The Glorious Revolution and Access to Parliament

Abstract

This paper shows that the Glorious Revolution of 1688 broadened access to Parliament for families needing rights to sell land in so-called estate bills. Bills were on average 14–27 percentage points more likely to be for gentry families and not aristocratic families in legislative sessions after the Revolution compared to sessions before. Regression and archival evidence suggest that parliamentary certainty was primarily responsible for improved access by altering families’ entry calculus and brokers’ recruitment of new business. More broadly, the paper provides insight into the ways in which political institutions affect access to and the provision of property rights.

The Glorious Revolution of 1688 resulted in the overthrow and exile of James II, who was replaced by William and Mary. The Revolution’s political and economic significance for England, and eventually Britain, is the subject of a large debate. The original, widely contested North-Weingast credible commitment theory predicted an improvement in the security of domestic property rights. The evidence generally indicates that such rights were secured long before 1688 (Clark 1996; Epstein 2000; Quinn 2001; Hoppit 2011; Jha 2015).¹ Cox re-formulates the North-Weingast framework and predicts the Revolution’s effect on property rights may depend on the degree of the sovereign’s influence over their provision prior to 1688 (Cox 2016, ch. 6). Consistent with this argument, some scholars have shown that rights in the domains of transportation and trade, where the monarch had significant influence before the Revolution, improved after 1688 (Bogart 2011; Pettigrew 2014).² However, recent studies using time series measures and techniques conclude the Revolution was unimportant for political institutions and lower-level legal institutions (Murrell 2017; Henriques and Palma 2022; Grajzl and Murrell 2021).

By studying the families that sought, and Members of Parliament (MPs) that worked on, so-called estate bills and acts of Parliament, this paper shows there was a measurable change in novel outcomes related to parliamentary organization and property rights after 1688. Estate bills dealt with families’ property rights to land, a fundamental asset in the preindustrial era. Most estate acts, which were bills approved by Parliament, removed legal restrictions created by entails in marriage and inheritance contracts (called family settlements) and allowed families to sell land. One interpretation is that the acts were economically important because they helped facilitate land market activity by lowering the transaction costs created by settlements (Bogart and Richardson 2009, 2011). They are also an excellent subject for a study of the effect of the Revolution because they were the single largest category of legislation during this period, which allows for a micro-level quantitative analysis. While previous work has documented that the Revolution improved the likelihood of estate bill success (passage by Parliament) and that Parliament worked on more bills after 1688, estate bills are generally missing from narratives of political change during this period.³

I use detailed evidence from the parliamentary record—titles of bills (projects introduced to Parliament) and acts (bills that passed the parliamentary process), the Journals of Parliament, and an archival sample of acts—as the main sources for the investigation. Using the titles of families named in bills, I first document that there was a broadening of access to gentry families, those referred to as esquires and gentlemen in legislative records, after the Revolution.⁴ There were an average of eight additional bills for gentry families per session after 1688. There was no change in the average number of bills for aristocratic families, as measured by families with noble titles or referred to as knights and baronets.

Second, I provide evidence on the mechanisms by which the Revolution broadened access for gentry families. A leading hypothesis is that the Revolution may have lowered the threshold of entry that families faced by making parliamentary meetings more certain after 1688. It is well documented that after 1688, Parliament met annually, its legislative calendar became standardized, and its legislative sessions were significantly longer on average (Hoppit 2017, ch. 2). Relatedly, more certainty and less royal influence in Parliament after 1688 may have facilitated the recruitment of new business by brokers, like MPs or other parliamentary agents who received fees from working on legislation (Cox 2016, ch. 6). Last, other factors, such as financial crises and opportunities or new types of family settlements, may have pushed more gentry families to access Parliament after 1688.⁵

I use probit regressions and draw on information from an archival sample of estate acts to provide evidence for these hypotheses. The weight of the evidence suggests that, while financial crises played some role in pushing families to Parliament, parliamentary certainty was central to gentry families’ access to Parliament in two ways: altering families’ entry calculus and brokers’ recruitment of new business. Probit models show bills were on average 7 to 9 percentage points more likely to be for gentry families and not aristocratic families in sessions adhering to a standard parliamentary calendar, which is measured by sessions beginning in autumn or winter and lasting longer than 95 days, compared to sessions that did not (Hoppit 2017, p. 43). I use committee composition as one measure of brokers’ business recruitment. I find that non-titled MPs were more prominent on gentry committees after 1688. Bills were about twice as likely on average to be for gentry families if they had relatively more non-titled committee members after 1688 compared to bills with relatively more non-titled committee members before 1688. In contrast, bills were about 5 to 8 percentage points more likely to be for gentry families in sessions when there were financial crises, as measured by a 1 standard deviation increase from the average interest rate.⁶ A random sample of 151 acts is consistent with the regression results. These show that all families in estate acts increasingly named new types of trustees, who can be considered brokers to Parliament, after 1688. These trustees tended to be attorneys or MPs.⁷

To the extent that estate acts allowed more land to be used in more efficient ways, improved gentry access may have contributed to or been a part of England’s fewer reversals and lower rates of decline in economic growth after 1700 (Broadberry and Wallis 2017). The new system of passing individual pieces of legislation after 1688 to address rights issues was not without its costs, such as those associated with rent-seeking. However, the ability to pass general reforms addressing all inheritance issues was legally, administratively, and politically challenging during this era. I discuss these issues later.

The paper contributes to our understanding of estate acts, the Glorious Revolution, and English political and legal change. First, it improves on previous work on estate bills by allowing them to be more fully integrated into narratives of political change with the Revolution. It does so by coding the same information as existing work and adding new information, such as if the bill was politically connected and the types of MPs that work on estate bills.⁸ I also systematically code novel information in acts, such as the types of trustees and the existence of legal restrictions preventing transactions, which can be applied to more acts in this period or later ones.⁹ This information allows me to document a compositional shift in estate acts and provide novel evidence on the mechanisms by which the gentry improved access that existing work has not systematically assessed.¹⁰

The findings provide evidence supporting political and economic theories of the Revolution’s significance and deepen our understanding of the political changes that did occur. The evidence suggests that constitutional change with the Revolution altered the brokers in the estate act system in a way predicted by Cox (2016), which adds novel evidence to support the idea that the Revolution was a political watershed and economically significant.¹¹ It is well known that new types of parliamentary agents emerged and Parliament adopted procedures to handle increased legislative business after 1688 (Hoppit 2017; Lambert 1971; Rydz 1979, ch. 2; William 1948, 1954). The findings here deepen our understanding of these developments by showing how the legislative system was different before and after 1688 on new margins, like committee composition and brokers named in acts.¹²

Last, the paper provides evidence of England’s political and legal development more broadly. The gentry and their economic mobility are argued to have been politically important during the Tudor era, the Civil Wars, and the Glorious Revolution (Brenner 1976; Tawney 1941; Heldring, Robinson, and Vollmer 2015; Acemoglu, Johnson, and Robinson 2005). In contrast to new wealth and upward mobility, I find that political and organizational change in Parliament and financial crises played a role in their participation in Parliament via estate acts. England also had multiple forums to provide rights and to contest them in court (McCloskey 1972; Bogart 2011; Klerman 2007). The existence of such pluralistic arrangements, excessively secure property rights, and early modern inheritance structures are seen as prohibitive for industrialization and economic development (Rosenthal 1990; Lamoreaux 2011; Harris 2004; Epstein 2000; Kuran 2004). In England, Parliament is viewed as the primary forum for undoing entails and addressing inheritance issues in settlements (English and Saville 1983; Bond 1964; Bogart and Richardson 2011).¹³ The findings suggest that the system would have remained closed to all but a political elite who were connected to the monarch without the Revolution (North, Wallis, and Weingast 2009; Cox 2016). The shift in access to a broader elite after 1688 appears to have been permanent (Bogart and Richardson 2010, table 12; Lambert 1971, p. 110).

BACKGROUND

Estate Acts and Settlements

Parliament, which was strengthened relative to the monarch after the Revolution, passed many types of acts. General acts, such as those regulating taxation, had a national scope. Local acts, as the name implies, had a local origin and a limited scope over local communities, individuals, or families.¹⁴ Estate acts are local, as are the more well-studied enclosure and transportation acts.¹⁵ Estate acts dealt with rights over land and assets, such as buildings built thereon. As recognized in the literature, estate acts concern the tenure of property as well as landownership and inheritance issues in wills, family settlements, and guardianships.¹⁶ Most estate acts dealt with settlements, which I overview next. Figure 1 shows the number of estate acts passed by Parliament from 1600 to 1830. After 1688, there was a notable change in the series in that Parliament passed significantly more acts on average. Estate acts were Parliament’s main quantitative output during this period (Hoppit 1996, table 3). Both motivate the choice to study the period from the Restoration of the monarchy with Charles II in 1660 to the end of William III’s reign in 1702.

Figure 1 ESTATE ACTS, 1600–1830

Note: Figure shows the number of estate acts passed from 1600 to 1830. Vertical line denotes 1688.

Source: Bogart and Richardson (2010, 2011).

A settlement is a generic name for a legal arrangement or transaction that is generally made by a family at marriage or by will. Settlements had three economic purposes: (1) provide income for individual family members; (2) land and property management; and (3) ensure the transmission of land and property within a family over generations. Regarding the second purpose, settlements granted powers to the main landholder, called the life tenant, or to a set of trustees. Examples of powers include making additional charges on the estate; making or renewing agricultural, building, mineral, and repairing leases; and selling, exchanging, or mortgaging land. Settlements specified an entail, a line of heirs who were customarily male, to ensure that land would stay in the family and provide for future generations. Over the course of the seventeenth century, settlements became increasingly restrictive and difficult to break or rewrite, through a new legal device that protected the entail. These so-called strict settlements became the norm and were used until the nineteenth century.¹⁷ Later, I discuss altering settlements.

One interpretation in the literature is that the rights structure in settlements raised the costs of conducting transactions in land markets. Estate acts, which could break or re-write settlements, therefore facilitated market activity.¹⁸ Settlements could increase the insecurity of transacting. Settlements were not public information, and unclear or ambiguous powers could make a conveyance (such as a sale or mortgage) less attractive to a third party. Because settlements provide for multiple family members, multiple interests could contest a potential transaction. Settlements could also prevent transactions if there was no family member with the legal authority (power) to carry out a transaction. The Online Appendix provides examples of such situations. I systematically use information from a random sample of acts to provide evidence of the existence of such legal restrictions.

Modifying a Settlement

A family had several venues to modify a settlement: Court of Common Pleas, Chancery, and Parliament. Details on the processes can be found in the Online Appendix. Given some value in modifying a settlement, for example, allowing land to be sold, a family’s choice of forum depended on the value of waiting to use the Court of Common Pleas, the jurisdiction of Chancery, and the costs of going to Chancery compared to Parliament (Priest and Klein 1984; Klerman 2007). It is generally agreed that, when there was jurisdictional overlap, Parliament was a more efficient or superior forum compared to Chancery (English and Saville 1983; Lambert 1971; Bogart and Richardson 2009).¹⁹ This superiority, however, would depend on parliamentary organization, which I show changed after the Revolution.

There was also forum shopping within Parliament during this period. From 1660 to 1702, a family could introduce a bill into either the House of Commons or the House of Lords. The Houses had a similar procedure for approving a bill: it would be read to the whole house twice, then sent to an ad hoc committee (which had the power to investigate the claims and modify the proposal), and a so-called reporter would then read the bill and any amendments to the whole house for approval. It would then be sent to the other house for a similar process.

Landownership in the Seventeenth Century

Estate acts were for landed families, who made up a small proportion of the population in England and Wales during the late seventeenth century, and their interests were primarily represented in Parliament. Revised estimates of King’s social tables indicate that family heads of traditional landowning groups numbered about 20,000 in total, or 10 percent of all families in 1688. There was inequality within the landed class. The lords or peerage, the smallest group, had an average yearly family income of £6,060, and the great landlords owned about 15–20 percent of cultivated land in England and Wales. In contrast, the larger group of gentry families, referred to as esquires and gentlemen and which I study in this paper, owned about 45–50 percent of cultivated land and had an average yearly income of £562 and £280, respectively (Lindert and Williamson 1982, table 2; Hoppit 2000, pp. 348–49).²⁰

Some have viewed the landed class to be in crisis after 1688 because of increased taxation and tight financial markets that led creditors to call in their debts or foreclose on estates with the growth of public borrowing (Habakkuk 1980). This crisis is traditionally linked to the development of the Country Tory Party, which had the landed gentry as its base and sought policies to protect their interests (Harris 1993; Horwitz 1977). I discuss how the development of parties may be related to estate bills further at the end of the next section.

I now turn to the simple conceptual framework of a family’s decision to petition Parliament for an estate act. The framework clarifies how the Glorious Revolution could influence the types of families seeking an estate act. Put simply, the framework relates politics in Parliament, a family’s social and economic status, and changing economic circumstances to a family’s decision to petition Parliament for an act. It also frames the goals and constraints of other interested parties, like MPs, who would exert effort to manage a bill through Parliament.

CONCEPTUAL FRAMEWORK AND THE GLORIOUS REVOLUTION

Decision to Introduce an Estate Bill

Consider a population of landholding families with a project that requires approval from Parliament to reorganize their rights. Each project has a benefit or value, B _i. Depending on the type of project, B would be the benefit of obtaining an act—for example, in terms of additional rents from a new lease structure in an urbanizing area, the reduction of debt payments, or additional income from selling low-return property. Different types of projects would also introduce other interests to an estate act, like leaseholders, creditors, or a potential purchaser. We can think of i as an ordered index of families with low-value to high-value projects. To petition Parliament, each family must pay some fixed cost F. Parliamentary fees, which were standardized in the 1690s, are one component of F (William 1954, p. 46). The fixed cost may also be endogenous to parliamentary politics, which I discuss next.

Parliament, comprised of the Lords and Commons, had to approve the bill during a legislative session. Each project has some probability p of passage. Members of the Commons and Lords had an incentive to sponsor, work on, and pass bills because they received fees. Working on legislation may have also improved their legislative experience or improved their standing with their constituents. They were constrained by the amount of time in a session and other legislative or parliamentary work.

For a family with project i, the expected payoff of an estate bill is pB _i − F and the payoff of not petitioning is 0. The family will petition if pB _i > F. The probability that a given family petitions for a bill is pr[B _i > F/p]. Assuming the distribution of projects is uniformly distributed from 0 to a high value (B _H), this implies that $pr[{B_i} > F/p] = (p{B_H} - F)/p{B_H} \equiv \hat B$ . With this assumption, $\hat B$ represents the proportion of families we observe coming to Parliament from all potential projects.

The probability that a family introduces a project for approval depends on its probability of passage p, the costs of introduction F, and the value of the act B. Though the model takes the three as given and exogenous, the framework helps conceptualize how the Glorious Revolution or other political and economic dynamics can affect the types of families observed in estate bills, which I empirically examine.²¹ Improved parliamentary meetings, financial crises and opportunities, de facto organization within Parliament, and legal change could all affect the types of families who choose to petition Parliament for an estate act.

Parliamentary Organization

The Revolution’s impact on parliamentary meetings was arguably central to improving its legislative output generally by increasing petitioners’ confidence in Parliament (Hoppit 1996, 2017, ch. 2). The Revolution significantly improved parliamentary meetings by requiring the monarch to call Parliament for funding and military purposes (Cox 2016). After 1688, legislative sessions met annually; they began to follow a standard calendar (beginning in autumn or winter), and nearly all met for an average of 100 days (Hoppit 2017, p. 43). The Revolution also improved the probability that estate bills were approved by mitigating conflict between the monarch and Parliament (Dimitruk 2018). All estate bills in my dataset had an average success (or approval) rate of 49 percent in sessions before 1688 and an average success (or approval) rate of 79 percent after 1688. The improvement in parliamentary meetings would increase p in the framework described previously and lower the threshold for entry, pulling families with relatively low-valued projects into Parliament. So long as there is some distribution of families ranging from low to high projects or families are ordered by their social rank (aristocracy with high-valued projects and gentry with low-valued projects), we would expect to see more low-valued projects introduced or gentry families in the proportion of families who introduce bills as sessions become standardized and the likelihood of success improves.

Relatedly, interested parties may have understood Parliament to be more sympathetic to gentry bills after 1688 because Parliament had established principles for addressing rights issues and could handle more bills with more time (Hoppit 2011). This may have decreased the opposition gentry families faced and, in turn, improved their access. I find gentry bills had fewer petitions compared to aristocratic bills on average, but there was no significant change in the difference after 1688.

Improved meetings may have also altered the recruitment strategies of MPs or other brokers. Parliament may have been akin to a royal forum in its provision of estate acts prior to the Revolution.²² MPs may have set a high cost of entry F or limited their services to protect their time because parliamentary meetings were unpredictable and short. The Court coalition, who were supporters of the monarch, often titled, and held a majority for most sessions, may have been key brokers and limited the supply of acts to fellow supporters of the monarch. After 1688, as meetings became standard and the probability of approval increased, the brokers who received fees from guiding bills through Parliament, such as MPs and attorneys, may have sought to expand their services (Cox 2016, ch. 6). We may therefore expect to see non-titled MPs or MPs who represent families’ constituencies work more on gentry bills after 1688 (Beckett 2014). We may also see attorneys or other brokers involved in acts after 1688.

Financial Markets, Legal Change, and De Facto Organization

In contrast to a political story of improved access to Parliament, financial markets may have pushed more gentry families into the population of families with projects or more gentry families to the upper tail of the distribution (B _H). The Revolution was a revolution in public finance. Tax receipts from 1690 to 1695 more than doubled the tax receipts from 1660 to 1685, a new land tax was introduced in 1694, and government borrowing increased dramatically after 1688 to finance the Nine Years’ War (Cox 2016, p. 49). Increased government borrowing may have led creditors to call in debts from landed families, decreased the supply of loanable funds, and led to high interest rates (Quinn 2001; Temin and Voth 2005; Sussman and Yafeh 2006; Habakkuk 1980). Thus, there may have been pressure on landholders to sell land because of an increase in their fixed charges, like taxes and interest rates (Habakkuk 1980). If the gentry were more sensitive to a financial shock, then there would be more gentry Bs in the distribution after 1688.²³ There were also new financial opportunities with the growth of the stock market after the Revolution (Cain and Hopkins 1986; Carlos, Fletcher, and Neal 2015). The new opportunities may have also pushed landowners to reorganize their assets and sell land for more fungible and higher-return assets. We would therefore expect to see the probability that a bill is for a gentry family correlated with measures of financial crises and opportunities. We may also find evidence of financial circumstances cited as motivations for acts.

There are two additional hypotheses. On the demand side, it is possible that the adoption of the strict settlement led more gentry families to Parliament during this period.²⁴ Second, the Revolution altered the de facto political organization of Parliament with the solidification of the two-party system. The Tory party may have been more favorable to the gentry because they tended to be the Tory Party’s base.²⁵ I do not find evidence that the adoption of the strict settlement or de facto political change explains the greater access by gentry families after 1688. These results are presented and discussed in the Online Appendix.

DATA AND GENTRY ACCESS TO PARLIAMENT

I use two main datasets in the analysis. The first is a repeated cross-section of bills introduced in 34 sessions of Parliament from 1660 to 1702.²⁶ For each bill, I hand-collect all information on its legislative process from the Journals of the House of Commons and the Journals of the House of Lords, which have ad hoc committee and reporter assignments as well as geographic information. There are 910 bills in the final bill dataset; 592 (65 percent) were approved by Parliament and became acts of Parliament. The second dataset is a set of full-text transcriptions of a 25 percent random sample of estate acts from the Parliamentary Archives (n = 151).²⁷ I also add time series variables and collect information on MPs from the History of Parliament research project. Table 1 reports summary statistics on variables. The Online Appendix details the dataset and the construction of other variables.

Table 1 SUMMARY STATISTICS OF ESTATE BILLS AND ACTS, 1660–1702

Notes: Column (I) reports the average and standard deviation for select variables in estate bill dataset. Column (II) reports the average and standard deviation for select variables in the random sample of acts. Panel A reports share of bills for a given family type for full sample of estate bills, n = 910, and the random sample of acts, n = 151. It also reports the share of aristocracy and gentry bills (n = 726) that vested land in trustees, were politically connected (MP bill), or had a majority party MP political connection. Panel B reports shares conditional on sample of aristocracy and gentry bills (n = 726). Panel B also reports shares conditional on bills with committee information (n = 563) and conditional on bills with committee and geographic information (n = 469). Panel C reports variables only found in estate act dataset. “=1” indicates the variable is an indicator equal to 1 and 0 otherwise.

Sources: Estate bill dataset and random sample of estate acts.

Family Type and Bill Characteristics

The main outcome to be explained is family type. Using the bill’s first reading or title, I classify each bill as for the nobility (such as for a duke, earl, lord, countess), titled gentry (baronet or knight), gentry (esquire or gentleman), professionals (such as a merchant), no rank, or other (such as infant, lunatic).²⁸ I make this a mutually exclusive category: if a bill names multiple interests from different categories, I classify the bill as for the “highest ranking” class.²⁹ Because titled gentry appear to be distinct as a group from the gentry, I classify the nobility and titled gentry as aristocracy. Most bills (81 percent) were for the aristocracy (nobility and titled gentry) and the gentry (Panel A, Table 1). I restricted the sample to these bills in the regression analysis to facilitate comparisons across bills (n = 726).

I classify bills as politically connected or not, which allows me to examine whether Parliament was favorable to families with certain political connections. A bill is considered politically connected if the named landholder is an MP or a person related to an MP. If politically connected, I also classify whether the family was with the majority party or coalition (Court, Whig, or Tory) using the History of Parliament database, Dimitruk (2021), and Bogart (2016). About 45 percent of all aristocratic and gentry bills were for an MP, and 14 percent of all aristocratic and gentry bills were affiliated with the majority party.

I add several bill-level characteristics to include as controls that capture other unobserved aspects of the family’s decision to introduce a bill to Parliament: if a bill vests land in a set of trustees, bill type, and house of entry. The patterns for bill type are broadly similar to previous studies (Bogart and Richardson 2010; McCahill 2013). See Online Appendix Section A3 and Table A1.

Commons’ Committees and Reporters

I use ad hoc committee and reporter assignments for information on the types of brokers in Parliament and for geographic information. Not all bills made it to the committee or reporter stage or have geographic information. The committee and reporter assignments are still valuable because they are perhaps the only surviving record of parliamentary activity on estate bills consistently recorded for every session during this period (Moore and Horwitz 1971). The results of the reporter analysis are similar to the committee analysis, so I omit discussion here. Table 1, Panel B, shows that 77 percent of gentry and aristocracy bills have Commons’ committee information (n = 563). More bills after 1688 have committee information, which is likely an outcome of the improvements in parliamentary proceedings after 1688 (Online Appendix Figure A1). About 72 percent of gentry and aristocracy bills have geographic information. About 64 percent of gentry and aristocracy bills have both committee and geographic information. This selection will likely lead to a non-random bias in the regression analysis.

For gentry and aristocracy bills with committee information, I calculate two measures: (1) the share of committee members who were not titled (not knights, baronets, or styled as lords) and (2) the share of committee members who were elected to constituencies in the same county as the family’s estates.³⁰ I refer to committee members representing constituencies in the same counties as land in estate bills as “constituency representatives” for simplicity. On average, 56 percent of committee members had no title, and 12 percent were constituency representatives (Table 1, Panel B).

Random Sample of Estate Acts

The second dataset is a set of transcriptions for a 25 percent random sample of estate acts from the Parliamentary Archives (n = 151).³¹ I use the acts for three pieces of information: the rental income value of the estate, the families’ motivations for an act and the types of settlement acts addressed, and the types of trustees as another indicator of parliamentary brokerage. I also code family types and the types of transactions, which show a similar distribution of types of families and transactions as the bill dataset (Online Appendix Table A1).

Table 1, Panel C (Column II), and Table A2 in the Online Appendix report summary statistics on variables only found in the sample of acts: the value of land, evidence of legal restrictions, and types of contracts addressed. Acts tended to address land with an average rental income value of about £400 per annum and a median of £185 per annum (n = 118 with value information). Both are significant sums of money and broadly similar to the annual income of gentry and titled families found in King’s social tables.³² Online Appendix Table A2 shows that 83 percent of acts explicitly cite some sort of legal restriction preventing the family from carrying out an economic transaction, which supports a transaction cost interpretation of settlements and estate acts (Bogart and Richardson 2009).

Time Series

Last, I add several time series variables.³³ To capture certainty of parliamentary meetings, I construct an indicator for sessions adhering to a normal parliamentary calendar based on Hoppit (2017, p. 43): “1688 […] established annual sessions and a predictable parliamentary calendar; sessions usually began in the autumn or winter, lasting an average of 100 days.” I code an indicator equal to 1 if a session began in autumn or winter (October–November or December–February) and lasted at least 95 days, and 0 otherwise.³⁴

I use three variables to test whether financial crises or opportunities pushed families into Parliament. I use two measures of financial crises: sessions meeting during war and the Dutch interest rate. War is correlated with high government interest rates and financial crises in England after 1688 (Quinn 2001; Sussman and Yafeh 2006). The Dutch interest rate (losrenten), which is a yield and ranges from 0.02 to 0.05, is from Gelderblom and Jonker (2011). There is no interest rate series for England covering 1660 to 1702, but the capital markets of Amsterdam and London were connected (Sussman and Yafeh 2006). As a proxy for financial opportunity, I use the value of East India Company commodity and treasure exports (in £10,000s) from Chaudhuri (1978).

Gentry Access

I now provide descriptive evidence that the gentry saw improved access to Parliament after 1688. The analysis uses probit models to explain their improved access. Figure 2 shows that the average number of bills for gentry families (lower gentry) increased by about 7 bills on average per session after 1688, from an average of 5 bills for gentry families before 1688 to 12 bills per session after. The number of bills for aristocratic families remained at a relatively constant average of 6 bills per session (Table A5 in the Online Appendix). The patterns point to Parliament becoming more efficient at processing bills where it took on more bills for gentry families. It did not become biased toward the gentry at the expense of others. The session of Parliament meeting in 1670 stands out for the large number of gentry bills (n = 19), which may be explained by the session’s unusual length of over 200 days for the pre-1688 period. This suggests parliamentary certainty may be a key mechanism underlying the permanent rise of the gentry because longer sessions became standard after 1688.

Figure 2 RISE OF GENTRY ESTATE BILLS AFTER 1688

Notes: Figure shows the number of estate bills introduced to Parliament for the nobility (such as earls, dukes), titled gentry (knights and baronets), and lower gentry (esquires and gentlemen).

Source: Author’s calculations from bill dataset.

Figure 3 and Online Appendix Table A6 indicate that acts after 1688 tended to have lower-valued projects. Kernel density estimates show the values of projects in the random sample of estate acts became more right-skewed after 1688, suggesting more low-valued projects were introduced (Figure 3).³⁵ Table A6 in the Online Appendix estimates and compares the total income addressed by gentry and aristocratic estate acts before and after 1688. It shows that Parliament addressed significantly more of the total gentry income through estate acts after 1688. Though gentry acts passed after 1688 tended to have lower income values compared to those passed before 1688, the parliamentary system processed more gentry acts after 1688.

Figure 3 CHANGE IN VALUE ESTATE ACTS AFTER 1688

Notes: Figure reports kernel density estimates of the annual rental income value of an estate from sample of acts passed before the Revolution (1660–1685) and acts passed after the Revolution (1688–1702). It excludes one act with a significant outlier rental value of £4,000.

Source: Author’s calculations from 118 estate acts with value information.

EMPIRICAL APPROACH AND RESULTS

I use probit models to examine whether changes to parliamentary organization, financial markets, or other factors increased the likelihood that a bill was for a gentry family after 1688. I also bring in evidence from the random sample of acts to deepen the discussion of the regression results. Replication files for results produced here can be found in Dimitruk (2023).

For a bill i in session t, I use probit models of the form:

(1) $$\Prob(Bil{l_{it}} = Gentry) = \alpha + \beta post1688 + \gamma {X_t} + \varphi {X_{it}} + {\varepsilon _{it}}.$$

The outcome is an indicator equal to 1 if a bill i in session t is for a gentry family and 0 for an aristocratic family. Because there may be unobserved correlation at the session level related to a family’s choice to introduce a bill, standard errors are clustered at the session level (n clusters = 34); X _t is a matrix of time series measures and X _it is a matrix of bill-level controls, such as bill type.

I am primarily interested in the estimate of β and if it can be explained by time series (X _t) or bill-level (X _it) controls. For example, based on Figure 2, we would expect β to be positive and significant. We may also expect the estimate of a parliamentary certainty variable to be positive and significant if it pulled more gentry landholders into Parliament by lowering the threshold of entry. If including the variable also decreases β, or makes it statistically insignificant, then we could interpret parliamentary certainty as the key mechanism leading to the rise of the gentry after 1688.

To test the hypothesis that the Revolution altered MPs’ recruitment of new types of business from gentry families, I use probit models of the form:

(2) $$\Pr ob(Bil{l_{it}} = Gentry) = \alpha + {\beta _1}post1688 + {\beta _2}CommitteeShar{e_{it}} + {\beta _3}(post1688 \times CommitteeShar{e_{it}}) + \gamma {X_t} + \varphi {X_{it}} + {\varepsilon _{it}}$$

for gentry and aristocracy bills with committee information. is either a bill’s share of committee members who are not titled or, for bills with committee and geographic information, a bill’s share of committee members who are constituency representatives.

This approach will provide descriptive evidence for the hypothesis that MPs recruited families more broadly into Parliament after 1688. The unobserved committee appointment process is likely more complicated. Families or their agents outside of Parliament may have requested specific MPs to attend the committee.³⁶ The structure of the empirical approach assumes that all committee and reporter assignments are the result of some unobserved recruitment decision by MPs, which led gentry bills to be introduced in the first place.

Other exercises reported in the Online Appendix establish robustness and support the main results. In addition to examining de facto political change and the adoption of the strict settlement, I also find evidence that Parliament as an organization, and not only the Commons, became more open to the gentry after 1688. A second exercise, which is similar to a break date analysis, examines whether there is a significant change in the series in 1688 or some other date (Bai and Perron 1998; Murrell 2017). Last, I reproduce the results for symmetric windows of sessions around 1688, omitting potential outlier sessions at the beginning and end of the period.

Parliament Meetings and Financial Markets

The evidence generally suggests that political and organizational change after 1688 was a relatively more important driver of broadening the estate act system to gentry families when compared to financial push factors like crises or investment opportunities. Political and organizational change broadened access in two ways. It altered a family’s entry calculus by lowering the threshold for entry. It also broadened MPs’ and other brokers’ recruitment of bills.

Table 2 reports the results of estimating Equation (1), where the outcome is an indicator equal to 1 if a bill is for a gentry family and 0 for an aristocratic family. All estimates report the average marginal effect (AME).³⁷ The estimates in Table 2 imply sessions after 1688 are between 14 and 27 percentage points more likely to have bills for gentry families and not aristocratic families than sessions before 1688. Column (1) only includes a post1688 indicator. Columns (2)–(8) include bill controls (bill type, house of entry, if the family was politically connected). Columns (3)–(8) include different time series variables to test the relevance of parliamentary certainty and financial crises or opportunities. The estimate for post1688 remains positive and significant across all specifications.

Table 2 PARLIAMENTARY CERTAINTY, FINANCIAL CHANGE, AND GENTRY BILLS AFTER 1688

Notes: Table reports average marginal effects (AMEs) of probit regressions where the outcome is an indicator equal to 1 if a bill is for a gentry family and 0 for an aristocratic family. Unit of observation is a bill-session (repeated cross-section of bills introduced across 34 sessions). All reported AMEs are for variables measured at the session-level. “=1” indicates the variable is an indicator equal to 1 and 0 otherwise. Post1688 is an indicator equal to 1 for sessions meeting after 1688 and 0 otherwise; calendar is an indicator equal to 1 if a session adheres to a standard parliamentary calendar and 0 otherwise; session begins in autumn or winter is an indicator if a session began in Oct.–Nov. or Dec.–Feb. and 0 otherwise; days ≥ 95 is an indicator if a session met for at least 95 days and 0 otherwise; conflict revenue is an indicator if there was conflict over revenue between the monarch and Parliament in a session and 0 otherwise; war is in an indicator if the session met during a war and 0 otherwise; EIC exports is the value of the EIC’s exports in a given year; losrenten rate is the Dutch interest rate. Bill controls are a set of bill level controls such as bill type, MP bill, and introduced to Commons. Majority Tory session denotes an indicator equal to 1 if the Tories held a majority in a session and 0 otherwise.

*** p<0.01, ** p<0.05, * p<0.10. Standard errors clustered at session-level (n = 34).

Sources: Estate bill dataset, Gelderblom and Jonker (2011), and Chaudhuri (1978).

The estimate of the variable measuring the certainty of Parliament’s meeting (calendar) is positive and significant. Bills introduced in sessions adhering to a standard calendar were about 9.3 percentage points more likely to be for gentry families compared to sessions that did not adhere to the standard calendar (Column (3)). It also explains some of the estimate of post1688. The size of the estimate for post1688 declines in Column (3) compared to Column (1). Columns (4) and (5) suggest that neither of the two component parts (a standard start month or session length) of the variable is driving the result. Rather, it was the combination of a standard start month and session length that pulled gentry families into Parliament. Column (6) suggests that uncertainty due to conflict over revenue between the monarch and Parliament, which led to the sudden closure of sessions, is not a direct mechanism leading to the selection of different family types.

Column (7) includes variables measuring financial crises (losrenten, war) and opportunity (EIC exports). Of the three, only the losrenten rate is positively related to the probability that bills are for gentry families and precisely estimated. Next, I interpret the estimate of losrenten in Column (8). The findings suggest that financial crises, as measured by higher interest rates, pushed gentry families into Parliament. While not precisely estimated, the estimate of the effect of war, which was related to large government borrowing, high rates, and tight credit markets, is positive. Similarly, the measure of financial opportunity, EIC exports, is positive but imprecisely estimated.

The results suggest that financial crises pushed gentry families into Parliament and that parliamentary certainty pulled families into Parliament, but neither can entirely explain improved gentry access after 1688 because the post1688 estimate remains positive and significant. In addition to the full set of bill controls, Column (8) includes the Parliament meeting variables (calendar and conflict) and the financial variables in a multivariate analysis. The estimate of post1688 remains positive and significant, as do the estimates of calendar and losrenten. This suggests that the Revolution had an independent effect, or an effect through a mechanism not yet addressed in the specification, on families’ access.

The estimates imply that parliamentary certainty has a slightly larger effect compared to financial crises. The AME of calendar from Column (8) implies that bills introduced in sessions that adhered to a normal parliamentary calendar were 7.6 percentage points more likely to be for gentry families and not aristocratic families compared to sessions that did not. I calculate the difference in the average probability bills are for gentry families for a 1 standard deviation increase from the average rate (from 3.8 to 4.3 percent) using the estimates in Column (8). Doing so implies that bills were about 5 percentage points more likely to be for gentry families when there were financial crises (Online Appendix Table A7). There is also evidence from the sample of acts that financial crises and new financial opportunities played a growing role in pushing families to Parliament after 1688, though the regression analysis suggests financial opportunity is not significant. Online Appendix Section A3 provides examples of these financial motivations or purposes, which include burdensome interest rate and tax charges, foreclosure by creditors, and opportunities where landholders sought to make loans.

Committees and Brokers

The previous section finds that gentry access after 1688 is related more directly to a family’s decision to introduce a bill: the pull of parliamentary certainty (p) and a push from financial market crises (B). The estimate of post1688 remains positive and significant in a multivariate analysis, which suggests that the Revolution had an independent effect on gentry access or affected gentry access through other channels. I now provide evidence that the Revolution may have improved gentry access by effectively lowering the costs of introduction F by broadening MPs’ recruitment strategies and the brokers linking families to Parliament.

As discussed, we might expect the Court coalition to be key brokers and to have limited the supply of acts to fellow supporters of the monarch to protect their time prior to 1688. Figure A4 in the Online Appendix shows that bills were more likely to be politically connected, and politically connected bills were more likely to be for the majority Court coalition before 1688.³⁸ Acts from 1660 to 1688 also commonly describe debts that were incurred because they fought for the monarch in the Civil Wars. These debts were incurred to pay fines levied on supporters of the monarch by parliamentarians.³⁹ After 1688, bills were less likely to be for the majority party, either Whig or Tory.

Table 3 provides further descriptive evidence that MPs broadened recruitment after 1688 by showing that gentry bills had a different composition of committee members after 1688.⁴⁰ All specifications in Table 3 include the variables capturing parliamentary certainty (calendar) and financial crises (losrenten rate), bill controls (such as bill type and house of entry), and time series controls (such as majority Tory sessions). I report the probit coefficients of the interaction terms in Columns (3)–(4) and Columns (7)–(8). All other estimates are AMEs. Columns (1)–(4) report results when including the share of non-titled committee members. Columns (5)–(8) report results for the share of constituency representative committee members.

Table 3 THE REVOLUTION AND ESTATE BILL COMMITTEE COMPOSITION

Notes: Table reports results of probit regression where the outcome is an indicator equal to 1 if a bill is for a gentry family and 0 for an aristocratic family. Unit of observation is a bill-session (repeated cross-section of bills introduced across 34 sessions). ^a Indicates probit coefficients; all other estimates are average marginal effects (AMEs). Columns (1)–(4) conditional on aristocracy or gentry bills with committee information (n = 563); Columns (5)–(8) conditional on aristocracy or gentry bills with committee and geographic information (n = 469). “=1” indicates the variable is an indicator equal to 1 and 0 otherwise. Post1688 is an indicator equal to 1 for sessions meeting after 1688 and 0 otherwise; share cm not titled is a bill’s share of committee members who are not titled; share cm const. is a bill’s share of committee members who are constituency representatives; calendar is an indicator equal to 1 if a session adheres to a standard parliamentary calendar and 0 otherwise; losrenten rate is the Dutch interest rate. Bill controls are a set of bill level controls such as bill type, MP bill, and introduced to Commons. Time series controls include Majority Tory session, war session, if there was conflict over revenue between the monarch and Parliament.

*** p<0.01, ** p<0.05, * p<0.10. Standard errors clustered at session-level (n = 34).

Sources: Estate bill dataset and Gelderblom and Jonker (2011).

The results generally support the idea that estate bills were more likely to be for gentry families after 1688 in the sample of bills with committee information. The AME for post1688 ranges from 11 percentage points to 39 percentage points. The estimates are biased upward because more bills after 1688 made it to the Commons’ committee stage and therefore have committee information.

Table 3 also shows that committees with relatively more non-titled members after 1688 were more likely to be for gentry families. Online Appendix Figure A5 shows that more MPs participated in committee work after 1688. Online Appendix Figure A6 shows that, on average, committees tended to have more non-titled members after 1688. The regression results suggest that these changes coincided with an evolution in the types of MPs that served on gentry committees. Columns (2)–(4) show that committees with relatively more non-titled members were more likely to be for gentry families on average. Column (3) indicates that the Revolution further strengthened this relationship: committees with relatively more non-titled members were even more likely to be for gentry families after 1688. I discuss interpreting the estimates further next. To control for the possibility that this relationship is a result of fewer titled MPs elected as opposed to different recruitment strategies, Column (4) includes a time series measure of the share of total titled MPs in a session. The interaction term remains positive and significant. Columns (5)–(8) repeat the analysis using the share of constituency representative committee members instead of non-titled committee members. Committees with constituency representatives were on average more likely to be for gentry families, but there is little evidence that the Revolution had an additional impact via committee composition.

These results suggest that parliamentary certainty may have also altered the incentives of these committee brokers. The estimate of calendar is generally positive but imprecisely estimated in most specifications. This could be due to the selected sample of bills with committee information and the proposed mechanism of certainty affecting both a family’s entry calculus and MPs’ recruitment strategies. Across all specifications, the losrenten rate is positive and significant.

The results support that political and organizational change was relatively more important compared to financial crises in broadening access to gentry families. The estimates from Column (4) imply that the non-titled committee composition variable has a larger effect on the likelihood that bills are for gentry families compared to financial crises. This difference can be seen in Figure 4, which shows the average probability bills are for gentry families for all values of ShareCommNotTitled before and after 1688. For a 1 standard deviation increase in the average share of non-titled committee members (from 0.56 to 0.72), the average probability bills are for gentry families doubles from 7.6 percentage points before 1688 to 17.1 percentage points after 1688 (Online Appendix Table A7, Panel B). In contrast, bills were 7.1 percentage points more likely to be for gentry families given a 1 standard deviation increase from the average losrenten rate.

Figure 4 NON-TITLED COMMITTEE MEMBERS AND PROBABILITY A BILL IS FOR A GENTRY FAMILY

Notes: Figure reports marginal effects (the average adjusted prediction of the likelihood a bill is for a gentry family for values of the share of non-titled committee members) by post1688.

Source: Author’s illustration using estimates from specification (4) in Table 3.

The costs of entry may have also been related to new brokers in the form of trustees named to manage estates that linked families to Parliament. Cox (2016, ch. 6) argues that the brokers involved in the provision of rights under the purview of the monarch would significantly change from courtiers to MPs and attorneys after 1688. Estate bills after 1688 were 21 percentage points (from about 12 to 33 percent) more likely to vest land in a set of trustees, who were (mostly) authorized to sell land, purchase new land, and raise money to support family members. Consistent with Cox (2016), Online Appendix Figure A7 uses the random sample of acts to show that both MPs and attorneys were increasingly used as trustees across all families (and not only the gentry) after 1688.

The evidence on the evolution of brokers or intermediaries (committee composition and trustees) deepens our understanding of how the Revolution changed Parliament’s legislative organization. We know that Robert Harper, an attorney and parliamentary agent who was often named as a trustee in estate acts, was a key player in estate acts in the first quarter of the eighteenth century (Lambert 1971, pp. 9–11). I find no obvious evidence of such specialization in my sample of acts, but it seems likely that the emergence of a specialist like Harper is related to the changes documented in this paper as well as the standardization of parliamentary procedures from 1699 to 1708.⁴¹ There may have been other types of intermediaries involved. Sir Robert Clayton, a prominent moneylender and broker for the gentry, sat on 22 estate bill committees after the Revolution and was named as a trustee on a gentry family’s act after 1688 (Whitley Estate Act, 5&6 W&M, c. 16). He and his partner, John Morris, are credited with integrating mortgages into banking practice when credit markets in general tended to be personalized, with borrowers relying on legal and economic networks to gain access to credit.⁴² Fully documenting these connections is one avenue for future work.

DISCUSSION AND CONCLUSION

Econometric and archival evidence indicates that political and organizational changes to parliamentary meetings with the Glorious Revolution were primarily responsible for improved gentry access to Parliament via estate bills. Making parliamentary meetings more certain lowered the threshold for entry and facilitated the recruitment of new business by MPs and other brokers. The evidence also indicates that financial crises, due to higher costs of taxes, increased interest rates, and creditors calling in debts, also played a role in pushing gentry families to Parliament.

Were there economic benefits to improving access for gentry families? Online Appendix Table A6 suggests that the acts addressed a significantly larger share of gentry income, but with limited national income effects. These patterns are consistent with the evidence that Britain saw fewer reversals of economic growth and a lower rate of so-called shrinking in the early eighteenth century compared to other preindustrial societies (Broadberry and Wallis 2017, tables 7 and 8).⁴³ Passing more gentry acts may have facilitated the process of Smithian growth in land and credit markets, as proposed by Broadberry and Wallis (2017), by allowing a wider range of transactions to take place. The findings here suggest that the parliamentary system would have remained closed to all but a political elite, who were connected to the monarch, without the Revolution.⁴⁴

The new system after 1688 was not without its costs. Passing individual acts as opposed to wholesale reform of inheritance institutions likely increased the value of individual MPs or other brokers and created opportunities for rent-seeking. Contemporary MPs remarked on the “mischiefs arising” from the increased passage of private bills generally (quoted in Hoppit 2000, p. 351). Others described MPs as “more like an assembly of trade or merchants than of legislators” and that “both public and private business came to market [in Parliament] and neither could be done unless paid for” (quoted in Horwitz 1977, p. 148). It seems that private legislation was a significant source of income for main intermediaries completely specialized in the legislation, like clerks of Parliament, while it was more of a side payment for those, such as MPs, working on one or two bills.⁴⁵

Costs of wholesale reform, like those undertaken in the nineteenth century in England and Europe to address the legal constraints created by settlements, appear to have been prohibitive during this era. Settlements and credit markets were decentralized, which made it difficult to write a general act to address the variety of family circumstances. Parliament employed long-standing principles of expropriation with compensation but was unwilling to produce general reforms due to potential disruptions or unintended consequences (Hoppit 2011, pp. 99–103). Decentralization also created vested interests that could oppose general reform.⁴⁶ Reforms were instead taken piecemeal, leading to interactions between local and general legislation (Innes 1993, p. 43). An Act of 1708 appears to have addressed the main legal restraint in seventeenth-century acts documented in Online Appendix Table A2. It allowed infants to make conveyances of estates without a private act of Parliament (7 Anne, c. 19).⁴⁷

Parliament passed thousands of estate acts, which undid entails and allowed a broad spectrum of landholders to sell land, in the eighteenth and early nineteenth centuries. This paper argues that this broad access to Parliament was a direct result of the political and organizational changes brought about by the Glorious Revolution of 1688. Before 1688, estate bills were primarily for aristocratic families who were politically connected and affiliated with the monarch. After 1688, acts were primarily for gentry families, who accessed Parliament through a wider network of brokers. The post-1688 estate act system was revolutionary and represented a significant break from that of the post-Restoration period.