You found out in a Tuesday meeting. Or from a calendar invite with a team on it you’d never heard of. Either way the shape is the same: the AP work is going, the reconciliations are going, your open staff seats aren’t being backfilled, and there’s now a Teams channel that lights up at 11pm with questions about coding rules you wrote three years ago and never documented.

And the question you’re asking, in the words most people actually use, is why the hell would anyone do this.

Here’s the straight answer. Accounting firms and corporate finance teams offshore work for two reasons that get mixed together and shouldn’t be: labour arbitrage, which is the old reason, and the collapse of the US accounting pipeline, which is the reason driving most of the decisions being made right now. The second one is the bigger driver in public accounting. It is also the one nobody says out loud in the all-hands, because “we cannot find people” sounds like an admission of failure and “we are optimising our delivery model” does not. Both are true at once, and if you only understand the cost story you will misread what your employer is doing and what it means for you.

What follows is the honest version: the shortage numbers, how the economics really work once you add back the parts the business case leaves off, what survives the trip, and what happens to a career ladder when somebody saws off the bottom rung. I’m not going to pretend this is good news for you. For a lot of accountants it’s bad news. But the logic isn’t stupid, and understanding it beats being angry at it.

The pipeline number that explains most of this

The AICPA publishes a report called Trends: A Report on Accounting Education, the CPA Exam, and Public Accounting Firms’ Hiring of Recent Graduates. It is the closest thing the profession has to a census of its own supply. The 2025 edition covers the 2023–24 academic year, and it counted 55,152 accounting bachelor’s and master’s degrees awarded in the United States, down 6.6% year over year. Bachelor’s came in at 40,817, down 3.3%. Master’s fell about 15%, to 14,335.

That’s the third consecutive drop; the two years before it fell 9.6% and 7.4%. Stack those and the supply of new accounting graduates has shrunk by roughly a fifth in three years. The licensure pipeline looks worse. New CPA Exam candidates hit 42,626 in 2023, itself inflated by people rushing to sit before the exam was redesigned. In 2024 it was 28,082. The first six months of 2025 produced 16,448. For scale, the AICPA’s earlier Trends report counted 72,271 unique CPA Exam candidates in 2021.

Now set that against demand. The Bureau of Labor Statistics puts employment of accountants and auditors at about 1.6 million in 2025, projects 5% growth through 2035, and estimates about 115,300 openings for accountants and auditors each year over the decade, most of them from retirements and people leaving the occupation entirely. Roughly 115,000 seats a year to fill. Around 41,000 new accounting bachelor’s graduates, not all of whom go into accounting, and fewer than 30,000 people a year now starting the CPA Exam.

One encouraging signal sits in the same report: accounting enrolment hit 266,506 in spring 2025, up 12.4% and the highest since 2020. Those students graduate in 2027 and 2028, which does nothing for a firm staffing a busy season in four months. When a managing partner says the decision is about capacity rather than cost, they are often telling the truth.

The cost math is real too, and smaller than the slide says

Do it properly, because the business case is almost always missing three lines. Start with what a US staff accountant actually costs. Robert Half’s 2026 Salary Guide puts the national midpoint for a staff accountant at $73,750. BLS reports benefits at 30.0% of total compensation for private industry workers as of June 2026, so that grosses up to roughly $105,000 in wages plus benefits alone. Software seats, desk space, recruiting fees and training push it higher, but $105,000 is a defensible floor.

Move five of those seats and the vendor quotes you half. The slide now shows about $264,000 saved. Here’s what comes off it in year one, using published midpoints for the people doing the absorbing:

• Manager time. Somebody has to run the offshore team. An accounting manager at the $113,000 midpoint grosses up to about $161,000 loaded. If 30% of their week goes to review, coordination and rework, that’s roughly $48,000.
• Senior capacity. Work comes back needing more review than it used to, and seniors absorb it. Three seniors at the $94,750 midpoint, each losing 10% of capacity, is about $41,000.
• Parallel running. You don’t switch a close over on a Friday. Two months of double-running while the offshore team learns the ledger costs about $88,000 in duplicated effort.

That leaves roughly $87,000 of the $264,000, before documenting processes nobody wrote down, the security and access review, vendor management time, travel, and the near-certainty that one of the five seats turns over inside twelve months and you pay the learning curve twice.

Those figures are illustrative, built from public midpoints rather than any one company’s books. Run your own. The shape holds either way: the first year of an offshoring programme usually costs more than it saves, and the saving only arrives in years two and three if the offshore team stays put. That’s the most common reason these projects get judged a failure. Finance books the saving in year one and it doesn’t show up. And the 40% to 60% cost reduction you’ll see quoted comes from the people selling the service; I’ve found no independent, audited dataset verifying it across firms, so treat it as a quote rather than a benchmark.

Which work travels, and which never does

This is where programmes are won or lost, and it has nothing to do with the capability of the people offshore. It’s about how much unwritten context a task carries. Work travels well when the input is defined, the rule is written down, and the output is checkable by someone who wasn’t there: AP and AR processing and invoice coding against a maintained chart; bank, intercompany and balance sheet reconciliations; standardised close tasks with a documented checklist and a fixed calendar; audit support schedules, tie-outs, confirmations, PBC chasing and workpaper prep; first-pass tax return preparation finalised onshore; fixed asset registers, lease schedules, expense report auditing, master data maintenance.

Work that doesn’t travel is work where the answer depends on something nobody wrote down:

• Judgment and estimates. Reserve adequacy, impairment triggers, revenue recognition on a messy contract, anything where the right answer is a defensible range rather than a number.
• Anything client-facing, including the informal version where a VP of Sales calls to ask whether a deal structure will fly.
• Institutional memory. Why the Q3 2022 accrual was booked that way. Which of three entities with nearly identical names is the real one. Which controller always sends the file late.
• Exception handling. The 5% of transactions that don’t fit the rule consume most of the time, and they’re exactly what a process document doesn’t cover.

The failure mode is predictable. A firm scopes the work by transaction volume, because volume is easy to measure, then finds the 80% it moved carried maybe 40% of the actual effort. The exceptions stayed behind. So did the judgment. And the people who handled both now review the offshore output as well.

The review burden, and what it does to the bottom rung

Ask any senior eighteen months into an offshoring programme what changed about their job and you get the same answer: I used to do work, now I check work.

That isn’t a complaint about quality. It’s structural. When a US staff accountant did the reconciliation, the reviewer could sample; that person sat twelve feet away, had the same context, and flagged the odd thing before it hit the file. When the same reconciliation comes from a team that has never met the AP clerk at the subsidiary, working from a document you wrote in a hurry, sampling isn’t enough. Review goes from spot-check to full check and the reviewer’s own capacity falls. It does improve, and a stable team on the same accounts for two years needs far less review. Most firms just don’t budget for the eighteen months in between, and the seniors absorbing it do so on top of a full workload. That’s where the resentment comes from, and it’s legitimate.

Here’s the part that should worry partners more than it does. The tasks that travel best offshore are, almost exactly, the tasks a first-year used to learn on. You learned what a plausible balance looks like by reconciling four hundred of them. You learned what a bad invoice smells like by coding thousands. Nobody designed that as a training programme, but it worked as one, and it produced the senior who can look at a trial balance and know within ten seconds that something is wrong without being able to say why.

Take that away and you have a firm that needs experienced reviewers and no longer manufactures them. Firms bridge the gap by poaching seniors. That works until every firm is doing it, and given the pipeline numbers, every firm is doing it. In five years the scarce, expensive resource isn’t the staff accountant. It’s the person with enough pattern recognition to supervise one.

Some firms are handling this on purpose: a smaller onshore cohort rotated through the harder work early, paired with the offshore team so they learn to review rather than produce. That’s a real answer. Most haven’t done it, because it costs money now and the pipeline problem belongs to somebody else in five years.

What separates the programmes that work

This is mainstream now. The AICPA’s 2023 National Management of an Accounting Practice survey, with responses from more than 1,100 firms, found roughly 30% already outsourcing domestically and about 25% offshoring internationally, with a further 12% planning to start. So the question is no longer whether, it’s whether a firm does it competently. The ones that do have most of these in common:

• They documented the process before moving it, not after. If the only copy of a procedure lives in one person’s head, you’re exporting a guessing game.
• They bought retention, not the cheapest rate. The saving lives in years two and three. A vendor with 40% annual attrition never lets you get there, and their rate card looks great.
• They resourced the review layer explicitly, with named people, protected time and a plan to taper it. Not “the seniors will absorb it.”
• They moved in waves, one process at a time, with a parallel run and a stated rollback point.
• They handled the compliance properly. For tax work, IRC §7216 and Treas. Reg. §301.7216-3 require specific taxpayer consent before return information goes to a preparer outside the United States, and §6713 penalties run to $250 per wrongful disclosure, capped at $10,000 a year. AICPA Code interpretation 1.700.040 sets a separate bar for confidential client information generally. Details in the FAQ below.
• They treated the offshore team as staff, with names, career paths and the same training. Teams treated as a cost line behave like a cost line.

The failures usually fail on the second and third. Cheapest rate, no review resourcing, saving booked in year one. Eighteen months later somebody declares that offshoring doesn’t work, brings it back, and the real lesson goes unlearned.

If it’s your department: what actually keeps you valuable

Not “learn Python.” Not “add value.” Specific things, in rough order of importance.

Get on the side of the work that doesn’t travel. Judgment, estimates, technical accounting positions, anything requiring you to take a position and defend it. If your week is mostly production, restructure it until it isn’t.

Own a process rather than run it. Process design, controls, documentation, the SOX narrative, the close calendar, the vendor relationship. Offshoring creates these roles, it doesn’t remove them. Somebody has to own the SLA, the quality metrics and the escalation path, and that person is more secure than before, not less.

Learn to review well, and fast. It’s a real, trainable skill and most people are mediocre at it: knowing where errors hide, sampling intelligently, giving feedback that stops a recurring mistake instead of fixing one instance. It’s the scarce skill in a firm with a hollow middle, which is the firm yours is becoming.

Get systems-fluent. Not a developer. Fluent. Whoever genuinely understands how the ERP, close tool and reporting stack hang together is who both teams have to come to.

Finish the CPA if you’re close. Sign-off authority, independence obligations and licensure still sit onshore. It’s a partial moat, not a permanent one: it protects the signature, not the production work underneath it. But with under 30,000 people a year starting the exam against 115,300 annual openings, it’s getting more valuable, not less.

Move toward the business. FP&A, commercial finance, business partnering, controllership at a smaller company where you do everything. These jobs are defined by proximity and context, which is exactly what doesn’t travel down a video call at 11pm.

And the unsentimental bit. If your employer moved your function offshore and gave you nothing but more review work and the same title, they’ve told you what your next five years look like. Read it as data. The market for an accountant who can review, design process and handle judgment work is strong right now, for exactly the pipeline reasons that caused your problem. That’s the one useful irony here.

AB7 Solutions runs finance and accounting outsourcing and KPO teams, so the firm-side reader is usually who we end up talking to. If that’s you, the useful conversation isn’t about rate cards. It’s about which processes are documented well enough to move, what the review layer will cost you in year one, and which parts of the function should stay exactly where they are. It isn’t free, and we’ll say so when the answer is “not yet.” Call +1 321 341 7733, or email ab@ab7solutions.com or director@ab7solutions.com. More at www.ab7solutions.com.

Questions people ask next

Do clients have to be told their work is being done offshore?

For tax, effectively yes. Section 7216 and Treas. Reg. §301.7216-3 require the taxpayer’s specific signed consent before return information is disclosed to a preparer outside the United States, and for individual returns Rev. Proc. 2013-14 prescribes the language: it must say federal law may not protect the information from further use or distribution, and the client can’t be required to sign as a condition of service. Breach carries criminal exposure of up to a year and a $1,000 fine under §7216. For non-tax engagements, AICPA interpretation 1.700.040 gives the member a choice: a confidentiality agreement with the service provider giving reasonable assurance of appropriate safeguards, or specific client consent. Many firms ask anyway, on the view that a client who finds out later is a client you lose.

Does offshoring reduce quality?

Not inherently, and people who insist it does are usually describing year one of a badly run programme. What reliably reduces quality is churn, thin documentation and unclear ownership of exceptions, and offshoring stresses all three at once. A stable team on the same accounts for three years, with a real review structure, produces work that’s hard to distinguish from onshore output. Getting to year three is the hard part.

Won’t AI make offshoring pointless anyway?

It changes what gets offshored more than whether. Automation and offshore teams compete for the same work: high-volume, rule-based, documented tasks. A firm that automates invoice coding doesn’t need to send it anywhere. So offshore scope drifts upward into work needing judgment, which raises the review burden rather than lowering it. Neither comes for work that depends on context, relationships and defensible professional judgment. That’s still the safe ground, and always has been.

Sources: AICPA & CIMA, Trends: A Report on Accounting Education, the CPA Exam, and Public Accounting Firms’ Hiring of Recent Graduates (2025 and 2023 editions), reported by the Journal of Accountancy and Accounting Today; US Bureau of Labor Statistics, Occupational Outlook Handbook: Accountants and Auditors and Employer Costs for Employee Compensation, June 2026; Journal of Accountancy, “Offshoring for CPA firms: The hows and whys” (Nov 2024), citing the AICPA 2023 National MAP Survey; The CPA Journal, “Considerations for Tax Return Preparers Outsourcing Overseas” (Aug 2025); IRS Rev. Proc. 2013-14; Robert Half 2026 Salary Guide.