WEALTH INEQUALITY IN POLITICAL PERSPECTIVE BEN ANSELL, UNIVERSITY OF

IBEI MAR 2018 1. WHAT IS WEALTH INEQUALITY?

2. WEALTH INEQUALITY IN THE 1ST DIMENSION - REDISTRIBUTION

3. WEALTH INEQUALITY IN THE 2ND DIMENSION - POPULISM WHAT IS WEALTH INEQUALITY?

Inequality in the STOCK of assets people have as opposed to their FLOW of income.

Can come from productive capital (investments) and non-productive capital (residential housing).

Sometimes has to be viewed as FUTURE claims on income - best example is pensions.

Much more unequally distributed than income and often in surprising ways. WEALTH vs INCOME

Standard cross-national income inequality patterns do not hold up with wealth.

Denmark is most unequal! HOUSING & POLITICS

Why might housing be especially important for politics?

Mass ownership of ‘capital’

But also massive differences in value of housing which have been amplified by housing booms and busts

Also forces us to think about age and place, as much as about incomes.

If an unproductive part of the economy matters so much for preferences what are macroeconomic consequences? 1ST DIMENSION POLITICS

Our standard hobbyhorse in Comparative Political Economy.

Generally we expect high-income people to desire lower taxes and spending than do low-income people.

Does this work the same way for wealth? And how would we distinguish this from income.

How is distribution of wealth produced and how does this vary across contexts? by 2010 before house prices had really risen again. With that caveat we can note that arestrictivefiscalpolicypost2010wasaccompaniedbymassivemonetarystimulus throughANTI-REDISTRIBUTION quantitative easing in the USA, the UK, and eventually the Eurozone and this in turn boosted house prices once more, thereby taking the edge off demands for an end to austerity, at least amongCYCLE the homeowning set. ANTI-REDISTRIBUTIONFigure 3: The Anti-Redistribution Cycle CYCLE Low demand for redistribution

Booming housing Weak supply markets of redistribution

Reliance on credit to maintain consumption Vote for Brexit in part a reflection of inability to compensate So far we have two stories. First an argument that rising income inequality producesfor rising rising wealth inequality. inequality, Those especially without in housing, expensive and especially houses where redis- tributionresented is minimal. status This quo is the effect of the supply of redistribution on credit demand. Credit expansion then further fuels the booming housing market, thereby weakening the demand for redistribution. In just those places where the supply of redistribution is low the demand for redistribution will remain low - this equilibrium being channeled through the housing market. In total, housing market booms cre- ate an ’anti-redistribution cycle’ - as seen in Figure 3 - low redistribution creates demand for credit in response to inequality, but this then produces asset booms that themselves reduce the demand for redistribution.1 Wealth inequalities also shape the coalitional dynamics of politics in surprising new ways. Wealth through homeownership is of course related closely to income - richer people are more likely to own houses and, among homeowners, to have

1There is an intriguing further connection between wealth inequality and the welfare state. Schwartz (2014) argues that countries with funded pension schemes (as opposed to PAYG systems funded from general taxation) rely in part on securitization of mortgages to provide the pool of borrowing that sustains private pension systems. In such countries wealth inequalities thus emerge from both private pensions entitlements and securitized housing assets.

8 SUPPLY OF REDISTRIBUTION → HOUSING

Ahlquist and Ansell (2017) argues that positional consumption drives borrowing.

As income inequality rises it creates demand for borrowing to close positional gaps.

Income inequality can then fuel wealth inequality. THEORY

Model of positional consumption - my consumption choices depend on my income net of taxes plus the difference between my net income and that of rich.

Latter drives up consumption as inequality widens, which pushes down savings and increases borrowing.

But… higher taxes reduce gap in net incomes, reducing positional consumption effect of rising pre-fisc inequality.

Use cumulative left-government as proxy for tax level/ 664 WORLD POLITICS

2

1

0 Long-Run (Std.) Change in Credit/GDP Germany US

(a) 7DDD 201834 244 201834 7DDD 5.0

2.5

0.0

–2.5

–5.0 Long-Run (Std.) Change in Credit/GDP –1 0 1 2 3 2001 Cumulative Left Government (Std.)

(b) 194274018344440C08:01:40 FIGURE 6 POSTERIOR PREDICTIVE DENSITY MEDIANSa

a Panel (a) shows posterior predictive density medians and 95 percent BCI for the long-run effect on credit of a change in top 1 percent income share equivalent to that observed in the United States from 1980 to 2000, for Germany and the US. Panel (b) shows posterior predictive density medians 20 for the long-run effect on credit of the same shock to inequality for all eighteen countries in our study . Triangles represent countries with posterior 95 percent BCI that do not contain 0.

quite wide, we do not plot BCI bars. Countries for which the 95 percent

/8C483 BCI for the posterior predictive long-run effect do not contain zero, are identifi ed using triangles; those for which 0 ∈95, are denoted using cir- cles. As expected, rising inequality is associated with greater credit in only those countries that had very low levels of left participation in gov- ernment since 1960. Inequality has no discernible long-run relationship with credit in most of the countries, though much of this uncertainty

is driven by the uncertainty in the error correction parameters, li . In-

7DDD 201834 24 201834 7DDD

738 . 738 ,D:0343 HOUSING → DEMAND FOR REDISTRIBUTION

Why do asset prices (especially but not only housing) affect individual preferences over public policies? Although house values and individual income are typically related they need not be. House price booms may be uncorrelated with both individual and aggregate labor market outcomes. EFFECTS ON PREFERENCES

LOWER TAXES: (a) Direct: land / property / inheritance / capital gains (b) Indirect: adopt income tax preferences of high-income LOWER SPENDING : (a) Reduced eligibility for means-tested benefits (b) Reduced willingness to pay taxes for spending (c) Reduced demand for social insurance because of ‘private insurance’ of nest egg. TYPES OF DATA

1. REGIONAL HOUSE PRICES: in many countries there is reliable contemporary data at a granular level on regional house prices. US has FHA data, UK has Land Registry. Allow apples to apples comparisons (though with limits).

2. INDIVIDUAL HOUSE PRICES: of course people don’t buy a ‘regional house’. Individual estimates tap into these important idiosyncrasies but rely on estimate’s accuracy. BHPS PANEL DATA

Panel dataset of 47,000 individuals from 1991 to 2006. Examine effects of both changes and levels in estimated housing values. DV: ∆ in (A) Support for Full Employment (B) Ideology Index IV: ∆ in self-estimate of house value in £10K TECHNIQUES: (A) Ordered Probit (B) Error Correction Model (C) Permanent / Transitory Inc. .5 .45 .4 .35 .3 Probability of Supporting Full Employment Full of Supporting Probability .25 100 150 200 250 300 House Price in £1,000s ThePoliticalEconomyofOwnership:HousingMarketsandtheWelfareState May2014

FIGURE 2. AttitudesISSP Towards Redistribution REDISTRIBUTION from the ISSP 2009 .5 .4 .3 .2 .1 Probability of Strongly Supporting Redistribution Neg. Equity Renters Low Equity Mid Equity High Equity House Equity Level Ansell, 2014 Grey: Right Wing; Black: Non Right Wing do not become substantially less supportive of redistri- cial spending as a percentage of GDP,spending on old- bution when they have high levels of equity. age pensions as a percentage of GDP, and spending To summarize, the analyses in this section of the ef- on unemployment benefits as a percentage of GDP. fects of homeownership and house prices on individual The total social spending measure includes not only social policy preferences show a powerful negative re- pensions and unemployment but also survivors’ bene- lationship that previous studies of preferences have not fits, incapacity benefits, health spending, family spend- uncovered. However, we are left with the question as ing, active labor market programs, and public housing. to whether these preferences matter substantively—do The pensions replacement rate is for a retired cou- political parties respond to the shifting views of voters? ple and the unemployment replacement rate variable measures the net replacement rate during periods of unemployment for a solo breadwinner with dependent HOUSING PRICES AND POLICY OUTCOMES family (Scruggs, 2004). These variables have the benefit AT THE MACRO LEVEL of capturing changes in policy generosity that are not simply functions of the macroeconomic climate but are In this section I move from examining how house prices direct policy changes. Finally, the social transfers mea- affect policy preferences to analyzing whether house sure from Armingeon et al. (2008) is defined as social price fluctuations alter policy outcomes. In particular, assistance grants and welfare benefits paid by general Iinvestigatewhetherthepatternofright-wingvoters government and has the best availability, dating back being most strongly affected by house prices manifests to the 1970s. itself at the national level through the actions of right- For independent variables I focus on the interactive wing parties. The findings in this section strongly sup- effect of house price appreciation and partisanship. For port these conjectures. Even controlling for changes house price appreciation I use the five-year percentage in standard macroeconomic factors including national change in real house prices (i.e., inflation adjusted), income, price inflation, unemployment, and interest taken from the Bank of International Settlements’ rates I find powerful effects of house prices on patterns house price data for 18 countries from 1970 to 2001. of government spending. In short, when house prices These housing data provide a country-specific level of are rising, right-wing governments appear to curtail house prices relative to 1970—note that this implies social spending programs further. I explore this po- that house price levels cannot be usefully compared litical effect of the housing cycle using cross-sectional cross-sectionally, though changes can, hence my use time-series data for 18 countries from 1975 to 2001. of the five-year percentage change.14 The mean of this Forthedependentvariablesinthissection,Iexamine variable is 12.8% (a compounded annual rate of around social spending policies from the OECD Social Spend- ing dataset from 1980 to 2001, as well as OECD data on social transfers as compiled by Armingeon et al. (2008) 14 Five-year changes are less volatile than one-year changes and less and data on pensions and unemployment replacement likely to cause endogeneity problems. In the supplementary material rates compiled by Scruggs (2004). The variables taken Ishowthatchangesinhousepricesarenotaffectedbychangesin from the OECD Social Spending dataset are total so- social spending or partisanship.

14 2ND DIMENSION POLITICS

The recent wave of elections suggests that economic class may be a less strong predictor of voting than before.

Many authors - e.g. Cas Mudde, Robert Ford - have argued this reflects value differences that are only loosely, or not at all, determined by the economy.

It expresses itself as values defined over ‘group’ and ‘grid’ rather than ‘greed’ (Kitschelt & Rehm).

What is connection to wealth and housing? PLACE POPULISM

House prices provide a way of measuring the ‘calculus of fortune’ - where prospered and where resentment festers. Long-lasting experiences shape world-views and outlooks on cosmopolitanism, diversity and globalization. The housing boom and bust reinforced this by solidifying residential pattern - ever harder to move. The base of populism becomes increasingly related to geography - captured by housing - as opposed to class. BREXIT BREXIT & PLACE

The Brexit vote was a bolt from the blue, even for those who were its advocates.

It exposed underlying divides across regions and demographics that cut across party lines.

David Goodhart has referred to the split as between ‘Somewheres’ and ‘Anywheres’ - highlights importance of place. DATA

I use Land Registry data on the sales price for every real estate transaction in / since 1996.

Use (logged) median prices at the Local Authority District (50,000-250,000 people) and ward (5-10,000).

Match to reported Remain vote at LAD. Also have subset of ward counts.

Use British Election Study data to match individuals by homeownership status to their LAD and Brexit intention. variable that measures the percentage change in the median (non-logged) house price in the LAD between 1996 and 2015. FigureLAD 1: Log Median PRICE House Prices and LEVEL the Remain Vote

80

60

40 % % Voting for Remain

20 £60k £160k 440k £1.2m Log Median House Price in 2015 (England And Wales)

Figures 1 and 2 give a good sense of both the distribution of the these housing variables and their relationship to the the Remain vote. In both figures, each LAD is weighted by the size of its electorate and a linear fit line is shown. In both cases, the relationship between house prices and the Remain vote is strongly positive. In LADs where median house prices were under £160,000, very few LADs voted to remain int the EU, whereas there was only one LAD with a median house price above £440,000 that voted to Leave. A similar pattern holds for house price growth, which is measured in nominal terms. The average LAD median house price in 1996 was £60,000, whereas by 2015 that had increased to £220,000, an increase of around 266%. Figure 2 shows that in LADs where prices increased by less than 200% very few districts voted remain whereas above around 350% it is almost all Remain LADs. These results are indicative of a close relationship between house prices and Brexit voting at the LAD level but how do they relate to one another and are they simply

6 FigureLAD 2: House PRICE Price Changes 1995-2016 CHANGE and the Remain Vote

80

60

% % Voting Remain 40

20 100 200 300 400 500 % Change in House Prices 1996-2015

functions of other regional factors or demographics differences across LADs? Tables 1 and 2 are linear regressions of the Remain vote, with standard errors clustered by region (England and Wales have ten such geographical regions). Table 1 begins in Model 1 by examining the bivariate effect of log median house prices (and thus mirrors Figure 1). Here we see a one point shift in the log price variable (an percent- age increase in prices by around 270%) is associated with an increase in the Remain vote share of around fourteen percent points. Note only is this a very sizeable effect but house prices also appear to explain around forty-four percent of the cross-LAD variation in the Brexit vote. One obvious explanation for the magnitude of the effect and closeness of fit is that house prices are simply picking up broader regional variation in the Brexit vote. Model 2 removes house prices and looks only at region dummies with the East region as the omitted category. Here we see the R2 drops to thirty-four percent but we do see the expected patterns - with much the highest Remain support followed by the more affluent South East and South West with the Midlands region particularly

7 pro-Brexit. The effect of house prices nonetheless appears to go beyond regional differences. Indeed, controlling for region the coefficient on log median house prices actually increases by a third. Here a log point increase in house prices is associated with atwentypercentpointhighersupportforBrexit.NotealsothattheR2 of this model that combines region and house price effects is now almost sixty percent. The changes in the regional dummies are also interesting. Unsurprisingly the ‘effect’ of London compared to the East reduces from eighteen points to six points. But we also see that regions that appeared very inherently pro-Brexit - for example the North East and - actually had higher baseline support for Remain, controlling for house prices, than did the South East and South West. Figure 3: Log House Price Levels by Region

East East Midlands London North East

80

60

40

20

North West South East South West Wales

80

60

40

20 £60k £160k 440k £1.2m £60k £160k 440k £1.2m

West Midlands Yorkshire and The

80

60

40

20 £60k £160k 440k £1.2m £60k £160k 440k £1.2m

Figure 3 demonstrates this robust pattern of house prices on the Brexit vote even controlling for region by breaking out the relationshop between these two variables across the ten regions. In every region the relationship between house prices and Remain support is positive across LADs in that region, with particularly strong

9 Figure 5: Housing in London

80 Hackney Camden HaringeyWandsworthIslington of London 70 Hammersmith and Fulham Richmond upon Thames KensingtonWestminster and Chelsea Tower Hamlets

Merton Kingston uponBarnet Thames Ealing 60 Waltham ForestBrent GreenwichEnfield Croydon Redbridge Harrow Newham 50 Bromley % % Voting Remain Sutton Hillingdon 40 BarkingBexley and

30 Havering 250 300 350 400 450 500 % House Price Change 1996-2015

In all we find that large LADs with high house prices and that experience high house price growth, limited populations in manufacturing and without a declining working class, and a high proportion of working-age population were the most likely to vote to Remain. The Leave victory was built on areas with limited housing wealth, stagnant house prices, a tradition of manufacturing that was nonetheless in decline and with high concentrations of retired people and children.

13 4AnalysisattheWardLevel

While full Brexit results are only available at the LAD level, two enterprising BBC researchers contacted a large number of wards after the referendum and have collected voting data at the ward level (around 5,000 people) for just over 1,000 of the nearly 8,000 wards in the UK. I conduct some preliminary analysis in this section on voting at the ward level, and again find the same positive relationship between house prices and Remain support,WARD even netting out average LEVEL LAD-level support. Table 3: Ward and LAD Price Levels: Remain Support

(1) (2) (3) (4)

Log Median Price (Ward) 15.72 15.72 9.698 10.80 (0.577) (1.726) (1.734) (1.551)

Log Median Price (LAD) 7.823 (2.762) Observations 1109 1109 1109 1109 LADsTable 4: Ward and LAD House Cluster Price SEs Levels Cluster and Changes SEs Fixed Effects Standard errors in parentheses (1) (2) (3) (4) Figure 6 demonstrates the overall correlation between the log of median house WardPriceChange1995-2016 0.091 0.049 0.022 0.012 prices at the ward level (note this(0.007) data is not the (0.009) LAD median) (0.007) and the (0.006) percentage voting Remain in that ward. Table 3 shows that the estimated effect of a log point Log Median Price (Ward) 9.196 8.049 9.846 increase in house prices is extraordinarily similar(1.786) to that found (1.736) at the (1.656) LAD level - around fifteen percent points. The effect becomes smaller but not dramatically so, Log Median Price (LAD) -5.080 even when we control for the LAD log median price. Indeed, here we see that a (4.649) log point shift in prices in both the ward and the LAD (i.e. all wards in the LAD increasingLAD Price by Change the same 1995-2016 amount) can be broken down into two 0.082 effects - a 9.6 percent (0.032) pointObservations increase in Remain support from 1109 the ward price 1109 increase 1109 and a 7.8 percent 1109 point shiftLADs from the LAD price increase. Cluster If anything SEs Cluster then, variationSEs Cluster within SEs LADs Fixed is FXslightly moreStandard important errors in parentheses than that across LADs, demonstrating that the local geography of Brexit is quite finely variegated.

14

17 Figure 7: Prices at the Ward Level and Remain Support: Conditional on LAD dummies

40

Moseley and Ordsall Bastwell BelfairsNelsonChiswick RiversideTurnhamChiswick Green Homefields 20 LadywoodGreenwich West Milkstone and DeeplishSparkbrook AstonShearSelly BrowRadford Oak and ParkPaddock ManninghamWernethCentralCrystal PalaceToller DrakeEdgbastonHeatonsPark Calder North BradfordLozells Moor andHitchin EastSpringfield HandsworthBeartonCentralCorporationHeatonDunkirk Southsea and LentonBlackheathMarket WestcombeSt Matthew'sSouthfield CentralCity St Mary'sRochdaleWashwoodAudley HeathPengeArboretumPrestonStCathedral Jude and ParkCatorHeatonsEmpressSherwoodEalingDundonaldTownKnellwood BroadwaySouth PeninsulaCloseShipley St AlphegeWollaton WharfedaleWest ColdhurstBordesleySouthClarence NorwoodStPalfrey Green PetersWollastonClockHoeWarfield HouseUpper Street andPedmore HarvestStourbridge NorwoodTrinityWalpole and Ride CleadonTownHitchinWimbledonUptonWingfieldDorridgeNortonSaddleworth and East andEast Park HockleyBoldon South Heath NorthumberlandPark QueensMilton Park ParkTottenhamNechellsSt. Peter'sBerridge HanoverHaleRochesterEastney MapperleyHistonandHighCrowthorneWilliam andWakefieldNorthBattenhallBowesUpton StreetWest Compton andand Elm MorrisLeigh StCraneswaterLaineImpington HillsideAbbeyGrove Thomas SouthNorthfieldBoothstownPeverellHangerBramhallPennBeardwoodEccles Hill and South EllenbrookWorsley withand WoodfordLammackHeswall KeighleyStDorkingLittle Peter'sThorntonCopersChalkwellThorpeKingsway Queen's HarwoodHarpurCentralBromley SouthRomford Cope HitchinHeathHamletGoldsmidWollaton Park TownBridgeGraiseleyBoston DeTownMarkhouseHighburyCastleClevelandWood CheadleParysPalmersMarshalswick BournvilleEastFleet CastleWest BlakenhallStreetHalesowenand West EalingSilhillGreenand KirbyLentonNorburyMarple SouthCommon Highgateand South AbbeyNorth WoldsThurstastonKemsing IrwellPleck RiversideSohoCastleSuttonOltonRochester LeytonstoneandGrovePertonHagleyCroftLittle MountCanonsActon Green BakerHandsworthEast SandhurstEastHydePetersfield GouldSouth RainhamCentralMossleyMertonBickley WernethCheadleBlythe Acton andBramhallSouthgate Wood CentralNewnhamUpminsterPark PattinghamWellington Hulme NorthNewbarnsBamford Green Regis SouthStreetlyWestRaynes andWinchmore PatshullKnowleLeigh Park Hill Little StHorton MillfieldPeter'sPenkridgeWoolwichCentralHarrowSt WensleyPeterColliers Highbury NorthCommonBruceTodmordenBrooklands on BarnfieldandStokeAshleyBaylis WoodsideWoodWrenthorpeEast theLeatherheadStEast FoldWakefield Squirrel'sGrovethe Mark'sWest Hill MiltonActonCleeve andNewnham HayleyandWaterfrontRiverParkStroud StokeBrentfordFreelandActonSouthgateLetchworthWarndon HounslowClaines Heath North WestDetlingOldCranhamGreen HawksHeadstoneAldridgeandSouthSkircoatWestCannockWarndonPinnerShirley AmblesideGreen TrussellCoulsdonHarlow ParksideOutwood andHall Emersonandand GreenWakefieldChestertonParish Alexandra SouthHampstead CentralHatchEastSharnbrook SouthHanborough Green Thurnham CradleyWest NorthParishMuswell HarpendenAshteadWestand North West EndWestParkHighburySaddleworthKimpton and GrasmereRoosecote Rural SouthSouth Town HillKinver SouthVillageHawcoat West EastNorth North Turton with Tockholes Heath TownWestFantWombourne GreatMoretonParkCantelowesTollingtonPlaistowHitchinChippingWoolwich HortonBroadWakefieldWheatonKentish HarmansLeaWestWest Stanley NorthFortune andWalsworthIsleworthStalybridgeSewell HorsleyHookArboretum BridgeGreen Queen'sTottenham Nortonand KentonDaltonKempstonSundridge HampsteadTownRiversideShortlands Churchdownand WallaseyWaterAston,RomseyBoxley East SaughallGreen ClaremontHill HorburyWashington LowerOutwoodWatlingSouth EastKelseyParkSouthHempsteadElthorneSelsdon VerulamBarnesBishopswood RuralGreenSyonCharlton FleetOakleyHerneMassiePurley LuddendenfootPenn BearstedSouthand Sitwell andTettenhallBrookfield ChislehurstandEastClaphamBromham SouthEden Hill Southand FulwellBallardsStMeriden Haverstoeand WigmoreMichael'sParkEaton WightwickOssett Common Lapleyand BiddenhamFrognalArbury and Fitzjohns BoughtonNethertonPerton MonchelseaStreathamFinsburyHeadstone WestLakesideStHillriseColeford ManorThomas'sWestboroughWestonEttingshall ThorntonHighgate HalesowenSt Parkand BrixtonRochesterAddiscombeHill HestonDalesBaldockStGeorge's Codsall MildmaySouthHeathPark BridgeChartPeter's ChapelAnstonCentral NetteswellCoptTulseAshtonandBelleSuttonDaltonLongdendaleSteppingBeaconCove StHill East East NorthPontefractSandonAndrew'sNorth SuttonMancroftSt OxtonWestSouthHillSwissValeParish End Wath andPensbyEasthamValenceStephen andStNorth HarpendenKingswinfordSt andGrange Michael'sHill ForestTrumpington WoodsettsHarrowCottage andSouthwoodPeter'sBelsize StanmoreAscot Bents andHeadcornHellaby SouthWales andHorstedWithdeanSouthWashington ThingwallMerryNorden LangleyCockfosters BaxendenWalthamNorthRyburn SouthScartho ParkMiddleton HillBaildonTrysull East and Seisdon Rotherham East HendonChathamRotherhamGillinghamHeathCheslyn CentralLyeChalveyMilton-under-WychwoodKendalBilbrook CentralWestHayesandGraveney StCrookhamHayEast WaddonHednesfordSouth Ackworth, Leonard'sStourbridgeWesternFairfieldHaroldNorthJunction Leyton OxenholmeCheadleSouthWashingtonEastKendalHighamHyde ShootersPenkridgeSandhurstHednesfordAudenshawLintonSimonsideColdharbourPenkridgeStaplehurstStHebburnBrickhillSouthHighamCoxheath OdihamKingswinfordCannandAndrew's WoodBoldonRowhillRaynersWestCleeveGodleyLarkhall BebingtonShirleyHighgateSpringsNorth Camden WalkdenHulmeGreenHarringayWestWimbleburyHipperholmeKnebworth Holderness FerndaleNorthSwaveseyLooseHill HylandsHallQueen South HillWestBreamMardenStandRoeandNorthThorntonCentral OssettWestoeBush CollierySouth St andLeenSouthBarnesLaneElmsallJohn's ChaddertonHeath RawnsleyWest North Michael'sEwshotNatlandRekendyke PettsLee Town NorthEast SouthDrayton HuntonEdith'sHillClaphamHartley ValleyPlymstock and and Wickham and Wood ParkandwithHarpenden Yaldingand Coven St Upton andWintneyLightcliffe TownPrimrose North SuttonMary'sWalland Radford Farlington KnollHarroldHeath VeseySouth Hill 0 TewkesburyHounslowHestonThamesmeadHighStreathamBushbury West StreetHounslowDownswoodPrior'sGreenhillClayton-le-MoorsUpper King'sRoystonCentralUniversityWeasteSopwell Haselbury SouthMooringsCannockWells ParkAshtonCrossShirleyEdmontonCippenhamFarnham MedlockRipleyHettonHeath PlymptonColeridgeFleetRuytonHeathHebburnWashington St ThurlowandHeanorLangley StStroodandHarpendenDinningtonandWalney SwindonBatchwood WestAnn'sRipleyShineyWaterloo HestonPerton Whitburn James'sWakefield CentralNoel SouthandCrookham SeedleySpotlandBrandwood StMarshalswickLow OthamShenleyValeandHooberAllingtonWestFailsworth AmblecotePlympton ClementPutnoe HobbayneChaddlewood SouthNorthMarehayVassallMeadowsWestKedermister ParkRow RotherCodicoteSouth HillDippons OvertonBaschurchSilksworthSaddleworthVillageGreetlandCentralAshtonWickersley DentonHealdWestPlymptonWaldersladePheasey East LeedsandBookhamWestandDoxfordRiseley LangleyEastPettitsHumberston Oval Vale ErleCadwellMarsdenEast GreenFalingeNorthHurstKersal WestCuxton Healey ParkPinner andSt ThorpeBookham StNorthWest Mary BedwardineStainlandStFarmMary'sBingley SouthandChad's and Halling South andLeesNew WalthamSutton Four Oaks RockBentleyEdgeley FerrySmallbridgeDavenportHindpoolAiredaleEdmonton andandLeasoweLiscardRoxbourneMonktonBloomsbury DarlastonCradleyCheadleEffordandHollowayLetchworthPonders and HounslowandYateleyCrown GreenChurchFerry Stockwell OwlsmoorGlyndonRoystonBromboroughandKnight'sEtchingOrmsgillFirgroveWombourneRoxeth BrighouseHaymillKendalBromleyCaleand WoodsideStCrohamCrofton, and Heath PrincesEndWashington Moreton NorthElmAnne's OldWoodWestBrinsworthFrystonLipson Green WollescotePontefractWestAltofts East Rainham HillParks HillandMeridian ParkAldridgeWardle BracknellOxleyBarmingLittleboroughCommonMarple RyhillParkCrouchSwintonand CollegeShawLynch WordsleyIstead EastSouthandRastrickPelsallKenley KeppelThe HayesPlymstock NorthandNorthCleeve andandSedgley NorthTownSouthBalsham NorthWhitwoodRyhope BroughtonCleadonStrood HillSunnyhurst RiseGreatSilverwood Wootton EndGreasby, and TownHeathEastWorth CatcliffeWaltonWest andLakesideSanderstead RishtonGrangeKeston Belmontand BarfordRural Harton Dunstone LittleboroughConey ParkValley CentralWalsallAlthamClatterbridge HallWinkfield Woodand Irby and Cranbourne West MarshBowling Biddickand BarkerendKempston andBrierleyHeneageWhiteleasRedhillSandridgeGillinghamRawmarshGospelReddish ElstowLangworthyAllKidbrookePallionLowerBensham VictoriaSaintsGipsyBullbrook HollingdeanEastcottsNorthHillCentralSouthburySowerbyFrattonCastleLordswoodHednesford ElmdonSouthwickandWarleyOakSwintonHanworth DroylsdenSouth YarboroughEdmonton NorwoodHaverstockBreretonArburyHillNorthStaple HodgeRushall-ShelfieldStewartby ChelsfieldBrockworthwith ManorCodnorandRainhamPrince'sSwintonClaughtonSawstonStDukinfieldHazelElland WheathampsteadBridge MilnrowHarrietshamFarnborough North StalybridgeDukinfield John TyeandPrimroseHornfairEastPriory ImminghamNortonKempston HillMilnshawandEggbuckland NorthEastHacton andCippenham GreenHydeCanonbury GroveStanmerand Wexham Houghton Kenton andSouthQuintonSouthCapstoneNorthowram Ashburton RavenhillPrittlewelland Blenheim HighlandsStalybridgeFreshney Waingroves CanesNewtonFortisPratts North West andNewhey andWestStBarnsbury Shirley LeaValleyGreen Green Peter'sBottomLenhamCrofton Park WybostonandHuncoatSt Shelf SuttonOswald'sErmine Trinity BirkenheadSpring Hill andWaterheadCheslynBalderstone TranmereHuntingtonEastLongthorntonWombourneWhiteHemsworthLuton MarshSidney WillenhallHayAlexandraEnfieldCastlefordBaldockKilburn RainhamandHartCunninghamEastFeatherstone CastleBoltonNorth andStocklandDevonportand Sussex SouthLyndonWayfield BartonBar Lock ChestertonLane StreathamSouthHaleMiddle RisedaleKirkholt ChesfieldTownSouthHathertonDroylsdenWestGooshays East DentonGoldingtonHill andCentralFellgate AcocksYardleyPerryMaltbyHitchwood,NewSouthway HarrowNormantonSevenEnd MiddletonGreen MawneysWestWensumWennington SaredonGreenUndercliffeCastletonBushParkBickenhillHolbornNorth Lydney FailsworthandBarrBrightonFernhill NorthSouthRoyton Copnor Willenhalland andGreenWestSisters andFair NorthWestMoorWeald WalneyRedbournHighamsCosham Town ClerkenwellGlasshoughtonHedworthKeighley Offa WilshamsteadSutcliffeEastandHilseaEast South ViewHeath HeywoodLiveseyWest NorthHopwoodCoventandNorth CromptonPark East Hoo with ImmanuelRural Garden HallPleasingtonNorthSutton Downs New Hall Netherton,Lavender WoodsideCauldwellBlakenallMarlborough FieldsWingfieldLetchworthRoystonChaddertonPeelCannockSelhurstColnbrookKendal andWildridingsKempstonCaledonianClaytonClaphamBirchillsSandhillChaseEnfieldHanworthSouthPlumstead TewkesburyBredbury StPalaceHitchin DentonFar Yateley AndrewsHaveringSouthMileNorth CannockWintonHeathfieldWilburySpring Edgware Ham St.Bedeand HornchurchCrossSouthwithLeamoreJubileeHighway Cranford SouthWest Luke'sCross andShort Oughton Fairweather EastSouthGreen EastThorntonValePoyleFallings Cannon Newtown Park CentralEastShoebury West BaffinsHeathRuralBunhill andEltham Park HeywoodWhitehallRegent'sHilland RoytonSouthchurchRomileyGreenIdle AllertonNorthHimley and NorthPark and Swindon SeacombeMillMottingham HillLetchworthLangleyLeatherheadCentralBidstonKursaalFeatherstoneGreatDarlastonWellingtonSudellSt MillUpper Pancras andand WalkdenWyrley EastTownSouthandHanworthCharles NorthGreat ChislehurstSt LadyGornalSouth EwoodDormersAldercarPark NorthBloxwich Greenford James andandEast PendleburyBushburyLandywoodWestPriestwood WyrleyWeoley BudsheadMargaretNorthDickens SomersStroodShoeburyness ShareshillAbbeyColney ParkandMiddletonBasfordCadishead MiddletonWells NorthWestEastWednesfieldAldershot Woodsetton RottingdeanNunneryPerivaleTown ErdingtonBrownhillsGreenTwydallOrpington North South Town Heath BrightonandEastwoodIrlamGornalPrentonSt EssingtonGarthPeninsula StephenPark SouthCoastalBishop's ParkCoulsdon FernhurstEast ShadsworthSwanleySouthallBrockmoor withFigge'sStGreenSouthQuarryPollards MoulsecoombMary'sManorCoseleyKnottingley KingsbrookWhitebirkGreatCastleWealdstone MarshTurkeyElmsallShepway St Bank Hilland Heaton BudeauxBritwellBilston EastHollandsBloxwichBromwichParkCherrywood Lakenham Abbey PensnettandStreetandBillesley andSouth Queensbury Wood North PatchamDudleySouthStandBowthorpe Wood South BevendeanEastHornseyLaurenceWindhillClifton NorthboroughSt Kirkby Wood James' North and Smith'sBarrow Wood IslandGreenfordReddishHolmwoodsLetchworth NorthBredburyOvendenNortholt BroadwayHollinwoodSouthallTongEarcroftHonicknowleElthamRainbowCattonNelson and Mandeville GrangeBroadwayKingsShepwayIllingworth NorthSouth Woodley Grove Hill ParkNortonElthamOffertonBulwell Greenford StreetNorth Meadowheadand West ForestMarsh Mixenden House Cricket GreenLondonLittleBlackwater ColneyStechford HultonEccleshill and andPaulsgroveWibsey CartertonHawley YardleyBilstonNorthfieldHangletonSt SouthNorthEast Ann'sQueensbury and Knoll NortholtCray WestCrayRavensbury Valley ValleyEndStCherry BigginEastHelier WestRoydsBounds HatchHillWednesfieldHintonBestwoodKeighleyValleyColdharbour OsterleyGreenLane West North and and Spring DarwinNew Grove Eltham KingshurstBrinnington and FordbridgeChelmsley and Central WoodAspleyBedfontFelthamLowerFieldway BilboroughWestLongbridgeChingford MordenLarkswoodWykeCrome GreenEndlebury WarndonFelthamCliftonBartley NorthTyburn South GreenSheldon King'sGorse HedgesBulwellNew Hill AddingtonHigher Croft -20 KingstandingOscott % % Voting Remain (conditional dummies) on LAD -40 -1 -.5 0 .5 1 Log Median Ward House Price (conditional on LAD dummies)

adds LAD dummies and here we see that ward levels remain important and ward changes becomes less substantively large but still statistically significant. Overall, there is ample reason to believe that both house price levels and changes at the ward level mattered substantially for the Brexit vote - even at this very disaggregated level of analysis.

16 HOUNSLOW

Chiswick Riverside TurnhamChiswick Green Homefields 70.00

60.00

BrentfordHounslow South Isleworth Syon Heston East HestonHounslow West CentralHounslow Heath Heston Central 50.00 Hounslow West

HanworthCranford Hanworth Park Percent Voting Remain 40.00 Osterley and Spring Grove BedfontFeltham West Feltham North 30.00 £165,000 £265,000 £445,000 £730,000 Log Median House Price (Ward) BRITISH ELECTION STUDY

I examine the panel completed just before the Referendum, which asked voting intention. Accurate sample - 51% support for Leave.

Data on housing tenure. Interact home-ownership with (log) LAD house price level.

Examine both Remain support and immigration attitudes. Figure 8: British Election Study: House Prices, Ownership and Supporting Remain

.7

.6

.5

.4 Probability of Supporting Remain Probability of Supporting

Non-owners Homeowners .3 £60K £160K £440K £1.2M Log Median House Price

cultural life. It is very apparent from Model 1 - a linear random effects model - that these variables have an extremely strong effect on Brexit vote intention. Moreover, the apparent effect of homeownership, conditional on house prices, vanishes. What is going on? Models 2 through 5 help to answer this question. In each case, the immigration attitudes measure - economic or cultural - is the dependent variable. Models 2 and 4 use LAD random effects and Models 3 and 5 use LAD fixed effects. In all four cases we see the result from Table 5 - a positive effect of LAD house prize for both non-homeowners and homeowners that is particularly strong for the latter group. Figure 9 demonstrates this effect graphically using the estimates from Model 4, examining views about immigration’s cultural impact. How should we interpret these findings? The high correlation between immi- gration attitudes and Brexit vote intention may reflect the fact that both essentially represent the same underlying set of values. Are these values prior to homeownership and choice of residential area? This is quite possible, although evidence of sorting by values in the UK suggests this is rather rare (Kaufmann and Harris, 2015). More

21 BES

In sum, living in high house price areas correlated with Remain support, especially for homeowners.

Remain support highly correlated with immigration attitudes. Place seems to affect both.

Remain base: young, high-income, female, in high- house price areas. Pro-migration.

Leave base: old, lower-income, male, low house- price areas. Anti-migration. TRUMP VOTE

Can see similar patterns using CCES 2016 data which codes individuals by zip-code.

Match 5 digit zip-code to Zillow ‘Zestimates’ of single family house prices in 2016 along with annual change since 2001.

Can see effect of ‘place’ at both individual level and at more aggregated Congressional District level.

But NO similar ownership ‘interaction’ effect to BES. DISTRICTS

20 IL17 OH06 IL12 MN08 IL15 NY21VA09MA01 MN07 OH13 OK02MO08 GA02FL11 MI10 10 MI05TX28MN01FL06 WI07 TX34 TX30 CA44 TX04FL12OH07 NY22 CA01 IL16NJ02FL17WI03AZ04OH04 VA03GA12GA08CT03TX01IL01NE03MI04TX27IN06 CA13 NY11 NY23 CA08 NY19CA10 TX15GA14AR01CT02KY01NJ03WI08FL13TX18AL04NY02NC11 NY24 IL02CA09OH09VA06VA05SC06OR04WA03TX13SC07KS02IN08UT02FL24TX14GA01MA09ID01 NY01 NY27 TX33CA03IL13FL20TX19CO03MD02WA10TX09MD01CA43CA05TN01KY02NJ06CA20MI07NY05CA34 CA37 NY08 NY26 MI13CA16 NJ01 CA41MA02AZ07CA40TX36WA06AZ01LA05SC05CA29NC10TX05FL08MO03OH05AR04IN01TN04TN06FL18CA02 CA06 WA05IA04CA35NJ10TX35 TX11SC03CO07MD05RI02MI01NJ09NC03MN06NC07 CA12 FL05 AL07FL02 IL18LA04HI02VA04FL15KS01NH02NC08CO04KS04VA02IA02MO04MD04MS01TN03AL03IA01NM03 CA21 AZ03MS02CA31NV01KY05CT01CO05NY15NC05OR02CA42NV04LA02OK03WI06TN07LA03GA10NJ08GA09TX06 FL21CA28 CA23 NY20 NM02CA04MD07GA04TX23CA32MO06GA03AL02IN04TX17VA01NE01AZ08NC13IN09GA05MA07CA19 CA17 0 WA04NC01CA36TX20GA13WV02 UT04AL01CT05WYALAKALNDALMTALNC06DEALME02IN02UT01WA02OK04KY04TX12WV01OR03CA15FL16IL07NY13NJ04 NY14 CA14 CA22TN09CA51FL03AZ02MO05NY18MO07CA38MI12AL05MI09WA08NV02 CA07 FL09 OR05IN03WI01MS04TX25IL03 OH15 NY09NY06 NY07 MO01FL14TN08IN07FL01MD06NJ12RI01NM01MI06IL04TX08MS03NH01NV03CA47CO02HI01NY04FL19 NC09 NY25 NC12TX16 CA25MA03VA07TX31CA53CA50OH08MA06CA24OH14LA06OH16CA27CA30FL22 FL23 TX29 AR02SC04CO06CO01MA04AZ05CA11 OH11 MI14FL10 MI02KY06MA08WA01WI02IL08TX10TX26 WI04 SC02OH10NC02ID02OK01UT03IL11OR01TN02CA26MN02AZ09MD03LA01MI08 TX21FL26WA07AZ06 OH03 CA46AR03OH02NJ05MA05WA09 CA18 KY03OK05MI03IA03IL14CA39MD08TN05 FL04FL07NE02 GA11 CA49SC01CA48 MN04NC04 FL25KS03NJ11NY16VA10 -10 MN05OH01 OH12CA52TX24MI11TX03NY03 AL06NY17 IL05VA11 TX32 CA33 NJ07WI05TX22CA45 GA07TX02IL10 VA08IL09 IN05 FL27MO02 Trump margin minus Romney margin MN03 IL06CT04 NY10 TX07 GA06 NY12

(controlling for Romney margin and state dummies) -20 -1.5 -1 -.5 0 .5 1 Log House Prices 2016 (controlling for Romney margin and state dummies)

Figure 2: Effect of Log House Prices on Trump Margin w. State Dummies

16 VOTE CHOICE

House prices: both levels (logged $2016 average by 5 digit zip) and changes (5 yr or 15yr) have negative correlation with Trump support.

Little sign of interaction effect with homeownership across specifications.

Robust to controlling for state / FIPS dummies, employment status, gender, age, race, income, education, and even Party ID. Table 3: 2016 Average House Price by 5 Digit Zip and Trump Support (1) (2) (3) (4) (5) (6) All All Metros Rural State FIPS House Price 2016 -0.41⇤⇤⇤ -0.43⇤⇤⇤ -0.39⇤⇤⇤ -0.83⇤⇤⇤ -0.32⇤⇤⇤ -0.17⇤⇤⇤ (0.05) (0.07) (0.07) (0.20) (0.09) (0.06)

Homeowner 0.36⇤⇤⇤ 0.34⇤⇤⇤ 0.57⇤⇤ 0.35⇤⇤⇤ 0.33⇤⇤⇤ (0.08) (0.08) (0.24) (0.08) (0.05)

Female -0.20⇤⇤⇤ -0.26⇤⇤⇤ 0.42⇤⇤ -0.21⇤⇤⇤ -0.24⇤⇤⇤ (0.06) (0.06) (0.19) (0.06) (0.04)

Year of Birth -0.02 ⇤⇤⇤ -0.02⇤⇤⇤ -0.03⇤⇤⇤ -0.02⇤⇤⇤ -0.02⇤⇤⇤ (0.00) (0.00) (0.01) (0.00) (0.00)

Education -0.25⇤⇤⇤ -0.26⇤⇤⇤ -0.18⇤⇤⇤ -0.25⇤⇤⇤ -0.27⇤⇤⇤ (0.02) (0.02) (0.06) (0.02) (0.02) Table 3: 2016 Average House Price by 5 Digit Zip and Trump Support Party ID 1.03⇤⇤⇤ 1.03⇤⇤⇤ 1.06⇤⇤⇤ 1.04⇤⇤⇤ 1.02⇤⇤⇤ (1)(0.02) (2) (0.02) (3) (0.06) (4) (0.02) (5) (0.01) (6) All All Metros Rural State FIPS HouseholdHouse Price Income 2016 -0.41⇤⇤⇤ -0.43 -0.01⇤⇤⇤ -0.39 -0.01⇤⇤⇤ -0.83 -0.02⇤⇤⇤ -0.32 -0.01⇤⇤⇤⇤⇤ -0.17-0.00⇤⇤⇤ (0.05)(0.01) (0.07) (0.01) (0.07) (0.03) (0.20) (0.01) (0.09) (0.01) (0.06) Observations 27779 24302 22454 1824 24302 23694 Homeowner 0.36⇤⇤⇤ 0.34⇤⇤⇤ 0.57⇤⇤ 0.35⇤⇤⇤ 0.33⇤⇤⇤ (0.08) (0.08) (0.24) (0.08) (0.05)

Female -0.20⇤⇤⇤ -0.26⇤⇤⇤ 0.42⇤⇤ -0.21⇤⇤⇤ -0.24⇤⇤⇤ (0.06) (0.06) (0.19) (0.06) (0.04) Table 4: 15 Year House Price Change by 5 Digit Zip and Trump Support Year of Birth(1) -0.02 (2)⇤⇤⇤ -0.02 (3)⇤⇤⇤ -0.03 (4)⇤⇤⇤ -0.02 (5)⇤⇤⇤ -0.02 (6)⇤⇤⇤ All(0.00) All Metros(0.00) (0.01) Rural (0.00) State (0.00) FIPS 15 Yr House Pr Ch. -0.49⇤⇤⇤ -0.44⇤⇤⇤ -0.39⇤⇤⇤ -0.90⇤⇤⇤ -0.28⇤ -0.22 Education(0.11) -0.25 (0.12)⇤⇤⇤ -0.26 (0.13)⇤⇤⇤ -0.18 (0.26)⇤⇤⇤ -0.25 (0.14)⇤⇤⇤ -0.27 (0.14)⇤⇤⇤ (0.02) (0.02) (0.06) (0.02) (0.02) Homeowner 0.38⇤⇤⇤ 0.36⇤⇤⇤ 0.47⇤⇤ 0.34⇤⇤⇤ 0.33⇤⇤⇤ Party ID 1.03(0.08)⇤⇤⇤ 1.03 (0.08)⇤⇤⇤ 1.06 (0.22)⇤⇤⇤ 1.04 (0.08)⇤⇤⇤ 1.02 (0.05)⇤⇤⇤ (0.02) (0.02) (0.06) (0.02) (0.01) Female -0.26⇤⇤⇤ -0.30⇤⇤⇤ 0.30 -0.26⇤⇤⇤ -0.27⇤⇤⇤ Household Income -0.01(0.06) -0.01(0.06) -0.02(0.21) -0.01 (0.06)⇤⇤ (0.05)-0.00 Table 6: Housing by 5(0.01) Digit Zip (0.01) and Trump (0.03) versus (0.01) Romney (0.01) YearObservations of Birth 27779 -0.02 24302⇤⇤⇤ -0.02 22454⇤⇤⇤ -0.03 1824⇤⇤⇤ -0.02 24302⇤⇤⇤ -0.02 23694⇤⇤⇤ (1)(0.00) (2) (0.00) (3) (0.01) (4) (0.00) (5) (0.00) (6) Trump Trump T-R T-R T-R T-R EducationRomney Vote 1.71⇤⇤⇤ -0.271.76⇤⇤⇤⇤⇤⇤ -0.28⇤⇤⇤ -0.26⇤⇤⇤ -0.27⇤⇤⇤ -0.27⇤⇤⇤ (0.08)(0.02) (0.07) (0.02) (0.07) (0.02) (0.02)

TablePartyHouse ID 4: Price 15 2016 Year House -0.43 Price⇤⇤⇤ 1.04-0.34 Change⇤⇤⇤⇤⇤⇤ 1.03 by-0.22 5⇤⇤⇤⇤⇤⇤ Digit1.18-0.20 Zip⇤⇤⇤⇤⇤⇤ and1.05 Trump⇤⇤⇤ Support1.03⇤⇤⇤ (0.07)(1)(0.02) (0.09) (2) (0.02) (0.04) (3) (0.04) (0.05) (4) (0.02) (5) (0.01) (6) All All Metros Rural State FIPS Household1515 Yr Yr House House Income Pr Pr Ch. Ch. -0.49⇤⇤⇤ -0.03-0.44⇤⇤⇤⇤⇤⇤ -0.02-0.39⇤⇤⇤⇤⇤⇤ -0.90-0.05⇤⇤⇤ -0.02-0.28 -0.13⇤⇤⇤⇤⇤ -0.01-0.22-0.09 (0.11)(0.01) (0.12) (0.01) (0.13) (0.03) (0.26) (0.01) (0.14)(0.08) (0.01) (0.14) (0.12) Observations 25298 22112 20705 1383 22112 21649 HomeownerHomeowner 0.31⇤⇤⇤ 0.380.30⇤⇤⇤⇤⇤⇤ 0.360.07⇤⇤⇤ 0.47 0.06⇤⇤ 0.34 0.09⇤⇤⇤ 0.33 0.07⇤⇤⇤ (0.08)(0.08) (0.07) (0.08) (0.06) (0.22) (0.06) (0.08) (0.06) (0.05) (0.06)

FemaleFemale -0.16⇤⇤⇤ -0.26-0.17⇤⇤⇤⇤⇤⇤ -0.300.02⇤⇤⇤ 0.30 0.02 -0.26 -0.00⇤⇤⇤ -0.27 -0.00⇤⇤⇤ (0.06)(0.06) (0.06)19 (0.06) (0.04) (0.21) (0.04) (0.06) (0.05) (0.05) (0.05)

YearYear of of Birth Birth -0.01 ⇤⇤⇤ -0.02-0.01⇤⇤⇤⇤⇤⇤ -0.020.01⇤⇤⇤⇤⇤⇤ -0.030.02⇤⇤⇤⇤⇤⇤ -0.020.02⇤⇤⇤ -0.020.02⇤⇤⇤ (0.00)(0.00) (0.00) (0.00) (0.00) (0.01) (0.00) (0.00) (0.00) (0.00) (0.00)

⇤⇤⇤ ⇤⇤⇤ ⇤⇤⇤ ⇤⇤⇤ ⇤⇤⇤ ⇤⇤⇤ EducationEducation -0.29 -0.27-0.29⇤⇤⇤ -0.28-0.22⇤⇤⇤ -0.26-0.22⇤⇤⇤ -0.27-0.24⇤⇤⇤ -0.27-0.23⇤⇤⇤ (0.02)(0.02) (0.02) (0.02) (0.02) (0.07) (0.02) (0.02) (0.02) (0.02) (0.02)

Party ID 0.81⇤⇤⇤ 0.81⇤⇤⇤ 0.07⇤⇤⇤ 0.07⇤⇤⇤ 0.08⇤⇤⇤ 0.08⇤⇤⇤ Party ID 1.04⇤⇤⇤ 1.03⇤⇤⇤ 1.18⇤⇤⇤ 1.05⇤⇤⇤ 1.03⇤⇤⇤ (0.02)(0.02) (0.02) (0.02) (0.01) (0.04) (0.01) (0.02) (0.01) (0.01) (0.01)

Household Income -0.02⇤⇤⇤ -0.03⇤⇤⇤ -0.04⇤⇤⇤ -0.04⇤⇤⇤ -0.05⇤⇤⇤ -0.05⇤⇤⇤ Household Income -0.03⇤⇤⇤ -0.02⇤⇤⇤ -0.05 -0.02⇤⇤⇤ -0.01 (0.01)(0.01) (0.01) (0.01) (0.01) (0.03) (0.01) (0.01) (0.01) (0.01) (0.01) ObservationsObservations 25298 24302 22112 24302 20705 24302 243021383 22112 22112 21649 22112

19

22 .5

.45 Homeowners

.4

.35 Non-Homeowners

.3 Probability of Voting for Donald Trump

.25 11 12 13 14 15 Log House Price 2016

Figure 6: Predicted Support for Trump by Homeownership

25 .05

0

-.05 on Support for Donald Trump

Marginal Effect of LogPrices 2016 House -.1

1 2 3 4 5 6 7 Ideology

Figure 3: Effect of Log House Prices on Trump Support by Ideology

.1

0

-.1 on Support for Donald Trump Marginal Effect of 15 Price House Yr Change -.2 1 2 3 4 5 6 7 Ideology

Figure 4: Effect of 15 Year House Price Change on Trump Support by Ideo- logy

21 REDISTRIBUTION REDUX

6

4

2

0 on Desire to Cut Spending -2 Estimated Effect of LogPrices House

-4 Strong D Weak D Lean D I Lean R Weak R Strong R Party ID

Figure 5: Effect of House Prices on Desire to Cut Spending by Party ID

References

Abramowitz, Alan. 2013. The Polarized Public?: Why American Govern- ment is So Dysfunctional. Pearson.

Algan, Yann, Sergei Guriev, Elias Papaioannou and Evgenia Passari. 2017. “The European trust crisis and the rise of populism.”.

Ansell, Ben. 2014. “The political economy of ownership: Housing markets and the welfare state.” American Political Science Review 108(2):383–402.

Becker, Sascha O, Thiemo Fetzer and Dennis Novy. 2017. “Who voted for Brexit? A comprehensive district-level analysis.” Economic Policy 32(92):601–650.

Bisgaard, Martin, Kim Mannemar Sønderskov and Peter Thisted Dinesen. 2016. “Reconsidering the Neighborhood Effect: Does Exposure to Resid- ential Unemployment Influence Voters’ Perceptions of the National Eco- nomy?” The Journal of Politics 78(3):719–732.

Blinder, Alan S and Mark M Zandi. 2010. How the great recession was brought to an end. Moody’s Economy. com.

22 Populist

Trump’s Midwest Republican Base Margin

Low Spending High Spending

Democratic Elites and Republican Never- Democratic base Trumpers

Liberal