An electoral calculus? Dual incentives and committee assignment in the UK’s mixed-member legislatures

ABSTRACT

Although mixed-member electoral systems offer an apparent opportunity to observe how different rules shape politicians’ behaviour, ‘contamination’ between the SMD and PR-list tiers has frequently confounded academic work. Investigating Scotland and Wales’ mixed-member legislatures by exploiting their different chamber sizes and an unusual dual candidacy prohibition in Wales, modelling of committee assignments uncovers a split finding. Controlling for membership of the lead governing party, list members have a higher committee workload than their constituency colleagues, and members with previous employment experience in justice and health are more likely to be assigned to the corresponding subject committee once elected. Elsewhere, expectations that members might seek assignments that best suit theorized re-election interests are not found. The hypothesized influence of electoral rules is strongly conditioned by the small size of the legislature in Wales.

KEYWORDS:

• Devolution
• legislative behaviour
• territorial politics
• mixed electoral systems
• legislative committees

Disclosure statement

No potential conflict of interest was reported by the author.

ORCID

Ed Gareth Poole http://orcid.org/0000-0002-9817-5014

Notes

1 Note dual candidacy has been banned in several legislatures elected by Mixed Member Majoritarian voting rules, including Korea, Taiwan, Thailand and Ukraine.

2 Examples include Welsh Labour’s ‘constituency twinning’ to select one female and one male candidate from seats that are geographically-proximate and of approximately equal ‘winnability’ (see Mitchell and Bradbury 2004).

3 See online annex for these four cases.

4 A 21st policy area (devolved taxes) was added by the Wales Act 2014.

5 During the first two Scottish parliamentary terms (1999–2007), most committees had seven, nine or eleven members; in the Third Session most had eight members.

6 See online annex for office-seeking motivations posited elsewhere in the literature.

7 See online annex for explanation for choice of newspapers.

8 See online annex for examples.

9 Strathkelvin and Bearsden constituency, won in 2003 by Dr Jean Turner.

10 E.g. “Withybush hospital ‘downgrade’: More than 700 protest at Welsh Government plans.” Wales Online, 18 June 2014.

11 See online annex.

12 Excludes the regional committees of the National Assembly (1999–2007) to which all Assembly Members were assigned.

13 Where exact dates of membership were not available (Wales 2003–2011), members were credited if their membership was recorded in The Wales Yearbook. See online annex for additional explanation.

14 See online annex for further analysis.

15 Previous election performance is not a perfect indicator of seat safety in subsequent elections, but it is a reasonable proxy and has been adapted as an explanatory variable in the literature (e.g. Heitshusen, Young, and Wood 2005; Ferrara, Herron, and Nishikawa 2005). See online annex for discussion of alternative operationalisations of the Safe Seat variable.

16 See online annex for results from an alternative operationalisation of this variable.

17 See online annex for historical explanation.

18 See online annex for additional explanations for excluding dual candidacy as a standalone variable in the model.

19 Instead of setting all other explanatory variables to particular values (such as their sample means or modes) to calculate marginal effects for the variable of interest, this method holds each of the other explanatory variables at their observed values for each observation in the data, calculates the marginal effect for each of these observations, then takes the mean average over all of these cases.