Remaking heaven in small-scale democracy: A randomised experiment to encourage contested elections...
-
Upload
marybeth-rice -
Category
Documents
-
view
214 -
download
0
Transcript of Remaking heaven in small-scale democracy: A randomised experiment to encourage contested elections...
Remaking heaven in small-scale democracy: A randomised experiment to encourage
contested elections and greater representativeness in English local councils
Matt Ryan, Gerry Stoker, Peter John, Alice Moseley, Oliver James,
Liz Richardson, and Matia Vannoni
What does Schattschneider’s heaven look like?
• Increased (equal) political participation
• We have seen plenty of experiments around enhancing main forms political participation… – Voting– Standing for office?– Communicating with a representative– Joining a movement/campaign
What do we know about recruitment?
• In the majority of cases a stimulus (communication) is needed to recruit a person to stand for office (Mcleod et al 1999).
• Recruiters tend to look first to relatively closed networks, and value people with the same characteristics that they themselves possess (Crowder-Meyer 2011, Brady et al. 1995; 1999).
• Logical to want confidence that the recruit will do a good job…close personal connection or often a family connection (Van Lieffringe, 2012)
• May also recruit from within any number of networks of ‘purpose’ where they can recognise that members have a shared identity (Lim 2008)…weak and seemingly innocuous relationships e.g. a friend of the family (Della Porta and Diani 2005)
…continued…
• Time-scarce…use crude heuristics – as proxy assume that education, income and labour-market position are related to political interest (Stromblad 2008).
• For recruit personal incentives key– Will I win? Will I make a difference? Can I gain
access/reward/status?
– Costs - Time to campaign/stand and do the job (Norris and Lovenduski 1994), being in the public eye (Lawless 2012)
Enabling intervention?
• Literature suggests interventions need to be centred on clear understandings of the fears and ambitions of potential recruits and practically enable capacity and confidence-building through training or reassurance mechanisms.
Coercive
Compulsory votingQuotas Other (downstream) affirmative action
Targeted mobilisation Training
Nudging Information
Facultative
Enhancing representation
Parish council• Rural Britain – well it can be cloudy like heaven
• Small communities (lowest tier of government for roughly 30% of the population of England) 9,500 parish councils and 95,000 councillors.
• 70,000 in habitants to less than 200 residents. 80% of parish council govern are as of less than 2,500 inhabitants
• Councillors are disproportionately old (only 8% under forty and just over 1% under thirty) and disproportionately white males (29% women and 4% black and minority ethnicities). *last census 2006
• Rarely hold contested elections = legitimacy problem.
May 7 Group: Hampshire, Northants, Suffolk, Surrey and Leics/Rutland
Sample
• 977 parishes• 818 clerks• 5 counties of Southern England• 51% (498) parishes received treatment letter
& briefing paper• 49% (479) parishes received control letter &
no briefing paper
Research design in a nutshell• Pilot in 2013-2014 to find out best practice > treatment for
2015• Randomization of 818 clerks to take account of shared
parishes• Sept 2014: Control got a general letter about recruitment• Sept 2014: Treatment got a letter & briefing paper from
researchers with inclusion of research findings + invited to training sessions with Stoker and John
• Nov 2014: training takes place (two crossovers)• Measurement of taking the treatment by a survey and
crosschecking websites• Outcomes measures by contested elections and seats, and
survey measures investigating the extent of activity to recruit candidates
818 Parish clerks randomised into 2 groups
410 clerks (representing 498 parishes) received enhanced letter,
background paper & invitation to briefing
session(treatment group)
408 clerks (representing 479
parishes) received
basic letter(control group)
Reminder about briefing sessions sent 2-3 weeks
before the events
Outcome measures:Was item of widening recruitment placed on meeting agenda? (Yes/ No)*
Were competitive elections held? (Yes/No)No of candidates; ratio of candidates to seats
Was a recruitment strategy created to encourage more candidates?How many/ what methods of recruitment were used to advertise vacant
positions? How many/ what methods of advertising were used to advertise
elections?
Did not attend briefing
415 parishes
Attended briefing
(2nd treatment)83 parishes
Experimental flow chart
We think there are some “elephants” in the room
• Elections cost money that could be spent on more practical things
• Asking existing councillors to recruit new councillors: Isn’t that a bit like asking turkeys to vote for Christmas
• Do “friends and neighbours” want to compete
Using social media• Identify groups
where residents interactonline.
• Not necessarily commensurate withParish boundaries
• Can highlight roleas a response toopinion on a localissue
Strong and weak ties
• The council tends to attract the same type of people because we are programmed to recruit through strong ties with people we know.
• It is worth considering ways in which we can target ‘weak ties’ – that is people we know of who may seem like good citizens but do not talk to everyday. One way of doing this is through a short ‘roadshow’ outside the school or supermarket where we have targeted conversations with people we might not normally ask.
Received letter Did not receive letter0
1
2
3
4
5
6
2.111.94
Candidate recruitment activity scores for treatment and control group
Candidate recruitment activity score = added item of widening recruitment to meeting agenda (1/0) + held event to recruit candidates (1/0) + advertised for
candidates in parish newsletter (1/0) + advertised for candidates on parish website (1/0) + advertised for candidates using social media (1/0) + advertised for candidates
using other media eg local radio/ newspaper (1/0)
Range 0-6
Received letter Did not receive letter0
1
2
3
4
5
6
7
2.86 3
Electoral activity scores for treatment and control group
Range = 0-7
Electoral activity score = added item of election to meeting agenda (1/0) + discussed election at council meeting (1/0) + held event to advertise election (1/0) + advertised
election in parish newsletter (1/0) + advertised election on parish website (1/0) + advertised election by poster (1/0) + advertised election by leaflet (1/0)
Received letter Did not receive letter40
42
44
46
48
50
52
54
5655
46
Placed an advert for candidates on parish or town council/ community website
Pearson chi2(1) = 2.8110 Pr = 0.094
%
Received letter Did not receive letter0
2
4
6
8
10
12
14
16
18
2019
11
Used social media to advertise for candidates
%
Pearson chi2(1) = 3.6324 Pr = 0.057
Received letter Did not receive letter0
5
10
15
20
25
30
1617
% of parishes where contested election held for treatment and control group
Received letter Did not receive letter0
1
2
3
4
5
6
7
8
9
10
7.9 7.8
Mean no of candidates for treatment and control group
Min = 0, Max = 44
Received letter Did not receive letter0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
10.908 0.919
Mean ratio of candidates to seats for treatment and control group
Marginally more candidates to seats in non-treated parishes, but non-significant
Mean difference = 0.011, t= 0.5093, p=0.6107. Range = 0-3.14
Other points
• If we make some assumptions it is possible to use a 2SLS model with training as the endogenous variable and random allocation as the instrument >>> no effect
• Might try CACE• Something odd with effects by county• Only treatment effect is the use of social
media - was stressed in the training sessions
Conclusions: why did it not work?
• Always realized it was a tough call because recruiting through the councils themselves, (but there was enthusiasm for the treatment in the sessions)
• We let the CALCs write to the control group – hence good survey results? But a short letter
• Treatment too weak?• What can be done in future? Too hard?
Resistance from incumbents Lack of salary/ payment
Having little in common with councillors Fear of being in the public eye
Fear of legal responsibilities Lack of experience in such roles
Reluctant to make decisions about local people known to them Red tape involved
Parish councils have limited power to make a difference Lack of knowledge about role
Being a councillor is time consuming
Perceived obstacles to candidate recruitment (view of clerks)
Role is time consuming (65%)
Lack of knowledge of parish councillor role (44%%)
Perception that parish councils have limited power to make a difference (44%)
341 responsesfrom 977 parishes(35% response rate)