Edemocracy experimentation: implications for research and teaching

1
E-democracy experimentation implications for
research and teaching

• Alexander H. Trechsel
• European University Institute

UNITAR State of the Art Workshop Session on
E-demcoracy Geneva, December 17 2007
2
Research and Teachingon E-democracy

• First prerequisite solid understanding of
  democracy tout court
• Second prerequisite theoretical approach to the
  study of democracy
• Whats new? -gt opportunity for using the object
  of our study as our tool
• For the first time the social sciences can
  conduct large scale experimentations

3
An example

• A study of democracy in Europe
• Focus electoral behavior in European elections
• Traditional approach European Elections Study
  (EES)
• Limits of the EES expensive, limited in its
  depth, increasingly difficult to get participants
• Still the EES is so far the only tool for
  measuring electoral behavior on a large,
  comparative scale in Europe
• Our research agenda add the e- to the EES in
  collaboration with the EES

4
Democracy in Europe

• Long list of challenges to democracy in Europe
  (Schmitter and Trechsel 2004, Kriesi 2007 etc.)
• In particular traditional mechanisms of
  representation break down
• trust ?
• electoral turnout ?
• party identification ?
• party cohesion ?
• party government ?
• guardian institutions ?
• etc.

5
Political offer

• Traditional cleavages ?
• Opaque policy positions ?
• Complex, multidimensional preference mash ?
• Fragmentation ?
• Unholy coalition-building ?
• Unpredictability ?
• -gt DIFFICULTIES FOR LARGE NUMBERS OF VOTERS TO
  IDENTIFY THOSE MOST LIKELY TO REPRESENT THEIR
  INTERESTS - gt DISAFFECTION FROM ELECTORAL POLITICS

6
The irony

• Voters get lost, but their political curiosity
  does not
• Thanks to ICTs opportunity for the social
  sciences to learn more about
• parties
• candidates
• public opinion
• political behavior
• campaign dynamics

7
The tool e-Profiling

• Started in the Netherlands in the mid.90s
  (Stemwijzer)
• Since spread to Switzerland, Bulgaria, Finland,
  Germany, Lithuania and many other places
• Ongoing research project on the Swiss case

8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
Match and mismatch
15
Some data from the Swiss case

• First trial in the Federal elections of 2003
• Candidates participation 50.3 percent
• Profiles 255000 (11.7 percent of voters)
• October 2007 Federal elections 2632 candidates
  profiles (85) and 841156 voters profiles
• Innovation questionnaire(s) added by us

16
Impact of smartvote (trial questionnaires on the
cantonal level)

• 67.2 the smartvote result was important for
  my decision
• 74.1 smartvote has influenced my decision
• 33.4 voted for unusual candidates

17
Additional features of smartvote

• Initial questionnaire (basic socio-demographic
  data, initial vote intentions, traditional voting
  behavior etc.)
• Smartvote profiling
• Final questionnaire (socio-demographic data,
  change from initial vote intentions etc.)
• Sampling according to known socio-demographic
  distribution and matching with profiles
• Cost very low (both candidates and voters
  provide the information for free)

18
The research agenda

• Target EP elections in May 2009
• Based on the 2007 Swiss smartvote research -gt
  development of a research project for profiling
  in the 2009 EP elections
• Very close coordination and collaboration with
  the EES

19
What can political science learn?1. On
political parties

• intra-party cohesion
• spatial distribution of parties
• - inter-party (EU-wide) congruence
• - EU vs. national level politics
• - comparison with manifesto data

20
What can political science learn?2. On Public
Opinion

• Mapping of detailed policy preferences within
  countries and across the EU27
• Mapping of saliency of policy preferences of
  citizens within countries and across the EU27
• Mapping of elite/citizen congruence (issue
  multidimensionality, European vs. national level
  issue salience etc.)

21
What can political science learn?3. On
Political Behavior

• Pre-smartvote and post-smartvote questionnaire -gt
  mobilization effect of smartvote? -gt importance
  of policy preferences vs. party identification
  model? -gt economic voting? -gt spatial models? -gt
  cleavages? etc.

22
What can political science learn?4. On
Campaigning

• Shifts in profile-matching over the campaign -gt
  campaign intensity, matching of campaign
  direction and profile matching
• Potential adaptation of party/candidate profiles
  when user-data is fed back -gt spatial models

23
What can political science learn?5. On
Methodology

• Experimental and novel data gathering on public
  and elite opinion, matching techniques,
  econometric models, online tools etc.
• Pre-vote online survey vs. post-electoral CATI
  survey (profiling vs. EES)

24
What can political science learn?6. On other
actors

• Idea have civil society organisations, national
  and EU elites and media exponents fill out the
  same questionnaire
• Match policy preference across actors

25
Next steps

• Decision taken to launch the project within the
  RSCAS EUDO framework
• Support from the RSCAS
• Currently technological solution scouting
  (development of international partnerships)
• Conceptual work (questionnaire, webdesign, media
  campaign etc.)

26
Training opportunity

• Need for an armada of researchers closely
  collaborating -gt EUI Ph.D researchers from all
  EU27
• Preparatory seminars on democracy offered at the
  EUI (in particular Mair/Trechsel seminar series)
• Data will offer very novel opportunities
• Development of tools for implementation in other
  electoral contexts

27
Thank you for your attention!

• Contact
• Alexander.Trechsel_at_eui.eu
• www.eui.eu
• www.eudo.eu