Professor David Farrell:

There are countries that do not have it. Greece is one such prominent example. Our political science colleagues in Greece found it so frustrating that they were unable to track what the hell was going on, whereas we had some funding and were able to get some sense of how voting was panning out in 2011 and 2016. That is a prominent example of bad practice.

When we cobbled together funding, and we begged, borrowed and stole from everyone we could for the studies in 2011 and 2016, there would eventually be particular conversations with agencies, which I will not name. They would tell us that they would give us funding for the study but that they required us to ask a bunch of questions so that if they were asked questions by public representatives or others about how their budget was being spent on such a thing, they would have to show that the questions that mattered to that agency were being asked. Unfortunately, that means that we must ask questions that are of no use to us whatever. There are questions from the 2011 and 2016 surveys that we do not make any use of but we were obliged to ask in order to get a few other questions in there.

It is crucial go after non-voters in the studies. Sampling is so important and is the practice in countries with a well-funded election study. In newspaper opinion polls, their polling agency will use an existing panel of respondents to track voting behaviour. By definition, their panel of respondents are people who are interested in politics who are more likely to vote. A properly funded election study specifically seeks the people who are very hard to find in a normal private opinion poll because one really wants to track those who are uninterested in politics. We have been unable to do that since 2007.

Dr. Reidy has commented on how to ensure that this is done properly and in a truly independent way. As social scientists or scientists, and as academics, we are required to follow ethical procedures. We must get ethical clearance for any research we do in advance which would specifically include survey works. We must abide by certain ethical standards and we must declare any conflicting interests that we may have. To secure funding, we will have had to secure the judgment of our peers to ensure that they are satisfied that this meets the best international standards. Ultimately, as academics we want to publish our research and that goes through a peer review process. If we write something substandard or which does not meet good scientific rigour, it will not get published and we must abide by those standards.

The Austrian election study has been particularly useful on the matter of votes at 16 years. When Austria moved to votes at 16 years, the election study which followed was able to focus on that and how it impacted on voting behaviour. I am a very strong supporter of votes at 16 years. There is good political science research that demonstrates that that is a better age to trap or secure voters. People are more likely to engage in voting practice in future elections if they start doing so early in their lives rather than at a later stage, such as 18 years. It is a very important development which Austria has taken. Other places are considering it and we should too.

Senator Boyhan asked us to say a little more on women and minorities. There are a few dimensions there. One matter, which may refer more to an electoral commission than the topic of today's meeting, is that a well-run electoral administration would do whatever it could to make it easier for people to vote. That would include experiments, something we see in other countries, such as being able to vote in advance of polling day, providing information in different languages - something that is done here but of which more could be done - making it easier for people with disabilities to vote, trying online voting - which is not very popular here - as well as trying other methods of voting that do not require turning up at the polling station. There are ways which could be explored to make the method of voting easier. That is one example of how to better engage certain communities that are not so actively engaged in politics.