Home affordability and after-tax income: why bank calculators understate burden in high-tax states

The 28/36 rule originated in FHA underwriting standards refined through the postwar period and codified in modern form during the 1980s thrift era. The choice of gross income over after-tax was deliberate. Gross is verifiable from W-2 line 1 or self-employment Schedule C; after-tax requires reading the full return, accounting for deductions and credits that vary year to year. Underwriters wanted a number they could pull from a single document and trust.

Conservatism was supposed to come from the 28% front-end ratio itself. In the 1980s, average effective federal-plus-state burden on a median household was around 22%. A buyer at 28% of gross paying housing was, in practice, paying about 36% of after-tax. That margin felt safe.

The convention has not aged well. Federal income tax brackets compressed and state burden bifurcated. Zero-income-tax states (TX, FL, NV, TN, WA, WY, SD, AK, NH partial) cluster at the low end. Coastal metros and the Northeast cluster at the high end, with NY, CA, and NJ at the top of the Tax Foundation 2026 ranking. The 28% gross front-end now means very different things in those two environments.

FHA went the other direction with its 31/43 cap, designed to expand homeownership access. Mathematically, 43% back-end on gross is roughly 60% of after-tax in California — past the threshold where most personal-finance frameworks consider housing sustainable. The FHA cap was not wrong when set, but the state-tax landscape it was set against has changed.

This is not a critique of underwriting. Underwriters need a number they can verify cheaply, and gross is that number. It is a mismatch between underwriter needs and borrower needs, and the post-tax adjustment is the borrower's responsibility, not the lender's.

The gap between gross and disposable income on $150k household income, single filer, 2026 federal brackets and FICA wage base ($176,100), with state-local burden from Tax Foundation 2026 data:

Tier 1 — Zero income tax states. TX, FL, NV, TN, WA, WY, SD, AK. Federal income tax effective rate on $150k after standard deduction sits around 17.5%. FICA adds 7.65% on the full amount. State-local burden runs 4-7% (mostly property tax and sales tax, with no state income component). Combined effective burden: 28-30%. After-tax disposable: roughly $105k-$108k of $150k gross.

Tier 2 — Middle tier. IL, PA, CO, OH, MI, IN, AZ, NC, GA, VA. Add state income tax of 3-5% to the Tier 1 baseline. State-local burden runs 8-11% combined. Combined effective burden: 32-34%. After-tax disposable: roughly $99k-$102k.

Tier 3 — High-tax metros. CA, NY, NJ, MA, OR, MN, HI. State income tax 6-13% (graduated, with top brackets in CA reaching 13.3%), high property tax (NJ averages 2.42%, highest in the US per Tax Foundation 2026), and city income tax in NYC adds 3-4%. State-local burden runs 13-16%. Combined effective burden: 38-40%. After-tax disposable: roughly $90k-$93k.

Numerical comparison — $150k gross household, conventional 28/36 DTI on gross:

• Maximum back-end housing + debt: $4,500/month
• TX (Tier 1): $4,500 ÷ ($108k ÷ 12) = 50% of after-tax
• IL (Tier 2): $4,500 ÷ ($102k ÷ 12) = 53% of after-tax
• CA (Tier 3): $4,500 ÷ ($93k ÷ 12) = 58% of after-tax

The same $4,500/month back-end that conforms to 36% of gross consumes 50% of after-tax in Texas but 58% in California. Same loan, same income — materially different lived experience.

A California buyer at 36% back-end is functionally equivalent to a Texas buyer paying 41% back-end measured on after-tax. Any sensible underwriter would flag 41% back-end as risky on the conventional rule. The gap explains why house-poor outcomes cluster geographically: borrowers in CA, NY, and NJ pass conforming DTI checks while sitting in functional 50%+ post-tax housing burden.

Tax Foundation's 2026 ranking is the canonical reference. New York reports 15.9% state-local burden as percent of personal income (highest in the US), New Jersey 49th overall on the State Tax Competitiveness Index, California 48th. Per capita state-local tax collections: NY $12,685, CA $10,319. These are not edge cases — they are the median experience for buyers in dense coastal labor markets.

The QuickUse home affordability calculator offers four DTI configuration modes:

• `conservative_28_36`: Fannie Mae traditional, 28% front / 36% back
• `standard_31_43`: FHA standard, 31% front / 43% back
• `lender_max`: whatever the underwriter currently approves, often 50% back-end with compensating factors
• `custom`: user-defined front_end_dti and back_end_dti

Standard advice is "stay conservative at 28/36." That advice is correct in Tier 1 states — aggressive everywhere else. Custom mode is where the post-tax adjustment lives.

Adjustment formula. Subtract a state-tier discount from the front-end cap. The exact magnitude depends on the gap between your state-local burden and the national 1980s baseline that 28/36 was calibrated against (about 22% combined effective burden):

• Tier 1 (~28-30% combined burden): no adjustment needed. 28% front-end is correct.
• Tier 2 (~32-34% combined burden): subtract 4 percentage points. Target 24% front-end.
• Tier 3 (~38-40% combined burden): subtract 6-8 percentage points. Target 20-22% front-end.

Back-end follows: keep an 8-point gap above front-end for non-housing debt. So Tier 3 buyers target 20% front / 28% back. Tier 2 buyers target 24% front / 32% back.

Calculator setup, California example. Buyer with $150k gross household income in CA, 720 FICO, $80k down. Default `conservative_28_36` setting outputs an affordable home price around $620k at current conventional rates. Switching to `custom` mode with front_end_dti = 22 and back_end_dti = 30 drops the affordable price to roughly $470k — a 24% reduction. The lower number is the post-tax sustainable price for that household. The gap between the two ($150k of purchase capacity) is exactly the after-tax adjustment that the default rule missed.

The 24% reduction sounds painful — it is also correct. The same household in Texas, running the same calculator with default 28/36, would land near $620k and the post-tax math would actually support that price because the after-tax disposable margin is wider. The calculator is doing its job; the convention is doing its job. The post-tax adjustment is the layer between them that the borrower has to add.

Anchor to actual outcome: the calculator's `closing_cash.user_has_enough` and `reserves.shortfall` outputs become more accurate when the front-end is calibrated to after-tax. A buyer who passes default 28/36 but fails custom 22/30 is exactly the buyer who ends up house-poor in 18-24 months when the property tax bill, insurance escrow shortfall, or roof replacement hits.

Three concrete scenarios where custom DTI changes the buying decision, not just the math.

Scenario 1: the lender pre-approval gap. Lender pre-approves $850k for a CA buyer at 36% back-end on gross. Custom 30% back-end (after-tax adjusted) gives $710k. The $140k gap is real. The lender's job is loan performance over the loan's expected life; the borrower's job is monthly survival without lifestyle compression — these are not the same target. Walk into the negotiation knowing your custom number, not the lender's max.

Scenario 2: dual-income with asymmetric tax burden. Spouse A residing in NJ paying combined 38% effective burden, Spouse B remote-working for an out-of-state company but residing in NJ. NJ taxes worldwide income for residents, so the underwriter sees combined gross while reality is symmetric burden against the worse state. Custom DTI must reflect residence-state taxation regardless of where the employer is. The remote-work tax fiction (employer state matters more than residence state) does not survive contact with state revenue departments.

Scenario 3: high-property-tax counties hidden inside a "manageable" state. Westchester NY, Bergen NJ, parts of Travis and Williamson counties in TX have effective property tax rates of 2-3%. The calculator's `property_tax_rate_annual_override` lets buyers model exact county rate. PITI on a $700k home in Bergen at 2.5% property tax = $1,458/month in property tax alone, before any P&I or insurance. That is 17% of $100k after-tax income just for the property tax line.

In all three scenarios, the calculator does the math correctly when given correct inputs. The post's job is teaching which inputs are correct given the buyer's actual geographic and tax reality. Conventional advice ("stay at 28/36") generalizes badly across the state burden distribution.

Several legitimate moves narrow the gross-vs-take-home gap without breaking any rules. None of them eliminate the gap entirely.

401(k) contributions. 2026 limit is $23,500 under age 50, $31,000 with catch-up at 50+. Maxing 401(k) on $150k gross drops federal-plus-state taxable income by $23,500 directly. Effective burden on remaining cash drops 2-4 percentage points depending on bracket. The tradeoff: 401(k) contribution is also unavailable for housing during the contribution year. It reduces both the burden and the disposable cash applied to mortgage.

HSA contributions. 2026 limits are $4,300 single / $8,550 family for HDHP-covered borrowers. Triple tax advantage (deductible at federal and most state levels, growth tax-free, withdrawal tax-free for qualified medical expenses). Maxing HSA reduces taxable income while preserving liquid medical reserve. Worth modeling for buyers on HDHP plans.

Mortgage interest deduction. Federal deduction up to $750k loan principal under TCJA. At 7% APR on a $750k loan, year-1 interest runs about $52,500. Federal deduction at the 24% bracket recovers roughly $12,600 in tax savings. State conformance varies: California conforms partially, several others do not. The state tax piece of the deduction is bounded by the SALT cap.

SALT cap. $10,000 federal deduction limit on combined state and local taxes (income + property), still in effect for 2026 under TCJA permanence. In CA, NY, and NJ, this means the bulk of state-local burden is non-deductible federally. The post-tax adjustment formula does not net out state tax deduction in high-tax states because the SALT cap eliminates most of the federal benefit. A Bergen NJ buyer paying $20,000 annual property tax + $15,000 state income tax can deduct only $10,000 of the $35,000 combined, with $25,000 staying as full federal-taxable income.

Net effect of stacking these legitimate reducers: a CA buyer maxing 401(k) and HSA, claiming mortgage interest deduction, still has effective combined burden around 32% versus 38-40% baseline without optimization. The custom front-end of 22-24% remains the right target — the optimizations narrow the gap, they do not close it.

How to drive the home affordability calculator with the after-tax adjustment baked in:

1. Input gross household income. The calculator requires gross. Do not pre-adjust to after-tax.

2. Set `state` (US 2-letter) for 50-state property tax lookup, or override with `property_tax_rate_annual_override` for a specific county rate (Bergen, Westchester, Travis, etc.).

3. Set `loan_type` to conventional, FHA, VA, or USDA based on your eligibility. This drives default DTI baseline if you stay on a non-custom mode.

4. Switch `dti_target` to `custom`.

5. Enter `custom_front_end_dti` based on state tier:

• Tier 1 states (TX, FL, NV, TN, WA, WY, SD, AK): 28%
• Tier 2 states (IL, PA, CO, OH, MI, IN, AZ, NC, GA, VA): 22-24%
• Tier 3 states (CA, NY, NJ, MA, OR, MN, HI): 20-22%

6. Enter `custom_back_end_dti` = front_end + 8. This preserves an 8-point gap for non-housing debt service (auto loans, student loans, credit card minimums).

7. Set `reserves_months_piti` based on state tier: 6 in Tier 3, 4 in Tier 2, 3 in Tier 1. Higher reserves in high-tax states reflect thinner post-tax buffer.

8. Read the output `recommended.max_home_price` and compare against your lender's pre-approval. The gap is your sustainability buffer.

9. Verify `closing_cash.user_has_enough` returns true. If false, the down payment + closing + reserves combination is too aggressive even at the adjusted price.

For modeling exact PITI at a specific rate and term, use the loan calculator. For comparing buy versus continued renting, the rent-vs-buy calculator runs the break-even analysis with the same after-tax frame.