The UK Party System and Party Politics Part 1: The electoral dimension Patrick Dunleavy Gv 311: British Politics course, Lecture 9 Michaelmas Term 2014-15 London School of Economics P. Dunleavy 2014
A party system is constituted by a. How voters behave b. Electoral institution effects, favouring one or more parties over others c. Party ideologies and cleavages d. Governance institutions favouring some parties e. How parties recruit elites and how they behave in office and opposition f. How much parties shape public policies
This week I cover 1. Duverger s Law, and how it completely ceased to apply in the UK 2. Current trends and characterizing the GB party system now 3. The role of disproportionality (measured as DV scores) in sustaining major party dominance at general elections 4. Tracking party system expansion with the effective number of parties (ENP)
Duverger s Law Plurality rule ( first past the post ) elections always produce/ tend to encourage the emergence of a two-party system. Formulated in 1955 at height of post-war re-growth period. Initially framed at national level, but that pattern soon collapsed Then re-framed to apply only within local electoral districts (constituencies) Supplementary Duverger s Hypothesis proportional systems encourage more parties to emerge and survive
Mechanical and psychological mechanisms in Duverger s Law Source: Dunleavy and Diwakar, 2013. Analysing multiparty competition in plurality rule elections, Party Politics, vol 19: p.855
Duverger s Law updated Andrew Cox s 1997 reformulation says for any electoral system the maximum number of parties per district N = M + 1 (where M is district magnitude). In plurality rule, M size = 1, so UK maximum number of (effective or major?) parties per constituency should = 2 Nationalization of parties is a separate issue driven by regional identities etc so Cox s theoretical max N for UK could be 646 x 2 = 1292. A pretty safe bet here!
Dickson and Scheve (partial) counter-theory Under plurality rule, a social group with 67%+ support in a constituency can split two ways, knowing they will still always beat the opposition Splitting majority vote is rational in maximizing the welfare of the majority of the majority MP closer to their view Implies we should never see a top party P 1 > 67% If opposition also splits, the majority social group may fragment further, yet still win
Total vote for all other parties in each electoral district The Crown diagram 100 90 Second national party 80 70 Double APT Top national party 60 50 40 Crown 30 67% 67% 20 Second national party Top national party wins local majority wins local majority 10 0-100 -90-80 -70-60 -50-40 -30-20 -10 0 10 20 30 40 50 60 70 80 90 100 Second national party ahead of top Top national party ahead of 2nd Top national party lead over second national party in each electoral district
Total vote for all other parties in each electoral district USA House of Representatives district-level outcomes 2006 election - USA is almost the only modern super-duvergerian two-party system 100 90 80 Double APT 70 60 50 Crown 67% 40 30 67% 20 10 0-100 -90-80 -70-60 -50-40 -30-20 -10 0 10 20 30 40 50 60 70 80 90 100 Democrats ahead of Republicans Republicans ahead of Democrats Republican lead over Democrats in each electoral district
A non-duvergerian pattern the Indian district-level outcomes in the 2004 general election. 43 parties in Lok Sabha, 18 in coalition government
Total vote share for all other parties (%) Constituency outcomes in the 1955 general election, in Great Britain - then a predominantly two-party system 100 90 80 70 Other MPs with a local majority 60 50 40 Labour MPs with a local majority 30 20 10 Conservative MPs with a local majority 0-100 -90-80 -70-60 -50-40 -30-20 -10 0 10 20 30 40 50 60 70 80 90 100 Conservatives lead over Labour in each local seat (%)
Constituency outcomes in the 2005 general election, Great Britain
Total vote share for all other parties in each seat (%) Constituency outcomes in the 2010 general election, in Great Britain 100 90 80 Other MPs with a local majority 70 60 50 67% 40 30 67% 20 Labour MPs with a local majority 10 0-100 -90-80 -70-60 -50-40 -30-20 -10 0 10 20 30 40 50 60 70 80 90 100 Conservative lead over Labour in each local seat (%) Conservative MPs with a local majority
onservative lead over Labour in each local seat (%) The patterns of constituency outcomes across four different regions in the 2010 general election South-east North-west 100 100 90 90 80 80 70 70 60 60 50 50 40 40 30 30 Labour MPs with a local majority 20 10 Conservative MPs with a local majority Labour MPs with a local majority 20 10 Conservative MPs with a local majority 0-100 -90-80 -70-60 -50-40 -30-20 -10 0 10 20 30 40 50 60 70 80 90 100 0-100 -90-80 -70-60 -50-40 -30-20 -10 0 10 20 30 40 50 60 70 80 90 100 100 100 90 90 80 80 70 70 60 60 50 50 40 40 30 30 Labour MPs with a local majority 20 10 Conservative MPs with a local majority Labour MPs with a local majority 20 10 Conservative MPs with a local majority 0-100 -90-80 -70-60 -50-40 -30-20 -10 0 10 20 30 40 50 60 70 80 90 100 0-100 -90-80 -70-60 -50-40 -30-20 -10 0 10 20 30 40 50 60 70 80 90 100 Conservative lead over Labour in each local seat (%) Scotland Wales
Comparing with other countries Election districts with a given number of parties receiving 1% or more of local votes United States, House of Representatives 2006 Indian general election 2004 Great Britain, general election 2005 One 7.8 0 0 Two 52.6 3.9 0 Three 29.0 13.8 3.5 Four 9.2 23.4 32.3 Five 0.7 26.7 41.1 Six 0.2 18.4 17.5 Seven 0.2 8.6 4.9 Eight 0.2 3.1 0.6 Nine or more 0 2.0 0 Total 100% 100% 100% No of cases 435 546 628 Rows consistent with Duverger s Law
The UK party system now 1. GB has not been a two-party system since 1974 - nor a 2.5 or 3 party system since 2000 2. In terms of voting Great Britain is a standard European multi-party system 3. Voters multi-partism has previously been artificially suppressed by plurality rule voting at general elections
Trends in the vote shares for the top two parties and for smaller parties, 1970 to 2010 100 90 80 70 60 50 40 80 79 30 20 10 50 50 62 40 46 53 48 45 35 30 0 1966 1970 1974 FEB 1974 OCT 1979 1983 1987 1992 1997 2001 2005 2010 Two party (Con plus Lab) Two party lead over Rest Liberal Democrats All other parties Note: The numbers in grey area here show the combined Conservative and Labour per cent support, minus the combined support for the Liberal Democrats and all other parties, that is the two party lead over the Rest.
Combined % Con and Lab votes The decline of the Labour and Conservatives two party system at general elections since 1950 100 90 1974 hung Parliament 1983 Labour slump: SDLP New Labour period 80 70 60 50 70 60 50 % of Seats where Lab vs Con are top 2 parties 40 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Source: Prof Ron Johnston, Bristol Election y ear 40
Party running top in 2014 Euro election votes, by local authority areas Lib Dem SNP Lab UKIP Source: House of Commons Library, European Parliament Elections 2014, Research Paper 14/32, 11 June 2014. http://www.parliament.uk/business/publica tions/research/briefing-papers/rp14-32/european-parliament-elections-2014 Con
State of the Parties since 2005 % vote Lab Con Lib UKIP Greens Rest Lab share Dem lead 2005 36 33 23 2 1 5 + 3 2010 29.5 37 23.5 3 1 6-7.5 2013 EP Polls (Nov 2014) 25.5 24 7 27.5 8 8 +1.5 33 32 8 15 5 7 + 1
The 2014 party system, in England Respect Greens Labour Lib Dems Conservative UKIP BNP left centre right
The English party system Respect Greens Labour Lib Dems Conservative UKIP BNP left centre The Scottish party system SSP Grn SNP Lab LD Con UKIP
Respect A lot has changed - the 2010 English party system Lib Dems Labour GREEN Conservative UKIP BNP left centre right - the 2010 Scottish party system Grn SNP Lab LD Con UKIP SSP left right BNP
Three key motors of party system change despite the non-reform of Westminster and local voting Class dealignment PR elections since 1997, introducing new voting systems, especially the Supplementary Vote, Additional Member System, and List PR broadening voters experience nationwide Multi-tier elections European elections, and devolved governments in Scotland, Wales and London (+ Northern Ireland) socalled cross-tier contagion effects
Per cent of each occupational class voting for main parties, general election 2010 Occupational class Conservative Labour Liberal Democrat Other parties Total Upper non-manual (AB) 39 26 29 7 100% Routine non-manual (C1) 39 28 24 9 100% Skilled manual (C2) 37 29 22 12 100% Unskilled manual/ not working (DE) 31 40 17 12 100% Source: Ipsos MORI (2010) How Britain Voted 2010. http://www.ipsos-mori.com/researchpublications/researcharchive/poll.aspx?oitemid=2613
DV (deviation from proportionality) score We calculate the differences in seats shares compared with votes shares for each party Add up all the scores ignoring + or - signs Divide by 2 to remedy doublecounting Gives DV score Note: Minimum DV score is 0% No maximum DV score unless all MPs go to a party with no votes at all, which is not a democracy Party Vote % Seats % Deviati on Con 35 45 +10 Lab 30 38 +8 Lib 20 7-13 Other 5 0-5 Total (Ignore = or -) 36 Deviation from Proportionality 18%
Deviation from proportionality scores, 1992-2012 2012 London 2011 Wales 2011 Scotland 2010 2009 Europe 2008 London 2007 Wales 2007 Scotland 2005 2004 Europe 2004 London 2003 Wales 2003 Scotland 2001 2000 London 1999 Europe 1999 Wales 1999 Scotland 1997 1992 17.4 22.7 20.7 22.4 21 0 5 10 15 20 25 Deviation from proportionality score Practicable minimum score for any voting system is around 4% Plurality election
DV score Deviation from proportionality scores, 1992-2012 12.1 2011 Wales 11.8 14.7 2010 2008 London 8.1 11.7 17.7 22.7 E 2007 Scotland 10.2 2004 Europe 13.7 14.8 20.7 E 2003 Wales 14.1 12.1 2001 1999 Europe 10.6 14.8 13.9 22.4 E 1999 Scotland 10.3 1992 17.4 21 0 5 10 15 20 25 Practicable minimum score for any voting system is around 4% Plurality election British AMS (PR) election
Local DV scores in the 2010 general election By Chris Hanretty (Univ of East Anglia) Local DV score % The darker the colour, the higher the DV score in the 30 seats around constituency X Durham South west Source: Democratic Audit blog, 5/10/2013 http://www.democraticaudit.com/?p=1496
Counting parties the ENP score We calculate the effective number of parties (ENP) by squaring the decimal vote shares, summing and dividing 1 by the sum The squaring process weights the contribution of large parties highly, and marginalizes that of small parties Here 1 divided by 0.312 = 3.21 parties Party Vote Vote sq Con.38 0.144 Lab.35 0.123 Lib.20.04 Others.07.005 Total 0.312
Effective number of party scores since 1992 2011 Wales 2010 2008 London 2007 Scotland 2004 Europe 2003 Wales 2001 1999 Europe 1999 Scotland 1992 0 1 2 3 4 5 6 7 Effective number of parties Plurality election Practicable minimum score for any voting system is around 1.5 parties
ENP Effective number of party scores, 1992-2012 2011 Wales 2010 3.5 3.5 3.6 4.4 2008 London 2007 Scotland 2004 Europe 2003 Wales 2001 2.63 3 3.4 4.57 4.59 4.65 4.2 4.7 5.8 1999 Europe 1999 Scotland 1992 3 3.3 3.06 2.21 0 1 2 3 4 5 Practicable 6 minimum 7 score for any voting system is around 1.5 parties Plurality election British AMS (PR) election
ENP Effective number of party scores, 1992-2012 2011 Wales 2010 2008 London 2007 Scotland 2004 Europe 2003 Wales 2001 1999 Europe 1999 Scotland 1992 2.21 2.63 3 3.4 3 3.3 3.06 3.5 3.5 3.6 4.4 4.57 4.59 4.65 4.2 4.7 4.3 5.1 5.8 6.3 0 1 2 3 4 5 6 7 Practicable minimum score for any voting system is Plurality election British AMS (PR) election Euro election around 1.5 parties
Total % vote share for all other parties in each local seat Outcomes of the European Parliament elections in Great Britain in 1999, 2004 and 2009, using regional list PR systems 100 90 80 70 60 Crown 50 67% 40 30 67% 20 10 0-100 -90-80 -70-60 -50-40 -30-20 -10 0 10 20 30 40 50 60 70 80 90 100 Labour ahead of Conservatives Conservatives ahead of Labour Conservative % lead over Labour in each local seat 2009 2004 1999
Thanks for listening Next week: Party System Ideology, Strategy, Governance and Policy-making aspects