A parent sits at the kitchen table after the children are asleep. A form on the phone asks for a document that is somewhere in the house, or perhaps in an office that will not open until morning.
The form is short. The service has been redesigned. Someone has probably counted the fields removed from the old version.
The parent must now find the document, work out whether it is the right one, and decide whether to risk submitting without it. Tomorrow's lunch break may become a telephone call.
This is an imagined scene. It describes a cost that many measures of efficiency are poorly equipped to see.
The organization has made its part of the work smaller. The work itself may have stayed the same size. Some of it may have grown.
It went somewhere.
The missing column
Think of a restaurant that saves money by asking customers to sort every piece of their own trash. Or an office that reduces telephone waiting times by directing every caller to a website whose instructions are difficult to follow.
These choices could be sensible. Some people prefer to do things themselves. A clear digital form can be a genuine improvement over a trip across town.
The question is whether the work became easier or merely changed hands.
An institution usually measures what happens inside its own boundary: minutes per transaction, staff hours, completed forms, cost per case. Those numbers matter. They are also an incomplete account of what happened.
Outside the boundary are other clocks. The person learning unfamiliar rules. The family member gathering documents. The employee leaving work to correct an error. The volunteer explaining the same instructions again.
When an organization closes its ledger at the door, it can mistake transferred effort for eliminated effort.
Three kinds of cost
Public administration researchers Donald Moynihan, Pamela Herd, and Hope Harvey gave this problem a useful vocabulary in their research on administrative burden.
They distinguish the effort required to learn what a service expects, the effort required to comply with its requirements, and the psychological cost of going through the process.
A form can take five minutes to complete and several evenings to understand. A request for one more document can mean transportation, childcare, and a lost hour of paid work. An unclear decision can leave someone wondering whether a basic need will be met.
These costs are unevenly distributed. A requirement that takes two people the same amount of time can still make very different demands on their lives. One can leave a desk at noon. Another works a shift with no paid time off. One knows the language of the institution. Another must learn it while the deadline approaches.
A rule that costs everyone one hour does not cost everyone the same.
The distinction has practical consequences. In April 2025, the U.S. Government Accountability Office reported that the federal instructions used to document the burden of benefit-related information collections did not direct agencies to discuss potential learning or psychological costs. Its review examined 51 selected information-collection requests. GAO warned that the omission could leave agencies with an incomplete picture of the burdens they impose.
Even a government trying to measure paperwork can miss the work that comes before the paperwork.
What the shortcut protects
There are reasons to ask people for evidence, confirm identities, or slow a decision down. A safeguard may protect public funds, personal safety, privacy, or the rights of another person. Removing every step would create other harms.
Good design takes those duties seriously. It also asks whether the same duty could be fulfilled with less expense to the person seeking help.
The difference is easiest to see at the handoff.
If someone supplies information that an organization already possesses and is authorized to use, why must they gather it again? If an error was made inside the institution, why does correcting it require the person outside to reconstruct the whole history? If a service offers self-service, can the person still reach someone when the ordinary path fails?
Each answer reveals where responsibility has been placed.
A shortcut becomes harmful when the institution keeps the saving and leaves someone else responsible for the trouble.
Follow the work home
Before celebrating a redesigned process, trace one person's path from the moment they recognize a need until that need is actually met.
Count the work on both sides of the counter. Notice what people must learn before they can begin, what they must collect, how many times they must explain themselves, and what they do when a step fails. Talk to people whose available time, language, mobility, and access differ from those of the design team.
Then ask a harder question: which of these burdens is truly necessary, and who has the authority to remove the rest?
This is useful far beyond public services. A small business can examine how it handles a return. A school can examine enrollment. A neighborhood organization can examine how a volunteer joins. A technology company can examine the distance between signing up and receiving help.
The second outcome of any project is what people can do afterward. There is another side to that accounting: what they are still being forced to do on its behalf.
A shorter internal process may be good news. It becomes progress when the person at the other end can feel the difference.
The next time a dashboard says the work got faster, ask which clock it stopped.
Then find the one still running.