The issue of measuring the political knowledge of voters is one that can be traced to the origins of modern political science. Without the ability to measure public opinion on a large scale, academics of the past worried that the electorate were apathetic and uninformed about government. Although modern public opinion measurement has improved at a rapid rate in the latter half of the 20th century, questions remain over how knowledgeable the electorate of a given democracy are.
Never one to back down from a challenge, Prof. Roger Scully and the 2016 Welsh Election Study sought to establish a measurement of political knowledge in Wales. In this blog, I present some initial findings and thoughts – and invite contributions from others in the comments section.
The WES 2016 provided respondents with the names of nine different political actors, three from each level of government that had legislative power in Wales (at the time); the National Assembly for Wales, Westminster, and the European Union (EU). For each name, they were given several different positions, and asked to match the name with the position. For the National Assembly level, respondents were asked to identify Rosemary Butler (Presiding Officer), Kirsty Williams (Leader of Welsh Liberal Democrats), and Mark Drakeford (Minister for Health and Social Services). At the Westminster level respondents were asked to identify John Bercow (Speaker of the House), Tim Farron (Leader of the Liberal Democrats), and Jeremy Hunt (Secretary of State for Health). For the EU level respondents were asked to identify Jean-Claude Juncker (President of the European Commission), Martin Schulz (President of the European Parliament), and Donald Tusk (President of the European Council). In previous election surveys around the world, this “Identifications’ test has proved to be an effective measure of political knowledge.
Here is how people did:
As you can see, the EU figures were the least well known, followed by the Welsh Assembly, with the 3 Westminster figures the best known.
Using respondents answers, we assigned them a standardised score that can be compared across different subgroups of society such as sex, age and income, but also political subgroups such as vote choice. This score can be a maximum of about 1.7 to a minimum of -1.7 (confusing scale I know, but this is still initial work).
Knowledge and Gender
On average, respondents who identified as male achieved a higher knowledge score than those who identified as female. An explanation of this likely lies in the structural barriers that women face to participating a political sphere still very much dominated by men. Work by Professor Roberta Guerrina has suggested that women face obstacles to obtain political knowledge: they are less likely to have jobs where it is ‘acceptable’ to talk politics, and less likely to discuss it among family and friends. Dr Peter Allen has also carried out work that points to women being actively put off by the political culture of Western democracies. This phenomena highlights the need for further study into gendered political engagement/disenfranchisement and the impact of women’s continued underrepresentation in politics (see previous Thinking Wales article).
Knowledge and Age
This is perhaps the most interesting graph, showing an apparently perfectly linear correlation between age and knowledge – the older you are, the better you performed in the knowledge test. There is no obvious explanation here, but it perhaps comes down to sources of information. Older respondents were more likely to read newspapers and watch news programmes, whereas younger respondents were more likely to use social media as their primary news source – perhaps these differing sources emphasise different types of information? Frankly I am not sure, so if you do have any ideas, please leave them in the comments!
Knowledge and Household Income
There is an upwards trend here – the higher the gross household income of the respondent, the higher the knowledge score. This is to be expected – those in high income households are more likely to have stayed in formal education longer, more likely to watch news programmes, more likely to vote, and generally just more politcally engaged.
And Finally…Knowledge and Regional Vote
Finally, we look at a political division in society – regional list vote at the National Assembly elections in 2016. Regional vote was picked as a preferable category to constituency vote, as it is thought to be a more ‘sincere’ vote. On average, Lib Dem voters achieved the best scores, followed by Plaid voters and voters of ‘other’ parties. UKIP voters performed the worst in the knowledge quiz, with Labour voters also achieving low average scores.
Given that Lib Dem voters are more likely to be graduates with higher incomes, and UKIP and Labour voters more likely to be older with lower incomes, this could be seen as a ‘repackaging’ of the information shown above. However, it is still interesting to see how these divisions play out in a political scenario and see which voters are the most/least politically engaged.
A Concluding Caveat
While an identifications test has shown to be a robust measure of knowledge, it could of course be the case that this test is biased. It may simply be the case that the questions were better suited to older, wealthier, Liberal Democrat voting men. Whilst we have no reason to think this would the case, we welcome any comments or suggestions people might have for us