Cambridge Voting Behaviour Survey
Jim Jepps
Cambridge Voting Behaviour Survey
4th May 2006
Contents
1. Introduction
2. Caution
3. Comparisons to what actually happened
a. voter turnout
b. vote distribution
4. Affinity
5. Non voters
6. Summary by Party
a. Conservatives
b. Green Party
c. Labour Party
d. Liberal Democrats
e. Respect
7. Thanks
8. Appendices
a. the questions asked
b. tables of data collected
c. notes for guidance (survey team)
1. Introduction
This report represents the findings of independent research on voting behaviour in the local elections on the 4th May 2006. The primary data was collected on Mill Road, in the Romsey Ward of Cambridge.
The aim of this investigation is two fold. Firstly to gain more detailed information about how people use their vote and secondly to grapple with why non-voters did not vote.
Whilst, by necessity, the sample size was relatively small (256 respondents) it is large enough for us draw tentative conclusions. There are a number of indicators from this study to suggest the following:-
2. Caution with statistics, or how to evaluate this data
Statistical surveys of this kind should never be treated as scientific fact, even when the survey group is far larger than with our modest study. This is a snap shot in time of what people said to our research team in Romsey Ward on the 4th May 2006.
This doesn’t mean the results aren’t interesting.
In order to help with transparency we are publishing the collated data, prior to interpretation in Appendix B and where we have interpreted results we will reference the specific table in order to aid those who would like to double check our results.
Some specifics to bear in mind about our study
We conducted this survey on Mill Road which is in the Romsey ward of Cambridge. This ward was chosen because there were five parties standing in the election (Labour, Conservative, Lib Dem, Green and Respect) which would allow us to develop an insight into the relationship between the three parties of Parliament and some of ‘the others’. This relied upon the votes of all the parties, and both Respect and Green votes in particular, holding up which, thankfully for the study, they did.
This does not make it a typical ward, but a ward where there were more voting options.
Another important factor to bear in mind is that the Lib Dems won this ward decisively and the Conservatives did very badly. Therefore voting behaviour is influenced by the specific local factors and does not represent any kind of national trend, except in passing.
For reference on 2006 local election
Results in Cambridge Evening News http://www.cambridge-news.co.uk/news/region_wide/2006/05/05/f83963b5-9dc2-4ee9-8be5-ed9b450163a8.lpf
Analysis in CEN http://www.cambridge-news.co.uk/news/region_wide/2006/05/05/fe7389a5-7bb2-44ca-9188-e4350c51ff3f.lpf
For an analysis of the Cambridge elections with reference to Romsey
“Cambridge; almost interesting” at
http://socialistunitynetwork.co.uk/activate/Election2006/cambridge01.htm
3. Comparisons to what actually happened
There are a number of discrepancies between the results of our survey and the actual results in the elections. The two key differences are turn out and vote distribution:-
A) Turnout (Appendix B. Table 3)
The turnout in Romsey ward was 37% and in our survey 50% of our respondents claimed to have voted. It is fair to suppose that those who do not wish to vote are also less likely to want to take part in a survey about voting, and there may also be a more general correlation between those who will take time to take part in a survey (of any nature) and those who will take time to vote.
It’s quite possible then that whilst we still did get a large number of respondents who did not vote this might have influenced the stated reasons on why they did not vote. For example, it is possible that we had a higher proportion of people who feel passionately that people should not vote as those who don’t care at all about elections may have been less willing to take part in the study.
B) Vote distribution (Appendix B. Table 1)
Our study’s voter distribution does not represent the actual vote in the Romsey ward, in fact there are some interesting discrepancies.
This is not overly troubling as the study was never attempting to find out what the result of the election was to be (as the election itself is a fool proof indicator of this) but it is worth noting. It could be that those who voted Labour were less willing to admit who they were voting for, and those who voted for the ‘minor’ were far more vocal, or proud, of the fact.
It does not take a great leap of imagination to suppose that those who were voting for the party of government felt some embarrassment about that fact, particularly at a time of great scandals at the top of the party – and that those who were voting for the parties of anti-establishment passions were more willing to wear their colours on their sleeve.
Whilst we may feel this is obvious on one level – it is not an objective certainty that Conservatives are less ashamed of their affiliation than Labour supporters, or that those voting for ‘fringe’ parties are willing to admit that they do so. Had the BNP or UKIP been standing in this ward we might have found their supporters unwilling to admit their affiliation to us, whilst still casting their vote for these parties at the ballot box.
Of course that is speculation and would take another study in a different area to ascertain.
4. Affinity
One core finding in this study has been a clear correlation, or affinity, between the voters of particular parties.
The Lib Dems did not have a low affinity with the voters of any party. In other words even those who did not vote for the Lib Dems found that party the least offensive alternative option, and would have been most likely to vote for them had their own party not stood.
This may well be due to the fact that the Lib Dems get the highest vote in this ward and are therefore the strongest performers, so it is not surprising that those who vote for other parties are tempted by the Lib Dems. Alternative explanations could be that the Lib Dems are more well placed to the centre of the political spectrum, appear less partisan or indeed are the least ‘offensive’ of the five parties.
If you examine Table 5 in Appendix B you can see that all the parties (bar Respect who place them second) place the Liberal Democrats as their closest affinity.
It is also part of a correlation between the three parties with strong representation in Parliament with a correspondingly strong affinity between the two more ‘non-establishment’ parties (Greens and Respect). In fact it was only Green voters that did not place Respect as the party they have least affinity with.
This could be seen as voters split into essentially ‘establishment’ and ‘non-establishment’ affinity groups. For example, both Respect and Greens placed Labour and Conservative as the parties they felt most distance from.
If we look at which voters would not have voted had their party not stood there are some surprising results.
We found that of all the parties it was the voters for the Greens and Conservatives who were least likely to have voted if their party had not stood. It was those who voted for Respect who were most likely to have cast their vote for another party rather than stay at home.
There are two conclusions you could draw from this.
This is perhaps understandable as Respect is the newest of the five parties and has yet to establish itself as a consistent political entity in the area.
There could also be another factor in that Romsey ward was held for some time by Labour. It is possible that those defectors to the Lib Dems have yet to develop deep ties with the party and are more likely to deliberate over who they vote for rather than having an inbuilt assumption in the way that those who habitually vote for the same party over a period of years will have.
So whilst voters for all the parties (bar Respect) placed the Lib Dems as the party they had most sympathy with, outside of their first choice, this affinity was also the most thinly spread. So if we compare the proportions of those who claimed to have voted for specific parties with the proportion of their second choices there are some interesting results.
The largest shift is that those choosing the Conservative candidates as their second choice was double that of those choosing them as their first choice, with Greens and Labour also seeing proportional increases in their second choices. The popularity of the Conservatives as second choice in this ward could be due to tactical considerations as the Conservatives were always going to have difficulties winning this seat.
The Liberal Democrats saw a significant drop in support, possibly indicating that as they had recently won over votes from other parties in this ward the well was running dry of those who would ever consider voting Lib Dem.
Respect saw the largest drop between those who selected it as first choice and those who selected it as second choice, a drop of more than a third. This correlates with the way that voters of other parties felt least affinity with Respect. In fact no respondents who voted Labour or Tory said they would have voted for Respect if their party of preference had not stood, and almost all of their second preferences came from those who voted Green.
This would indicate that despite Respect achieving a very credible vote they polarise opinion. Whilst everyone would consider voting Lib Dem only those in the ‘anti-establishment sphere’ would consider voting Respect.
This is not the case with the Greens, who despite getting no second preferences from Conservative voters, were considered an acceptable alternative to voters from other parties.
5. Non voters (see Appendix B. Table 7 for reference)
One of the things we were keen to investigate in this study were the reasons why people do not vote.
By taking qualitative data from non-voters about their reasons for not voting we were able to indicate what the core factors were in effecting turnout.
We found that the largest section of people of voting age who did not vote were simply unable to vote at this election. The reasons for this ranged from the fact that they are not UK citizens, not registered (due to moving, being students, etc), there were no elections in their area to other similar reasons. This largest section of non-voters is essentially disenfranchised rather than apathetic and it could indicate that the electoral roll is not in a fit state to introduce compulsory voting (as some are discussing currently)
The second largest group of those who did not vote, choose, consciously, not to do so. Their reasons ranged from “They are all the same”, “It doesn’t make any difference”, “I don’t vote out of principle” and simply that they did not care about politics.
The third group did not vote because of lack of information. Many people did not know there was an election taking place, others had had no literature from candidates and did not feel they could vote for people without first assessing the candidates policies.
A significant group of non-voters were people who were not choosing consciously not to vote but simply did not see voting as a priority. Many of these respondents answered that they were too busy or didn’t have time to vote. One respondent told us that he did not vote because he works nights.
We can say though that whilst it might have been inconvenient for this group to vote if they had given the election more importance they would have taken the time to (particularly as the polls had extended opening this year of 7 am to 10 pm). This is not to say they are wrong to feel this way, only that these people do not feel sufficiently engaged with the political process to go out of their way to vote.
There were a small number of people who felt disenfranchised by the electoral system and were not voting because they felt that first past the post meant that their vote would be wasted in an area that was ‘sewn up’ by one political party.
If we choose to use these figures to make suggestions to increase turnout it would seem that the three most important tasks would be to:-
1. the electoral role needs serious attention if it is to adequately represent those who should be eligible to vote.
2. attempt to engage the electorate to help them feel their views make a difference (or perhaps, more radically, ensure voters views actually make a difference)
3. ensure that everyone knows that an election is taking place and that information about that election is widely available.
4. students are most likely not to be registered to vote at their university
town’s address. By either making a large push to register them to vote
postally, or to have dual registration so they can vote where they live during
term time would significantly cut the number of disenfranchised voters.
6. Summary by party
We have sorted the core information related elsewhere in this report into party sections for ease of use.
a. Conservative
In the Romsey ward the Conservative vote was ‘a long way from top’. However, many voters seem to have considered voting Conservative as their second preference, far more proportionally than any other party.
This may well indicate a number of voters did not vote for the Tories on the basis that they were unlikely to win, rather than because they oppose their political positions. This tactical voting appears to have primarily benefited the Lib Dems, the clear winners in Romsey ward.
Unsurprisingly the Conservative are placed well within the ‘establishment sphere’ where their voters feel more affinity with the other parties with strong representation in Parliament (including Labour) than they did with the Greens and Respect. In fact not one respondent who voted Conservative placed either of the ‘non-establishment’ parties as their second preference.
Perhaps more surprisingly Conservative voters were also among the most tribal, or loyal voters, who were the most likely of all five parties not to have voted at all if the Conservatives had not stood a candidate.
b. Green Party
Both Green and Respect voters appear to be far more proud, and willing to admit their affiliation, than other parties.
Those who voted Green were the only group of voters that did not place Respect as the party they felt least affinity with (Greens placed them second after Lib Dems). The Green vote was clearly part of a non-establishment sphere with Respect, placing the two parties of government (Labour and Conservative) as their least preferred alternatives. Although it must be stated that those who voted Labour or Lib Dem would prefer to vote Green than Conservative in a way that they would not consider voting for Respect.
Green voters were the most tribal, or loyal of voters, with almost 38% of them claiming they would not have voted if the Greens had not stood (compared to 18% of Lib Dems for instance) this would indicate that the Green Party standing boosted turnout. It may also indicate that the fact that the Greens have stood in this ward consistently over a number of years means they have developed a layer of people who regard themselves as ‘Green voters’ rather than simply choosing the Greens when it comes to the election, despite the fact that these candidates are often paper candidates with no significant election campaign.
There are strong indications that if Respect and the Greens had stood one candidate each in the multi-vote ward (where people had two votes) the Green vote would have increased by possibly 98 votes or more bringing them close to beating a Labour candidate.
c. Labour
Labour voters are less willing to admit who they vote for than voters of other parties. It does not take a great leap of imagination to suppose that those who were voting for the party of government felt some embarrassment about that fact, particularly at a time of great scandals at the top of the party – and that those who were voting for the parties of anti-establishment passions were more willing to wear their colours on their sleeve, so to speak.
Whilst we may feel this is obvious on one level – it is not an objective certainty that Conservatives are less ashamed of their affiliation than Labour supporters.
It will come as no surprise that the Labour Party was placed well within the ‘establishment sphere’ of the parties with strong representation in Parliament. No respondents who voted Labour placed Respect as their second preference (which was not the case the other way round).
It is interesting to note that the two parties that came first and second in this ward (Lib Dems and Labour) placed their nearest rival as the party they would be most likely to vote for if their first choice had not stood. So whilst the activists of the parties may consider the other the most dangerous threat, the voters for those parties actually consider the two to be reasonably close together.
d. Liberal Democrats
The Lib Dems won Romsey ward decisively at the election and so the data we gathered in this study has to be viewed in this context. For example if you examine Table 5 in Appendix B you can see that all the parties (bar Respect who place them second) place the Liberal Democrats as their closest affinity.
This may well be due to the fact that the Lib Dems get the highest vote in this ward and are therefore the strongest performers, so it is not surprising that those who vote for other parties are tempted by the Lib Dems. Alternative explanations could be that the Lib Dems are more well placed to the centre of the political spectrum, appear less partisan or indeed are the least ‘offensive’ of the five parties.
It may well be the case that the Lib Dems were beneficiaries of Conservative tactical voting. Where those who would normally vote Conservative because they live in a ward where their first preference has little chance of winning voted for their preferred alternative instead.
Whilst many voters for other parties considered voting Lib Dem as their second choice these affinities appear relatively thinly spread. We might be able to account for this because Romsey ward was held for some time by Labour. It is possible that some of those defectors to the Lib Dems have yet to develop deep ties with the party and are more likely to deliberate over who they vote for rather than having an inbuilt assumption in the way that those who habitually vote for the same party over a period of years will have.
When we look at the statistics in a slightly different way we realise that whilst many voters voted for the Lib Dems, when we come to choosing the Lib Dems as second preference there is a dramatic drop by 26%. This possibly indicates that as many of these voters are recently won over the well is beginning to run dry of those who would consider voting Lib Dem. However, bearing in mind the Lib Dems won more than 40% of the vote in this ward it probably should not lose them too much sleep that their potential for winning further converts is slowing down.
e. Respect
Respect voters were among the most willing to admit who they were voting for, and therefore among the most proud and vocal of their affiliation.
Respect, not surprisingly, is placed well within the non-establishment sphere of the voters in our survey, feeling the closest affinity with the Green Party and the least affinity with the two parties that form governments in this country, the Conservatives and Labour.
Respect voters were the least tribal in their voting behaviour by a very long way, with only one respondent saying that they would not have voted had Respect not stood. This would indicate that the fact that Respect stood had little impact upon the turnout in this ward.
This is perhaps understandable as Respect is the newest of the five parties and has yet to establish itself as a consistent political entity in the area. Although Respect stood at the last general election, this is the first time they contested a local election in this ward and may well not yet have the profile and recognition that they may gain in future years.
Only Green voters did not place Respect as the party they felt least affinity to, with no Labour or Tory respondents saying they would have considered voting Respect. This might indicate that Respect is the “Marmite option”. With those strongly supporting it on the one hand and few others lending it support, almost all of their second preferences came from those who voted Green. Respect saw the largest drop from those who actually voted Respect and those who would have considered it as a second option (a massive drop of 41%, far higher than the only other party to lose support this way, the Lib Dems, who dropped 26%)
This would indicate that despite Respect achieving a very credible vote they polarise opinion. Whilst all the voter groups would consider voting Lib Dem only those in the ‘anti-establishment sphere’ consider voting Respect.
There are strong indications that if Respect and the Greens had stood one candidate each in the two vote ward the Respect vote would have increased 89 votes or more, significantly enhancing their vote.
7. Thanks
Finally I’d like to thank everyone who took part in the survey, everyone who gallantly stood in the baking hot sun indulging my statistical obsession (Carol, Rich, Marc, Stuart, Reuben, Salman and Laura), to Rich who helped turn the first draft of this report into the second draft and all those who helped turn second draft into final product (Richard, Tim and Dave) of course all those who help to distribute and publish this report.
Jim Jepps
Appendix A
Questions asked
After the respondent told our researcher that they were willing to answer questions about their voting behaviour they would be asked two questions
The first question was if they had voted (or were intending to vote) who were they voting for?
If the respondent replied that they were not going to vote they were asked what was the reason for this. Researchers were asked to keep the response simple (preferably one sentence) to aid compilation of results.
If the respondent had voted we then asked who they would have voted for if the party they had voted for had not stood in this election.
Appendix C contains some details of policy on how certain responses were to be
dealt with.
Appendix B
Tables of data collected
Core data on voting behaviour (Table 1.)
|
Voted for |
Would have voted for |
|
||||||
|
|
Conservative |
Green |
Labour |
Lib Dem |
Respect |
None |
Survey % |
Actual % in Romsey ward |
|
Conservative |
n/a |
0 |
2 |
6 |
0 |
5 |
8.9% |
10.9% |
|
Green |
2 |
n/a |
5 |
4 |
7 |
11 |
20.0% |
11.1% |
|
Labour |
3 |
5 |
n/a |
8 |
0 |
5 |
17.8% |
27.6% |
|
Lib Dem |
7 |
12 |
9 |
n/a |
3 |
7 |
34.4% |
40.2% |
|
Respect |
2 |
6 |
4 |
5 |
n/a |
1 |
18.9% |
10.1% |
|
Total |
14 |
23 |
20 |
23 |
10 |
29 |
|
|
Table 1a.
|
|
% distribution of parties voted for |
% distribution of second choices (excluding would not votes) |
Actual % in Romsey ward |
|
Conservative |
8.9% |
15.5% |
10.9% |
|
Green |
20.0% |
25.6% |
11.1% |
|
Labour |
17.8% |
22.2% |
27.6% |
|
Lib Dem |
34.4% |
25.6% |
40.2% |
|
Respect |
18.9% |
11.1% |
10.1% |
There were also the following responses of those who used, or would use their vote for a number of parties due to the multi-vote ward, this data is essentially unusable due to the low number of respondents (except perhaps to note that there was no regular pattern of vote distribution, e.g. there was no consistent Green and Lib Dem vote) Table 2.
|
Voted for |
Would have voted for |
|
LD |
G & R |
|
G & R |
None |
|
LD & R |
Con |
|
R & G & L |
None |
|
Lab |
LD & R |
|
G & L |
LD |
|
L & LD |
R |
|
G & LD |
Con |
|
L & LD |
G |
|
LD & R |
L |
Total number of Respondents Table 3
|
Multi party vote |
10 |
Our turnout |
50% |
|
Single party vote |
119 |
Actual turnout |
37% |
|
Did not vote |
127 |
|
|
|
Total respondents |
256 |
|
|
It’s worth noting that there is a probably connection between those who will take time to vote and those who will take time to answer a questionnaire on the street. Also those who are not interested in voting may well be particularly uninterested in answering questions about voting. It is these two connected factors that probably go some way to answering why there was a disparity between the number of the survey’s respondents who claimed to have voted and the reality.
Affinity of political parties; Table 4
|
Affinity pair
|
Number of respondents
|
|
|
Conservative / Green |
2 |
Low. |
|
Conservative / Labour |
5 |
Mid |
|
Conservative / Lib Dem |
13 |
High |
|
Conservative / Respect |
2 |
Low. |
|
Green / Labour |
10 |
Mid |
|
Green / Lib Dem |
16 |
High |
|
Green / Respect |
13 |
High |
|
Labour / Lib Dem</ | ||
