Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› jpam-writing-style
jpam-writing-style
GitHub用于撰写或润色JPAM风格的政策分析稿件，强调清晰、校准和决策导向。确保政策主张与证据匹配，避免过度推断，提升跨学科可读性，明确研究局限，使非专业读者也能理解核心发现与政策含义。
Trigger Scenarios
起草摘要、引言或政策启示部分
审稿人指出过度推断、结论被掩盖或难以理解
需根据证据校准政策建议强度
简化术语以适应跨学科APPAM受众
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill jpam-writing-style -g -y
SKILL.md
Frontmatter
{
"name": "jpam-writing-style",
"description": "Use when drafting or polishing the prose of a Journal of Policy Analysis and Management (JPAM) manuscript — a clear policy implication stated without overclaiming, legible to economists, political scientists, public-management scholars, and practitioners. Improves clarity and calibration; it does not change the analysis."
}
Writing Style (jpam-writing-style)
The JPAM house voice is clear, calibrated, and decision-oriented. A JPAM paper earns its place by stating a policy implication the design actually supports — neither buried in hedges nor inflated beyond what was identified. Write for a reader who may be an economist, a political scientist, a public- management scholar, or a practitioner, and who wants to know: what did you find, how credible is it, and what should we do about it.
When to trigger
- Drafting the abstract, introduction, or the policy-implications / conclusion section
- A reviewer said the paper "overclaims," "buries the finding," or "isn't legible to non-economists"
- Calibrating the strength of the policy recommendation to the evidence
- Tightening jargon for the cross-disciplinary APPAM audience
The JPAM introduction arc
Policy problem & decision → what's known and the gap → the policy/program studied → the credible design (named as a tool) → the headline effect in policy units → the cost-benefit / distributional takeaway → the calibrated implication → brief roadmap. The reader should know the finding and what it means for
policy by the end of the first page.
Calibrated implication (the JPAM signature)
- State the implication, then bound it. "This program raised Y by Z; at a cost of C per recipient, the benefit-cost ratio is R" — followed by the scope conditions that govern whether it transfers.
- Match the verb to the design. A local RD estimate supports a claim about the threshold population, not the whole eligible pool. An ITT supports a claim about offering the program, not receiving it.
- Say what you cannot conclude. Naming the limits is persuasive at JPAM, not weak; an honest boundary protects the credible core from the overclaiming critique.
- No policy advocacy beyond the evidence. Recommend within what the estimate, cost-benefit, and external-validity conditions support.
Cross-disciplinary legibility
- Define field-specific jargon; do not assume the econ, PS, or PA reader shares your vocabulary.
- Lead with substance over method; the estimator is a tool, not the headline.
- Keep the abstract concrete: question, policy/program, design, finding, implication.
Checklist
- Finding and policy implication clear by the end of page one
- Implication calibrated to the estimand (local/ITT/LATE stated, not over-generalized)
- Scope conditions for transfer stated; limits named honestly
- Effects in policy-relevant units in the prose, not only standardized
- Jargon defined for a cross-disciplinary audience
- No advocacy beyond what the design + cost-benefit support
Anti-patterns
- Overclaiming: a sweeping policy recommendation from a local or imprecise estimate
- The opposite failure: burying a credible finding under so many hedges no implication survives
- Leading with the estimator instead of the policy question and finding
- Field jargon that locks out the other APPAM disciplines
- A conclusion that pivots to advocacy the evidence does not license
- An abstract that names the method but never the policy takeaway
Overclaim vs. calibrated (illustrative contrasts)
- Overclaim: "Our results show that this program should be adopted nationwide." → Calibrated: "Among the threshold population our RD identifies, the program raised Y by Z at a benefit-cost ratio of R; whether this extends to the full eligible pool depends on [scope condition]."
- Overclaim: "The policy works." → Calibrated: "Offering the program (ITT) raised Y by Z; because take-up was T%, the effect on participants is larger but less precisely estimated."
- Over-hedged (the opposite failure): burying a credible effect under so many caveats that no implication survives — JPAM wants the calibrated claim stated, not dissolved.
Calibration anchors (hedged)
- At JPAM, naming what you cannot conclude strengthens the paper — it is the credibility that earns the policy recommendation you can make.
- The Point/Counterpoint forum aside, JPAM is not an advocacy outlet; recommendations live inside the envelope of the estimand, the cost-benefit result, and the external-validity conditions.
- Write the abstract so a policymaker learns the finding and the takeaway; a method-only abstract misses JPAM's audience.
Output format
【Implication (calibrated)】what to do, bounded by the design
【Estimand match】local/ITT/LATE reflected in the verb? [Y/N]
【Scope + limits】transfer conditions and honest boundaries stated
【Units】effects in policy-relevant terms in prose? [Y/N]
【Legibility】jargon defined for econ/PS/PA readers? [Y/N]
【Next】jpam-transparency-and-data
Supplementary resources
../../resources/worked-examples/01-introduction.md— before→after JPAM introduction in house style../../resources/exemplars/library.md— JPAM papers that state calibrated implications
Version History
- 1839142 Current 2026-07-05 13:53
