Entering voluntary disclosure? What a state-rate-only exposure estimate misses
Edvin Givargis Published 8 minute read
The short answer
The local and district layer, and the miss matters because nobody estimates multistate sales tax exposure out of idle curiosity: the number exists to price voluntary disclosure, to set reserves, and to size the escrow in a transaction. A state-rate-only estimate is a reasonable first pass because the state rate is public, uniform, and easy to pull, but it treats forty or fifty states as if each had one sales tax rate when in fact a large share of them layer city, county, transit-district, or special-purpose-district tax on the state rate, and a smaller but still significant number make the local government responsible for administering and collecting its own share rather than folding it into a single state-run return. Skipping that layer does two kinds of damage at once. It understates the dollar exposure, concentrated precisely in the states where local rates are high or ubiquitous, which means a voluntary disclosure priced on the state rate walks into the agreement underfunded. And in the states that administer local tax separately, it hides a compliance obligation, a separate registration, a separate return, sometimes a separate remittance address, that a state-only analysis will never reveal. The error runs in the other direction too: an estimate padded with a blanket cushion instead of real local rates overstates the liability, and in a deal that overstatement becomes purchase price sitting idle in escrow for years, earning an escrow account's rate instead of the return the seller's capital ordinarily earns, held against an exposure that was never that large. A defensible estimate treats the local layer as its own line of inquiry, not as a rounding error absorbed into the state rate.
The decisions the estimate prices: voluntary disclosure and the deal escrow
An exposure estimate is a decision document, and the two decisions it most often prices leave no room for a soft number. The first is voluntary disclosure. Whether to enter a state's program, in which states, and in what order turns on the arithmetic the estimate supplies: the full exposure with penalties and an unlimited lookback on one side, the cost under the program's limited lookback and penalty relief on the other. An estimate that misses the local layer distorts that arithmetic state by state, and it misses the scope question entirely in the states where a state-level agreement does not reach self-administered local taxes, so the company signs, pays, and believes itself finished while the local exposure keeps aging. The second is the transaction. In a purchase or sale, the diligence exposure schedule becomes the indemnity negotiation, and the negotiation becomes an escrow or holdback sized to the number on the schedule. Both directions of error cost real money. An understated schedule leaves the buyer collecting the difference from the seller after closing, on worse terms and with interest running. An overstated one parks the seller's capital in an escrow account for the survival period when a defensible estimate would have released it at closing to earn whatever that capital earns everywhere else. Precision in the rate layer is not pedantry; it is the difference between an escrow sized to the exposure and an escrow sized to the preparer's shortcut.
Why the shortcut is tempting and where it fails
A multistate exposure estimate usually starts from a sales-by-state report, because that is the data a company's accounting or tax software already produces, and a table of statewide rates, because that is the number every state publishes prominently and updates predictably. Multiplying the two together produces a clean, defensible-looking number for every state in a single pass, and for a state with no local sales tax at all, or with a local tax so uniform and so folded into standard published combined rates that the distinction barely matters, the shortcut is close enough to the truth to be useful. The trouble is that this description fits a minority of states. In most states with a general sales tax, cities, counties, school districts, transit authorities, and special districts impose their own additional percentage, and the combined rate that actually applies at a given address can run well above the statewide rate, varying block by block within the same state and sometimes within the same city, depending on which special districts overlap a given location. An estimate built on the statewide rate alone is not a conservative floor; it is simply the wrong number for every sale made in a jurisdiction with a local add-on, and because those jurisdictions include some of the largest population centers in the states with the highest sales volumes, the aggregate miss tends to land where the exposure was already largest.
Where local administration multiplies the miss
The dollar gap is the smaller problem. The larger one is that a handful of states do not merely add a local percentage to a state-administered tax; they make local jurisdictions independently responsible for imposing, and in some cases collecting, their own sales tax, with their own registration and filing mechanics layered on top of, or sometimes entirely apart from, the state system. California's local sales and use tax under its Bradley-Burns framework and its additional voter-approved district taxes under the state's Transactions and Use Tax Law both ride on the state return, but the combined rate that applies to a given delivery address still depends on which cities, counties, and special districts the address sits within, and treating "California" as a single rate erases that variation even where a single registration covers it. Colorado is the sharper example: a substantial number of Colorado municipalities are home-rule cities that administer and collect their own local sales tax directly, under authority the state constitution gives them, rather than through the state's centralized system, which means a seller with nexus in Colorado at the state level can still owe a separate local registration, a separate local return, and a separate local remittance in every self-collecting home-rule city where it makes sales, entirely apart from whatever the state-level analysis concluded. Louisiana runs a parish-based local sales tax system under its Uniform Local Sales Tax Code, historically administered by local collectors rather than a single state agency, with its own registration and filing footprint that a state-rate-only model has no way to capture. Alabama addresses part of the same problem from the other direction: recognizing that its patchwork of local sales and use taxes was a genuine barrier for remote sellers, it built a Simplified Sellers Use Tax program that lets an eligible remote seller collect and remit a single flat rate covering state and local tax through one registration, which is itself an acknowledgment, from inside the state's own tax administration, that the ordinary local system is complicated enough to need a workaround. None of these mechanics show up in a report that multiplies taxable sales by a single statewide percentage.
What a defensible estimate has to add
Correcting for this does not require abandoning the state-rate estimate as a starting point; it requires treating it as exactly that, a starting point, and adding two layers of work before calling the number final. The first layer is rate-level: for every state where nexus has already been established, pull combined state-and-local rates at the level of the addresses where sales actually occur, not a single statewide average, because the gap between the statewide rate and the highest local combined rate in a state can itself be a meaningful percentage of the sale price, and because sales concentrated in high-local-rate metropolitan areas will skew the true blended rate well above the statewide figure. The second layer is administrative: for every state with any form of local self-administration, separately confirm whether local registration, filing, and remittance are handled through the state's system or require independent local action, because the answer changes both the size of the compliance project and its timeline; a seller that resolves state-level exposure but never registers with a self-collecting home-rule city has not finished the exposure analysis, it has finished half of it and mistaken the half for the whole. Where voluntary disclosure is part of the remediation plan, the same distinction matters again, because a state-level voluntary disclosure agreement does not automatically extend to a self-administering local jurisdiction, and a company that assumes it does can walk out of a state-level resolution still exposed at the local level for the same sales it just cleaned up at the state level.
Practice notes
Build every multistate sales tax exposure estimate in two passes, not one: a state-rate pass to size the project and prioritize states, followed by a combined-rate pass, using address-level or at minimum ZIP-level rate data, before any number goes into a client-facing exposure figure. Flag every state with home-rule or self-administered local sales tax as its own line item in the workpaper, separate from the state-level nexus conclusion, because the two questions, does the state require registration, and does a local jurisdiction inside that state separately require it, do not always have the same answer. Where the underlying sales data is a summary report rather than address-level detail, say so in the assumptions and treat the resulting number as a floor rather than a final figure, the same discipline the underlying estimate should already apply to the transaction-count prong of economic nexus. And before closing out a voluntary disclosure or registration project as complete, confirm state by state whether the resolution reaches self-administered local taxes or stops at the state line, because the gap between those two outcomes is exactly the gap a rate-only estimate was built to hide. The same discipline sizes deal escrows: when the estimate feeds an indemnity negotiation, the combined-rate pass is what defends the number in both directions, against the buyer who wants the escrow bigger and the seller whose capital should be working somewhere better.
This article states the law as of September 17, 2026
Statutes, rates, thresholds, and agency practice change, and a different set of facts can change the answer. Before acting on anything discussed above, contact G&G State Tax Group to confirm what has changed since this was written and how the rules apply to a specific situation.
G&G State Tax Group, LLC is a state and local tax advisory firm. The firm provides state and local tax consulting and representation in state and local tax controversies. G&G does not prepare or file tax returns, perform attest services, or provide bookkeeping, and is not a CPA firm.