§0 · What's already been said, and what we're adding

Wage-timing and high-cost credit have substantial academic literature. Stephens (2003, "3rd of tha Month") documented consumption-cycle effects of pay frequency. Shapiro (2005) showed related effects in food-assistance recipients. Zeldes (1989) and Jappelli & Pistaferri (2010) frame liquidity constraints in consumption decisions. Melzer (2011), Skiba & Tobacman, and Morse have published on welfare effects of payday lending access. The CFPB's Making Ends Meet series provides canonical descriptive evidence on household liquidity. The CFPB has also published extensively on payday lending, overdraft fees, and BNPL — all of which appear in §5 below.

What this investigation adds is twofold. First, a per-paycheck discounted-cash-flow analysis with a small editorial credit spread, applied at the individual worker level rather than aggregate. Second, a parallel bonus-cadence analysis that distinguishes the legal instrument (creditor claim vs. expectation) from the structurally-similar cash-flow extraction. The bonus case in particular has been less well-explored in the literature; the §3 contingent- creditor analysis is, as far as we can tell, original to this essay.

The Payroll Lag

If you get paid every two weeks, you are extending your employer an interest-free loan worth roughly one week of your salary. Continuously. The loan never pays down — every payday it resets to the same balance and starts accruing again. You did not sign anything to agree to this. The HR teams that administer it rarely frame it as a loan. It is one of the most universal invisible financial transactions in American economic life — touching nearly every worker, every pay period — and the convention that produces it is older than the technology that would make it unnecessary.

Aggregate US figure at current rates: $20.7B of foregone interest per year, flowing from workers to employers as working capital. This is the direct number. It is not the largest number in this story.

§2 · The mechanism

You earn your wages continuously — every minute on the clock adds to what you're owed. Your employer pays you in arrears, on a schedule. The gap between "earned" and "paid" is float. On any given day, the average balance of unpaid-but-earned wages is half a pay period.

• Weekly cadence: ~3.5 days of your salary, on average, sitting unpaid.
• Biweekly: ~7 days.
• Semi-monthly: ~7.6 days.
• Monthly: ~15 days.

Frequency is the knob. Monthly pay floats roughly 4× what weekly pay does, for the identical job. Your hourly rate, your title, your benefits — none of those change. The cadence of when the bank deposit hits is set by an HR convention that nobody, on either side, treats as a financial decision.

§3 · What this instrument actually is

Let me name it precisely. You are an unsecured creditor of your employer for the value of those earned-but-unpaid wages, compensated at 0%.

That phrase deserves to be unpacked.

Unsecured.

If your employer goes bankrupt mid-pay-period, you stand in line with the other unsecured claims for those wages. In a Chapter 7 you are a priority unsecured creditor up to a statutory cap (currently $17,150 per 11 U.S.C. §507(a)(4), adjusted by the Judicial Conference every three years for CPI; was $15,150 prior to the April 1, 2025 adjustment), which is better than ordinary trade debt but worse than a secured lender. You can lose part or all of it.

Demand loan.

The balance is due to you in full on payday. You can't accelerate it. You can't refinance it. You have no covenants, no collateral, no rate.

Comp'd at 0%.

You receive no interest, no fee, no equity, no consideration of any kind for extending this credit. As of June 2026, real rates are modestly positive (TIPS 10-year around 2%), so foregone interest costs the worker a slightly negative real return. In the real-rate-negative environment that characterized roughly 2009–2022, the lag was materially worse — float lost ground to inflation faster than the (near-zero) compensation.

This is not a "zero-coupon" instrument — a zero-coupon bond yields its discount at maturity. Yours doesn't yield anything. It is the rare financial position that combines credit risk, inflation risk, and opportunity cost without compensation for any of them.

The mirror image is your employer's position. They are being financed at 0% for working capital they would otherwise have to obtain from credit markets — and that cost is meaningfully above the risk-free rate. As a directional proxy, the ICE BofA US Investment Grade Corporate OAS (FRED BAMLC0A0CM) sat around 0.76% as of May 2026; the High-Yield Corporate OAS (FRED BAMLH0A0HYM2) typically runs 300–500 bps. These are secondary-market bond spreads, not bank revolver pricing or commercial paper rates — actual financing costs vary by firm size, collateral, and bank relationship, with smaller private firms paying meaningfully more. The directional point is robust to the proxy choice: the same dollar of float a worker extends at 0% is a dollar an investment-grade firm would pay roughly ~4.4% to replace at market (risk-free + spread), the high-yield firm roughly ~7-9%, and a small private firm without bond-market access more still.

One sharpening that matters: in the paycheck case, the receivable is a legal entitlement — earned wages are owed throughout the period under federal and state wage law, and you are an unsecured creditor of your employer for every hour worked but unpaid.

In the bonus case (the calculator's sibling mode), your creditor status is contingent on three things: whether the bonus is structured as discretionary (employer's pure judgment) or earned (tied to objective performance criteria); the reason for separation (voluntary, for cause, or without cause); and state law. In many states for many bonus structures, you have no enforceable claim during the accrual year and no claim if you leave or are dismissed before payout. California treats earned bonuses as wages regardless of termination cause — AB 692 (effective Jan 1, 2026) further restricts stay-or-pay retention clauses on involuntary separation. New York is moving in that direction (S6775, 2025–26). For the worst-case slice — discretionary bonus, voluntary departure, weak-protection state — you are the holder of an expectation, not a creditor. The lag-tax math runs on the brief post-vest window (typically ~60 days) and the number is genuinely small. But the extraction during the accrual year, in the worst-case slice, is not interest you are owed — it is interest you cannot even charge for, because you have no legal standing to. The employer still accrues the bonus liability on the balance sheet throughout the year and uses it as working capital. The §7 asymmetry sharpens: in the paycheck case they price what you have a claim to and underpay; in the bonus case (worst-case slice) they price what you have no claim to at all. The calculator's bonus-mode mirror callout makes this concrete.

§5 · Who actually pays for this

When I ran the numbers, the surprise wasn't the foregone interest. It was what the foregone interest isn't — and what infrequent pay leaves workers without buffer reaching for instead.

Pay cadence sets the size of the cash-flow gap between income deposits. Rent is due on the 1st; your paycheck arrives on the 5th. Your kid's prescription is $80 today; payday is Thursday. A worker with three months of savings absorbs the gap from a buffer. A worker without one borrows.

The credit they reach for is not priced at 5%.

• Payday loans: typically ~391% APR per the Consumer Federation of America methodology (Pew Charitable Trusts publishes parallel state-by-state data); states without rate caps (Texas, Nevada, and others) commonly see effective APRs of 600%+. ~$2.4 billion in fees per year drained from ~12 million borrowers in the ~26 states where the product remains legal (Center for Responsible Lending, Down the Drain, February 2025). Average $15-per-$100 two-week fee structure compounds to the APR.
• Overdraft / NSF fees: ~$12.1 billion in 2024 across US banks + credit unions (Financial Health Network; CFPB 2024 data spotlight). Dropped to $5.83B in 2023 when several large banks voluntarily cut fees, then rebounded after the CFPB overdraft rule was repealed by Congress in 2025.
• Buy-now-pay-later: 53.6 million users in 2023, $45.2 billion in originations (CFPB BNPL Market Report, December 2025). Pay-in-four BNPL is nominally interest-free when paid on time — late fees are small (~0.18% of GMV in 2023), so this is not "fees" in the same predatory sense as payday or overdraft. The harm vector is debt accumulation as workers use BNPL as bridge financing for the same cash-flow gaps that drive the other two.

These are the products the market has built to bridge cash-flow gaps. The empirical question — how much of their use is attributable to pay cadence specifically rather than low wages in general — has been worked on for a decade in the academic literature; this essay's argument turns on it, and Methodology cites where the evidence stands. The directional finding is consistent: longer pay cycles compound cash-flow stress for workers without buffer, and workers under cash-flow stress reach for high-cost credit.

The float you extend to your employer costs you near-zero in foregone interest. The bridge credit you reach for costs you 300%+. The payroll lag tax is not the interest you didn't earn. It is the payday loan you took to make rent because payday was Friday and rent was due Tuesday.

The effect compounds on both ends. ADP Research Institute data — which covers payroll for more than 26 million US workers across 500,000+ employers (~1 in 6 workers nationally; ~1 in 5 in the private sector), the broadest sample available — finds 48% of US workers paid biweekly (the modal cadence) and 43% saying they are unhappy with the frequency of their pay (ADP Research, More Paydays, More Fairness, HR Experience Survey, 20,000 working adults, June 2022 – June 2023). The dominant preference among dissatisfied workers is more frequent pay — especially among lower-income workers, women, and younger workers, who report living paycheck-to- paycheck more often. Industry splits track expected: construction, trade, and manufacturing skew weekly (NY Labor §191 already mandates it for "manual workers"); salaried sectors skew biweekly or monthly. Lower-wage workers are also more likely to use payday and overdraft credit (CFPB Making Ends Meet in 2024, November 2024; CFPB Developments in the Paycheck Advance Market, 2024: of earned-wage-access users surveyed, the majority fell under federal poverty guidelines and over 80% were hourly or gig workers, paying effective ~109.5% APR despite the product being marketed as "free or low-cost"). The cadence convention that produces the smallest direct float per person sits alongside the largest induced borrowing per person — and a material share of the workers exposed have told the largest payroll processor in the country they want it changed.

The bonus-cadence version has a different shape but the same logic — the convention produces a behavior whose cost dwarfs the float itself. Bonus-eligible workers do not typically bridge cash-flow gaps with payday loans, but they do stay in jobs they want to leave because the bonus cliff is months away (the retention trap), and — depending on bonus structure, separation cause, and state — they may bear the forfeiture if their employer ends the relationship before payout. A worker laid off in October has worked nine months for a bonus the firm captures, at least under the most common structures (discretionary, with a "must be employed on payout date" clause) and in states with weaker protections. California already treats earned bonuses as wages owed regardless of termination cause; AB 692 (effective Jan 1, 2026) restricts stay-or-pay retention clauses on involuntary separation. New York is moving via S6775 (2025–26). Outside those jurisdictions and for discretionary structures, the legal framing says it was not owed; the economic substance is harder to defend. Different population, different harm mechanism, same structural extraction. (The §8 levers below address the paycheck side; the bonus side has its own stack.)

§7 · The asymmetry

On the employer side, accrued wages are a well-understood corporate finance line item. They appear on the balance sheet as a current liability. CFOs treat them as spontaneous, interest-free working capital — funding that scales automatically with the size of the workforce, costs nothing to obtain, never has to be rolled over, and substitutes one-for-one for capital that would otherwise be priced at the firm's marginal short-term borrowing rate. Textbooks teach this. It is not hidden from anyone with a finance degree.

On the worker side, this same balance — the same dollars, the same convention — has no name. There is no line on your paystub for "wages you have earned but I haven't paid you yet." There is no employer disclosure of "average float we held from your paycheck this quarter." There is no consumer-finance education category that names this instrument or computes its cost. The CFPB and FINRA publish on payday lending, overdraft, BNPL, and budgeting — all downstream of the lag — but none treat the lag itself as a discrete financial position the worker holds.

The same fact is taught as working-capital strategy on one side and is invisible on the other. This is a structural information asymmetry in the formal economic sense (Akerlof, The Market for Lemons, 1970), produced because employers who shroud the actuarial price out-compete employers who reveal it (Gabaix & Laibson, Shrouded Attributes, 2006). The wage-lag case is a recurring-relationship instance of the same mechanism.

Both ends of this asymmetry sharpen in the bonus case. The accrued liability is named explicitly as a retention tool on the firm side — not just working capital but a behavior-shaping mechanism. The same dollars that are "accrued comp expense" on the firm's books are "maybe" on the worker's mental balance sheet. They price what you have a claim to and underpay; they also price what you have no claim to and use it.

They price it. You can't. They've heard you ask. They've chosen not to deliver.

§8 · The case for change

The infrastructure to settle payroll daily — or per shift, or continuously — exists today. Cheap deserves precision.

The strongest honest argument for the lag is cash-flow matching. Many businesses do not receive revenue smoothly. A restaurant has lumpy nightly cash; a construction firm gets paid on 30–60–90-day milestone draws; a card-heavy retailer waits 2–3 days for processor settlement; a seasonal business is cash-rich half the year and cash-poor the other half; a pre-revenue startup has no inflows at all. Payroll, by contrast, is a continuous obligation — workers earn every hour. The lag between earning and payment is, for these firms, a working-capital buffer that smooths revenue volatility against payroll commitments. Without it, employers in this bucket would need to hold more cash reserves or borrow short-term capital at market rates. The lag is genuinely doing economic work for them.

The weaker, more commonly cited argument is operational complexity. Modern payroll isn't a single transaction. It is a stack: pre-tax deductions (401(k), HSA, FSA), insurance premiums, tax withholding across federal, state, and sometimes local jurisdictions, garnishments, leave accrual, multi-state compliance, end-of-cycle filings. None of that has simplified since 1985. Labor and G&A are usually meaningful line items on the income statement, and a CFO has real reasons to be cautious about doubling pay-cycle frequency on paper.

Both arguments have limits. The operational-complexity case is largely platform-amortized for employers at scale — at any employer running a modern provider (ADP, Workday, Gusto, Paychex, Rippling) the per-cycle work is largely automated, and the HR team's time does not scale proportionally with cycle frequency (qualitative observation; the "per-cycle labor cost is fixed per worker per year" claim is not measured to a published benchmark). The marginal additional cost of going from biweekly to weekly is primarily compute and per-employee provider fees, not 2× the HR team's time. The HR-time lever that mattered in 1985 is meaningfully reduced in 2026 for employers at scale.

The cash-flow-matching case is harder to dismiss, and this essay should not dismiss it. What it does not justify is paying the worker zero for the financing. A bank would not lend that working capital at 0%. A factoring company would not buy a firm's receivables at face value. The lag is genuinely useful to the employer in the cash-flow-matching case — but the legitimacy of the use does not transfer to the legitimacy of the pricing. The worker is providing real, economically valuable financing and getting nothing in return.

The honest framing is therefore heterogeneous. For a local daycare collecting tuition on the 1st of each month — smooth monthly revenue, predictable cash flow — the payroll lag is pure rent extraction with no working-capital justification. For a small construction firm waiting on a quarterly milestone draw (an illustrative archetype; NAICS data on construction-industry AR aging documents the pattern), the lag is real working capital, but it is being financed at 0% by people who never agreed to that price. One business needs the float; the other just captures it. The lag is a one-size-fits-all convention; the economic justification for it varies enormously. Policy should discriminate accordingly.

The remaining real operational cost is concentrated in three places: small employers without modern platforms, per-cycle provider fees (where the platform vendor extracts the public-rail savings as private-rail rent), and multi-jurisdiction filing edge cases. The policy levers below engage with each.

The per-transaction friction that originally justified longer pay cycles — paper checks, manual reconciliation, ACH cost, batch-processing windows — disappeared in stages through the late 20th century. The default never updated. The convention that exists today is one the conditions that produced it stopped producing decades ago, and nobody on either side re-opened the negotiation. Here is what re-opening looks like.

A note on what these levers fix. Five address the paycheck-cadence lag — wages owed under wage law, paid in arrears at employer-set cadence. The sixth (working-capital backstop) engages the cash-flow-matching steelman directly by providing an alternative financing source, so the worker contribution stops being load-bearing for employers with legitimate revenue volatility. The bonus-cadence stack is different because the underlying instrument can be different (expectation rather than creditor claim, depending on structure and state). The bonus-side levers worth surfacing in a parallel conversation: mandatory interest accrual once a bonus is approved (post-approval, pre-payout); pro-rata vesting on involuntary separation, so a mid-year layoff does not let the firm capture the worker's accrued labor at zero; bonus-pool variability disclosure at firm level (year-over-year and percentage-of-revenue/profit) so workers can actually price the expectation; and treatment of involuntary-separation forfeitures as wage theft — California is partly there already (AB 692 restricts stay-or-pay retention agreements starting January 1, 2026; CA labor law already treats earned bonuses as wages regardless of termination cause), and New York's S6775 would extend wage-law protection to earned bonus compensation. Broader state adoption is the natural next move. Same math, different problem, different solutions. The full bonus case is its own investigation.

§9 · Methodology + sources

Full per-pay-period DCF, term-structure Treasury curve, and mean-reverting inflation projection are documented on the methodology page. Sources used in this investigation, tiered by verifiability: