Hefyd ar gael yn: Cymraeg

Size Matters: Making the National assembly more effective

Electoral Reform Society,

Posted on the 26th March 2014

2. Executive Summary & Introduction
  1. Executive Summary
  2. Introduction
10. Bibliography
  1. Bibliography


By Richard Wyn Jones Chair, UK’s Changing Union

In some important ways, the story of Welsh devolution is a remarkable success story. From uncertain beginnings at the time of the 1997 referendum, our institutions of devolved government have rapidly gained widespread public support and legitimacy. Indeed, according to the most recent public attitudes evidence, fully 80% of the population now support some form of devolution. By contrast, only 9% continue to hanker after the abolition of the National Assembly with an equally meagre 9% supporting outright independence. That the electorate have embraced ‘home rule’ despite the manifest inadequacies of the successive constitutional dispensations visited on Wales serves to make the transformation in attitudes since the 1990s even more striking.

The UK government’s establishment of the Commission on Devolution in Wales, coupled with its enlightened decision to collaborate with all the mainstream political parties in Wales in drawing up its wide-ranging terms of reference, has created a unique opportunity to finally place Welsh devolution on a truly stable and sustainable basis.

The UK’s Changing Union partnership has embraced this opportunity by submitting its own evidence to the Commission, by commissioning background reports, and by encouraging debate across civil society. What has become ever clearer through this process of deliberation and engagement is that, however political inconvenient it may be to say this, Wales will not have the system of devolved government it deserves until the size of the National Assembly is properly addressed.

Truly effective, accountable government requires a powerful, properly resourced legislature to hold the executive to account. The evidence suggests that the capacity constraints resulting from the fact that the National Assembly for Wales has only 60 Members is a real barrier to good government.

60 Assembly Members compared to the 108 of the Northern Ireland Assembly, the 129 Members of the Scottish Parliament and the 1400+ Member of Parliament and Peers at Westminster.

This paper makes a reasoned case for increasing the number of Assembly Members. In our view, what one might term the Silk process will remain incomplete until this nettle is finally grasped.

In making this case we are of course fully aware that the instinctive reaction of many will be dismiss out of hand any argument for more politicians. This is not an easy let alone a fashionable case

to make. All we can reasonably ask is that critics first consider the evidence put forward in the following pages. We firmly believe that when considered in an objective manner, the evidence that size matters, and that the number of Assembly Members should be increased is compelling.

The paper has been produced in collaboration with Electoral Reform Society Cymru. It will be submitted as evidence to both the Commission on Devolution in Wales and the Commission on Public Service Governance and Delivery in Wales. Our hope is that its publication will also catalyse and inform a wider debate across civil society in Wales about this most fundamental of issues.

Richard Wyn Jones
Chair, UK’s Changing Union

The UK Changing Union Partnership

The UK’s Changing Union project is a joint initiative between the Wales Governance Centre at Cardiff University, the Institute of Welsh Affairs, and Cymru Yfory/Tomorrow’s Wales. The project seeks to explore and debate the future of the Union and the Welsh devolution settlement through research and engagement with relevant public and civil society stakeholders.

It also seeks to involve young people and wider civil society in constitutional debate, disseminating the project’s research findings in an accessible format and holding events in which organisations and individuals can understand what constitutional change might mean for them.

The project began work in January 2012 and will run for three years. It is funded by the Joseph Rowntree Charitable Trust and the Nuffield Foundation. Additional resources to the project are provided by the partner organisations.

The Nuffield Foundation is an endowed charitable trust that aims to improve social well-being in the widest sense. It funds research and innovation in education and social policy and also works to

build capacity in education, science and social science research. The Nuffield Foundation has funded this project, but the views expressed are those of the authors and not necessarily those of the Foundation. More information is available at www.nuffieldfoundation.org

The project has responded to the remit of the Silk Commission by forming three working groups on:

  • Legal jurisdiction
  • Finance and funding
  • Scrutiny, capacity, accountability and powers

In parallel with this work the project has also convened a Forum on the Changing Union. This meets bi-annually to bring together experts in their fields drawn from the four countries of the UK. It is central to the project as a whole and contributes to the new thinking required to imagine the future shape of constitutional relationships within the British Isles. Each Forum has a different

focus. Participants include academics, politicians, civil servants and other practitioners. The Forum produces discussion papers and reports on proceedings that are all available, with our submissions to the Silk Commission, on our website – www.ukchangingunion.org

Project Steering Group

  • Professor Richard Wyn Jones, Wales Governance Centre at Cardiff University (UKCU Project Chair)
  • Hywel Ceri Jones, Wales Governance Centre at Cardiff University
  • Dr Rebecca Rumbul, Wales Governance Centre at Cardiff University Geraint Talfan Davies, Institute of Welsh Affairs
  • Gerald Holtham, Institute or Welsh Affairs John Osmond, Institute of Welsh Affairs Lee Waters, Institute of Welsh Affairs
  • Peter Price, Cymru Yfory/Tomorrow’s Wales
  • David Llewellyn Davies, Cymru Yfory/Tomorrow’s Wales Michelle Matheron, Cymru Yfory/Tomorrow’s Wales

Project Coordinator

  • Lleu Williams, UK’s Changing Union

Working Group Chairs

  • Emyr Lewis, Wales Governance Centre at Cardiff University (Chair – Working Group on Jurisdiction and Legal Matters)
  • Cynog Dafis, Cymru Yfory/Tomorrow’s Wales
    (Chair – Working Group on Powers, Capacity, Scrutiny and Accountability) and

The Electoral Reform Society Cymru

  • Steve Brooks, Director, Electoral Reform Society Cymru
  • Owain Llyr ap Gareth, Campaigns and Research Officer, Electoral Reform Society Cymru

Executive Summary & Introduction

Executive Summary

The number of AMs in Cardiff Bay has long been a source of controversy. Yet there has been little objective analysis of the number of elected representatives the National Assembly needs to function effectively. In 2004 the cross-party Richard Commission recommended an increase from 60 to 80 members. It argued that the Assembly did not have enough members to populate its Committees

or undertake its scrutiny functions effectively. It warned that a move to full legislative powers, as it recommended, would make the scrutiny gap even more intolerable.

This paper provides an analysis of the evidence, going beyond arbitrary proposals that the size should be 60 or 80 or some other number plucked from the air, to come up with a reasoned case. It marshals the evidence that is available in relation to:

  • The experience of the operation of the National Assembly since it began work in 1999, and especially the role of its Committees in holding the Welsh Government to account and scrutinising
  • The operation of similar small nation and ‘regional’ legislatures elsewhere, especially within the EU, and in Australia and

Leaving aside Ministers and other office holders, only 42 of the present 60 AMs are available to hold the Welsh Government to account and scrutinise legislation. This compares with 113 in the Scottish Parliament and 522 at Westminster. It means AMs have to attend multiple committees. They are always in a hurry, constantly moving from one meeting or issue to another. Many say that they do not always have time to read, let alone reflect properly on their documents ahead of meetings. The Assembly’s Remuneration Board has recommended improving the research and support services available to AMs to help them overcome these problems. However, this is a second best option to actually increasing the number of AMs to at least bring their number closer into line with the other devolved legislatures in Edinburgh and Belfast.

The present allocation of elected representatives between local authorities, the National Assembly and Welsh MPs and Peers at Westminster is poorly distributed. More could be made of existing resources if we had fewer MPs and Peers at Westminster, and fewer paid councillors, but a greater number of AMs in Cardiff Bay.

A comparative analysis of equivalent small nation and ‘Regional’ legislatures elsewhere in the world indicates that 60 members are extremely few for a legislative Assembly that also provides an Executive. The analysis finds that for an institution with the National Assembly’s functions, at least 100 representatives is the norm.

A history of the half-century leading to democratic devolution in 1999 illustrates the arbitrary way in which the number of 60 Members came about. In all previous proposals the recommended membership never fell below 75, and generally assumed a figure of around 100.

To bring it into line with the capacity of other comparable legislatures the National Assembly should have around 100 members. This would add approximately £10.1 million to the Assembly’s current annual running costs of £49.5 million, which is a small price to pay for the benefits that would accrue.


Whether the National Assembly has enough elected representatives to do its job effectively is a concern that has been raised since the birth of democratic devolution. While the Scottish Parliament has 129 members and the Northern Ireland Assembly 108, the National Assembly only has 60 members. Nine of Wales’ 22 local councils have more representatives.

At its inception, the Assembly was often derided as a ‘talking shop’. It had responsibility only for secondary legislation and was established as a corporate body akin to a local authority in which powers are vested in the elected councillors collectively. In contrast, the Scottish Parliament and Northern Ireland Assembly had separate executives and legislatures and enjoyed full legislative powers in the areas of their competence. This pattern of asymmetrical devolution, in which Wales was dealt a much poorer hand than its Celtic neighbours, was compounded by the Assembly’s smaller size. The small number of its elected members mirrored its more limited functions.

In the first two terms, the Assembly gradually took on a more parliamentary character. It separated the executive from the legislature, abandoned its corporate body status and, with the 2006 Wales Act, acquired the potential for gaining full legislative powers. This was approved in the 2011 referendum.

These developments now raise the question of whether 60 AMs are enough to provide adequate oversight and accountability, thus ensuring responsible government.

In 2004 the cross-party Richard Commission recommended an increase from 60 to 80 members. It argued that the National Assembly did not have enough members to populate its Committees or undertake its scrutiny functions effectively. It warned that a move to full legislative powers, as it recommended, would make the scrutiny gap even more intolerable.

The recommendations of the current Silk Commission on devolving taxation powers make the question of the National Assembly’s capacity to do its job properly even more urgent. The more powers it wields – including those that will have an impact on Welsh citizens’ wallets – the more important it is that a strong central Executive is held to account. Welsh Ministers and the legislation they produce should be scrutinised properly. Viable alternative policies need to be produced. Yet there is an insufficient number of AMs and not enough time for them to develop the necessary expertise to be effective.

Indeed, the first report of the Silk Commission on fiscal devolution came very close to conceding this point. Noting that the question of the Assembly’s size was not strictly part of its remit, nonetheless, it went into some detail about the need to address the Assembly’s scrutiny deficiency. As the report stated, “Good scrutiny means good legislation, and good legislation pays for itself.” (Commission on Devolution in Wales, Empowerment and Responsibility: Financial Powers to Strengthen Wales, para 8.4.31, p 136, November).

There has been little objective analysis of the number of elected representatives the Assembly needs to function effectively. While the Richard Commission recommended 80 members, it did not explain why it had chosen this particular number. It merely stated that 60 members were insufficient.

On what basis should we judge there to be an adequate number of AMs? On a crude population basis for instance, if the National Assembly were to have an equivalent number of representatives as there are MSPs in the Scottish Parliament it would have 75 members. However, if the same calculation were made in relation to the Northern Ireland Assembly, there would be 183 politicians in the Senedd in Cardiff Bay. This paper provides an analysis of the evidence, going beyond arbitrary proposals that the size should be 60 or 80 or some other number plucked from the air, to come up with a reasoned case. It marshals the evidence that is available in relation to:

  • The experience of the operation of the National Assembly since it began work in 1999, and especially the role of its Committees in holding the Welsh Government to account and scrutinising A great deal of information on these matters can be gleaned from the work of the National Assembly’s Remuneration Board which was established as a statutory body in October 2010. As well as deciding the pay and allowances of AMs the Board also monitors the role and duties of AMs and the staff support they need. (In 2008 the National Assembly Commission established an independent Review Panel to look at all aspects of financial support available to Ams The Panel’s report Getting it Right for Wales (July 2009) recommended that the automatic link between the pay of AMs and MPs should be broken and that a statutory independent body be established to set future salary levels and to monitor and review other financial support available for AMs, including allowances for travel, pensions, office costs, and staff support. An Assembly Measure was duly approved in July 2010 and the new Remuneration Board under the chairmanship of Rt Hon George Reid (a former Presiding Officer in the Scottish Parliament) met for the first time on 1 October)
  • The operation of similar small nation and ‘regional’ legislatures in the rest of the UK, across the EU, and within Australia and

Chapter 3 examines the debate within the National Assembly itself. While most AMs are wary of pronouncing publicly on the issue, most are in favour of increasing their number. In successive reports the Assembly’s Remuneration Board has recommended improving the research and support services available to AMs to help them overcome the problems they are encountering in scrutinising legislation and effectively holding the Welsh Government to account. Yet this is a second best option to actually increasing the number of AMs to at least bring their number closer into line with the other devolved legislatures in Edinburgh and Belfast.

Chapter 4 examines the opportunity costs of retaining the status quo. Where democratic representation and scrutiny are concerned, size certainly matters. The notion ‘more politicians cost more’ misses the value that enhanced democratic scrutiny can bring. Focusing on the cost of change underestimates the cost of staying the same. To make a balanced assessment of how many AMs we need to run the Assembly effectively, to develop policy, and to hold the Welsh Government to account, we should have full knowledge of the facts. There are wholly predictable consequences of opting for an inadequate number of elected representatives.

Chapter 5 details the cost of more AMs in terms of their pay and staffing. The costs are compared with those of representatives elsewhere in the UK, allowing us to assess the relative value of more AMs.

This examination suggests strongly that the present allocation of elected representatives between local authorities, the National Assembly and Welsh MPs and Peers at Westminster is poorly distributed.

More could be made of existing resources if we had fewer MPs and Peers at Westminster, or fewer paid councillors, but a greater number of AMs in Cardiff Bay.

Chapter 6 examines the numbers of elected politicians that currently serve similarly sized regions and countries elsewhere in the world. The average number in these institutions is 80 but their electoral ratios (members per head of population) are considerably lower than in Wales. A comparison with equivalent institutions in terms of their powers and the populations they serve indicates the National Assembly should have 100 members.

Chapter 7 describes the numbers of elected members proposed for different models of devolution in Wales at various times during the second half of the 20th Century. There were six detailed proposals in that period and all recommended a greater number than the 60 contained in the 1998 Wales Act – a number arrived at as a consequence of arbitrary political manoeuvring.

At the outset of its first report, the Silk Commission noted that “Polling repeatedly shows that there has been a growing acceptance over the last 15 years that devolved government has become a fixture of the constitution.” (Silk Commission First Report, Empowerment and Responsibility: financlia powers to strengthen Wales, 1.2.1., November 2012.).

Democratic devolution is now demonstrably the settled will of the people of Wales. Yet the history shows that the politicians have been some way behind the people in forging our institutions. The Welsh devolution process has edged forwards in crab-like steps, more in response to the political pressures of the day than to a desire to put in place a coherent, unified and stable structure of the kind that the Richard Commission recommended. The Silk Commission now provides us with another opportunity to establish a stable constitution for Wales that will endure. Part of this must be a National Assembly with sufficient members to undertake their role effectively.


Too Few in Number?

There is a broad consensus amongst AMs in the National Assembly that they are too few in number to properly hold the Welsh Government to account or to scrutinise the growing volume of legislation for which they are responsible. However, for fear of the media and the voters, most AMs hold this view privately. Very few are ready to make the case in public.

An exception has been the Presiding Officer Rosemary Butler. In her evidence to the Silk Commission she argued that the number of AMs should be increased from 60 to 80. (Rosemary Butler AM, Letter to Paul Silk, Chairman of the Commission on Devolution, 28 February).

She pointed out that, nearly a decade ago, the cross-party Richard Commission on the Powers and Electoral Arrangements of the National Assembly made a similar recommendation. It said that were the Assembly to gain the legislative powers it recommended, 80 members were essential if it was to discharge its roles of holding the Government to account, making laws for Wales effectively, and representing the people of Wales.

Of course, following the referendum in 2011 the recommendations made by the Richard Commission in relation to the Assembly’s legislative powers have been largely implemented. However, those relating to its size have not. As Rosemary Butler put it:

“I believe that the expectations evident on Members today, let alone any extension of the institution’s powers and responsibilities, makes the case for more Members more compelling than ever.” (Ibid)

In its March 2011 report Fit for Purpose, the Assembly’s Remuneration Board said it had heard evidence from a broad range of AMs of the increasing pressures on their time as the powers of the Assembly had developed. Theoretically, 47 of the 60 AMs are available to undertake scrutiny after the deduction of the 13 Ministers and Deputy Minsters. However, the Presiding Officer and Deputy Presiding Officer restrict themselves to the Business Committee, the Constitutional and Legal Affairs Committee, and the Committee for Scrutiny of the First Minster. In addition, the Conservative and Plaid leaders and the Labour Chief Whip generally do not attend committees. So, in practice, the actual number of AMs available for scrutiny work is 42.

This number is extremely small when compared with the other UK legislatures, as shown in Table 1 below. In Wales both the number and percentage of members available to undertake scrutiny and legislative functions is substantially lower than at Westminster, Holyrood, or Stormont. As Rosemary Butler stated in her evidence to the Silk Commission:

“… the pressure on Members of the Assembly is very different to those of its larger counterparts elsewhere in the UK. Necessarily, most Assembly committees have very broad remits, certainly stretching beyond the portfolio of a single Minister, and are responsible for the scrutiny of legislation, policy and finance within those remits. In addition to these high committee demands, the majority of Members will be active every week in plenary – again, in contrast to larger parliaments where the opportunity to question or participate in statements or debate will come along less frequently. The multiple roles, as office holders and party spokespeople, which many Members must inevitably assume within a small legislature, add to the pressure on Member capacity and bring distinct institutional challenges.” (Ibid)

Table 1: Members of UK legislatures available to undertake scrutiny and legislative functions

Total number of seats Members available Percentage
UK Parliament 650 522 85%
Scottish Parliament 129 113 88%
Northern Ireland Assembly 108 92 85%
National Assembly for Wales 60 42 70%

Note: Table based on Remuneration Board of the National Assembly for Wales report, Fit for purpose, para. 142, 2011. The comparative statistics were calculated by the National Assembly Members’ Research Service using information provided on the websites of the Governments and Legislatures in question.

The relatively small number of AMs has a major impact on the operation of the Committees, where most scrutiny is undertaken. (Hansard Society, Assembly Line? The Experiences and Development of new Assembly Members, p 28-9) There are simply too few members to populate these committees, and they have insufficient time to develop the specialist expertise that will optimize their effectiveness. On occasion, this can work in a perverse way in the opposite direction. On occasion, Ministers have exited the Welsh Government only to find themselves on policy committees scrutinising proposals introduced by themselves whilst still part of the Executive. (One example was former Finance Minister Andrew Davies. He stood down from the Cabinet in Autumn 2009 and subsequently served as a backbencher on a number of Assembly committees. In these roles, he participated in an examination of proposed legislation concerning the Rights of the Child, about which he previously had participated in Cabinet discussions. He also served on the Finance Committee which, for example, questioned UK Treasury Minister Danny Alexander about the Barnett formula in which he had been previously involved as a Minister in detailed inter-governmental talks.) That such situations arise is a consequence of the fact that there are simply not enough backbench AMs to go around.

In the third Assembly, 13 AMs were members of four or more Committees, specialising on a broad range of issues as a result. This reduced scrutiny to a level of tinkering around the edges, resulting in a dominant Executive that was able to push through its agenda with relative ease. In the current Fourth Assembly there has been a drastic cut in the number of committees coupled with a removal of the separation between legislative and non-legislative scrutiny. The result has been fewer committees with a wider range of work in each. However, this has not reduced the range and scope of issues with which AMs have to deal.

Much of the pressure on AMs still comes from the number of committees they have to attend. There are currently nine AMs on three or more committees, including the Business Committee. Table 2 shows the number of committees attended by 44 backbench AMs between September 2012 and June 2013, as surveyed by the Assembly Commission. The highest number of meetings attended by one backbencher was 87.

Table 2: Committee meetings attended by backbench AMs, September 2012-June 2013

Committee Meetings Backbench AMs attending
10 to 20 2
20 to 30 8
30 to 40 9
40 to 50 8
50 to 60 4
60 to 70 9
70 to 80 4

Source: National Assembly Commission internal statistics.

Assembly Members find themselves in a weekly cycle of committee work, demanding a high level of specialised policy, legislative, financial and procedural expertise. These are timetabled around two plenary sessions where a high level of attendance and participation is the norm. AMs are always in a hurry, constantly moving from one meeting or issue to another. Many say that they do not always have time to read, let alone reflect properly on their documents ahead of meetings. (These views have been expressed by AMs in response to surveys and interviews carried out by the Remuneration Board of the National Assembly: see Annual Reports for 2011, 2012 and 2013) The inevitable result of this pressure and stress is to reduce the quality of scrutiny and inhibit the production of alternative policy proposals.

All this was the case even before the National Assembly gained primary powers as a result of the 2011 referendum. For instance, the third Assembly Commission’s Legacy Report stated that the work of committees between 2007 and 2011 were characterised by the following features:

  • A higher volume of work – plenary sat for longer and committees were more active than in previous Assemblies; use of the Research Service rose inexorably; the legislative programme grew steadily; and the development of legislative proposals by individual Assembly Members and committees generated a huge amount of demanding work for researchers, lawyers and clerks and longer hours for supporting
  • Greater complexity and sophistication in the services required by Members throughout the legislative process and in fulfilling their scrutiny role in committee; the procedural advice required from Legislation Office and Chamber staff; the legislative drafting and advice delivered by the Legal Service; specialist legal translation; and the combination of policy and procedural advice demanded by independent scrutiny and other (National Assembly for Wales, Assembly Commission Legacy Report 2007 to 2011, March 2011)

Pressures on the capacity of the small number of AMs to engage effectively with their responsibilities has led the Remuneration Board to seek to compensate by increasing their allowances for hiring support staff and undertaking research. In the current 2013-14 financial year the staffing allowance for each AM has been increased to £89,890 to allow them to appoint three staff members to support their research, policy and constituency work. In addition, the Board has stated it is giving consideration to awarding a further increase in future years to allow each AM to appoint a Senior Adviser:

“From evidence submitted by the Presiding Officer and others on the impact of the Assembly gaining full legislative powers, we recognise the increased responsibilities for Members in formal Assembly Business in a legislature of only 60 Members. We consulted on a proposal to establish a Senior Advisor position to provide high-level support to individual Members on matters of strategic significance locally and nationally, and in relation to formal Assembly business and policy. Our justification for this proposal is clear and we remain firmly of the view that the capacity of the Assembly and its Members should be enhanced in order to meet their complex new responsibilities. Whilst many welcomed this proposal and felt it would be of real benefit to them as Members and to the Assembly as a legislature, it did not receive across-the-board support. We therefore do not intend to proceed with this proposal for the current financial year but shall leave it on the table for consideration in 2014.”

(Remuneration Board of the National Assembly for Wales, Annual report 2012-13, 82-84, July 2013.)

This is further evidence that with just 60 members the National Assembly does not have the capacity to engage effectively with its duties. As the Remuneration Board adds, “We should look to enhance the strategic capacity of an Assembly tasked with scrutinising a very substantial volume of policy and legislative work.” (Ibid, para 65.) Given its remit, the only option the Remuneration Board has is to come up with proposals for increasing the support staff and research funding available to AMs. Yet this is second best to increasing the number of AMs themselves.

One argument for putting more resources into support services is that it is not so much the number of politicians that is important, but rather how well prepared they are – that is to say, it is better to have 60 well-briefed and well-informed AMs than 80 or 100 ill-informed or badly briefed AMs. But this is to miss the point that time and pressure is a major factor, however well briefed an AM might be. As Anna Nichol, a Special Adviser in the Welsh Government during the 2007-11 One Wales coalition, said:

“There’s only so much that better research and briefing by support staff can achieve when Members themselves don’t have a chance to read papers or prepare for committees. I’ve seen Ministers run rings around AMs in committee because they know that they don’t have time to read briefings properly, however well-researched they are. AMs can read out pertinent questions, but they’re easy for Ministers to deflect when it’s clear the committee member doesn’t understand the question themselves.”

(E-mail communication, 12 September 2013.)

Putting an emphasis on support services as an alternative to more AMs is also to miss the wider democratic imperative, and the case for more effective civic engagement across Wales. Given that it now has equivalent legislative responsibility why should Welsh democracy be less well served than is the case at Holyrood or Stormont? And given that 80 or 100 AMs would have the same back-up resources as the present 60 AMs, why should they be any less informed? Certainly, a greater number of AMs would have a greater opportunity of getting on top of their job of holding the government to account and providing more effective scrutiny of legislation than is presently the case.

Why size matters

Costs of Ineffective Scrutiny

The most common argument against more politicians is that it will add costs at a time of austerity. However, there are hidden costs involved in remaining at the present 60 members. These are the costs of ineffective scrutiny, ineffective opposition, ineffective checks and balances, and a general lack of accountable government.

The powerful Cabinet and executive that now operates within the Welsh system requires equally powerful scrutiny. Effective scrutiny also opens up alternatives to policies proposed by government. But if AMs and their research staff do not have the capacity to develop a depth of knowledge, their alternatives are less likely to have traction. This itself can be costly in terms of impoverished political debate and the calibre of the manifestoes presented to the electorate. It might be suggested that in the UK Parliament relatively little policy development comes from MPs at Westminster. The political parties tend to rely on aligned think tanks and ad hoc policy groups among whom elected representatives are a small minority, to generate new policy thinking. In Wales this simply isn’t the case. Welsh civil society is poorly served by organisations that generate policy. There are relatively few think tanks and those few that are aligned to the parties are poorly resourced. Overwhelmingly policy development in relation to the production of the Welsh manifestos comes from within the party groups inside the Assembly.

Moreover, there is evidence that more effective scrutiny – that would result from having more members – would save money. In a recent study the Swedish economist Per Petterson-Lidbom found that, for every new legislator, there was a corresponding cut of 0.5 per cent in spending. He explains that this is due to better efficiency through better scrutiny of complex public policy fields. As he puts it, “more legislators can better control a budget”. (Journal of Public Economics, 96, Issues 3–4, 269-278, 2012.) Greater scrutiny capacity brings greater efficiency in spending. Such savings may not be inevitable, but Petterson-Lidbom’s work does undermine the argument that some have put forward, to the effect that larger legislatures may lead to greater government spending. (There are also a number of different possible variables according to other academic studies in other countries. For example, research into Swiss Cantons points to the size of cabinet as having an effect on spending, and the importance of an equivalent of an Office for Budgetary Responsibility in promoting efficiency. See Schaltegger A. and Feld L., ‘Do large cabinets favour large government? Evidence on the fiscal commons problem for Swiss cantons’, Journal of Public Economics, 93, 35-47, 2008)

Cabinet structure, scrutiny and oversight

To understand the impact the size of the National Assembly has on effective scrutiny we need to understand the relationship between the legislature and executive within the overall culture of the institution. As a result of proportional representation, Cardiff Bay operates a more consensual style of governance compared with the majoritarian culture in Westminster. Table 3 summarises the differences between the two, as analysed by the political scientist Arendt Lijphart.

Table 3: Lijphart’s classification of political institutions*

Westminster model Consensus Model
Executive Single party Cabinet system Multi-party coalition

Cabinet  Government

Balance of powers Executive dominance Executive-Legislative balance
Electoral system Majoritarian Multi-party system

*Adapted and simplified to include only the issues raised in this paper.

Wales lies somewhere between these ideal types of governance. Our semi-proportional system (where a third of Assembly is elected on the proportional regional list) allows for a greater level of party competition than is the case with the Welsh MPs elected by Westminster’s majoritarian first- past-the-post system. In the absence of a second chamber, the pluralism promoted by proportional representation becomes more important in order to provide checks and balances and promote better scrutiny.

However, in terms of the executive and the centralisation of government power, the National Assembly has moved from one side of the spectrum to the other. The original 1998 Government of Wales Act established the National Assembly as a corporate body, with the Assembly as a whole taking an active part in formulating policy under a consensus model. The 2006 Act changed this to a Westminster model of centralised Executive dominance. Legislative power is now exercised by the Cabinet, with the rest of the Assembly Members responsible for scrutinising it.

The present Assembly is split between the governing party and the opposition, with 30 on each side. During the previous ‘One Wales’ coalition, 41 members were either from Labour or Plaid Cymru, leaving only 19 AMs from a non-governing party. This did not provide for a strong opposition or effective scrutiny. Currently, the Labour Party has a technical majority. However, it is likely that coalition arrangements will continue to be a regular occurrence, as happened between 2000-3 and 2007-11. Therefore, we can expect in future relatively weak opposition within the Assembly as a direct result of the small number of AMs remaining outside the governing parties.

The overall result is to strengthen the executive and weaken the legislature. Following Lijphart’s model, the limited capacity of the Assembly has resulted in a move from ‘Executive-legislative balance’ to ‘Executive dominance’. As a result of the 2006 Government of Wales Act, and confirmed by the 2011 referendum, the Assembly now has primary legislative powers. Yet it now also has a more limited capacity to amend and repeal laws and hold the Executive properly to account.

The workload of the Committees can only grow and strain the system further as the Assembly gains a wider range of powers. The danger is of an over-mighty Executive with too few AMs to hold it to account effectively.


Co-option of experts

Some evidence to the Silk Commission has suggested co-opting Assembly Associates to work as non-voting members of Committees. They would be experts who would be able to advise AMs and question witnesses. Advocates note that co-option is used extensively in local government.

However, such a system of appointees runs counter to the British parliamentary tradition of direct

representative democracy, a tradition which has been followed in the creation of the devolved institutions in Cardiff, Edinburgh and Belfast. The tradition reflects the inherent dangers in the co-option of ‘experts’ as an alternative to more elected representatives. ‘Experts’ are not politically neutral and should serve elected representatives rather than displace them. Technocracy is not the best solution to a scrutiny gap. Creating Assembly Associates would be a move away from parliamentary custom and practice and could undermine the authority of the National Assembly in the eyes of the electorate.

Furthermore, ‘Assembly Associates’ would not be cost-free. The costs of the kind of Associates likely to make a significant contribution to the Assembly’s deliberations would be likely to exceed those of AMs themselves. Any additional expenditure on improving scrutiny should be deployed in strengthening democratic provision.

Arguments for a second chamber

Some have suggested that another route to plugging the Welsh scrutiny gap would be the creation of a Second Chamber, either provided for within Wales or amongst Welsh MPs at Westminster. Prior to the 2011 Referendum, the True Wales ‘No’ campaign claimed that a unicameral system was inherently dangerous. They said that the Legislative Competence Order system that existed at the time was more desirable because the Welsh Affairs Select Committee functioned as a de facto second chamber.

Yet the function of the Welsh Select Committee was not to scrutinise Welsh legislation, but rather to consider whether the Legislative Competence Order which would provide the Assembly with a specific power should be made. Moreover, the system was cumbersome and inefficient, lacked clarity and blurred the division of responsibilities between Cardiff Bay and Westminster. Sharing power at different levels is particularly difficult for a model of governance guided by precedent and custom, such as the UK.

Another suggestion is that Welsh Peers form a Committee in the House of Lords to as an upper chamber charged with revising legislation made by the National Assembly. This would have all the disadvantages of the same task being carried out by Welsh members in the House of Commons, but with the added difficulty that peers are unelected. It is hard to imagine the National Assembly agreeing to its democratic legitimacy being diluted through such a device.

It has also been suggested by some that a Second Chamber be established at Cardiff Bay. If this were to mean a mini House of Lords for Wales, this would require the election of a significant number of ‘Welsh Senators’. The Australian Regional Assemblies use two chambers. The closest in size to Wales, Western Australia with 2.4 million citizens, has an upper chamber of 36 members which, including the lower chamber of 59 adds up to 95 legislators in total. This suggests that at a minimum a Welsh upper house would require between 30 and 40 senators. However, it is surely difficult to justify such a second chamber when we could create additional Assembly Members to do the job, without all the attendant costs of a new governmental chamber. Further, a second chamber would still not obviate the argument for expanding the membership of the primary chamber. And in any case there are many examples of successful unicameral parliaments around the world, including in New Zealand, Israel, Denmark and Sweden.

This chapter has identified the opportunity costs, in terms of scrutiny and democratic accountability that will be incurred as the National Assembly expands the scope and use of its legislative powers. The Executive dominance of the Assembly has increased substantially since 1999, and especially since the 2011 referendum approved moving to full legislative powers. As a consequence there are real and predictable costs, in terms of both capacity and democratic representation, in the National Assembly staying with just 60 members. In these vital considerations, size does matter.

Relative costs and value in levels of Welsh governance

Current Expenditure

If cost is to be raised as a barrier to increasing the size of the Assembly, it is logical to look at the relative costs of the different levels of governance in Wales. There is a strong case that, as more powers accrue to the National Assembly, the numbers of elected representatives distributed between local government, Cardiff Bay and Westminster are skewed towards the wrong places.

The current expenditure on AMs, including their pay, expenses, support staff and equipment is £13.5m per year. (Assembly Commission Budget 2012-13, p 3, available at http://www.assemblywales.org/gen-ld8713-e.pdf). Therefore, the annual average cost per Assembly Member is £225,000. By comparison, the average expenditure per representative in other legislatures is as follows:

Legislature Average expenditure per representative
House of Commons £590,000
House of Lords £130,000
European Parliament £1,790,0002

(Hansard House of Lords, Answer to written question)

Under the UK Government’s proposals to equalise the size of Welsh constituencies to those of the rest of the UK, the number of MPs would have fallen to 30. As these proposals were deferred, the Welsh position remains anomalous among the devolved territories. (It is important to note that the legislation to cut the number of Welsh MPs by a quarter, from 40 to 30, has only been deferred – to 2017 – and so remains on the statute book. It has not been put into  force  in this  parliament because of a refusal by the Liberal Democrats in the coalition to support it because of the failure of the AV Unless the new UK Government after the 2015 general election  brings  in  new  legislation,  the reduction in the total of MPs for the UK to 600, including the reduction in the number of Welsh MPs,  will  take effect from 2018 – that is, ie in time for the UK general election in 2020 and the Assembly elections in 2021.) Scotland cut the number of its Westminster MPs from 72 to 59 in 1999. The ratio of Welsh MPs per head of population is 1:76,586, whereas the average MP per head of population in the UK as a whole is 1:97,202.

While the number of AMs is very small, there is a relatively high number of Welsh MPs per head, when compared internationally. Table 4 shows the number of representatives per head of population in the largest nation-states within the European Union. It will be seen that while the UK has the highest number of representatives per head, when considered separately Welsh MPs score even higher. If the UK as a whole followed the ratio for Welsh MPs, there would be 825 MPs in the House of Commons.

At an average of £590,000 a year, the cost of an MP is equivalent to that of 2.6 Assembly Members. Presuming these average expenditures across the UK are also reflected among Welsh MPs (and there appears little reason to think they would vary greatly), cutting ten MPs would, theoretically, save £5.9 million a year. Of course, as things stand these savings would simply go directly to the Treasury through a reduction in the Vote for the costs of Parliament. It would be up to the Welsh Government of the day to make a case to the Treasury that the funding should be adjusted to allow the savings to come to Wales as part of an adjustment to the Welsh block.

Table 4: Ratio of Members per population in EU legislatures

Lower Chamber size Population Electoral Ratio

(1 member per head of population for the lower chamber)

Spain 350 46,073,000 131,637
Italy 630 60,498,000 96,028
France 577 64,848,000 112,388
Germany 598 81,757,000 136,717
UK 650 63,181,778 97,202
Wales –
Westminster 40 3,063,456 76,586

Although we have given indicative figures for both savings and additional costs below, it is worth noting that savings may be lower than the theoretical figures, as some fixed costs will remain.

Similarly, the figures for additional costs may be less than we have stated, as there are sure to be some economies of scale. However, we believe that some broad indication is needed in this debate.

Under the current UK government an additional 125 peers had been elevated by June 2013, with the likelihood of more being appointed in order to comply with a pledge in the Coalition Agreement to ensure a House of Lords that reflected the general election vote share in 2010. (Crick, M, ‘Cameron promises to cram yet more peers into the Lords’)

David Cameron has appointed peers at a faster rate than any previous Prime Minister in history.

Given that this pledge sets a precedent, if the vote share were radically different in 2015 and a different government formed, it is plausible that new peers would be created with the aim of addressing the difference again. Extrapolating from the £130,000 annual expenditure per peer, the cost of 125 new peers is £16.25 million a year. The annual expenditure of these 125 new peers alone is equivalent to the annual expenditure of 72 Assembly Members.

Wales has 1,254 councillors compared with 1,222 in Scotland. This is despite Scotland having a population of 5.3 million compared with 3.1 million in Wales. The ratio of councillors per head in Wales is 1:2,742, and 1:4,270 in Scotland. For the UK as a whole, the ratio is 1:2,860 (E Bort, R McAlpine, G Morgan, The Silent Crisis, Jimmy Reid Foundation, page 8,).

The relatively large number of Welsh councillors is being investigated by the Welsh Government’s Commission on Public Service Governance and Delivery, due to report by the end of 2013. It has been mooted that the 22 Welsh county councils could be cut by half, with some suggesting they should be coterminous with the seven Welsh health authorities.

The basic salary of a councillor is £13,175. If the number of councillors in Wales were cut by half, there could be an annual saving of £8.3million, without taking into account expenses or additional wages. This is possibly double the cost of 20 more Assembly Members with all their staff and expenses and would almost match the cost of 40 more AMs on the same basis.

Cutting the number of councils by half would also reduce the number of Council Leaders, which would be a further saving of £522,000. If we take the average council cabinet to be eight members with a medium pay scale of £28,780, this would be yet another saving of £2.5 million.

Table 5 compares the relative extra costs of AMs, MPs, Lords and Councillors in light of the various proposals for increasing or decreasing them. The level of representation in local government and at Westminster indicates that the Welsh electorate are over-represented at both these levels. On the other hand, they are significantly under-represented in the National Assembly.

Table 5: Comparative costs of elected members

Annual cost
20 additional AMs £4.5m
40 additional AMs £9m
125 Peers £16.25m
10 MPs £5.9m
632 Councillors (basic rate) £8.3m

These admittedly broad-brush figures are shown to put the costs of more Assembly Members in their proper context of the democratic representation in Wales at different levels of governance. They are not necessarily an argument for reducing the numbers of councillors or MPs. But they do demonstrate the relative value for money of a larger number of AMs.

As quoted at the outset, the annual average cost per AM, including pay, expenses, support staff and equipment, is £225,000. On top of this there are various allowances, including additional funding for party groups, and the costs of the Assembly Commission which need to be taken into account. These are itemised in detail in the Appendix. The current annual National Assembly budget is £49.5 million. Our calculations find that adding 20 more AMs would increase this amount to £54.8 million. Adding 40 AMs to make a chamber of 100 members would increase the cost to £59.6 million.

Given the benefits that would flow in terms of increased accountability, better scrutiny of legislation, and enhanced policy development, an extra £10.1 million for 40 more AMs, would be a small price to pay. Moreover, our examination of the relative value of elected representatives at the different levels of Welsh governance suggests that the present allocation between local authorities, the National Assembly and Welsh MPs and Peers at Westminster is poorly distributed. More could be made of existing resources if we had fewer MPs and Peers at Westminster, and fewer paid councillors, but a greater number of AMs in Cardiff Bay.

Size of the National Assembly in an international context

Regional Authority Index

The analysis that follows draws upon the Regional Authority Index, a study that examined the relative autonomy of Regions within 42 countries (L Hooghe, G Marks, A H Schakel, The Rise of Regional Authority: A Comparative Study of 42 Democracies, Routledge: 2009. The analysis presented here simplifies the index – as we are only interested in the extent of autonomy and not in the scoring of the sharing of power in federal states, which the index also measures. In this chapter, the term ‘region’ is used for empirical clarity as defining geographical areas that lie between the ‘nation-state’ level and the ‘local’, such as local authorities within the UK.) The index gives a total score of between 0-15 for the following characteristics:

  • Institutional depth: The extent to which regional government is autonomous (scores of 0-3).
  • Policy scope: The range of policies for which a regional government is responsible (0-4).
  • Fiscal autonomy: The extent to which a regional government can independently tax its population (0-4).
  • Representation: The extent to which a regional government is endowed with an independent legislature and executive (0-4).

Table 6 gives the score for autonomous regions within Europe, North America and Australia. It shows that even within different unitary and federal structures within these jurisdictions, the score for self-rule remains fairly constant, ranging from 12-15 (and mostly at 13 or 14). This demonstrates that, for all their other differences, in terms of their respective powers it is possible to compare them in a meaningful way. Putting a changing Wales into the Index shows how the Silk Commission proposals on taxation will place Wales within the range of regional autonomy across the democracies. We will also see how Wales compares with other Regions that have a similar level of autonomy.

These scores can also be used to gauge changes over time, with an international trend towards increasing regional authority. The scores in the Regional Authority Index date from 2006. So it is necessary to score Wales from the same date, prior to the Government of Wales Act of 2007 coming into force.

Table 6: Relative autonomy for selected regions in Europe, North America, and Australia

Institutional Depth Policy Scope Fiscal Autonomy Representation Total Score
Italy Regions 3 3 3 4 13
Spain – Historical Nations
Basque Country 3 3 4 4 14
Catalunya 3 3 3 4 13
Galicia 3 3 3 4 13
Andalucia 3 3 3 4 13
Navarre 3 3 4 4 14
Spain Autonomous Communities  










German Lander 3 3 2 4 12
Scotland 3 3 3 4 13
Northern Ireland 2 3 0 3 8
Australia (States) 3 4 2 4 13
Canada (provinces) 3 4 4 4 15
USA (states) 3 3 4 4 14

Table 7 describes the increasing regional autonomy of Wales across this period. It also looks ahead to assume Wales achieving some control over minor taxes and borrowing powers as a result of the Silk Commission recommendations being implemented by, say 2020, plus the possibility of control over income tax within circumscribed limits as well as devolution of police and criminal justice.

Some of these scores are open to interpretation. For instance, on policy scope there are some grey areas when applied to developments in Wales. However, the Index does capture the fact that

adopting the Silk recommendations would see Wales approaching the norm for regional authorities in other nation-states.

Of course, it is difficult to make direct comparisons. When examining the size of legislatures their different executive structures – presidential, cabinet or corporate – need to be taken into account. With just 25 members the London Assembly is very small. However, it does not have executive functions. These are reserved to the mayor Boris Johnson, who is directly elected outwith the Assembly. As such, the London Assembly does not produce a strong Cabinet Government, in contrast with Wales where the Cabinet is recruited from within the Assembly as a whole. The same is true, for example, for Italian regional assemblies.

Table 7: Levels of Welsh autonomy by timeline

Year Institutional Depth Policy Scope Fiscal Autonomy Representation Total Score
Wales 2006 1999–06 3 2 0 3 8
Wales 2007 2007-11 3 2 0 4 9
Wales 2011 2011- 3 3 0 4 10
Wales with limited tax base  


3 3 2 4 12
Wales with Income Tax powers (within limits), control of police, and powers over own institutional arrangements ?2020 3 3 3 4 13

Nevertheless, it is certainly true that, by all international comparisons, the National Assembly is a small legislature. Table 8 lists European national parliaments by the size of their respective

legislatures, their electoral ratio, and population. It can be seen that the average size of the lower chambers of the parliaments in the countries of the EU is 267 members. If we take the countries of 10 million or less the average is 166, while those serving populations of between 1 and 6 million people is not much less, at 142.

Therefore, even excluding second chambers, it is clear that by comparison the National Assembly for Wales is a remarkably small institution. When looking at the electoral ratio – that is, how many elected representatives per head of population – Wales is again shown to be far less well represented than the EU average.

However, if we were to focus on smaller countries with populations below 10 million it is clear that the Welsh Assembly is by far the least representative in terms of elected representatives per head, with a ratio of one member per 51,058 people. The average in the EU as a whole on this score is one per 39,100. For countries up to 10 million it is one per 22,122, and for those in the population range of 1-6 million it is one per 23,566.

Table 8: European parliaments by size and electoral ratio

EU Country Size Lower House Size Upper house Population Electoral Ratio
Member:Population (1:)
Malta 65 416,000 6,400
Luxembourg 60 506,000 8,433
Cyprus 80 804,000 10,050
Estonia 101 1,340,000 13,267
Slovenia 90 40 2,049,000 15,761
Latvia 100 2,239,000 22,390
Lithuania 141 3,287,000 23,312
Ireland 166 60 4,476,000 19,805
Finland 200 5,363,000 26,815
Slovakia 150 5,430,000 36,200
Denmark 179 5,546,000 30,983
Bulgaria 240 7,534,000 31,391
Austria 183 62 8,388,000 34,236
Sweden 349 9,378,000 26,871
Hungary 386 10,000,000 25,906
Czech Republic 200 81 10,517,000 37,427
Portugal 230 10,637,000 46,247
Belgium 150 71 10,883,000 49,244
Greece 300 11,305,000 37,683
Netherlands 150 75 16,612,000 73,831
Romania 412 176 21,431,000 36,447
Poland 460 100 38,187,000 68,191
Spain 350 266 46,073,000 74,793
Italy 630 321 60,498,000 63,615
UK 650 760 61,990,000 43,964
France 577 348 64,848,000 70,105
Germany 598 69 81,757,000 122,574
Average 267 187 18,540,518 39,109

As shown in Table 9, Scotland and Northern Ireland have substantially more representatives than Wales in terms of their absolute number. Their electoral ratios are also more favourable than the ratio for Wales. Scotland’s Parliament is more than double the size of the National Assembly, and despite having less than two-thirds Wales’ population Northern Ireland is approaching double.

Table 9: Scotland and Northern Ireland - population and representation

Size of Chamber Population Electoral Ratio Member: Population
Scotland 129 5,295,000 41,046
Northern Ireland 108 1,810,863 16,767
Wales 60 3,063,456 51,058

As the Regional Authority Index illustrates, the National Assembly is approaching the level of autonomy that Scotland was allocated in 1999. This again raises the question of whether it has enough representatives to fulfill its new functions effectively.

Table 10 provides the same statistics for European countries with strong regional legislatures within larger unitary or federalised countries. The Spanish figures are split into three types, reflecting different levels of devolution. In the more autonomous ‘historical nations’, the average electoral ratio is about one elected representative for every 40,500. Where there is a substantially higher population, such as in Catalunya, Andalusia and Valencia, the electoral ratio is higher than in other Spanish regions and in comparison with the National Assembly. In terms of the size of these legislatures, there are also substantially greater numbers of members in comparison to Wales and the other regions. For instance, Catalunya elects 135 members, Andalusia 109 and Valencia 99.

Table 10: European Regions – population and representation

Size: Lower chambers Population Electoral Ratio
Member:Population (1:)
Spain – Historical nationalities
Catalunya 135 7,565,603 56,042
Galicia 75 2,778,913 37,052
Basque 75 2,155,546 28,741
Andalusia 109 8,424,102 77285
Aragon 67 1,277,471 19067
Balearic Islands 59 1,106,049 18747
Canary Islands 60 2,117,519 35,292
Valencian Community 99 5,111,706 51,633
Asturias 45 1,076,896 23,931
Cantabria 39 593,121 15,208
Castille – La Mancha 49 2,095,855 42,773
Castile and Leon 83 2,558,463 30,825
Extremadura 65 1,097,744 16,888
La Rioja 33 308,968 9,363
Murcia 45 1,470,069 32,668
Baden-Wurttemberg 139 10,786,227 77,598
Bavaria 187 12,595,891 67,357
Berlin 141 3,520,061 24,964
Brandenburg 88 2,495,635 28,359
Bremen 83 660,999 7,963
Hamburg 121 1,788,000 14,776
Hesse 118 6,066,000 51,406
Mecklenburg-Vorpommern 71 1,639,000 23,084
Lower Saxony 152 7,914,000 52,066
North Rhine-Westphalia 237 17,837,000 75,262
Rhineland-Palatinate 101 3,999,000 39,594
Saarland 51 1,018,000 19,961
Saxony 132 4,143,000 31,386
Saxony-Anhalt 97 2,331,000 24,031
Schleswig-Holstein 69 2,833,000 41,058
Thuringia 88 2,231,000 25,352
Italy – regions
Abruzzo 40 1,343,000 33,575
Aosta Valley 35 128,000 3,657
Apulia 70 4,091,000 58,443
Basilicata 30 587,680 19,589
Calabria 50 2,009,227 40,185
Campania 60 5,83,3891 97,232
Emilia-Romagna 50 4,429,766 88,595
Friuli-Venezia Giultia 60 1,23,5270 20,588
Lazio 70 5,789,993 82,714
Liguria 40 1,613,369 40,334
Lombardy 80 10,010,865 125,136
Marche 40 1,568,835 39,221
Molise 30 478,488 15,950
Piedmont 60 4,467,914 74,465
Sardinia 85 1,673,981 19,694
Sicily 90 5,043,480 56,039
Tuscany 55 3,763,939 68,435
Umbria 30 989,965 32,999
Veneto 60 4,960,661 82,678

All the German Lander have more legislators than the National Assembly, except Saarland which has about a third of the population of Wales. Berlin has a similar population than Wales, yet its legislature is almost two-and-a-half times the size of the Welsh Assembly. Most Lander have smaller electoral ratios than that of the National Assembly. Two have a similar, although slightly larger, ratios – Hesse and Lower Saxony. These have at least twice the population of Wales, at six and almost eight million respectively. They also have substantially more legislators at 118 and 152. All the Lander with fewer representatives per head serve populations of more than 10 million people.

The regions of Italy have an average electoral ratio similar to that of Wales at around 1:52,000. They also have a smaller number of legislators – on average 54.5 legislators. But with the exception of two regions – Aosta Valley and Trentino-Alto Adige/Sudtirol – the executive is comprised of a directly elected regional President, who then selects a cabinet from outside the council. As such, the executive is not created from the legislature. These institutions do not have executive functions as the Assembly does, but more of an oversight function similar to the London Assembly. Given that the National Assembly has just 42 backbenchers (and even fewer from opposition parties) to conduct scrutiny and hold the government to account, the Italian regions have a much greater capacity to fulfill these functions. In general, these figures further confirm that the National Assembly is a remarkably small institution in terms of its size. Where institutions are smaller, they have less autonomy or, as is the case in Italy, they exercise oversight over a president rather than draw the executive from among their own membership.

In terms of electoral ratio, the National Assembly is also less favourably situated than most European regions, again with the exception of Italy. Where regions have fewer representatives per head of population, they have much higher populations and a larger legislature in terms of its number of members. Table 11 illustrates the size of legislatures, population, and representation ratio of the provinces of Australia and Canada.

Table 11: Regional populations and representation in Australia and Canada

Lower House Upper House Population Electoral Ratio Member:population
New South Wales 93 42 7,314,100 54,179
Queensland 89 4,584,600 51,512
South Australia 47 22 1,658,100 24,030
Tasmania 25 15 512,100 12,802
Victoria 88 40 5,649,100 44,134
Western Australia 59 36 2,451,400 25,804
Alberta 87 3,645,000 41,897
British Columbia 85 4,400,000 51,764
Manitoba 57 1,208,000 21,193
New Brunswick 55 751,157 13,657

and Labrador







Nova scotia 52 921,727 17,726
Ontario 107 13,505,900 126,223
Prince Edward Island  






Quebec 125 8,050,000 64,400
Saskatchewan 58 1,033,000 17,810

While the high ratio of members to population in Australia is similar to Wales, Australian states have a substantially higher number of representatives across both chambers. Western Australia, for example, serves 2.5 million citizens and has 95 representatives across its two houses.

Canada’s provincial legislatures are notably small. Nonetheless, on the whole the provincial Assemblies still have a higher level of representation than Wales. Where it is similar, such as British Columbia, the province will have a substantially higher absolute number of representatives than Wales. Canadian provinces and Australian states with fewer representatives than Wales all have substantially smaller populations.

The average size of the elected institutions across all the polities analysed here is 80.3 elected representatives, while the electoral ratio of representatives per head of population is 1:40,262. This includes many that have a far smaller population than Wales. Of those with populations of less than 1 million, the average size is 88.2 representatives per legislature, while their electoral ratio is 1:45,983. If we exclude those with fewer powers than Wales, such as the less autonomous Spanish regions, and also the Italian regions, the average size per legislature comes out at 104 elected representatives, with an electoral ratio of 1:41,661.

Table 12 shows the ratio of Assembly Member per head of population for a National Assembly of different sizes:

Table 12: AMs in National Assembly according to different numbers per head of population

Number of AMs Population per AM
60 51,058
80 38,293
100 30,634
120 25,528
150 20,423

Moving to a National Assembly of 80 members would give Wales an electoral ratio of around the same level as Scotland, and close to the EU average. However, this would still be a low level for a small nation-state within the EU. To achieve the mean level for this category the National Assembly would require around 140 Assembly Members. Compared with other devolved institutions, Wales has a relatively small number of elected representatives. Even when we include all the smaller regions, those with fewer powers, and those that do not have executive functions, the average number is 76. When we exclude those regions, the average is 104.

This analysis of sub-national legislatures suggest that 80 members would be relatively small for a legislature such as the National Assembly that represents more than three million people. Indeed, the comparison suggests that 100 AMs would be closer to the norm.

Different sizes at different times

A short history of recent devolution proposals

A short history of recent devolution proposals for Wales shows how arbitrary was the process that led to our having 60 Assembly Members. During the half-century that preceded the establishment of the National Assembly in 1999, six detailed proposals for Welsh devolution were put forward.

Four of them were contained in Bills presented to the House of Commons, in 1955, 1967, 1977, and 1996. One was put forward by the Royal Commission on the Constitution in 1973. Another was proposed by a Welsh Labour Policy Commission in 1995. All varied in their recommendations about the number of members that should be elected. However, without exception they all recommended a greater number of representatives than the 60 AMs contained in the 1998 Wales Act.

In February 1955 the Merthyr Labour MP S.O. Davies, supported by a number of other Welsh MPs, published his Private Member’s Bill for a Welsh parliament. Its centrepiece was a parliament in Cardiff composed of 72 senators, two for each of the 36 Welsh constituencies. Their powers would be similar to those exercised by the Northern Ireland Stormont, leaving responsibilities for defence, foreign policy and overseas trade with Westminster. On St David’s Day 1967 Emlyn Hooson, MP for Montgomery and Leader of the Welsh Liberal Party, introduced a Government of Wales Bill proposing a senate of 88 members.

It was noteworthy that, when the Kilbrandon Commission on the Constitution (1969-1973) came to consider the size of its proposed legislative assembly for Wales, it was influenced by consideration of the small size of the former Stormont Parliament in Northern Ireland. That had 52 members – not far short of the number currently in the National Assembly – though, of course, they were serving a population half the size of Wales. Kilbrandon’s recommendation was for a minimum of 108 members for Wales, elected by the Single Transferable Vote proportional system, with at least three for each of the 36 constituencies. As it stated:

“The Northern Ireland House of Commons had 52 members. This number was acknowledged to be too small. The field from which Ministers had to be selected was too restricted; and when Ministerial appointments had been made there remained too few backbenchers.. The precise number of members needed would depend on the extent of the devolved functions, the geographical area covered and the method of working, but we think something of the order of a hundred would be about right.”

(Royal Commission on the Constitution 1969 – 1973, Volume I, Report, para 789. The Commission’s Minority report, which proposed Assemblies for the regions of England as well as Wales, Scotland and Northern Ireland, also recommended chambers with 100)

The eventual legislation that attempted to put the Kilbrandon proposals into effect departed from this recommendation, mainly on the grounds that it would require the introduction of proportional representation (Our Changing Democracy: Devolution to Scotland and Wales, November 1975, Cmnd. 7348, para 177.). Instead, it proposed 72 Members initially, but eventually rising to about 75, using the First Past the Post electoral system. For the first election, the 72 members would comprise two for each of the 36 Welsh Parliamentary constituencies. For later elections the Boundary Commission for Wales would divide the seats into single member constituencies for the Assembly, according to the following formula:

  • Parliamentary constituencies whose electorate was less than 75 per cent of the average would be given one Assembly
  • Parliamentary constituencies whose electorate was not less than 75 per cent and not more than 125 per cent of the average would be divided into two seats and given two Assembly
  • Parliamentary constituencies whose electorate was more than 125 per cent of the average would be divided into three seats and given three Assembly

The White Paper calculated that this system would give the Assembly about 75 members. It was an arbitrary number, devised to ensure that Assembly Members were elected by First Past the Post and returned from single member constituencies.

In 1996 a Government of Wales Bill that recommended 100 members for a Welsh Parliament, was presented to the House of Commons in the name of John Marek, then Labour MP for Wrexham, and supported by prominent MPs across all four parties in Wales and beyond (Government of Wales Bill, HMSO, 10 December). In an Explanatory Memorandum the recommended size of the Parliament (described as a Senedd) was explained in the following terms:

“The number of members in the Senedd is governed by a number of factors: the number of persons serving as Ministers and assistant Ministers; the number of back-benchers from the government party available to serve on the important departmental committees which the Bill requires Ministers to consult before deciding upon the details of any legislative proposals; the demands of the Regional Committees on members’ time and the requirement of equal opportunities for women members as a result of the provision requiring the political parties nominating at least two men and two women in each constituency; and the adoption in the Bill of two adjoining Westminster constituencies to constitute a single Senedd constituency returning five members each so as to facilitate the operation of proportional representation on the Single Transferable Vote system.Hence, provision was made in the Bill for 100 full-time members. A Senedd with less than 90 members would certainly be insufficient to support an effective government and the degree of democratic input provided for in the draft.”

(Parliament for Wales Campaign, Power to the People of Wales: Government of Wales Bill 1997, Explanatory Memorandum, para 16.10.)

Meanwhile, in the mid 1990s Labour began resurrecting its own Welsh devolution policy. An internal Policy Commission, chaired by Executive member Ken Hopkins, secretary to the Rhondda constituency, brought before the party two options in its document Shaping the Vision:

  • Westminster-style ‘first-past-the-post’ – eighty Assembly Members, two each from the 40 Westminster constituencies in
  • A German-style additional member system (AMS) – incorporating a further twenty ‘top-up seats’ allocated on a proportionate basis, so providing 100 members (Welsh Labour Party, Shaping the Vision: a Report on the Powers and Structure of the Welsh Assembly.)

The first option was chosen by a vote at Welsh Labour’s annual conference in autumn 1995. However, within less than a year this recommendation fell foul of Scottish politics. In July 1996, just eight months before coming to power in May the following year, Tony Blair visited Scotland determined to remove from the election agenda Labour’s promise of tax varying powers for the proposed Scottish Parliament. The mechanism was to be a referendum in which the people of Scotland would be asked two questions, one on whether they supported a Parliament and, crucially, a second on whether they agreed that it should have the power to vary income tax by 3p in the pound, either up or down.

By placing the tax decision directly in the hands of the Scottish people, the referendum would remove it from the immediate general election debate. The commitment to a Scottish referendum was therefore driven entirely by the exigencies of Scottish politics and the impact it was feared they might have on Labour’s electoral prospects in the United Kingdom as a whole. However, the most immediate effect was on Wales. It was soon appreciated that it would be impossible to have a referendum in Scotland without also having one in Wales.

For many in Wales, including the Shadow Secretary of State for Wales Ron Davies, Tony Blair’s referendum ploy came out of the blue. Yet within days Ron Davies was locked in negotiations with Blair on the help he would need if he were to lead the Welsh Labour Party through the forthcoming general election and into a referendum. There was one pivotal requirement. Blair would have to lean on the party in Wales to reverse its conference decision eight months before, in October 1995, in favour of first past the post for a Welsh Assembly, and opt instead for some variation of proportional representation. Davies argued that without a commitment to PR it would be very difficult for Labour to win a referendum in Wales. Certainly, without PR it would be impossible to persuade the Welsh Liberal Democrats and Plaid Cymru to campaign for a Yes vote.

There was a further argument that doubtless weighed with Blair. Labour was already committed to the Additional Member system of PR for elections to a Scottish Parliament, the result of protracted negotiations that had taken place during the previous decade within the Scottish Convention. It would surely look anomalous for the party to fight for a Yes vote in parallel referendums in Scotland and Wales for institutions that would be elected by different voting systems.

In any event, Ron Davies won his concession and went on to lead the Yes campaign to the wafer thin majority in the referendum that was held a little over a year later, in September 1997. There is little doubt, given the closeness of the result, that without the PR commitment the referendum would have been lost.

Labour’s commitment to proportional representation for the National Assembly seemed grudging. In fact, compared with the Scottish Parliament and the Northern Ireland Assembly the concession was minimal. Just 33 per cent of the Assembly membership were to be elected by a proportional system, compared with 43 per cent in Scotland and 100 per cent in Northern Ireland – see Table 13.

Table 13: Size and Electoral Systems of Devolved Bodies in UK

UK Devolved Bodies Constituency Members Regional Members Total Members Ratio FPTP: List Ratio Member: Population
Wales 40 20 60 67% : 33% 1: 51,058
Scotland 73 56 129 57% : 43% 1: 41,046
Northern Ireland Elected by STV system 108 n/a 1: 16,767

Source: Richard Commission report, page 229, 2004.

Along with fears of perceived cost, the relatively small concession to proportionality in Wales helps explain why Wales was given only 60 members. Indeed, in an unguarded moment Rhodri Morgan confessed that the Welsh system had been devised to ensure that Labour would achieve a majority of seats in at least three elections out of four (See John Osmond (Ed.), Welsh Politics Come of Age: Responses to the Richard Commission, IWA, 2005, page Experience so far has signally failed to fulfil Rhodri Morgan’s expectations. In 1999 Labour won only 28 of the 60 seats and in 2007 just 26. In 2003 and 2011, in more favourable political circumstances, it barely won half). When the Wales Bill was being put together following the referendum in September 1997 Ron Davies did consider increasing the size of the Assembly.

Another option was to include a clause in the Bill enabling the Assembly itself to increase its size at a later date. However, Ron Davies judged that, although he favoured approaches along these lines himself, opposition amongst Welsh Labour backbench MPs was too strong. It was calculated that about half were in favour of devolution and half against. Any departure from the commitments made in the White Paper that were the basis for the referendum would have provoked opposition and endangered the passage of the Bill. (It was noteworthy, however, that the Bill that eventually emerged from the parliamentary process did depart significantly from the White Paper in grafting a Cabinet system on the goal government-style executive committee structure that had been originally).

Of course, within just five years the Richard Commission was to argue that 80 members would be required. However, apart from the fact that 80 would be better than 60 – on the grounds of workload and scrutiny, a point that is well made in the report – the Commission did not examine the issue in the round, other than to point out that both Scotland and Northern Ireland had considerably more members than Wales. Furthermore, it gave no consideration to what might be an appropriate number for a fully legislative National Assembly, with a separation of powers between the Executive and Legislature akin to the Westminster Parliament, and with the prospect of the devolution of fiscal and criminal justice powers as well.


Good legislation pays for itself

There is no doubt that, with just 60 AMs in the National Assembly, our democracy in Wales is under powered, over stretched and under strain. With the Silk Commission recommending far-reaching and ambitious changes to fiscal powers and responsible government in Wales, it is perverse to be afraid of advocating an Assembly with an adequate membership to do the job properly.

The brief history of the half-century leading to democratic devolution in 1999 illustrated the arbitrary way in which the number of 60 Members of the Assembly came about. In the previous proposals for an Assembly the recommended membership never fell below 75, and generally assumed a figure of around 100.

The international perspective further emphasises that 60 members are extremely few for a legislative Assembly that also provides an Executive. The analysis of small to medium nation-states and Regional Assemblies across the Western world points to the conclusion that for an institution with the National Assembly’s functions, at least 100 representatives is the norm.

The report has also analysed the cost to the quality and effectiveness of our democracy in sticking to the National Assembly’s current size. Failing to do any job properly can be costly. That applies with especial force to the work of democracy. It can result in poorly designed policies, inadequate legislation, wasteful expenditure and poor outcomes.

From this perspective, the focus needs to be on the number of AMs from parties other than the governing party/ies. The main burden of effective scrutiny falls upon them. Here, the contrast with other legislatures is at its sharpest. Moreover, the split among different opposition parties means that the benefits of team working – with subject areas divided among the members of the team – are almost non-existent at committee level. This makes the shortage of capacity in the National Assembly especially acute.

Meanwhile, the National Assembly has increased its responsibilities substantially over the past decade and, following the recommendations of the Silk Commission, they are likely to increase even more in the coming years.

Our main conclusion is to recommend that, to bring it into line with the capacity of other comparable legislatures – not least the other devolved UK institutions at Stormont and Holyrood
– the National Assembly should have around 100 members. We calculate that in total, this would add approximately £10.1 million to the current annual running costs of £49.5 million. We judge this a small price to pay for the democratic benefits that would accrue.

A more effective democratic structure should be a vital part of meeting the need for a more mature democracy that better serves the people of Wales. Democracy on a shoestring undermines effective and efficient governance, and in the end also results in higher costs due to inadequate scrutiny. As the Silk Commission succinctly put it in its first report and quoted at the outset, “Good scrutiny means good legislation, and good legislation pays for itself.”


Additional costs of more AMs

The essential case for increasing the number of AMs is to more effectively hold the Welsh Government to account, improve the scrutiny of legislation, and enhance capacity for developing alternative policies. There is also a strong democratic case for increased resources being spent on AMs and the party groups rather than enhancing the role of the Assembly Commission. The direct salaries, pension and office costs of increasing the number of AMs was itemised in Chapter 5. Here we outline the additional necessary expenditure that would be incurred, including for party groups, Assembly Commission staff, security, and office space. Where there is uncertainty about the range in costs, we have weighed towards the upper limit to allow for unforeseen contingencies.

Party Groups

Funding is already available for Assembly Party Groups with three or more AMs, administration, research, and policy development as follows:

  • A group of between 3 and 10 AMs – £127,390
  • A group of between 10 and 15 AMs – £199,048

In addition, groups of 10 members or more are entitled to an additional £30,866 per 5 AMs who are not represented in the Welsh Government. (National Assembly for Wales Remuneration Board, Determination of Member’s Pay and Allowances, December 2012.)

For the purpose of this exercise we use the current allocation and thresholds for funding and the current representation of parties according to size. For 20 more Assembly Members, following the current share of the seats, half would be shared among the opposition parties. We allocate four each to the Conservatives, making 18 AMs, and to Plaid Cymru, increasing to 15. Two would be given to the Liberal Democrats whose AMs would rise to seven. In these circumstances, the Liberal Democrats would retain the same funding as they do now. Both the Conservatives and Plaid Cymru would gain £30,866 for each additional 5 additional members, and so would gain £30,866 each. There would be an additional cost of £61,732. Under the same allocation and with the same thresholds, the Conservatives would gain a further £30,866, for 40 additional members, with the Lib Dems and Plaid just missing out on the thresholds for additional funding. So the additional cost would be £92,598.

It should be noted that, in practice, these costs would change depending on future elections and the electoral make-up of the National Assembly. Given that there are currently three groups not represented in Government, this costing will be on the higher end of possible outcomes. The costs of additional groups funding based on the makeup of fourth Assembly are as follows:

  • 20 additional AMs – £61,732
  • 40 additional AMs – £92,598

Assembly Commission staff

One of the reasons for increasing the number of AMs is to ensure that there are more AMs to go around to sit on committees. If the decision was taken to increase the number of committees there need not be a significant cost increase, but a reallocation of resources. After all, the number of committees has fluctuated since the Assembly’s inception.

Because there is currently a lack of capacity within the parties in the National Assembly to undertake much of their own research the Commission’s Members Research Service, which supports AMs in their research, already has a generous number of 34 staff to make up a little for the shortfall. The increase in staffing for the party groups that would arise from a greater number of AMs would lead to a greater capacity for their own research, resulting in a reduction in their over-reliance on the Members Research Service. Accordingly, if the number of AMs were increased there would be no need to increase the number of Members Research staff.

Security, IT and general costs

Security staff costs have been between £1.5m and £1.7m, and a staff of 64 has been required for Ty Hywel and the Senedd. As well as the effects of economies of scale, any new offices would not require the same level of security as the Senedd and Ty Hywel as they would most likely be public office suites. The costs of additional security staff should therefore be minimal. An additional 10 staff would be a reasonable estimate, which would cost about £250,000.

To cover this and other contingencies, such as IT support and general costs, we put £500,000.

Office space

Ty Hywel is the home of Assembly Commission Staff, Assembly Members and their staff and the Welsh Government. It has a floor space of 11,583 metres (38,002 feet). It is currently near 100 per cent occupancy. (Apirose brochure, p 4). There would therefore need to be additional office space procured outside Ty Hywel to make room for the new AMs. Ty Hywel is rented by the Commission at a cost of £2.3 million a year.

Space allocation for staff and AMs based in Ty Hywel is at the discretion of the Commission, but extra space is needed. We estimate that each new AM would need space for an additional member of staff. (In practice, many AMs would house all their staff in their Constituency offices, while some would use Ty Hywel to house more staff. This number seems a fair number from which to estimate the amount of office space required – in reality, AMs could fit more staff into the office if that is their wish). In addition, if the funding for party groups continues in its current form, this will result in an additional 0.5 staff member per AM. Therefore for every new AM we would envisage space needed for an additional 2.5 staff (including the Member themselves).

The Welsh Government allocates 18m² per employee at their Cardiff headquarters (p 25). Office costs on the National Assembly estate are £189.41 per m². This in turn works out as follows (we also work out the 30 additional Commission staff we have allocated):

2.5 AM + staff requires 45m² X £189.41 per m² = £8,523.45 per additional Assembly Member.

Combining these two costs together, the office space costs of additional AMs are as follows:

• Additional 20 AMs – £170,469
• Additional 40 AMs – £340,938

New constituency office space

Each AM is allowed to claim up to £5,000 as a start-up cost for a new office. If all new AMs were to use this maximum budget (which is unlikely as many would share offices with MPs for example), the costs would be as follows:

• 20 new AMs – £100,000
• 40 new AMs – £200,00

Space in the Senedd Chamber

The Chamber already has room to house 20 more AMs so this would come at no additional cost. Some thought would be required to house 100 AMs. However, any changes need not be drastic or require any ‘grand designs’. Some of the clerks could be moved to make space for more AMs. Also, the laptops currently used by AMs will be due for renewal at some time. If they were to be allocated tablets rather than laptops, this would free up desk space and allow for the desks to be narrower.

Another row of desks could then be put in, which would certainly accommodate a further 20 AMs.


Taking into account the costs of each AM (including, salary, staff, offices), as well as the funding to parties, support and security costs, office space, the annual cost breakdown is as follows:

Individual AMs costs (inc. staff etc) (£) New Office Start up budget Additional funding to groups (£) Support & Security (£) Additional Office Space (£) TOTAL ADDITIONAL COST
20 new AMs 4,500,000 100,000 61,732 500,000 170,469 £5,332,201
40 new AMs 9,000,000 200,000 92,598 500,000 340,938 £10,133,536

The current annual National Assembly budget is £49.5 million. The rounded costs per annum for more AMs would be as follows:

  • 80 AMs: £54.8 million 100
  • AMs: £59.6 million



  • All Wales Convention, Report (2010)
  • Assembly Commission Budget 2012-13, p. 3, available at http://www.assemblywales.org/gen-ld8713-e.pdf
  • Rachel Banner, Vote ‘No’ to Assembly Failings, Daily Post Feb 28 2011, available at http://www.dailypost.co.uk/news/north-wales-news/2011/02/28/vote-no-to-assembly-failings-says-true-wales-rachel-banner-55578-28246758/
  • Baqir, R, ‘Districting and Government Spending’, Journal of Political Economy, 110, (2002) Bort, E., McAlpine, R, Morgan, G., The Silent Crisis, Jimmy Reid Foundation, 2012.
  • Carman, C., Shephard, M., ‘Committees in the Scottish Parliament’ in Jeffrey C., et al The Scottish Parliament 1999-2009: The First Decade (Edinburgh, Luath Press: 2009),
  • Commission on Devolution in Wales, Empowerment and Responsibility: Financial Powers to Strengthen Wales, November 2012.
  • Crick, M., ‘Cameron promises to cram more peers into the Lords’, from http://blogs.channel4.com/michael-crick-on-politics/cameron-promises-to-cram-yet-more-peers-into-the-lords/2138
  • Gay, O., Wales and Devolution, House of Commons Research Paper 60/97: 1997, p 10 available at http://www.parliament.uk/briefing-papers/RP97-60.pdf
  • Gilligan, T. W., and Matsuaka, J. G., ‘Deviation from Constituent Interests: The Role of Legislative Structure and Political Parties in the States’ from Economic Inquiry 33, No. 3, 1995.
  • Gilligan, T. W., Matsuaka, J. G., ‘Fiscal Policy Legislature Size and Political Parties: Evidence from State and Local Governments in the First Half of the 20th Century’, National Tax Journal, Vol. 54, No. 1, March 2001
  • Hansard House of Lords, Answer to written question: http://www.publications.parliament.uk/pa/ld201213/ldhansrd/text/130108w0003.htm
  • Hooghe, L., Marks, G., Schakel, A. H., The Rise of Regional Authority: A Comparative Study of 42 Democracies, Routledge, 2009. Hansard Society, Assembly Line? The Experiences and Development of new Assembly Member , 2013.
  • Huckle, T., Supreme Court points Direction for Future Devolution, 20 February 2013, available at https://www.iwa.wales/click/2013/02/supreme-court-points-direction-for-devolution/
  • National Assembly for Wales Remuneration Board, Fit for Purpose, March 2011. National Assembly for Wales Remuneration Board, Annual Report 2012-13, July 2013.
  • National Assembly for Wales Independent Review Panel, Getting it Right for Wales: An independent review of the current arrangements for the financial support of Assembly Members, 2009.
  • Osmond J. (Ed.), Welsh Politics Come of Age – Responses to the Richard Commission, IWA, 2005.
  • Pettersson-Lidbom P., ‘Does the Size of the Legislature Affect the Size of Government? Evidence from Two Natural Experiments’, Journal of Public Economics, Volume 96, Issues 3–4, 2012.
  • Rawlings, R., Delineating Wales, University of Wales Press, 2003.
  • Roderick, V., Wales Devolution Briefing, 1997, available at http://www.bbc.co.uk/news/special/politics97/devolution/wales/briefing/westminster.shtml
  • Royal Commission on the Constitution 1969 – 1973, Volume I, Report.
  • Schaltegger A., Feld L., ‘Do large cabinets favour large government? Evidence on the fiscal commons problem for Swiss cantons’, Journal of Public Economics, 93, 2008.
  • Stigler, G. J., ‘The size of Legislatures’, Journal of Legal Studies 5 No. 1, 1976.
  • Thomas A., Roberts O., Assembly Committee Structures, National Assembly for Wales Research Paper, April 2013.

Read more publications...