The cost of gathering material for a non-fiction book falls on the author, paid out of the advance. That arrangement quietly decides which subjects get investigated in depth.
The advance is a budget, not a salary
An advance is meant to cover the author's living expenses during the writing and every expense of the research, with no separate allowance for either.
It is also paid in installments spread across signing, delivery and publication, so the money arrives slowly while the costs arrive early.
An author who underestimates the research spends their living expenses on travel, which is a common reason books arrive late.
Access has a price list
Archives charge for reproduction, courts charge for records, and specialized databases carry subscription fees that individuals rarely hold.
Interviews mean travel, and travel to a distant state or country for a single conversation is routine in reported work.
Transcription, translation and a researcher's time then add costs that scale with how much material the book gathers.
Legal review is a real line item
Books about living people or active institutions are read by lawyers before publication, and the publisher's review does not always cover the author's exposure.
Where a subject is likely to object, the reporting standard has to be higher, which means more documents, more corroboration and more time.
The cost of doing that properly is one reason some investigative subjects are pursued by news organizations rather than by individual authors.
The economics select the subjects
Books that can be researched from home, or from a single collection, are cheaper to produce than books requiring years of field reporting.
That difference does not appear in a catalog, but it shapes what proposals authors write, because an author has to be able to survive the project.
Subjects that are expensive to report and uncertain to sell fall into a gap where nobody can fund the work.
Fellowships and institutions fill part of the gap
Grants, university appointments, journalism centers and residencies fund reporting that advances alone would not support, and much serious American non-fiction depends on them.
These sources have their own priorities and application cycles, which influences the timing and framing of the books that receive support.
The result is a category whose depth in any given area reflects the funding available for reporting it, not just reader interest.