In 2019, the American Institute of CPAs counted roughly 1.9 million active accountants in the United States. By 2024, that number had fallen by 340,000. One in six accountants had walked away from the profession in five years.
To understand how extraordinary that is, consider the investment those people made. Each completed a four-year degree. Most completed a fifth year. They sat for a four-part exam with a pass rate below 50%. They completed two years of supervised experience. And then they looked at what the profession was actually asking them to do, and they quit.
The industry calls this a "talent crisis." That framing is wrong. A talent crisis implies the talent does not exist. The talent exists. It is choosing to go somewhere else.
Nobody leaves a profession because the salary is too low if the work is meaningful.
Doctors work punishing hours for years before they earn serious money, and medical school applications have never been higher. The work is hard but it matters. Accountants are leaving because the work is easy and it does not matter. Matching a bank feed to an invoice is not intellectually stimulating. Filing a VAT return is not fulfilling. The daily experience of most accountants is tedium punctuated by deadline stress.
The real problem is the work itself
The accounting profession employs over ten million people globally. They are not uneducated. To qualify requires three to seven years of study, supervised practice, and examination. After all of that training, what does the profession actually spend its time doing?
Data entry. Document matching. Bank reconciliation. Transaction coding. VAT box assignment. Filing preparation. These tasks consume 60-80% of a typical accountant's working hours. Not advisory. Not strategy. Not the high-judgment work the training was designed for.
The profession has convinced itself this is necessary. That manual review ensures accuracy. That human eyes on every transaction prevent errors. This is a story the profession tells itself to justify the status quo. Manual review does not ensure accuracy. It ensures employment. The error rate in manually processed accounting is 2-5%. That is not a precision instrument. That is a human being getting tired at 4pm and miscoding a supplier.
The pyramid trap
Consider how a traditional accounting firm operates. At the bottom: juniors doing reconstruction. In the middle: managers checking the juniors' work. At the top: partners reviewing the reviewed reviews.
This is a pyramid built on a foundation of low-value work. Remove the foundation and the pyramid collapses. Which is exactly why nobody inside the profession is incentivised to remove it. The juniors need the work so they have jobs. The managers need the juniors so they have people to manage. The partners need the whole structure so they have leverage on their time.
The partnership track is the profession's most effective retention tool, and it works precisely because it is an illusion. For every partner, there are roughly eight juniors. The pyramid narrows by design. Most will never reach the top. The firm needs you to believe in it long enough to generate five, seven, ten years of surplus before you either make it or leave.
The work you trained for is the reward for surviving the work you were never trained for. That is not a career path. It is a hazing ritual with a pension.
Where the money actually goes
When a firm charges EUR100 an hour, roughly EUR30 goes to the person doing the work. Another EUR20 goes to the partner's profit share. And EUR50 goes to overhead: the office, the hierarchy, the management, the software licences, the insurance, the billing system. Half the fee is the cost of the firm existing as a firm.
A junior generates two to three times their salary in revenue. The surplus does not go to the client in lower fees. It does not go to the accountant in higher pay. It goes to the firm.
If you are an accountant reading this, do a simple calculation. Take the number of clients you serve. Estimate what the firm charges each of them annually. Add it up. Then look at your salary. The gap between those two numbers is the cost of a structure that exists to organise work that should not require a human being to perform.
AI is not the threat. The firm is.
Most commentary frames AI as the threat to accountants. That is wrong. AI is the liberation. The threat has always been the firm structure itself, which wastes professional talent on reconstruction, extracts the surplus, and calls it a career.
When AI automates the reconstruction, the firm structure loses its purpose. The juniors doing data entry are not needed. The managers supervising them are not needed. The overhead built for that headcount is not needed. What remains is the human judgment that the training was designed for: advisory, tax strategy, client relationships, the moments when a qualified professional makes a difference.
The 340,000 who left were not wrong. They saw what the profession was and decided they deserved better. They were right. The question now is whether the profession will transform before it loses the rest.
Michael Cutajar, CPA — Founder of Accora.