Legibility and the Firm

I recently took a sales course for my job. Some of the discussion there prompted some thoughts about ideas I've read/heard and how they relate to the modern business world.


I initially came across the concept of "legibility" in a book review of Seeing Like A State by James C. Scott. Although I haven't read the full book, I've read an essay by the author to get a further grasp on this concept. It begins with the following paragraph:

State naming practices and local, customary naming practices are strikingly different. Each set of practices is designed to make the human and physical landscape legible, by sharply identifying a unique individual, a household, or a singular geographic feature. Yet they are each devised by very distinct agents for whom the purposes of identification are radically different. Purely local, customary practices, as we shall see, achieve a level of precision and clarity—often with impressive economy—perfectly suited to the needs of knowledgeable locals. State naming practices are, by contrast, constructed to guide an official “stranger” in unambiguously identifying persons and places, not just in a single locality, but in many localities using standardized administrative techniques.

Legibility, in Scott's usage, refers to the way governments put standardized labels on everything from roads to people. He gives an example of how a road between the towns of Guilford and Durham might be known as the Durham road to people on one end and the Guildford road to people on the other end but the state maps will label it as Route 77 to avoid confusion (since many towns around Durham will have a "Durham road"). The conventional name is more useful to locals, since it tells them where the road goes, while the standardized number is more useful to officials governing a wide area since it removes ambiguity.

Over time there has been a trend towards making things more tractable to distant officials. This has included everything from land surveys (which came up a lot a couple of posts ago) and even permanent last names:

The invention of permanent inherited patronyms was, along with the standardization of weights and measures, uniform legal codes, and the cadastral land tenure survey, a vital technique in modern statecraft. It was, in nearly every case, a state project designed to allow officials to identify unambiguously the majority of its citizens. The armature of the modern state: tithe and tax rolls, property rolls, conscription lists, censuses, deeds, birth, marriage and death certificates recognized in law were inconceivable without some means of fixing an individual’s identity and linking him or her to a kin group.

I also thought this point from the essay was interesting, about the way that measures taken to increase legibility can have unintended consequences:

The quest for legibility, when joined to state power, is not merely an “observation.” By a kind of fiscal Heisenberg principle, it has the capacity the change the world it observes. The window and door tax established in France under the Directory and abolished only in 1917 is a striking case in point. Its originator must have reasoned that the number of windows and doors in a dwelling was almost perfectly proportionate to the dwelling’s size. Thus a tax assessor need only walk around the house counting the windows and doors to estimate its size. As a simple expedient, it was a brilliant stroke, but not without consequences. Peasant dwellings were subsequently designed or renovated with the formula in mind so as to have as few apertures as possible! While the fiscal losses could be recouped by raising the tax per opening, the effects on the long term health of the rural population lasted for than a century.

Of relevance to this post is the point that this concept of legibility is applicable to corporations not just governments:

Intervention in society (or nature) for whatever purpose (e.g. delivering welfare benefits to those with particular disability or keeping watch on political enemies) requires creating the mapping or optics necessary to legibility. In Seeing Like a State, and as a student of politics, I concentrate on state-making and government. Nevertheless, as I endeavor to make clear, large-scale capitalism is just as much an agency of homogenization, uniformity, grids, and heroic simplification as the state, with the difference that, for capitalists, simplification must pay. The profit motive compels a level of simplification and tunnel vision that, if anything, is more heroic that the early scientific forest of Germany. In this respect, the conclusions I draw from the failures of modern social engineering are as applicable to market-driven standardization as they are to bureaucratic homogeneity.

Scott's essay is part of a magazine issue that includes some response essays. A reponse by Donald Boudreaux raises the point that making things more legible, while it can lead to a loss of specialized local knowledge, can also facilitate positive interactions (not just for the officials driving the increased legibility) over a wider area. For example, standardized units of measure facilite trade in markets and social security numbers have been a useful form of identification for the financial industry, not just for the government that assigned them.

James Scott’s fascinating essay reminds us that trade-offs are inescapable. Much useful local knowledge is indeed lost when names and other descriptors of roads, rivers, and even family lines are changed from their original, localized, and idiosyncratic expressions into synoptic, standardized expressions that are meaningful to a larger and more diverse group of people. In return, though, something useful is gained — namely, greater coordination among larger numbers of people.

Here's another review of Scott's book.

The Firm

One of the papers I read recently on irrigation (by Alston and Smith) included this statement:

From the insight that transaction costs define observed economic outcomes in important ways (Coase 1937) came the recognition that the costs of specifying, enforcing, and exchanging rights to property are an essential example of the deep significance of transaction costs.

I decided to read this paper by Coase, titled "The Nature of the Firm" since I'd also heard about it elsewhere, such as in a podcast I listened to recently (incidentally featuring Donald Boudreaux who was quoted above).

This paper is trying to arrive at a useful technical definition of a firm. It starts off by explaining that in economic theory the use of resources and actions by different participants in the market are coordinated by price signals, without relying on any central director. However, within a firm there is someone directing things:

If a workman moves from department Y to department X, he does not go because of a change in relative prices, but because he is ordered to do so.


As D.H. Robertson points out, we find "islands of conscious power in this ocean of unconscious co-operation..." But in view of the fact that it is usually argued that co-ordination will be done by the price mechanism, why is such organization necessary? Why are there these "islands of conscious power"? Outside the firm, price movements direct production, which is co-ordinated through a series of exchange transactions on the market. Within a firm, these market transactions are eliminated and in place of the complicated market structure with exchange transactions is substituted the entrepreneur-co-ordinator, who directs production.

This leads to the observation that,

the distinguishing mark of the firm is the supersession of the price mechanism. ... [A] distinction between the allocation of resources in a firm and the allocation in the [capitalist] economic system...

And the central question of the paper is that,

it is surely important to enquire why co-ordination is the work of the price mechanism in one case and of the entrepreneur in another.

Some reasons that Coase brings forward are the costs associated with price discovery and the difference in efficiency between standing contracts compared to one-time contracts:

The main reason why it is profitable to establish a firm would seem to be that there is a cost of using the price mechanism. ... that of discovering what the relevant prices are. ... The costs of negotiating and concluding a separate contract for each exchange transaction which takes place on a market must also be taken into account.

This raises the implication that if it becomes easier to make short-term and one-off contracts (e.g. gig economy, smart contracts on the blockchain) then maybe big companies will become less common and more people will work as independent contractors.

Because "it is always possible to revert to the open market" if the organizing authorities of a firm fail to add effeciency, the transactions taking place within a firm need to be justified by having a efficiency gain from smoother coordination than what can be achieved by price signals. The potential to capture a monopoly provides an incentive for companies to grow, but this is countered by a loss of efficiency with increasing size:

as a firm gets larger, ... the costs of organizing additional transactions within the firm may rise. Naturally, a point must be reached where the costs of organizing an extra transaction within the firm are equal to the costs involved in carrying out the transaction in the open market

As I understand it, this approach of marginal analysis of transaction costs (internal and external) to determine the theoretically optimal boundary of a firm is the enduring legacy of Coase's 1937 paper:

The question always is, will it pay to bring an extra exchange transaction under the organising authority?

Bringing These Concepts Together

Having introduced these concepts, I want to spend this final section of this post thinking about how they fit together and apply to the modern business world. In a company, management needs to allocate labour and resources more efficiently than they could fill the same functions by going to the open market. One thing that makes this goal possible is transaction costs. For example, hiring a consultant and hiring an employee both take up time in soliciting proposals/resumes, evaluating them, and negotiating with the selected candidate. However, a consultant will generally do a task and write a report and then move on, while an employee can be assigned a whole series of tasks and reports; the upfront investment of time in finding a candidate is spread out over more output in the case of the employee, lowering the transaction cost. So as long as a company has enough work for someone with particular skills, it's probably more efficient to have them as an employee than as a consultant. Another factor that can help management meet their goals is detailed knowledge of the company. In the case of small businesses, or large businesses that are still led by a founder or by someone who rose through the ranks and knows the business intimately, this knowledge comes from "local" experience. In the case of large companies led by professional managers who have MBAs rather than a background in the specific field at hand, however, the business operations need a high level of legibility to provide this knowledge.

Tools for making business processes more legibile include ERP software. From what I've seen, the types of effects discussed by James Scott are certainly present in this domain. Although it makes business processes more visible to management through the dashboards and automated reports available in the software, the uniformity imposed (across different product lines, office locations, etc.) can lead to losses in utility. This is because reality is fractally complex and the map is not the territory. Any system of legibility (i.e. simplified, standardized labels) will have cases where it doesn't capture the full picture. To satisfy an ERP system, people in one office might have to waste time collecting information they don't need, while people in another office might have nowhere to enter information they do need. When individuals have more discretion, they can tweak things to meet the "local" needs, but when uniformity is imposed, some details are inevitably lost in the reduced granularity.

More generally, for all the benefits metrics/KPIs can bring to companies (and other organizations), they are not without drawbacks, as Jerry Muller has written (and discussed on a podcast). These include things like an incentive to game metrics, a loss of focus on anything not captured by the metrics, the transaction cost (which fits in well with this post) of tracking them, and their tendency to stifle innovation:

Compelling people in an organisation to focus their efforts on a narrow range of measurable features degrades the experience of work. Subject to performance metrics, people are forced to focus on limited goals, imposed by others who might not understand the work that they do. Mental stimulation is dulled when people don’t decide the problems to be solved or how to solve them, and there is no excitement of venturing into the unknown because the unknown is beyond the measureable. The entrepreneurial element of human nature is stifled by metric fixation. (emphasis added)

Things that are unknown (especially unknown unknowns) don't fit easily into categorization schemes, so they are inherently resistent to legibility. But this is also the domain where some of the greatest rewards (and risks) can be found.

What is the alternative to an excessive pursuit of legibility? I would say at least part of the answer is a respect for local knowledge and a robust network of relationships. Although price signals generally aren't used within a firm, as Coase discussed, that doesn't mean other forms of decentralized information sharing and feedback can't be. Consider this image of an artificial neural network from Wikipedia:

Neural network structure diagram

Imagine a sales team that functioned like this. Someone represented by a red circle might get a request from a customer. Maybe they have dealt with that customer before, or it's in their geographic territory, or they represent a product line that the customer thinks is the best solution for them (whether or not it actually is). In a large enough company, an individual sales person might not know all the product lines that are available. But if he/she can talk to various people (represented by the blue circles) who have local knowledge about different divisions of the company, hopefully one of them can pass the request to the appropriate product specialist (represented by the green circles). If there is positive feedback when a request makes its way to the right person, over time those relationships will be strengthened so that the frontline salespeople have strong connections to the specialists dealing with products that are the best fit for the customers they encounter. Other approaches would be to split the company up if internal transaction costs (not necessarily in money but in time and general frictional inefficiencies) are higher than just going to the wider marketplace, or to take steps to make the sales process more legible and structure it into silos such that each salesperson will be aware of all the products that might be relevant to any customers he/she is likely to encounter. But if the silos are too rigid, then some unforeseen (i.e. unknown and hard to measure as mentioned above) opportunities may be missed; thus I still think robust networks of relationships are a good thing to promote within a business or other organization.