jep-balance-and-objectivity
GitHub用于审查JEP风格文章是否保持平衡与客观，确保公平呈现竞争观点、区分事实与政策主张，并诚实承认不确定性。避免偏袒或倡导性语调，提升学术合成的严谨性与中立性。
Trigger Scenarios
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill jep-balance-and-objectivity -g -y
SKILL.md
Frontmatter
{
"name": "jep-balance-and-objectivity",
"description": "Use when a Journal of Economic Perspectives (JEP) article must treat competing views and policy debates fairly — synthesis, not advocacy — and acknowledge uncertainty honestly. Audits balance and tone; defer evidence presentation to jep-evidence-without-equations and structure to jep-narrative-arc."
}
Balance & Objectivity (jep-balance-and-objectivity)
When to trigger
- The draft argues for one side of a live debate and slights the other
- It reads as a policy advocacy piece or an op-ed, not a synthesis
- Uncertainty and dissenting evidence are downplayed to sharpen the conclusion
- The author has skin in the game (their own theory/result is the "right" answer)
The JEP objectivity bar
JEP is rigorous synthesis by economists for economists, not advocacy. A JEP article must give a fair account of competing views, weigh the evidence rather than the author's preferences, and be honest about what is uncertain or unknown. The reader should trust that they are getting the state of the debate, not a brief for one position. This is the line that separates JEP from a well-written op-ed.
Balancing moves
- Steelman the other side. State the strongest version of competing views and the best evidence for them — not a strawman that's easy to knock down.
- Weigh evidence, not advocates. Let the weight and consistency of studies drive the verdict (mirrors
jep-evidence-without-equations), and say plainly where the evidence is mixed. - Separate positive from normative. Distinguish "what the evidence shows" from "what policy should do." Where you offer a policy view, label it as a judgment, give the trade-offs, and note reasonable people who'd disagree.
- Disclose your stake. If you are summarizing a debate you are a party to, acknowledge it and lean harder toward fairness; let opposing work speak in its own terms.
- Mark uncertainty. "The evidence is suggestive but not settled" is a feature; resist the urge to round debate up to consensus.
- Avoid loaded framing. Neutral language for contested terms; don't smuggle the conclusion into the labels.
The advocacy red flags (and the fix)
| Red flag | Fix |
|---|---|
| Only one side's evidence appears | Add the strongest opposing studies and weigh them |
| Policy recommendation stated as if it followed mechanically from facts | Separate the positive finding from the normative call; show trade-offs |
| Dissent dismissed in a clause | Give the dissent a fair hearing and say why it does/doesn't move the verdict |
| Confident verdict on a genuinely open question | Report the disagreement and what would resolve it |
| Author's own result framed as the settled answer | Disclose the stake; present competing estimates neutrally |
Checklist
- Competing views stated in their strongest form (steelmanned)
- Verdict driven by weight/consistency of evidence, not by the author's side
- Positive findings separated from normative/policy judgments
- Policy views (if any) labeled as judgments, with trade-offs and dissenters noted
- Author's own stake disclosed where relevant
- Uncertainty and open questions stated, not rounded to consensus
- Framing and labels neutral on contested terms
Worked vignette (illustrative)
An author who has spent a career arguing that a policy works writes the JEP synthesis of that policy. The balanced version opens by disclosing the stake, presents the strongest evidence on both sides (including the studies that find null or negative effects), and concludes: "On balance the evidence leans positive for short-run outcomes, but the long-run and general-equilibrium effects remain genuinely contested." That last sentence — refusing to round a live debate up to consensus — is what makes an editor trust the piece as synthesis rather than as the author's brief.
Anti-patterns
- A one-sided brief that an editor will read as an op-ed, not a synthesis
- Strawmanning the opposing view so the favored view wins easily
- Presenting a policy preference as a direct, value-free implication of the data
- Burying dissent or uncertainty to make the conclusion look cleaner
- Using the author's own work as the yardstick of truth without disclosure
- Loaded terminology that pre-decides the debate
Output format
【Debate】[the contested question]
【Competing views (steelmanned)】side A: … side B: …
【Verdict basis】weight/consistency of evidence: [...]
【Positive vs. normative】separated? policy labeled as judgment? [Y/N]
【Author stake】disclosed if relevant? [Y/N]
【Uncertainty stated】open questions flagged? [Y/N]
【Advocacy check】reads as synthesis, not op-ed? [Y/N]
【Next step】jep-editor-strategy
Version History
- 1839142 Current 2026-07-05 13:34
