Representative Government

The dependent variable of this study is representative government. I define it as a characteristic of regimes in which a freely elected assembly makes law, has effective oversight of the executive, and is in session regularly. Unlike democracy, it does not require either a wide franchise or an elected executive. Thus, although nearly all democracies have a representative government, not all representative governments are democracies. Before the twentieth century, most representative governments were constitutional monarchies. In such cases, a hereditary monarch administered the central bureaucracy, conducted foreign policy, and had the right to veto laws, while an unelected “upper house” shared responsibility for lawmaking. Representative government is an element of a regime, not its sole defining characteristic, so it can be present in a range of constitutional forms.

Representative government is qualitatively different from monarchical government, in which parliaments, if they exist, are constitutionally subordinate to the crown and serve at its pleasure. Both constitutionalism (Ertman Reference Ertman1997, 28–34) and representation, or “representative governance” (Boucoyannis Reference Boucoyannis2015), can exist in regimes with subordinate parliaments; thus, they are not terminologically equivalent to representative government. Before the Glorious Revolution, England had a constitutional regime with a representative parliament, but its government was monarchical. Only after the Glorious Revolution did England develop representative government (Cox Reference Cox2016; Roberts Reference Roberts1966). Over the past three hundred years, much more has changed, but many of these changes, like the eclipse of the monarchy and the House of Lords, constituted a deepening of representative government, while many others, like the progressive extension of the franchise, were accomplished through representative government. The origins of representative government, therefore, warrant close attention.

Structural Theories

Two structural variables, economic modernization and the distribution of wealth, dominate contemporary debates over regime change. Neither explains the Glorious Revolution. Representative government preceded the Industrial Revolution by a half-century; moreover, early modern England was unexceptional in its level of urbanization (Epstein Reference Epstein and Epstein2004, 10) and income equality (Milanovic, Lindert, and Williamson Reference Milanovic, Lindert and Williamson2011). These factors might account for the expansion of the franchise in nineteenth-century England but not for the rise of representative government in seventeenth-century England.

More relevant are geography and state capacity. As an island, England was unusually secure. According to Hintze (Reference Hintze1975, 199), it “needed no standing army, at least not one of Continental proportions, but only a navy” and, as a result, “developed no absolutism.” Furthermore, it had an unusually unitary state. The Norman Conquest of 1066 strengthened governmental authority throughout England, rationalizing and centralizing the political system. Parliament originated as an instrument of royal power over the provinces, and its usefulness to the Crown proved critical to its persistence through the medieval period (Boucoyannis Reference Boucoyannis2015). However, neither geography nor state capacity can explain the transition from monarchical to representative government after the Glorious Revolution. Representative government arose in England only after centuries of “no standing army,” “no absolutism,” a unitary state, and parliaments.

A third important factor is taxation. The early modern English state was highly dependent on revenue from trade, especially wool and cloth exports. Whenever it wanted to raise rates, it “had to make concessions to the magnates” and be “accountable to the squires, urban representatives, and merchants” (Levi Reference Levi1988, 112). As a result, the Crown became increasingly dependent on Parliament. Attempts to reassert its autonomy first led to the Civil War and then to the permanent diminution of its authority in the Glorious Revolution. Thus, contestation over taxation is critical to explaining change (see also Ertman Reference Ertman1997, chap. 4). It is not, however, causally sufficient. States have always struggled to raise revenue from elites, but rarely has that struggle resulted in representative government. Even when opposition coalesces around a parliament, authoritarian leaders have many tools with which to subvert it. For centuries, hybrid regimes have been able to win over recalcitrant elites by offering them either local autonomy or centralized patronage. Why was the English Crown unable to co-opt its elites into a stable authoritarian coalition, and what drove elites to demand representative government?^{Footnote 1}

Weingast’s Theory

Weingast (1997; see also 2005) offers an institutionalist alternative to structural approaches. In his model, a state, or “sovereign,” divides the economic surplus between itself and two groups of citizens. The state can increase its payoffs by transgressing the rights of citizens, and it can defeat a challenge from one group of citizens. If the two groups cooperate to resist the state, they will overcome it, but their cooperation is not assured. This game has three plausible equilibria. First, the state can ally with one group and transgress against the other; this is typical under authoritarianism. Second, if the state does not expect the two groups ever to coordinate, it will transgress against both. Third, if the state expects that the two groups will coordinate in response to any transgression, regardless of which group it targets, it will refrain from transgressing against either. According to Weingast, this last equilibrium establishes the “foundations of democracy and the rule of law.” This is not easy to achieve: it requires that the two groups agree to what constitutes a transgression and to respond to any transgression jointly, even if it is directed at only one of them. Only under unusual circumstances will they be able to reach such an agreement. If the opportunity arises, however, they can negotiate terms in the form of a constitution that explicitly delineates the powers of the sovereign. This new consensus will be “self-enforcing,” in that it effectively deters the sovereign from further transgressions.

Weingast’s theoretical framework has several critical flaws. First, it requires implausible and ad hoc assumptions about the willingness of citizen groups to commit to each other’s defense. These groups start out as rivals: the state allies with one and transgresses against the other. At some point, the state gets greedy, and it transgresses against its erstwhile ally. This opens a window of opportunity. If the two groups coordinate, they can negotiate a constitution that specifies the conditions under which they will unite against future transgressions. According to Weingast, the state will be properly chastened, and the new equilibrium will be stable. But why should it be? The new equilibrium does not change the character of the state: it deters, not reforms, the state. A clever sovereign, acting strategically, has every incentive to encourage one of the two groups to defect from the new constitutional settlement. It can try to make a new ally of its former victim, or it can recommit to its former ally. In either case, it can offer that group of citizens a larger share of the economic surplus than the status quo, and it can obtain that surplus by transgressing against the other group. A group that accepts this offer can be confident that the sovereign will fulfill its end of the bargain for the same reason that the initial arrangement should have been stable: a state that transgresses simultaneously against both groups risks revolution.

If solving the coordination problem between citizen groups is not sufficient to create a stable constitution, the only way to induce lasting cooperation between them is to add some factor that is exogenous to the model. In a discussion of the “implications of the model,” Weingast (1997, 251), writes that “limits become self-enforcing when citizens hold these limits in high enough esteem that they are willing to defend them.” Subsequently, he emphasizes the importance of “citizen values” for maintaining democracy. Whence such esteem and such values? The suggestion that actors in a rational choice model should have non-instrumental preferences over strategies should set off alarm bells. In fact, the model implies the opposite conclusion. It assumes that the two groups have divergent interests and preferences (248), despite their momentary willingness to coordinate against the state. They were rivals in the past, when the state favored one group to the detriment of the other, and their divergent interests and preferences ensure that they will remain rivals in the future. Normative consensus is therefore unlikely. “Esteem” and “values” are a deus ex machina, an ad hoc device to bridge a fundamental disconnect in the argument.

Another lacuna in Weingast’s model is its failure to establish any logical relationship between constitutional settlements and the “foundations of democracy.” The basic concern of the two citizen groups is to prevent the state from transgressing their rights. Those rights need not be democratic; indeed, for the sake of the model, Weingast operationalizes them as the terms of division of an economic surplus. When citizen groups coordinate against the state, they are simply trying to redistribute payoffs, not create a specific regime type. Weingast assumes that their constitutional settlement will be democratic, but this assumption has no theoretical basis. To be “self-enforcing,” the constitutional settlement need only specify the conditions under which citizen groups will challenge transgressions. Perhaps democratic rights are uniquely suited for solving coordination problems, but perhaps not. In a section on ethnic conflict, Weingast argues that the Netherlands has attained constitutional stability through the principle of proportionality, which distributes resources to different groups according to their population. This is not a specifically democratic right; it is illiberal. Nevertheless, it is “a constructed focal point from which deviations are easy to police” (1997, 257).

A final problem with the model is its neglect of agent motives. Weingast assumes that a state’s decision to transgress the rights of its citizens is limited only by the threat of rebellion. In reality, states come in different types, with different underlying preferences. One sovereign might have goals that can only be achieved through the expansion of its constitutional authority; another might have goals that necessitate concessions. The latter type may be rare, but if one does come to power, as occurred in the Glorious Revolution, its actions may have far-reaching consequences.

Weingast’s Empirics

The flaws in Weingast’s theory have critical implications for his interpretation of the Glorious Revolution. He begins by describing its citizen groups: “By the end of the [seventeenth] century, two coalitions had formed, called Tories and Whigs. The latter were more focused on commercial activities. They favored secure property rights, low and stable taxes on economic activity, and … explicit limits on sovereign behavior.” Tories were on the opposite side of these issues and allied with the Crown. Throughout the seventeenth century, the Crown “transgressed significant rights of the Whigs while retaining support of the Tories” (Weingast Reference Weingast1997, 252).

What altered this equilibrium was royal overreach. First, James “sought to raise a standing army in time of peace without Parliamentary consent” (Weingast Reference Weingast2005, 100), an “intelligent gamble” to make himself “one of the most powerful monarchs in Europe” (Weingast Reference Weingast1997, 252n17). Then, he disenfranchised Tories and Whigs alike in his Parliament-packing scheme of 1687. A “united political nation” (253) rebelled against him, and he fled the country. Whigs and Tories then negotiated over the terms of a constitutional settlement. They reached a “new focal consensus about rights” that “listed a set of actions that any future king who failed to honor risked causing a coup” (Weingast Reference Weingast2005, 101). These provisions, codified in the Bill of Rights, created a stable coordination equilibrium. If the state transgressed against either Whigs or Tories, both would retaliate, and shared expectations of this outcome deterred defection by the state. Thus, England’s new constitution was self-enforcing. Per Weingast (101), “The clear rules concerning parliamentary legislation created for the first time an explicit separation of powers system, and with it, a huge change in the incentives of government.”

In this account, the settlement was a novel solution to the problem of state transgression. Yet support for “explicit limits on sovereign behavior” (Weingast Reference Weingast1997, 252) was not new, nor was the Glorious Revolution the first time that the English had imposed such limits. A recognizably Whiggish point of view, lacking the appellation but committed to the same set of principles, coalesced in the late 1620s in Parliament (Cogswell Reference Cogswell1989; Cust Reference Cust1987; Sommerville Reference Sommerville1999). After several years of frustration with this faction’s demands, Charles I decided in 1629 to govern without Parliament. For eleven years of “personal rule,” the king committed numerous abuses of royal power, and public discontent grew. In 1640, a rebellion in Scotland, combined with a desperate shortage of revenue, left Charles with no choice but to call on Parliament. By then, however, the king had alienated all but his closest allies. Some elites, much like the later Whigs, sought to create new institutional constraints on the power of the crown; other elites, much like the later Tories, simply sought to enforce existing boundaries. The former faction held the upper hand in Parliament, and it pushed hard for royal concessions. The result was a series of laws that sharply limited the prerogative powers of the Crown (Adamson Reference Adamson2007, 505–6).

The laws created by the Long Parliament before the Civil War constituted no less of a constitutional settlement than the Bill of Rights in the Glorious Revolution. Indeed, much of the Bill of Rights echoed grievances raised by Parliament against Charles in 1640 (Murrell Reference Murrell2017, 45; Schwoerer Reference Schwoerer1981, chap. 4). As De Krey (Reference De Krey2008, 764) writes, “The English state of 1689 … reflected the goals of parliamentary government that had been advocated over the previous fifty years.” It is problematic, therefore, that Weingast and other prominent scholars (Pincus Reference Pincus2009; Pincus and Robinson Reference Pincus, Robinson, Galiani and Sened2014) neglect English history before the Restoration of 1660. Rather than focusing narrowly on the case of “success” in the Glorious Revolution, we should be asking a broader question: What was it about the structure of English society that predisposed it toward representative government in both 1640–42 and 1688–89? The answer, as explained next, lies in the structure of its rural economy.

Political-Economic Foundations

Political leaders from time immemorial have created patronage networks that are deeply invested in the survival of their regime. Sometimes these networks break down, but rarely does that result in representative government. Instead, a new coalition of rent-seekers, no less authoritarian than the last, rises to wrest control over the state. Such was often the case in early modern Europe. Dynastic intrigues and aristocratic rebellions led not to regime change but to the replacement of one monarchical government by another.

What made England different? Weingast (1997, 252) notes that the reform-minded Whigs were “more focused on commercial activities” than Tories; however, this is a structural factor, exogenous to his model. Moreover, the importance of merchants should not be overestimated. Seventeenth-century England was not a sufficiently commercialized economy for merchants to have constituted the bulk of opposition to the Crown. England’s elite was more diverse than that of most other countries, but it had a clear hierarchy of power and prestige that favored land. At the start of the Long Parliament in 1640, only 64 of 527 members elected to the Commons were merchants (Antler Reference Antler1972, 155). Even in the eighteenth century, landowners “dominated Parliament” and “controlled with the aid of the crown’s patronage almost the whole policy-making and executive machinery of central and local government” (Mingay Reference Mingay1963, 10–11; see also ibid., 3–5, 12–13; Hoppit Reference Hoppit and Hoppit2003, 84). In the seventeenth century, no resistance to the Crown could succeed without the organized leadership of the landed gentry. Thus, the earlier question must be revised: What made rural England different?

The connection between rural structure and representative government is at the heart of one of the most influential works of historical sociology, Barrington Moore’s Social Origins of Dictatorship and Democracy. Moore (Reference Moore1966, 430, emphasis deleted) argues that one of the crucial “prerequisites” of democracy is “a turn toward an appropriate form of commercial agriculture.” Specifically, agriculture must be capitalist: it must rely on free labor that is paid wages at a market rate. When the rural sector is freed from traditional social hierarchies, economic modernization leads to democratization. In contrast, when agriculture is labor-repressive, rural elites are “likely to fuse with the monarchy at a later point in search of political support” (435).

The logic underlying Moore’s argument has recently resurfaced in a vibrant strand of research on “economic autonomy” (Arriola Reference Arriola2013; Bellin Reference Bellin2002; Handley Reference Handley2008; McMann Reference McMann2006; Radnitz Reference Radnitz2010). Defined as “the ability to earn a living independent of the state” (McMann Reference McMann2006, 28), economic autonomy has important political implications. When actors receive special privileges from the state, such as a legal monopoly on the importation or production of a particular good, they are keen to maintain the political status quo: any change in leadership would threaten their interests. In contrast, when actors are economically autonomous, they may have much to benefit from reform. A regime that maintains traditional hierarchies and favors well-connected clients will limit their opportunities. Their interest is in equal rights under the law, open access to markets, and voice in the political system—if not for all citizens, then at least for all elites. Above all, they would benefit from a system that allows for the formation of coalitions of economically autonomous outsiders against privileged, rent-seeking insiders.

What regime, then, should economically autonomous elites prefer? A second literature, one that compares opportunities for rent-seeking across regime types, provides an unambiguous answer (Ekelund and Tollison Reference Ekelund and Tollison1981; Lake Reference Lake1992; McGuire and Olson Reference McGuire and Olson.1996; Wagner Reference Wagner1966). Although its focus is on authoritarianism and democracy, its theoretical logic applies equally well to monarchical and representative government. It suggests that representative government has three key advantages. First, it entails electoral competition, which, even with a narrow franchise, encourages politicians to favor the interests of the majority of voters against small numbers of rent-seeking regime insiders. Second, it has low barriers of entry for regime outsiders, which increases the likelihood that political entrepreneurs will try to mobilize voters on the basis of their broad-based but diffuse interest in economic autonomy. Finally, it involves many politicians in policy making, which increases the transactions costs of rent-seeking and therefore decreases its viability. These claims are supported by an enormous literature, both qualitative and quantitative, on the negative relationship between electoral competition and corruption. Representative government is no panacea, but for elites who are threatened by rent-seeking and state predation, it is far better than monarchical government. Thus, I hypothesize that economically autonomous elites constitute a key constituency for representative government.

Economic Autonomy in Rural England

England’s transition to agrarian capitalism began much earlier than that of any other country (Dimmock Reference Dimmock2014, 205–32; van Zanden Reference van Zanden and Prak2000, 84–85; Wallis, Colson, and Chilosi Reference Wallis, Colson and Chilosi2018). During that transition, English landowners consisted of two ideal-types: those whose relationship with their tenants was based on traditional forms of land tenure and manorial custom, and those who leased their land to yeoman farmers at market rates under the jurisdiction of common law (Brown Reference Brown2015, 169–71). Over time, the latter would become dominant. In the seventeenth century, however, the country was more evenly divided between the two.

The transition to agrarian capitalism had important implications for state and society. As the gentry enclosed their fields and relinquished their feudal rights, they altered the political conditions for the accumulation of their wealth. They “ceased to rely for their survival on the exercise of extra-economic forms of surplus extraction, such as the direct use of force over the peasantry (as in France),” in favor of “a system of property relations in which landlords relied upon rents accrued from market-dependent commercial tenants” (Kennedy Reference Kennedy2008, 88–89). With the rise of market compulsion, their need for state coercion declined. As a result, landed elites became increasingly economically autonomous. Economic autonomy, in turn, was a permissive condition for political autonomy. Individuals who opposed any aspect of royal policy, such as religious edicts or foreign affairs, needed to be autonomous of the Crown to be able to resist its authority.

Economic autonomy was also an active cause of landowner support for representative government. Beginning in 1629, Charles ruled without parliaments and, consequently, without parliamentary grants of taxation. To raise revenue and increase his authority, he undertook several initiatives that threatened the property rights of agrarian capitalists. In each case, they had good reason to believe that Parliament would better represent their interests and enact reforms that would ensure their autonomy. First, many improving landlords owned property that had been seized from monasteries in the mid-sixteenth century. The Crown sold that land under a tenurial status, “knight-service,” that carried certain feudal obligations. These obligations no longer served any pressing societal need, but the Crown could exploit them to raise revenue. This became a major point of contention between Crown and Parliament in the late sixteenth and early seventeenth centuries (Bell Reference Bell1953, chap. 7; Reid Reference Reid1995, 238–42).

Second, and also pertaining to the sale of monastery land, many improving landlords owned impropriations, the right to collect tithes to provide for local members of the clergy. Charles and Archbishop Laud sought to recover these impropriations and raise tithe rates, the effect of which would have been to undermine property rights and erode profits from agricultural improvement (Davies Reference Davies1992, 85; Hill Reference Hill1956, 153–59, 340–43; James Reference James1941, 3–7). At stake was not only landlords’ economic autonomy but also their social autonomy, in the form of local control over the Anglican Church. Third, Charles revived sixteenth-century statutes against enclosure and depopulation. He appointed Laud to head commissions to investigate landowners who had converted tillage to pasture and then used prerogative courts to coerce composition fines from violators. Commissioners were more interested in raising revenue than reversing agrarian capitalism, but their actions nevertheless constituted an attack on property rights and were deeply unpopular (Beresford Reference Beresford and Fisher1961; Reid Reference Reid1995, 258–59; Tate Reference Tate1967, 124–27; Tawney Reference Tawney2015, 176–77).

A final contentious issue was legal reform. Alongside the common law, the Crown maintained a separate system of prerogative courts under its own control. It used these courts, which included the notorious Star Chamber, to enforce its “fiscal feudalism” and persecute those who challenged its authority (Alford Reference Alford2011; see also Prest Reference Prest1992). Capitalist landlords not only opposed these exactions but also benefited from the strengthening of common law. To make their holdings profitable, they had to overcome customary restrictions on the use and transfer of land (e.g., French and Hoyle Reference French and Hoyle2007, 172, 294–95). Common law was an important means to this end: it enabled lawsuits and agreements that extinguished customary rights and allowed landlords to buy, sell, and rent their land at market prices (Brooks Reference Brooks and Whittle2013, 198–99; Hoyle Reference Hoyle1987, 48–51; Simpson Reference Simpson1986, chap. 7).

Traditional manorial landlords, in contrast, had no such stake in common law. They had tenants with customary tenures, and disputes over such tenures were adjudicated in the court of the manor. In that court, which stood apart from common law, a representative of the manorial lord served as judge, and the lord appointed jurors from village elites (Brooks Reference Brooks2009, 251–54, 345–48). As a result, manorial lords had significant influence over court proceedings. They could not force tenants to give up their customary rights, but they had every incentive to exert pressure in favor of their own interests when the legal situation was ambiguous (e.g., French and Hoyle Reference French and Hoyle2007, 145–53, 169–71). Critically, manorial courts were part of the royal jurisdiction (Brooks Reference Brooks2009, 248). Though not under the direct control of the Crown, they owed their legal legitimacy to it. Thus, traditional manorial landlords were locked into an institutional framework that depended on a defense of the monarchy and traditional social order (Anderson Reference Anderson2005, 249).

Bases of Civil War Allegiance

To assess the logic of economic autonomy, I focus on the outbreak of the English Civil War. This case provides an ideal test of the theory, a natural experiment of sorts. Between Charles’s flight from London in March 1642 and the start of fighting in October 1642, no territorial line divided the two sides. Only their respective capitals were their exclusive domain: Oxford for the Crown and London for Parliament. Thus, activists from each side were free to canvass for their cause throughout the English countryside in the spring and summer of 1642. Per Newman, “Neither the King nor the Parliament possessed the means of coercion: they were reliant upon argument and exhortation until their supporters were animated enough to provide the men and the means to translate the political stalemate into armed confrontation” (1993, 257–58).

A number of idiosyncratic factors affected partisanship, including ideological commitments, personal loyalties, and pressure from family members and social superiors (Newman Reference Newman1993, chaps. 1–2). If these factors were sufficient to explain how individuals chose sides in the Civil War, there would be no reason to expect geographic clusters of alignment. Instead, partisanship would be distributed randomly. Not until after October 1642, when the rival armies went on the march and began to establish exclusive zones of administrative control, should distinct patterns have emerged on the map. Conversely, if the Civil War had structural causes, partisanship should have been distributed nonrandomly. That is, individuals’ consciences should have been biased by their material interests.

Of course, not all structural causes are economic. Religion, too, demands consideration. Since the reign of Elizabeth, the Church of England had accommodated a wide range of practices and theologies, from conservative “high church” Anglicanism, which did not stray far from Catholicism, to the more radical “low church” Puritanism, which was based on Calvinism. Charles cared little about doctrinal controversies, but as the divinely anointed head of the Church, he viewed religious nonconformity as a direct challenge to his rule. Thus, “the king marked out evangelical Puritanism as a popularist threat to hierarchy and authority, which elevated individualism over deferential principles” (Davies Reference Davies1992, 10). Charles and Archbishop Laud imposed new rules for services and the sabbath, persecuted ministers who refused to conform, and required that Scots use the same prayer book as the English. In response, England polarized and Scotland rebelled. At the outbreak of the Civil War, most high church Anglicans were Royalists, most Puritans were Parliamentarians (Blackwood Reference Blackwood1978, 63–65; 1996, 203–4; Reference Blackwood2001, 159–60; Cliffe Reference Cliffe1969, 343–49; Fletcher Reference Fletcher1975, 231–63; Hunt Reference Hunt1983; Stoyle Reference Stoyle1994, chap. 10), and Scotland allied with Parliament.

Religious conflict was critical to the outbreak of the Civil War, but there was more to it than that. First, religious resistance depended on economic autonomy. Whatever their motives, individuals would be reluctant to oppose the Crown if their wealth and status depended on its favor and authority. Second, the established church was an integral part of royal power. Toward both Church and Parliament, Charles’s goal was the same: to assert his divine right and to ensure the obedience of his subjects (Davies Reference Davies1992, 11–18, 299–301). For him and his loyal followers, “Puritan” became an epithet that they applied indiscriminately, regardless of religious belief, to whomever opposed him. For resisters, in contrast, “Protestantism was synonymous with putting down the over-mighty clergymen, with keeping ... free of Roman-Spanish control, and, best of all, … with preserving the freedom of men of property from the interference of a centralizing government” (Finlayson Reference Finlayson1983, 118; see also ibid., chap. 2; Adamson Reference Adamson2007, 35, 518). Thus, religious policy and secular politics were inseparable.

Finally, and most importantly, religious belief was endogenous to economic change. Puritanism, a “hotter sort of Protestantism,” propagated unevenly through the English countryside in the late sixteenth and early seventeenth centuries. Capitalist landlords were drawn to the individualism and egalitarianism of Calvinism, whereas traditional landlords held to a high church doctrine that legitimated customary social and economic hierarchies. In regions with enclosed fields, as was typical of capitalist agriculture, reformist ideas spread rapidly; in regions with communal field systems, as was typical of traditional manorialism, they did not (Hopcroft Reference Hopcroft1997; Zaret Reference Zaret1985, 47–53). Thus, the rise of capitalism fed the rise of Puritanism (Tawney Reference Tawney2015). Religious belief mattered greatly, but it was not causally independent. To explain the structure of allegiance in the English Civil War, one must begin with capitalism.

My interpretation owes much to the insights of political Marxist historians and historical sociologists. They argue that the interests of capitalists, both urban and rural, were in direct conflict with the interests of the Crown and its dependent rentiers (Brenner Reference Brenner1993; Dimmock Reference Dimmock2014; Hill Reference Hill1980; Kennedy Reference Kennedy, Haynes and Wolfreys2007; Lachmann Reference Lachmann2000; Stone Reference Stone1965). In this view, the Civil War was a clash between two different types of elites: one that derived its wealth from rents, or “politically constituted property,” and one that derived its wealth from open market competition. However, none of these scholars undertakes a broad empirical analysis of the economic interests and political allegiances of landowners in rural England. This is a problematic omission. One cannot simply assume that agrarian capitalists around the country sided with Parliament against the Crown; evidence is needed.

Operationalizing Economic Autonomy

The independent variable of this study, the economic autonomy of rural elites, is not easy to operationalize. One approach is qualitative, based on written records and archaeological findings. This kind of analysis makes a distinction between “wood pasture” and “open pasture.” In the former, fields were enclosed, marked out by lines of trees or hedges, and customary rights had been extinguished. In the latter, fields were unenclosed, and tenants still had rights to the commons. The other method to operationalize economic autonomy is quantitative: it assembles data on farm size, labor productivity, and occupational types. These metrics provide valuable information because agriculture is subject to economies of scale. Under the pressure of market discipline and the need for productive efficiency, capitalist farms tended to be larger in size and hire fewer laborers per acre, whereas customary farms had a high proportion of husbandmen to landless laborers.

Both qualitative and quantitative measures reach the same conclusion: land use in the English countryside differed sharply by region. The north and west, or “highland zone,” was mainly open pasture and sheep-corn farming; the south and east, or “lowland zone,” was mainly wood pasture and dairying (Thirsk Reference Thirsk and Thirsk1967, 4; see also Everitt Reference Everitt and Thirsk1967, 462–64; Hopcroft Reference Hopcroft1999, chap. 4; Kitsikopoulos Reference Kitsikopoulos2000). The underlying cause of this variation was the uneven progress of the agrarian transition: “In 1700 small-scale agrarian capitalism predominated in the south-east…. In contrast, in the north-west family farms continued to predominate in 1700” (Shaw-Taylor Reference Shaw-Taylor2012, 57–58). Numerous studies of individual counties and regions reach the same conclusion, contrasting the backwardness of rural northern England (Appleby Reference Appleby1975; Gregson Reference Gregson1989; Searle Reference Searle1986), Wales (Dodd Reference Dodd1952, 14–29), and Cornwall (Coate Reference Coate1963, 11–14) with the rapid progress of agrarian capitalism around London and in East Anglia. Thus, if economic autonomy determined allegiance in the Civil War, the country should have divided between a Parliamentarian southeast and a Royalist northwest.

This hypothesis requires several caveats. First, patterns of land use do not always reflect the organization of agricultural production (Davie Reference Davie1991). I employ land use as an imperfect indicator of economic status, not as a causal variable in its own right. Second, the southeast and northwest were not internally homogeneous (Hochberg Reference Hochberg1984; Holmes Reference Holmes1974, 10). Pockets of customary agriculture remained in parts of the southeast, and capitalist agriculture made some inroads in parts of the northwest. Where possible, I exploit this variance to generate additional tests of the theory. Third, no administrative boundary or geographical feature sharply divides the southeast from the northwest. Where the two regions meet, in the English midlands, was an intermediate zone for the social organization of agricultural production. As such, it should have held an intermediate position on the political spectrum, either more internally divided or more determinedly neutral than the southeast and northwest.

Patterns of Allegiance in the Civil War

The political geography of England at the outset of the Civil War confirms my hypothesis. In March 1642, when mobilization orders from both sides were delivered to county elites, they did not land on a blank slate of public opinion. Instead, economic interests shaped political attitudes. By October 1642, when the first shots were fired, there had emerged a clear pattern of regional alignment: “Support for Parliament came from the economically advanced south and east of England, the King’s support from the economically backward areas of the north and west” (Hill Reference Hill1980, 119–20). The best evidence on the north of England comes from Newman (1993, 266), who has compiled data on Royalist regimental colonels. He concludes, “Yorkshire, the north-east, and the north-west … [were] a Royalist stronghold from 1642 to 1644.” Similarly, Morrill (Reference Morrill1979, 79; see also Phillips Reference Phillips1978), in a survey of research on the Civil War in northern England, finds that “the number of royalist gentry considerably outnumbered the number of parliamentarian gentry” in five of the six northern counties.

Royalism’s other stronghold was the west. Crown recruiters had some success in the western Midlands, particularly Herefordshire (Hutton Reference Hutton1999, chap. 1). The strongest response, however, came from Wales, where popular opinion was “solidly royalist,” with a “unique unity of political purpose in most communities” (Bowen Reference Bowen2007, 250, 253; see also Stoyle Reference Stoyle2005, 11–14, 27–29). Welsh soldiers thus formed the backbone of the Royalist infantry (Hutton Reference Hutton1999, 15–17). Finally, in the extreme southwest corner of England was Cornwall. It was not as well affected to the Crown as Wales, because it had suffered from disproportionately heavy taxation under the prerogative rule of Charles I. Nevertheless, “relations between landlords and tenants in Cornwall retained much that was patriarchal and medieval in character” (Coate Reference Coate1963, 8–9), like in Wales and northern England. Consequently, when forced to choose between royal traditionalism and parliamentary radicalism, most of the Cornish gentry cast their lot with the Crown (Duffin Reference Duffin1996, 193–98, 202; Stoyle Reference Stoyle2005, 32–33, 37–43).

In the south and east of England, the balance shifts. East Anglia, “the early improving region par excellence” (Bryer Reference Bryer2006, 376), was the heartland of support for Parliament (Blackwood Reference Blackwood1996, 197–98; Reference Blackwood2001, 140–43, 150–54; Everitt Reference Everitt1960, 11–15; Holmes Reference Holmes1974, 34–62; Hunt Reference Hunt1983, 294–95, 306–7). In the winter of 1642–43, Parliament sought to create four regional commands for its army. Only the Eastern Association, which comprised the East Anglian counties of Essex, Suffolk, Norfolk, Cambridgeshire, and the Isle of Ely, plus Hertfordshire, managed to overcome internal squabbling and form an effective military force. In 1644, it fielded the strongest army for Parliament, and in 1645, it became the core of the New Model Army (Everitt Reference Everitt1960, 16–17, 28–34; Holmes Reference Holmes1974, 1–2, 62–67). Holmes (Reference Holmes1974) attributes the success of the Eastern Association to the involvement of Parliament in its creation, but his account begs the question: What made East Anglian members of Parliament so willing to subordinate their local interests to the national cause (Smith Reference Smith1975, 713)? Their actions make sense only in the context of relatively high elite support for Parliament and opposition to the Crown.

In parts of the south and east where agrarian capitalism was less pervasive or a more recent arrival, Parliamentarian gentry sometimes found themselves in the minority (Fleming Reference Fleming1981, 30–31; Hughes Reference Hughes1987, 161–62; Stoyle Reference Stoyle1994, 138–41; Wood Reference Wood1937, 33–35). To secure their counties, they had to reach out to men of lower social standing. One such group was cloth makers, whom I address elsewhere (Reference NariznyNarizny n.d.). The most natural partner for the rural gentry, however, was yeomen farmers. As the middle stratum in the rural economy, they owed their prosperity to agrarian capitalism, and they had close ties to the owners of the estates on which they rented farmland. As a consequence, they were one of the most Parliamentarian of social classes in England (Blackwood Reference Blackwood1993, 112; Skipp Reference Skipp1978, 105–6; Underdown Reference Underdown1973, 39–40; Wood Reference Wood1937, 34; see also Manning Reference Manning1991).

Counties in the “intermediate zone,” where northwest met southeast, exhibited the most internal variation. As such, they provide fertile ground for theory testing. Few historians have undertaken the difficult task of trying to identify agricultural organization and political allegiance from village to village and manor to manor, but the few that have reveal striking patterns. The seminal work is by Underdown (Reference Underdown1985), who examines the southwestern counties of Somerset, Wiltshire, and Dorset. He finds that “the most solidly parliamentarian region was the dairying and cloth-making country of north Wiltshire and north Somerset,” whereas “support for the Royalists was strongest in the downlands, in Blackmore Vale and the adjoining parts of south Somerset, and in the hill country to the west” (165). Although Underdown does not use the term “agrarian capitalism,” it is the essential difference between these two political geographies. Royalist England was “roughly coterminous with the surviving areas of open-field arable husbandry” and “the old conception of the stable, harmonious village community based on deference”; Parliamentarian England was “the pasture and woodland areas linked to an expanding market economy” (18).

Research from several other counties confirms this conclusion. Warwickshire had two main regions: Feldon, in which “the nucleated village was the typical community; society was close knit, traditional and highly manorialized,” and Arden, a forest that was “slowly cleared and settled by individuals or families rather than by communities,” in which “a traditional open-field system had never existed” and “agricultural improvement was vital in maintaining a steady population increase” (Hughes Reference Hughes1987, 4). These economic differences had predictable political consequences: Feldon was Royalist, whereas Arden was Parliamentarian (151–52, 157). Gloucestershire was divided between the Vales of Berkeley and Gloucester, which had wood-pasture farms and dairying, and the Cotswolds, in which sheep-corn farming prevailed (Warmington Reference Warmington1997, 12–17, 26–27; see also Rollison Reference Rollison1992, 142, 147). Not surprisingly, “the bulk of the wealth and influence in the Vales was for Parliament” (Warmington Reference Warmington1997, 41), while “the Parliamentarians peter out as the Cotswold escarpment rises” (42). In Sussex, the west and coastal plain had mixed farming, while the east and upland weald had wood-pasture farms (Fletcher Reference Fletcher1975, 11–12). The result was a factional rivalry in which “the parliamentarian oligarchy imposed a firm hold on the eastern end of Sussex” (255), whereas the west was “a neutralist, even an incipiently royalist, countryside” (267). The only county to present “worrying inconsistencies” is Devon (Stoyle Reference Stoyle1994, 155). In that case, however, “the apparent correspondence between pastoralism and Parliamentarianism may well owe more to the local distribution of the cloth industry than to anything else” (156).

Another source of evidence is prosopographical research on changes in familial wealth. The typical rural capitalist gentleman, an “improving landlord,” enclosed and invested in his land to increase its value. Freed from the strictures of custom, he would charge the maximum rent that the market would bear. Over time, his income would increase, while that of traditional manors would stagnate or decline. Of course, individual circumstances vary: some capitalists might make unwise investments and ruin their fortunes, and some traditional landlords might derive income from other sources. Nevertheless, I expect an increase in the wealth of rural elites to be correlated with support for Parliament.

The available evidence supports this hypothesis. In the Long Parliament of 1640, 103 members who experienced “upward economic mobility” in the two decades before the Civil War became Parliamentarians, whereas only 39 in that category became Royalists. Furthermore, only 15 members with “downward economic mobility” became Parliamentarians, whereas 44 became Royalists (Antler Reference Antler1972, 155–56; Reference Antler1975). The same correlation holds at the local level. In Leicestershire, three of eight distinguished families of Royalists “were in financial difficulties by the time the war arrived,” whereas “of the families of equal antiquity who were rising to replace this redundant caucus, most were parliamentarian” (Fleming Reference Fleming1981, 31). In Cheshire, “there was a high incidence of Royalism … amongst those who can be shown to have been in financial troubles in the 1620s and 1630s” (Morrill Reference Morrill1974, 71; see also Dore Reference Dore1966, 21). In Yorkshire, “resistance to the Crown … was far more closely associated with growing prosperity than with economic decline.… Few [Parliamentarians] had court connections or offices of profit, but many were improving gentry who had increased their estates since the beginning of the century” (Cliffe Reference Cliffe1969, 360, 361; see also ibid., 333, 351–55). In Kent, Royalist activists “were either heads or younger sons of impoverished Kentish families, with little stake in the land themselves, and no interest in … careful estate management” (Everitt Reference Everitt1966, 101). In sum, wherever historians research this question, they reach the same conclusion: the rising, capitalist gentry favored Parliament, and the declining, traditional gentry favored the Crown.^{Footnote 2}

The Glorious Revolution

The Civil War did not result in a stable constitutional settlement. Three years into the conflict and desperate for it to end, Parliament reformed its military in ways that increased its battlefield effectiveness but weakened its political accountability. The New Model Army quickly defeated the Royalists; then it turned on its master. It purged opposition from Parliament, leaving only an unrepresentative “rump,” and its commander, Oliver Cromwell, became Lord Protector. Throughout the 1650s, England was in political disarray; the only source of stability was Cromwell’s control over the army. When he died in 1658, struggles for power among senior officers threatened a return to chaos. In 1659, therefore, General George Monck decided to put an end to the game. He marched his soldiers into London and imposed a return to the status quo ex ante bellum: the restoration of both the monarchy and Parliament. Charles I had been beheaded in 1649; thus, his son, Charles II, became king.

In the early years of the Restoration, fears of renewed conflict induced restraint on both sides (McInnes Reference McInnes1982, 379–81). As time passed, however, the relationship between Crown and Parliament became increasingly contentious. Religion played a particularly divisive role: Parliament sought to exclude Charles II’s younger brother, James II, from the line of royal succession on account of his Catholicism. Charles defeated this effort, and James became king on his brother’s death in 1685. Public opinion initially accepted the transition, but James declined in popularity as he sought to secure toleration for Catholics and Protestant dissenters. As resistance mounted, he undertook increasingly authoritarian measures to consolidate his power. James II adopted the opposite position of Charles I on religious dissent, so the 1680s was not simply a rerun of the 1630s, but religion and constitutional politics were once again intertwined.

Across the English Channel, the stadholder of the Dutch Republic, William of Orange, saw these events as both a crisis and an opportunity (Claydon Reference Claydon2002). At the time, the Dutch Republic faced an existential threat. France had nearly conquered it in 1672 and appeared likely to try again soon. William desperately needed an alliance with England, but James, as a committed Catholic and aspiring absolutist, favored France. How, then, to sway the English to the Dutch side? Fortunately for William, he had a dynastic connection to England. His mother was the daughter of Charles I, and his wife Mary was the eldest daughter of Charles II and, until the birth of James’s son in June 1688, next in the line of succession. William had no claim against James’s right to rule, but with the support of Mary and her sister Anne, he could represent himself as a legitimate stakeholder in the English constitution.

To save his homeland, William devised a daring plan: he would form a massive flotilla, load it with the Dutch Republic’s best troops, land in England, and march into London. It is not known whether William secretly intended to seize the throne from James or only to force him into submission, but his overriding purpose was clear: to bring England into war against France. To this end, William would not only have to outmatch James on the battlefield but also win over English public opinion. Thus, he mounted a sophisticated propaganda campaign that cast him as the defender of English liberties and Protestantism against royal despotism. When William landed at Torbay in November 1688, he had both military superiority and a substantial share of public opinion on his side. At first, James rallied his forces to fight, but desertions and low morale broke his confidence. In late December, he threw the royal seals into the Thames and fled to France, never to return.

Weingast (1997, 253; see also 2005, 100–1; North, Wallis, and Weingast Reference North, Wallis and Weingast2009, 186) represents these events as follows: “The Tories joined the Whigs to form a united political nation against the king, forcing him to flee and inviting in a new monarchy.” This interpretation is in error on two counts. First, neither Whigs nor Tories, much less a “united political nation,” forced James out. From the defeat of the Monmouth Rebellion in July 1685 until William’s arrival in November 1688, no one openly opposed James’s right to rule. Only a small number of plotters actively worked against him, and they did so surreptitiously. Their famous “invitation” for William to invade was not only anonymous but also unpublicized (Hosford Reference Hosford1976, 78–79). Moreover, the plotters in no way represented the “united political nation.” Instead, the “really substantial politicians whom James had offended declined to become involved” (Speck Reference Speck1988, 220). For the vast majority of the English ruling class, resistance to James was entirely passive: when asked to serve, they sat on their hands or dragged their feet (Hosford Reference Hosford2008). Only on William’s arrival in Torbay did disaffection with James come to a boil. It was not the English who forced out James II; it was William and his army.

Second, Tories did not support the revolution. Some welcomed William’s arrival, but they did so in the belief that he intended only to rein in James, not to usurp him. For Tories, the king’s flight to France was a “shattering blow” (Miller Reference Miller, Cruickshanks and Black1988, 14). The only evidence that one can adduce for Weingast’s claim that “the Tories joined the Whigs … [in] inviting in a new monarchy” is that Tories participated in the Convention of January 1689, which resolved to offer the crown to William. However, they sought to block that outcome (Beddard Reference Beddard and Beddard1991, 73–94; Horwitz Reference Horwitz1977, 10–11; Miller Reference Miller, Cruickshanks and Black1988, 15). After losing the vote, they accepted the majority’s decision, but they had little choice: James was already gone, England was under the occupation of a foreign army, and Whigs were enthusiastically collaborating with it. For Tories to have continued to oppose William’s rule would have been politically suicidal (Beddard Reference Beddard and Beddard1991, 92). They did not “unite against” James and compromise with the Whigs: they were defeated (Beddard Reference Beddard and Beddard1991; Cruickshanks Reference Cruickshanks2012, 120–21; Miller Reference Miller, Cruickshanks and Black1988, 18–19). This is apparent in Weingast’s (1997, 252) own characterization of the two factions’ policy preferences. On all four of the issues on which they differed—commerce, taxation, foreign policy, and royal prerogative—Whigs achieved, and Tories lost, their desiderata. Moreover, with the Toleration Act of 1689, Whigs broke the monopoly of the Tories’ beloved Church of England. This law not only dampened a long-standing source of societal conflict but also weakened the traditional social hierarchy and undermined an important instrument of monarchical power.

In sum, the revolution settlement was not a negotiated compromise between Whigs and Tories, nor was it a clever solution to a coordination problem. Rather, it was imposed on Tories by the Dutch and Whigs. It resolved instability in the English constitution not because it created novel institutions, but because it changed the political balance of power. That change had both agential and structural causes. On the side of agency, the preferences of the monarch mattered tremendously. James’s overriding goal, toleration for Catholics, required him to subordinate Parliament to the Crown. William’s overriding goal, to defeat France, required him to subordinate the Crown to Parliament (Claydon Reference Claydon2002; McInnes Reference McInnes1982, 390, 392; Pincus and Robinson Reference Pincus, Robinson, Galiani and Sened2014). Only with the enthusiastic support of Parliament would William be able to mobilize England effectively for war. Whigs, who had economic and sectarian interests in a conflict with France, were willing to provide this support, but it came at a price: they insisted that Parliament meet frequently, supervise every expenditure, and take the initiative in legislation. William resisted the erosion of his prerogative powers, but whenever he had to choose between antagonizing Parliament with a defense of his constitutional rights and placating Parliament to ensure funding for war, he chose the latter. William’s successor, Anne, had no ambition to retake the initiative, and her energies were sapped by severe health problems (McInnes Reference McInnes1982, 391). By the time of her death in 1714, representative government was firmly entrenched.

On the side of structure, there is a clear continuity over time in the ideological character and coalitional basis of support for Parliament. It arose in opposition to royal prerogative in the late 1620s, rallied against Charles I in the Civil War, sought to protect the privileges of Parliament during the Restoration monarchies of Charles II and James II, and insisted on constraining William in the Glorious Revolution and its aftermath. Support for Parliament did not come from an ever-changing coalition of opportunists seeking the spoils of office. Instead, it was a stable feature of the English political system over nearly a century, excepting only the chaotic Interregnum.

A limiting factor in this account is a lack of direct evidence about the socioeconomic basis of the Glorious Revolution. The type of analysis conducted for the outbreak of war in 1642 is simply not possible for 1688. Unlike the Civil War, which was preceded by six months of nationwide mobilization, lasted more than three years, and subsequently reignited twice, the Glorious Revolution featured a confrontation between two standing armies that lasted only a few weeks. In terms of public opinion, one can find scattered reports of declarations of support for William, but that is all. The most detailed of the few local studies of the Glorious Revolution is unable to provide relevant information on more than a handful of individuals (Hosford Reference Hosford1976). Even if better data were available, however, it would not be dispositive. During the initial phase of the Glorious Revolution, in late 1688, there was little reason for England to divide along partisan lines. William represented himself not as a claimant to the throne but as the savior of Protestantism and a defender of traditional rights. He did not call for constitutional reform, much less representative government, but rather for James to accept the supremacy of the Church of England and convene a session of Parliament. As long as he adhered to this line, he appealed to Whigs and Tories alike. Only once the unthinkable happened—James fled and William demanded the throne—did a fundamental break from the Restoration settlement become possible. At that point, as demonstrated in the foregoing discussion of the Convention of January 1689, England divided predictably.

Despite the inherent limits of the available evidence, a comparison between 1642 and 1688 is suggestive. In 1642, the economically autonomous elite consisted mainly of agrarian capitalists, allied with a small but wealthy set of unregulated industries and unchartered merchants. Opposing them were traditional manorial landowners, allied with a small but wealthy set of regulated industries and chartered merchants. The two sides were relatively evenly balanced; thus, both chose to fight rather than back down. As it happened, Parliament was strong enough to win the Civil War, but only barely so, and it lost control to the New Model Army. By 1688, much had changed. The conversion of traditional agriculture to agrarian capitalism had continued apace, encouraged by policies enacted by Parliament during the Civil War and Interregnum (Brooks Reference Brooks and Whittle2013; Hill Reference Hill1997, chap. 5; James Reference James1930, 78–90, 343), while unregulated industries and unchartered merchants expanded rapidly (Brenner Reference Brenner1993, 711–13; James Reference James1930, 131–82, 342). As a result, the economically autonomous elite became an increasingly dominant majority.

This shift in the societal balance of power may explain why James II in 1688 behaved so differently than Charles I in 1642. With London controlled by hostile forces, both kings had the option of heading north to raise support and “force the English ruling classes to choose sides in a prolonged civil war” (Pocock Reference Pocock and Schwoerer1992, 57–58). Yet only Charles did so. Perhaps James was less resolute than his father, another notch in the column of agency. Certainly the presence of an army of Dutch veterans, a factor with which his father never had to contend, weighed heavily. However, James must have realized that William could not afford a long conflict over England, one that would drain the resources of the Dutch Republic and leave it exposed to France. Perhaps, then, the decisive difference between 1642 and 1688 was the changed societal balance of power. In 1642, Charles could count on a large reserve of traditional manorial landlords to rally to him, because England’s transition to agrarian capitalism was far from complete. In 1688, with the transition more advanced, James could not.

English society did not stand still after the Glorious Revolution. Instead, the political triumph of economically autonomous elites generated a positive feedback loop. As Stasavage (Reference Stasavage2002) demonstrates, the newly dominant Whigs used Parliament to reform the economic and legal system. This contributed to advances in agrarian and merchant capitalism, which eventually sparked the Industrial Revolution and provided abundant opportunities for the further enlargement and enrichment of the economically autonomous elite. Throughout the eighteenth century, a period of massive expansion in economic legislation, Parliament “remained extremely cautious about enhancing the interests of one person or group, for fear of eroding interests elsewhere” (Hoppit Reference Hoppit and Hoppit2003, 515). Thus, representative government and economic autonomy continued to reinforce each other over time.