Requisite Inefficiency
Viability Needs Strategic Waste
The theory of Requisite Inefficiency was born around 2012. There were a couple of blog posts and several nerdy talks in London, Hull and Brussels. Later on, in 2020, a more accessible version made it into a section in one of the chapters of Essential Balances. Now, observing the actions of DOGE, I thought the time was good to resurface it in an extended and refined form.
DOGE, which, judging by its actions, should be spelled out as the Demonic Order for Government Extermination, is not there to improve efficiency.1 But if we imagine that it genuinely works on efficiency, something every government needs, and succeeds in that, the result might be just as damaging but not visible so soon.
All big organizations are inefficient. Government organizations excel at this, but so do all big companies, especially in sectors where most of the work is administration, such as banking, insurance, and utilities. Inefficiency is a natural tendency, just like entropy. While entropy is the tendency towards disorder, inefficiency often results from actions to improve order.
When inefficiencies become obvious and start creating problems that nobody can ignore anymore, they invite efficiency actions — to remove waste in lean-speak.
Contrary to the announced intentions, most efficiency initiatives increase waste by keeping those working on it busy and funded. The funding will stop if they actually deal with the inefficiency, but most often, they are not able to deal with it. Yet, some brave ones make a visible change. The change looks like a scoop from a pile of sand:
It is as if they imagine the inefficiency layer as a layer of stones on top.
In fact, it looks more like this:
And importantly, not all inefficacy stones are to be thrown away. Some inefficiency stones are gems.
Formal organizations are not unique in needing a healthy amount of inefficiency to remain viable. All living and social systems won't last long without it.
How can that be?
Let's start with some examples, moving from small to big and then through time, from ancient to recent. Then, we’ll review some strategies, where there will be more examples.
From Cells to Empires
Our health depends on the highly inefficient functioning of B-cells. Cell mutation is generally harmful. It leads to cancer and genetic disorders. But it can also be a good thing. When an organism is infected, it relies on the immune system to restore its healthy state. Since the body cannot know who will invade it, it needs to generate diversity internally, which happens with the B-cells. The process is called somatic hypermutation. Only 10-20% are selected and used from these hypermutated B-cells. Our life depends on this inefficiency.
When ants find food, they bring a piece to the nest, leaving a pheromone trail. If other ants stumble upon such a trail, they tend to follow it, and when they reach the food source, take a piece and send it back to the nest, reinforcing the pheromone trail. This trail is transient. If not enforced, it evaporates. Quite inefficient. It's like owning a house with poorly insulated walls, requiring constant heating to maintain warmth. However, with ants, this inefficiency has an important function. When ants don't follow a pheromone trail, they ramble around, exploring for new food sources. The pheromone trail's evaporation rate contributes to maintaining the balance between exploration and exploitation.
Sexual reproduction is inefficient compared to asexual reproduction. It is energetically costly; it passes on only half of an individual’s genes to each offspring and requires the production of males who don’t bear offspring directly. Nonetheless, such inefficiencies promote genetic recombination, leading to genetic diversity. Such variation is crucial for adapting to changing environments and resisting pathogens. The latter is probably the very reason sexual reproduction took off.2
Sexual reproduction is a strategy similar to that of B-cells and for a similar reason. It is just at another level of organization.
Using diversity to deal with uncertainty can be seen in any living system, biological and social. In human social systems, this widespread strategy has been employed since ancient times.
One such practice, notable for its apparent inefficiency, is for farmers to scatter their plots of land. Such is still the case with the potato farmers in the district of Cuyocuyo in southern Peru.3 Each household owns numerous small pieces of land, ranging from 10 to 20 or even more, which they keep scattered over a large area. Consolidating these parcels may seem like the most sensible thing to do. Even more so considering that they are at an altitude between 3,300 and 5,000 meters above sea level, and traveling from one plot to another is difficult. Yet, this inefficiency is a smart risk strategy. Since their strips are scattered, the risk of various hazards (insect blight, theft, or frost) is spread, and the probability of getting something from their land yearly is greater.4
Strategies with requisite inefficiencies are ancient and extend beyond farming to the social order of polities. The Achaemenid Persian Empire, founded in 550 BC, developed one of the most sophisticated administrative systems of the ancient world. To manage their vast territory, they constructed a network of roads, created a postal system, and standardized their currencies and measurements. Persia was able to collect taxes and mobilize resources across diverse regions.
However, their highly efficient system was less adaptable to sudden military pressures or economic shocks. In contrast, nomadic groups (Scythians and later the Mongols) in the Asian steppes operated with inefficient but flexible command structures. When Darius I of Persia launched a campaign against the Scythians, their evasive guerrilla tactics led to a protracted campaign that ultimately forced Darius to withdraw.
Ten centuries after the fall of the Achaemenid Empire, another empire, the Venetian Republic, reached a new level of government efficiency.
Since Venice had no land or natural resources, it relied on trade. But no land also meant no feudal system. Venice experimented with new forms of government and new ways to make the administration efficient. The Venetian Republic successfully acquired and maintained extensive territories, which included the entire Adriatic coast, numerous islands in the Ionian Sea, and several cities in the Eastern Mediterranean and the Black Sea. With well-coordinated trade and administration machines, it dominated Mediterranean commerce for centuries.
In contrast, Genoa, their main rival, had a fragmented political and commercial structure. Unlike the centralized Venetian oligarchy, power was distributed among competing merchant families. That inefficiency, however, resulted in innovation. Genoese merchants experimented with early financial instruments (such as letters of credit and bills of exchange), which helped them adapt more rapidly to sudden market changes.
As a result of the loss of the spice trade, decline of the textile industry, rising cost and competition (e.g., in publishing) and other external shocks, Venice declined. Genoa survived these external shocks5 and even prospered in the 16th century.
From cells and insects, I went straight to farmers and polities, skipping the individual human being, so let's zoom back to individuals.
Many people today — especially knowledge workers — are pressed to increase output per unit of time. That brought a tsunami of self-help books, techniques, and tools on time management and productivity. During the pandemic, it led to the proliferation of productivity tools. Before the pandemic, there were fewer than a dozen personal knowledge management (PKM) tools; by the end, this number exceeded 80. A real explosion. And that's not counting the tools supporting a single technique.
But no matter how good these technologies are, they can't help with the productivity paradox:
The technologies we use to try to “get on top of everything” always fail us, in the end, because they increase the size of the “everything” of which we’re trying to get on top.6
Probably the most widespread time-management technique is the Pomodoro technique. I use it occasionally and can confirm that it works well. What it does is that it keeps you focused. It's one of many techniques supporting the popular belief that the greatest results are achieved when you are in the flow. Yet, many discoveries happen when interruptions break the flow. The likelihood of serendipitous events is influenced by one's working mode, and being attuned to such signals often conflicts with maintaining the flow and keeping distractions shut.7
The second problem with personal productivity and time management is that striving for efficiency may maximize every timebox, but some inefficiency is needed to make better choices about what to put in those timeboxes.
Strategies
We are instilled with the idea of fitness. A low-quality key copy won't fit the keyhole and will not work — making its production a waste. Everything created should serve its intended purpose. Isn't that nature's way? It’s all about survival of the fittest. That's what Darwin said (or how he is massively interpreted). Natural selection.
Yet, it’s not the fittest who survive but the viable, those sufficiently adapted to maintain their autopoiesis. And it is not a matter of natural selection. It is a matter of natural drift. In a similar way, we have cultural drift.8 What is irrelevant in routine cases becomes an important resource during disruption.
The ways that cultural practices adapt to the vicissitudes of situated action are a source of variability in performance, but are often considered to be formally irrelevant to the accomplishment of the task. However, this variability in "task irrelevant" dimensions may be a resource for adaptive processes when routine activity is disrupted.9
There are patterns of these inefficient, “task-irrelevant” dimensions, and patterns invite attempts at classification. Since these are behavior patterns (in which I also include adaptation), we can call them strategies. They are all systemic, yet they look different at different levels.
Component-level strategies
If we split component by type and function, we get the following matrix:
Redundancy
There are several “inefficient” strategies that have proven evolutionarily successful. One is the reserve of variety,10 a good example of which is genetic redundancy. Some genes with the same function are duplicated so that if there is a mutation in one of them, the other can successfully carry out that function. We have established similar strategies in life. We keep a spare tire in our cars.11 When the main server goes down, we have mechanisms to automatically switch to a “spare” one to ensure service continuity. This functional redundancy involves having two or more components of the same type capable of performing the same function. The spare components remain inactive during regular operations.
Cultural practices using redundancy range from simple to very sophisticated. It is intuitively expected that they will become more sophisticated over time, but this is not always the case.
A simple contemporary redundancy practice is the reserve players in football (aka soccer in the US), where substitutions can be made to respond to injuries or fatigue, or they can be a matter of tactics.
A more sophisticated ancient practice was the Celtic Trimarcisia cavalry tactic. It was used by the Gauls during their invasion of Greece in the 3rd century BCE. Each horseman had two servants who provided support during battle. If a rider or horse fell, a servant brought a replacement horse. If the rider died, the servant took his place. If both were lost, a reserve rider stepped in.
Degeneracy
A component can be of a different type, but when needed, it may have a temporary shift of function. In other words, to deal with uncertainty, different components have evolved to provide the same functions when needed. In biology, this phenomenon is called degeneracy and is observed in genes, proteins, and metabolism. A societal instance is when the army mobilizes during disasters, such as floods and earthquakes.12 In some cities, taxis and fire trucks are rigged out with essential paramedic equipment. Their drivers are trained to provide first aid. These taxis and fire trucks can shift their function and serve as ambulances when needed.
Multifunctionality
The human hands evolved for grasping, but they can also write, play the piano and acquire more functions without being structurally different and without losing their previous functions.
The human mouth handles eating, breathing, and speaking. Development of speech required the larynx to descend lower in the throat, enabling complex vocalizations. This change increased the risk of choking, as food and air share a common path, making it easier for food to enter the airway. Humans are more prone to choking compared to other mammals, with thousands dying annually from choking in the US.13 Multifunctionality is not just inefficient but dangerous.
Exaptation
A shift in function can last for an extended period, either complementing or entirely replacing the initial function. In evolutionary studies, this is called exaptation. A classic example is bird feathers. Their primary function was to regulate temperature. Birds used to use their feathers to trap air in cold weather. When it became too warm, they could cool their bodies by lifting their feathers. Later, feathers evolved to attract mates, catch insects, and, finally, for flight.
Another example of exaptation is the lungs of some basal fish, which evolved to control buoyancy.
A third example is in mammals, where some of the jawbones abandoned their initial function and evolved to improve hearing.
Even when the mechanisms of natural selection are artificially created (e.g., using genetic algorithms), the most successful strategy can use exaptation and show short-term inefficiency.14
Similar examples can be seen also in the evolution of technology. In 1945, a chocolate bar in the pocket of Percy Spencer at Raytheon started to melt when it was exposed to microwaves from an active radar set. That was the invention of the microwave oven.
In biological exaptation, an organ adopts a new function, usually abandoning its initial one. In technology exaptation, discovering a new function does not lead to abandoning the initial one, if it’s still useful, thus increasing the individual and collective variety of humans even more.
Systemic strategies
There are three systemic strategies with requisite inefficiency: diversity, exploration and agility.
Diversity
Diversity is the strategy to deal with uncertainty applied by the adaptive immune system, sexual reproduction, and Peruvian potato farmers scattering their plots of land, as described earlier.
The practice of distributing lands that remained intact in southern Peru was once widespread in Europe. In certain regions, it used to be more sophisticated than that of the Peruvian potato farmers. A rotation system was applied not only to crops but also to ownership. Instead of permanent ownership, there was an adaptive system. And the plots were not just rotated. Their number varied as the families grew and shrunk.15
Then came the state and killed all that. The bureaucrats needed to unify it. What mattered was not what worked or what was fair but what could be understood and managed.
These state simplifications, the basic givens of modern statecraft, were, I began to realize, rather like abridged maps. They did not successfully represent the actual activity of the society they depicted, nor were they intended to; they represented only that slice of it that interested the official observer. They were, moreover, not just maps. Rather, they were maps that, when allied with state power, would enable much of the reality they depicted to be remade. Thus a state cadastral map created to designate taxable property-holders does not merely describe a system of land tenure; it creates such a system through its ability to give its categories the force of law.16
The state adopts simplifications to improve governance efficiency. Then, its administration expands in response to new challenges, leading to inefficiencies. They invite efficiency initiatives that either fail, display superficial success to justify the perpetuation of cancerous internal structures or achieve real success, but a short-term one since it includes eliminating the inefficiencies essential for viability.
Exploration
Exploration is everywhere: from ants rambling to discover new food sources to human individuals trying out different things to get to know themselves and find their calling (a lot of time spent acquiring a skill never to be used again looks highly inefficient), organizations allowing exploratory time for their employees to stimulate innovation, and science progressing more often via curiosity-driven than market-driven research.
For the last point, let's be clear that it's not about minor innovations here and there. Apart from the serendipitous discoveries mentioned earlier, there are equally significant ones that result from exploratory research.
ARPANET, in the 1960s, was an experimental network that allowed exploration and resulted in the protocols we use on the Internet today.
Then, in the 1970s, the invention of monoclonal antibodies, driven by immune system curiosity, led to the development of life-saving drugs. Monoclonal antibodies target specific cells or molecules, improving treatments for conditions like lymphoma and some autoimmune diseases. This exploratory, curiosity-driven research created a market of 237 61 billion US dollars in 2023 and is expected to reach half a trillion in the next 5 years.17
And let’s not forget that exploration is a characteristic element of reinforcement learning currently used by most LLMs. It plays a prominent role in Claude and DeepSeek. Prioritizing reinforcement learning over supervised learning was one reason DeepSeek achieved results comparable to those of OpenAI at a lower cost. The right type of inefficiency at one level can lead to higher efficiency at another.
Agility
Agility is now over-associated with Agile and abused by zealots and consultants, so let's try to think of it here only in terms of structural flexibility and the ability to change quickly.
A historical example of an agile organization is the Hanseatic League, a federation of merchant guilds and towns between the 13th and 17th centuries.
The Hansa’s structure was always ambiguous. The League was not a state, though it was state-like in many of its actions. It was unique in the history of Northern Europe because unlike kingdoms or counties, it had no formal territory. It lacked a central decision making body, or a robust mechanism to enforce its policies. And even as a formal League, no definitive list of its members has ever been found.18
While inefficient in decision-making and unified action, the Hanseatic League had adaptability to local conditions, flexibility to change trade routes and authority to sign treaties with foreign rulers.
Another example is the Prussian Auftragstaktik, or mission-type tactics. Although it has inefficiencies such as miscoordination, misinterpretation of orders, and conflicting actions — which are not present in conventional military structures — it is adaptable at lower levels, and this has proven effective in rapidly evolving battlefield situations.
We already saw two other historical examples in the first section. One was Scythians. The structural flexibility of their society and military command structures had advantages over bigger and more efficiently organized Persians. The other example was Genoa's fragmented merchant families. Unlike Venice, Genoa didn't have coordinated administration and efficient trade management. Yet, their fragmentation resulted in experimenting, having diverse approaches to trade and innovating with financial instruments, which made them more resilient to external shocks. Both examples show that it's not just a matter of agility but also diversity and exploration. When it comes to socio-technical systems, we can think of each strategy as a strategy dimension since it is often a combination.
Inefficiencies
What is the source of inefficiency in each of these strategies?
Redundancy expends resources on unused backups. This inefficiency ensures reliability by offering replacements if a component fails.
Degeneracy employs less-optimized components for shared tasks. While this approach is inefficient, it allows for flexibility by enabling different components to substitute for one another.
Multifunctionality sacrifices specialization for versatility. By enabling a single component, like the human hand, to perform diverse tasks — such as grasping, writing, or playing instruments — it reduces the need for multiple specialized parts and makes the system more adaptable to changes in the environment.
Exaptation depends on structures that are not perfectly suited for their new purposes. Yet, such inefficiency promotes innovation by adapting existing elements for fresh roles.
Diversity generation wastes resources on maintaining a wide variety of components or outcomes, many of which may not be immediately useful. This inefficiency ensures the system is prepared for unknown future challenges by having a pool of diverse options available to draw from when needed.
Exploration wastes effort on random or inefficient actions that don’t provide immediate benefits. This inefficiency allows the system to discover new opportunities or solutions that wouldn’t be found through purely goal-directed behavior.
Agility reduces efficiency due to coordination costs and a lack of specialization. This inefficiency enables the system to adapt to changing conditions or unexpected shocks quickly.
Efficiency Norms
To evaluate something as efficient (or effective) is a matter of norms. Norms change.
In the closing years of the 19th century in the small town of Jena, more innovation and advancement of human thought were achieved in a decade than in the preceding and following centuries. It was an intellectual accelerator where philosophers, poets and scientists engaged in an intense, collaborative dialogue that redefined individuality and creative expression. It will suffice to list them: Hegel, Fichte, Goethe, Schiller, the von Humboldt brothers, the Schlegel brothers, and Novalis. For the majority of them, these were the most productive years. Nowadays, it comes as a surprise that these highly influential thinkers and creators not only knew each other but what they created was the result of their intensive collaboration. What's truly surprising is that it all happened in Jena, of all places.19
Why was Jena such a gravitation force and accelerator?
At the end of the 18th century, Germany was highly fragmented, and censorship was difficult to enforce. Jena was an extreme case. By that time, the Jena University was controlled by four different Saxon dukes. By the norms of the highly centralized France and England, that was both inefficient and ineffective. By contemporary norms, we may call these "favorable conditions for permissionless innovation."
Crisis Controversy
Nothing demands efficiency more urgently than a crisis. The now famous Lean concept of "waste", was a result Toyota's efforts to address resource scarcity in Japan following World War II. The Toyota Production System (TPS) identified eight types of waste, or muda: overproduction, waiting, transportation, overprocessing, inventory, motion, defects, and underutilized talent. By systematically eliminating these, TPS aimed to reduce costs, improve quality, and increase productivity.
But, as we saw, some inefficiency is needed when there is uncertainty. Yet, in times of crisis or uncertainty, the natural response is to reduce the overall inefficiency and, with it, the part that might be what the organization or society needs to survive the crisis.
Situations of crisis and uncertainty are complex situations. Best practices don’t work since the behavior is emergent and there are no obvious causal relations. You need to act and make sense of how the system reacts to that. One strategy the Cynefin framework recommends is running parallel, save-to-fail experiments. That is certainly inefficient and requires capacity and risk-taking. The former is unavailable due to crisis-driven efficiency measures. The latter is suppressed by crisis-induced conservative attitudes.
A crisis also needs new ideas. They come from something that so far was underutilized, something that did not fit the mainstream. Yet, efficiency measures are usually not equipped to distinguish normal stones from gems, and the sand pile gets cleaned of all kinds of stones.
Continuous efficiency improvement is much needed in every large organization. I'm not sure if another organization should manage that, and surely not one like DOGE, but whatever the structure and technology, it can only bring value if it has a way to protect the needed inefficiency.
Related posts:
DOGE seems to be a part of a different agenda. See Democracy Is Done: The Rise of Corporate Monarchy by
(Substack) and The strange political philosophy motivating Musk (Financial Times).Hamilton, W. D., Axelrod, R., & Tanese, R. (1990). Sexual reproduction as an adaptation to resist parasites (a review). Proceedings of the National Academy of Sciences of the United States of America, 87(9), 3566–3573. https://doi.org/10.1073/pnas.87.9.3566
Goland, C. (1993). Field Scattering as Agricultural Risk Management: A Case Study from Cuyo Cuyo, Department of Puno, Peru. Mountain Research and Development.
They employ various practices, including crop rotation, but not arbitrarily. For instance, the most sought-after varieties, known for their high dry matter content or flouriness, are cultivated on the upper terraces, where there is a low risk of flooding. See this article by Mike Jackson for more details. That article is also the source of the thumbnail image used for this essay.
In 2019, Cuyocuyo was recognized as the first biodiversity zone in Peru.
Some shocks were not the same, like the wars with the Ottoman Empire.
Burkeman, O. (2023). Four thousand weeks: Time management for mortals (First paperback edition). Picador/Farrar, Straus, and Girous.
See Random, Living, Error and The Rea Model.
Hutchins, E. (2010). Enaction, Imagination, and Insight. In J. Stewart, O. Gapenne, & E. A. Di Paolo (Eds.), Enaction (pp. 425–450). The MIT Press. https://doi.org/10.7551/mitpress/9780262014601.003.0016
Requisite in Requisite Inefficiency is picked to resonate with the Law of Requisite Variety. The direct relation with it can be seen in earlier, nerdier descriptions. For a detailed, non-technical introduction to the LORV, see Stimuli and Responses.
Now, the trend is not to have one.
If we zoom in on the case of a shift of function, we can see that some of them are based on the history of events. There is a specialization, a readiness to carry out a specific additional function. In other words, it is a prepared shift of function for an expected kind of adverse event. These are not helpful when there is a completely unexpected change in the environment or when the effects of a known kind of event have not been experienced in recent times. The spread of Covid-19 showed that very clearly. Some organizations were able to meet the unexpected stimuli with novel responses, while others relied on routine business continuity practices, most of which turned out to be inadequate.
See statistics.
In Complexity, A Guided Tour, Melanie Michell described an evolutionary algorithm for a robot cleaning soda cans in a 10x10 grid. The evolved strategy included an apparently inefficient action — sometimes not picking up a can immediately — and, in this way, using it as external memory. This approach scored higher than the best human-designed strategy.
That's how the evolution of structural coupling works. And although successful practices stabilize, different conditions make different paths of congruence.
Scott, J. C. (2020). Seeing like a state: How certain schemes to improve the human condition have failed (Veritas paperback edition). Yale University Press. https://yalebooks.yale.edu/book/9780300078152/seeing-like-a-state/
Ahmed, A. (n.d.). The rise and fall of the Hanseatic League. Retrieved March 27, 2025, from https://worksinprogress.co/issue/the-rise-and-fall-of-the-hanseatic-league/
For a detailed account of these events, see
Wulf, A. (2022). Magnificent Rebels: The First Romantics and the Invention of the Self. John Murray Publishers Ltd. https://www.johnmurraypress.co.uk/titles/andrea-wulf-2/magnificent-rebels/9781529392760/