The paradox of reduced commitments

You cut your list and your output went up

You have eight projects running. You know this because you can list them if pressed, though the list takes a moment to assemble because three of them have been sitting untouched for weeks despite technically being "active." You spend your days switching between them. Monday morning you push project A forward by forty-five minutes before a meeting about project B pulls you away. After the meeting you have eleven minutes before lunch so you answer three emails about project C. After lunch you have what should be a two-hour block, but you spend the first twenty minutes remembering where you left off on project D and the last fifteen minutes feeling guilty about project E, which you have not touched since last Thursday. By Friday, you have touched everything and finished nothing. You tell people you are busy. You are not busy. You are fragmented.

Then you cut the list from eight to three. Your calendar develops something you have not seen in years: empty space. Blocks of ninety minutes with nothing scheduled. It feels wrong. It feels lazy. It feels like you are falling behind. You resist the urge to fill the space. And then you finish the first project in half the time it would have taken when eight projects were competing for your attention. You finish the second project the following week. By the end of the month, you have completed more things than you completed in the previous three months combined.

Your output went up despite doing less. Welcome to the paradox. It is not a paradox at all once you understand the mathematics. It is a predictable consequence of how throughput actually works in any system with limited resources — including you.

The throughput paradox

The intuition that more commitments produce more output is wrong, and it is wrong in a specific, mathematically demonstrable way. The error is confusing starts with finishes. Starting a project is not output. Finishing a project is output. And the relationship between the number of things you start and the number of things you finish is not linear. It is inversely related once you exceed your capacity threshold.

The formal articulation comes from John D.C. Little, who proved in 1961 what is now called Little's Law: the average number of items in a system (L) equals the average arrival rate (lambda) multiplied by the average time each item spends in the system (W). Rearranged: lead time equals work-in-progress divided by throughput rate. W = L / lambda.

This equation has a direct implication that most people miss. If your throughput rate is roughly constant — and for a single human it is, because you have a fixed number of productive hours per day — then increasing WIP (work-in-progress) increases lead time proportionally. Double your active projects and each one takes twice as long to finish. Triple them and each one takes three times as long. The relationship is mechanical, not motivational. It does not matter how hard you try. The math does not care about your willpower.

But it gets worse than the math alone predicts, because human throughput rate is not actually constant across different WIP levels. It degrades as WIP increases, due to a phenomenon that manufacturing engineers discovered decades before productivity writers started writing about it: context switching costs.

The evidence is overwhelming

Gloria Mark, a professor of informatics at the University of California, Irvine, has spent over two decades studying attention in the workplace. Her research, synthesized in her 2023 book Attention Span, establishes several findings that bear directly on the commitment paradox. First, the average knowledge worker switches tasks every three minutes and five seconds. Second, after an interruption, it takes an average of twenty-three minutes and fifteen seconds to return to the original task at the same level of engagement. Third — and this is the critical finding — the cost of switching is not just the transition time. It includes a "residue" effect, first described by Sophie Leroy in 2009, where attention from the previous task lingers and contaminates performance on the current task. You are not just losing the minutes of switching. You are degrading the quality of every minute that follows the switch.

Stephen Monsell's task-switching research, published across multiple studies from the late 1990s through the 2010s, quantified the cost more precisely. Even when subjects knew in advance which task was coming next and had time to prepare, the switch cost persisted. It was not a surprise penalty. It was a structural penalty — an inherent cost of reconfiguring mental resources from one task context to another. The more different the tasks, the higher the cost. The more tasks in the rotation, the higher the cumulative cost. Monsell's findings suggest that the effective throughput of a person juggling five distinct projects is not 100% divided by five (20% per project). It is closer to 5-10% per project, because 50-75% of total capacity is consumed by switching overhead.

David Anderson, who brought Kanban methodology from Toyota's manufacturing floors to knowledge work in the mid-2000s, built an entire management framework around this insight. His core prescription is the WIP limit: an explicit cap on the number of items allowed to be in progress simultaneously. Anderson's data from hundreds of software teams showed that reducing WIP limits — often dramatically — consistently increased throughput, decreased lead time, and improved quality. Teams that fought the idea, insisting they needed to work on more things to deliver more things, reversed their position once the metrics proved them wrong. The pattern was so consistent that Anderson made WIP limits the foundational practice of the Kanban method, preceding all other improvements.

Greg McKeown's Essentialism (2014) frames the same principle from the strategic level. McKeown argues that the undisciplined pursuit of more — more projects, more commitments, more opportunities — produces less total contribution. The word "priority" was singular until the twentieth century. It meant the one thing that came first. McKeown's central claim is that restoring its singularity is the highest-leverage move available to most people. When everything is a priority, nothing is. When you protect the vital few, their impact compounds.

Warren Buffett reportedly described his investment approach using what is sometimes called the "20-slot" rule: imagine you have a punch card with only twenty slots, and every investment decision you make for the rest of your life punches one slot. You would be enormously careful about what you chose. Buffett's point was not about investing. It was about the quality premium that comes from constraint. When you cannot do everything, you must choose. When you must choose, you choose better. And when you choose better, each choice receives the resources it needs to succeed.

Why fewer commitments produce more output

The mechanism has four interlocking components, and all four contribute to the paradox.

Reduced context switching. With three projects instead of eight, you switch contexts less frequently within a day and across a week. Each project gets longer unbroken sessions. Gloria Mark's research shows that performance in a focused session increases nonlinearly with session length — the twentieth minute of a focused block is more productive than the fifth minute, because depth of processing increases with sustained attention. Fewer projects mean longer sessions mean deeper processing per session.

Reduced decision overhead. With eight active projects, you face a prioritization decision every time you sit down to work: which one do I advance right now? That decision consumes time and willpower. Barry Schwartz's The Paradox of Choice (2004) demonstrated that more options increase decision fatigue and decrease satisfaction with the eventual choice. With three projects, the decision is trivial. Often there is no decision at all — you simply continue what you were doing yesterday. The cognitive resources that were being consumed by meta-work (deciding what to work on) are freed for actual work.

Reduced guilt and anxiety. Every active commitment you are not currently advancing generates a background process of guilt, worry, and rumination. David Allen's Getting Things Done framework calls these "open loops" — commitments that consume mental bandwidth regardless of whether you are actively working on them. Eight projects mean five to seven open loops humming at all times. Three projects mean zero to two. The freed bandwidth is not trivial. It is the difference between a mind that can think clearly and a mind that is perpetually distracted by its own backlog.

Increased quality per unit of effort. When a project receives adequate resources — enough time, enough attention, enough creative energy — the output quality is higher. Higher quality means less rework, fewer false starts, and fewer cycles of revision. A project completed at 90% quality ships once. A project completed at 60% quality because it was starved of attention ships, gets criticized, and requires a second round of work that would not have been necessary if the first round had been given adequate resources. The rework cycle consumes more total time than doing it right once.

These four effects compound. Less switching means more depth. Less decision overhead means more execution. Less guilt means more clarity. More quality means less rework. The net effect is that three projects, each receiving adequate resources, produce more finished output than eight projects each receiving scraps.

The experiment you should run

Theory is useful. Data from your own life is conclusive. Here is the experiment.

For one month, cut your active commitments to a number that is roughly 40% below your current count. If you have eight active projects, cut to five. If you have twelve, cut to seven. If you have five, cut to three. The specific number matters less than the act of deliberate reduction.

The cut must be real. Archive the deferred commitments. Remove them from your task management system's active view. Communicate to stakeholders that these items are paused, not abandoned. Set a calendar reminder to revisit them in thirty days. Then stop thinking about them. If you catch yourself checking on a deferred project, redirect. The experiment requires a genuine reduction in cognitive load, not a cosmetic one.

Track two things during the month. First, count finished outputs: projects completed, deliverables shipped, milestones reached. Not tasks started. Not hours worked. Finished things. Second, note the quality of what you finish — not with a vague feeling, but with a concrete assessment. Did this require rework? Did you ship it at the standard you intended? Would you be comfortable showing it to someone whose judgment you respect?

At the end of the month, compare these numbers to the month before the cut. In almost every case I have seen — and the manufacturing, software, and knowledge-work literature backs this consistently — the reduced-WIP month produces more finished output at higher quality despite fewer active commitments. The data makes the case better than any argument could.

If the data does not support the paradox in your specific situation, that is also useful information. It means either the cut was not deep enough (you are still above your capacity threshold), the cut was not real (you kept working on deferred items), or your particular bottleneck is not WIP-related. All three of those diagnoses point toward specific next steps that are more useful than continuing to run at maximum load and hoping the results improve.

The Third Brain

Your externalized knowledge system and AI assistant become powerful instruments for making the throughput paradox visible in your own data.

The core challenge with the reduced-commitments experiment is that your intuition will fight the data. You will feel less productive during the first week because your calendar has gaps and your task list is shorter. Your identity may be entangled with busyness — many people's is. An AI system with access to your output logs can provide the objective evidence your intuition will not.

Track your completed outputs in a structured format: date, project, deliverable, time invested, quality assessment. Feed this to your AI at the end of each week. Ask it to compare weekly throughput across the reduced-WIP period versus the previous period. Ask it to calculate average lead time per project — how many days from start to finish. Ask it to flag the ratio of finished items to started items.

The AI can also surface a pattern that is difficult to see from inside the experiment: the relationship between your WIP count and your completion rate across longer time horizons. If you have three months of data, your AI can show you that weeks with fewer active projects consistently produced more completions — or that the effect appears only below a specific threshold. That threshold is your personal WIP limit, and knowing it precisely is one of the most valuable pieces of self-knowledge you can possess.

The AI does not need to motivate you to do less. It needs to show you the data that proves less produced more. Once you have seen the graph — WIP on the x-axis, completed output on the y-axis, the curve bending downward past a certain point — the paradox stops being a paradox. It becomes your operating policy.

The courage to have empty space

The deepest resistance to reducing commitments is not logical. It is emotional. Empty space on a calendar feels like waste. A short task list feels like underperformance. Saying "I am only working on three things" feels like an admission that you cannot handle more. In a culture that equates busyness with importance and overcommitment with ambition, choosing to do fewer things requires a kind of courage that no productivity framework can give you.

But this entire phase has been building toward a single recognition: capacity is finite, and pretending otherwise produces suffering, not results. You measured your actual capacity in Measure your actual capacity. You learned that sustainable pace outperforms sprint pace in Sustainable pace over sprint pace. You confronted the cost of overcommitment in The cost of overcommitment. You learned to say no in Saying no to protect capacity. And in Capacity for growth and maintenance, you saw that growth and maintenance both require dedicated capacity — you cannot fund growth with the scraps left over from an overfull maintenance load.

The paradox of reduced commitments is the culmination of that argument. It is not enough to understand that capacity is finite. You must act on that understanding by deliberately constraining your commitments to a level your capacity can actually serve. The reward for that constraint is not less output. It is more output, higher quality, shorter lead times, and a mind that is clear enough to do its best work on the things that actually matter.

The next lesson is the capstone of this phase, and it reframes everything you have learned about capacity planning as something larger than a productivity technique. Capacity planning, done honestly, is a form of integrity — an alignment between what you say you will do and what you can actually do. The paradox you explored here is the most vivid proof of that principle: the most productive version of you is not the one who takes on everything, but the one who chooses carefully and delivers fully.