Reconciling Two Offers to One Comparable Take-Home Number

You have two written offers. They differ on base, bonus, equity, city, and state, and the answer to "which one pays more" changes depending on which of those you look at first. This guide carries one worked case through every adjustment - bonus probability, equity vesting, benefits value, cost-of-living, and state tax - and lands on two comparable annual take-home numbers you can defend to a spouse, a recruiter, or yourself at 2 a.m. It is for anyone holding or expecting an offer who needs to answer before the clock runs out.

Most pages you will find do one of two things: hand you a blank weighted-scoring matrix and tell you to trust your gut, or run a cost-of-living calculator that stops at pre-tax salary. Neither shows the arithmetic where the answer actually flips. This one exposes the four naive comparisons that reverse the winner, and shows the intermediate numbers at every fork, including the wrong turns.

The case: two offers that look tied on paper

Here is the worked example carried through the whole guide. Offer A is in New York; Offer B is in a no-income-tax city. On the headline base, they are close, which is exactly the situation where naive comparisons mislead.

On base alone, A wins by $10,000. On headline equity, B wins by $80,000. Both readings are wrong, and they point in opposite directions. The rest of this guide converts these raw numbers into two figures that mean the same thing.

The order of operations matters and is genuinely contested. Some sources apply the cost-of-living index first to produce an "equivalent salary," then tax it. I apply tax first, then the cost index, because tax is levied on nominal income - that is the order money actually leaves your paycheck. Whichever you choose, write it down.

Discount each bonus to what it actually pays

A target bonus is not income until it pays out, so book it at expected value: target multiplied by your honest probability of payout. This single move corrects the most common way two offers appear tied when they are not.

The dossier's rule of thumb: a $20,000 target bonus with a 70% likelihood adds $14,000 to expected annual earnings. Apply your own probability, informed by what the recruiter says about historical payout, not the number in the letter.

For the case, suppose you judge Offer A's bonus pays out 80% of years and Offer B's pays out 90% (smaller, more reliable):

• Offer A: $24,000 x 0.80 = $19,200 expected
• Offer B: $15,000 x 0.90 = $13,500 expected

Signing bonuses are one-time. Offer A's $10,000 signing bonus is real money, but it does not belong in the annual comparison. If you want to account for it, amortize it over your expected tenure (say, $10,000 over 3 years is $3,333/year) and label it clearly, or set it aside as a first-year sweetener.

Annualize equity, then probability-weight the cliff

Divide the total grant by the vesting term to get an annual figure, then treat the year-one portion as conditional, because the standard schedule pays nothing if you leave before the cliff. Equity is the line item where the headline number lies most.

The dominant schedule is four years with a one-year cliff: 0% for the first twelve months, 25% at month twelve, then 1/48th (2.08%) each month after. The load-bearing fact in the whole guide is this one: eleven months and thirty days versus twelve months and zero days equals $0 versus 25% of your grant. Year-one equity is not guaranteed income. It is a bet on your own tenure.

For the case:

• Offer A: $80,000 / 4 = $20,000/year naive. Year-one slice is $20,000, at risk until the cliff.
• Offer B: $160,000 / 4 = $40,000/year naive. Year-one slice is $40,000, at risk until the cliff.

Now apply the cliff discount. If you are highly confident you will stay past year one at both companies (say 90%), the year-one figures become $18,000 and $36,000. If you think there is a real chance you leave Offer B early - a startup, a role you are unsure about - drop that probability to 60%, and Offer B's year-one equity falls to $24,000. That is the insight the headline hides: a smaller grant with certain tenure can out-pay a larger grant you might walk away from before it vests.

For public-company RSUs, apply an additional volatility discount of typically 70 to 80% of the current price, because the shares can fall before you sell. For this case I will keep both grants at face value but note that if Offer B were a private startup, its equity carries liquidity risk that face value ignores entirely.

Value the benefits the base number hides

Three benefits convert cleanly to dollars, and together they realistically move total compensation by about 30%. An offer that wins on base by 5% can lose once you price them in, because employer-paid benefits are untaxed value the salary line never shows.

The aggregate anchor from the dossier: across the economy, wages and salaries average $32.60 per hour with benefit costs averaging $14.01, roughly 30% of total compensation. You are not pricing all of that - much of it is legally mandated and identical across offers - but the discretionary slice is real.

For the case, assume Offer A matches 50% of the first 6% and Offer B matches 100% of the first 4%:

• Offer A match: 0.03 x $160,000 = $4,800
• Offer B match: 0.04 x $150,000 = $6,000

Health premium: suppose both employers pay 80% of a $9,000 annual premium, so each contributes about $7,200. This is roughly equal, so it will not move the ranking - but if one paid 70% and the other 90%, that gap is worth well over a thousand dollars and belongs in the total.

PTO is where people double-count. Adding the full dollar value of PTO on top of a salary that already assumes paid weeks off books the same money twice. The clean move is to compare only the difference in PTO between the offers. If Offer A gives 15 days and Offer B gives 25, the 10-day difference on a $150,000 salary is worth about $5,760 to Offer B - and that is all you should count.

Sum, tax, and adjust for where the money is spent

Sum base plus expected bonus plus annualized equity plus benefits to get gross total comp, then apply state tax, then divide by the cost-of-living index. Each of the last two steps can flip the answer on its own.

Here is the case summed to pre-adjustment totals (year-one equity discounted at 90% for both, benefits counting match and the PTO difference of $5,760 to A which we will say gives more PTO):

On gross total comp, the offers are within about 1%. This is the moment the naive reader stops and calls it a wash. It is not a wash, because A's dollars are earned in a high-tax, high-cost city and B's are not.

State tax, and why you must name the source

Sources disagree on tax burden, so you must state which you used. Tax Foundation (2022) ranks New York highest at 15.9% and Alaska lowest at 4.6%; WalletHub (2026) ranks Hawaii highest at about 13% and Alaska lowest at 4.9%. Seven states levy no income tax: Alaska, Florida, Nevada, South Dakota, Tennessee, Texas, and Wyoming.

The methodologies differ - one measures collections over income, the other models a household - so an offer memo that cites one without the other is unfalsifiable. Note too that a no-income-tax state is not automatically cheaper: it can recoup revenue through property and sales tax, so compare total state and local burden, not the income-tax line alone.

For the case, estimate combined federal, state, and local effective rates. Suppose Offer A in New York lands at a 34% effective rate and Offer B in the no-income-tax city at 26% (federal only, plus sales and property baked into cost of living):

• Offer A after tax: $207,760 x (1 - 0.34) = $137,122
• Offer B after tax: $205,500 x (1 - 0.26) = $152,070

The 1% gross tie has become a 10% after-tax lead for Offer B. This is the "which job offer pays more after tax" step that COL-only calculators skip entirely.

Cost of living, weighted to your real housing share

Now divide each after-tax number by the city's cost index against a baseline of 100. The dominant named source is the C2ER Cost of Living Index (formerly ACCRA), published quarterly since 1968, built from nearly 100,000 data points across 400 cities, where the average of all participating places equals 100. Online city comparisons are fee-based at $7.95; the BEA Regional Price Parities are a free alternative, also indexed to 100.

Housing is where these adjustments live or die. It is almost always the widest-varying category: NYC's housing index is 317, over three times the national average, while Memphis is 62. Food, by contrast, ranges only 90 to 120. The national housing share of spend is about 33%, but renters in expensive cities often spend 45 to 55%. If you use the blended composite index at the national housing weight while paying half your income in rent, you understate how badly a high-rent city hits you.

For the case, suppose Offer A's city carries a composite index of 187 but you would spend 50% of income on housing, so you re-weight the index upward to an effective 210. Offer B's city carries an index of 98. Purchasing-power-equivalent take-home:

• Offer A: $137,122 / 2.10 = $65,296 in national-baseline dollars
• Offer B: $152,070 / 0.98 = $155,173 in national-baseline dollars

The gap is now enormous, and it is driven almost entirely by New York housing. Offer A never recovers from the combination of high tax and a housing-heavy cost index. The base-only reader who took Offer A for the extra $10,000 would have been wrong by a factor that dwarfs the base difference.

Run the procedure on your own case

Follow these eight steps in order on your two real offers. Each produces a named intermediate number, so you can show your work and find where a colleague's estimate differs from yours.

Refolk can shortcut the front of this. When it writes and tailors your application from your own history, it already holds the base, level, and metro for every role you applied to, so Refolk can line up your live offers against what comparable roles pay before you start filling in probabilities.

How this goes wrong: the four flips and four more

Eight failure modes reverse or distort the answer, and four of them will flip the winner outright. Re-run your comparison under each before you sign.

• Base-only comparison. Picks the higher salary and ignores equity, bonus, benefits, and tax. A $160k base "beats" a $150k base carrying $40k of vesting equity. Check: did you sum all four components before ranking?
• Headline-equity trap. Counts the full multi-year grant as year-one income. A "$200k equity" offer looks like +$200k/year when it is $50k/year over four years and $0 before the cliff. Check: divide by the term, flag the cliff separately.
• Full-vest-assumed. Treats year-one equity as guaranteed when day 364 versus day 365 is $0 versus 25% of the grant. Check: multiply year-one equity by your honest probability of staying twelve months.
• Index-without-tax. Uses a COL calculator that stops at pre-tax salary, making a no-income-tax city look equal to a high-tax one. Check: apply state tax before or alongside the cost index, never skip it.
• Wrong tax source. Quotes one ranking as fact when sources diverge; Tax Foundation puts NY highest, WalletHub puts Hawaii highest. Check: name the source and year in the offer memo.
• Housing weight left at the national average. Uses 33% housing when you spend 50%, making a high-cost destination look affordable. Check: set housing share to your real rent-to-income ratio.
• Bonus at target, not expected. Books a 100%-target bonus every year. Two offers tie on paper but one pays out 60% of years. Check: multiply target by realized-payout probability.
• PTO double-counting. Adds PTO dollar value on top of a salary that already assumes paid weeks off. Check: exclude PTO, or compare only the difference between offers.

The failure that surprises people most is the seniority one. If your two offers are actually different levels - say a Senior role at one company and a Staff role at another - the ladder step and its equity-refresh cadence may be a larger driver than any city adjustment. In Refolk's index there are 5.6 Senior Software Engineers for every Staff Software Engineer in the US, and the two bands concentrate in different metros, New York versus San Francisco. If you normalize cities first, you can hide that you are comparing two different rungs.

number: 5.6
label: Senior Software Engineers per Staff Software Engineer in the US
note: From Refolk's index. If your two offers are different levels, the rung difference can outweigh the city difference.