Does religion count for religious parliamentary representation? Evidence from Early Day Motions

ABSTRACT

The article addresses one facet of the representation puzzle, namely substantive minority representation in the UK House of Commons. It examines whether a religious Jewish and Muslim minority background stimulates politicians from these backgrounds to address issues of concern for Jewish and Muslim minority groups in Early Day Motions (EDMs), and compares the effects from identity-based and institutional predictors. The study draws upon previous studies that used low-cost parliamentary activities to assess the impact of gender and ethnic minority identities on the representation of women and ethnic minorities, employing quantitative content analysis and time-series cross-sectional data analysis to examine the content of EDMs sponsored by members of parliament from Jewish and Muslim background (plus a control group) between 1997 and 2012. The analyses test for the effects of religious background and institutional predictors on the likelihood of referring to minority issues. They show that identity-based predictors such as a religious background are vastly inferior to institutional factors, including a legislative role, representing a constituency with a significant proportion of minority population, and the length of parliamentary service, in determining such references.

KEYWORDS:

• parliamentary representation
• religious minority
• Early Day Motion
• text analysis
• time-series cross-sectional analysis

Acknowledgements

The author would like to express gratitude to Prof Philip Cowley, Dr Caitlin Milazzo and Dr Stuart Fox, as well as the participants of the Nottingham Postgraduate Conference ‘Politics in Crisis?’ and the 2013 EPOP Annual Conference for their helpful feedback and comments on earlier versions of this paper.

Note on Author

Ekaterina Kolpinskaya is an Associate Lecturer in Quantitative Methods, College of Social Sciences and International Studies, University of Exeter, UK, email: [email protected]

Notes

1. The sample includes all Jewish and Muslim politicians elected to the House of Commons between 1997 and 2012 (Cracknell, 2012; Janner & Taylor, 2008). Additionally, a control group of 25 non-minority MPs is selected using stratified random sampling. It includes the following characteristics of MPs: (1) white ethnic background; (2) Christian or secular religious background; (3) representing ‘minority’ and ‘non-minority'; and (4) being elected from the Labour, Conservative or Liberal Democrat parties between 1997 and 2012.

2. The motions sponsored by the MPs in the sample are a more appropriate source of data. That is because sponsoring a motion implies a stronger commitment to its cause, rather than signing it, because it is more labour intensive and requires an MP to invest more time and effort (House of Commons, 2010). The terms ‘to sponsor’ and ‘to table’ an EDM are used interchangeably.

3. The analysis does not account for the effect from the party. Although it is the main predictor of behaviour in high-cost parliamentary activities such as voting (Cowley, 2002), it is not as relevant in low-cost activities such as tabling EDMs. This has been supported by the early stages of this study. Including the party in the model harms the model fit, making the results of the analysis inconclusive. That is why the model includes the institutional predictors that have proven more relevant for low-cost parliamentary activities – a legislative role, the proportion of religious minority groups, and the length of parliamentary service. Using these predictors vastly improved the fit of the models. These predictors partly capture the effect of party that cannot be added to the model alongside these predictors because it leads to over-controlling.

4. Three types of constituency – ‘Jewish’, ‘Muslim' and ‘White’ – are defined on the basis of community watch-lists and the proportion of minority population recorded in the 2001 and 2011 censuses. The interest of Jewish and Muslim minorities in certain constituencies is noted in community-affiliated election campaigns, such as The Jewish Chronicle’s Election 2010 (Jewish Chronicle, 2010), and the Muslim Council of Britain's Muslim Vote (Muslim Vote, 2010). Jewish and Muslim presence in a constituency is considered significant if it is twice as large as the average proportion of minority population across the country – 1.5 per cent or more Jews and 10 per cent or more Muslims for England, and 0.6 per cent of Jews and 2 per cent of Muslims for Scotland (ONS, 2003, 2012). Although the latter criterion to defining a ‘minority’ constituency is arbitrary, it indicates the areas densely populated by British Jews and Muslims in relation to the country's average. These constituencies are coded as ‘minority’ constituencies, while the rest are coded as ‘non-minority’ seats.

The constituencies represented by 74 selected MPs are observed over 14 parliamentary sessions (i.e. 621 observations). Three hundred and twenty-five ‘minority’ and 296 ‘non-minority’ constituencies are included in the analysis. Observing constituencies over time allows controlling for the change of population, including the significance of a Jewish and/or Muslim presence.

5. Although PPSs are not parliamentary leaders as such, for the purposes of this empirical study the differences between them and other MPs in leadership roles (e.g. ministers) are far less substantial and significant than the differences between them and backbench MPs when it comes to their behaviour in relation to EDMs and Parliamentary Questions for written answers (WPQs). In so far as PPSs are bound by institutional constraints, e.g. collective agreement (Cabinet Office, 2011, p. 23), which lead them consistently to behave more similarly to ministers and other MPs in leadership roles than to backbench MPs, it makes sense to categorise them with the former in this context.

6. To account for the change of legislative roles, this study observes 74 MPs selected over 14 parliamentary sessions, adjusting their legislative roles when they are promoted or demoted. Overall, the behaviour of 345 backbenchers and 276 MPs in leadership roles is considered in the analysis.

7. ‘Minority issues’ are operationalised using indicators that reflect on the issues of concern for Jewish and Muslim minority groups. ‘British Jewry’ and ‘British Muslims’ focus on internal community development, their needs and interests. ‘Middle East’ and ‘South Asia’ draw upon the engagement of British Jews and Muslims with their countries of origin. Finally, ‘Immigration’ is a concept of general concern that focuses upon the legislation, institutions, and actors involved in drafting and implementing immigration legislation.

8. The analysis is conducted on 74 groups (MPs) multiplied by 14 time periods (sessions). This gives 1036 observations, of which 617 are valid and 419 observations are missing. The missing values are the result of the uneven length of parliamentary service of the MPs in the sample. When the data are declared to be TSCS data, the values for MPs elected in the 2001 general election, for instance, are considered missing for the earlier sessions and omitted. Similarly, for MPs who stood down and/or lost an election, the values for the following sessions are considered missing and omitted from the analysis.

9. A random effects model is preferred to a fixed effects model because it is feasible to assume that the unobserved effect is uncorrelated with each explanatory variable. Additionally, the IVs are relatively stable over time. For example, a religious background is time-constant, whereas the shares of minority population in constituencies and legislative roles change infrequently. Finally, random effects models are preferred to fixed effects models in pooled logistic regressions, which are used in the analysis (Wooldridge, 2013).

10. In dependent variables, ‘0' stands for no references to minority issues, and ‘1' for one or more references to a minority issue per session. In the ‘religious background’ variables (IVs), ‘1' stands for having a Jewish or Muslim, background, and ‘0' for not having such a background. In the ‘legislative role’ variable (IV), ‘1' stands for holding a leadership role, and ‘0' for being a backbencher. In the ‘minority’ constituency variable (IV), ‘1' stands for representing a ‘minority’ constituency, and ‘0' for representing a ‘non-minority constituency’. The length of parliamentary experience is a continuous variables measured in the number of parliamentary sessions in which an MP has served.

11. For instance, Holocaust memorial days (EDM 287, 2009–10; EDM 558, 2009–10), Holocaust Education Trust Auschwitz trips (EDM 1032, 2007–8; EDM 1317, 2010–12) and the ‘Never again for anyone’ initiative (EDM 1360, 2010–12).

12. For instance, ‘ … all progressive people to invoke the spirit of Cable Street again today by challenging and defeating the rise of racism and fascism at home and abroad’ (EDM 2705, 2005–6) and ‘ … the Government …  the police service and local government to prevent … attacks against places of sacred or religious significance, which cause[d] such pain and incite[d] fear and division’ (EDM 343, 2005–6).

13. Although there have not been any Muslim parliamentarians elected from a constituency with a significant Jewish population, there have been a few non-minority and Jewish parliamentarians representing ‘minority’ constituencies, such as Oona King (Bethnal Green and Bow), Robert Halfon (Harlow), Harry Cohen (Leyton and Wanstead), Sir Gerald Kaufman (Manchester Gorton), Mike Gapes (Ilford South), Stephen Timms (East Ham), Jack Straw (Blackburn), Terry Rooney and David Ward (Bradford East).

14. For instance, the invasion in Iraq increased criticism of the government's foreign policy (BBC, 2006, 12 August), raised concerns over the dangers of involvement in the Middle East for British troops and was linked with the 2005 attacks in London (Ipsos Mori, 2005, 23 July). It triggered the growth of xenophobia and mistrust of migrant-origin persons, especially Muslims, putting their regions of origin – South Asia and the Middle East – in the spotlight and causing the increasing ‘securitisation’ of minority issues.

15. All eight Labour Muslim members have been elected from constituencies with a significant Muslim population: Rushanara Ali (Bethnal Green and Bow), Khalid Mahmood (Birmingham Perry Bar), Shabana Mahmood (Birmingham Ladywood), Sadiq Khan (Tooting), Yasmin Qureshi (Bolton South East), Mohammad Sarwar (Glasgow Central and before then Glasgow Gowan), Anas Sarwar (Glasgow Central) and Shahid Malik (Dewsbury).