Cross-cutting issues,intraparty dissent and party strategy: The issue of European integration in the House of Commons

Abstract

When do legislative opposition parties use wedge issues to attack the government? In this article, I focus on the issue of European integration and its impact on party strategy in the 1992–1997 British House of Commons. Utilizing both voting and rhetorical data, the analysis reveals that both government and opposition were split on the issue, and thus the opposition was not able to use it. This, I argue, stems from the complexity of the issue, i.e. the fact that it combines redistributive cleavages with pre- and post-material ones, which cannot be suppressed by party leaders. The results demonstrate the importance of taking into consideration both the government’s and the opposition’s cohesion in modeling party strategies. Further, the combination of voting and rhetorical data adds to our understanding of the dimensionality and structure of partisan ideologies in Europe.

We have won more elections than any party in Britain because we are the most united and the most determined party in Britain. Over the same period Labour has been the most divided and the most undisciplined party. That is why they have lost, and lost, and lost, and lost again.

Sir Norman Fowler, Conservative Party Chairman, 1992 Party Conference.

Keywords

Wedge issues cross-cutting Euroskepticism opposition text-analysis

Introduction

The wedge issues literature posits that opposition party leaders use these issues strategically (Jeong et al., 2011; Miller and Schofield, 2003; Riker, 1986; Schattschneider, 1960). In this literature, a wedge issue is defined as ‘[A] policy issue dividing one party into two or more camps without disturbing the other party’ (Seo, 2011: 827, emphasis added). Based on work from McKelvey and others (McKelvey, 1976; McKelvey and Schofield, 1987), the literature also holds that politics is multidimensional and that this characteristic results in coalition instability. In turn, outside parties can take advantage of this instability by bringing a previously ‘passive’ issue (but one that is becoming more important) into the fore of political agenda, in an attempt to split the ruling coalition. In this article, I expand on this limited definition of a wedge issue and demonstrate that the emergence of cross-cutting issues that are complex can lead to completely different political dynamics.

I analyze issue dynamics in the 1992–1997 session of the British House of Commons. I focus on the topic of European integration, which scholars have characterized as orthogonal or only partly related to the traditional, national level left–right dimension (Hix, 1999; Hix and Lord, 1997; Hooghe and Marks, 2009; Hooghe et al., 2000), and as a result can be used as a wedge issue. Similar to Parsons and Weber (2011), I characterize this issue as a cross-cutting, double-wedged one. However, in contrast to Parsons and Weber (2011) who differentiate between the effect this issue has on parties after electoral victories and losses, I argue that the effect of electoral victory is limited and that the divisiveness of the issues affects both winners and losers in a similar fashion. The issue of European integration is divisive for both main parties because it is complex. Specifically, it involves political aspects that are rooted in multiple cleavages: pre- and post-materialist cleavages as well as an economic materialist one (Inglehart, 1977). This leads to low levels of party cohesion over the issue and makes it ineffective as a strategic weapon for the opposition. Thus, this case demonstrates that not only cross-party dynamics but also intraparty dynamics, especially the opposition’s party cohesion, are crucial to our understanding of the strategic use of issues as political ‘wedges.’

Scholars of Europeanization have thus far used qualitative methods or quantitative analysis of manifestos/surveys to gauge the dimensionality of the issue.¹ Here, I analyze fine-grained data on legislative voting and rhetoric, in order to map parties and party-factions on the issue of European Integration. While the analysis of parliamentary votes shows that parties are split on the issue, the analysis of parliamentary rhetoric enables me to demonstrate that party affiliation, the level of opposition to the European project and party status all have a nuanced effect on the stances of legislators on the issue. To foreshadow, the analysis demonstrates that nationalistic, sovereignty and fiscal aspects of Europeanization unite Euroskeptic factions from both parties against Europhiles. Second, that Labour Europhiles highlight egalitarian and cooperational aspects of Europeanization, while Conservative Europhiles are limited to discussing procedural aspects of it. These findings contextualize the ‘cross-cuttingness’ of the European integration issue and shed light on its complexity (i.e. its relevance to multiple political cleavages). Finally, they are helpful in explaining the current difficulties political leaders are facing in Europe, as they try to save the European project.

Issues as oppositional opportunities

The literature specifies two necessary conditions for the use of wedge issues. The first is the multidimensionality of the agenda. Clearly, if parliamentary politics only takes place along the strict lines of government versus opposition, then wedge issues do not exist. The second condition is that the issue in question divides one party into a number of factions. With regard to wedge-issues and party strategy, the theoretical expectation is that the opposition party will attack – ‘party leaders are not oblivious to the majority party’s intramural disputes. They marshal their forces to mount a fight that could help divulge a serious fissure on the other side of the aisle’ (Seo, 2011: 828). In terms of legislative votes, this assault should manifest itself as a series of votes in which the opposition votes cohesively against the government and is joined by defectors from the government’s party.

Figure 1 presents two possible ideal points’ distributions over a two-dimensional space in a two-party chamber. Without loss of generality, I assume that the horizontal dimension is the traditional right–left continuum and the vertical dimension reflects a nascent issue. In both panels, government and opposition parties are clearly distinguishable along the right–left dimension. Panel a presents the basic scenario described in the wedge issues literature. In it, the opposition is cohesive along the vertical dimension, but the government is not. Specifically, a group of government ‘rebels’ occupy the first quadrant and are distant from the government ‘loyalists.’ Moreover, it is possible to draw a line cutting the vertical dimension, separating the opposition and a group of government rebels from the government loyalists. Thus, this dimension presents an opportunity for the opposition. In contrast, panel b presents a double wedge. In particular, both the government and the opposition are split on this dimension. As a result, the opposition cannot use the vertical dimension to attack the government.

Estimating voting dimensionality in the House of Commons

As the starting point of this research, I model the House of Commons’ political dimensionality using parliamentary votes (divisions). This is important since scholars of legislative ideology and dimensionality have used models of legislative votes as their workhorse over the last three decades. The ‘industry’ of ideal point estimation research has focused mostly on the American Congress (e.g. Clinton et al., 2004; Cox and Poole, 2002; Desposato, 2006; Snyder and Groseclose, 2000), but scholars have increasingly used optimal classification (OC) to analyze legislators’ ideal points in other legislatures (Hix et al., 2006; Londregan, 2002; Rosenthal and Voeten, 2004).

Spirling and McLean (2007) argue that using OC to estimate parliamentarians’ preferences is problematic and will erroneously place extreme governing party members (rebels) near moderate opposition members. They demonstrate their claim using 1997–2001 data. The problem stems, arguably, from the fact that in some votes coalition rebels vote sincerely against the government, and opposition members almost always vote strategically against it (i.e. opposition members vote against the government regardless of their ideology). This contention overlooks two important facts: first, that opposition MPs do not always vote cohesively. There are cases where opposition members rebel against their own leadership and vote sincerely on the government’s bills. Second, while in most cases a unidimensional model is sufficient to explain voting behavior, consistent rebellions on specific issues imply that the agenda is not unidimensional. Thus, as I demonstrate in this section, the reason that OC places coalition rebels near opposition members stems from the fact that the binary choice nature of voting data cannot account for the multidimensionality of the agenda.

I analyze voting data from the 1992–1997 House of Commons. Since most votes in the House of Commons take place according to strict party lines, I focus only on the 245 dissenting votes during this session. Seventy-seven of those are related to European integration. To clarify, dissenting votes are those divisions where one or more Labour or Conservative MPs voted against their party whip or the wishes of their front bench.²

Figure 2 presents the results of a two-dimensional OC scaling algorithm. Inspection of the figure reveals two things.³ The figure shows clearly that European integration does not fall into the government-opposition dichotomy. While there is a clear distinction between Conservative and Labour MPs along the horizontal dimension, some MPs from both parties occupy a similar area along the vertical dimension. These MPs are members of Euroskeptic factions. While both parties’ frontbenchers were generally supportive of European integration (and in most cases Labour leaders requested their MPs to abstain), members of both the Tory and the Labour backbench rebelled against their own frontbench.

In order to visualize the difference between voting on European integration and on other issues, it is useful to compare Rice voting cohesion scores (Rice, 1925). As proposed by Rice, cohesion on a single division is calculated as: $Rice = \frac{| YES - NO |}{YES + NO}$ . This comparison is shown in Figure 3. While the parties are cohesive on other issues, cohesion scores’ distributions of votes on European integration have large variances (with mean values of 0.69 and 0.64 for Labour and Tories, respectively). Thus, both parties were facing low levels of cohesion on the issue of European integration, but were very cohesive on other dimensions.

To sum up, the issue of European integration is important in explaining voting behavior in the House of Commons, in addition to the traditional left–right continuum. While both parties had voted cohesively on other issues, they were clearly split on European integration. Based on the distribution of ideal points in Figure 2, it is evident that both parties were vulnerable to the European integration issue. This situation is similar to the picture presented in panel b of Figure 1. Thus, the voting data suggests that Labour was not able to ‘attack’ the Conservatives on European integration. The reason for this lies in the complexity of the issue, that is, the fact that it brings to the fore aspects of three different cleavage types: pre-materialist, materialist and post-materialist (Inglehart, 1977).

European integration as a complex issue

Using Inglehart’s (1977) terminology, we can characterize the European integration issue as encapsulating pre-materialist cleavages such as nationalism and identity, materialist cleavages such as economic distribution, and post-materialist cleavages such as green politics. For example, Hix (1999) argues that it is composed of two orthogonal cleavages: the integration-independence cleavage and the right–left one. The former captures questions of ‘more’ or ‘less’ integration among European states, and the latter has a libertarian-authoritarian component and an intervention-free market component. Hooghe and Marks (2009) posit that the traditional economic left–right continuum affects positioning only on European policies that involve redistribution and regulating capitalism. In addition, the European project reinforces a ‘pre-material/industrial’ dimension, in which national identity is the dominant component. Similarly, Hooghe et al. (2000) argue that an important factor related to parties’ attitude toward European integration is their position in a Traditional/Authoritarian/Nationalism (TAN) versus a Green/Alternative/Libertarian dimension.

Further, in terms of party competition, students of European integration make two important, complementary claims. First, some scholars (Aspinwall, 2002; Hooghe and Marks, 2009) find a robust, inverted U-shape pattern of support for Europeanization. This pattern describes a phenomenon whereby mainstream parties, located near the center of the left–right continuum generally support Europeanization, while parties at the extreme oppose it. According to the second claim, advanced by Hix and Lord (1997) and Hix (1999), mainstream parties are threatened by the national sovereignty aspect of Europeanization. As a result of this threat, these parties attempt to stifle partisan competition relating to aspects of Europeanization that are not subsumed by the national left–right continuum. Thus, the second claim highlights that parties are non-unified actors. They are broad coalitions of interests and views, specifically with regard to stances on Europeanization. This logic is taken a step further by Parsons and Weber (2011), who connect the issue of European integration to the vast wedge-issue literature.

Overall, these works highlight the complexity of the issue of European integration and the challenges it poses for political parties. First, the inverted U-shape pattern underscores the difference between mainstream and extreme parties. Second, because of the complexity of the issue, the level of support for Europeanization varies widely within mainstream parties. As a result of this variation, it has the potential to create internal party fractures and serve as a wedge issue by opponents. However, note that Europeanization poses a problem both for right- and left-wing mainstream parties, since these parties are a broad umbrella of interests. For right-wing parties, higher capital mobility and increased economic competition are attractive, but the possibility of a central European government that can regulate markets on a larger scale is problematic. Moreover, parties near the TAN pole, such as Conservative parties, value national identity and the political sovereignty of the nation state. Thus, within the right, neo-liberals’ interests clash with nationalistic views. For left-wing parties, Europeanization poses a threat because it impinges on the power of the welfare state and reduces its ability to regulate the economy. However, it also creates a potential for regulated capitalism as well as environmental and human rights policies at the European level. In the UK, this dynamic led to the formation of a multitude of political groups within mainstream parties, emphasizing a variety of Euroskeptic views.

In the UK, the inherent complexity of the issue of European integration led to the formation of a multitude of political groups within mainstream parties, emphasizing a variety of Euroskeptic views. I now turn to reviewing these ideological groups.

Varieties of Euroskepticism in the UK

The Conservative Party

The President of the Commission, Mr. Delors, said at a press conference the other day that he wanted the European Parliament to be the democratic body of the Community, he wanted the Commission to be the Executive and he wanted the Council of Ministers to be the Senate. No. No. No.

Margaret Thatcher, 30 October 1990, speech to the College of Europe, Bruges

From 1992 to 1997, most Tory frontbenchers were either Europhiles or neutral towards Europe. As part of the government, they took part in the process of negotiating the details of European Union (EU) integration. However, Euroskepticism was widespread among backbenchers. It was expressed by a number of legislative groups. The Friends of Bruges Group, which in the early 1990s was rumored to be associated with over 130 backbenchers (Forster, 2002), was composed mainly of Thatcherite-Nationalists. They viewed European monetary union as a threat to British nationhood and emphasized ‘defence of the popular sovereignty of the British people against the incursions of Brussels and against the drive for further integration’ (Gifford, 2010: 332). Economically, Thatcherites feared that European interventionist policies could threaten Britain’s newly deregulated economy. Therefore, the majority of the group opposed the European exchange rate mechanism (ERM).

The Fresh Start Group had its origin in two Early Day Motions; the first prompted by the Danish rejection of the Maastricht treaty in June 1992 and the second tabled on September 1992 after Black Wednesday, i.e. the Sterling’s withdrawal from the European ERM. The group’s main goals were opposition to the ratification of the Maastricht treaty, securing a referendum on the treaty and the issue of single currency (Forster 2002: 86).

Prominent Conservative MPs argued against the single currency. Norman Lamont, the former Chancellor of the Exchequer, maintained that the monetary union would only give a spurious appearance of economic stability. In his view, given the differences in economic structures between EU member states, the convergence pact would only enforce inappropriate policies on states experiencing the greatest difficulties as a result of being in the monetary union. Similarly, Jon Redwood, the Conservative spokesman on Trade and Industry, argued that the Maastricht convergence criteria would hurt jobs and raise taxes, but more importantly, his main argument was not primarily economic. Rather, it was ‘about the effect of single currency in displacing British national identity with a spurious European identity’ (George, 1998: 11). According to Redwood, joining the European currency would also erode British parliamentary sovereignty because decision making powers would be transferred to a non-elected European Bank.

The Labour Party

Since the Labour Party had gradually moved toward a pro-European stance by the late 1980s (Daniels, 1998), the formal opposition did not criticize the Maastricht agreement very much. When they did, it was mostly for not signing the Social Chapter. For Labour Europhiles, the union was seen as having the potential to ‘[t]ranscend national rivalries and to bring the power of public action to bear in a world otherwise dominated by unaccountable multinational companies and global capital’ (George, 1998: 15).

In contrast to this view, many backbenchers were still staunchly Euroskeptic. Labour Euroskeptics argued that the British economy differs in many ways from those of the continental European countries, and that these differences make it particularly risky for UK to join the union. The ‘right’ of the Labour Party echoed Conservative rhetoric, with MPs such as Peter Shore and Austin Mitchell advancing arguments regarding the loss of sovereignty and national autonomy. For other Labour Euroskeptics, the main problem with the treaty was that it imposed a deflationary policy that would make it impossible for a future Labour government to deliver on its commitments. Similarly, it was argued that the Maastricht version of the monetary union committed governments to driving down wages and social spending. In terms of monetary policy, Euroskeptics argued it was a dangerous experiment, a ‘leap in the dark.’

Partisan rhetoric and European integration

Thus far I have provided evidence for the existence of overall agenda multidimensionality in the 1992–1997 session of the House of Commons. I have used voting data to suggest that both main parties were divided on the emergent issue of European integration. Based on the European and the UK political parties’ literature, I have also argued that the issue is complex, bringing to the fore a number of important political sub-issues/frames. These can be grouped mainly into materialist and pre-materialist cleavages. In turn, this issue complexity leads to low levels of party cohesion on European integration, and makes it ineffective as an oppositional weapon. In this section, I provide additional evidence for the inability of the opposition to take advantage of the issue of European integration and present the ways in which the issue of European Integration cuts through both parties.

Analysis of rhetoric

While an analysis of voting suggests that the opposition was divided over the issue of European integration, there are important reasons for analyzing rhetoric. First, the dynamics of partisan parliamentary rhetoric is different. Even if the opposition does not mount a legislative assault on the government, it can still join forces with Tory Euroskeptics and use its floor speeches to highlight the problematic aspects of Europeanization. Especially since the majority does not have strict rhetorical agenda control, legislative speeches can be used to contribute to ruling parties’ splits (Tzelgov, Forthcoming). Second, the analysis of rhetoric adds context and detailed information regarding the structure and nature of partisan splits. Third, rhetorical data allow us to test hypotheses which cannot be tested by relying only on voting data.

At the most general level, using European integration as a wedge issue implies that Labour had been more likely to discuss European integration than the Conservative party. This reasoning is grounded in the selective emphasis approach (Budge and Farlie, 1983), which underlies the Comparative Manifestos Project (Budge et al., 2001). According to this view, parties do not take opposing positions on the same issue (e.g. more versus less social spending), but rather choose to focus on issues on which they think they have an advantage.

To test whether Labour was more likely to discuss European integration, I compare the partisan proportion of utterances on that issue to the partisan proportion of utterances over other topics in the 1992–1997 period. Overall, Labour and Conservative MPs have made 547,059 utterances. Of those, 11,847 were on European integration. This yields a list of 535,212 utterances on all other issues. A Welch t-test comparing the proportions of Labour and Conservative speakers across both sets of issues does not support the Labour European integration wedge issue strategy expectation. Labour MPs were not more likely to bring up the issue than Conservative MPs.⁴

Next, I examine party and faction rhetoric on European integration. If, as the wedge issues’ literature predicts, Labour had been using the issue to attack the divided Conservative party, then we should see a cohesive rhetorical assault by Labour. However, if, as I suggest, ideological fissures within Labour had prevented it from using the issue strategically, these divides should be discernible in speeches. Most importantly, the divides should reflect our prior expectations and knowledge regarding the position of factions over European integration, rather than stem from random lexical variation.

Inter-party expectation

Given that both main parties are constrained in their approach to European integration by previous interests, and that some economic aspects of European integration have been subsumed by the traditional right–left continuum (Hooghe and Marks, 2009), the first expectation is the following:

H1: In its European integration rhetoric, the Labour party (as a whole) should emphasize social, employment and production aspects. These reflect the party’s overall objection to the refusal of the Conservative government to sign Maastricht’s Social chapter, as well as traditional Labour concerns.

Note that this is the only expectation regarding difference between parties as a whole. Both Europhile and Euroskeptic Labour MPs are committed to lowering unemployment, and both see social aspects of European integration as shaving high priority. Moreover, relative to Tories, Labour MPs are more concerned with the state’s ability to influence and plan economic production.

Inter-faction expectations

Two issues are expected to differentiate Euroskeptics from Europhiles: nationhood and the economy. With regard to the first aspect, the theoretical literature suggests that both Tory and Labour skeptics should use nationalistic frames in their European integration rhetoric. Economically, Euroskeptic factions from both parties decried the monetary convergence (Euro) aspects of European integration.

H2: Relative to Europhiles, Euroskeptics are more likely to emphasize nationalistic aspects of Europeanization and highlight aspects of monetary convergence.

Intraparty expectations

Here, I explicate the expectations regarding the differences within the two parties, by comparing Euroskeptics and Europhiles. While some of these expectations reflect the overall Euroskeptic/Europhile divide, they also deal with additional, intraparty divides.

H3a: Relative to Labour Europhiles, Labour Euroskeptic MPs are likely to emphasize nationalism and highlight monetary convergance. In contrast, Labour Europhiles, supportive of the notion of a supranational accountable government, should emphasize the cooperational aspects of European integration.

H3b: Relative to Tory Europhiles, Tory Euroskeptic MPs are likely to emphasize nationalism and discuss monetary convergence. In addition, since Tory Europhiles are mostly frontbenchers who are involved in the negotiations, their rhetoric should reflect technical and functional aspects of European integration negotiations.

Data and methodology

Recent computational advances and methodological innovations in political science have led to a growing volume of research using ‘words as data,’ describing the agenda of political institutions and estimating ideological party positions (e.g. Diermeier et al., 2011; Grimmer, 2010; Quinn et al., 2010). Here, I use a similar approach in analyzing the debates of the House of Commons.

Records of House of Commons debates (Hansard) are found online. The records were converted to XML format. Each day of debates is structured hierarchically, with sentences nested within utterances, which are in turn nested within discussions. I use Python scripts to extract only debates whose title includes stems that indicate they address the issue of European integration.⁵ Using these data, I generate a vocabulary based on the 5000 most common words. Leveraging the nested structure of the data, I then generate a matrix of utterance by stem, where each entry represents the number of times a stem i appears in utterance d. These entries are tagged by details identifying the utterance by date of the discussion as well as by the name of the speaking MP, his partisan affiliation and whether he is a Euroskeptic or not.⁶ Next, I remove stop-words from the data as well as stems that have low term frequency-inverted document frequency values.⁷ This yields a 11,807 × 4000 matrix, which is the input used for the statistical analysis.

To classify the corpus into topics, I use a latent Dirichlet allocation (LDA) model (Blei et al., 2003). The LDA is a hierarchical Bayesian model, in which each document (here, legislative utterance), is modeled as a mixture over an underlying set of topics. Thus, every political utterance exhibits topics with different proportion.

In the context of LDA, the topics are hidden variables and the observed variables are the documents’ words. In terms of inference, we use the LDA to compute the posterior distribution of the topics given the words. More formally, the topics are β_1: _K , where β_K is the distribution of words in the topic. Topic proportions for the dth document are θ_d, where θ_d_, _k is the proportion of topic k in document d. In addition, document topic assignments are z_d, and z_d_, _n is the topic assignment for the nth word of document d. Finally, the words for document d are w_d, where w_d_, _n is the nth word in document d. The implied joint distribution of the hidden and observed variables is:

p (β_{1 : K}, θ_{1 : D}, z_{1 : D}, ω_{1 : D}) = Π_{k = 1}^{K} p (β_{k}) Π_{d = 1}^{D} p (θ_{d}) (Π_{n = 1}^{N} p (z_{d, n} | θ_{d}) p (ω_{d, n} | β_{1 : K}, z_{d, n}))

And the posterior distribution is:

p (β_{1 : K}, θ_{1 : D}, Z_{1 : D} | ω_{1 : D}) = \frac{p (β_{1 : K}, θ_{1 : D}, Z_{1 : D}, ω_{1 : D})}{p (ω_{1 : D})}

The full results of a 22 topic LDA model are presented in Table A1, which can be found in the online appendix.⁸ To illustrate some of the LDA results, under the title ‘Monetary Convergence’ we find stems like: ‘currenc,’ ‘single,’ ‘converg,’ ‘monetari,’ etc. Under the title ‘Elections,’ we find stems like: ‘govern,’ ‘scotland,’ ‘repres,’ ‘elect,’ etc., and under the title ‘Fiscal Policies,’ we find stems like: ‘Monei,’ ‘cost,’ ‘tax,’ and ‘public.’ Note that there are a number of topics that do not convey any relevant information for analytical purposes. These topics are ‘Procedural,’ ‘Q&A,’ ‘Party Politics’ and ‘Parliamentary.’ I therefore drop utterances that were classified under any of these topics. Hence, the data set used for hypotheses testing includes 8236 utterances, distributed over 18 topics.

Results

It is useful to think of the topics extracted using the LDA model as frames of reference to the complex issue of European integration. In order to provide a simple test of my hypotheses, I focus on the log odds-ratios of using these frames by political groups (i.e. a logit transformation). Specifically, if we denote as $π_{i} = \frac{p_{i, k}}{1 - p_{i, k}}$ the odds of group i using a specific topic k, then:

θ_{i, j}^{k} = \log \frac{π_{i, k}}{π_{j, k}}

is the logged odds-ratio of using a topic k for the two groups.

Figures 4 and 5 present the results of this model applied to test the first two hypotheses, respectively comparing parties’ and factions’ rhetoric on Europeanization.

The results presented in Figure 4 support H1. Labour is significantly more likely (by 112%) to discuss employment/social aspects of European integration, as well as being more likely to emphasize the production and agriculture frame (see column 4 on the bottom of Table A1, a frame characterized by stems such as: ‘agricultur,’ ‘govern,’ ‘produc,’ ‘import,’ ‘cap,’ etc.)

Figure 5 presents the logged odds-ratios estimates of Euroskeptics frame use relative to Europhile frame use (factions are aggregated across parties). H2 is therefore also supported. As expected, nationalistic/sovereignty and Monetary Convergence frames characterize Euroskeptics’ discourse. Specifically, Subsidiarity⁹ (see column 5 on the top of Table A1, a frame characterized by stems such as: ‘subsidiar,’ ‘cultur,’ ‘languag,’ ‘differ’ etc.) and Citizenship emerge as topics that are used much more by Euroskeptics, followed by the issues of Monetary Convergence and the Powers of the European Parliament (a topic capturing mainly the transfer of powers from the British parliament to the European one.) The differences in these effects are large in magnitude. For example, Euroskeptics are approximately 167% more likely to discuss Subsidiarity and 118% more likely to discuss Citizenship than their fellow Europhile MPs. In addition, Euroskeptics are 58% more likely to discuss Monetary Convergence than Europhiles.

Next, Figure 6 presents intraparty comparisons between Euroskeptics and Europhiles within the Labour and Conservative parties, respectively (H3a).¹⁰

First, Labour Euroskeptics are significantly more likely than Labour Europhiles to discuss the Subsidiarity frame (by 68%) and the citizenship frame (a frame characterized by stems such as (‘citizen,’ ‘nation,’ ‘countri’ etc). Second, compared to Europhiles, Labour Euroskeptics are significantly more likely to discuss the problematic aspect of Monetary Convergence. As expected, Labour Europhiles are significantly more likely to discuss intergovernment cooperation between the UK and Europe.

Similar to the trends we find in the Labour Party, Tory Euroskeptics are more likely than their party peers to use nationalistic/sovereignty frames of European integration (e.g. Subsidiarity and Citizenship, by 278% and 129% respectively) as well as Monetary Convergence. In contrast, Europhiles are approximately 200% more likely to discuss EU meetings and 150% more likely to discuss EU policies. This is due to their participation in EU negotiations as government frontbenchers.

These results generally support my theoretical expectations. I find that party membership affects the rhetoric of Labour MPs in a predictable way. Regardless of their level of support for the European project, they are all more likely to use social frames, i.e. Employment/Social and Production and Agriculture frames of the issue relative to Tories. Similarly, Euroskeptics from both sides of the aisle are more likely to use nationalistic/sovereignty frames relative to Europhile MPs. Finally, the rhetoric of Tory Europhiles is mostly influenced by their membership in government, as is evident by their relatively high likelihood of using functional frames.

While informative, these binary comparisons cannot succinctly summarize the overall variation in the 18 frames/components of the European integration issue. In order to reduce the dimensionality of these data and place factions/parties on a unified scale, I use principal component analysis. Based on the preceding analysis, we should expect two meaningful components to account for the variance in frame use. These components should represent ideological differences between Euroskeptics and Europhiles, as well as between the parties.

Figure 7 summarizes the principal component analysis results using varimax rotation. Factions and parties are placed on a two-dimensional space, corresponding to the component (for example, the PHILE faction in the figure represents the weighted factors’ scores of Europhiles, aggregated across both parties. LAB is the weighted aggregation of factors’ scores for Labour’s Europhile and Euroskeptic factions.). The arrows represent the respective frames’ loadings on each dimension.

Along the horizontal axis (a dimension capturing 70% of the variance) of Figure A2 in the online appendix, the frames that have the lowest (negative) loadings on this component are Fiscal Policies (taxation and spending), two national/sovereignty frames: Subsidiarity and Citizenship. These are followed by the Maastricht and Monetary convergence frames. In contrast, two functional/procedural frames (EU Meetings and Financial Reports) are the frames with the highest positive loadings, followed by UK–EU Economic Cooperation. This dimension clearly captures the Euroskeptic/Europhile distinction. On the one hand, Euroskeptic factions’ scores indicate the nationalistic views and objections to Maastricht. On the other, Europhile factions highlight their views of working within the European project. I characterize this as a ‘Nationalism & Anti-Fiscal Integration vs. European Governance’ dimension.

Along the vertical axis (22% of the variance), the frames with the highest loadings are procedural frames such as EU policies, Maastricht Technical and EU meetings. On the negative pole, the strongest frames are Employment/Social, Production & Agriculture, Elections, and UK–EU economic Cooperation. Two of these frames (Social and Production) capture traditional Labour concerns. The other two (Elections and UK–EU Cooperation) capture Labour’s vision of an integrated Europe. This dimension separates Europhiles from both parties. I characterize this dimension as ‘Egalitarianism & Cooperation vs. European Government Procedures.’

Discussion and conclusions

A number of findings emerge from the analysis of voting and rhetoric in the 1992–1997 session of the House of Commons. First, using a two-dimensional OC model, I show that MPs from the ruling Conservative party and the Labour opposition were clustered in similar coordinates along a dimension that is oblique to the traditional left–right continuum. These ‘rebels’ are generally Euroskeptics, members of both the government and the opposition who voted (sincerely) against the wishes of their respective parties’ leadership. Thus, this dimension captures the fact that the emergent issue of European integration is a double-wedge issue, cutting across both parties. This suggests that in terms of voting, the Labour opposition was not likely to attack the Conservatives over the issue of Europeanization, and that a wedge issue strategy was not a feasible one.

I then use an LDA model to classify all the utterances spoken by MPs in debates over European integration. This enables me to compare partisan rhetoric across parties and party-factions. I test hypotheses regarding the relative rhetorical emphases of parties and factions, and by implication, their positions on the multiple dimensions of the European integration issue. On the one hand, the findings I obtain from the rhetorical analysis reinforce the results obtained from analysis of voting. Put succinctly, the issue of European integration is cross-cutting, that is, it cuts across both main parties. However, the analysis of rhetoric demonstrates exactly how both the government and opposition are split, and consequentially, why the opposition cannot use the Europeanization as a wedge issue, as the wedge issue literature posits. In addition, the results of the rhetorical analysis reveal that MPs choices of frames are a function of an interaction involving their partisan affiliation, their support for European integration and their party seniority. Further, as other scholars have noted, I find empirical evidence of the complexity of the European integration issue, involving mainly pre-materialist (nationalistic) and materialist (economic) and to a lesser extent post-materialist (cooperational) aspects.

Specifically, partisan rhetorical emphases were shown to be conditional on party affiliation: Labour MPs are highly likely to address European integration using social frames, regardless of their level of support for the European project. Second, regardless of party affiliation, Euroskeptic MPs are highly likely to frame European integration in terms of British nationhood, citizenship and political sovereignty. They are also likely to emphasize the problematic aspects of monetary convergence. Third, Tory Euroskeptics are likely to criticize European integration by emphasizing fiscal frames. Fourth, Labour Europhiles are likely to address Europeanization by emphasizing intragovernmental cooperation. Finally, Conservative Europhiles’ discourse of Europeanization is characterized by functional/procedural frames, reflecting their frontbench status and their role in European integration negotiations.

Given the complexity of the European integration issues, principal component analysis techniques were helpful in reducing the dimensionality of the rhetorical data and place parties/factions on a common scale. The main dimension within the issue of European integration was found to cut across both parties, pitting Euroskeptic nationalists and opponents of monetary unification versus Europhiles from both parties, who are concerned with different visions of working within the EU. The fact that nationalistic frames are characteristic of skeptics from both traditional right and left is in accordance with recent literature (Halikiopoulou et al., 2012).

The second dimension was found to separate Europhiles, with Labour MPs focusing on egalitarian (i.e. social, electoral, cooperational) aspects of Europeanization, while Tories were limited to focusing on functional/procedural rhetoric, due to their participation in government.

These results complement the extant literature on wedge-issue politics. Most importantly, they demonstrate clearly that cross-cutting, complex issues cannot be used in wedge-issues tactics by the opposition. Within the context of European party politics, this article shows that even when the party system is composed mainly of two large mainstream parties, these parties host both supporters and opponents of the European project, and thus the ‘depoliticization’ (Hooghe and Marks, 2009) of the issue within mainstream parties is impossible. Finally, in contrast to Parsons and Weber (2011), this detailed analysis of party rhetoric shows that in the case of a complex, cross-cutting issue, party leadership is not always able to muffle the issue and suppress dissent.

This article seeks to contribute to a number of debates. First, it adds to our understanding of the strategic use (or lack thereof) of wedge issues by opposition parties. When opposition parties are themselves threatened by new issues, the opposition-government dynamic alone cannot explain party strategy. In these cases, extant political ideologies interact with the positions all legislators occupy on the new dimension of conflict. Whether these new positions are related to their preferences on other dimensions is an empirical question. In the case of European integration, while some new economic aspects were related to it and could be explained by positions on the left–right continuum, others were not. In particular, pre-materialist concerns were found to be oblique to the left–right scale. This configuration prevented the opposition from mounting a wedge issue assault on the government.

Second, the article demonstrates the advantages of combining quantitative analysis of votes and rhetoric in order to improve our understanding of parliamentary parties’ strategies, and the dimensionality of the political agenda. While studying votes has become the workhorse of legislative studies, adding analysis of rhetoric can shed light on the intricacies of political ideas and on an additional arena of oppositional politics.

Finally, the paper adds to our understanding of the issue of European integration. It is important to understand the complexity of the issues underlying the political conflict in Europe, especially when the future of the Union seems unclear. The findings highlight the difficulties European leaders are facing as they try to form new coalitions and handle an issue composed of important economic, nationalistic and post-materialist aspects.

Footnotes

Acknowledgements

I thank Burt Monroe, Phil Schrodt, and the conference participants for their suggestions.

Funding

Parts of this research were supported by National Science Foundation Doctoral Dissertation Improvement Grant 1024190, and BCS Grant 07-14688.

Notes

References

Aspinwall

(2002) Preferring Europe: Ideology and national preferences on European integration. European Union Politics 3(1): 81–111.

Blei

Jordan

(2003) Latent Dirichlet allocation. Journal of Machine Learning Research 54(1): 209–228.

Budge

Farlie

(1983) Party competition – selective emphasis or direct confrontation? An alternative view with data. In: Daadler

Mair

(eds) Western European Party Systems, Beverly Hills, CA: Sage.

Budge

Klingemann

Bara

(2001) Mapping Policy Preferences: Estimates for Parties, Electors & Governments 1945–1998, Oxford: Oxford University Press.

Clinton

Jackman

Rivers

(2004) The statistical analysis of roll call data. American Political Science Review 98(2): 355–370.

Cox

Poole

(2002) On measuring partisanship in roll-call voting: The U.S. House of Representatives, 1977–1999. American Political Science Review 46(3): 477–489.

Daniels

(1998) From hostility to constructive engagement: The Europeanisation of the Labour party. West European Politics 21(1): 72–96.

Desposato

(2006) The impact of electoral rules on legislative parties: Lessons from the Brazilian Senate and Chamber of Deputies. Journal of Politics 68(4): 1015–1027.

Diermeier

Godbout

(2011) Language and ideology in Congress. British Journal of Political Science 42(1): 33–55.

10.

Forster

(2002) Euroscepticism in Contemporary British Politics, London: Routledge.

11.

George S (1998) The intellectual debate in Britain on the European Union (Research and policy paper no. 5). Paris: Notre Dame.

12.

Gifford

(2010) The UK and the European Union: Dimensions of sovereignty and the problem of Euroskceptic Britishness. Parliamentary Affairs 63(2): 321–338.

13.

Grimmer

(2010) A Bayesian hierarchical topic model for political texts: Measuring expressed agendas in Senate press releases. Political Analysis 18(1): 1–25.

14.

Halikiopoulou

Kyriaki

Vasilopoulou

(2012) The paradox of nationalism: The common denominator of radical right and radical left Euroscepticism. European Journal of Political Research 51(4): 504–539.

15.

Heppell

Hill

(2008) The Conservative Party leadership election of 1997: An analysis of the voting motivations of Conservative parliamentarians. British Politics 3(1): 63–91.

16.

Hix

(1999) Dimensions and alignments in European Union politics: Cognitive constraints and partisan responses. European Journal of Political Research 35(1): 69–106.

17.

Hix

Lord

(1997) Political Parties in the European Union, London: Palgrave Macmillan.

18.

Hix

Noury

Roland

(2006) Dimensions of politics in the European Parliament. American Journal of Political Science 50(2): 494–511.

19.

Hooghe

Marks

(2009) A postfunctionalist theory of European integration: From permissive consensus to constraining dissensus. British Journal of Political Science 39(1): 1–23.

20.

Hooghe

Marks

Wilson

(2000) Does left/right structure party positions on European integration? Comparative Political Studies 35(8): 965–989.

21.

Inglehart

(1977) The Silent Revolution – Changing Values and Political Styles among Western Publics, New York: Princeton University Press.

22.

Jeong

Miller

Schofield

(2011) Cracks in the opposition: Immigration as a wedge issue for the Reagan coalition. American Journal of Political Science 55(3): 511–525.

23.

Londregan

(2002) Legislative Institutions and Ideology in Chile’s Democratic Transition, New York, NY: Cambridge University Press.

24.

MacKay

(2003) Information Theory, Inference, and Learning Algorithms, Cambridge, UK: Cambridge University Press.

25.

McAllister

Studlar

(2000) Conservative Euroscepticism and the referendum party in the 1997 British General Election. British Journal of Political Science 6(3): 359–371.

26.

McKelvey

(1976) Intransitivities in multidimensional voting models and some implications for agenda control. Journal of Economic Theory 12(3): 472–482.

27.

McKelvey

Schofield

(1987) Generalized symmetry conditions at the core. Econometrica 55(4): 923–933.

28.

Miller

Schofield

(2003) Activists and partisan realignment in the U.S. American Political Science Review 97(2): 245–260.

29.

Parsons

Weber

(2011) Cross-cutting issues and party strategy in the European Union. Comparative Political Studies 44(4): 383–411.

30.

Quinn

Monroe

Colaresi

(2010) How to analyze political attention with minimal assumptions and costs. American Journal of Political Science 54(1): 209–228.

31.

Rice

(1925) The behavior of legislative groups. Political Science Quarterly 40(1): 60–72.

32.

Riker

(1986) The Art of Political Manipulation, New Haven, CT: Yale University Press.

33.

Rosenthal

Voeten

(2004) Analyzing roll calls with perfect spatial voting. American Journal of Political Science 48(3): 620–632.

34.

Schattschneider

(1960) The Semi-Sovereign People: A Realist’s View of Democracy in America, New York, NY: Holt, Rinehart and Winston.

35.

Seo

(2011) Wedge-issue dynamics and party position shifts Chinese exclusion debates in the post-reconstruction US Congress, 1879–1882. Party Politics 17(6): 823–847.

36.

Snyder

Groseclose

(2000) Estimating party influence in Congressional roll-call voting. American Journal of Political Science 44(2): 187–205.

37.

Spirling

McLean

(2007) UK OC OK? Interpreting optimal classification scores for the U.K. House of Commons. Political Analysis 15(1): 161–185.

38.

Tzelgov

(Forthcoming) Damned if you do and damned if you don’t: rhetorical heresthetic in the Israeli Knesset. Party Politics.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

6.13 MB

0.00 MB

0.15 MB