The distinguishing feature is where the investment risk sits. The employer sets money aside in a plan trust, invests it, and stays on the hook for every promised benefit no matter how those investments perform — a bad decade for the portfolio is the employer's problem to solve. In a defined contribution plan such as a 401(k), the employee's balance is simply whatever their own chosen investments produced, so a bad decade is the employee's problem. That single difference drives most of the rest of the rules.
Most formulas multiply three things: a benefit percentage set by the plan (often somewhere around 1% to 2%), years of credited service, and a pay figure, commonly the average of your highest-earning or final few years. Because the formula rewards long tenure and rising pay, these plans suit employees who stay with one employer for decades far better than employees who change jobs often, since job-hopping restarts the years-of-service count each time. That, plus a funding commitment stretching decades into the future, is why most private-sector employers stopped offering new plans in the 1980s and 1990s. They remain common in state and local government, public education, federal civilian and military service, and some unionized industries.
Employers cannot fund a plan on instinct. ERISA imposes minimum funding standards, and an actuary periodically estimates how much the plan needs today to cover benefits already earned, using assumptions about investment returns, employee turnover, and how long retirees will live. The plan's funded status — assets divided by the estimated value of benefits owed — is the routine health metric. An underfunded plan is not automatically a failing one, but it signals that larger employer contributions are coming.
Private-sector plans are generally insured by the Pension Benefit Guaranty Corporation, a federal corporation created by ERISA in 1974, which runs separate insurance programs for single-employer and multiemployer (typically union-negotiated) plans. If a covered plan fails, the PBGC pays benefits up to legal limits that depend on your age and the type of plan, so a failure rarely means losing everything, though a highly paid participant with a large promised benefit can see a reduction. The PBGC does not insure defined contribution plans — there is no promised benefit to guarantee, only whatever balance already sits in the account — and government and church plans sit outside the ERISA framework that created the PBGC, so they are not covered either.
Vesting decides whether the promise is actually yours. Plans commonly use a cliff schedule (0% until a set number of years of service, then 100%) or a graded schedule (an increasing percentage each year). Leave before you vest and the employer-funded benefit is forfeited entirely, which makes the vesting date a real number to look up before accepting another job.
Two wrinkles are worth knowing. A cash balance plan is legally a defined benefit plan — the employer still funds it and bears the investment risk — but it expresses the benefit as a hypothetical account balance that grows with annual pay credits and interest credits, so it looks and feels like a 401(k) to the employee. Almost nobody calls a cash balance plan a pension, which is one reason the formal name is the more useful label. And at retirement, many plans offer a choice between a lifetime monthly annuity, often with survivor options for a spouse, and a one-time lump sum the retiree then has to invest and manage — an election that is usually irreversible.
