Withholding documents is dangerous. Producing too many is not.
That asymmetry is the whole story, and it is worth stating plainly because most discovery strategy is built on top of it without anyone saying so.
A defendant who withholds responsive material is exposed. There is a motion, a hearing, a record of what was asked and what came back, and a judge with a long memory. A defendant who produces five million documents has, on the face of it, complied. The volume is the compliance. Challenging it means arguing that the other side gave you too much, which is a difficult motion to write and a worse one to argue.
Sanctions for producing in a deliberately unusable form do exist, and courts have occasionally reached them. But they are rare, they require showing intent, and they arrive years after the production that caused the problem. Set against the near-certainty of a sanction for withholding, the incentives are not close.
The cost moves
Volume does not destroy evidence. It relocates cost.
Every document produced is a document somebody has to look at, and the party who chose the volume is not the party who pays to read it. That is the transfer, and it runs one way.
The asymmetry underneath it is financial rather than legal. Defense counsel is paid to review, by the hour, as incurred — the reading is the product. Plaintiff counsel funds the same reading out of a contingency, against a recovery that may be years away and may not come. Two firms doing identical work, one billing it and one financing it.
A production sized to be reasonable can still be sized to be expensive. Both things are true at once, and only one of them shows up in a motion.
Reasonable is not the same as helpful
The standard a defendant has to meet is a reasonable search. It is not a promise that what matters will be easy to find in the result.
This is the part that surprises people new to large cases. The useful documents are almost always in there. Nobody had to hide them. They are sitting inside a set large enough that finding them requires either months of reading or knowing precisely what to ask for — and the party that assembled the set knows exactly which of those two you are equipped to do.
Delay compounds it. A production that takes eight months to work through is eight months in which the case does not develop, memories do not improve, and the pressure to resolve builds on the side that is paying to wait.
Reading faster is the wrong answer
The instinct is to add reviewers, extend hours, and get through it. It is the wrong instinct, and it is the response the volume was chosen to provoke.
Speed does not change the ratio. A team that reads twice as fast through a set where one document in four thousand matters is still spending most of its attention on the other three thousand nine hundred and ninety-nine. The cost transfer is intact; it just happens sooner.
The answer is to make the volume irrelevant to the question being asked.
Measure it before you read it
A production has a shape, and the shape is knowable before anyone opens a document. Which custodians, over which months, in what proportion. Where the density falls off. Which date ranges are thin next to the ones that matter, and which people are absent from a set where you expected them.
That analysis takes a day and answers a different question from review: not “what is in here” but “what is missing, and what would I move to compel.” It is also the only part of the exercise where the defendant’s choices become visible rather than invisible.
Then ask for a proposition rather than a subject. “Everything about the recall” is a request that gets you volume, and the volume is a legitimate answer to it. “The documents showing who approved the label change and when” is a request that either gets you the documents or gets you a fight worth having, and the fight is on a record that a judge can act on.
What this is really about
None of this makes a large production small. It changes who controls what the production costs you.
A team that reads in the order documents arrived has accepted the other side’s sequencing, the other side’s sizing, and the other side’s economics. A team that measures first, asks by proposition, and stops when a proposition is established has made volume into a fact about the case rather than a weapon in it.
The defendant chose how much to send. You choose what you spend on it.