A Stronger Second Chamber? Assessing the Impact of House of Lords Reform in 1999 and the Lessons for Bicameralism



The UK's second chamber is generally seen as a historical curiosity and, as Labour governments post-1997 have failed to introduce elected members, continues to be seen as ‘unreformed’. But a ‘first stage’ reform of the House of Lords in 1999 removed most of its hereditary members, leaving it almost wholly made up of appointed life peers. Although many at the time claimed that this would neuter the upper house, since most of those evicted were Conservatives, it instead seems to have had the opposite effect. This article examines whether the 1999 reform strengthened the House of Lords, how we can assess this and why it might have happened. The article concedes that measuring legislative influence is notoriously difficult. The analysis it constructs seeks to overcome this by using a range of indicators on two dimensions: the assertiveness of the House of Lords to use its powers, and the executive's responses to the chamber. This analysis indicates that the 1999 changes strengthened the Lords against the government and, in doing so, strengthened parliament as a whole. The most obvious consequence is that our understanding of British politics now needs revision. But if the Lords was strengthened by reform this also demonstrates a need to revise established comparative politics theories of bicameralism – particularly that of Arend Lijphart, which would have predicted the reverse. The recent developments in Britain validate George Tsebelis' approach of recognising ‘partisan veto players’ and the critical role of partisan balance in inter-cameral relations. But they also demonstrate that a more nuanced understanding of both legitimacy and legislative influence is needed in order to understand second chambers' (and indeed first chambers') de facto power.

The House of Lords has long been seen as a curiosity. As British democracy advanced, the second chamber was slow to adapt and remains, alongside the Canadian Senate, one of only two unelected second chambers in the major democracies. Largely thanks to its curious composition, the Lords in the twentieth century was generally timid about using its (not insubstantial) formal powers. British politics scholars described it as having ‘a little public profile but no actual power’ (Shell, 1993, p. 335), while comparativists concluded that Britain had ‘extremely feeble bicameralism [which] ... shades into unicameralism’ (Sartori, 1994, p. 188).

Since the early twentieth century the House of Lords has perpetually been considered ‘unreformed’, with those changes that did take place – reduction in its powers in 1911 and 1949, and the addition of life peers in 1958 – seen as partial and long overdue. At root is the fact that the ‘second chamber constituted on a popular instead of hereditary basis’ promised by the 1911 Parliament Act was never put into effect. Since Labour won power in 1997 Lords reform has remained on the political agenda, and the chamber tends only to be discussed in relation to its future reform.

Yet in 1999 a change was made that in retrospect appears significant. Presented as a ‘first stage’ to reform, the House of Lords Act removed the great majority of hereditary peers. Consequently 667 members were stripped of their rights to sit. This left a chamber made up largely of life peers, appointed by successive prime ministers, alongside 26 Church of England bishops (Shell, 2007). The change made the chamber both far smaller and far more politically balanced. Although a compromise allowed 92 hereditary peers to remain (most of them elected by their colleagues), the second chamber's dominance by the hereditaries, and consequently by the Conservative party, was at an end.

This article examines the significance of the 1999 reform, and concludes that – contrary to many predictions at the time – it did not weaken the House of Lords, but strengthened it. This finding is of significant importance to British politics – suggesting that a change in power relations between Lords and Commons, and between the executive and parliament, is underway. But it also has consequences for our understanding of legislative influence more broadly, and of how bicameralism works. On the former it demonstrates that our assessment of legislative impact is often too simplistic. On the latter, established theories would suggest that the 1999 reform should either have weakened or have had no effect on the power of the House of Lords. Our finding to the contrary hence shows that these theories need refinement.

The article begins by considering two possible perspectives on the 1999 reform, based on the views of different commentators and scholars: first, that the Lords would be weakened by it; and second that it would be strengthened. The article then asks how we can reliably assess changes to the chamber's power, and sets out two broad hypotheses. After a brief discussion of sources and methods, it assesses the evidence on each hypothesis. It ends with conclusions for British politics, and for comparative politics based on the seemingly anomalous case of the UK.

Context: Two Conflicting Interpretations of the 1999 Reform

The removal of the hereditary peers from the House of Lords could clearly be argued on principle. It was widely accepted that inheriting seats in the legislature was anachronistic. However there were also clear political motives for Labour to pursue reform. Immediately before the 1999 Act, 301 of the 640 hereditary peers registered to attend the House took the Conservative whip and only 19 took the Labour whip. There were also 217 hereditary ‘Crossbenchers’ who took no whip, but were suspected of Conservative leanings.1

Labour's reform could therefore easily be characterised as politically driven. The new government saw an unjustified potential obstacle to its legislative programme, and understandably wished to remove it. Tony Blair gained confidence from the overwhelming 1997 victory, which gave him a 179-seat Commons majority. To the Conservatives, who stood to lose their dominance of the upper house, Lords reform was an ‘act of class spite’, intended to reduce the checks on Blair's dominance as prime minister.2 In debate on the bill Conservative frontbenchers suggested that Labour ‘seem[ed] to want, effectively, one-chamber government, where the executive are in control’ and that the reformed chamber would be ‘a pliant, supplicant body’.3 The Conservative-leaning Daily Telegraph claimed it was ‘becoming increasingly clear that Labour is driven by the desire to control the Lords as it does the Commons’.4 Even some commentators on the left expressed concern: Hugo Young (1999, p. 16) in The Guardian alleged that ‘he calls it reform, but all Blair wants is an impotent second chamber’. From the right Peter Hitchens (1998, p. 13) feared Lords reform could signal ‘the road to the British Reich’.

These bleak predictions for executive accountability post-1999 were echoed by some scholars. One of the more considered analyses of Labour's constitutional reform programme (including, but not restricted to, Lords reform) was conducted by Matthew Flinders (2005). Flinders applied the well-established framework of Arend Lijphart (1984; 1999) to analyse the impact of reform on concentration of executive power. Lijphart's now classic treatment counterposed two archetypal models of ‘majoritarian’ and ‘consensus’ democracy, with Britain the prime exemplar of the former. His analysis was based on ten factors, of which bicameralism was one. Consensus democracies enjoyed strong bicameralism, while majoritarian democracies were at best weakly bicameral, or else unicameral.

Flinders (2005, p. 81) concluded that the 1999 reform moved Britain from a system of ‘between medium strength and weak bicameralism’ to one ‘which could now be better described as simply weak bicameralism or even a “one and a half chamber system” ’. His reasoning used Lijphart's original analysis, which based bicameral strength on two dimensions: symmetry/asymmetry (of power), and congruence/incongruence (of membership). On the first dimension second chambers' formal power varies widely – from a delay power of two months over bills (e.g. Spain) to an absolute veto and co-equality with the lower house (e.g. Australia) (see Russell, 2000; Tsebelis and Money, 1997). The Lords' powers, allowing delay of most bills by around a year, are moderate to strong. But as Lijphart (1999, p. 206) notes, ‘the actual political importance of second chambers depends not only on their formal powers’. Importantly, he suggests that ‘second chambers that are not directly elected lack the democratic legitimacy, and hence the real political influence, that popular election confers’ (p. 206). On this basis Lijphart demoted British bicameralism to the ‘asymmetrical’ category. Where the old House of Lords scored better was on his second dimension. This classified parliaments with two chambers similar in composition as ‘congruent’, and those with chambers differing in composition as ‘incongruent’– especially if they over-represented minorities. The House of Lords, where ‘the overrepresented minority ... [was], of course, the nobility’ (Lijphart, 1984, p. 98), qualified as incongruent. Following reform this led Flinders (2005, p. 81) to conclude that ‘the removal of all but 92 of the overwhelmingly Conservative hereditary peers ... has actually increased the asymmetrical [sic] nature of the balance of powers between the two houses’.5 He therefore suggested that the British parliament should now score just 1.75 on Lijphart's scale from 1.0 (unicameralism) to 4.0 (strong bicameralism) – rather than the 2.5 it was originally awarded. In other words, parliament had become significantly more executive-dominated.

Although many predicted a weakening of the House of Lords post-1999, there were also other voices. Most obvious were the claims from Labour ministers. As articulated in debates on the Bill by Baroness Jay, then Leader of the House of Lords, these claimed that the hereditary peers' removal would make the Lords ‘more legitimate’.6 Whether such statements were sincere or not, scholars who watched the House closely lent them some credibility, suspecting that the new chamber might gain new confidence to use its powers. Donald Shell (2000, p. 308) noted: ‘Given that all the members of the transitional House have either been themselves created peers or have been elected members, it is reasonable to suppose that a new boldness might be shown by it’. By the time the new Lords gathered in late 1999, even some previous critics saw new possibilities. The Daily Telegraph suggested that ‘with most of the hereditaries gone (and with them the Conservatives' inbuilt majority), the Upper House should feel free to flex its muscles’.7

The proposal that reform would make the chamber ‘more legitimate’ raised one of the most contested concepts in political science (Barker, 1990; Beetham, 1991; Saward, 1991; Weber, 1947) and was politically controversial. Reform did not change the Lords' position against Lijphart's criteria, which adopted a conventional understanding that legitimacy comes with election. This would imply that appointed members are no better than hereditary members. But some who argued that legitimacy would grow linked this also to another key factor: the change in the chamber's party balance. Flinders (and others) had assumed that removing hundreds of Conservative parliamentarians would make life easier for Labour in government. But ending Conservative dominance and creating a chamber more representative in party terms could also make peers feel more legitimate and confident.

At least as important is that party balance also had the clear potential to affect the internal dynamics of the Lords. Reform left no party having a majority among its members and thus, in the language of George Tsebelis (2002), increased the number of ‘partisan veto players’ in the chamber. Although Tsebelis would not view the Lords normally as being an ‘institutional veto player’, since it lacks an absolute veto over most legislation, his analysis captures the new plurality of influence inside the chamber. Previously Conservative dominance meant the third party, the Liberal Democrats, were as marginal in the Lords as in the Commons. But this changed after reform. Tsebelis explicitly departed from Lijphart by giving far more emphasis to partisan actors. Specifically he suggested that changes in party balance can result in an institution's effective power varying over time.

We therefore have two competing predictions about the impact of reform on the de facto power of the House of Lords, supported (at least in part) by two competing theories. On one side, following Flinders/Lijphart and many opponents of reform, it would weaken the chamber and strengthen executive dominance. On the other, following Shell/Tsebelis and Labour's official line, it would strengthen the chamber. The remainder of this article tests these competing predictions, and consequently the associated theories.

Hypotheses: How Would We Know if the Lords was Stronger?

Ten years after most hereditary peers departed from the House of Lords, it should be possible to make a systematic, albeit initial, assessment of the effects reform had on parliamentary power. We have two broad opposing claims: that reform either weakened or strengthened the second chamber. There is of course also a third possibility, that reform had no effect.

Superficially, the question of whether the Lords has been strengthened appears quite straightforward. But this in fact requires us to recognise an increase or decrease in the chamber's effective, or de facto, power. There has been no change in its formal constitutional position, which provides the same ability to obstruct legislation as before. Most ordinary bills can be delayed for around one year, after which they can be passed under the Parliament Acts by the House of Commons alone. In contrast the chamber can delay financial legislation for only a month, while retaining an absolute veto over those (roughly one-third of) bills that start their passage in the Lords, and over secondary legislation.

In the past the Lords' composition, and related legitimacy problems, meant that these powers were in practice little used. As Shell (1999, p. 203) notes, the chamber exercised ‘restraint, perhaps to such a degree that some would say [it had] embraced a voluntary impotence’. Only three times between 1911 and 1999 was the Commons forced to use the Parliament Acts to pass legislation without the Lords' consent. Generally, when conflict occurred, the second chamber backed down. Similarly, throughout the entire twentieth century it rejected delegated legislation only once. In fact a series of conventions became established over time, informally limiting the chamber's powers. Most notably the ‘Salisbury convention’ required it not to reject completely any bill implementing a government manifesto commitment.

The first and clearest indication of growing House of Lords strength would therefore be a greater readiness to use its existing powers, and an associated breakdown in conventions. This provides our first hypothesis:

Hypothesis 1: Post-1999 the House of Lords is more inclined than previously to use its powers.

This might be expected to be the beginning and end of our analysis. But such an approach would significantly oversimplify the nature of parliamentary power. Legislative studies scholars accept that influence often comes not through exercising formal powers, but through the very existence of these powers causing changes in executive behaviour. As Michael Mezey (1990, p. 155) suggests, it is possible that ‘all government proposals appear to pass the legislature unopposed and unamended simply because the government avoids introducing those things which may provoke legislative resistance, or because it has acceded to changes privately’. In particular, in the Commons, it is acknowledged to be the relationship between the government and its own backbenchers that matters most (King, 1976). These forms of influence concern not so much the legislature's behaviour, but the executive's responses, which are principally determined in private. This makes them ‘difficult to nail down empirically’ (Mezey, 1990, p. 155). Nonetheless this challenge must be tackled if parliamentary influence is to be assessed. A second hypothesis is therefore:

Hypothesis 2: Post-1999 the government is more responsive to the Lords, and takes greater steps to avoid confrontation.

Each of these hypotheses has a counter-hypothesis, reflecting the counter-claim that reform weakened the Lords, and an associated null hypothesis of no change. Each is also quite general, and requires breaking down into more measurable factors. This detail is reserved for the main empirical section of the article. Before that a summary is given of methods and sources, and some of the practical obstacles to such an analysis.

Methodological Issues

What follows draws from a range of data concerning the post-1999 House of Lords, including items from the public record and material provided by parliamentary officials. It also draws from a database compiled by the author of the over 1,200 divisions (i.e. votes) in the chamber since reform in 1999, questionnaire surveys of members in 2005 and 2007 and 70 interviews with members and officials.8

Despite this breadth of information there are two clear obstacles to a comparative analysis of the Lords' influence over time. First is the paucity of past research and data about the chamber. Even the House of Commons has been subject to fairly limited empirical research, but the Lords, being considered politically unimportant, has received significantly less attention. One exception was a study of the House in the 1988–9 parliamentary session (Shell and Beamish, 1993), which documented the legislative process in some detail. Other studies (notably Baldwin, 1995; Morgan, 1975) have focused on earlier periods in less detail. Aside from this, the only sources available are limited official records and raw data, references in more general studies about parliament, and political memoirs.

The bigger issue in assessing changes over time, however, is that the dynamics of the Lords depend greatly on the broader political context. In particular, the chamber has historically presented greater challenges to Labour than Conservative governments. Given that the most reliable data available come from the Thatcher period, this amplifies the difficulties. It is generally accepted that there was some strengthening of the Lords following the 1958 Life Peerages Act (which in effect rescued it from oblivion). This may have been boosted by Harold Wilson's failed attempt at reform ten years later. In the 1980s Margaret Thatcher's governments faced some Lords resistance, partly in response to Thatcher's large Commons majority and partly to her brand of conservatism (Baldwin, 1985; Shell, 1992b). But the extent of measurable conflict with the Commons was still far smaller than in the 1960s or 1970s (Shell, 1992a). Thus if the Lords proved more assertive post-1997 (or post-1999) this could simply reflect the change in government party control. But a comparison with the chamber's behaviour under Labour in the 1960s or 1970s is also problematic, not just because of the unavailability of data, but also because these periods differ from the contemporary context in important ways. For example, Labour's 1974–9 majority in the Commons was small or negative, in contrast to the huge majorities enjoyed by Tony Blair. (The last time Labour enjoyed such a majority was 1945 – before the addition of life peers.) In earlier periods political agendas were also very different. For example, one of the biggest clashes between government and the Lords 1974–9 was over nationalisation in the Aircraft and Shipbuilding Industries Bill: a long way from modern-day controversies over, for example, foundation hospitals or anti-terrorism legislation.

These obstacles demonstrate that an analysis of changes to the House of Lords over time requires caution. But they should not preclude such an analysis altogether. The discussion that follows is careful to take such difficulties into account.

Analysis: Is the House of Lords Stronger or Weaker Following Reform?

We now return to the two broad hypotheses above: that the post-1999 Lords is more prepared to make use of its powers and that it is being taken more seriously by government. This section draws on several sources of evidence in order to inform our assessment on each hypothesis. The overall balance of evidence is then weighed up in the subsequent section.

The Chamber's Use of Its Powers

As already indicated, the House of Lords, while being far from ‘co-equal’ with the House of Commons, enjoys formal power significantly greater than many other second chambers. It can block some policies – bills starting in the Lords, and secondary legislation – indefinitely, while others – most bills starting in the Commons – it can delay for around a year. The chamber's previous reputation as weak originated not because its powers were small in comparative terms, but because they were so little used. If reform strengthened the chamber we would expect to see more readiness to use these powers. Obviously if the chamber were weakening, we should see the reverse.

Lords Defeats.  The most visible indicator of the chamber's power is its ability to inflict government defeats. It does not generally vote on government bills in their entirety at second and third reading (unlike the Commons) so most ‘divisions’ are on amendments to bills, and each bill may therefore suffer several defeats. These defeats may or may not then be overturned when the bill returns to the House of Commons. This is the most obvious place to start looking for an increase in the chamber's de facto power.

Table 1 shows the total number of defeats, and the average number per calendar year, for each parliament since October 1974. The 1997–2001 parliament is broken into two periods: before and after reform. This provides a relatively short time series, including only one complete session post-reform. We see that the number of defeats per year in the two most recent parliaments is significantly higher than in those immediately preceding, but there is also a clear party effect. Under Conservative governments 1979–97 the number of defeats per year never exceeded twenty, but it rose once Labour re-entered power. The figures for 2001–7 are more similar to, indeed rather lower than, those for 1974–9.

Table 1. Government Defeats in the House of Lords by Parliamentary Session
 Defeats (n)Average defeats per calendar year*
  1. Source:House of Lords (1998).

  2. *As parliaments tend not to comprise exact years this has been calculated from a daily average during the (calendar) length of the parliament scaled up by 365 days.

‘Old’ House of Lords  
 1974–9 (Wilson/Callaghan)34375
 1979–83 (Thatcher)4511
 1983–7 (Thatcher)6215
 1987–92 (Thatcher/Major)7215
 1992–7 (Major)6212
 1997–9 (Blair)7027
‘New’ House of Lords  
 1999–2001 (Blair)3824
 2001–5 (Blair)24563
 2005–7 (Blair/Brown)10743

The arithmetic leading to defeat in the chamber is now very different to that under former Conservative and Labour governments. Shell (1992a) calculates that, of active peers in 1975–6, 149 were Labour, 30 were Liberals and 292 were Conservatives. The Conservatives were clearly dominant and Labour could be defeated relatively easily, even if supported by Liberal peers. Now the Conservatives and Labour each have around 200 peers, and the balance of seats is held by around 70 Liberal Democrats and 200 independent ‘Crossbenchers’.9 Analysis of voting records shows that the Conservatives can rarely defeat the government on their own, and in practice the third party usually decides the outcome of votes. Of all 283 defeats during 1999–2005 there were just twelve occasions when the government lost despite having Liberal Democrat support (Russell and Sciara, 2007). (The Crossbenchers are far less often influential, due to low voting turnout and lack of cohesion.) So while at the moment the change may not be obvious, because the Lords used to defeat Labour more, and continues to do so, the new arithmetic will almost certainly become more apparent should the Conservatives return to power. They potentially face a hostile coalition of Labour and Liberal Democrats, meaning that we can expect defeats to remain high, rather than returning to the low levels seen in the 1980s.

Lords Policy Insistences.  Defeats in the Lords are not the end of the story, since these may later be overturned by government in the Commons. Except in the unusual circumstances where the Parliament Acts are used, both chambers must agree legislation in identical terms. Following defeat the Commons must (re)consider the bill, which must then return again to the Lords if the Lords' amendments have been rejected, and so on. This process, colloquially known as ‘ping-pong’, continues until agreement between the chambers is reached.

Table 2 shows data on the number of bills per parliament since October 1974 where the Lords has insisted on at least one amendment, and in each case the number of iterations of this process. This presents a rather different picture to Table 1. While there is a clear distinction between the Conservative years and the Blair years, there is also a clear distinction between the Blair years and the 1974–9 Labour government. As Shell (1992a, pp. 146–8) notes, in the three years from October 1974 there were over 400 Lords amendments rejected by the Commons and ‘in almost every case the House of Lords agreed not to insist on its amendment’. As shown, during that whole parliament there were only four bills where the Lords insisted on change (in two cases once, in one case twice and in one case three times). Similarly during the entire period 1964–70 there was only one Lords insistence (Morgan, 1975), and in the 1940s ‘though government defeats on bills were numerous ... the Lords did not insist on their amendments if the Commons resisted them’ (Jennings, 1957, p. 429). All of this is in stark contrast to the record of the Lords since 2001. In the first session alone there were insistences on seventeen bills, including two where the Lords insisted four times. This appears to indicate a more assertive chamber.

Table 2. Number of Bills per Parliament where the House of Lords has Insisted on Its Amendments 1974–2007
ParliamentNumber of insistencesTotal
  1. Source:1974–2005 drawn from the Clerk of the Parliament's evidence to the Joint Committee on Conventions (2006, pp. Ev 97–8), supplemented from the author's database of divisions.

  2. *Two years only.


Since 1997, clashes between the government and the Lords have become familiar. The two bills facing four insistences in 2001–5 were the European Parliamentary and Local Elections (Pilots) Bill (where the Lords sought to restrict the piloting of all-postal voting), and the Prevention of Terrorism Bill (where it inflicted eighteen defeats over the detention of terrorist suspects, ending with a record-breaking 32-hour parliamentary sitting). In the first two years of the 2005 parliament there were again four rounds of insistences on both the Identity Cards Bill and the Corporate Manslaughter and Corporate Homicide Bill. But it must be noted that there were also four rounds of insistences on the European Parliamentary Elections Bill in the 1997–8 parliament (on the issue of ‘open’ vs. ‘closed’ lists), immediately before reform. Not just reform, but also other factors including the size of Blair's majority and the nature of legislation, may therefore have helped fuel the growth in Lords assertiveness.

Conventions Breaking Down?  As already noted, the cautiousness of the old House of Lords had become part of established convention. The best-known example was the ‘Salisbury’ (or ‘Salisbury-Addison’) convention about manifesto bills. But in practice the Lords did not only restrain itself from defeating manifesto bills at second or third reading; it rarely rejected government bills in their entirety at all. Similarly, its formal veto over secondary legislation conventionally was not used. The last time such a measure was defeated before 1999 was 1968. Another indication that the chamber was strengthening would therefore be for these conventions to start crumbling.

There were signs immediately after the 1999 reform that this would happen, when the Conservative Leader in the Lords, Lord Strathclyde (1999), declared the convention on secondary legislation ‘dead’. But this was not, as might have been expected, followed through with a string of defeats. There was one defeat on secondary legislation in 2000, on arrangements for the London Mayoral elections, and another in 2007, on the siting of the Manchester ‘supercasino’. Visible change has therefore been limited. But it has been enough to encourage government caution. At least once, over a measure to restrict trial by jury in 2005, the government has withdrawn a piece of secondary legislation (in this case already tabled) in anticipation of defeat. The Conservatives remain reluctant to defeat such legislation, despite Lord Strathclyde's threat – fearing a new precedent that could be used against them on return to power. But the Liberal Democrats are significantly less restrained, and defeat is possible even without Conservative frontbench support – as happened in the casino vote.10

Similarly to secondary legislation, there is some evidence that conventions governing primary legislation are crumbling, but this remains limited. In contrast to a complete lack of second reading defeats in the 1970s (and 1980s) there have been a small number of such moves since 1999, though none were on manifesto bills.11 In 2004 the House unprecedentedly voted at second reading to send the Constitutional Reform Bill to a specially convened select committee – this significantly delayed the legislation, but it ultimately passed. More importantly the Lords has twice used a wrecking tactic at second reading to completely block a bill, in 2000 and 2007, both times again on limiting access to trial by jury.12 Government ambitions in this area consequently remain thwarted. The Liberal Democrats have expressed particular hostility to the conventions on primary legislation, their Lords leader having dubbed the Salisbury convention ‘the last refuge of legislative scoundrels’.13 The party's representatives, enjoying their pivotal role in the upper house, claim that such conventions, in the words of then leader Charles Kennedy, were made for ‘absolutely different political circumstances’, i.e. when the chamber was disproportionately Conservative.14 In fact the political balance in the Lords is now, as Liberal Democrats frequently remind the government, more proportional to general election votes than that in the House of Commons.

Although there have been few actual breaches of conventions, the 2007 survey of peers found these under significant pressure. As Table 3 shows, there was general support for the convention of not blocking government bills at second reading, but far less so at third reading, and less still for the convention on secondary legislation. In all cases we see a clear division between government and non-government peers.

Table 3. Peers' Attitudes to Key Conventions (% Agreeing)
 LabConLib DemCrossbenchAll
  1. Notes:N = 357–70 (Lab 96–100, Con 104–8, Lib Dem 43–5, Crossbench 103–6, others 9–11). Figures show percentage saying ‘strongly agree’ or ‘agree’ on a five-point Likert scale.

  2. Source:2007 peer survey.

The House of Lords should not block government bills at second reading8962585666
The House of Lords should not block government bills at third reading6529293741
The House of Lords should not block delegated legislation541251523

Without comparable data from earlier periods we cannot gauge the extent to which peers' attitudes have changed. But the government clearly believes this to be the case, as demonstrated by its establishment in 2006 of a joint committee of both chambers specifically to review the Lords conventions. In evidence to the committee, Leader of the House of Commons Jack Straw suggested that it was essential to ‘pin down what these conventions are’ because ‘we have seen ... in the last 10 years, the more that the composition of [the Lords] is changed and “modernised” the greater the appetite is [there] for an assertion of powers’ (Joint Committee on Conventions, 2006, p. Ev 13). The committee concluded that ‘the Salisbury-Addison convention has changed since 1945, and particularly since 1999. ... The continued validity of the original Salisbury-Addison convention is clearly contested by the Liberal Democrats’ (Joint Committee on Conventions, 2006, pp. 31–2).

Peers' Confidence to Demand Policy Change.  Finally on Hypothesis 1, we can investigate directly whether peers are feeling more confident post-1999. This has been done using member surveys, and here the results seem clear. In 2007 a large majority of members (86 per cent, n= 353) believed that ‘peers' confidence to demand policy change’ had increased since reform. A similarly large group (85 per cent, n= 369) agreed that the chamber had ‘become more assertive’ since 1999.15

An important reason for this appears to be peers' perception of the chamber's legitimacy following the hereditaries' departure. In 2007 three-quarters of members (76 per cent, n= 335) believed the 1999 reform had increased ‘the legitimacy of the House of Lords’. This in itself is perhaps not surprising: it may be convenient for peers to think so. But these views were shared by a majority of MPs in a 2004 survey (57 per cent, n= 195). Here the majority believing the Lords to be ‘more legitimate’ post-reform was greatest among Labour MPs (75 per cent, n= 107), which is of course consistent with the government's earlier official line.16 And support for a more assertive Lords is found not only at the elite level but also in public surveys. For example, around two-thirds of respondents in a 2005 poll believed that if bills had ‘little public support’ it was ‘justified for the House of Lords to block’ them.17

There have now been several occasions when disgruntled Labour MPs have actively encouraged the Lords to intervene in government policy. This marks a distinct departure from the past. For example, in 2003 former Health Secretary Frank Dobson wrote to peers urging them to vote against foundation hospitals. Similarly, in 2005 former Foreign Secretary Robin Cook claimed that a Labour rebellion in the Commons had ‘sent the House of Lords a message’ to challenge the government over anti-terrorist legislation.18 It is reasonable that peers' confidence should be boosted by such encouragement from members in the elected house. Both peers and MPs' attitudes will be influenced, in turn, by their knowledge of public opinion.

My examination of Hypothesis 1 has tried to quantify as far as possible the extent to which the post-1999 Lords is inclined to make greater use of its powers. The evidence is not wholly decisive, but appears to support the hypothesis. It is certainly more consistent with a strengthening than a weakening of the House of Lords. Defeats have risen, peers are backing down less in confrontations with government and established conventions are under pressure. Peers also appear to feel more confident – bolstered in part by the views of MPs and the public. All of this lends greater support to the Shell/Tsebelis position than to that of Flinders/Lijphart. To be conclusive, observation over a longer period (particularly under a Conservative government) would clearly be necessary. But we also know already that such quantifiable measures cannot provide a full picture of legislative influence. It is for this reason that we now turn, more briefly, to Hypothesis 2.

The Government's Response to the House of Lords

The most obvious place to look for a strengthening or weakening of the House of Lords was the chamber itself. But the complex nature of parliamentary influence means further evidence is also relevant. It is important in particular to look at the responses of the executive to the post-1999 chamber. On this there are again several sources of evidence from which we can draw.

Government Responses to Lords Defeats.  We know that the Lords is regularly defeating the government. This helps demonstrate its assertiveness but does not necessarily indicate real policy power, since the Commons may subsequently reject the Lords' position. Where it does we have seen that the Lords now seems less inclined to back down than it was in the past. One fairly obvious change in government response may therefore be a greater inclination to offer concessions following Lords defeats.

A detailed study of all such defeats 1999–2006 shows that, although the Commons has the power to overturn them, it does not routinely do so (Russell and Sciara, 2008). Of 274 unique defeats (i.e. excluding ‘ping-pong’), only 42 per cent were rejected outright. In a further 17 per cent of cases some very minor concessions were given, while the remaining 40 per cent (i.e. 110 unique defeats) saw the Lords' position wholly or substantially accepted. Furthermore, and contrary perhaps to expectations, the Lords was more likely to ‘win’ in confrontations over major than minor policy matters. Trial by jury, as already indicated, is one such example. Others all relate to important policies within bills, rather than bills in their entirety. These include repeated blocking by the Lords of a proposed new offence of religious hatred, the rejection of plans to merge criminal justice inspectorates and, in 2008, government abandoning its proposals to allow terrorist suspects to be held without trial for up to 42 days.

The government is therefore making significant and regular policy concessions following Lords defeats. There are various reasons why it may do so. On small matters it may concede on technical arguments, or simply in order to avoid legislative delay. On larger matters it may be reluctant to reverse defeats where the Lords has media and/or public opinion on its side. In some such cases it may even be risky to return bills to the Commons, where Labour backbenchers have become more rebellious (Cowley, 2002; 2005). On religious hatred, for example, the government was actually defeated in the Commons when it tried to overturn a Lords amendment. The same might well have happened on 42 days' detention, where the Commons had previously passed the policy by a majority of only nine – but the government chose instead to drop the policy. A more assertive Lords can therefore drive, and be driven by, changes to the crucial ‘intra-party’ mode of legislative relations that Anthony King (1976) identified in the Commons. In the 1970s, when Labour had a small or negative majority, there were similar pressures. Government was defeated several times in the Commons on Lords amendments, and accepted various other Lords amendments to avoid Commons defeat, though the overall level of concessions to the Lords is unquantified (Baldwin, 1995). The result was ‘not so much ... a clash between the two Houses of Parliament as a clash between government and Parliament facilitated by the House of Lords’ (Baldwin, 1995, p. 223). The fact that neither main party controls the Lords, combined with declining party cohesion in the Commons, means this situation may now become the norm.

Government Negotiation to Avoid Lords Defeat.  Government defeats in the Lords may be seen as a breakdown of the system. Faced with a chamber where it has no majority, it seems more rational for the government to negotiate. Indeed, negotiated outcomes may be at least as important to the Lords' power. In the 2005 survey of peers 85 per cent of respondents (n= 377) agreed that ‘the chamber's real influence is exerted as much through persuasion as through defeats’. In 2007, when asked which forms of policy influence exercised by the chamber were most important, one in three peers felt government defeats were ‘very important’. But more than half said the same about ‘government bringing forward its own amendments under pressure from peers’ (see Table 4).

Table 4. Peers' Views on the Importance of Different Processes to the Lords' Policy Influence (%)
 Very importantImportantNot very importantNot at all important
  1. Notes:The question was ‘How important do you think each of the following are in terms of how the House of Lords influences the policy process?’ N = 361–70 (Lab 97–9, Con 104–10, Lib Dem 44–5, Crossbench 100–4, other 11–2).

Government defeats in the House of Lords336170
Government bringing forward its own amendments under pressure from peers544510
Peers exerting pressure behind the scenes to influence policy before it is announced2652212

Numerous amendments are made to bills in the Lords – most without division, and most in the name of ministers. In the 2006–7 session, for example, 1,911 amendments were agreed (House of Lords, 2008). But this alone tells us little about the chamber's power. Large numbers of government amendments could indicate that regular concessions are being made to the Lords (making it influential). But they could also reflect government introducing poorly drafted legislation and amending it itself later in parliament (implying the opposite). It is a labour-intensive exercise to trace the origins of amendments and thus to quantify influence. A case study of the Identity Cards Bill (2005–6) found that more than half the government amendments of substance made in parliament responded to non-government concerns at earlier stages, and most resulted from pressure in the Lords (Russell and Johns, 2007). Detailed studies are lacking on earlier periods, though Nicholas Baldwin (1995) did find evidence of negotiation on several bills in the 1970s. A study of ministerial memoirs from this period, however, finds no mention of negotiation with the Lords, indicating that it was not a dominant theme.19 Concession in negotiation with the Lords therefore appears important, but we do not know how important, or how this compares with the past.

Government Attempts to Anticipate and Accommodate the Opinion of the Lords.  Even negotiation over amendments lies at the more measurable end of influence – as it concerns public reversals in government policy positions after bills are published. Other parliamentary influence may occur yet further ‘upstream’. As Table 4 shows, a large majority of peers believe that exertion of pressure behind the scenes before government policy is announced is important. This is obviously very difficult to capture, but there are nonetheless clear signs of changes to government behaviour post-1999.

Following Labour's election in 1997 a new procedure was established at the instigation of the Lords Chief Whip, Lord Carter, with respect to approving bills for introduction to parliament. This required ministers to provide the Cabinet's Legislative Programme committee with a written ‘Lords Handling Strategy’, detailing the likely points of contention in the upper house, and plans for dealing with them. Over time this procedure has become more onerous, with bill team civil servants encouraged to make early contact with the government Chief Whip's staff in the Lords to devise a joint strategy. This may include holding concessions in reserve for discussion in the Lords, which sometimes means withholding these when points are raised in the lower house, much to the frustration of MPs. In extremis the new procedure means that if ministers cannot convince colleagues that a measure will be passed by the upper house, it may even fail to be introduced to parliament.

While this process appears no more than sensible preparation, there is no evidence that similar procedures were followed by earlier Labour governments. Interviews with surviving cabinet ministers from 1974 to 1979 indicate that they gave little or no consideration then to Lords' concerns. One interviewee recalled that the Lords was no more than ‘a nuisance’, and another that, given the government's narrow majority, ‘if you could get [a bill] through the House of Commons you didn't worry about the House of Lords’. These views are consistent with Baldwin's (1995) interviews, and with the almost complete lack of mention of the Lords (other than its reform) in ministerial memoirs of the 1960s and 1970s.20 Notably the diary of Richard Crossman (1979), who was Leader of the House of Commons 1966–8 and responsible for the government's legislative programme, never refers to the Lords in this context – despite thirteen mentions of the Cabinet's Legislation Committee, which he chaired. In part earlier Labour governments were preoccupied with managing the Commons. But there was also clearly an expectation (justified by the evidence above) that the Lords would back down in the face of Commons opposition, which it is now less inclined to do.

A final new development is the open consultation meetings that ministers now hold in the Lords on bills before they are formally debated. Between February 2003 (the first point for which data are available) and May 2008, 89 such meetings were advertised to peers. Increasingly they have begun to involve Commons, as well as Lords, ministers. In early 2008 these included, for the first time, three meetings hosted by senior cabinet members.

The evidence on Hypothesis 2 lends support to, and strengthens, the conclusions reached on Hypothesis 1. It is of course difficult to quantify how much government has changed its behaviour to accommodate the post-1999 House of Lords: legislative studies scholars emphasise the challenge of this kind of assessment. Nonetheless, we know that the government is now giving in regularly to Lords' demands on legislation, and perhaps more so than in the past. More clear is that government has both changed its internal process for approving bills to anticipate the Lords' response, and is facilitating early discussion with peers to avoid legislative conflict. These are new departures not seen during previous periods of Labour government.

Parliamentary Influence, the ‘New’ House of Lords and British Politics

This study has drawn together a range of evidence to assess the role of the House of Lords since 1999. Hypothesis 1, which asserted that the Lords would have more confidence to use its powers after reform, found some support. This was reinforced by the evidence on Hypothesis 2, that the government would change its behaviour to be more responsive to the policy demands of the ‘new’ post-reform House of Lords. In both cases conclusions must remain tentative, and indeed patterns of behaviour may change in future years. But they support the views held by peers, by ministers and increasingly by the media and outside observers that the second chamber is gaining in assertiveness and influence. In short nowhere do we find evidence of the more ‘pliant, supplicant’ chamber, subservient to the government in the Commons, that some sceptics predicted before reform. Those commentators who thought (and perhaps those Labour politicians who hoped) that removing numerous opposition members would weaken the Lords seem to have been proved very wrong.

It was acknowledged at the start of this investigation that legislative influence is complex and multilayered, and its assessment is difficult. The discussion has reinforced this point. The most measurable and visible sign of Lords impact is government defeats, of which there have been more than 400 since 1999. The chamber also now appears more reluctant to back down when confronting the government. But one of the most important changes is that the government's own procedures now explicitly anticipate and seek to avoid conflict with the Lords. All things being equal these government initiatives should result in the number of Lords defeats falling, rather than rising. Confrontation between government and the Lords may therefore increase as the chamber gains greater confidence, but may also fall. Less confrontation would not necessarily indicate a decline in the chamber's influence on policy. The period 1999–2008 has clearly been one of adjustment, so a new equilibrium, based on greater government flexibility, may in future be found. An assumption that parliamentary influence is measurable through visible conflict has long led the House of Commons, and many other legislative chambers, to be dismissed as marginal to the policy process. This study helps illustrate why such assumptions can be simplistic, and often incorrect. As some defenders of parliament have emphasised, the institution creates an essential framework of power that places limits on policy negotiations between other actors (Judge, 1993). Recent changes to the House of Lords have shifted this defining framework in the UK.

The purpose of this article was not to claim that the House of Lords is a dominant policy actor, or that it has become more powerful than the House of Commons. Indeed the discussion has demonstrated not only that legislative influence cannot be easily assessed, but also that the policy pressure generated by each chamber is inextricably linked to the other. The Lords, where the government lacks a majority and where the third party and independents hold the balance of power (Russell and Sciara, 2007), operates with knowledge of, and often in collaboration with, forces in the Commons. Opposition parties may be able to extract concessions in the Commons in the knowledge that they can push policy points to defeat if necessary in the Lords. The government backbench in the Commons, already more restive following Blair's outsized majorities and controversial policy positions, may also work openly or covertly with the Lords. A more assertive Lords therefore strengthens the Commons, both in its ‘opposition’ and ‘intra-party’ modes (King, 1976). A more assertive Commons, in turn, strengthens the negotiating power of the Lords.

All of this is very different to the classic ‘Westminster model’ that Lijphart (1984; 1999) and others have described. In this system the second chamber was weak, coalition government was rare and an executive with a partisan majority in the House of Commons faced little impediment in achieving its policy goals. Lords reform has empowered the third party (and independents), and created a forum in which government faces more pressure to negotiate. It has increased the leverage of actors in the first, as well as the second, chamber. The implications for British politics are thus potentially far-reaching, pushing Britain in the direction of more pluralistic and consensual policy making. These dynamics are likely to endure, including under any future Conservative government, at least until the Lords is further reformed. But if reform proceeds the most likely outcome is a second chamber elected on a proportional basis: which would make these dynamics more powerful still.

Implications for Bicameral Theory

This article has concluded that reform in 1999 strengthened, not weakened, the House of Lords, and that British politics has become more ‘consensual’ as a result. Such a conclusion is problematic, however, since a strict application of Lijphart's theory of majoritarian and consensus democracies (Flinders, 2005) suggested that the reverse would happen. Our results therefore expose problems with the original theory. These are found on both the ‘congruence’ (composition) and ‘symmetry’ (power) dimensions of bicameralism.

The most important misunderstanding is over the importance and role of party balance. On Lijphart's first dimension, Flinders judged that the removal of large numbers of Conservative hereditary peers would make the Lords more congruent with the Commons, due to the class differences between the chambers converging. But in fact this also advantaged new political forces that previously lacked influence, putting the Liberal Democrats (and occasionally Crossbenchers) in a pivotal position. This situation is well explained by Tsebelis (2002), who would see these as new ‘partisan veto players’ in the chamber. Tsebelis sought to correct deficiencies in Lijphart's original theory by emphasising the importance of political party balance to the de facto power of institutions. The case study of the House of Lords demonstrates the strength and importance of this analysis.

The second difficulty demonstrated by the Lords case relates to the ‘symmetry’ dimension. Here the chamber has not changed its formal powers, but seems more willing to use them. This concurs with Lijphart's important point that de facto and formal constitutional power may differ. It also confirms his judgement that legitimacy is important. But it shows, as is also evident from other cases such as the German Bundesrat, that a dichotomous view whereby unelected chambers are considered illegitimate, while directly elected chambers are legitimate, is too simplistic. The House of Lords is now perceived by members of both parliamentary chambers as more legitimate than it was, despite its unelected basis, and enjoys some public support. The Lords may not equal the legitimacy of the Commons, and certainly continues to use its powers with caution. But it does have important features that the Commons lacks. Most obvious is its new and more proportional party balance (in contrast to the Commons' over-representation of the governing party), but also a presence among its membership of numerous ‘experts’ and independent members. In this ‘anti-political’ age, interventions by such a chamber to constrain single-party governments will often have public support. We can thus conclude two things on legitimacy: first, that this is better understood as a continuous than as a dichotomous variable: it is a matter of more or less legitimacy, not of absolutes; and second, that legitimacy can be influenced by other factors aside from democratic election, including party balance. Indeed party balance, which was largely overlooked by Lijphart, is important not just to one, but to both, of his dimensions of bicameralism. It is central to de facto bicameral strength.

The House of Lords has long been an unusual institution, which does not fit neatly into the categories by which modern-day politics is understood. As this study shows, it can continue to surprise us. But if its oddness forces us to question our assumptions and refine our theories of how political institutions work, this is surely to the good.

About the author

Meg Russell is Reader in British and Comparative Politics and Deputy Director of the Constitution Unit at University College London. She is the author of Reforming the House of Lords: Lessons from Overseas (Oxford University Press, 2000), Building New Labour: The Politics of Party Organisation (Palgrave, 2005) and numerous papers, particularly on parliament and parliamentary reform. She is currently pursuing work on the policy impact of the British parliament and the application of social psychology to parliamentary behaviour, under an ESRC Fellowship.

Meg Russell, Constitution Unit, Department of Political Science, University College London, 29–30 Tavistock Square, London WC1H 9QU, UK; email: meg.russell@ucl.ac.uk


  • This research was funded by the ESRC under research grants RES-000-23-0597 and RES-063-27-0163. I am grateful to Martha Spurrier for research assistance, to the House of Lords authorities and Julian Dee for providing access to various data, and to the anonymous referees for their helpful comments.

  • 1

    Miers and Brock (1993) found that during the 1988–9 parliamentary session the votes of unaligned Crossbench peers (including life peers) divided roughly 60 : 40 in favour of the Conservative government.

  • 2

    Liam Fox, Conservative spokesman on the constitution, quoted in The Independent, 12 October 1998, p. 2.

  • 3

    Liam Fox, speaking at the second reading of the Bill, House of Commons, Hansard, 1 February 1999, col. 609.

  • 4

    Leader article, 22 June 1999, p. 23.

  • 5

    Flinders' use of the term ‘asymmetrical’ here is somewhat confusing, as in Lijphart's terms what changed was the congruence of the two chambers. Flinders is therefore referring to the combined effects of reform on two dimensions (as he suggests, making the Lords relatively weaker), rather than to Lijphart's symmetry dimension.

  • 6

    House of Lords, Hansard, 14 October 1998, col. 925.

  • 7

    Leader article, 1 December 1999, p. 29.

  • 8

    Response rates to the surveys were 396 members in 2005 and 381 in 2007 – in both cases exceeding 50 per cent. Respondents were representative in terms of political party, length of time in the House and hereditary and life peers. There were 60 interviews with members, who were a representative sample in these same terms.

  • 9

    Numbers in the Lords fluctuate constantly, due to deaths and new appointments. On 6 October 2008 there were 213 Labour members, 201 Conservative members, 74 Liberal Democrats, 203 Crossbenchers, 26 bishops and 15 others (House of Lords website, http://www.parliament.uk).

  • 10

    Here the Conservatives did not apply the whip, and only 36 Conservatives voted (split 30 : 6 against the government). Other opponents were 50 Liberal Democrats, 26 Crossbenchers, 11 Labour rebels, 3 bishops and 3 others.

  • 11

    See Joint Committee on Conventions (2006). Prior to 1998 there was only one bill rejected at second reading: the War Crimes Bill in 1990. There were two further instances of defeat at second reading in 1998 (European Parliamentary Elections Bill) and 1999 (Sexual Offences [Amendment] Bill), but here the Lords' action was a cooperative gesture when the bill was reintroduced a second time, allowing government to pass it quickly under the Parliament Acts.

  • 12

    On the Criminal Justice Mode of Trial (No. 2) Bill, and the Fraud (Trials without a Jury) Bill, respectively.

  • 13

    Lord McNally, House of Lords, Hansard, 6 June 2005, col. 760.

  • 14

    House of Commons, Hansard, 17 May 2005, cols 50–1.

  • 15

    In each case responses were on a five-point Likert scale. Figures quoted reflect those saying ‘increased a lot’/‘increased a little’ or ‘agree strongly’/‘agree’.

  • 16

    ESRC-funded survey conducted by UCL Constitution Unit.

  • 17

    MORI poll for UCL Constitution Unit, May 2005: 1,007 valid respondents, with results adjusted to be demographically and politically representative. Several scenarios were given, with 66 per cent of respondents saying that Lords intervention was justified at least ‘sometimes’ on manifesto bills, and 64 per cent on non-manifesto bills, lacking public support. Sixty-eight to seventy per cent said the same about bills where many Labour MPs had rebelled.

  • 18

    Newsnight, 28 February 2005.

  • 19

    The sources examined were Benn (1988a; 1988b; 1990); Callaghan (1987); Castle (1980); Crossman (1979); Healey (2006); Jenkins (2006); also Thomas (1985).

  • 20

    For sources see Note 19 above.