| 2.4 Deliberative opinion polling |
| |
| Overview |
| |
| Deliberative opinion polling
is the brainchild of the American political scientist James Fishkin, a professor
at the University of Texas 23 A national random sample of between 250-600 citizens
is brought together to discuss and debate a particular issue. Carefully
balanced briefing material is provided and the citizens are given the opportunity
to question competing experts and politicians. At the end of the process
and after much small-group discussion, the citizens are polled in detail.
The resulting survey offers a representation of the considered judgements
of the wider population. The first deliberative opinion poll was run in
1994 in the UK on the subject of crime. It was sponsored by Channel 4 and
The Independent. Since then Fishkin has run a small number of deliberative
opinion polls in both the United States and the United Kingdom. |
| |
| Taking the widely used opinion
poll as his starting point, Fishkin argues that the information they provide
is limited because individuals respond to questions that they may have given
little or no thought to. A deliberative opinion poll, on
the other hand, allows space for deliberation and debate before opinions
are elicited and recorded. An ordinary opinion poll models what the public
thinks, given how little it knows. A deliberative opinion poll models what
the public would think, if it had an adequate chance to think about the
questions at issue. Equally, deliberative opinion polls move beyond focus
groups because of their sheer size and increased statistical representativeness.
Most of the deliberative opinion polls to date have been run in conjunction
with television companies - Fishkin sees the interest of the media as fundamental
in creating both a more informed 'civic journalism' and a more informed
wider public. |
| |
| Description of procedure |
| |
| Stage one: setting up the
process |
| |
- An independent facilitating organisation
is commissioned by an agency (media company, public agency etc.) to
run a deliberative opinion poll on a specific topic.
- A number of committees involving relevant
stakeholders are set up to provide advice, oversee the process, select
witnesses and provide background information for the chosen citizens.
- A statistically representative sample
of citizens of between 250-600 is selected - typically citizens are
paid a small fee for participating.
- The opinions of the citizens are sought
before the process - this acts as baseline data against
which changes in opinion can be measured.
- Background material is provided to the
citizens
|
| |
| Stage two: bringing the citizens
together |
| |
- The citizens are brought together for
2-4 days to hear evidence, to question experts and politicians and to
discuss the issues amongst themselves
- Much of the deliberation between citizens
occurs in small groups. Here the citizens discuss what they have learnt
and devise questions for the witnesses.
- At the end of the process the citizens
are once again polled.
- A report is compiled by the independent
facilitating organisation - this includes the changes in opinion after
deliberation.
- The media is invited to most of the sessions
in order that the deliberative opinion poll can become part of a wider
process of public debate.
|
| |
| Case Studies |
| |
| United Kingdom |
| |
| Topics |
| 1994 'Rising Crime: What can
we do about it?' |
| 1995 The UK's future in Europe |
| 1996 The future of the monarchy |
| 1997 Voting intentions for the
British General Election |
| |
| |
| United States of America |
| |
| Topics |
| 1996 National Issues Convention
for Presidential candidate selection |
| 1996 Polls conducted by three
public utilities in Texas |
| |
| National Issues Convention (N
IC) |
| |
| The National Issues Convention
held in the United States in 1996 was part of a wider project which attempted
to increase informed citizen participation in the selection of presidential
candidates. The project was co-sponsored by all eleven of the United States'
Presidential Libraries, by the Public Broadcasting Service (a national television
company) and the University of Texas at Austin. A number of foundations
and companies also supported the event. To ensure neutrality and the participation
and support of a wide range of stakeholders, a series of committees were
established. The Review Committee reviewed the briefing material for accuracy
and balance; the Technical Review Committee checked the survey research,
sampling and questionnaire design; and the National Advisory Board oversaw
the whole project. |
| |
| The Convention was televised
on PBS and became the focal point of a wider project involving a series
of local forums which used the same materials and discussed the same topics.
The NIC offered an opportunity for a representative sample of citizens to
engage with each other and with experts and politicians on a range of subjects.
From an original sample of 1,534 households, 914 responded to initial
interviews (the baseline) and from this sample, 460 individuals participated
in the weekend Convention. This was a 36% response rate. It is interesting
that many of the participants at the National Convention would never attend
local town meetings. On race, gender, religion, income, political affiliation
and previous voting, the citizens matched both the baseline and census data. |
| |
| The Convention allowed citizens
the opportunity to move beyond simple political rhetoric into discussions
about substantive areas of policy such as foreign aid, the budget deficit,
education and training, etc. Over the course of the weekend opinions changed
markedly as citizens listened to evidence, questioned experts and politicians,
and deliberated amongst themselves. These changes in opinion did not conform
to ideological stereotypes. In the evaluation of the proceedings, 73% of
the participants gave the process the highest possible rating; high ratings
were also given for small group discussions. There was a high level of press
and media interest - most commentators saw it as a rare opportunity to see
and hear a representative sample of the public coming to terms with major
political and social issues. |
| |
| Public Utility Polling in Texas |
| |
| In Texas, as in other states,
the public utilities are required to consult their customers as part of
their Integrated Resource Planning. They are expected to seek their views
on how they should provide power to customers and the necessary environmental
and other tradeoffs that will occur. The standard approach is to conduct
a normal opinion poll, town meetings or focus groups, or a combination of
these procedures. It is widely recognised that all three of these procedures
are limited - standard opinion polls elicit comments on issues that the
public has little interest or knowledge; town meetings tend to be dominated
by special interest groups; and focus groups are too small to be representative.
Deliberative opinion polling was seen as a promising method of achieving
an informed public input into the policy making process and as such three
utility companies - Central Power and Light; West Texas Utilities; and South
West Electric Power - commissioned Fishkin in 1996 to run three separate
deliberative opinion polls involving a representative sample of their customers. |
| |
| The process had a significant
effect on citizens' views and although the three different polls were held
in different locations and with different witnesses, a sharp increase in
support for investment in energy conservation occurred in all three cases.
The political impact of the process has also been significant. The recommendations
of the citizens were followed in detail by the three utility companies and
the process was supported by most of the major stakeholders such as the
Environment Defence Fund. In response to the process, the regulatory authority
- The Public Utility Commission of the State of Texas - changed its rules
to require that in future a representative sample of the
utilities' customers should be consulted through a deliberative process
in which it receives accurate and balanced information. |
| |
| Summary of Issues |
| |
| Representativeness: A deliberative
poll uses a random sample of the population and therefore better represents
the population as a whole than small citizens juries or consensus
conferences. However, the status of the outcome still raises some difficult
questions since the participants are an unallocated group of representatives
of the wider citizen body. Furthermore, the sample is still not big enough
to ensure that particular minorities within the population will not be overlooked. |
| |
| Consensus: Deliberative
opinion polls do not aim at the production of consensus. They are therefore
a good means of measuring the diversity of public opinion. |
| |
