Elia Stebach reviews Julia Cagé and Thomas Piketty’s A History of Political Conflict: Elections and Social Inequalities in France, 1789-2022 (Belknap Press – Harvard University Press, 2025, 848 p. Translated by Steven Rendall).
Recent decades have witnessed the unravelling of France’s bipolar politics, once among the most entrenched in the world. The old duopoly, pitting socialists and communists against liberals and Gaullist conservatives, has fractured under the rise of the far right. Since the 2022 parliamentary elections, three blocs (left and green parties, Macron’s ruling coalition, and the radical right) have commanded roughly equal shares of the electorate.
In their A History of Political Conflict, Julia Cagé and Thomas Piketty trace the roots of France’s departure from bipolarity. Drawing on a remarkable digitization effort that recovered archival voting records, candidates’ affiliations, and socioeconomic data back to 1789 across France’s 36,000 communes, they argue that the bi- or tripartite character of the electoral landscape results from the relative weight of two determinants of voting behavior: socioeconomic status and geography.
The novelty of the data used, combined with the breadth of an inquiry spanning over two centuries and the whole national territory, places the authors in methodologically pioneering territory. Gone are post-electoral surveys, too recent for such a long-run analysis, and qualitative studies alike. The authors are upfront about their limited engagement with existing research: their work aims to open a new methodological front. Commendably, all digitized and processed data has been made publicly available.
Mapping two centuries of voting behavior
The first quarter of the book is devoted to testing the authors’ approach and analytical categories through a geo-economic sociology of France since 1789. Their findings are broadly consistent with historical consensus: socioeconomic and territorial inequalities declined steadily throughout the 20th century before reversing modestly; rural territories experienced significant relative demographic decline, particularly after WWII; Catholicism lost much of its social weight from the mid-twentieth century onward.
Source: Cagé and Piketty (p. 105)
Such a quantitative approach has its limits. Because the socioeconomic data used are administrative in origin, they miss the once-considerable weight of informal economy in rural France, which may introduce marginal bias into findings on the wealth distribution of communes. The spatial classification system, where each commune is assigned a fixed category (metropolis, suburb, town, or village) for the entire period, sits uneasily with the spatial expansion of urban areas since the 19th century. Yet the broad sociopolitical trends the book identifies are robust enough to survive these limitations.
Cagé and Piketty’s focus on electoral poles risks overlooking non-voters and those who cast blank or invalid ballots. The book’s second quarter turns to this gap and shows that voting has grown increasingly segregated along socio-economic and spatial lines in recent decades. This pattern holds for both voter registration, which is not automatic in France, and turnout, though the two follow different logics.
On registration, the share of citizens missing from the electoral rolls has held remarkably steady over the past two centuries, at around 10%. Village enrolment has always outpaced that of large cities, but this urban-rural gap has recently widened to historic extremes: urban citizens are now up to twenty times more likely to be unregistered than their rural counterparts. Cagé and Piketty attribute part of this gap to measurable socio-economic factors, such as growing divergences in age structure, residential stability and land ownership, but they emphasize the likely role of harder-to-quantify causes, particularly the widening gap in local life embeddedness as urban life atomizes.
Turnout followed a more uneven path. During the French Revolution, between 30% and 60% of eligible voters turned out in local elections. Interestingly, rural areas outpaced cities despite the concentration of polling stations in departmental prefectures, often hours away from rural hamlets. The authors attribute this gap to an overwhelming power of political factors in determining voting behaviour: turnout tracked support for the regime, since opponents preferred to abstain rather than travel and lose a day’s work voting against revolutionary plebiscites. From the Revolution’s outset, rural populations rejected the Ancien Régime more sharply and showed a stronger egalitarian, pro-democratic appetite than their urban counterparts.
Geo-economic determinants grew more powerful over the course of the Revolution. After the conservative takeover of 1794, rural turnout began to track local property structures: constituencies that had seen the largest transfers of ecclesiastic property since 1789 recorded turnout well above what pre-1794 voting patterns would have predicted. The authors trace this pattern to the economic policies of the early revolutionaries. The Church lands confiscated and sold off at the Revolution’s outset, which peasants had expected to receive, were passed overwhelmingly to the rising bourgeoisie. The following sense of betrayal struck hardest in the formerly ecclesiastic territories, where rural hopes had run highest. Conscription into the revolutionary armies weighed most heavily on the peasantry there as well, since counter-revolutionary forced were better organised in formerly ecclesiastic areas. Such geo-economic determinants, as well as individual-level socio-economic factors, grew more influential over the nineteenth century, alongside this century’s rise in inequalities.
From the late nineteenth century, as multi-party democracy took hold, turnout among registered voters stabilised near 80% for nearly a century. Class-based and territorial gaps in participation shrank faster than inequalities and stayed modest throughout the entire period. That equilibrium has shattered in recent decades: not only has turnout collapsed, especially in non-presidential elections, but the geo-social stratification of voting has grown faster than inequalities and reached historically unprecedented levels. In 1848, income, property, municipality size, and urban area size together explained only around 25% of the variance in turnout between communes; today, those same factors account for roughly 70%, with lower-income and urban registered voters the most affected.
Cagé and Piketty contend that the structural decline of left-leaning parties, from around 40% to 30% of votes cast over the past three decades, is partly attributable to a comparatively much greater rise in abstention among their core socio-demographic constituencies. This erosion has allowed the formerly unified liberal-conservative-nationalist camp to split into two distinct poles without fatally undermining its prospects of holding power.
Explaining bi- and tripolarity
Piketty and Cagé suggest France’s recent tripolarization owes as much to differential participation as to the rise of spatial factors as key determinants of voting behavior, alongside economic ones. Across every legislative election between 1848 and 2022, they identified two structural voting patterns: a spatial divide, with right-wing candidates consistently overperforming in rural areas and left-wing ones in cities; and an economic cleavage, with the left-wing vote historically determined by poverty. Whether the political field takes a bi- or tripolar form, they argue, depends on the relative weight of these two factors: a third pole tends to emerge when the economic cleavage loses prominence to the spatial one. This typically occurs when institutional or identitarian debates gain salience.
In the 19th century, the prominence of debates over nationalism and the new regime’s nature relegated economic issues to the background. The left’s programmatic offer, aimed primarily at the industrial working class, struggled to reach the rural poor: at equal levels of poverty, the left performed roughly three times better in urban areas than in rural ones. Conversely, the question of regime type, republic or authoritarian rule, divided the centre from the right: wealthy urban and rural classes, who benefited from the liberal republican order, mostly backed the centre, while rural lower classes supported the authoritarian right.
This tripartite equilibrium broke down at the turn of the 20th century. Gradual salarization of the rural world and the expansion of public services across the territory brought urban and rural workers’ social conditions closer together. The left shifted its economic platform, from property tax to a more rural-friendly progressive income tax, passed in 1914. Meanwhile, the republican regime’s consolidated victory, together with the salience of economic questions raised by WW1 indebtedness and the 1929 crisis, accelerated the coalescence of richer voters into a unified pole opposing the left’s social demands. From then on, the territorial cleavage became more marginal to communes’ political leaning. This bipolarization process peaked in the second half of the 20th century, as cross-bloc cooperation grew rare and intra-bloc electoral fluidity rose on both sides.
“In the electric social and political atmosphere of the late 1960s and the 1970s, the government, first under de Gaulle, then Pompidou, and then Giscard, felt obliged to give guarantees to salaried employees and labor unions, hoping to delay as long as possible the Left’s electoral victory. This driving dialectic made a decisive contribution to the success of the Trente Glorieuses and, beyond that, to the movement toward greater social equality and greater economic prosperity that was established in the twentieth century. (p. 493)”
Throughout this century, far-right currents, which polled between 5% and 15%, formed an established radical wing of the right-of-centre bloc. They drew mainly on executives and self-employed voters attracted by their staunchly anti-socialist stances, alongside niche publics like colonial war veterans. From the 1990s, and especially after the 2008 crisis, the National Front retreated toward a less economically liberal platform and made rapid inroads among rural lower-middle class, particularly property owners hit hard by the crisis. Within two decades, the sociology of Le Pen’s electorate had reversed, with the party now overperforming in small towns and villages. This sociological detachment from the traditional right-of-centre bloc ultimately reproduced nineteenth-century tripartition: lower and middle classes split left-right along a rural-urban axis, while a new upper-class-dominated centrist bloc combined the moderate centre-right with the centre-left’s social-liberal wing.
Source: Cagé and Piketty (p. 347)
Discussion
Some caveats deserve mention. Contrary to what Piketty and Cagé’s framing might suggest, voting patterns in France are not as sharply divided along sociological lines as in some other countries. For readability, the book consistently conflates individual voters with the average vote of their commune. Yet even in the most homogeneous constituencies, it remains rare for a single bloc to exceed 50% of the vote. When the authors call the 2022 Ensemble [Macron’s coalition] vote “one of the most ‘bourgeois’ observed for two centuries,” (8), a more accurate formulation would note that the Ensemble vote “showed the strongest correlation between communal deviation from the mean and average local income.”
Commune-level data also fails to capture dynamics like the growing residence-workplace disconnect or accelerating residential segregation, both of which may inflate the apparent weight of spatial components in political behavior. Fuller engagement with existing research might have clarified these grey areas, but the book’s introductory nature easily justifies such omissions.
Where the lack of engagement with existing scholarship is most costly, however, in the book’s explicitly political dimension. Piketty and Cagé, self-identified left-wing intellectuals, punctuate their analysis with strategic prescriptions: left-leaning parties, they argue, should seek a majority by reconquering the rural poor rather than courting social-liberal voters, by reasserting economic issues at the centre of their discourse. This argument follows naturally from their data but appears to sit in uncomfortable tension with the broader body of research, where recent studies of far-right voters point to a shrinking red-brown swing electorate (i.e. Gonthier, 2025). Given what this contradiction implies for the future of European democracies, it is perhaps the most fertile question future research might explore.
These limitations do not undermine the book’s two central findings. First, that the act of voting is more than ever segregated across individuals’ socio-spatial characteristics. Second, that the socio-economic cleavage which dominated twentieth-century voting patterns in France is increasingly sharing its explanatory power with spatial factors. Perhaps most importantly, Piketty and Cagé have successfully opened a new research agenda in electoral history, as evidenced by the rapid proliferation of quantitative socio-spatial approaches in French research. With the book’s recent English translation attracting considerable attention, there is reason to think this approach may become a lasting framework in the field.
Elia Stebach is an Assistant Editor at the Review of Democracy and prospective electoral studies researcher.
