What Is Pay Equity? A Small Business Guide

Nearly everything written about pay equity is written for a company with a compensation team, a statistician, and a consultant on retainer. This guide is pay equity for a small business instead: what it actually means, how it differs from the pay gap you keep reading about, what the law requires, and a method for checking your own numbers that works at your size.

The big-company advice is to run a regression analysis across your workforce. If your whole company fits around one table, that is a mathematical joke, and being told to run one is the fastest way to conclude that pay equity is somebody else's problem.

It is not. The Equal Pay Act has no minimum headcount, and small businesses are precisely where pay inequity accumulates, because we make pay decisions informally, in the moment, and never write down why. Then there is the rule that catches everyone: if you find a gap, you cannot fix it by cutting anybody's pay.

I build FirstHR, which is where compensation records and job classifications live. One caveat that matters: this is discrimination law, I am not a lawyer, and this is general information rather than legal advice.

What Is Pay Equity?

Pay equity means compensating employees equally for work of equal value, so that any difference in pay is explained by a legitimate job-related factor rather than by a characteristic like sex or race.

Two things in that definition do the work. Equal value, not identical pay: seniority and performance are real and the law knows it. And explainable: the test is not whether people are paid the same, it is whether you can say why they are not, with evidence, to somebody skeptical.

That is the whole discipline. If you can explain every pay difference in your company out loud, you have pay equity. If you cannot, you have a problem you have not looked at yet.

Internal Pay Equity vs External Pay Equity

Internal pay equity is how people inside your business are paid relative to each other. External pay equity is how those numbers compare to what the wider market pays for the same work. Only the first one carries a legal obligation. The second decides whether your offer gets accepted.

The two fail in different ways and neither covers for the other. You can pay everyone in a role the identical number and still lose all of them, because the number sits well under market. You can also pay competitively and still have two people in the same job separated by nine thousand dollars for no stated reason.

A full compensation analysis covers both halves: the market rate for each role, taken from published wage survey and benchmarking data, and the spread between your own employees doing equal work. Run the internal comparison first. It is the half with legal exposure attached, and it costs an afternoon.

Equity, Equality, and the Gap

Pay equity, pay equality, and the pay gap are three different things. They get confused constantly, and the confusion is why employers either panic or dismiss the subject entirely.

The third term is the one to be clear about. According to the US Census Bureau (Income in the United States: 2025), women working full time, year round had median earnings equal to 83.9 percent of what men earned, up from 80.6 percent the year before.

That 83.9 percent ratio, a gap of about 16 cents on the dollar, is a population-wide average across every job in the economy, driven mostly by which occupations and industries people end up in and how many hours they work. It is a real phenomenon, and it is not a measurement of your business.

You cannot fix the national pay gap. Nobody is asking you to, and no regulator will ever hold you to it. What you can fix, what you are legally obliged to fix, is whether two people doing the same work for you are paid differently for a reason you cannot name.

That is a much smaller, much more tractable problem, and it is entirely within your control. It is the good news buried under a mountain of discouraging statistics.

What the Law Requires

Two federal statutes set the floor, and a state layer on top of them is frequently stricter.

Three implications worth pulling out, because they are the ones small businesses get wrong.

Job titles do not protect you. If two people do substantially the same work, calling one of them a Senior Coordinator does not make the pay difference defensible. Courts look at what people actually do. Creative titling is one of the more common small business instincts and it does not work.

All pay counts, not just salary. Bonuses, commissions, and benefits are all in scope. A base salary that is identical, paired with a bonus that is not, is still a pay difference.

The all-pay rule reaches further than people expect: variable compensation is in scope too, so a bonus scheme applied unevenly carries the same risk as an uneven salary.

The Equal Pay Act has no headcount minimum. It operates through the FLSA, which reaches most employers. Title VII kicks in at fifteen employees, according to the EEOC. Many state equal pay laws apply from your very first hire, so the assumption that you are too small to be covered is usually wrong.

You Cannot Level Down

If you discover a pay gap, you may not fix it by reducing anyone's pay. This is the single most consequential operational fact on this page, and most articles bury it in a footnote.

Sit with what that means financially, because it changes how you should think about the whole exercise. If you find that you have been underpaying someone by $8,000 relative to a comparator doing equal work, your options are: raise them by $8,000. That is the list.

Every shortcut fails. You cannot cut the other person. Freezing them and waiting for the gap to close over three years leaves the gap, and your exposure, in place the whole time. Splitting the difference still involves a cut.

The rule has a second-order consequence that nobody says out loud: the cost of pay inequity only ever goes up. Every year you do not look, the gap persists, the back-pay exposure grows, and the eventual remediation gets more expensive. Waiting is not neutral. Waiting is the expensive option.

Where Gaps Come From

Nobody sets out to underpay women. Pay inequity at a small business is not a decision; it is an accumulation. Here is exactly where it comes from, and which causes will actually defend you.

Look at the top four. Not one of them involves malice, and every one of them is a normal thing a reasonable owner does.

In my experience, negotiation is the biggest: two candidates, same role, one asks for more and gets it, and you now have a pay difference with no job-related basis. It felt like a market outcome at the time. It is not a defense.

Salary history is the second, and it is worse than it looks, because it means you are importing somebody else's pay inequity into your business and making it yours. A woman underpaid at her last job, offered a percentage above her old salary, arrives underpaid at yours. A growing number of states now ban asking for exactly this reason.

And the one-time exception is the quiet one. You made a call five years ago, for a reason that made sense then, and nobody wrote it down. It is now a permanent, unexplainable difference, and the person who made it cannot reconstruct why.

Pay Equity Examples

Six situations that come up constantly at small companies, with the verdict on each. The verdict sits in the middle column, and the reason on the right is the part that generalizes: the defensible cases rest on something written down before the pay decision, and the rest on something that happened in a room.

Two of those six come out clearly defensible, and both only because a document exists. That is the whole distinction in practice: the same pay difference is a legitimate business decision or an unexplained gap depending entirely on whether anybody wrote down the reason at the time.

How to Check Your Own Pay

To check your own pay, group people by the work they actually do, compare the median pay within each group, and write down why any difference exists. This is the section the competition cannot write, because they are selling software that does regression analysis, and the honest answer is that you do not need it and it will not work at your size.

Regression fails at your size not because it is too sophisticated but because it is statistically invalid. A regression model needs enough observations (here, people) per variable to produce a meaningful result.

With six people in a job group, a model testing sex while controlling for tenure, performance, and location estimates at least five coefficients (the values the model has to solve for) from six data points. It will output a number with a confidence interval so wide it means nothing. Worse, it will output it confidently.

Whatever the method, the data has to be retrievable, and for most small businesses it does not currently exist in one place. Compensation, job classification, and tenure all belong in the employee record, which is what an HRIS (human resources information system) is for. When the analysis feels impossible, the usual reason is that the data is scattered, not that the method is hard.

Comparing medians is not a downgrade. At small-business scale it is the correct method, because the real work is not the arithmetic, which takes twenty minutes. The real work is step five: asking, honestly, why each difference exists, and finding out how many of them you cannot answer.

The five tabs below run that sequence in order, and the last two are the ones that matter a year later: what you paid to close each gap, and a dated record that the audit happened at all.

Documenting the Reasons

An undocumented reason is not a defense. It is a memory, and memories are worth nothing in a wage claim filed three years after the decision.

The four statutory defenses, and the only four, are: a seniority system, a merit system, a system measuring quantity or quality of production, and another factor other than sex.

Note the word system appearing three times. A system is a thing that exists, that is written down, and that is applied consistently. It is not a recollection of who you thought was doing well.

The discipline is not the audit. The audit is a once-a-year cleanup. The discipline is writing down the reason at the moment you make the decision, when the reason is fresh and honest, rather than reconstructing it years later when you have every incentive to reconstruct it favorably. A reconstructed rationale reads exactly like what it is.

Fixing What You Find

You ran the analysis and found a gap you cannot justify. The fix is to raise the underpaid employee to parity, and here is how that compares with every alternative employers reach for.

Read the first row and then read the second, because the whole decision lives in the gap between them. There is one lawful remedy and it costs money. Every alternative that does not cost money is either prohibited outright or amounts to leveling down slowly.

The exposure behind that last row is statutory. The Equal Pay Act is enforced through the Fair Labor Standards Act, so a claim reaches back two years, or three where the violation was willful, and a successful employee recovers the unpaid wages plus an equal amount again in liquidated damages, along with attorney fees.

The practical advice is uncomfortable and I am going to give it anyway: if you find an unjustifiable gap, close it now, in one step, and document that you did. Partial remediation leaves the gap in place while creating a paper trail proving you knew about it, which is genuinely the worst of both worlds.

Preventing It Structurally

Everything above is remedial. Prevention happens upstream, before the pay decision is made, and it is a boring structural change rather than an insight.

The pay range item is the one I would prioritize above everything else in this article. If a range exists before the negotiation begins, negotiation cannot create inequity.

A candidate who pushes hard gets the top of the range rather than an arbitrary premium, and the person who did not push is paid from the same range. That single structural change removes what I have found to be the largest source of small business pay inequity, at essentially no cost.

Consistent job classification is the other half, and it is what makes the analysis possible at all. If your roles are clearly defined and consistently titled, grouping comparable employees is trivial. If everyone has a bespoke title invented at hire, you cannot even begin. The job description is the artifact that does this work, and writing one properly is both better hiring and better compliance.

The State Layer

Federal law is the floor. A growing number of states go considerably further, and the obligations follow your employee's work location rather than your office.

The two to check first are the salary history ban and pay range disclosure, because they change your actual hiring process rather than just your risk profile.

Salary history bans and pay range rules are both now in force in a meaningful number of states, and both apply based on where the employee works. That means a remote hire in a new state can change what you are allowed to say in an interview and what you must put in a job posting.

The State Equal Pay Acts Employers Ask About Most

Three state statutes generate more questions than the rest combined, and each one is stricter than the federal floor in a way that changes what you have to be able to prove about your own pay decisions. If you employ anyone in these states, the state standard is the one that governs.

Put that wording next to the federal standard. Federal law asks whether two jobs are substantially equal. The California Equal Pay Act and the Colorado Equal Pay for Equal Work Act both ask whether the work is substantially similar, which is a wider net and pulls more of your people into a single comparison group.

Coverage thresholds move as well. Colorado reaches an employer with a single employee in the state, and the Illinois Equal Pay Act covers every employer with anyone gainfully employed in Illinois. One remote hire in any of these states can put that person, and your pay decisions about them, under the stricter standard.

At the federal level, separate pay data collection through EEO-1 Component 2 is not currently required, and the regulatory position has moved more than once. Check the EEOC data collection page for the current status rather than trusting any article, including this one, on a point that changes.

Common Mistakes

Six recurring failures, and one of them is a violation committed while trying to fix a violation.

The leveling-down mistake deserves the most emphasis because it is made in good faith by employers actively trying to do the right thing. You discover a gap, you have a budget, and cutting the overpaid person seems like the fair and affordable fix. It is a violation.

The higher rate is the legal standard, and the lower rate must rise to it. There is no version of this where you save money.

Never looking is the mistake that costs the most, because it lets every other one compound. The exposure grows quietly with every hire and every raise. Your next step is the first check: group your roles by the work people actually do, compare the medians, and write down the reason for every difference you find.

Frequently Asked Questions