Measuring the diversity of political engagement in Wales

In this post, we’re going to take a closer look at how we developed what we call our Political Diversity metric.

This is our measure which makes use of data on the range of political choices on offer to voters in distinct Welsh geographies, as well as the extent to which voters are engaging with these choices, to offer a value for the strength of political diversity in that area.

Why Did We Build the Political Diversity Index?

Last September, the Welsh Government published a tender for a research project into socio-economic barriers to participation in Welsh politics. The brief: “to consider how social-economic factors impact on an individual’s ability to get involved in local democracy and stand for elected office in devolved elections in Wales.”

As part of our response, we decided to look for a way of measuring the diversity of political participation in Wales. We wanted to go beyond basic indices like voter demography and turnout to create a more fine-grained picture of what voting in or standing for local elections looks like on the ground. 

Our desire to develop our metric was motivated by the suspicion that political engagement, both in theory or practice was not homogenous in Wales and being able to count – and account for – this was crucial to understanding the engagement or lack thereof of Wales’ communities. It is also reasonable to assume that the reasons for this heterogeneity are diverse themselves and very much tied to the context of place.

This call for research into the manner in which socio-economic factors impinge upon democratic participation, be it in the form of voting or standing for office, arrived against a background of a recent national election that has already been dubbed the ‘least representative ever’ (The Economist, July 24). The evidence is that in the UK the national electorate is becoming increasingly less representative of the voting age population of the home countries. This process whereby the demography of voters does not represent the population as a whole is also underway in the national and local elections in Wales. In part at least this process is being understood as reflecting what is perceived to be a deepening problem of political disengagement amongst the UK electorate (UK Parliament, 2022).

Just as important in the context of the this tender was the detailed evidence that the candidates elected to office in county and local councillor elections in Wales (and the UK more widely) are far from representative of the population they are called to represent. Specifically, the Welsh local government candidate survey shows elected councillors to be older, have more years of education, a different pattern of employment and/or much more likely to be retired than the general population. Indeed people under 35 are more noticeable by their absence than their presence and people under 25 are, relatively speaking, simply absent from local councils. 

We decided that we would design our metric to use data from local elections, sourced from the invaluable Democracy Club database. This would be a good place to start, for three reasons. Second, because this allows us to study political participation at the most granular geographic scale, and finally, because becoming a local council member represents – in theory – one of the most accessible on-ramps to a political career. 

Unfortunately (although for good reason) these data are absent personal details about the candidates that might be revealing – age, income etc.

As we surveyed the data from the most recent council elections, we were immediately struck by one outstanding fact. Of the 753 council wards in Wales, there were 72 where only a single candidate stood for election. In a further 159 seats, there were just two. In roughly 30% of seats, voters were faced with extremely limited – or even zero – choice as to the identity of their elected representatives. 

What problems might this pose for political participation?

Well, on the one hand, fewer candidates means fewer opportunities for people to get involved with politics through campaigning, standing for election and representing their local community. This makes it more difficult for newcomers to engage with electoral politics.

On the other, an absence of choice – or even of a political contest – is likely to mean that many voters feel that their views are not represented by the available candidates. This could lead to lower turnout and reduced likelihood of political participation by individuals from the local area. 

How Did We Build the Political Diversity Index?

Clearly levels of public engagement as reflected by voter turnout and the representativeness of the population who offer themselves for political office are connected, but they are not the same thing. 

It seems reasonable, therefore, to suggest that the range of political choice available in a given electoral ward – measured by the number of candidates from distinct parties who stand for a particular election – would be a useful indicator when considering patterns of political engagement across Wales. Mapping how this figure fluctuates across different political geographies would help understand not only the breadth of political perspectives represented in that ward, but also give some sense of the range of opportunities available to people from that area looking to get involved in the electoral system.

However, this measure on its own would only tell half the story. It is one thing for voters to have access to a range options, but if there are many candidates standing who nobody is voting for, does that constitute meaningful diversity? One way of measuring this factor would be by analysing the percentage of votes received by each candidate in a given ward. If votes are relatively evenly distributed across several candidates, then this suggests that voters are engaged with a broad range of political offerings; if, however, there are many candidates but votes are largely divided between one or two individuals, this indicates that whilst choice is available, most voters are not choosing to engage with it.

By combining these two measures into a single metric, we created our Political Diversity Index – a score which reflects both the choices available to voters in a given electoral ward and the degree to which voters are actually engaging with those choices. Wards which perform well on both aspects of the Index (i.e., high number of choices, relatively even vote share) will score high, and vice versa. The Index also incorporates other important factors, such as voter turnout, and contains controls for population density, to ensure that we aren’t simply measuring the fact that some wards contain more people than others.

The result is a novel metric which allows us to see at a glance what political engagement and participation looks like in different council wards across Wales. More work needs to be done to establish the best way to interpret this data, but there are a view important takeaways from our early analysis.

Cardiff

Even relative to other urban areas, Cardiff stands out for its high levels of political diversity. This is emphasised by the fact that removing Cardiff from the overall dataset substantially reduces the strength of the correlation between Political Diversity and population density. This indicates that Cardiff has a highly distinctive political culture, and political engagement strategies there will and should look very different to the rest of the country.

Anglesey/Ynys Môn

Another notable outlier region relative to the rest of Wales is Anglesey/Ynys Môn, which demonstrates consistently high scores relative to its levels of population density and urbanisation. Without more research, we cannot speculate as to why this might be the case, but it certainly warrants further investigation. Certainly, if the region has recently implemented any specific policies with regard to political engagement this would suggest they had created an impact.

The urban/rural divide

 The most striking pattern observable in the data is the distinction between the urban south, which tends to score high, and rural Wales, where scores are generally low across the board. This reflects the fact that urban political cultures are relatively more vibrant and open, with a wider range of views being represented and voted for.

It also suggests that political participation will look quite different depending on where in the country an individual is based. A young person in Cardiff who is interested in getting involved with electoral politics will be able to choose between a relatively larger range of parties. Urban wards also tend to have a higher rate of candidate turnover, perhaps because council seats in these areas tend to be seen as stepping stones towards a political career in Cardiff or Westminster. Whether this is a good or a bad thing for local politics is a separate question; from the point of view of an aspiring candidate, it increases the likelihood that they will have an opportunity to stand for office sooner rather than later. 

Compare this with the experience of a young person hoping to stand in Gwynedd or Ceredigion. They are far more likely to face a situation where one or two parties hold an effective monopoly over local politics, and where local seats are held by long-term incumbents. Under these circumstances, an individual’s options for political engagement are less likely to be shaped by their personal convictions and aspirations, and more likely to depend on access to social networks and relationships.

It is worth observing at this point that Wales’s Political Diversity Index map contains many greyed-out constituencies where no results are given. These are areas where political engagement is so low (both in terms of number of candidates and voter turnout) that precise voting figures cannot be published without incurring the risk that individual voting choices could be inferred from the results. This fact, as much as anything else, underlines the difference in political cultures that exists between urban and rural regions. 

Independent Candidates

Independent candidates are treated in our data as though each one represented a distinct political party. We believe that this is a fair representation of their contribution to a culture of political diversity; however, it should be noted that since independent candidates, especially those campaigning in niche issues, can tend to receive relatively low shares of the vote, they are likely to have a double-edged effect on a ward’s overall Political Diversity Index score. In short, independent candidates will raise PDI scores insofar as they represent a distinct choice for voters, but will reduce scores where voters are not engaging with them.

In these elections, of the 3,292 candidates, 710 stood as independents, with a further 19 standing for “Independent” parties with a unique registration.

Before we start drawing conclusions about what the data might tell us about the experience of independent candidates, it is first necessary to understand what role we think they should play in Wales’s political culture. If we believe that there is a lack of diversity amongst Wales’s political class (whether in terms of demography, socio-economic background or political orientation), then do we think this is best addressed by making it easier for independents to compete, by making existing political parties more accessible, or by supporting the formation of new parties? Without consensus around these questions, it is difficult to say what kinds of outcome would be desirable with respect to independence candidates.

Age

It is a well-established psephological fact that older people tend to turn out to vote in higher numbers. We have not attempted to account for this in the current model, but with additional time and funds it would be possible to do so.

Household Income

Perhaps unsurprisingly, unitary authorities with higher average household income (according to ONS’ indices of multiple deprivation) also tend to score higher for Political Diversity.

A final note on the research project. We weren’t chosen by the Welsh Government to complete this contract. We did, however feel that we had a lot to offer by thinking about the question. As well as designing a tool for quantifying the very phenomena the project was trying to engage with, a tool based on freely available data, we also designed a survey that attempted to get to the heart of understanding the phenomena of engaging with politics from the perspective of a Welsh citizen.

It was essential that we focus both upon the reasons behind the shortfall in the absolute number of candidates – remembering that almost 1 in 10 local council seats in 2022 had only 1 candidate (Democracy Club, 2022). As well as responding to the data which suggest that elected candidates do not reflect the wider population with respect to the data we do have that points to the socio-economic status of elected members in Welsh County and local councils.

The boundaries of the discussions surrounding socioeconomic impacts upon electoral engagement are well understood, models of their impacts upon the decision to stand less so.

In designing our research we therefore decided that it is important to frame the research in terms of a conceptual framework and evidence base that can be used to both understand the situation and derive practical interventions.  Focussing upon those factors that can impact the perceived self-efficacy of the individual with respect to the decision to stand for office. We do this because self-efficacy theory has a long history of successful interventions to change outcomes with respect to issues as diverse as academic achievement, health interventions and computer training. More specifically we are seeking to build upon the prior research using this approach that has explored political engagement in multiple European countries (Vechionne, et al. 2014 – https://doi.org/10.1111/bjop.12067) and demonstrated a robust validated instrument that measures political self-efficacy and isa predictive of engagement with politics.

We proposed therefore to conduct an on-line survey directed particularly towards students in school aged sixteen and over. This constituency would have been sampled using a multipart survey instrument that would provide a) a score for political self-efficacy, b) a score for the level of engagement with politics and b) questions designed to identify their sources of information e.g. chat, social media, scramming print etc. Using postcode as a indicator of socio-economic standing we would then test for relationship between political self-efficacy, political activity, modes of accessing information and socio-economic status. The additional benefit of the self-efficacy model is that this approach can be used to identify impactful interventions that will most directly address the needs identified by the survey.