Seeing Like a State

From The Jolly Contrarian
Jump to navigation Jump to search
The Jolly Contrarian’s book review service™

North Korea.jpg
A weekly stakeholder work-stream check-in call, yesterday.

Index: Click to expand:

Comments? Questions? Suggestions? Requests? Insults? We’d love to 📧 hear from you.
Sign up for our newsletter.

Seeing Like a State: How Certain Schemes to Improve the Human Condition Have FailedJames C. Scott

No battle — Tarutino, Borodino, or Austerlitz — takes place as those who planned it anticipated. That is an essential condition.

—Tolstoy, War and Peace

Business, and government, suffers from a kind of physics envy.

Rory Sutherland, citing Paul Ormerod

In sum, the legibility of a society provides the capacity for large-scale social engineering, high-modernist ideology provides the desire, the authoritarian state provides the determination to act on that de­sire, and an incapacitated civil society provides the leveled social ter­rain on which to build.

James C. Scott

This one, with Jane JacobsThe Death and Life of Great American Cities, goes to the top of JC’s 2020 lockdown re-reads. It was published in 1998, so it’s a bit late to get excited — but while it addresses the “high modernism” of 20th Century government, the read-across to the capitalist market economy, and beyond that into the interior workings of any large corporation — are you reading, boss?[1] — shrieks from every page. These are profound ideas we all should recognise, and which could transform the effectiveness of what we all do, but — being, well, citizens of a “prostrate civil society” — either we can’t or we won’t.

Exactly why there is this collective affliction of wilful blindness to our administrative compulsion is a great, unexplored topic of our age. That so many, great and small, have so much to lose by exploring it may explain the mystery.

Seeing Like a State takes as its thesis how well-intended patrician government can, in some circumstances, lead to utter disaster. While Scott’s examples are legion, one could — and some do — criticise him for his anecdotal approach: he has curated examples that best fit his thesis, and it therefore suffers from insoluble confirmation bias. That may be true, but I don’t think it matters, for Scott’s thesis is so familiar, so plausible and its exhortations so consistent with other theories in adjacent fields,[2] that it is hard to be bothered by a lack of empirical rigour. This stuff all stands to reason. Data is not its value: Scott’s narrative is its value, as a counter-narrative to modern statist (and corporate) orthodoxy — that some gilded superman, sitting at the top of the heap magically pulls levers for the betterment of all — and that, in itself, is valuable and enlightening.

In any case, Scott does not say that top-down bureaucratic disaster is inevitable, but notes the same four conditions are present wherever we find it: a will to bend nature — and the polity — to the administrator’s agenda; a “high modernist” ideology that holds that that all problems can be anticipated and solved in time with the necessary organisation, application and empirical rigour; an authoritarian state, with machinery to impose its ideological modernist vision; and a subjugated citizenry (or staff) without the means (or inclination) to resist the machinations of the administrator.

These qualities, of course, pertain in any autocratic polity. Stalinist Russia, Maoist China and latter-day North Korea fit the pattern exactly. But so do most modern multinational corporations. If you are interested in how not to run one, Seeing Like a State is worth a close read.

Legibility: the administrative ordering of nature and society

Any government must be able to “read” and thus “get a handle on” — hence, “make legible” — and so administrate the vast sprawling detail and myriad of interconnections between its citizens, lands and resources. It does this by, in its “statey” way, narratising a bafflingly complex system into a thin, idealistic model: it assigns its citizens permanent identities (in the middle ages, literally, by giving them surnames: now, identity cards and the, er, chips that are shortly to be implanted in our foreheads); it decrees standard weights and measures for all times and places (we may have proceeded by local customs and conventions);[3] commissions cadastral surveys of the land so it can collect taxes; it records land holdings, registers births, deaths and marriages, imposes conventions of language and legal discourse designs cities and transport networks: in effect, to create a standard grid that could be measured, monitored and understood from the bird’s eye view of city hall. A population that legible is manipulable.

This cost of this legibility is abridgement: it represents only the slice of society that interests the administrator, which would be harmless enough those measures did not in turn impact how citizens interact with each other and their environment. But, as we know they do. Citizens account for their income to optimise their tax position. When adminstrators levied a window tax — reasoning that the number of windows is proportionate to the size of a building, and therefore a fair proxy — citizens redesign houses to have as few windows as possible, notwithstanding adverse consequences to the general health of the population. Modern society is shot through with similar arbitrary rules wherever government interacts with its citizenry. Through their combined effect society comes to be remade to suit the administrator, but not always in ways the administrator might have had in mind. Society is the archetypal system: arbitrarily diverting its natural stocks and flows only creates other feedback loops.

Scott is persuasive that we lose something critical when we simplify in our yen for clear description, which state officials cannot but do. Trying to covert local customs — “a living, negotiated tissue of practices which are continually being adapted to new ecological and social circumstances — including, of course, power relations” — to unalterable laws loses the subtlety and scope for micro-adjustments that these customs, if left to themselves, continually experience.

In other words, you lose something special when you atomise a complex system. Emergent properties vanish. It is a poorer, less productive thing.

High modernist ideology

This yen to regularise often comes with a “muscle-bound” self-confidence that the state can expand production, better satisfy human needs and master nature (including human nature) and centrally configure social order “commensurate with the scientific understanding of natural laws”. This is the “high-modernist” view. It translates to a rational, ordered, geometric (hence “legible”) view of a world which depends on the benign guiding vision of the state to bring about big projects.

Now those infinitesimal interconnections and illegible relations are not just “invisible” to the state programme but inimical to it. Natural forests are replaced with grid-planted Norway spruce: swathes of the unwanted ecosystem — which provide a richness and benefit to participants in that ecosystem which the state cannot “see” — are rejected because they don’t fit the model. But they can play valuable and vital roles in the ecosystem — even for the Norway spruce.

A new term, Waldsterben (“forest death”), entered the German vocab­ulary to describe the worst cases. An exceptionally complex process in­volving soil building, nutrient uptake, and symbiotic relations among fungi, insects, mammals, and flora — which were, and still are, not en­tirely understood — was apparently disrupted, with serious conse­quences. Most of these consequences can be traced to the radical sim­plicity of the scientific forest.[4]

The deterministic belief that the “illegible” details — in this case, literally, “in the weeds” — don’t matter will eventually come back to haunt you. “Nature,” as Dr. Ian Malcolm put it in Jurassic Park, “finds a way”.

But the high modernist believes the future is solvable, and the certainty of that better future justifies the disruption and “short-term side-effects” of one’s grand plan to get there.

The alternative — which terrifies the high-modernist — is an iterative, ground-up, organic management by those on the ground, best placed and best incentivised to use their judgment and experience to best solve their own problems and improve the general lot as they personally perceive it. Their “read” of the landscape will be necessarily far richer and more detailed than the state’s. This is both far more effective for society, and far scarier for administrators: they have less control over progress, less sight of it, (therefore) less to do, and a harder job justifying the rent they extract (in a government, this is called a “tax”; in a corporation, it is executive compensation) for providing their “vital” administration.[5]

Once the desire for comprehensive urban planning is established, the logic of uniformity and regimentation is well nigh inexorable. Cost effectiveness contributes to this tendency. Every concession to diversity is likely to entail an increase in time and budgetary cost.

Another cost of the high modernist ideology that seeks to regularise and unitise is diversity in the things so regularised. That diversity and inclusion is the cause célèbre du jour, in the public and private sectors, hardly falsifies this observation. It just sharpens the irony, since the typical approach to delivering diversity chimes with this desire for narratising legibility and high-modernism.

Diversity ought, you’d think, to be hard to pin down, its manifestations being naturally — well — diverse. Diversity is the very benefit that accrues from the range of our differences; the interplay of our unique perspectives.

But, to get a handle it, organisations must make diversity legible. They do this by defining it in a strikingly limited and homogenous way. They gather data from their staff on that limited metric — to make it more legible, so that the organisation can propagate statistics about its “improving” diversity. Thus, “diversity” as the administration knows it is a formalised, homogenised, parameterised and regularised proxy of diversity, and no attention is paid to how this proxy diversity affects the behaviour of people in the organisation, for good or ill.

In any case, the point is clear: if imposed proxies can prompt the wealthy to restructure their tax affairs and French peasants to fill in their windows, so can it prompt those in a commercial organisation to behave in similarly counterproductive ways. There is an argument that whole segments of the infrastructure have developed for precisely that reason. Legal included.

An authoritarian state and prostrate civil society

Scott’s last two criteria are opposite sides of the same coin: an authoritarian state that is able to coerce the society it manages to bring its high modernist ideals to bear, and a subjugated population that cannot resist it.

Scott was writing in 1998, a few years after the collapse of communism, when Francis Fukuyama and others were declaring the end of history, all battles won and so forth, so was a little shoe-shuffly about this. He needn’t have been. Not only have we seen the return of authoritarian governments and prostrate populations — for posterity, I write from the ninth month of a government-mandated nationwide lockdown — but both the authoritarian disposition amongst the executive class and the supine one amongst the general population have always been a feature of the corporate sector.

Every “meaningful”[6] aspect of your performance and your role is, at some level, reduced to a parameterised data point: ID, location, salary, rank, position, performance, reporting line, holiday entitlement, sick-leave, service catalog, objectives. All of that work you do: the subtle analysis, the advocacy, the creative solutions — all is, in the eyes of the executive, reduced to a grade, a rank and a number.

As for the high modernist ideal, well, this entire site is a paean to that, but “strategy” as we mutely receive it, seems entirely predicated on a reductionist ideology that we can solve all conundrums in our landscape and then proceed sedately and without the need to be troubled by turbulent subject matter experts thereafter.

Given our recent history you would think our overlords ought to know better than that.


Speaking of subject matter experts brings us nicely to Scott’s closing, where he ruminates on the concept, missing from high modernist canon, of metis. This is hard to describe — folk wisdom, knowhow, Odyssean cunning — but in the corporate world it struck me as most resembling expertise. Ingenuity, problem-solving, lateral thinking; smarts for figuring out what to do on the fly if you are in a jam. This is something that the high modernist programme seeks to abolish — the theory being that loose cannon employees wandering around making snap decisions is potentially catastrophic, and jams of this sort can and should be avoided by appropriate planning and the right algorithm: thus, subject matter experts aren’t needed.

There are two interesting observations here. The first is that metis is much more efficient than an algo, even if you can find one to work. You could — if you accept the reductionist stance — solve any problem with the right calculations, but the necessary data and processing power would be huge. Practical knowledge, on the other hand — metis — is “as economical and accurate as it needs to be, no more and no less, for addressing the problem at hand.”

This is the difference, says Scott, between Red Adair[7] and an articled clerk. There are some skills you cannot acquire except through experience: learning to sail, ride a bike, or play a musical instrument or negotiate a commercial contract. You could spend as much time as you like with textbooks, but you will never master that kind of skill until you have done enough practical rehearsal. This is where the meatware knocks the chatbots into a cocked hat, and always will.

Which brings us to the last connection: to complexity theory, systems analysis and normal accidents theory. And, for that matter, superforecasters. All of these come to the same conclusion: if you are dealing with complex systems, especially tightly-coupled ones with non-linear interactions, you cannot solve these with algorithms, no matter how much data and no matter how sophisticated is your conceptual scheme. The only way to manage these risks is with experts on the ground, whom you empower to exercise judgment and make provisional decisions, which they can adjust as a situation unfolds. That is, with their metis. If with your high-modernist schema you have eliminated metis from your operation, you may carry on in times of peace and equability, but come the revolution, you are stuffed.

See also


  1. Boss: “Yes, JC, I am. Now, get your coat.”
  2. Charles Perrow’s Normal Accidents theory; systems theory as expounded by Donella H. Meadows, Thomas Kuhn’s The Structure of Scientific Revolutions
  3. It is said Chinese farmers gauged distance by “the time it takes to boil rice”, which provides a different, and more practical means of comprehending how far away you are.
  4. Scott, 20.
  5. It is, of course, a heresy to question it, but is any CEO really worth the hundred times the average employee the firm pays for him?
  6. Meaningful to them, not to you.
  7. Younger readers may not remember this legend of the fire-fighting community.