Why Are Party Systems Collapsing inthe Most Developed Countries onEarth? -- Theories and EvidenceThomas Ferguson

The Stylized Facts:Most European center-left parties, in decline foryears, now collapsing. 1. France: 2017, Socialist Party presidential candidate gets 6.4%of vote after party won presidency and controlled Senate andlower house and most regions in 2012. 2. Germany: SPD vote share halved since 1998; in 2017 down5.2% to barely 20% from previous election; sinking in polls since. 3. Netherlands: Dutch Labor Party vote in 2017 down 19 pointsfrom previous election. 4. Greece: Pasok vote in 2009 was 43.9; in 2015 6.3%. 5. Spain: Socialist Party vote declines from 43.9% in 2008 to22.6% in 2016. 6. Italy: Democratic Party and allies won 37% in 2008; in 201823%. 7. Sweden: Social Democrats in 1994 won 45.2%; in 2018, 28.4.

Two Ways “Populist” Parties RiseUS and UK: Populist wings grow strongly, mostly within existingmajor national parties, though in UK UKIP was briefly influential. InItaly, populist leaders also take over a major national party.Elsewhere, most new populist forces organize as new parties.Nearly all are openly right-wing, though Five Star in Italy claims totranscend left/right divisions.Left populist parties grew episodically on the European periphery;but only two big movements exist in major countries: The UK LaborParty and the Sanders Movement in the US.

4 Broad Explanations: 1. “Cultural Backlash” – Early treatments ofPopulism traced it to value conflicts arising frommodernization; psychological shocks and fears ofthe “Other”; flatly denied economic forces muchimportance, e.g., Inglehart and Norris, 2016. Strongly argued by many for 2016 election (PoliSci consensus emphasizes race, gender in Trumpvote; flat denials of econ influencecontinue[Krugman, 2018, citing PS studies]).

2. Economic Pressures Arising FromGlobalization1. UK: Becker, S. O., Fetzer, T., & Novy, D. (2016).Who Voted for Brexit ? A Comprehensive DistrictLevel Analysis. CAGE Working Paper 305.2. US: Autor, et al. (2017) – Imports; string of otherpapers, some critical.3. Germany: Dippel et al., 2016, Südekum, 2017.4. Algan et al., 2017 – Unemployment in many EUcountries related to PopulismCf. also INET Plenary Sessions, Edinburgh, October2017.

3. Statistical Studies of Financial Crisis and GreatRecession:Fin Crises Advantage Right Wing Parties1. de Bromhead, Eichengreen, O’Rourke 2013 –Pre-WWII crises benefit right wing Parties,though extent varies with conditions -- how longthe slump continues, WWI, and pol traditions2. Funke et al., 2016 – Confirm the benefit to onlythe Right

4.Piketty: Brahmin Left and MerchantRight Old system: “In the 1950s-1960s, the vote for “left-wing”(socialist-labour-democratic) parties was associatedwith lower education and lower income voters. Thiscorresponds to what one might label a “class-based”party system: lower class voters from the differentdimensions (lower education voters, lower incomevoters, etc.) tend to vote for the same party orcoalition, while upper and middle class voters from thedifferent dimensions tend to vote for the other party orcoalition.”

New System:Since the 1970s-1980s, “left-wing” vote has graduallybecome associated with higher education voters,giving rise to what I propose to label a “multipleelite” party system in the 2000s-2010s: high-educationelites now vote for the “left”, while high income/highwealth elites still vote for the “right” (though less andless so).I.e. the “left” has become the party of the intellectualelite (Brahmin left), while the “right” can be viewedas the party of the business elite (Merchant right).I show that the same transformation happened inFrance, the US and Britain.

Problems With the Statistical Studies:The exceptions are of overwhelmingimportance:Pre-war: New Deal; Blum Gov’t in France;post-2008: Obama twice elected.Completely unexplained in the stat studies.Censored sample before WWII: elections inmany countries were tightly controlled: leftistsurges meant end of the regime, e.g., Ebertand SPD; postwar is straightforward:Left parties act Right

Problems With Piketty– Simply false to claim that Right Parties representthe business elite, while left parties the highlyeducated: he recognizes that median voteraccounts are way off, why then fixate on precisevoting totals as the explanation? The Gilens andPage result for US, now also found in Germany.– “Education” is systematically misunderstood inthe Information Age and the triumph of fiscalausterity.

2016 and the Trump Era:One Picture Worth 1000s of WordsData for Ferguson, Jorgensen, Chen, NTON0.1SANDERS0

2012: Support for Candidates Full Sampleand Big Business Only Percentage of FirmsContributingFerguson, Jorgensen, Chen anPaulPerrySantorumPawlenty% All% Big Bus234123325442N 23,59057771416181030182112N 777

Formal Campaign Money is Only A Slice ofthe Spectrum of Political MoneyFigure After Ferguson, Jorgensen, and Chen, 20171. Paymentsto Lawyersfor Services(After Stigler,See Text)Substantial,But Unknown2. Payments toPolitical FiguresMany Hundreds ofMillions of DollarsIncludes CertainDirectors Fees,Speaking Fees,Book Contracts;Some “Research”and Philanthropic“Advice” FromConsultants3. Foundationsand CharitableGrantsMany NotPolitical; SomeThat Do GoThrough ThinkTanks 296 Billion inTotal Giving in2006; Perhaps 3to 5% MightCount as BroadlyPolitical4. LobbyingLegal Definition IsVery Narrow2010 On theRecord TotalsApprox. 3.5Billion. Refers toWashington, D.C.Lobbying in Statesand Cities AlsoLarge5. Think TanksRapid GrowthEspecially Since1970sIn 2005 Major D.C.Based Think TanksSpent Approx 411MillionMany More NowOutside Washington,D.C.Not Included inEstimate6. FormalCampaignSpendingTotalExpenditures onFederalCampaigns Only 5.2 Billion in2008; State andLocal SpendingHeavy, Too7. Value ofStock Tips,IPOs ToPoliticalFigures“EventAnalysis”Studies SuggestVery Large inCertain PeriodsSee Text8. PublicRelationsSpendingSomeCertainlyAffects Politics

Linear Models of Legislative Elections: U.S. House 2012; All Such ElectionsFor Which We Have Data, Including France (!) Look Roughly Like This2012: Pseudo-R Sq .779; Bayesian Latent Spatial Instrumental Regression, Ferguson, Jorgensen,Chen 2016

House 0y0. 1980x , r 0.813-0.5921.26 x , r2y 0.813-0.49yy-0.491.24 x , r 0.8151.31 x , r2y 0.846-0.628y-0.6280.50y21.3 x , r 0.845% Dem - % GOP% Dem - % GOP0.25y21.28 1984x , r 0.819-0.61921.28 x , r 0.819199021.3 x, r 0.845198819921994-1.0221.0y-0.492 0.997x , r 0.736y-0.674 1.18 1994x , r 0.80419920.51.022y-0.492 0.997 x , r 0.736y-0.674 1.18 x , r 0.8040.00.5-0.50.0-1.0-0.519982000-1.0221.0y-0.616 1.3 x, r 0.823y-0.571 1.19 2000x , r 0.82719980.51.022y-0.616 1.3 x , r 0.823y-0.571 1.19 x , r 0.8270.00.5-0.50.0-1.0-0.520042006-1.0221.0y-0.536 1.13 2004x , r 0.804y-0.429 1.06 2006x , r 0.8140.5221.0y-0.536 1.13 x , r 0.804y-0.429 1.06 x , r 0.8140.00.5-0.50.0-1.0-0.520102012-1.0221.0y-0.548 1.01 2010x , r 0.78y-0.459 1.02 2012x , r 0.7650.5221.0y-0.548 1.01 x , r 0.78y-0.459 1.02 x , r 0.250.500.751.00-1.00.00-0.619198821.31 1986x , r 0.846-0.592y21986-0.592198421.24 x, r 0.8151982PercentTwo 0.50Party Total0.75DEM1. x, r 0.7251990-0.53221.1 x , r 0.7251996y-0.496y21.08 1996x , r 0.833-0.49621.08 x , r 0.8332002y-0.594y21.17 2002x , r 0.809-0.59421.17 x , r 0.8092008y-0.488yMoney1.00DEM Percent Two Party Total Money21.07 2008x , r 0.771-0.48821.07 x , r 0.771

Linear Model: Senate 1980 to 20141980.0100500-50-100y1982.0-15.5 0.582 x , r 0.3091984.0100500-50-100-51.8 1.05 x, r 0.833y-50.6 1.13 x, r 0.746-35.2 0.898 x , r 0.847% Dem - % GOP2y-45.4 0.942 x , r 0.7721992.01994.02y2y-33.9 0.737 x, r 0.7381998.02y-42.9 0.897 x, r 0.742100500-50-100y-66.7 1.26 x, r 0.776100500-50-100yy2002.02000.02-41.6 0.874 x, r 0.901y2004.002-47.1 0.937 x, r 0.7492006.0y2-46.3 0.945 x , r 0.802-51.6 1.06 x, r 0.76y33.2 0.353 x, r 0.6342010.02012.02-40.5 0.846 x, r 0.786y2-44.4 0.872 x , r 0.6772-54.6 0.896 x, r 0.7482550752y2014.0y2-45.6 0.856 x, r 0.74822008.0100500-50-1002-28.8 0.819 x, r 500-50-1001982.122-41.3 0.827 x, r 0.613 y100Dem Percent Two Party Total Money2

Conclusion: Globalization Generates Or Intensifies ExistingPressures toward Dual Economies1. Increases in Income Inequality; Wealth Inequality2. Pressures to Lower Tax Rates, Esp. on High Incomes3. Restructuring of Jobs, Careers, Consequent on Reorganization ofFirms.4. Permanent Fiscal Squeezes over Long Periods of Time5. Laissez Faire for Most Citizens, But State Guarantees andSupport for FTE(Temin, Storm, Lazonick, Ferguson, Jorgensen, Chen – All INETWorking Papers; David Weil on Job Fissuring

The European Union Enters the Danger Zone;When Businesses BailEuropean Union, SocialDemocrats, AmericanDemocrats –Weaker Econ Appeals,Identity Politics

So the Issue is Top Down Led or Bottom Up LedMovements for “Change”:Crucial Question is Alignments Within BusinessCommunity:US, UK, distinctive in that they both have very large FreeMarket Fundamentalist Blocs;Contrast Macron, German Situation; Also ParliamentaryCoalitions Harder to OrganizeMovements Against Globalization Have Succeeded WithStrong Support FromFree Market Fundamentalists; Stance of the Rest ofBusiness Becomes Crucial As These Movements AssumePower

Claims that economicissues did not affectvoting patterns in the2016 election are false.Ferguson, Page,Rothschild, Chang,and Chen, 2018:

Predictors of 2016 presidential election vote choice.Right TrackLimit ImportsRacial ResentmentModern SexismWhiteACA rty IDConstantObservationsPseudo-R2Issues AffectingGeneral ElectionVotingNote: DV for each column is 2016 votechoice, with 0 vote for Clinton and1 vote for Trump. Cells are logitcoefficients with standard errors inparentheses. *** p 0.01, ** p 0.05, *p 0.1

Predictors of General Election Vote ChoiceIssues AffectingGeneral ElectionVotingRight TrackLimit ImportsRacial ResentmentModern SexismWhite-4Pseudo-R2 .52-20246

Weighted Racial Resentment Means 2008 overall mean 3.46 GOP mean 3.84 Dem mean 3.16 2012 overall mean 3.52 GOP mean 3.99 Dem mean 3.11 2016 overall mean 3.19 GOP mean 3.78 Dem mean 2.63 GOP primary voter mean 3.84 Trump primary voter mean 3.99Racial ResentmentFour Question ScaleRepeated FromElection to ElectionMeans in Both Parties

Weighted Modern Sexism Means 2008 overall mean 2.54 GOP mean 2.60 Dem mean 2.36 2012 overall mean 2.48 GOP mean 2.71 Dem mean 2.26 2016 overall mean 2.34 GOP mean 2.66 Dem mean 2.03 GOP primary voter mean 2.65 Trump primary voter mean 2.74Modern Sexism Scale

Problem of Money in Politics is Problem ofMoney in Society (Think Inequality)Figure After Ferguson and Johnson, 2013

New System: Since the 1970s-1980s, "left-wing" vote has gradually become associated with higher education voters, giving rise to what I propose to label a "multiple- elite" party system in the 2000s-2010s: high-education elites now vote for the "left", while high income/high- wealth elites still vote for the "right" (though less and .