Financial Innovations and Political Development: Evidence from Revolutionary England

Home , Broad Hinton, English Civil War, English Revolution, John Harris (Royalist), John Selden

Saumitra Jha∗ Stanford University

October 2012

Abstract The English Parliament’s struggle for supremacy in the seventeenth century was crucial for the development of representative government in the English-speaking world, yet its lessons continue to be debated. This paper provides the first systematic evidence on the determinants of individuals’ decisions to join the coalition for revolutionary reform. The paper employs a novel micro-dataset on the endowments of each member of the Long Parliament (1640-60) that initiated England’s institutional transformation and finds that the key determinants of support for reform were overseas interests and other factors over which the executive enjoyed discretion under the existing constitution. Further, investment in newly available shares in overseas companies appears central in fostering support for reform, chiefly among those lacking prior overseas interests. The paper argues that the innovation of shares allowed new investors to take advantage of emerging economic opportunities overseas, aligning their interests with overseas traders. However, since these shared opportunities were heavily exposed to executive discretion, financial innovation broadened support for parliamentary control of government.

JEL codes: O10, F10, K00, P10, N23 Keywords: Institutions, Civil War, Conﬂict, Constitutional Reform, Economic Growth, Financial Markets, Trade, Political Economy, Revolution

∗Address: [email protected]; Graduate School of Business, 655 Knight Way, Stanford CA 94305. This is a revised version of paper circulating as “Financial Innovations and Political Develop- ment”. I owe particular thanks to Susan Athey and Avner Greif as well as Elhanan Helpman, four anonymous referees and Ran Abramitsky, Amrita Ahuja, Ken Arrow, Dan Bogart, Ernesto Dal Bo, Ann Carlos, Mauricio Drelichman, Jeffry Frieden, Sean Gailmard, Oscar Gelderblom, Claudia Goldin, Luigi Guiso, Eric Hilt, Kimuli Kasara, Peter Koudijs, Timur Kuran, David Laitin, Jessica Leino, Kris Mitch- ener, Aprajit Mahajan, Noel Maurer, Ted Miguel, Pedro Miranda, James Robinson, Matthias Schündeln, Carmit Segal, Jordan Siegel, David Stasavage, Nathan Sussman, Barry Weingast and seminar partici- pants at Clio, PacDev, Berkeley, Harvard, NYU, Penn, Princeton and Stanford. Zac Peskowitz provided excellent research assistance. This research benefits from articles made available prior to publication by the History of Parliament Trust.

1 1 Introduction

The Father of Parliaments, which ﬁrst rendered Parliaments supreme, and has since set the world upon the chase of Parliaments . . . - Thomas Carlyle (1845, pg.316), referring to the Long Parliament.

The seizure of executive authority by Parliament from the Crown in the years spanning England’s Civil War (1642-48) was arguably central for the development of representative government in the English-speaking world.1 Before the Long Parliament of 1640-1660, the king of England enjoyed independent sources of finance and possessed “sovereignty” rights over foreign policy, including rights to declare war, to collect customs and to charter monopolies on most goods and innovations introduced from abroad. The king called and dismissed Parliament at will. Between 1629 and 1640, no parliament sat in England. The summoning of parliament in 1640 initiated a remarkable process of institutional change. Through both legislative and ultimately violent means, the “Long” Parliament acquired rights to convene without royal approval, to control state finance and to direct foreign policy and war: in other words, Parliament acquired the Crown’s sovereignty rights (Howat, 1974, Smith, 2003). These institutional changes have been since recognized as the beginning of a path of experimentation that left England with the one of the world’s most enduring, and most commonly imitated, traditions of representative government.2 With the acquisition of executive authority, the Parliamentary victors of the Civil War instituted dramatically new policies. Large investments in a particular set of public goods– the Navy– supported the expansion of England’s international trade. These policies are credited with London’s emergence as the most important trading hub of Europe by the early eighteenth century and Britain’s acquisition of an empire that spread the influence of its institutions– including the common law, parliamentary paramountcy in government, and the Bill of Rights– around the world (Rodger, 2004, de la Escosura, ed, 2004, Ormrod, 2003, Ferguson, 2002). Given the importance of this episode for political development in Britain and elsewhere, the key puzzle, among the most famous and enduring in political economy–

1For example, Friedrich Hayek (1960, pg.160) suggests:“Out of the extensive and continuous discussion . . . during the Civil War, there gradually emerged all the political ideals which were thenceforth to govern English political evolution.” See also Murrell (2009). 2The transfer of the rights over state ﬁnance and foreign policy to Parliament was intermittently chal- lenged following the Restoration of the monarchy in 1660, until being consolidated following the Glorious Revolution of 1688. This has led to debate about the relative importance of the Civil War and Glori- ous Revolution in England’s subsequent development (see e.g. Murrell, 2009). However it is relatively uncontroversial that the activities of the Long Parliament made subsequent reforms possible (Pincus, 2009).

2 how a novel, broad and ultimately successful coalition in favor of representative government was formed in seventeenth century England– has been a focus of much debate ever since (Hobbes, 1682, Hayek, 1960, Moore, 1966, Acemoglu, Johnson and Robin- son, 2005b). Contemporary explanations for England’s revolutionary reforms include the Protestant ethic, the emergence of a commercial “middle” class seeking to protect its new wealth or a coalition formed to defend property rights in response to excessive executive greed (North and Weingast, 1989, World Bank, 2002, Acemoglu, Johnson and Robinson, 2005b, Rajan and Zingales, 2003, de Lara, Greif and Jha, 2008, Murrell, 2009). This paper provides the first direct evidence on the importance of different endowments on individuals’ decisions to join the coalition in favor of Parliamentary control of government during the English Civil War. It uses a novel micro-dataset collected by the author on the economic and social endowments of each of the 548 members of the Long Parliament (1640-1660) that initiated England’s revolutionary reforms, and finds that a range of domestic wealth endowments show little effect on support or opposition to Parliamentary takeover of executive authority. This suggests that constituencies formed around the defence of existing domestic property, whether old or newly acquired, do not appear to have played a major role in the initiation of England’s transition to representative government. Instead, the paper documents that, consistent with a simple theoretical framework, support for Parliamentary control of government was most impacted by those areas over which the executive had greater discretion under the existing constitution, and thus were most likely to change with a change in regime. These included the Crown’s “sovereignty” rights over foreign policy, overseas trade and colonies, which governed the return on new opportunities for wealth that accompanied the discoveries of direct sea routes to the New World and Asia. Further, the paper shows that a financial innovation–the introduction of shares in new overseas joint stock companies– played a significant role in transforming fragmented interests into a broad coalition. The paper finds evidence of an alignment effect– that shares, while having little effect on those with existing mercantile endowments, had a robust influence on the propensity to support constitutional reform among non-merchants (who were mainly sedentary landowners). In fact, shares appear to have been pivotal in swaying non-merchants in favor of reform, changing support for Parliamentary control of government from a minority position to one enjoying majority support. Non-merchant shareholders encompass a number of the central organizers of the Parliamentary rebellion, who far from being extremists, would have been slightly more likely to support the Crown had they not invested in overseas ventures.

3 The paper uses two complementary means to assess the causal effect of shares. First, the paper compares shareholders to non-investors matched along a range of those endowed characteristics that might plausibly lead to departures from the mean-variance efficient financial portfolio where every agent invests in shares (Markowitz, 1959). The paper finds that even under a lower bound estimate, the effect of shares was sufficient to push the support of the majority of MPs from Crown to Parliament. The second approach exploits a plausibly exogenous shock to the propensity to invest overseas: the spike in nationwide enthusiasm for foreign expeditions among agents who became free to control their finances just after Francis Drake’s successful circumnavigation and raid on Spain’s silver fleet. This regression discontinuity design also shows consistent increases in the propensity to support parliamentary control of government among shareholders. Beyond the stark choice to support Crown or parliamentary control of government during the violence of the Civil War, the paper traces a direct link between overseas shareownership and support for constitutional reform throughout the lifecycle of the early struggle for parliamentary supremacy, exploiting surviving records that indicate support for reform during the initial attempts to change the constitution through legislative means, at the eve of the outbreak of violent hostilities, and in the parliament of post-Civil War victors that implemented dramatic investments in England’s navy in defence of trade. A consistent picture emerges: the introduction of shares appears to lead non-merchants to make similar political decisions to those with prior overseas trade interests, creating a majority coalition that favored revolutionary reform. The paper interprets these results as reflecting the role played by shares in aligning the interests of sedentary non-merchants, who otherwise had few opportunities to invest overseas, with merchants, who already had access to such opportunities, in favor of constitutional reforms to acquire the Crown’s newly valuable overseas rights. Because the rights needed to profit from overseas investment belonged to the executive, the introduction of shares aligned the incentives of a broad coalition in favor of constitutional reforms aimed at seizing control of these rights. These rights were then used to increase England’s public investments in favor of policies, such as an expanding Navy, that enhanced the value of overseas investments and placed England at the center of world trade. Beyond shedding light on a pivotal moment in England’s political and economic development, this paper relates to a number of important literatures in finance, institutional and development economics. Much blame for under-development around the world has been attributed to a failure to align the incentives of disparate interest groups or “constituencies” in favor of political reform and beneficial public policies such as the reduction of barriers to entry in education, accessing credit and engaging in trade (Benmelech and

4 Moskowitz, forthcoming, Rajan, forthcoming). A growing body of evidence suggests that societies with different pre-existing ethnic and social interests are more prone to civil conflict, provide fewer public goods and suffer diminished growth trajectories (e.g. Alesina, Baqir and Easterly, 1999, Alesina and La Ferrara, 2005). An important “political economy” tradition has emerged that emphasizes the role played by disparate economic constituencies in impeding development and instead harnessing institutional reforms and regulations to create barriers to entry and protect rents (e.g. Benmelech and Moskowitz, forthcoming, Haber and Perotti, 2010, Rajan and Ramcharan, 2008, Perotti and von Thadden, 2006, Acemoglu, Johnson and Robinson, 2005a). Looking within one country does not allow the relative importance of England’s pre- existing institutions to be assessed, however the use of detailed individual data does permit measurement of the relative importance of different constituencies in a process of institutional change that proved crucial for England’s development. Unlike a number of papers in this literature, where interests form around differences in wealth, and shocks to existing wealth provide the impetus for change, this paper argues that an important role appears to have been played by shocks to future opportunities for wealth, in combination with financial mechanisms that allowed disparate groups to share in those opportunities, in building coalitions for reform. With the introduction of financial mechanisms to share in future opportunities for growth and wealth potentially much easier to influence through policy than attempting to reshape long-lived institutions driven by historical accidents or existing distributions of wealth and capital, the paper provides a more hopeful message for policymakers than many existing lessons drawn from England’s experience. This paper also relates to an important literature on privatization.3 The potential role that can be played by financial instruments in creating broad constituencies in favor of private property rights has found significant resonance in theoretical and policy circles, most notably motivating voucher privatization in the post-transition Czech Re- public and in Russia (Boycko, Shleifer and Vishny, 1994). These theories are backed by some evidence that underpricing of privatized assets is used by right-wing governments to consolidate domestic support in favor of lower taxation (Biais and Perotti, 2002, Jones, Megginson, Nash and Netter, 1999) and that allocating shares may shift political allegiances in favor of less redistribution (e.g. Duca and Saving, 2008, Kaustia and Torstila, 2008). This paper complements this work by examining an environment where the introduction of shares created access to new opportunities for future wealth, rather than implying an increase in individuals’ current assets. In England, the introduction of shares appears to have actually led to support for higher taxation, funding policies, such as the

3Megginson and Netter (2001) and Haber and Perotti (2010) provide useful overviews.

5 expansion of the Navy, that favored these new investment opportunities.4 Similarly, this paper builds upon an important literature on the role of news and attention in mitigating the extent of the “stockholding puzzle”– deviations in individuals’ investments from the optimal market portfolio, which should include shares (e.g. Guiso and Jappelli, 2005, Barber and Odean, 2006). In common with the mature stockmarkets that are the main focus of this literature, the paper finds that in a nascent market, a “news” shock common to elites generated significant increases in share ownership that in turn led to the strengthening of representative political institutions. The paper also contributes to a prominent debate, beginning with La Porta, Lopez-de Silanes, Shleifer and Vishny (1998), on whether countries with common law legal origins enjoy better investor protections that encourage diffuse corporate ownership (Haber and Perotti, 2010). This paper documents that in England, the originating nation of the common law tradition, the causal relationship went in the opposite direction: the diffusion of share ownership encouraged both constitutional reform and improved legal protections for investors. The potential positive feedback seen in England between diffuse corporate ownership, representative government and legal protections suggests an answer to why nations that transplant only one of these institutions, such as the common law, often develop along very different trajectories. Section 2 provides a brief background and a simple theoretical framework to understand the relation between endowments and choices that motivates the empirical methodology. Section 3 introduces the new micro dataset on each member of Parliament collected for this study (with further details in a Data Appendix). Section 4 presents the results and Section 5 concludes. A detailed Historical Appendix draws on new statistical evidence to document the relevant historical and constitutional context.

2 Background and empirical strategy

The English Civil War began as a struggle between the monarch and members of the Long Parliament over the “prerogative”: the rights of the executive (see Historical Appendix for details). As a major turning point in the political development of England and of parliamentary supremacy in government, the lessons of the English Civil War have been debated by many of the most prominent political economists ever since. For Karl Marx and his intellectual successors, England’s Civil War was the “First Bourgeois Revolution”

4Eﬀective taxation in fact rose considerably after the Civil War and Glorious Revolution. This diﬀerence is consistent with agents funding increased military action in support of their new interests overseas, rather than a broad attempt to limit redistribution, as one might expect if a coalition of wealthholders was driving institutional change (Brewer, 1989).

6 and as such an inspiration for revolutions in France and beyond (Stone, 1985).5 Among modern economists, the seminal contribution of North and Weingast (1989) argues that the Civil War and the Glorious Revolution of 1688 took place to protect the existing wealth and property rights of property owners who were faced by the unusual avarice of Stuart monarchs. The successful removal of kings in both cases yielded a credible threat that then allowed future rulers to commit not to expropriate property, leading to dramatic financial and fiscal development in England. Alternative interpretations by Rajan and Zingales (2003) and Acemoglu, Johnson and Robinson (2005b) also stress the importance of property rights, contending that the struggle was instigated by newly- enriched groups, whether a newly commercialized gentry in the former case, or merchants who had benefited from the rise in Atlantic trade in the latter.6 In all three modern interpretations, a constituency of wealth holders emerged for institutional reform that was able to obtain improved protection of domestic wealth and property rights, leading both to representative government and to economic growth. However, these theories have proven difficult thusfar to reconcile with a body of indirect empirical evidence that suggests that domestic property rights were already relatively secure in seventeenth century England, at least for the wealthy, and did not experience much change thereafter. For example, Clark (1996) finds no appreciable reduction on the interest rates on land due to the Glorious Revolution, as might be expected with a lower risk of expropriation. In fact, he shows that property rights over freehold land had been relatively strong throughout the seventeenth century. Similarly, examining the lending history of a London financier, Quinn (2001) fails to find a fall in interest rates on private capital over this period. Sussman and Yafeh (2002) reveal that the interest rates on government debt responded more to Dutch capital markets and England’s wars than to any reduced political risk from the Glorious Revolution.7 Since the central contest in the Civil War was between King and Parliament, theories

5On the role of the English Revolution on inspiring the French, see Pincus (2009). A prominent alternative to the Marxian view, associated chiefly with Whig historians, sees England’s revolution in the seventeenth century as being a response to challenges to liberty and property rights by an increasingly absolutist executive rather than about class struggle (Macaulay, 1898). A third “Revisionist” perspective has pointed to particular episodes of amity between King and Parliament in the years leading up to the Civil War to challenge the notion of broad economic or political struggle, and instead revives old views that religious differences, such as Protestant ideology, were central in what was once called the “Puritan Revolution” (Gardiner, 1883, Morrill, 1993). 6The hypothesis that newly commercialized rural landowners were the major supporters of reform finds resonance with interpretations by Barrington Moore (1966) and Tawney (1941). See also Brenner (1993), who emphasizes the role of new merchants. 7See also de Lara et al. (2008) and Murrell (2009). These effects may not have been uniform, however. Once Parliament consolidated its control, charters for England’s transport infrastructure and organization became more secure (Bogart, 2009).

7 about the identity, motivation or organization of groups responsible for England’s constitutional development should ﬁnd validation in the observable history of who supported and opposed the monarchy among England’s elected representatives. A particularly valuable feature of using these data is that the political allegiances of virtually all members were publicly revealed by their actions during the Civil War. Parliamentary fence-sitters on the eve of the Civil War were forced to choose between accepting the conﬂicting sum- mons of the King to Oxford and of Parliament to Westminster. By 1644, every living parliamentarian could be associated with one side or another (Brunton and Pennington, 1954) (see also Data Appendix).8

2.1 A simple theoretical framework

A simple theoretical framework can shed light on how to interpret the relationship between endowments, property rights and political choices. Suppose that the expected utility for a member of parliament (MP) can be summarized by the following additive relationship: X Ui = βjxij + ur (1) j

where xij are predetermined individual endowments, βj represent the rates of return to

xij and ur contains other orthogonal factors that inﬂuence expected utility in a state of the world r. Suppose that the support of an individual agent increases the chances of victory by an amount s > 0. Suppose that each agent believes that with probability µ, 9 Parliament (P) will win the struggle against Royal authority (R). Let βz, z ∈ {P,R} denote the rate of return on endowment j in the state of the world where either the monarchy (R) or Parliament (P) won. Then the agent’s problem is to choose to support

8The well-documented allegiances of Parliamentarians differs from other public figures. The loyalties of local leaders were confounded by both local power politics and the presence of occupying armies. Even London, often seen as the epicenter of Parliamentary power, initially had a strong Royalist presence among its leadership, including a Royalist Lord Mayor. 9Naturally, we expect µ to be affected by other agents’ choices. For plausible specifications of the multi-agent game, we would expect multiple values of µ to be consistent with equilibrium. However, as shown below, the specific realisation of µ is irrelevant for an agent’s decision, as long as µ ± s is interior. This condition: that there is some uncertainty about whether Parliament or monarchy wins regardless of an individual agent’s choices– makes sense in the historical context.

8 parliamentary or monarchical control: " ! ! X X max (µ + s) xijβj|P + uP + (1 − µ − s) xijβj|R + uR, , z∈{P,R} j j ! !# X X (µ − s) xijβj|P + uP + (1 − µ + s) xijβj|R + uR, j j

The optimal choice implies a cut-off strategy: an agent will choose to support Parliament if the value from supporting Parliament exceeds that of supporting the monarchy. Sub- tracting the values above reveals that an agent will choose to support political reform if: ! X s xij[βj|P − βj|R] + (uP − uR) > 0 (2) j The inequality (2) establishes that, given the linear utility specification above, a sufficient condition for an agent’s decision to support political reform to be invariant to the agent’s exposure to any particular endowment xij is that βj|P = βj|R. In other words, support for political reform will be unaffected by an endowment if the value of that endowment is the same regardless of regime. This will occur when there are believed to be secure property rights for that endowment. Furthermore, in this formulation, the condition above is in fact also necessary for irrelevance of an endowment if MPs believe that their choice will have a non-zero effect on the outcome (s > 0). This condition lends itself to an empirical test: if s = 0, it implies that all endowments are irrelevant at the same time, and thus a joint test of the significance of all endowments should be zero. The regressions below are sufficiently significant to reject this test. The inequality (2) also implies that those endowments that play the biggest role in determining an MP’s support or opposition to political reform will be those most subject to change in value based on the identity of the regime. These are likely to be endowments that fell within the “prerogative” rights of the king under the existing constitution– including those linked to overseas opportunities, trade, religion and royal patronage (see Historical Appendix). All of these could be expected to change with Parliament’s seizure of control over these rights.

Inequality (2) further yields an implicit condition on the minimum uP − uR required

9 for support for Parliamentary control. The probability of supporting Parliament is: ! X P{P } = F sxij[βj|P − βj|R] (3) j where F (·) is the cumulative density function of uP −uR. Assuming that uz are normal or uniform, Equation (3) can be estimated using standard probit or OLS respectively. The coefficients γj from such a regression identify γj = s(βj|P − βj|R). Note that s cannot be identified, but must be non-negative.10 Thus, the sign of expected changes in the value of the endowment can be inferred by inspecting the coefficients, but marginal effects are necessary to assess the relative magnitude of such changes.

2.2 Estimating the eﬀect of shareholding

One class of assets that might sway an individual’s propensity for constitutional reform are shares in overseas joint stock companies. Joint stock companies were introduced into England in the late sixteenth century to take advantage of new opportunities in the New World and Asia. Prior to the joint stock company, traders were organized in “regulatory companies” that were similar to medieval merchant guilds. Merchants in regulatory companies gained the freedom, often after long apprenticeships, to engage in a particular trade on their own account or in small partnerships. In contrast, agents of a joint-stock company traded on behalf of that firm, which had a unified management, and ownership– and thus risk– distributed among its often numerous shareholders.11 Thus, for the first time, joint stock companies enabled sedentary agents to take advantage of potentially

10In this simple specification, s is assumed to be the same for each agent. In a more general specification, we could imagine that an agent’s effect on the outcome of political struggle is a function of his endowments, and make s a function of Xij. As long as the effect of support does not depend on whether the agent supports the royalists or parliament and is distributed independently conditional on Xij, this generalisation would change the structural interpretation of the coefficients γj, but not effect the sign or irrelevance conditions described above, and we can still make inference on the marginal effects. 11A near-contemporary account describes the motivations behind the founding of the mysterie and companie of the Merchants adventurers for the discoverie of regions, dominions, islands and places unknown in 1552: And whereas many things seemed necessary to bee regarded in this so hard and difficult a matter, they first made choyse of certaine grave and wise persons in maner of a Senate or companie, which should lay their heads together, and give their judgements and provide things requisite and profitable for all occasions: by this companie it was thought expedient that a certaine summe of money should publiquely bee collected to serve for the furnishing of so many shippes. And lest any private man should bee too much oppressed or charged, a course was taken, that every man willing to bee of the societie, should disburse the portion of twentie and five pounds a piece: so that in a short time by this means the sume of six thousand pounds being gathered, the three shippes were bought (Hakluyt, 1589)[pp.267].

10 highly lucrative overseas opportunities, without themselves specializing in navigation and commerce. Shares therefore may have aligned the interests of a broad coalition in favor of political reforms aimed at enhancing the value of overseas investments. The empirical section will provide estimates of Equation (3) and, in particular, measure the effect of prior shareholding in overseas companies on support for political reform. If, as is suggested by the historical record, investments overseas were highly subject to executive discretion (see Historical Appendix), and these investments were also likely to be encouraged under the new regime, then a relatively larger effect of shareholding on the decision to support increased parliamentary control should be expected. Equation (3) will be estimated under two sets of assumptions. First, if the selection of investors into investment in shares was uncorrelated with subsequent political decisionmaking, estimates can be made of the average effect of shareholding on support for political reform among shareholders. In fact, this condition would be satisfied in the benchmark canonical model of portfolio choice (Markowitz, 1959, Sharpe, 1964): assuming that investors are aware of all assets, there are no transaction costs, and there are no uninsurable risks (such as accumulated human capital), all agents seeking the mean-variance efficient portfolio should choose to hold the market portfolio. The market portfolio would include shares in joint stock companies, as soon as they are introduced. Differences in risk preference will affect the allocation of assets between risky and non- risky assets, but not the particular set of risky assets, such as shares in joint stock companies (Guiso and Jappelli, 2005). Thus in a frictionless environment, there should be no systematic selection bias among those who choose to invest in shares, and OLS will provide unbiased estimates of the average effect of the introduction of shares on support for parliamentary reform among investors.12 Relaxing the strong assumptions underlying the canonical model allows us to delin- eate the relevant channels through which selection biases may occur and motivates the identification strategy. First, it is plausible, particularly in a nascent stockmarket environment, that even fully rational agents cannot invest in particular assets that they do not know exist (Merton, 1987). If investors face fixed costs in gaining access to information about particular assets, or more generally, in purchasing them, there will be a correlation between the possession of liquid assets, or wealth, and investment in stocks.13 This may also lead to local geographical concentrations among shareholders (Coval and Moskowitz,

12Furthermore, if investors’ utility functions obey constant absolute risk aversion, investors will invest the same amount in stock. Even with constant relative risk aversion, asset shares will be independent of wealth- the poor and rich will hold the same proportion of each asset, including stock, but in diﬀering amounts (Guiso and Jappelli, 2005). 13Indeed, such a correlation appears empirically conﬁrmed among modern US and European populations (Guiso, Haliassos and Jappelli, 2003).

11 1999, Zhu, 2002), motivating the inclusion of a set of controls for the geographical constituency of the MP, including distance to London, already the major financial center, and fixed effects for each county. While shares were often divisible during this period, so that a number of individuals could get together to buy a single share, liquidity constraints and fixed costs in acquiring information may also be an issue. However, the rich historical and biographical records available for each MP allow the development of a set of controls for wealth endowments. These include whether an agent was the eldest son (and thus the heir, particularly important during this period during which primogeniture was customary), whether the agent inherited land or manors and whether the agent’s father had an aristocratic title, a baronetcy or a knighthood. Differences in human capital endowments can present an important source of non- insurable risk. Prior to the introduction of shares in England, individuals seeking to invest overseas had to join a regulatory company and trade on their own account or in small partnerships. Those who possessed such human capital were naturally already exposed to foreign opportunities, and risks from changes in Crown foreign policy. In contrast, the introduction of shares allowed sedentary non-merchants to gain such exposure for the first time. Thus to test whether shareholding aligned the incentives of disparate groups, a comparison can be made of the effect for merchants and non-merchants. If shareholding aligned incentives for non-merchants to support political reforms in order to pursue overseas opportunities, then a greater effect of shareholding on reform for non-traders should be expected than for those with mercantile backgrounds, who could already invest overseas in the absence of shares.

3 Data

Surprisingly, despite the importance of this episode, this is, to the author’s best knowledge, the ﬁrst paper to systematically gather, digitize and analyze the endowments and political choices of the individual MPs who initiated England’s constitutional reforms. As summarized in Table 1, these data were collected from a number of diﬀerent primary and secondary sources. First, biographies of each member of the Long Parlia- ment, drawing in particular from compilations by Keeler (1954), Brunton and Pennington (1954), the History of Parliament Trust (forthcoming) and the Dictionary of National Biography, were used to construct a range of endowment measures, including inherited wealth and rank and whether an individual was apprenticed into a merchant company.14

14The House of Commons consisted of representatives of 249 constituencies including the chartered boroughs, the 59 counties and the Universities of Oxford and Cambridge. The franchise was limited to

12 The Data Appendix includes an example of such an entry. Each individual, and their father and father-in-law, were matched to the data on trading interests from Rabb’s 1967 lists of all investors in overseas companies mentioned in the founding charters, patent rolls and subsequent transfer books of the major overseas trading companies founded in England between 1575 and 1630. In total, Rabb provides names of 6,336 investors mentioned during this period.15 These investor lists were further supplemented and extended to 1640, where possible, using biographical information and the charters of the Saybrook and Providence Island companies. A set of controls for other endowments that may have played some role in influencing these political choices were also collected. For example, religion has played an important role in the historiography of the Civil War. A combination of two proxies can be used to capture the effect of Puritanism. First, biographical data identifies individuals who attended Puritan seminaries or colleges that had strong Puritan ties.16 An MP’s education at such institutions may be interpreted as an indicator of Puritan preferences. To capture religious preferences among those who did not attend such institutions, data was gathered on active Puritan ministers and Catholic recusants in the area each MP represented from diocesan records.17 Parliamentary constituencies were further matched to a set of geographical and other historical data that capture the impact of royal and noble influence on particular MPs, differences in the preferences of boroughs and counties, cities, towns and ports, as well as other forms of regional preferences (see Data Appendix for further details).

4 Results

As the descriptive statistics in Table 2 reveal, shareholders constituted 21 percent of members of the Long Parliament, or 116 members. Of shareholders, 76 percent invested in unproﬁtable companies, with 59 percent investing solely in unproﬁtable ventures. Share- holders had similar wealth endowments to non-shareholders, including similar proportions of MPs who were heirs, inherited ties to the royal court, had fathers with titles or had

a relatively few aﬄuent “burgesses” in towns and owners of freehold land worth 20 shillings a year in the counties. Birth dates that were not known were imputed where possible from the individuals’ entry into colleges, inns, completion of apprenticeships and dates of knighthoods. 15Some uncertainties about identity faced by Rabb were resolved. Remaining uncertainties stem mainly from common names–such as Thomas Smith– which should constitute pure measurement error. See Rabb (1967, chp.3) and the Data Appendix for a detailed discussion of each company. 16For example, Emmanuel College, Cambridge, was founded in 1584 in order to render “as many possible ﬁt for the administration of the Divine Word and Sacraments.” (Porter, 1958, pg. 238). 17These were at the ancient diocese level (McGrath, 1967). To get estimates at the county level, a uniform distribution of ministers and recusants per unit of area in a diocese was assumed.

13 inheritances of manors or any landed estate. MPs who had been educated in Puritan colleges or seminaries do not seem to have been more likely to invest. Shareholders do however differ in two salient dimensions. First, as expected, shareholder MPs came of age on average eight years closer to the time of Drake’s voyages, which demonstrated the potential profits and feasibility of direct English trade to the Indies. Second, 23 percent of shareholder MPs came from mercantile backgrounds, while those with mercantile background represented 14 percent of the sample as a whole. As we shall demonstrate however, the effect of shareholding is largest on those without existing mercantile backgrounds. Despite the broad similarities between the endowments of shareholders and non- shareholders, there are large differences in their political decisions. Three in four shareholder MPs supported the expansion of Parliamentary control in the Civil War, compared to around half of non-shareholders. Shareholder MPs also were more likely to support Parliament against the Crown at other points in the lifecycle of the struggle for which roll call evidence survives: they were around half as likely as non-shareholders to vote against the conviction of the Charles’ chief advisor, the Earl of Strafford, in 1640, more likely to offer a loan to defend Parliament in London at the beginning of the war in 1642, and less likely to be purged for favoring compromise with the King, allowing them to sit in the Rump Parliament. The summary statistics also reveal important similarities and significant differences between Royalists and supporters of parliamentary reform. First, virtually all measures of endowed wealth appear similar both among both populations, suggesting that the Civil War was not fought to protect domestic wealth. Secondly, supporters of reform appear to disproportionately come from mercantile and Puritan backgrounds, and less likely to possess inherited court connections. This is consistent with the model, as overseas property, religion and court patronage were all areas under which the Crown had broad discretion under the existing constitution and were most likely to change in value with a change in regime (see Historical Appendix). Figure 2 reveals that while the delegations from certain areas, most notably East Anglia, seemed more inclined to support Parliament, there is also significant within- county and constituency variation in support for reform among MPs, and there are discernable differences in the geographical patterns of investment in shares and in support for reform. This suggests that regional differences, while potentially important, are not a complete explanation of the relationship between shares and support for reform (Figures 1 and 2). Table 3 presents regression results on the propensity to invest in overseas shares

14 among MPs. Note that merchants, even with similar wealth endowments, were more likely to invest. Further, there was an inter-generational persistence in propensity to invest; these factors will be controlled for in all regressions. However, apart from some non-robust evidence for a relative lack of investment among the children of nobles, a range of measures of domestic wealth endowments appear to have little effect on an MP’s propensity to hold shares. Religion too seems to have little effect across specifications. Despite individual wealth endowments having little influence on shareholding, Table 3 suggests that the propensity to hold shares does appear to differ with the constituency of the MP. MPs from more densely populated counties, from constituencies outside the Crown lands and representing ports were more likely to invest, though the distance from London appears to have little consistent effect. These results are consistent with the literature on local biases of investors (Coval and Moskowitz, 1999, Zhu, 2002): that shareholding occurred among groups more familiar with opportunities overseas, not just in London, but across the country. A set of controls, including fixed effect specifications that compare MPs from the same county, will address this variation. Table 4 presents regressions assessing the role of endowments and shares on the probability of support for Parliamentary control of government during the Civil War, se- quentially adding controls for personal and constituency characteristics (columns 1-5), omitting the county of Middlesex, which included London and its environs (column 6) and comparing MPs representing constituencies within the same county (column 7). Share ownership is associated with a twenty percentage point rise in the probability of support for Parliament, an effect that is strongly significant and remarkably unchanged with the addition of controls for wealth, religion and geographical features. Consistent with the theoretical framework, the other individual endowments that show strong effects on support or opposition to political reform are those over which the executive wielded discretion before the Civil War: mercantile interests, religion and inherited ties to court (Columns 1-7). Naturally, the value of inherited ties to the royal court would be deeply affected with a change in regime and the table reveals consistent evidence that those with such ties were more likely to support the Crown. Further, the positive effects of the two proxies for Puritanism are also consistent with the theoretical framework, since following the Reformation, the Crown possessed prerogative control over the newly-formed Anglican Church.18

18Recent accounts revive the view that the “Puritan Revolution” was motivated by religious differences, seeing the Civil War as a skirmish in the broader Wars of Religion (Morrill, 1993). However, it is unlikely that this provides a complete explanation. As the Grand Remonstrance, issued by Parliament on the eve of hostilities, suggests, MPs favored a “profitable” war in the Spanish West Indies instead of direct conflict with Catholic Spain. Religion was also seen as an effective propaganda tool. As the contemporary lawyer and MP, John Selden, wrote:

15 In contrast, a range of measures of endowments of domestic wealth appear to have little effect on support for reform, including the MP’s status as an heir or the inheritance of a manorial estate, that would be a strong determinant of whether an MP was considered a member of the “gentry”. Other wealth endowment measures appear to have no effect, with one exception– that the sons of aristocrats were less likely to rebel. There is also little evidence in these data to support the theory that MPs from families of newly commercialized gentry who acquired land during and after the reign of the Tudors– which spanned the dissolution of monasteries during the Reformation– were more likely to support Parliamentary control as suggested by Tawney (1941),Moore (1966) and Rajan and Zingales (2003). These results instead suggest that support for parliamentary control of government was unaffected by endowments of new or old wealth, and thus that it is unlikely the Civil War was fought primarily to defend domestic property.19 It could be that MPs’ support for Parliamentary control was shaped more by the interests of their constituency than by their individual investments in shares. Indeed, it may be that the effect of shareholding may be capturing the fact that shareholder MPs were more likely to represent ports newly-enriched by Atlantic trade (Acemoglu et al., 2005b). Another possibility is that the effect of shareholding is capturing the preferences of MPs who represented dense populated counties or regions, such as East Anglia, that had different legacies of law and institutions from the medieval period that made them more likely to support Parliamentary control (Fischer, 1989). However, adding a range of controls for constituencies of representation has little effect on the coefficient of shareholding, and there is no evidence that MPs representing towns or ports were more likely to support Parliamentary control (Columns 2-7).20 Though MPs representing constituencies closer to London appear to be more likely to rebel, these MPs were not more likely to invest in shares (Table 3). Even comparing MPs representing constituencies from within each of the 52 historic counties in England and Wales does not appear to diminish the effect of shareholding (Column 7). Despite employing a set of wealth and geographical controls that address the main

the very Arcanum of pretending religion in all wars is that something may be found out in which all men may have interest. In this the groom has as much interest as the lord. Were it for land, one has one thousand acres and the other but one; he would not venture so far as he that has a thousand. But religion is equal to both. Had all men land alike, then all men would say they fought for land (Hill, 1961)[pg. 105].

19It is interesting to note that, even though the lands of Royalists were confiscated during the hostilities, the vast majority were able to regain their lands upon the payment of a “compounding” fine– a percentage of a year’s income. 20Controlling for measures of per capita county wealth gleaned from the Tudor lay subsidies (Sheail, 1998) also has no effect on the results, but contains gaps that significantly reduce the sample size (results not shown).

16 sources of likely selection, there still might be an unobserved driver of shares, such as unobserved wealth, that might impact both share ownership and the propensity to support Parliament. Following Altonji, Elder and Taber (2005), the bottom panel of Table 4 estimates the potential bias due to selection under the assumption that the relationship between shareownership and our set of controls is merely representative of the relationship between shares and unobserved drivers of shareownership. Since the measures of wealth and geographical factors were specifically gathered to reduce the bias on shares, the assumption that they are instead merely representative is likely to be weak, leading to a lower bound estimate of the effect of shares. The estimated bias is significant at the 10% level for the more parsimonious specifications, though becomes insignificant by adding the large set of county fixed effects. Regardless, the implied lower bound on the effect of shares appears to be quite stable: for the preferred specification with the richest control set, including county fixed effects, the lower bound suggests that shareownership results in a 12.5 percentage point increase in the probability of supporting Parliamentary control of government. Further, the shift in shareownership due to unobservables would have to be 2.5 times as great as the shift due to this richest control set to eliminate the effect of shares, suggesting that at least part of this effect is real.21 Figure 3 provides a complementary approach to assessing the sensitivity of our results to selection bias, following Imbens (2003). The figure compares the partial correlations between the observed covariates, shareownership and support for Parliament, relative to the thresholds necessary for an omitted binomial variable to reduce the effect of shareholding to insignificance at conventional levels. Consistent with Table 3 and the robustness of the effect of shares across specifications in Table 4, the coefficient on shareownership is largely insensitive to the inclusion or deletion of the wealth of included observable covariates. The variables that are most correlated with shareholding are geographic constituency characteristics, such as population density and representation of a port constituency which, however, show weak correlations with the propensity to support Parliament. In fact, as the figure reveals, no observable covariate is sufficiently correlated with either shareownership or support for Parliament to eliminate the effect of shares at the conventional 5% significance level. Even at the more stringent 1% level, only the positive correlation between shareholding and inherited court connections (which would conversely imply stronger support for the Crown) is strong enough to have that effect. Table 5 adds an interaction term between shareholding and mercantile endowments

21These results are also consistent with that from nearest neighbor propensity score matching, with estimated eﬀects [and clustered bootstrapped standard errors] of shares on the propensity to support Parliament among shareholders of 0.129 [0.084] and 0.217 [0.062], matching on personal and additionally on constituency characteristics respectively.

17 to the specifications in Table 4. While the point estimates on shareholding show a slight increase with the addition of this interaction term, there is also a robust, strongly significant and offsetting negative interaction effect between shareholding and existing mercantile interests. These results are consistent with an alignment effect: by providing non-merchants with the opportunity to benefit from overseas trade and expansion, the effect of shares on support for parliamentary control should be greater relative to those with existing mercantile endowments, who already enjoyed such opportunities and a relatively strong propensity to support Parliament. Figure 4 illustrates this effect, plotting the distribution of predicted probabilities by share investment and overseas trading interests. Notice that merchants who had previously invested in overseas shares do not seem to be very different from merchants who had not done so: both were later likely to support Parliament against the Crown. However, while the median non-merchant was actually slightly more likely to support the Crown in the absence of share investment, non-merchant shareholders reveal broad support across the distribution for parliamentary control. Even though there is no significant effect of a range of measures of domestic wealth on support for political reform, it still might be the case that insecure domestic property rights were crucial in the decisions of agents to support political reform and that the effect of shareholding occurs not through the alignment of interests across groups in favor of control over sovereignty rights, but rather due to a desire to protect newly acquired wealth from investments in profitable overseas companies Acemoglu et al. (2005b). However the accumulation of new wealth does not appear to be driving these results. Table 4(Panel B) compares shareholders who only invested in companies that were unprofitable prior to the Civil War to otherwise similar MPs. The effect of shares in unprofitable companies appears to be actually slightly stronger, and once again the effect is mainly on those without pre-existing mercantile interests. Though this result may appear somewhat counter-intuitive, it is in fact consistent with the theoretical framework: those with profitable overseas investments under the existing regime have something to lose (a −βj|R) with constitutional change relative to those with non-performing investments under the existing regime. We can use the lower bound estimates of the effect of shares using the full set of controls with county fixed effects in Table 4 and in Table 5 to perform a counterfactual exercise. As discussed above, MPs were faced with the stark choice of supporting the Crown or Parliament during the Civil War. Since this choice is symmetric– not making a choice is not an option– those pushed over the 50% probability threshold of supporting Parliament due to holding shares can be thought of as likely switchers of allegiance from

18 the Crown. If we subtract the lower bound estimate of the effect of shares of a 12.5% increase from the predicted probability that shareholders support Parliament, the median MP had a probability of support for Parliament of 43.6%, and thus actually was more likely to support the Crown rather than Parliament. In fact, a majority of 58.6% of MPs would have been likely to support Crown authority. With the addition of the lower bound effect of shares, the median MP had a 56.7% probability of supporting Parliamentary authority, with an implied associated majority of 59.0% in favor of constitutional reform. While such a majority would not of course have been a necessary condition for Parliamentary reforms to take place, and these counterfactual exercises are at best suggestive, the extent of these swings even assuming the lower bound effect of shares does indicate that shares played an important, and likely pivotal, role in broadening the coalition for reform. While the effect of shares does appear to generate a shift in the support of the median MP towards Parliament, we can also extend our counterfactual exercise to examine the extensive margin: those individual MPs that were likely to have been pushed over the threshold towards support for constitutional reform due to shares. The number of those pushed over the threshold ranges from 20 using the lower bound estimate to 29 (or 5.4% of all MPs) using the conventional estimates derived from Table 4. Table 6 further unpacks these figures by using the lower bound and conventional estimates from Table 4 with the full set of controls and county fixed effects to isolate the compliant switchers: shareholder MPs who ultimately supported Parliament but, all else equal, were actually more likely to have supported the Crown in the absence of shares. As the table suggests, the majority of these MPs appear to have otherwise lacked mercantile interests. While a few were involved in the profitable new trading companies, like the East India company and the Levant company, a large majority were invested in unprofitable colonization projects, with shares bringing acres of land in the New World, but whose charters and investments were subject to Crown expropriation (see Historical Appendix). Close to half were also involved in companies, particularly the Providence Island company, that made violent privateering incursions into profitable Spanish trade routes in the West Indies.22 A remarkable feature of Table 6 is that despite looking at members who otherwise would have supported the Crown in the absence of shares, the table contains three of the famous “Five Members” that were identified by the King to be ringleaders of Par-

22In fact, in 1643, the newssheet Mercuricus Civicus claimed that the struggle “was conceived (some say) near Banbury and shaped in Gray’s Inn Lane, where the undertakers for the Island of Providence did meet and plot it” (Kupperman, 1993)[pg.6]. Though an extreme interpretation, it is still indicative of contemporary perceptions.

19 liament’s legislative challenges to his prerogative rights. The King’s illegal armed entry into Parliament to arrest the Five for treason in January 1642, which later led to the mobilization of the London militia in defense of Parliament, is often seen as a major step towards the outbreak of violent conflict (Hexter, 1941, see also Historical Appendix). Of the three MPs– Denzil Holles, John Pym and William Strode– John Pym in particular is seen by historians as a major figure in initially organizing the coalition in favor of Parliamentary control (Hill, 1961, Hexter, 1941).23 Yet even the lower bound estimates suggest these leaders of the opposition might have actually favored support for the Crown in the absence of shares. Instead of being radical extremists, it may be that the leaders of the push for constitutional reform were instead effective at creating a coalition precisely because their interests were aligned with both sedentary landowners and overseas traders. Shares may have played an important role in this. Table 7 presents results using the same sets of controls to estimate the effect of shares on other indicators of support for Parliamentary control of government over the lifecycle of the struggle. The first panel provides evidence on the MP’s decision of the sole legislative vote from the period on which individual roll calls survive: that of whether an MP voted for or against the “attainder” of the King’s chief advisor, the Earl of Strafford, in 1641.24 The second panel indicates support on the eve of hostilities, where evidence exists of who subscribed to a loan for the defense of the London Parliament in 1642.25 The third panel assesses whether the MP sat in the “Rump” Parliament of Civil War victors that implemented many of England’s changes in foreign policy from 1648-53 (please see Figure 8 and the Historical Appendix). The broad picture that emerges is that at least as early as the peaceful legislative push for reform in 1640-42, shareholders were consistently more likely than non-shareholders to oppose executive authority. This continued across the lifecycle of the struggle. Further, where merchants exhibit significant differences to non-merchants, shares appear to align

23The remaining two of the Five were John Hampden and Arthur Haselrig, both proprietors of the Saybrook Company, and apparently strongly connected to the Providence Island venture. Shareholding does not however appear to be pivotal in shifting their support in these estimates. Another notable name in the list is Oliver St John, John Hampden’s lawyer and Oliver Cromwell’s brother-in-law, who played an key role in the Council of State under the Rump Parliament (See also Hillmann, 2008). 24Unlike later in the seventeenth century, records of MP’s votes in Parliament were not systematically kept, with one exception. In April 1641, Parliament voted to convict the King’s chief advisor and architect of many of his policies, Thomas Wentworth, the Earl of Strafford. Those who voted against the conviction were seen as supporters of royal authority and their names were anonymously posted in Westminster Yard. 25In June, 1642, claiming that the “King (seduced by wicked Counsel) intends to make War against his Parliament”, the Long Parliament passed an ordinance soliciting a loan to “uphold the Power and Privileges of Parliament”. Lists survive of the subscribers who were promised the return of their funds with 8% interest, and that “no Man’s affection [to Parliament and its privileges] shall be measured by the Proportion of his offer . . . ”, but rather by the act of participation (Firth and Rait, eds, 1911)[pg.6-9].

20 non-merchant choices with merchants. The fact that both non-merchant shareholders and merchants were signiﬁcantly more likely to sit in the Rump Parliament suggests a direct link between the genesis of the coalition for reform and the subsequent dramatic changes in England’s public investments in the Navy and its foreign policy that occurred upon its success.

4.1 Drake’s voyage as an exogenous shock to investment

Next we consider the possibility that the OLS estimates, rather than being upper bounds, are actually underestimates. This is due to the presence of a classic hold up problem– with the Crown able (and as the Historical Appendix reveals, willing) to expropriate the returns from overseas investments, some individuals may have been motivated by the new access to overseas opportunities provided by shares to support reform, who also faced incentives to wait until after the reform to actually invest. This would lead to a downward bias in the measured effect. To assess whether this is the case, we can exploit an arguably exogenous shock to information about the potential benefits from overseas investments that resulted from Drake’s unlikely and remarkable feat of navigation. In September 1580, the Pelican, the sole survivor of a fleet of five ships that had sailed three years earlier, moored at Plymouth harbor in England.26 Her captain, Francis Drake, had achieved an unlikely success– the circumnavigation of the world, direct trade with the Spice Islands, and a raid on treasure ships in the Pacific that caught Spain entirely unprepared. Drake’s voyage and the charts of ports, watering places and trade routes that he constructed meant that for the first time English traders could break into Portuguese and Spanish monopolies in Eastern trades (Andrews, 1967). Not only did Drake’s voyage change the feasibility of English trade, it also amply demonstrated the scale of profits to be had from trade and plunder. The details of Drake’s voyage, being technically illegal, was kept a closely-guarded secret. This changed in 1585, when Drake successfully raided Spanish ports in the Atlantic as well, demonstrating their vulnerability to English attacks. Knighted by Elizabeth I, Drake’s extraordinary achievements made him a national hero.27 Whereas previous English attempts for form joint stock ventures for overseas exploration, as late as Gilbert in 1583, had struggled to find outside financing (Williamson, 1946), suddenly a cohort of Englishmen and women were inspired by Drake’s voyages to

26The Pelican later gained further fame under her new name: the Golden Hind. 27The Spanish Ambassador, Bernardino de Mendoza, cautioned Philip II to destroy all English and French ships that might enter the Paciﬁc as: “. . . at present there is hardly an Englishman who is not talking of undertaking the voyage, so encouraged are they by Drake’s return . . . everybody wants a share in the [next] expedition. . . (Rabb, 1967)[pg.20]”

21 invest in endeavors overseas (Andrews, 1967).28 The eﬀects were not conﬁned to London. As a clergyman in the West Country town of Exeter, Thomas Hooker, wrote in 1585, Drake’s exploits:

inﬂamed the whole country with a desire to adventure unto the seas, in hope of like success, [so] that a great number prepared ships, mariners and soldiers and travelled every place where any proﬁt might be had . . . (Andrews, 1964)[pg.4].

Research on the “stockholding puzzle” emphasizes a number of factors, beyond wealth, that foster increased stockholding in contemporary stockmarkets. A common theme is that individuals buy stock when they are paying attention. High profile events, or stocks with extreme one day returns, attract new investors, both by advertising the stock and by generating social feedback: individuals invest more if they know those around them are investing as well (Hong, Kubik and Stein, 2004, Grullon, Kanatas and Weston, 2004, Guiso and Jappelli, 2005, Barber and Odean, 2006). Inexperienced investors are particularly influenced by bursts of high returns (Greenwood and Nagel, 2006). All of these features appear to parallel the nationwide enthusiasm that swept England in 1585. There was a perception of extremely high returns to Drake’s voyages which were ultimately high profile events that hit everyone at the same time (Andrews, 1967). This may have created a “social effect”. Those coming of age in 1585, in particular, were both inexperienced and free for the first time to deploy resources, thus more likely to be paying attention to these new opportunities.29 As Figure 5 reveals, Drake’s voyages did indeed act as a shock to the propensity to invest in joint stock by the cohort of Long Parliament MPs who became adults at this time. The effect diminishes almost monotonically among subsequent generations, however, reflecting the failure of these ventures to make profits, due in large part to Crown policy and predation (see Historical Appendix). Furthermore, as Figure 6 reveals, other endowments, with one possible exception, show insignificant trends and do not reveal significant jumps before and after 1585.30

28The women included Elizabeth I herself, who took her substantial proceeds from Drake’s voyage and invested them in joint stock ventures rather than buying back Crown lands that she had been selling to service debt. 29The age of majority– the age at which heirs were able to write legally enforceable contracts and were entitled to dispose of their inheritances– has evolved in England over time. By custom, the general age of majority for socage tenure, the most common form of tenure prior to 1660 (and the only form thereafter), was ﬁfteen (James, 1960) 30That exception: that MPs who came of age after 1585 appear less likely to inherit land, would, if anything be expected to result in a reduced incentive to support reform, if the property rights/ wealth story is correct. Regardless, controls for these and other wealth endowments will be included in all regression tables.

22 Table 8 augments the regressions on the propensity to invest in shares to include polynomial age controls, revealing a consistent zero direct effect of age on the propensity to hold shares across specifications (Columns 1-5). However, consistent with Figure 5, there is a jump in the propensity to invest among those that came of age after 1585, that declines among those who came of age in the years thereafter (Columns 6-9). MPs that came of age just after 1585 were around 30-45 percentage points more likely to invest in shares compared to those who came of age just before. We can use this jump in investment to compare support for political reform among those investors who invested because they came of age just after Drake’s voyages and otherwise would not have invested. This regression discontinuity is “fuzzy” as not everyone who came of age after this time deterministically invested, suggesting that two-stage least squares estimates are appropriate. As the bottom panel of Table 8 reveals, the instruments are robust and strongly jointly significant across a range of specifications.31 Table 9 presents regressions of our four different indicators of support for reform on shares, both including quadratic controls for age (Cols 1-4) and implementing the regression discontinuity design (Cols 5-9). First note that the quadratic controls for age do not reveal an independent effect on any of our measures of support for reform. The effect of shares is also robust to the inclusion of age controls. This suggests that age effects are unlikely to be inducing the robust relationship we see between shareholding and support for political reform. Furthermore as the regression discontinuity results reveal, relative to those who came of age just before Drake’s return, those who invested because they came of age just after Drake’s voyage were, once again, more likely to support Parliamentary reform during the Civil War and reveal consistent patterns across the early lifecycle of the struggle. As the regression discontinuity depends on the relatively small number of MPs who came of age prior to 1585 to identify a separate trend, some specifications are not precisely estimated and the effects should be interpreted with caution. It is nevertheless reassuring to find consistent point estimates across specifications, many of which are significant at conventional levels. It could be that those that came of age in 1585, instead of being motivated by non- performing assets that they acquired overseas through shareholding were instead were inspired by Drake to be ‘dreamers’ with strong tastes for risk that they managed to

31We can also compare investment among those MPs from the cohorts most likely to be different: those that came of age the same regime: comparing among Elizabethans, Stuarts, and those who came of age under the “11 Years Tyranny” from 1629-1630 when no Parliament sat in England. The latter cohort effect also allows us to assess whether increased shareholding came about due to attempts by companies to “bribe” sitting MPs. However, the cohort effects are insignificant and do not substantively effect the results (results not shown).

23 maintain despite years of subsequent disappointment in the performance of their investments.32 However once again, the interaction effects suggest that the effects were mainly on non-merchants, consistent with the role of shares in aligning interests, rather than being reflective of a cohort-wide taste for risk.

4.2 Alternative channels

It is possible to implement further tests of the two key alternative channels– domestic wealth and ideology– that have been mooted as explanations for the successful development of a coalition in favor of political reform in England. It could be, for example, that our measures of endowed domestic wealth are simply not precise, and lack of precision, rather than strength of property rights is responsible for a lack of an effect on support for reform. Table 10 (Panel A) tests whether these endowment measures have any effect on that subsample for which measures exist of an MP’s income at the time of the Civil War. Though it appears that there are differences in contemporaneous income between shareholders and non-shareholders, these are mainly due to merchant shareholders, and the effect for non-merchants cease to be significant in the preferred specification with full controls. There are also no discernable differences among shareholders who invested because they came of age just after Drake’s voyages. Further, Table 10 (Panel A) shows that inheritances of landed estates and titles are all strong determinants of contemporaneous income. This suggests that our measures of endowed wealth do capture important variation in contemporary income and it is not imprecision that is leading to a lack of a domestic wealth effect. A second possibility is that the effect of shareholding is capturing unobserved pre- existing differences in political allegiances that led individuals to oppose the Court and happened to also invest in shares. This “ideology” story would suggest that in the years before the Civil War, anti-monarch shareholders would be also less likely to attend court or work for the Crown. In contrast, if the theoretical framework above is correct, then prior to the stark decision to support or oppose constitutional reform in the Long Parlia- ment, individuals with endowments most subject to executive control, including shareholders, would paradoxically face a greater incentive to secure their property through investing in client relations with the Royal court. Table 10 (Panel B) examines the effect of shareholding on the acquisition of court positions. Here the potential for endogeneity is significant: some courtiers may have invested in joint stock since they had more access to royal charters and royal patronage,

32Such a possibility would run counter to work by Malmendier and Nagel (2007), who ﬁnd that investors weight more recent experiences more heavily.

24 or alternatively, the ruler may have attempted to buy off the most recalcitrant members of his opposition. However, as the RD specifications suggest, shareholders who invested overseas due to Drake’s voyages were more likely to acquire court positions, with a greater effect for non-merchants than for those with existing trading interests. Thus, it appears that shareholders did attempt to work within the existing constitutional system to secure overseas property: prior to the Long Parliament, shareholders were not consistently opposed to the Court, and many assumed court roles.33 In contrast, there is no evidence that those endowed with greater domestic property cultivated more ties to court, suggesting once again that domestic property rights were relatively secure.

5 Discussion

With English political institutions strongly influencing the institutional design of many nations around the world, the question of how a novel, broad, and ultimately successful coalition in favor of parliamentary supremacy in government emerged in England has proved to be among the most enduring in political economy and institutional economics. This paper provides the first direct evidence on the determinants of individuals’ decisions to join the coalition for revolutionary reform. The paper documents the contrast between the relative lack of importance of domestic wealth and the significance of overseas interests in support for reform. By highlighting the empirical and constitutional distinction between the relative security of domestic property rights and the executive’s discretion over rights overseas, the paper reconciles theories that suggest that England’s reforms were driven by those seeking to protect property (e.g. Acemoglu et al., 2005b, Rajan and Zingales, 2003, North and Weingast, 1989) with a lack of evidence of change in existing property rights in the seventeenth century (e.g. Murrell, 2009, Sussman and Yafeh, 2002, Clark, 1996). The paper also sheds new light on how a constituency in favor of parliamentary supremacy was created. Through the innovative use of shares in joint stock companies, access to new overseas opportunities expanded beyond traders to encompass other groups with initially disparate interests. In fact, shares appear to have played an important, likely pivotal, role in generating a parliamentary majority in favor of reform, as well as

33Indeed, the historical record suggests that a high degree of collegiality between monarch and parliament could exist when the monarch ceded prerogative control over foreign policy to Parliament, as occurred in 1624 (Smith, 1999). Charles, in concert with parliament, explicitly tied parliamentary grants of taxation to the declaration of war with Spain. In this session of Parliament, 35 statutes were passed as opposed to only one in 1621. This anomaly is often cited by Revisionist historians as evidence against broad socio-economic explanations for England’s revolution (e.g. Morrill, 1993). However, it is also completely consistent with our framework.

25 shifting the allegiances of key members that provided leadership to the parliamentary coalition.34 Further, the role of shares in aligning incentives may have gained in importance following the war. The development of active stockmarkets that occurred between the Civil War and Glorious Revolution (Neal and Quinn, 2001, Carlos, Key and Dupree, 1998, Carruthers, 1999) made it possible for both winners and losers of the Civil War to reallocate their investments towards companies benefiting from England’s new regime and assertive overseas policies, encouraging a consensus that favored the institutionalized control of national policies by parliament rather than the Crown (Carlos and Jha, in progress).35 The new “political parties” that emerged between the Civil War and Glorious Revolution transcended old landowner-merchant distinctions, instead consolidating their memberships within joint stock companies and bargaining over the burdens of state finance (Carruthers, 1999, Stasavage, 2003). It is likely that the emergence of secondary financial markets weakened the link between different endowments and opposing interests, allowing a consolidation of England’s revolutionary reforms. Thus, rather than becoming an limited oligarchy of Civil War victors, England could continue its path towards broadly representative government and, ultimately, democracy. Instead of yielding another example of wealth shocks generating a new political coalition, England’s experience suggests two tantalizing alternative lessons. First, broad support for institutional reforms was fostered by combining new opportunities for future wealth with a financial mechanism to share that wealth. By allowing a broad range of individuals to share in new opportunities overseas, financial innovation appears to have aligned initially disparate interests, fostering majority support for reform. Second, there appeared to have been a “political multiplier” effect to the development of financial markets. The development of markets to trade endowments that earlier shaped political interests appears to have made politics less conflictual, consolidating reform. Financial innovations appear to have been central in aligning disparate interests that led to one of the first, and among the most enduring, traditions of representative government in the world. 34Joint-stock companies may have been key in organizing the coalition. Joint-stock companies were governed by directors selected by a system of voting in proportion to the value of each investors’ shares (Scott, 1912). Large joint stock ventures, such as the Providence Island company, may have facilitated the selection of opposition leaders. Similarly, the leaders of dissent in the early Stuart Par- liaments include the Earl of Southampton and Sir Edwin Sandys, who were first brought together while serving as directors of the Virginia and Bermuda companies (Rabb, 1998, Hill, 1961). 35The post-Civil War history of the major joint stock companies–the East India company, the Royal Africa company, the Bank of England and the South Sea company–is a history of companies issuing new stock to accommodate new MPs in Parliament. See Scott (1912).

26 6 Historical Appendix

This section gives the contextual basis necessary to motivate the interpretation used in the paper. As discussed above, joint-stock companies were introduced into England in the late 16th century to take advantage of new opportunities for trade with the New World and Asia. This section tabulates why joint-stock ventures failed prior to the Civil War, emphasizing the role in these failures of the monarch’s constitutional control over overseas investments. The section highlights the monarch’s relative lack of discretion over domestic property and contrasts this with evidence that the executive was exploiting these overseas rights to increasingly live without parliamentary oversight. The section describes attempts by a broad coalition to wrest control of the monarch’s overseas rights, first through parliamentary means and ultimately through Civil War. Finally the section describes the constitutional changes that occurred, and documents the massive increase in overseas public investments in the Royal Navy that supported England’s subsequent expansion and growth. Though the potential profits to be had overseas were large, English investors overseas faced two key problems that meant that virtually none of the joint-stock companies made profits prior to the English Civil War. Table 11 summarizes the position of the major English joint-stock companies on the eve of the Long Parliament (1640). As the table suggests, the first problem was that the Spanish and Portuguese, and later the Dutch, had organized national defences of their lucrative monopolies over Atlantic and East Asian trade and violently resisted English entry. English merchants were forced to trade with the New World, indirectly, through Spain and Portugal. Without a more aggressive foreign policy providing support for those seeking to break Spanish and Portuguese monopolies overseas, London was unlikely to emerge as anything more than a regional center of trade and commerce. The second problem was that the right to declare war and, in fact all rights to overseas commerce and territory were owned by the King. Unlike domestic property rights, which were governed by the common law, and were relatively secure from Crown expropriation, foreign trade in early modern England was governed by civil law, administered by the Crown in the Admiralty courts.36 Prior to the Civil War, English rulers also had the right to revise customs rates and “impose” customs upon newly-introduced goods, as commerce was believed to be protected and maintained by the king’s foreign policy (Gras, 1912).37

36Domestic property rights were governed by the common law, a set of precedents that put limits on the Crown’s ability to inﬂuence domestic judicial decisions (Burgess, 1992). Along with access to a jury of their peers, wealth holders were themselves responsible for local enforcement and for collecting taxes. They often refused to cooperate with the Crown bureaucracy. Compared to other contemporary states, the monarch’s ability to expropriate wealth through courts or via taxation in early Stuart England was remarkably light. There was ﬂagrant tax-evasion, particularly in response to policies wealth-holders did not support. As Sir Walter Raleigh admitted in 1601: “our estates that be £30 or £40 in the Queen’s books are not the hundredth part of our wealth.” (Smith, 1999). 37As the judge, Sir Henry Yelverton is noted as stating in parliament in 1610 (precedents from previous parliaments that he referred to are in parentheses): Impositions may be layd upon merchant strangers (13 E 4). But the merchants of England trade not by the comon lawe of the land, but by the lawe of nations . . . We are where the common lawe cannot judge. The merchant hathe no remedy agaynst hym

27 Though limited by both legal institutions and parliament in his ability to expropriate wealth within the country, the king’s prerogative was firmly established beyond England’s coasts.38 Sovereignty rights over war, colonies and customs enabled the Stuart kings, James I and Charles I, to extract much of the residual gains from England’s emerging commerce overseas. As Table 11 suggests, those joint stock companies that were not targeted by foreign attacks and enjoyed temporary profits soon lost these profit sources through the assertion of the Crowns’ sovereignty rights–either through rising customs charges or the revocation of their charters.39 Shareholders in these companies were left with claims on hundreds of acres of undeveloped territory overseas with the potential to be valuable in the future.40 In contrast, as Figure 7 reveals, overseas customs revenue accruing to the Crown rose from insignificance to providing the majority of Crown revenues on the eve of the Civil War. In fact, the combination of a peaceful foreign policy and the accompanying rise in customs revenue from England’s indirect trade with the New World through Spanish and Portuguese ports were allowing the king increasingly to live independent of Parlia- ment. Both James I (1603-25) and Charles I (1625-1648) succeeded in living without Parliamentary subsidy for 11-year periods (Smith, 1999). Customs revenue meant that England’s kings, though still relatively poor and weak in their ability to extract resources domestically, were becoming increasingly enriched over time.41

that spoyles at sea. He is not under the protection of the lawe, thoe under the protection of the King . . . He is under the jurisdiction of the King by the lawe of nations, (6 R 2), Protection. (Gardiner, ed, 1862)[pg.87]

38Overseas rights also included control over innovation. At this time, most technical progress was occurring outside England, particularly in the more advanced centers of France, the United Provinces and Venice (Rapp, 1975). Thus introduction of “new industries”, such as cannon-founding, sugar- refining, soap and papermaking, into England was considered a component of foreign policy, and the king had the prerogative right to assign monopoly patents and thus derive revenues in lieu of what he might have received in customs. An important example of the king’s prerogative role in industry is Alderman Cockayne’s Project- an attempt by James I in 1614 to dye and dress English cloth domestically. At this time, such “finishing” was being done in Holland. The Cockayne project collapsed when the Dutch responded by embargoing English traders- another instance where foreign policy was critical for economic development in this period. 39Not surprisingly, then, the probate records of individuals’ estates reveal no discernable evidence that the wealth of the business community in England rose in any part of the distribution until the trade boom of the 1660s, after the Civil War (Grassby, 1970). 40A typical offer showing the overseas assets that accompanied shares comes from a solicitation for the Newfoundland company (Whitbourne, 1622): And whosoeuer will aduenture £200, may not only haue a whole Harbour, Bay, or Road to himselfe in fee for euer, and foure thousand acres of land, and woods thereunto adioyning on the North side of Trinity Bay; but also a conuenient place, and ground to build Stages and houses, and for drying of fish fit for a Ships fishing voyage of 160 . . . And whosoeuer shall aduenture lesse then £100 be it £80, £60, £50, £30 or £20 shall haue lesse proportionably . . .

41That the monarch was the major beneﬁciary of expanded overseas trade, and no measurable rise in mercantile wealth occurred until after the Civil War suggests that the process of enrichment of Atlantic traders suggested by Acemoglu et al. (2005b), while potentially playing a role in creating a constituency

28 Attempts to bargain over the control over rights over customs and foreign policy played a pivotal role in Parliamentary debates from 1603 to 1625 (Ashton, 1967, Rabb, 1998). These culminated in the “Great Contract” of 1611, an attempt to exchange the king’s authority over independent means of revenue in exchange for assured parliamentary grants of direct taxes. Bargaining foundered, however, chiefly because parliament wanted control over customs, and the king was counselled not to surrender “the fairest flowers for profit and command in all his garland (Hill, 1961).” Though his resources were growing, a simultaneous shock due to invasion by the Scots and rebellion by the Irish was sufficient to overwhelm the king’s finances and, in a position of weakness, he was forced to summon the Long Parliament of 1640. Members of the Long Parliament, in different subsets, sat as England’s representatives for the next 20 years. During this time, England underwent a dramatic institutional transformation. The institutional changes of the Long Parliament began before the Civil War, through the passage of Acts of Parliament that bore the signature of the King. These included passage of the Triennal Act in 1641, guaranteeing that parliament must be called at least every three years and could not be dismissed without its own consent. The royal prerogative courts were also abolished and Parliament passed bills that deposed the king’s chief councillors, such as the Earl of Strafford and Archbishop Laud, and sought control over future appointments. Further attempts to institute parliamentary authority over remaining Crown rights, including over foreign policy, finance and the armed forces led the king to illegally enter the House of Commons to arrest “Five Members” considered the ringleaders of the Parliamentary opposition in January 1942. Parliament summoned the London Trained Bands in its defense, leading the King to abandon the city and later that year to raise his war banner in defense of his prerogative.42 In combination with other financial reforms, Parliamentary control of the customs revenue flowing through London enabled it to out-spend the king during the Civil War and played a crucial role in bringing about a parliamentary victory by 1648 (Rodger, 2004). As Figure 8(a) suggests, during and after this period, England began a remarkable naval buildup. As Figure 8(b) shows, at the end of the Civil War, the Royal Navy was unremarkable relative to its rivals. However, in 1648, England’s new rulers–the “Rump” Parliament of victors of the Civil War–immediately embarked upon a series of foreign policy initiatives designed to wrest control of overseas commerce. These initiatives, including the Navigation Acts and wars with the Dutch and Spanish, paved the way for a boom in English overseas trade that began in the 1660s (Davis, 1973). By the end of the seventeenth century, England’s position as the pre-eminent naval power of Europe, which it would maintain for the next two centuries, was assured. for reform during the Glorious Revolution, is unlikely to have played that role in the Civil War. 42As Charles I himself later provides as his reason for going to war: For although I can be content to eclipse my own beams to satisfy their fears . . . yet I will never consent to put out the sun of sovereignty to all posterity and succeeding kings, whose just recovery of their rights from unjust usurpations and extortions shall never be prejudiced or obstructed by any act of mine . . . (Charles I and Gauden, 1649)[pp.48-49] This description appears to match the condition for the failure of the Political Coase Theorem outlined by Fearon (1996).

29 The constitution of England also began a period of experimentation that led to increased parliamentary control. Rule by the Rump Parliament was followed by a constitutional protectorate under Cromwell, that fell into crisis amid competing military factions at his death.43 Ultimately the Long Parliament was reconvened, inviting Charles II back to England in 1660 as long as all laws passed by the Long Parliament prior to the Civil War (i.e. that bore Charles I’s signature) would remain in force. Since these implied that Parliament controlled the state finances, Crown policies disagreeable to Parliament, such as the Treaty of Dover with France, had to be conducted in secret and lacked credibility. Thus foreign policy remained consistent following the Restoration, with re-affirmation of the Navigation Acts and continued commercial wars with the Dutch. Intermittent attempts by the monarch to reassert independent authority continued until Parliament’s increased executive control was clarified and made explicit following the Glorious Rev- olution of 1688. By 1714, George I of Hanover, 57th in line to the throne and lacking a working knowledge of English, was anointed more by the Settlement Act, enacted by Parliament in 1701, than by God. By 1714, Britain had already begun to assume the role of market hub for a trading empire spanning the world. Amsterdam, the hub of Atlantic trade through much of the late seventeenth century, was already being reduced to re-exporting goods entering British ports (Ormrod, 2003, O’Brien, 1988). While the empires of Spain and Portugal entered long-term decline, and France was heading toward increased executive control, Britain was on a trajectory toward a commercial and colonial empire abroad and the makings of liberal democracy at home.

7 Data Appendix

An advantage of examining the Long Parliament is that its members were deemed his- torically important enough that over the centuries, a series of pushes have been made to gather the original sources that document the lives of individual members. Thus, the lives of members of the Long Parliament have been remarkably well-documented in biographical sources. This study in particular simply uses the classification as Royalist or “Rebel” of Brunton and Pennington (1954), who based their decision to classify a member as a Royalist based upon a series of factors, chiefly attendance at the King’s Oxford Parlia- ment, disablement as a Royalist from the Westminster Parliament and Royalist fines. In the absence of the data compiled here, nor the use of contemporary econometrics, the work of Brunton and Pennington however found no systematic economic or social differences between Royalists and Parliamentarians, a result that may have dissuaded systematic follow up until now. This paper combined the classification used by Brunton and Pennington with the contemporary biographies that were compiled by Mary Frear Keeler (Keeler, 1954). Here is a typical entry from the 548 members of the Long Parliament, Keeler documents:

43In fact, both US and English versions of the Bill Rights were heavily inﬂuenced by England’s ﬁrst written constitution, the Instrument of Government of 1653, that governed the early Protectorate. The Cromwellian Protectorate too continued the expansionist foreign policies of the Rump Parliament, engaging in a “Western Design” against the Spanish West Indies that seized Jamaica, and providing a permanent charter to the East India Company (Howat, 1974).

30 Piers Edgcombe (1609-1667) of Mount Edgcombe, Cornwall, near Plymouth, was a member for Camelford 1640-22 January 1644. He was the head of an ancient and distinguished house. His grandfather, Piers (1536-1607), son of Richard Edgcombe, had been sheriff and M.P. and been interested in Sir Richard Grenville’s projects for exploring the distant seas and in developing the mines of western Britain and Ireland. Sir Richard Edgcombe (d. 1639), father of the MP, was a member of the Council for New England, a JP, a sheriff and DL in Cornwall, MP for Grampound and Bossiney. He was one of the royalists who supported the cause of Buckingham against the attacks of Eliot’s faction in the early years of King Charles. Piers, the eldest son of Sir Richard by his second wife, Mary, daughter of Sir Thomas Coteeles of London, was aged eleven in 1620. He studied at St. John’s, Cambridge, matriculating in 1626, and was probably the “Perseus Esgaimb” who was admitted to study at Leyden University in June 1629. In the preceding year, while still under age, he was successful among five candidates at the parliamentary election at Newport, Cornwall. In 1636 he married Mary, daughter of John Glanville of Broad Hinton, Wilts, receiving with her a portion of £3000, and three years later succeeded to the Edgcombe estate. He had some responsibilities for the defenses of Plymouth in 1639. In 1640 Edgcombe was elected in the Spring and the fall as MP for Camelford. He voted against the attainder of Strafford, and absented himself at the outbreak of hostilities. The House ordered that he be brought up in custody in November 1642, but did not vote for his disablement until 1644. After fighting for the king, he surrendered his command and in 1647 arranged to compound for his estates. He served again in parliament after the Restoration, and died on 6 January 1666/7. Edgcombe’s will, dated 12 May 1666, was proved 14 May 1667. His epitaph describes him as “a pattern to posterity, and an hon- our to an age he lived in; a master of languages and sciences, a lover of the king and church, which he endeavoured to support to the utmost of his power and fortune.” Of Edgcombe’s fortune there are numerous evidences. Carew commented on the excellence of his mansion at Mount Edgcombe. His father’s properties at the time of his death included thirteen manors and other holdings in Devon and Cornwall. Piers owned furthermore some Hampshire properties inherited from his mother’s family, and he bought a Sussex manor in 1641. His compounding papers reveal a yearly income of over £1200, but it may have been nearer £2000 and one observer reported it to be £3000. His fine, set first at a tenth, £2513, was afterwards changed because of an undervaluation and also because of the intercession of Lord Fairfax on his behalf. Edgcombe planned a portion of £5000 for his eldest daughter when he made his will in 1666. Piers Edgcombe was a brother of Richard Edgcombe, a brother-in-law of William Glanville and a relative through his grandmother of Alexander Lut- trell, all fellow parliament men of 1640.” (pg 163-164) For each entry of this kind, I merged this data with that of Brunton and Penning-

31 ton1954, who classified Piers as both a Royalist and a Straffordian (one who both voted and was present during the attainder of Thomas Wentworth, the Earl of Strafford in 1640). Pier’s claim to be of ancient family (and thus bearing a coat of arms prior to the reign of the Tudors- a check of the “emergent gentry” hypothesis of Rajan and Zingales (2003), Moore (1966) and Tawney (1941)) can be checked using Debrett’s and Burke’s Peerage, as well as by checking the entries for his parents and grandparents using the Dictionary of National Biography and the History of Parliament Trust. His other biographical information, including status as heir, attendance at particular colleges (in this case not a predominantly Puritan seminary, as classified by Keeler (1954),McGrath (1967) or Porter (1958)) and the number of inherited manors are directly coded. His father’s entry in the Dictionary of National Biography, the History of Parliament Trust and genealogy.com is used to assess other endowed variables as well as giving Piers an entry of 13 inherited manors. Piers’s income is estimated as a simple average of all available estimates, with aggregate wealth measures put into per year terms. His shareholding and that of his father and father-in-law are matched to the lists of all shareholders in major joint stock companies compiled in the Appendix by Theodore Rabb 1967 and extended and corrected based on these biographies and primary sources.44 We find that Pier’s father was invested in the New England company and the Royal Mines, though he himself lacked such investments. His constituency of Camelford is matched to geographic, noble and Crown demesne data drawing mainly upon the author’s constructed GIS of historic constituencies in England, the Historical Atlas of England by Falkus and Gillingham, eds (1987), estimates of population change by Wrigley (1985), diocesan records on Pu- ritans and Catholics listed in McGrath (1967)) and documentation of Elizabethan ports by Willan (1968). Among companies, the classification of profitability is drawn from company balance sheets in the relevant volumes of Scott (1910).

References Acemoglu, Daron, Simon Johnson, and James Robinson, “Institutions as the fundamental cause of long-run growth,” in Phillippe Aghion and Steven Durlauf, eds., Handbook of economic growth, North-Holland, 2005, chapter 6, pp. 385–472. , , and , “The rise of Europe: Atlantic trade, institutional change and economic growth,” American Economic Review, June 2005. Alesina, Alberto and Eliana La Ferrara, “Ethnic diversity and economic performance,” Journal of Economic Literature, 2005, 43 (3), 762–800. , R. Baqir, and William Easterly, “Public goods and ethnic divisions,” Quarterly Journal of Economics, 1999, 114 (4), 1243–1284. Altonji, Joseph G., Todd E. Elder, and Christopher R. Taber, “Selection on Observed and Unobserved Variables: Assessing the Eﬀectiveness of Catholic Schools,” Journal of Political Economy, February 2005, 113 (1), 151–184.

44An example of these corrections: Rabb has three Framlingham Gawdys, perhaps under the assumption that no one man would invest in joint stock companies over the entire 1575-1630 period that he covers. In fact, Gawdy, a nonageneraion at the time of the Long Parliament, yet one its most active members, was a single man. Use of the detailed biographical information available to us allows such corrections.

32 Andrews, Kenneth R., Elizabethan privateering: English privateering during the Span- ish war, 1585-1605, Cambridge, UK: Cambridge University Press, 1964. , Drake’s Voyages: a reassessment of their place in Elizabethan maritime expansion, London: Weidenfeld and Nicolson, 1967. Ashton, Robert, “The parliamentary agitation for free trade in the opening years of the reign of James I,” Past and Present, December 1967, (38), 40–55. Barber, Brad M. and Terrance Odean, “All that Glitters: The Effect of Attention and News on the Buying Behavior of Individual and Institutional Investors,” Review of Financial Studies, 2006. Benmelech, Efraim and Tobias Moskowitz, “The political economy of financial regulation: evidence from US usury laws in the 19th century,” Journal of Finance, forthcoming. Biais, Bruno and Enrico Perotti, “Machiavellian Privatization,” American Economic Review, 2002, 92 (1), 240–258. Bogart, Dan, “Did the Glorious Revolution contribute to the Transport Revolution: evidence from investment in roads and rivers,” mimeo, UC-Irvine, 2009. Boycko, Maxim, Andrei Shleifer, and Robert Vishny, “Voucher privatization,” Journal of Financial Economics, 1994, 35 (2), 249–266. Brenner, Robert, Merchants and revolution: commercial change, political conflict and London’s overseas traders, 1550-1653, Princeton, NJ: Princeton University Press, 1993. Brewer, John, The sinews of power: war and the English state, 1688-1783, London: Unwin Hyman, 1989. Brunton, Douglas and David H. Pennington, Members of the Long Parliament, London: George Allen and Unwin, 1954. Burgess, Glenn, The politics of the ancient constitution: an introduction to English political thought, 1603-1642, University Park: Pennsylvania State University Press, 1992. Carlos, Ann M. and Saumitra Jha, “Stockmarkets and political development: evidence from England,” in progress, in progress. , Jennifer Key, and Jill L. Dupree, “Learning and the creation of stock market institutions: evidence from the Royal African and Hudson’s Bay Companies, 1670- 1700,” Journal of Economic History, June 1998, 58 (2), 318–344. Carlyle, Thomas, Oliver Cromwell’s letter and speeches, 1897 reprint ed., Vol. 3, Lon- don: Chapman and Hall, 1845. Carruthers, Bruce, City of Capital: Politics and Markets in the English Financial Revolution, Princeton, 1999. Charles I and John Gauden, Eikon Basilike: the portraiture of His Sacred Majesty in his solitude and sufferings, 1966 ed., Ithaca, NY: Folger Shakespeare Library: Cornell University Press, 1649. Clark, Gregory, “The political foundations of modern economic growth: England, 1540-1800,” Journal of interdisciplinary history, Spring 1996, 26 (4), 563–588.

33 Coval, Joshua and Tobias Moskowitz, “Home bias at home: local equity preference in domestic portfolios,” Journal of Finance, December 1999, 65 (6), 2045–2073. Davis, Ralph, English overseas trade, 1500-1700, London: Macmillan, 1973. de la Escosura, Leandro Prados, ed., Exceptionalism and industrialisation: Britain and its European Rivals: 1688-1815, Cambridge: Cambridge U. Press, 2004. de Lara, Yadira Gonzalez, Avner Greif, and Saumitra Jha, “The Administrative Foundations of Self-Enforcing Constitutions,” American Economic Review, May 2008, 98 (2), 105–109. Duca, John V. and Jason L. Saving, “Stock Ownership and Congressional Elections: The Political Economy of the Mutual Fund Revolution,” Economic Inquiry, 2008, (46), 454–479. Falkus, Malcolm E. and John B. Gillingham, eds, Historical atlas of Britain, New York: Crescent books, 1987. Fearon, James D., “Bargaining over objects that inﬂuence future bargaining power.,” October 1996. Paper presented APSA general meetings. Ferguson, Niall, Empire: The rise and decline of the British World Order and the Lessons for Global Power, London: Basic Books, 2002. Firth, C. H. and R. S. Rait, eds, Acts and Ordinances of the Interregnum, 1642-1660, http://www.british-history.ac.uk/report.aspx?compid=55730, 1911. Fischer, David Hackett, Albion’s seed: four British folkways in America, Oxford: Oxford University Press, 1989. Gardiner, Samuel Rawson, History of England from the accession of James I to the outbreak of the Civil War, 1603-1642 1883. , ed., Parliamentary debates in 1610, edited, from the notes of a member of the House of Commons, Vol. 81, London: Camden society, 1862. Gras, N.S.B., “Tudor “Book of Rates”: a chapter in the history of the English customs,” Quarterly Journal of Economics, August 1912, 26 (4), 766–775. Grassby, Richard, “The personal wealth of the business community in seventeenth- century England,” Economic History Review, August 1970, 23 (2), 220–234. Greenwood, Robin and Stefan Nagel, “Inexperienced investors and bubbles,” 2006. Stanford GSB working paper. Grullon, Gustavo, George Kanatas, and James P. Weston, “Advertising, breadth of ownership and liquidity,” Review of Financial Studies, 2004, 17, 439–461. Guiso, Luigi and Tullio Jappelli, “Awareness and stockmarket participation,” Review of Finance, 2005, 9, 537–567. , Michalis Haliassos, and Tullio Jappelli, “Stockholding in Europe: where do we stand and where do we go?,” Economic Policy, April 2003, pp. 123–70. Haber, Stephen and Enrico Perotti, “The Political Economy of Finance,” Journal of Economic Literature, forthcoming 2010.

34 Hakluyt, Richard, Voyages, Everyman 1962 ed., Vol. 1, London: Dent, 1589. Hayek, Friedrich, The constitution of liberty, Chicago: University of Chicago Press, 1960. Hexter, J. H., The reign of King Pym, Cambridge, MA: Harvard University Press, 1941. Hill, Christopher, The century of revolution:1603-1714, New York, NY: Routledge, 1961. Hillmann, Henning, “Mediation in Multiple Networks: Elite mobilization before the English Civil War,” American Sociological Review, June 2008, 73, 426–54. Hobbes, Thomas, Behemoth or the Long Parliament, 1889 ed., London: Simpkin, Marshall and co., 1682. Hong, Harrison, Jeffrey D. Kubik, and Jeremy Stein, “Social interaction and stock market participation,” Journal of Finance, February 2004, 59 (1), 137–163. Howat, G.M.D., Stuart and Cromwellian foreign policy, New York: St. Martin’s Press, 1974. Imbens, Guido W., “Sensitivity to Exogeneity Assumptions in Program Evaluation,” American Economic Review, May 2003, 93 (2), 126–132. James, T.E., “The Age of Majority,” American Journal of Legal History, January 1960, 4 (1), 22–33. Jones, Steven L., William L. Megginson, Robert C. Nash, and Jeffry M. Netter, “Share Issue Privatizations as Financial Means to Political and Economic Ends,” Journal of Financial Economics, 1999, 53, 217–253. Kaustia, Markku and Sami Torstila, “Political Preferences and Stock Market Par- ticipation,” Helsinki School of Economics working paper, 2008. Kupperman, Karen Ordahl, Providence Island, 1630-41, Cambridge: Cambridge Uni- versity Press, 1993. Macaulay, Thomas B., History of England, Vol. 6, London: Longmans, Green and co., 1898. Malmendier, Ulrike and Stefan Nagel, “Depression babies: do macro-economic experiences affect risk-taking?,” August 2007. Stanford GSB. Markowitz, Harry, Portfolio selection: efficient diversification of investments, New York: John Wiley, 1959. Mary Frear Keeler, The Long Parliament:1640-1641, a biographical study of its members, Philadelphia, PA: American Philosophical Society, 1954. McGrath, Patrick, Papists and Puritans under Elizabeth I, New York: Walker and co., 1967. Megginson, William L. and Jeffry M. Netter, “From State to Market: A survey of empirical studies on Privatization,” Journal of Economic Literature, 2001, 39 (2), 321–389. Merton, Robert, “A simple model of capital market equilibrium with incomplete information,” Journal of Finance, 1987, (42), 483–510.

35 Moore, Barrington, Social origins of dictatorship and democracy: lord and peasant in the making of the modern world, 1993 edition ed., Boston: Beacon, 1966. Morrill, John S., The nature of the English Revolution, New York, NY: Longman, 1993. Murrell, Peter, “Design and evolution in institutional development: the insignificance of the English Bill of Rights,” December 2009. mimeo, Maryland. Neal, Larry and Stephen Quinn, “Networks of information, markets and institutions in the rise of London as a financial centre, 1660-1720,” Financial history review, April 2001, 8, 7–26. North, Douglass C. and Barry R. Weingast, “Constitutions and commitment: evolution of institutions governing public choice,” Journal of Economic History, December 1989, 49, 803–32. O’Brien, Patrick K., “The political economy of British taxation, 1660-1815,” Economic History Review, 1988, 2 (41). and Philip A. Hunt, “The rise of a fiscal state in England, 1485-1815,” Historical research, June 1993, 66 (160), 129–176. and , “England, 1485-1815,” in Richard Bonney, ed., The rise of the fiscal state in Europe, Oxford: Oxford University Press, 1999. Ormrod, David, The rise of commercial empires: England and the Netherlands in the Age of Mercantilism, 1650-1770, Cambridge, UK: Cambridge University Press, 2003. Perotti, Enrico C. and Ernst-Ludwig von Thadden, “The Political Economy of Corporate Control and Labor Rents,” Journal of Political Economy, February 2006, 114 (1), 145–174. Pincus, Steven C. A., 1688: The First Modern Revolution, Yale University Press, 2009. Porta, Rafael La, Florencio Lopez de Silanes, Andrei Shleifer, and Robert W. Vishny, “Law and Finance,” Journal of Political Economy, December 1998, 106 (6), 1113–1155. Porter, Harry Culverwell, Reformation and reaction in Tudor Cambridge, Cambridge, UK: Cambridge University Press, 1958. Quinn, Stephen, “The Glorious Revolution’s effect on British private finance: a micro- history,” Journal of Economic History, 2001, 61 (3), 593–615. Rabb, Theodore K., Enterprise and empire: merchant and gentry investment in the expansion of England, 1575-1630, Cambridge,MA: Harvard University Press, 1967. , Jacobean Gentleman: Sir Edwin Sandys, 1561-1629, Princeton, NJ: Princeton Uni- versity Press, 1998. Rajan, Raghuram and Rodney Ramcharan, “Landed interests and financial underdevelopment in the United States,” September 2008. mimeo, Chicago GSB. Rajan, Raghuram G., “Competitive Rent Preservation, Reform Paralysis, and the Per- sistence of Underdevelopment,” American Economic Journal: Macroeconomics, forthcoming.

36 and Luigi Zingales, Saving capitalism from the capitalists: unleashing the power of financial markets to create wealth and spread opportunity, New York: Crown business, 2003. Rapp, Richard T., “The unmaking of the Mediterranean trade hegemony: international trade rivalry and the commercial revolution.,” Journal of Economic History, September 1975, 35 (3), 499–525. Roberts, Stephen K., ed., History of Parliament: the House of Commons, 1640-1660, London: History of Parliament Trust, forthcoming. Rodger, N. A. M., The command of the ocean: a naval history of Britain, 1649-1815, Vol. 2, London: Penguin, 2004. Scott, William Robert, The constitution and finance of English, Scottish and Irish joint-stock companies to 1720: companies for foreign trade, colonization, fishing and mining, Vol. 2, Cambridge, UK: Cambridge University Press, 1910. , The constitution and finance of English, Scottish and Irish joint-stock companies to 1720: the general development of the joint-stock system to 1720, Vol. 1, Cambridge, UK: Cambridge University Press, 1912. Sharpe, William, “Capital Asset Prices: A theory of market equilibrium under conditions of risk,” Journal of Finance, 1964, 19, 425–442. Sheail, John, The regional distribution of wealth in England as indicated in the 1524/5 lay subsidy returns, Vol. 1 and 2, Chippenham, Wiltshire: List and Index Society, 1998. Smith, David L., The Stuart Parliaments, 1603- 1689, London: Arnold, 1999. , “Politics in early Stuart Britain, 1603-1640,” in Barry Coward, ed., A companion to Stuart Britain, Oxford, UK: Blackwell, 2003, chapter 14. Stasavage, David, Public debt and the birth of the democratic state: France and Great Britain, 1688-1789, Cambridge: Cambridge U. Press, 2003. Stone, Lawrence, “The Bourgeois Revolution of Seventeenth Century England Revis- ited,” Past and Present, November 1985, (109), 44–54. Sussman, Nathan and Yishay Yafeh, “Constitutions and commitment: evidence on the relation between institutions and the cost of capital.,” working paper, Hebrew University of Jerusalem 2002. Tawney, R.H., “The Rise of the Gentry, 1558-1640,” Economic History Review, 1941, 11, 1–38. Whitbourne, Richard, “A discourse containing a louing inuitation . . . for . . . Nevv- found-land, lately vndertaken,” Early English Books Online 1622. Willan, T.S., Studies in Elizabethan foreign trade, New York, NY: Augustus Kelley, 1968. Williamson, James A., “Richard Hakluyt,” in Edward Lynam, ed., Richard Hakluyt and his successors, London: Hakluyt Society, 1946, chapter 1. World Bank, World Development Report: Building Institutions for Markets, New York, NY: Oxford University Press, 2002.

37 Wrigley, E. Anthony, “Urban growth and agricultural change: England and the Con- tinent in the early modern period,” Journal of interdisciplinary history, Spring 1985, 15 (4), 683–728. Zhu, Ning, “The Local Bias of Individual Investors,” mimeo, Yale, February 2002.

38 (a) MP or father investor (b) MP investor

Figure 1: Constituency-wise distribution of Long Parliament shareholders (1640). NB: Durham had no representation until 1678

% MPs in constituency 0% - 20% 20.1% - 40% 40.1% - 60% 60.1% - 80%

80.1% - 100% % MPs in county 0% - 20% 20.1% - 40% 40.1% - 60% 60.1% - 80% 80.1% - 100% )" London (Holborn)

Figure 2: % MPs rebelling in each constituency, England and Wales, 1642- 44 NB: Durham had no representation until 1678.

5 percent reduction 0.18 1 percent reduction

0.16

0.14

0.12

0.1

0.08

0.06 Partial R−squared Parliamentary Control 0.04

0.02 PurMinPC Puritan NRegCo CourtCon DistancetoConst lnPopDen BoroughInManor3 Demesne 0InherLand HeirUrban00FNobleFSirBtCastlesPtAtlant Port 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 Partial R−squared Shareholding

Figure 3: Assessing sensitivity to an unobserved covariate. Partial correlations between observed personal and constituency covariates, shares and rebellion relative to thresholds necessary for a single unobserved binomial covariate to reduce the eﬀect of shares on support for Parliament to insigniﬁcance at the 5% and 1% levels.

40 t n1 ermvn vrgs sn h idw17-60 h e iedntsidvdascoming individuals denotes line red The 1570-1630. window 1585. the in using age averages, of moving year 10 on ﬁts 5: Figure box the median, the depicts line old central noble, The father seminary. baronet, or range. or college inter-quartile knight the puritan father depicts a age heir, age, at ties, investment, attendance share court on and inherited gentry, (0) royalist manors, or inherited (1) land, rebel of inherited regression probit a upon based merchants 4: Figure rbblt frblinb hrhligaogmrhnsadnon- and merchants among shareholding by rebellion of Probability rbblt fM netetb g fmjrt fe 1585 after majority of age by investment MP of Probability Probability of owning shares

−.5 0 .5 1 group, by rebellion of probabilities predicted of distribution the depict boxplots These −2

Pr(rebel) .2 .4 .6 .8 1 o-hrhlessaeodr o-hrhlesshareholders non-shareholders shareholders non-shareholders ecat non-merchants merchants 0 Decades MPofageafterDrake(1585) local average 41 quarticfit 2 4 2 ahrinvestor, father , Quartic suc:oncluain ae pnOBinadHn 19,1999), (1993, Hunt and O’Brien upon based calculations own (source: 7: Figure 1570-1630. window the using averages, moving year 10 on fits 6: Figure g rfie n nomnsbfr n fe rk’ voyages. Drake’s after and before endowments and profiles Age h ieo vrescsosrvne ro oteCvlWr(1642). War Civil the to prior revenues customs overseas of rise The

Attended puritan seminary Inherited a manor Endowed court connections 1000s sterling, 1600 base −.5 0 .5 1 0 .2 .4 .6 .8 1 0 .5 1 −2 −2 −2 0 10 20 30 40 1550 local average local average local average 0 0 0 Decades MPofageafterDrake(1585) Decades MPofageafterDrake(1585) Decades MPofageafterDrake(1585) 1560 2 2 2 quartict quartict quartict 1570 crown assetsales overseas customsrevenue 4 4 4

Port constituency Inherited land Distance to London 1580 −.2 0 .2 .4 0 .5 1 1.5 0 .5 1 1.5 2 −2 −2 −2 local average local average local average 0 0 0 1590 Decades MPofage after Drake(1585) Decades MPofage after Drake (1585) Decades MPofage after Drake (1585) 42 year 2 2 2 1600 quartic quartic quartic t t t 4 4 4 1610

Borough constituency Pre−Tudor gentry Merchant apprenticeship 0 .5 1 1.5 0 .2 .4 .6 .8 1 −.5 0 .5 1 −2 −2 −2 1620 oigaverages moving local average local average local average 0 0 0 direct taxes Decades MPofage after Drake (1585) Decades MPofage after Drake (1585) Decades MPofage after Drake (1585) 1630 2 2 2 quartic quartic quartic t t t 1642 ). 4 4 4 Quartic iwytruhti eid osisfo h rcdn eidaecasfiduigtedfiiinabove. definition the using classified are period preceding the from ships so period, (1997,2004). this through Rodger midway in line lists the ship of Ships from calculations own source: 8: Figure h mrec fteRylNv olwn h ii War. Civil the following Navy Royal the of emergence The : > 0 50 100 150 200 0 50 100 150 200 0 os&5+gn.Tenvlcasfiaino siso h ie a introduced was line” the of “ships of classification naval The guns. 50+ & tons 500 1640 1500

1650 Civil War b h oa ayadisrvl,1640-1715 rivals, its and Navy Royal The (b) a asisi h oa ay 1518-1750 Navy, Royal the in Warships (a) 1660 1550 1670 riesShipsoftheline Total warships Cruisers ehrad Denmark France Netherlands England (Britain) 1680 1600 1690 43

Year Glorious Rev Year 1700 1650 Civil War 1710 1720 Cruisers 1700 1730 ships : 1740 1750 1750 > 0 osburthen; tons 100 Table 1: Data sources

Variable Original source Secondary source Individual data Straffordian List posted in Westminster Yard Brunton & Pennington (1954) Member of parliamentary Commons lists (1642-48), Compounding Keeler (1954), Brunton & Pennington coalition Committee (1642-48) (1954), DNB (2007) Sat in Rump Parliament (1648) Commons lists (1642-48) Brunton & Pennington (1954) Sat in Cavalier Parliament Commons lists (1642-48) Brunton & Pennington (1954) Proportion of compounding fine Commons lists (1642-48), Compounding Keeler (1954) (for royalists) Committee (1642-48) Date of disablement from Commons lists (1642-48) Keeler (1954) Parliament Loan for the Cause 1642 Commons lists (1642-48) Keeler (1954) Loan for the State 1640 Commons lists (1642-48) Keeler (1954) Holder of royal office Biographies with multiple sources Keeler (1954), DNB (2007), HPT Investors in (failed) joint stock Charters (1575-1640), Company books Scott (1912), Rabb (1998), Keeler company (1575-1640) (1954), Willan (1968b) Father investor in (failed) joint Charters (1575-1640), Company books Scott (1912), Rabb (1998), Keeler stock company (1575-1640) (1954), Willan (1968b) Member of merchant guild/ Charters (1575-1640), Company books Scott (1912), Rabb (1998), Keeler regulatory co. (1575-1640) (1954), Willan (1968b) Heir Biographies with multiple sources Keeler (1954), DNB (2007), HPT Inherited manors Biographies with multiple sources Keeler (1954), DNB (2007), HPT Family wealth from land Biographies with multiple sources Keeler (1954), DNB (2007), HPT Family wealth from commerce Biographies with multiple sources Keeler (1954), DNB (2007), HPT Armigerous before Tudor dynasty Biographies with multiple sources Keeler (1954), DNB (2007), HPT

Father knight or baronet Biographies with multiple sources Keeler (1954), DNB (2007), HPT Father noble Biographies with multiple sources Keeler (1954), DNB (2007), HPT Age, Date of birth & death Biographies with multiple sources Keeler (1954), DNB (2007), HPT Attended puritan college/ Emmanuel & Sidney Sussex College & Keeler (1954), DNB (2007) seminary Lincolns Inn members lists

Constituency data Borough Commons lists (1642-48) Keeler (1954), DNB (2007) Elizabethan port Customs revenue lists (late 16C) Willan (1968) Town above 5000 Estimates: towns above 5000 in Wrigley (1985) population (1600) Puritan ministers Diocesan records (1600)- assigned to Usher (1910) county in ratio of area in diocese In royal demesne Traditional crown lands (in 1415) Falkus and Gillingham (1987) Castle in constituency Castles extant in 1415 Falkus and Gillingham (1987) Population of county 1834 enumeration abstract (estimates) Rickman (1834), Wrigley and Schonfield County / Borough wealth Tudor lay subsidy assessments (1524-25, Sheail (1998) 1545) Number of taxpayers in Tudor lay subsidy assessments (1524-25, Sheail (1998) constituency 1545)

Abbreviations: HPT: History of Parliament Trust (forthcoming), DNB: Oxford Dictionary of National Biography

44 Table 2: Summary statistics, by shareholding and by support for reform

Non-investors Investors Welch Royalist Reformist Welch Obs Mean SD Mean SD (Prob>|t|) Mean SD Mean SD (Prob>|t|) Outcomes Supported Parliament in Civil War (1642-48) 534 0.51 0.50 0.75 0.44 0.000 - - - - - Supported Crown Advisor (Strafford) (1640) 548 0.14 0.34 0.05 0.22 0.002 0.26 0.44 0.01 0.12 0.000 Contributed to London Defense (1642) 548 0.34 0.47 0.52 0.50 0.001 0.11 0.32 0.60 0.49 0.000 Amount for London Defense (1642) 547 90.05 222.39 155.30 242.36 0.010 31.40 139.60 165.70 267.89 0.000 Sat in Rump Parliament (1648-53) 535 0.23 0.42 0.31 0.46 0.124 0.00 0.07 0.44 0.50 0.000 Log(income) (wills/ fines) (1640-50) 270 6.72 1.12 7.11 1.29 0.047 6.77 1.16 6.87 1.18 0.523 Held court office (before 1640) 548 0.24 0.43 0.33 0.47 0.067 0.37 0.49 0.17 0.38 0.000 Individual endowment controls

45 Father investor 548 0.23 0.42 0.31 0.46 0.111 0.27 0.44 0.24 0.43 0.474 Merchant (apprenticed) 548 0.11 0.32 0.23 0.42 0.005 0.07 0.26 0.19 0.39 0.000 Gentleman (inherited a manor) 548 0.72 0.45 0.69 0.46 0.564 0.73 0.44 0.69 0.46 0.315 Heir 545 0.74 0.44 0.71 0.45 0.577 0.72 0.45 0.74 0.44 0.737 Father noble 548 0.10 0.30 0.05 0.22 0.071 0.12 0.33 0.06 0.24 0.021 Father knight or baronet 548 0.35 0.48 0.34 0.48 0.925 0.37 0.48 0.32 0.47 0.272 Landed prior to Tudor dynasty 548 0.21 0.40 0.16 0.36 0.194 0.22 0.42 0.18 0.38 0.209 Inherited tie to royal court 548 0.28 0.45 0.29 0.46 0.823 0.37 0.48 0.23 0.42 0.001 Constituency distance to London (100km) 548 0.86 0.91 0.71 0.85 0.090 1.05 0.91 0.66 0.85 0.000 Religious endowment controls Attended Puritan seminary or college 548 0.16 0.36 0.16 0.37 0.869 0.11 0.31 0.19 0.39 0.009 Puritan ministers per 10000 in county 548 0.43 0.70 0.62 0.84 0.030 0.29 0.59 0.61 0.81 0.000

Years of age after Drake's Voyage (1585) 536 31.18 12.33 23.46 10.56 0.000 33.19 11.90 26.88 11.93 0.000 NB: The column ``Welch'' provides t-tests of equality of means, allowing unequal variances. Table 3: Regression: Probability of Holding Overseas Shares

(1) (2) (3) (4) (5) (6) (7)

No Omitting County Personal Constituency controls Middlesex FE Probit Probit Probit dF/dX dF/dX OLS dF/dX OLS OLS OLS Merchant (apprenticed) 0.171** 0.169** 0.169** 0.109* 0.113* 0.111* 0.113* [0.070] [0.073] [0.070] [0.066] [0.064] [0.065] [0.064] Gentleman (inherited a manor) 0.013 0.011 0.012 0.020 0.016 0.021 0.001 [0.039] [0.059] [0.057] [0.058] [0.055] [0.056] [0.059] Father investor 0.072* 0.072* 0.073* 0.066* 0.066* 0.055 [0.038] [0.037] [0.043] [0.039] [0.040] [0.044] Inherited land 0.015 0.017 -0.005 0.003 -0.009 0.001 [0.060] [0.061] [0.061] [0.057] [0.058] [0.065] Heir -0.012 -0.010 -0.022 -0.022 -0.017 -0.016 [0.059] [0.059] [0.059] [0.056] [0.057] [0.057] Father knight or baronet -0.005 -0.005 -0.017 -0.017 -0.020 -0.023 [0.038] [0.038] [0.037] [0.037] [0.039] [0.043] Father noble -0.107* -0.117* -0.114** -0.121* -0.123* -0.095 [0.058] [0.066] [0.051] [0.062] [0.062] [0.072] Landed prior to Tudor dynasty -0.029 -0.027 -0.042 -0.040 -0.038 -0.039 [0.045] [0.043] [0.047] [0.045] [0.046] [0.052] Inherited tie to royal court 0.040 0.040 0.066* 0.061 0.054 0.046 [0.036] [0.036] [0.040] [0.038] [0.039] [0.045] Constituency dist to London (100km) -0.034 -0.033 0.038 0.037 0.045* -0.039 [0.024] [0.022] [0.028] [0.026] [0.026] [0.056] Puritan education 0.040 0.041 0.032 0.033 0.036 0.038 [0.049] [0.047] [0.044] [0.043] [0.043] [0.043] Borough constituency -0.139*** -0.119*** -0.125*** -0.105** [0.051] [0.043] [0.044] [0.050] Port (1603) 0.176*** 0.175*** 0.164*** 0.191*** [0.053] [0.050] [0.049] [0.069] Urban pop.> 5000 (1600) -0.028 -0.043 -0.068 -0.094 [0.081] [0.096] [0.094] [0.094] Constituency within royal demesne (1415) -0.132*** -0.139*** -0.146*** -0.119 [0.042] [0.050] [0.051] [0.108] Castle in constituency (1415) 0.002 -0.011 -0.010 0.153 [0.072] [0.073] [0.073] [0.122] Puritan ministers per 10000 in county 0.036 0.036 0.035 [0.035] [0.039] [0.039] Log. population density of county (1600) 0.158*** 0.182*** 0.236** [0.043] [0.042] [0.092] Observations 548 545 545 545 545 536 545 (Pseudo) R-squared 0.04 0.10 0.13 0.12 0.15 0.14 0.26 Robust standard errors, clustered at county level; * significant at 10%; ** 5%; *** 1%

46 Table 4: Regression: Supporter of Parliament in Civil War (1642-48) I

(1) (2) (3) (4) (5) (6) (7) Omitting County FE No controls Personal Constituency Middlesex Probit Probit Probit dF/dX dF/dX OLS dF/dX OLS OLS OLS Shareholder in joint stock 0.213*** 0.205*** 0.189*** 0.197*** 0.179*** 0.182*** 0.207*** [0.045] [0.046] [0.042] [0.045] [0.042] [0.043] [0.044] Merchant (apprenticed) 0.214** 0.183** 0.168** 0.203** 0.176** 0.179** 0.150* [0.089] [0.090] [0.080] [0.085] [0.073] [0.074] [0.078] Gentleman (inherited a manor) 0.032 0.030 0.028 0.021 0.021 0.022 0.042 [0.041] [0.048] [0.043] [0.048] [0.042] [0.044] [0.046] Father investor -0.005 -0.003 -0.008 -0.002 -0.002 -0.019 [0.048] [0.044] [0.049] [0.045] [0.046] [0.048] Inherited land 0.007 0.001 0.006 -0.002 -0.002 0.008 [0.054] [0.049] [0.056] [0.051] [0.053] [0.057] Heir -0.001 0.003 0.002 0.006 0.005 0.007 [0.055] [0.049] [0.054] [0.049] [0.049] [0.050] Father knight or baronet -0.058 -0.050 -0.075 -0.061 -0.060 -0.084 [0.057] [0.053] [0.062] [0.055] [0.056] [0.057] Father noble -0.166* -0.151* -0.157* -0.143* -0.142* -0.143 [0.085] [0.079] [0.086] [0.077] [0.079] [0.093] Landed prior to Tudor dynasty -0.017 -0.017 -0.014 -0.017 -0.017 -0.029 [0.066] [0.060] [0.063] [0.056] [0.056] [0.064] Inherited tie to royal court -0.109* -0.105* -0.115* -0.111* -0.110* -0.118* [0.064] [0.060] [0.066] [0.061] [0.063] [0.066] Constituency dist to London (100km) -0.123*** -0.114*** -0.071* -0.068** -0.071** 0.059 [0.027] [0.024] [0.038] [0.034] [0.035] [0.076] Puritan education 0.168** 0.155** 0.148* 0.137* 0.137* 0.112 [0.073] [0.070] [0.078] [0.071] [0.073] [0.080] Borough constituency 0.075 0.070 0.075 0.023 [0.061] [0.053] [0.055] [0.063] Port (1603) -0.068 -0.066 -0.065 -0.069 [0.064] [0.055] [0.054] [0.071] Urban pop.> 5000 (1600) -0.075 -0.053 -0.048 -0.042 [0.118] [0.099] [0.102] [0.135] Constituency within royal demesne (1415) -0.046 -0.052 -0.048 0.021 [0.063] [0.058] [0.061] [0.133] Castle in constituency (1415) -0.013 0.002 0.003 -0.069 [0.069] [0.065] [0.066] [0.115] Puritan ministers per 10000 in county 0.101*** 0.082** 0.082** [0.039] [0.031] [0.031] Log. population density of county (1600) 0.129 0.076 0.050 [0.093] [0.050] [0.123] Observations 534 531 531 531 531 522 531 (Pseudo) R-squared 0.04 0.10 0.13 0.12 0.15 0.14 0.26 AET Bias 0.055** 0.081* 0.060 0.082 [0.026] [0.042] [0.037] [0.052] Implied Lower Bound 0.134 0.098 0.122 0.125 Ratio: Unobservables : Observables 3.469 2.200 3.019 2.512 Robust standard errors, clustered at county level; * significant at 10%; ** 5%; *** 1%; Altonji, Elder, Taber (AET) Bias, Lower Bounds assume selection on observables equal that on unobservables (with standard errors calculated by bootstrapping clusters with 100 replications).

47 Table 5: Regression: Supporter of Parliament in Civil War (1642-48) II

(1) (2) (3) (4) (5) (6) (7) No Omitting + County Personal Constituency controls Middlesex FE Probit Probit Probit dF/dX dF/dX OLS dF/dX OLS OLS OLS Panel A: All investors Shareholder in joint stock 0.264*** 0.254*** 0.242*** 0.244*** 0.231*** 0.233*** 0.270*** [0.043] [0.045] [0.045] [0.045] [0.043] [0.044] [0.044] Shareholder x merchant -0.321** -0.300** -0.270** -0.323** -0.289** -0.301** -0.358*** [0.135] [0.138] [0.126] [0.135] [0.122] [0.137] [0.124] Merchant (apprenticed) 0.296*** 0.265*** 0.259*** 0.286*** 0.271*** 0.277*** 0.264*** [0.067] [0.073] [0.072] [0.072] [0.069] [0.070] [0.068] Gentleman (inherited a manor) 0.025 0.019 0.02 0.011 0.014 0.015 0.034 [0.042] [0.048] [0.043] [0.048] [0.042] [0.043] [0.045] Observations 534 531 531 531 531 522 531 (Pseudo) R-squared 0.05 0.11 0.14 0.13 0.16 0.15 0.27 Panel B: Investors solely in unprofitable joint stock companies Unprofitable shareholder 0.284*** 0.289*** 0.280*** 0.269*** 0.257*** 0.258*** 0.271*** [0.050] [0.054] [0.056] [0.058] [0.057] [0.058] [0.060] Unprofitable JSC x merchant -0.487*** -0.475*** -0.477*** -0.481*** -0.464*** -0.481*** -0.470*** [0.070] [0.077] [0.145] [0.076] [0.134] [0.140] [0.143] Merchant (apprenticed) 0.315*** 0.281*** 0.266*** 0.289*** 0.266*** 0.272*** 0.246*** [0.065] [0.072] [0.070] [0.075] [0.072] [0.073] [0.076] Gentleman (inherited a manor) 0.025 0.024 0.023 0.017 0.019 0.019 0.039 [0.043] [0.049] [0.044] [0.048] [0.043] [0.044] [0.048] Observations 534 531 531 531 531 522 531 (Pseudo) R-squared 0.05 0.11 0.14 0.13 0.16 0.15 0.26

Robust standard errors, clustered at county of constituency level; * significant at 10%; ** 5%; *** 1%; Personal controls include: Heir, Father -investor, -knight or baronet and -noble, Inherited land, inherited tie to royal court, Puritan education, landed prior to Tudor dynasty, constituency distance to London; Constituency controls include: Urban population>5000, borough, port, royal demesne, noble castles, puritan ministers per capita in county, log. population density of county.

48 Table 6: Parliamentary supporters for whom shares predicted to be pivotal

Name Constituency Predicted Prob Mercantile/ Joint Stock Company (Support for Gentry Parliament) endowments? Incursion into Colonization New Trades Spanish/ Portuguese monopolies Robert Harley Herefordshire 0.504 Virginia Nathaniel Stephens Gloucestershire 0.518 Gentry East India Edward Stephens Tewkesbury, Gloucestershire 0.536 Gentry East India Denzil Holles Dorchester, Dorset 0.598 Gentry Dorchester Henry Darley Northallerton, Yorkshire 0.600 Gentry Providence Island Mass Bay Arthur Ingram Callington, Cornwall 0.603 Merchant Guiana Virginia Thomas Walsingham Rochester, Kent 0.604 Gentry Virginia East India John Pym Tavistock, Devon 0.604 Gentry Providence Island Saybrook

49 John Wylde Worcestershire 0.618 Privateering East India Peregrine Pelham Hull, Yorkshire 0.620 Merchant Virginia John Hippisley Cockermouth, Cumberland 0.622 Privateering John Fenwick Cockermouth, Cumberland 0.663 Virginia John Downes Arundel, Sussex 0.664 Virginia John Browne Dorset 0.674 Gentry Privateering Dorchester, Mass Bay, New East India, Levant England, Newfoundland

Oliver St John Totness, Devon 0.681 Providence Island Virginia Africa (Gynney Bynney) William Strode Berealston, Devon 0.689 Dorchester Edmund Fowell Ashburton, Devon 0.692 Privateering John Rolle Truro, Cornwall 0.697 Merchant Levant John Harris Launceston, Cornwall 0.714 Gentry Virginia Benjamin Rudyard Wilton, Wiltshire 0.714 Providence Island Above the line: shareholder MPs who actually rebeled but were likely to support the Crown in the absence of shares, based upon the lower bound share effect with personal, constituency controls and county fixed effects from Table 4. Below the line: additional switchers based on the conventional estimate. Endowments are coded ``merchant'' if the MP was apprenticed in a merchant company as a child; ``gentry'': if inherited a manor. Table 7: Regression: Indicators of Support for Reform

(1) (2) (3) (4) (5) (6) (7) (8) (9) No Omitting Omitting OLS Personal Constituency County FE Personal Constituency County FE controls Middlesex Middlesex

Panel A: Supporter of Crown Advisor (Strafford) (1640) Shareholder in joint stock -0.070*** -0.060*** -0.055** -0.055** -0.052* -0.074*** -0.074** -0.068** -0.077** [0.022] [0.021] [0.022] [0.024] [0.027] [0.027] [0.030] [0.030] [0.034] Shareholder x merchant 0.070* 0.106* 0.079 0.141** [0.041] [0.054] [0.052] [0.060] Merchant (apprenticed) -0.129*** -0.119*** -0.098*** -0.109*** -0.093*** -0.142*** -0.133*** -0.134*** -0.137*** [0.019] [0.029] [0.026] [0.028] [0.029] [0.032] [0.033] [0.034] [0.034] Gentleman (inherited a manor) -0.011 0.014 0.01 0.02 0.03 0.016 0.013 0.022 0.034 [0.035] [0.046] [0.044] [0.044] [0.048] [0.046] [0.044] [0.043] [0.047] Observations 548 545 545 536 545 545 545 536 545 R-squared 0.03 0.08 0.09 0.10 0.21 0.08 0.09 0.10 0.21 AET Bias -0.017* -0.028* -0.040** -0.036* -0.054** [0.009] [0.014] [0.019] [0.020] [0.025] Implied Lower Bound -0.053 -0.032 -0.015 -0.019 0.002 Ratio: Unobservables : Observables 4.107 2.123 1.380 1.530 0.958 Panel B: Contributor to Defense of London (1642) Shareholder in joint stock 0.167*** 0.159*** 0.145*** 0.143*** 0.173*** 0.168*** 0.161*** 0.160*** 0.183*** [0.049] [0.047] [0.049] [0.050] [0.055] [0.049] [0.051] [0.052] [0.055] Shareholder x merchant -0.046 -0.089 -0.098 -0.054 [0.114] [0.113] [0.115] [0.111] Merchant (apprenticed) 0.109 0.014 -0.007 0.011 -0.005 0.029 0.022 0.043 0.012 [0.069] [0.065] [0.060] [0.059] [0.064] [0.086] [0.079] [0.078] [0.079] Gentleman (inherited a manor) 0.03 0.042 0.042 0.042 0.044 0.041 0.04 0.04 0.042 [0.045] [0.050] [0.048] [0.049] [0.056] [0.051] [0.049] [0.049] [0.056] Observations 548 545 545 536 545 545 545 536 545 R-squared 0.03 0.12 0.13 0.13 0.26 0.12 0.14 0.13 0.26 AET Bias 0.016 0.035 0.063* 0.044 0.059 [0.011] [0.023] [0.036] [0.029] [0.044] Implied Lower Bound 0.151 0.119 0.082 0.099 0.114 Ratio: Unobservables : Observables 10.296 4.549 2.304 3.260 2.922 Panel C: Served in the Rump Parliament (1648-53) Shareholder in joint stock 0.057 0.061 0.069 0.07 0.092** 0.121*** 0.123*** 0.127*** 0.143*** [0.042] [0.040] [0.041] [0.043] [0.044] [0.042] [0.039] [0.040] [0.041] Shareholder x merchant -0.303*** -0.295*** -0.330*** -0.281** [0.105] [0.103] [0.104] [0.115] Merchant (apprenticed) 0.144** 0.113 0.081 0.064 0.079 0.211*** 0.177** 0.171** 0.168** [0.070] [0.075] [0.074] [0.073] [0.078] [0.076] [0.075] [0.075] [0.082] Gentleman (inherited a manor) -0.005 -0.016 -0.016 -0.013 -0.017 -0.027 -0.024 -0.022 -0.025 [0.027] [0.037] [0.037] [0.038] [0.040] [0.039] [0.038] [0.039] [0.041] Observations 535 532 532 523 532 532 532 523 532 R-squared 0.02 0.04 0.05 0.04 0.12 0.05 0.06 0.05 0.13 AET Bias 0.019 0.021 0.018 0.004 0.013 [0.012] [0.018] [0.025] [0.018] [0.032] Implied Lower Bound 0.038 0.040 0.051 0.066 0.079 Ratio: Unobservables : Observables 3.019 2.845 3.838 16.547 6.827 Robust standard errors, clustered at county of constituency level; * significant at 10%; ** 5%; *** 1%; Personal controls include: Heir, Father - investor, -knight or baronet and -noble, Inherited land, inherited tie to royal court, Puritan education, landed prior to Tudor dynasty, constituency distance to London; Constituency controls include: Urban population>5000, borough, port, royal demesne, noble castles, puritan ministers per capita in county, log. population density of county. Altonji, Elder, Taber (AET) Bias, Lower Bounds assume selection on observables equal that on unobservables (with standard errors calculated by bootstrapping clusters with 100 replications). A normalized shift in the distribution of unobservables would have to be Ratio: Unobservables: Observables times the shift due to the included observables to explain away the entire effect of shares.

50 Table 8: Regression: Eﬀects of Drake’s Voyages on Shareholding

(1) (2) (3) (4) (5) (6) (7) (8) (9)

No Omitting Omitting OLS Personal +Constituency +County FE Personal +Constituency +County FE controls Middlesex Middlesex Merchant (apprenticed) 0.116 0.123* 0.082 0.071 0.075 0.129* 0.087 0.075 0.079 [0.070] [0.071] [0.064] [0.064] [0.063] [0.070] [0.063] [0.063] [0.063] Gentleman (inherited a manor) 0.013 0.014 0.017 0.022 0.005 0.020 0.022 0.028 0.012 [0.034] [0.053] [0.051] [0.052] [0.056] [0.052] [0.050] [0.052] [0.055] Adult after 1585 0.415*** 0.421*** 0.418*** 0.337** [0.122] [0.116] [0.115] [0.130] (Adult after 1585) x Age -0.243 -0.133 -0.077 -0.333 [0.303] [0.330] [0.329] [0.316] (Adult after 1585) x Age2 -0.416*** -0.353** -0.334** -0.466*** [0.142] [0.156] [0.155] [0.163] Age (decades) -0.013 0.003 -0.020 -0.016 -0.043 0.240 0.107 0.060 0.298 [0.053] [0.054] [0.052] [0.053] [0.057] [0.299] [0.322] [0.321] [0.321] Age2 (decades2) -0.012 -0.015* -0.010 -0.010 -0.006 0.401*** 0.343** 0.324** 0.458*** [0.009] [0.009] [0.009] [0.009] [0.010] [0.142] [0.157] [0.155] [0.161] Observations 536 533 533 525 533 533 533 525 533 R-squared 0.08 0.09 0.14 0.12 0.24 0.11 0.15 0.13 0.24 Joint F-test (1585 variables) 47.07 39.95 41.44 27.86 Probability> F 0.000 0.000 0.000 0.000 Robust standard errors, clustered at county of constituency level; * significant at 10%; ** 5%; *** 1%; Columns cumulate the following controls. Personal controls include: Heir, Father -investor, -knight or baronet and -noble, Inherited land, inherited tie to royal court, Puritan education, landed prior to Tudor dynasty, constituency distance to London; Constituency controls add: Urban population>5000, borough, port, royal demesne, noble castles, puritan ministers per capita in county, log. population density of county.

51 Table 9: Age and Regression Discontinuity Estimates

(1) (2) (3) (4) (5) (6) (7) (8) (9)

+ Constituency, + Constituency, Omitting + Constituency, Personal Personal Personal Personal + Constituency County FE County FE Middlesex County FE

OLS OLS OLS OLS 2SLS-RD 2SLS-RD 2SLS-RD 2SLS-RD 2SLS-RD Panel A: Supported Parliament in Civil War (1642-1648) Shareholder in joint stock 0.148*** 0.158*** 0.191*** 0.212*** 0.549* 0.540* 0.478 0.468 0.173 [0.044] [0.050] [0.048] [0.051] [0.307] [0.312] [0.354] [0.340] [0.204] Shareholder x merchant -0.233* -0.332** 0.082 0.156 0.189 0.269 [0.129] [0.130] [0.440] [0.478] [0.533] [0.460] Merchant (apprenticed) 0.132* 0.122 0.214*** 0.232*** 0.064 0.036 0.050 0.047 0.037 [0.073] [0.073] [0.070] [0.070] [0.082] [0.168] [0.173] [0.184] [0.169] Gentleman (inherited a manor) 0.037 0.055 0.029 0.047 0.036 0.040 0.036 0.035 0.069* [0.043] [0.047] [0.043] [0.045] [0.043] [0.047] [0.046] [0.048] [0.041] Age(decades) -0.024 -0.054 -0.025 -0.057 -0.026 -0.028 -0.035 -0.038 -0.059 [0.054] [0.066] [0.054] [0.066] [0.048] [0.050] [0.054] [0.053] [0.039] Age2 (decades2) -0.009 -0.004 -0.008 -0.004 -0.002 -0.002 -0.002 -0.002 -0.003 [0.009] [0.011] [0.009] [0.011] [0.005] [0.006] [0.006] [0.006] [0.006] Observations 519 519 519 519 519 519 519 511 519 R-squared 0.16 0.29 0.16 0.30 Panel B: Supporter of Crown Advisor (Strafford) (1640) Shareholder in joint stock -0.040* -0.033 -0.052* -0.054 -0.277 -0.201 -0.176 -0.191 -0.197** [0.024] [0.031] [0.029] [0.037] [0.194] [0.180] [0.186] [0.176] [0.096] Shareholder x merchant 0.065 0.134** 0.281* 0.377** 0.295 0.455* [0.044] [0.066] [0.154] [0.162] [0.189] [0.235] Merchant (apprenticed) -0.101*** -0.076*** -0.123*** -0.119*** -0.066 -0.178*** -0.201*** -0.175** -0.228*** [0.028] [0.028] [0.031] [0.035] [0.046] [0.062] [0.066] [0.073] [0.081] Gentleman (inherited a manor) 0.013 0.029 0.016 0.032 0.015 0.029 0.032 0.039 0.040 [0.046] [0.049] [0.046] [0.048] [0.047] [0.044] [0.043] [0.043] [0.042] Age(decades) -0.020 0.010 -0.019 0.011 -0.009 -0.007 -0.009 -0.012 0.013 [0.032] [0.027] [0.032] [0.028] [0.031] [0.031] [0.035] [0.033] [0.020] Age2 (decades2) 0.008 0.004 0.008 0.003 0.003 0.005 0.005 0.005 0.002 [0.006] [0.006] [0.006] [0.006] [0.005] [0.004] [0.005] [0.004] [0.005] Observations 533 533 533 533 533 533 533 525 533 R-squared 0.09 0.22 0.09 0.22 Panel C: Contributor to Defense of London (1642) Shareholder in joint stock 0.120** 0.131** 0.120** 0.138** 0.918*** 0.793*** 0.767** 0.742** 0.235 [0.052] [0.060] [0.053] [0.060] [0.262] [0.290] [0.318] [0.315] [0.266] Shareholder x merchant -0.003 -0.043 -0.524 -0.721 -0.746 -0.161 [0.118] [0.124] [0.390] [0.440] [0.484] [0.471] Merchant (apprenticed) -0.030 -0.032 -0.029 -0.018 -0.134 0.077 0.172 0.189 0.020 [0.068] [0.066] [0.091] [0.087] [0.083] [0.171] [0.170] [0.178] [0.178] Gentleman (inherited a manor) 0.046 0.051 0.046 0.050 0.024 0.010 0.006 0.005 0.037 [0.050] [0.057] [0.050] [0.057] [0.049] [0.049] [0.046] [0.047] [0.052] Age(decades) -0.006 -0.015 -0.006 -0.015 0.017 0.006 0.006 0.002 -0.003 [0.051] [0.050] [0.050] [0.049] [0.042] [0.048] [0.054] [0.054] [0.055] Age2 (decades2) -0.011 -0.008 -0.011 -0.008 -0.002 -0.003 -0.004 -0.004 -0.009 [0.008] [0.008] [0.008] [0.008] [0.004] [0.005] [0.005] [0.005] [0.007] Observations 533 533 533 533 533 533 533 525 533 R-squared 0.14 0.29 0.14 0.29 Panel D: Served in the Rump Parliament (1648-53) Shareholder in joint stock 0.066 0.091* 0.116** 0.132*** 0.208 0.224** 0.164 0.216** 0.269* [0.043] [0.047] [0.044] [0.045] [0.170] [0.110] [0.101] [0.100] [0.139] Shareholder x merchant -0.268** -0.244* -0.584 -0.593 -0.745 -0.632 [0.120] [0.131] [0.378] [0.413] [0.458] [0.474] Merchant (apprenticed) 0.077 0.041 0.167** 0.120 0.062 0.284* 0.265* 0.289* 0.263 [0.070] [0.072] [0.079] [0.085] [0.062] [0.153] [0.155] [0.164] [0.174] Gentleman (inherited a manor) -0.011 -0.013 -0.020 -0.019 -0.006 -0.030 -0.026 -0.025 -0.024 [0.039] [0.042] [0.040] [0.043] [0.038] [0.043] [0.042] [0.043] [0.044] Age(decades) 0.071 0.080 0.070 0.078 0.097*** 0.102*** 0.089*** 0.099*** 0.129*** [0.052] [0.063] [0.052] [0.063] [0.026] [0.021] [0.022] [0.024] [0.030] Age2 (decades2) -0.013 -0.015 -0.013 -0.015 -0.015*** -0.016*** -0.015*** -0.017*** -0.021*** [0.008] [0.010] [0.008] [0.010] [0.005] [0.003] [0.004] [0.004] [0.004] Observations 520 520 520 520 520 520 520 512 520 R-squared 0.03 0.12 0.04 0.12

Robust standard errors, clustered at county of constituency level; * significant at 10%; ** 5%; *** 1%; Successive columns cumulate the following: Personal controls include: Heir, Father -investor, -knight or baronet and -noble, Inherited land, inherited tie to royal court, Puritan education, landed prior to Tudor dynasty, constituency distance to London; Constituency controls include: Urban population>5000, borough, port, royal demesne, noble castles, puritan ministers per capita in county, log. population density of county. First stage for IV in Table 3. Excluded instruments include: Of age after 1585 and interactions with age and age2, weighted using GMM optimal weighting matrix as in Hayashi (2000)

52 Table 10: Regression: Alternative channels- Income and Ideology

(1) (2) (3) (4) (5) (6) (7) (8) (9) + Constituency, + Constituency, Omitting + Constituency, Personal Personal Personal Personal + Constituency County FE County FE Middlesex County FE OLS OLS OLS OLS 2SLS-RD 2SLS-RD 2SLS-RD 2SLS-RD 2SLS-RD Panel A: Log. Contemporaneous income Shareholder in joint stock 0.486** 0.383* 0.363* 0.264 0.950 0.858 0.062 0.062 0.452 [0.205] [0.202] [0.208] [0.183] [0.738] [0.995] [0.864] [0.864] [1.008] Shareholder x merchant 0.987*** 1.263** 0.341 0.924 0.924 1.202 [0.359] [0.572] [1.203] [1.062] [1.062] [1.244] Merchant (apprenticed) 0.472 0.840** 0.048 0.274 0.353 0.219 0.254 0.254 0.245 [0.317] [0.403] [0.301] [0.412] [0.409] [0.403] [0.394] [0.394] [0.497] Gentleman (inherited a manor) 0.081 -0.054 0.083 -0.061 0.063 0.066 0.098 0.098 -0.075 [0.135] [0.182] [0.137] [0.184] [0.140] [0.140] [0.127] [0.127] [0.151] Age(decades) -0.372 -0.405 -0.312 -0.310 -0.291** -0.279* -0.358** -0.358** -0.276* [0.249] [0.255] [0.251] [0.252] [0.137] [0.165] [0.174] [0.174] [0.141] Age2 (decades2) 0.067* 0.071* 0.057 0.058 0.062* 0.059 0.062* 0.062* 0.055** [0.039] [0.038] [0.040] [0.038] [0.032] [0.038] [0.035] [0.035] [0.027] Father investor 0.158 0.048 0.157 0.045 0.102 0.107 0.157 0.157 0.019 [0.146] [0.175] [0.140] [0.166] [0.172] [0.171] [0.142] [0.142] [0.173] Inherited land 0.785*** 0.942*** 0.790*** 0.975*** 0.779*** 0.781*** 0.830*** 0.830*** 0.980*** [0.285] [0.324] [0.292] [0.329] [0.276] [0.281] [0.275] [0.275] [0.287] Heir 0.266 0.393* 0.251 0.347 0.231 0.229 0.235 0.235 0.331 [0.184] [0.223] [0.180] [0.216] [0.202] [0.199] [0.191] [0.191] [0.236] Father knight or baronet 0.438*** 0.368** 0.454*** 0.389*** 0.432*** 0.438*** 0.368*** 0.368*** 0.393*** [0.117] [0.144] [0.117] [0.142] [0.121] [0.119] [0.096] [0.096] [0.126] Father noble 0.759** 1.105*** 0.772** 1.067*** 0.749** 0.755** 0.755*** 0.755*** 1.073*** [0.305] [0.400] [0.301] [0.394] [0.315] [0.306] [0.287] [0.287] [0.355] Landed prior to Tudor dynasty 0.024 -0.013 0.019 -0.017 0.039 0.036 0.007 0.007 -0.014 [0.180] [0.204] [0.176] [0.201] [0.171] [0.168] [0.137] [0.137] [0.166] Inherited tie to royal court 0.053 -0.048 0.038 -0.051 0.041 0.038 0.026 0.026 -0.055 [0.132] [0.156] [0.133] [0.150] [0.135] [0.132] [0.127] [0.127] [0.138] Constituency dist to London (100km) 0.064 -0.004 0.069 -0.013 0.074 0.075 0.175 0.175 -0.01 [0.092] [0.170] [0.093] [0.165] [0.089] [0.089] [0.127] [0.127] [0.141] Observations 265 265 265 265 265 265 265 265 265 R-squared 0.25 0.52 0.27 0.53 Panel B: Courtier before Long Parliament Shareholder in joint stock 0.066 0.067 0.030 0.031 0.470*** 0.528*** 0.528*** 0.559*** 0.794*** [0.049] [0.060] [0.055] [0.064] [0.145] [0.113] [0.144] [0.134] [0.160] Shareholder x merchant 0.192* 0.216* -0.345 -0.392 -0.452* -0.558** [0.106] [0.124] [0.225] [0.248] [0.275] [0.260] Merchant (apprenticed) -0.111** -0.091 -0.176*** -0.161*** -0.162*** -0.044 -0.009 0.011 0.055 [0.047] [0.057] [0.038] [0.050] [0.056] [0.087] [0.093] [0.100] [0.106] Gentleman (inherited a manor) -0.027 -0.035 -0.020 -0.030 -0.033 -0.043 -0.041 -0.035 -0.056 [0.035] [0.039] [0.035] [0.040] [0.038] [0.040] [0.036] [0.036] [0.042] Age(decades) 0.020 0.024 0.021 0.026 0.000 0.011 0.020 0.027 0.064** [0.041] [0.052] [0.042] [0.053] [0.025] [0.018] [0.023] [0.021] [0.029] Age2 (decades2) -0.011 -0.012 -0.011 -0.013 -0.002 -0.004 -0.006 -0.007* -0.010** [0.007] [0.009] [0.008] [0.009] [0.005] [0.004] [0.005] [0.004] [0.005] Inherited tie to royal court 0.488*** 0.484*** 0.487*** 0.482*** 0.475*** 0.478*** 0.464*** 0.473*** 0.473*** [0.037] [0.042] [0.038] [0.043] [0.038] [0.038] [0.037] [0.038] [0.042] Observations 533 533 533 533 533 533 533 525 533 R-squared 0.29 0.36 0.3 0.37 Robust standard errors, clustered at county of constituency level; * significant at 10%; ** 5%; *** 1%; Personal controls include: Heir, Father -investor, -knight or baronet and -noble, Inherited land, inherited tie to royal court, Puritan education, landed prior to Tudor dynasty, constituency distance to London; Constituency controls include: Urban population>5000, borough, port, royal demesne, noble castles, puritan ministers per capita in county, log. population density of county. First stage for IV in Table 3. Excluded instruments include: Of age after 1585 and interactions with age and age2, weighted using GMM optimal weighting matrix, as in Hayashi (2000)

53 Table 11: Major overseas joint-stock companies, 1575-1640 and representation in the Long Parliament

Chartered/ Investors in Long Name formed Members+Parliament (1640) Profitable?* Foreign attacks? Royal intervention Father East Indies Self Father in-law Drake's circumnavigation 1577 8 0 1 0 Yes Spanish Elizabeth major beneficiary Cavendish's ventures 1586 4 0 0 0 No Spanish East India Co 1599 1318 23 38 38 Initially Dutch Crown raises customs Courteen's East Indies Co 1635 7 2 0 2 No Dutch Charles shareholder Africa trades Senegal Adventurers 1588 22 0 1 1 Unknown Portuguese Gynney & Bynney Co 1618 38 1 1 2 No Portuguese Charles shareholder Nicholas Crispe & Co 1630 3 1 0 0 Yes French, Dutch Patentees were courtiers Central / South America incursion Other privateering 1581 1051 28 22 17 No (on avg) Spanish Guiana Co / Amazon Co 1584 105 7 9 12 No Spanish Raleigh executed (Spanish infl.) Drake's 1587 voyage 1587 21 1 1 0 Yes Spanish Elizabeth shareholder Fenton's voyage to Brazil 1592 42 0 2 1 No Portuguese Bermuda Co 1612 177 2 8 13 Initially Crown raises customs Providence Island Co 1630 23 13 5 5 No Spanish New endeavours within Europe Muscovy Co / Greenland Adv** 1555 211 3 3 4 Yes Dutch Irish co (Munster & Londonderry) 1586 762 6 14 18 Yes Charter revoked Levant Co** 1592 572 9 14 14 Yes Spanish, Barbary Crown raises customs Northwest passage exploration Frobisher's Voyages 1576 121 1 8 6 No Other NWP ventures 1584 311 3 21 20 No Hudson's NWP venture 1610 23 1 1 3 No North America Gilbert's enterprises 1578 147 1 4 3 No Gosnold's voyage 1602 5 0 0 1 Yes Weymouth's voyage 1605 5 0 1 1 Yes Other New England ventures*** 1606 70 15 12 12 No Virginia Co (Roanoke) 1606 (1584) 1671 42 77 69 No Newfoundland/ N. Scotia / Canada Co 1610 58 7 0 4 No French Crown surrenders colonies Baffin's NWP backers 1615 8 0 0 1 No Plymouth Co 1620 50 4 2 0 No Dorchester/ Massachusetts Bay Co 1623 123 13 1 0 No Crown threatens charter += Lower bound;*=Profitable for shareholders prior to Long Parliament; **= switched to regulatory company structure upon confirmation of profitability;***=Northern VA Co, New England Council and delegated patents, including Saybrook Co and Maine Exploration. For sources: see Appendix.

Research Paper No. 2093

Sharing the Future: Financial Innovation and Innovators in Solving the Political Economy Challenges of Development

Saumitra Jha Graduate School of Business, Stanford University

January 2012

R ESEARCH P APER S ERIES

Electronic copy available at: http://ssrn.com/abstract=2001039 Sharing the Future: Financial Innovation and Innovators in Solving the Political Economy Challenges of Development

Saumitra Jha∗ Stanford Graduate School of Business

October, 2011

Abstract The failure to align the incentives of self-interested groups in favor of beneficial reform is often considered a major cause of persistent underdevelopment around the world. However, much less is known about strategies that have been successful at overcoming such political economy challenges. One approach that holds much promise, and in fact appears to have had some historical success, is the provision of financial assets that align the interests of winners and potential losers from reform by providing claims on the future. This paper analyzes the role of financial instruments as a means for fostering broad political coalitions that favor beneficial reforms. It takes as a departure point the benchmark theory of portfolio choice, in which all agents hold the same (market) portfolio and thus all beneficial reforms are adopted. It then analyzes a range of historical cases in which innovative financial assets, often introduced by technocratic reformers, have succeeded at making politics less conflictual over time, focusing on three revolutionary states that subsequently led the world in economic growth: England, the early United States and Meiji Japan. The paper draws upon the theory and the historical cases to assess the promise of finance in solving political economy challenges in contemporary settings.

JEL codes:G00, N20, O10 Keywords: ﬁnance, shares, bonds, reform, revolution, growth

∗Address: [email protected]. This is a version of a paper presented at the July 2011 Inter- national Economic Association World Congress at a session on “Institutions and the History of Economic Development”. I would like to thank Susan Athey, Jon Bendor, Michael Bordo, Gary Cox, Alex Debs, Jim Fearon, Avner Greif, Justin Grimmer, Steve Haber, Chad Jones, Keith Krehbiel, Timur Kuran, Philip Lipscy, Debin Ma, Helen Milner, Kris Mitchener, Dan Posner, James Robinson, Gerard Roland, Dick Sylla and Barry Weingast for very useful discussions. I am grateful to Nick Eubank and Fiona Wilkes for terriﬁc research assistance.

Electronic copy available at: http://ssrn.com/abstract=2001039 1 Introduction

It is a familiar story. A new leader assumes authority in a country that has recently discovered that it has fallen behind its neighbours in economic development. He brings in foreign expertise, announcing a series of modernizing reforms that adopt what is perceived as best development practice and promises much in the way of social benefits. These include the establishment of schools and competitive exams based upon a modern curriculum, the abolition of sinecure positions in government, and the expansion of a system of railways and banks. However the reforms threaten the interests of existing elites. Within four months of the start of the reforms, these elites stage a coup, imprison the executive and rescind the changes. The country faces years of political unrest, civil war and revolution, at great human cost, before reforms occur. Though the failure of the “Hundred Days Reforms”, attempted by the Guangxu Em- peror of China in 1898 (Emperor Kuang Hsu, 1900), does not occupy a prominent place in most modern development economics textbooks, the outcome would not come as a surprise to many. In fact, much blame for persistent under-development around the world has been attributed to a failure to align the incentives of disparate interest groups in favour of political reform and broadly beneficial public policies (Acemoglu, John- son, and Robinson, 2005a, Rajan, 2006). From Latin America, where interests forming around wealth disparities are blamed for weak contemporary institutions (Engerman and Sokoloff, 2000), to African and other developing countries, where polarization along ethnic lines leads to lower growth and greater conflict (eg Montalvo and Reynal-Querol, 2005), to reduced public goods provision in ethnically diverse jurisdictions in the United States (eg Alesina, Baqir, and Hoxby, 2004), to restrictions on financial market development by incumbents seeking to protect their rents (eg Benmelech and Moskowitz, 2010, Haber and Perotti, 2010), a common theme of historical political economy is that within- society differences in endowments– of wealth, ethnicity or of political rights– can lead to conflicting interests that shape the patterns of economic growth and development we see today. These works have greatly deepened our understanding of the importance of political economy to impeding development. Yet the policy prescriptions are less clear. Since differences in endowments create the conflicts of interests central to political economy, then in order to change incentives, one often has to change the endowments of the agents– in fact, to homogenize them. Two homogenizing policies are often suggested in policy and academic circles: partition along ethnic lines as a solution for development in areas facing ethnic conflict, and redistributive policies that foster a “middle class” in areas with inequities in wealth. However, there is some debate about whether ethnically-

2 based partitions actually do lower conflict (eg Sambanis and Schulhofer-Wohl, 2009), and only recently has empirical research begun to shed light on the conditions under which partition may occur peacefully or result in ethnic cleansing (Jha and Wilkinson, 2011). Further, even in ethnically homogeneous societies, the redistribution of endowments to create a middle class will likely be blocked by the potential losers from such redistributive changes. In the political economy tradition, therefore, the policy outlook is grim. It is often historical accidents or external shocks that change endowments that are credited with changing support for beneficial reform. Yet, the problem of aligning the incentives of disparate groups in favour of broad reform and less conflictual politics is very old. Understanding how these problems have been solved in the past may have much to tell us about developing effective prescriptions for contemporary policy. This paper highlights the role that innovations in finance can play and have played in making endowments fungible and aligning interests. The paper takes as a departure point the benchmark theory of portfolio choice, in which all agents hold the same (market) portfolio and thus all beneficial reforms are adopted. It then analyzes the role of finance, deployed often by technocratic problem-solvers, in fostering reforms in three revolutionary states that subsequently led the world in economic growth: England, the early United States and Meiji Japan. In doing so, the paper seeks to draw conclusions for contemporary policy.

2 Complete markets solve the political economy problem.

It is useful to begin with two theoretical benchmarks. First consider an environment with no functioning markets. Let us consider a society with j ∈ J (1×N) agents, each choosing to support or oppose a reform. Each agent possesses endowments of i ∈ I different assets, j W = {ωj1 . . . ωjI }. These “assets” are broadly defined to include physical assets (e.g. land, physical capital), financial assets (e.g. shares in firms or bonds) and human capital characteristics (e.g. endowed social ties, skills and education or ethnicity). Assume that each individual places an initial value of vji on each endowment i, and that she is risk averse and cares only about ex post utility. Let utility increase in the mean rate of return but decrease in the variance (i.e. the risk). The ex post proportional rate of return for individual j on asset i is given by R˜ji. A simple characterization of this is a slightly

3 modiﬁed Markowitz-Tobin utility speciﬁcation:

X δj X U = E( R˜jiω v ) − V ar( R˜jiω v ) (1) j ji ji 2V j ji ji i i

V j denotes the total value of the individual j’s initial endowments in different types of j P assets: V = i ωjivji, and δj > 0. For simplicity, let us define a socially beneficial reform P P r as improving a weighted utilitarian social welfare function j αjUj|r > j αjUj|−r, with P some welfare weights αj, j αj = 1.

Each individual will support the reform only if Uj|r ≥ Uj|−r. Furthermore, an individual will support reform if the overall rate of return to an individual’s endowment increases for a given level of risk, or if the riskiness of the individual’s portfolio of endowments falls for a given rate of return. Whether the reform is adopted will also depend on the allocation of political decision rights to each individual. Let us define a subset N P 1 of decisionmakers g ∈ {0, 1} such that the reform passes if N g ≥ 2 . Thus, in a pure dictatorship, g = {0, 1, 0 ... 0}N , while in a majoritarian democracy under universal franchise g = {1,..., 1}N . The classic political economy problem then stems from the differences in rates of return to individuals’ endowments and the structure of political decisionmaking. For an arbitrary allocation of political decision rights, unless at least half of the decisionmakers g are made better off by the reform, the socially-beneficial reform will not be adopted for the broader population. In the benchmark case above, with no markets, individuals’ interests are shaped entirely by their endowments. Now let us consider the opposite extreme: there are markets for all assets– in other words, there are no uninsurable risks. Now, individuals’ subjec-

tive value of endowments are pinned down by their ex ante equilibrium prices pi, i.e. vji = pi, ∀i, j. The problem above then reduces to the familiar Markowitz portfolio selection problem of choosing that asset portfolio that minimizes the overall variance for a given level of expected return. The efficient portfolio will be the one that eliminates risk unique to each asset. The variance of the overall portfolio can be reduced by diversification: choosing negatively correlated assets. In fact, if individuals are aware of all assets, and there are no transaction costs, each individual will choose to hold the same (market) portfolio of risky assets (Merton, 1987). Depending on risk preference (which may change with wealth), individuals may also hold a greater or small proportion of risk-free assets, but the composition of the risky asset portfolio will not change.1 Consider what happens to a beneficial political reform in these circumstances. Now,

1Risk-free assets, if they exist, would by definition not be contingent on the outcome of the political reform, and thus will not affect the cutoff values to support reform defined by Uj|r − Uj|−r = 0.

4 for any allocation of ex ante political decision rights, from dictatorship to democracy, all beneficial reforms will pass. This is because all individuals, including all decisionmakers, will have the same allocations of state-contingent assets in their portfolio.2 Thus, complete markets can guarantee the enactment of socially beneficial reforms.3 Naturally, the assumptions underlying our modification of the canonical model of portfolio choice are extreme. Not only do we assume the absence of transaction costs, we are expanding our notion of assets beyond those commonly considered financial, such as common stocks, to a broad set of individual characteristics, including ethnicity and other forms of human capital, that are often considered uninsurable (Guiso, Haliassos, and Jappelli, 2003). Yet, this is where financial innovation may help. What is important is not that an endowment can be traded– you may ask, “how would one trade ethnicity?”– but rather a claim on the returns on that endowment can be traded– in other words, is it possible to sell or share the returns on having a particular ethnic identity? We will provide examples of historical situations in Japan and India where it appears that it was. Fundamentally, the theoretical analysis above suggests that we can re-cast what seem like insuperable political economy challenges into potentially solvable problems of reducing transaction costs and creating means for broad groups to invest in shared objectives. This works by giving political decisionmakers the same portfolio of risky assets as the broader population. The creation of financial assets that make fungible the future returns on endowments naturally reduces the political conflicts of interests that may have otherwise formed around differences in those endowments. Thus there can be a political economy multiplier to the more direct effects of financial innovation and the development of financial markets on economic growth (Jha, 2008a). A further reason that financial mechanisms may be even more effective than strategies that stress redistribution or partition is that they can be used to align the incentives of disparate groups by enabling them to share the future. As we shall see, because financial contingent claims allow a sharing of future returns for new opportunities, without nec- essarily requiring the politically more difficult redistribution of existing assets, financial mechanisms may be easier to implement even in politically contentious environments. Even while financial mechanisms can make politics less conflictual in theory, a ques-

2 P P P P j αjUj|r > j αjUj|−r ≡ NUr j αj > NU−r j αj ≡ Ur > U−r, ∀j. 3Note that there may remain conflictual interests on policies that are not Pareto improving– such as straight redistribution. The discussion resonates with important work exploring the role of complete financial markets in allowing agents to perfectly hedge the risks of pro-redistribution political candidates by writing insurance contracts (eg Matozzi, 2010) as well as the role played by privatization in fostering opposition to redistributive politics (see Haber and Perotti (2010) for a useful overview). It also resonates with emerging work on the value of having policymakers hold financial assets in making sovereign debt credible (Cox, 2011).

5 tion that still remains is whether policymakers will implement such financial reforms in reality. In fact, often financial mechanisms have been designed or restricted by incumbents to benefit themselves (Haber and Perotti, 2010). Yet, there are also key examples where financial innovation has been harnessed successfully by technocratic reformers faced with similar problems of aligning incentives of disparate groups and with social welfare objectives in mind. It is to draw lessons from these examples for contemporary policymakers seeking to achieve socially beneficial objectives, whether or not these are in the policymakers’ own short-term interests, to which we now turn.

3 Three Revolutions

Since at least the seventeenth century, the leading nations of the world– the Dutch Re- public, England, the United States and Japan– have been economies where a financial revolution preceded economic growth (Sylla, 2002). In each of these cases, the expansion of finance coincided with a period of remarkable, even revolutionary, political development. In this section we will draw up on the historical examples of the latter three to examine the generalizable lessons that can be drawn from the role played by financial innovation in solving three tough political economy problems: to create a broad coalition in favour of parliamentary control of government in the case of revolutionary England, to align the interests of armed groups and investors against disunity and violent conflict in the aftermath of the American revolution, and to align the interests of the main potential losers of Meiji Japan’s push towards modernization- the samurai. While in England’s case, the alignment of political interests caused by financial innovation may have been an unintended consequence, in the US and Japan, financial innovation was an avowed policy of technocratic reformers (Sylla, 2002). Respectively learning from the experience of first England and then the US, these reformers sought to align the incentives of disparate groups towards the support of national, rather than sectarian, goals.

3.1 Innovation and aligned interests in Revolutionary England

In 1552, inspired by the discovery of the New World and learning from organizational ideas developed in Italy, the English created their ﬁrst joint stock company: the mysterie and companie of the Merchants adventurers for the discoverie of regions, dominions, islands and places unknown. Initially motivated by the high degrees of risk associated with oceanic exploration, the joint stock company was innovative for England in a number of ways. First, large numbers of individuals, particularly non-merchants, could invest in shares and overseas opportunities for the ﬁrst time. Second, unlike the traditional

6 overseas “regulatory companies” where merchants gained the freedom after long apprenticeships to trade on their own account or in small partnerships, now trade was done on behalf of the company. Third, England’s early joint stock companies were run by courts of directors who were elected by votes allocated in proportion to the votes held. These latter two factors meant that joint stock companies had a system of governance designed to accommodate larger groups of investors. The agency problems of such a governance arrangement were somewhat mitigated by the fact that managers were often those most heavily invested themselves. The ships of the first joint stock company, having sailed North rather than South in their search for the Indies, ended up in Russia. Rechristened the Muscovy Company, it and other early joint stock companies later switched back to the regulatory company form once it was established the new trade was profitable and less risky. It was unclear whether the joint stock company would survive as a form of organization. Yet enthusiasm for overseas joint stock ventures in England received a nationwide boost due to the unlikely circumnavigation of the world by Sir Francis Drake (1577-80) and his later successful raid on the Spanish silver fleets (1585) (Jha, 2008a, Rabb, 1967). Though initially small as a proportion of the total English population, joint stock investment was considerably more commonplace among political elites: close to half of the members of parliament that sat between 1575 and 1640 were investors in joint stock companies. The majority of these investors did not come from mercantile backgrounds. However, despite widespread belief in the new opportunities posed by the New World and Asia, England’s joint-stock companies virtually all failed to make profits for their investors before the English Civil War, as the Crown was able to use its rights over customs and over charters to extract much of the benefits from England’s overseas expansion. Overseas customs went from insignificant levels to more than half of the Crown’s revenue on the eve of the English Civil War in 1642 (Jha, 2008a)). It is arguable that the Civil War resulted in large part from conflict over the king’s rights over these new opportunities, rights which would alter the future distribution of bargaining power (Fearon, 1996, Jha, 2008a). Using novel data on the asset holdings of each Member of Parliament and their decisions during the English Civil War, as well as the exogenous shock to the propensity to hold shares among the cohort that came of age at the time of Drake’s return, Jha (2008a) finds that shares in these joint stock ventures appear to have played an important role in aligning incentives of non-merchant members of parliament with existing mercantile interests in favour of parliamentary control of government. Under a lower bound estimate of a 12.5% increase in the propensity to support parliament due to shares, the

7 introduction of shares led to a swing from a majority of 58.6% of MPs supporting continued monarchical rule to 59.0% supporting parliamentary control. Among those who were the “compliant switchers”– those that supported Parliament because of their share investments– were three of the ‘Five Members’ singled out by the king as the ringleaders of the Parliamentary opposition (Jha, 2008a). Rather than extremists, the leaders of the Parliamentary cause during the Civil War were only moderately inclined to revolution and might have actually supported the King in the absence of shares. These shares did not provide wealth, but they did provide the opportunity for future wealth from a common expansionist national policy in support of England’s colonization and trading ventures overseas. Indeed, upon seizing power, the Rump Parliament of the victors of the Civil War, in which previous investors in overseas shares held prominent roles, began to invest heavily in England’s Navy, which grew from unremarkable to the largest in the world (Figure 1, reproduced from Jha (2008a)). A remarkable feature of England’s post-Civil War history was that Britain was to push both political power and joint-stock company holdings beyond the confines of monarchical and oligarchic government. By creating companies that encompassed both merchants and non-merchants, joint stock companies in England aligned disparate interests in a manner arguably different from the large overseas ventures of the Dutch Republic and Italian states on one hand and Spain on the other, where merchants and courtiers respectively were sufficiently wealthy not to need to accommodate new social groups in order to share risks. Instead, because the coalition that took power from the king depended on its strength on the alignment of interests and the contributions of a large group that encompassed merchants and non-merchants, not only were there no individuals who were powerful enough to act unilaterally to return to the pre-Civil War state of executive control, the decisionmakers in power had interests that were broadly aligned with those of other wealth holders, and thus broadly lacked the incentive to do so.4 Thus the new coalition of rulers did not face as severe a ‘executive moral hazard’ problem with respect to their own wealth.5 Instead, a broad consensus spanning both non-merchant and merchant circles appears to have emerged both that England should pursue an aggressive (and expensive) naval and foreign policy expansion and that ultimate control over national policy should remain under the control of the majority of wealthholders, as represented by Parliament, not the

4Even the one possible exception, Oliver Cromwell, was highly constrained in his ability to act unilaterally. The constitution of the Protectorate would in fact form the basis for the Revolutionary Settlement following the Glorious Revolution. 5Expropriation of the disenfranchised, in contrast accelerated in this period, with a rise of Enclosure and Clearance Acts.

8 monarch. The 1688 Revolution was later called “Glorious” precisely because it was virtually uncontested and bloodless in England. Why then did English politics become less conflictual after the Civil War? Not only were the incentives of England’s new ruling coalition aligned with that of other wealth holders, even those who were not initially part of the ruling coalition could benefit from its policies through finance. It is likely that the development of active asset markets that occurred between the Civil War and Glorious Revolution allowed both the winners and losers of the reforms to reallocate their portfolios in favour of those investments benefiting from an assertive foreign policy, and thus new coalitions that spanned a large number of initially disparate interests could be amassed. Thus, England’s government remained that of the wealthy, but not a stable subsection of the wealthy, nor was it closed to entrants.In fact, companies successively issued new stock to accommodate new MPs to Parliament (Scott, 1912). The first political parties– Whig and Tory– transcended traditional cleavages– town versus country, landed versus merchant– and instead coalesced around investments in emergent joint stock companies (Carruthers, 1999). The consolidation of Parliamentary control and stability of foreign policy following the Glorious Revolution led to a boom in public finance of England’s wars (Stasavage, 2003). Holding bonds in the national debt became an accepted asset class, likely further consolidating national support for policies that would finance that debt. By 1718, close to 5% of Eng- land’s wealth may have been held in financial assets (Scott, 1912). Though debate was heated, particularly over the burdens of state finance, it was much less conflictual than before. As suggested by the theory, it is likely that the emergence of secondary markets in shares weakened the link between endowments and opposition to political reform. Thus, rather than becoming an oligarchy of overseas investors, post-Civil War England began a gradual process towards broadly representative government and, ultimately, democracy. What else do we learn from England’s experience? First, nuancing a number of important political economy interpretations of England’s revolution, that have hypothesized that changes to the existing distribution of wealth drove political change by creating a newly enriched “middle” group (Acemoglu, Johnson, and Robinson, 2005b, Moore, 1966), it instead appears that shares aligned incentives by providing a means through which potentially anyone could avail of future opportunities. The remarkably asymmetric weight that people place on losses versus gains is commonplace in behavioural economics. Yet even abstracting from behavioural phenomena, the challenges of taxation in environments, like seventeenth century England, and many contemporary developing countries, where individuals have hidden information about their extent of their resources are of first order importance. With the spread of assets,

9 both stocks and bonds, that would gain in value with the success of England’s post-Civil War expansionist national policies and wars, however, the incentive to cooperate with taxation also was likely higher. The seizure of control by the majority of wealthholders, rather than by a single, potentially capricious, individual, of England’s foreign policies is likely to have generated a more credible willingness to fund the fiscal-military state, both through current taxation and future taxation (i.e. bonds), that was to secure Eng- land’s naval and colonial preeminence in the 18th century. With parliamentary control of government, England also enjoyed a remarkable increase in compliance over taxes. A further lesson from England’s experience may be found in the timing of the opening of secondary markets. The lack of secondary markets until the 1660s for most overseas shares and assets may have been a reason that a broad coalition was able to develop. These assets, from trade rights to the Guinea coast to land rights in Virginia– though of great possible future value, were of little value at the time, and also difficult to sell. The constraints on secondary sales and forward contracts of these assets may have been important for reducing overly-concentrated ownership. This may have broadened initial support for reform and policy change, with the subsequent development of secondary markets then allowing to losers from England’s reforms to also invest and benefit from its new policies. In contrast, the timing of some modern privatizations, that allowed the immediate presence of secondary markets (as in Russia) (Boycko, Shleifer, and Vishny, 1997) or restricted the secondary market but permitted the ability to make forward contracts (as in the Czech Republic) may have undermined attempts to use privatization to create constituencies favouring the strengthening of general property rights in those countries.

3.2 Innovators: post-Revolutionary US and Japan

England’s metamorphosis from monarchical to parliamentary control of government, and the accompanying changes in its financial system took a century of experiment and revolution to effect. However, England’s subsequent military successes in the long eighteenth century at challenging nations, such as France, that were considerably larger, more pop- ulous and resource-rich, made its system of finance one of great interest to technocratic problem-solvers, such as Alexander Hamilton, seeking to implement and consolidate revolutionary reforms elsewhere (Wood, 2009)[p.103]. Though historical accounts of the US and elsewhere credit the foresight of national “founding fathers”, many benchmark models of political economy and institutional change have no role for leadership in institutional design. Those endowed with political decisionmaking capabilities or with wealth are often designated “elites”, and defined by their

10 shared endowments. Yet, a parallel empirical literature suggests that many of the most important changes in policies that affect economic growth and development have been implemented by leaders, often dictators with limited constraints on their ability to make policy (Jones and Olken, 2005). Ironically, the role of policymakers and leaders as institutional mechanism designers has been relatively neglected, despite the fact that aligning the incentives of disparate groups is arguably the central problem in the political economy of development. A key reason for this is that the political economy tradition holds that policymakers should be expected to act in their own best interest. Given that a policymaker’s best interest will often lie in maintaining her position, it is only when reforms benefit her per- sonally will reforms occur. Similarly, other elites, anticipating the actions of a reforming policymaker who might undermine their interests, would act to prevent such a person from gaining authority in the first place. Once again, rather than leadership, shocks to endowments are required to effect political change. To escape the policy straitjacket placed by the logic of political economy, it is useful to recall that innovative ideas can themselves act as shocks. Though technological change is often the crucial driver of neoclassical models of growth, innovation has not occupied a prominent role in historical political economy. Along with technological change, the discovery of a new opportunity– such as those in the New World– or the introduction of a new method of organization– such as the joint stock company– can itself alter political interests. Sometimes, as in England, these new ideas and opportunities have unanticipated consequences that result in a change in interests and thus in reform. At other times they are actually old ideas, adopted and adapted by institutional designers from elsewhere. In the United States and Japan, the very unfamiliarity of innovative ideas may have constituted an opportunity for innovative leadership– as potential losers may not be able to accurately predict the outcome and act to block reforms.

3.3 Revolutionary America, 1790

Mistakes do happen, even when the most foresighted founding fathers are confronted with new ideas. For example, Thomas Jeﬀerson would lament his role in orchestrating the “Compromise of 1790” between James Madison and Alexander Hamilton as the biggest political mistake of his life (Ellis, 2000)[p.51]. According to his own account, Jeﬀerson invited the two to dinner to resolve

two of the most irritating questions that can ever be raised among [the Congress]: 1. The funding of the public debt; and 2. The ﬁxing on a more

11 central residence. . . And, in time it has become probable that unless they can be reconciled by some plan of compromise, there will be no funding bill agreed to; our credit . . . will burst and vanish, and the States separate, to take care every one of itself.6

Indeed, Jefferson’s fear that the individual states might separate loomed large for good reason. Levi Allen, one of the two brothers who helped establish Vermont, sought to make a separate commercial treaty with the British in Canada and to assist in Vermont’s secession (Wood, 2009)[p.112]. There were similarly active discussions by a number of New Englanders who had pre-Revolutionary trading ties with Britain to secede. Spanish attempts to secure the Western hinterland involved closing the Mississippi at New Orleans and offering trade licenses and bribes to settlers in Kentucky and Tennessee, including the commander of the US Army, James Wilkinson (Wood, 2009)[p.113-114]. Anti-Federalists like James Winthrop of Massachusetts considered the difficulties faced by the nascent republic in overcoming the classic political economy challenge of governing such diverse, dispersed interests, and reached a grim conclusion:

The idea of an uncompounded republick, on an average, one thousand miles in length, and eight hundred in breadth, and containing six millions of white inhabitants all reduced to the same standard of morals, or habits, and of laws, is in itself an absurdity, and contrary to the whole experience of mankind. The attempt made by Great-Britain to introduce such a system, struck us with horrour, and when it was proposed by some theorist that we should be represented in parliament, we uniformly declared that one legislature could not represent so many diﬀerent interests for the purposes of legislation and taxation. 7

Yet looking to England’s example of a funded public debt, the pro-federalist Hamilton saw an opportunity where many others saw a liability: the 25 million dollars of debt incurred by the states during the Revolutionary war, and owed mainly to Revolutionary war veterans. While the representatives of the main debtor states such as Massachusetts, Connecticut and South Carolina, saw an opportunity to share the burdens of this debt with taxpayers elsewhere, Virginia, Maryland and Georgia had already paid much of their war obligations (Wood, 2009)[pg141]. Over the famous dinner at Jeﬀerson’s in 1790, Hamilton agreed to support the permanent settlement of the capital in the South on the Potomac and compensate for Virginia’s additional tax burden in exchange to a weakening of Madison’s objections over the assumption of state debt.

6 Thomas Jeﬀerson to James Monroe, June 1790, (Foley, 1900)[p. 59]”. 7Agrippa, no. 4, 3. December 1787, in Herbert Storing (1981), ed. The Complete Anti-Federalist, Vol 1, Chapter 8, Doc. 21, University of Chicago Press.

12 Whether or not the Compromise of 1790 was actually critical to the passage of both bills is subject to some debate (Clinton and Meirowitz, 2004). Regardless, it appears that Hamilton’s victory on Assumption in 1790 may have created a coalition critical for the success of his further reforms, including the creation of the Bank of the United States.8 The Bank of the United States and its branches stimulated the competitive chartering of other banks and joint stock companies by the legislatures of individual states. The number of chartered joint stock companies experienced a boom, rising from seven in the entire colonial history of the United States (1607- 1776) to 28 between 1781-91 and 295 between 1791-1800. Between 1800 and 1830, the ten northeastern US states alone issued more than 3,500 corporate charters (Sylla, 2008). In New York City, the proportion of households holding stock may have risen from 6% in 1790 to 11% by 1826, despite a period of rapid population growth (Hilt and Valentine, 2011). While the empirical link between Assumption and the generation of a coalition of support for Hamilton’s other reforms is a subject of ongoing research (Jha, in progress), there are reasons to believe that it was Assumption that created a coalition that provided support for Hamilton’s Federalist program. Federal and state war debts totalled $67 million, mostly owed in the form of backpay to war veterans, an important organized group with the geographical coverage and resources to likely to play an important role in the subsequent legislative politics or disunion of the United States. At the same time, speculators had bought up the war bonds from a number of veterans, creating some concentration of ownership of the new US public debt (Wright, 2008). While both groups previously would have depended on state legislatures for their payments, now many, particularly in the erstwhile debtor states, had an interest in support the funding and the policies of a Federal government that backed (and could also default) on this debt. In particular, it is likely that the interests of the war veterans became aligned in reducing political conflict and in supporting the integrity of the union. Once again, two important but very different constituencies thus gained similar interests. The fact that these debts could be traded in the emergent market for US securities may have expanded these constituencies beyond veterans and first movers. Indeed preliminary results reveal a robust relation between a Congressman’s decision to vote for the Bank Bill and the proportion of bondholders in his district, even comparing Congressmen from the same state, with the same party affiliation and similar ideal points. To Hamilton’s legislative opponent, Thomas Jefferson, the causal effect of Assumption was clear. Writing three years after the Compromise, Jefferson argued that Assumption:

8For excellent discussions of Hamilton’s reforms, see Sylla, Wright, and Cowen (2009) and Sylla (2008).

13 was unjust in itself, oppressive to the States, and was acquiesced in merely from a fear of discussion. While our government was still in its most infant state, it enabled Hamilton so to strengthen himself by corrupt services to many that he could afterwards carry his bank scheme, and every measure he proposed in deﬁance of all opposition. In fact it was the principal ground whereon was reared up that speculating phalanx, in and out of Congress, which has since been able to give laws to change the political complexion of the United States”9

Jefferson’s mistake and Hamilton’s innovative use of financial instruments to generate a political constituency for reform arguably had lasting and profound effects on the unity and the financial development of the United States. The Federalists’ control over Amer- ican government would not last, but Hamilton’s change of the “political complexion” of the United States, particularly in the creation of a “speculating phalanx” of financial institutions and interests appears to have persisted, surviving Jefferson’s terms as presi- dent and even Andrew Jackson’s veto of the charter of the Bank of the United States in 1832.10 Hamilton’s reforms did not result in a dominant central government- in fact the competitive chartering of state banks to compete with the branches of the Bank of the United States may have instead acted ultimately to also strengthen within-state interests (Haber, North, and Weingast, 2008). Regardless, what the veterans, speculators and local state elites that became the assetholders of both the Federal and the state banks were likely to share was a natural interest in lowering the threat of local violent conflict and in peaceful, legislative resolution of disagreements between the states.

3.4 Revolutionary Japan, 1867-1880

The assumption of debts owed to war veterans in the democratic context of the post- revolutionary United States was to be a strategy also used in the feudal dictatorship of post-revolutionary Japan. The fall of the Tokugawa Bakufu in 1867-68, precipitated in large part by attempts by the US and other countries to ‘open Japan’, brought about the end of over seven hundred years of warrior rule. Though often considered highly

9Thomas Jefferson (1793), in Foley (1900) (page 61). 10Jefferson would remain convinced of the importance of Assumption, writing near the end of his life: . . . and so the Assumption was passed, and twenty millions of stock divided among the favored States, and thrown in as pabulum to the stock-jobbing herd. This added to the number of votaries to the Treasury and made its Chief the master of every vote in the Legislature which might give to the government the directions suited to his political views.– Thomas Jefferson (1818) The Anas. ix, 92, Ford edition, i, 161

14 homogeneous today, Japan, like the early US republic, had strong regional differences, divided into feudal domains, some of which enjoyed a large range of local autonomy from the Shogunate. The Boshin War of 1868, though fought to “restore” the rule of the Meiji emperor, was largely a domainal war between Tokugawa feudatories in the north- east and an alliance of domains, mainly from the Southwest. It was unclear whether the Meiji restoration would simply replace the Tokugawa-led shogunate with a shogunate led by the domains of Satsuma and Choshu, or whether Japan would once again disintegrate (Jansen, 2000). Where Japan also differed from the American Republic and England, was the presence of a hereditary caste system. At the bottom of the hierarchy, ritually unclean professions were the exclusive domain of an untouchable caste, while the samurai atop enjoyed the exclusive right to administrative and military positions in exchange for hereditary stipends of rice. Samurai were legally and custom-bound not to accept other professions or to inter-marry with commoners (heimin) (Jansen, 2000). Lower-ranking samurai, once more urban bureaucrats than men-at-arms, had begun to remilitarize in response to the forced opening up by the United States and other countries. A number had returned from action in the Boshin War. Instead of being rewarded, however, the 1,800,000 samurai would be the likeliest losers of the modernization of Japan. General conscription off all male Japanese was introduced in 1873, abolishing the exclusive rights of the samurai to the military. Large numbers of samurai retainers became unemployed. As Figure 2 suggests, soon afterwards began a series of armed rebellions by declassed samurai, the largest of which, the Satsuma Rebellion of 1877- 78, often considered a civil war, cost U42 million, or 80% of the government’s budget, required the mobilization of 68,000 troops and led to the loss of life of 20,000 ex-samurai and 6,000 government soldiers (Vlastos, 1989). Despite the presence of a class of elite and militarized potential losers and despite the strong regional cleavages, Japan did not disintegrate. Instead, as the historian Mar- ius Jansen (2000)[pg. 335] concludes: “Japan, which began the Meiji period as one of the modern world’s most fractured polities, emerged within a generation as one of its most centralized states”. Japan’s caste distinctions too rapidly diminished. It is further commonly accepted that the reforms of the early Meiji period– including the abolition of samurai privileges, the development of a modern banking system and introduction of private ownership of land– also laid the basis for Japan’s remarkable catch up in economic growth. How then did the administrators of the Meiji era build support for their reforms? Jha and Mitchener (in progress) explore the role played by a series of reforms, in

15 addition to a novel use of deflationary monetary policy in aligning the incentives of the most dangerous potential losers to reform with that of the Meiji state: the declassed samurai. In 1871, the central government abolished the domains, pensioning off their feudal lords, and, like the US republic, took over responsibility for paying the samurai stipends. However, these stipends absorbed close to a third of the government’s budget. Amidst the period of violent samurai protest (Figure 2), the government responded by an innovative package of reforms. First, the samurai’s rice stipends were compulsorily commuted into interest-bearing bonds in 1876. 310,971 ex-samurai received public bonds worth U113 million (Harootunian, 1960). At the same time, the National Bank Regula- tions were modified so that samurai bonds could be exchanged for stock in newly-opening branches of the National Bank. Bank owners were required to capitalize the bank using 80% government (i.e. samurai) bonds. The remaining 20%, in currency, could come from the heimin (commoner) class. Subsequent to these reforms, Matsukata Masayoshi was appointed to the post of Finance Minister in September 1881. He reversed a long-period of inflation that had eroded the value of the 7% samurai bonds to less than two-thirds of their face value by instead implementing a dramatic period of monetary and fiscal tight- ening known as the “Matsukata Deflation” (Figure 2), with 36% of paper yen taken out of circulation (Vlastos, 1989, Jansen, 2000). This macro policy led to a transfer of wealth from debtors (mainly farmers) to bondholders (mainly samurai). Matsukata was clearly aware of the threat to the Meiji state posed by the declassed samurai even after the failure of the Satsuma Rebellion. As he noted in 1883, “If the government remained an onlooker to the plight of the samurai, it would have certainly meant that the government did not understand the relationship between peace and rebellion.”11 Jha and Mitchener (in progress) perform a difference-in-difference analysis, exploiting idiosyncratic pre-Meiji heterogeneity in the average stipend level per samurai in domains and the timing of the Deflation to examine the role played by the Meiji reforms on bank formation, the reduced propensity for samurai rebellion, rises in “debtor” protests, and for the change from violence towards to peaceful political mobilization (juyu minken “popular rights” parties) in favour of representative government. Though there is much empirical work that remains to be done, the qualitative evidence suggests that there was a strong relationship between the reforms and the alignment of samurai interests in favour of modernization and commercial development. The change in banking regulations in 1876 led to a dramatic expansion of bank branches, increasing from seven to around 150 new banks within two years- so many, in fact, that the govern-

11Masayoshi, Matsukata (1883) “Memo Explaining the Way to Eliminate Bank Notes”, cited in Ha- rootunian (1960)[pg.440].

16 ment called a halt to future expansion (Yamamura, 1967). In 1878, 29,360 samurai and nobles controlled U30, 580, 000 in bank stock, compared to U8, 870, 000 held by 4,730 commoners (Harootunian, 1960)[pg.440]. Though commoners would play an increasing role in the banks (Yamamura, 1967), in 1882, samurai still owned three-quarters of the stock of Japan’s banks (Harootunian, 1960). As one might expect among a class of newly-minted bankowners, political risk in the form of violent samurai rebellions appear to have ceased. Instead the samurai played a central role in peaceful political protests (jiyu minken) that helped propel Japan towards the framing of a constitution and its first national elections in 1890. The reforms and the Deflation was not without losers– “debtor’s parties” emerged, and there were 108,850 bankruptcies in 1885 alone (Vlastos, 1989). Though Japan’s reforms appear to have succeeded by using financial instruments and monetary policy to allow the group most likely to impede reforms to share in the future, this may have been at the expense of those who posed less of a threat.

4 Discussion

Our two examples of technocratic leadership using financial innovations to solve political economy problems– the early US republic and Japan– are interesting not just because these became among the fastest growing states of their time following their financial development, but also because they show that financial innovations can work in both a democratic environment, where the challenges faced by technocratic leaders were on building legislative support for reform, and in an emergent caste-ridden dictatorial state, where the threats posed were mainly of violent resistance. In fact, the effect of financial innovation in aligning incentives of groups by allowing them to share in the future, in the case of both England and Japan, may have played an important role in their movement towards representative government. In seventeenth century England, the introduction of the joint stock company allowed non-merchants to share in the opportunities available to overseas merchants through their accumulated human capital– access to overseas trade and skills at navigation. In the early US, the assumption of state debts may have aligned the incentives of a “speculating phalanx” of individuals opposing violent inter-state conflict that became broader as finance expanded. In Japan, shared stock investment by declassed samurai and non- samurai in banks may have played an important role in aligning interests in favour of broader political representation and economic growth. The failure of the Hundred Day’s Reforms in China may in contrast have been, in part, due to failure by Chinese reformers to realize the importance of finance as a means to reconcile the interests of incumbent

17 elites to reform. The social heterogeneity in the Japanese example is particularly useful to consider in contemporary developing societies. Though the status of the samurai as an endoga- mous caste or “ethnicity” had been legally abolished during the first generation of the transition, the Japanese explicitly provided ethnically-based financial assets and an opportunity for samurai and non-samurai to share in the future through the capitalization of long-lived institutions- Japanese banks. In doing so, Japanese reformers appear to have implemented, rapidly and at a large scale, the creation of long-lived institutions that parallel those that supported small inter-ethnic equity partnerships and sustained inter- ethnic cooperation among Hindus and Muslims for centuries.12 However, the Japanese use of financial claims on a shared future instead of ethnic specialization as a means of aligning interests towards cooperation may have been more effective at undermining social and ethnic distinctions. More broadly, there are three useful ideas to consider when attempting to develop policies that may successfully address the gravest political economy challenges of development. First to escape the policy straitjacket of political economy, it is useful to recall that innovative ideas can also act as shocks. Along with technological change, the discovery of a new opportunity or the introduction of a new method of organization or an innovative process to a society can itself alter political interests. Sometimes these new ideas have unanticipated consequences that result in a change in interests and thus in reform. At other times they are actually old ideas, adopted and adapted from elsewhere. As we saw in the cases of revolutionary England, the United States and Japan, the very unfamiliarity of innovative ideas may constitute an opportunity– as potential losers may not be able to accurately predict the outcome and act to block reforms. Beyond the indirect role of technological change in changing interests that lead to political reforms, a particular class of ideas and approaches may be particularly valuable in generating new and broader pro-reform constituencies: those that improve the ability of disparate groups to share the value of future opportunities. Financial mechanisms– shares, bonds and their derivatives– can allow individuals to share ownership claims on both current and future revenue streams, not just of asset classes we commonly consider financial, but even of human capital and ethnicity. In fact, since with complete markets and no transaction costs, agents all hold the same (market) portfolio, we can reduce the

12Though Indian attempts to develop joint-stock companies go back as early as Tipu Sultan’s in the late 18th century, most trade in the Indian Ocean continued to be done, often in large convoys, but through relatively small partnerships until modern times (Kuran and Singh, 2010). This may explain why long-lived Indian institutions supporting inter-ethnic cooperation tended to take diﬀerent and often non-ﬁnancial forms (Jha, 2008b).

18 gravest challenges of political economy to the challenge of creating markets and reducing transaction costs. There is much work to be done: theoretically on the most politically viable path to implement such reforms, and empirically and experimentally, on the effect that the introduction of novel financial instruments, such as those that share ethnic risks, might have on making politics less conflictual and consolidating reform. However, this paper has sought to provide important examples where reforms were consolidated, even in the face of strong opposition, by technocratic leaders who tried new (and sometimes old) financial approaches to political development. By facilitating the process through which individuals can credibly share the gains of a broadly beneficial post-reform future, financial innovations, done right, hold much promise for leaders seeking to solve the political economy problems of development.

References

Acemoglu, D., S. Johnson, and J. Robinson (2005a): “Institutions as the fundamental cause of long-run growth,” in Handbook of economic growth, ed. by P. Aghion, and S. Durlauf, chap. 6, pp. 385–472. North-Holland.

(2005b): “The rise of Europe: Atlantic trade, institutional change and economic growth,” American Economic Review.

Alesina, A., R. Baqir, and C. Hoxby (2004): “Political jurisdictions in heteroge- neous communities,” Journal of Political Economy, 112(2), 348–395.

Benmelech, E., and T. Moskowitz (2010): “The political economy of ﬁnancial regulation: evidence from US usury laws in the 19th century,” Journal of Finance, 65(3), 1029–1073.

Boycko, M., A. Shleifer, and R. Vishny (1997): Privatizing Russia. MIT Press, Cambridge, MA.

Carruthers, B. (1999): City of Capital: Politics and Markets in the English Financial Revolution. Princeton.

Clinton, J., and A. Meirowitz (2004): “Testing explanations of strategic voting in legislatures: A reexamination of the Compromise of 1790,” American Journal of Political Science, 48(4), 675–689.

Cox, G. W. (2011): “Sovereign debt, political stability and bargaining eﬃciency,” mimeo, Stanford.

Ellis, J. J. (2000): Founding Brothers. Alfred Knopf, New York.

19 Emperor Kuang Hsu (1900): The Emperor Kuang Hsu’s reform decrees, 1898. North China Herald Oﬃce, Reprinted from the “North China Daily News”, Shanghai.

Engerman, S. L., and K. L. Sokoloff (2000): “Institutions, factor endowments and paths of development in the New World,” Journal of Economic Perspectives, 14(3), 217–32.

Fearon, J. D. (1996): “Bargaining over objects that inﬂuence future bargaining power.,” Paper presented APSA general meetings.

Foley, J. P. (ed.) (1900): The Jeﬀersonian cyclopedia: A comprehensive collection of the views of Thomas Jeﬀerson. Funk and Wagnalls, New York and London.

Goldsmith, R. W. (1983): The Financial Development of Japan, 1868-1977. Yale University Press, New Haven.

Guiso, L., M. Haliassos, and T. Jappelli (2003): “Stockholding in Europe: where do we stand and where do we go?,” Economic Policy, pp. 123–70.

Haber, S., D. North, and B. Weingast (eds.) (2008): Political Institutions and Financial Development. Stanford University Press, Stanford, CA.

Haber, S., and E. Perotti (2010): “The Political Economy of Finance,” Journal of Economic Literature.

Harootunian, H. D. (1960): “The Economic Rehabilitation of the Samurai Class in the early Meiji Period,” The Journal of Asian Studies, 19(4), 433–444.

Hilt, E., and J. Valentine (2011): “Democratic Dividends: Stockholding, Wealth and Politics in New York, 1791-1826,” working paper 17147, Cambridge, MA.

Jansen, M. B. (2000): The Making of Modern Japan. The Belknap Press of Harvard University, Cambridge.

Jha, S. (2008a): “Shares, coalition formation and political development: evidence from 17th century England,” GSB Research Paper 2005, Stanford Graduate School of Busi- ness, Stanford CA.

(2008b): “Trade, institutions and religious tolerance: evidence from India,” GSB Research Paper 2004, Stanford Graduate School of Business, Stanford CA.

(in progress): “‘That Speculating Phalanx’: war veterans, ﬁnance and nation- building in the early US Republic,” in progress.

Jha, S., and K. Mitchener (in progress): “Swords into Bank Shares: Finance, conﬂict and political reform in Meiji Japan,” in progress.

Jha, S., and S. Wilkinson (2011): “Veterans, organizational skill and ethnic cleansing: evidence from the Partition of South Asia,” SSRN working paper 1739812.

20 Jones, B., and B. Olken (2005): “Do Leaders Matter? National Leadership and Growth Since World War II,” Quarterly Journal of Economics, 120(3), 835 –864.

Kuran, T., and A. Singh (2010): “Economic Modernization in Late British India: Hindu-Muslim diﬀerences,” working paper 53, Duke ERID.

Matozzi, A. (2010): “Policy uncertainty, electoral securities and redistribution,” Inter- national Economic Review, 51(1), 45–71.

Merton, R. (1987): “A simple model of capital market equilibrium with incomplete information,” Journal of Finance, (42), 483–510.

Montalvo, J. G., and M. Reynal-Querol (2005): “Ethnic polarization, potential conﬂict and civil wars,” American Economic Review, 95(3), 796–816.

Moore, B. (1966): Social origins of dictatorship and democracy: lord and peasant in the making of the modern world. Beacon, Boston, 1993 edition edn.

Rabb, T. K. (1967): Enterprise and empire: merchant and gentry investment in the expansion of England, 1575-1630. Harvard University Press, Cambridge,MA.

Rajan, R. G. (2006): “The persistence of underdevelopment: constituencies and competitive rent preservation,” manuscript.

Sambanis, N., and J. Schulhofer-Wohl (2009): “What’s in a Line? Is Partition a Solution to Civil War?,” International Security, 34(2), 82–118.

Scott, W. R. (1912): The constitution and ﬁnance of English, Scottish and Irish joint- stock companies to 1720: the general development of the joint-stock system to 1720, vol. 1. Cambridge University Press, Cambridge, UK.

Stasavage, D. (2003): Public debt and the birth of the democratic state: France and Great Britain, 1688-1789. Cambridge U. Press, Cambridge.

Sylla, R. (2002): “Financial systems and economic modernization,” Journal of Eco- nomic History, 62(2), 277–291.

(2008): “Comparing the UK and US Financial Systems, 1790-1830,” in The Evolution of Financial Markets and Institutions from the Seventeenth Century to the Present, ed. by J. Atack, and L. Neal, pp. 290–338. Cambridge University Press, Cam- bridge.

Sylla, R., R. E. Wright, and D. J. Cowen (2009): “Alexander Hamilton, Central Banker: Crisis Management during the U.S. Financial Panic of 1792,” Business History Review, 83, 61–86.

Vlastos, S. (1989): “Opposition Movements in Early Meiji, 1868-1885,” in The Cam- bridge History of Japan: The Nineteenth Century, Vol 5. Cambridge University Press, Cambridge.

21 Wood, G. (2009): Empire of Liberty: A History of the Early Republic, 1789-1815. Oxford University Press, New York.

Wright, R. E. (2008): One Nation Under Debt: Hamilton, Jeﬀerson and the history of what we owe. McGraw-Hill, New York.

Yamamura, K. (1967): “The Role of the Samurai in the Development of Modern Bank- ing in Japan,” Journal of Economic History, 27(2), 198–220.

22 iwytruhti eid osisfo h rcdn eidaecasfiduigtedfiiinabove. definition the using classified are period preceding the from ships so period, this through midway line Rodger. the N.A.M. of by compiled Ships lists ship from calculations own source: 1: Figure h mrec fteRylNv olwn h ii War. Civil the following Navy Royal the of emergence The : > 0 50 100 150 200 0 50 100 150 200 0 os&5+gn.Tenvlcasfiaino siso h ie a introduced was line” the of “ships of classification naval The guns. 50+ & tons 500 1640 1500

Year Glorious Rev Year 1700 1650 Civil War 1710 1720 Cruisers 1700 1730 1740 ships : 1750 1750 > 0 osburthen; tons 100 hmu,a ie n(odmt,18)p 21]). 1983)[pg (Goldsmith, in cited as Shimbun, 2: Figure rc eesadsmriplce neryMiiJapan. Meiji early in policies samurai and levels Price Wholesale prices (1885=100) 60 70 80 90 100 110 120 130 140 1865

Meiji Restoration and Boshin War 1870

Conscription Act Samurai Rebellions 24 1875 Year

Samurai Stipends converted to Bonds 1880 Deflation Matsukata 1885 suc:teAsahi the (source: