History, Institutions and Economics

I have read and enjoyed several series of novels set in the British navy during the Napoleonic wars, most recently one by Naomi Novik that departs a little further from history than its predecessors by providing the British and their enemies with dragons. The internal structure and the associated rules and customs of the navy seem very strange to a modern eye, yet it was a strikingly successful institution.

One feature likely to catch an economist's eye was prize money. If a naval vessel captured a legitimate prize, an enemy warship or merchantman, and brought it back to port, the vessel and its contents were sold and the money distributed among those responsible. One large chunk went to the captain, another was distributed among his officers, a third among the crew, a fourth to the admiral under whose orders he was operating.

Another feature of the system was the role of patronage, political influence both within the navy and outside it in the career of an officer, especially a young officer. The critical step of promotion from lieutenant to captain depended in part on performance, in particular on the opinion of the captain under whom a lieutenant was serving. But it depended also on things that seem, to us, irrelevant.

One of Patrick O'Brien's novels contains a conversation between Maturin, one of his protagonists, and a friend, a young officer of aristocratic birth. The officer has been having an affair with the separated wife of a high naval official and wants to know whether he should live openly with her. Maturin's response is that, moral issues aside, it might be imprudent for him to offend a powerful official and so risk his future career. His friend replies that he has considered that matter but his family controls a significant number of seats in both houses of parliament and he thinks their influence will be sufficient to balance that of the man he will be offending.

Neither party sees anything strange in either the assumption that giving personal offense to someone within the bureaucracy will make it harder for a competent officer to be promoted or that having a politically influential family will make it easier; that is just part of how the system works. It was a system that produced extraordinarily successful results, a navy that, from the late 18^th century to the early 20^th won almost every ship to ship or fleet to fleet battle it fought at anything close to even odds.¹

A third feature was the seniority system. Once a lieutenant was promoted to captain, his future rank depended only on how long he survived. His name was on the list of captains, the list was ordered by strict seniority, and the next captain to be promoted to admiral would be the one at the top of the list. When two or more captains were working together it was the senior who commanded. That provided an unambiguous rule for allocating command, since every captain knew where he was on the list and knew, or could readily find out, where any other captain was. But it was a rule that had nothing to do with the relative competence of two officers of the same nominal rank.

Promotion beyond captain was entirely determined by seniority; what the officer got to do with his rank was not. An insufficiently competent captain who made it to admiral would end up as an admiral of the yellow, an admiral without a fleet, effectively retired on half pay. A sufficiently competent captain could be assigned particularly important duties, including the command of a group of ships with the temporary position of commodore — provided none of the other captains in the squadron was senior to him. A sufficiently incompetent captain could end up without a ship, on half pay with no chance of prize money. In peacetime, when there was no shortage of competent captains, a minor failing might do it.

The Institutional Revolution by Douglas W. Allen looks at the institutions of the British navy as part of a bigger puzzle about institutions of the premodern world. The central features, as he describes them, were patronage and purchase, the two methods by which many civil and some military offices were allocated. He explains them as premodern solutions to the agency problem, the problem faced by a principal in controlling his agent. If the king needs taxes collected, how does he make sure that all the money goes to him instead of into the pocket of the tax collector? If he wants to win battles, how does he make sure that the resources devoted to that purpose don’t get siphoned off by the chain of people between him and the battlefield? How does he make sure that the officer commanding an army or ship does his best, even at serious risk of getting killed in the process?

As Allen describes it, put in my words not his, there are three different solutions to that problem.

1. Employment: The agent is paid, either by time or output. If he does not do what the principal wants he is fired. This is the standard modern solution in both private and government employment.

2. Property: The agent is given a property right in his position, structured in such a way that the revenue he gets depends, at least in part, on how well he achieves the principal’s goals. The property right may or may not be transferable.

3. Patronage: The agent is someone the principal trusts to act in his interest. That trust may be lost if it becomes clear that the agent does not deserve it.

Each solution has requirements that can be satisfied better in some contexts than in others, hence which gets used will vary both from one activity to another and over time. The first, employment, depends on the principal’s ability to monitor the agent well enough to know how many hours he is working, whether he is doing what the principal wants him to do, whether success or failure is due to the agent or other causes. The second, property, requires the principal to find some bundle of rights to give the agent that makes it in his interest to do what his principal wants him to do. The third, patronage, requires a potential agent who the patron trusts, either for reasons such as kinship or because losing that trust will be costly enough so that the agent will not be tempted to serve his own interest at the expense of his patron’s.

Allen argues that, prior to about the late 18^th century, there were serious problems with using the first mechanism, especially for the military and the civil service. Nature played a large role in most productive processes, making it hard to tell how much of the outcome was due to what the agent did, how much to natural conditions. There were few accurate ways of measuring time or coordinating activities. The lawyer who was supposed to spend two hours negotiating a contract for you might spend the time waiting for the other party’s lawyer, delayed by a bad road, lame horse or broken down coach. Hence pre-modern institutions relied much more heavily on the second and third.

Consider the case of the pre-modern British navy. Prize money was a property solution. The admiralty wanted captains to have an incentive to capture enemy merchant ships, defeat and capture enemy warships, even at risk to their lives. Most of the relevant decisions were made by the captain, so he got the largest part of the reward, but other people, including the admiral whose orders determined what opportunities the captain had to earn prize money, got some of it. A pattern that shows up in the novels, and presumably in the real history, is an admiral who puts an unusually competent and aggressive captain in places where he is likely to encounter enemy warships not because he likes the captain but because he hopes to profit from successful encounters.

Allen argues that prize money was an imperfect property solution because capturing a warship was much riskier, more likely to get the captain killed, than capturing a merchant ship, but prize money was awarded for both. One puzzle he does not consider is why the navy did not solve the problem of misaligned incentives by lowering the prize money awarded for merchant ships or raising it for warships, which should have been easy enough to do.

Allen offers the imperfect alignment of incentives, such as the temptation for a captain in a fleet action to hang back and let other ships and their captains take the risk, as a reason why the property solution had to be supplemented with elements of the other two systems. The admiralty had detailed information on what a captain did through a system of three different logs, one by the captain, one by his first lieutenant, one by the sailing master, the ship’s senior warrant officer. A captain whose career showed him to be incompetent or too inclined to go after merchant ships and avoid warships might end up spending the rest of his career on shore with no ship, hence no opportunity for prize money. A captain who declined a clear opportunity for combat with a ship of the same class was subject to trial by court martial; one admiral ended up convicted and executed for failing to pursue the enemy fleet after an engagement.

The property system was carried further in the army, where officer positions were purchased and could be resold. The price was much higher than justified by the salary, presumably because of the opportunity for loot and, in the case of an important victory, rewards from a grateful crown. A successful officer such as Marlborough could end up wealthy.

One argument for allowing someone to buy a position as an officer was that the better he did the job the more money he would make, hence the more competent bidder would be willing to offer a higher price. It was an imperfect system since it was based on expectation, not performance, and depended on the candidate having access to enough money to buy the position. But since the pattern was to start with the lowest officer’s rank and then move up step by step, all but the first transaction would be made after the bidder had some evidence of how well he could expect to do. Permitting resale gave an officer who discovered that he was not doing as well as he expected an incentive to give up the position and, since what was sold was a position in a specific unit, gave an officer who planned to retire an incentive to maintain the quality of the unit he commanded..

Many offices in the bureaucracy whose income depended, often through what we would today view as bribery, on the job the office holder did, were also treated as private property. A version of this approach common in classical antiquity was tax farming. The Emperor sells the right to collect taxes from a province to the highest bidder; it is then up to him to collect the taxes. If the tax farmer has only bought the right for one year he has no incentive to do his tax collecting in ways that maintain the ability of the province to pay taxes in future years, just as someone who rents a house for one year has no incentive to make sure it is in good condition at the year’s end. In both cases there may be ways of reducing the problem — a deposit for the renter that forfeits if the house needs repair, an imperial officer with the job of collecting complaints of extortion by the tax farmer. And in both cases it may, despite problems, be the least bad way of achieving the objective. Tax farming was used in England but abandoned earlier than other elements of the purchase system.

The other solution to the principle/agent problem in the pre-modern system as Allen describes it was patronage. The king appoints a high officer he trusts, with the implicit threat that if things go badly enough the officer’s position, perhaps his life, are at risk. The king cannot judge the officer by how well things go day to day or even month by month, but if the navy consistently loses battles instead of winning them, especially if the reason is that the ships are in poor condition and under-crewed because the money that was supposed to pay for those things went somewhere else, the official in charge may lose his position, possibly his head. The person appointed is chosen not mainly for his competence but because he is someone the king trusts. The high official cannot do everything himself so he again appoints subordinates, chosen to be people he trusts. As do they.

Allen argues that the system of patronage explains the existence, the nature and the eventual end of the aristocracy. The job of the medieval aristocracy had been largely military; it was, as someone put it, an army quartered on the people. That changed in the 16^th century. From then until the early 19^th century the aristocracy, much of it new members, had as its primary and very lucrative function running the government, largely through a system of patronage. It had become a bureaucracy quartered on the people.

Someone who wanted patronage offices had to make a variety of costly investments to put him in the pool of potential candidates, investments that would lose much of their value if he lost the trust of those above them. That included investments in social capital, an education and a network of relationships that made him part of the upper class and would be worthless if that class rejected him. It included physical capital, a country estate scaled to host large social events for fellow aristocrats, provided with extensive lands designed for show not agricultural output. It included the willingness to prove commitment to aristocratic values of honor by risking his life in a duel in response to being accused of a failure of honor. It included the would-be aristocrat abandoning the commercial activities that had made him the money needed to become an aristocrat, hence having no fallback position if he failed to maintain the trust of his superiors.

All of this Allen views as a bonding system, the requirement to be considered for a patronage office, a way of making failure to live up to obligations to one’s patron costly enough so it would be unlikely to happen. I find this part of his argument less convincing than the general point about patronage. Requiring a bond makes sense, but why should the bond take the form of burning money, studying Latin and Greek instead of modern languages and skills useful in future life, constructing lavish estates inconveniently located far from population centers, devoting large areas of potentially productive land to show? The system could instead have required the would-be aristocrat to buy and provision a warship and donate it to the navy or purchase government bonds that paid below market interest rates, expecting to be repaid over time in the revenue from patronage offices.

One explanation may be that some of the apparently wasteful investment was relationship specific capital. A large country estate was not merely a bonding device, it was a place for doing some of the things that it was the job of the aristocrat to do as the leading figure in his local community, things that required gathering, sometimes feeding and hosting, substantial numbers of people.

Another explanation of apparently wasteful expenditure is that the revenues from government offices made aristocrats rich, as Allen extensively documents, and rich people like to show off their wealth, buy status among their peers, by extravagant spending. Jeff Bezos’ superyacht, his flight to the edge of space, his Venice wedding, were not bonding capital for patronage offices.

One of the nice things about Allen’s theory is that it neatly explains the end of the aristocracy. As technological progress made possible a modern bureaucracy to replace the patronage system the aristocracy lost its function, its political power and those features of aristocratic culture that, by Allen’s account, had functioned to bond potential recipients of patronage. The aristocrats were still rich with wealth accumulated through past offices, some of their descendants still are, but they no longer make up a distinct class distinguished by culture and lifestyle, they no longer duel. Most of their country estates, if still standing, belong not to their descendants but to the British government.

One weakness in the system of patronage is that it depended on the man at the top, the king, making the right choices, appointing high officers who will do a good job of running the kingdom for him through the men they appoint. If it fails at the top …

The Dutch in the Medway
1664-72

If wars were won by feasting,
0r victory by song,
Or safety found in sleeping sound,
How England would be strong!
But honour and dominion
Are not maintained so.
They're only got by sword and shot,
And this the Dutchmen know!

The moneys that should feed us
You spend on your delight,
How can you then have sailor-men
To aid you in your fight?
Our fish and cheese are rotten,
Which makes the scurvy grow -
We cannot serve you if we starve,
And this the Dutchmen know!

Our ships in every harbour
Be neither whole nor sound,
And, when we seek to mend a leak,
No oakum can be found;
Or, if it is, the caulkers,
And carpenters also,
For lack of pay have gone away,
And this the Dutchmen know!

Mere powder, guns, and bullets,
We scarce can get at all;
Their price was spent in merriment
And revel at Whitehall,
While we in tattered doublets
From ship to ship must row,
Beseeching friends for odds and ends -
And this the Dutchmen know!

No King will heed our warnings,
No Court will pay our claims -
Our King and Court for their disport
Do sell the very Thames!
For, now De Ruyter's topsails
Off naked Chatham show,
We dare not meet him with our fleet -
And this the Dutchmen know!

(Rudyard Kipling)

---

Past posts, sorted by topic

My web page, with the full text of multiple books and articles and much else

A search bar for text in past posts and much of my other writing

1. With the possible exception of the War of 1812.