National Assembly for Wales

Welcome to this public session of the Senedd's Standards of Conduct Committee. We are continuing with our inquiry into individual Member accountability, and this is our fourth evidence session. We are very pleased to welcome Professor Jonathan Tonge from the University of Liverpool, so welcome, Jonathan. If you'd like to give us a short introduction to yourself and your work on recall.

Thanks very much to all members of the committee for the invitation; it's great, even if I'm speaking on an issue on which many of the measures, you hope, are never put into effect, which is a slightly strange position. In my own work on recall, what I've done is examine the Westminster recall petition system to see how effective it is, to see what could be modified and to look at what tweaks could usefully be made, particularly, I think, to the method of petitioning that goes on. Obviously, in this particular evidence session, I've also been looking at the possibilities in terms of by-elections and whether they would be viable if such sanctions were taken against a Member of the Senedd. So, those are the sorts of areas that I was going to cover: sanctions, petitions and by-elections.

Brilliant, thank you very much. I'll start off, then, with some questions around the triggering of a petition, if I may. What do you think are the best circumstances to actually trigger a recall petition? We know, in Westminster, they have three essential grounds: it's anything less than a 12-month prison sentence, given that a 12-month prison sentence automatically disqualifies a Member; there's the suspension of at least 10 sitting days from Parliament; or a Member being convicted of an expense offence. Do you think that those three areas are broadly the right place to be?

Yes, I think the criteria for triggering a recall petition are, actually, fine, and they've worked pretty well. I've been surprised at the number of recall petitions that have been triggered. When the Act was finally passed at the start of 2016, I wasn't envisaging so many Members of Parliament, if you like, caught in the net, but I think the actual criteria are very fair. I think there is a question to be asked over the 10-day cliff edge. In other words, if you get a suspension of less than 10 days, then you can lose rights and privileges within the House of Commons, in the same way that you, as a committee, can take away such rights and privileges within the Senedd. It's a bit of a cliff edge, though, because after 10 days, a recall petition is automatically triggered. On the evidence thus far, that's pretty brutal; there's only one MP who has actually survived that process subsequent to it.

So, I do wonder whether there can be a more tapered area, rather than just this 10-day cliff edge, after which we're straight into recall petition and by-election territory, which is very, very difficult for the miscreant to survive. Some people may think, 'Well, that's fine, the miscreant shouldn't survive', but I do wonder, and I don't want to be facetious, whether a sin bin approach to the Parliament might possibly work, in which you have another level of sanctions that applies if someone's suspended for, say, between 10 and 30 days, which doesn't necessarily trigger a recall petition, and then, perhaps, after 30 days, recall petitions are considered. That's just a thought. I just think you go from one relatively light area of sanctions below 10 days, and then, after a 10 working day suspension, you are deep into difficulty as a Member of Parliament, for sure, at Westminster. 

That's really interesting. Have you got any thoughts around how that tapered area between 10 and, say, 30 days might work in practice?

In terms of the types of 'crimes', in terms of what a Senedd Member's misdemeanours would be, it's hard to say directly what that would be; it would be for the judgment of you, obviously, as a standards committee. But I'd advise you to be cognisant of the implications of the recall system, because so far the recall system has been very effective—it's worked in all bar one case in removing the offender. But there doesn't seem to be much of a graduated process in respect of that.

I think the other point to make is that, of course, a Member of the Senedd should have the right of appeal to the Senedd as a whole body. I think, in the interest of fairness, that should apply. I do think that should probably be to avoid at least the accusation or the optics that might suggest political parties at play. But it could be usefully a two-thirds weighted majority vote. So, in the 96-Member Senedd coming down the track, it depends how you'd apply that. Would it be all those present and voting, which would be quite a high hurdle? Sixty-five people would have to support the sanction if all 96 were present. Should it just be a weighted majority vote of those attending, or should it be of the entire Senedd? That's an important consideration as well. But, obviously, I think there has to be a right of appeal. And that right of appeal should take two forms. One should be against the verdict itself, and secondly, given the importance, it should be against the length of sanction that is imposed by the committee. I think that ensures fair play all round.

Thank you. That's very useful for us. And just to take you back to this idea of a tapered approach, would you be encouraging us as a committee to look at the range of sanctions that we have on offer to us currently, to see whether they need to be broadened in order to fill that tapered area?

You could have a list of graduated punishments associated with the length of suspension from the Senedd. As I say, it's a fairly blunt cut-off at present, the way that Westminster applied it. My strong suspicion is that when Westminster passed the legislation, for sure it was popular. If you look at public opinion, so many constituents wanted right of recall, either—and this was not implemented—directly via constituents themselves, or for Parliament to initiate a right of recall. So, it was bowing to public opinion, to popular opinion, when it was introduced, but the 10 sitting day figure—a suspension over that can trigger such drastic procedures—I've always thought was actually quite arbitrary. I've yet to see the deliberation, or I've yet to see clear evidence as to how that 10-day figure was actually arrived at. You go from one extreme, like I say, of relatively—and I'm not undermining the importance of the removal of privileges, speaking rights, et cetera, from the Senedd—light sanctions to very drastic, draconian sanctions beyond the 10-day procedure. There might be some conduct of an MS that merits a sanction of some form of suspension from the Senedd for up to 30 days, but might not automatically generate a recall petition, which potentially, when we come on to by-elections, might actually be an outright removal petition, in effect, so the sanction is even more draconian in that sense. I'm sure there would be a list of misdemeanours that might be accommodated with a, say, 10-day to 30-day removal from the Senedd—the sin bin approach, as I've termed it—that falls somewhat short of actual recall. That's not to remove recall at all. I think recall is very good, it has been very healthy for democracy. It's simply that you've got the opportunity in the Senedd to introduce perhaps a more nuanced policy than that which was introduced at Westminster.

Thank you. If I can focus now on the interface between recall and disqualification. Currently in the Senedd, the same as in Westminster, any prison sentence that's 12 months or longer would trigger disqualification. Do you think that the matters that disqualify Members need to be looked at and possibly changed again, or do you see that a recall system can sit alongside that fairly neatly as it is?

I think it would sit alongside fairly neatly. I think it would be a hard job to explain to the public in Wales, ‘Well, actually, for the Westminster Parliament, a custodial sentence of 12 months automatically disqualifies, and a custodial sentence up to 12 months triggers recall, but we're not going to implement that.’ I actually think the public would understand why custodial sentences should trigger either outright removal or a recall process. So I would leave that alone. The same with expenses and the way that expenses issues can trigger sanction. I actually think that's been very successful, I think the public understand it, and I think it resonates with the public. I would leave that alone. I would say the only area to adapt—. I don't like the term ‘Westminster model’ in many ways, but if you want to use that term, if you want to adapt the Westminster model, it is more in, like I say, a graduated area of sanctions, not removing the automatics, as it were, that you already have in place at Westminster. Does that make sense? I hope it makes sense.

Just before I bring Pred in, just to clarify, you feel that the current system of a custodial sentence of 12 months or longer is where we should be for disqualification, and that any custodial sentence less than 12 months could fall under recall. 

Yes, I think that's fair and appropriate. 

Just briefly on this, what about suspended sentences?

At the moment, the current legislation says it's any custodial sentence, even if suspended. That might be a slightly strange one, because a suspended sentence, in many ways, although it's classed legally as a custodial sentence, is not a custodial sentence in that the miscreant is not put in custody. So you might want to say that, in Wales, if the sentence is only—again, using that term advisedly—suspended, then perhaps you could be more lenient. There's a case for that. I wouldn't go overboard in advocating that case for sure, but I think it's possible. I think it would be legitimate to distinguish between a custodial case and not, as in ‘custodial’ being actually going to prison. I suppose the issue with that is—and again, I don't want to be glib—with the current sentencing policies and the desire not to send people to prison, a lot of sentences perhaps are suspended. But my own overarching view is that what is fixed in terms of prison sentences, and in relation to expenses as well, I would leave alone. I think that part of what Westminster introduced is transparent, people understand it and people understand the reasons for it.

One final question from me before I pass on to other Members. We've been focusing so far in our discussions with you on what would be perceived as misbehaviour, wrongdoings, but at the heart of a recall mechanism is of course the issue of democracy. And so I wondered whether you had a view—. Here in the Welsh Parliament, because of the element of proportional representation that we have, and our list system, we have in the past had issues where a Member has been elected on that list for a particular party, rather than under their own name, and then during the course of that Parliament term, has switched from the party that they were elected for to another party, and sometimes to another party again. Do you think that that is something that should be looked at as an area for recall, to give the electorate the ability to continue having a representative from the party that they actually selected on that ballot paper, or do you see that as being a separate issue?

I do see it as an area for recall—a separate area of recall, obviously, because it's not about a misdemeanour in any sort of legal/criminal sense, but it is—. If you've got a party list system and someone stood on a party label and then switches, the electorate may struggle to understand the legitimacy of that. It is, in many ways, anti-democratic. I see the difficulties. What happens, for example, if an independent was to switch to a political party? But, yes, I do see that defecting to another political party is grounds for recall on the basis of a democratic principle that if you're elected to a particular political party, and, particularly in what is coming in terms of a closed list PR system, to me, it makes sense to implement recall procedures to give the electorate a chance. Obviously, as we're going to come on to the method of how you have by-elections under closed list PR, which I'm very happy to come on to whenever you want, yes, I would initiate recall procedures in the event of someone changing their party. 

Thank you. I'll hand over now to Peredur Owen Griffiths. Pred. 

Thank you very much. Very interesting, what you've had to say already, and maybe, following that thought that you talked about there, in that we're going to have 16 paired constituencies on a PR system, list system, going forward, how does it work, in your view? How would a recall petition work, and how would a potential by-election, or what method would you advocate using?

Well, I think you've got two broad options. One is the recall petition followed by a by-election—I'll come on to the details of how you could have by-elections—or you can have, in effect, a removal petition. A 'recall petition' would actually be the wrong title in many ways, because it will be the removal of the miscreant and replacement. Now, the easiest route is simply, 'Why should the party be punished if this is the responsibility of the individual?' If this is wrongdoing by the individual, then the obvious and perhaps easiest solution is to remove the offending individual, subject to a petition—again, I'll come back to that in a second—and replace them with a co-opted member from the same political party, because it's not the party that's at fault here, it's the individual who has—

In that instance, is that the next person on the list, effectively? 

Yes, the next person on the list. That's the most straightforward way to do it. In terms of the petition itself, there are different ways of doing the recall petition to trigger that process of co-option. Overall, I'm not a massive fan of co-option. My main area of research is the Northern Ireland Assembly, and the level of co-option without the recall system is phenomenally high, but there are special circumstances, obviously, in Northern Ireland. I think the 2017-22 Northern Ireland Assembly had 29 out of 90 Members that were co-opted at one point, which does raise fundamentals about democracy. 

But, anyway, getting back to the issue, there are different ways then of how to exercise the recall petition. You could just have the 10-per-cent-of-the-electorate threshold to sign the petition and then that triggers automatic removal. There's a case for having a higher threshold, because we're talking here about outright removal of the miscreant, albeit replaced by someone from the same party, whereas, obviously, for the Westminster system, that miscreant has the option potentially of fighting a by-election, not that most do so far. And so, there would be a case for raising the threshold required to remove someone outright up to 20 per cent, for example. That figure is slightly plucked out of the air, I would concede, but do you have a higher threshold in terms of the percentage of the electorate required to sign a removal petition as distinct from a recall petition? You could keep it at 10 per cent. It would be fairly straightforward, then: either you get 10 per cent of the electorate sign the petition, or you don't, in which case the person stays in place. If more than 10 per cent sign, you simply replace that Member of the Senedd.

There are other mechanisms available. You could have a straightforward, if you like, mini-referendum, on whether the person stays or goes. You could put it to the electorate and, over a specified period—again, I'll come back to that in a second—you could ask the electorate in that constituency to say, 'Yes or no—does this person stay or go?' There's a certain amount of democracy in that, so that would be another option.

Another option would be to have by-elections. It's not impossible to have PR-STV by-elections. Ireland's managed it for more than 100 years. A hundred and thirty-four—I'll just double-check the figures—sorry, 138 by-elections since 1923, when they were introduced in Ireland under PR-STV. It's not a case that you simply get the original party re-elected. You get almost half—in 67 of those cases, there's been a change of party at the by-election. But there are obvious disadvantages to having STV elections and, again, there's a risk that you punish the party rather than necessarily the individual, which I don't think is the point of the exercise. And another alternative, which I'm sure you won't go for, but I'll just raise it, you could—it would probably confuse the electorate as well—have a first-past-the-post by-election. But mixing that in a PR-STV system, I'm not advocating it; I'm just floating it very, very gently as a possibility.

So, there are different options. You can have the referendum option, the recall option with by-election, you can have the co-option approach. Removal with co-option looks the most straightforward to me, but, you know, there are a range of alternatives, each of which do have some merit.

And with your removal co-option option, could that be done on one ballot paper?

Yes.

I mean, really, you're just putting two names on a ballot paper: the person who is subject to removal and the next person on whichever party list. The other thing, going back to what Vikki was saying earlier, if that person left that particular party and went to another party, I'm assuming—but you might correct me if I'm misrepresenting you—that, from the original party list that they were elected on, it should be the next person on that list that should be ticked.

I think that's a very neat idea. I think you'd have to think carefully about the phraseology on the ballot paper: would it be, if you like, a head-to-head, 'Should X be replaced by Y?', or do you give two separate votes of, 'Should X be removed?', and then you affirm for Y. The problem would be if you didn't get a clear and decisive result. I mean, it's a question of how you would phrase that on the ballot paper then: 'The proposal is to replace, to remove, Member'—and their name—with, 'Do you agree or disagree?' You'd have to think about the phraseology. But it's certainly doable, what you suggest, yes—a straightforward replacement.

I'm not necessarily advocating for anything; I'm just trying to tease some ideas out, really, and see how they would work in practice.

And then whether you'd bother having a threshold or whether, you know—again, I don't want to sound facetious—if only 50 people bothered voting, as long as there was an affirmation in favour of what you propose, would that suffice, or would you need a 10 per cent threshold? That's to be formulated, I think.

If we were going for a recall petition, in the same way as Westminster, as in, 'We wish to recall this person', is 10 per cent sufficient, and bearing in mind that we're going to be looking at paired constituencies, going forward, something that came up previously is would you need 10 per cent in both constituencies or across the constituencies.

I'm quite a fan of just 10 per cent of the electorate signing a recall petition to trigger a by-election, but, obviously, potentially you've got a more serious sanction here because the person who's being removed will not have the opportunity to fight a by-election if it's just a removal petition. So, I think there's a case for raising the threshold up from 10 per cent because you are removing this person wholesale; they're not going to be given the opportunity to, in their view, clear their name with the electorate, as it were. My only note of caution on that is if you look at the percentages that have signed recall petitions thus far, they vary enormously. Obviously, one didn't make it, in the North Antrim case—particular politics pertaining to Northern Ireland there—but only 9.4 per cent signed that particular case in North Antrim, whereas it's been as high as 27.6 per cent, in Peterborough. You're averaging around the 18 per cent mark, so if you raise the percentage too high you might not get—. And if it's the aim of the committee broadly, given what has gone on, to remove the offender, then I wouldn't raise the threshold too high because a lot of people might not bother signing the petition.

Again, I can come back to how you can improve turnout in recall petitions in a moment, if you want, because I do think that's where the Westminster legislation has been fundamentally flawed. I'd be wary of raising it much above 10 per cent—that would be my advice—perhaps to 15 per cent, cognisant of the fact that you are removing this person outright, whereas a recall petition gives that person still the opportunity to fight on in a by-election.

Based on what you're saying there as well, as you talked a little bit about how you'd get more people to potentially engage, is that by using proxy and postal, or other ways? What were your thought processes around that?

Postal, for sure. I think where the Westminster legislation was flawed was that it permitted a very variable number of petition stations, which made very little sense. You can have up to 10 petition stations or petition places or signing stations—whatever term you want to use—for recall petitions. So, the first thing I'd do in terms of those who want to sign it in person is use the maximum number of petition stations. There is a correlation between the number of stations that have been used and the percentage of the electorate signing it. I know it's a small 'n', so I'm wary of overusing this statistic, but I've looked at the data. The number of petition stations is varied without explanation. North Antrim only had three and, frankly, that, in my view, was inexcusable in a huge constituency like that. Brecon and Radnorshire used six; Peterborough, 10; Rutherglen and Hamilton, seven; Wellingborough, nine. Well, why? If you look at the North Antrim case, there were 53 polling stations on election day in that constituency, and yet they used three petition stations. Now, you can turn around and say, 'Yes, but there were six weeks to sign the petition', but why would you have 20-mile distances on average between each petition station? So, that's the first thing—I'd standardise the number of petition stations.

Secondly, yes, I think it's a really good point you raise about postal voting—postal petitioning, I should say. Yes, it does help. If you look at the Rutherglen and Hamilton West recall petition, the bulk of signatures were by post. So, postal voting makes very good sense. I wouldn't take away the right for someone to go to a petition station, although there is controversy with that because when you go into a polling station on election day, people don't know how you're going to vote; it's a secret thing. If you go to a petition station—and this was an issue in North Antrim with the sectarian geography of that constituency—people know what political act you're about to engage in, that you're going to sign a petition to unseat an elected representative. So, in that sense, it's a trickier issue. So, I'd strongly advocate postal voting.

I'm not as keen on proxy voting. If you've got several weeks to sign, do you need a proxy? But, if you want to keep proxy voting, so be it.

I'd also shorten, for reasons of cost, the length of time that the petition station is open, but potentially have longer hours—and again, these aren't fully regulated in the Westminster legislation; they're partially regulated, the opening hours the petition station is open—more evening openings, perhaps, for people who are at work, but you could close it earlier, not at six weeks. It's pretty expensive. It can be up to—

So, closing early would be once you've reached 10 per cent.

Yes, exactly. I mean, the bulk of signatures take place in the first two weeks. In fact, the vast bulk take place in the first week. In Rutherglen and Hamilton West, it was done after two weeks and yet it had to stay open, superfluously, for another four weeks. There's no point. They were already at 10 per cent. It made no sense. There were another four weeks. Well, to prove what?

Just some thoughts around—. We'll be looking at paired constituencies here. Potentially, in Scotland, they're looking at potentially regional, but that's different legislation. Here it will be paired constituencies across multiple county council boundaries. How does that work? From a practical point of view of managing, if you are getting more stations to be able to go to, open for longer, it adds a lot to the financial cost to the electorate to keep that open.

Yes, I recognise that. Like I say, it's an expensive business. The most expensive recall petition so far has been £0.5 million, and I think Peterborough cost £166,000. They're not cheap, having recall petitions, so it's a balance between a cost-cutting—. You don't want to cost cut on democracy, but you want to keep things realistic. In terms of who should shoulder the burden of those costs, yes, that's an interesting one, which I confess I've not thought through as to where the burden should be shared. Should it be shared across those different council areas? One would assume it would, but that's something that—mea culpa here—I need to think about more deeply, as to where the costs should lie. Just to emphasise one point, although whilst I would close things off to save money after two weeks—there's no point carrying on—I do like and respect the secrecy rules that were implemented by Westminster. There's no commentary allowed whatsoever from any individual or any campaigning group on the progress of the petition. That would undermine it and it would unfairly potentially disadvantage the person who's trying to cling on, again, for want of a better phrase. I fully understand the secrecy rules with that. You can't have a running commentary on it.

Thank you, Chair. That brings me to the end of my questions.

Thank you, Pred. I'll bring in Mark Drakeford now, then. Mark.

Chair, thank you. Just two questions from me, Professor Tonge: first of all, whether you've done any work around the issue of expenses in the petition system. As you just said, there are expenses for public authorities, but there are expenses that can be incurred by campaign groups and individuals trying to promote the idea of signing a petition. Are the Westminster rules good enough? Do we just pick them up and drop them into the Welsh context, or do you think the committee should do some work of our own on making sure that how money can be spent, expenditure limits and so on, are tailored to the circumstances we would face in Wales?

I think the expenditure limits are reasonable that were introduced at Westminster—. I think it ranged from between £500 to, I think, £10,000 that any campaigning group could spend. In reality, the parties haven't spent that much on campaigning to oust a particular alleged miscreant, so if the parties want to spend that money, then I think that's reasonable for them to spend it on campaigning to get greater public awareness of the petition. It seems to me a reasonable campaigning stance to be able to take. It hasn't cost the public a great deal of money thus far, other than the quite significant costs of having permanent staff, or diverting staff, to actually staff the petition stations themselves. In terms of the political parties, obviously that ceiling could be raised with inflation, but, in terms of the amount they've spent, it's been rather modest. I mean, there hasn't been full-on campaigning, partly because the parties are limited in what they can do; all they can do is try and raise public awareness that the petition exists. But, as I say, they're not allowed to provide a commentary; you can't get the political-party-style 'x winning here' and lots of bar charts. They're not allowed to do that under the legislation, which I think is quite right in terms of preserving the secrecy of how the ballot is progressing.

So, I think parties have found themselves fairly hamstrung in what they can do anyway, and that might explain why they've not, thus far, spent a lot of money campaigning on the recall process. I think also there hasn't been much fear, to be quite honest, of—. I think partly the other reason is parties know that the threshold's going to be reached. As I say, the North Antrim case was very, very exceptional, due to the particular politics of Northern Ireland. The 10 per cent threshold has been passed pretty comfortably ever since. I think the closest was Wellingborough, which was about 13, 14 per cent, but got over the line. So, they've not had to do that much. If there was to be a tight contest with more controversial sanctions, I suppose parties might get more involved in campaigning then.

Thank you very much. And my second question is about the two-stage process. In your retain-and-replace option, wouldn't it be simpler just to do away with the petition stage, given, as you say, the petitions have all been reached in any case, and given that, before you get to the triggers for a petition, someone has got to have gone through a whole series of stages internally within the Senedd before they get to that, so you already know that there is a prima facie case that they've done something pretty unusual and exceptional, why not go straight to a retain/replace ballot with a threshold of, say, 20 per cent in there, to make sure that there's a reasonable proportion of the electorate taking part in such a choice?

I wouldn't argue against that, as long as the electorate is involved. Initially, when you—. Sorry, when you made those comments, I thought you meant just that the Senedd themselves could remove the miscreant, which would remove the electorate from the process entirely, and I think that was always the issue. When the Recall of MPs Act was being debated, first of all, there were some opponents of it outright, but the criticism that came from some re that Act was, 'Well, the electorate isn't sufficiently involved here; they can't initiate any recall procedures themselves.' I'd personally prefer it that yourselves as a Senedd committee can—you're the ones to initiate the procedure, but, at some point, the electorate has got to have a view on this.

So, get the electorate involved at the recall stage. If it is an outright removal, if you're not going for a by-election model, as I said earlier, I think there's a case for raising the threshold, but I'd be cautious about raising the threshold as high as 20 per cent, because you might find it a struggle to get 20 per cent of the electorate to turn out and sign the thing. So, I'd say 15 per cent might be the sort of area, the threshold you're looking at in terms of removal, which gives the electorate a say, gives a reasonable proportion of the electorate the right to remove that person. It's a higher percentage, so you can justify it. It's a higher percentage for outright removal than it is just for recall—under the Westminster system, that's 10 per cent, but that person is still in the game, as it were, with the option to fight the by-election. So, I think 15 per cent is the mark that I would land on, but, obviously, it's up to yourselves. I think 15 per cent has a certain logic to it. A higher threshold, but I wouldn't start making the threshold insurmountable. If you look, with a 20 per cent threshold, a couple of the petitions would not have got through with a 20 per cent threshold; I've got the figures in front of me. Well, Brecon and Radnorshire, if you want the example from Wales, 18.9 per cent of the electorate signed that petition, so you can argue that's a warning against the 20 per cent threshold.

Thank you.

Okay. Thank you, Mark. I'll bring in Natasha Asghar now. Natasha.

Thank you, Chair, and thank you so much, Professor Tonge. I have one question. Obviously, you've mentioned the Northern Ireland model, we've spoken about Westminster, we heard from our Scottish counterpart earlier this morning, and we've also been privy to various other legislatures and how they actually carry out their recall systems, for example, when it comes to sanctions et cetera. I just wanted to ask you, from your research, which political institution do you feel has it right, if anyone does indeed have it right, because, obviously, we want to have the best system in place. You mentioned the tapered system earlier on, that's something that is quite new, but I just wanted to ask you: is there any political institution from your research, from your experience, that you think has just nailed it and got it right?

I don't think any institution has got it completely right, and it's not often these words might cross my lips, but I actually think Westminster has got this broadly right. It has issues of detail that I don't like, such as the ad hoc number of petition stations that can be used. I think Westminster has got it right, but, obviously, in the Westminster model, having a by-election is a lot easier than when you're using a closed list party/STV system. I personally wouldn't rule out by-elections for yourselves, but I understand the difficulty; it is more complex under STV. I mean, if you look at the models, certain parts of the United States, they have recall, but the rules there are pretty arbitrary, and they've not been hugely successful.

Northern Ireland doesn't have the system at all. Northern Ireland's just co-option. They do have the right of expelling an Assembly member, for sure, but that's just a straightforward co-option, and the electorate is not involved in that process. There's a vote of the Northern Ireland Assembly; the electorate aren't involved in the process in Northern Ireland. To remove a Minister in Northern Ireland, you've got to have a cross-community vote, so it's particularly difficult, given Northern Ireland's politics.

So, I do think you've got a great opportunity here to, if you like, correct the details that Westminster got wrong, whilst retaining the 80 per cent plus that Westminster got right here. I do think it's a wonderful opportunity to have. You cannot create a 100 per cent perfect recall system, there'll always be anomalies, and I think what you're moving towards probably is a removal system. Would you call it a recall system or removal? It depends. The phraseology can be important here, but, if you're co-opting a party replacement, I think people will see the logic of it, it just needs explaining that it's the miscreant that's being punished, it's not the party. Why should the political party be punished? So, I think you've got a great opportunity.

Professor Tonge—sorry, Chair, just a sub-question to that—you mentioned that there are some things that we need to watch out for, and I took note of all the things that you mentioned previously. But, as with every rule, every law, there are certain things called loopholes, which I'm sure we're all aware of. Are there any particular loopholes you think we should be wary of when we do go further with this?

I'm trying to think—. It's easy to co-opt a replacement for political parties who have got several representatives in a multi-member system. The difficulty will be—. Let's say it's a party that is relatively unpopular in that constituency and only has one elected representative, and they're removed. Then supporters of that political party are, in effect, disenfranchised. What do you do? Do you get a replacement? They've only, according to the electorate, been entitled to one representative, that representative is removed. Does the party then produce a list of reserves, none of whom were elected? Yes, okay, you can do that, but I think there's difficulty with that.

I think, in terms of independents, there's no reserve list. What happens if an independent Member is elected to the Senedd and they then are removed? Where does the reserve list come from with that? Do you just leave the seat vacant? That would be the obvious possibility. You could have a by-election. I don't think you can mix and match. In terms of parity to the electorate, you've got to decide are you going to use co-options from party lists or are you going to go for by-elections, but if there's no party list because of that independent person, then, potentially, you're looking at a seat vacant. So, again, it shows the difficulty of creating a perfect system.

Thank you so much, Professor Tonge. Thank you so much, Chair.

Thank you, Natasha. Professor Tonge, your evidence has been really invaluable to us this morning. We thank you for taking the time to come and speak with us, and I think you've laid down the gauntlet to us now. We could say that one mark of success for us as a committee is if we get referenced later on in your academic years as having been the Parliament that has really, truly got this right. That's at least something that I, as Chair of the committee, will aim for. So, thank you so much for your session with us this morning. A copy of the transcript will be provided as soon as possible to you, so that it can be checked for factual accuracy.

Thank you very much for the opportunity. Thank you. It's much appreciated.

Thank you, Professor Tonge.

Okay. So, I propose, in accordance with Standing Order 17.42(vi), to resolve to exclude the public from the remainder of the meeting. Are Members content to agree the motion? Members are. In that case, now, we will continue in private.