Changing Careers at 40: The Playbook That Respects Your Bills

The honest answer: changing careers at 40 is usually survivable, but only if you treat it as a cash-flow problem first and a purpose problem second. Very few people fail at a mid-career switch because they picked the wrong field. They fail in month seven of a nine-month transition, when the savings run out and the only available move is the fastest job rather than the right one.

So this is a playbook about money: the pay dip, health coverage the day the old badge stops working, the cost of pausing retirement contributions, and how many months of runway a given pay cut demands. It also says where the public data stops and where the retraining industry starts filling the gap with numbers it collected about itself.

Forty is not late, and the tenure data shows why

The Bureau of Labor Statistics Employee Tenure release of September 2024, covering January 2024, put median tenure with a current employer at 3.9 years across all wage and salary workers. For workers aged 45 to 54 the median was 7.0 years, and for those aged 55 to 64 it was 9.6 years. Half of workers in their late forties and early fifties had been with their employer seven years or less.

The story that a 40 year old is abandoning a lifelong track is therefore usually wrong. What is genuinely different at 40 is the balance sheet: a mortgage priced against the current salary, possibly children, and an employer health plan that several people depend on. The risk is not reputational. It is liquidity.

The pay dip: what the data supports and what it does not

There is no clean national statistic for what happens to earnings when somebody voluntarily changes field at 40. Nobody collects it. What does exist is the Bureau of Labor Statistics Worker Displacement release of August 2026, which tracked 3.3 million long-tenured workers who lost jobs held at least three years between January 2023 and December 2025. As of January 2026, 66.1 percent were reemployed. Of the 1.6 million back in full-time wage and salary jobs, about 49 percent were earning as much or more than at the job they lost, against about 62 percent in the previous survey in January 2024.

Roughly half of experienced workers who move under pressure take a pay cut, and that share worsened between the two most recent surveys. Age also changes the odds of getting back in at all: the same release showed 72.9 percent of displaced workers aged 25 to 54 reemployed, against 57.3 percent of those aged 55 to 64.

Displaced workers are not voluntary career changers, who choose their timing and usually prepare, so treat this as a floor on plausible outcomes rather than a forecast. The defensible conclusion is narrow: plan for a pay cut, because the coin flip is roughly even for experienced workers who move.

How long the dip lasts is the question nobody can answer with United States public data. There is no federal series tracking the earnings path of mid-career switchers, so anyone quoting a precise recovery period is citing a private dataset with an interest in the answer. Budget for twenty-four months rather than twelve.

Health insurance is the item that ends transitions early

The day employment ends, so does the employer premium subsidy. Under COBRA, as described by the United States Department of Labor in its employee guide to health benefits, a plan may charge up to 102 percent of the plan total cost of coverage: the former employee share, plus the employer share, plus a 2 percent administrative charge. The employer need not contribute anything. Coverage after termination runs up to 18 months, with an election period of at least 60 days, at least 45 days for the first payment, and a 30 day grace period thereafter.

The number people get wrong is the premium. A payroll deduction is often a quarter or a third of the true cost. Get the COBRA figure in writing before resigning, not after.

The Department of Labor also notes that losing job-based coverage triggers a special enrollment right: generally 30 days to request enrollment in a spouse plan, and 60 days for Marketplace coverage. A spouse plan is frequently cheaper than COBRA, and Marketplace coverage may be cheaper still, because household income usually falls during a transition. Check subsidy eligibility rather than assuming it.

Retirement contributions: the quiet cost

Pausing contributions does not feel like spending. It is. For 2026 the Internal Revenue Service set the 401(k) elective deferral limit at 24,500 dollars, with an additional 8,000 dollars of catch-up contributions permitted from age 50. A 40 year old has ten years before catch-up becomes available, so the pause is not easily made up inside the annual limits.

The employer match is the part that does not come back. Suppose 500 dollars a month with a 250 dollar match, paused for twelve months. That is 9,000 dollars not invested. At a nominal 6 percent annual growth rate, used purely as an illustration and not a forecast, 9,000 dollars left uninvested for 25 years is roughly 38,600 dollars of foregone balance. That belongs in the transition budget as a line item, not as a vague regret.

Check the vesting schedule on the employer match before setting a leaving date. Unvested employer money is a pure loss, and vesting cliffs are often only months away.

The runway calculation

This is the part most career-change writing skips. The formula:

Cash required = (months with no income x monthly essential spending) + (months with no income x monthly health premium) + training cost + (months at reduced pay x monthly shortfall)

Three scenarios run against one baseline household, with the assumptions stated so they can be replaced with real numbers: current salary 84,000 dollars, take-home pay 5,200 dollars a month, essential monthly spending 4,200 dollars, and a health premium of 650 dollars a month once the employer subsidy stops. Every figure in the table is arithmetic from those assumptions, not a survey result.

The gap between column A and column C is the whole argument. A lateral move costs about three and a half months of essential spending. Full retraining with a fifteen month income gap and a 35 percent landing salary costs about twenty-five months, which for most households is more than they have. That difference is decided before anybody chooses a field.

The decision rule

Use this one, and use it before handing in notice:

1. Write down four inputs: months with no income, monthly essential spending, monthly health premium, and the pay cut in the target role.
2. Run the formula, then add three months of essential spending as a buffer, because the job search is the part that slips.
3. If that total cannot be covered from cash savings without touching retirement accounts and without adding revolving credit card debt, the plan is not ready. Change the plan, not the buffer.

Changing the plan almost always means moving left across the table: retrain while still employed, take an internal transfer, or accept a smaller first step rather than the target role. Those moves are less satisfying to describe. They are also the ones that finish.

Be sceptical of retraining outcome claims

Retraining is sold on outcome statistics, and that market is poorly policed.

The clearest illustration is the Consumer Financial Protection Bureau action against the coding school BloomTech and its chief executive in April 2024. The Bureau found the school had advertised job placement rates of 71 to 86 percent within six months of graduation, while its own internal figures were closer to 50 percent and in some cases as low as 30 percent. The chief executive had also publicized a 100 percent placement rate for a cohort that turned out to consist of one student.

That is one company, but the structural problem is general. There is no federal reporting requirement for non-degree career training outcomes. The main voluntary standard, the Council on Integrity in Results Reporting, defines graduation rates, employment at 90, 180 and 360 days, median salaries and employment type, and requires third-party audit. It is voluntary, and a school that dislikes its numbers simply does not join, so the available outcome data is biased towards programs confident of their results.

Practical tests before paying anybody:

• Ask what counts as a placement. Freelance, part-time and unrelated jobs are often folded into one headline figure.
• Ask for the denominator. Rates quoted on job seekers exclude everyone who gave up, which is the group worth knowing about.
• Ask for median salary, not average, and ask for the cohort size.
• Ask whether the figures are audited by anyone outside the company.
• Treat a refusal to answer any of these as the answer.

Where the outlook data helps

The Bureau of Labor Statistics Employment Projections release of August 2026, covering 2025 to 2035, projects total employment rising from 170.3 million to 176.2 million, growth of about 3.5 percent. Healthcare support was the fastest growing occupational group at 13.3 percent, followed by healthcare practitioners and technical at 8.0 percent, life, physical and social science at 7.5 percent, community and social service at 7.4 percent, and computer and mathematical at 7.3 percent.

Read growth rates carefully. A fast-growing small group generates fewer openings than a slow-growing large one, and most hiring replaces people who left rather than filling new positions. Growth rates also say nothing about entry conditions for a 40 year old with no track record, which is the actual question.

On pay, the Bureau of Labor Statistics Occupational Employment and Wage Statistics release of May 2026, covering May 2025, put the median hourly wage across all occupations at 24.51 dollars, with 52.54 dollars for computer and mathematical occupations, 41.60 dollars for healthcare practitioners and technical occupations, and 60.83 dollars for management. Those are medians across everybody in the occupation, including people with twenty years in it. A career changer enters near the bottom of that distribution, and budgeting against the published median is the most expensive mistake in this exercise.

Where these figures come from

• Bureau of Labor Statistics, Employee Tenure news release, September 2024
• Bureau of Labor Statistics, Worker Displacement news release, August 2026
• Bureau of Labor Statistics, Employment Projections news release, August 2026
• Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2026 release
• United States Department of Labor, An Employee Guide to Health Benefits Under COBRA
• Internal Revenue Service, retirement plan contribution and catch-up limits for 2026
• Consumer Financial Protection Bureau, action against BloomTech Inc., April 2024
• Council on Integrity in Results Reporting, published reporting standards

The part that does not resolve neatly

Running the numbers does not make the decision for you. It tells you which version of the decision you can afford, and for many households the honest output of the table is that full retraining is not available this year. That is a constraint, not a failure of nerve, and funding it with credit anyway is how a career change becomes a debt problem that outlasts the career.

Some people run the arithmetic, build the runway, execute properly, and still land in a field they like less than the one they left. The money can be managed. The outcome cannot.

General information, current as of September 2026, and not personalised advice.