“I Intend Therefore to Prorogue”: the effects of political conflict and the Glorious Revolution in parliament, 1660–1702

Abstract

1. Introduction

How to set up and change property rights is a fundamental problem for societies.¹ For example, for land to be traded today, there are titling agencies and registries to securely transfer ownership. States and governments play a key role in this system, but can be impeded by political barriers and influenced by political factors.² This paper studies how political barriers to changing property rights were removed in seventeenth century English parliament. The lessons from this particular period in English history loom large in the economic history, political economy, and development literatures as well as for policy makers today.

Specifically, I study how politics and conflict influenced parliament’s passage of legislation that reorganized property rights in land markets, so-called estate bills. Land was not easily transferred or improved because there were complex, opaque restrictions from England’s inheritance system and norms. Parliament provided estate bills as a legislative service to its land and property-holding constituents so they could reallocate their rights and use resources in new ways. An Act passed in 1677, for example, was titled “An Act to enable Sir Edward Hungerford Knight of the Bath, to make Leases for Years of Hungerford House in the Strand.” Hungerford cited that the “Houses Buildings and Shoppes are too old and ruinous” and that with the rights to lease land “encouragment might be given to those who should undertake the rebuilding.”³ After improving the buildings, Hungerford was granted the right in 1678 to hold a market on the property, which was in the heart of London, three days a week. He eventually sold the property and the market rights.⁴ Estate bills were also quantitatively important. They were the single largest category of parliament's legislative output (successful bills) from 1660 to 1714, making up the majority of personal Acts during this period.⁵

Estate bills were numerous but they were not always successful in being approved by parliament and the monarch. Figure 1 shows changes in the success rate of estate bills from 1660 to 1702. A “successful bill” was approved by the House of Commons, the House of Lords, and received the royal assent from the monarch to become an Act of Parliament. The era begins after the restoration of the monarchy in 1660, includes the regime change with the Glorious Revolution of 1688 and ends with the death of William III in 1702. Gaps show years when parliament did not meet. The success rate was low on average, about 40 percent, before 1688. After 1688, however, the rate shows a marked increase to an average of about 80 percent. This paper studies why the success rate of estate bills was low before the Glorious Revolution and improved thereafter.

Figure 1.

Success rate of Estate Bills: 1660–1702. Notes: Author’s calculations from estate bill dataset. Figure shows improvements in the success rate of estate bills (reorganization of property rights) by years bills were introduced. A “successful bill” was approved by House of Commons, House of Lords, and monarch (became an Act of Parliament). Parliament did not meet in 1672, 1676, 1682–1684, and 1686–1688.

Open in new tabDownload slide

I use a new dataset on estate bills to document that the sudden prorogation (closure) of a session of parliament by the monarchs caused estate bills to fail until 1690. The dataset contains detailed information on each bill’s “legislative life-cycle” collected from the Journals of the two Houses.⁶ I code sessions that were suddenly prorogued from the monarchs’ speeches in parliament and from the historiographies. I first provide descriptive evidence that the sudden closure of sessions by the monarch caused estate bills to fail. I also show that sudden prorogations were more common before 1690. I then use probit and OLS models to estimate the effect of sudden prorogation on estate bill success.

The estimates show that bills were about 19 percentage points more likely to fail in these sessions. The sudden closure of parliamentary meetings had several economic effects. The main effect was that bills entering in sessions suddenly prorogued were “scarred”: they had a significantly lower long-term success rate. Bills that were eventually successful were delayed by about 8 months and there is some evidence to suggest new bills were displaced in sessions meeting after one that was suddenly prorogued.

My approach addresses several identification issues. I include a rich set of bill-level and time-varying controls to account for possible correlation between sessions that were suddenly prorogued and other factors that could lead to bill failure. I provide evidence on the timing of bill introduction and passage in parliament to show that sudden prorogations were unexpected. Bills do not appear to be introduced earlier on average nor did parliament legislate on bills more quickly in sessions that were suddenly prorogued.

In the last section, I examine why sudden prorogations occurred more often before the Glorious Revolution. I document that the monarch suddenly ended sessions of parliament to interfere with parliamentary proceedings unrelated to estate bills. The proceedings centered on the crown–parliament relationship: (1) conflict that stopped the passage of revenue bills, (2) conflict in parliament with a monarch’s policies, and (3) the monarch’s independence from parliament. The three political variables are positively correlated with sudden prorogations under the late Stuarts before the Glorious Revolution, but not under William and Mary and William III who reigned after 1688. The evidence shows that political and constitutional changes with the Glorious Revolution made sudden prorogations less likely largely by decreasing the likelihood that both conflict delaying revenue bills and conflict with policies occurred. Further econometric analysis of estate bill success shows that conflict delaying revenue significantly hindered parliament’s passage of estate bills under the late Stuarts.

There are several contributions of this paper. First, it contributes to our understanding of parliaments and state capacity. Functioning legislatures and states with robust capacity are key elements of growth and development. They can secure property rights from expropriation by the executive and provide public goods and services.⁷ I provide new evidence on how a parliament’s provision of an economic service legislation and functioning improved during this era. This adds an important historical example to understanding how conflict can hinder state capacity and how removing political barriers can improve its capacity. This paper also improves upon recent historical work studying functioning parliaments. I study a parliament’s legislative output, which provides insight to how a parliament that meets can help explain urban economic growth.⁸

Second, the analysis presents new evidence to how the Glorious Revolution was important for English development. The literature is extensive and follows several threads.⁹ Work argues changes with 1688 asserted parliament’s supremacy by constraining the executive, which served as the foundations of England’s political and economic ascent until the Industrial Revolution.¹⁰ The changes could have impacts for public finance because it improved credibility of the government.¹¹ It could also influence policy-making in general because it shifted power to parliament.¹² This paper is the first to provide causal evidence on how political and constitutional changes with the Glorious Revolution were important for parliament’s legislative position by making sudden closures by the monarch less likely. It is also the first to provide evidence consistent with Cox (2012), who argues that parliament’s new designs of revenue bills and fiscal policies improved the government’s fiscal credibility and were important for the development of the English state. This paper shows these changes were also important for allowing parliament to pass legislation.

Last, this paper contributes to our understanding of property rights change. For England, property rights in land were relatively secure after the restoration of the monarchy, but they could not be changed easily and could not be adjudicated in the common law system.¹³ Parliament would operate as a forum where individuals, families, and communities could reorganize their property rights en masse to use land and resources in new ways throughout the preindustrial era.¹⁴ The removal of political barriers during this era provides insight to one reason parliament became an adaptable institution providing legislation to changing economic conditions until the nineteenth century.¹⁵

2. Background

2.1 Property Rights to Land and Estate Bills

Land and permanent fixtures, such as buildings and mines, were bound by a set of inheritance institutions and norms, known as strict settlements.¹⁶ A strict settlement was a contract that (1) ensured that a family's estate would descend across generations and (2) provided for extended family members. To carry out these goals, land and resources belonged to a family trust. The trust ensured the estate followed a succession across generations and provided for extended family. To ensure the transmission of land across generations, a property holder assigned control of the estate to an heir, usually an eldest son, and dictated that on the death of the heir, the estate would descend to an unborn heir (creating a tenant in tail). The trust made the current holder a beneficiary and a life tenant with restricted powers over land use. The life tenant in turn held the land in trust for a set of beneficiaries who had rights to income from the land (estates). The beneficiaries could include his wife, children, unborn descendants, future heirs, and charities.

There were a variety of ways the strict settlement increased costs of land use and trade. First, life tenants or landholders could not easily convey or transfer ownership of settled land to third parties. A buyer of settled land may not be secure in his or her title because settlements were not public records and any remote family member with an estate could dispute the purchase far into the future. Second, life tenants could also not easily use or improve their resources, like mines, timbers, or buildings because their powers could be complex or unknown. Third, if an economic opportunity for improvement arose or there was an interested third party, the rights and powers could not immediately be changed to respond to the new opportunity: “No part of the settlement could be broken” unless by Act of Parliament, when the tenant in tail came of age (turned 21), or by a legal process called a common recovery.¹⁷

Parliament provided estate acts as a service to its constituents land and property-holding families.¹⁸ Estate acts could legally remedy any problem created by a strict settlement. They could authorize transactions and help landholders “convey property to other persons [..] put resources to new and better resources.”¹⁹ Parliament served as a forum where life tenants could respond to changing economic opportunities by the beginning of the eighteenth century.²⁰ After a series of procedural changes beginning in 1702, the parliamentary process for passing estate acts was streamlined, relatively inexpensive, and revealed information pertinent to securely transferring ownership of settled property.²¹ How the parliamentary system worked from 1660 to 1702, a crucial era in English political and economic history, however, is largely unknown.

2.2 Crown and Parliament in the Seventeenth Century

Crown–parliament politics influenced how parliament passed estate bills from 1660 to 1702. The Triennial Act of 1664 broadened the monarch’s royal prerogatives that controlled parliamentary meetings.²²Parliaments were called and dissolved, requiring general elections. A session of parliament formally began when the monarch called it and ended when the monarch decided to prorogue it. I focus on legislative activity in sessions of parliament in this paper because estate bills had to be passed during single sessions. When the monarch prorogued (ended) a session, any unfinished or still active bills automatically failed.

Both parliaments and sessions were generally called by the monarch to receive revenue. From 1660 to 1702, the monarchs had two main sources of revenue: hereditary revenues and parliamentary grants of revenue. Parliament settled a monarch’s hereditary or ordinary revenue at the beginning of his or her reign by passing a bill. Monarchs’ independence from parliament depended on their income from ordinary revenue as well as the terms of the bill. Charles II and James II were granted ordinary revenues from indirect sources of customs and excise for life.²³ Parliament settled ordinary revenue from customs and excise on William and Mary after the Glorious Revolution, but significantly altered the terms after the first several sessions of the new reign. The second source of income was from additional revenue bills called parliamentary grants or supply bills. These were often requested to fund war efforts. Both hereditary revenue bills and supply bills were legislated on during single sessions.

The monarch opened a session of parliament by making a speech in parliament and requesting “supply.” Once convened, parliament as an organization had legislative, judicial, and administrative functions that were essential to the workings of the English state. While in session, parliament began legislating on the requested revenues, the House of Lords heard court cases as the highest appellate court, and both Houses began legislative work on two general classifications of bills. They received petitions for and introduced (1) private bills, for example estate bills, and (2) public bills, for example regarding religion and trade.²⁴

Parliament faced a problem when passing estate bills because it did not have a preset or fixed amount of time to work. The monarch decided when to end a session, which was, in theory, unknown. Landholders, their lawyers and agents worked with MPs and peers to decide if and when to introduce estate bills as well as when to work on estate bills to pass a multi-layered process.

This process is summarized in figure 2. The process required attention and approval by both Houses of Parliament and had to receive the royal assent by the monarch before becoming an Act of Parliament. Estate bills could enter in either the House of Commons or the House of Lords during this era. The method of approval was identical in both Houses. The legislative process formally has twelve-steps. After a first and second public reading in the first house, a bill was sent to committee, where the committee would hear all concerned parties and make amendments (stages 1–3). It was then reported on from committee and amendments could be made. After a third public reading, the bill was approved by the first house (stages 4–5) and was sent to the second house for a similar process (stages 6–10). Any amendments made in the second house would be approved by the first house (stage 11). After the legislature’s approval, a bill received the royal assent from the monarch and became an Act (stage 12). The monarch could also withhold the royal assent and veto an estate bill, which was only done four times during this era.

Figure 2.

Legislative process. Notes: Stages of estate bill passage. The process required approval by both Houses of Parliament and had to receive the royal assent by the monarch before becoming an Act of Parliament. Estate bills could be introduced in either house, but the process was similar. Bills were approved by a committee where amendments could be made and the whole house. After the legislature’s approval, a bill received the royal assent from the monarch and became an Act. Source: http://www.parliament.uk/about/how/laws/passage-bill/.

Open in new tabDownload slide

2.3 Estate Bills: Success and Failure

I collected new information on all of the intermediate stages of estate bills’ passage, their “legislative life-cycle” from the Journals of the House of Lords and Journals of the House of Lords to provide insight to how parliament solved its problem.²⁵ Table 1 shows basic summary information on estate bills. From 1660 to 1702, most bills (67 percent) passed the entire legislative process and became Acts of Parliament (Panel A of table 1), but figure 1 shows how success varied under each monarch and improved markedly after 1688. Estate bills could fail because they were rejected by parliament. Panel B shows parliament rejected 26 percent of failed bills as recorded in the Journals. A rejection is when the Journals record that the bill was rejected or failed by a vote.²⁶

Figure 3 shows an example of estate bills’ “legislative life-cycle” in a session of parliament under William III.  Appendix Section A.2 provides details on the data collection and coding. The figure plots the stage of the legislative life-cycle (1–10) as a function of the ordered days in the session.²⁷ Stage 1 is when a bill was introduced; Stage 5 is when the bill was sent to the second house. Stages 2 and 7 are when the bill was sent to committee in either of the two Houses. Stage 10 is when the bill received the royal assent. An estate bill’s progress in a session of parliament over time is represented as a line (its “legislative path”). The slope shows how quickly parliament passed a bill. Acts of Parliament, successful bills, reached stage 10 and are full lines. The failed bills have shorter paths and are marked by bold lines.

Figure 3.

Normal Session, Session 33 (20 October 1696–16 April 1697). Notes: Author’s calculations from estate bill dataset. There were ten stages in legislative process as recorded in Journals. Stage 1 is bill introduction. Stage 10 is approval (royal assent) by the monarch. Full lines (end at Stage 10) denote successful bills (Acts of Parliament). Bold lines denote failed bills. Slope of line represents speed of passage.

Open in new tabDownload slide

The session in figure 3 met for about 6 months and estate bills moved relatively quickly through parliament. On average, bills took 1.6 months to pass all ten stages. Sets of estate bills were given the royal assent on the same day, which only occurred on five times during the session (the five “points” at stage 10). On each of those days, William III also signed at least one supply bill or an important public measure.²⁸ The evidence supports that parliament endogenously passed estate bills with supply or public bills.

Parliament rejected the failed bills during this session (bold lines in figure 3), possibly because a bill did not meet legal standards.²⁹ Table 2 summarizes information on the failed bills: a reference name, its house of origin, the stage and date of failure, reason for failure if recorded by the Journals, and if the bill became an Act in a later session. About 85 percent of the failed bills were rejected or have evidence of political or legal influence, e.g., a petition or counsel was heard. Only one bill was rejected, re-introduced, and passed to become an Act in the next session, which began 7 months later. The bill took 13 months to pass.

The patterns suggest that lawyers, MPs’, and peers’ decisions to introduce and pass estate bills were endogenous to when they passed supply bills or public measures, and possibly expect the session to end. Estate bills were never given the royal assent by themselves: they were always approved with supply or public bills. About 80 percent of sessions were prorogued when the monarch approved a supply bill.³⁰ Historians note that parliament used the power of supply in a strategy known as “redress before supply” where parliament presented its grievances to the monarch, expecting redress, and not passing a supply bill until the grievances were acknowledged or remedied. They could also use the strategy to pass other bills. This was a tactic acknowledged by Charles II, James II, and William.³¹

This strategy was not always effective.³² In particular, conflict and debates in parliament that were exogenous to estate bills could make the monarch more likely to directly interfere with parliamentary proceedings by suddenly proroguing the session. The monarchs’ speeches in parliament are illustrative of the decision to suddenly end a session of parliament. Charles II, for example, admonished the House of Commons when he suddenly prorogued the sixteenth session of parliament. He explained his reasoning to the House of Lords:

I have received an Address of such a Nature from the House of Commons, as I cannot but resent very highly, from the ill Consequences I have lived to see from such Addresses. I intend therefore to prorogue them for some short Time, in Hopes they will consider better what they ought to do at their Return. I have chosen to tell this to you first, because I would have you know I am very well satisfied with the dutiful Behaviour of this House; and you will by that Time be more enabled to give Me your Advice.³³

Hoppit (1996) hypothesizes that political tensions and conflict led to the “sudden and unexpected end of a session” that killed bills, such as estate bills, which otherwise would have passed.³⁴ As an example, figure 4 shows the legislative life-cycle for estate bills in a session of parliament that was suddenly prorogued under Charles II. The bills appear to be “cut-off” when the session ended. It is also puzzling that two bills (marked by x’s) were introduced only 3 days before the session ended, when there was presumably not enough time to pass the entire legislative process. Table 3 presents information on failed bills during this session. In contrast to the “normally prorogued session”, it shows that a majority of estate bills were still active when the session was prorogued. It also shows that failed bills never returned and did not become Acts of Parliament. Bills that were re-introduced were successful, but it took about a year and a half for to become an Act. I return to these patterns when discussing the economic effects of sudden prorogation below.

Figure 4.

Sudden prorogation in Session 15 (13 October 1675–22 November 1675). Notes: Author’s calculations from estate bill dataset. There were ten stages in legislative process as recorded in Journals. Stage 1 is bill introduction. Stage 10 is approval (royal assent) by the monarch. Full lines (end at Stage 10) denote successful bills (Acts of Parliament). Bold lines and x’s denote failed bills. Slope of line represents speed of passage.

Open in new tabDownload slide

3. The Effects of Sudden Prorogation: Empirical Analysis

I used the Journals of the House of Lords and the historical literature to code which sessions were suddenly prorogued. The Journals of the House of Lords record the monarch’s speeches given at the end of every session. The content of the speeches sometimes reflects that the decision was sudden. I also read detailed historiographies of each session, found in Ogg (1962), Horwitz (1977), and Smith (1999).³⁵ Panel A of table 4 shows 44 percent of sessions that had an estate bill enter were suddenly prorogued.

I estimate equation 1 using a probit model and thus assume $εit$ is distributed standard normal. For comparison and robustness I also use OLS and cluster standard errors on sessions of parliament.³⁶ The goal is to consistently estimate the causal effect of sudden prorogations on parliament’s functioning (the success of individual estate bills). There are several potential concerns when estimating equation 1.

First, unobservable factors could be correlated with a monarch’s decision to suddenly end the session. Landholders’ agents, MPs, and peers decided if to introduce a bill, when to introduce it, and which house to introduce it in. This decision may be a function of unobserved characteristics, such as the legal or economic quality of the bill or family resources or connections. I include X_it to help mitigate bias from the selected sample.

I include X_t to control for any time-varying observable factors that could also be correlated with bill success or failure. They are summarized in Panel B of table 4. Examples of such factors include changes in England’s state capacity, sessions meeting during war years, and House of Commons majority political party. Details on sources and discussion can be found in  Appendix Section A.3. X_t includes ConflictHouses_t, an indicator for sessions where there is recorded conflict between the two Houses. It also includes a year trend to control for any trend in bill success over time; WarSession_t, an indicator for sessions meeting during war years; DeficitRatio_t, the deficit ratio as a measure of England’s fiscal capacity; and MajorityParty_t, a categorical variable for majority party in the House of Commons during a parliament. X_it is a matrix of bill-level characteristics that help control for the effects of omitted variable bias at the bill-level. It also includes an indicator for bills that repeated in any session. Selected variables in X_it are shown in Panel C of table 4. Details on coding and the data can be found in  Appendix Section A.1.

Second, SP_t will not be consistently estimated if landholders or MPs somehow expected the session to end suddenly and decided to introduce or pass bills differently in these sessions. The historical literature suggests the decision was unanticipated. The description of “prorogation events” is stark. For a session meeting under Charles II, for example, Ogg (1962) describes the prorogation as “suddenly interrupting” the proceedings.³⁷ For a session meeting during the Exclusion Crisis, the Black Rod’s “ominous tap at the door” signaled the monarch’s decision.³⁸Horwitz (1977) describes that “the actual prorogation [of the second session under William and Mary] came as a surprise.”³⁹

There is some evidence to suggest that sessions that were suddenly prorogued had a different composition of bills. Table  B.1 shows more bills naming a member of the nobility had more estate bills, and fewer non-estate bills entered in sessions that were suddenly prorogued. More estate bills also entered the House of Lords in normally prorogued sessions. Including the detailed bill characteristics will help account for any bias this has on the estimate of SP_t. I investigate this issue further below.

A last main concern is that the monarch may have decided to suddenly prorogue a session of parliament to end estate bills themselves. This is unlikely. The monarchs had the power to veto individual estate bills, which was used only four times during this era. It was unlikely that the monarch was using the power of prorogation as another tool to end estate bills. Monarchs were aware that a sudden prorogation would have real effects on parliament’s legislative output.⁴⁰ Regarding estate bills, however, the Journals and historical literature provide little evidence to suggest it was done intentionally to end the estate bills themselves. The decision to suddenly prorogue a session was not random but was a function of political conflict and developments in a specific session of parliament. In particular, sudden prorogation interrupted parliament’s legislative or administrative initiatives when they conflicted with the monarch’s revenue bills and conflicted with the monarch’s other “policies” or interests. This is the focus of Section 4.

3.1 Results & Economic Implications

Figure 5 graphically shows the relationship between estate bill success and sudden closure of parliament. It shows years in which there was a sudden prorogation of a session relative to the success rate of estate bills. There is a clear negative correlation between sudden prorogations and estate bill success. There is a higher concentration of sessions that were suddenly prorogued prior to and immediately after the Glorious Revolution in 1688, capturing the period of growing political instability after 1665 under Charles II through the first few sessions of William and Mary’s new reign. Sudden prorogations no longer occurred after 1690.

Figure 5.

Sudden closures of sessions of parliament, 1660–1702. Notes: Figure shows years when monarch suddenly closed a session of parliament relative to success rate of estate bills. Sudden closures (prorogations): author’s calculations from Journals, Ogg (1962), Horwitz (1977), and Smith (1999). Vertical lines mark years where a session was suddenly prorogued. Success rate: author’s calculations from estate bill dataset. A “successful bill” was approved by House of Commons, House of Lords, and Monarch (became an Act of Parliament). Success rate is calculated by years in which an estate bill was introduced. Average success rate is conditional on years parliament met and had estate bills introduced. Parliament did not meet in 1672, 1676, 1682–1684, and 1686–1688.

Open in new tabDownload slide

Table 5 presents results from estimating equation 1. Columns (1–3) report the average marginal effect (AME) from the probit model. Columns (4–6) report OLS estimates. For the probit specifications, column (1) reports the raw correlation between sudden prorogation and estate bill success. Column (2) introduces political variables contained in X_t and column (3) includes bill-level covariates in X_it. For the OLS specifications, column (4) includes political variables and column (5) clusters standard errors on sessions of parliament. Column (6) includes bill-level covariates and clusters standard errors on sessions of parliament.

The results show that the monarch’s decision to suddenly prorogue a session of parliament had a significant effect on parliament’s passage of estate bills. Estate bills were about 19 percentage points more likely to fail if a session was suddenly prorogued. The negative effect is robust across all specifications and the magnitude is little changed by the specification. Comparing the AME to the OLS estimates suggests that the independence assumption is reasonable and leads to downward bias (towards zero) in the estimated AME.⁴¹

The results also provide insight to other factors influencing estate bill success. The last two rows of table 5 report the F-statistic for a test that the bill type and landholder characteristics are jointly equal to zero. Bill legal type appear to be important for bill success, while the evidence for landholder type is mixed. Table  B.2 in Appendix Section B provides the estimates of the political variables. There is no evidence to support that sessions with conflict between the two houses is associated with bill failure. A higher deficit ratio, when the government was less fiscally prudent or had lower capacity, is associated with lower estate bill success. Bills were also more likely to fail when the opposition had a majority in the House of Commons relative to the Court party. There is evidence that bills were more likely to succeed in sessions meeting during war years, sessions with more time, and if they entered the House of Lords.

The historical literature suggests the conflict that led to a sudden prorogation was largely unknown to landholders or their agents deciding to introduce a bill. Does the data further support this? If landholders or their agents somehow expected the conflict and sudden prorogation, we might observe more bills being introduced earlier in sessions that were suddenly prorogued. Figure  B.2 compares the average number of bills entering in 10-day intervals by sessions that were normally prorogued to those that were sudden for the first 50 days. As it is important for the estimate of SP_t, we do not observe obvious evidence that bills were introduced earlier in sessions that were suddenly prorogued, perhaps because the landholders or MPs expected conflict to occur.

Second, if parliament or agents were anticipating the end of a session, we may see bills being passed more quickly in these sessions. Table 6 shows there is no statistical difference between how quickly parliament passed estate bills in the first eight stages of the legislative life-cycle in sessions that were suddenly prorogued compared to those that were not. Bills moved more quickly through the final two stages in sessions that were suddenly prorogued. For bills that were able to make it this far in the legislative process, this suggests that parliament may have been anticipating the monarch in parliament and moved the bills quickly to be ready for the royal assent. The average slope of all bills is not statistically different in the two types of sessions.

Figure 5 shows that prorogations stopped in 1690. It is possible that a new system evolved or immediately changed after 1690 that would make “normal sessions” after the Glorious Revolution different than “normal sessions” before. If this happened, the estimate of SP_t will be biased downwards because of an unobserved factor influencing parliament’s passage of estate bills under the new regime.

The data suggest that the system may have slowly evolved after 1690. Figure  B.3 shows that the number of estate bills entering parliament slowly increases after a period of volatility from 1689 to 1693. Figures  B.4 and  B.5 shows that more bills were introduced in the first 25–50 days in sessions meeting under Charles II (1660–1681), James II (1685–1688). After the Glorious Revolution, bills appear to uniformly enter during the first 100 days of a session under William and Mary (1689–1694). Under William III (1694–1702), however, bills are mostly introduced after the fiftieth day of the session. To address this source of bias, I estimate equation 1 using data from sessions meeting from 1660 to 1694. For comparison, I also estimate equation 1 for sessions meeting from 1660 to 1690.

Columns (7) and (8) of table 5 report the most clean estimates from the two subsamples. Column (7) reports the estimate of SP_t from sessions meeting from 1660 to 1694 and column (8) reports the estimate from sessions meeting from 1660 to 1690. Estate bills were about 23 percentage points more likely to fail in sessions that were suddenly prorogued from 1660 to 1694. From 1660 to 1690, estate bills were 45 percentage points more likely to fail in sessions that were suddenly prorogued. The estimates both increase relative to the results from the full-sample, suggesting that the results in columns (1)–(6) serve as a lower bound of the effect of sudden prorogations on parliament’s passage of estate bills.

As a second exercise, I use the information from the Journals on reason for bill failure as an outcome variable. This is motivated by the pattern that less bills were rejected by parliament in the session that was suddenly (Table  A.3 in Appendix Section A.2). We would expect bills failing in sessions that were suddenly prorogued to have significantly less rejections. I coded three different outcomes: (1) bills that were rejected, (2) bills rejected or failed in committee, and (3) bills that were rejected or had a petition or counsel heard in its last stage. I use probit and OLS models to test this hypothesis. Table  B.4 in Appendix Section B reports the results and a discussion. The statistical evidence is less robust, but the coefficients are negative and supports the hypothesis that bills were less likely to be outright rejected, and therefore still active, if a session was suddenly prorogued (figure 6).

Figure 6.

Bills not introduced earlier in sessions suddenly closed. Notes: Author’s calculations from estate bill dataset. Figure compares the average number of bills introduced in the first 50 days of a session in 10-day increments in sessions suddenly closed and those that were not. NP: sessions not suddenly closed (prorogued). SP: sessions suddenly closed (prorogued).

Open in new tabDownload slide

Table 7 shows the economic implications of the sudden closure of parliament on three dimensions: scarring, delay, and displacement of new bills. Bills that failed because of sudden prorogation could be passed in a later session and can therefore be thought of as “projects.”⁴² A project could have one of four outcomes: (1) fail after one attempt, (2) fail after multiple attempts, (3) succeed after one attempt, and (4) succeed after multiple attempts. Table  A.1 in Appendix Section A.1 presents the number of projects that have these different outcomes. There were 833 projects (as opposed to 954 bills). Most of the projects (68 percent) succeeded in their first and only session of parliament. Most projects that ultimately failed did so in their first and only session of parliament.

Panel A of table 7 shows that a sudden prorogation had scarring effects. Projects first entering in a session that was suddenly prorogued were significantly less likely to succeed and ever become an Act of Parliament by about 15 percent. If a bill was successful, it was still delayed by about 7 months (Panel B). Panel C provides evidence to suggest that new bills were displaced in sessions after a sudden prorogation. Estimates from negative binomial and poisson regressions of counts of new bills on a lag of sessions that were suddenly prorogued show that there were fewer new bills in a session meeting after a sudden prorogation. OLS estimates of shares in sessions of parliament also show this effect. In total, the evidence in table 7 shows that main economic impact of sudden prorogation was that most bills failed and were not successful again.

4. Political Conflict, Sudden Prorogations, and the Glorious Revolution

This section examines why the monarch suddenly prorogued sessions of parliament. I first show that three political factors are statistically associated with sudden prorogations before 1688. Second, I discuss political and constitutional changes made with the Glorious Revolution that made sudden prorogations less likely after 1688.⁴³ I then analyze the direct impact of political conflict on estate bill success.

4.1 Why Did Sudden Prorogations Occur?

Three factors likely led the monarch to suddenly prorogue a session and interrupt parliamentary proceedings: (1) a monarch’s independence from parliament (IndirectRevenue), (2) conflict influencing parliament’s passage of revenue bills (ConflictRevenue), and (3) conflict with the monarch’s “policies” (ConflictMonarch) explained below. The three political factors focus on the crown–parliament relationship, which was one of the dominant political themes of the era.⁴⁴ The economic history and historical literature argues the Glorious Revolution resolved or altered the nature of this conflict and made the monarch more dependent on parliament.⁴⁵

Table  A.3 in Appendix Section A.3 summarizes the three variables and sources. IndirectRevenue_t is a series of indirect revenues (customs and excise) and serves as a proxy for the monarch’s financial independence from parliament. Indirect sources were dependent on the state of trade and could make the monarchs less dependent on calling parliament or keeping it in session to receive parliamentary grants of revenues. ConflictRevenue_t is an indicator for sessions where there was conflict in a session that prevented or stalled the passage of a revenue bill. For example, parliament investigated the administration of finances rather than passing a new revenue bill during a session under Charles II in 1669.⁴⁶ Under William and Mary, in 1689, there was disagreement in parliament over the settlement of the ordinary revenue, slowing down the process and making amendments that “were not to William’s taste.”⁴⁷

ConflictMonarch_t is an indicator where there was conflict in parliament with a monarch’s interests or policies. It occurred when there was disagreement in parliament regarding a royal policy or legislation was initiated that was divergent from royal interests. For example, under Charles II, parliament began work on a bill that would influence the succession of the monarchy by requiring heirs to be raised as Protestant. Under William III, parliament had failed to act on most of the king’s recommendations regarding trade and poor relief.⁴⁸

Table 8 presents correlations between SP_t, sessions that were suddenly prorogued, and the more fundamental sources of conflict and tension between the crown and parliament. Panel A shows the correlations for sessions meeting under Charles II and James II. Panel B shows the correlations for sessions meeting after the Glorious Revolution under William and Mary and William III. Under Charles II and James II, the patterns show that the three political variables are positively associated with sessions that were suddenly prorogued. IndirectRevenue_t and ConflictRevenue_thave the strongest relationship with SP_t. ConflictMonarch_t also has a positive association with the monarch’s decision to suddenly prorogue a session of parliament. It is less strong than the tension and conflict over revenue. The sessions in which conflict with the monarch did not lead to sudden prorogation were because Charles had yielded in order to obtain supplies.⁴⁹ Panel B shows correlations after the Glorious Revolution under William and Mary and William III. Sudden prorogations are less likely to occur in general and ConflictRevenue is the only variable still positively associated with sessions that were suddenly prorogued.

Table 9 shows OLS estimates of the conflict variables on sudden prorogations. Columns (1–3) include only the conflict variables. Column (4) introduces other political controls described in Section 3. Column (5) includes a post-1688 indicator. Both ConflictRev and ConflictMonarch remain positively associated with sessions that were suddenly prorogued. Different from the correlations in table 8, IndirectRevenue is negatively associated with sudden prorogations except in column (5), where it is no longer significant when including the Glorious Revolution indicator. The estimate of the post-1688 indicator shows that sudden prorogations were less likely to occur after the Glorious Revolution.

The patterns in table 9 are insightful to the effects of the Glorious Revolution. First, conflict delaying a revenue bill, ConflictRevenue remains strongly associated with sessions that were suddenly prorogued. After the Glorious Revolution, the monarch suddenly closed sessions because of debates surrounding the settlement of the new monarchs’ ordinary revenue. The question of settlement raised more “fundamental questions about the future relationship between the crown and parliament” among the law-making MPs.⁵⁰ It was an open issue for the first several sessions of William and Mary’s reign: that William and Mary did not have “settled revenues over the first two years of their reign alter[ed] their bargaining relationship” with parliament.⁵¹ As argued by Cox (2012), parliament also engineered supply bills differently after the Glorious Revolution. Parliament began using time limits on tax revenue bills that parliament could terminate, increasing the monarchs’ present value of revenues sanctioned by parliament.⁵² They also legislated bills such that government debt was funded on parliament’s terms.⁵³ The new constitutional changes regarding the ordinary revenues and new types of supply bills appears to have resolved conflict sufficiently that it no longer led to sudden prorogations after 1690.

IndirectRev_t is not associated with SP_t once the Glorious Revolution dummy is included. This result complements the evidence on ConflictRevenue. The historical literature suggests that this is because the settlement of the ordinary revenue in 1690 “shortened the financial leash of the monarch.”⁵⁴ In 1660 parliament settled tax revenue from customs and excise on Charles II and his heirs for life. James II also received this ordinary revenue in 1685. Parliament, however, did not settle the ordinary revenue on William and Mary until 1690 because it could not agree on the terms. When it did pass the legislation, parliament made significant constitutional changes to the terms, only settling the excise for life and the customs for a term of 4 years. Revenues from excise for their lives were granted; the customs, however, was limited to 4 years.⁵⁵

The results also show that ConflictMonarch_t loses some of its impact on SP_t after the Glorious Revolution indicator is included. The evidence shows that conflict with monarch was more frequent under the late Stuarts when parliament was in session, and could lead the monarch to interfere with parliamentary proceedings. The findings support arguments by Roberts (1993) and Pincus (2009) that the Glorious Revolution changed the nature of a fundamental type of conflict and tension in England’s government. Under Charles II, conflict with the monarch’s policies were largely driven by religious policy and altering succession of the monarchy.⁵⁶ Groups in parliament, for example, worked on bills to influence and alter the succession of the monarchy. The most well-known example of this is the Exclusion Crisis. From 1679 to 1681, a growing opposition in parliament initiated legislation that would exclude James from becoming monarch because of his conversion to Catholicism. After the Glorious Revolution conflicts with the monarchy centered on different types of legislative and policy issues.⁵⁷ In 1699 for example, parliament “had failed to act on most of the king [William III]’s recommendations” by not passing legislation to pay off debts incurred by war, to provide relief of the poor, and to advance trade.⁵⁸ Crucially, while conflicts with the monarch and regarding supply did occur after 1688, they did not necessarily lead to sudden prorogations.

Last, there is evidence that the Glorious Revolution had an impact on making sudden prorogations less likely in addition to the constitutional and political changes that affected the specific conflict variables (Column (5)). This could embody new political norms in England’s government. The only formal constitutional change was the adoption of the Bill of Rights.⁵⁹ Perhaps there was an additional effect of the Glorious Revolution through who was elected or larger societal changes.⁶⁰ This is an important avenue for future work.

4.2 The Effects of Political Conflict: Empirical Analysis

Table 10 presents results from estimating equation 2. Columns (1–5) report the average marginal effect of a probit model; columns (6–8) report OLS estimates.

The results highlight the impact that conflict over revenue bills had on parliament’s passage of estate bills under Charles II and James II. The estimates are negative and little changed by the specification. Estate bills were about 20 percentage points more likely to fail in sessions in which there was conflict over revenue bills under Charles II and James II. The estimates underscore the importance of resolving this conflict with the Glorious Revolution in improving parliament’s functioning. Complementary to the results in table 9, there is evidence to suggest that the Glorious Revolution had an additional effect on improving parliament’s passage of estate bills beyond resolving or changing the effect of conflict (columns (3) and (6)). The bottom two rows of columns (5) and (8) present test statistics for the hypothesis that the Glorious Revolution had an impact on estate bill success beyond constitutional and political changes specific to the conflict variables. Specifically, row (a) tests the hypothesis that the coefficient on ConflictRevenue (π₁) is equal to the coefficients on the Glorious Revolution dummy plus the coefficient on the interaction term with ConflictRevenue (π₄ + π₅). Row (b) presents the test statistic for the coefficients on both ConflictRevenue and ConflictMonarch. Both sets of test statistics reject that the coefficients are equivalent. This supports that the Glorious Revolution had an additional impact not captured by the specific interaction terms that made estate bills more likely to succeed.

Figure 7 shows the counterfactual success rate of estate bills had the Glorious Revolution not occurred and a “Stuart-like” system continued until 1702. Specifically, it shows how well the model fits the data from 1660 to 1689 from estimating the effect of ConflictRevenue_t on estate bill success. From 1689 to 1702, it shows the forecasted (out of sample predicted success rate). The model predicts that the success rate of estate bills would have been lower had the Glorious Revolution not occurred. Economically, this implies that political barriers would have continued to make it difficult to use parliament as a forum to change and reorganize property rights in preindustrial England.

Figure 7.

If Glorious Revolution did not occur: counterfactual estate bill success rate, 1689–1702. Notes: Glorious Revolution allowed parliament to effectively reorganize property rights to land. This figure shows the counterfactual success rate relative to the actual success rate of estate bills. A “successful bill” was approved by House of Commons, House of Lords, and Monarch (became an Act of Parliament). Solid line shows changes in the actual success rate from 1660 to 1702. I construct the counterfactual success rate of estate bills in two steps. I first estimate equation (1) with fiscal conflict and political and bill-level controls on the subsample of estate bills entering parliament from 1660 to 1688. From 1660 to 1688, the dashed line shows the fitted success rate from this estimation. Second, I predict individual bill success using equation (1) for estate bills entering parliament from 1689 to 1702. I then take yearly averages to obtain the counterfactual success rate (dashed line) after 1688.

Open in new tabDownload slide

5. Conclusion

This paper provides new evidence that conflict in England’s government at the end of the seventeenth century prevented the reorganization of property rights through parliament. Specifically, it shows that the monarchs’ decision to suddenly prorogue, or close, sessions of parliament made estate bills, a legislation that reorganized property rights, more likely to fail. The evidence shows that political and constitutional changes with the Glorious Revolution of 1688 resolved or changed the nature of key conflicts in government, making the monarch more dependent on parliament and sudden prorogations less likely to occur. These changes in turn improved parliament’s passage of estate acts.

Infrequent and sporadic meetings of parliament were endemic across preindustrial Europe.⁶¹ While this can be used as evidence that monarchs were less constrained, this paper provides evidence that a parliament that meets can also provide services and public goods that help coordinate economic activity.⁶² The evidence shows how the Glorious Revolution was important for credibility in government by improving parliament’s position via the monarch and allowing it to operate.⁶³ This era in English economic history is often used as an important historical example for understanding the connection between property rights and economic development because property rights became more secure by decreasing expropriation risk.⁶⁴ This paper shows how the Glorious Revolution was important to a different aspect of property rights change in England that is also important for understanding economic development. Property rights in England during this era were complex and not easy to change, preventing land and property from being used as a mobile factor of production, traded in land markets, or to support other transactions. Changes after 1688 removed political barriers to parliament’s operation, allowing it to become an adaptable institution and forum where individuals, families, and communities could reorganize their property rights to changing economic opportunities.

Acknowledgements

I thank Dan Bogart, Gary Richardson, and Stergios Skaperdas for critical feedback, encouragement, and guidance. I am also grateful to Nikolaus Wolf and participants at the Fast Track meeting for valuable comments and suggestions. I thank Kim Makuch and Sebastian Hohmann for detailed and thoughtful comments on earlier drafts, Mayuri Chaturvedi, Amihai Glazer, Walker Hanlon, Alexander Luttman, Marco Martinez, Cody Nehiba, Paulina Oliva, Stephen Quinn, Jean-Laurent Rosenthal, Kevin Roth, Thales A. Zamberlan Pereira, as well as seminar and conference participants at EHES 2017 Conference, Queen’s University Belfast, University of Warwick, UC Davis, UC Irvine, Caltech, RCI Institutions Workshop, and University of Tampere for useful suggestions at various stages. Funding from the All-UC Economic History Summer Research grant provided financial assistance. Russell Frankston, Charlotte Densham, and Anna Dimitruk provided excellent research assistance. All errors are my own.

Conflict of interest statement. None declared.

Footnotes

References