jpube-data-analysis

Data Analysis (jpube-data-analysis)

When to trigger

• You are estimating a behavioral elasticity, take-up rate, or moral-hazard parameter
• The analysis uses administrative or register microdata with disclosure constraints
• You need to map estimates into welfare (DWL, MVPF, sufficient statistics)
• Robustness, heterogeneity, or measurement concerns are unresolved

What JPubE analysis looks like

Public economics at JPubE is typically built on large administrative or register data — tax records (IRS/SOI), social-insurance files (UI/DI/SSA), health-program data (Medicaid/CMS), or whole-population European registers — because credible policy elasticities need population-scale variation around kinks, notches, and reform dates. The analysis should convert clean identification into a policy-relevant quantity, not stop at a coefficient.

Analysis norms

• Estimate the policy parameter directly. Recover the taxable-income / labor-supply elasticity, the take-up or crowd-out rate, or the insurance-vs.-moral-hazard wedge that the welfare argument needs.
• Sufficient statistics & MVPF. Where you claim a welfare verdict, show the mapping from estimated responses to the formula explicitly, state the primitives held fixed, and propagate standard errors (delta method or bootstrap) into the welfare object.
• Respect disclosure and licensing. Restricted tax/health/register data require formal access and output clearance; document cell-size suppression and the access path, and supply programs even when microdata cannot be shared.
• Measurement honesty. Top-coding, income definitions, real-vs.-nominal, and program-rule coding drive public-finance results; document each choice.
• Heterogeneity that matters for policy. By income, eligibility margin, or jurisdiction — heterogeneity is the input to optimal nonlinear policy, not a fishing expedition.
• Robustness. Bin width and excluded region (bunching), bandwidth (RD/RKD), estimator choice (staggered DID), functional form, and sensitivity of the welfare conclusion to key elasticities.

Execution bridge (StatsPAI / Stata MCP)

Run the battery, don't just enumerate it. Full map:

[execution-with-mcp](../../../shared-resources/empirical-methods/execution-with-mcp.md). JPubE is public economics — tax/transfer/program designs; DiD/IV/RDD and bunching are central, magnitudes in policy units.

• Many outcomes / specifications: romano_wolf (step-down FWER) or benjamini_hochberg.
• OVB sensitivity: oster_delta / sensemakr.
• Inference: wild_cluster_bootstrap (few clusters), twoway_cluster / conley.
• Re-fit off one handle: audit_result(result_id) lists missing checks + the exact

suggest_function for each.

• Exhibits: etable / did_summary_to_latex from the handle — no retyped numbers.

Decisive checks in the body, exhaustive battery in the appendix.

[JF execution walkthrough](../../../Journal-of-Finance-Skills/resources/worked-examples/02-execution-walkthrough.md).

Checklist

• [ ] The estimated object is the parameter the welfare claim needs
• [ ] Sufficient-statistics / MVPF mapping shown with primitives stated and SEs propagated
• [ ] Data access, licensing, and disclosure (cell suppression) documented
• [ ] Income / program-rule / top-coding definitions stated and tested
• [ ] Heterogeneity tied to a policy margin, not data-mined
• [ ] Robustness covers the design-specific tuning parameters
• [ ] The welfare conclusion's sensitivity to key elasticities is reported

Anti-patterns

• Reporting a regression coefficient and never converting it to a welfare quantity
• An MVPF / sufficient-statistic number with no standard error or sensitivity analysis
• Ignoring disclosure rules when describing restricted administrative data
• Subgroup splits with no multiple-testing discipline presented as "heterogeneity"

Worked example: from elasticity to a welfare number (illustrative)

A DI-reform evaluation recovers a labor-supply response to a benefit cut, then builds the MVPF: the mechanical fiscal saving is the denominator; the behavioral fiscal externality (induced earnings → recovered taxes, minus crowd-out) adjusts the numerator, giving MVPF ≈ 0.8 (illustrative). The skill's norms then bind: state the primitives held fixed (no GE wage response, fixed program rules); propagate the elasticity's SE through the MVPF by the delta method, reporting a CI on the welfare object; and show how MVPF moves if the key elasticity sits at the high or low end of the literature. The welfare statistic with its uncertainty — not the bare elasticity — is the deliverable.

Calibration table: estimate → welfare object

| Estimated object | Welfare mapping | Watch for |

|------------------|-----------------|-----------|

| Taxable-income elasticity | Marginal DWL / optimal top rate | Mean reversion, income shifting |

| Take-up / benefit response | MVPF numerator + denominator | Crowd-out onto other programs |

Evidence pass for Journal of Public Economics

Treat this skill as an executable review pass, not a prose hint. First lock the policy instrument, affected margin, identification design, and welfare or incidence interpretation; then judge whether the current manuscript answers the venue's real reader: public economists who ask whether policy design, fiscal incidence, or welfare interpretation is credible.

• Do the pass: Audit the research design before polishing prose: unit of analysis, comparison set, uncertainty, sensitivity, missingness, and reproducibility must be visible.
• Return a ledger: give claim / evidence / risk / manuscript location rows, so the next agent can edit rather than rediscover the issue.
• Sibling guard: compare against JDE for development policy, JIE for cross-border policy, AEJ Economic Policy for broad policy readership; if a sibling owns the contribution, recommend re-routing before polishing format.
• Stop condition: do not give submission-ready advice until the pack's resources/official-source-map.md has been checked for volatile rules and the manuscript has one concrete fix for the largest venue-specific risk.

Output format

【Data】source + restricted? + disclosure handled? [Y/N]
【Policy parameter】elasticity / take-up / crowd-out / moral-hazard wedge
【Welfare mapping】DWL / MVPF / sufficient stat — SEs propagated? [Y/N]
【Measurement choices】[income def, top-coding, rule coding, ...]
【Heterogeneity】policy margin: [...]
【Robustness done】[bin/bandwidth/estimator/sensitivity, ...]
【Next step】jpube-tables-figures