procure the software themselves, use the system

in a state or federal electorate office, or simply call the information through to staffers

in such an office. Some major party councillors double up as political staffers in state

or federal member’s offices, within which area their council is located. In the case of

Feedback electoral conferences at the organisational level of the party can purchase

and retain the database (this is not the case with the ALP’s Electrac). This generally

only occurs in conferences unrepresented at state and/or federal level. This

represents a final access point for councillors, however again resource limitation is a

hindrance of usage.

Differences between state and federal operation of databases are primarily driven by

the party’s incumbency at the time. As highlighted in the previous section,

incumbent governments have far greater capacity to effectively operate the database

system. The major difference in operation of databases when comparing state and

federal levels, barring issues of incumbency, is the more limited staffing pool at the

state level. Federal electorate offices are staffed by three fulltime ‘Electorate Officers’

plus a casual staffing allowance. State electorate offices are generally staffed by no

more than two electorate officers (in some cases one) plus a small casual allowance.

Although state electorates are considerably smaller, the loss of the additional staffer

does reduce an office capacity to utilise the database system. In the case of Feedback,

Peter van Onselen, Wayne Errington: Electoral Databases

Page 14

given that it is primarily operated out of the federal wing of the party (GMS) and that

it originated as a result of federal MP investigations, means that there has been a

greater emphasis and training for Feedback at the federal level when compared with

the state parties. This variability of emphasis on Feedback has cut across incumbency

stages at state levels. It remains to be seen whether this will alter when the federal

Liberal Party slips into Opposition. In the case of the ALP there was a marked

transfer of control from federal to state with regard to its database system when the

ALP lost Government at the federal level in 1996 (most particularly in states with

Labor Governments).

Feedback is a private company owned and operated by the Liberal Party, with an

independent staff, however one that can only offer adequate assistance with the use

of the GMS. The Feedback organisation provides training in database management

and telephone support to Coalition staff. This support is currently mostly provided

by the Government Members Secretariat, a taxpayer funded organisation that is

overseen by the Department of Finance. There is thus a good deal of public subsidy

involved in the maintenance and effective operation of political party databases. As

well as giving the government an advantage over the opposition, the level of

resources required to make the databases effective helps to entrench the two-party

system, despite a political environment favouring minor parties in ways outlined by

Marsh (1995).

Members that have managed to be elected without Feedback, often in marginal seats,

are sceptical of the value it adds. The new crop of party members, combined with

MPs in the lower and upper houses with central party experience as part of their pre-parliamentary backgrounds are likely to be more supportive of the new systems6. A

permanent factor political databases need to overcome is the inertia of members in

safe seats. While their want to maintain a solid margin for the purposes of both pre-selection and to avoid the seats slipping toward the opposition is always present,

Peter van Onselen, Wayne Errington: Electoral Databases

Page 15

such members are generally more sceptical about Feedback. The role of Senators in

the Feedback process is therefore important for such seats - particularly where

maintaining the senate vote is concerned. This is an inhibiting factor re senators

performing duty and patron roles in their allocated marginal seats.

The Target: Identifying Swinging Voters

Databases are used both by individual members and party organisations in their

campaigns. The strategy was summarised in the 2002 Feedback review of the

previous parliamentary term: ‘…informing voters of Federal Government initiatives

in an area where you [the MP] have these identified they have an interest. These

mail-outs may be small in number but will have a big impact in the long run’

(Feedback Document 1). Since its inception, Feedback has identified over 500,00

federal voting preferences, of which over 203 000 were identified as swinging voters,

and 75,500 as soft (weak or persuadable major party voters) or minor party voters.

So, on average, each Coalition user (an MP or candidate) identified 2, 204 swinging

voters (Feedback Document 1).

Once these swinging voters are identified, they are further broken down according to

the issues with which they have been tagged on the database. For example, an MP

may have identified about a hundred swinging voters who have indicated a strong

interest in the aged care portfolio. When the government (or opposition) makes a

policy announcement in that area, the GMS will send to each MP a generic letter

outlining the advantages of the policy. The MP can then send the information only to

those constituents identified as both swinging voters and interested in that particular

issue, gaining the maximum possible value from the postal allowance and other

resources expended. Further, constituents not interested in that particular issue are

not irritated by superfluous mail from their local MP. This strategy forms part of the

6 New MPs are increasingly likely to have experience in the party bureaucracy (van Onselen 2000).

Peter van Onselen, Wayne Errington: Electoral Databases

Page 16

continuous campaigning in which the major political parties currently engage,

serving as one aspect of incumbency advantage7.

The weakness of gathering information from contact initiated by the constituent is

that while this provides Feedback with issues, depth of concern, and organisational

affiliation, it is unusual for such contact to reveal voting preferences. For example, an

ALP voter is unlikely to reveal their voting preference when seeking assistance from

their local Coalition MP. Instead, voting preferences are more systematically

gathered through telephone canvassing or door-knocking. It is for this reason that

while the parties have a significant number of voters identified on an issue basis,

such identification does not readily translate into swinging voter understanding.

MPs are able to use issue-based knowledge to design issue-based direct mail-outs.

However, awareness of swinging voter status is significantly more valuable in

targeting a mail-out because costs can be reduced. It should be noted that where

sitting members are concerned the issue of cost is primarily one of the tax-payer

funded allowance mail-outs. The costs incurred are a combination of printing,

postage and the time consuming tasks of mail-out preparation. The first two forms of

costs are carried by taxpayer-funded allowances. MPs are conscious of keeping such

costs within a reasonable frame to avoid unwanted attention at Senate Estimates

hearings. It is the third cost of mail-out preparation that engages the political staff of

an MP, thereby limiting the alternative functions that they could be performing to the

benefit of the local Member. It can therefore be seen that on all accounts the MP has

an interest in keeping costs to a minimum.

Information from the databases also allows telephone canvassing to be targeted

towards filling in gaps in the database rather than wasting calls on voters who

already identify with a political party. Party pollsters are also provided with

feedback information so as to gain better data through the conducting of their polls.

7 The contents of such letters embarrassed former ALP Member for Eden-Monaro, Jim Snow, when their contents made the local media. He had sent quite different letters to supporters of the local timber industry and

Peter van Onselen, Wayne Errington: Electoral Databases

Page 17

A strong ALP tag will ensure that the constituent receives no contact from the Liberal

member or candidate (and therefore probably re-enforce such voters' negative

perception of the party).8 Strong Liberal voters on the other hand, are targeted for

requests for donations, party membership, and volunteer help. This aspect of the

database is yet to be fully realised. Since elected as Federal Treasurer of the Liberal

Party Malcolm Turnbull has publicly spoken of the party’s need to engage Liberal

voters for political donations. Turnbull has suggested this is becoming increasingly

important in the face of ever equalling out of big business donations to the two major

parties (Sydney Morning Herald. 3 December 2001).

Similarly, once constituents with strong party identification are excluded,

doorknocking the entire electorate becomes a more realistic goal for the candidate.

When an area is chosen for doorknocking, if possible on the basis of its high

proportion of swinging voters, the database can provide, for each household, a

number of issues for discussion, as well as pointers for fruitful small-talk such as club

membership or occupation. The system can print out the list of households on one

side of the road at a time, providing the member or candidate with a list of any

contacts to the office the constituent has made, or any issue of concern they have

been identified as having. In such situations members typically doorknock with the

assistance of office staff or volunteers. The helper is thereby able to record

constituent particulars while the local member freely engages in discussion. The

constituent is likely unaware that the content of their discussion will be logged into a

party database which will follow them as long as they remain on the electoral role.

Targeted Communication and the Nature of Political Representation

While electoral databases undoubtedly assist in the professionalisation of marginal

seat campaigning, these comparatively unknown information systems raise a range

of questions about marginal seat campaigning, and its advantages and disadvantages

conservationists before the 1990 election (Australian Financial Review. 2 March 1990).

Peter van Onselen, Wayne Errington: Electoral Databases

Page 18

for the operation of democracy. Do databases contribute to the marginalisation of

large numbers of voters on the basis that they can be identified as strongly

supporting a political party? Does the targeting of campaigns towards swinging

voters skew public policy towards the wants of a tiny minority of the electorate?

These questions strike at the very heart of representative democracy.

Whether or not one approves of the use of the database to tailor political

communication depends largely on one’s view of what political leadership is all

about in a democracy. A major party candidate for office has an enormous amount of

information about voters: their position and strength of conviction on a range of

issues, occupation, membership of political and community organisations. Leaving

aside the ethical considerations of the use of the information in the databases

(discussed below), the question of the way in which it should be used depends on

whether representative democracy is better served through MPs who closely reflect

the views of their electorate, or those who seek to lead public opinion toward more

effective and fairer policies. Applying these well-known delegate and trustee models

of representation to Australia, Emy commented that despite many politicians from

both major parties styling themselves as delegates, the most common type of

representation in Australian politics is what he called the ‘politico.’ This is the term

Emy used to describe the attempt to balance local representation with the need of the

party to win office at the national level (Emy 1974: 482). Contemporary studies of

representation in Australia have closely followed the typology established by Emy

(see, for example, Sawer and Zappala 2001: 5). Clearly databases assist politicians in

fulfilling this balance of local and national concerns. As described above, candidates

for individual seats can highlight those aspects of party policy of most interest to

voters in their seat.

8 A small number of other constituents will be tagged as 'no future contact' because they have proved to be unreasonable in their dealings with the office, or in personal dealing with the member or candidate.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 19

On the one hand, political parties have a wealth of quantitative and qualitative

information on public opinion with which to assist in the formulation of policies. On

the other hand, the system is designed to assist the major parties in skewing this

policy formulation process toward the views of a smaller and smaller number of

electors. A positive way of putting this is that public resources are not wasted trying

to persuade voters who have no intention of changing their vote. In reality, however,

a growing majority of campaign resources are being targeted towards a small

number of voters.

It might be said that the down-side of this type of communication is that election

campaigns are increasingly being fought over the votes of a smaller and smaller

number of voters. Indeed, the guidelines for the use of Feedback suggest a rule of

thumb for the entry of information about a constituent: 'Always ask yourself while

tagging information "Is this information going to be useful in a campaign?'''

(Feedback Document 2). However, databases may assist to allay concerns about the

representation of ‘oppressed groups’ (Phillips 2001: 30), since databases are compiled

from contact with self-selected constituents who invariably have some sort of

problem with government services. Instead, the database marginalises voters

wedded to the major parties.

As suggested by the name Feedback, the purpose of political databases is to allow

communication between politicians and the electorate. The fact that these databases

have been invented to serve the interests of political parties should not blind us to

the possibility that they may serve a wider public good, particularly in the efficient

way they transmit information between MPs and thousands of constituents. The

number of swinging and weakly identifying voters has been steadily increasing since

the 1960s. In 1967, only 11% of voters failed to identify with one of the major political

parties. That number had increased to 30% in 1990 (Chaples 1997: 358). Further,

while the actual number of swinging voters (or those weakly identifying with

parties) in each seat may be relatively small, finding out exactly who they are is not

Peter van Onselen, Wayne Errington: Electoral Databases

Page 20

easy. Voters may be reluctant to tag themselves as strongly identifying with an

opposition political party to that of the MPs office. This limits the number of

constituents totally excluded from consideration by candidates through the database.

While the number of swinging voters actively targeted by marginal seats campaigns

is relatively small in order to save on mailing costs, these people serve as proxies for

many other voters with similar concerns, who may have had no contact with their

MP's office.

The targeting of political communication at the electorate level serves to complement

the message of the central party machine, where the majority of resources (in media

advertising) are spent. This point is underlined by another contemporary trend in

modern campaigning, the emphasis on the party leader. Indeed, with media

coverage of politics increasingly centred upon the activities of party leaders, the

tightly targeted messages described above are becoming more important. As much

time as the present Prime Minister spends talking to radio talkback callers, he can

only speak personally to a limited number of voters. The combination of emphasis on

the leaders, and reduced time given to broadcast news media coverage of politics

creates a less politically-literate electorate. Many voters would be hard-pressed to

name a federal politician aside from the Prime Minister, Opposition Leader and

Treasurer. This impersonal trend in political campaigning thus finds its polar

opposite in the targeted, personal political communication facilitated by political

databases.

Databases also provide the aggregation of electorate-wide data so that the most

important issues for the electorate can be readily identified. This system allows the

early identification of burgeoning issues, such as voter concern about particular

legislation or local issues. This information is of interest to the party organisation and

parliamentary leadership as they make judgements about the popularity of policy

and legislative proposals. Again, this efficient dissemination of information serves to

improve the functioning of representative democracy. Targeting political

Peter van Onselen, Wayne Errington: Electoral Databases

Page 21

communication towards specific groups (by sending letters on youth issues to young

people, seniors' issues to seniors etc.) simply makes that communication more

relevant to the recipient. From the point of view of an individual elector, the

databases ensure that their MP retains a list of issues in which the constituent is

interested, as well as a record of correspondence. This allows MPs to attain a

comprehensive and accurate picture of public opinion in their electorate, instead of

relying on a gut feeling of views in the electorate. Thus, despite their tendency to

allow political parties to ignore a substantial section of the electorate (those with

strong party alignment), there are a number of positives to come from the

development of political databases.

The Use and Abuse of Information

In addition to these wider questions about representative democracy, the use of

political databases raises ethical and legal questions of the handling of information

by political parties. Because political parties are private organisations, and because

the major parties have no interest in public scrutiny of their databases, Electrac and

Feedback have come under remarkably little scrutiny from parliament and the

media. The very fact that private information, such as a health problem, becomes a

small cog in a political campaign would no doubt upset many people were they

made aware of it. Indeed, fear of media coverage of a Big Brother-style database

ensures the subject is not publicly discussed by the parties. For example, instructions

for Feedback include 'ensure that constituents cannot read the computer screen if

Feedback is open.' (Feedback Document 2). Each of the major parties in Australia

uses a single database at state and federal level. Compared to the United States,

where the decentralised major political parties, as well as private campaigns for

ballot initiatives, have ensured the development of a lively political database

industry9, Australia's market is very small. Public consciousness of the databases is

therefore likely to remain low.

9 A number of companies promote databases to assist in political campaigning in the United States. See, for example www.aristotle.com

Peter van Onselen, Wayne Errington: Electoral Databases

Page 22

When constituents contact an MP, are they dealing with an officer of the parliament,

or a member of a political party? In order for their MP to assist with a problem with,

for example, a child support case, a member of the public may hand over information

they would prefer is kept private. In their role as a Member of Parliament

representing a particular citizen, the politician (or their staff) contacts the

bureaucracy (say, the Child Support Agency), or the relevant minister’s office10 to

discuss the constituent’s problem. In order to provide the highest level of assistance,

as much detail as possible is gathered about the problem. Just how much of this

detail is recorded on the database, and how much of that information is down-loaded

to the central computer, will vary from office to office, and from case to case. Of

course, party organisations are only interested in the raw data about voters, such as

their level of interest in a range of issues. Nevertheless, the potential for abuse of this

information is clear.

Commonwealth privacy legislation is designed to prevent the misuse of personal

information by private organisations (Privacy Amendment (Private Sector) Act 2000).

Such legislation mitigates against the collection of information without an

individuals consent (amongst other limitations). Political parties however are exempt

from such legislative requirements where their activities are ‘in connection with an

election, a referendum, or other participation in the political process’ (Privacy

Amendment (Private Sector) Act 2000). In effect therefore, party databases can

collect, log and use information in such a way, were the database not controlled by a

political party, as to contravene the Privacy Amendment (Private Sector) Act 2000.

Political party exemption from the aforementioned Act has been described by the

legal community as ‘a surprise’ given it had ‘never previously been raised during the

extensive consultations over the legislation’ (Dixon, 2001). Further, given that party

databases routinely collect information in a manner contrary to privacy legislation

10 Some government departments have staff assigned to dealing with such inquiries from MPs.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 23

(barring the fact they are affiliated to political parties), it is high time closer attention

were paid to the apparent conflict of interest in political parties determining where

and when they receive exemptions to privacy legislation.

One obvious reason why political parties rarely publicly discuss the databases is

their effort to keep the operation of the system secret from the opposition party (even

though the two systems are quite similar). Apart from the strictures of party secrecy,

though, there appears to be little thought given by either party organisation to the

ethical ramifications of the handling of personal information. Party databases are a

powerful and invasive political tool. Public opinion surveys both within Australia

and overseas consistently show that an overwhelming majority of citizens are

concerned about invasion of privacy (Dixon 2001:NP). Surveys also consistently

indicate high levels of voter distrust towards politicians. Political parties’ exemption

from privacy legislation therefore demands greater scrutiny.

Conclusion

The operation of political databases is central to the way in which Australians are

represented by the major political parties. To put it bluntly, the system allows the

major parties to treat voters who strongly identify with either major party,

particularly against their own, with contempt. For the major political parties, the

trade-off between the usefulness of the system and its role in the distortion of public

policy barely rates consideration. There is, of course, a wider debate within political

parties as to whether the general shift towards targeting marginal seats is corrosive

of the sound public policy-making required to maintain a solid nation-wide primary

vote. However, the sheer effectiveness of the marginal seats strategies of both parties

(particularly while in government) in recent decades ensures that tools such as

electoral databases are an entrenched part of the management of political campaigns.

On balance, the positive elements of electoral databases, chiefly a systematised flow

of information between voters and their representatives, have the potential to

outweigh the negative elements.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 24

Electoral databases would be much less effective were there not considerable public

resources devoted to their smooth operation, both in MPs offices and the backup

provided by the Government Members Secretariat. This aspect of their operation

alone justifies public scrutiny of their operation. It is therefore essential that the

operation of electoral databases is more extensively discussed, and that a set of

principles be developed by the parliament to ensure that privacy is respected.

Further, attention needs to be drawn to the problem of MPs gathering personal

information about their constituents, which is in turn used for party political gain. At

minimum, the public is entitled to know about the collection and intrusion into their

personal data by political parties. Once the public is aware of the operation of party

databases, the wider ramifications of data collection and use can be debated. Equally

the wider operation of political parties can thus be more fully discussed.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 25

References

Australian Bureau of Statistics (2001), Commonwealth Census.

Australian Government and Politics (2002). www.elections.uwa.edu.au [accessed 3.3.2003].

Bishop, Mark (1999), author interview.

Blaemire, Robert (2001), ‘Database Management: One Size Does Not Fit All,’ Campaigns and Elections, November, V22, no 9, pp. 45-46.

Chaples, E (1997), ‘The Australian Voters’, in R. Smith (Ed), Politics in Australia, Third Edition. Allen & Unwin, Sydney.

Dixon, T (2001), ‘Australia’s New Privacy Legislation’ at www.bakercyberlawcentre.org/Articles/Cyberspace_May_2001/Privacy_2001 [accessed 21.4.2003]

Emy, Hugh (1974), The Politics of Australian Democracy. Macmillan, Melbourne.

Feedback Document 1, Confidential letter to MP

Feedback Document 2, Confidential internal party report

Marsh, Ian (1995), Beyond the Two Party System. Cambridge University Press, Melbourne.

Payne, Marise (2002), author interview,

Phillips, Anne (2001), ‘Representation Renewed’, in M.Sawer and G.Zappala (eds), Speaking for the People: Representation in Australian Politics. Melbourne University Press, Melbourne.

Ray, Robert (2001), author interview.

Sawer, Marian and Gianni Zappala (2001), ‘Introduction: Representation in Australian Politics,’ in M.Sawer and G.Zappala (eds), Speaking for the People: Representation in Australian Politics. Melbourne University Press, Melbourne.

Subject A (2002), Confidential interview with Liberal Party Organisational Member.

Tierney, John (2002), author interview.

Peter van Onselen, Wayne Errington: Electoral Databases

Page 26

Van Onselen, P.M. (2000), ‘Elective Bicameralism and Party Representation: A Study of the 39th Parliament of Australia’, in Australasian Political Studies Conference Proceedings, Australian National University, Canberra.

Van Onselen, P.M. (2001), ‘Major Party Senators: A Study of Party Expectations’, paper presented at Australasian Political Studies Conference, Griffith University, Brisbane.

Weller, Patrick & Liz Young (2000), ‘Political Parties and the Party System: Challenges for effective governing’, in M.Keating, J.Wanna, P.Weller (eds), Institutions on the Edge? Capacity for Governance. Allen & Unwin, Sydney.