Linking Political Disengagement and Process Assessments with implications for Political Representati - PowerPoint PPT Presentation

Loading...

PPT – Linking Political Disengagement and Process Assessments with implications for Political Representati PowerPoint presentation | free to view - id: bcedf-ZDc1Z



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

Linking Political Disengagement and Process Assessments with implications for Political Representati

Description:

Linking Political Disengagement and Process Assessments ... a matter of substantial concern across popular and academic political analysis ... – PowerPoint PPT presentation

Number of Views:19
Avg rating:3.0/5.0
Slides: 27
Provided by: ccar
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Linking Political Disengagement and Process Assessments with implications for Political Representati


1
Linking Political Disengagement and Process
Assessments (with implications for Political
Representation)
Dr Robert Johns Lecturer Department of
Government University of Strathclyde robert.johns_at_
strath.ac.uk
Dr Christopher Carman John Anderson Senior
Research Lecturer Department of
Government University of Strathclyde christopher.c
arman_at_strath.ac.uk
and
  • Paper presented at Impact of Electoral Reform on
    Democratic Engagement in Scotland Conference,
    University of Glasgow, 23 February 2008

2
Political Disengagement and Turnout
  • Political (dis)engagement is a matter of
    substantial concern across popular and academic
    political analysis
  • Political science has invested a great deal of
    time and attention in explaining why people
    participate (vote, donate, campaign, protest,
    wear badges, etc) and the circumstances that are
    more (or less) likely to induce participation
  • Particularly in the UK, since the 2001 election,
    what gets voters to the polls - and what keeps
    them away - is a prominent issue

3
Modelling turnout in the UK
  • Clarke et al. (2004) Pattie, Seyd Whiteley
    (2003) Bromley, Curtice Seyd (2004) test the
    standard theories predicting turnout finding
    mixed support for the different models
  • Most support for general incentives and
    cognitive mobilisation models
  • Little support for civic volunteerism and
    social capital models
  • Socio-demographic variables (except age) of
    limited utility in predicting likelihood of voting

4
Likely Voters
  • If we think of the trial heat polls that are
    being conducted in Texas, Ohio and Pennsylvania
    right now, we know that survey firms segregate
    respondents into various categories of likely
    voters.
  • Each firm devises a screen to put respondents
    in categories of how likely they are to vote.
  • For our purposes, lets think of three categories
    of likely and unlikely voters
  • Hyper-engaged (VOTERS) - people committed to
    voting and vote on a regular basis
  • Non-engaged, (NON-voters) - people who have
    permanently switched off and will not vote
  • Moderately-engaged, Fence Sitters - those
    individuals who may not always participate but
    feel that they should and will if the costs are
    low, they feel it matters and (we argue) think
    the processes are fair

5
The Influence of Process Perceptions
  • Previous research has established that in many
    contexts, perceptions of process matter to
    individuals
  • Tylers (1994, 2000, 2006) work on procedural
    justice has demonstrated that individuals are
    more (or less) accepting of outcomes that they
    disagree with if they think the processes by
    which that outcome was generated were fair (or
    unfair).
  • Hibbing and Theiss-Morse (2001) have convincingly
    demonstrated that people evaluate institutions
    and political outcomes not only according to
    policy but also according to process
  • People who see a smaller gap between the
    processes they want and those they think they are
    getting are more inclined to support the
    political system

6
Process and Turnout Models
  • The vast majority of turnout models do not
    include or consider process evaluations.
  • Is this a terrible oversight?
  • No - as a general rule process (or how fair
    or unfair) elections are is not at all salient
    and therefore does not provide any leverage in
    predicting why people may or may not turnout for
    elections
  • And in the UK context, election fairness can be
    (and is) tied to preferences for electoral
    systems rather than electoral administration

7
Process and Electoral Participation
  • There may be cases where electoral administration
    issues are suddenly catapulted to high public
    prominence and become salient concerns (e.g., US
    2000, Scotland 2007)
  • When this happens we expect that perceptions of
    the fairness of elections (whether they are
    administered in a politically neutral fashion
    that ensures that all votes have an equal chance
    of being counted) may influence the degree to
    which individuals (dis)engage with the political
    system

8
3 May 2007
  • Process and electoral administration suddenly
    matter.
  • Very high rates of rejected ballots across
    Scotland, with especially high rates in Glasgow
    and Lothians
  • Several possible influences on the absolute and
    relative number of rejected ballots (see Carman,
    Mitchell Johns forthcoming Electoral Studies)
  • The questions posed by journalists and pundits
    over and over How did the election debacle
    effect public confidence in elections? Does it
    matter for what the public thinks about
    elections? What will be the longstanding effects?

9
From the political science side
  • How durable are perceptions of electoral
    fairness? Do they attenuate over time?
  • Will perceptions of electoral processes have a
    lasting influence on political participation?

10
Scottish Election Study, 2007
  • Funded by the ESRC
  • ESRC provided additional funding for an
    emergency third wave to track public
    perceptions of the problems associated with the
    May elections
  • This paper uses data from Wave 1 (April, 2007)
    and Wave 3 (Dec, 2007)
  • Panel data
  • Wave 1 N1,872
  • Wave 3 N1,166
  • Survey administered by YouGov using Scottish
    Panel
  • Data available online http//www.scottishelection
    study.org.uk

11
Thinking back to the 2007 election to the
Scottish Parliament, how fair would you say it
was?
  • Freq. Percent Cum.
  • -------------------------------------------------
    --
  • not at all fair 131 12.17
    12.17
  • not very fair 320 29.74
    41.91
  • quite fair 496 46.10
    88.01
  • very fair 129 11.99
    100.00
  • -------------------------------------------------
    --
  • Total 1,076 100.00

12
Thinking back to the 2007 election to the
Scottish Parliament, how fair would you say it
was?
13
As a comparison
Q How fair would you say are/were the following
elections?
Summary Fair Westminster 76 Scottish
Parliament 75 May 2007 Election 57
Summary Unfair Westminster 24 Scottish
Parliament 24 May 2007 Election 43
T-tests show no difference between the
Westminster and Scottish Parliament vars but
clear differences between those and the 2007
election measure
14
Thinking of your own ballot in the Scottish
Parliament election, how confident are you that
it was counted and not rejected?
Just to provide a sense for some of the responses
  • Freq. Percent
    Cum.
  • -------------------------------------------------
    ---------
  • very confident 414 49.29
    49.29
  • quite confident 249 29.64
    78.93
  • not that confident 86 10.24
    89.17
  • not at all confident 59 7.02
    96.19
  • i didnt vote 3 0.36
    96.55
  • dont know 29 3.45
    100.00
  • -------------------------------------------------
    ---------
  • Total 840 100.00

Calculated for Respondents who claimed to have
voted in Wave 2
15
There is also disagreement about who is to blame
for the big increase in rejected ballot papers.
Based on the information you have heard, who do
you think was most to blame?
and since we are always interested in blaming
someone
  • Freq. Percent Cum.
  • -------------------------------------------------
    --------------------------
  • voters themselves
    266 22.81 22.81
  • the people who counted the votes
    20 1.72 24.53
  • the people who designed the electronic
    181 15.52 40.05
  • the scottish office at westminster
    252 21.61 61.66
  • the scottish executive at holyrood
    224 19.21 80.87
  • the political parties in scotland
    82 7.03 87.91
  • donít know
    141 12.09 100.00
  • -------------------------------------------------
    --------------------------
  • Total
    1,166 100.00

16
All things considered, what effect have the
problems in 2007 had on the likelihood of you
voting in future Scottish Parliament elections?
Did the debacle have any effect?
  • Freq.
    Percent
  • -------------------------------------------------
    ------------
  • makes no difference
    984 87.94
  • makes me slightly less likely to vote
    92 8.22
  • makes me much less likely to vote
    43 3.84
  • -------------------------------------------------
    ------------
  • Total
    1,119 100.00

What if we break this out by voters and
non-voters
17
By Voted All things considered, what effect have
the problems in 2007 had on the likelihood of you
voting in future Scottish Parliament elections?
T-tests show a significant difference between
voters and non-voters (t9.68, plt.000)
This, however, is a highly problematic question.
We should be very cautious in interpreting it.
We will get to another way of looking at this
in one minute. First, lets look at predicting
the how fair variable
18
Regressing Fairness Perceptions on
predictors(ordered logit with robust standard
errors)
19
Fairness Perceptions and Disengagement
  • To the main question what influence, if any, do
    perceptions of procedural fairness have on
    disengagement? (7 months later)
  • As a DV, use question Suppose that there was an
    election for the Scottish Parliament tomorrow.
    On a scale from 0 to 10, where 0 means very
    unlikely and 10 means very likely, how likely
    is it that you would vote in that election?
  • NOTE this was the first question asked in Wave 3
    - before respondents knew what survey was about
    so framing effects should be minimal or
    nonexistent
  • We control for most of the main theoretical
    models that predict turnout
  • Also control for whether R voted (self-report)
    and their self-reported likelihood of voting from
    Wave 1

20
Regression (OLS) of Self-Reported Likelihood to
Vote in future elections in perceptions of
election fairness and controls
21
Fairness of Election Effects
not at all fair 9.727098 not very fair
10.03731 quite fair 10.34752
very fair 10.65773
Predicted values
(all vars at mean)
22
What if we include an indicator for ballots with
abbreviated instructions (residence in Glasgow
and Lothians)?
23
Evaluations of Process Matter
  • It would seem that perceptions of the fairness of
    elections can matter under particular
    circumstances
  • They, so far, seem somewhat durable
  • After 7 months and controlling for a large number
    of possible predictors of likelihood of voting,
    still finding effects
  • At this stage difficult to know about 2011,
    though the 2011 elections will be framed by media
    in terms of 2007
  • In thinking about (and modelling) political
    disengagement, we must consider process
    evaluations

24
Linking Political Disengagement and Process
Assessments (with implications for Political
Representation)
Dr Robert Johns Lecturer Department of
Government University of Strathclyde robert.johns_at_
strath.ac.uk
Dr Christopher Carman John Anderson Senior
Research Lecturer Department of
Government University of Strathclyde christopher.c
arman_at_strath.ac.uk
and
Paper presented at Impact of Electoral Reform on
Democratic Engagement in Scotland Conference,
University of Glasgow, 23 February 2008
25
Likely to vote model what if we control for
perceptions of fairness of other elections?
Linear regression Number of obs 643
R-squared 0.4592
--------------------
--------------------------------------------------
--------
Robust wv3_likelye Coef. Std. Err.
t Pgtt 95 Conf. Interval -------------
--------------------------------------------------
-------------- pre_likelye .3834452
.0686361 5.59 0.000 .2486594
.5182309 general_fair .0382685 .0605513
0.63 0.528 -.0806406 .1571776 fl_elec200r
.2295544 .0799162 2.87 0.004
.0726171 .3864916 fl_elec200t .2110626
.0766651 2.75 0.006 .0605097
.3616155 fl_dutytov2 .2837141 .1080964
2.62 0.009 .0714372 .4959909
interest .2187815 .0842641 2.60 0.010
.053306 .384257 cons_likede
.0318502 .0314928 1.01 0.312 -.0299945
.0936948 lab_likedie .0136191 .0294918
0.46 0.644 -.0442961
.0715342 ld_likedise -.0085542 .0312375
-0.27 0.784 -.0698975 .0527891 snp_likedi
e .0011187 .0205207 0.05 0.957
-.0391794 .0414167 grn_likedie .0357805
.0297974 1.20 0.230 -.0227348
.0942958 women .1781971 .1307668
1.36 0.173 -.0785993 .4349935
univ -.2599585 .132248 -1.97 0.050
-.5196636 -.0002534 age .0115151
.0060896 1.89 0.059 -.0004435
.0234736 voted 1.03385 .3097567
3.34 0.001 .4255578 1.642142
stv_voter .1636895 .1604594 1.02
0.308 -.1514163 .4787953 moderates
-.6720626 .5325906 -1.26 0.207
-1.71795 .3738244 econ_imprv .0769253
.0750049 1.03 0.305 -.0703673
.2242179 _cons 1.096676 1.010463
1.09 0.278 -.8876441 3.080996 ------------
--------------------------------------------------
----------------
26
Predicting All things considered, what effect
have the problems in 2007 had on the likelihood
of you voting in future Scottish Parliament
elections?
Ordered logistic regression
Number of obs 689 Wald chi2(18)
127.93 (p.000) Pseudo R2 0.2769 Log
pseudolikelihood -225.988 --------------------
--------------------------------------------------
--------
Robust elec2007_it Coef. Std. Err.
z Pgtz 95 Conf. Interval -------------
--------------------------------------------------
-------------- pre_likelye -.0602812
.065398 -0.92 0.357 -.188459
.0678966 general_fair .6189786 .1353965
4.57 0.000 .3536064 .8843508 fl_dutytov2
-.3716754 .1575355 -2.36 0.018
-.6804393 -.0629115 toobusyvt_2 -.1815118
.142551 -1.27 0.203 -.4609065
.097883 interest -.3475863 .2048827
-1.70 0.090 -.749149 .0539764 cons_liked
e .0585767 .0575119 1.02 0.308
-.0541445 .171298 lab_likedie .1608751
.0622299 2.59 0.010 .0389067
.2828435 ld_likedise -.066492 .0830927
-0.80 0.424 -.2293506 .0963666 snp_likedi
e -.0202406 .0567665 -0.36 0.721
-.1315009 .0910197 grn_likedie -.1172026
.0731736 -1.60 0.109 -.2606203
.026215 women -.2955207 .3361094
-0.88 0.379 -.9542829 .3632415
univ .3102623 .2989478 1.04 0.299
-.2756646 .8961892 age -.0024317
.0132182 -0.18 0.854 -.0283388
.0234754 voted -1.493652 .4620464
-3.23 0.001 -2.399246 -.5880577
stv_voter -.0116598 .3697521 -0.03
0.975 -.7363605 .7130409 moderates
.7245931 .5133178 1.41 0.158 -.2814913
1.730678 econ_imprv -.1231783 .1926835
-0.64 0.523 -.5008309 .2544744
notconfvote .8660806 .3485735 2.48
0.013 .182889 1.549272 ------------------
--------------------------------------------------
--------- /cut1 -.6296758 1.161443
-2.906063 1.646711
/cut2 1.258297 1.143759
-.9834303 3.500024 ----------------------------
--------------------------------------------------
About PowerShow.com