brycewang-stanford REStud Literature Positioning (restud-literature-positioning)
When to trigger
- The related-work discussion is a list of citations rather than an argument
- The user cannot name the single closest paper and what theirs adds
- A referee or seminar comment was "how is this different from [X]?"
- The contribution is original (per
restud-topic-selection) but its location in the literature is unclear
What REStud positioning must do
REStud is general-interest and weights theory and applied work equally, so the positioning has two jobs at once:
- Confront the closest papers honestly. Name the 3–5 nearest results and state precisely what each one already established and what it could not.
- Make the marginal contribution legible to a non-specialist on either side. Because the handling editor and referees may sit on the theory or the applied side (the Joint Managing Editors span IO/applied econometrics, micro theory, and information economics), an applied paper should be legible to a theorist and a theory paper to an applied reader. Models that defined whole literatures appeared here first — Mirrlees (1971) on optimal taxation, Stiglitz (1974) on sharecropping incentives — so the register is "this changes how the field models X," not "this adds new data to X."
A REStud-quality positioning paragraph is an argument, not a literature dump: "Paper A established X but assumed Y; paper B relaxed Y but only for setting Z; we do W, which neither could, because [our model / design / data]."
Building the positioning
Step 1 — The nearest-neighbor map
List the closest papers in two rings:
- Inner ring (3–5): papers a referee will say you must beat or extend. For each: what it shows, its key assumption or limitation, and the one sentence distinguishing yours.
- Outer ring: the broader literatures the result speaks to (this is where general interest is demonstrated). Two to three strands, each with a canonical anchor citation.
Step 2 — The "why now / why not before" clause
State why the contribution was not made earlier: a method that did not exist, data that was not available, a model that was not tractable, or a fact no one had documented. This is the spine of an original-contribution claim.
Step 3 — Confront, do not bury
Cite the closest competitor in the introduction, in the main text, not in a footnote. Referees who suspect you are hiding the nearest paper turn hostile.
Step 4 — Canonical anchors
REStud referees expect the foundational theory or methods references to be present (e.g., the seminal model your framework extends, or the identification-method papers your design relies on). Missing the obvious canonical citation reads as not knowing the field.
Checklist
- Inner-ring papers (3–5) named, each with "what it could not do"
- One-sentence distinguisher for each inner-ring paper
- Outer-ring strands identified (demonstrating general interest)
- "Why now / why not before" clause written
- Closest competitor cited in the introduction, not a footnote
- Canonical theory / method anchors present
- Positioning is an argument, not a list
Anti-patterns
- A "related literature" paragraph that is a string of "(Author, Year; Author, Year)" with no argument
- Hiding the single closest paper to inflate apparent novelty — referees find it and it costs credibility
- Over-claiming ("the first paper to ...") when an honest reader can name a precedent
- Positioning only within the narrow subfield, so the general-interest claim is unsupported
- Citing your own working papers as the main precedent while ignoring the field's anchor results
Output format
【MARGINAL CONTRIBUTION】<one sentence, relative to the literature>
【INNER RING】[paper — what it could not do — your distinguisher] x3-5
【OUTER RING STRANDS】<2-3 literatures + anchor cites>
【WHY-NOT-BEFORE CLAUSE】<one sentence>
【CANONICAL ANCHORS PRESENT】yes / missing: [...]
【NEXT SKILL】restud-identification (empirical) | restud-theory-model (theory)