Abstract

The British Westminster parliament is frequently dismissed as a weak policy actor, in the face of dominant executive power. But through analysis of 4361 amendments to 12 government bills, and over 120 interviews, we suggest six reasons for doubting the orthodox view. These fall into three groups: overstating government success in making amendments, overstating non-government failure, and overlooking parliamentary influence before and after the formal passage of bills. We demonstrate that Westminster in fact has substantial influence in the policy process, not readily visible through commonly published data. Uncovering influence requires careful tracking of amendments, but also qualitative analysis of actors' motivations and the power of ‘anticipated reactions’. Because Westminster is often seen as being at the weak end of a comparative spectrum of parliamentary influence, these results are important for demonstrating both the dynamics of British politics, and of parliamentary systems more broadly.

Introduction

In a recent prize-winning book on policy failures in British government, two well-known political scientists claimed that ‘[a]s a legislative assembly the parliament of the UK is, much of the time, either peripheral or totally irrelevant. It might as well not exist’ (King and Crewe, 2013, p. 361). This arresting statement is consistent with common perceptions of the Westminster parliament. Among political commentators, Westminster has in the last few years been described for example as ‘a legislature on its knees’, ‘an elaborate rubber-stamp’ or even ‘God's gift to dictatorship’.1 While legislatures in parliamentary systems are frequently dismissed as weak actors in the policymaking process, Westminster is often presented as an extreme case. For comparative scholars, it has classically been seen as at the opposite end of a spectrum when contrasted to the powerful US Congress. Indeed some see the ‘Westminster model’ as shorthand for a dominant executive and compliant legislature (Lijphart, 1999).

Yet while such descriptions are commonplace, the underlying assumptions about Westminster are largely untested empirically. This article hence asks whether the UK executive really dominates the legislative process, or whether there is instead parliamentary influence at play that is not widely noticed or understood. Based on study of 4361 amendments to 12 government bills, and an extensive programme of interviews, our findings challenge the popular view. By suggesting that Westminster is more influential than is widely recognised, our results have important implications for the study of British politics. As no study of this scale has to date (to our knowledge) been carried out in any other domestic parliament, they also contribute to broader understanding of parliamentary systems.

The article begins with an overview of the literature on parliamentary influence, after which we briefly describe our methods. We then present summary data on amendment outcomes for the 12 case study bills, which initially appear to confirm executive dominance. In the sections that follow, however, we offer six ‘reasons for doubt’, indicating why these figures cannot be taken at face value. They fall into three categories: overstating government success, overstating non-government failure, and focusing only on the formal stages of the legislative process. Having argued that on all six counts Westminster is more influential than first appears, we turn briefly to what types of policy changes it helps to bring about. We end with a short conclusion.

Parliamentary influence at Westminster and beyond

Despite the centrality of legislatures in democratic states, scholars note that ‘the prevailing view among political scientists over the last few decades has been that parliaments play a marginal role in the policymaking process’ (Martin and Vanberg, 2011, p. 4). Michael Mezey's (1979) well-known classification judged the US Congress to be an ‘active’ legislature, but most parliaments in Europe to be at best ‘reactive’ (the alternatives being ‘minimal’ or ‘marginal’ legislatures, found primarily in less well-established democracies). Norton (1998a) disaggregated the ‘reactive’ category, pointing out that in some countries (e.g. Sweden) parliamentary bodies were relatively strong, while in others (e.g. Ireland) they were relatively weak.

Within such league tables, the British Westminster parliament has conventionally been placed towards the bottom, contrasted unfavourably to both the strength of the US Congress (Mezey, 1979), and other European parliamentary systems (Martin and Vanberg, 2011). In line with this, Norton's survey concluded that the British parliament had only ‘a limited capacity to affect the outcome of public policy’ (1998b, p. 41). Richardson and Jordan (1979) memorably described Britain as a ‘post-parliamentary democracy’, with protestations from others that the UK remained a fundamentally ‘parliamentary state’ (Judge, 1993) finding little support. Hence the mainstream view, as reflected in a popular British textbook, is that the House of Commons is ‘misunderstood if viewed as a legislator’ (Moran, 2005, p. 196).

But empirically testing the strength of legislatures, and explicitly ‘questioning the Mezey question’ is challenging. Arter (2006, p. 245) notes a previous ‘propensity to conflate “legislative capacity” and “legislative performance”’, which ‘has fed a tendency to accept rather than confront … legislative stereotypes’. In a search for parliamentary performance measures, the legislative process is one obvious place to look. David Olson (1994, p. 84) famously observed that ‘[i]n most democratic legislatures’ there is something approaching ‘a “90 per cent rule”: the cabinet proposes at least 90 per cent of the legislative agenda, and at least 90 per cent of what it proposes is adopted’. Studies contrasting the success of government versus non-government bills in parliaments in Europe do generally find executive bills dominate (e.g. Arter, 1985; Maurer, 1999; Capano and Giuliani, 2001; Pettai and Madise, 2006; Zubek, 2011). But such measures are necessarily crude, taking no account for example of the importance of different bills, and crucially of the amendments made to them during their parliamentary passage. A finer-grained picture may be achieved by considering the success of government and non-government amendments (e.g. Damgaard and Jensen, 2006; Kerrouche, 2006)—but this raises similar questions about the substantiveness of amendments, and also about the true origins of government amendments.

One early study at Westminster addressed such questions. Griffith's (1974),Parliamentary Scrutiny of Government Bills painstakingly analysed amendments tabled over three sessions in the late 1960s and early 1970s. This found that among the amendments passed, those proposed by government were dominant, but there was hidden non-government influence. For example, at least 15 per cent of government amendments agreed at House of Commons report stage responded to policy demands from backbench or opposition MPs at the previous stage. Overall Griffith (1974, p. 256) concluded that parliament's impact was ‘by no means negligible’, but that concessions were made ‘largely on [the government's] own terms’.

But by now Griffith's study is several decades out of date. It also had some methodological weaknesses—for example failing to track amendments consistently between the Commons and the Lords. In the intervening years, some developments may have weakened parliament—most notably, growing policy complexity. But changes at Westminster itself may well have had the opposite effect. In the Commons, these include creation of specialist departmental ‘select committees’ for executive oversight in the 1970s (Drewry, 1985), which have gradually grown in reputation and strength (Kelso, 2009; Benton and Russell, 2012; Hazell et al., 2012), and declining party cohesion (Cowley, 2002, 2005). In the Lords, reform in 1999 removed most peers who had inherited their seats, leaving a largely appointed and more party-balanced chamber, more confident to use its powers (Shell, 2007; Russell, 2013).

Subsequent empirical analyses of Westminster legislative impact have been more limited, however, either in terms of scope or methods. Thompson (2013) conducted a major study of the Commons committee stage 2000–2010, identifying almost 1000 occasions when a minister responded to concerns by promising at least to consider policy change. Russell and Sciara (2008) studied defeats in the post-1999 House of Lords, finding that around 40 per cent were ultimately accepted by the government and House of Commons. Brazier et al. (2008) analysed the passage of five case study bills, suggesting that parliament was indeed influential, but made no attempt at quantification. Hence no definitive two-chamber study of Westminster's impact on the legislative process has been conducted in recent times.

Detailed amendment analyses have meanwhile been applied in several other settings, most notably the European Parliament. Here, various studies have accounted, for example, for policy substantiveness of different amendments, and the links between those at one legislative stage and the next (e.g. Kreppel, 1999; Tsebelis et al., 2001; Kasack, 2004; Häge and Kaeding, 2007). At the domestic level, a study of the Scottish Parliament demonstrated that roughly a quarter of substantive executive-sponsored amendments were inspired by non-executive parliamentarians (Shephard and Cairney, 2005). A more limited amendment analysis in the Irish Oireachtas drew some similar conclusions (O'Dowd, 2010). Overall, a gap exists not only with respect to detailed study at of the amendment process at Westminster, but also in other national domestic parliaments.

Detailed amendment analyses can clearly answer important questions about parliamentary influence on legislation. But, while very labour-intensive, they may still tell only half the story. Legislative studies scholars have long emphasised that the ‘reactive’ influence of parliaments can be less important than their ‘preventive’ influence (Blondel, 1970, p. 78, p. 82), in line with the classic ‘rule of anticipated reactions’. Recent European Parliament studies (e.g. Häge and Kaeding, 2007) have shown this to be the case. Hence, ‘significant changes in policy wrought by parliament … [may] be achieved not in parliament, but in advance of a measure being submitted to it’ (Arter, 1985, p. 68). This is readily expressed in the language of principal-agent analysis, now often applied to parliaments (e.g. Strøm et al., 2003). As McGann (2006, pp. 454–455) suggests, ‘if the legislature was to find an agent that perfectly implemented its wishes, then [it might appear] … that the legislature was a rubber stamp’. In parliamentary systems like that at Westminster, where the executive depends on the legislature for its own survival, such dynamics seem particularly likely.

If a lack of conflict does not necessarily signal parliamentary quiescence, the British parliament could be either weak or strong. A detailed amendment analysis, in particular tracing the origin of government amendments, can help to tease this out. But while quantitative analysis is necessary, it is also insufficient. A fuller picture, allowing for ‘anticipated reactions’, requires qualitative analysis as well. Should such an analysis show Westminster to be influential, the same may well also be true of other European domestic parliaments.

Methods and data

The research reported here represents the largest study of Westminster's legislative process since Griffith (1974). We built on his methods, and those of subsequent authors in other settings, to apply amendment analysis across the whole Westminster legislative process—taking in all stages in both the Commons and the Lords. This largely quantitative exercise was supplemented by other documentary research, and interviews.

We chose 12 case study government bills to analyse in detail, listed in Figure 1. Seven were drawn from the 2005–2010 parliament, when Labour was in government, and five from the first session of the 2010–2015 parliament, under a Conservative–Liberal Democrat coalition. Although it is impossible with such a small sample to be fully representative, we selected bills against a range of key variables, including chamber of introduction, sponsoring government department, length, and degree of controversy (in terms of number of government defeats in the House of Lords).

Figure 1.

Case study bills.

Figure 1.

Case study bills.

Our documentary research focused on the amendments proposed during the bill's formal passage through both chambers, plus careful reading of Hansard debates and relevant committee reports. Amendments at Westminster can be tabled at five initial legislative stages (Commons committee, Commons report, Lords committee, Lords report and Lords third reading).2 An unlimited number of ‘ping pong’ stages between the two chambers may then follow, in the event that they disagree. We recorded information about every amendment proposed at each stage—amounting to 4361 in total. These ranged from 47 amendments on the Identity Documents Bill to 1076 on the Welfare Reform Bill.

As well as basic data about each amendment—including its text, legislative stage and party affiliation of its sponsors—we used debates to determine more qualitative features; for example, the amendment's legislative effect and what the minister (in the case of non-government amendments) said in response. Each amendment was also coded for its policy substantiveness (further discussed below), and outcome. Crucially, we traced and recorded whether an amendment that failed to be agreed at one stage was agreed subsequently in identical or very similar form. While these judgements introduced some subjectivity, we worked from a detailed coding scheme and each researcher's codes were checked by another.

To supplement these data, we conducted over 120 semi-structured interviews with key actors on the bills, including ministers, opposition frontbenchers, backbenchers from across the parties, civil servants, parliamentary officials and representatives of outside pressure groups. This helped us to interpret our quantitative data, as well as to uncover other, more hidden, forms of influence.

The appearance of executive dominance

The outcomes of the 4361 amendments are presented in Table 1. Just over one-fifth were agreed (964, or 22 per cent), while a small minority (135, 3 per cent) were rejected by parliament. By far, the largest group (3262, 75 per cent) were neither agreed nor disagreed. In the Commons this was often due to not being selected for debate or called by the Speaker or committee chair. In the Lords, the presiding officer has no such power, but it was common for an amendment, having been tabled, not to be moved. In both chambers many amendments were also withdrawn, either before or after debate. These actions can indicate that the proposer has been satisfied by the minister's answer, as further discussed below.

Table 1

Amendments and their outcomes, by House

House Commons
 
Lords
 
Total
 
Agreed on division 29 1.5 35 1.4 64 1.5 
Agreed without divisiona 371 19.6 529 21.4 900 20.6 
Negatived on division 87 4.6 38 1.5 125 2.9 
Negatived without division 0.4 0.1 10 0.2 
Not selected 227 12.0 n/a n/a 227 5.2 
Not called 774 41.0 n/a n/a 774 17.7 
Not moved 17 0.9 1096 44.3 1113 25.5 
Withdrawn 346 18.3 585 23.7 931 21.3 
Withdrawn prior to debate 30 1.6 187 7.6 217 5.0 
Total 1889 100 2472 100 4361 100 
Of agreed, overturned or overturning 90 4.8 122 4.9 212 4.9 
House Commons
 
Lords
 
Total
 
Agreed on division 29 1.5 35 1.4 64 1.5 
Agreed without divisiona 371 19.6 529 21.4 900 20.6 
Negatived on division 87 4.6 38 1.5 125 2.9 
Negatived without division 0.4 0.1 10 0.2 
Not selected 227 12.0 n/a n/a 227 5.2 
Not called 774 41.0 n/a n/a 774 17.7 
Not moved 17 0.9 1096 44.3 1113 25.5 
Withdrawn 346 18.3 585 23.7 931 21.3 
Withdrawn prior to debate 30 1.6 187 7.6 217 5.0 
Total 1889 100 2472 100 4361 100 
Of agreed, overturned or overturning 90 4.8 122 4.9 212 4.9 

aIncludes 83 amendments agreed alongside others on a division during ping-pong.

In assessing influence, we first took into account that 212 of the agreed amendments made no difference to the final bill text, because some amendments were wholly reversed by others. This applies most often with respect to government defeats in the Lords, which can subsequently be overturned in the Commons. Such amendments were initially agreed, but cannot be considered ultimately successful. Likewise, an amendment whose sole purpose was to overturn an earlier one is not a real change, as it merely returns the bill to its previous form. We thus exclude these ‘overturned’ and ‘overturning’ amendments from our subsequent figures, which leaves 4149 in total—of which 752 (18 per cent) ultimately succeeded.

A key question is obviously who proposed the successful amendments. Breaking down amendment outcomes by sponsor appears very strongly to confirm executive dominance. Of these 4149 amendments, 775 (19 per cent) were sponsored or co-sponsored by a government minister, while 3374 came entirely from non-government parliamentarians (defined as parliamentarians who are not ministers, including government backbenchers, opposition and non-party members). Table 2 shows their relative success rates. Although 94 per cent of government amendments were ultimately successful, just 24 non-government amendments (fewer than 1 per cent) were agreed and not overturned. No government amendment was actively disagreed by parliament.

Table 2

Amendments and their outcomes, by sponsor (excluding overturned and overturning)

Outcome Government amendments Non-government amendments Total 
Agreed (with or without division) 728 24 752 
Negatived (with or without division) 135 135 
Not moved, withdrawn, etc. 47 3215 3262 
Total 775 3374 4149 
% agreed 93.9 0.7 18.1 
Outcome Government amendments Non-government amendments Total 
Agreed (with or without division) 728 24 752 
Negatived (with or without division) 135 135 
Not moved, withdrawn, etc. 47 3215 3262 
Total 775 3374 4149 
% agreed 93.9 0.7 18.1 

If anything, in fact, Table 2 understates the typical success of government amendments, because one of our bills was an outlier. Of the 47 government amendments neither agreed nor disagreed, 44 were on the Public Bodies Bill, where the government took the unusual step of dropping many of its own amendments, replacing them with redrafted proposals. If this bill is excluded the success rate of government amendments is 99.5 per cent, and that of non-government amendments is 0.6 per cent.

Figures such as these are sometimes quoted as demonstrating that the executive dominates the legislative process. At first glance they do suggest that Westminster is weak, but the following sections outline six reasons for doubting this interpretation.

Overstating government success

The above figures suggest that almost all successful amendments were made on the government's initiative. Yet they give a misleading impression by overstating government success, for two reasons.

Reason 1: most government amendments have little substance

First, not all amendments are equal. While some propose a substantive change to policy, many others do not. These differences are explicitly recognised in the government's own internal classification of amendments. This is unpublished, but used behind-the-scenes in Whitehall to determine which ones require policy clearance from other ministers.

The classification of amendments by substantiveness is also widely applied in academic studies (e.g. Kreppel, 1999; Tsebelis and Kalandrakis, 1999). We used a similar scheme to that deployed by Shephard and Cairney (2005), to code every amendment on a three-point scale. The least substantive category was typographical or consequential amendments. These either made only cosmetic or lexical changes (e.g. renumbering, or correcting drafting errors) or provided consistency following other, more substantive amendments (e.g. adding references throughout the bill to a new clause or definition). Neither type can be considered to have policy substance of their own. The middle category comprised clarificatory amendments, which did not significantly alter the intention of a provision, but simply limited its interpretation. Substantive amendments were those that would actually alter the effect of the legislation.

As shown in Table 3, around two-thirds of amendments were substantive, but there was a clear divide between those proposed by ministers and non-government parliamentarians. Most government amendments (56 per cent) fell into the least substantive category, while less than a third made substantive policy changes. For non-government amendments the reverse was true: almost three quarters were substantive, and just 16 per cent typographical or consequential.

Table 3

Amendments by substantiveness and sponsor (excluding overturned and overturning)

 Government Non-government Total 
Typographical/consequential 435 550 985 
Clarificatory 111 421 532 
Substantive 229 2403 2632 
Total 775 3374 4149 
% Substantive 29.5 71.2 63.4 
 Government Non-government Total 
Typographical/consequential 435 550 985 
Clarificatory 111 421 532 
Substantive 229 2403 2632 
Total 775 3374 4149 
% Substantive 29.5 71.2 63.4 

Government amendments tend to be non-substantive for several reasons. By the time a bill is published, most of the government's preferred policy will be settled. But the rush to introduction may leave some loose ends. Far more than other parliamentary actors, the government has responsibility for ensuring that legislation is legally watertight, and it also has the specialist capacity to spot and correct errors, particularly via the ‘parliamentary counsel’ who draft government bills. Some post-introduction tidying up therefore occurs. In addition, when the government does make substantive amendments (for whatever reason—as discussed below), it will carefully identify all the necessary consequential amendments. In contrast, non-government actors are usually interested only in debating substantive policy and are unlikely to identify minor technical deficiencies. When proposing substantive policy changes, non-government parliamentarians accompany these with only the most obvious consequential amendments—knowing that in the (unlikely) event that their amendment is agreed, government lawyers will attend to the detail.

Reason 2: most substantive government amendments respond to parliamentary pressure

So the raw total of government amendments overstates the extent to which government amends its own legislation, as two-thirds have little policy substance. But the fact remains that nearly all agreed amendments are sponsored by ministers.

The next question is thus how many government amendments respond to pressure from other parliamentarians, rather than being at the government's sole initiative. Griffith's (1974) work demonstrated that this can be the case. A typical scenario is for an opposition member or government backbencher to propose an amendment that the minister initially resists. But the minister, if unable to convince its proposer of the government's position, may indicate a willingness to reconsider. Consequently, the proposer will often let the amendment drop, in the hope of a government amendment at a later stage. This process can occur repeatedly, for example with an initial amendment at Commons committee stage, subsequent non-government amendments thereafter, and an ultimate government climbdown in the House of Lords. Tracing this process carefully across all legislative stages enables us to see how many government amendments in fact represent concessions to other parliamentarians.

There are also now other possible sources of parliamentary influence, not present in Griffith's day. A recent study of the Commons select committees found that their recommendations are often taken up by government (Benton and Russell, 2012). Analysis by the Lords Delegated Powers and Regulatory Reform Committee (2005)—which reviews each bill for the appropriateness of powers delegated to ministers—likewise indicates that a high proportion of its recommendations go on to be accepted. Sometimes government accepts committee recommendations directly, while at others this follows non-government amendments pressed on ministers supporting the committee's case.3 To assess these dynamics, we reviewed all relevant committee reports in the same session(s) as the bill, comparing their recommendations with subsequent government amendments.4 In addition, where a government amendment responded to neither non-government amendments nor committee recommendations, we found a handful of cases where it responded to a concern expressed during earlier parliamentary debates on the bill.

Our results on these three forms of influence are presented in Table 4. Of the 728 ultimately successful government amendments, 297 (41 per cent) could be traced to some form of parliamentary pressure, and this applied to a clear majority (60 per cent) of substantive government amendments. Just 77 successful government amendments (11 per cent) were both substantive and untraceable to parliamentary pressure.

Table 4

Ultimately successful government amendments by whether responding or not

Reason for government amendment Unsubstantive*
 
Substantive
 
Total
 
Not responding to any identifiable parliamentary pressure 354 66.3 77 39.7 431 59.2 
Responding to a non-government amendment (only) 108 20.2 62 32.0 170 23.4 
Responding to a select committee recommendation (only) 20 3.7 14 7.2 34 4.7 
Responding to both of the above 44 8.2 36 18.6 80 11.0 
Responding to debate only 1.5 2.6 13 1.8 
Total 534 100 194 100 728 100 
Total % responding  33.7  60.3  40.8 
Reason for government amendment Unsubstantive*
 
Substantive
 
Total
 
Not responding to any identifiable parliamentary pressure 354 66.3 77 39.7 431 59.2 
Responding to a non-government amendment (only) 108 20.2 62 32.0 170 23.4 
Responding to a select committee recommendation (only) 20 3.7 14 7.2 34 4.7 
Responding to both of the above 44 8.2 36 18.6 80 11.0 
Responding to debate only 1.5 2.6 13 1.8 
Total 534 100 194 100 728 100 
Total % responding  33.7  60.3  40.8 
*

Typographical/consequential or clarificatory (see Table 3).

The commonest source of parliamentary influence was an earlier non-government amendment. This accounted for 51 per cent of substantive government amendments (and 28 per cent of non-substantive amendments, most of which were consequential). For example, the Further Education and Training Bill enabled educational providers to be granted powers to award ‘foundation degrees’. Non-government amendments at several stages sought to ensure that students could progress to a full degree at a higher education institution. The government ultimately conceded, with an amendment requiring progression arrangements to be reviewed prior to granting an institution foundation degree awarding powers. This was a fairly typical pattern, repeated across most of the case study bills.

The casual observer might assume that non-government amendments acceptable to ministers would simply be allowed to pass. But there are several reasons why the government prefers to table its own amendments. The most obvious is political, i.e. not to be seen conceding to the opposition. But, in practice, the most important reasons are procedural. Ministers require policy clearance, often including from other Whitehall departments, before accepting an amendment, and getting this necessarily takes time. Even if clearance can be achieved, government must also ensure that amendments are legally sound, so parliamentary counsel prefer to redraft.5 The official guidance warns ministers, when faced with an amendment which is acceptable in policy terms, to ‘accept the amendment in principle (which he or she can only do if clearance has been given) and offer to table improved wording to meet the intended aim at a later stage’ (Cabinet Office, 2012, p. 151). Hence, ministers very rarely accept a non-government amendment as it stands: there were only 16 such cases across the 12 bills.6 The remaining eight successful non-government amendments followed (mostly Lords) defeats which the Commons did not overturn—though even here consequential government amendments were often added.

Although less common, 50 substantive government amendments (26 per cent of the total) were traceable to select committee recommendations. Many responded to the Delegated Powers and Regulatory Reform Committee, by making changes to future delegated powers. But amendments were traceable to various other committees as well. One of the most high-profile was the introduction of a complete ban on smoking in public places through the Health Bill. The government's original intention, included in its recent election manifesto, had been to introduce a ban with certain exemptions (e.g. for bars and pubs not serving food). But shortly before the bill's publication the Health Select Committee launched an inquiry into the matter, and recommended—part-way through its Commons passage—that the bill be amended to introduce a comprehensive ban. Committee members tabled such an amendment (which was withdrawn), and MPs overwhelmingly agreed a government amendment with the same effect.

A handful of government amendments responded only to debate, without earlier non-government amendments or committee recommendations. We classified such cases with caution, to avoid overstating the case. One example occurred on the Identity Documents Bill, where the government made report stage amendments introducing quite specific safeguards, in response to concerns raised by a government backbencher in Commons committee.

We therefore see that while the majority of government amendments lack policy substance, the majority that do have substance follow clear parliamentary pressure for policy change. We cannot ascribe such pressures solely to parliament, as it often provides the forum to negotiate changes that have been called for by others, e.g. through the media and interest groups. This notably occurred on the Public Bodies Bill, when the government removed clauses allowing it to sell off public forests, following a noisy public campaign alongside amendments from opposition and backbench members of the Lords. Without the external campaign, policy might well not have changed. But equally, campaigners might well have failed without ministers having to respond formally to amendments.

We likewise cannot assume that those substantive government amendments untraceable to parliamentary pressure (i.e. the minority) reflect straightforward ministerial changes of heart. These too can be driven by external events. For example, eight government amendments to the Welfare Reform Bill responded to a vote in the Scottish Parliament (on a ‘consent motion’, regarding powers delegated to Scottish ministers), while government amendments to two bills responded to the outcome of a referendum on further devolution in Wales. These clearly do not show Westminster parliamentary influence, but neither do they show the executive driving the process.

Overstating non-government failure

The previous reasons for doubt concerned overstating the government's success at amending its own legislation. Many government amendments were traceable to pressure from non-government parliamentarians. Yet there were only 250 such cases across the 12 bills, while there were 3350 non-government amendments which parliament did not agree, and which might thus be assumed to have ‘failed’. We now turn our attention to these, arguing that taking the official figures at face value risks overstating non-government failure, for two reasons.

Reason 3: there may be several non-government amendments on the same issue

First, multiple non-government amendments are often tabled on the same issue. This kind of duplication does not occur with respect to government amendments, which almost invariably succeed at their first attempt.

There are several causes of such duplication. One we have already seen, in terms of similar amendments pursued at multiple stages. Non-government parliamentarians use this strategy when wishing to maintain pressure on an issue, commonly until either an adequate ministerial response is given or the bill runs out of time. Most amendments are proposed at committee stage (on our bills the figure was 64.5 per cent), with the remainder proposed at the report stage, at third reading in the Lords, or during ‘ping-pong’. Our linking of related amendments at different stages found that a minority were in packages pursued only at a single stage. Most of the remainder spanned two, three or four stages, while a handful lasted longer—with two strands of amendments spanning 13 and 14 stages. Both of these longest examples were on issues which began in the Commons and moved to the Lords, were pursued by a combination of opposition and government backbench parliamentarians and resulted in repeated Lords defeats. Both ultimately also forced government climbdowns: on the degree of compulsion in the identity card scheme (Identity Cards Bill), and the extension of the new corporate manslaughter law to deaths in prison and police custody (Corporate Manslaughter and Corporate Homicide Bill). In the latter case, 46 ‘unsuccessful’ non-government amendments at various stages led to 10 ultimately agreed government amendments.

As well as similar amendments recurring at multiple stages, duplication also often occurs between different non-government actors at a single legislative stage—usually when opposition parties and government backbenchers move amendments in an uncoordinated way.7 For example, at Commons committee stage on the Energy Bill, three separate amendments were tabled by Conservative, Liberal Democrat and Labour MPs, all calling for an ‘emissions performance standard’ for power plants. The government subsequently tabled its own amendments, which all three non-government packages could thus be seen as helping to bring about.

A less common type of duplication is when non-government parliamentarians table several options, rather than a single preferred amendment. Thus, on the Savings Accounts and Health in Pregnancy Grant Bill, which abolished the Child Trust Fund from January 2011, the opposition's preferred option was to retain the scheme, but it tabled three additional options: to delay abolition until 2014, 2016 or another unspecified date (all failed to save the policy). Likewise on the Welfare Reform Bill an opposition peer tabled four alternative amendments on frequency of benefit payments, explaining that she was ‘offering the Minister a number of courses of mitigating action’ (again, all of her proposals failed).8

Reason 4: many non-government amendments do not aim to change the bill

The aforementioned scenarios show how simple comparisons between the success rate of government and non-government proposals overstate non-government failure. But more importantly, many seemingly ‘failed’ non-government amendments were never intended to be agreed in the first place. Non-government parliamentarians have many motivations for proposing amendments, besides trying to change the text of a bill—as is clear from our interview evidence in particular.

The most common among these are ‘probing’ amendments, which are explicitly understood at Westminster as not intended to be voted upon. They are largely a product of parliamentary procedure: because most stages in the legislative process are structured around amendments, this is the most effective way of securing debate on an issue. Interviewees described such amendments as being ‘a form of question’, or ‘a hook for a debate’—equivalent to ‘highlighting something’. Probing amendments are used to get ministers to explain policy in the bill more clearly, and hence to hold government to account. As one MP put it,

If there's an issue that you're concerned about and you want to discuss, or you want to find out why the minister wants to do it a particular way, then you put down a probing amendment that you have no intention ever of pushing to a vote, that often is defective—so that if it was accepted into the bill it would make the bill make no sense, or it's very poorly drafted—but the whole point of it is it just gives you the opportunity to raise that as an issue.

Another motivation for amendments results from the public nature of debates on legislation on the record. Amendments provide an opportunity for parliamentarians to ‘signal’ alternative policy positions to those outside, including the media, pressure groups and their own parties—again with no realistic expectation that such amendments will be agreed. One interviewee explained that, on such amendments, ‘you know you're never going to win them, but you want to make the political point’. Likewise, amendments can be used as an attempt to shame political opponents. For example, a Labour parliamentarian told us how post-2010 some amendments were proposed simply to ‘get the Tories to vote against something good and decent’ or to ‘show up the Lib Dems’.

Even where some amendments are sincerely meant, others may be tabled simply to maximise the chance that they are voted on in prime time. These ‘filler’ amendments are designed to be debated or withdrawn as necessary, as a means of speeding up or slowing down debate. Other sincere amendments may form part of what one MP described as a ‘very long game’ where she was ‘trying to change some of the climate of the debate’, rather than achieving immediate change. Lastly, even non-government parliamentarians seeking actual change may be satisfied by ministerial assurances on the record about implementation, without words being added to the bill itself. Primary legislation often specifies only the broad framework of policy, with the details left to secondary legislation or guidance. On several occasions on the case study bills parliamentarians withdrew amendments following assurances that changes to these would be made. Such policy ‘wins’ are clearly not visible through the figures on agreed amendments.

Amendments therefore have a wide range of aims. The most obvious motivation may be to change the bill, but while this motivates virtually all government amendments, non-government amendments have much more varied purposes. As one MP told us, ‘big amendments that you would actually like to see accepted into the bill’ are ‘far more in a minority’ than might be expected. Taking amendment outcomes as a sign of non-government ‘failure’ substantially overstates the case.

Parliament's influence at other policy stages

These first four reasons all raise significant doubts about the interpretation of amendment figures for the parliamentary legislative process. The final two are less obviously tangible and measurable, but at least as important. They relate, respectively, to parliament's influence before legislation is formally introduced, and after it has reached the statute book.

Reason 5: parliament influences policy before the formal legislative process begins

As noted above, legislative studies scholars emphasise the ‘preventive’ role of parliaments in conditioning what executives propose (Blondel, 1970, p. 78). Our fifth reason for doubting executive dominance relates to this crucial power of anticipated reactions. Given their dependence on retaining the Commons' confidence, ministers cannot risk seriously alienating their own backbench MPs (King, 1976; Cowley, 2002). They likewise seek to avoid defeat in the Lords—where they lack a partisan majority (Russell, 2013). Increasingly, actors in Whitehall also think through the likely responses from select committees (Benton and Russell, 2012). Such factors underlie the processes discussed above, in terms of why government makes concessionary amendments. But the key time for government to consider anticipated reactions, and avoid unnecessary conflict, is instead before a bill is actually introduced.

Such forces were widely acknowledged by our interviewees. For example, one senior government source told us that ‘ministers do take account of what they know the view of parliament is when they draft the bill’. A former minister likewise suggested that ‘the Commons is able to set the parameters for a bill quite closely’ and that ‘the government can get a sense of what will and will not be acceptable’.

Prior to every bill's publication, the responsible civil service team is now required to draft a parliamentary ‘handling strategy’ (Cabinet Office, 2012). Typically this identifies the parliamentarians likely to be interested in the bill (particularly in the Lords), the issues that they may raise, the government's response and any concessions that could be offered. The handling strategy is examined by the cabinet committee for Parliamentary Business and Legislation which considers bills for introduction. When parliamentarians know that a bill is in preparation, there is often active lobbying behind-the-scenes to shape its content. For example, some interviewees claimed that the coalition's proposed reduction of housing benefit for the long-term unemployed was omitted from the Welfare Reform Bill following pre-introduction pressure from Liberal Democrat parliamentarians.

But this kind of early influence was not limited to the period of coalition government post-2010, and nor was it entirely preventive. Parliamentarians can also be instrumental in helping to bring about legislation. Other authors have documented how government may pursue legislative ideas previously proposed in private members' bills (Brazier and Fox, 2010), and this was one early source of pressure for the smoking ban in the Health Bill. Select committees can also exert influence at this early stage. Benton and Russell (2012) judged that around 1 in 10 select committee reports were ‘agenda-setting’, giving the example of the Public Administration Select Committee's long campaign to persuade ministers to put core civil service values on a statutory footing (ultimately implemented in the Constitutional Reform and Governance Act 2010). Similarly, Hindmoor et al. (2009) judged that 20 out of 93 policy proposals within government education bills 1997–2005 were either identical or similar to earlier recommendations by the relevant select committee. Among our case study bills, the introduction of the Saving Gateway Accounts Bill may have been influenced by the Treasury Select Committee's strong support for the scheme.

But our biggest example of positive agenda-setting by parliamentarians occurred with respect to the Labour government's Corporate Manslaughter and Corporate Homicide Bill. This created a new offence of corporate manslaughter, long called for by trade unions, and by the relatives of those killed in various disasters. Ministers and civil servants clearly had doubts about the merits of such legislation, but it was repeatedly demanded by Labour MPs. A reliable source told us in interview that the government was ‘under quite a lot of pressure’, and although the Prime Minister himself ‘didn't want corporate manslaughter legislation, for a variety of reasons’, this was nonetheless ultimately introduced. It went on to reach the statute book in a further strengthened form.

Reason 6: parliament influences policy after the legislative process is complete

Our research found various specific examples of legislative amendments sowing seeds for later parliamentary influence. For example on the Savings Accounts and Health in Pregnancy Grant Bill, which the coalition used to abolish Labour's Child Trust Fund, a cross-party group of MPs pursued an amendment to retain such a scheme for children in local authority care. The bill was passed without this amendment, but the government subsequently launched such a scheme, citing the amendment's lead sponsor as influencing the decision.9 In other cases, agreed amendments provided explicit opportunities for later influence, for example by increasing government's requirement to report to parliament, or strengthening parliamentary oversight of future secondary legislation. When such regulations are subsequently drafted, ministerial assurances given during the initial legislative debates may well be taken into account.10

More broadly, parliament has well-established mechanisms to oversee policy implementation, including ministerial questions, backbench and opposition day debates and select committees. Around half of departmental select committee reports review progress in policy implementation (Benton and Russell, 2012). Thus once a bill has passed, parliamentarians retain an oversight role and have various opportunities to influence the detail of implementation, and/or make recommendations for future changes.

On what is parliament influential?

These six reasons for doubt suggest that Westminster is far more influential on government legislation than is often recognised. Yet questions may remain over what types of change it helps to bring about. As just indicated, some amendments concern improved oversight or future reporting, rather than triggering policy reversals. If this were typical of parliament's traceable influence, our case would perhaps be less convincing.

On some of the 12 bills—for example the coalition's Saving Accounts and Health in Pregnancy Grant Bill or Identity Documents Bill—parliament made very few changes. Yet on others—as we have already seen—it helped to bring about significant policy shifts on highly political matters. These included effectively preventing the introduction of identity cards (Identity Cards Bill), introducing a total rather than partial smoking ban in public places (Health Bill), including deaths in custody within the remit of the new offence of corporate manslaughter (Corporate Manslaughter and Corporate Homicide Bill) and removing the government's power to sell off public forests and abolish numerous ‘quangos' (Public Bodies Bill). Alongside these high-profile examples were many smaller policy shifts, such as making recipients of Carer's Allowance eligible for saving gateway accounts (Saving Gateway Accounts Bill), retaining some benefits for care home residents (Welfare Reform Bill) and widening ‘carbon capture and storage’ beyond coal-fired electricity generation (Energy Bill).

Nonetheless, many changes wrought by parliament were also relatively minor and procedural. To get a complete picture, we thus assigned every ultimately successful amendment to one of five ‘themes’: two relating to policy change, two to procedural change and one comprising purely technical amendments. Policy changes were those affecting the implementation of public policy itself, and we distinguished between amendments purely changing its timing and those changing its actual substance. Likewise, we distinguished between procedural changes relating to future delegated legislation and other procedural requirements, such as requiring parliamentary reports.11

The results are given in Table 5, which shows that two-thirds (67 per cent) of ultimately successful amendments implemented a change in policy, while just 14 per cent made procedural changes and 20 per cent purely technical changes. Looking only at ultimately successful amendments that were attributable to parliamentary influence (which here include the small number of such non-government amendments), around three quarters (76 per cent) implemented policy change, while most of the remainder were procedural. If non-substantive amendments are filtered out, around two-thirds implemented policy change. We can thus dismiss concerns that parliamentary influence is largely on minor procedural matters. Of 321 amendments traceable to parliamentary influence, 131 were changes of substance, of which 89 related to actual policy change.

Table 5

Ultimately successful amendments by theme

Theme Total
 
Total with parliamentary influence
 
Substantive with parliamentary influence
 
Policy change: timing 18 2.4 1.2 1.5 
Policy change: other 485 64.5 239 74.5 87 66.4 
Procedure: regulations 75 10.0 57 17.8 32 24.4 
Procedure: other 27 3.6 17 5.3 10 7.6 
Technical 147 19.5 1.2 0.0 
Total 752 100 321 100 131 100 
Theme Total
 
Total with parliamentary influence
 
Substantive with parliamentary influence
 
Policy change: timing 18 2.4 1.2 1.5 
Policy change: other 485 64.5 239 74.5 87 66.4 
Procedure: regulations 75 10.0 57 17.8 32 24.4 
Procedure: other 27 3.6 17 5.3 10 7.6 
Technical 147 19.5 1.2 0.0 
Total 752 100 321 100 131 100 

These figures of course focus only on the changes wrought when a bill is actually in parliament, and so exclude pre-introduction (anticipated reactions) and post-passage (implementation, etc.) influence. The latter forms of influence are if anything probably more important, and apply in terms of high-level strategic priorities, not just legislative detail.

Conclusion

This article has presented summary results from the largest study of the legislative process at Westminster for 40 years. The resulting evidence suggests that the British parliament is far less dominated by the executive than is often assumed, and that it exerts a significant influence on government policy. Despite the various dismissive statements at the start of the article (just some of the many we could have cited), the empirical evidence simply does not back these up.12

A key reason for Westminster's reputation as weak is that it exhibits few visible signs of conflict with the executive: government defeats in the House of Commons are very rare, and few non-government amendments are passed in either chamber. In contrast, many hundreds of government amendments are agreed in both each year. This creates an impression not only of parliamentary weakness, but of executive dominance, with ministers freely changing their own legislation even after it has been introduced. But a more careful look at the amendment process shows that this is a misleading picture. The majority of government amendments have little substance, while most substantive changes made inside parliament in fact follow pressure from non-government parliamentarians. Meanwhile, many non-government amendments serve purposes other than seeking immediate policy change, such as ‘probing’, signalling to groups outside parliament, embarrassing the government or creating policy markers for the future. Published figures thus give a false impression of non-government failure, as well as of government success. We have shown that significant changes to legislation were wrought in parliament, both under the Labour government 2005–2010 and the Conservative/Liberal Democrat coalition.

Perhaps most importantly, however, parliament influences the shape of legislation long before it is introduced, and continues to influence policy implementation and future development once it has been formally passed. As Judge (1993) emphasised years ago, parliament sets the political parameters within which policy must be negotiated, even where the detailed discussions largely go on between government departments and interest groups. But this was a difficult argument to make without clear, empirical evidence of parliamentary policy influence. In exploring more tangibly the power of parliament, it is clearly not sufficient to scour the official record or trace changes made within the institution. Because parliament's greatest power is one of ‘anticipated reactions’, detecting this requires discussion with insiders (particularly from the executive) about behind-the-scenes negotiations. Grasping both the visible and hidden powers of parliament therefore requires a thorough and detailed mixed methods approach, which is very resource intensive.

Our scope has been limited to the UK parliament. But if these dynamics apply at Westminster, which is notoriously ‘executive-dominated’, they are surely present in most other parliamentary systems. Indeed they seem likely to characterise such systems. As cited above, numerous studies of other parliaments—throughout Europe and beyond—have suggested that executive bills and executive-backed amendments dominate. But these figures alone do not demonstrate parliamentary acquiescence. The key distinction between presidential and parliamentary systems is of course that executives in the latter are agents of parliament, rather than having independent electoral accountability (Lijphart, 1992). Where ministers depend on parliament for their survival, it should be no surprise that they seek to implement parliamentarians' policy wishes, and refrain from proposing policies that parliament will not accept. Indeed the ‘fusion’ rather than separation of executive/legislative power in such systems makes it difficult to say who is controlling whom. The visible amendments implemented inside parliament are just the final stages of a process that begins and ends with parliamentary accountability. Measuring legislative influence through looking for conflict is fundamentally to misunderstand how such systems work.

Acknowledgements

This work was funded by the Nuffield Foundation (grant number OPD/38858), and draws on earlier work funded by the ESRC (grant RES-063-27-0163). The views expressed are those of the authors and not necessarily of the Foundation. More information is available at www.nuffieldfoundation.org. We are grateful to Meghan Benton for her contribution to these projects, and to the anonymous reviewers for their comments.

References

Arter
D.
(
1985
) ‘
The Nordic Parliaments: Patterns of Legislative Influence
’,
West European Politics
 ,
8
,
55
70
.
Arter
D.
(
2006
) ‘
Questioning the Mezey Question: An Interrogatory Framework for the Comparative Study of Legislatures
’,
Journal of Legislative Studies
 ,
12
,
462
482
.
Benton
M.
Russell
M.
(
2012
) ‘
Assessing the Impact of Parliamentary Oversight Committees: The Select Committees in the British House of Commons
’,
Parliamentary Affairs
 ,
66
,
772
797
.
Blondel
J.
(
1970
) ‘
Legislative Behaviour: Some Steps Towards a Cross-National Measurement
’,
Government and Opposition
 ,
5
,
67
85
.
Brazier
A.
Fox
R.
(
2010
) ‘
Enhancing the Backbench MP's Role As a Legislator: The Case for Urgent Reform of Private Members Bills
’,
Parliamentary Affairs
 ,
63
,
201
211
.
Brazier
A.
Kalitowski
S.
Rosenblatt
G.
(
2008
)
Law in the Making: Influence and Change in the Legislative Process
 ,
London
,
Hansard Society
.
Cabinet Office
. (
2012
)
Guide to Making Legislation (June 2012 edn)
 ,
London
,
The Stationery Office
.
Capano
G.
Giuliani
M.
(
2001
) ‘
Governing Without Surviving? An Italian Paradox: Law-Making in Italy, 1987–2001
’,
Journal of Legislative Studies
 ,
7
,
13
36
.
Cowley
P.
(
2002
)
Revolts and Rebellions: Parliamentary Voting Under Blair
 ,
London
,
Politico's
.
Cowley
P.
(
2005
)
The Rebels: How Blair Mislaid His Majority
 ,
London
,
Politico's
.
Damgaard
E.
Jensen
H.
(
2006
) ‘
Assessing Strength and Weakness in Legislatures: The Case of Denmark
’,
Journal of Legislative Studies
 ,
12
,
426
442
.
Delegated Powers and Regulatory Reform Committee
(
2005
)
Sessions 2003–04 and 2004–05: The Work of the Committee (Eighteenth Report of Session 2004–05)
 ,
London
,
House of Lords
, accessed at .
Drewry
G.
, ed (
1985
)
The New Select Committees: A Study of the 1979 Reforms
 ,
Oxford
,
Clarendon
.
Griffith
J. A. G.
(
1974
)
Parliamentary Scrutiny of Government Bills
 ,
London
,
Allen and Unwin
.
Häge
F. M.
Kaeding
M.
(
2007
) ‘
Reconsidering the European Parliament's Legislative Influence: Formal vs. Informal Procedures
’,
Journal of European Integration
 ,
29
,
341
361
.
Hazell
R.
Chalmers
M.
Russell
M.
(
2012
) ‘
Pre-Appointment Scrutiny Hearings in the British House of Commons: All Bark, or Some Bite?
’,
Journal of Legislative Studies
 ,
18
,
222
241
.
Hindmoor
A.
Larkin
P.
Kennon
A.
(
2009
) ‘
Assessing the Influence of Select Committees in the UK: The Education and Skills Committee, 1997–2005
’,
Journal of Legislative Studies
 ,
15
,
71
89
.
Judge
D.
(
1993
)
The Parliamentary State
 ,
London
,
Sage
.
Kasack
C.
(
2004
) ‘
The Legislative Impact of the European Parliament Under the Revised Co-Decision Procedure: Environmental, Public Health and Consumer Protection Policies
’,
European Union Politics
 ,
5
,
241
260
.
Kelso
A.
(
2009
)
Parliamentary Reform at Westminster
 ,
Manchester
,
Manchester University Press
.
Kerrouche
E.
(
2006
) ‘
The French Assemblée Nationale: The Case of a Weak Legislature?
’,
Journal of Legislative Studies
 ,
12
,
336
365
.
King
A.
(
1976
) ‘
Modes of Executive–Legislative Relations: Great Britain, France and West Germany
’,
Legislative Studies Quarterly
 ,
1
,
11
36
.
King
A.
Crewe
I.
(
2013
)
The Blunders of Our Governments
 ,
London
,
Oneworld
.
Kreppel
A.
(
1999
) ‘
What Affects the European Parliament's Legislative Influence? An Analysis of the Success of EP Amendments
’,
Journal of Common Market Studies
 ,
37
,
521
537
.
Lijphart
A.
ed. (
1992
)
Parliamentary Versus Presidential Government
 ,
Oxford
,
Oxford University Press
.
Lijphart
A.
(
1999
)
Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries
 ,
New Haven
,
Yale University Press
.
Martin
L. W.
Vanberg
G.
(
2011
)
Parliaments and Coalitions: The Role of Legislative Institutions in Multiparty Governance
 ,
Oxford
,
Oxford University Press
.
Maurer
L. M.
(
1999
) ‘
Parliamentary Influence in a New Democracy: The Spanish Congress
’,
Journal of Legislative Studies
 ,
5
,
24
45
.
McGann
A.
(
2006
) ‘
Social Choice and Comparing Legislatures: Constitutional Versus Institutional Constraints
’,
Journal of Legislative Studies
 ,
12
,
443
461
.
Mezey
M. L.
(
1979
)
Comparative Legislatures
 ,
Durham, NC
:
Duke University Press
.
Moran
M.
(
2005
)
Politics and Governance in the UK
 ,
Basingstoke
,
Palgrave
.
Norton
P.
(
1998a
) ‘
The Institution of Parliaments
’. In
Norton
P.
(ed.),
Parliaments and Governments in Western Europe
 ,
London
,
Frank Cass
.
Norton
P.
(
1998b
) ‘
Old Institution, New Institutionalism? Parliament and Government in the UK
’. In
Norton
P.
(ed.),
Parliaments and Governments in Western Europe
 ,
London
,
Frank Cass
.
O'Dowd
J.
(
2010
) ‘
Parliamentary Scrutiny of Bills
’. In
MacCarthaigh
M.
Manning
M.
(eds)
The Houses of the Oireachtas: Parliament in Ireland
 ,
Dublin
,
Institute of Public Administration
.
Olson
D. M.
(
1994
)
Democratic Legislative Institutions: A Comparative View
 ,
Armonk, NY
,
M E Sharpe
.
Pettai
V.
Madise
U.
(
2006
) ‘
The Baltic Parliaments: Legislative Performance from Independence to EU Accession
’,
Journal of Legislative Studies
 ,
12
,
291
310
.
Richardson
J. J.
Jordan
A. G.
(
1979
)
Governing Under Pressure: The Policy Process in a Post-Parliamentary Democracy
 ,
Oxford
,
Martin Robertson
.
Russell
M.
(
2013
)
The Contemporary House of Lords: Westminster Bicameralism Revived
 ,
Oxford
,
Oxford University Press
.
Russell
M.
Sciara
M.
(
2008
) ‘
The Policy Impact of Defeats in the House of Lords
’,
British Journal of Politics and International Relations
 ,
10
,
571
589
.
Shell
D.
(
2007
)
The House of Lords
 ,
Manchester
,
Manchester University Press
.
Shephard
M.
Cairney
P.
(
2005
) ‘
The Impact of the Scottish Parliament in Amending Executive Legislation
’,
Political Studies
 ,
53
,
303
319
.
Strøm
K.
Bergman
T.
Müller
W.
, eds (
2003
)
Delegation and Accountability in Parliamentary Democracies
 ,
Oxford
,
Oxford University Press
.
Thompson
L.
(
2013
) ‘
More of the Same or a Period of Change? The Impact of Bill Committees in the Twenty-First Century House of Commons
’,
Parliamentary Affairs
 ,
66
,
459
479
.
Tsebelis
G.
Kalandrakis
A.
(
1999
) ‘
The European Parliament and Environmental Legislation: The Case of Chemicals
’,
European Journal of Political Research
 ,
36
,
119
154
.
Tsebelis
G.
Jensen
C. B.
Kalandrakis
A.
Kreppel
A.
(
2001
) ‘
Legislative Procedures in the European Union: An Empirical Analysis
’,
British Journal of Political Science
 ,
31
,
573
599
.
Zubek
R.
(
2011
) ‘
Negative Agenda Control and Executive–Legislative Relations in East Central Europe, 1997–2008
’,
Journal of Legislative Studies
 ,
17
,
172
192
.
1
Respectively: Chris Huhne, ‘Cleaning up the House’, Guardian (website), 27 January 2009, http://www.theguardian.com/commentisfree/2009/jan/27/lords-labour; Ken Macdonald, ‘Day One: Judges Put Power to the Test’, Times, 5 October 2009, http://www.thetimes.co.uk/tto/law/columnists/article2049400.ece; Simon Jenkins, ‘This House of Commons is God's Gift to Dictatorship’, Guardian, 1 November 2006, http://www.theguardian.com/commentisfree/2006/nov/01/comment.politics1.
2
These obviously occur in a different order for bills starting in the Lords. In both cases amendment is possible at third reading in the Lords, but not the Commons.
3
It should be noted that Westminster select committees do not usually conduct legislative scrutiny—the Commons committee stage of bills instead takes place in temporary, nonspecialist, Public Bill Committees. The select committees are permanent and specialist, but largely conduct executive oversight and investigations.
4
We looked at all reports from the departmental select committee shadowing the department that sponsored the bill, and crosscutting committees such as the Joint Committee on Human Rights and Lords Delegated Powers and Regulatory Reform Committee. In a small number of additional cases influence from committee reports in earlier sessions was spotted through comments made in debate. These were always checked against the relevant committee report.
5
Sometimes this desire for control appears absurd. For example, on one bill a backbench amendment to add the words ‘in writing’ was withdrawn and replaced at the next legislative stage by an identically worded government amendment.
6
Even in these cases parliamentary counsel may well have redrafted, as ministers occasionally allow proposers to meet with counsel when their amendment is acceptable, and to move the redraft themselves. Ministers also sometimes add their name to such amendments: but our figures (in line with the official record) classify all amendments signed by a minister as government amendments.
7
Often uncoordinated amendments at one stage are followed by more coordinated joint non-government amendments at subsequent stages, if government does not concede.
8
HL Deb 10 Oct 2011 cGC429.
9
HC Deb 22 Mar 2011 c835.
10
Both chambers can ultimately reject the government's secondary legislation, and such measures are sometimes withdrawn in order to avoid defeat, particularly in the Lords (Russell, 2013).
11
Here amendments forming part of a package were all given the same theme—so consequential amendments could count as either policy or procedural, while all amendments coded as ‘technical’ are part of a purely technical package.
12
Indeed King and Crewe do include evidence of parliamentary victories over the executive, but this goes unacknowledged. They cite the complete ban on smoking in public places introduced through Labour's Health Bill approvingly as ‘[o]ne of the Blair government's enduring accomplishments' (2013, p. 18); and likewise celebrate the abandonment of the compulsory identity cards scheme. Both of these changes—as seen here—were forced by parliament.