jhr-literature-positioning

Literature Positioning (jhr-literature-positioning)

When to trigger

• The paper has a policy question but the contribution over prior work is fuzzy
• Prior estimates conflict with yours and need reconciliation
• The introduction reads like a general literature survey

JHR positioning bar

JHR explicitly values whether a manuscript reconciles its results with prior

published work. Positioning must therefore do more than cite related papers: it

should explain why estimates differ across samples, designs, populations, periods,

or mechanisms.

Checklist

• Name the closest prior estimates and their design.
• State whether your contribution is new variation, new data, new population,

new mechanism, stronger reconciliation, or policy interpretation.

• Compare magnitude, not only sign and significance.
• Identify which prior result your analysis confirms, revises, or contradicts.
• Preview any comparative estimation you run to bridge prior results.

Prior-estimate bridge

For each closest paper, write:

[Prior paper] estimates [magnitude] for [population/design]. Our estimate differs because [sample,
policy environment, outcome, or design]. We bridge the difference by [comparative specification].

This makes the JHR reconciliation expectation visible in the literature section rather than deferring it

to the appendix.

Reconciliation map

Build a compact map for the introduction team:

Prior result | Population | Design | Estimate | Why comparable | Your bridge

Use the map to decide which papers deserve main-text comparison. A paper belongs in the first-page or

second-page narrative only if it supplies a benchmark estimate, a competing mechanism, or a policy context

that changes interpretation. Broader background can move later or disappear.

When estimates conflict, avoid the weak phrase "differences may reflect context." Say which context,

which sample, or which design component produces the difference, and point to the table that tests it.

Vignette: bridging conflicting minimum-wage estimates

Illustrative positioning problem (all magnitudes invented): your state-panel

design finds a teen-employment elasticity of -0.12; the closest published

county-pair design reports -0.03.

1. Decompose the gap before writing: restricting your sample to border counties

moves -0.12 to -0.06; matching the prior paper's 2005-2014 window moves it

to -0.04. So roughly two-thirds of the gap is comparison-group choice, the

rest is period.

1. Write the introduction sentence from the decomposition: "Our larger

elasticity reflects statewide rather than border-county comparisons; under

the border-pair design our estimate is statistically indistinguishable from

the prior one."

1. Park the full bridge table in the Online Appendix and cite it at that

sentence.

1. Close the loop in the conclusion: say which prior estimate now looks like a

special case of yours, and what evidence would discriminate between the

remaining explanations.

This is the JHR pattern: the literature section quantifies disagreement instead

of narrating it.

Field benchmarks referees carry in their heads

Position against the benchmark literatures a JHR referee in each field will

expect you to know, hedging where canon is contested:

• Labor: minimum-wage employment debate, returns to experience and tenure,

monopsony evidence.

• Education: returns to schooling, teacher and school value-added, class

size, school-choice lotteries.

• Health: Medicaid and insurance expansions, early-life health shocks.
• Discrimination: audit and correspondence studies, decomposition methods

and their critiques.

• Retirement: Social Security claiming, default and auto-enrollment

effects.

• Development: schooling-and-earnings microdata work, cash transfers.

Missing the canonical comparison invites a referee to do the reconciliation for

you — on less favorable terms.

Positioning sentences to avoid at JHR

• "To our knowledge, no paper has..." without saying why the gap matters.
• "Differences may reflect context" with no named sample or design component.
• A paragraph of citations with no magnitudes attached to any of them.
• Claiming to "extend" a literature whose estimates you never compare to yours.

Output format

[Closest prior estimate] citation + magnitude/design
[Your delta] data / design / population / mechanism / reconciliation
[Why estimates differ] ...
[Policy implication] ...
[Next step] jhr-identification-strategy