Methods

Made by AI agents, with a person directing and checking the work.

How this was made

AI agents found and downloaded the public data, ran the analysis, made the charts and wrote the text. A person set the questions, directed the work and checked the results. Every number traces to a listed source file, automated checks run before each build, and corrections are logged in Amendments. No surveys, interviews or site visits were done.

Approach

Energy burden is total home energy spending divided by total income for a group of homes, as in the DOE LEAD Tool. We used LEAD's 2022 Kansas census-tract file, which splits each tract's homes by income (as a share of the federal poverty level), owner or renter, building age, building type and main heating fuel. Program facts come from the agencies that run the programs. Scenario costs multiply published per-home figures by local counts.

Automated checks (latest run)

[PASS] 1 blanks: tract_table                                     27 rows (expect 27); blank: []; dup: []
[PASS] 1 blanks: out/scenarios.csv                               4 rows
[PASS] 1 blanks: out/lead_burden_by_fpl.csv                      5 rows
[PASS] 1 blanks: out/lowincome_burden_by_area.csv                3 rows
[PASS] 1 blanks: out/summary_numbers.csv                         13 rows
[PASS] 2 totals: LEAD tracts vs LEAD county file                 tracts 49,605 vs county 49,605
[PASS] 2 totals: tract_table vs raw                              49,605 vs 49,605 (rounding tolerance 27)
[PASS] 2 totals: LEAD 2022 vs ACS 2023 households                LEAD 49,605 vs ACS 49,939 (-0.7%, tolerance 3%, different survey years)
[PASS] 2 totals: summary households                              49,605
[PASS] 3 reality: LEAD elec vs EIA-861 EKC 2022 bill             LEAD Douglas $1,262/yr vs EKC $1,550/yr, ratio 0.81 (tolerance 0.75-1.25) FLAG: >15% gap, noted in report limits
[PASS] 3 reality: KS LIEAP reach vs national 17.2% (Urban 2023)  22.8% vs 17.2% (tolerance 10 pts)
[PASS] 4 assumptions.md                                          all items present (the author to review)
[PASS] 5 MOE/flags                                               MOE present for all ACS estimates; 13/27 tracts flagged
[PASS] 6 cleanup                                                 no 0-byte files; 29 raw files read-only
[PASS] 7 provenance                                              all data files in sources.md; all outputs have sidecars
[PASS] 8 history timeline                                        timeline vs newest out file out/scenarios.csv.meta.txt
[PASS] 9 citations                                               15 cited sources all in sources.md; unused: []
[PASS] 10 site                                                   placeholders []; AI note on every page: True

Assumptions

Data

• Energy burden = sum of annual energy spending (electricity + gas + other fuel) / sum of household income, over LEAD cells (DOE LEAD method). LEAD 2022 update = ACS 2018-2022 5-year PUMS, calibrated to EIA-861/EIA-176 2022.
• Income bands use the LEAD FPL file (0-100, 100-150, 150-200, 200-400, 400%+ of the federal poverty level). LIEAP eligibility = at or below 150% FPL (KS DCF).
• LEAD publishes no margins of error. Rule: a tract value based on fewer than 300 low-income (<200% FPL) LEAD households is flagged few_lowincome_hh and never headlined.
• ACS context columns carry ACS 90% margins of error; median income with CV (MOE/1.645/estimate) > 0.3 is flagged income_moe_gt_cv30 and not headlined.
• Student-heavy tract = undergraduates (ACS B14001_008) >= 30% of population 3+. Flag only; values kept.
• Area groups: a tract is "Lawrence" if >= 80% of its land area is in the Lawrence place boundary (TIGER 2023), "Lawrence edge and rural" if 15-80%, else "Rural".
• Count of households above 6% uses LEAD cell averages (a cell is income band x tenure x building age x building type x heating fuel in a tract); it is an approximation of the household distribution.

Program figures

• LIEAP reach in Kansas = 43,000 households served in 2025 (DCF) / households at or below 150% FPL in LEAD 2022 counties file (188,328). Mixed years; labelled approximate.
• LIEAP average benefit $680 (DCF, 2025). Weatherization cost per home $7,475 (KHRC 2025 plan average) to $8,550 (DOE PY2025 average cap). Savings $283/yr per home from the 2008 retrospective evaluation (nominal, not inflated).

Scenarios

• A: +1,000 LIEAP households x $680. B: 6% percentage-of-income cap for <=150% FPL cells; yearly cost = sum of max(0, energy - 6% x income), +/-20% band; excludes admin cost. C/D: weatherization per 100 homes and for all owner households <=200% FPL (2,460).
• Scenarios are illustrations, not recommendations. No field work, surveys or interviews.

Dates

• Data retrieved 2026-10-25. Analysis period: LEAD 2022, ACS 2023 5-year (2019-2023), program years 2025-2026.