Analysis of the Republican Party popular vote in Georgia county congressional elections of 1876 suggests that Charles Tilly's (1978) model of interest-based collective action would be useful if embedded in the dynamic model of political processes and mechanisms that Tilly (2007) proposes. Specifically, class (petit bourgeois), status (black), and party (liberal Republican) interests explain 25 percent of the variance in the election returns. Adding a racial-change variable increases the

more »

... d variance to 32 percent but fails to distinguish the yeoman and freedman constituencies and the process through which the Democratic Redeemers divided and conquered the opposition in the process of " de-democratization" (ibid.). By embedding the structural analysis in the analysis of process (quantitatively and qualitatively), we can appreciate how yeoman and freedman constituencies experienced contract/convict labor differently and expressed opposition to Redeemers in qualitatively different ways, ultimately facilitating divide-and-conquer efforts. Once they lost federal allies, Republicans in Confederate states were hardpressed to resist Democratic Redeemers. Democrats reestablished the uneasy alliance of Appalachian and cotton-belt whites while effectively dividing and conquering black opposition. Although most Confederate states could claim a white majority (Degler 1974: 193) and could use state government authority or local Ku Klux Klan or militia coercion to undermine opposition in blackmajority counties (Campbell 1877; Woodward 1951; Wynne 1986), the Democrats ultimately refused to tolerate islands of opposition and instead disenfranchised blacks and imposed oppressive one-party rule, enforcing labor 134 Social Science History discipline through contract labor and convict labor systems that eventually sustained a population of unfree white as well as black labor (Schwartz 1976) . The South could not accommodate demands for free land and free labor (Foner 1974 (Foner [1970 ), because, like the Party of Order in 1848 France (Marx 1978 (Marx [1852), it lacked a hegemonic capitalist class that could provide the foundation for a republican form of government. Thus the prospects for a moderate Republican (or even Democratic) Party were nil. Instead, a Radical Republican Party, representing the expressed interests of black freedmen and the imputed interests of white yeomen-essentially, a petit bourgeois agrarian class-fought for a revolutionary redistribution of property and the celebration of the rights of labor. Ultimately, the revolutionary forces were defeated, and the prospects for a democratic South were foreclosed by the rise of the Democratic Redeemer Party representing the interests of land and trade. This was a dominant but not a hegemonic class, a fledging capitalist class largely without banking and insurance interests and thoroughly dependent on the government to protect itself from workers, on the one hand, and from its own impotence and ambition, on the other. In other words, the problems of labor discipline and honor among thieves inspired repressive and conspiratorial efforts that defied the logic of republican capitalist institutions as they developed elsewhere in the United States (Moore 1967; Burawoy 1985; Bensel 1990).1 Nevertheless, counterrevolutionary, reactionary politics and oppressive labor relations were not inevitable (Honey 1993). There were times and places where it appeared that revolution was possible, and we can learn much about the prospects for social justice and democratic politics by looking at those places where, against the most overwhelming odds, black and white Republicans or third-party challengers opposed the Democratic Redeemers. Here we shall focus our attention on the white yeoman constituency of the Seventh Congressional District (in the northwestern corner of Georgia, on the Cumberland Plateau) and the black freedman Atlantic coastal constituency of the First Congressional District (in the southeastern corner of Georgia) in contrast to the rest of Georgia, particularly the black/cotton-belt county of Oglethorpe in what became the Eighth Congressional District (in northeast-central Georgia). These three districts were selected because they represent interesting differences within the range of local Georgia politics that correspond to major regional differences (on negative case methodology, see Hogan 1990: 18; Emigh 1997: 653-60).