You did everything right. Clean financials, strong revenue trajectory, debt service coverage well above the threshold. The loan officer was cordial. And then came the denial — or the counteroffer at half the amount, with a personal guarantee requirement that wasn’t in the published terms.

You assume you weren’t good enough. That your numbers fell short somewhere. That the man who got approved ahead of you must have had a better application. The ask gap myth would tell you that you simply didn’t ask for enough.

He didn’t. His application was worse. And that’s the part nobody is talking about.

The Experiment That Reframes Everything

In a job market paper presented at the American Economic Association’s 2026 annual meeting, economist Youngjoo Jung designed the experiment the lending discrimination field has needed for decades.

Here’s what she did:

The headline finding: applications with male names were approved 6.2 percentage points more often than identical applications with female names.

That’s consistent with the broader research on how lending discrimination operates showing persistent gender gaps in lending. But Jung’s study goes somewhere no previous study has gone.

Where the Gap Actually Lives

The 6.2-point gap didn’t distribute evenly across all loan quality levels. It concentrated in one specific place.

On high-quality applications — the ones that performed well and repaid on time — there was no statistically significant gender gap. Male and female names were approved at roughly the same rate when the underlying loan was genuinely strong.

The entire gap came from low-quality applications. Loans that later defaulted. Loans with late repayment histories. Loans that a rigorous credit analysis would have flagged.

When the application was bad, loan officers approved it anyway — but only if the name at the top was male.

This is not a story about women being held to an unfair standard. It’s a story about men being held to no standard at all on the applications that matter most.

Why This Distinction Changes the Conversation

The traditional framing of lending bias goes like this: women face higher rejection rates, tighter terms, more collateral demands. All true. All documented. The revenue-approval gap shows that women outperform on the loans they do receive.

But the traditional framing implies that the fix is “be more lenient with women.” Lower the bar. Extend more credit. That framing invites backlash — and it misidentifies the problem.

Jung’s experiment reveals a different mechanism entirely:

This reframes every conversation about lending discrimination. The problem isn’t that the door is too hard to open for women. It’s that the door is propped open for men who shouldn’t be walking through it.

The Male Approval Premium

Call it what it is: a male approval premium. An unearned advantage that kicks in precisely when the numbers don’t support approval.

Think about what this means in practice.

You’re a woman business owner with a borderline application — maybe your business is early-stage, or you had one rough quarter, or your industry is cyclical. The loan officer evaluates your application against the actual risk. Your numbers are soft, so you get denied. Fair enough.

A man with the exact same borderline application sits down the next day. Same soft numbers. Same risk profile. But the loan officer reads the application differently. Not consciously. Not maliciously. The bias is implicit — a pattern 87% of loan officers carry without knowing it.

Something about the application “feels right.” The officer gives the benefit of the doubt. The loan gets approved.

Then it defaults.

This isn’t hypothetical. Jung’s data shows this pattern across 720 loan officers evaluating real applications with known outcomes. The loans that the male approval premium pushed through were the ones that performed worst.

Three Mechanisms Driving the Premium

Jung’s analysis, together with research from the University of Washington Foster School of Business, points to three cognitive mechanisms that create the male approval premium:

1. The Competence Assumption

When a male applicant presents marginal financials, loan officers unconsciously interpret the weakness as situational — a bad quarter, a tough market, a temporary setback. The assumption is that the borrower is fundamentally competent and will recover.

When a female applicant presents the same marginal financials, officers are more likely to interpret the weakness as diagnostic — a signal of the borrower’s actual capability. Same numbers, different story.

This is the attribution asymmetry at work. A man’s dip in revenue is “market conditions.” A woman’s identical dip is “business viability.” The loan officer isn’t lying. They genuinely believe they’re reading the application objectively. The bias lives in the interpretation layer, not the data layer.

Research on credit score miscalibration documents a parallel version of this: the same credit metric carries different weight depending on the applicant’s gender. A 680 FICO on a male application reads as “solid enough.” A 680 on a female application reads as “concerning.”

2. The Risk Narrative Gap

Loan officers build narratives around applications. They have to — a spreadsheet of ratios doesn’t make a credit decision. Someone has to decide what the numbers mean. And that’s where the story diverges.

For male applicants on the border, the narrative tends toward “growth potential” and “entrepreneurial drive.” The weak quarter becomes evidence of resilience. The thin margins become evidence of aggressive reinvestment. The loan officer sees a trajectory.

For female applicants on the same border, the narrative tilts toward “sustainability concerns” and “limited scalability.” The weak quarter becomes a warning sign. The thin margins become evidence of a business that can’t generate enough profit. The loan officer sees a ceiling.

A Springer Nature study on owner vs. manager gender in credit markets found that even the gender of a business’s general manager — not the owner — affects lending outcomes. The bias doesn’t require a woman to be sitting across the table. It activates from a name on a form.

This means gender primes the entire narrative construction process. Before the loan officer has read the third line of the application, the name has already shaped the story they’re about to tell themselves about what the numbers mean.

3. The Downside Blindness

On strong applications, the default is approval regardless of gender. The decision is easy. The numbers carry themselves.

But on weak applications, the decision requires judgment — and judgment is where bias lives.

Male applicants trigger less perceived downside risk in the officer’s mind. The officer underweights the probability of default. The approval feels “close enough.” The internal justification sounds like: “He seems capable of turning this around.”

For female applicants, the officer’s risk perception is better calibrated to reality — which means women are getting the accurate assessment, and men are getting the inflated one.

Here’s the irony: the loan officer’s instinct about the female applicant is right. The application is borderline. It probably shouldn’t be approved. The problem is that the same correct instinct disappears when the name changes to a man’s. The officer isn’t biased against women on bad applications — they’re biased for men on bad applications. And the loans prove it.

What the CFPB Already Knew

This experimental evidence doesn’t exist in a vacuum. A CFPB pilot study on differential treatment in small business lending had already documented that women and minority business owners received less information about loan products, less encouragement to apply, and less favorable treatment during the application process — even in matched-pair tests.

The CFPB study showed the how. Jung’s experiment shows the where. The bias isn’t uniform. It isn’t a flat tax on every application. It’s a thumb on the scale that lifts marginal male applications over the line while leaving equivalent female applications exactly where they are.

The Federal Reserve’s Small Business Credit Survey reinforces this from the demand side: women business owners report lower satisfaction with lender interactions, receive smaller amounts relative to what they request, and are more likely to be offered a different product than the one they applied for.

Why “Just Apply More” Doesn’t Fix This

A common response to lending bias data is to tell women to apply to more lenders. And yes, the lender shopping strategy is a real tactic that works. But it addresses a symptom, not the mechanism.

If the male approval premium is concentrated in low-quality applications, then shopping more lenders with a strong application won’t encounter much gender bias at all — Jung’s data shows the gap is negligible on good applications. You were going to get approved anyway.

And if your application genuinely has weaknesses, shopping more lenders just means running into the same asymmetry at every stop: men with your exact profile getting approved, you getting denied, and nobody documenting the discrepancy.

The advice also ignores the cost of applying. Every application takes time, documentation, and a credit inquiry. For women running businesses — often with less administrative support and tighter operating margins — the application itself is a tax. Telling women to apply five times to get the outcome a man gets in one is not equality. It’s a workaround dressed up as empowerment.

The real fix requires structural change at the lender level. But you can’t wait for structural change. So here’s what works right now.

Two people reviewing financial documents at a table

What You Can Actually Do

Know where the premium activates

The male approval premium kicks in on borderline applications. If your application has any soft spots — thin credit history, less than two years in business, revenue that dipped recently, an industry the lender doesn’t know well — you’re in the zone where gender matters most.

That means your preparation has to be disproportionate to your application’s weakness. A strong applicant can walk in with standard documentation. A borderline applicant who is also a woman needs to arrive over-prepared.

Eliminate subjective judgment wherever possible

Every piece of your application that requires the loan officer to interpret rather than calculate is a surface where bias can operate.

The goal is to turn a judgment call into a math problem. Math has less room for “feels right.”

Apply to lenders with structured processes

Not all lenders have equal exposure to the male approval premium. Institutions with standardized scoring models, committee-based approvals, and demographic-blind review stages leave less room for individual officer bias.

Capital advisors like Lendesca can help you identify lenders whose approval processes rely on structured criteria rather than individual officer discretion — the kind of process where your numbers speak louder than your name.

Ask these questions before applying:

Document the asymmetry if you encounter it

If you’re denied on a borderline application (and for the full playbook, see what to do when your business loan gets denied):

  1. Request the specific adverse action reasons in writing. You’re entitled to them under ECOA.
  2. Ask what would change the decision. Get concrete thresholds: “We need to see DSCR above 1.25” or “We need 24 months of operating history.” Write them down.
  3. If you know of a similar business with a male owner that was approved by the same lender, note it. You don’t need to file a complaint right away, but you need the record.
  4. File a complaint with the CFPB if you believe the denial was influenced by gender. Even if nothing comes of your individual complaint, aggregate complaint data is one of the few tools regulators have.

Strengthen the weak spots before you apply

If your application is borderline, don’t just document it better — improve it. Before you walk in:

Frequently Asked Questions

Does this mean women who get denied actually had bad applications?

Not necessarily. The male approval premium activates on borderline and weak applications — that’s where the gender gap concentrates. But “borderline” is a wide category. It includes early-stage businesses with strong trajectories, seasonal businesses with uneven revenue, and experienced operators entering new industries. Many borderline applications represent perfectly reasonable lending risks. The point is that men in the same borderline category get the benefit of the doubt. Women don’t.

If algorithmic lending removes officer bias, is fintech the answer?

Algorithms eliminate the specific mechanism Jung identified — one human officer making a subjective call. But they introduce their own problems. Models trained on historical approval data learn the male approval premium and encode it. The Regulation B debiasing trap covers why “just use an algorithm” creates new legal and practical complications. The best approach is structured human decision-making: scoring models that constrain judgment plus committee review that diffuses it.

What about women-focused lending programs?

Programs that target women borrowers address the access gap but don’t fix the mechanism. They’re essential — and they’re also a workaround. The structural problem is that mainstream lending has a male approval premium baked into its judgment layer. Programs that route women to separate pipelines don’t change the mainstream pipeline. They create a parallel one.

The Bigger Picture

Jung’s experiment doesn’t just reframe the gender lending gap. It reframes the cost of bias itself.

The male approval premium isn’t free. Those extra approvals on bad applications create real losses — defaults, charge-offs, recovery costs. The lender pays. The guarantor pays. The borrower pays. The only party that doesn’t pay is the bias itself, which persists because its costs are diffuse and its benefits feel like good judgment.

Meanwhile, every woman who was denied on an application that a man with worse numbers got approved on — she pays too. She bootstraps at a higher cost. She grows slower. She exits smaller. And the economy loses the output of businesses that should have been funded but weren’t, while absorbing the losses of businesses that shouldn’t have been funded but were.

The data is no longer ambiguous. The bias isn’t that lenders are too hard on women. It’s that they’re too easy on men. And the applications where they’re too easy are exactly the ones where rigor matters most.

Your application isn’t the problem. The grading curve is.

This article draws on Youngjoo Jung’s December 2025 job market paper, “Loan Officers’ Gender Bias: Evidence from a Randomized Experiment,” presented at the AEA 2026 annual meeting. Additional sources include the CFPB pilot study on differential treatment, University of Washington Foster School lending inequities research, the Federal Reserve Small Business Credit Survey, and a Springer Nature study on gender in credit markets.

HerCapital is an independent publication. All analysis is editorial. All franchise trademarks referenced on this site belong to their respective owners. HerCapital is not affiliated with, endorsed by, or sponsored by any franchisor, lender, or financial institution mentioned in our coverage.