Starting in the late 1990s, a strong political split took hold in the United States. White rural voters, who had previously voted pretty similarly to white urban voters in presidential elections, started drifting apart and became much more loyal to Republican candidates. Footnote 1 Many other countries have long had rural-urban political divides, but historically both American parties drew support from cities and countryside alike. Up until about 25 years ago, many rural places were still competitive politically, and where one party was dominant it often was the Democrats as much as Republicans. Since then, this divide between rural and urban voters has appeared all over the country, making social and political polarization worse (Brown, Mettler, and Puzzi Reference Brown, Mettler and Puzzi2021). And because of how uniquely American politics works institutionally, this shift has given Republicans more power than their share of votes might suggest (Rodden Reference Rodden2019).
Lately, political scientists studying America have started unpacking this place-based split with lots of detail (e. g., Mc Kee Reference Mc Kee2008). Most work so far looks at how individual rural folks think about their lives and politics – focusing on their unique rural identity or values (e. g., Cramer Reference Cramer2016; Gimpel et al. Reference Gimpel, Lovin, Moy and Reeves2020; Lunz Trujillo Reference Lunz Trujillo2022; Nelsen and Petsko Reference Nelsen and Petsko2021; Diamond Reference Diamond2023). This research has done well putting the rural-urban gap on political science’s radar as a key part of today’s polarization debate and made some important points we use here to form ideas. But scholars still haven’t really nailed down why this nationwide divide popped up when it did or why it kept growing (except for some exceptions like Rodden Reference Rodden2019; Hood and Mc Kee Reference Hood and Mc Kee2022 for the South).
By pointing out what’s missing in these studies, we’re adding to research that asks why new political splits appear when they do-and then stick around. Societies always have differences between groups but only some end up politically important-and which ones can change over time. Sometimes religion or ethnicity matter for politics; other times they don’t; sometimes class lines shape party loyalties instead. So why did America go from having no real partisan gap between city folks and country folks to having a big one in just a few decades? And why has it gotten so deep?
To tackle these questions we lean on theories from comparative politics that explain links between social divisions and parties. We especially build on Seymour Martin Lipset and Stein Rokkan’s (Reference Lipset and Rokkan1967) idea that both economic shifts plus growing state centralization are key forces. We take a developmental look-that is looking back further than most studies-by exploring patterns since the 1970s in rural versus urban voting behavior. That helps us show when exactly certain processes helped create this divide over time. Our findings suggest multiple factors worked together across different moments.
Descriptively speaking: while back in the 1970s rural counties resembled cities on several social economic demographic fronts-they’ve since grown very different from each other. Then we link those trends to how people voted for president over time: through the ’90s into early 2000s people living in shrinking economies or areas losing population leaned Republican first. Then later-in late 2000s-places with more less-educated residents or lots of evangelical churches per person swung GOP too – both traits commoner outside cities than inside them. Also higher anti-Black racism among white rural folks especially during 2010s helped lock down Republican control there too. By going through polarization step by step-economic changes first sparking resistance to national policies pushed by centralized government-the country ended up with a major political fault line running right along urban-rural lines.
Political Differences Growing Across Time & Place
Figure 1 shows how many votes went to Republicans nationally by county type using presidential elections data-with “rural” & “urban” defined by Office of Management & Budget standards used also by USDA which factor not only population density but social/economic ties too. Footnote 2 Between 1970s-1990s city & countryside voters moved mostly together supporting same presidents at similar rates-even if country folks were just slightly more likely Republican fans. Both supported Nixon Reagan strongly – Reagan especially during his huge ’84 win – yet also backed Bill Clinton heavily. But starting around year 2000 sharp split emerged as rural voters increasingly backed Republicans while city dwellers preferred Democrats. The gap grew fast-from just two points difference in ’92 election-to twenty-one points by election of 2020. Footnote 3
Figure 1 The divide between rural & urban presidential voting
This widening gap shows up everywhere across regions too-as figure 2 shows. Of course everyone knows about big changes in South where many counties switched from solidly Democratic earlier last century toward stronger Republican support. Less obvious is that since around ’96 South developed a sharp split within itself-a real difference between its cities & countryside-a gap hitting twenty points by election day in 2020(Hood & Mc Kee Reference Hood & Mc Kee2022). Every region shifted like this at roughly same times Midwest now even bigger gap than South at twenty-one points West close behind with nineteen. Northeast had biggest gap already back then but grew slower though still fourteen point wide now.
Figure 2 The divide between city & country voting by region
These regional voting shifts are just one sign of this new geographic cleavage. The urban-rural vote split also widened nearly everywhere among states. Many mostly-rural states flipped toward Republicans across offices – governors legislatures congress members. We focus mainly here on presidential votes using counties as units. In spite of seeing trends across both sides our main spotlight remains on rural areas cause that’s where biggest changes happened. And even if fewer Americans live rurally now(see figure4), US institutions give those less populated areas outsized influence(Rodden Reference Rodden2019)-meaning their shift matters big-time.
Building A Developmental Viewpoint
Despite growing attention recently few studies try explaining how place-based trends changed over longer time frames. Ethnographic work plus large surveys reveal several things-rural folks often feel resentment against city types(Jacobs & Munis Ref Jacobs&Munis22; Cramer16); anti-black racism shapes white rural attitudes(Nelsen&Petsko21); and white rurals stick with GOP after controlling for other stuff(Scala&Johnson17). Others point out sorting effects(e. g., Bishop08, Gimpel&Hui15)but detailed voter registration analysis suggests sorting explains only little(Martin&Webster20). Footnote 4 Yet most research uses recent cross-section data focused narrowly on current views limiting answers about when/how these factors gained power.
First off such work kinda assumes the divide lasted forever-which isn’t true-it formed only past quarter-century. Historically culture/politics divided cities/rurals sometimes-but usually varied locally instead nationally. Today’s nationwide clear-cut durable split begs “why when?”Second existing work treats divide like rooted solely inside innate “rural” traits-but back then they voted closer to city folk. Explanations based only on today must explain timing Why attitudes suddenly count Thirdly modern studies tend isolate single causes(economy despair/status threat/racism etc.) ignoring how multiple forces might combine sequentially. An example would be old economic shocks stoking resentment fueling later fights over environment/social policy.
So our developmental framework takes broader look. Paul Pierson argues understanding politics needs “moving pictures,”not snapshots(attending historical development sequence) rather than isolated causes(Orren&Skowronek04; Thelen&Streeck05). We want see what trends shaped divides when/how differing effects unfolded causing next steps(Eric Schickler16).
Plus our approach stresses complexity not single cause. Multiple interacting historical processes unfold slowly. We explore timing/sequencing each factor scholars found relevant before-all helping fill gaps left by recent narrow analyses.
Thinking About Place-Based Trends
America saw huge economic/social/political shake-ups last decades affecting places differently. Did these shifts spark partisan splits along urban/rural lines? How might timing/interactions affect politics?
We lean heavily here on Lipset&Rokkan classic idea positing two main drivers. First is economy-related conflict over resources/products/jobs originally tied European industrialization/global trade. Footnote 6 We update focus post-industrial capitalism rising since late twentieth century highlighting growth concentrated mostly inside cities fueled knowledge economy growth. Want see if such economic gaps drove party preferences along urban/rural lines. Second driver is “nationalizing dimension”-people far away pushing back against elites’ state centralizing efforts resisting bureaucracy/machine standardizing(Lipset&Rokkan67 footnote7). Although they wrote about suffrage expansion decades ago, we see national power manifesting today through extension civil rights laws enforcing equality regardless race/gender/etc. and environmental regulations/gun control/immigration rules. As Dem Party came seen pushing these policies late ’90s-early ’00s others felt imposed upon leading resistance. By analyzing both processes’ interaction sequentially(Lipset&Rokkan67 p34), we get better sense emerging cleavages timing.
While nationalizing trends began earlier we expect disruption caused first economically hurting rurals made them open switching parties. Then later nationalization efforts mobilized resistance driving stronger GOP turn. Rurals resented outsider government telling them what do org mobilizations followed tightening cleavages(Footnote 8).
Left Behind
It ‘ s common wisdom that rurals feel ” left behind ” economically. The ethnographies support it(Cramer16, Wuthnow18), showing distinct resentment towards perceived urbane elites. Clearly no broad study yet detailing why alienation arose exactly when nor linking directly to GOP rise historically. US absorbed rurals into nation economy before via ag programs(Farm Security Admin, Tennessee Valley Authority etc.). Last thirty years upheaval reversed trend with free trade agreements e. g. NAFTA93 opening China imports early00’s causing plant closures widespread job losses rurally. Technological advances cut farming/mining jobs. Moreover deregulated finance wrecked small local banks harming credit(Barton22). All combined entrenched uneven development job/population growth stark divergent Florida&King19.
Figure3 displays jobs per capita showing stagnation rurally vs booms urbansince200094%jobgrowth nationwidenetto urbans. Poploss hit47%ruralloss15%urbancountiesdespitepopaging(Wuthnow18 ). This signals huge worsening fortunes for countrysides.(Fig4 Population stagnation graphs complement.)Jobs weighted populaid helpful conservative econ proxy non-city places despite fewer persons overall population dropped drastically
Figure3 Employment per capita overtime
Figure4 Raw population growth overtime
Job/pop loss tied politically. Choi et al21, Wright20, and Autor et al20show trading exposed whites swing republican. White identity/status fear activated(Baccini&Weymouth21 )Clinton-Dems embrace free trade/deregulation triggered defection. GOP wins remain strong centers GDP rich counties(Muro et al20, Muro&Liu16)
Pop loss spurs grievance closing businesses/services/schools/hospitals triggering anger/frustration esp conservatives vocal(Wuthnow18 ), pop loss studied little US polit, but comparativists link it populist right rises(Dancygier et al22 ). Spatial inequality traps residents away(Patana22 )with backlash radical right support expected. Pop growth drop fostered partisanship switch countrysides. Reason couldn’t be better GOP policy addressing decline but rather felt abandoned led backlash national issues rise after lost faith dems.
Education Gap
Diverging conditions produced education importance. Most youth leave countrysides seeking schooling rarely return(Wuthnow18 ). Fewer low ed stay rurally. Socially education shapes issue views(Kitschelt&Rehm19)and may fuel perceptions imposed priorities nationally. People college grads cluster urbans socially liberal(Dems)while rurals lagged behind. Going beyond economics education signals cultural divisions(liberals vs conservatives )triggering resentments(Lipset&Rokkan67).
Figure5 charts college degree attainment showing near parity70’sthen rapid divergence35%college grads urbans vs21% rurals latest census. Urban progressives align Dem while lower ed rurales initially Democrats flipped post2008 GOP due growing ideological distance tied natl policies including guns/environment/social issues opposing imposed elite will. Divergence feeds into wider politic cleavage emphasizing division own terms.
Figure5 Education levels diverging Rural vs Urban overtime
Organization Mobilization
Groups mobilize resentments creating narratives shaping votes. Scholars highlight religous groups Koch affiliates gun groups police fraternal order all boosting Repub shift(Skocpol refs ; Margolis refs ; Lacombe ref ). Evangelicals particularly significant aligning strongly red socially conservative causes forming Christian Right surprise1980 landslide moral majority awakening giant organizing1994 election mobilizations abortion gay rights conservative judges push continuing midwest south influence. Trump16 less visible formal org role evangelicals yet high personal support remains(Wilcox18)
Using religious census data evangelical church density stable heavily skewed towards rurales nearly threefold versus urbans. Footnote10Their presence offers organizational base fostering conservative activism amplifying existing tensions post-economic decline once seen dem abandonement. Resentment fuels opposition against Dem promoted agendas viewed imposed culturally alien. Anticipate denser evangelical counties vote harder GOP increasingly post mid90’s.
Racism And Racial Threat
Race remains defining politic force despite historic Obama election08 Trump victory16 racially charged campaign. White partisan gaps widen racial resentment predicting affiliation/votes controlling others(Enders Scott19).
Where does racism fit geographically? Nelsen/Petsko19 find anti-black linked with rural consciousness cross-sectionally but timeline unclear. From ANES geo-coded county files back to76 ask whether govt should help Black Americans scale measuring racial animus correlate traditional scales(r=.66, p=.001)good proxy though limited span questions asked starting86. Footnotes11-12
Figure6 Racial animus scores Rural vs Urban whites Overtime rescaled zero hundred higher means more support govt aid blacks76-20 Urban whites liberalized faster diverging modest ten point gap recent years Footnotes13
Meanwhile demographic diversity increased since immigration reforms86 spurring ethnic minorities including Latinos grow fast overall especially urbans(Lichter refs ). Figures7-8 track nonwhite/Hispanic shares rising sharper suburbs/cities yet some gains rurales too diversity pace varies sharply fueling potential threat/contact dynamics.
Competing theories exist Contact theory predicts intergroup exposure reduces prejudice whereas Threat hypothesis expects increasing diversity sparks competition backlash activating racism. Jardina19, Baccini&Weymouth21 show threat boosts candidate choice.
We expect race related backlash highest amongst white rurales opposing expanding minority shares fueling greater Repub support particularly post08 Obama Trump elections. Anticipate anti-black racist attitudes stronger predictor dividing rurals/republican alignment versus ethnic threat weaker effect declining lately correlating older era predivided period.
Wrapping Up
<Ｐ Ｃｌａｓｓ＝＂ｐ＂＞In summary small town/city differences70’ widened greatly90 onward appearing almost separate worlds facing divergent challenges. Shifts led many longtime Dem affiliated countrysiders feeling forsaken flipping Republican multiple reasons acting sequentially. First Econ distress triggered vote switch second cultural-nationalist backlash against perceived elites promoting values alien locals helped cement division reinforced organizations boosting turnout. Finally polar opposite camps entrenched represent major American electoral faultline.<／Ｐ＞ ＜Ｈ２ ｃｌａｓｓ＝＂Ａ＂＞ Methods And Data ＜／Ｈ２＞ ＜Ｐ ｃｌａｓｓ＝＂ｐ＂＞ We test our ideas using two strategies. County-level regressions analyze relationship between vote share %GOP president w/economic educational religious demographics. Interactions test differential effects across metro/nonmetro zones. Data pooled into three periods capturing baseline emergence intensification allowing formal tests change coefficients(STATA suest/Wald tests). Year-region fixed effects control unobserved confounders robust SE clustered county level applied. For racial threat theory use individual ANES survey merged geo-county info restricted ICPSR file limited sample non-Hispanic whites examining racial animus/local Latino share effects controlling demographics estimating logistic models predicting likelihood GOP vote choices. Interactions check variation metro/nonmetro periods analyzed separately.<／Ｐ＞ ＜Ｈ３ ｃｌａｓｓ＝Ｂ＞ Dependent Variables ＜／Ｈ３＞ ＜Ｐ ｃｌａｓｓ＝＂ｐ＂＞ County models dependent variable Republican vote share scaled zero-one denominator includes all votes accounting third-party anti-establishment influence. Data source David Leip Atlas. US election results. Logistic regressions estimate individual GOP voting coded one else zero among non-Hispanic whites.[footnotes]＜／Ｐ＞ ＜Ｈ３ ｃｌａｓｓ＝Ｂ＞ Independent Variables ＜／Ｈ３＞ ＜Ｐ ｃｌａｓｓ＝Ｐ＞ Rurality uses OMB definition metro/non-metro based pop density integration adjusted consistently using93 delineations [footnotes]. Left Behind measured via logged percent job/population growth derived Census Bureau BEA databases. Education measured percent adults ≥ four-year degrees interpolated Census. Decennial. Religion focuses evangelical congregation counts weighted pop ASARB census controls Catholic congregations. Racism proxies include ANES racial animus question plus local Hispanic share representing ethnic threat controlling age income gender education. Pop age controlled older pct≥65%. Race ethnicity pct non-Hispanic white included.[footnotes]<／Ｐ＞ ＜Ｈ２ ｃﾗＳＳ＝Ａ＞ Regression Results Political Effects Of Trends ＜／H２＞ ＜Ｐ ｃlass=”page_paginator_text_all_pages_block_content_header__title____primary____text primary_large dark text-md text-center mb-md mb-md–mb-lg mx-auto md: text-left font-serif font-normal text-base italic max-w-screen-lg sm: max-w-[95vw] overflow-hidden truncate___black___dark: text-white tracking-tight leading-[32px] uppercase mt-[12px] sm:[30px] md:[28px]" ＞Our analysis reveals various complex relationships showing distinct/concentrated factors shaping voter behavior across place/time. Above simple model confirms significant rising impact of ‘rurality’ correlating increased repub support third period despite controls implying interactions matter greatly.[continued]</Ｐ＞..Source link
