The lightship in economics

Beginning with the 5th edition of Economics: An Introductory Analysis, Samuelson (1961, pp. 192–193) used the lighthouse as an example of pure public good. Even after the publication of Coase (1974), however, Samuelson continued to use the lighthouse as an example of a good that required public provision. The lighthouse still appears as an example of a public good in the 19^th edition of his textbook (see Samuelson and Nordhaus 2009, p. 37).

As Brubaker (1975, p. 151) puts it, “An essential and neglected fact is that no matter how costly individual exclusion may be after the creation of a collective good, prior to consummation of a contract exclusion of the group, and consequently also of the individual, is always very easily accomplished. The producer simply holds no inventories. He works strictly on the basis of pre-paid orders. Lacking a firm contract with the community there will be no collective good for any of its members”. Leeson (2007b) employs a similar self-enforcing mechanism to explain how the extension of credit by producers increased the costs of theft and raised the benefits of trade among middlemen in late precolonial Africa. In terms with which Brubaker might agree, credit is a pre-contractual mechanism of excludability because credit allowed producers to trade with goods that did not yet exist, minimizing the benefits of theft, and incentivizing future repeated dealings. Thus, credit was a pre-contractual mechanism to produce a public good privately, in this case self-governance, namely by eliciting a demand for trade among the middlemen. Analogous to the case being made herein, pre-payments for the construction of lightships allowed its producers to minimize free riding and incentivized payers to reveal their demands for lightship services.

In 1836, an act of Parliament nationalized all remaining privately owned lighthouses in England (Hardy 1895, p. 41); see also Coase (1974).

According to Clarke (2016, p. 32), pilotage refers to “the art of taking a vessel from one place to another in sight of land and providing ships with safe passage onto rivers and harbours or through dangerous waters”.

Historians argue that since Trinity House was responsible for mariners’ pensions, it was reluctant to bear too much risk. As such, its monopoly power allowed it to shift risks to private parties (Lindberg 2013, p. 546).

The 1834 report of the House Select Committee (1834, pp. XXXVII–XXXVIII) reported 55 lighthouses and floating lights operated by Trinity House for 1832 and 14 privately operated lighthouses in England and Wales. Taylor (2001, p. 758) writes that by 1835, the number of private lighthouses had fallen to 10. Ireland and Scotland had their own lighthouse authorities (the Commissioners of Irish Lights and the Commissioners of the Northern Lighthouses), which were separate from Trinity House’s operations in England and Wales.

We distinguish between for-profit, private lighthouses and non-profit, private lighthouses specifically to focus our discussion on the former, rather than the latter. Although lighthouses were provided by hermits and religious orders on a voluntary basis, they were operated on a non-profit basis (see Van Zandt 1993, p. 59). Block and Barnett (2009) also argue that purely private production was possible for charitable purposes. Because the main dispute in the literature is on the profitability of lighthouses in the absence of state enforcement of the collection of light dues, we do not focus on non-profit lighthouses.

Our argument also builds on a theme that Coase already had developed in his 1961 paper, “The British Post Office and the Messenger Companies,” which illustrates how, in an attempt to preserve its monopoly, the Post Office tried to obstruct the entry of private entrepreneurs into the provision of postal services.

We use the term “rivalry” here to mean mutual exclusivity of consumption and/or production; we do not use it interchangeably with “competition” or “contestability”.

In any case, patents for lighthouse operation came with a radius of exclusive operation within which no competitor could install itself.

In navigational distance, the Nore lightship was 48 miles away from London (Chandler, Diston and Adams 1788, p. 113). The route between the Nore and the Port of London took roughly three hours (House of Commons 1796, p. 46).

Modern lightships under the Trinity House system are no longer manned. According to Clarke (2016, p. 150), the majority of manned lightships was decommissioned in the 1970s and 1980s. The last manned lightship in England, the Inner Dowsing, was decommissioned in 1991.

Clarke (2016, p. 38) reports that the origins of modern lightships can be traced back to ancient Roman galleys, known as liburnae, used by the Roman navy for raids, but also provided lighted beacons to deter pirates. As he states further, the “vessels carried on their mastheads iron baskets in which fire was built serving as a signal when a friendly vessel was sighted”. However, their functions as lightships were limited because ancient sailors tried not to sail at night. It is unclear whether Avery or Hamblin were aware of their historical provenance.

Lightships were very popular for the navigation of Canada’s Saint Lawrence seaway (Cloutier and Charest 2016) and the American eastern coastline (Webster 2000). This is a telling element given that frontier economies like Canada and the United States would be have been blank slates in terms of technologies available to adopt, which may explain their greater relative popularity in North America.

Coase (1974, p. 366) excluded “floating lights” from his presentation of the numbers.

By 1834, all 13 lightships were owned by Trinity House (House of Commons 1834, p. 344).

Most of the secondary sources that discuss the event rely nearly exclusively on the information provided by John Whormby (1746 [1861]). Whormby was the clerk of Trinity House who wrote a history of the Corporation in which he gave his retelling of the Nore lightship episode as he lived it. Most of it is condescending towards Hamblin and Avery.

Some secondary sources mention 1732 (e.g., Adams 1870, p. 254; Gattie 1890, p. 156) and some even place it in 1733 (Cotton 1818, p. 77), while others are vague about the date (Adams and Woodman 2013, p. 87). But the advertisements published in the Daily Courant (October 12, 1731), the Universal Spectator and Weekly Journal (October 16, 1731), the Daily Post (October 12, 1731), and the London Journal (September 18, 1731; October 16, 1731) all make it clear that the ship became operational in 1731. Hamblin mentions in an ad submitted on October 10 that he had “this morning moor[e]d a complete ve[ss]el of about 100 tons” (London Journal, October 16, 1731). The only secondary source that gives 1731 as the starting date is Stevenson (1959, p. 139) for whom we have confirmed every assertion. A July 22nd, 1731, article in the Grubb Street Journal mentions a ship placed at the buoy at the Nore on order of Trinity House (this is echoed in the Caledonian Mercury of July 26, 1731). However, the article is unclear about what was occurring, but we know that, given its virulent reaction (see next section) to the Nore lightship, Trinity House could not have ordered the placing of the first 19-ton lightship. The possibility exists of a journalistic error. However, it is true that notice of a first ship was placed by Hamblin for advertising purposes. Secondary sources confirm it. Robert Peirce Cruden (1843, p. 412) pointed out that on the August 9th 1731 (a date that contradicts the July Grubb Street Journal ad), Hamblin moored a 19-ton vessel at the Nore (called the Experiment), which he had replaced by an additional ship (called the Good Intention) of 100 tons on October 10th 1731. The mooring of the Experiment apears to have been meant as a marketing ploy. The elements contained in Cruden’s account are confirmed by the September and October ads placed in the Daily Post, Daily Courant, Universal Spectator and Grubb Street Journal. As such, the 19-ton Experiment operated from either late July 1731 to early October 1731 or from early August 1731 to early October 1731.

This is confirmed by the clerk of Trinity House, John Whormby, in his 1746 history of the corporation. He pointed out that Trinity House apprehended, “very justly” in in his opinion, that Avery would go “round the kingdom with lights of that kind, to the danger and confusion of navigation, as well as to the prejudice of fixed lights” (Whormby 1746 [1861], pp. 135–146; emphasis added).

The Well refers to Dudgeon Shoals off the coast from Cromer in East Anglia where a lighthouse already existed (Stevenson 1959, p. 136).

Few sources provide details about how that was the case. The one exception is a laudatory article in the Newcastle Courant on August 21, 1731. The article refers clearly to the first, smaller, lightship as a “great advantage to Navigation in general, as well a Means to preserve the Lives and Fortune of his Majesty’s Trading and Maritime Subjects from the Dangers of the Nore Sand”. However, that article was a copy of the August 11, 1731 article published in London’s Daily Courant.

The voluntary payment component of the lightship at the Nore often is noted in passing (Cruden 1843, p. 412; Pask 1884, pp. 604–605). In fact, Trinity House complained explicitly about the voluntary nature of payments through subscription books.

The official historians of Trinity House refer vaguely to expenses of £2,000 on the “project” (Adams and Woodman 2013, p. 87). Many sources mention “fitting out”, but they all represent secondary sources. As we document in Appendix, it is likely that the estimates combine operating costs (i.e., variable costs) and fixed (i.e.,construction) costs. If the secondary sources are correct that the figure of £2,000 concerns only the construction and fitting-out of the Experiment and Good Intention, then the operating cost estimate derived using the number of workers and gallons of fuel consumed is the most accurate (see Appendix).

They also were more expensive to build: Chambers’s Encyclopaedia (1891, p. 623) placed the construction cost of a lighthouse between £5,000 and £10,000, while a lightship cost £9,000 during the late 19th century.

Revenues from smaller ships are omitted. Moreover, we do not include revenues from the public subscriptions, which were collected ex ante.

Including the lightship at the Nore, which had, by that time, passed into the hands of Trinity House. The rate schedules probably were vestiges of Avery and Hamblin’s experiment because the schedules were taken from the last establishment that practiced price discrimination in 1832 (House of Commons 1834, p. 341). The other lighthouses were the Castor and Lowestoft lighthouses (which charged a flat fee of four pence for all ships and then an additional 16 pence for every 100 tons) and the Winterton lighthouse (which charged 6 pence for ships under 100 tons and 12 pence for all ships above 100 tons) (Anonymous 1768, p. 21–22). All other lights had a flat rate per ton or per ship.

Lai et al. (2008a, b) are the only contributors to the lighthouse literature who emphasize the importance of the point regarding price discrimination, arguing that it makes private provision more viable–an argument that has not been made elsewhere. However, Koyama (2012) advanced the same argument for the same period of British history, but for prosecution associations (private conflict adjudicators) because it elicited greater participation by poorer individuals. What is more important, is that it is worth emphasizing that free riding can be endogenous. If free riding was to drive the enterprise into losses, the marginal user might free ride less so as to keep the operation afloat. A system of price discrimination would have been able to cater to that marginal user and thus permit sustainable production (we are thankful to Robert Whaples for this insight).

Omitting colliers, which were major sources of potential revenues.

The secondary sources tend to mention that the patent specified ships’ ability to distinguish light sources based on their colors. That is not the case. Hamblin’s petition makes no clear mention of color (National Archives 1730, SP 36/17/81). It merely speaks of the ability to distinguish one light from another. However, the lightship–being a novelty–would have distinguished itself from lighthouses, making it easier for Avery and Hamblin to differentiate their product and collect fees from users. Moreover, the vague wording of the petition, by not mentioning the lightship, meant that the innovation could be copied by others. Patents define products, not markets.

From the Attorney General’s report to the King regarding the petition made by Trinity House to revoke the patent.

More details on the legal technicalities raised could not be found. However, by virtue of the petition of Trinity House (National Archives 1731, PC 1/5/3), combined with the report of the Attorney General and Solicitor General (National Archives 1732a, PC 1/5/4), we can see that revocation was argued on two bases. The first was that the invention was not new, the second that public subscriptions with more than five individuals was not permissible under the patent. The former seems to have been the basis that the special committee of the Privy Council (National Archives 1732b, PC 1/5/5) used to justify the revocation. It made no mention of the latter point to justify revocation.

This supports a point made by Mancur Olson (1965 [1971, p. 33, emphasis in original) in The Logic of Collective Action, that “if at any level of purchase of the collective good, the gain to the group exceeds the gain to any individual, then there is a presumption that the collective good will be provided, for then the gain to the individual exceeds the total cost of providing the collective good to the group”.

A similar mechanism existed in the United States, where lotteries financed some of the lighthouses of the colonial era (Dolin 2016, p. 60) even if few private lighthouses existed at the beginning of the Republic. Moreover, given that the subscriptions were solicited at coffeehouses, it also worth pointing out that, according to Stringham (2015) and Kingston (2007, 2014), coffeehouses often used ostracism and social norms to enforce complex contractual arrangements that otherwise may have given room for opportunistic behavior. Members who did not honor their arrangements would be blacklisted and, through multilateral punishments, other club members would shun relations with them. Most of the coffeehouses mentioned by Hamblin in his advertisement were merchant houses that could have acted to ostracize non-payers. They also could offer a reputational premium through the discipline of continuous dealings. For example, Lloyds coffeehouse, where individuals involved in the insurance market met, was known for being “a hub for information about… the reputations of market participants” (Kingston 2007, p. 380; see also Kingston 2014).

As we document in Appendix, with such economies and with the relaxing of some of upwardly biased operating costs estimates, it is impossible for potential revenues to have been less than costs.

The article identified Avery as the proprietor.

Hardy (1895, p. 32) argues that there were similar proposals as early as 1623, but he provides few details on these proposals.

Later, the Attorney General said that he thought that Hamblin and Avery merely intended to diversiy the colors of the lights by making them burn blue or green (Adams and Woodsman 2013, p. 87).

Coyne and Leeson (2004, p. 237) have coined the term “evasive entrepreneurship” to describe such activity. Avery and Hamblin, while engaging in productive entrepreneurship, also were expending effort and resources to evade being excluded from the coastal lighting market by Trinity House.

Stevenson (1959, p. 139) makes it clear that the patent they obtained was unusual. The normal patent empowered its holder to collect dues for a term of years from every vessel. The type of patent received by Avery and Hamblin “allowed an individual to exploit an invention for 14 years” (ibid.). One secondary source asserts that once they had obtained their patent, Avery and Hamblin petitioned for the “power of levying payment” (Anonymous 1865, p. 624). However, no other source mentions that aspect of the petition. In fact, other sources highlight the “voluntary” nature of the payment system (Cruden 1843, p. 412; Pask 1884, pp. 604–605). Moreover, one of the key complaints of Trinity House in its November 1731 petition is the system of public subscriptions upon which the financing of the Nore lightship rested (National Archives. PC 1/5/3 1731).

The accounts of the reactions can be found in multiple sources. They all offer the same timeline and storyline with varying degrees of detail (see Anonymous 1865, p. 624; Renard 1867, pp. 193–195; Adams 1870, p. 254; Gattie 1890, pp. 156–159; Hardy 1895, pp. 72–75; Stevenson 1959, pp. 138–140).

Adams and Woodman (2013, p. 87) mention that Hamblin attempted to open another lightship (date undisclosed) with a Yarmouth pilot at the Cockle Gat, but that Trinity House turned them down, preferring the existing lighthouse.

This line of reasoning explains why Avery’s negotiation of a license with Trinity House to keep the lightship operating resulted in private operators of lighthouses joining hands with Trinity House to protest the lightship’s patent.

Krause (2015) provides an example lighting provision with buoys in the case of a present-day suburb of Buenos Aires. While we have focused on the case of the lightship as an example, buoys and other seamarks should enter future studies as well.

Bertrand’s argument also aligns with the broader portrait of rent seeking between different interest groups that were competing for the rents from the British Crown (see Zahedieh 2010). As such, lighthouse operators can be seen as one of the many groups participating in that wider historical struggle.

Of that amount, £11.25 came from superannuated pensions to seamen–an expense that Avery and Hamblin would not have needed to cover (House of Commons 1834, p. 346). To remain conservative, we did not adjust for this minor issue.

The same numbers were reported in a House of Commons document (House of Commons 1816, Vol XIX, pp. 149–173). That probably is the source of Cotton’s numbers.

The lowest level of expenses was in 1813, at £479.8 and the highest was in in 1808, at £840.6.

The lowest level of expenses after removing collection costs was in 1813, at £357.9 and the highest was in in 1808, at £694.3.

These numbers refer to large vessels. Willan (1934 [1964], pp. 103–105) points to lower costs, but the numbers therein apply to barges, lighters and trows.

Ralph Davis (1962, pp. 373, 378) reports a wide array of prices–few of which include the cost of fitting-out ships with masts, spars, ropes and the critical pieces of equipment for lightships, such as anchors and lanterns. He reports only construction costs. Armstrong (1991, p. 80) argues that fitting out a ship inflated total costs by somewhere between 50% and 100%. Those percentages probably are too low for a lightship, given the special equipment it requires. The lanterns were not inexpensive and the anchors were specifically designed (mushroom anchors) to limit drifting. As such, we used the high bound of the fitting-out cost when using Craig's (1971, p. 129) estimate. 

The first sailed at 19 tons and acted as an experiment (the ship was christened Experiment), while the latter weighed at 100 tons (Cruden 1843, p. 412; Grubb Street Journal July 22nd 1731).

On another page of the House report (1834, p. 340), oil consumption is given at 193 gallons rather than 199 gallons. The difference would reduce operating costs by £2.4.

The quantity of oil to be burned would be roughly constant.

Seamen were paid the same wages as construction laborers; the mate was paid the wage of a skilled craftsmen, while the master was paid twice the wage of a skilled craftsmen. Converted to 1832 pounds, those wages are higher than the ones reported by the House of Commons in 1834. As such, we believe that our estimate is upwardly biased. We also should not assume size of the crew in 1832 was the same as in 1731. Hiring only a master and two seamen brings the cost to £256.48.

Wherries are excluded.