How long is not how good

The order is an order of mechanisms

median age in use, by kind
  desks and tables           9
  storage and shelving       8
  printing and scanning      7
  seating                    7
  cables and adapters        6
  lamps and lighting         6
  screens and stands         5
  input devices              5
  other                      5
  audio                      4

Read down that list and it is not a league table of quality. It is a list sorted by how many ways a thing can stop being used.

A desk can stop for one reason: the room. A screen can stop for five: a better one, the work, the room, a failure, or somebody else needing it. The record shows exactly that — two stops in twenty-four desks, twenty-one in forty-one screens.

Things last in proportion to how few things can end them. That is not a property of the objects. It is a property of the categories.

Nothing here evaluates anything

No brand is recorded anywhere in this inventory. No price. No specification, no condition assessment, no note of whether anybody liked anything.

So there is no sense in which this record could rank objects even if it wanted to, and the figures above are not the beginning of an order of merit that fuller data would complete. They are counts of how long things have been where they are, in eleven places, by kind.

A reader who comes away thinking that desks are a better purchase than screens has read a fact about surfaces not improving as though it were advice.

The cables are the proof

Cables and adapters sit fifth in the list above, with a median in-use age of six years — the same as lamps, and higher than screens.

The same cables are the largest category in the cupboard, the largest share of the breakages, and the largest group of things replaced by something better. Thirty-four of the sixty cables in the record have stopped being used.

Both facts are true at once. A cable that survives its first couple of years goes on indefinitely, and cables stop being used more often than anything else in the inventory.

Any order of merit built from one of those two figures would be contradicted by the other, which is the clearest demonstration available that the duration figures are not a quality measure.

What the record would need to say anything about quality

A brand or model for every object, so that like could be compared with like.

A price, so that duration could be weighed against cost.

A condition assessment at a fixed point, by somebody applying the same standard, rather than each owner's own sense of whether a thing works.

A date for every stop, so that duration meant something precise.

A population that was not assembled from memory, so that the short-lived objects were not systematically missing.

Five things, none of which this record has, and the first two of which it deliberately refuses — because a record with brands and prices in it becomes a record about what to buy, and this one exists partly as a demonstration that the question what became of it can be asked without ever asking what was it.

The two kinds that prove it from opposite ends

Desks: nine years, the top of the list. Twenty-four objects, two stops, both because the room changed. Nothing replaces a desk and nothing breaks one. The nine years is the age of twenty-two unfinished lives in a category with almost no way to end.

Printing and scanning: seven years, fourth on the list. Eight objects, six stopped, all six by the same change of work, and all six carried out of the building. The seven years rests on the two survivors.

One kind scores high because nothing can end it; the other scores high on two objects while its category was being retired wholesale. Neither figure says anything about whether a desk or a printer is a good thing to own.

Where the list would mislead most

Cables, at six years, above screens and input devices.

Sixty cables in the record. Thirty-four have stopped — more than any other kind. Twenty-four are in cupboards, three of them broken. Fifteen stopped for no reason anybody can give, which is the largest single cell anywhere in the inventory. Six broke, which is a third of all the breakage in the record.

And the ones still in use have a median age of six years, with the oldest at eleven.

A reader taking the duration list as a league table would conclude that cables outlast screens. The rest of the record says cables are the most frequently retired, most frequently broken and most frequently forgotten thing on a desk.

The one thing duration does establish

That the equipment in these eleven workplaces is old.

A median of six years in use, twenty-two desks averaging nine, three objects past fifteen. Whatever else is true, this is not a stock that turns over quickly, and the picture of desk equipment as something constantly refreshed does not describe these rooms.

That is a statement about eleven workplaces and about nothing else. Eleven places are not a sample, the next eleven could be entirely different, and the final section of this record sets out at length what these figures can and cannot be made to carry.