2.1 The early business corporations: the East India Companies

The first large-scale long-lasting joint-stock business corporations were the English East India Company (EIC), established in 1600, and the Dutch East India Company (VOC), established in 1602. Both engaged in long-distance oceanic trade with Asia around the Cape of Good Hope, and both combined the legal concept of a separate entity with the financial scheme of joint-stock investment. This consolidation of the joint-stock with the corporate entity enabled a lock-in of the investments. The corporation which already acquired early in its history the features of a separate pool of assets and collective governance added now also the lock-in feature. The partitioning of assets into the EIC and VOC was not designed in order to provide owner shielding from creditors.

The EIC's most distinctive attribute, as evident in its first charter in 1600, was its incorporation as ‘one body corporate and politic’. As a separate legal entity, it had a full set of legal capacities and privileges to own land, litigate in court and hold franchises such as monopoly. The charter made no mention of joint-stock finance or of limiting the liability of its members (Harris, Reference Harris2020). For the limited liability attribute to have become an issue in the early history of the EIC, it would have had to be financed by both equity and debt. However, debt was initially only a marginal source of corporate finance, due to the high level of uncertainty in the early stages of trade, the high level of on-going risk, the absence of tangible collaterals, and the non-existence of commercial banks or bond markets of the magnitude that could provide long-distance finance. The company was financed almost exclusively by equity, while commercial credit was used only on a small scale, as an integral part of commodity transactions. Thus, there was no significant conflict between shareholders and creditors, and no need to define the liability of shareholders for corporate debt. Some degree of asset-partitioning was evident from the fact that the company was a separate legal entity, but this should not be confused with limited liability in the modern sense.

The EIC started using long-term debt finance only during its challenging years before the Glorious Revolution. The 1685 balance sheet recorded a debt of £783,890 (some of it short-term) and an equity of £1,703,422 (Chaudhuri, Reference Chaudhuri1978: 424). After the Revolution, the company became a financial intermediary, issuing bonds to the investing public, and a lender to the government. At this stage, in theory, it could have become insolvent, and the question of the relative priority of bondholders over shareholders – and related questions of how to limit the personal liability of its shareholders to company debts – could have become an issue. But they did not. The EIC's relationship with the government may have made the company ‘too big to fail’.

The VOC was chartered 2 years after the EIC, in 1602, by the States-General, the Federal Assembly of the Dutch Republic. The charter positioned the company as an entity separated from the state and other parties with the capacity to own property, transact and hold privileges (such as a monopoly or franchise) issued by the state. Unlike the EIC, the VOC's charter also addressed the financial scheme.Footnote ⁴ It contained permission to issue a public offering of shares, granted all residents of the Netherlands the right to subscribe to them, and locked-in the capital thus raised (6,424,588 guilders) for 10 years.

Was the liability of VOC shareholders limited in the modern sense? Prominent economic historians of the company (Gelderblom et al., Reference Gelderblom, De Jong and Jonker2013) have argued that, by 1623, the VOC was a limited liability corporation.Footnote ⁵ When the Middelburg Chamber postponed paying import duties in 1611, the Zeeland tax collectors did not seize the VOC property but threatened the directors with imprisonment for debt instead. From this, these authors concluded that the directors bore personal liability for company debt, and may have inferred that passive shareholders (participanten) enjoyed limited liability. In 1623, it was resolved that the text of bonds would be rewritten to explicitly exclude creditors' recourse to the signatories' person or property.

I think that this interpretation of historical facts is misjudged. The VOC liability regime is confusing because, unlike the EIC, it had two classes of shareholders: the aforementioned passive participanten and the active bewindhebbers. The passive shareholders were entitled only to a share in the profits. The VOC also had two levels of governance, six city Chambers and a united company.Footnote ⁶ The active shareholders served by their status as the governors of their respective city Chambers and could be appointed as the united company's directors (members of the Herren XVII). The liability of the passive shareholders of the VOC was not defined in the charter or anywhere else (de Jongh, Reference de Jongh and Koppell2011). Clause 42 of the VOC Charter specifically exempted directors, who were also the active shareholders, from salary debts. The charter does not refer to the event of insolvency and dissolution.

The VOC was initially financed exclusively by equity investment. When resorting to debt finance, each VOC City Chamber borrowed according to its needs and at the discretion of its directors (active bewindhebbers). The survey of VOC bonds by Gelderblom et al. (Reference Gelderblom, De Jong and Jonker2013) shows that City Chamber directors contracted loans on their personal account, pledging their person and goods. This personal borrowing may have resulted from the fact that the chambers, unlike the VOC itself, were not viewed as separate legal entities. The question of whether passive shareholders had limited liability was muted by the fact that active shareholders assumed personal liability. In 1617, the directors centralized borrowing and resolved that henceforth they – and not the City Chambers – would make borrowing decisions. In 1622, bonds were no longer issued and signed by the directors, but by the bookkeepers, and they no longer carried the signatories' customary guarantee of person and goods. But this happened because the VOC was a legal personality and the directors ceased borrowing personally, acting instead as VOC agents borrowing on its behalf. There is no evidence that the liability regime applying to passive shareholders in the event of insolvency and dissolution was defined in 1602, in 1623 or at any later stage, or that the liability attribute of the VOC was separated from the legal personality attribute and conceptualized as a limited liability attribute in the modern sense.

Over the following years, the VOC's resort to longer-term debt finance increased because, unlike the EIC, it did not raise additional equity investment beyond the initial public offering of 1602. However, the mid-life VOC was backed by the state and was never under real risk of insolvency. When the VOC became insolvent in 1799, the Dutch government revoked its charter and nationalized the company's territorial possessions in Asia, its assets and debts (Sinninghe and van de Vrugt, Reference Sinninghe, van de Vrugt, Solinge, Gepken-Jager and Timmerman2005). Limitation of liability, as we understand it today, was irrelevant throughout its entire history.

In short, the two largest corporations of the 17th and 18th centuries, the EIC and the VOC, had thousands of shareholders each. Their paid-up capital was in the millions and their shares were traded successfully in high volumes on the Amsterdam and London Stock Exchanges – and all of this was achieved without limited liability being one of their key features.

2.2 The corporate economy begins to expand

The Bank of England was formed in 1694 by an Act of Parliament and a charter. Neither defined the Bank as having limited shareholder liability, yet this did not prevent the company from opening public subscriptions for a capital of £1,200,000. The offering was fully subscribed, with 1,520 shareholders contributing from £25 to £10,000 each (Bank of England, 2018). In 1720, two new marine insurance corporations were formed, the London Assurance Company and the Royal Exchange Assurance Corporation. The two were fully incorporated, but their shareholders did not have limited liability. From the 1760s, further insurance companies were established in the fire and life insurance subsectors. These were unincorporated and their shareholders were not shielded by a corporate entity.

The fastest-growing corporate sector of 18th-century Britain was canals. In 1766, bills were introduced in Parliament for two schemes, the Trent and Mersey Canal and the Staffordshire and Worcestershire Canal. Both were conceived as joint-stock corporations (Harris, Reference Harris2000). In the following decades, no fewer than 122 acts were passed for the incorporation of joint-stock companies for the construction of canals, and more than £17 million was raised by canal companies in the period before 1814. The Grand Junction Canal, the largest of all, required 18 times more capital than the Staffordshire and Worcestershire (£1,800,000 compared to £100,000) (Ward, Reference Ward1974: 29–30, 43–46).Footnote ⁷ Canal Acts typically incorporated the canals as legal entities but made no reference whatsoever to the liability of shareholders.Footnote ⁸ The financial model relied on equity investment, initially from local investors, provincial financial markets and exchanges, and, over time, from the national market and the London Stock Exchange.

The main instrument for raising additional capital downstream, which was often needed as the canal projects came in over budget, was by additional calls on the original shares. For this purpose, the nominal value of canal shares was high, while the initially called-up capital could be as little as 10% of the nominal capital. Bank loans were rare. Some of the companies borrowed from public investors by issuing bonds, sometimes convertible into stock and usually secured by the future toll income. Hence, in most scenarios, the shareholders bore the risks of loss of investment. Thus, once again, limited liability in the modern sense did not feature here. Even if debt finance had been used extensively, in the event of insolvency, creditors could have collected the uncalled capital from shareholders, which was often much greater than the called-up capital.

2.3 Jurists, courts and legal discourse around corporations

On the level of legal discourse and legal doctrine, historians have offered three indications that a doctrine of limited liability already existed in the late-18th century (DuBois, Reference DuBois1938). First, shareholders could not, in practice, be arrested for the debts of the company in which they held shares; thus, they could not be forced to pay its debts. Second, an Act of 1662 confirmed the exemption of shareholders of certain trade corporations from bankruptcy procedures by holding that they were not traders.Footnote ⁹ Third, it can be inferred from a judgment of 1671 that, when a corporation was not authorized to make further calls upon its members, the debts of the corporation could not be collected from them.Footnote ¹⁰ My interpretation of these three doctrinal pieces is that it was the legal personality attribute of the corporation that provided the rationale for them.

The first general statute of Parliament that dealt with joint-stock companies was the famous Bubble Act of 1720. This statute was enacted amid the wild days of the South Sea Bubble, in which numerous joint-stock companies were floated and speculative trade in company shares reached an all-time peak. This Act prohibited the formation of undertakings that pretended to act as joint-stock business corporations without obtaining a Charter or Act of Incorporation:

Note that it prohibited companies from pretending to have legal personality, joint-stock or transferable shares. But there is no mention of pretending to have limited liability for shareholders. Companies that bypassed state incorporation and turned directly to investors in the market did not view limited liability as a lucrative attribute that could attract investors to their companies.Footnote ¹² I take this striking absence as further support for my claim that, in 1720, limited liability as a separate attribute had not yet come into being.

In his comprehensive digest of mid-18th-century English law, Blackstone (Reference Blackstone1765: 462) wrote:

Limited liability is not mentioned by Blackstone – either by this name or even as an idea. It is also entirely absent from the most significant English book on corporations of the 18th century, Kyd's (Reference Kyd1793) Treatise on the Law of Corporations. Indeed, no scenarios of insolvency of corporation or claims of creditors are even discussed, so the absence is not just of the term or the concept, but of the entire circumstances in which equity holders and creditors are in conflict and the liability of shareholders can be defined.

The inexistence of the concept of limited liability in legal works published in Period I and its rise in the transition to Period II is illustrated in the following digital text analysis of the most exhaustive collection of legal treatises of American and British law, Gale's Making of Modern Law: Legal Treatises 1800–1926 database (Figure 1).

Source: Gale's The Making of Modern Law.

Figure 1. Occurrence of relevant terms in The Making of Modern Law database.

The terms ‘limited liability’ picked-up gradually in the first decades of the 19th century. The term ‘winding up’ that indicates the ability to dissolve insolvent companies also picked up simultaneously. The terms ‘company’ and ‘corporation’ that were used as control group show up frequently from the beginning of the century. A terminological control ‘shareholder liability’ was not used by contemporaries throughout the century.Footnote ¹³ As the coverage of the database begins only in 1800, a Google Ngram Viewer analysis (on all Google Books) was run for the period 1600–1900. No mention was found before 1773 and only a handful of mentions were found in books published in the period to 1820.

Limited liability in the modern sense kicks-in only when the corporation is not only insolvent but is also being dissolved because of this. In 16th-century British law, the Crown could create a corporation at will (through a charter) and could also revoke a charter at will, thereby terminating the corporation. But by the 17th century, the Crown could no longer arbitrarily dissolve corporations; and misuse of charters had to be reviewed by the court through the writs of quo warranto or scire facias before a charter could be rescinded. By the 18th century, the main means of liquidating a corporation was through special act of Parliament (Blackstone, Reference Blackstone1765), the effect of which was noted by Kyd (Reference Kyd1793: 516):

This statement comes nowhere close to reflecting limited liability. Insolvency is not mentioned as grounds for dissolution; the dissolution is not viewed as a collection measure on behalf of creditors; and the rationale for extinguishing the debts is based on the death of the corporate legal entity. Indeed, until the 19th century, bankruptcy procedures, both in Britain and the US, did not apply to corporations, only to individuals. Insolvent corporations could only be dealt with by Parliament. The affairs of the York Buildings Corporation exemplify the impossibility of swiftly dissolving an insolvent corporation. This entity was first incorporated by a letter patent issued by King Charles II in 1665 and was reincorporated after the Glorious Revolution in 1690. During the South Sea Bubble of 1720, it became involved in high-risk speculations, including lottery schemes. From then on, its financial state was shaky, at best; and its creditors, sensing it was insolvent, attempted to collect on their debts. But despite numerous litigations, it took an Act of Parliament to dissolve the company in 1829, over a century after the first signs of financial insolvency (Cummings, Reference Cummings1980; Murray, Reference Murray1883). At no stage during these century-long proceedings could the issue of personal liability of shareholders be discussed, because the corporation was not dissolved and its inability to pay its debts in full was not ascertained.

2.4 Legal personality is not limited liability

In sum, the corporation went through a fundamental organizational transformation around 1600. It became a for-profit entity with joint-stock capital, lock-in of capital and transferable shares. Within a few decades, business corporations with thousands of shareholders and an active market for corporate shares were thriving, and dominating long-distance Eurasian trade (Harris, Reference Harris2020). They did not need, and hence did not develop, owner shielding in the form of limitation on shareholders' liability. Nor did the Bank of England and the insurance corporations established during the post-1688 financial revolution. This was also the case with the canal corporations of the 18th century transport revolution.

As I have demonstrated, the period between 1600 and 1800 simply predates the emergence of the attribute of limited liability in the modern sense of owner shielding. Corporations then were not unlimited, but they were not limited either – hence, the continuum of limited to unlimited liability had not yet been formulated. One should not read my argument as claiming the attribute of limited liability existed but was not yet conceptualized in this period. What was absent was not conceptualization of limited liability, but the relevance of owner shielding from creditors. What may have misled some scholars are the manifestations of the separate legal personality of the corporation. The implication of the legal personality attribute of the corporation was that the corporation formed asset partitioning, in the sense of owning assets and governing them. But without debt finance, without a procedure for dissolving insolvent corporations, without the legal ability to determine whether shareholders would bear liability in insolvency, owner shielding was a non-issue and limited liability in the modern sense could not yet exist. In this era in which the legal personality attribute created asset partitioning, the level of liability of shareholders could not be tuned. The conflict of interest and related agency problem between creditors and equity investors was not yet created. The distinction between adjusting and non-adjusting creditors was not yet formed. One could not yet envision a liability bearing shareholder or a veil piercing that would subject shareholders to liability towards creditors.

3.1 Unincorporated and unlimited

The unincorporated company was a new organizational form created in Britain in the second half of the 18th century. This form was a bypass, created out of necessity, given the reluctance of the Crown and Parliament (and the opposition of vested interests) to incorporate new business corporations following the South Sea Bubble (Harris, Reference Harris2000). Unincorporated companies were formed by shrewd lawyers at the service of businesspersons in the realm of private law. These lawyers used legal devices, such as private contracts, the common law partnership and particularly the trust, to hold assets in a single pool, turning the company's directors into trustees that governed these assets (DuBois, Reference DuBois1938). Such companies acquired the financial attributes of joint-stock capital and transferable shares, but they could not enjoy the status of a separate legal personality (which could only be provided by the state in the realm of public law). These were the first large firms that did not use the corporate entity as a platform and consequently the first whose shareholders had direct personal liability with respect to third parties.

Because their shareholders were not shielded as a side effect of the separate legal personality of the entity, unincorporated insurance companies often limited the liability of shareholders towards insured persons contractually through clauses in the insurance policies. But shareholders were fully exposed to tort liabilities and to contractual liabilities that were not explicitly limited in written contracts. The trust device was also used to shield shareholders as trust beneficiaries from debts created by the trustees, but courts applied partnership law to unincorporated companies in order to overcome the shielding effect of the trust. Thus, shareholders in unincorporated companies were liable for the company's debts insofar as these were not limited in contracts with third parties. It became evident that, in the absence of supporting legislation, a joint-stock company could be formed with or without limited liability. Entrepreneurs could consider whether to invest the effort and money required to apply for full incorporation or settle for the unincorporated and unlimited form. Hence, limited liability had taken its first step towards being a separate and desirable attribute.

3.2 Drafting incorporation acts

Somewhat similarly, the shift from incorporation by charters to incorporation by private and special Acts of Parliament created a way of defining the limitation of shareholders' liability separately from legal personality. In the chartering era, the Crown used a template that included a standard legal personality clause (like the ‘body corporate and politic’ clause in the EIC charterFootnote ¹⁴) but made no reference to shareholder liability. When incorporation moved to Parliament, drafting was openly negotiated between entrepreneurs, parliamentary agents and MPs, meaning that limitation of liability could be included in incorporation acts. Explicit clauses dealing with shareholder liability gradually became more common.

Toward the end of the 18th century, limited shareholder liability became a declared motive of entrepreneurs who petitioned for incorporation, claiming it was essential for the success of their undertakings.Footnote ¹⁵ In his Trader's and Manufacturer's Compendium, Montefiore (Reference Montefiore1804) wrote that incorporation was indeed sought ‘principally for the purpose of exempting the shareholders from any responsibility as partners’ (p. 235). Thus, unlike the 18th century canal companies, by the early-19th century, some level of limitation of liability was an important motive for seeking incorporation.

The issue then filtered back to charters. In 1825, the Bubble Act of 1720 was repealed. This statutory repeal legalized, subject to the common law, the formation of unincorporated companies (Harris, Reference Harris1997). The second section of the 1825 Act gave the Crown discretion to grant charters without full limited liability.Footnote ¹⁶ This, according to Attorney-General Copley, the drafter of the repeal Act, would make law officers more willing to grant charters, and would encourage promoters to apply for charters rather than for Parliamentary Acts of incorporation.Footnote ¹⁷ We also learn in retrospect, from the fact that in 1825, a specific legislative arrangement had to be made to legalize the granting of charters without full limited liability, that limitation of liability had not been considered an attribute dealt-with by charters until then.

3.3 The general incorporation acts

General incorporation legislation created the opportunity (and need) to define the attributes of corporations, including their liability regime. Here we shift our attention to the US, because the earliest of these acts were drafted and passed there. New York was the first jurisdiction in the world to enact general incorporation law in 1811, initially only for manufacturing companies (Hilt, Reference Hilt2008). The Act Relative to Incorporation for Manufacturing Purposes of 1811 served as a model for similar acts enacted in New Jersey in 1816 and Connecticut in 1837 (Cadman, Reference Cadman1949; Kessler, Reference Kessler1948). Its terms included a low maximum capitalization of $100,000, a short life of only 20 years and a shareholders' liability provision (Section 7):

This Act has been positioned by some historians of US corporations as the first declaration of general limited liability, albeit for manufacturing companies (Wright, Reference Wright, Irwin and Sylla2010). But Howard (Reference Howard1938) convincingly shows that the case law that interpreted and applied it did not view it as creating full limited liability in the modern sense. In the cases of Slee v. Bloom (litigated between 1816 and 1823) and Briggs v. Penniman (litigated between 1819 and 1826), the judges were uncertain how to apply Section 7. Eventually, Judge Woodworth of New York's Court of Error held that: ‘Every stockholder, in a company of this description, incurs the risk of not only losing the amount of stock subscribed, but is also liable for an equal sum, provided the debts due and owing at the time of dissolution, are of such magnitude as to require it.’ This is what Howard (Reference Howard1938) justly terms ‘double liability’. Thus, while liability was perceived in the 1811 Act as an attribute separate from legal personality, a mandatory double liability regime was set.

In the first half of the 19th century, Massachusetts granted the highest number of corporate charters of any American state. The charter of the first Massachusetts manufacturing corporation, granted in 1789, contained no reference to individual liability. This was often the case in late 18th-century charters, but some provided that if a judgment against the corporation remained unresolved for 6 months after its dissolution, it should be satisfied out of the private property of the members. Several early 19th-century charters stipulated that if a judgment could not be collected from a corporation, individual shareholders were personally liable. Hence while Massachusetts refused to grant limited liability to any manufacturing corporation in the early-19th century, it did separate the attribute of shareholders' liability from that of the corporation's separate legal personality. This did not, however, create universal limited liability in the modern sense.

3.4 Corporate liquidation procedures

As we have seen, in the era of incorporation by charters and special Acts of Incorporation, dissolution could only be realized by a revocation of the charter or repeal of the act. Creditors could not take an insolvent corporation to court and liquidate it. In the UK, a procedure for the winding-up of companies was first made possible with the Joint Stock Companies Winding-Up Act 1844,Footnote ¹⁸ enacted immediately after the Joint Stock Companies Act 1844,Footnote ¹⁹ which had made available general incorporation by registration, thus also rendering a winding-up procedure necessary. The procedure applied to registered companies was similar to the bankruptcy procedure applied to individuals. Now, for the first time, with permission from the Chancery Court, creditors could dissolve insolvent companies to collect from their shareholders. The dissolution of registered companies was thus connected to the earlier doctrines applying to the dissolution of partnerships (Getzler and Macnair, Reference Getzler, Macnair, Brand, Costello and Osborough2005).

In the US, the Constitution made bankruptcy a federal issue and incorporation a state law issue. This fact made the US path to corporate liquidation different from the British one. The short-lived Bankruptcy Acts enacted by Congress in 1800 and 1841 did not apply to corporate entities, and there was significant opposition to such application on constitutional grounds. It was argued that, as corporations were created by states, they should also be liquidated (exclusively) by states. But the growing interstate nature of railroad corporations weakened this constitutional view. Corporations were permitted for the first time to take advantage of the 1867 Bankruptcy Act, which covered all legal persons, including corporate persons. Corporate insolvency became a major issue in the 1870s, during a wave of railroad companies' failures. The Bankruptcy Act provided corporate liquidation procedures via federal courts and state judiciary used equity receivership tools to avoid the liquidation of large concerns mid-way through construction of essential railroad lines (Skeel, Reference Skeel2001). This opening of the road for the dissolution of insolvent corporations brought the question of the shareholders' limited liability regime to the centre stage.

3.5 Debt finance

Debt was not a significant component of corporate finance in Period I and in the transition to Period II and thus cannot explain the transition. Furthermore, the late creation of corporate insolvency and bankruptcy procedures in Britain and the US can be explained by the low demand for such mechanisms in the corporate economy of Period I. Debt only gradually became a (secondary) source for financing the large infrastructure projects of the transport revolution. The loans for river and turnpike projects were more municipal than private and were secured by tolls. Canals and railroads were financed initially by locally floated equity investment. Often, shares in canals and even railways were traded locally on provincial stock exchanges (Baskin et al., Reference Baskin, Baskin and Miranti1999; Evans, Reference Evans1936; Ward, Reference Ward1974). Debt finance was only sought when investment was required from non-locals at an informational disadvantage, such as London investors for canals in Yorkshire, or for US railroads (Chandler, Reference Chandler1954). To reduce risks, the debt was raised after equity had been raised locally in the region served by the route in question, and once the real estate of the project could be offered as a security.

The financial sector was the exception in the sense that both banks and insurance firms had creditors as part of their normal course of business, policy holders and depositors. Initially, insurance was organized in the Lloyds marketplace in which underwriters assumed personal liability and in unincorporated companies with unlimited liability. Banking was mostly organized as family banks and partnerships with personal liability. The exposure to personal liability placed the financial sector in the forefront in Period II, earlier in the US in which the financial sector switched to the corporate form around the turn of the19th century and later in Britain beginning in the second quarter of the century (Kingston, Reference Kingston2007).

The corporate bond market emerged only in Period II. The total number of corporate debentures listed in 1860 on the UK Stock Exchanges was just nine, all in railway ventures, and their nominal par value was just £5.8 million (Coyle and Turner, Reference Coyle and Turner2013). The corporate bond market grew slowly, and by the 1880s was still composed mostly of railway bonds. The real par value of bonds and the par value of bonds per GDP reached a peak in 1909. The small (compared to equity) and slow-growing (in real terms and compared to GDP) corporate bond market in such a late period simply cannot explain the transition from Period I to Period II and the separation of the limited liability attribute. The separation of the liability attribute is not a story of shareholders calling for a limitation of their liability in the face of growing need for debt finance. The intensification of a conflict between creditors and equity investors did not present itself in Britain or the US before the late 19th century. It might explain the second transition but not the first.

4.1 Unlimited liability jurisdictions

As we have seen, the development of the conceptualization of limited shareholder liability as a distinct attribute emerged out of the conceptualization of unlimited liability. State jurisdictions realized they could select where to position themselves on the liability continuum, and some opted for unlimited liability.

In America, Massachusetts, unlike New York, opted to position itself at the unlimited liability end. Did this negatively affect economic development, as many have suggested? The short answer is ‘no’. Its unlimited liability regime did not prevent Massachusetts from being the most advanced corporate economy in the US in terms of the number of charters issued to manufacturing. Following a wave of business failures in 1829, the Massachusetts Act Defining the General Powers and Duties of Manufacturing Corporations was passed in 1830. The law imposed a charter clause that limited the liability of shareholders to the unpaid balance on the shares (Hilt, Reference Hilt, Collins and Margo2016). A century later, no less a corporate law authority than Dodd (Reference Dodd1948) observed that the textile industry in New York, a jurisdiction with general limited liability (double liability) for manufacturing companies since 1811, grew no faster or larger than in Massachusetts, where it was introduced a full two decades later. In other words, the leading industry of the time (textiles) in a leading industrial state (Massachusetts) seems not to have been adversely affected by holding on to a regime of unlimited liability during this crucial pre-1830 industrialization phase. This suggests that limitation of shareholders' liability was not a major factor, given the financial structure of the companies and the available winding-up procedures. Massachusetts and other New England states gradually adopted and extended limited liability laws. But there is no evidence, from the current state of research, that they did so because unlimited liability had retarded their economic growth.

4.2 General limited liability

In Britain, the Limited Liability Act 1855 did not impose limited liability in the modern sense. It did not even put forward a default limited liability rule. It allowed incorporators to decide how much shareholders would personally owe creditors by determining how much of the nominal capital was not to be paid up initially. No single liability model derived from this Act. The Companies Act 1862 similarly stated:

The 1862 Act did not set a default rule that incorporators could opt out of either. Liability had to be drafted by the incorporators into the Memorandum. This enabled the incorporators – and later, the directors – to design the desired level of liability by determining the level of nominal capital and varying the percentage of that capital that was actually paid up. This drafting also opened the way for the use of reserve liability, to which we will turn later.

4.3 Liability for unpaid balance

The common practice of issuing only partly paid company shares, which began with canals and railway companies, continued also after the enactment of the aforementioned 1855 and 1862 Acts. It was initially a response to gradual cash-flow needs of construction projects, but was eventually also used as security for corporate creditors. While in some canal and railway companies only 10–20% of the capital was initially called up, the gap narrowed as the 19th century progressed and paid-up capital reached 50–70% by the third quarter of the century (Acheson et al., Reference Acheson, Turner and Ye2012; Jefferys, Reference Jefferys1946; Shannon, Reference Shannon1933). In Harris (Reference Harris2013), I concluded that by 1897, the average paid-up capital of our sample was 70% (Harris, Reference Harris2013: 34–35).Footnote ²¹ It follows that, as late as the turn of the 20th century, the practice of issuing shares that were not paid in full was still common in Britain. As long as this practice continued, limited liability in the modern sense was therefore not yet the dominant practice.

4.4 Divergent liability regimes in banking

In Britain, the divergent continuum of liability regimes was manifested in banking more than in any other sector. Acheson et al. (Reference Acheson, Hickson and Turner2010) provide the most comprehensive survey of liability regimes in the banking sector in England, Scotland and Ireland. A handful of banks were state-chartered: Bank of England (est. 1694), Bank of Ireland (1783), and the three Scottish banks – Bank of Scotland (1695), Royal Bank of Scotland (1727) and British Linen Bank (1746). But these banks were not explicitly chartered with the limited liability attribute, because their establishment predated its separation from the personality attribute. Up until the mid-1820s, these banks enjoyed the privilege of being the only note-issuing banks in England and Ireland to be organized as corporations. Other banks were excluded from this organizational form and were not permitted to have more than six partners. In this way, they were also excluded from the unincorporated joint-stock form.

Parliament permitted the incorporation of unlimited liability joint-stock co-partnership banks in Ireland by means of the Banking Copartnership Regulation Act (1825), and in England by the Banking Copartnerships Act (1826). An 1826 statute legalized the unlimited joint-stock banks in Scotland.Footnote ²² The liability of shareholders in such unlimited banks was joint and several. I interpret these Acts as a manifestation of the separation between the personality and liability attributes of the corporation. They also reveal that the older chartered banks were at some point re-conceptualized as having a shareholder-limited liability attribute separately from their legal personality attribute. For example, by 1825, the Bank of England (which, as we have seen, was not established with limited liability in mind) was understood to have this attribute. By 1826, unlimited liability joint-stock banking had become available throughout the UK. Banks were excluded from the initial Joint Stock Companies Act 1844 and the Limited Liability Act of 1855, but by 1858, they were allowed to incorporate by registration with limited liability. This enabling legislation abolished the distinction between chartered banks and registered limited-liability banks.

The study conducted by Acheson et al. (Reference Acheson, Hickson and Turner2010) illustrates my contention regarding the divergent liability continuum by showing that the banking sector in Britain did not converge into one single liability regime in the 19th century, but rather came to enjoy a choice of four regimes (see Table 1):

Table 1. Liability regimes in UK banking, 1849–1889

Source: Adapted from Acheson et al. (Reference Acheson, Hickson and Turner2010: 250).

Hence, more than one liability option existed, and more than one option was selected during each of the three time-points surveyed by Acheson et al. (Reference Acheson, Hickson and Turner2010). We can observe that absolute unlimited liability had disappeared by 1889, but there is no sign of convergence to limited liability in the modern sense during the period covered in their study. Questions for historical empirical researchers remain. For now, we can conclude that the banking sector in England, Scotland and Ireland seems to have functioned for a significant time without limited liability in the modern sense.

4.5 Reserve liability

The fourth and last type of bank liability to appear in UK banking in the 1880s was reserve liability. Enacted in the aftermath of the City of Glasgow Bank's collapse, the Companies Act 1879 intended to facilitate the conversion of existing unlimited joint-stock banks to limited liability companies. The main innovation was the creation of ‘reserve liability’ – a sum of capital that could be called only upon insolvency and liquidation of the company. Unlike uncalled capital, which could be used by the directors of solvent companies to complete projects, reserved capital was only used to protect creditors. Again, reserve liability offered a continuum rather than a single liability regime. Many banks selected double or triple liability, and others had different multiples of nominal par-value of the shares.

Reserve liability was also common in the US. Some states imposed double liability on all corporate shareholders, while many imposed it particularly on banks. This requirement was initially imposed by clauses in specific charters and then in state general statutes, and even state constitutions. In 1863, the Federal Government entered the field of banking regulation with the National Banking Act, which stipulated that ‘each shareholder shall be liable to the amount of the par value of the shares held by him, in addition to the amount invested in such shares’. Most states implemented the federal legislation in their own laws, making shareholders in banks bear proportionate double liability (Blumberg, Reference Blumberg1985).

Proponents of the evolutionary-convergence-to-efficiency hypothesis may view Period II as a brief and unremarkable stage during which inefficient liability configurations were rejected in favour of the most efficient organizational regime – limited liability in the modern sense. Yet this is an instructive period, in which for more than a century different configurations of shareholder liability along a continuum were selected by different jurisdictions, sectors and companies, some of which performed very well under regimes that were different from limited liability in the modern sense. A period of growing economies, expanding sectors and successful enterprises at the height of Western industrialization cannot be dismissed as a mere hiatus. We can gain important historical and policy insights from the liability experimentations made in this period.