paradoxically—contributes to an individ-ual’s sense of place-based resentment.

jealousy

Similarly, future work shouldexamine the extent to which urbanites harbor resentmenttoward suburbs and the political ramifications of thosefeelings

such as intra-state urban-rural divides

resentment might not have held its predictivepower, as modeled in this paper.

over urbanAmerica’s disparate political influence and anti-ruralstereotypes (e.g., “those backward hicks”) have longbeen pervasive in America and many other Western so-cieties

Ourresults, which show that place resentment was onlyconsistently predictive of vote choice for rural voters

that place-based animus fluctuates with thegeographic context of the election.

“have become so ideolog-ically inbred that we don’t know, can’t understand, andcan barely conceive of ‘those people’ who live just a fewmiles away”

and, therefore, islikely an important component of the emergent urban-rural divide in American politics.

Figure

larger sums of money that U.S. Senate electionsgarner and how that excess money translates intoheightened levels of voter exposure to campaign re-lated communications across various media

there was a 13.5 point decrease in the prob-ability of voting for the Democratic House candidate

Figure

Place-based resentmentpredicts electoral outcomes in the U.S. House, Senate, andgubernatorial contests, but only among the rural subset ofthe American population

The interaction termcaptures variability in the effect of place-based resentmentamong self-identifying urbanites, suburbanites, and ru-ralites. Given that interaction term in the model, wepresent the results of the marginal effect of place-basedresentment.

cities are often shorthand for peoplewho are not white.

Finally, we include a multi-faceted measure of racialresentment. 8 Racial attitudes imbue the American psy-che, and distinguishing race from other dispositions is acomplicated endeavor

vote choice in the2018 Congressional midterm elections and the variouscontests held for state governor in that year.

Figure

Figure

among those who identify asliving in “very rural” areas.

using the difference in feelingthermometer ratings of the Democratic and Republicanparties as the dependent measure;

Evaluationsof the Parties

(nearly 17% higher, in fact)

rural areas have an outsized voice inAmerican politics due to institutions like the Senate andElectoral College

the appreciation forpeople who choose to stay in communities unlike theirown;

the deservingness of differentcommunities in respondent’s state;

The distribution of political power within re-spondent’s state;

self-report depending on how each respondentunderstands their community in relation to others

mong all respondents, place-based re-sentment will predict vote choice in the U.S. House,Senate, and Gubernatorial Elections. Respondent’s placeof residence will moderate the directionality of resentfulattitudes toward specific parties running for office.

have trans-formed into a form of place-based resentment

This raises the question of whether placed-based attitudes are distinct from racial ones

not everyone embodies place identity and, ofthose that do, not everyone harbors place resentment

individuals rely on their own sense of place in orderto understand their personal stakes and justify their po-litical beliefs, because they live “not just anywhere, [but]somewhere in particular”

their place-based ingroup as beingunjustly deprived relative to the “others.”

hese identities provide apowerful heuristic for how Americans interpret politicalphenomenon that affect those spaces

Neighbors are engaged in a common enterpriseoften despite religion or partisanship; they are tied to-gether by the fact that they live near one another and sharea set of common experiences

nique because of eachplace’s inherent peculiarity.

filter political information

creating narratives about the types of communities thatare winning or losing, and who is to blame.

By quantitatively measuring place-basedattitudes, we are able to distinguish place-based resent-ment from other prior dispositions, such as feelings ofracial animosity, which are highly spatialized

within state-and district-level races.

living in different placesare drawn together because of how think about com-munities unlike their own

build social co-hesion

We conclude that placematters in the construction of individuals’ political beliefsbecause place—the social connection individuals feel to aparticular locale—is highly amenable to politicization.

. Rural resentment is a powerful predictor of vote choice in bothelection years examined

d racial resentmen

we study the extent to which Americans feel animus toward communities that aregeographically distinct from their own and whether these feelings explain Americans’ attitudes toward the two majorpolitical parties and self-reported vote choice.

location in space structures the propa-gation of opinion, while factors such as size and density shape its diversity

social context in which people share their likes and dislikes, values, andbeliefs

In someplaces the conflict may be primarily economic—revolving around the way ruralproperty is overvalued and taxed, for example, while in others the separation is morecultural and moral.

. If we take two voters who are of thesame race, religion, age, education level, income, sex, marital status, and report-ing the same level of religious commitment, and one is living in the central city,and another lying well outside a metro area, there will be a difference in politicalparty affiliation.

This is a dubious conclusion, however, given that com-positional characteristics are themselves distributed unevenly in space, likely theresult of features associated with disparate settlement and socialization patterns.

Similarly, the probability of identifying as a leaningDemocrat declines from .18 to .16 and the probability of identifying as a leaningRepublican increases from .15 to .18. The probability of being an independent risesmodestly from .09 to .10.

independent

than it is on assessing the impact ofplace location net of whatever influences these controls exert.

concentration

independents

education

Population Density (logged scale)

Consider Casper, Wyoming (82601) along I-25 in eastern Wyoming.Though it is two hundred miles from Denver, the community measures nearly 1,700persons per square mile, which is more than double the median in the sample. Thereare also locations that are close to major cities but are lightly populated, includingneighborhoods near Phoenix, Arizona, and in Anchorage, Alaska.

Localesat approximately this distance include Woodbridge, Virginia, and Bowie, Maryland,both suburbs of Washington, DC; and Chesterfield, Missouri, a western suburb ofSt. Louis. These are middle-range suburbs, not bordering the core city but not lyingat the fringe of the metropolitan area, either.

less

A respondent who lives in close proximity to a citythis size is considered less rural than one living farther away.

we use the population density of the locale in which the respondent resides.We calculate the population density based on a ten-mile radius around the centroidof the respondent’s ZIP Code.

we situate each respondent at or near thecentroid of their ZIP Code of residence.

Lower density areas are expected to adhere to morallytraditionalist positions, controlling for compositional characteristics of the popula-tion and also accounting for distance to the nearest city.

Small scale set-tlement encourages religious adherence and traditional views of morality becauseit accentuates group life among those with common beliefs over acting as an indi-vidual

Habitual ways of thought and behavior are upheld and perpetuatedover long periods of time.

Consequently, values in these places are more interdepend-ent and distinct

Left leaning, or “progressive” ideologiesare typically those more accommodating of new modes of thought and behaviorin social and political life.

These measures gauge the place-based differ-ences in the number of people an individual might meet within a routine work-day.

the urban–ruralpartisan gap has more than a simple source rooted in racial composition, economicconditions, age differences, or religious background.

Studies document the unconventionality ofurban life, running contrary to tradition in multiple domains

Goods and ser-vices that were once considered luxuries become necessities as more people come todesire what is advertised

The physicaldistance between the two locations is itself influential in the observed difference inpolitical values

Urban–rural differences in opinion may exist asa consequence of the separation of two populations from each other. As distanceincreases, so will the divergence in viewpoint.

Distance captures the degree of isola-tion of two populations and figures prominently in explaining species differentia-tion across the landscape.

the values and traditions of that bygone life may still influence contem-porary beliefs and behaviors

behavioral path dependence,” whichoccurs when “ideas, norms, and behaviors [are] passed down...[and] interact withinstitutions, reinforcing each other over time

self-reliance and traditionalism

more individuals moved to cities, their common economic inter-ests drove class consciousness and created political unity within urban and ruralpopulations respectively.

With the rise of the city came the rise of manufacturing and thedecline of agriculture.

dominated by sectional interests defined by the “greatcrop regions, founded on climate and

first, the geo-graphic distance between small towns and major central cities, and second, differ-ences in population concentration.

Among some groups, theurban–rural divide is wider (e.g., whites) than among other groups (e.g., blacks)

Third, the concentration and density of urban Democrats have reinforced theirparty loyalty and progressive-leaning over time with a similar development occur-ring among geographically dispersed Republicans, thereby heightening the diver-gence in political preference by location

We find that sizable urban–ruraldifferences persist even after accounting for an array of individual-level characteris-tics that typically distinguish them

to understand the urban–rural fissure that hasbeen so noticeable in recent elections

This might be the case not just for developmentof political attitudes in the US South but also in other arenaswithin American politics and elsewhere in the world

an im-pact on political habit whose influence has not worn awayeven yet.

As affirmative support, we showed that greaterprevalence of slavery predicts more conservative (for manyyears more Democratic) presidential vote shares, higher ratesof radical violence, and decreased wealth concentrated inblack farms in the decades after Reconstruction

(i) partisan identification,(ii) attitudes on affirmative action, (iii) levels of racial re-sentment, and (iv) attitudes toward blacks

What these correlations show is that children withracially conservative parents in 1965 are more likely to beracially conservative themselves at least through age 50,which is evidence of intergenerational socialization.

which measured the racial attitudes of anational probability sample of high school senior studentsin 1965 along with their parents

10 percentage point increase in proportion slave leads to a1.8 percentage point drop in the percent of whites whoidentify as Democrat today (95% confidence interval:[22.7, 21.0]). Where mechanization grew rapidly, with0.06 more tractors per 100,000 acres (90th percentile), thesame change in proportion slave leads to only a 0.2 per-centage point decrease in the percent Democrat (95%confidence interval: [21.1, 0.06]).

the incentives for whites tointerfere in the labor market with such tactics should lessen

in comparison to white farmers, blackfarmers in former high-slave areas were significantly worseoff than those in other areas of the South. They were morelikely to be under tenancy agreements and less likely to owntheir own farm.

is greater in counties that had highslave proportions in 1860,

but also to suppresstheir mobility and wages

does not trace itsorigins to this time period.

In both states,perhaps surprisingly, there is little evidence of a strongrelationship between slavery and vote choice, even in anelection that focused so heavily on the issue.

differences in white viewsappear to organize around the density of slavery more stronglyafter the Civil War.

suggesting that there is some decay in these geographicallybased relationships over time

they were threat-ening because they were an important provider of labor and,in the post-emancipation environment, they could leave.

poor whites were complicitwith the landowning elite and would engage in and supportviolent acts toward blacks, even though such violence couldpresumably also lower white wages

emancipation brought blacks some freedomover the amount of labor they supplied, and many ex-slaveschose to work for themselves rather than for the white rulingclass

evidence for intergenerational transmission of racialattitudes.

institutionalpath dependence and intergenerational socialization

bypromoting racially targeted violence, anti-black norms, and,to the extent legally possible, racist institutions.

Proportion slave, 1860

the effects are in the oppositedirection as statistical discrimination theory would predict.

discrimination against blacks because they are, on average,poorer than whites

butthose differences are fairly constant across proportion slavein the county.

For geographic sorting to explain our results, patterns ofmobility into (and out of ) the former slaveholding areaswould have to differ from non-slaveholding areas

Proportion slave, direct effect

Indeed, the correlationbetween percent slave in 1860 and percent black in 2000 is0.77

Cotton suitability

Second, because counties may have had different norms aboutrace, we include controls for (vi) the proportion of total pop-ulation in 1860 that is free black.

but we use the differ-ence in case slavery has an overall effect on racial groupthermometer ratings

but may also reflect beliefs on policy issuesclosely related to race, including redistribution

for example, income gaps betweenblacks and whites, urban-rural differences, and other con-textual and individual-level factors

should have decayed more) in areaswhere the incentives for anti-black attitudes faded earlier

ntergenerational socialization

reacted more sharply toemancipation by curtailing blacks’ rights and oppressing newlyfreedmen and their mobility

the historical persistence of attitudes orig-inating in slavery and (ii) contemporary factors, includingcontemporary demographics and geographic mobility.

if the politics of the South revolves aroundany single theme, it is that of the role of the black belts”

makingour position quite distinct from much of the existing publicopinion literature.

abruptly increasing black wages, raising la-bor costs, and threatening the viability of the Southern plan-tation economy

Du Bois (1935)

that un-dermined Southern whites’ political and economic power.

We show thatthese differences are robust to accounting for a variety offactors, including geography and mid-nineteenth-centuryeconomic and social conditions.

weshow that whites who currently live in counties that hadhigh concentrations of slaves in 1860 are today on averagemore conservative and express colder feelings toward Af-rican Americans than whites who live elsewhere in theSouth.

which in turn have been passeddown locally across generations

Following the Civil War, Southern whites faced political and economicincentives to reinforce existing racist norms and institutions to maintain control over the newly freed African Americanpopulation.

Whites who currently live in Southern counties that had highshares of slaves in 1860 are more likely to identify as a Republican, oppose affirmative action, and express racial re-sentment and colder feelings toward blacks

we would expect the negative effects of shalegas on a wide range of environmental and energy issues.

if anything, were less likely to have Republican mem-bers of Congress.

(i) the effects of shale resources on theHouse of Representatives election outcome, (ii) the narrowversus broad effects of shale on environmental voting, and(iii) the legislative voting pattern in districts that remainedin the hands of congressional members of the same party.

artisanship of legislator in the pre-boom period(Rep. in 2004, models 4 and 5)

1.042

Shale # post-boom

This exercise suggests that shaleand no-shale units moved in the same direction and to thesame degree in their trends in the pre-boom period.

We include New York inour analysis because the new fracking moratorium is en-dogenous to state politics.

There was no substantialredistricting in the districts in our sample between 2003 and2011.

our case is thus one of uncertainty rather than in-terpretability or identification

Again, a pro-environment House Representative would vote “no.” For alist of example bills, see table A3

Shale # Post-boomijkt

Thus, it is possible that elected officials changedtheir behavior in 2012 in response to new district boundaries.

If, all thingsconsidered, the public reacts to access to shale gas by increas-ing demand for environmental protection against the putativenegative effects of fracking, then pro-environmental candi-dates should see their electoral fortunes improve. On the otherhand, if economic concerns dominate, then the expected elec-toral effect should favor anti-environmental candidates.

“behavioral” causal mechanism(officials change their voting behavior) and a “selection” causalmechanism (officials with different preferences win elections)

Elected officials might respond to the availability of frackingwhen in office by changing their support for pro- or anti-fracking policies

Where shale gas is avail-able, the industry has an incentive to support the anti-environmental camp and strengthen the pro-fracking sen-timents of the population.

as economic interestsprevail

the salience of environmental issues among pro-environmental voters and political elites

cre-ates demand for environmental regulation.

While some of the rents goto the fossil fuel industry, some also go to landowners whosell access to extractors and to the public through royaltiesand taxes.

as the unexpected resource windfallincreases the political and economic clout of the energy in-dustry, thus reducing the competitiveness of pro-regulationpolitical candidates.

Thereport made energy companies aware of a potentially hugewindfall.

we detail the policy effects of natural resourcewindfalls in a polarized society

pro-shale candidates are more likely to winelections and also hold a set of correlated anti-environmentalpolicy preferences that they carry with them into office.

26–35 percentagepoints more likely to elect Republicans relative to the timetrend in neighboring districts.

electoral fortunes of pro-environmentaland anti-environmental candidates

the difference is 18 percentage points on votesthat are specifically about energy and drilling.

Third, we only consider direct neighbors along theborder in our main specifications, as these are the most likelyto have equal propensities of falling on either side of the bound-ary

is consideredconditionally independent of potential outcomes

such as regulation of oceans,transport, and wildlife.

throughthe changed incentives of elected officials and preferences ofthe local population.

shale gas development has divided neighbors, spurredlawsuits and sown deep mistrust

conflicts over natural resources can affect a broader range ofregulatory issues through the selection of elected officialswith correlated policy preferences

changed the votingrecord of House Representatives on environmental policy relative to neighboring districts without access. Votes become15–20 percentage points less likely to be in favor of the environment

Political dispositions—espe-cially party—moderate its uptake as do contextualfactors outside movement control.

By changing norms and affecting how people thinkabout teaching children, social movements may yieldlong-term attitudinal changes in the future public

Our measures in con-trast capture behaviors: consumption patterns and col-lective action choices that we corroborate with othertypes of data

Ourresults suggest both that movement concepts are storedin long-term memory and that the politics of socializingchildren is a topic even those without young childrencare about

the public, and not justparents, has a stake in crafting the nation’s futurethrough the socialization of children

Whilethese estimates are imprecise due to small sample size,it appears Democrats are moved to support curricularmaterials focused on issues of racism and discrimina-tion when primed with BLM, while Republicans andindependents are not

Republicans

We expectthat compared to the control, priming BLM shouldmatter only among Democrats

We test whether exposure to a mere mentionof BLM increases selection of The Hate U Give(1) relative to the other three books (0)

how did white Americansthink about exposing kids to progressive race conceptsrelated to the police, discrimination, and white privi-lege in public schools?

Peaceful protests mayhave opened opportunities for our white parent sampleto include their children in movement politics andincreased the likelihood they would do so for thefirst time.

The propor-tion of peaceful protests has no relationship to in-homeactivities—but it is positively associated with engagingin public-facing actions

then, parents whose workforce hoursdecreased during the pandemic—and presumably,whose caregiving hours increased—were more likelyto engage in progressive race-related parenting thanthose with consistent employmen

For each respondent, we create a variable—pro-portion peaceful protest—indicating the share of BLMprotest events defined as peaceful within 25 miles oftheir zip code between May 25, 2020, and our surveyfielding.

I Am Enough, a bookfeaturing themes of diversity, moved up in its rankingby four spots and remained a bestseller for 18 weeks

31% of our sample are first-timers:they report engaging in one of these race-focusedactivities for the first time during this period

These actions arethose that are publicly observable outside the home andmost require resources or opportunities coordinatedwith others

ompleted this action prior to May2020.

differences in how social-izing agents’ engaged with children

alking about race with children is normativelygood

They pro-vide tips about books to buy, television shows to watch,and ways to start and lead conversations with childrento shape their racial attitudes.

but this content was alsodisproportionately progressive in valence.

Conservative content was again absent from the random sampleused to assess interrater reliability. Few posts in the data receivedthis mark

progressive or conservative theme

These considerations may shapefuture behavior for some people persistently, and forothers when reminded

In the short term, thediscussion may have altered child-rearing behaviorsand practices.

it decreased to 681

shows that following Floyd’s murder onMay 25, posts on these topics increased dramatically