300,000+ Accountants Left the Profession (2019-2022) and CPAs Command Record Wage Premiums—Is This Crisis or Market Correction?

General
1

I’ve been following the Bureau of Labor Statistics data closely, and the numbers are staggering: over 300,000 accountants left the profession between 2019-2022—that’s a 17% drop from the peak workforce. At the same time, CPAs are now commanding record wage premiums, with starting salaries jumping 30% in some markets and firms hiking pay by 14% on average just to compete for talent.

The question everyone’s debating: Is this a crisis or a market correction?

The Crisis Narrative

The alarm bells are loud:

The crisis view says we’re hemorrhaging talent, burning out the remaining staff, and heading toward a quality collapse. Firms literally can’t scale because they can’t hire. Tax season becomes a nightmare every year. Client service suffers.

The Market Correction View

But here’s the counterargument I keep hearing: maybe the profession was oversupplied with low-skill transactional work that’s now being automated. The exodus eliminated roles that shouldn’t have existed in the post-AI era. What’s left are high-value advisory positions where CPAs can command appropriate compensation for their expertise.

Consider this: 87% of finance leaders now prioritize candidates with data analytics, financial modeling, and ERP expertise over traditional bookkeeping skills (Robert Half 2026 Salary Guide). The profession is shifting from “enter transactions” to “interpret data and advise strategy.”

The Beancount Question

This is where it gets interesting for our community. Does plain text accounting accelerate this transition or democratize the profession?

Accelerate argument: Beancount automation (Python importers, BQL queries, Git workflows) eliminates exactly the low-skill manual entry work that’s disappearing. If you can code your accounting workflows, you don’t need to hire three entry-level staff—you need one technical accountant. This increases the premium on technical skills and makes the CPA license less relevant.

Democratize argument: Beancount lowers barriers to entry. Anyone can learn plain text accounting without $50K in credentials. A software developer can pick up double-entry bookkeeping in a weekend and build their own financial system. This creates more competition, not less.

The Career Strategy Dilemma (2026)

For someone entering the field today, what’s the ROI calculation?

CPA path: 150-hour education requirement, $20K-$50K additional tuition, one year of lost wages, exam fees, but capture 5-15% wage premium and access to certain roles.

Technical skills path: Learn Python, Git, automation tools, Beancount workflows—potentially higher ROI without formal credential, but limited to certain types of clients/roles.

Questions for the Community

  1. For those who left the profession (or considered it): What drove your decision? Was it burnout, better opportunities elsewhere, or lack of compensation?

  2. For those who stayed: Is the labor shortage an opportunity (charge more, clients desperate) or a threat (can’t hire staff, can’t scale)?

  3. For the technically skilled: Do you see Beancount as making you MORE competitive (you can do more with less) or LESS competitive (you’re competing with AI, not other humans)?

  4. For future outlook: Is the 300K exodus a one-time pandemic disruption, or the start of a secular decline in the profession?

I’m genuinely curious where this community lands on the crisis-vs-correction debate. Are we witnessing the profession’s evolution into something more technical and valuable, or watching it collapse under unsustainable pressures?

2

Crisis AND Correction—We’re Living Both Simultaneously

Mike, your framing is spot-on, but I think the answer is: it’s BOTH. The crisis is real for firms trying to operate under the old model. The correction is real for the profession’s long-term evolution. And we’re stuck in the painful transition zone.

The CPA Perspective from My Firm

I run a 15-person CPA firm in Chicago. Here’s what the shortage looks like from my seat:

The Crisis Part:

  • I’ve had two staff accountant positions open for 14 months. Not exaggerating. FOURTEEN MONTHS.
  • The candidates I do see? They’re choosing tech companies over accounting firms. Same skill level (Python, data analysis), but tech offers $85K starting vs our $65K.
  • I’ve had to turn away three potential clients this quarter because I literally don’t have the capacity. That’s $45K in annual recurring revenue I couldn’t take.
  • My senior staff are working 60-hour weeks during tax season (again). Burnout is real. One told me last week she’s considering leaving for a corporate FP&A role with better hours.

The Correction Part:

  • The clients I DID take on? I’m charging 30% more than I did in 2019. And they’re paying it because they have no other options.
  • I’m eliminating transactional work entirely. If a client just needs data entry, I refer them to a bookkeeper. I only take advisory clients now—tax strategy, entity structuring, financial planning.
  • My effective hourly rate has gone from $175 to $275 in three years. The math works: fewer clients × higher rates × more automation = better margins.

The Beancount Angle: Acceleration, Not Democratization

I disagree with the “democratization” argument. Here’s why:

Beancount is making ME more valuable, not creating more competitors:

  • I’ve automated 80% of my bookkeeping clients using Beancount + Python importers. What used to take 20 hours/month now takes 4 hours.
  • But here’s the key: my CLIENTS can’t do this. They don’t have time to learn Beancount. They don’t want to write Python scripts. They want a professional they trust.
  • The “software developer picks up Beancount in a weekend” scenario? Sure, they can track their OWN finances. But can they advise a $2M small business on multi-state tax nexus? Can they design a cost segregation study? Can they defend an IRS audit? NO.

Technical skills COMPLEMENT the CPA license, not replace it:

  • My value proposition is: “I’m a CPA who can code, so I can give you real-time financial dashboards, automated compliance reports, and strategic tax planning.”
  • A software developer’s value proposition is: “I can code.” Great for their own finances. Useless for professional services.

The ROI Calculation You Asked About

You asked: CPA path vs technical skills path?

My answer: DO BOTH.

The highest-earning professionals in 2026 are technically skilled CPAs. Here’s the math:

  • Traditional CPA (no tech skills): $85K-$120K salary
  • Software developer (no accounting knowledge): $90K-$140K salary
  • CPA + Python/Beancount skills: $140K-$200K salary OR $200K+ as solo practitioner

The combination is rare. That’s why it’s valuable.

What Would Reverse the Exodus?

You asked what would bring talent back. My controversial take: Nothing should.

The profession NEEDED to shed the low-skill transactional work. We were training thousands of people to do manual data entry, then wondering why they were miserable and left for better jobs. Good riddance.

What we need instead: Attract a different TYPE of person:

  • Computer science students who are also interested in business
  • Data analysts who want to understand financial systems deeply
  • Automation-minded professionals who see accounting as a workflow engineering problem

The accounting programs that are thriving in 2026? They’re the ones teaching Python + GAAP, not just debits and credits.

The Small Business Reality Check

Bob, I’m curious: how does the shortage look from your bookkeeping practice? Are you seeing more clients because CPAs are too expensive / unavailable? Or are you also struggling to scale?

3

Living the “Opportunity” Side—But It’s Lonely at the Top

Alice, you asked how the shortage looks from my bookkeeping practice. Short answer: I’m drowning in opportunity and can’t hire fast enough to handle it.

The Bookkeeper’s Gold Rush (2026 Edition)

Here’s my situation in Austin:

Client demand is INSANE:

  • I have 12 qualified leads in my inbox right now. TWELVE. I can maybe take on 2-3.
  • Referrals from CPAs like Alice who are “eliminating transactional work”? That’s half my new business.
  • Small businesses who used to do their own bookkeeping in QuickBooks? They’re giving up and hiring me because they can’t find staff who know how to use it.

But I can’t scale:

  • I’ve been trying to hire a junior bookkeeper for 8 months. Got ONE qualified applicant. They chose a corporate accounting job instead (better benefits, no client management stress).
  • The “accountants leaving the profession” stat? I’m convinced half of them went to corporate jobs, not out of finance entirely. Makes sense—same pay, better hours, no tax season hell.

So I’m stuck at 22 clients when I could handle 40+ with the right staffing.

The Beancount Lever: Automation as Survival Strategy

Alice, your “80% automation” number resonates. Here’s my version:

Before Beancount (2021):

  • 15 clients × 18 hours/month average = 270 hours/month
  • That’s 67.5 hours/week. Unsustainable.

After Beancount + Python importers (2024-2026):

  • 22 clients × 8 hours/month average = 176 hours/month
  • That’s 44 hours/week. Livable.

The automation isn’t optional anymore—it’s survival. Without Beancount, I’d still be stuck at 15 clients working 70-hour weeks.

The “Crisis vs Correction” Question from Ground Level

Mike, you asked if this is crisis or correction. From where I sit: It’s a correction that FEELS like a crisis because we’re in the messy middle.

Why correction:

  • The profession was absolutely oversupplied with low-value work. I remember 2015-2018 when I’d see job postings for “$14/hour bookkeeper to enter receipts.” That was insulting and unsustainable.
  • Automation (Beancount, but also tools like Dext, Receipt Bank, even QuickBooks’ AI categorization) is eliminating exactly those jobs. GOOD. They shouldn’t have existed.
  • What’s left? Higher-value work: cash flow analysis, financial advisory, tax planning support, fraud detection, system design.

Why it feels like crisis:

  • We’re in the 5-year transition period where: (a) the low-skill jobs are disappearing, (b) demand for high-skill work is exploding, but (c) the talent pipeline hasn’t adjusted yet.
  • Small businesses are SUFFERING. They can’t afford a $275/hour CPA for everything. They can’t find a $60/hour bookkeeper. They’re stuck doing it themselves badly, or using AI tools that give them 95% accuracy but they don’t catch the 5% errors.

The Wage Premium Question: Real, But Narrow

You mentioned CPAs commanding 5-15% wage premium. Here’s what I’m seeing at the bookkeeper level:

2019: Typical bookkeeper in Austin: $45K-$55K salary
2026: Same role: $65K-$75K salary (33-36% increase!)

But here’s the catch: it’s ONLY for technically skilled bookkeepers.

If you can ONLY do QuickBooks data entry? You’re competing with AI and offshore labor at $20/hour.

If you can build Beancount importers, write SQL queries, design accounting systems, understand Git workflows? You’re earning $70K+ and getting recruited constantly.

The Career Advice Dilemma

Sarah (newbie_accountant), I’m curious about your perspective. You’re exactly the profile Alice and I are trying to hire: software developer who’s learning accounting.

If you were considering a career in this field (DevOps → bookkeeping/accounting), what would make it attractive? What would keep you in tech instead?

Because honestly, the profession NEEDS people like you. But we’re competing with tech salaries and better work-life balance. How do we win?

The Uncomfortable Truth About “Democratization”

Alice said Beancount is acceleration, not democratization. I’ll go further:

Beancount is creating a NEW professional class, not eliminating the old one.

The “software developer learns Beancount in a weekend” scenario? That person becomes either:

  1. A sophisticated DIY user (manages own finances, maybe a few rental properties)
  2. A bookkeeper/accountant who happens to code (my career path)

But they don’t replace CPAs. They don’t replace traditional bookkeepers (most small businesses still want QuickBooks because their bank integrates with it).

What Beancount creates: a THIRD category—technically skilled financial professionals who charge premium rates for automation + advice.

That’s Alice’s $275/hour. That’s my $85/hour (bookkeeper rates). That’s not available to the “$14/hour data entry” person.

The Existential Question: Is This Sustainable Long-Term?

Here’s what keeps me up at night: I’m 38 years old. Can I sustain this pace until 65?

Even with automation, I’m at 44 hours/week of billable work + 10 hours/week of business development, admin, learning new tools. That’s 54 hours/week.

If I can’t hire, can’t scale, can’t take vacations without client panic… is this a good 30-year career? Or am I grinding toward burnout just like the people who left the profession?

I don’t have the answer. But I think the “correction” narrative only works if we figure out the SUSTAINABILITY question.

4

Why I’m NOT Leaving Tech (Yet)—And What Would Change My Mind

Bob, you asked what would make the accounting profession attractive to someone like me (DevOps engineer learning Beancount for personal finance). Honest answer: right now, it wouldn’t. But let me explain why, because I think it reveals something important about the exodus.

The Salary Math Doesn’t Work

Alice mentioned the salary tiers:

  • Traditional CPA: $85K-$120K
  • Software developer: $90K-$140K
  • CPA + Python skills: $140K-$200K

Here’s my reality in Portland:

  • My current DevOps salary: $155K
  • Benefits: Full remote, unlimited PTO (actually usable), stock options, $3K/year learning budget, no client management stress
  • Hours: Genuinely 40 hours/week. Sometimes 45 during deployments. Never 60.

To switch to accounting/bookkeeping, I’d need:

  • Minimum $140K (that’s a $15K pay cut I’d consider for work I genuinely love)
  • Similar work-life balance (I have a life outside work; I’m not giving that up)
  • Intellectual challenge (automation and system design, not repetitive data entry)

Bob’s offer of $70K for technically skilled bookkeeper? That’s a $85K pay cut. Alice’s “$200K+ as solo practitioner”? That requires CPA license ($50K + 2 years) + years building client base + client management stress I explicitly avoid in tech.

The math doesn’t work.

What Tech Workers Value That Accounting Doesn’t Offer (Yet)

Alice said accounting programs need to attract “computer science students who are also interested in business.” I WAS that student. Here’s why I chose tech over accounting:

1. Work-life balance is non-negotiable

  • Bob mentioned 54 hours/week as “sustainable.” In tech, that’s called “crunch mode” and triggers burnout alerts.
  • “Tax season hell” is a recruiting DISASTER for my generation. We grew up watching our parents burn out. Not interested.

2. Learning is built into the job

  • In tech: “Here’s $3K/year for conferences, courses, books. Please stay current.”
  • In accounting (from what I see): “Learn the new tax code on your own time, we’re billing 60 hours this week.”

3. Remote work is standard

  • 90% of tech jobs are remote-first post-COVID.
  • Accounting? Still seeing “hybrid 3 days/week in office” requirements.

4. Prestige and identity

  • Sad but true: “I’m a DevOps engineer at [startup]” gets more social cachet than “I’m a bookkeeper.”
  • Among my peers, accounting is seen as boring, old-school, for people who couldn’t code.

The “Democratization” Debate: I’m Living It

Alice and Bob said Beancount creates a “third category” of technically skilled financial professionals, not democratization.

I disagree. Here’s my experience:

I started tracking my finances in Beancount 6 months ago. Zero accounting background. Within 3 weeks I:

  • Built Python importers for 4 bank accounts
  • Set up Git workflow with daily commits
  • Created custom BQL queries for spending analysis
  • Generated tax reports (basic, but functional)

Could I do this for a $2M business? No. Could I advise on multi-state tax nexus? No.

But could I replace a $45K/year entry-level bookkeeper for a small business? Probably yes.

And here’s the scary part for the profession: there are thousands of tech workers like me who COULD pivot into accounting if the economics made sense.

We’re not, because tech pays better. But if that changes (layoffs, market crash, AI replacing software jobs), you’ll see an influx of technically skilled people who can pick up Beancount/accounting in 3-6 months.

That IS democratization. And it’s a threat to the “$85/hour bookkeeper” tier Bob is in.

The ROI Calculation from My Seat

Mike asked: CPA path vs technical skills path?

From where I sit, there’s a THIRD path no one’s talking about:

Tech career + Beancount for personal finance mastery = Financial independence without credential burden

This is the FIRE community’s answer:

  • Earn $150K+ in tech (high savings rate)
  • Use Beancount to track finances with forensic detail (optimize tax strategy, track net worth, project FI date)
  • Retire at 40-45 (or downshift to part-time consulting)
  • NEVER become a professional accountant

This path offers:

  • Higher income (tech salaries > accounting salaries for equivalent skill level)
  • Better work-life balance (40 hours vs 54+ hours)
  • Personal financial mastery (Beancount skills applied to own wealth)
  • Earlier retirement (FIRE math works better at $150K than $85K)

Why would I trade this for the “accountant who can code” career path?

What Would Actually Attract Me to Accounting

Bob asked: “How do we win?”

Here’s what would make me consider the switch:

1. Competitive total compensation

  • Salary: $130K-$150K for experienced (3-5 year) technically skilled bookkeeper/accountant
  • Benefits: Full remote, real PTO, professional development budget
  • Equity: Profit-sharing or partnership track (like law firms)

2. Sustainable work-life balance

  • 40-45 hours/week MAX, year-round
  • If “tax season hell” is unavoidable, compensate with 3-month summer sabbaticals
  • No expectation of 60-hour weeks as “normal”

3. Intellectual challenge

  • Position as “financial systems engineer” not “bookkeeper”
  • Focus on automation, system design, workflow optimization
  • Minimize repetitive client management

4. Prestige rebranding

  • Market the role as elite technical position (like “DevOps engineer” or “data scientist”)
  • Highlight the intersection of finance + code + strategy
  • Show career paths that DON’T require CPA license but still reach $200K+

5. Faster credentialing path

  • If CPA license is genuinely necessary, lobby for reform:
    • Accept CS degree + accounting bootcamp instead of 150-hour requirement
    • Accept 5 years experience in lieu of formal education
    • Create “technical accountant” designation separate from CPA

The Uncomfortable Question for This Community

Alice said the profession NEEDS to attract people like me. But here’s my question:

Does the accounting profession actually WANT to change enough to attract tech workers?

Because what I’m describing (full remote, 40-hour weeks, $140K salaries, prestige rebranding) would require:

  • Fundamentally rethinking firm economics (fewer hours billed per person, but higher rates)
  • Changing industry culture (tax season hell → sustainable year-round pacing)
  • Competing directly with tech companies for talent (matching benefits, flexibility, compensation)

Is the profession willing to do this? Or will it continue operating on the “long hours, client service, pay your dues” model until it can’t find anyone willing to play that game?

I’m genuinely asking. Because right now, I’m using Beancount to track my path to FIRE at 45. The accounting profession could recruit me—but only if it’s willing to evolve.